* [PATCH 01/11] generic: test recovery of extended attribute updates
2023-12-31 19:59 ` [PATCHSET v13.0 1/3] fstests: adjust tests for xfs parent pointers Darrick J. Wong
@ 2023-12-27 13:46 ` Darrick J. Wong
2023-12-27 13:46 ` [PATCH 02/11] xfs/206: filter out the parent= status from mkfs Darrick J. Wong
` (9 subsequent siblings)
10 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-27 13:46 UTC (permalink / raw)
To: zlang, djwong
Cc: fstests, catherine.hoang, allison.henderson, guan, linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
Fork generic/475 to test recovery of extended attribute modifications
and log recovery.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
tests/generic/1834 | 93 ++++++++++++++++++++++++++++++++++++++++++++++++
tests/generic/1834.out | 2 +
2 files changed, 95 insertions(+)
create mode 100755 tests/generic/1834
create mode 100644 tests/generic/1834.out
diff --git a/tests/generic/1834 b/tests/generic/1834
new file mode 100755
index 0000000000..7910a40545
--- /dev/null
+++ b/tests/generic/1834
@@ -0,0 +1,93 @@
+#! /bin/bash
+# SPDX-License-Identifier: GPL-2.0
+# Copyright (c) 2022-2024 Oracle, Inc. All Rights Reserved.
+#
+# FS QA Test No. 1834
+#
+# Test log recovery with repeated (simulated) disk failures. We kick
+# off fsstress on the scratch fs to exercise extended attribute operations,
+# then switch out the underlying device with dm-error to see what happens when
+# the disk goes down. Having taken down the fs in this manner, remount it and
+# repeat.
+#
+. ./common/preamble
+_begin_fstest shutdown auto log metadata eio recoveryloop attr
+
+# Override the default cleanup function.
+_cleanup()
+{
+ cd /
+ rm -f $tmp.*
+ $KILLALL_PROG -9 fsstress > /dev/null 2>&1
+ _dmerror_unmount
+ _dmerror_cleanup
+}
+
+# Import common functions.
+. ./common/dmerror
+
+# Modify as appropriate.
+_supported_fs generic
+
+_require_scratch
+_require_dm_target error
+_require_command "$KILLALL_PROG" "killall"
+
+echo "Silence is golden."
+
+_scratch_mkfs >> $seqres.full 2>&1
+_require_metadata_journaling $SCRATCH_DEV
+_dmerror_init
+_dmerror_mount
+
+args=('-z' '-S' 'c')
+
+# Do some directory tree modifications, but the bulk of this is geared towards
+# exercising the xattr code, especially attr_set which can do up to 10k values.
+for verb in unlink rmdir; do
+ args+=('-f' "${verb}=50")
+done
+for verb in creat mkdir; do
+ args+=('-f' "${verb}=2")
+done
+for verb in getfattr listfattr; do
+ args+=('-f' "${verb}=3")
+done
+for verb in attr_remove removefattr; do
+ args+=('-f' "${verb}=4")
+done
+args+=('-f' "setfattr=20")
+args+=('-f' "attr_set=60") # sets larger xattrs
+
+while _soak_loop_running $((50 * TIME_FACTOR)); do
+ ($FSSTRESS_PROG "${args[@]}" $FSSTRESS_AVOID -d $SCRATCH_MNT -n 999999 -p $((LOAD_FACTOR * 4)) >> $seqres.full &) \
+ > /dev/null 2>&1
+
+ # purposely include 0 second sleeps to test shutdown immediately after
+ # recovery
+ sleep $((RANDOM % 3))
+
+ # This test aims to simulate sudden disk failure, which means that we
+ # do not want to quiesce the filesystem or otherwise give it a chance
+ # to flush its logs. Therefore we want to call dmsetup with the
+ # --nolockfs parameter; to make this happen we must call the load
+ # error table helper *without* 'lockfs'.
+ _dmerror_load_error_table
+
+ ps -e | grep fsstress > /dev/null 2>&1
+ while [ $? -eq 0 ]; do
+ $KILLALL_PROG -9 fsstress > /dev/null 2>&1
+ wait > /dev/null 2>&1
+ ps -e | grep fsstress > /dev/null 2>&1
+ done
+
+ # Mount again to replay log after loading working table, so we have a
+ # consistent XFS after test.
+ _dmerror_unmount || _fail "unmount failed"
+ _dmerror_load_working_table
+ _dmerror_mount || _fail "mount failed"
+done
+
+# success, all done
+status=0
+exit
diff --git a/tests/generic/1834.out b/tests/generic/1834.out
new file mode 100644
index 0000000000..5efe2033b5
--- /dev/null
+++ b/tests/generic/1834.out
@@ -0,0 +1,2 @@
+QA output created by 1834
+Silence is golden.
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 02/11] xfs/206: filter out the parent= status from mkfs
2023-12-31 19:59 ` [PATCHSET v13.0 1/3] fstests: adjust tests for xfs parent pointers Darrick J. Wong
2023-12-27 13:46 ` [PATCH 01/11] generic: test recovery of extended attribute updates Darrick J. Wong
@ 2023-12-27 13:46 ` Darrick J. Wong
2023-12-27 13:47 ` [PATCH 03/11] xfs/122: update for parent pointers Darrick J. Wong
` (8 subsequent siblings)
10 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-27 13:46 UTC (permalink / raw)
To: zlang, djwong
Cc: fstests, catherine.hoang, allison.henderson, guan, linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
Filter out the parent pointer bits from the mkfs output so that we don't
cause a regression in this test.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
tests/xfs/206 | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/tests/xfs/206 b/tests/xfs/206
index cb346b6dc9..f7f12ff1f9 100755
--- a/tests/xfs/206
+++ b/tests/xfs/206
@@ -64,7 +64,8 @@ mkfs_filter()
-e "s/\(sunit=\)\([0-9]* blks,\)/\10 blks,/" \
-e "s/, lazy-count=[0-9]//" \
-e "/.*crc=/d" \
- -e "/^Default configuration/d"
+ -e "/^Default configuration/d" \
+ -e 's/, parent=[01]//'
}
# mkfs slightly smaller than that, small log for speed.
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 03/11] xfs/122: update for parent pointers
2023-12-31 19:59 ` [PATCHSET v13.0 1/3] fstests: adjust tests for xfs parent pointers Darrick J. Wong
2023-12-27 13:46 ` [PATCH 01/11] generic: test recovery of extended attribute updates Darrick J. Wong
2023-12-27 13:46 ` [PATCH 02/11] xfs/206: filter out the parent= status from mkfs Darrick J. Wong
@ 2023-12-27 13:47 ` Darrick J. Wong
2023-12-27 13:47 ` [PATCH 04/11] populate: create hardlinks " Darrick J. Wong
` (7 subsequent siblings)
10 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-27 13:47 UTC (permalink / raw)
To: zlang, djwong
Cc: fstests, catherine.hoang, allison.henderson, guan, linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
Update test for parent pointers.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
tests/xfs/122.out | 3 +++
1 file changed, 3 insertions(+)
diff --git a/tests/xfs/122.out b/tests/xfs/122.out
index 89f7b735b0..55138218dd 100644
--- a/tests/xfs/122.out
+++ b/tests/xfs/122.out
@@ -98,6 +98,8 @@ sizeof(struct xfs_fsop_ag_resblks) = 64
sizeof(struct xfs_fsop_geom) = 256
sizeof(struct xfs_fsop_geom_v1) = 112
sizeof(struct xfs_fsop_geom_v4) = 112
+sizeof(struct xfs_getparents) = 96
+sizeof(struct xfs_getparents_rec) = 24
sizeof(struct xfs_icreate_log) = 28
sizeof(struct xfs_inode_log_format) = 56
sizeof(struct xfs_inode_log_format_32) = 52
@@ -107,6 +109,7 @@ sizeof(struct xfs_legacy_timestamp) = 8
sizeof(struct xfs_log_dinode) = 176
sizeof(struct xfs_log_legacy_timestamp) = 8
sizeof(struct xfs_map_extent) = 32
+sizeof(struct xfs_parent_name_rec) = 16
sizeof(struct xfs_phys_extent) = 16
sizeof(struct xfs_refcount_key) = 4
sizeof(struct xfs_refcount_rec) = 12
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 04/11] populate: create hardlinks for parent pointers
2023-12-31 19:59 ` [PATCHSET v13.0 1/3] fstests: adjust tests for xfs parent pointers Darrick J. Wong
` (2 preceding siblings ...)
2023-12-27 13:47 ` [PATCH 03/11] xfs/122: update for parent pointers Darrick J. Wong
@ 2023-12-27 13:47 ` Darrick J. Wong
2023-12-27 13:47 ` [PATCH 05/11] xfs/021: adapt golden output files " Darrick J. Wong
` (6 subsequent siblings)
10 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-27 13:47 UTC (permalink / raw)
To: zlang, djwong
Cc: fstests, catherine.hoang, allison.henderson, guan, linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
Create some hardlinked files so that we can exercise parent pointers.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
common/populate | 38 ++++++++++++++++++++++++++++++++++++++
src/popdir.pl | 11 +++++++++++
2 files changed, 49 insertions(+)
diff --git a/common/populate b/common/populate
index 8097151919..83cd0eb5db 100644
--- a/common/populate
+++ b/common/populate
@@ -464,6 +464,44 @@ _scratch_xfs_populate() {
cp --reflink=always "${SCRATCH_MNT}/REFCOUNTBT" "${SCRATCH_MNT}/REFCOUNTBT2"
fi
+ # Parent pointers
+ is_pptr="$(_xfs_has_feature "$SCRATCH_MNT" parent -v)"
+ if [ $is_pptr -gt 0 ]; then
+ echo "+ parent pointers"
+
+ # Create a couple of parent pointers
+ __populate_create_dir "${SCRATCH_MNT}/PPTRS" 1 '' --hardlink --format "two_%d"
+
+ # Create one xattr leaf block of parent pointers
+ nr="$((blksz * 2 / 16))"
+ __populate_create_dir "${SCRATCH_MNT}/PPTRS" ${nr} '' --hardlink --format "many%04d"
+
+ # Create multiple xattr leaf blocks of large parent pointers
+ nr="$((blksz * 16 / 16))"
+ __populate_create_dir "${SCRATCH_MNT}/PPTRS" ${nr} '' --hardlink --format "y%0254d"
+
+ # Create multiple paths to a file
+ local moof="${SCRATCH_MNT}/PPTRS/moofile"
+ touch "${moof}"
+ for ((i = 0; i < 4; i++)); do
+ mkdir -p "${SCRATCH_MNT}/PPTRS/SUB${i}"
+ ln "${moof}" "${SCRATCH_MNT}/PPTRS/SUB${i}/moofile"
+ done
+
+ # Create parent pointers of various lengths
+ touch "${SCRATCH_MNT}/PPTRS/vlength"
+ local len_len
+ local tst
+ local fname
+ ln "${SCRATCH_MNT}/PPTRS/vlength" "${SCRATCH_MNT}/PPTRS/b"
+ for len in 32 64 96 128 160 192 224 250 255; do
+ len_len="${#len}"
+ tst="$(perl -e "print \"b\" x (${len} - (${len_len} + 1))")"
+ fname="v${tst}${len}"
+ ln "${SCRATCH_MNT}/PPTRS/vlength" "${SCRATCH_MNT}/PPTRS/${fname}"
+ done
+ fi
+
# Copy some real files (xfs tests, I guess...)
echo "+ real files"
test $fill -ne 0 && __populate_fill_fs "${SCRATCH_MNT}" 5
diff --git a/src/popdir.pl b/src/popdir.pl
index e89095aafe..0104957a3c 100755
--- a/src/popdir.pl
+++ b/src/popdir.pl
@@ -17,6 +17,7 @@ GetOptions("start=i" => \$start,
"dir=s" => \$dir,
"remove!" => \$remove,
"help!" => \$help,
+ "hardlink!" => \$hardlink,
"verbose!" => \$verbose);
@@ -35,6 +36,7 @@ Options:
--format=str printf formatting string for file name ("%08d")
--verbose verbose output
--help this help screen
+ --hardlink hardlink subsequent files to the first one created
EOF
exit(1) unless defined $help;
# otherwise...
@@ -56,12 +58,21 @@ if ($file_pct < 0) {
$file_pct = 100;
}
+if ($hardlink) {
+ $file_pct = 100;
+ $link_fname = sprintf($format, $start);
+}
+
for ($i = $start; $i <= $end; $i += $incr) {
$fname = sprintf($format, $i);
if ($remove) {
$verbose && print "rm $fname\n";
unlink($fname) or rmdir($fname) or die("unlink $fname");
+ } elsif ($hardlink && $i > $start) {
+ # hardlink everything after the first file
+ $verbose && print "ln $link_fname $fname\n";
+ link $link_fname, $fname;
} elsif (($i % 100) < $file_pct) {
# create a file
$verbose && print "touch $fname\n";
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 05/11] xfs/021: adapt golden output files for parent pointers
2023-12-31 19:59 ` [PATCHSET v13.0 1/3] fstests: adjust tests for xfs parent pointers Darrick J. Wong
` (3 preceding siblings ...)
2023-12-27 13:47 ` [PATCH 04/11] populate: create hardlinks " Darrick J. Wong
@ 2023-12-27 13:47 ` Darrick J. Wong
2023-12-27 13:47 ` [PATCH 06/11] xfs/{018,191,288}: disable parent pointers for this test Darrick J. Wong
` (5 subsequent siblings)
10 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-27 13:47 UTC (permalink / raw)
To: zlang, djwong
Cc: fstests, catherine.hoang, allison.henderson, guan, linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
Parent pointers change the xattr structure dramatically, so fix this
test to handle them. For the most part we can get away with filtering
out the parent pointer fields (which xfs_db decodes for us), but the
namelen/valuelen/attr_filter fields still show through.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
common/rc | 4 +++
tests/xfs/021 | 15 +++++++++--
tests/xfs/021.cfg | 1 +
tests/xfs/021.out.default | 0
tests/xfs/021.out.parent | 62 +++++++++++++++++++++++++++++++++++++++++++++
5 files changed, 80 insertions(+), 2 deletions(-)
create mode 100644 tests/xfs/021.cfg
rename tests/xfs/{021.out => 021.out.default} (100%)
create mode 100644 tests/xfs/021.out.parent
diff --git a/common/rc b/common/rc
index 969ff93de7..0898ac08eb 100644
--- a/common/rc
+++ b/common/rc
@@ -3419,6 +3419,8 @@ _get_os_name()
_link_out_file_named()
{
+ test -n "$seqfull" || _fail "need to set seqfull"
+
local features=$2
local suffix=$(FEATURES="$features" perl -e '
my %feathash;
@@ -3454,6 +3456,8 @@ _link_out_file()
{
local features
+ test -n "$seqfull" || _fail "need to set seqfull"
+
if [ $# -eq 0 ]; then
features="$(_get_os_name),$FSTYP"
if [ -n "$MOUNT_OPTIONS" ]; then
diff --git a/tests/xfs/021 b/tests/xfs/021
index 9432e2acb0..ef307fc064 100755
--- a/tests/xfs/021
+++ b/tests/xfs/021
@@ -67,6 +67,13 @@ _scratch_mkfs_xfs >/dev/null \
echo "*** mount FS"
_scratch_mount
+seqfull=$0
+if _xfs_has_feature $SCRATCH_MNT parent; then
+ _link_out_file "parent"
+else
+ _link_out_file ""
+fi
+
testfile=$SCRATCH_MNT/testfile
echo "*** make test file 1"
@@ -108,7 +115,10 @@ _scratch_unmount >>$seqres.full 2>&1 \
echo "*** dump attributes (1)"
_scratch_xfs_db -r -c "inode $inum_1" -c "print a.sfattr" | \
- sed -e '/secure = /d' | sed -e '/parent = /d'
+ perl -ne '
+/\.secure/ && next;
+/\.parent/ && next;
+ print unless /^\d+:\[.*/;'
echo "*** dump attributes (2)"
@@ -124,10 +134,11 @@ s/info.hdr/info/;
/hdr.info.uuid/ && next;
/hdr.info.lsn/ && next;
/hdr.info.owner/ && next;
+/\.parent/ && next;
s/^(hdr.info.magic =) 0x3bee/\1 0xfbee/;
s/^(hdr.firstused =) (\d+)/\1 FIRSTUSED/;
s/^(hdr.freemap\[0-2] = \[base,size]).*/\1 [FREEMAP..]/;
-s/^(entries\[0-2] = \[hashval,nameidx,incomplete,root,local]).*/\1 [ENTRIES..]/;
+s/^(entries\[0-[23]] = \[hashval,nameidx,incomplete,root,local]).*/\1 [ENTRIES..]/;
print unless /^\d+:\[.*/;'
echo "*** done"
diff --git a/tests/xfs/021.cfg b/tests/xfs/021.cfg
new file mode 100644
index 0000000000..73b127260c
--- /dev/null
+++ b/tests/xfs/021.cfg
@@ -0,0 +1 @@
+parent: parent
diff --git a/tests/xfs/021.out b/tests/xfs/021.out.default
similarity index 100%
rename from tests/xfs/021.out
rename to tests/xfs/021.out.default
diff --git a/tests/xfs/021.out.parent b/tests/xfs/021.out.parent
new file mode 100644
index 0000000000..1af1301061
--- /dev/null
+++ b/tests/xfs/021.out.parent
@@ -0,0 +1,62 @@
+QA output created by 021
+*** mkfs
+*** mount FS
+*** make test file 1
+# file: <TESTFILE>.1
+user.a1
+user.a2--
+
+*** make test file 2
+1+0 records in
+1+0 records out
+# file: <TESTFILE>.2
+user.a1
+user.a2-----
+user.a3
+
+Attribute "a3" had a 65535 byte value for <TESTFILE>.2:
+size of attr value = 65536
+
+*** unmount FS
+*** dump attributes (1)
+a.sfattr.hdr.totsize = 53
+a.sfattr.hdr.count = 3
+a.sfattr.list[0].namelen = 16
+a.sfattr.list[0].valuelen = 10
+a.sfattr.list[0].root = 0
+a.sfattr.list[1].namelen = 2
+a.sfattr.list[1].valuelen = 3
+a.sfattr.list[1].root = 0
+a.sfattr.list[1].name = "a1"
+a.sfattr.list[1].value = "v1\d"
+a.sfattr.list[2].namelen = 4
+a.sfattr.list[2].valuelen = 5
+a.sfattr.list[2].root = 0
+a.sfattr.list[2].name = "a2--"
+a.sfattr.list[2].value = "v2--\d"
+*** dump attributes (2)
+hdr.info.forw = 0
+hdr.info.back = 0
+hdr.info.magic = 0xfbee
+hdr.count = 4
+hdr.usedbytes = 84
+hdr.firstused = FIRSTUSED
+hdr.holes = 0
+hdr.freemap[0-2] = [base,size] [FREEMAP..]
+entries[0-3] = [hashval,nameidx,incomplete,root,local] [ENTRIES..]
+nvlist[0].valuelen = 8
+nvlist[0].namelen = 2
+nvlist[0].name = "a1"
+nvlist[0].value = "value_1\d"
+nvlist[1].valueblk = 0x1
+nvlist[1].valuelen = 65535
+nvlist[1].namelen = 2
+nvlist[1].name = "a3"
+nvlist[2].valuelen = 10
+nvlist[2].namelen = 16
+nvlist[3].valuelen = 8
+nvlist[3].namelen = 7
+nvlist[3].name = "a2-----"
+nvlist[3].value = "value_2\d"
+*** done
+*** unmount
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 06/11] xfs/{018,191,288}: disable parent pointers for this test
2023-12-31 19:59 ` [PATCHSET v13.0 1/3] fstests: adjust tests for xfs parent pointers Darrick J. Wong
` (4 preceding siblings ...)
2023-12-27 13:47 ` [PATCH 05/11] xfs/021: adapt golden output files " Darrick J. Wong
@ 2023-12-27 13:47 ` Darrick J. Wong
2023-12-27 13:48 ` [PATCH 07/11] xfs/306: fix formatting failures with parent pointers Darrick J. Wong
` (4 subsequent siblings)
10 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-27 13:47 UTC (permalink / raw)
To: zlang, djwong
Cc: fstests, catherine.hoang, allison.henderson, guan, linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
These tests depend heavily on the xattr formats created for new files.
Parent pointers break those assumptions, so force parent pointers off.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
common/xfs | 15 +++++++++++++++
tests/xfs/018 | 4 ++++
tests/xfs/191 | 3 +++
tests/xfs/288 | 4 ++++
4 files changed, 26 insertions(+)
diff --git a/common/xfs b/common/xfs
index f53b33fc54..88fa6fb55a 100644
--- a/common/xfs
+++ b/common/xfs
@@ -1816,3 +1816,18 @@ _xfs_discard_max_offset_kb()
$XFS_IO_PROG -c 'statfs' "$1" | \
awk '{g[$1] = $3} END {print (g["geom.bsize"] * g["geom.datablocks"] / 1024)}'
}
+
+# Adjust MKFS_OPTIONS as necessary to avoid having parent pointers formatted
+# onto the filesystem
+_xfs_force_no_pptrs()
+{
+ # Nothing to do if parent pointers aren't supported by mkfs
+ $MKFS_XFS_PROG 2>&1 | grep -q parent=0 || return
+
+ if echo "$MKFS_OPTIONS" | grep -q 'parent='; then
+ MKFS_OPTIONS="$(echo "$MKFS_OPTIONS" | sed -e 's/parent=[01]/parent=0/g')"
+ return
+ fi
+
+ MKFS_OPTIONS="$MKFS_OPTIONS -n parent=0"
+}
diff --git a/tests/xfs/018 b/tests/xfs/018
index 73040edc92..7d1b861d1c 100755
--- a/tests/xfs/018
+++ b/tests/xfs/018
@@ -111,6 +111,10 @@ attr32l="X$attr32k"
attr64k="$attr32k$attr32k"
echo "*** mkfs"
+
+# Parent pointers change the xattr formats sufficiently to break this test.
+# Disable parent pointers if mkfs supports it.
+_xfs_force_no_pptrs
_scratch_mkfs >/dev/null
blk_sz=$(_scratch_xfs_get_sb_field blocksize)
diff --git a/tests/xfs/191 b/tests/xfs/191
index 7a02f1be21..e2150bf797 100755
--- a/tests/xfs/191
+++ b/tests/xfs/191
@@ -33,6 +33,9 @@ _fixed_by_kernel_commit 7be3bd8856fb "xfs: empty xattr leaf header blocks are no
_fixed_by_kernel_commit e87021a2bc10 "xfs: use larger in-core attr firstused field and detect overflow"
_fixed_by_git_commit xfsprogs f50d3462c654 "xfs_repair: ignore empty xattr leaf blocks"
+# Parent pointers change the xattr formats sufficiently to break this test.
+# Disable parent pointers if mkfs supports it.
+_xfs_force_no_pptrs
_scratch_mkfs_xfs | _filter_mkfs >$seqres.full 2>$tmp.mkfs
cat $tmp.mkfs >> $seqres.full
source $tmp.mkfs
diff --git a/tests/xfs/288 b/tests/xfs/288
index aa664a266e..60fb9360f4 100755
--- a/tests/xfs/288
+++ b/tests/xfs/288
@@ -19,6 +19,10 @@ _supported_fs xfs
_require_scratch
_require_attrs
+# Parent pointers change the xattr formats sufficiently to break this test.
+# Disable parent pointers if mkfs supports it.
+_xfs_force_no_pptrs
+
# get block size ($dbsize) from the mkfs output
_scratch_mkfs_xfs 2>/dev/null | _filter_mkfs 2>$tmp.mkfs >/dev/null
. $tmp.mkfs
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 07/11] xfs/306: fix formatting failures with parent pointers
2023-12-31 19:59 ` [PATCHSET v13.0 1/3] fstests: adjust tests for xfs parent pointers Darrick J. Wong
` (5 preceding siblings ...)
2023-12-27 13:47 ` [PATCH 06/11] xfs/{018,191,288}: disable parent pointers for this test Darrick J. Wong
@ 2023-12-27 13:48 ` Darrick J. Wong
2023-12-27 13:48 ` [PATCH 08/11] common: add helpers for parent pointer tests Darrick J. Wong
` (3 subsequent siblings)
10 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-27 13:48 UTC (permalink / raw)
To: zlang, djwong
Cc: fstests, catherine.hoang, allison.henderson, guan, linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
The parent pointers feature isn't supported on tiny 20MB filesystems
because the larger directory transactions result in larger minimum log
sizes, particularly with nrext64 enabled:
** mkfs failed with extra mkfs options added to " -m rmapbt=0, -i nrext64=1, -n parent=1," by test 306 **
** attempting to mkfs using only test 306 options: -d size=20m -n size=64k **
max log size 5108 smaller than min log size 5310, filesystem is too small
We don't support 20M filesystems anymore, so bump the filesystem size up
to 100M and skip this test if we can't actually format the filesystem.
Convert the open-coded punch-alternating logic into a call to that
program to reduce execve overhead, which more than makes up having to
write 5x as much data to fragment the free space.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
tests/xfs/306 | 9 ++++-----
1 file changed, 4 insertions(+), 5 deletions(-)
diff --git a/tests/xfs/306 b/tests/xfs/306
index b57bf4c0a9..152971cfc3 100755
--- a/tests/xfs/306
+++ b/tests/xfs/306
@@ -23,6 +23,7 @@ _supported_fs xfs
_require_scratch_nocheck # check complains about single AG fs
_require_xfs_io_command "fpunch"
_require_command $UUIDGEN_PROG uuidgen
+_require_test_program "punch-alternating"
# Disable the scratch rt device to avoid test failures relating to the rt
# bitmap consuming all the free space in our small data device.
@@ -30,7 +31,8 @@ unset SCRATCH_RTDEV
# Create a small fs with a large directory block size. We want to fill up the fs
# quickly and then create multi-fsb dirblocks over fragmented free space.
-_scratch_mkfs_xfs -d size=20m -n size=64k >> $seqres.full 2>&1
+_scratch_mkfs_xfs -d size=100m -n size=64k >> $seqres.full 2>&1 || \
+ _notrun 'could not format tiny scratch fs'
_scratch_mount
# Fill a source directory with many largish-named files. 1k uuid-named entries
@@ -49,10 +51,7 @@ done
$XFS_IO_PROG -xc "resblks 16" $SCRATCH_MNT >> $seqres.full 2>&1
dd if=/dev/zero of=$SCRATCH_MNT/file bs=4k >> $seqres.full 2>&1
$XFS_IO_PROG -c "fsync" $SCRATCH_MNT/file >> $seqres.full 2>&1
-size=`_get_filesize $SCRATCH_MNT/file`
-for i in $(seq 0 8192 $size); do
- $XFS_IO_PROG -c "fpunch $i 4k" $SCRATCH_MNT/file >> $seqres.full 2>&1
-done
+$here/src/punch-alternating $SCRATCH_MNT/file
# Replicate the src dir several times into fragmented free space. After one or
# two dirs, we should have nothing but non-contiguous directory blocks.
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 08/11] common: add helpers for parent pointer tests
2023-12-31 19:59 ` [PATCHSET v13.0 1/3] fstests: adjust tests for xfs parent pointers Darrick J. Wong
` (6 preceding siblings ...)
2023-12-27 13:48 ` [PATCH 07/11] xfs/306: fix formatting failures with parent pointers Darrick J. Wong
@ 2023-12-27 13:48 ` Darrick J. Wong
2023-12-27 13:48 ` [PATCH 09/11] xfs: add parent pointer test Darrick J. Wong
` (2 subsequent siblings)
10 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-27 13:48 UTC (permalink / raw)
To: zlang, djwong
Cc: Allison Henderson, Catherine Hoang, fstests, catherine.hoang,
allison.henderson, guan, linux-xfs
From: Allison Henderson <allison.henderson@oracle.com>
Add helper functions in common/parent to parse and verify parent
pointers. Also add functions to check that mkfs, kernel, and xfs_io
support parent pointers.
Signed-off-by: Allison Henderson <allison.henderson@oracle.com>
Signed-off-by: Catherine Hoang <catherine.hoang@oracle.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
[djwong: add license and copyright, dont _fail tests immediately, make
sure the pptr-generated paths match the dir-generated paths]
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
common/parent | 209 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++
common/rc | 3 +
common/xfs | 12 +++
3 files changed, 224 insertions(+)
create mode 100644 common/parent
diff --git a/common/parent b/common/parent
new file mode 100644
index 0000000000..3f7ac4ca35
--- /dev/null
+++ b/common/parent
@@ -0,0 +1,209 @@
+#! /bin/bash
+# SPDX-License-Identifier: GPL-2.0
+# Copyright (c) 2022-2024 Oracle and/or its affiliates. All Rights Reserved.
+#
+# Parent pointer common functions
+#
+
+#
+# parse_parent_pointer parents parent_inode parent_pointer_name
+#
+# Given a list of parent pointers, find the record that matches
+# the given inode and filename
+#
+# inputs:
+# parents : A list of parent pointers in the format of:
+# inode/generation/name_length/name
+# parent_inode : The parent inode to search for
+# parent_name : The parent name to search for
+#
+# outputs:
+# PPINO : Parent pointer inode
+# PPGEN : Parent pointer generation
+# PPNAME : Parent pointer name
+# PPNAME_LEN : Parent pointer name length
+#
+_parse_parent_pointer()
+{
+ local parents=$1
+ local pino=$2
+ local parent_pointer_name=$3
+
+ local found=0
+
+ # Find the entry that has the same inode as the parent
+ # and parse out the entry info
+ while IFS=\/ read PPINO PPGEN PPNAME_LEN PPNAME; do
+ if [ "$PPINO" != "$pino" ]; then
+ continue
+ fi
+
+ if [ "$PPNAME" != "$parent_pointer_name" ]; then
+ continue
+ fi
+
+ found=1
+ break
+ done <<< $(echo "$parents")
+
+ # Check to see if we found anything
+ # We do not fail the test because we also use this
+ # routine to verify when parent pointers should
+ # be removed or updated (ie a rename or a move
+ # operation changes your parent pointer)
+ if [ $found -eq "0" ]; then
+ return 1
+ fi
+
+ # Verify the parent pointer name length is correct
+ if [ "$PPNAME_LEN" -ne "${#parent_pointer_name}" ]
+ then
+ echo "*** Bad parent pointer:"\
+ "name:$PPNAME, namelen:$PPNAME_LEN"
+ fi
+
+ #return sucess
+ return 0
+}
+
+#
+# _verify_parent parent_path parent_pointer_name child_path
+#
+# Verify that the given child path lists the given parent as a parent pointer
+# and that the parent pointer name matches the given name
+#
+# Examples:
+#
+# #simple example
+# mkdir testfolder1
+# touch testfolder1/file1
+# verify_parent testfolder1 file1 testfolder1/file1
+#
+# # In this above example, we want to verify that "testfolder1"
+# # appears as a parent pointer of "testfolder1/file1". Additionally
+# # we verify that the name record of the parent pointer is "file1"
+#
+#
+# #hardlink example
+# mkdir testfolder1
+# mkdir testfolder2
+# touch testfolder1/file1
+# ln testfolder1/file1 testfolder2/file1_ln
+# verify_parent testfolder2 file1_ln testfolder1/file1
+#
+# # In this above example, we want to verify that "testfolder2"
+# # appears as a parent pointer of "testfolder1/file1". Additionally
+# # we verify that the name record of the parent pointer is "file1_ln"
+#
+_verify_parent()
+{
+ local parent_path=$1
+ local parent_pointer_name=$2
+ local child_path=$3
+
+ local parent_ppath="$parent_path/$parent_pointer_name"
+
+ # Verify parent exists
+ if [ ! -d $SCRATCH_MNT/$parent_path ]; then
+ echo "$SCRATCH_MNT/$parent_path not found"
+ else
+ echo "*** $parent_path OK"
+ fi
+
+ # Verify child exists
+ if [ ! -f $SCRATCH_MNT/$child_path ]; then
+ echo "$SCRATCH_MNT/$child_path not found"
+ else
+ echo "*** $child_path OK"
+ fi
+
+ # Verify the parent pointer name exists as a child of the parent
+ if [ ! -f $SCRATCH_MNT/$parent_ppath ]; then
+ echo "$SCRATCH_MNT/$parent_ppath not found"
+ else
+ echo "*** $parent_ppath OK"
+ fi
+
+ # Get the inodes of both parent and child
+ pino="$(stat -c '%i' $SCRATCH_MNT/$parent_path)"
+ cino="$(stat -c '%i' $SCRATCH_MNT/$child_path)"
+
+ # Get all the parent pointers of the child
+ parents=($($XFS_IO_PROG -x -c \
+ "parent -f -i $pino -n $parent_pointer_name" $SCRATCH_MNT/$child_path))
+ if [[ $? != 0 ]]; then
+ echo "No parent pointers found for $child_path"
+ fi
+
+ # Parse parent pointer output.
+ # This sets PPINO PPGEN PPNAME PPNAME_LEN
+ _parse_parent_pointer $parents $pino $parent_pointer_name
+
+ # If we didnt find one, bail out
+ if [ $? -ne 0 ]; then
+ echo "No parent pointer record found for $parent_path"\
+ "in $child_path"
+ fi
+
+ # Verify the inode generated by the parent pointer name is
+ # the same as the child inode
+ pppino="$(stat -c '%i' $SCRATCH_MNT/$parent_ppath)"
+ if [ $cino -ne $pppino ]
+ then
+ echo "Bad parent pointer name value for $child_path."\
+ "$SCRATCH_MNT/$parent_ppath belongs to inode $PPPINO,"\
+ "but should be $cino"
+ fi
+
+ # Make sure path printing works by checking that the paths returned
+ # all point to the same inode.
+ local tgt="$SCRATCH_MNT/$child_path"
+ $XFS_IO_PROG -x -c 'parent -p' "$tgt" | while read pptr_path; do
+ test "$tgt" -ef "$pptr_path" || \
+ echo "$tgt parent pointer $pptr_path should be the same file"
+ done
+
+ echo "*** Verified parent pointer:"\
+ "name:$PPNAME, namelen:$PPNAME_LEN"
+ echo "*** Parent pointer OK for child $child_path"
+}
+
+#
+# _verify_parent parent_pointer_name pino child_path
+#
+# Verify that the given child path contains no parent pointer entry
+# for the given inode and file name
+#
+_verify_no_parent()
+{
+ local parent_pname=$1
+ local pino=$2
+ local child_path=$3
+
+ # Verify child exists
+ if [ ! -f $SCRATCH_MNT/$child_path ]; then
+ echo "$SCRATCH_MNT/$child_path not found"
+ else
+ echo "*** $child_path OK"
+ fi
+
+ # Get all the parent pointers of the child
+ local parents=($($XFS_IO_PROG -x -c \
+ "parent -f -i $pino -n $parent_pname" $SCRATCH_MNT/$child_path))
+ if [[ $? != 0 ]]; then
+ return 0
+ fi
+
+ # Parse parent pointer output.
+ # This sets PPINO PPGEN PPNAME PPNAME_LEN
+ _parse_parent_pointer $parents $pino $parent_pname
+
+ # If we didnt find one, return sucess
+ if [ $? -ne 0 ]; then
+ return 0
+ fi
+
+ echo "Parent pointer entry found where none should:"\
+ "inode:$PPINO, gen:$PPGEN,"
+ "name:$PPNAME, namelen:$PPNAME_LEN"
+}
diff --git a/common/rc b/common/rc
index 0898ac08eb..f760eedc26 100644
--- a/common/rc
+++ b/common/rc
@@ -2656,6 +2656,9 @@ _require_xfs_io_command()
echo $testio | grep -q "invalid option" && \
_notrun "xfs_io $command support is missing"
;;
+ "parent")
+ testio=`$XFS_IO_PROG -x -c "parent" $TEST_DIR 2>&1`
+ ;;
"pwrite")
# -N (RWF_NOWAIT) only works with direct vectored I/O writes
local pwrite_opts=" "
diff --git a/common/xfs b/common/xfs
index 88fa6fb55a..f77d4639b9 100644
--- a/common/xfs
+++ b/common/xfs
@@ -1831,3 +1831,15 @@ _xfs_force_no_pptrs()
MKFS_OPTIONS="$MKFS_OPTIONS -n parent=0"
}
+
+# this test requires the xfs parent pointers feature
+#
+_require_xfs_parent()
+{
+ _scratch_mkfs_xfs_supported -n parent > /dev/null 2>&1 \
+ || _notrun "mkfs.xfs does not support parent pointers"
+ _scratch_mkfs_xfs -n parent > /dev/null 2>&1
+ _try_scratch_mount >/dev/null 2>&1 \
+ || _notrun "kernel does not support parent pointers"
+ _scratch_unmount
+}
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 09/11] xfs: add parent pointer test
2023-12-31 19:59 ` [PATCHSET v13.0 1/3] fstests: adjust tests for xfs parent pointers Darrick J. Wong
` (7 preceding siblings ...)
2023-12-27 13:48 ` [PATCH 08/11] common: add helpers for parent pointer tests Darrick J. Wong
@ 2023-12-27 13:48 ` Darrick J. Wong
2023-12-27 13:48 ` [PATCH 10/11] xfs: add multi link " Darrick J. Wong
2023-12-27 13:49 ` [PATCH 11/11] xfs: add parent pointer inject test Darrick J. Wong
10 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-27 13:48 UTC (permalink / raw)
To: zlang, djwong
Cc: Allison Henderson, Catherine Hoang, fstests, catherine.hoang,
allison.henderson, guan, linux-xfs
From: Allison Henderson <allison.henderson@oracle.com>
Add a test to verify basic parent pointers operations (create, move, link,
unlink, rename, overwrite).
Signed-off-by: Allison Henderson <allison.henderson@oracle.com>
Signed-off-by: Catherine Hoang <catherine.hoang@oracle.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
[djwong: test the xfs_io parent -p argument too]
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
doc/group-names.txt | 1
tests/xfs/1851 | 116 +++++++++++++++++++++++++++++++++++++++++++++++++++
tests/xfs/1851.out | 69 ++++++++++++++++++++++++++++++
3 files changed, 186 insertions(+)
create mode 100755 tests/xfs/1851
create mode 100644 tests/xfs/1851.out
diff --git a/doc/group-names.txt b/doc/group-names.txt
index fec6bf71ab..4676825faf 100644
--- a/doc/group-names.txt
+++ b/doc/group-names.txt
@@ -83,6 +83,7 @@ nfs4_acl NFSv4 access control lists
nonsamefs overlayfs layers on different filesystems
online_repair online repair functionality tests
other dumping ground, do not add more tests to this group
+parent Parent pointer tests
pattern specific IO pattern tests
perms access control and permission checking
pipe pipe functionality
diff --git a/tests/xfs/1851 b/tests/xfs/1851
new file mode 100755
index 0000000000..6cfc7cce79
--- /dev/null
+++ b/tests/xfs/1851
@@ -0,0 +1,116 @@
+#! /bin/bash
+# SPDX-License-Identifier: GPL-2.0
+# Copyright (c) 2022-2024 Oracle and/or its affiliates. All Rights Reserved.
+#
+# FS QA Test No. 1851
+#
+# simple parent pointer test
+#
+
+. ./common/preamble
+_begin_fstest auto quick parent
+
+# get standard environment, filters and checks
+. ./common/parent
+. ./common/filter
+
+# Modify as appropriate
+_supported_fs xfs
+_require_scratch
+_require_xfs_sysfs debug/larp
+_require_xfs_parent
+_require_xfs_io_command "parent"
+
+# real QA test starts here
+
+# Create a directory tree using a protofile and
+# make sure all inodes created have parent pointers
+
+protofile=$tmp.proto
+
+cat >$protofile <<EOF
+DUMMY1
+0 0
+: root directory
+d--777 3 1
+: a directory
+testfolder1 d--755 3 1
+file1 ---755 3 1 /dev/null
+$
+: back in the root
+testfolder2 d--755 3 1
+file2 ---755 3 1 /dev/null
+: done
+$
+EOF
+
+_scratch_mkfs -f -n parent=1 -p $protofile >>$seqres.full 2>&1 \
+ || _fail "mkfs failed"
+_check_scratch_fs
+
+_scratch_mount >>$seqres.full 2>&1 \
+ || _fail "mount failed"
+
+testfolder1="testfolder1"
+testfolder2="testfolder2"
+file1="file1"
+file2="file2"
+file3="file3"
+file1_ln="file1_link"
+
+echo ""
+# Create parent pointer test
+_verify_parent "$testfolder1" "$file1" "$testfolder1/$file1"
+
+echo ""
+# Move parent pointer test
+mv $SCRATCH_MNT/$testfolder1/$file1 $SCRATCH_MNT/$testfolder2/$file1
+_verify_parent "$testfolder2" "$file1" "$testfolder2/$file1"
+
+echo ""
+# Hard link parent pointer test
+ln $SCRATCH_MNT/$testfolder2/$file1 $SCRATCH_MNT/$testfolder1/$file1_ln
+_verify_parent "$testfolder1" "$file1_ln" "$testfolder1/$file1_ln"
+_verify_parent "$testfolder1" "$file1_ln" "$testfolder2/$file1"
+_verify_parent "$testfolder2" "$file1" "$testfolder1/$file1_ln"
+_verify_parent "$testfolder2" "$file1" "$testfolder2/$file1"
+
+echo ""
+# Remove hard link parent pointer test
+ino="$(stat -c '%i' $SCRATCH_MNT/$testfolder2/$file1)"
+rm $SCRATCH_MNT/$testfolder2/$file1
+_verify_parent "$testfolder1" "$file1_ln" "$testfolder1/$file1_ln"
+_verify_no_parent "$file1" "$ino" "$testfolder1/$file1_ln"
+
+echo ""
+# Rename parent pointer test
+ino="$(stat -c '%i' $SCRATCH_MNT/$testfolder1/$file1_ln)"
+mv $SCRATCH_MNT/$testfolder1/$file1_ln $SCRATCH_MNT/$testfolder1/$file2
+_verify_parent "$testfolder1" "$file2" "$testfolder1/$file2"
+_verify_no_parent "$file1_ln" "$ino" "$testfolder1/$file2"
+
+echo ""
+# Over write parent pointer test
+touch $SCRATCH_MNT/$testfolder2/$file3
+_verify_parent "$testfolder2" "$file3" "$testfolder2/$file3"
+ino="$(stat -c '%i' $SCRATCH_MNT/$testfolder2/$file3)"
+mv -f $SCRATCH_MNT/$testfolder2/$file3 $SCRATCH_MNT/$testfolder1/$file2
+_verify_parent "$testfolder1" "$file2" "$testfolder1/$file2"
+
+# Make sure that parent -p filtering works
+mkdir -p $SCRATCH_MNT/dira/ $SCRATCH_MNT/dirb/
+dira_inum=$(stat -c '%i' $SCRATCH_MNT/dira)
+dirb_inum=$(stat -c '%i' $SCRATCH_MNT/dirb)
+touch $SCRATCH_MNT/gorn
+ln $SCRATCH_MNT/gorn $SCRATCH_MNT/dira/file1
+ln $SCRATCH_MNT/gorn $SCRATCH_MNT/dirb/file1
+echo look for both
+$XFS_IO_PROG -c 'parent -p' $SCRATCH_MNT/gorn | _filter_scratch
+echo look for dira
+$XFS_IO_PROG -c 'parent -p -n dira' -c "parent -p -i $dira_inum" $SCRATCH_MNT/gorn | _filter_scratch
+echo look for dirb
+$XFS_IO_PROG -c 'parent -p -n dirb' -c "parent -p -i $dirb_inum" $SCRATCH_MNT/gorn | _filter_scratch
+
+# success, all done
+status=0
+exit
diff --git a/tests/xfs/1851.out b/tests/xfs/1851.out
new file mode 100644
index 0000000000..99a9d42892
--- /dev/null
+++ b/tests/xfs/1851.out
@@ -0,0 +1,69 @@
+QA output created by 1851
+
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1
+
+*** testfolder2 OK
+*** testfolder2/file1 OK
+*** testfolder2/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder2/file1
+
+*** testfolder1 OK
+*** testfolder1/file1_link OK
+*** testfolder1/file1_link OK
+*** Verified parent pointer: name:file1_link, namelen:10
+*** Parent pointer OK for child testfolder1/file1_link
+*** testfolder1 OK
+*** testfolder2/file1 OK
+*** testfolder1/file1_link OK
+*** Verified parent pointer: name:file1_link, namelen:10
+*** Parent pointer OK for child testfolder2/file1
+*** testfolder2 OK
+*** testfolder1/file1_link OK
+*** testfolder2/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link
+*** testfolder2 OK
+*** testfolder2/file1 OK
+*** testfolder2/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder2/file1
+
+*** testfolder1 OK
+*** testfolder1/file1_link OK
+*** testfolder1/file1_link OK
+*** Verified parent pointer: name:file1_link, namelen:10
+*** Parent pointer OK for child testfolder1/file1_link
+*** testfolder1/file1_link OK
+
+*** testfolder1 OK
+*** testfolder1/file2 OK
+*** testfolder1/file2 OK
+*** Verified parent pointer: name:file2, namelen:5
+*** Parent pointer OK for child testfolder1/file2
+*** testfolder1/file2 OK
+
+*** testfolder2 OK
+*** testfolder2/file3 OK
+*** testfolder2/file3 OK
+*** Verified parent pointer: name:file3, namelen:5
+*** Parent pointer OK for child testfolder2/file3
+*** testfolder1 OK
+*** testfolder1/file2 OK
+*** testfolder1/file2 OK
+*** Verified parent pointer: name:file2, namelen:5
+*** Parent pointer OK for child testfolder1/file2
+look for both
+SCRATCH_MNT/gorn
+SCRATCH_MNT/dira/file1
+SCRATCH_MNT/dirb/file1
+look for dira
+SCRATCH_MNT/dira/file1
+SCRATCH_MNT/dira/file1
+look for dirb
+SCRATCH_MNT/dirb/file1
+SCRATCH_MNT/dirb/file1
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 10/11] xfs: add multi link parent pointer test
2023-12-31 19:59 ` [PATCHSET v13.0 1/3] fstests: adjust tests for xfs parent pointers Darrick J. Wong
` (8 preceding siblings ...)
2023-12-27 13:48 ` [PATCH 09/11] xfs: add parent pointer test Darrick J. Wong
@ 2023-12-27 13:48 ` Darrick J. Wong
2023-12-27 13:49 ` [PATCH 11/11] xfs: add parent pointer inject test Darrick J. Wong
10 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-27 13:48 UTC (permalink / raw)
To: zlang, djwong
Cc: Allison Henderson, Catherine Hoang, fstests, catherine.hoang,
allison.henderson, guan, linux-xfs
From: Allison Henderson <allison.henderson@oracle.com>
Add a test to verify parent pointers while multiple links to a file are
created and removed.
Signed-off-by: Allison Henderson <allison.henderson@oracle.com>
Signed-off-by: Catherine Hoang <catherine.hoang@oracle.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
tests/xfs/1852 | 69 ++++
tests/xfs/1852.out | 1002 ++++++++++++++++++++++++++++++++++++++++++++++++++++
2 files changed, 1071 insertions(+)
create mode 100755 tests/xfs/1852
create mode 100644 tests/xfs/1852.out
diff --git a/tests/xfs/1852 b/tests/xfs/1852
new file mode 100755
index 0000000000..a5a9c75552
--- /dev/null
+++ b/tests/xfs/1852
@@ -0,0 +1,69 @@
+#! /bin/bash
+# SPDX-License-Identifier: GPL-2.0
+# Copyright (c) 2022-2024 Oracle and/or its affiliates. All Rights Reserved.
+#
+# FS QA Test No. 1852
+#
+# multi link parent pointer test
+#
+. ./common/preamble
+_begin_fstest auto quick parent
+
+# get standard environment, filters and checks
+. ./common/parent
+
+# Modify as appropriate
+_supported_fs xfs
+_require_scratch
+_require_xfs_sysfs debug/larp
+_require_xfs_parent
+_require_xfs_io_command "parent"
+
+# real QA test starts here
+
+# Create a directory tree using a protofile and
+# make sure all inodes created have parent pointers
+
+protofile=$tmp.proto
+
+cat >$protofile <<EOF
+DUMMY1
+0 0
+: root directory
+d--777 3 1
+: a directory
+testfolder1 d--755 3 1
+file1 ---755 3 1 /dev/null
+: done
+$
+EOF
+
+_scratch_mkfs -f -n parent=1 -p $protofile >>$seqresres.full 2>&1 \
+ || _fail "mkfs failed"
+_check_scratch_fs
+
+_scratch_mount >>$seqres.full 2>&1 \
+ || _fail "mount failed"
+
+testfolder1="testfolder1"
+file1="file1"
+file1_ln="file1_link"
+
+echo ""
+# Multi link parent pointer test
+NLINKS=100
+for (( j=0; j<$NLINKS; j++ )); do
+ ln $SCRATCH_MNT/$testfolder1/$file1 $SCRATCH_MNT/$testfolder1/$file1_ln.$j
+ _verify_parent "$testfolder1" "$file1_ln.$j" "$testfolder1/$file1"
+ _verify_parent "$testfolder1" "$file1" "$testfolder1/$file1_ln.$j"
+done
+# Multi unlink parent pointer test
+for (( j=$NLINKS-1; j<=0; j-- )); do
+ ino="$(stat -c '%i' $SCRATCH_MNT/$testfolder1/$file1_ln.$j)"
+ rm $SCRATCH_MNT/$testfolder1/$file1_ln.$j
+ _verify_no_parent "$file1_ln.$j" "$ino" "$testfolder1/$file1"
+done
+
+# success, all done
+status=0
+exit
diff --git a/tests/xfs/1852.out b/tests/xfs/1852.out
new file mode 100644
index 0000000000..a13664d481
--- /dev/null
+++ b/tests/xfs/1852.out
@@ -0,0 +1,1002 @@
+QA output created by 1852
+
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.0 OK
+*** Verified parent pointer: name:file1_link.0, namelen:12
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.0 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.0
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.1 OK
+*** Verified parent pointer: name:file1_link.1, namelen:12
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.1 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.1
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.2 OK
+*** Verified parent pointer: name:file1_link.2, namelen:12
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.2 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.2
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.3 OK
+*** Verified parent pointer: name:file1_link.3, namelen:12
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.3 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.3
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.4 OK
+*** Verified parent pointer: name:file1_link.4, namelen:12
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.4 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.4
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.5 OK
+*** Verified parent pointer: name:file1_link.5, namelen:12
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.5 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.5
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.6 OK
+*** Verified parent pointer: name:file1_link.6, namelen:12
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.6 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.6
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.7 OK
+*** Verified parent pointer: name:file1_link.7, namelen:12
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.7 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.7
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.8 OK
+*** Verified parent pointer: name:file1_link.8, namelen:12
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.8 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.8
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.9 OK
+*** Verified parent pointer: name:file1_link.9, namelen:12
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.9 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.9
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.10 OK
+*** Verified parent pointer: name:file1_link.10, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.10 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.10
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.11 OK
+*** Verified parent pointer: name:file1_link.11, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.11 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.11
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.12 OK
+*** Verified parent pointer: name:file1_link.12, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.12 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.12
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.13 OK
+*** Verified parent pointer: name:file1_link.13, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.13 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.13
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.14 OK
+*** Verified parent pointer: name:file1_link.14, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.14 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.14
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.15 OK
+*** Verified parent pointer: name:file1_link.15, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.15 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.15
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.16 OK
+*** Verified parent pointer: name:file1_link.16, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.16 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.16
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.17 OK
+*** Verified parent pointer: name:file1_link.17, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.17 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.17
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.18 OK
+*** Verified parent pointer: name:file1_link.18, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.18 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.18
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.19 OK
+*** Verified parent pointer: name:file1_link.19, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.19 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.19
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.20 OK
+*** Verified parent pointer: name:file1_link.20, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.20 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.20
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.21 OK
+*** Verified parent pointer: name:file1_link.21, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.21 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.21
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.22 OK
+*** Verified parent pointer: name:file1_link.22, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.22 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.22
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.23 OK
+*** Verified parent pointer: name:file1_link.23, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.23 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.23
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.24 OK
+*** Verified parent pointer: name:file1_link.24, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.24 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.24
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.25 OK
+*** Verified parent pointer: name:file1_link.25, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.25 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.25
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.26 OK
+*** Verified parent pointer: name:file1_link.26, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.26 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.26
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.27 OK
+*** Verified parent pointer: name:file1_link.27, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.27 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.27
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.28 OK
+*** Verified parent pointer: name:file1_link.28, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.28 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.28
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.29 OK
+*** Verified parent pointer: name:file1_link.29, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.29 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.29
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.30 OK
+*** Verified parent pointer: name:file1_link.30, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.30 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.30
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.31 OK
+*** Verified parent pointer: name:file1_link.31, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.31 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.31
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.32 OK
+*** Verified parent pointer: name:file1_link.32, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.32 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.32
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.33 OK
+*** Verified parent pointer: name:file1_link.33, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.33 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.33
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.34 OK
+*** Verified parent pointer: name:file1_link.34, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.34 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.34
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.35 OK
+*** Verified parent pointer: name:file1_link.35, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.35 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.35
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.36 OK
+*** Verified parent pointer: name:file1_link.36, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.36 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.36
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.37 OK
+*** Verified parent pointer: name:file1_link.37, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.37 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.37
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.38 OK
+*** Verified parent pointer: name:file1_link.38, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.38 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.38
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.39 OK
+*** Verified parent pointer: name:file1_link.39, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.39 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.39
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.40 OK
+*** Verified parent pointer: name:file1_link.40, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.40 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.40
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.41 OK
+*** Verified parent pointer: name:file1_link.41, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.41 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.41
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.42 OK
+*** Verified parent pointer: name:file1_link.42, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.42 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.42
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.43 OK
+*** Verified parent pointer: name:file1_link.43, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.43 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.43
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.44 OK
+*** Verified parent pointer: name:file1_link.44, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.44 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.44
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.45 OK
+*** Verified parent pointer: name:file1_link.45, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.45 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.45
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.46 OK
+*** Verified parent pointer: name:file1_link.46, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.46 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.46
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.47 OK
+*** Verified parent pointer: name:file1_link.47, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.47 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.47
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.48 OK
+*** Verified parent pointer: name:file1_link.48, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.48 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.48
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.49 OK
+*** Verified parent pointer: name:file1_link.49, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.49 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.49
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.50 OK
+*** Verified parent pointer: name:file1_link.50, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.50 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.50
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.51 OK
+*** Verified parent pointer: name:file1_link.51, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.51 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.51
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.52 OK
+*** Verified parent pointer: name:file1_link.52, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.52 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.52
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.53 OK
+*** Verified parent pointer: name:file1_link.53, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.53 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.53
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.54 OK
+*** Verified parent pointer: name:file1_link.54, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.54 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.54
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.55 OK
+*** Verified parent pointer: name:file1_link.55, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.55 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.55
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.56 OK
+*** Verified parent pointer: name:file1_link.56, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.56 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.56
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.57 OK
+*** Verified parent pointer: name:file1_link.57, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.57 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.57
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.58 OK
+*** Verified parent pointer: name:file1_link.58, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.58 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.58
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.59 OK
+*** Verified parent pointer: name:file1_link.59, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.59 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.59
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.60 OK
+*** Verified parent pointer: name:file1_link.60, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.60 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.60
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.61 OK
+*** Verified parent pointer: name:file1_link.61, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.61 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.61
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.62 OK
+*** Verified parent pointer: name:file1_link.62, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.62 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.62
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.63 OK
+*** Verified parent pointer: name:file1_link.63, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.63 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.63
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.64 OK
+*** Verified parent pointer: name:file1_link.64, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.64 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.64
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.65 OK
+*** Verified parent pointer: name:file1_link.65, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.65 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.65
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.66 OK
+*** Verified parent pointer: name:file1_link.66, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.66 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.66
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.67 OK
+*** Verified parent pointer: name:file1_link.67, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.67 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.67
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.68 OK
+*** Verified parent pointer: name:file1_link.68, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.68 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.68
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.69 OK
+*** Verified parent pointer: name:file1_link.69, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.69 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.69
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.70 OK
+*** Verified parent pointer: name:file1_link.70, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.70 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.70
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.71 OK
+*** Verified parent pointer: name:file1_link.71, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.71 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.71
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.72 OK
+*** Verified parent pointer: name:file1_link.72, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.72 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.72
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.73 OK
+*** Verified parent pointer: name:file1_link.73, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.73 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.73
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.74 OK
+*** Verified parent pointer: name:file1_link.74, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.74 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.74
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.75 OK
+*** Verified parent pointer: name:file1_link.75, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.75 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.75
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.76 OK
+*** Verified parent pointer: name:file1_link.76, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.76 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.76
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.77 OK
+*** Verified parent pointer: name:file1_link.77, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.77 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.77
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.78 OK
+*** Verified parent pointer: name:file1_link.78, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.78 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.78
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.79 OK
+*** Verified parent pointer: name:file1_link.79, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.79 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.79
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.80 OK
+*** Verified parent pointer: name:file1_link.80, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.80 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.80
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.81 OK
+*** Verified parent pointer: name:file1_link.81, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.81 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.81
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.82 OK
+*** Verified parent pointer: name:file1_link.82, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.82 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.82
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.83 OK
+*** Verified parent pointer: name:file1_link.83, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.83 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.83
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.84 OK
+*** Verified parent pointer: name:file1_link.84, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.84 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.84
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.85 OK
+*** Verified parent pointer: name:file1_link.85, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.85 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.85
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.86 OK
+*** Verified parent pointer: name:file1_link.86, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.86 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.86
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.87 OK
+*** Verified parent pointer: name:file1_link.87, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.87 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.87
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.88 OK
+*** Verified parent pointer: name:file1_link.88, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.88 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.88
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.89 OK
+*** Verified parent pointer: name:file1_link.89, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.89 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.89
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.90 OK
+*** Verified parent pointer: name:file1_link.90, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.90 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.90
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.91 OK
+*** Verified parent pointer: name:file1_link.91, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.91 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.91
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.92 OK
+*** Verified parent pointer: name:file1_link.92, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.92 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.92
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.93 OK
+*** Verified parent pointer: name:file1_link.93, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.93 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.93
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.94 OK
+*** Verified parent pointer: name:file1_link.94, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.94 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.94
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.95 OK
+*** Verified parent pointer: name:file1_link.95, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.95 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.95
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.96 OK
+*** Verified parent pointer: name:file1_link.96, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.96 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.96
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.97 OK
+*** Verified parent pointer: name:file1_link.97, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.97 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.97
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.98 OK
+*** Verified parent pointer: name:file1_link.98, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.98 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.98
+*** testfolder1 OK
+*** testfolder1/file1 OK
+*** testfolder1/file1_link.99 OK
+*** Verified parent pointer: name:file1_link.99, namelen:13
+*** Parent pointer OK for child testfolder1/file1
+*** testfolder1 OK
+*** testfolder1/file1_link.99 OK
+*** testfolder1/file1 OK
+*** Verified parent pointer: name:file1, namelen:5
+*** Parent pointer OK for child testfolder1/file1_link.99
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 11/11] xfs: add parent pointer inject test
2023-12-31 19:59 ` [PATCHSET v13.0 1/3] fstests: adjust tests for xfs parent pointers Darrick J. Wong
` (9 preceding siblings ...)
2023-12-27 13:48 ` [PATCH 10/11] xfs: add multi link " Darrick J. Wong
@ 2023-12-27 13:49 ` Darrick J. Wong
10 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-27 13:49 UTC (permalink / raw)
To: zlang, djwong
Cc: Allison Henderson, Catherine Hoang, fstests, catherine.hoang,
allison.henderson, guan, linux-xfs
From: Allison Henderson <allison.henderson@oracle.com>
Add a test to verify parent pointers after an error injection and log
replay.
Signed-off-by: Allison Henderson <allison.henderson@oracle.com>
Signed-off-by: Catherine Hoang <catherine.hoang@oracle.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
tests/xfs/1853 | 85 ++++++++++++++++++++++++++++++++++++++++++++++++++++
tests/xfs/1853.out | 14 +++++++++
2 files changed, 99 insertions(+)
create mode 100755 tests/xfs/1853
create mode 100644 tests/xfs/1853.out
diff --git a/tests/xfs/1853 b/tests/xfs/1853
new file mode 100755
index 0000000000..d5a5050e00
--- /dev/null
+++ b/tests/xfs/1853
@@ -0,0 +1,85 @@
+#! /bin/bash
+# SPDX-License-Identifier: GPL-2.0
+# Copyright (c) 2022-2024 Oracle and/or its affiliates. All Rights Reserved.
+#
+# FS QA Test No. 1853
+#
+# parent pointer inject test
+#
+. ./common/preamble
+_begin_fstest auto quick parent
+
+# get standard environment, filters and checks
+. ./common/filter
+. ./common/inject
+. ./common/parent
+
+# Modify as appropriate
+_supported_fs xfs
+_require_scratch
+_require_xfs_sysfs debug/larp
+_require_xfs_io_error_injection "larp"
+_require_xfs_parent
+_require_xfs_io_command "parent"
+
+# real QA test starts here
+
+# Create a directory tree using a protofile and
+# make sure all inodes created have parent pointers
+
+protofile=$tmp.proto
+
+cat >$protofile <<EOF
+DUMMY1
+0 0
+: root directory
+d--777 3 1
+: a directory
+testfolder1 d--755 3 1
+$
+: back in the root
+testfolder2 d--755 3 1
+: done
+$
+EOF
+
+_scratch_mkfs -f -n parent=1 -p $protofile >>$seqres.full 2>&1 \
+ || _fail "mkfs failed"
+_check_scratch_fs
+
+_scratch_mount >>$seqres.full 2>&1 \
+ || _fail "mount failed"
+
+testfolder1="testfolder1"
+testfolder2="testfolder2"
+file4="file4"
+file5="file5"
+
+echo ""
+
+# Create files
+touch $SCRATCH_MNT/$testfolder1/$file4
+_verify_parent "$testfolder1" "$file4" "$testfolder1/$file4"
+
+# Inject error
+_scratch_inject_error "larp"
+
+# Move files
+mv $SCRATCH_MNT/$testfolder1/$file4 $SCRATCH_MNT/$testfolder2/$file5 2>&1 \
+ | _filter_scratch
+
+# FS should be shut down, touch will fail
+touch $SCRATCH_MNT/$testfolder2/$file5 2>&1 | _filter_scratch
+
+# Remount to replay log
+_scratch_remount_dump_log >> $seqres.full
+
+# FS should be online, touch should succeed
+touch $SCRATCH_MNT/$testfolder2/$file5
+
+# Check files again
+_verify_parent "$testfolder2" "$file5" "$testfolder2/$file5"
+
+# success, all done
+status=0
+exit
diff --git a/tests/xfs/1853.out b/tests/xfs/1853.out
new file mode 100644
index 0000000000..736f6dec00
--- /dev/null
+++ b/tests/xfs/1853.out
@@ -0,0 +1,14 @@
+QA output created by 1853
+
+*** testfolder1 OK
+*** testfolder1/file4 OK
+*** testfolder1/file4 OK
+*** Verified parent pointer: name:file4, namelen:5
+*** Parent pointer OK for child testfolder1/file4
+mv: cannot stat 'SCRATCH_MNT/testfolder1/file4': Input/output error
+touch: cannot touch 'SCRATCH_MNT/testfolder2/file5': Input/output error
+*** testfolder2 OK
+*** testfolder2/file5 OK
+*** testfolder2/file5 OK
+*** Verified parent pointer: name:file5, namelen:5
+*** Parent pointer OK for child testfolder2/file5
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 1/2] common/fuzzy: stress directory tree modifications with the dirtree tester
2023-12-31 19:59 ` [PATCHSET v13.0 2/3] xfs: detect and correct directory tree structures Darrick J. Wong
@ 2023-12-27 13:49 ` Darrick J. Wong
2023-12-27 13:49 ` [PATCH 2/2] scrub: test correction of directory tree corruptions Darrick J. Wong
1 sibling, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-27 13:49 UTC (permalink / raw)
To: zlang, djwong; +Cc: linux-xfs, guan, fstests
From: Darrick J. Wong <djwong@kernel.org>
Stress test the directory tree corruption detector by racing it with
fsstress.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
tests/xfs/1864 | 38 ++++++++++++++++++++++++++++++++++++++
tests/xfs/1864.out | 2 ++
tests/xfs/1865 | 38 ++++++++++++++++++++++++++++++++++++++
tests/xfs/1865.out | 2 ++
4 files changed, 80 insertions(+)
create mode 100755 tests/xfs/1864
create mode 100644 tests/xfs/1864.out
create mode 100755 tests/xfs/1865
create mode 100644 tests/xfs/1865.out
diff --git a/tests/xfs/1864 b/tests/xfs/1864
new file mode 100755
index 0000000000..d00bcb28b4
--- /dev/null
+++ b/tests/xfs/1864
@@ -0,0 +1,38 @@
+#! /bin/bash
+# SPDX-License-Identifier: GPL-2.0
+# Copyright (c) 2023-2024 Oracle. All Rights Reserved.
+#
+# FS QA Test No. 1864
+#
+# Race fsstress and directory tree structure corruption detector for a while to
+# see if we crash or livelock.
+#
+. ./common/preamble
+_begin_fstest scrub dangerous_fsstress_scrub
+
+_cleanup() {
+ _scratch_xfs_stress_scrub_cleanup &> /dev/null
+ cd /
+ rm -r -f $tmp.*
+}
+_register_cleanup "_cleanup" BUS
+
+# Import common functions.
+. ./common/filter
+. ./common/fuzzy
+. ./common/inject
+. ./common/xfs
+
+# real QA test starts here
+_supported_fs xfs
+_require_scratch
+_require_xfs_stress_scrub
+
+_scratch_mkfs > "$seqres.full" 2>&1
+_scratch_mount
+_scratch_xfs_stress_scrub -x 'dir' -s "scrub dirtree" -t "%dir%"
+
+# success, all done
+echo Silence is golden
+status=0
+exit
diff --git a/tests/xfs/1864.out b/tests/xfs/1864.out
new file mode 100644
index 0000000000..472f56323a
--- /dev/null
+++ b/tests/xfs/1864.out
@@ -0,0 +1,2 @@
+QA output created by 1864
+Silence is golden
diff --git a/tests/xfs/1865 b/tests/xfs/1865
new file mode 100755
index 0000000000..098891536c
--- /dev/null
+++ b/tests/xfs/1865
@@ -0,0 +1,38 @@
+#! /bin/bash
+# SPDX-License-Identifier: GPL-2.0
+# Copyright (c) 2023-2024 Oracle. All Rights Reserved.
+#
+# FS QA Test No. 1865
+#
+# Race fsstress and directory tree structure repair for a while to see if we
+# crash or livelock.
+#
+. ./common/preamble
+_begin_fstest online_repair dangerous_fsstress_repair
+
+_cleanup() {
+ _scratch_xfs_stress_scrub_cleanup &> /dev/null
+ cd /
+ rm -r -f $tmp.*
+}
+_register_cleanup "_cleanup" BUS
+
+# Import common functions.
+. ./common/filter
+. ./common/fuzzy
+. ./common/inject
+. ./common/xfs
+
+# real QA test starts here
+_supported_fs xfs
+_require_scratch
+_require_xfs_stress_online_repair
+
+_scratch_mkfs > "$seqres.full" 2>&1
+_scratch_mount
+_scratch_xfs_stress_online_repair -x 'dir' -s "repair dirtree" -t "%dir%"
+
+# success, all done
+echo Silence is golden
+status=0
+exit
diff --git a/tests/xfs/1865.out b/tests/xfs/1865.out
new file mode 100644
index 0000000000..9f2fecad3f
--- /dev/null
+++ b/tests/xfs/1865.out
@@ -0,0 +1,2 @@
+QA output created by 1865
+Silence is golden
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 2/2] scrub: test correction of directory tree corruptions
2023-12-31 19:59 ` [PATCHSET v13.0 2/3] xfs: detect and correct directory tree structures Darrick J. Wong
2023-12-27 13:49 ` [PATCH 1/2] common/fuzzy: stress directory tree modifications with the dirtree tester Darrick J. Wong
@ 2023-12-27 13:49 ` Darrick J. Wong
1 sibling, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-27 13:49 UTC (permalink / raw)
To: zlang, djwong; +Cc: linux-xfs, guan, fstests
From: Darrick J. Wong <djwong@kernel.org>
Make sure that we can fix directory tree loops and multiply-owned dirs.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
tests/xfs/1866 | 122 ++++++++++++++++++++++++++++++++++++++++
tests/xfs/1866.out | 19 ++++++
tests/xfs/1867 | 133 ++++++++++++++++++++++++++++++++++++++++++++
tests/xfs/1867.out | 25 ++++++++
tests/xfs/1868 | 121 ++++++++++++++++++++++++++++++++++++++++
tests/xfs/1868.out | 21 +++++++
tests/xfs/1869 | 157 ++++++++++++++++++++++++++++++++++++++++++++++++++++
tests/xfs/1869.out | 32 +++++++++++
tests/xfs/1870 | 146 ++++++++++++++++++++++++++++++++++++++++++++++++
tests/xfs/1870.out | 30 ++++++++++
tests/xfs/1871 | 78 ++++++++++++++++++++++++++
tests/xfs/1871.out | 2 +
12 files changed, 886 insertions(+)
create mode 100755 tests/xfs/1866
create mode 100644 tests/xfs/1866.out
create mode 100755 tests/xfs/1867
create mode 100644 tests/xfs/1867.out
create mode 100755 tests/xfs/1868
create mode 100644 tests/xfs/1868.out
create mode 100755 tests/xfs/1869
create mode 100644 tests/xfs/1869.out
create mode 100755 tests/xfs/1870
create mode 100644 tests/xfs/1870.out
create mode 100755 tests/xfs/1871
create mode 100644 tests/xfs/1871.out
diff --git a/tests/xfs/1866 b/tests/xfs/1866
new file mode 100755
index 0000000000..280c33da3e
--- /dev/null
+++ b/tests/xfs/1866
@@ -0,0 +1,122 @@
+#! /bin/bash
+# SPDX-License-Identifier: GPL-2.0
+# Copyright (c) 2023-2024 Oracle. All Rights Reserved.
+#
+# FS QA Test No. 1866
+#
+# Functional testing for online fsck of a directory loop that is not accessible
+# from the root directory.
+#
+. ./common/preamble
+_begin_fstest auto online_repair
+
+# Import common functions.
+. ./common/filter
+. ./common/inject
+. ./common/fuzzy
+. ./common/populate
+
+# real QA test starts here
+
+# Modify as appropriate.
+_supported_fs xfs
+_require_xfs_db_command "link"
+_require_xfs_db_command "unlink"
+_require_scratch
+_require_xfs_stress_online_repair
+
+prepare_fs() {
+ _scratch_mkfs >> $seqres.full
+ _scratch_mount
+ __stress_scrub_check_commands "%dir%" '' '' 'scrub dirtree'
+
+ # Begin by creating the following directory tree:
+ # root["A"]->A
+ # A["B"]->B
+ # B["C"]->C
+ mkdir -p "$SCRATCH_MNT/A/B/C"
+
+ root_inum="$(stat -c '%i' "$SCRATCH_MNT/")"
+ a_inum="$(stat -c '%i' "$SCRATCH_MNT/A")"
+ b_inum="$(stat -c '%i' "$SCRATCH_MNT/A/B")"
+ c_inum="$(stat -c '%i' "$SCRATCH_MNT/A/B/C")"
+
+ echo "root: $root_inum; a: $a_inum; b: $b_inum; c: $c_inum" >> $seqres.full
+
+ # Next, we complete the loop by creating C["A"]->A and deleting root["A"]->A.
+ # Directory tree is now:
+ # A["B"]->B
+ # B["C"]->C
+ # C["A"]->A
+ _scratch_unmount
+
+ root_gen=$(_scratch_xfs_get_metadata_field core.gen "inode $root_inum")
+ a_gen=$(_scratch_xfs_get_metadata_field core.gen "inode $a_inum")
+ b_gen=$(_scratch_xfs_get_metadata_field core.gen "inode $b_inum")
+ c_gen=$(_scratch_xfs_get_metadata_field core.gen "inode $c_inum")
+
+ _scratch_xfs_db \
+ -c "echo before root $root_inum" -c "inode $root_inum" -c 'print core.nlinkv2' -c "ls" \
+ -c "echo before C $c_inum" -c "inode $c_inum" -c 'print core.nlinkv2' -c "ls" \
+ -c "echo before A $a_inum" -c "inode $a_inum" -c 'print core.nlinkv2' -c "parent" \
+ >> $seqres.full
+
+ _scratch_xfs_db -x \
+ -c "inode $c_inum" -c "link -i $a_inum A" \
+ -c "inode $root_inum" -c "unlink A" \
+ >> $seqres.full
+
+ _scratch_xfs_db \
+ -c "echo after root $root_inum" -c "inode $root_inum" -c 'print core.nlinkv2' -c "ls" \
+ -c "echo after C $c_inum" -c "inode $c_inum" -c 'print core.nlinkv2' -c "ls" \
+ -c "echo after A $a_inum" -c "inode $a_inum" -c 'print core.nlinkv2' -c "parent" \
+ >> $seqres.full
+}
+
+simple_online_repair() {
+ echo "check root"
+ $XFS_IO_PROG -c "scrub dirtree $root_inum $root_gen" $SCRATCH_MNT
+ echo "check A"
+ $XFS_IO_PROG -c "scrub dirtree $a_inum $a_gen" $SCRATCH_MNT
+ echo "check B"
+ $XFS_IO_PROG -c "scrub dirtree $b_inum $b_gen" $SCRATCH_MNT
+ echo "check C"
+ $XFS_IO_PROG -c "scrub dirtree $c_inum $c_gen" $SCRATCH_MNT
+
+ echo "repair root"
+ $XFS_IO_PROG -x -c "repair dirtree $root_inum $root_gen" $SCRATCH_MNT
+ echo "repair A"
+ $XFS_IO_PROG -x -c "repair dirtree $a_inum $a_gen" $SCRATCH_MNT
+ echo "repair B"
+ $XFS_IO_PROG -x -c "repair dirtree $b_inum $b_gen" $SCRATCH_MNT
+ echo "repair C"
+ $XFS_IO_PROG -x -c "repair dirtree $c_inum $c_gen" $SCRATCH_MNT
+
+ echo "check root"
+ $XFS_IO_PROG -c "scrub dirtree $root_inum $root_gen" $SCRATCH_MNT
+ echo "check A"
+ $XFS_IO_PROG -c "scrub dirtree $a_inum $a_gen" $SCRATCH_MNT
+ echo "check B"
+ $XFS_IO_PROG -c "scrub dirtree $b_inum $b_gen" $SCRATCH_MNT
+ echo "check C"
+ $XFS_IO_PROG -c "scrub dirtree $c_inum $c_gen" $SCRATCH_MNT
+}
+
+# Part 1: Use raw ioctls to detect the loop and fix it.
+prepare_fs
+_scratch_mount
+simple_online_repair
+_check_scratch_fs
+_scratch_unmount
+
+# Part 2: Use xfs_scrub to detect the loop and fix it.
+prepare_fs
+_scratch_mount
+_scratch_scrub &>> $seqres.full
+echo "xfs_scrub returned $?" >> $seqres.full
+_check_scratch_fs
+_scratch_unmount
+
+# success, all done
+status=0
+exit
diff --git a/tests/xfs/1866.out b/tests/xfs/1866.out
new file mode 100644
index 0000000000..b6b08aea7f
--- /dev/null
+++ b/tests/xfs/1866.out
@@ -0,0 +1,19 @@
+QA output created by 1866
+check root
+check A
+Corruption detected.
+check B
+Corruption detected.
+check C
+Corruption detected.
+repair root
+Metadata did not need repair or optimization.
+repair A
+repair B
+Metadata did not need repair or optimization.
+repair C
+Metadata did not need repair or optimization.
+check root
+check A
+check B
+check C
diff --git a/tests/xfs/1867 b/tests/xfs/1867
new file mode 100755
index 0000000000..2c34b56503
--- /dev/null
+++ b/tests/xfs/1867
@@ -0,0 +1,133 @@
+#! /bin/bash
+# SPDX-License-Identifier: GPL-2.0
+# Copyright (c) 2023-2024 Oracle. All Rights Reserved.
+#
+# FS QA Test No. 1867
+#
+# Functional testing for online fsck of a directory loop that is accessible
+# from the root directory.
+#
+. ./common/preamble
+_begin_fstest auto online_repair
+
+# Import common functions.
+. ./common/filter
+. ./common/inject
+. ./common/fuzzy
+. ./common/populate
+
+# real QA test starts here
+
+# Modify as appropriate.
+_supported_fs xfs
+_require_xfs_db_command "link"
+_require_xfs_db_command "unlink"
+_require_scratch
+_require_xfs_stress_online_repair
+
+prepare_fs() {
+ _scratch_mkfs >> $seqres.full
+ _scratch_mount
+ __stress_scrub_check_commands "%dir%" '' '' 'scrub dirtree'
+
+ # Begin by creating the following directory tree:
+ # root["A"]->A
+ # A["B"]->B
+ # B["C"]->C
+ # C["D"]->D
+ mkdir -p "$SCRATCH_MNT/A/B/C/D"
+
+ root_inum="$(stat -c '%i' "$SCRATCH_MNT/")"
+ a_inum="$(stat -c '%i' "$SCRATCH_MNT/A")"
+ b_inum="$(stat -c '%i' "$SCRATCH_MNT/A/B")"
+ c_inum="$(stat -c '%i' "$SCRATCH_MNT/A/B/C")"
+ d_inum="$(stat -c '%i' "$SCRATCH_MNT/A/B/C/D")"
+
+ echo "root: $root_inum; a: $a_inum; b: $b_inum; c: $c_inum; d: $d_inum" >> $seqres.full
+
+ # Next, we complete the loop by creating D["B1"]->B. Directory tree is now:
+ # root["A"]->A
+ # A["B"]->B
+ # B["C"]->C
+ # C["D"]->D
+ # D["B1"]->B
+ _scratch_unmount
+
+ root_gen=$(_scratch_xfs_get_metadata_field core.gen "inode $root_inum")
+ a_gen=$(_scratch_xfs_get_metadata_field core.gen "inode $a_inum")
+ b_gen=$(_scratch_xfs_get_metadata_field core.gen "inode $b_inum")
+ c_gen=$(_scratch_xfs_get_metadata_field core.gen "inode $c_inum")
+ d_gen=$(_scratch_xfs_get_metadata_field core.gen "inode $d_inum")
+
+ _scratch_xfs_db \
+ -c "echo before root $root_inum" -c "inode $root_inum" -c 'print core.nlinkv2' -c "ls" \
+ -c "echo before D $d_inum" -c "inode $d_inum" -c 'print core.nlinkv2' -c "ls" \
+ -c "echo before B $b_inum" -c "inode $b_inum" -c 'print core.nlinkv2' -c "parent" \
+ >> $seqres.full
+
+ _scratch_xfs_db -x \
+ -c "inode $d_inum" -c "link -i $b_inum B1" \
+ >> $seqres.full
+
+ _scratch_xfs_db \
+ -c "echo after root $root_inum" -c "inode $root_inum" -c 'print core.nlinkv2' -c "ls" \
+ -c "echo after D $c_inum" -c "inode $d_inum" -c 'print core.nlinkv2' -c "ls" \
+ -c "echo after B $a_inum" -c "inode $b_inum" -c 'print core.nlinkv2' -c "parent" \
+ >> $seqres.full
+}
+
+simple_online_repair() {
+ echo "check root"
+ $XFS_IO_PROG -c "scrub dirtree $root_inum $root_gen" $SCRATCH_MNT
+ echo "check A"
+ $XFS_IO_PROG -c "scrub dirtree $a_inum $a_gen" $SCRATCH_MNT
+ echo "check B"
+ $XFS_IO_PROG -c "scrub dirtree $b_inum $b_gen" $SCRATCH_MNT
+ echo "check C"
+ $XFS_IO_PROG -c "scrub dirtree $c_inum $c_gen" $SCRATCH_MNT
+ echo "check D"
+ $XFS_IO_PROG -c "scrub dirtree $d_inum $d_gen" $SCRATCH_MNT
+
+ echo "repair root"
+ $XFS_IO_PROG -x -c "repair dirtree $root_inum $root_gen" $SCRATCH_MNT
+ echo "repair A"
+ $XFS_IO_PROG -x -c "repair dirtree $a_inum $a_gen" $SCRATCH_MNT
+ echo "repair D"
+ $XFS_IO_PROG -x -c "repair dirtree $d_inum $d_gen" $SCRATCH_MNT
+ echo "repair B"
+ $XFS_IO_PROG -x -c "repair dirtree $b_inum $b_gen" $SCRATCH_MNT
+ echo "repair C"
+ $XFS_IO_PROG -x -c "repair dirtree $c_inum $c_gen" $SCRATCH_MNT
+ echo "repair D"
+ $XFS_IO_PROG -x -c "repair dirtree $d_inum $d_gen" $SCRATCH_MNT
+
+ echo "check root"
+ $XFS_IO_PROG -c "scrub dirtree $root_inum $root_gen" $SCRATCH_MNT
+ echo "check A"
+ $XFS_IO_PROG -c "scrub dirtree $a_inum $a_gen" $SCRATCH_MNT
+ echo "check B"
+ $XFS_IO_PROG -c "scrub dirtree $b_inum $b_gen" $SCRATCH_MNT
+ echo "check C"
+ $XFS_IO_PROG -c "scrub dirtree $c_inum $c_gen" $SCRATCH_MNT
+ echo "check D"
+ $XFS_IO_PROG -c "scrub dirtree $d_inum $d_gen" $SCRATCH_MNT
+}
+
+# Part 1: Use raw ioctls to detect the loop and fix it.
+prepare_fs
+_scratch_mount
+simple_online_repair
+_check_scratch_fs
+_scratch_unmount
+
+# Part 2: Use xfs_scrub to detect the loop and fix it.
+prepare_fs
+_scratch_mount
+_scratch_scrub &>> $seqres.full
+echo "xfs_scrub returned $?" >> $seqres.full
+_check_scratch_fs
+_scratch_unmount
+
+# success, all done
+status=0
+exit
diff --git a/tests/xfs/1867.out b/tests/xfs/1867.out
new file mode 100644
index 0000000000..88fbb85e50
--- /dev/null
+++ b/tests/xfs/1867.out
@@ -0,0 +1,25 @@
+QA output created by 1867
+check root
+check A
+check B
+Corruption detected.
+check C
+Corruption detected during cross-referencing.
+check D
+Corruption detected during cross-referencing.
+repair root
+Metadata did not need repair or optimization.
+repair A
+Metadata did not need repair or optimization.
+repair D
+Corruption still detected during cross-referencing.
+repair B
+repair C
+Metadata did not need repair or optimization.
+repair D
+Metadata did not need repair or optimization.
+check root
+check A
+check B
+check C
+check D
diff --git a/tests/xfs/1868 b/tests/xfs/1868
new file mode 100755
index 0000000000..7436343c0c
--- /dev/null
+++ b/tests/xfs/1868
@@ -0,0 +1,121 @@
+#! /bin/bash
+# SPDX-License-Identifier: GPL-2.0
+# Copyright (c) 2023-2024 Oracle. All Rights Reserved.
+#
+# FS QA Test No. 1868
+#
+# Functional testing for online fsck of a directory chain that is not
+# accessible from the root directory.
+#
+. ./common/preamble
+_begin_fstest auto online_repair
+
+# Import common functions.
+. ./common/filter
+. ./common/inject
+. ./common/fuzzy
+. ./common/populate
+
+# real QA test starts here
+
+# Modify as appropriate.
+_supported_fs xfs
+_require_xfs_db_command "link"
+_require_xfs_db_command "unlink"
+_require_scratch
+_require_xfs_stress_online_repair
+
+prepare_fs() {
+ _scratch_mkfs >> $seqres.full
+ _scratch_mount
+ __stress_scrub_check_commands "%dir%" '' '' 'scrub dirtree'
+
+ # Begin by creating the following directory tree:
+ # root["A"]->A
+ # A["B"]->B
+ # B["C"]->C
+ mkdir -p "$SCRATCH_MNT/A/B/C"
+
+ root_inum="$(stat -c '%i' "$SCRATCH_MNT/")"
+ a_inum="$(stat -c '%i' "$SCRATCH_MNT/A")"
+ b_inum="$(stat -c '%i' "$SCRATCH_MNT/A/B")"
+ c_inum="$(stat -c '%i' "$SCRATCH_MNT/A/B/C")"
+
+ echo "root: $root_inum; a: $a_inum; b: $b_inum; c: $c_inum" >> $seqres.full
+
+ # Next, we sever the tree by deleting root["A"]->A. Directory tree is now:
+ # A["B"]->B
+ # B["C"]->C
+ _scratch_unmount
+
+ root_gen=$(_scratch_xfs_get_metadata_field core.gen "inode $root_inum")
+ a_gen=$(_scratch_xfs_get_metadata_field core.gen "inode $a_inum")
+ b_gen=$(_scratch_xfs_get_metadata_field core.gen "inode $b_inum")
+ c_gen=$(_scratch_xfs_get_metadata_field core.gen "inode $c_inum")
+
+ _scratch_xfs_db \
+ -c "echo before root $root_inum" -c "inode $root_inum" -c 'print core.nlinkv2' -c "ls" \
+ -c "echo before C $c_inum" -c "inode $c_inum" -c 'print core.nlinkv2' -c "ls" \
+ -c "echo before A $a_inum" -c "inode $a_inum" -c 'print core.nlinkv2' -c "parent" \
+ >> $seqres.full
+
+ _scratch_xfs_db -x \
+ -c "inode $root_inum" -c "unlink A" \
+ >> $seqres.full
+
+ _scratch_xfs_db \
+ -c "echo after root $root_inum" -c "inode $root_inum" -c 'print core.nlinkv2' -c "ls" \
+ -c "echo after C $c_inum" -c "inode $c_inum" -c 'print core.nlinkv2' -c "ls" \
+ -c "echo after A $a_inum" -c "inode $a_inum" -c 'print core.nlinkv2' -c "parent" \
+ >> $seqres.full
+}
+
+simple_online_repair() {
+ echo "check root"
+ $XFS_IO_PROG -c "scrub dirtree $root_inum $root_gen" $SCRATCH_MNT
+ echo "check A"
+ $XFS_IO_PROG -c "scrub dirtree $a_inum $a_gen" $SCRATCH_MNT
+ echo "check B"
+ $XFS_IO_PROG -c "scrub dirtree $b_inum $b_gen" $SCRATCH_MNT
+ echo "check C"
+ $XFS_IO_PROG -c "scrub dirtree $c_inum $c_gen" $SCRATCH_MNT
+
+ echo "repair C"
+ $XFS_IO_PROG -x -c "repair dirtree $c_inum $c_gen" $SCRATCH_MNT
+ echo "repair root"
+ $XFS_IO_PROG -x -c "repair dirtree $root_inum $root_gen" $SCRATCH_MNT
+ echo "repair A"
+ $XFS_IO_PROG -x -c "repair dirtree $a_inum $a_gen" $SCRATCH_MNT
+ echo "repair B"
+ $XFS_IO_PROG -x -c "repair dirtree $b_inum $b_gen" $SCRATCH_MNT
+ echo "repair C"
+ $XFS_IO_PROG -x -c "repair dirtree $c_inum $c_gen" $SCRATCH_MNT
+
+ echo "check root"
+ $XFS_IO_PROG -c "scrub dirtree $root_inum $root_gen" $SCRATCH_MNT
+ echo "check A"
+ $XFS_IO_PROG -c "scrub dirtree $a_inum $a_gen" $SCRATCH_MNT
+ echo "check B"
+ $XFS_IO_PROG -c "scrub dirtree $b_inum $b_gen" $SCRATCH_MNT
+ echo "check C"
+ $XFS_IO_PROG -c "scrub dirtree $c_inum $c_gen" $SCRATCH_MNT
+}
+
+# Part 1: Use raw ioctls to detect the chain and fix it.
+prepare_fs
+_scratch_mount
+simple_online_repair
+_check_scratch_fs
+_scratch_unmount
+
+# Part 2: Use xfs_scrub to detect the chain and fix it.
+prepare_fs
+_scratch_mount
+_scratch_scrub &>> $seqres.full
+echo "xfs_scrub returned $?" >> $seqres.full
+_check_scratch_fs
+_scratch_unmount
+
+# success, all done
+status=0
+exit
diff --git a/tests/xfs/1868.out b/tests/xfs/1868.out
new file mode 100644
index 0000000000..f4f444ed52
--- /dev/null
+++ b/tests/xfs/1868.out
@@ -0,0 +1,21 @@
+QA output created by 1868
+check root
+check A
+Corruption detected.
+check B
+Corruption detected during cross-referencing.
+check C
+Corruption detected during cross-referencing.
+repair C
+Corruption still detected during cross-referencing.
+repair root
+Metadata did not need repair or optimization.
+repair A
+repair B
+Metadata did not need repair or optimization.
+repair C
+Metadata did not need repair or optimization.
+check root
+check A
+check B
+check C
diff --git a/tests/xfs/1869 b/tests/xfs/1869
new file mode 100755
index 0000000000..188bc0adc8
--- /dev/null
+++ b/tests/xfs/1869
@@ -0,0 +1,157 @@
+#! /bin/bash
+# SPDX-License-Identifier: GPL-2.0
+# Copyright (c) 2023-2024 Oracle. All Rights Reserved.
+#
+# FS QA Test No. 1869
+#
+# Functional testing for online fsck of a multiply-owned directory that is
+# accessible from the root directory.
+#
+. ./common/preamble
+_begin_fstest auto online_repair
+
+# Import common functions.
+. ./common/filter
+. ./common/inject
+. ./common/fuzzy
+. ./common/populate
+
+# real QA test starts here
+
+# Modify as appropriate.
+_supported_fs xfs
+_require_xfs_db_command "link"
+_require_xfs_db_command "unlink"
+_require_scratch
+_require_xfs_stress_online_repair
+
+prepare_fs() {
+ _scratch_mkfs >> $seqres.full
+ _scratch_mount
+ __stress_scrub_check_commands "%dir%" '' '' 'scrub dirtree'
+
+ # Begin by creating the following directory tree:
+ # root["A"]->A
+ # A["B"]->B
+ # B["C"]->C
+ # C["D"]->D
+ # root["Z"]->Z
+ # Z["Y"]->Y
+ mkdir -p "$SCRATCH_MNT/A/B/C/D" "$SCRATCH_MNT/Z/Y"
+
+ root_inum="$(stat -c '%i' "$SCRATCH_MNT/")"
+ a_inum="$(stat -c '%i' "$SCRATCH_MNT/A")"
+ b_inum="$(stat -c '%i' "$SCRATCH_MNT/A/B")"
+ c_inum="$(stat -c '%i' "$SCRATCH_MNT/A/B/C")"
+ d_inum="$(stat -c '%i' "$SCRATCH_MNT/A/B/C/D")"
+ z_inum="$(stat -c '%i' "$SCRATCH_MNT/Z")"
+ y_inum="$(stat -c '%i' "$SCRATCH_MNT/Z/Y")"
+
+ echo "root: $root_inum; a: $a_inum; b: $b_inum; c: $c_inum; d: $d_inum" >> $seqres.full
+ echo "root: $root_inum; z: $z_inum; y: $y_inum" >> $seqres.full
+
+ # Next, we create the multiply-owned directory by creating Y["C1"]->C.
+ # Directory tree is now:
+ # root["A"]->A
+ # A["B"]->B
+ # B["C"]->C
+ # C["D"]->D
+ # root["Z"]->Z
+ # Z["Y"]->Y
+ # Y["C1"]->C
+ _scratch_unmount
+
+ root_gen=$(_scratch_xfs_get_metadata_field core.gen "inode $root_inum")
+ a_gen=$(_scratch_xfs_get_metadata_field core.gen "inode $a_inum")
+ b_gen=$(_scratch_xfs_get_metadata_field core.gen "inode $b_inum")
+ c_gen=$(_scratch_xfs_get_metadata_field core.gen "inode $c_inum")
+ d_gen=$(_scratch_xfs_get_metadata_field core.gen "inode $d_inum")
+ z_gen=$(_scratch_xfs_get_metadata_field core.gen "inode $z_inum")
+ y_gen=$(_scratch_xfs_get_metadata_field core.gen "inode $y_inum")
+
+ _scratch_xfs_db \
+ -c "echo before root $root_inum" -c "inode $root_inum" -c 'print core.nlinkv2' -c "ls" \
+ -c "echo before Y $y_inum" -c "inode $y_inum" -c 'print core.nlinkv2' -c "ls" \
+ -c "echo before B $b_inum" -c "inode $b_inum" -c 'print core.nlinkv2' -c "ls" \
+ -c "echo before C $c_inum" -c "inode $c_inum" -c 'print core.nlinkv2' -c "parent" \
+ >> $seqres.full
+
+ _scratch_xfs_db -x \
+ -c "inode $y_inum" -c "link -i $c_inum C1" \
+ >> $seqres.full
+
+ _scratch_xfs_db \
+ -c "echo before root $root_inum" -c "inode $root_inum" -c 'print core.nlinkv2' -c "ls" \
+ -c "echo before Y $y_inum" -c "inode $y_inum" -c 'print core.nlinkv2' -c "ls" \
+ -c "echo before B $b_inum" -c "inode $b_inum" -c 'print core.nlinkv2' -c "ls" \
+ -c "echo before C $c_inum" -c "inode $c_inum" -c 'print core.nlinkv2' -c "parent" \
+ >> $seqres.full
+}
+
+simple_online_repair() {
+ echo "check root"
+ $XFS_IO_PROG -c "scrub dirtree $root_inum $root_gen" $SCRATCH_MNT
+ echo "check A"
+ $XFS_IO_PROG -c "scrub dirtree $a_inum $a_gen" $SCRATCH_MNT
+ echo "check B"
+ $XFS_IO_PROG -c "scrub dirtree $b_inum $b_gen" $SCRATCH_MNT
+ echo "check C"
+ $XFS_IO_PROG -c "scrub dirtree $c_inum $c_gen" $SCRATCH_MNT
+ echo "check D"
+ $XFS_IO_PROG -c "scrub dirtree $d_inum $d_gen" $SCRATCH_MNT
+ echo "check Z"
+ $XFS_IO_PROG -c "scrub dirtree $z_inum $z_gen" $SCRATCH_MNT
+ echo "check Y"
+ $XFS_IO_PROG -c "scrub dirtree $y_inum $y_gen" $SCRATCH_MNT
+
+ echo "repair D"
+ $XFS_IO_PROG -x -c "repair dirtree $d_inum $d_gen" $SCRATCH_MNT
+ echo "repair root"
+ $XFS_IO_PROG -x -c "repair dirtree $root_inum $root_gen" $SCRATCH_MNT
+ echo "repair A"
+ $XFS_IO_PROG -x -c "repair dirtree $a_inum $a_gen" $SCRATCH_MNT
+ echo "repair B"
+ $XFS_IO_PROG -x -c "repair dirtree $b_inum $b_gen" $SCRATCH_MNT
+ echo "repair C"
+ $XFS_IO_PROG -x -c "repair dirtree $c_inum $c_gen" $SCRATCH_MNT
+ echo "repair D"
+ $XFS_IO_PROG -x -c "repair dirtree $d_inum $d_gen" $SCRATCH_MNT
+ echo "repair Z"
+ $XFS_IO_PROG -x -c "repair dirtree $z_inum $z_gen" $SCRATCH_MNT
+ echo "repair Y"
+ $XFS_IO_PROG -x -c "repair dirtree $y_inum $y_gen" $SCRATCH_MNT
+
+ echo "check root"
+ $XFS_IO_PROG -c "scrub dirtree $root_inum $root_gen" $SCRATCH_MNT
+ echo "check A"
+ $XFS_IO_PROG -c "scrub dirtree $a_inum $a_gen" $SCRATCH_MNT
+ echo "check B"
+ $XFS_IO_PROG -c "scrub dirtree $b_inum $b_gen" $SCRATCH_MNT
+ echo "check C"
+ $XFS_IO_PROG -c "scrub dirtree $c_inum $c_gen" $SCRATCH_MNT
+ echo "check D"
+ $XFS_IO_PROG -c "scrub dirtree $d_inum $d_gen" $SCRATCH_MNT
+ echo "check Z"
+ $XFS_IO_PROG -c "scrub dirtree $z_inum $z_gen" $SCRATCH_MNT
+ echo "check Y"
+ $XFS_IO_PROG -c "scrub dirtree $y_inum $y_gen" $SCRATCH_MNT
+}
+
+# Part 1: Use raw ioctls to detect the multi-parent dir and fix it.
+prepare_fs
+_scratch_mount
+simple_online_repair
+_check_scratch_fs
+_scratch_unmount
+
+# Part 2: Use xfs_scrub to detect the multi-parent dir and fix it.
+prepare_fs
+_scratch_mount
+_scratch_scrub &>> $seqres.full
+echo "xfs_scrub returned $?" >> $seqres.full
+_check_scratch_fs
+_scratch_unmount
+
+# success, all done
+status=0
+exit
diff --git a/tests/xfs/1869.out b/tests/xfs/1869.out
new file mode 100644
index 0000000000..a7ea4c2223
--- /dev/null
+++ b/tests/xfs/1869.out
@@ -0,0 +1,32 @@
+QA output created by 1869
+check root
+check A
+check B
+check C
+Corruption detected.
+check D
+Corruption detected during cross-referencing.
+check Z
+check Y
+repair D
+Corruption still detected during cross-referencing.
+repair root
+Metadata did not need repair or optimization.
+repair A
+Metadata did not need repair or optimization.
+repair B
+Metadata did not need repair or optimization.
+repair C
+repair D
+Metadata did not need repair or optimization.
+repair Z
+Metadata did not need repair or optimization.
+repair Y
+Metadata did not need repair or optimization.
+check root
+check A
+check B
+check C
+check D
+check Z
+check Y
diff --git a/tests/xfs/1870 b/tests/xfs/1870
new file mode 100755
index 0000000000..c4a32de061
--- /dev/null
+++ b/tests/xfs/1870
@@ -0,0 +1,146 @@
+#! /bin/bash
+# SPDX-License-Identifier: GPL-2.0
+# Copyright (c) 2023-2024 Oracle. All Rights Reserved.
+#
+# FS QA Test No. 1870
+#
+# Functional testing for online fsck of a directory loop that is inaccessible
+# from the root directory and has subdirectories.
+#
+. ./common/preamble
+_begin_fstest auto online_repair
+
+# Import common functions.
+. ./common/filter
+. ./common/inject
+. ./common/fuzzy
+. ./common/populate
+
+# real QA test starts here
+
+# Modify as appropriate.
+_supported_fs xfs
+_require_xfs_db_command "link"
+_require_xfs_db_command "unlink"
+_require_scratch
+_require_xfs_stress_online_repair
+
+prepare_fs() {
+ _scratch_mkfs >> $seqres.full
+ _scratch_mount
+ __stress_scrub_check_commands "%dir%" '' '' 'scrub dirtree'
+
+ # Begin by creating the following directory tree:
+ # root["A"]->A
+ # A["B"]->B
+ # B["C"]->C
+ # C["D"]->D
+ # D["E"]->E
+ mkdir -p "$SCRATCH_MNT/A/B/C/D/E"
+
+ root_inum="$(stat -c '%i' "$SCRATCH_MNT/")"
+ a_inum="$(stat -c '%i' "$SCRATCH_MNT/A")"
+ b_inum="$(stat -c '%i' "$SCRATCH_MNT/A/B")"
+ c_inum="$(stat -c '%i' "$SCRATCH_MNT/A/B/C")"
+ d_inum="$(stat -c '%i' "$SCRATCH_MNT/A/B/C/D")"
+ e_inum="$(stat -c '%i' "$SCRATCH_MNT/A/B/C/D/E")"
+
+ echo "root: $root_inum; a: $a_inum; b: $b_inum; c: $c_inum; d: $d_inum; e: $e_inum" >> $seqres.full
+
+ # Complete the loop by creating D["B1"]->B and severing A["B"]->B. Directory
+ # tree is now:
+ # root["A"]->A
+ # B["C"]->C
+ # C["D"]->D
+ # D["E"]->E
+ # D["B1"]->B
+ _scratch_unmount
+
+ root_gen=$(_scratch_xfs_get_metadata_field core.gen "inode $root_inum")
+ a_gen=$(_scratch_xfs_get_metadata_field core.gen "inode $a_inum")
+ b_gen=$(_scratch_xfs_get_metadata_field core.gen "inode $b_inum")
+ c_gen=$(_scratch_xfs_get_metadata_field core.gen "inode $c_inum")
+ d_gen=$(_scratch_xfs_get_metadata_field core.gen "inode $d_inum")
+ e_gen=$(_scratch_xfs_get_metadata_field core.gen "inode $e_inum")
+
+ _scratch_xfs_db \
+ -c "echo before root $root_inum" -c "inode $root_inum" -c 'print core.nlinkv2' -c "ls" \
+ -c "echo before A $d_inum" -c "inode $a_inum" -c 'print core.nlinkv2' -c "ls" \
+ -c "echo before D $d_inum" -c "inode $d_inum" -c 'print core.nlinkv2' -c "ls" \
+ -c "echo before B $b_inum" -c "inode $b_inum" -c 'print core.nlinkv2' -c "parent" \
+ >> $seqres.full
+
+ _scratch_xfs_db -x \
+ -c "inode $d_inum" -c "link -i $b_inum B1" \
+ -c "inode $a_inum" -c "unlink B" \
+ >> $seqres.full
+
+ _scratch_xfs_db \
+ -c "echo before root $root_inum" -c "inode $root_inum" -c 'print core.nlinkv2' -c "ls" \
+ -c "echo before A $d_inum" -c "inode $a_inum" -c 'print core.nlinkv2' -c "ls" \
+ -c "echo before D $d_inum" -c "inode $d_inum" -c 'print core.nlinkv2' -c "ls" \
+ -c "echo before B $b_inum" -c "inode $b_inum" -c 'print core.nlinkv2' -c "parent" \
+ >> $seqres.full
+}
+
+simple_online_repair() {
+ echo "check root"
+ $XFS_IO_PROG -c "scrub dirtree $root_inum $root_gen" $SCRATCH_MNT
+ echo "check A"
+ $XFS_IO_PROG -c "scrub dirtree $a_inum $a_gen" $SCRATCH_MNT
+ echo "check B"
+ $XFS_IO_PROG -c "scrub dirtree $b_inum $b_gen" $SCRATCH_MNT
+ echo "check C"
+ $XFS_IO_PROG -c "scrub dirtree $c_inum $c_gen" $SCRATCH_MNT
+ echo "check D"
+ $XFS_IO_PROG -c "scrub dirtree $d_inum $d_gen" $SCRATCH_MNT
+ echo "check E"
+ $XFS_IO_PROG -c "scrub dirtree $e_inum $e_gen" $SCRATCH_MNT
+
+ echo "repair root"
+ $XFS_IO_PROG -x -c "repair dirtree $root_inum $root_gen" $SCRATCH_MNT
+ echo "repair A"
+ $XFS_IO_PROG -x -c "repair dirtree $a_inum $a_gen" $SCRATCH_MNT
+ echo "repair E"
+ $XFS_IO_PROG -x -c "repair dirtree $e_inum $e_gen" $SCRATCH_MNT
+ echo "repair B"
+ $XFS_IO_PROG -x -c "repair dirtree $b_inum $b_gen" $SCRATCH_MNT
+ echo "repair C"
+ $XFS_IO_PROG -x -c "repair dirtree $c_inum $c_gen" $SCRATCH_MNT
+ echo "repair D"
+ $XFS_IO_PROG -x -c "repair dirtree $d_inum $d_gen" $SCRATCH_MNT
+ echo "repair E"
+ $XFS_IO_PROG -x -c "repair dirtree $e_inum $e_gen" $SCRATCH_MNT
+
+ echo "check root"
+ $XFS_IO_PROG -c "scrub dirtree $root_inum $root_gen" $SCRATCH_MNT
+ echo "check A"
+ $XFS_IO_PROG -c "scrub dirtree $a_inum $a_gen" $SCRATCH_MNT
+ echo "check B"
+ $XFS_IO_PROG -c "scrub dirtree $b_inum $b_gen" $SCRATCH_MNT
+ echo "check C"
+ $XFS_IO_PROG -c "scrub dirtree $c_inum $c_gen" $SCRATCH_MNT
+ echo "check D"
+ $XFS_IO_PROG -c "scrub dirtree $d_inum $d_gen" $SCRATCH_MNT
+ echo "check E"
+ $XFS_IO_PROG -c "scrub dirtree $e_inum $e_gen" $SCRATCH_MNT
+}
+
+# Part 1: Use raw ioctls to detect the loop and fix it.
+prepare_fs
+_scratch_mount
+simple_online_repair
+_check_scratch_fs
+_scratch_unmount
+
+# Part 2: Use xfs_scrub to detect the loop and fix it.
+prepare_fs
+_scratch_mount
+_scratch_scrub &>> $seqres.full
+echo "xfs_scrub returned $?" >> $seqres.full
+_check_scratch_fs
+_scratch_unmount
+
+# success, all done
+status=0
+exit
diff --git a/tests/xfs/1870.out b/tests/xfs/1870.out
new file mode 100644
index 0000000000..8274c6602c
--- /dev/null
+++ b/tests/xfs/1870.out
@@ -0,0 +1,30 @@
+QA output created by 1870
+check root
+check A
+check B
+Corruption detected.
+check C
+Corruption detected.
+check D
+Corruption detected.
+check E
+Corruption detected during cross-referencing.
+repair root
+Metadata did not need repair or optimization.
+repair A
+Metadata did not need repair or optimization.
+repair E
+Corruption still detected during cross-referencing.
+repair B
+repair C
+Metadata did not need repair or optimization.
+repair D
+Metadata did not need repair or optimization.
+repair E
+Metadata did not need repair or optimization.
+check root
+check A
+check B
+check C
+check D
+check E
diff --git a/tests/xfs/1871 b/tests/xfs/1871
new file mode 100755
index 0000000000..760259d18b
--- /dev/null
+++ b/tests/xfs/1871
@@ -0,0 +1,78 @@
+#! /bin/bash
+# SPDX-License-Identifier: GPL-2.0
+# Copyright (c) 2023-2024 Oracle. All Rights Reserved.
+#
+# FS QA Test No. 1871
+#
+# Race rename and directory tree structure corruption detector for a while to
+# exercise the dirtree code's directory path invalidation and its ability to
+# handle unlinked directories.
+#
+. ./common/preamble
+_begin_fstest scrub dangerous_fsstress_scrub
+
+# Import common functions.
+. ./common/filter
+. ./common/fuzzy
+. ./common/inject
+. ./common/xfs
+
+# real QA test starts here
+_supported_fs xfs
+_require_scratch
+_require_xfs_stress_scrub
+
+_scratch_mkfs > "$seqres.full" 2>&1
+_scratch_mount
+__stress_scrub_check_commands "%dir%" '' '' 'scrub dirtree'
+
+parentA="$SCRATCH_MNT/a"
+parentB="$SCRATCH_MNT/b"
+child="$parentA/c/d/e/f/g/h/i/j/k/l/m/n/o/p"
+unlinked="$SCRATCH_MNT/unlinked"
+
+mkdir -p "$parentA" "$parentB" "$child" "$unlinked"
+
+# Find handle info for the child so that we can scrub by handle
+child_inum="$(stat -c '%i' "$child")"
+_scratch_unmount
+child_gen=$(_scratch_xfs_get_metadata_field core.gen "inode $child_inum")
+_scratch_mount
+
+# Queue up a bunch of scrub requests per invocation
+ioargs=()
+for ((i = 0; i < 100; i++)); do
+ ioargs+=('-c' "scrub dirtree $child_inum $child_gen")
+done
+
+renamer() {
+ # Make sure the scrubber handles unlinked directories correctly
+ # by squatting on an empty directory
+ cd "$unlinked"
+ rm -r -f "$unlinked"
+
+ # Bounce the second level directory between parents to stress the
+ # invalidation detector
+ while [ -e $RUNNING_FILE ]; do
+ mv "$parentA/c" "$parentB/"
+ mv "$parentB/c" "$parentA/"
+ done
+}
+
+RUNNING_FILE="$SCRATCH_MNT/run"
+touch $RUNNING_FILE
+renamer &
+
+# Exercise the directory tree scrubber in two ways -- scrubbing the lowest
+# subdir by handle, and running xfs_scrub on the entire fs.
+while _soak_loop_running $((10 * TIME_FACTOR)); do
+ $XFS_IO_PROG "${ioargs[@]}" "$SCRATCH_MNT"
+ XFS_SCRUB_PHASE=5 _scratch_scrub -n >> $seqres.full
+done
+rm -f $RUNNING_FILE
+wait
+
+# success, all done
+echo Silence is golden
+status=0
+exit
diff --git a/tests/xfs/1871.out b/tests/xfs/1871.out
new file mode 100644
index 0000000000..24331e63d5
--- /dev/null
+++ b/tests/xfs/1871.out
@@ -0,0 +1,2 @@
+QA output created by 1871
+Silence is golden
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 1/1] xfs/122: update for vectored scrub
2023-12-31 19:59 ` [PATCHSET 3/3] xfs_scrub: vectorize kernel calls Darrick J. Wong
@ 2023-12-27 13:49 ` Darrick J. Wong
0 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-27 13:49 UTC (permalink / raw)
To: zlang, djwong; +Cc: linux-xfs, guan, fstests
From: Darrick J. Wong <djwong@kernel.org>
Add the two new vectored scrub structures.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
tests/xfs/122.out | 2 ++
1 file changed, 2 insertions(+)
diff --git a/tests/xfs/122.out b/tests/xfs/122.out
index 55138218dd..5d14386518 100644
--- a/tests/xfs/122.out
+++ b/tests/xfs/122.out
@@ -121,6 +121,8 @@ sizeof(struct xfs_rtrmap_root) = 4
sizeof(struct xfs_rud_log_format) = 16
sizeof(struct xfs_rui_log_format) = 16
sizeof(struct xfs_scrub_metadata) = 64
+sizeof(struct xfs_scrub_vec) = 16
+sizeof(struct xfs_scrub_vec_head) = 32
sizeof(struct xfs_swap_extent) = 64
sizeof(struct xfs_sxd_log_format) = 16
sizeof(struct xfs_sxi_log_format) = 80
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 1/1] design: document new name-value logged attribute variants
2023-12-31 20:03 ` [PATCHSET v13.0 1/2] xfs-documentation: document attr log item changes for parent pointers Darrick J. Wong
@ 2023-12-27 14:07 ` Darrick J. Wong
0 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-27 14:07 UTC (permalink / raw)
To: djwong, darrick.wong; +Cc: linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
In preparation for parent pointers, we added a few new opcodes for
logged extended attribute updates. Document them now.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
.../journaling_log.asciidoc | 60 ++++++++++++++++++--
1 file changed, 54 insertions(+), 6 deletions(-)
diff --git a/design/XFS_Filesystem_Structure/journaling_log.asciidoc b/design/XFS_Filesystem_Structure/journaling_log.asciidoc
index daf9b225..c91fbb6a 100644
--- a/design/XFS_Filesystem_Structure/journaling_log.asciidoc
+++ b/design/XFS_Filesystem_Structure/journaling_log.asciidoc
@@ -730,10 +730,21 @@ of file block mapping operation we want.
.Extended attribute update log intent types
[options="header"]
|=====
-| Value | Description
-| +XFS_ATTRI_OP_FLAGS_SET+ | Set a key/value pair.
-| +XFS_ATTRI_OP_FLAGS_REMOVE+ | Remove a key/value pair.
-| +XFS_ATTRI_OP_FLAGS_REPLACE+ | Replace one key/value pair with another.
+| Value | Description
+| +XFS_ATTRI_OP_FLAGS_SET+ | Associate an attribute name with the
+given value, creating an entry for the name if necessary.
+| +XFS_ATTRI_OP_FLAGS_REMOVE+ | Remove an attribute name and any
+value associated with it.
+| +XFS_ATTRI_OP_FLAGS_REPLACE+ | Remove any value associated with an
+attribute name, then associate the name with the given value.
+| +XFS_ATTRI_OP_FLAGS_NVREMOVE+ | Remove the specific name and value
+from the attribute structure. The name and value must already exist.
+| +XFS_ATTRI_OP_FLAGS_NVSET+ | Associate an attribute name with the
+given value. The name and value must not exist in the attribute structure.
+A name associated with a different value will not be removed.
+| +XFS_ATTRI_OP_FLAGS_NVREPLACE+ | Remove a specific name and value from
+the attribute structure, then associate a specific name with a given value.
+The two names and values need not be the same.
|=====
The ``extended attribute update intent'' operation comes first; it tells the
@@ -747,11 +758,17 @@ through the complex update will be replayed fully during log recovery.
struct xfs_attri_log_format {
uint16_t alfi_type;
uint16_t alfi_size;
- uint32_t __pad;
+ uint32_t alfi_new_value_len;
uint64_t alfi_id;
uint64_t alfi_ino;
uint32_t alfi_op_flags;
- uint32_t alfi_name_len;
+ union {
+ uint32_t alfi_name_len;
+ struct {
+ uint16_t alfi_old_name_len;
+ uint16_t alfi_new_name_len;
+ };
+ };
uint32_t alfi_value_len;
uint32_t alfi_attr_filter;
};
@@ -764,6 +781,11 @@ order, not big-endian like the rest of XFS.
*alfi_size*::
Size of this log item. Should be 1.
+*alfi_new_value_len*::
+For NVREPLACE, this is the length of the new xattr value, and +alfi_value_len+
+contains the length of the old xattr value.
+For all other opcodes, this field must be zero.
+
*alfi_id*::
A 64-bit number that binds the corresponding ATTRD log item to this ATTRI log
item.
@@ -778,6 +800,13 @@ The operation being performed. The lower byte must be one of the
*alfi_name_len*::
Length of the name of the extended attribute. This must not be zero.
The attribute name itself is captured in the next log item.
+This field is not defined for the NVREPLACE opcode.
+
+*alfi_old_name_len*::
+For NVREPLACE, this is the length of the old name.
+
+*alfi_new_name_len*::
+For NVREPLACE, this is the length of the new name.
*alfi_value_len*::
Length of the value of the extended attribute. This must be zero for remove
@@ -789,6 +818,25 @@ name.
Attribute namespace filter flags. This must be one of +ATTR_ROOT+,
+ATTR_SECURE+, or +ATTR_INCOMPLETE+.
+For a SET or REPLACE opcode, there should be two regions after the ATTRI intent
+item. The first region contains the attribute name and the second contains the
+attribute value.
+
+For a REMOVE opcode, there should only be one region after the ATTRI intent
+item, and it will contain the attribute name.
+
+For an NVSET or NVREMOVE opcode, there should be one or two regions after the
+ATTRI intent item. The first region contains the attribute name. The second
+region, if present, contains the attribute value.
+
+For an NVREPLACE opcode, there should be between two and four regions after the
+ATTRI intent item. The first region contains the attribute name to remove.
+The second region contains the attribute name to create. If +alfi_value_len+
+is nonzero, the third region contains the attribute value to remove. If
++alfi_new_value_len+ is nonzero, the next region seen contains the attribute
+value to create. This could be the third region if there was no value to
+remove, or it could be the fourth region.
+
[[ATTRD_Log_Item]]
=== Completion of Extended Attribute Updates
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 1/1] design: document the parent pointer ondisk format
2023-12-31 20:03 ` [PATCHSET v13.0 2/2] xfs-documentation: document parent pointers Darrick J. Wong
@ 2023-12-27 14:07 ` Darrick J. Wong
0 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-27 14:07 UTC (permalink / raw)
To: djwong, darrick.wong; +Cc: linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
Add parent pointers to the ondisk format documentation.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
.../allocation_groups.asciidoc | 4 +
.../extended_attributes.asciidoc | 94 ++++++++++++++++++++
2 files changed, 98 insertions(+)
diff --git a/design/XFS_Filesystem_Structure/allocation_groups.asciidoc b/design/XFS_Filesystem_Structure/allocation_groups.asciidoc
index c91a06bf..bd825207 100644
--- a/design/XFS_Filesystem_Structure/allocation_groups.asciidoc
+++ b/design/XFS_Filesystem_Structure/allocation_groups.asciidoc
@@ -469,6 +469,10 @@ space mappings allowed in data and extended attribute file forks.
Metadata directory tree. See the section about the xref:Metadata_Directories[
metadata directory tree] for more information.
+| +XFS_SB_FEAT_INCOMPAT_PARENT+ |
+Directory parent pointers. See the section about xref:Parent_Pointers[parent
+pointers] for more information.
+
|=====
*sb_features_log_incompat*::
diff --git a/design/XFS_Filesystem_Structure/extended_attributes.asciidoc b/design/XFS_Filesystem_Structure/extended_attributes.asciidoc
index 19bff70f..6f905a1b 100644
--- a/design/XFS_Filesystem_Structure/extended_attributes.asciidoc
+++ b/design/XFS_Filesystem_Structure/extended_attributes.asciidoc
@@ -90,6 +90,7 @@ A combination of the following:
| +XFS_ATTR_SECURE+ | The attribute's namespace is ``secure''.
| +XFS_ATTR_INCOMPLETE+ | This attribute is being modified.
| +XFS_ATTR_LOCAL+ | The attribute value is contained within this block.
+| +XFS_ATTR_PARENT+ | This attribute is a parent pointer.
|=====
.Short form attribute layout
@@ -911,6 +912,99 @@ Log sequence number of the last write to this block.
Filesystems formatted prior to v5 do not have this header in the remote block.
Value data begins immediately at offset zero.
+[[Parent_Pointers]]
+== Directory Parent Pointers
+
+If this feature is enabled, each directory entry pointing from a parent
+directory to a child file has a corresponding back link from the child file
+back to the parent. In other words, if directory P has an entry "foo" pointing
+to child C, then child C will have a parent pointer entry "foo" pointing to
+parent P. This redundancy enables validation and repairs of the directory tree
+if the tree structure is damaged.
+
+Parent pointers are stored in a private namespace within the extended attribute
+structure. The attribute name contains the following binary structure, and
+the attribute value contains the directory entry name.
+
+[source, c]
+----
+struct xfs_parent_name_rec {
+ __be64 p_ino;
+ __be32 p_gen;
+ __be32 p_namehash;
+};
+----
+
+*p_ino*::
+Inode number of the parent directory.
+
+*p_gen*::
+Generation number of the parent directory.
+
+*p_namehash*::
+The directory name hash of the directory entry name in the parent.
+
+=== xfs_db Parent Pointer Example
+
+Create a directory tree with the following structure, assuming that the
+XFS filesystem is mounted on +/mnt+:
+
+----
+$ mkdir /mnt/a/ /mnt/b
+$ touch /mnt/a/autoexec.bat
+$ ln /mnt/a/autoexec.bat /mnt/b/config.sys
+----
+
+Now we open this up in the debugger:
+
+----
+xfs_db> path /a
+xfs_db> ls
+8 131 directory 0x0000002e 1 . (good)
+10 128 directory 0x0000172e 2 .. (good)
+12 132 regular 0x5a1f6ea0 12 autoexec.bat (good)
+xfs_db> path /b
+xfs_db> ls
+8 8388736 directory 0x0000002e 1 . (good)
+10 128 directory 0x0000172e 2 .. (good)
+15 132 regular 0x9a01678c 10 config.sys (good)
+xfs_db> path /b/config.sys
+xfs_db> p a
+a.sfattr.hdr.totsize = 64
+a.sfattr.hdr.count = 2
+a.sfattr.list[0].namelen = 16
+a.sfattr.list[0].valuelen = 12
+a.sfattr.list[0].root = 0
+a.sfattr.list[0].secure = 0
+a.sfattr.list[0].parent = 1
+a.sfattr.list[0].parent_ino = 131
+a.sfattr.list[0].parent_gen = 3772462576
+a.sfattr.list[0].parent_namehash = 0x5a1f6ea0
+a.sfattr.list[0].parent_name = "autoexec.bat"
+a.sfattr.list[1].namelen = 16
+a.sfattr.list[1].valuelen = 10
+a.sfattr.list[1].root = 0
+a.sfattr.list[1].secure = 0
+a.sfattr.list[1].parent = 1
+a.sfattr.list[1].parent_ino = 8388736
+a.sfattr.list[1].parent_gen = 1161632072
+a.sfattr.list[1].parent_namehash = 0x9a01678c
+a.sfattr.list[1].parent_name = "config.sys"
+----
+
+In this example, +/a+ and +/b+ are subdirectories of the root. A regular file
+is hardlinked into both subdirectories, under different names. Directory +/a+
+is inode 131 and has an entry +autoexec.bat+ pointing to the child file.
+Directory +/b+ is inode 8388736 and has an entry +config.sys+ pointing to the
+same child file.
+
+Within the child file, notice that there are two parent pointers in the
+extended attribute structure. The first parent pointer tells us that directory
+inode 131 should have an entry +autoexec.bat+ pointing down to the child; the
+second parent pointer tells us that directory inode 8388736 should have an
+entry +config.sys+ pointing down to the child. Note that the name hashes are
+the same between each directory entry and its parent pointer.
+
== Key Differences Between Directories and Extended Attributes
Directories and extended attributes share the function of mapping names to
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [NYE PATCHRIVER 2/4] xfs: online repair part 2
@ 2023-12-31 18:18 Darrick J. Wong
2023-12-31 19:32 ` [PATCHSET v13.0 1/7] xfs: design documentation for online fsck, " Darrick J. Wong
` (17 more replies)
0 siblings, 18 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 18:18 UTC (permalink / raw)
To: Chandan Babu R, Christoph Hellwig
Cc: xfs, greg.marsden, shirley.ma, konrad.wilk, Matthew Wilcox,
Dave Chinner, Catherine Hoang, fstests, Zorro Lang,
Carlos Maiolino, Kent Overstreet
Hi everyone,
This is part 2 of online repair. This river combines the directory
parent pointer work that I took over from Allison after she departed
with a bunch of new online repair functionality that enables rebuilding
of directory trees and correction of problems with the tree structure.
As a consequence, this river needs closer review than the previous
river. I've changed the ondisk format of the parent pointers to make it
easier to correct the directory structure, but that required some
alterations to the attr log intent items to support looking up xattrs by
name and value.
AFAICT this part is stable enough for review-- I've been running a few
testvm host systems with parent pointers enabled for about 6 months, and
haven't noticed any problems. With this part added, there's a dramatic
drop in the number of uncorrectable directory fields reported by the
fuzz test suite.
--D
^ permalink raw reply [flat|nested] 185+ messages in thread
* [PATCHSET v13.0 1/7] xfs: design documentation for online fsck, part 2
2023-12-31 18:18 [NYE PATCHRIVER 2/4] xfs: online repair part 2 Darrick J. Wong
@ 2023-12-31 19:32 ` Darrick J. Wong
2023-12-31 20:42 ` [PATCH 1/4] docs: update the parent pointers documentation to the final version Darrick J. Wong
` (3 more replies)
2023-12-31 19:33 ` [PATCHSET v13.0 2/7] xfs: retain ILOCK during directory updates Darrick J. Wong
` (16 subsequent siblings)
17 siblings, 4 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 19:32 UTC (permalink / raw)
To: djwong; +Cc: linux-xfs
Hi all,
This series updates the design documentation for online fsck to reflect
the final design of the parent pointers feature as well as the
implementation of online fsck for the new metadata.
If you're going to start using this code, I strongly recommend pulling
from my git trees, which are linked below.
This has been running on the djcloud for months with no problems. Enjoy!
Comments and questions are, as always, welcome.
--D
kernel git tree:
https://git.kernel.org/cgit/linux/kernel/git/djwong/xfs-linux.git/log/?h=online-fsck-design
---
.../filesystems/xfs-online-fsck-design.rst | 349 +++++++++++++++-----
1 file changed, 260 insertions(+), 89 deletions(-)
^ permalink raw reply [flat|nested] 185+ messages in thread
* [PATCHSET v13.0 2/7] xfs: retain ILOCK during directory updates
2023-12-31 18:18 [NYE PATCHRIVER 2/4] xfs: online repair part 2 Darrick J. Wong
2023-12-31 19:32 ` [PATCHSET v13.0 1/7] xfs: design documentation for online fsck, " Darrick J. Wong
@ 2023-12-31 19:33 ` Darrick J. Wong
2023-12-31 20:43 ` [PATCH 1/7] xfs: Increase XFS_DEFER_OPS_NR_INODES to 5 Darrick J. Wong
` (6 more replies)
2023-12-31 19:33 ` [PATCHSET v13.0 3/7] xfs: name-value xattr lookups Darrick J. Wong
` (15 subsequent siblings)
17 siblings, 7 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 19:33 UTC (permalink / raw)
To: djwong
Cc: Catherine Hoang, Allison Henderson, catherine.hoang,
allison.henderson, linux-xfs
Hi all,
This series changes the directory update code to retain the ILOCK on all
files involved in a rename until the end of the operation. The upcoming
parent pointers patchset applies parent pointers in a separate chained
update from the actual directory update, which is why it is now
necessary to keep the ILOCK instead of dropping it after the first
transaction in the chain.
As a side effect, we no longer need to hold the IOLOCK during an rmapbt
scan of inodes to serialize the scan with ongoing directory updates.
If you're going to start using this code, I strongly recommend pulling
from my git trees, which are linked below.
This has been running on the djcloud for months with no problems. Enjoy!
Comments and questions are, as always, welcome.
--D
kernel git tree:
https://git.kernel.org/cgit/linux/kernel/git/djwong/xfs-linux.git/log/?h=retain-ilock-during-dir-ops
xfsprogs git tree:
https://git.kernel.org/cgit/linux/kernel/git/djwong/xfsprogs-dev.git/log/?h=retain-ilock-during-dir-ops
---
fs/xfs/libxfs/xfs_defer.c | 6 ++-
fs/xfs/libxfs/xfs_defer.h | 8 +++-
fs/xfs/scrub/rmap_repair.c | 16 -------
fs/xfs/scrub/tempfile.c | 2 +
fs/xfs/xfs_dquot.c | 41 ++++++++++++++++++
fs/xfs/xfs_dquot.h | 1
fs/xfs/xfs_inode.c | 98 ++++++++++++++++++++++++++++++++------------
fs/xfs/xfs_inode.h | 2 +
fs/xfs/xfs_qm.c | 4 +-
fs/xfs/xfs_qm.h | 2 -
fs/xfs/xfs_symlink.c | 6 ++-
fs/xfs/xfs_trans.c | 9 +++-
fs/xfs/xfs_trans_dquot.c | 15 ++++---
13 files changed, 156 insertions(+), 54 deletions(-)
^ permalink raw reply [flat|nested] 185+ messages in thread
* [PATCHSET v13.0 3/7] xfs: name-value xattr lookups
2023-12-31 18:18 [NYE PATCHRIVER 2/4] xfs: online repair part 2 Darrick J. Wong
2023-12-31 19:32 ` [PATCHSET v13.0 1/7] xfs: design documentation for online fsck, " Darrick J. Wong
2023-12-31 19:33 ` [PATCHSET v13.0 2/7] xfs: retain ILOCK during directory updates Darrick J. Wong
@ 2023-12-31 19:33 ` Darrick J. Wong
2023-12-31 20:45 ` [PATCH 01/14] xfs: require XFS_SB_FEAT_INCOMPAT_LOG_XATTRS for attr log intent item recovery Darrick J. Wong
` (13 more replies)
2023-12-31 19:33 ` [PATCHSET v13.0 4/7] xfs: Parent Pointers Darrick J. Wong
` (14 subsequent siblings)
17 siblings, 14 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 19:33 UTC (permalink / raw)
To: djwong; +Cc: Allison Henderson, catherine.hoang, allison.henderson, linux-xfs
Hi all,
Directory parent pointers are stored as namespaced extended attributes
of a file. Because parent pointers can consume up to 267 bytes of
space and xattr names are 255 bytes at most, we cannot use the usual
attr name lookup functions to find a parent pointer. This is solvable
by introducing a new lookup mode that checks both the name and the
value of the xattr.
Therefore, introduce this new lookup mode. Because all parent pointer
updates are logged, we must extend the xattr logging code to capture the
VLOOKUP variants, and restore them when recovering logged operations.
These new log formats are protected by the sb_incompat PARENT flag, so
they do not need a separate log_incompat feature flag.
If you're going to start using this code, I strongly recommend pulling
from my git trees, which are linked below.
This has been running on the djcloud for months with no problems. Enjoy!
Comments and questions are, as always, welcome.
--D
kernel git tree:
https://git.kernel.org/cgit/linux/kernel/git/djwong/xfs-linux.git/log/?h=pptrs-attr-nvlookups
xfsprogs git tree:
https://git.kernel.org/cgit/linux/kernel/git/djwong/xfsprogs-dev.git/log/?h=pptrs-attr-nvlookups
xfsdocs git tree:
https://git.kernel.org/cgit/linux/kernel/git/djwong/xfs-documentation.git/log/?h=pptrs-attr-nvlookups
---
fs/xfs/libxfs/xfs_attr.c | 52 ++++--
fs/xfs/libxfs/xfs_attr.h | 32 +++-
fs/xfs/libxfs/xfs_attr_leaf.c | 45 ++++-
fs/xfs/libxfs/xfs_da_btree.h | 10 +
fs/xfs/libxfs/xfs_format.h | 6 +
fs/xfs/libxfs/xfs_log_format.h | 29 +++
fs/xfs/xfs_attr_item.c | 349 ++++++++++++++++++++++++++++++++++------
fs/xfs/xfs_attr_item.h | 2
fs/xfs/xfs_xattr.c | 15 ++
9 files changed, 460 insertions(+), 80 deletions(-)
^ permalink raw reply [flat|nested] 185+ messages in thread
* [PATCHSET v13.0 4/7] xfs: Parent Pointers
2023-12-31 18:18 [NYE PATCHRIVER 2/4] xfs: online repair part 2 Darrick J. Wong
` (2 preceding siblings ...)
2023-12-31 19:33 ` [PATCHSET v13.0 3/7] xfs: name-value xattr lookups Darrick J. Wong
@ 2023-12-31 19:33 ` Darrick J. Wong
2023-12-31 20:48 ` [PATCH 01/18] xfs: Expose init_xattrs in xfs_create_tmpfile Darrick J. Wong
` (17 more replies)
2023-12-31 19:33 ` [PATCHSET v13.0 5/7] xfs: fsck for parent pointers Darrick J. Wong
` (13 subsequent siblings)
17 siblings, 18 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 19:33 UTC (permalink / raw)
To: djwong
Cc: Mark Tinguely, Allison Henderson, Dave Chinner, Darrick J. Wong,
catherine.hoang, allison.henderson, linux-xfs
Hi all,
This is the latest parent pointer attributes for xfs. The goal of this
patch set is to add a parent pointer attribute to each inode. The
attribute name containing the parent inode, generation, and directory
offset, while the attribute value contains the file name. This feature
will enable future optimizations for online scrub, shrink, nfs handles,
verity, or any other feature that could make use of quickly deriving an
inodes path from the mount point.
At this point, Allison is moving on to other things, so I've merged her
patchset into djwong-dev for merging.
Updates since v12 [djwong]:
Rebase on 6.5-rc and update the online fsck design document.
Updates since v11 [djwong]:
Rebase on 6.4-rc and make some tweaks and bugfixes to enable the repair
prototypes. Merge with djwong-dev and make online repair actually work.
Updates since v10 [djwong]:
Merge in the ondisk format changes to get rid of the diroffset conflicts
with the parent pointer repair code, rebase the entire series with the
attr vlookup changes first, and merge all the other random fixes.
Updates since v9:
Reordered patches 2 and 3 to be 6 and 7
xfs: Add xfs_verify_pptr
moved parent pointer validators to xfs_parent
xfs: Add parent pointer ioctl
Extra validation checks for fs id
added missing release for the inode
use GFP_KERNEL flags for malloc/realloc
reworked ioctl to use pptr listenty and flex array
NEW
xfs: don't remove the attr fork when parent pointers are enabled
NEW
directory lookups should return diroffsets too
NEW
xfs: move/add parent pointer validators to xfs_parent
Updates since v8:
xfs: parent pointer attribute creation
Fix xfs_parent_init to release log assist on alloc fail
Add slab cache for xfs_parent_defer
Fix xfs_create to release after unlock
Add xfs_parent_start and xfs_parent_finish wrappers
removed unused xfs_parent_name_irec and xfs_init_parent_name_irec
xfs: add parent attributes to link
Start/finish wrapper updates
Fix xfs_link to disallow reservationless quotas
xfs: add parent attributes to symlink
Fix xfs_symlink to release after unlock
Start/finish wrapper updates
xfs: remove parent pointers in unlink
Start/finish wrapper updates
Add missing parent free
xfs: Add parent pointers to rename
Start/finish wrapper updates
Fix rename to only grab logged xattr once
Fix xfs_rename to disallow reservationless quotas
Fix double unlock on dqattach fail
Move parent frees to out_release_wip
xfs: Add parent pointers to xfs_cross_rename
Hoist parent pointers into rename
Questions comments and feedback appreciated!
Thanks all!
Allison
If you're going to start using this code, I strongly recommend pulling
from my git trees, which are linked below.
This has been running on the djcloud for months with no problems. Enjoy!
Comments and questions are, as always, welcome.
--D
kernel git tree:
https://git.kernel.org/cgit/linux/kernel/git/djwong/xfs-linux.git/log/?h=pptrs
xfsprogs git tree:
https://git.kernel.org/cgit/linux/kernel/git/djwong/xfsprogs-dev.git/log/?h=pptrs
fstests git tree:
https://git.kernel.org/cgit/linux/kernel/git/djwong/xfstests-dev.git/log/?h=pptrs
xfsdocs git tree:
https://git.kernel.org/cgit/linux/kernel/git/djwong/xfs-documentation.git/log/?h=pptrs
---
fs/xfs/Makefile | 3
fs/xfs/libxfs/xfs_attr.c | 15 +-
fs/xfs/libxfs/xfs_attr.h | 10 +
fs/xfs/libxfs/xfs_attr_leaf.c | 6 -
fs/xfs/libxfs/xfs_attr_sf.h | 1
fs/xfs/libxfs/xfs_da_format.h | 33 +++
fs/xfs/libxfs/xfs_format.h | 4
fs/xfs/libxfs/xfs_fs.h | 2
fs/xfs/libxfs/xfs_fs_staging.h | 66 +++++++
fs/xfs/libxfs/xfs_log_format.h | 1
fs/xfs/libxfs/xfs_log_rlimit.c | 43 +++++
fs/xfs/libxfs/xfs_ondisk.h | 4
fs/xfs/libxfs/xfs_parent.c | 368 +++++++++++++++++++++++++++++++++++++++
fs/xfs/libxfs/xfs_parent.h | 155 ++++++++++++++++
fs/xfs/libxfs/xfs_sb.c | 4
fs/xfs/libxfs/xfs_trans_resv.c | 326 +++++++++++++++++++++++++++++------
fs/xfs/libxfs/xfs_trans_space.c | 121 +++++++++++++
fs/xfs/libxfs/xfs_trans_space.h | 25 +--
fs/xfs/scrub/attr.c | 4
fs/xfs/scrub/dir_repair.c | 2
fs/xfs/scrub/orphanage.c | 5 -
fs/xfs/scrub/parent_repair.c | 3
fs/xfs/scrub/symlink_repair.c | 2
fs/xfs/scrub/tempfile.c | 2
fs/xfs/xfs_attr_item.c | 42 ++++
fs/xfs/xfs_attr_list.c | 25 ++-
fs/xfs/xfs_inode.c | 199 +++++++++++++++++----
fs/xfs/xfs_inode.h | 2
fs/xfs/xfs_ioctl.c | 146 +++++++++++++++
fs/xfs/xfs_iops.c | 15 +-
fs/xfs/xfs_parent_utils.c | 161 +++++++++++++++++
fs/xfs/xfs_parent_utils.h | 20 ++
fs/xfs/xfs_super.c | 14 +
fs/xfs/xfs_symlink.c | 28 ++-
fs/xfs/xfs_trace.c | 1
fs/xfs/xfs_trace.h | 76 ++++++++
fs/xfs/xfs_xattr.c | 8 +
fs/xfs/xfs_xattr.h | 2
38 files changed, 1806 insertions(+), 138 deletions(-)
create mode 100644 fs/xfs/libxfs/xfs_parent.c
create mode 100644 fs/xfs/libxfs/xfs_parent.h
create mode 100644 fs/xfs/libxfs/xfs_trans_space.c
create mode 100644 fs/xfs/xfs_parent_utils.c
create mode 100644 fs/xfs/xfs_parent_utils.h
^ permalink raw reply [flat|nested] 185+ messages in thread
* [PATCHSET v13.0 5/7] xfs: fsck for parent pointers
2023-12-31 18:18 [NYE PATCHRIVER 2/4] xfs: online repair part 2 Darrick J. Wong
` (3 preceding siblings ...)
2023-12-31 19:33 ` [PATCHSET v13.0 4/7] xfs: Parent Pointers Darrick J. Wong
@ 2023-12-31 19:33 ` Darrick J. Wong
2023-12-31 20:53 ` [PATCH 01/22] xfs: check dirents have " Darrick J. Wong
` (21 more replies)
2023-12-31 19:34 ` [PATCHSET v13.0 6/7] xfs: detect and correct directory tree problems Darrick J. Wong
` (12 subsequent siblings)
17 siblings, 22 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 19:33 UTC (permalink / raw)
To: djwong; +Cc: catherine.hoang, allison.henderson, linux-xfs
Hi all,
This series implements online checking and repair for directory parent
pointer metadata. The checking half is fairly straightforward -- for
each outgoing directory link (forward or backwards), grab the inode at
the other end, and confirm that there's a corresponding link. If we
can't grab an inode or lock it, we'll save that link for a slower loop
that cycles all the locks, confirms the continued existence of the link,
and rechecks the link if it's actually still there.
Repairs are a bit more involved -- for directories, we walk the entire
filesystem to rebuild the dirents from parent pointer information.
Parent pointer repairs do the same walk but rebuild the pptrs from the
dirent information, but with the added twist that it duplicates all the
xattrs so that it can use the atomic extent swapping code to commit the
repairs atomically.
This introduces an added twist to the xattr repair code -- we use dirent
hooks to detect a colliding update to the pptr data while we're not
holding the ILOCKs; if one is detected, we restart the xattr salvaging
process but this time hold all the ILOCKs until the end of the scan.
For offline repair, the phase6 directory connectivity scan generates an
index of all the expected parent pointers in the filesystem. Then it
walks each file and compares the parent pointers attached to that file
against the index generated, and resyncs the results as necessary.
If you're going to start using this code, I strongly recommend pulling
from my git trees, which are linked below.
This has been running on the djcloud for months with no problems. Enjoy!
Comments and questions are, as always, welcome.
--D
kernel git tree:
https://git.kernel.org/cgit/linux/kernel/git/djwong/xfs-linux.git/log/?h=pptrs-fsck
xfsprogs git tree:
https://git.kernel.org/cgit/linux/kernel/git/djwong/xfsprogs-dev.git/log/?h=pptrs-fsck
---
fs/xfs/Makefile | 2
fs/xfs/libxfs/xfs_attr.c | 39 +
fs/xfs/libxfs/xfs_attr.h | 2
fs/xfs/libxfs/xfs_bmap.c | 38 -
fs/xfs/libxfs/xfs_bmap.h | 3
fs/xfs/libxfs/xfs_dir2.c | 2
fs/xfs/libxfs/xfs_dir2.h | 2
fs/xfs/libxfs/xfs_parent.c | 107 +++
fs/xfs/libxfs/xfs_parent.h | 18 +
fs/xfs/scrub/attr.c | 2
fs/xfs/scrub/attr_repair.c | 502 ++++++++++++++++
fs/xfs/scrub/attr_repair.h | 4
fs/xfs/scrub/common.h | 1
fs/xfs/scrub/dir.c | 352 +++++++++++
fs/xfs/scrub/dir_repair.c | 567 +++++++++++++++++-
fs/xfs/scrub/findparent.c | 10
fs/xfs/scrub/findparent.h | 10
fs/xfs/scrub/inode_repair.c | 41 +
fs/xfs/scrub/listxattr.c | 94 +++
fs/xfs/scrub/listxattr.h | 13
fs/xfs/scrub/nlinks.c | 71 ++
fs/xfs/scrub/nlinks.h | 3
fs/xfs/scrub/nlinks_repair.c | 2
fs/xfs/scrub/orphanage.c | 42 +
fs/xfs/scrub/orphanage.h | 3
fs/xfs/scrub/parent.c | 695 ++++++++++++++++++++++
fs/xfs/scrub/parent_repair.c | 1316 +++++++++++++++++++++++++++++++++++++++++-
fs/xfs/scrub/readdir.c | 57 ++
fs/xfs/scrub/readdir.h | 3
fs/xfs/scrub/scrub.c | 2
fs/xfs/scrub/trace.c | 1
fs/xfs/scrub/trace.h | 217 +++++++
32 files changed, 4113 insertions(+), 108 deletions(-)
^ permalink raw reply [flat|nested] 185+ messages in thread
* [PATCHSET v13.0 6/7] xfs: detect and correct directory tree problems
2023-12-31 18:18 [NYE PATCHRIVER 2/4] xfs: online repair part 2 Darrick J. Wong
` (4 preceding siblings ...)
2023-12-31 19:33 ` [PATCHSET v13.0 5/7] xfs: fsck for parent pointers Darrick J. Wong
@ 2023-12-31 19:34 ` Darrick J. Wong
2023-12-31 20:59 ` [PATCH 1/4] xfs: teach online scrub to find directory tree structure problems Darrick J. Wong
` (3 more replies)
2023-12-31 19:34 ` [PATCHSET v13.0 7/7] xfs: vectorize scrub kernel calls Darrick J. Wong
` (11 subsequent siblings)
17 siblings, 4 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 19:34 UTC (permalink / raw)
To: djwong; +Cc: linux-xfs
Hi all,
Historically, checking the tree-ness of the directory tree structure has
not been complete. Cycles of subdirectories break the tree properties,
as do subdirectories with multiple parents. It's easy enough for DFS to
detect problems as long as one of the participants is reachable from the
root, but this technique cannot find unconnected cycles.
Directory parent pointers change that, because we can discover all of
these problems from a simple walk from a subdirectory towards the root.
For each child we start with, if the walk terminates without reaching
the root, we know the path is disconnected and ought to be attached to
the lost and found. If we find ourselves, we know this is a cycle and
can delete an incoming edge. If we find multiple paths to the root, we
know to delete an incoming edge.
Even better, once we've finished walking paths, we've identified the
good ones and know which other path(s) to remove.
If you're going to start using this code, I strongly recommend pulling
from my git trees, which are linked below.
This has been running on the djcloud for months with no problems. Enjoy!
Comments and questions are, as always, welcome.
--D
kernel git tree:
https://git.kernel.org/cgit/linux/kernel/git/djwong/xfs-linux.git/log/?h=scrub-directory-tree
xfsprogs git tree:
https://git.kernel.org/cgit/linux/kernel/git/djwong/xfsprogs-dev.git/log/?h=scrub-directory-tree
fstests git tree:
https://git.kernel.org/cgit/linux/kernel/git/djwong/xfstests-dev.git/log/?h=scrub-directory-tree
---
fs/xfs/Makefile | 2
fs/xfs/libxfs/xfs_fs.h | 4
fs/xfs/libxfs/xfs_health.h | 4
fs/xfs/scrub/common.h | 1
fs/xfs/scrub/dirtree.c | 938 +++++++++++++++++++++++++++++++++++++++++
fs/xfs/scrub/dirtree.h | 171 +++++++
fs/xfs/scrub/dirtree_repair.c | 824 ++++++++++++++++++++++++++++++++++++
fs/xfs/scrub/health.c | 1
fs/xfs/scrub/ino_bitmap.h | 37 ++
fs/xfs/scrub/orphanage.c | 6
fs/xfs/scrub/orphanage.h | 8
fs/xfs/scrub/repair.h | 4
fs/xfs/scrub/scrub.c | 7
fs/xfs/scrub/scrub.h | 1
fs/xfs/scrub/stats.c | 1
fs/xfs/scrub/trace.c | 4
fs/xfs/scrub/trace.h | 270 ++++++++++++
fs/xfs/scrub/xfarray.h | 1
fs/xfs/xfs_health.c | 1
fs/xfs/xfs_inode.c | 2
fs/xfs/xfs_inode.h | 1
21 files changed, 2284 insertions(+), 4 deletions(-)
create mode 100644 fs/xfs/scrub/dirtree.c
create mode 100644 fs/xfs/scrub/dirtree.h
create mode 100644 fs/xfs/scrub/dirtree_repair.c
create mode 100644 fs/xfs/scrub/ino_bitmap.h
^ permalink raw reply [flat|nested] 185+ messages in thread
* [PATCHSET v13.0 7/7] xfs: vectorize scrub kernel calls
2023-12-31 18:18 [NYE PATCHRIVER 2/4] xfs: online repair part 2 Darrick J. Wong
` (5 preceding siblings ...)
2023-12-31 19:34 ` [PATCHSET v13.0 6/7] xfs: detect and correct directory tree problems Darrick J. Wong
@ 2023-12-31 19:34 ` Darrick J. Wong
2023-12-31 21:00 ` [PATCH 1/3] xfs: reduce the rate of cond_resched calls inside scrub Darrick J. Wong
` (2 more replies)
2023-12-31 19:50 ` [PATCHSET v13.0 1/6] xfsprogs: retain ILOCK during directory updates Darrick J. Wong
` (10 subsequent siblings)
17 siblings, 3 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 19:34 UTC (permalink / raw)
To: djwong; +Cc: linux-xfs
Hi all,
Create a vectorized version of the metadata scrub and repair ioctl, and
adapt xfs_scrub to use that. This is an experiment to measure overhead
and to try refactoring xfs_scrub.
If you're going to start using this code, I strongly recommend pulling
from my git trees, which are linked below.
This has been running on the djcloud for months with no problems. Enjoy!
Comments and questions are, as always, welcome.
--D
kernel git tree:
https://git.kernel.org/cgit/linux/kernel/git/djwong/xfs-linux.git/log/?h=vectorized-scrub
xfsprogs git tree:
https://git.kernel.org/cgit/linux/kernel/git/djwong/xfsprogs-dev.git/log/?h=vectorized-scrub
fstests git tree:
https://git.kernel.org/cgit/linux/kernel/git/djwong/xfstests-dev.git/log/?h=vectorized-scrub
---
fs/xfs/libxfs/xfs_fs.h | 10 ++
fs/xfs/libxfs/xfs_fs_staging.h | 32 ++++++++
fs/xfs/scrub/common.h | 25 ------
fs/xfs/scrub/scrub.c | 168 ++++++++++++++++++++++++++++++++++++++++
fs/xfs/scrub/scrub.h | 64 +++++++++++++++
fs/xfs/scrub/trace.h | 78 ++++++++++++++++++-
fs/xfs/scrub/xfarray.c | 10 +-
fs/xfs/scrub/xfarray.h | 3 +
fs/xfs/scrub/xfile.c | 10 +-
fs/xfs/scrub/xfs_scrub.h | 2
fs/xfs/xfs_ioctl.c | 50 ++++++++++++
11 files changed, 415 insertions(+), 37 deletions(-)
^ permalink raw reply [flat|nested] 185+ messages in thread
* [PATCHSET v13.0 1/6] xfsprogs: retain ILOCK during directory updates
2023-12-31 18:18 [NYE PATCHRIVER 2/4] xfs: online repair part 2 Darrick J. Wong
` (6 preceding siblings ...)
2023-12-31 19:34 ` [PATCHSET v13.0 7/7] xfs: vectorize scrub kernel calls Darrick J. Wong
@ 2023-12-31 19:50 ` Darrick J. Wong
2023-12-31 23:02 ` [PATCH 1/1] xfs: Increase XFS_DEFER_OPS_NR_INODES to 5 Darrick J. Wong
2023-12-31 19:50 ` [PATCHSET v13.0 2/6] xfsprogs: name-value xattr lookups Darrick J. Wong
` (9 subsequent siblings)
17 siblings, 1 reply; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 19:50 UTC (permalink / raw)
To: djwong, cem
Cc: Catherine Hoang, Allison Henderson, catherine.hoang, linux-xfs,
allison.henderson
Hi all,
This series changes the directory update code to retain the ILOCK on all
files involved in a rename until the end of the operation. The upcoming
parent pointers patchset applies parent pointers in a separate chained
update from the actual directory update, which is why it is now
necessary to keep the ILOCK instead of dropping it after the first
transaction in the chain.
As a side effect, we no longer need to hold the IOLOCK during an rmapbt
scan of inodes to serialize the scan with ongoing directory updates.
If you're going to start using this code, I strongly recommend pulling
from my git trees, which are linked below.
This has been running on the djcloud for months with no problems. Enjoy!
Comments and questions are, as always, welcome.
--D
kernel git tree:
https://git.kernel.org/cgit/linux/kernel/git/djwong/xfs-linux.git/log/?h=retain-ilock-during-dir-ops
xfsprogs git tree:
https://git.kernel.org/cgit/linux/kernel/git/djwong/xfsprogs-dev.git/log/?h=retain-ilock-during-dir-ops
---
libxfs/libxfs_priv.h | 2 ++
libxfs/xfs_defer.c | 6 +++++-
libxfs/xfs_defer.h | 8 +++++++-
3 files changed, 14 insertions(+), 2 deletions(-)
^ permalink raw reply [flat|nested] 185+ messages in thread
* [PATCHSET v13.0 2/6] xfsprogs: name-value xattr lookups
2023-12-31 18:18 [NYE PATCHRIVER 2/4] xfs: online repair part 2 Darrick J. Wong
` (7 preceding siblings ...)
2023-12-31 19:50 ` [PATCHSET v13.0 1/6] xfsprogs: retain ILOCK during directory updates Darrick J. Wong
@ 2023-12-31 19:50 ` Darrick J. Wong
2023-12-31 23:02 ` [PATCH 01/11] xfs: allow newer INCOMPAT/RO_COMPAT feature bits to protect ATTRI log items Darrick J. Wong
` (10 more replies)
2023-12-31 19:50 ` [PATCHSET v13.0 3/6] xfsprogs: Parent Pointers Darrick J. Wong
` (8 subsequent siblings)
17 siblings, 11 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 19:50 UTC (permalink / raw)
To: djwong, cem
Cc: Allison Henderson, catherine.hoang, linux-xfs, allison.henderson
Hi all,
Directory parent pointers are stored as namespaced extended attributes
of a file. Because parent pointers can consume up to 267 bytes of
space and xattr names are 255 bytes at most, we cannot use the usual
attr name lookup functions to find a parent pointer. This is solvable
by introducing a new lookup mode that checks both the name and the
value of the xattr.
Therefore, introduce this new lookup mode. Because all parent pointer
updates are logged, we must extend the xattr logging code to capture the
VLOOKUP variants, and restore them when recovering logged operations.
These new log formats are protected by the sb_incompat PARENT flag, so
they do not need a separate log_incompat feature flag.
If you're going to start using this code, I strongly recommend pulling
from my git trees, which are linked below.
This has been running on the djcloud for months with no problems. Enjoy!
Comments and questions are, as always, welcome.
--D
kernel git tree:
https://git.kernel.org/cgit/linux/kernel/git/djwong/xfs-linux.git/log/?h=pptrs-attr-nvlookups
xfsprogs git tree:
https://git.kernel.org/cgit/linux/kernel/git/djwong/xfsprogs-dev.git/log/?h=pptrs-attr-nvlookups
xfsdocs git tree:
https://git.kernel.org/cgit/linux/kernel/git/djwong/xfs-documentation.git/log/?h=pptrs-attr-nvlookups
---
db/attrset.c | 4 +
libxfs/xfs_attr.c | 52 +++++++++++++-----
libxfs/xfs_attr.h | 32 ++++++++++-
libxfs/xfs_attr_leaf.c | 45 +++++++++++++--
libxfs/xfs_da_btree.h | 10 +++
libxfs/xfs_format.h | 6 ++
libxfs/xfs_log_format.h | 29 +++++++++-
logprint/log_redo.c | 138 ++++++++++++++++++++++++++++++++++++++---------
logprint/logprint.h | 6 +-
9 files changed, 263 insertions(+), 59 deletions(-)
^ permalink raw reply [flat|nested] 185+ messages in thread
* [PATCHSET v13.0 3/6] xfsprogs: Parent Pointers
2023-12-31 18:18 [NYE PATCHRIVER 2/4] xfs: online repair part 2 Darrick J. Wong
` (8 preceding siblings ...)
2023-12-31 19:50 ` [PATCHSET v13.0 2/6] xfsprogs: name-value xattr lookups Darrick J. Wong
@ 2023-12-31 19:50 ` Darrick J. Wong
2023-12-31 23:05 ` [PATCH 01/32] xfs: add parent pointer support to attribute code Darrick J. Wong
` (31 more replies)
2023-12-31 19:51 ` [PATCHSET v13.0 4/6] xfsprogs: fsck for " Darrick J. Wong
` (7 subsequent siblings)
17 siblings, 32 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 19:50 UTC (permalink / raw)
To: djwong, cem
Cc: Allison Henderson, Darrick J. Wong, Mark Tinguely, Dave Chinner,
catherine.hoang, linux-xfs, allison.henderson
Hi all,
This is the latest parent pointer attributes for xfs. The goal of this
patch set is to add a parent pointer attribute to each inode. The
attribute name containing the parent inode, generation, and directory
offset, while the attribute value contains the file name. This feature
will enable future optimizations for online scrub, shrink, nfs handles,
verity, or any other feature that could make use of quickly deriving an
inodes path from the mount point.
At this point, Allison is moving on to other things, so I've merged her
patchset into djwong-dev for merging.
Updates since v11 [djwong]:
Rebase on 6.4-rc and make some tweaks and bugfixes to enable the repair
prototypes. Merge with djwong-dev and make online repair actually work.
Updates since v10 [djwong]:
Merge in the ondisk format changes to get rid of the diroffset conflicts
with the parent pointer repair code, rebase the entire series with the
attr vlookup changes first, and merge all the other random fixes.
Updates since v9:
Reordered patches 2 and 3 to be 6 and 7
xfs: Add xfs_verify_pptr
moved parent pointer validators to xfs_parent
xfs: Add parent pointer ioctl
Extra validation checks for fs id
added missing release for the inode
use GFP_KERNEL flags for malloc/realloc
reworked ioctl to use pptr listenty and flex array
NEW
xfs: don't remove the attr fork when parent pointers are enabled
NEW
directory lookups should return diroffsets too
NEW
xfs: move/add parent pointer validators to xfs_parent
Updates since v8:
xfs: parent pointer attribute creation
Fix xfs_parent_init to release log assist on alloc fail
Add slab cache for xfs_parent_defer
Fix xfs_create to release after unlock
Add xfs_parent_start and xfs_parent_finish wrappers
removed unused xfs_parent_name_irec and xfs_init_parent_name_irec
xfs: add parent attributes to link
Start/finish wrapper updates
Fix xfs_link to disallow reservationless quotas
xfs: add parent attributes to symlink
Fix xfs_symlink to release after unlock
Start/finish wrapper updates
xfs: remove parent pointers in unlink
Start/finish wrapper updates
Add missing parent free
xfs: Add parent pointers to rename
Start/finish wrapper updates
Fix rename to only grab logged xattr once
Fix xfs_rename to disallow reservationless quotas
Fix double unlock on dqattach fail
Move parent frees to out_release_wip
xfs: Add parent pointers to xfs_cross_rename
Hoist parent pointers into rename
Questions comments and feedback appreciated!
Thanks all!
Allison
If you're going to start using this code, I strongly recommend pulling
from my git trees, which are linked below.
This has been running on the djcloud for months with no problems. Enjoy!
Comments and questions are, as always, welcome.
--D
kernel git tree:
https://git.kernel.org/cgit/linux/kernel/git/djwong/xfs-linux.git/log/?h=pptrs
xfsprogs git tree:
https://git.kernel.org/cgit/linux/kernel/git/djwong/xfsprogs-dev.git/log/?h=pptrs
fstests git tree:
https://git.kernel.org/cgit/linux/kernel/git/djwong/xfstests-dev.git/log/?h=pptrs
xfsdocs git tree:
https://git.kernel.org/cgit/linux/kernel/git/djwong/xfs-documentation.git/log/?h=pptrs
---
db/attr.c | 67 +++++
db/attrshort.c | 51 ++++
db/metadump.c | 325 +++++++++++++++++++++++-
db/namei.c | 335 +++++++++++++++++++++++++
db/sb.c | 2
include/handle.h | 1
include/libxfs.h | 1
include/xfs_inode.h | 6
io/parent.c | 527 ++++++++++++++-------------------------
libfrog/Makefile | 2
libfrog/fsgeom.c | 6
libfrog/getparents.c | 348 ++++++++++++++++++++++++++
libfrog/getparents.h | 36 +++
libfrog/paths.c | 183 ++++++++++++++
libfrog/paths.h | 27 ++
libhandle/handle.c | 7 -
libxfs/Makefile | 3
libxfs/init.c | 7 +
libxfs/libxfs_api_defs.h | 16 +
libxfs/libxfs_priv.h | 6
libxfs/util.c | 14 +
libxfs/xfs_attr.c | 15 +
libxfs/xfs_attr.h | 10 -
libxfs/xfs_attr_leaf.c | 6
libxfs/xfs_attr_sf.h | 1
libxfs/xfs_da_format.h | 33 ++
libxfs/xfs_format.h | 4
libxfs/xfs_fs.h | 2
libxfs/xfs_fs_staging.h | 66 +++++
libxfs/xfs_log_format.h | 1
libxfs/xfs_log_rlimit.c | 43 +++
libxfs/xfs_ondisk.h | 4
libxfs/xfs_parent.c | 369 +++++++++++++++++++++++++++
libxfs/xfs_parent.h | 155 +++++++++++
libxfs/xfs_sb.c | 4
libxfs/xfs_trans_resv.c | 324 ++++++++++++++++++++----
libxfs/xfs_trans_space.c | 121 +++++++++
libxfs/xfs_trans_space.h | 25 +-
logprint/log_redo.c | 81 ++++++
man/man2/ioctl_xfs_getparents.2 | 227 +++++++++++++++++
man/man8/xfs_db.8 | 9 +
man/man8/xfs_io.8 | 30 +-
man/man8/xfs_spaceman.8 | 7 -
mkfs/proto.c | 60 +++-
mkfs/xfs_mkfs.c | 31 ++
repair/attr_repair.c | 25 +-
repair/phase6.c | 14 +
scrub/common.c | 41 +++
spaceman/Makefile | 4
spaceman/file.c | 7 +
spaceman/health.c | 53 +++-
spaceman/space.h | 3
52 files changed, 3247 insertions(+), 498 deletions(-)
create mode 100644 libfrog/getparents.c
create mode 100644 libfrog/getparents.h
create mode 100644 libxfs/xfs_parent.c
create mode 100644 libxfs/xfs_parent.h
create mode 100644 libxfs/xfs_trans_space.c
create mode 100644 man/man2/ioctl_xfs_getparents.2
^ permalink raw reply [flat|nested] 185+ messages in thread
* [PATCHSET v13.0 4/6] xfsprogs: fsck for parent pointers
2023-12-31 18:18 [NYE PATCHRIVER 2/4] xfs: online repair part 2 Darrick J. Wong
` (9 preceding siblings ...)
2023-12-31 19:50 ` [PATCHSET v13.0 3/6] xfsprogs: Parent Pointers Darrick J. Wong
@ 2023-12-31 19:51 ` Darrick J. Wong
2023-12-31 23:13 ` [PATCH 01/18] xfs: create a blob array data structure Darrick J. Wong
` (17 more replies)
2023-12-31 19:51 ` [PATCHSET v13.0 5/6] xfs: detect and correct directory tree problems Darrick J. Wong
` (6 subsequent siblings)
17 siblings, 18 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 19:51 UTC (permalink / raw)
To: djwong, cem; +Cc: catherine.hoang, linux-xfs, allison.henderson
Hi all,
This series implements online checking and repair for directory parent
pointer metadata. The checking half is fairly straightforward -- for
each outgoing directory link (forward or backwards), grab the inode at
the other end, and confirm that there's a corresponding link. If we
can't grab an inode or lock it, we'll save that link for a slower loop
that cycles all the locks, confirms the continued existence of the link,
and rechecks the link if it's actually still there.
Repairs are a bit more involved -- for directories, we walk the entire
filesystem to rebuild the dirents from parent pointer information.
Parent pointer repairs do the same walk but rebuild the pptrs from the
dirent information, but with the added twist that it duplicates all the
xattrs so that it can use the atomic extent swapping code to commit the
repairs atomically.
This introduces an added twist to the xattr repair code -- we use dirent
hooks to detect a colliding update to the pptr data while we're not
holding the ILOCKs; if one is detected, we restart the xattr salvaging
process but this time hold all the ILOCKs until the end of the scan.
For offline repair, the phase6 directory connectivity scan generates an
index of all the expected parent pointers in the filesystem. Then it
walks each file and compares the parent pointers attached to that file
against the index generated, and resyncs the results as necessary.
The last patch teaches xfs_scrub to report pathnames of files that are
being repaired, when possible.
If you're going to start using this code, I strongly recommend pulling
from my git trees, which are linked below.
This has been running on the djcloud for months with no problems. Enjoy!
Comments and questions are, as always, welcome.
--D
kernel git tree:
https://git.kernel.org/cgit/linux/kernel/git/djwong/xfs-linux.git/log/?h=pptrs-fsck
xfsprogs git tree:
https://git.kernel.org/cgit/linux/kernel/git/djwong/xfsprogs-dev.git/log/?h=pptrs-fsck
---
libxfs/Makefile | 2
libxfs/libxfs_api_defs.h | 4
libxfs/xfblob.c | 156 ++++
libxfs/xfblob.h | 26 +
libxfs/xfile.c | 11
libxfs/xfile.h | 1
libxfs/xfs_attr.c | 39 +
libxfs/xfs_attr.h | 2
libxfs/xfs_bmap.c | 38 -
libxfs/xfs_bmap.h | 3
libxfs/xfs_dir2.c | 2
libxfs/xfs_dir2.h | 2
libxfs/xfs_parent.c | 107 +++
libxfs/xfs_parent.h | 18
man/man2/ioctl_xfs_scrub_metadata.2 | 20 -
man/man8/xfs_admin.8 | 8
repair/Makefile | 6
repair/globals.c | 1
repair/globals.h | 1
repair/listxattr.c | 271 +++++++
repair/listxattr.h | 15
repair/phase2.c | 39 +
repair/phase6.c | 118 +++
repair/pptr.c | 1303 +++++++++++++++++++++++++++++++++++
repair/pptr.h | 17
repair/strblobs.c | 211 ++++++
repair/strblobs.h | 24 +
repair/xfs_repair.c | 11
scrub/phase6.c | 75 ++
29 files changed, 2481 insertions(+), 50 deletions(-)
create mode 100644 libxfs/xfblob.c
create mode 100644 libxfs/xfblob.h
create mode 100644 repair/listxattr.c
create mode 100644 repair/listxattr.h
create mode 100644 repair/pptr.c
create mode 100644 repair/pptr.h
create mode 100644 repair/strblobs.c
create mode 100644 repair/strblobs.h
^ permalink raw reply [flat|nested] 185+ messages in thread
* [PATCHSET v13.0 5/6] xfs: detect and correct directory tree problems
2023-12-31 18:18 [NYE PATCHRIVER 2/4] xfs: online repair part 2 Darrick J. Wong
` (10 preceding siblings ...)
2023-12-31 19:51 ` [PATCHSET v13.0 4/6] xfsprogs: fsck for " Darrick J. Wong
@ 2023-12-31 19:51 ` Darrick J. Wong
2023-12-31 23:18 ` [PATCH 1/6] xfs_db: add link and unlink expert commands Darrick J. Wong
` (5 more replies)
2023-12-31 19:51 ` [PATCHSET 6/6] xfs_scrub: vectorize kernel calls Darrick J. Wong
` (5 subsequent siblings)
17 siblings, 6 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 19:51 UTC (permalink / raw)
To: djwong, cem; +Cc: linux-xfs
Hi all,
Historically, checking the tree-ness of the directory tree structure has
not been complete. Cycles of subdirectories break the tree properties,
as do subdirectories with multiple parents. It's easy enough for DFS to
detect problems as long as one of the participants is reachable from the
root, but this technique cannot find unconnected cycles.
Directory parent pointers change that, because we can discover all of
these problems from a simple walk from a subdirectory towards the root.
For each child we start with, if the walk terminates without reaching
the root, we know the path is disconnected and ought to be attached to
the lost and found. If we find ourselves, we know this is a cycle and
can delete an incoming edge. If we find multiple paths to the root, we
know to delete an incoming edge.
Even better, once we've finished walking paths, we've identified the
good ones and know which other path(s) to remove.
If you're going to start using this code, I strongly recommend pulling
from my git trees, which are linked below.
This has been running on the djcloud for months with no problems. Enjoy!
Comments and questions are, as always, welcome.
--D
kernel git tree:
https://git.kernel.org/cgit/linux/kernel/git/djwong/xfs-linux.git/log/?h=scrub-directory-tree
xfsprogs git tree:
https://git.kernel.org/cgit/linux/kernel/git/djwong/xfsprogs-dev.git/log/?h=scrub-directory-tree
fstests git tree:
https://git.kernel.org/cgit/linux/kernel/git/djwong/xfstests-dev.git/log/?h=scrub-directory-tree
---
db/namei.c | 374 +++++++++++++++++++++++++++++++++++
include/xfs_inode.h | 4
libfrog/scrub.c | 5
libxfs/libxfs_api_defs.h | 6 +
libxfs/xfs_fs.h | 4
libxfs/xfs_health.h | 4
man/man2/ioctl_xfs_bulkstat.2 | 3
man/man2/ioctl_xfs_fsbulkstat.2 | 3
man/man2/ioctl_xfs_scrub_metadata.2 | 14 +
man/man8/xfs_db.8 | 20 ++
scrub/phase5.c | 271 ++++++++++++++++++++++++-
scrub/repair.c | 13 +
scrub/repair.h | 2
spaceman/health.c | 4
14 files changed, 711 insertions(+), 16 deletions(-)
^ permalink raw reply [flat|nested] 185+ messages in thread
* [PATCHSET 6/6] xfs_scrub: vectorize kernel calls
2023-12-31 18:18 [NYE PATCHRIVER 2/4] xfs: online repair part 2 Darrick J. Wong
` (11 preceding siblings ...)
2023-12-31 19:51 ` [PATCHSET v13.0 5/6] xfs: detect and correct directory tree problems Darrick J. Wong
@ 2023-12-31 19:51 ` Darrick J. Wong
2023-12-31 23:19 ` [PATCH 01/10] xfs: introduce vectored scrub mode Darrick J. Wong
` (9 more replies)
2023-12-31 19:59 ` [PATCHSET v13.0 1/3] fstests: adjust tests for xfs parent pointers Darrick J. Wong
` (4 subsequent siblings)
17 siblings, 10 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 19:51 UTC (permalink / raw)
To: djwong, cem; +Cc: linux-xfs
Hi all,
Create a vectorized version of the metadata scrub and repair ioctl, and
adapt xfs_scrub to use that. This is an experiment to measure overhead
and to try refactoring xfs_scrub.
If you're going to start using this code, I strongly recommend pulling
from my git trees, which are linked below.
This has been running on the djcloud for months with no problems. Enjoy!
Comments and questions are, as always, welcome.
--D
kernel git tree:
https://git.kernel.org/cgit/linux/kernel/git/djwong/xfs-linux.git/log/?h=vectorized-scrub
xfsprogs git tree:
https://git.kernel.org/cgit/linux/kernel/git/djwong/xfsprogs-dev.git/log/?h=vectorized-scrub
fstests git tree:
https://git.kernel.org/cgit/linux/kernel/git/djwong/xfstests-dev.git/log/?h=vectorized-scrub
---
io/scrub.c | 371 ++++++++++++++++++++++++++++++----
libfrog/fsgeom.h | 6 +
libfrog/scrub.c | 124 +++++++++++
libfrog/scrub.h | 4
libxfs/xfs_fs.h | 10 +
libxfs/xfs_fs_staging.h | 32 +++
man/man2/ioctl_xfs_scrubv_metadata.2 | 168 +++++++++++++++
man/man8/xfs_io.8 | 51 +++++
scrub/phase1.c | 2
scrub/phase2.c | 93 +++++++--
scrub/phase3.c | 84 ++++++--
scrub/repair.c | 354 ++++++++++++++++++++------------
scrub/scrub.c | 348 ++++++++++++++++++++++++--------
scrub/scrub.h | 19 ++
scrub/scrub_private.h | 62 ++++--
15 files changed, 1422 insertions(+), 306 deletions(-)
create mode 100644 man/man2/ioctl_xfs_scrubv_metadata.2
^ permalink raw reply [flat|nested] 185+ messages in thread
* [PATCHSET v13.0 1/3] fstests: adjust tests for xfs parent pointers
2023-12-31 18:18 [NYE PATCHRIVER 2/4] xfs: online repair part 2 Darrick J. Wong
` (12 preceding siblings ...)
2023-12-31 19:51 ` [PATCHSET 6/6] xfs_scrub: vectorize kernel calls Darrick J. Wong
@ 2023-12-31 19:59 ` Darrick J. Wong
2023-12-27 13:46 ` [PATCH 01/11] generic: test recovery of extended attribute updates Darrick J. Wong
` (10 more replies)
2023-12-31 19:59 ` [PATCHSET v13.0 2/3] xfs: detect and correct directory tree structures Darrick J. Wong
` (3 subsequent siblings)
17 siblings, 11 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 19:59 UTC (permalink / raw)
To: zlang, djwong
Cc: Allison Henderson, Catherine Hoang, fstests, catherine.hoang,
allison.henderson, guan, linux-xfs
Hi all,
These are the test adjustments that are required for parent pointers.
There's also a few new tests to ensure that the GETPARENTS ioctl (and
file extent can cross an rtgroup boundary.
The best advantage for rtgroups will become evident later when we get to
adding rmap and reflink to the realtime volume, since the geometry
constraints are the same for rt groups and AGs. Hence we can reuse all
that code directly.
If you're going to start using this code, I strongly recommend pulling
from my git trees, which are linked below.
This has been running on the djcloud for months with no problems. Enjoy!
Comments and questions are, as always, welcome.
--D
kernel git tree:
https://git.kernel.org/cgit/linux/kernel/git/djwong/xfs-linux.git/log/?h=pptrs
xfsprogs git tree:
https://git.kernel.org/cgit/linux/kernel/git/djwong/xfsprogs-dev.git/log/?h=pptrs
fstests git tree:
https://git.kernel.org/cgit/linux/kernel/git/djwong/xfstests-dev.git/log/?h=pptrs
xfsdocs git tree:
https://git.kernel.org/cgit/linux/kernel/git/djwong/xfs-documentation.git/log/?h=pptrs
---
common/parent | 209 +++++++++
common/populate | 38 ++
common/rc | 7
common/xfs | 27 +
doc/group-names.txt | 1
src/popdir.pl | 11
tests/generic/1834 | 93 ++++
tests/generic/1834.out | 2
tests/xfs/018 | 4
tests/xfs/021 | 15 +
tests/xfs/021.cfg | 1
tests/xfs/021.out.default | 0
tests/xfs/021.out.parent | 62 +++
tests/xfs/122.out | 3
tests/xfs/1851 | 116 +++++
tests/xfs/1851.out | 69 +++
tests/xfs/1852 | 69 +++
tests/xfs/1852.out | 1002 +++++++++++++++++++++++++++++++++++++++++++++
tests/xfs/1853 | 85 ++++
tests/xfs/1853.out | 14 +
tests/xfs/191 | 3
tests/xfs/206 | 3
tests/xfs/288 | 4
tests/xfs/306 | 9
24 files changed, 1839 insertions(+), 8 deletions(-)
create mode 100644 common/parent
create mode 100755 tests/generic/1834
create mode 100644 tests/generic/1834.out
create mode 100644 tests/xfs/021.cfg
rename tests/xfs/{021.out => 021.out.default} (100%)
create mode 100644 tests/xfs/021.out.parent
create mode 100755 tests/xfs/1851
create mode 100644 tests/xfs/1851.out
create mode 100755 tests/xfs/1852
create mode 100644 tests/xfs/1852.out
create mode 100755 tests/xfs/1853
create mode 100644 tests/xfs/1853.out
^ permalink raw reply [flat|nested] 185+ messages in thread
* [PATCHSET v13.0 2/3] xfs: detect and correct directory tree structures
2023-12-31 18:18 [NYE PATCHRIVER 2/4] xfs: online repair part 2 Darrick J. Wong
` (13 preceding siblings ...)
2023-12-31 19:59 ` [PATCHSET v13.0 1/3] fstests: adjust tests for xfs parent pointers Darrick J. Wong
@ 2023-12-31 19:59 ` Darrick J. Wong
2023-12-27 13:49 ` [PATCH 1/2] common/fuzzy: stress directory tree modifications with the dirtree tester Darrick J. Wong
2023-12-27 13:49 ` [PATCH 2/2] scrub: test correction of directory tree corruptions Darrick J. Wong
2023-12-31 19:59 ` [PATCHSET 3/3] xfs_scrub: vectorize kernel calls Darrick J. Wong
` (2 subsequent siblings)
17 siblings, 2 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 19:59 UTC (permalink / raw)
To: zlang, djwong; +Cc: linux-xfs, guan, fstests
Hi all,
Historically, checking the tree-ness of the directory tree structure has
not been complete. Cycles of subdirectories break the tree properties,
as do subdirectories with multiple parents. It's easy enough for DFS to
detect problems as long as one of the participants is reachable from the
root, but this technique cannot find unconnected cycles.
Directory parent pointers change that, because we can discover all of
these problems from a simple walk from a subdirectory towards the root.
For each child we start with, if the walk terminates without reaching
the root, we know the path is disconnected and ought to be attached to
the lost and found. If we find ourselves, we know this is a cycle and
can delete an incoming edge. If we find multiple paths to the root, we
know to delete an incoming edge.
Even better, once we've finished walking paths, we've identified the
good ones and know which other path(s) to remove.
If you're going to start using this code, I strongly recommend pulling
from my git trees, which are linked below.
This has been running on the djcloud for months with no problems. Enjoy!
Comments and questions are, as always, welcome.
--D
kernel git tree:
https://git.kernel.org/cgit/linux/kernel/git/djwong/xfs-linux.git/log/?h=scrub-directory-tree
xfsprogs git tree:
https://git.kernel.org/cgit/linux/kernel/git/djwong/xfsprogs-dev.git/log/?h=scrub-directory-tree
fstests git tree:
https://git.kernel.org/cgit/linux/kernel/git/djwong/xfstests-dev.git/log/?h=scrub-directory-tree
---
tests/xfs/1864 | 38 +++++++++++++
tests/xfs/1864.out | 2 +
tests/xfs/1865 | 38 +++++++++++++
tests/xfs/1865.out | 2 +
tests/xfs/1866 | 122 ++++++++++++++++++++++++++++++++++++++++
tests/xfs/1866.out | 19 ++++++
tests/xfs/1867 | 133 ++++++++++++++++++++++++++++++++++++++++++++
tests/xfs/1867.out | 25 ++++++++
tests/xfs/1868 | 121 ++++++++++++++++++++++++++++++++++++++++
tests/xfs/1868.out | 21 +++++++
tests/xfs/1869 | 157 ++++++++++++++++++++++++++++++++++++++++++++++++++++
tests/xfs/1869.out | 32 +++++++++++
tests/xfs/1870 | 146 ++++++++++++++++++++++++++++++++++++++++++++++++
tests/xfs/1870.out | 30 ++++++++++
tests/xfs/1871 | 78 ++++++++++++++++++++++++++
tests/xfs/1871.out | 2 +
16 files changed, 966 insertions(+)
create mode 100755 tests/xfs/1864
create mode 100644 tests/xfs/1864.out
create mode 100755 tests/xfs/1865
create mode 100644 tests/xfs/1865.out
create mode 100755 tests/xfs/1866
create mode 100644 tests/xfs/1866.out
create mode 100755 tests/xfs/1867
create mode 100644 tests/xfs/1867.out
create mode 100755 tests/xfs/1868
create mode 100644 tests/xfs/1868.out
create mode 100755 tests/xfs/1869
create mode 100644 tests/xfs/1869.out
create mode 100755 tests/xfs/1870
create mode 100644 tests/xfs/1870.out
create mode 100755 tests/xfs/1871
create mode 100644 tests/xfs/1871.out
^ permalink raw reply [flat|nested] 185+ messages in thread
* [PATCHSET 3/3] xfs_scrub: vectorize kernel calls
2023-12-31 18:18 [NYE PATCHRIVER 2/4] xfs: online repair part 2 Darrick J. Wong
` (14 preceding siblings ...)
2023-12-31 19:59 ` [PATCHSET v13.0 2/3] xfs: detect and correct directory tree structures Darrick J. Wong
@ 2023-12-31 19:59 ` Darrick J. Wong
2023-12-27 13:49 ` [PATCH 1/1] xfs/122: update for vectored scrub Darrick J. Wong
2023-12-31 20:03 ` [PATCHSET v13.0 1/2] xfs-documentation: document attr log item changes for parent pointers Darrick J. Wong
2023-12-31 20:03 ` [PATCHSET v13.0 2/2] xfs-documentation: document parent pointers Darrick J. Wong
17 siblings, 1 reply; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 19:59 UTC (permalink / raw)
To: zlang, djwong; +Cc: linux-xfs, guan, fstests
Hi all,
Create a vectorized version of the metadata scrub and repair ioctl, and
adapt xfs_scrub to use that. This is an experiment to measure overhead
and to try refactoring xfs_scrub.
If you're going to start using this code, I strongly recommend pulling
from my git trees, which are linked below.
This has been running on the djcloud for months with no problems. Enjoy!
Comments and questions are, as always, welcome.
--D
kernel git tree:
https://git.kernel.org/cgit/linux/kernel/git/djwong/xfs-linux.git/log/?h=vectorized-scrub
xfsprogs git tree:
https://git.kernel.org/cgit/linux/kernel/git/djwong/xfsprogs-dev.git/log/?h=vectorized-scrub
fstests git tree:
https://git.kernel.org/cgit/linux/kernel/git/djwong/xfstests-dev.git/log/?h=vectorized-scrub
---
tests/xfs/122.out | 2 ++
1 file changed, 2 insertions(+)
^ permalink raw reply [flat|nested] 185+ messages in thread
* [PATCHSET v13.0 1/2] xfs-documentation: document attr log item changes for parent pointers
2023-12-31 18:18 [NYE PATCHRIVER 2/4] xfs: online repair part 2 Darrick J. Wong
` (15 preceding siblings ...)
2023-12-31 19:59 ` [PATCHSET 3/3] xfs_scrub: vectorize kernel calls Darrick J. Wong
@ 2023-12-31 20:03 ` Darrick J. Wong
2023-12-27 14:07 ` [PATCH 1/1] design: document new name-value logged attribute variants Darrick J. Wong
2023-12-31 20:03 ` [PATCHSET v13.0 2/2] xfs-documentation: document parent pointers Darrick J. Wong
17 siblings, 1 reply; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:03 UTC (permalink / raw)
To: djwong, darrick.wong; +Cc: linux-xfs
Hi all,
This patch documents the changes to the ondisk format of attr log items
that are needed to support directory parent pointers.
If you're going to start using this code, I strongly recommend pulling
from my git trees, which are linked below.
Comments and questions are, as always, welcome.
kernel git tree:
https://git.kernel.org/cgit/linux/kernel/git/djwong/xfs-linux.git/log/?h=pptrs-attr-nvlookups
xfsprogs git tree:
https://git.kernel.org/cgit/linux/kernel/git/djwong/xfsprogs-dev.git/log/?h=pptrs-attr-nvlookups
xfsdocs git tree:
https://git.kernel.org/cgit/linux/kernel/git/djwong/xfs-documentation.git/log/?h=pptrs-attr-nvlookups
---
.../journaling_log.asciidoc | 60 ++++++++++++++++++--
1 file changed, 54 insertions(+), 6 deletions(-)
^ permalink raw reply [flat|nested] 185+ messages in thread
* [PATCHSET v13.0 2/2] xfs-documentation: document parent pointers
2023-12-31 18:18 [NYE PATCHRIVER 2/4] xfs: online repair part 2 Darrick J. Wong
` (16 preceding siblings ...)
2023-12-31 20:03 ` [PATCHSET v13.0 1/2] xfs-documentation: document attr log item changes for parent pointers Darrick J. Wong
@ 2023-12-31 20:03 ` Darrick J. Wong
2023-12-27 14:07 ` [PATCH 1/1] design: document the parent pointer ondisk format Darrick J. Wong
17 siblings, 1 reply; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:03 UTC (permalink / raw)
To: djwong, darrick.wong; +Cc: linux-xfs
Hi all,
This patch documents the parent pointers tree feature.
If you're going to start using this code, I strongly recommend pulling
from my git trees, which are linked below.
Comments and questions are, as always, welcome.
kernel git tree:
https://git.kernel.org/cgit/linux/kernel/git/djwong/xfs-linux.git/log/?h=pptrs
xfsprogs git tree:
https://git.kernel.org/cgit/linux/kernel/git/djwong/xfsprogs-dev.git/log/?h=pptrs
fstests git tree:
https://git.kernel.org/cgit/linux/kernel/git/djwong/xfstests-dev.git/log/?h=pptrs
xfsdocs git tree:
https://git.kernel.org/cgit/linux/kernel/git/djwong/xfs-documentation.git/log/?h=pptrs
---
.../allocation_groups.asciidoc | 4 +
.../extended_attributes.asciidoc | 94 ++++++++++++++++++++
2 files changed, 98 insertions(+)
^ permalink raw reply [flat|nested] 185+ messages in thread
* [PATCH 1/4] docs: update the parent pointers documentation to the final version
2023-12-31 19:32 ` [PATCHSET v13.0 1/7] xfs: design documentation for online fsck, " Darrick J. Wong
@ 2023-12-31 20:42 ` Darrick J. Wong
2023-12-31 20:42 ` [PATCH 2/4] docs: update online directory and parent pointer repair sections Darrick J. Wong
` (2 subsequent siblings)
3 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:42 UTC (permalink / raw)
To: djwong; +Cc: linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
Now that we've decided on the ondisk format of parent pointers, update
the documentation to reflect that.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
.../filesystems/xfs-online-fsck-design.rst | 91 +++++++++++---------
1 file changed, 50 insertions(+), 41 deletions(-)
diff --git a/Documentation/filesystems/xfs-online-fsck-design.rst b/Documentation/filesystems/xfs-online-fsck-design.rst
index 827fcd49fe6d5..8fb0fc30f3fa4 100644
--- a/Documentation/filesystems/xfs-online-fsck-design.rst
+++ b/Documentation/filesystems/xfs-online-fsck-design.rst
@@ -4464,10 +4464,11 @@ reconstruction of filesystem space metadata.
The parent pointer feature, however, makes total directory reconstruction
possible.
-XFS parent pointers include the dirent name and location of the entry within
-the parent directory.
+XFS parent pointers contain the information needed to identify the
+corresponding directory entry in the parent directory.
In other words, child files use extended attributes to store pointers to
-parents in the form ``(parent_inum, parent_gen, dirent_pos) → (dirent_name)``.
+parents in the form ``(parent_inum, parent_gen, dirent_name_hash) →
+(dirent_name)``.
The directory checking process can be strengthened to ensure that the target of
each dirent also contains a parent pointer pointing back to the dirent.
Likewise, each parent pointer can be checked by ensuring that the target of
@@ -4475,8 +4476,6 @@ each parent pointer is a directory and that it contains a dirent matching
the parent pointer.
Both online and offline repair can use this strategy.
-**Note**: The ondisk format of parent pointers is not yet finalized.
-
+--------------------------------------------------------------------------+
| **Historical Sidebar**: |
+--------------------------------------------------------------------------+
@@ -4518,8 +4517,54 @@ Both online and offline repair can use this strategy.
| Chandan increased the maximum extent counts of both data and attribute |
| forks, thereby ensuring that the extended attribute structure can grow |
| to handle the maximum hardlink count of any file. |
+| |
+| For this second effort, the ondisk parent pointer format as originally |
+| proposed was ``(parent_inum, parent_gen, dirent_pos) → (dirent_name)``. |
+| The format was changed during development to eliminate the requirement |
+| of repair tools needing to to ensure that the ``dirent_pos`` field |
+| always matched when reconstructing a directory. |
+| |
+| There were a few other ways to have solved that problem: |
+| |
+| 1. The field could be designated advisory, since the other three values |
+| are sufficient to find the entry in the parent. |
+| However, this makes indexed key lookup impossible while repairs are |
+| ongoing. |
+| |
+| 2. We could allow creating directory entries at specified offsets, which |
+| solves the referential integrity problem but runs the risk that |
+| dirent creation will fail due to conflicts with the free space in the |
+| directory. |
+| |
+| These conflicts could be resolved by appending the directory entry |
+| and amending the xattr code to support updating an xattr key and |
+| reindexing the dabtree, though this would have to be performed with |
+| the parent directory still locked. |
+| |
+| 3. Same as above, but remove the old parent pointer entry and add a new |
+| one atomically. |
+| |
+| 4. Change the ondisk xattr format to |
+| ``(parent_inum, name) → (parent_gen)``, which would provide the attr |
+| name uniqueness that we require, without forcing repair code to |
+| update the dirent position. |
+| Unfortunately, this requires changes to the xattr code to support |
+| attr names as long as 263 bytes. |
+| |
+| 5. Change the ondisk xattr format to ``(parent_inum, hash(name)) → |
+| (name, parent_gen)``. |
+| If the hash is sufficiently resistant to collisions (e.g. sha256) |
+| then this should provide the attr name uniqueness that we require. |
+| Names shorter than 247 bytes could be stored directly. |
+| |
+| In the end, it was decided that the hash collisions of #5 were not a |
+| serious issue because the directory/attr btree can handle multiple |
+| identical extended attribute keys. |
+| Reusing the dirent name hash instead of sha256 is much faster and would |
+| result in a more compact ondisk format. |
+--------------------------------------------------------------------------+
+
Case Study: Repairing Directories with Parent Pointers
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
@@ -4568,42 +4613,6 @@ The proposed patchset is the
<https://git.kernel.org/pub/scm/linux/kernel/git/djwong/xfs-linux.git/log/?h=pptrs-online-dir-repair>`_
series.
-**Unresolved Question**: How will repair ensure that the ``dirent_pos`` fields
-match in the reconstructed directory?
-
-*Answer*: There are a few ways to solve this problem:
-
-1. The field could be designated advisory, since the other three values are
- sufficient to find the entry in the parent.
- However, this makes indexed key lookup impossible while repairs are ongoing.
-
-2. We could allow creating directory entries at specified offsets, which solves
- the referential integrity problem but runs the risk that dirent creation
- will fail due to conflicts with the free space in the directory.
-
- These conflicts could be resolved by appending the directory entry and
- amending the xattr code to support updating an xattr key and reindexing the
- dabtree, though this would have to be performed with the parent directory
- still locked.
-
-3. Same as above, but remove the old parent pointer entry and add a new one
- atomically.
-
-4. Change the ondisk xattr format to ``(parent_inum, name) → (parent_gen)``,
- which would provide the attr name uniqueness that we require, without
- forcing repair code to update the dirent position.
- Unfortunately, this requires changes to the xattr code to support attr
- names as long as 263 bytes.
-
-5. Change the ondisk xattr format to ``(parent_inum, hash(name)) →
- (name, parent_gen)``.
- If the hash is sufficiently resistant to collisions (e.g. sha256) then
- this should provide the attr name uniqueness that we require.
- Names shorter than 247 bytes could be stored directly.
-
-Discussion is ongoing under the `parent pointers patch deluge
-<https://www.spinics.net/lists/linux-xfs/msg69397.html>`_.
-
Case Study: Repairing Parent Pointers
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 2/4] docs: update online directory and parent pointer repair sections
2023-12-31 19:32 ` [PATCHSET v13.0 1/7] xfs: design documentation for online fsck, " Darrick J. Wong
2023-12-31 20:42 ` [PATCH 1/4] docs: update the parent pointers documentation to the final version Darrick J. Wong
@ 2023-12-31 20:42 ` Darrick J. Wong
2023-12-31 20:42 ` [PATCH 3/4] docs: update offline parent pointer repair strategy Darrick J. Wong
2023-12-31 20:43 ` [PATCH 4/4] docs: describe xfs directory tree online fsck Darrick J. Wong
3 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:42 UTC (permalink / raw)
To: djwong; +Cc: linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
Update the case studies of online directory and parent pointer
reconstruction to reflect what they actually do in the final version.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
.../filesystems/xfs-online-fsck-design.rst | 58 +++++++++++---------
1 file changed, 31 insertions(+), 27 deletions(-)
diff --git a/Documentation/filesystems/xfs-online-fsck-design.rst b/Documentation/filesystems/xfs-online-fsck-design.rst
index 8fb0fc30f3fa4..f5ba59b335f8e 100644
--- a/Documentation/filesystems/xfs-online-fsck-design.rst
+++ b/Documentation/filesystems/xfs-online-fsck-design.rst
@@ -4572,8 +4572,9 @@ Directory rebuilding uses a :ref:`coordinated inode scan <iscan>` and
a :ref:`directory entry live update hook <liveupdate>` as follows:
1. Set up a temporary directory for generating the new directory structure,
- an xfblob for storing entry names, and an xfarray for stashing directory
- updates.
+ an xfblob for storing entry names, and an xfarray for stashing the fixed
+ size fields involved in a directory update: ``(child inumber, add vs.
+ remove, name cookie, ftype)``.
2. Set up an inode scanner and hook into the directory entry code to receive
updates on directory operations.
@@ -4582,35 +4583,34 @@ a :ref:`directory entry live update hook <liveupdate>` as follows:
pointer references the directory of interest.
If so:
- a. Stash an addname entry for this dirent in the xfarray for later.
+ a. Stash the parent pointer name and an addname entry for this dirent in the
+ xfblob and xfarray, respectively.
- b. When finished scanning that file, flush the stashed updates to the
- temporary directory.
+ b. When finished scanning that file or the kernel memory consumption exceeds
+ a threshold, flush the stashed updates to the temporary directory.
4. For each live directory update received via the hook, decide if the child
has already been scanned.
If so:
- a. Stash an addname or removename entry for this dirent update in the
- xfarray for later.
+ a. Stash the parent pointer name an addname or removename entry for this
+ dirent update in the xfblob and xfarray for later.
We cannot write directly to the temporary directory because hook
functions are not allowed to modify filesystem metadata.
Instead, we stash updates in the xfarray and rely on the scanner thread
to apply the stashed updates to the temporary directory.
-5. When the scan is complete, atomically swap the contents of the temporary
- directory and the directory being repaired.
+5. When the scan is complete, replay any stashed entries in the xfarray.
+
+6. Atomically swap the contents of the temporary directory and the directory
+ being repaired.
The temporary directory now contains the damaged directory structure.
6. Reap the temporary directory.
-7. Update the dirent position field of parent pointers as necessary.
- This may require the queuing of a substantial number of xattr log intent
- items.
-
The proposed patchset is the
`parent pointers directory repair
-<https://git.kernel.org/pub/scm/linux/kernel/git/djwong/xfs-linux.git/log/?h=pptrs-online-dir-repair>`_
+<https://git.kernel.org/pub/scm/linux/kernel/git/djwong/xfs-linux.git/log/?h=pptrs-fsck>`_
series.
Case Study: Repairing Parent Pointers
@@ -4620,8 +4620,9 @@ Online reconstruction of a file's parent pointer information works similarly to
directory reconstruction:
1. Set up a temporary file for generating a new extended attribute structure,
- an `xfblob<xfblob>` for storing parent pointer names, and an xfarray for
- stashing parent pointer updates.
+ an xfblob for storing parent pointer names, and an xfarray for stashing the
+ fixed size fields involved in a parent pointer update: ``(parent inumber,
+ parent generation, add vs. remove, name cookie)``.
2. Set up an inode scanner and hook into the directory entry code to receive
updates on directory operations.
@@ -4630,34 +4631,37 @@ directory reconstruction:
dirent references the file of interest.
If so:
- a. Stash an addpptr entry for this parent pointer in the xfblob and xfarray
- for later.
+ a. Stash the dirent name and an addpptr entry for this parent pointer in the
+ xfblob and xfarray, respectively.
- b. When finished scanning the directory, flush the stashed updates to the
- temporary directory.
+ b. When finished scanning the directory or the kernel memory consumption
+ exceeds a threshold, flush the stashed updates to the temporary file.
4. For each live directory update received via the hook, decide if the parent
has already been scanned.
If so:
- a. Stash an addpptr or removepptr entry for this dirent update in the
- xfarray for later.
+ a. Stash the dirent name and an addpptr or removepptr entry for this dirent
+ update in the xfblob and xfarray for later.
We cannot write parent pointers directly to the temporary file because
hook functions are not allowed to modify filesystem metadata.
Instead, we stash updates in the xfarray and rely on the scanner thread
to apply the stashed parent pointer updates to the temporary file.
-5. Copy all non-parent pointer extended attributes to the temporary file.
+5. When the scan is complete, replay any stashed entries in the xfarray.
-6. When the scan is complete, atomically swap the attribute fork of the
- temporary file and the file being repaired.
- The temporary file now contains the damaged extended attribute structure.
+6. Copy all non-parent pointer extended attributes to the temporary file.
+
+7. Atomically swap the attribute fork of the temporary file and the file being
+ repaired.
+ The temporary file now contains the old extended attribute structure
+ containing the damaged parent pointers.
7. Reap the temporary file.
The proposed patchset is the
`parent pointers repair
-<https://git.kernel.org/pub/scm/linux/kernel/git/djwong/xfs-linux.git/log/?h=pptrs-online-parent-repair>`_
+<https://git.kernel.org/pub/scm/linux/kernel/git/djwong/xfs-linux.git/log/?h=pptrs-fsck>`_
series.
Digression: Offline Checking of Parent Pointers
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 3/4] docs: update offline parent pointer repair strategy
2023-12-31 19:32 ` [PATCHSET v13.0 1/7] xfs: design documentation for online fsck, " Darrick J. Wong
2023-12-31 20:42 ` [PATCH 1/4] docs: update the parent pointers documentation to the final version Darrick J. Wong
2023-12-31 20:42 ` [PATCH 2/4] docs: update online directory and parent pointer repair sections Darrick J. Wong
@ 2023-12-31 20:42 ` Darrick J. Wong
2023-12-31 20:43 ` [PATCH 4/4] docs: describe xfs directory tree online fsck Darrick J. Wong
3 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:42 UTC (permalink / raw)
To: djwong; +Cc: linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
Now update how xfs_repair checks and repairs parent pointer info.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
.../filesystems/xfs-online-fsck-design.rst | 79 +++++++++++++++-----
1 file changed, 58 insertions(+), 21 deletions(-)
diff --git a/Documentation/filesystems/xfs-online-fsck-design.rst b/Documentation/filesystems/xfs-online-fsck-design.rst
index f5ba59b335f8e..04e445d79f2d4 100644
--- a/Documentation/filesystems/xfs-online-fsck-design.rst
+++ b/Documentation/filesystems/xfs-online-fsck-design.rst
@@ -4672,26 +4672,54 @@ files are erased long before directory tree connectivity checks are performed.
Parent pointer checks are therefore a second pass to be added to the existing
connectivity checks:
-1. After the set of surviving files has been established (i.e. phase 6),
+1. After the set of surviving files has been established (phase 6),
walk the surviving directories of each AG in the filesystem.
This is already performed as part of the connectivity checks.
-2. For each directory entry found, record the name in an xfblob, and store
- ``(child_ag_inum, parent_inum, parent_gen, dirent_pos)`` tuples in a
- per-AG in-memory slab.
+2. For each directory entry found,
+
+ a. If the name has already been stored in the xfblob, then use that cookie
+ and skip the next step.
+
+ b. Otherwise, record the name in an xfblob, and remember the xfblob cookie.
+ Unique mappings are critical for
+
+ 1. Deduplicating names to reduce memory usage, and
+
+ 2. Creating a stable sort key for the parent pointer indexes so that the
+ parent pointer validation described below will work.
+
+ c. Store ``(child_ag_inum, parent_inum, parent_gen, name_hash, name_len,
+ name_cookie)`` tuples in a per-AG in-memory slab.
3. For each AG in the filesystem,
- a. Sort the per-AG tuples in order of child_ag_inum, parent_inum, and
- dirent_pos.
+ a. Sort the per-AG tuple set in order of ``child_ag_inum``, ``parent_inum``,
+ ``name_hash``, and ``name_cookie``.
+ Having a single ``name_cookie`` for each ``name`` is critical for
+ handling the uncommon case of a directory containing multiple hardlinks
+ to the same file where all the names hash to the same value.
b. For each inode in the AG,
1. Scan the inode for parent pointers.
- Record the names in a per-file xfblob, and store ``(parent_inum,
- parent_gen, dirent_pos)`` tuples in a per-file slab.
+ For each parent pointer found,
- 2. Sort the per-file tuples in order of parent_inum, and dirent_pos.
+ a. Validate the ondisk parent pointer.
+ If validation fails, move on to the next parent pointer in the
+ file.
+
+ b. If the name has already been stored in the xfblob, then use that
+ cookie and skip the next step.
+
+ c. Record the name in a per-file xfblob, and remember the xfblob
+ cookie.
+
+ d. Store ``(parent_inum, parent_gen, name_hash, name_len,
+ name_cookie)`` tuples in a per-file slab.
+
+ 2. Sort the per-file tuples in order of ``parent_inum``, ``name_hash``,
+ and ``name_cookie``.
3. Position one slab cursor at the start of the inode's records in the
per-AG tuple slab.
@@ -4700,28 +4728,37 @@ connectivity checks:
4. Position a second slab cursor at the start of the per-file tuple slab.
- 5. Iterate the two cursors in lockstep, comparing the parent_ino and
- dirent_pos fields of the records under each cursor.
+ 5. Iterate the two cursors in lockstep, comparing the ``parent_ino``,
+ ``name_hash``, and ``name_cookie`` fields of the records under each
+ cursor:
- a. Tuples in the per-AG list but not the per-file list are missing and
- need to be written to the inode.
+ a. If the per-AG cursor is at a lower point in the keyspace than the
+ per-file cursor, then the per-AG cursor points to a missing parent
+ pointer.
+ Add the parent pointer to the inode and advance the per-AG
+ cursor.
- b. Tuples in the per-file list but not the per-AG list are dangling
- and need to be removed from the inode.
+ b. If the per-file cursor is at a lower point in the keyspace than
+ the per-AG cursor, then the per-file cursor points to a dangling
+ parent pointer.
+ Remove the parent pointer from the inode and advance the per-file
+ cursor.
- c. For tuples in both lists, update the parent_gen and name components
- of the parent pointer if necessary.
+ c. Otherwise, both cursors point at the same parent pointer.
+ Update the parent_gen component if necessary.
+ Advance both cursors.
4. Move on to examining link counts, as we do today.
The proposed patchset is the
`offline parent pointers repair
-<https://git.kernel.org/pub/scm/linux/kernel/git/djwong/xfsprogs-dev.git/log/?h=pptrs-repair>`_
+<https://git.kernel.org/pub/scm/linux/kernel/git/djwong/xfsprogs-dev.git/log/?h=pptrs-fsck>`_
series.
-Rebuilding directories from parent pointers in offline repair is very
-challenging because it currently uses a single-pass scan of the filesystem
-during phase 3 to decide which files are corrupt enough to be zapped.
+Rebuilding directories from parent pointers in offline repair would be very
+challenging because xfs_repair currently uses two single-pass scans of the
+filesystem during phases 3 and 4 to decide which files are corrupt enough to be
+zapped.
This scan would have to be converted into a multi-pass scan:
1. The first pass of the scan zaps corrupt inodes, forks, and attributes
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 4/4] docs: describe xfs directory tree online fsck
2023-12-31 19:32 ` [PATCHSET v13.0 1/7] xfs: design documentation for online fsck, " Darrick J. Wong
` (2 preceding siblings ...)
2023-12-31 20:42 ` [PATCH 3/4] docs: update offline parent pointer repair strategy Darrick J. Wong
@ 2023-12-31 20:43 ` Darrick J. Wong
3 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:43 UTC (permalink / raw)
To: djwong; +Cc: linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
I've added a scrubber that checks the directory tree structure and fixes
them; describe this in the design documentation.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
.../filesystems/xfs-online-fsck-design.rst | 121 ++++++++++++++++++++
1 file changed, 121 insertions(+)
diff --git a/Documentation/filesystems/xfs-online-fsck-design.rst b/Documentation/filesystems/xfs-online-fsck-design.rst
index 04e445d79f2d4..29e123189d303 100644
--- a/Documentation/filesystems/xfs-online-fsck-design.rst
+++ b/Documentation/filesystems/xfs-online-fsck-design.rst
@@ -4780,6 +4780,127 @@ This scan would have to be converted into a multi-pass scan:
This code has not yet been constructed.
+.. _dirtree:
+
+Case Study: Directory Tree Structure
+^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
+
+As mentioned earlier, the filesystem directory tree is supposed to be a
+directed acylic graph structure.
+However, each node in this graph is a separate ``xfs_inode`` object with its
+own locks, which makes validating the tree qualities difficult.
+Fortunately, non-directories are allowed to have multiple parents and cannot
+have children, so only directories need to be scanned.
+Directories typically constitute 5-10% of the files in a filesystem, which
+reduces the amount of work dramatically.
+
+If the directory tree could be frozen, it would be easy to discover cycles and
+disconnected regions by running a depth (or breadth) first search downwards
+from the root directory and marking a bitmap for each directory found.
+At any point in the walk, trying to set an already set bit means there is a
+cycle.
+After the scan completes, XORing the marked inode bitmap with the inode
+allocation bitmap reveals disconnected inodes.
+However, one of online repair's design goals is to avoid locking the entire
+filesystem unless it's absolutely necessary.
+Directory tree updates can move subtrees across the scanner wavefront on a live
+filesystem, so the bitmap algorithm cannot be applied.
+
+Directory parent pointers enable an incremental approach to validation of the
+tree structure.
+Instead of using one thread to scan the entire filesystem, multiple threads can
+walk from individual subdirectories upwards towards the root.
+For this to work, all directory entries and parent pointers must be internally
+consistent, each directory entry must have a parent pointer, and the link
+counts of all directories must be correct.
+Each scanner thread must be able to take the IOLOCK of an alleged parent
+directory while holding the IOLOCK of the child directory to prevent either
+directory from being moved within the tree.
+This was made possible only recently with locking changes in Linux 6.5.
+
+The scanning process uses a dirent hook to detect changes to the directories
+mentioned in the scan data.
+The scan works as follows:
+
+1. For each subdirectory in the filesystem,
+
+ a. For each parent pointer of that subdirectory,
+
+ 1. Create a path object for that parent pointer, and mark the
+ subdirectory inode number in the path object's bitmap.
+
+ 2. Record the parent pointer name and inode number in a path structure.
+
+ 3. If the alleged parent is the subdirectory being scrubbed, the path is
+ a cycle.
+ Mark the path for deletion and repeat step 1a with the next
+ subdirectory parent pointer.
+
+ 4. Try to mark the alleged parent inode number in a bitmap in the path
+ object.
+ If the bit is already set, then there is a cycle in the directory
+ tree.
+ Mark the path as a cycle and repeat step 1a with the next subdirectory
+ parent pointer.
+
+ 5. Load the alleged parent.
+ If the alleged parent is not a linked directory, abort the scan
+ because the parent pointer information is inconsistent.
+
+ 6. For each parent pointer of this alleged ancestor directory,
+
+ a. Record the parent pointer name and inode number in the path object
+ if no parent has been set for that level.
+
+ b. If an ancestor has more than one parent, mark the path as corrupt.
+ Repeat step 1a with the next subdirectory parent pointer.
+
+ c. Repeat steps 1a3-1a6 for the ancestor identified in step 1a6a.
+ This repeats until the directory tree root is reached or no parents
+ are found.
+
+ 7. If the walk terminates at the root directory, mark the path as ok.
+
+ 8. If the walk terminates without reaching the root, mark the path as
+ disconnected.
+
+2. If the directory entry update hook triggers, check all paths already found
+ by the scan.
+ If the entry matches part of a path, mark that path and the scan stale.
+ When the scanner thread sees that the scan has been marked stale, it deletes
+ all scan data and starts over.
+
+Repairing the directory tree works as follows:
+
+1. Walk each path of the target subdirectory.
+
+ a. Corrupt paths and cycle paths are counted as suspect.
+
+ b. Paths already marked for deletion are counted as bad.
+
+ c. Paths that reached the root are counted as good.
+
+2. If the subdirectory is either the root directory or has zero link count,
+ delete all incoming directory entries in the immediate parents.
+ Repairs are complete.
+
+3. If the subdirectory has exactly one path, set the dotdot entry to the
+ parent and exit.
+
+4. If the subdirectory has at least one good path, delete all the other
+ incoming directory entries in the immediate parents.
+
+5. If the subdirectory has no good paths and more than one suspect path, delete
+ all the other incoming directory entries in the immediate parents.
+
+6. If the subdirectory has zero paths, attach it to the lost and found.
+
+The proposed patches are in the
+`directory tree repair
+<https://git.kernel.org/pub/scm/linux/kernel/git/djwong/xfs-linux.git/log/?h=scrub-directory-tree>`_
+series.
+
+
.. _orphanage:
The Orphanage
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 1/7] xfs: Increase XFS_DEFER_OPS_NR_INODES to 5
2023-12-31 19:33 ` [PATCHSET v13.0 2/7] xfs: retain ILOCK during directory updates Darrick J. Wong
@ 2023-12-31 20:43 ` Darrick J. Wong
2023-12-31 20:43 ` [PATCH 2/7] xfs: Increase XFS_QM_TRANS_MAXDQS " Darrick J. Wong
` (5 subsequent siblings)
6 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:43 UTC (permalink / raw)
To: djwong
Cc: Allison Henderson, Catherine Hoang, catherine.hoang,
allison.henderson, linux-xfs
From: Allison Henderson <allison.henderson@oracle.com>
Renames that generate parent pointer updates can join up to 5
inodes locked in sorted order. So we need to increase the
number of defer ops inodes and relock them in the same way.
Signed-off-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Catherine Hoang <catherine.hoang@oracle.com>
[djwong: have one sorting function]
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/libxfs/xfs_defer.c | 6 +++++-
fs/xfs/libxfs/xfs_defer.h | 8 +++++++-
fs/xfs/xfs_inode.c | 27 ++++++++++++++++++---------
fs/xfs/xfs_inode.h | 2 ++
4 files changed, 32 insertions(+), 11 deletions(-)
diff --git a/fs/xfs/libxfs/xfs_defer.c b/fs/xfs/libxfs/xfs_defer.c
index f77b2eaaa1b0d..8788f9f3f19ec 100644
--- a/fs/xfs/libxfs/xfs_defer.c
+++ b/fs/xfs/libxfs/xfs_defer.c
@@ -1096,7 +1096,11 @@ xfs_defer_ops_continue(
ASSERT(!(tp->t_flags & XFS_TRANS_DIRTY));
/* Lock the captured resources to the new transaction. */
- if (dfc->dfc_held.dr_inos == 2)
+ if (dfc->dfc_held.dr_inos > 2) {
+ xfs_sort_inodes(dfc->dfc_held.dr_ip, dfc->dfc_held.dr_inos);
+ xfs_lock_inodes(dfc->dfc_held.dr_ip, dfc->dfc_held.dr_inos,
+ XFS_ILOCK_EXCL);
+ } else if (dfc->dfc_held.dr_inos == 2)
xfs_lock_two_inodes(dfc->dfc_held.dr_ip[0], XFS_ILOCK_EXCL,
dfc->dfc_held.dr_ip[1], XFS_ILOCK_EXCL);
else if (dfc->dfc_held.dr_inos == 1)
diff --git a/fs/xfs/libxfs/xfs_defer.h b/fs/xfs/libxfs/xfs_defer.h
index e3cf81bafca3e..c9a1fe3fe363e 100644
--- a/fs/xfs/libxfs/xfs_defer.h
+++ b/fs/xfs/libxfs/xfs_defer.h
@@ -77,7 +77,13 @@ extern const struct xfs_defer_op_type xfs_swapext_defer_type;
/*
* Deferred operation item relogging limits.
*/
-#define XFS_DEFER_OPS_NR_INODES 2 /* join up to two inodes */
+
+/*
+ * Rename w/ parent pointers can require up to 5 inodes with deferred ops to
+ * be joined to the transaction: src_dp, target_dp, src_ip, target_ip, and wip.
+ * These inodes are locked in sorted order by their inode numbers
+ */
+#define XFS_DEFER_OPS_NR_INODES 5
#define XFS_DEFER_OPS_NR_BUFS 2 /* join up to two buffers */
/* Resources that must be held across a transaction roll. */
diff --git a/fs/xfs/xfs_inode.c b/fs/xfs/xfs_inode.c
index 71640afc3a8ee..6ff3d2cab5802 100644
--- a/fs/xfs/xfs_inode.c
+++ b/fs/xfs/xfs_inode.c
@@ -440,7 +440,7 @@ xfs_lock_inumorder(
* lock more than one at a time, lockdep will report false positives saying we
* have violated locking orders.
*/
-static void
+void
xfs_lock_inodes(
struct xfs_inode **ips,
int inodes,
@@ -2804,7 +2804,7 @@ xfs_sort_for_rename(
struct xfs_inode **i_tab,/* out: sorted array of inodes */
int *num_inodes) /* in/out: inodes in array */
{
- int i, j;
+ int i;
ASSERT(*num_inodes == __XFS_SORT_INODES);
memset(i_tab, 0, *num_inodes * sizeof(struct xfs_inode *));
@@ -2826,17 +2826,26 @@ xfs_sort_for_rename(
i_tab[i++] = wip;
*num_inodes = i;
+ xfs_sort_inodes(i_tab, *num_inodes);
+}
+
+void
+xfs_sort_inodes(
+ struct xfs_inode **i_tab,
+ unsigned int num_inodes)
+{
+ int i, j;
+
+ ASSERT(num_inodes <= __XFS_SORT_INODES);
+
/*
* Sort the elements via bubble sort. (Remember, there are at
* most 5 elements to sort, so this is adequate.)
*/
- for (i = 0; i < *num_inodes; i++) {
- for (j = 1; j < *num_inodes; j++) {
- if (i_tab[j]->i_ino < i_tab[j-1]->i_ino) {
- struct xfs_inode *temp = i_tab[j];
- i_tab[j] = i_tab[j-1];
- i_tab[j-1] = temp;
- }
+ for (i = 0; i < num_inodes; i++) {
+ for (j = 1; j < num_inodes; j++) {
+ if (i_tab[j]->i_ino < i_tab[j-1]->i_ino)
+ swap(i_tab[j], i_tab[j - 1]);
}
}
}
diff --git a/fs/xfs/xfs_inode.h b/fs/xfs/xfs_inode.h
index 8f0dccb0361d7..4826155ad9147 100644
--- a/fs/xfs/xfs_inode.h
+++ b/fs/xfs/xfs_inode.h
@@ -623,6 +623,8 @@ int xfs_ilock2_io_mmap(struct xfs_inode *ip1, struct xfs_inode *ip2);
void xfs_iunlock2_io_mmap(struct xfs_inode *ip1, struct xfs_inode *ip2);
void xfs_iunlock2_remapping(struct xfs_inode *ip1, struct xfs_inode *ip2);
void xfs_bumplink(struct xfs_trans *tp, struct xfs_inode *ip);
+void xfs_lock_inodes(struct xfs_inode **ips, int inodes, uint lock_mode);
+void xfs_sort_inodes(struct xfs_inode **i_tab, unsigned int num_inodes);
static inline bool
xfs_inode_unlinked_incomplete(
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 2/7] xfs: Increase XFS_QM_TRANS_MAXDQS to 5
2023-12-31 19:33 ` [PATCHSET v13.0 2/7] xfs: retain ILOCK during directory updates Darrick J. Wong
2023-12-31 20:43 ` [PATCH 1/7] xfs: Increase XFS_DEFER_OPS_NR_INODES to 5 Darrick J. Wong
@ 2023-12-31 20:43 ` Darrick J. Wong
2023-12-31 20:43 ` [PATCH 3/7] xfs: Hold inode locks in xfs_ialloc Darrick J. Wong
` (4 subsequent siblings)
6 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:43 UTC (permalink / raw)
To: djwong; +Cc: Allison Henderson, catherine.hoang, allison.henderson, linux-xfs
From: Allison Henderson <allison.henderson@oracle.com>
With parent pointers enabled, a rename operation can update up to 5
inodes: src_dp, target_dp, src_ip, target_ip and wip. This causes
their dquots to a be attached to the transaction chain, so we need
to increase XFS_QM_TRANS_MAXDQS. This patch also add a helper
function xfs_dqlockn to lock an arbitrary number of dquots.
Signed-off-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/xfs_dquot.c | 41 +++++++++++++++++++++++++++++++++++++++++
fs/xfs/xfs_dquot.h | 1 +
fs/xfs/xfs_qm.h | 2 +-
fs/xfs/xfs_trans_dquot.c | 15 ++++++++++-----
4 files changed, 53 insertions(+), 6 deletions(-)
diff --git a/fs/xfs/xfs_dquot.c b/fs/xfs/xfs_dquot.c
index 8703495c2fdc6..bc1893a4b6738 100644
--- a/fs/xfs/xfs_dquot.c
+++ b/fs/xfs/xfs_dquot.c
@@ -1363,6 +1363,47 @@ xfs_dqlock2(
}
}
+static int
+xfs_dqtrx_cmp(
+ const void *a,
+ const void *b)
+{
+ const struct xfs_dqtrx *qa = a;
+ const struct xfs_dqtrx *qb = b;
+
+ if (qa->qt_dquot->q_id > qb->qt_dquot->q_id)
+ return 1;
+ if (qa->qt_dquot->q_id < qb->qt_dquot->q_id)
+ return -1;
+ return 0;
+}
+
+void
+xfs_dqlockn(
+ struct xfs_dqtrx *q)
+{
+ unsigned int i;
+
+ BUILD_BUG_ON(XFS_QM_TRANS_MAXDQS > MAX_LOCKDEP_SUBCLASSES);
+
+ /* Sort in order of dquot id, do not allow duplicates */
+ for (i = 0; i < XFS_QM_TRANS_MAXDQS && q[i].qt_dquot != NULL; i++) {
+ unsigned int j;
+
+ for (j = 0; j < i; j++)
+ ASSERT(q[i].qt_dquot != q[j].qt_dquot);
+ }
+ if (i == 0)
+ return;
+
+ sort(q, i, sizeof(struct xfs_dqtrx), xfs_dqtrx_cmp, NULL);
+
+ mutex_lock(&q[0].qt_dquot->q_qlock);
+ for (i = 1; i < XFS_QM_TRANS_MAXDQS && q[i].qt_dquot != NULL; i++)
+ mutex_lock_nested(&q[i].qt_dquot->q_qlock,
+ XFS_QLOCK_NESTED + i - 1);
+}
+
int __init
xfs_qm_init(void)
{
diff --git a/fs/xfs/xfs_dquot.h b/fs/xfs/xfs_dquot.h
index 956272d9b302f..677bb2dc9ac91 100644
--- a/fs/xfs/xfs_dquot.h
+++ b/fs/xfs/xfs_dquot.h
@@ -223,6 +223,7 @@ int xfs_qm_dqget_uncached(struct xfs_mount *mp,
void xfs_qm_dqput(struct xfs_dquot *dqp);
void xfs_dqlock2(struct xfs_dquot *, struct xfs_dquot *);
+void xfs_dqlockn(struct xfs_dqtrx *q);
void xfs_dquot_set_prealloc_limits(struct xfs_dquot *);
diff --git a/fs/xfs/xfs_qm.h b/fs/xfs/xfs_qm.h
index f5993012bf98f..6e09dfcd13e25 100644
--- a/fs/xfs/xfs_qm.h
+++ b/fs/xfs/xfs_qm.h
@@ -136,7 +136,7 @@ enum {
XFS_QM_TRANS_PRJ,
XFS_QM_TRANS_DQTYPES
};
-#define XFS_QM_TRANS_MAXDQS 2
+#define XFS_QM_TRANS_MAXDQS 5
struct xfs_dquot_acct {
struct xfs_dqtrx dqs[XFS_QM_TRANS_DQTYPES][XFS_QM_TRANS_MAXDQS];
};
diff --git a/fs/xfs/xfs_trans_dquot.c b/fs/xfs/xfs_trans_dquot.c
index f5e9d76fb9a2f..833a65be05705 100644
--- a/fs/xfs/xfs_trans_dquot.c
+++ b/fs/xfs/xfs_trans_dquot.c
@@ -368,24 +368,29 @@ xfs_trans_mod_dquot(
/*
* Given an array of dqtrx structures, lock all the dquots associated and join
- * them to the transaction, provided they have been modified. We know that the
- * highest number of dquots of one type - usr, grp and prj - involved in a
- * transaction is 3 so we don't need to make this very generic.
+ * them to the transaction, provided they have been modified.
*/
STATIC void
xfs_trans_dqlockedjoin(
struct xfs_trans *tp,
struct xfs_dqtrx *q)
{
+ unsigned int i;
ASSERT(q[0].qt_dquot != NULL);
if (q[1].qt_dquot == NULL) {
xfs_dqlock(q[0].qt_dquot);
xfs_trans_dqjoin(tp, q[0].qt_dquot);
- } else {
- ASSERT(XFS_QM_TRANS_MAXDQS == 2);
+ } else if (q[2].qt_dquot == NULL) {
xfs_dqlock2(q[0].qt_dquot, q[1].qt_dquot);
xfs_trans_dqjoin(tp, q[0].qt_dquot);
xfs_trans_dqjoin(tp, q[1].qt_dquot);
+ } else {
+ xfs_dqlockn(q);
+ for (i = 0; i < XFS_QM_TRANS_MAXDQS; i++) {
+ if (q[i].qt_dquot == NULL)
+ break;
+ xfs_trans_dqjoin(tp, q[i].qt_dquot);
+ }
}
}
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 3/7] xfs: Hold inode locks in xfs_ialloc
2023-12-31 19:33 ` [PATCHSET v13.0 2/7] xfs: retain ILOCK during directory updates Darrick J. Wong
2023-12-31 20:43 ` [PATCH 1/7] xfs: Increase XFS_DEFER_OPS_NR_INODES to 5 Darrick J. Wong
2023-12-31 20:43 ` [PATCH 2/7] xfs: Increase XFS_QM_TRANS_MAXDQS " Darrick J. Wong
@ 2023-12-31 20:43 ` Darrick J. Wong
2023-12-31 20:44 ` [PATCH 4/7] xfs: Hold inode locks in xfs_trans_alloc_dir Darrick J. Wong
` (3 subsequent siblings)
6 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:43 UTC (permalink / raw)
To: djwong
Cc: Allison Henderson, Catherine Hoang, catherine.hoang,
allison.henderson, linux-xfs
From: Allison Henderson <allison.henderson@oracle.com>
Modify xfs_ialloc to hold locks after return. Caller will be
responsible for manual unlock. We will need this later to hold locks
across parent pointer operations
Signed-off-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Catherine Hoang <catherine.hoang@oracle.com>
[djwong: hold the parent ilocked across transaction rolls too]
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/xfs_inode.c | 12 +++++++++---
fs/xfs/xfs_qm.c | 4 +++-
fs/xfs/xfs_symlink.c | 6 ++++--
3 files changed, 16 insertions(+), 6 deletions(-)
diff --git a/fs/xfs/xfs_inode.c b/fs/xfs/xfs_inode.c
index 6ff3d2cab5802..910764bf8810f 100644
--- a/fs/xfs/xfs_inode.c
+++ b/fs/xfs/xfs_inode.c
@@ -769,6 +769,8 @@ xfs_inode_inherit_flags2(
/*
* Initialise a newly allocated inode and return the in-core inode to the
* caller locked exclusively.
+ *
+ * Caller is responsible for unlocking the inode manually upon return
*/
int
xfs_init_new_inode(
@@ -895,7 +897,7 @@ xfs_init_new_inode(
/*
* Log the new values stuffed into the inode.
*/
- xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL);
+ xfs_trans_ijoin(tp, ip, 0);
xfs_trans_log_inode(tp, ip, flags);
/* now that we have an i_mode we can setup the inode structure */
@@ -1114,8 +1116,7 @@ xfs_create(
* the transaction cancel unlocking dp so don't do it explicitly in the
* error path.
*/
- xfs_trans_ijoin(tp, dp, XFS_ILOCK_EXCL);
- unlock_dp_on_error = false;
+ xfs_trans_ijoin(tp, dp, 0);
error = xfs_dir_createname(tp, dp, name, ip->i_ino,
resblks - XFS_IALLOC_SPACE_RES(mp));
@@ -1164,6 +1165,8 @@ xfs_create(
xfs_qm_dqrele(pdqp);
*ipp = ip;
+ xfs_iunlock(ip, XFS_ILOCK_EXCL);
+ xfs_iunlock(dp, XFS_ILOCK_EXCL);
return 0;
out_trans_cancel:
@@ -1175,6 +1178,7 @@ xfs_create(
* transactions and deadlocks from xfs_inactive.
*/
if (ip) {
+ xfs_iunlock(ip, XFS_ILOCK_EXCL);
xfs_finish_inode_setup(ip);
xfs_irele(ip);
}
@@ -1260,6 +1264,7 @@ xfs_create_tmpfile(
xfs_qm_dqrele(pdqp);
*ipp = ip;
+ xfs_iunlock(ip, XFS_ILOCK_EXCL);
return 0;
out_trans_cancel:
@@ -1271,6 +1276,7 @@ xfs_create_tmpfile(
* transactions and deadlocks from xfs_inactive.
*/
if (ip) {
+ xfs_iunlock(ip, XFS_ILOCK_EXCL);
xfs_finish_inode_setup(ip);
xfs_irele(ip);
}
diff --git a/fs/xfs/xfs_qm.c b/fs/xfs/xfs_qm.c
index 4f357cb6de748..c25d917487f0e 100644
--- a/fs/xfs/xfs_qm.c
+++ b/fs/xfs/xfs_qm.c
@@ -835,8 +835,10 @@ xfs_qm_qino_alloc(
ASSERT(xfs_is_shutdown(mp));
xfs_alert(mp, "%s failed (error %d)!", __func__, error);
}
- if (need_alloc)
+ if (need_alloc) {
+ xfs_iunlock(*ipp, XFS_ILOCK_EXCL);
xfs_finish_inode_setup(*ipp);
+ }
return error;
}
diff --git a/fs/xfs/xfs_symlink.c b/fs/xfs/xfs_symlink.c
index 06df5522db7a5..307a2b7e5426e 100644
--- a/fs/xfs/xfs_symlink.c
+++ b/fs/xfs/xfs_symlink.c
@@ -173,8 +173,7 @@ xfs_symlink(
* the transaction cancel unlocking dp so don't do it explicitly in the
* error path.
*/
- xfs_trans_ijoin(tp, dp, XFS_ILOCK_EXCL);
- unlock_dp_on_error = false;
+ xfs_trans_ijoin(tp, dp, 0);
/*
* Also attach the dquot(s) to it, if applicable.
@@ -216,6 +215,8 @@ xfs_symlink(
xfs_qm_dqrele(pdqp);
*ipp = ip;
+ xfs_iunlock(ip, XFS_ILOCK_EXCL);
+ xfs_iunlock(dp, XFS_ILOCK_EXCL);
return 0;
out_trans_cancel:
@@ -227,6 +228,7 @@ xfs_symlink(
* transactions and deadlocks from xfs_inactive.
*/
if (ip) {
+ xfs_iunlock(ip, XFS_ILOCK_EXCL);
xfs_finish_inode_setup(ip);
xfs_irele(ip);
}
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 4/7] xfs: Hold inode locks in xfs_trans_alloc_dir
2023-12-31 19:33 ` [PATCHSET v13.0 2/7] xfs: retain ILOCK during directory updates Darrick J. Wong
` (2 preceding siblings ...)
2023-12-31 20:43 ` [PATCH 3/7] xfs: Hold inode locks in xfs_ialloc Darrick J. Wong
@ 2023-12-31 20:44 ` Darrick J. Wong
2023-12-31 20:44 ` [PATCH 5/7] xfs: Hold inode locks in xfs_rename Darrick J. Wong
` (2 subsequent siblings)
6 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:44 UTC (permalink / raw)
To: djwong
Cc: Allison Henderson, Catherine Hoang, catherine.hoang,
allison.henderson, linux-xfs
From: Allison Henderson <allison.henderson@oracle.com>
Modify xfs_trans_alloc_dir to hold locks after return. Caller will be
responsible for manual unlock. We will need this later to hold locks
across parent pointer operations
Signed-off-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Catherine Hoang <catherine.hoang@oracle.com>
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/xfs_inode.c | 14 ++++++++++++--
fs/xfs/xfs_trans.c | 9 +++++++--
2 files changed, 19 insertions(+), 4 deletions(-)
diff --git a/fs/xfs/xfs_inode.c b/fs/xfs/xfs_inode.c
index 910764bf8810f..606d0aec9b450 100644
--- a/fs/xfs/xfs_inode.c
+++ b/fs/xfs/xfs_inode.c
@@ -1370,10 +1370,15 @@ xfs_link(
if (xfs_has_wsync(mp) || xfs_has_dirsync(mp))
xfs_trans_set_sync(tp);
- return xfs_trans_commit(tp);
+ error = xfs_trans_commit(tp);
+ xfs_iunlock(tdp, XFS_ILOCK_EXCL);
+ xfs_iunlock(sip, XFS_ILOCK_EXCL);
+ return error;
error_return:
xfs_trans_cancel(tp);
+ xfs_iunlock(tdp, XFS_ILOCK_EXCL);
+ xfs_iunlock(sip, XFS_ILOCK_EXCL);
std_return:
if (error == -ENOSPC && nospace_error)
error = nospace_error;
@@ -2783,15 +2788,20 @@ xfs_remove(
error = xfs_trans_commit(tp);
if (error)
- goto std_return;
+ goto out_unlock;
if (is_dir && xfs_inode_is_filestream(ip))
xfs_filestream_deassociate(ip);
+ xfs_iunlock(ip, XFS_ILOCK_EXCL);
+ xfs_iunlock(dp, XFS_ILOCK_EXCL);
return 0;
out_trans_cancel:
xfs_trans_cancel(tp);
+ out_unlock:
+ xfs_iunlock(ip, XFS_ILOCK_EXCL);
+ xfs_iunlock(dp, XFS_ILOCK_EXCL);
std_return:
return error;
}
diff --git a/fs/xfs/xfs_trans.c b/fs/xfs/xfs_trans.c
index 12d45e93f07d5..81337fdb89ab8 100644
--- a/fs/xfs/xfs_trans.c
+++ b/fs/xfs/xfs_trans.c
@@ -1430,6 +1430,8 @@ xfs_trans_alloc_ichange(
* The caller must ensure that the on-disk dquots attached to this inode have
* already been allocated and initialized. The ILOCKs will be dropped when the
* transaction is committed or cancelled.
+ *
+ * Caller is responsible for unlocking the inodes manually upon return
*/
int
xfs_trans_alloc_dir(
@@ -1460,8 +1462,8 @@ xfs_trans_alloc_dir(
xfs_lock_two_inodes(dp, XFS_ILOCK_EXCL, ip, XFS_ILOCK_EXCL);
- xfs_trans_ijoin(tp, dp, XFS_ILOCK_EXCL);
- xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL);
+ xfs_trans_ijoin(tp, dp, 0);
+ xfs_trans_ijoin(tp, ip, 0);
error = xfs_qm_dqattach_locked(dp, false);
if (error) {
@@ -1484,6 +1486,9 @@ xfs_trans_alloc_dir(
if (error == -EDQUOT || error == -ENOSPC) {
if (!retried) {
xfs_trans_cancel(tp);
+ xfs_iunlock(dp, XFS_ILOCK_EXCL);
+ if (dp != ip)
+ xfs_iunlock(ip, XFS_ILOCK_EXCL);
xfs_blockgc_free_quota(dp, 0);
retried = true;
goto retry;
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 5/7] xfs: Hold inode locks in xfs_rename
2023-12-31 19:33 ` [PATCHSET v13.0 2/7] xfs: retain ILOCK during directory updates Darrick J. Wong
` (3 preceding siblings ...)
2023-12-31 20:44 ` [PATCH 4/7] xfs: Hold inode locks in xfs_trans_alloc_dir Darrick J. Wong
@ 2023-12-31 20:44 ` Darrick J. Wong
2023-12-31 20:44 ` [PATCH 6/7] xfs: don't pick up IOLOCK during rmapbt repair scan Darrick J. Wong
2023-12-31 20:44 ` [PATCH 7/7] xfs: unlock new repair tempfiles after creation Darrick J. Wong
6 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:44 UTC (permalink / raw)
To: djwong
Cc: Allison Henderson, Catherine Hoang, catherine.hoang,
allison.henderson, linux-xfs
From: Allison Henderson <allison.henderson@oracle.com>
Modify xfs_rename to hold all inode locks across a rename operation
We will need this later when we add parent pointers
Signed-off-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Catherine Hoang <catherine.hoang@oracle.com>
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/xfs_inode.c | 45 +++++++++++++++++++++++++++++++++------------
1 file changed, 33 insertions(+), 12 deletions(-)
diff --git a/fs/xfs/xfs_inode.c b/fs/xfs/xfs_inode.c
index 606d0aec9b450..88e0e93ded2e4 100644
--- a/fs/xfs/xfs_inode.c
+++ b/fs/xfs/xfs_inode.c
@@ -2806,6 +2806,21 @@ xfs_remove(
return error;
}
+static inline void
+xfs_iunlock_rename(
+ struct xfs_inode **i_tab,
+ int num_inodes)
+{
+ int i;
+
+ for (i = num_inodes - 1; i >= 0; i--) {
+ /* Skip duplicate inodes if src and target dps are the same */
+ if (!i_tab[i] || (i > 0 && i_tab[i] == i_tab[i - 1]))
+ continue;
+ xfs_iunlock(i_tab[i], XFS_ILOCK_EXCL);
+ }
+}
+
/*
* Enter all inodes for a rename transaction into a sorted array.
*/
@@ -3115,8 +3130,10 @@ xfs_rename(
* Attach the dquots to the inodes
*/
error = xfs_qm_vop_rename_dqattach(inodes);
- if (error)
- goto out_trans_cancel;
+ if (error) {
+ xfs_trans_cancel(tp);
+ goto out_release_wip;
+ }
/*
* Lock all the participating inodes. Depending upon whether
@@ -3127,18 +3144,16 @@ xfs_rename(
xfs_lock_inodes(inodes, num_inodes, XFS_ILOCK_EXCL);
/*
- * Join all the inodes to the transaction. From this point on,
- * we can rely on either trans_commit or trans_cancel to unlock
- * them.
+ * Join all the inodes to the transaction.
*/
- xfs_trans_ijoin(tp, src_dp, XFS_ILOCK_EXCL);
+ xfs_trans_ijoin(tp, src_dp, 0);
if (new_parent)
- xfs_trans_ijoin(tp, target_dp, XFS_ILOCK_EXCL);
- xfs_trans_ijoin(tp, src_ip, XFS_ILOCK_EXCL);
+ xfs_trans_ijoin(tp, target_dp, 0);
+ xfs_trans_ijoin(tp, src_ip, 0);
if (target_ip)
- xfs_trans_ijoin(tp, target_ip, XFS_ILOCK_EXCL);
+ xfs_trans_ijoin(tp, target_ip, 0);
if (wip)
- xfs_trans_ijoin(tp, wip, XFS_ILOCK_EXCL);
+ xfs_trans_ijoin(tp, wip, 0);
/*
* If we are using project inheritance, we only allow renames
@@ -3152,10 +3167,13 @@ xfs_rename(
}
/* RENAME_EXCHANGE is unique from here on. */
- if (flags & RENAME_EXCHANGE)
- return xfs_cross_rename(tp, src_dp, src_name, src_ip,
+ if (flags & RENAME_EXCHANGE) {
+ error = xfs_cross_rename(tp, src_dp, src_name, src_ip,
target_dp, target_name, target_ip,
spaceres);
+ xfs_iunlock_rename(inodes, num_inodes);
+ return error;
+ }
/*
* Try to reserve quota to handle an expansion of the target directory.
@@ -3169,6 +3187,7 @@ xfs_rename(
if (error == -EDQUOT || error == -ENOSPC) {
if (!retried) {
xfs_trans_cancel(tp);
+ xfs_iunlock_rename(inodes, num_inodes);
xfs_blockgc_free_quota(target_dp, 0);
retried = true;
goto retry;
@@ -3395,12 +3414,14 @@ xfs_rename(
xfs_dir_update_hook(src_dp, wip, 1, src_name);
error = xfs_finish_rename(tp);
+ xfs_iunlock_rename(inodes, num_inodes);
if (wip)
xfs_irele(wip);
return error;
out_trans_cancel:
xfs_trans_cancel(tp);
+ xfs_iunlock_rename(inodes, num_inodes);
out_release_wip:
if (wip)
xfs_irele(wip);
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 6/7] xfs: don't pick up IOLOCK during rmapbt repair scan
2023-12-31 19:33 ` [PATCHSET v13.0 2/7] xfs: retain ILOCK during directory updates Darrick J. Wong
` (4 preceding siblings ...)
2023-12-31 20:44 ` [PATCH 5/7] xfs: Hold inode locks in xfs_rename Darrick J. Wong
@ 2023-12-31 20:44 ` Darrick J. Wong
2023-12-31 20:44 ` [PATCH 7/7] xfs: unlock new repair tempfiles after creation Darrick J. Wong
6 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:44 UTC (permalink / raw)
To: djwong; +Cc: catherine.hoang, allison.henderson, linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
Now that we've fixed the directory operations to hold the ILOCK until
they're finished with rmapbt updates for directory shape changes, we no
longer need to take this lock when scanning directories for rmapbt
records.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/scrub/rmap_repair.c | 16 +---------------
1 file changed, 1 insertion(+), 15 deletions(-)
diff --git a/fs/xfs/scrub/rmap_repair.c b/fs/xfs/scrub/rmap_repair.c
index 9ece83704518d..c6bb90fa43cca 100644
--- a/fs/xfs/scrub/rmap_repair.c
+++ b/fs/xfs/scrub/rmap_repair.c
@@ -584,23 +584,9 @@ xrep_rmap_scan_inode(
struct xrep_rmap *rr,
struct xfs_inode *ip)
{
- unsigned int lock_mode = 0;
+ unsigned int lock_mode = xrep_rmap_scan_ilock(ip);
int error;
- /*
- * Directory updates (create/link/unlink/rename) drop the directory's
- * ILOCK before finishing any rmapbt updates associated with directory
- * shape changes. For this scan to coordinate correctly with the live
- * update hook, we must take the only lock (i_rwsem) that is held all
- * the way to dir op completion. This will get fixed by the parent
- * pointer patchset.
- */
- if (S_ISDIR(VFS_I(ip)->i_mode)) {
- lock_mode = XFS_IOLOCK_SHARED;
- xfs_ilock(ip, lock_mode);
- }
- lock_mode |= xrep_rmap_scan_ilock(ip);
-
/* Check the data fork. */
error = xrep_rmap_scan_ifork(rr, ip, XFS_DATA_FORK);
if (error)
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 7/7] xfs: unlock new repair tempfiles after creation
2023-12-31 19:33 ` [PATCHSET v13.0 2/7] xfs: retain ILOCK during directory updates Darrick J. Wong
` (5 preceding siblings ...)
2023-12-31 20:44 ` [PATCH 6/7] xfs: don't pick up IOLOCK during rmapbt repair scan Darrick J. Wong
@ 2023-12-31 20:44 ` Darrick J. Wong
6 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:44 UTC (permalink / raw)
To: djwong; +Cc: catherine.hoang, allison.henderson, linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
After creation, drop the ILOCK on temporary files that have been created
to stage a repair.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/scrub/tempfile.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/fs/xfs/scrub/tempfile.c b/fs/xfs/scrub/tempfile.c
index 93d8a6b68f442..9e0f1d311118b 100644
--- a/fs/xfs/scrub/tempfile.c
+++ b/fs/xfs/scrub/tempfile.c
@@ -145,6 +145,7 @@ xrep_tempfile_create(
xfs_qm_dqrele(pdqp);
/* Finish setting up the incore / vfs context. */
+ xfs_iunlock(sc->tempip, XFS_ILOCK_EXCL);
xfs_setup_iops(sc->tempip);
xfs_finish_inode_setup(sc->tempip);
@@ -160,6 +161,7 @@ xrep_tempfile_create(
* transactions and deadlocks from xfs_inactive.
*/
if (sc->tempip) {
+ xfs_iunlock(sc->tempip, XFS_ILOCK_EXCL);
xfs_finish_inode_setup(sc->tempip);
xchk_irele(sc, sc->tempip);
}
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 01/14] xfs: require XFS_SB_FEAT_INCOMPAT_LOG_XATTRS for attr log intent item recovery
2023-12-31 19:33 ` [PATCHSET v13.0 3/7] xfs: name-value xattr lookups Darrick J. Wong
@ 2023-12-31 20:45 ` Darrick J. Wong
2023-12-31 20:45 ` [PATCH 02/14] xfs: allow newer INCOMPAT/RO_COMPAT feature bits to protect ATTRI log items Darrick J. Wong
` (12 subsequent siblings)
13 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:45 UTC (permalink / raw)
To: djwong; +Cc: catherine.hoang, allison.henderson, linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
The XFS_SB_FEAT_INCOMPAT_LOG_XATTRS feature bit protects a filesystem
from old kernels that do not know how to recover extended attribute log
intent items. Make this check mandatory instead of a debugging assert.
Fixes: fd920008784ea ("xfs: Set up infrastructure for log attribute replay")
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/xfs_attr_item.c | 5 +++--
1 file changed, 3 insertions(+), 2 deletions(-)
diff --git a/fs/xfs/xfs_attr_item.c b/fs/xfs/xfs_attr_item.c
index d7ebb54a03870..c023962141556 100644
--- a/fs/xfs/xfs_attr_item.c
+++ b/fs/xfs/xfs_attr_item.c
@@ -469,6 +469,9 @@ xfs_attri_validate(
unsigned int op = attrp->alfi_op_flags &
XFS_ATTRI_OP_FLAGS_TYPE_MASK;
+ if (!xfs_sb_version_haslogxattrs(&mp->m_sb))
+ return false;
+
if (attrp->__pad != 0)
return false;
@@ -542,8 +545,6 @@ xfs_attri_recover_work(
XFS_DA_OP_LOGGED;
args->owner = args->dp->i_ino;
- ASSERT(xfs_sb_version_haslogxattrs(&mp->m_sb));
-
switch (attr->xattri_op_flags) {
case XFS_ATTRI_OP_FLAGS_SET:
case XFS_ATTRI_OP_FLAGS_REPLACE:
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 02/14] xfs: allow newer INCOMPAT/RO_COMPAT feature bits to protect ATTRI log items
2023-12-31 19:33 ` [PATCHSET v13.0 3/7] xfs: name-value xattr lookups Darrick J. Wong
2023-12-31 20:45 ` [PATCH 01/14] xfs: require XFS_SB_FEAT_INCOMPAT_LOG_XATTRS for attr log intent item recovery Darrick J. Wong
@ 2023-12-31 20:45 ` Darrick J. Wong
2023-12-31 20:45 ` [PATCH 03/14] xfs: check opcode and iovec count match in xlog_recover_attri_commit_pass2 Darrick J. Wong
` (11 subsequent siblings)
13 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:45 UTC (permalink / raw)
To: djwong; +Cc: catherine.hoang, allison.henderson, linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
Log recovery (which can include replaying ATTRI intent items) occurs on
rw and ro mounts. Dirty logs containing these log items must be
protected from being replayed by older kernels. The log incompat
feature XFS_SB_FEAT_INCOMPAT_LOG_XATTRS provides this protection.
However, adding this flag to the filesystem introduces performance
problems of its own -- each time we do, we must force the log and write
the primary superblock before writing any ATTRI log items. This was ok
when the only users were developers using the debug knob, but this sucks
for regular users. We'd like to avoid that.
If a filesystem has ro-compat or incompat feature bits set that weren't
defined at the time that ATTRI log items were defined, then any kernel
that doesn't know about ATTRI items will reject that filesystem. This
provides the same protection as the log-incompat feature, but at a much
lower cost because most ro-compat and incompat features are set on a
permanent basis.
Avoid the performance hit by detecting these feature bits and skipping
the xfs_add_incompat_log_feature calls.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/libxfs/xfs_attr.c | 6 +++++-
fs/xfs/libxfs/xfs_attr.h | 23 +++++++++++++++++++++++
fs/xfs/libxfs/xfs_format.h | 6 +++++-
fs/xfs/xfs_attr_item.c | 3 ++-
fs/xfs/xfs_xattr.c | 10 +++++++++-
5 files changed, 44 insertions(+), 4 deletions(-)
diff --git a/fs/xfs/libxfs/xfs_attr.c b/fs/xfs/libxfs/xfs_attr.c
index b002ddd5f05a2..2e5550ab1454f 100644
--- a/fs/xfs/libxfs/xfs_attr.c
+++ b/fs/xfs/libxfs/xfs_attr.c
@@ -885,9 +885,13 @@ xfs_attr_defer_add(
struct xfs_da_args *args,
unsigned int op_flags)
{
-
struct xfs_attr_intent *new;
+ /* ATTRI log items must be protected from older kernels */
+ if (args->op_flags & XFS_DA_OP_LOGGED)
+ ASSERT(xfs_attri_can_use_without_log_assistance(args->dp->i_mount) ||
+ xfs_sb_version_haslogxattrs(&args->dp->i_mount->m_sb));
+
new = kmem_cache_zalloc(xfs_attr_intent_cache, GFP_NOFS | __GFP_NOFAIL);
new->xattri_op_flags = op_flags;
new->xattri_da_args = args;
diff --git a/fs/xfs/libxfs/xfs_attr.h b/fs/xfs/libxfs/xfs_attr.h
index e4f55008552b4..273e8dff76c07 100644
--- a/fs/xfs/libxfs/xfs_attr.h
+++ b/fs/xfs/libxfs/xfs_attr.h
@@ -620,4 +620,27 @@ void xfs_attr_intent_destroy_cache(void);
int xfs_attr_sf_totsize(struct xfs_inode *dp);
+/*
+ * Decide if this filesystem has a new enough permanent feature set to protect
+ * attri log items from being replayed on a kernel that does not have
+ * XFS_SB_FEAT_INCOMPAT_LOG_XATTRS set.
+ */
+static inline bool
+xfs_attri_can_use_without_log_assistance(
+ struct xfs_mount *mp)
+{
+ if (!xfs_sb_is_v5(&mp->m_sb))
+ return false;
+
+ if (xfs_sb_has_incompat_feature(&mp->m_sb,
+ ~(XFS_SB_FEAT_INCOMPAT_FTYPE |
+ XFS_SB_FEAT_INCOMPAT_SPINODES |
+ XFS_SB_FEAT_INCOMPAT_META_UUID |
+ XFS_SB_FEAT_INCOMPAT_BIGTIME |
+ XFS_SB_FEAT_INCOMPAT_NREXT64)))
+ return true;
+
+ return false;
+}
+
#endif /* __XFS_ATTR_H__ */
diff --git a/fs/xfs/libxfs/xfs_format.h b/fs/xfs/libxfs/xfs_format.h
index ec25010b57797..8b952909ce1e2 100644
--- a/fs/xfs/libxfs/xfs_format.h
+++ b/fs/xfs/libxfs/xfs_format.h
@@ -390,7 +390,11 @@ xfs_sb_has_incompat_feature(
return (sbp->sb_features_incompat & feature) != 0;
}
-#define XFS_SB_FEAT_INCOMPAT_LOG_XATTRS (1 << 0) /* Delayed Attributes */
+/*
+ * Log contains ATTRI log intent items which are not otherwise protected by
+ * an INCOMPAT/RO_COMPAT feature flag.
+ */
+#define XFS_SB_FEAT_INCOMPAT_LOG_XATTRS (1 << 0)
/*
* Log contains SXI log intent items which are not otherwise protected by
diff --git a/fs/xfs/xfs_attr_item.c b/fs/xfs/xfs_attr_item.c
index c023962141556..c95cef827179c 100644
--- a/fs/xfs/xfs_attr_item.c
+++ b/fs/xfs/xfs_attr_item.c
@@ -469,7 +469,8 @@ xfs_attri_validate(
unsigned int op = attrp->alfi_op_flags &
XFS_ATTRI_OP_FLAGS_TYPE_MASK;
- if (!xfs_sb_version_haslogxattrs(&mp->m_sb))
+ if (!xfs_sb_version_haslogxattrs(&mp->m_sb) &&
+ !xfs_attri_can_use_without_log_assistance(mp))
return false;
if (attrp->__pad != 0)
diff --git a/fs/xfs/xfs_xattr.c b/fs/xfs/xfs_xattr.c
index 1920ca49b08d6..5246539ad2174 100644
--- a/fs/xfs/xfs_xattr.c
+++ b/fs/xfs/xfs_xattr.c
@@ -33,6 +33,13 @@ xfs_attr_grab_log_assist(
{
int error = 0;
+ /*
+ * As a performance optimization, skip the log force and super write
+ * if the filesystem featureset already protects the attri log items.
+ */
+ if (xfs_attri_can_use_without_log_assistance(mp))
+ return 0;
+
/*
* Protect ourselves from an idle log clearing the logged xattrs log
* incompat feature bit.
@@ -76,7 +83,8 @@ static inline void
xfs_attr_rele_log_assist(
struct xfs_mount *mp)
{
- xlog_drop_incompat_feat(mp->m_log, XLOG_INCOMPAT_FEAT_XATTRS);
+ if (!xfs_attri_can_use_without_log_assistance(mp))
+ xlog_drop_incompat_feat(mp->m_log, XLOG_INCOMPAT_FEAT_XATTRS);
}
static inline bool
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 03/14] xfs: check opcode and iovec count match in xlog_recover_attri_commit_pass2
2023-12-31 19:33 ` [PATCHSET v13.0 3/7] xfs: name-value xattr lookups Darrick J. Wong
2023-12-31 20:45 ` [PATCH 01/14] xfs: require XFS_SB_FEAT_INCOMPAT_LOG_XATTRS for attr log intent item recovery Darrick J. Wong
2023-12-31 20:45 ` [PATCH 02/14] xfs: allow newer INCOMPAT/RO_COMPAT feature bits to protect ATTRI log items Darrick J. Wong
@ 2023-12-31 20:45 ` Darrick J. Wong
2023-12-31 20:45 ` [PATCH 04/14] xfs: make xfs_attr_set require XFS_DA_OP_REMOVE Darrick J. Wong
` (10 subsequent siblings)
13 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:45 UTC (permalink / raw)
To: djwong; +Cc: catherine.hoang, allison.henderson, linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
Check that the number of recovered log iovecs is what is expected for
the xattri opcode is expecting.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/xfs_attr_item.c | 27 +++++++++++++++++++++++++++
1 file changed, 27 insertions(+)
diff --git a/fs/xfs/xfs_attr_item.c b/fs/xfs/xfs_attr_item.c
index c95cef827179c..b6b558f4501ad 100644
--- a/fs/xfs/xfs_attr_item.c
+++ b/fs/xfs/xfs_attr_item.c
@@ -708,6 +708,7 @@ xlog_recover_attri_commit_pass2(
const void *attr_value = NULL;
const void *attr_name;
size_t len;
+ unsigned int op;
attri_formatp = item->ri_buf[0].i_addr;
attr_name = item->ri_buf[1].i_addr;
@@ -726,6 +727,32 @@ xlog_recover_attri_commit_pass2(
return -EFSCORRUPTED;
}
+ /* Check the number of log iovecs makes sense for the op code. */
+ op = attri_formatp->alfi_op_flags & XFS_ATTRI_OP_FLAGS_TYPE_MASK;
+ switch (op) {
+ case XFS_ATTRI_OP_FLAGS_SET:
+ case XFS_ATTRI_OP_FLAGS_REPLACE:
+ /* Log item, attr name, attr value */
+ if (item->ri_total != 3) {
+ XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp,
+ attri_formatp, len);
+ return -EFSCORRUPTED;
+ }
+ break;
+ case XFS_ATTRI_OP_FLAGS_REMOVE:
+ /* Log item, attr name */
+ if (item->ri_total != 2) {
+ XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp,
+ attri_formatp, len);
+ return -EFSCORRUPTED;
+ }
+ break;
+ default:
+ XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp,
+ attri_formatp, len);
+ return -EFSCORRUPTED;
+ }
+
/* Validate the attr name */
if (item->ri_buf[1].i_len !=
xlog_calc_iovec_len(attri_formatp->alfi_name_len)) {
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 04/14] xfs: make xfs_attr_set require XFS_DA_OP_REMOVE
2023-12-31 19:33 ` [PATCHSET v13.0 3/7] xfs: name-value xattr lookups Darrick J. Wong
` (2 preceding siblings ...)
2023-12-31 20:45 ` [PATCH 03/14] xfs: check opcode and iovec count match in xlog_recover_attri_commit_pass2 Darrick J. Wong
@ 2023-12-31 20:45 ` Darrick J. Wong
2023-12-31 20:46 ` [PATCH 05/14] xfs: allow xattr matching on name and value for local/sf attrs Darrick J. Wong
` (9 subsequent siblings)
13 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:45 UTC (permalink / raw)
To: djwong; +Cc: catherine.hoang, allison.henderson, linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
Quite a few patches from now, we're going to change the parent pointer
xattr format to encode as much of the dirent name in the xattr name as
fits, and spill the rest of it to the xattr value. To make this work
correctly, we'll be adding the ability to look up xattrs based on name
/and/ value.
Internally, the xattr data structure supports attributes with a zero
length value, which is how we're going to store parent pointers for
short dirent names. The parent pointer repair code uses xfs_attr_set to
add missing and remove dangling parent pointers, so that interface must
be capable of setting an xattr with args->value == NULL.
The userspace API doesn't support this, so xfs_attr_set currently treats
a NULL args->value as a request to remove an attr. However, that's a
quirk of the existing callers and the interface. Make the callers of
xfs_attr_set to declare explicitly that they want to remove an xattr.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/libxfs/xfs_attr.c | 10 +++++-----
fs/xfs/xfs_xattr.c | 5 +++++
2 files changed, 10 insertions(+), 5 deletions(-)
diff --git a/fs/xfs/libxfs/xfs_attr.c b/fs/xfs/libxfs/xfs_attr.c
index 2e5550ab1454f..2de3f6ad36601 100644
--- a/fs/xfs/libxfs/xfs_attr.c
+++ b/fs/xfs/libxfs/xfs_attr.c
@@ -926,6 +926,7 @@ xfs_attr_set(
struct xfs_mount *mp = dp->i_mount;
struct xfs_trans_res tres;
bool rsvd = (args->attr_filter & XFS_ATTR_ROOT);
+ bool is_remove = args->op_flags & XFS_DA_OP_REMOVE;
int error, local;
int rmt_blks = 0;
unsigned int total;
@@ -950,7 +951,7 @@ xfs_attr_set(
args->op_flags = XFS_DA_OP_OKNOENT |
(args->op_flags & XFS_DA_OP_LOGGED);
- if (args->value) {
+ if (!is_remove) {
XFS_STATS_INC(mp, xs_attr_set);
args->total = xfs_attr_calc_size(args, &local);
@@ -984,7 +985,7 @@ xfs_attr_set(
if (error)
return error;
- if (args->value || xfs_inode_hasattr(dp)) {
+ if (!is_remove || xfs_inode_hasattr(dp)) {
error = xfs_iext_count_may_overflow(dp, XFS_ATTR_FORK,
XFS_IEXT_ATTR_MANIP_CNT(rmt_blks));
if (error == -EFBIG)
@@ -997,8 +998,7 @@ xfs_attr_set(
error = xfs_attr_lookup(args);
switch (error) {
case -EEXIST:
- if (!args->value) {
- /* if no value, we are performing a remove operation */
+ if (is_remove) {
xfs_attr_defer_add(args, XFS_ATTRI_OP_FLAGS_REMOVE);
break;
}
@@ -1010,7 +1010,7 @@ xfs_attr_set(
break;
case -ENOATTR:
/* Can't remove what isn't there. */
- if (!args->value)
+ if (is_remove)
goto out_trans_cancel;
/* Pure replace fails if no existing attr to replace. */
diff --git a/fs/xfs/xfs_xattr.c b/fs/xfs/xfs_xattr.c
index 5246539ad2174..2339e3fcfb384 100644
--- a/fs/xfs/xfs_xattr.c
+++ b/fs/xfs/xfs_xattr.c
@@ -122,6 +122,11 @@ xfs_attr_change(
use_logging = true;
}
+ if (args->value)
+ args->op_flags &= ~XFS_DA_OP_REMOVE;
+ else
+ args->op_flags |= XFS_DA_OP_REMOVE;
+
error = xfs_attr_set(args);
if (use_logging)
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 05/14] xfs: allow xattr matching on name and value for local/sf attrs
2023-12-31 19:33 ` [PATCHSET v13.0 3/7] xfs: name-value xattr lookups Darrick J. Wong
` (3 preceding siblings ...)
2023-12-31 20:45 ` [PATCH 04/14] xfs: make xfs_attr_set require XFS_DA_OP_REMOVE Darrick J. Wong
@ 2023-12-31 20:46 ` Darrick J. Wong
2023-12-31 20:46 ` [PATCH 06/14] xfs: preserve NVLOOKUP in xfs_attr_set Darrick J. Wong
` (8 subsequent siblings)
13 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:46 UTC (permalink / raw)
To: djwong; +Cc: catherine.hoang, allison.henderson, linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
Add a new NVLOOKUP flag to signal that the caller wants to look up an
extended attribute by name and value. This only works with shortform
and local attributes. Only parent pointers need this functionality
and parent pointers cannot be remote xattrs, so this limitation is ok
for now.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/libxfs/xfs_attr_leaf.c | 45 ++++++++++++++++++++++++++++++++++-------
fs/xfs/libxfs/xfs_da_btree.h | 4 +++-
2 files changed, 40 insertions(+), 9 deletions(-)
diff --git a/fs/xfs/libxfs/xfs_attr_leaf.c b/fs/xfs/libxfs/xfs_attr_leaf.c
index 3face870b4dac..f7a4839e93e3f 100644
--- a/fs/xfs/libxfs/xfs_attr_leaf.c
+++ b/fs/xfs/libxfs/xfs_attr_leaf.c
@@ -508,10 +508,12 @@ xfs_attr3_leaf_read(
*/
static bool
xfs_attr_match(
- struct xfs_da_args *args,
- uint8_t namelen,
- unsigned char *name,
- int flags)
+ const struct xfs_da_args *args,
+ uint8_t namelen,
+ const unsigned char *name,
+ unsigned int valuelen,
+ const void *value,
+ int flags)
{
if (args->namelen != namelen)
@@ -519,6 +521,23 @@ xfs_attr_match(
if (memcmp(args->name, name, namelen) != 0)
return false;
+ if (args->op_flags & XFS_DA_OP_NVLOOKUP) {
+ if (args->valuelen != valuelen)
+ return false;
+ if (args->valuelen && !value) {
+ /* not implemented for remote values */
+ ASSERT(0);
+ return false;
+ }
+ if (valuelen && !args->value) {
+ /* caller gave us valuelen > 0 but no value?? */
+ ASSERT(0);
+ return false;
+ }
+ if (valuelen > 0 && memcmp(args->value, value, valuelen) != 0)
+ return false;
+ }
+
/* Recovery ignores the INCOMPLETE flag. */
if ((args->op_flags & XFS_DA_OP_RECOVERY) &&
args->attr_filter == (flags & XFS_ATTR_NSP_ONDISK_MASK))
@@ -537,6 +556,10 @@ xfs_attr_copy_value(
unsigned char *value,
int valuelen)
{
+ /* vlookups already supplied the attr value; don't copy anything */
+ if (args->op_flags & XFS_DA_OP_NVLOOKUP)
+ return 0;
+
/*
* No copy if all we have to do is get the length
*/
@@ -761,6 +784,7 @@ xfs_attr_sf_findname(
base += size, i++) {
size = xfs_attr_sf_entsize(sfe);
if (!xfs_attr_match(args, sfe->namelen, sfe->nameval,
+ sfe->valuelen, &sfe->nameval[sfe->namelen],
sfe->flags))
continue;
break;
@@ -929,6 +953,7 @@ xfs_attr_shortform_lookup(xfs_da_args_t *args)
for (i = 0; i < sf->hdr.count;
sfe = xfs_attr_sf_nextentry(sfe), i++) {
if (xfs_attr_match(args, sfe->namelen, sfe->nameval,
+ sfe->valuelen, &sfe->nameval[sfe->namelen],
sfe->flags))
return -EEXIST;
}
@@ -956,6 +981,7 @@ xfs_attr_shortform_getvalue(
for (i = 0; i < sf->hdr.count;
sfe = xfs_attr_sf_nextentry(sfe), i++) {
if (xfs_attr_match(args, sfe->namelen, sfe->nameval,
+ sfe->valuelen, &sfe->nameval[sfe->namelen],
sfe->flags))
return xfs_attr_copy_value(args,
&sfe->nameval[args->namelen], sfe->valuelen);
@@ -1008,7 +1034,7 @@ xfs_attr_shortform_to_leaf(
nargs.total = args->total;
nargs.whichfork = XFS_ATTR_FORK;
nargs.trans = args->trans;
- nargs.op_flags = XFS_DA_OP_OKNOENT;
+ nargs.op_flags = XFS_DA_OP_OKNOENT | XFS_DA_OP_NVLOOKUP;
nargs.owner = args->owner;
sfe = &sf->list[0];
@@ -1212,7 +1238,7 @@ xfs_attr3_leaf_to_shortform(
nargs.total = args->total;
nargs.whichfork = XFS_ATTR_FORK;
nargs.trans = args->trans;
- nargs.op_flags = XFS_DA_OP_OKNOENT;
+ nargs.op_flags = XFS_DA_OP_OKNOENT | XFS_DA_OP_NVLOOKUP;
nargs.owner = args->owner;
for (i = 0; i < ichdr.count; entry++, i++) {
@@ -2509,14 +2535,17 @@ xfs_attr3_leaf_lookup_int(
if (entry->flags & XFS_ATTR_LOCAL) {
name_loc = xfs_attr3_leaf_name_local(leaf, probe);
if (!xfs_attr_match(args, name_loc->namelen,
- name_loc->nameval, entry->flags))
+ name_loc->nameval,
+ be16_to_cpu(name_loc->valuelen),
+ &name_loc->nameval[name_loc->namelen],
+ entry->flags))
continue;
args->index = probe;
return -EEXIST;
} else {
name_rmt = xfs_attr3_leaf_name_remote(leaf, probe);
if (!xfs_attr_match(args, name_rmt->namelen,
- name_rmt->name, entry->flags))
+ name_rmt->name, 0, NULL, entry->flags))
continue;
args->index = probe;
args->rmtvaluelen = be32_to_cpu(name_rmt->valuelen);
diff --git a/fs/xfs/libxfs/xfs_da_btree.h b/fs/xfs/libxfs/xfs_da_btree.h
index 7a004786ee0a2..1bcb291150eb5 100644
--- a/fs/xfs/libxfs/xfs_da_btree.h
+++ b/fs/xfs/libxfs/xfs_da_btree.h
@@ -94,6 +94,7 @@ typedef struct xfs_da_args {
#define XFS_DA_OP_REMOVE (1u << 6) /* this is a remove operation */
#define XFS_DA_OP_RECOVERY (1u << 7) /* Log recovery operation */
#define XFS_DA_OP_LOGGED (1u << 8) /* Use intent items to track op */
+#define XFS_DA_OP_NVLOOKUP (1u << 9) /* Match local attr on name+value */
#define XFS_DA_OP_FLAGS \
{ XFS_DA_OP_JUSTCHECK, "JUSTCHECK" }, \
@@ -104,7 +105,8 @@ typedef struct xfs_da_args {
{ XFS_DA_OP_NOTIME, "NOTIME" }, \
{ XFS_DA_OP_REMOVE, "REMOVE" }, \
{ XFS_DA_OP_RECOVERY, "RECOVERY" }, \
- { XFS_DA_OP_LOGGED, "LOGGED" }
+ { XFS_DA_OP_LOGGED, "LOGGED" }, \
+ { XFS_DA_OP_NVLOOKUP, "NVLOOKUP" }
/*
* Storage for holding state during Btree searches and split/join ops.
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 06/14] xfs: preserve NVLOOKUP in xfs_attr_set
2023-12-31 19:33 ` [PATCHSET v13.0 3/7] xfs: name-value xattr lookups Darrick J. Wong
` (4 preceding siblings ...)
2023-12-31 20:46 ` [PATCH 05/14] xfs: allow xattr matching on name and value for local/sf attrs Darrick J. Wong
@ 2023-12-31 20:46 ` Darrick J. Wong
2023-12-31 20:46 ` [PATCH 07/14] xfs: restructure xfs_attr_complete_op a bit Darrick J. Wong
` (7 subsequent siblings)
13 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:46 UTC (permalink / raw)
To: djwong; +Cc: catherine.hoang, allison.henderson, linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
Preserve the attr-value lookup flag when calling xfs_attr_set. Normal
xattr users will never use this, but parent pointer fsck will.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/libxfs/xfs_attr.c | 8 ++++----
1 file changed, 4 insertions(+), 4 deletions(-)
diff --git a/fs/xfs/libxfs/xfs_attr.c b/fs/xfs/libxfs/xfs_attr.c
index 2de3f6ad36601..d1f228c67857f 100644
--- a/fs/xfs/libxfs/xfs_attr.c
+++ b/fs/xfs/libxfs/xfs_attr.c
@@ -945,11 +945,11 @@ xfs_attr_set(
/*
* We have no control over the attribute names that userspace passes us
* to remove, so we have to allow the name lookup prior to attribute
- * removal to fail as well. Preserve the logged flag, since we need
- * to pass that through to the logging code.
+ * removal to fail as well. Preserve the logged and vlookup flags,
+ * since we need to pass them through to the lower levels.
*/
- args->op_flags = XFS_DA_OP_OKNOENT |
- (args->op_flags & XFS_DA_OP_LOGGED);
+ args->op_flags &= (XFS_DA_OP_LOGGED | XFS_DA_OP_NVLOOKUP);
+ args->op_flags |= XFS_DA_OP_OKNOENT;
if (!is_remove) {
XFS_STATS_INC(mp, xs_attr_set);
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 07/14] xfs: restructure xfs_attr_complete_op a bit
2023-12-31 19:33 ` [PATCHSET v13.0 3/7] xfs: name-value xattr lookups Darrick J. Wong
` (5 preceding siblings ...)
2023-12-31 20:46 ` [PATCH 06/14] xfs: preserve NVLOOKUP in xfs_attr_set Darrick J. Wong
@ 2023-12-31 20:46 ` Darrick J. Wong
2023-12-31 20:46 ` [PATCH 08/14] xfs: use helpers to extract xattr op from opflags Darrick J. Wong
` (6 subsequent siblings)
13 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:46 UTC (permalink / raw)
To: djwong; +Cc: catherine.hoang, allison.henderson, linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
Reduce the indentation in this function by flattening the nested if
statements. We're going to add more code later to this function later,
hence the early cleanup. No functional changes.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/libxfs/xfs_attr.c | 10 +++++-----
1 file changed, 5 insertions(+), 5 deletions(-)
diff --git a/fs/xfs/libxfs/xfs_attr.c b/fs/xfs/libxfs/xfs_attr.c
index d1f228c67857f..1d4cd8a8ac9ab 100644
--- a/fs/xfs/libxfs/xfs_attr.c
+++ b/fs/xfs/libxfs/xfs_attr.c
@@ -421,11 +421,11 @@ xfs_attr_complete_op(
bool do_replace = args->op_flags & XFS_DA_OP_REPLACE;
args->op_flags &= ~XFS_DA_OP_REPLACE;
- if (do_replace) {
- args->attr_filter &= ~XFS_ATTR_INCOMPLETE;
- return replace_state;
- }
- return XFS_DAS_DONE;
+ if (!do_replace)
+ return XFS_DAS_DONE;
+
+ args->attr_filter &= ~XFS_ATTR_INCOMPLETE;
+ return replace_state;
}
static int
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 08/14] xfs: use helpers to extract xattr op from opflags
2023-12-31 19:33 ` [PATCHSET v13.0 3/7] xfs: name-value xattr lookups Darrick J. Wong
` (6 preceding siblings ...)
2023-12-31 20:46 ` [PATCH 07/14] xfs: restructure xfs_attr_complete_op a bit Darrick J. Wong
@ 2023-12-31 20:46 ` Darrick J. Wong
2023-12-31 20:47 ` [PATCH 09/14] xfs: validate recovered name buffers when recovering xattr items Darrick J. Wong
` (5 subsequent siblings)
13 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:46 UTC (permalink / raw)
To: djwong; +Cc: catherine.hoang, allison.henderson, linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
Create helper functions to extract the xattr op from the ondisk xattri
log item and the incore attr intent item. These will get more use in
the patches that follow.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/libxfs/xfs_attr.h | 5 +++++
fs/xfs/xfs_attr_item.c | 16 ++++++++++------
2 files changed, 15 insertions(+), 6 deletions(-)
diff --git a/fs/xfs/libxfs/xfs_attr.h b/fs/xfs/libxfs/xfs_attr.h
index 273e8dff76c07..ca51b93873bb9 100644
--- a/fs/xfs/libxfs/xfs_attr.h
+++ b/fs/xfs/libxfs/xfs_attr.h
@@ -529,6 +529,11 @@ struct xfs_attr_intent {
struct xfs_bmbt_irec xattri_map;
};
+static inline unsigned int
+xfs_attr_intent_op(const struct xfs_attr_intent *attr)
+{
+ return attr->xattri_op_flags & XFS_ATTRI_OP_FLAGS_TYPE_MASK;
+}
/*========================================================================
* Function prototypes for the kernel.
diff --git a/fs/xfs/xfs_attr_item.c b/fs/xfs/xfs_attr_item.c
index b6b558f4501ad..80499811a9f72 100644
--- a/fs/xfs/xfs_attr_item.c
+++ b/fs/xfs/xfs_attr_item.c
@@ -308,6 +308,12 @@ xfs_attrd_item_intent(
return &ATTRD_ITEM(lip)->attrd_attrip->attri_item;
}
+static inline unsigned int
+xfs_attr_log_item_op(const struct xfs_attri_log_format *attrp)
+{
+ return attrp->alfi_op_flags & XFS_ATTRI_OP_FLAGS_TYPE_MASK;
+}
+
/* Log an attr to the intent item. */
STATIC void
xfs_attr_log_item(
@@ -466,8 +472,7 @@ xfs_attri_validate(
struct xfs_mount *mp,
struct xfs_attri_log_format *attrp)
{
- unsigned int op = attrp->alfi_op_flags &
- XFS_ATTRI_OP_FLAGS_TYPE_MASK;
+ unsigned int op = xfs_attr_log_item_op(attrp);
if (!xfs_sb_version_haslogxattrs(&mp->m_sb) &&
!xfs_attri_can_use_without_log_assistance(mp))
@@ -524,8 +529,7 @@ xfs_attri_recover_work(
args = (struct xfs_da_args *)(attr + 1);
attr->xattri_da_args = args;
- attr->xattri_op_flags = attrp->alfi_op_flags &
- XFS_ATTRI_OP_FLAGS_TYPE_MASK;
+ attr->xattri_op_flags = xfs_attr_log_item_op(attrp);
/*
* We're reconstructing the deferred work state structure from the
@@ -546,7 +550,7 @@ xfs_attri_recover_work(
XFS_DA_OP_LOGGED;
args->owner = args->dp->i_ino;
- switch (attr->xattri_op_flags) {
+ switch (xfs_attr_intent_op(attr)) {
case XFS_ATTRI_OP_FLAGS_SET:
case XFS_ATTRI_OP_FLAGS_REPLACE:
args->value = nv->value.i_addr;
@@ -728,7 +732,7 @@ xlog_recover_attri_commit_pass2(
}
/* Check the number of log iovecs makes sense for the op code. */
- op = attri_formatp->alfi_op_flags & XFS_ATTRI_OP_FLAGS_TYPE_MASK;
+ op = xfs_attr_log_item_op(attri_formatp);
switch (op) {
case XFS_ATTRI_OP_FLAGS_SET:
case XFS_ATTRI_OP_FLAGS_REPLACE:
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 09/14] xfs: validate recovered name buffers when recovering xattr items
2023-12-31 19:33 ` [PATCHSET v13.0 3/7] xfs: name-value xattr lookups Darrick J. Wong
` (7 preceding siblings ...)
2023-12-31 20:46 ` [PATCH 08/14] xfs: use helpers to extract xattr op from opflags Darrick J. Wong
@ 2023-12-31 20:47 ` Darrick J. Wong
2023-12-31 20:47 ` [PATCH 10/14] xfs: always set args->value in xfs_attri_item_recover Darrick J. Wong
` (4 subsequent siblings)
13 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:47 UTC (permalink / raw)
To: djwong; +Cc: catherine.hoang, allison.henderson, linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
Strengthen the xattri log item recovery code by checking that we
actually have the required name and newname buffers for whatever
operation we're replaying.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/xfs_attr_item.c | 58 +++++++++++++++++++++++++++++++++++++++---------
1 file changed, 47 insertions(+), 11 deletions(-)
diff --git a/fs/xfs/xfs_attr_item.c b/fs/xfs/xfs_attr_item.c
index 80499811a9f72..b9a9bb7c1ebad 100644
--- a/fs/xfs/xfs_attr_item.c
+++ b/fs/xfs/xfs_attr_item.c
@@ -712,22 +712,20 @@ xlog_recover_attri_commit_pass2(
const void *attr_value = NULL;
const void *attr_name;
size_t len;
- unsigned int op;
-
- attri_formatp = item->ri_buf[0].i_addr;
- attr_name = item->ri_buf[1].i_addr;
+ unsigned int op, i = 0;
/* Validate xfs_attri_log_format before the large memory allocation */
len = sizeof(struct xfs_attri_log_format);
- if (item->ri_buf[0].i_len != len) {
+ if (item->ri_buf[i].i_len != len) {
XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp,
item->ri_buf[0].i_addr, item->ri_buf[0].i_len);
return -EFSCORRUPTED;
}
+ attri_formatp = item->ri_buf[i].i_addr;
if (!xfs_attri_validate(mp, attri_formatp)) {
XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp,
- item->ri_buf[0].i_addr, item->ri_buf[0].i_len);
+ attri_formatp, len);
return -EFSCORRUPTED;
}
@@ -756,31 +754,69 @@ xlog_recover_attri_commit_pass2(
attri_formatp, len);
return -EFSCORRUPTED;
}
+ i++;
/* Validate the attr name */
- if (item->ri_buf[1].i_len !=
+ if (item->ri_buf[i].i_len !=
xlog_calc_iovec_len(attri_formatp->alfi_name_len)) {
XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp,
- item->ri_buf[0].i_addr, item->ri_buf[0].i_len);
+ attri_formatp, len);
return -EFSCORRUPTED;
}
+ attr_name = item->ri_buf[i].i_addr;
if (!xfs_attr_namecheck(attr_name, attri_formatp->alfi_name_len)) {
XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp,
- item->ri_buf[1].i_addr, item->ri_buf[1].i_len);
+ attri_formatp, len);
return -EFSCORRUPTED;
}
+ i++;
/* Validate the attr value, if present */
if (attri_formatp->alfi_value_len != 0) {
- if (item->ri_buf[2].i_len != xlog_calc_iovec_len(attri_formatp->alfi_value_len)) {
+ if (item->ri_buf[i].i_len != xlog_calc_iovec_len(attri_formatp->alfi_value_len)) {
XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp,
item->ri_buf[0].i_addr,
item->ri_buf[0].i_len);
return -EFSCORRUPTED;
}
- attr_value = item->ri_buf[2].i_addr;
+ attr_value = item->ri_buf[i].i_addr;
+ i++;
+ }
+
+ /*
+ * Make sure we got the correct number of buffers for the operation
+ * that we just loaded.
+ */
+ if (i != item->ri_total) {
+ XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp,
+ attri_formatp, len);
+ return -EFSCORRUPTED;
+ }
+
+ switch (op) {
+ case XFS_ATTRI_OP_FLAGS_REMOVE:
+ /* Regular remove operations operate only on names. */
+ if (attr_value != NULL || attri_formatp->alfi_value_len != 0) {
+ XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp,
+ attri_formatp, len);
+ return -EFSCORRUPTED;
+ }
+ fallthrough;
+ case XFS_ATTRI_OP_FLAGS_SET:
+ case XFS_ATTRI_OP_FLAGS_REPLACE:
+ /*
+ * Regular xattr set/remove/replace operations require a name
+ * and do not take a newname. Values are optional for set and
+ * replace.
+ */
+ if (attr_name == NULL || attri_formatp->alfi_name_len == 0) {
+ XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp,
+ attri_formatp, len);
+ return -EFSCORRUPTED;
+ }
+ break;
}
/*
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 10/14] xfs: always set args->value in xfs_attri_item_recover
2023-12-31 19:33 ` [PATCHSET v13.0 3/7] xfs: name-value xattr lookups Darrick J. Wong
` (8 preceding siblings ...)
2023-12-31 20:47 ` [PATCH 09/14] xfs: validate recovered name buffers when recovering xattr items Darrick J. Wong
@ 2023-12-31 20:47 ` Darrick J. Wong
2023-12-31 20:47 ` [PATCH 11/14] xfs: use local variables for name and value length in _attri_commit_pass2 Darrick J. Wong
` (3 subsequent siblings)
13 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:47 UTC (permalink / raw)
To: djwong; +Cc: catherine.hoang, allison.henderson, linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
Always set args->value to the recovered value buffer. This reduces the
amount of code in the switch statement, and hence the amount of thinking
that I have to do. We validated the recovered buffers, supposedly.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/xfs_attr_item.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/fs/xfs/xfs_attr_item.c b/fs/xfs/xfs_attr_item.c
index b9a9bb7c1ebad..fbf392ab2dc5b 100644
--- a/fs/xfs/xfs_attr_item.c
+++ b/fs/xfs/xfs_attr_item.c
@@ -545,6 +545,8 @@ xfs_attri_recover_work(
args->name = nv->name.i_addr;
args->namelen = nv->name.i_len;
args->hashval = xfs_da_hashname(args->name, args->namelen);
+ args->value = nv->value.i_addr;
+ args->valuelen = nv->value.i_len;
args->attr_filter = attrp->alfi_attr_filter & XFS_ATTRI_FILTER_MASK;
args->op_flags = XFS_DA_OP_RECOVERY | XFS_DA_OP_OKNOENT |
XFS_DA_OP_LOGGED;
@@ -553,8 +555,6 @@ xfs_attri_recover_work(
switch (xfs_attr_intent_op(attr)) {
case XFS_ATTRI_OP_FLAGS_SET:
case XFS_ATTRI_OP_FLAGS_REPLACE:
- args->value = nv->value.i_addr;
- args->valuelen = nv->value.i_len;
args->total = xfs_attr_calc_size(args, &local);
if (xfs_inode_hasattr(args->dp))
attr->xattri_dela_state = xfs_attr_init_replace_state(args);
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 11/14] xfs: use local variables for name and value length in _attri_commit_pass2
2023-12-31 19:33 ` [PATCHSET v13.0 3/7] xfs: name-value xattr lookups Darrick J. Wong
` (9 preceding siblings ...)
2023-12-31 20:47 ` [PATCH 10/14] xfs: always set args->value in xfs_attri_item_recover Darrick J. Wong
@ 2023-12-31 20:47 ` Darrick J. Wong
2023-12-31 20:47 ` [PATCH 12/14] xfs: log NVLOOKUP xattr removal operations Darrick J. Wong
` (2 subsequent siblings)
13 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:47 UTC (permalink / raw)
To: djwong; +Cc: catherine.hoang, allison.henderson, linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
We're about to start using tagged unions in the xattr log format, so
create a bunch of local variables in the recovery function so we only
have to decode the log item fields once.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/xfs_attr_item.c | 25 ++++++++++++++-----------
1 file changed, 14 insertions(+), 11 deletions(-)
diff --git a/fs/xfs/xfs_attr_item.c b/fs/xfs/xfs_attr_item.c
index fbf392ab2dc5b..19fc535b4a22b 100644
--- a/fs/xfs/xfs_attr_item.c
+++ b/fs/xfs/xfs_attr_item.c
@@ -709,9 +709,11 @@ xlog_recover_attri_commit_pass2(
struct xfs_attri_log_item *attrip;
struct xfs_attri_log_format *attri_formatp;
struct xfs_attri_log_nameval *nv;
- const void *attr_value = NULL;
const void *attr_name;
+ const void *attr_value = NULL;
size_t len;
+ unsigned int name_len = 0;
+ unsigned int value_len = 0;
unsigned int op, i = 0;
/* Validate xfs_attri_log_format before the large memory allocation */
@@ -740,6 +742,8 @@ xlog_recover_attri_commit_pass2(
attri_formatp, len);
return -EFSCORRUPTED;
}
+ name_len = attri_formatp->alfi_name_len;
+ value_len = attri_formatp->alfi_value_len;
break;
case XFS_ATTRI_OP_FLAGS_REMOVE:
/* Log item, attr name */
@@ -748,6 +752,7 @@ xlog_recover_attri_commit_pass2(
attri_formatp, len);
return -EFSCORRUPTED;
}
+ name_len = attri_formatp->alfi_name_len;
break;
default:
XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp,
@@ -757,15 +762,14 @@ xlog_recover_attri_commit_pass2(
i++;
/* Validate the attr name */
- if (item->ri_buf[i].i_len !=
- xlog_calc_iovec_len(attri_formatp->alfi_name_len)) {
+ if (item->ri_buf[i].i_len != xlog_calc_iovec_len(name_len)) {
XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp,
attri_formatp, len);
return -EFSCORRUPTED;
}
attr_name = item->ri_buf[i].i_addr;
- if (!xfs_attr_namecheck(attr_name, attri_formatp->alfi_name_len)) {
+ if (!xfs_attr_namecheck(attr_name, name_len)) {
XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp,
attri_formatp, len);
return -EFSCORRUPTED;
@@ -773,8 +777,8 @@ xlog_recover_attri_commit_pass2(
i++;
/* Validate the attr value, if present */
- if (attri_formatp->alfi_value_len != 0) {
- if (item->ri_buf[i].i_len != xlog_calc_iovec_len(attri_formatp->alfi_value_len)) {
+ if (value_len != 0) {
+ if (item->ri_buf[i].i_len != xlog_calc_iovec_len(value_len)) {
XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp,
item->ri_buf[0].i_addr,
item->ri_buf[0].i_len);
@@ -798,7 +802,7 @@ xlog_recover_attri_commit_pass2(
switch (op) {
case XFS_ATTRI_OP_FLAGS_REMOVE:
/* Regular remove operations operate only on names. */
- if (attr_value != NULL || attri_formatp->alfi_value_len != 0) {
+ if (attr_value != NULL || value_len != 0) {
XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp,
attri_formatp, len);
return -EFSCORRUPTED;
@@ -811,7 +815,7 @@ xlog_recover_attri_commit_pass2(
* and do not take a newname. Values are optional for set and
* replace.
*/
- if (attr_name == NULL || attri_formatp->alfi_name_len == 0) {
+ if (attr_name == NULL || name_len == 0) {
XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp,
attri_formatp, len);
return -EFSCORRUPTED;
@@ -824,9 +828,8 @@ xlog_recover_attri_commit_pass2(
* name/value buffer to the recovered incore log item and drop our
* reference.
*/
- nv = xfs_attri_log_nameval_alloc(attr_name,
- attri_formatp->alfi_name_len, attr_value,
- attri_formatp->alfi_value_len);
+ nv = xfs_attri_log_nameval_alloc(attr_name, name_len,
+ attr_value, value_len);
attrip = xfs_attri_init(mp, nv);
memcpy(&attrip->attri_format, attri_formatp, len);
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 12/14] xfs: log NVLOOKUP xattr removal operations
2023-12-31 19:33 ` [PATCHSET v13.0 3/7] xfs: name-value xattr lookups Darrick J. Wong
` (10 preceding siblings ...)
2023-12-31 20:47 ` [PATCH 11/14] xfs: use local variables for name and value length in _attri_commit_pass2 Darrick J. Wong
@ 2023-12-31 20:47 ` Darrick J. Wong
2023-12-31 20:48 ` [PATCH 13/14] xfs: log NVLOOKUP xattr setting operations Darrick J. Wong
2023-12-31 20:48 ` [PATCH 14/14] xfs: log NVLOOKUP xattr nvreplace operations Darrick J. Wong
13 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:47 UTC (permalink / raw)
To: djwong; +Cc: catherine.hoang, allison.henderson, linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
If high level code wants to do a deferred xattr remove operation with
the NVLOOKUP flag set, we need to push this through the log.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/libxfs/xfs_attr.c | 2 ++
fs/xfs/libxfs/xfs_log_format.h | 1 +
fs/xfs/xfs_attr_item.c | 18 ++++++++++++++++++
3 files changed, 21 insertions(+)
diff --git a/fs/xfs/libxfs/xfs_attr.c b/fs/xfs/libxfs/xfs_attr.c
index 1d4cd8a8ac9ab..76cc9753cb116 100644
--- a/fs/xfs/libxfs/xfs_attr.c
+++ b/fs/xfs/libxfs/xfs_attr.c
@@ -904,6 +904,8 @@ xfs_attr_defer_add(
new->xattri_dela_state = xfs_attr_init_replace_state(args);
break;
case XFS_ATTRI_OP_FLAGS_REMOVE:
+ if (args->op_flags & XFS_DA_OP_NVLOOKUP)
+ new->xattri_op_flags = XFS_ATTRI_OP_FLAGS_NVREMOVE;
new->xattri_dela_state = xfs_attr_init_remove_state(args);
break;
default:
diff --git a/fs/xfs/libxfs/xfs_log_format.h b/fs/xfs/libxfs/xfs_log_format.h
index d4531060b6b49..bf648b75194ef 100644
--- a/fs/xfs/libxfs/xfs_log_format.h
+++ b/fs/xfs/libxfs/xfs_log_format.h
@@ -1043,6 +1043,7 @@ struct xfs_icreate_log {
#define XFS_ATTRI_OP_FLAGS_SET 1 /* Set the attribute */
#define XFS_ATTRI_OP_FLAGS_REMOVE 2 /* Remove the attribute */
#define XFS_ATTRI_OP_FLAGS_REPLACE 3 /* Replace the attribute */
+#define XFS_ATTRI_OP_FLAGS_NVREMOVE 4 /* Remove attr w/ vlookup */
#define XFS_ATTRI_OP_FLAGS_TYPE_MASK 0xFF /* Flags type mask */
/*
diff --git a/fs/xfs/xfs_attr_item.c b/fs/xfs/xfs_attr_item.c
index 19fc535b4a22b..bd4bb931da2ae 100644
--- a/fs/xfs/xfs_attr_item.c
+++ b/fs/xfs/xfs_attr_item.c
@@ -492,6 +492,7 @@ xfs_attri_validate(
case XFS_ATTRI_OP_FLAGS_SET:
case XFS_ATTRI_OP_FLAGS_REPLACE:
case XFS_ATTRI_OP_FLAGS_REMOVE:
+ case XFS_ATTRI_OP_FLAGS_NVREMOVE:
break;
default:
return false;
@@ -561,6 +562,9 @@ xfs_attri_recover_work(
else
attr->xattri_dela_state = xfs_attr_init_add_state(args);
break;
+ case XFS_ATTRI_OP_FLAGS_NVREMOVE:
+ args->op_flags |= XFS_DA_OP_NVLOOKUP;
+ fallthrough;
case XFS_ATTRI_OP_FLAGS_REMOVE:
attr->xattri_dela_state = xfs_attr_init_remove_state(args);
break;
@@ -734,6 +738,16 @@ xlog_recover_attri_commit_pass2(
/* Check the number of log iovecs makes sense for the op code. */
op = xfs_attr_log_item_op(attri_formatp);
switch (op) {
+ case XFS_ATTRI_OP_FLAGS_NVREMOVE:
+ /* Log item, attr name, optional attr value */
+ if (item->ri_total != 3 && item->ri_total != 2) {
+ XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp,
+ attri_formatp, len);
+ return -EFSCORRUPTED;
+ }
+ name_len = attri_formatp->alfi_name_len;
+ value_len = attri_formatp->alfi_value_len;
+ break;
case XFS_ATTRI_OP_FLAGS_SET:
case XFS_ATTRI_OP_FLAGS_REPLACE:
/* Log item, attr name, attr value */
@@ -808,12 +822,16 @@ xlog_recover_attri_commit_pass2(
return -EFSCORRUPTED;
}
fallthrough;
+ case XFS_ATTRI_OP_FLAGS_NVREMOVE:
case XFS_ATTRI_OP_FLAGS_SET:
case XFS_ATTRI_OP_FLAGS_REPLACE:
/*
* Regular xattr set/remove/replace operations require a name
* and do not take a newname. Values are optional for set and
* replace.
+ *
+ * Name-value remove operations must have a name, do not
+ * take a newname, and can take a value.
*/
if (attr_name == NULL || name_len == 0) {
XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp,
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 13/14] xfs: log NVLOOKUP xattr setting operations
2023-12-31 19:33 ` [PATCHSET v13.0 3/7] xfs: name-value xattr lookups Darrick J. Wong
` (11 preceding siblings ...)
2023-12-31 20:47 ` [PATCH 12/14] xfs: log NVLOOKUP xattr removal operations Darrick J. Wong
@ 2023-12-31 20:48 ` Darrick J. Wong
2023-12-31 20:48 ` [PATCH 14/14] xfs: log NVLOOKUP xattr nvreplace operations Darrick J. Wong
13 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:48 UTC (permalink / raw)
To: djwong; +Cc: catherine.hoang, allison.henderson, linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
If high level code wants to do a deferred xattr set operation with the
NVLOOKUP flag set, we need to push this through the log.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/libxfs/xfs_attr.c | 2 ++
fs/xfs/libxfs/xfs_log_format.h | 1 +
fs/xfs/xfs_attr_item.c | 8 +++++++-
3 files changed, 10 insertions(+), 1 deletion(-)
diff --git a/fs/xfs/libxfs/xfs_attr.c b/fs/xfs/libxfs/xfs_attr.c
index 76cc9753cb116..a900e184f6566 100644
--- a/fs/xfs/libxfs/xfs_attr.c
+++ b/fs/xfs/libxfs/xfs_attr.c
@@ -898,6 +898,8 @@ xfs_attr_defer_add(
switch (op_flags) {
case XFS_ATTRI_OP_FLAGS_SET:
+ if (args->op_flags & XFS_DA_OP_NVLOOKUP)
+ new->xattri_op_flags = XFS_ATTRI_OP_FLAGS_NVSET;
new->xattri_dela_state = xfs_attr_init_add_state(args);
break;
case XFS_ATTRI_OP_FLAGS_REPLACE:
diff --git a/fs/xfs/libxfs/xfs_log_format.h b/fs/xfs/libxfs/xfs_log_format.h
index bf648b75194ef..2ac520a18e909 100644
--- a/fs/xfs/libxfs/xfs_log_format.h
+++ b/fs/xfs/libxfs/xfs_log_format.h
@@ -1044,6 +1044,7 @@ struct xfs_icreate_log {
#define XFS_ATTRI_OP_FLAGS_REMOVE 2 /* Remove the attribute */
#define XFS_ATTRI_OP_FLAGS_REPLACE 3 /* Replace the attribute */
#define XFS_ATTRI_OP_FLAGS_NVREMOVE 4 /* Remove attr w/ vlookup */
+#define XFS_ATTRI_OP_FLAGS_NVSET 5 /* Set attr with w/ vlookup */
#define XFS_ATTRI_OP_FLAGS_TYPE_MASK 0xFF /* Flags type mask */
/*
diff --git a/fs/xfs/xfs_attr_item.c b/fs/xfs/xfs_attr_item.c
index bd4bb931da2ae..1ae2ecfe780c2 100644
--- a/fs/xfs/xfs_attr_item.c
+++ b/fs/xfs/xfs_attr_item.c
@@ -493,6 +493,7 @@ xfs_attri_validate(
case XFS_ATTRI_OP_FLAGS_REPLACE:
case XFS_ATTRI_OP_FLAGS_REMOVE:
case XFS_ATTRI_OP_FLAGS_NVREMOVE:
+ case XFS_ATTRI_OP_FLAGS_NVSET:
break;
default:
return false;
@@ -554,6 +555,9 @@ xfs_attri_recover_work(
args->owner = args->dp->i_ino;
switch (xfs_attr_intent_op(attr)) {
+ case XFS_ATTRI_OP_FLAGS_NVSET:
+ args->op_flags |= XFS_DA_OP_NVLOOKUP;
+ fallthrough;
case XFS_ATTRI_OP_FLAGS_SET:
case XFS_ATTRI_OP_FLAGS_REPLACE:
args->total = xfs_attr_calc_size(args, &local);
@@ -738,6 +742,7 @@ xlog_recover_attri_commit_pass2(
/* Check the number of log iovecs makes sense for the op code. */
op = xfs_attr_log_item_op(attri_formatp);
switch (op) {
+ case XFS_ATTRI_OP_FLAGS_NVSET:
case XFS_ATTRI_OP_FLAGS_NVREMOVE:
/* Log item, attr name, optional attr value */
if (item->ri_total != 3 && item->ri_total != 2) {
@@ -822,6 +827,7 @@ xlog_recover_attri_commit_pass2(
return -EFSCORRUPTED;
}
fallthrough;
+ case XFS_ATTRI_OP_FLAGS_NVSET:
case XFS_ATTRI_OP_FLAGS_NVREMOVE:
case XFS_ATTRI_OP_FLAGS_SET:
case XFS_ATTRI_OP_FLAGS_REPLACE:
@@ -830,7 +836,7 @@ xlog_recover_attri_commit_pass2(
* and do not take a newname. Values are optional for set and
* replace.
*
- * Name-value remove operations must have a name, do not
+ * Name-value set/remove operations must have a name, do not
* take a newname, and can take a value.
*/
if (attr_name == NULL || name_len == 0) {
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 14/14] xfs: log NVLOOKUP xattr nvreplace operations
2023-12-31 19:33 ` [PATCHSET v13.0 3/7] xfs: name-value xattr lookups Darrick J. Wong
` (12 preceding siblings ...)
2023-12-31 20:48 ` [PATCH 13/14] xfs: log NVLOOKUP xattr setting operations Darrick J. Wong
@ 2023-12-31 20:48 ` Darrick J. Wong
13 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:48 UTC (permalink / raw)
To: djwong; +Cc: Allison Henderson, catherine.hoang, allison.henderson, linux-xfs
From: Allison Henderson <allison.henderson@oracle.com>
(Formerly titled "xfs: Add new name to attri/d" and described as
follows:
This patch adds two new fields to the atti/d. They are nname and
nnamelen. This will be used for parent pointer updates since a
rename operation may cause the parent pointer to update both the
name and value. So we need to carry both the new name as well as
the target name in the attri/d.)
If high level code wants to do a deferred xattr nvreplace operation with
the NVLOOKUP flag set, we need to push this through the log. This log
item records the old name/value pair and the new name/value pair, and
completely replaces one with the other. Parent pointers will need this
ability to handle rename moving a child file between parents.
Signed-off-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
[djwong: reworked to handle new disk format]
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/libxfs/xfs_attr.c | 16 +++
fs/xfs/libxfs/xfs_attr.h | 4 -
fs/xfs/libxfs/xfs_da_btree.h | 6 +
fs/xfs/libxfs/xfs_log_format.h | 27 ++++-
fs/xfs/xfs_attr_item.c | 207 +++++++++++++++++++++++++++++++++++-----
fs/xfs/xfs_attr_item.h | 2
6 files changed, 230 insertions(+), 32 deletions(-)
diff --git a/fs/xfs/libxfs/xfs_attr.c b/fs/xfs/libxfs/xfs_attr.c
index a900e184f6566..c6621af9554c2 100644
--- a/fs/xfs/libxfs/xfs_attr.c
+++ b/fs/xfs/libxfs/xfs_attr.c
@@ -425,6 +425,20 @@ xfs_attr_complete_op(
return XFS_DAS_DONE;
args->attr_filter &= ~XFS_ATTR_INCOMPLETE;
+ if (xfs_attr_intent_op(attr) != XFS_ATTRI_OP_FLAGS_NVREPLACE)
+ return replace_state;
+
+ /*
+ * NVREPLACE operations require the caller to set the old and new names
+ * and values explicitly.
+ */
+ ASSERT(args->new_namelen > 0);
+
+ args->name = args->new_name;
+ args->namelen = args->new_namelen;
+ args->hashval = xfs_da_hashname(args->name, args->namelen);
+ args->value = args->new_value;
+ args->valuelen = args->new_valuelen;
return replace_state;
}
@@ -903,6 +917,8 @@ xfs_attr_defer_add(
new->xattri_dela_state = xfs_attr_init_add_state(args);
break;
case XFS_ATTRI_OP_FLAGS_REPLACE:
+ if (args->op_flags & XFS_DA_OP_NVLOOKUP)
+ new->xattri_op_flags = XFS_ATTRI_OP_FLAGS_NVREPLACE;
new->xattri_dela_state = xfs_attr_init_replace_state(args);
break;
case XFS_ATTRI_OP_FLAGS_REMOVE:
diff --git a/fs/xfs/libxfs/xfs_attr.h b/fs/xfs/libxfs/xfs_attr.h
index ca51b93873bb9..b4e8ecee3e039 100644
--- a/fs/xfs/libxfs/xfs_attr.h
+++ b/fs/xfs/libxfs/xfs_attr.h
@@ -510,8 +510,8 @@ struct xfs_attr_intent {
struct xfs_da_args *xattri_da_args;
/*
- * Shared buffer containing the attr name and value so that the logging
- * code can share large memory buffers between log items.
+ * Shared buffer containing the attr name, new name, and value so that
+ * the logging code can share large memory buffers between log items.
*/
struct xfs_attri_log_nameval *xattri_nameval;
diff --git a/fs/xfs/libxfs/xfs_da_btree.h b/fs/xfs/libxfs/xfs_da_btree.h
index 1bcb291150eb5..93fcf49ab79dc 100644
--- a/fs/xfs/libxfs/xfs_da_btree.h
+++ b/fs/xfs/libxfs/xfs_da_btree.h
@@ -54,11 +54,15 @@ enum xfs_dacmp {
*/
typedef struct xfs_da_args {
struct xfs_da_geometry *geo; /* da block geometry */
- const uint8_t *name; /* string (maybe not NULL terminated) */
+ const uint8_t *name; /* string (maybe not NULL terminated) */
+ const uint8_t *new_name; /* new attr name */
int namelen; /* length of string (maybe no NULL) */
+ int new_namelen; /* new attr name len */
uint8_t filetype; /* filetype of inode for directories */
void *value; /* set of bytes (maybe contain NULLs) */
+ void *new_value; /* new xattr value (may contain NULLs) */
int valuelen; /* length of value */
+ int new_valuelen; /* length of new attr value */
unsigned int attr_filter; /* XFS_ATTR_{ROOT,SECURE,INCOMPLETE} */
unsigned int attr_flags; /* XATTR_{CREATE,REPLACE} */
xfs_dahash_t hashval; /* hash value of name */
diff --git a/fs/xfs/libxfs/xfs_log_format.h b/fs/xfs/libxfs/xfs_log_format.h
index 2ac520a18e909..285a0a089df24 100644
--- a/fs/xfs/libxfs/xfs_log_format.h
+++ b/fs/xfs/libxfs/xfs_log_format.h
@@ -115,11 +115,13 @@ struct xfs_unmount_log_format {
#define XLOG_REG_TYPE_BUD_FORMAT 26
#define XLOG_REG_TYPE_ATTRI_FORMAT 27
#define XLOG_REG_TYPE_ATTRD_FORMAT 28
-#define XLOG_REG_TYPE_ATTR_NAME 29
+#define XLOG_REG_TYPE_ATTR_NAME 29
#define XLOG_REG_TYPE_ATTR_VALUE 30
#define XLOG_REG_TYPE_SXI_FORMAT 31
#define XLOG_REG_TYPE_SXD_FORMAT 32
-#define XLOG_REG_TYPE_MAX 32
+#define XLOG_REG_TYPE_ATTR_NEWNAME 33
+#define XLOG_REG_TYPE_ATTR_NEWVALUE 34
+#define XLOG_REG_TYPE_MAX 34
/*
* Flags to log operation header
@@ -1045,6 +1047,7 @@ struct xfs_icreate_log {
#define XFS_ATTRI_OP_FLAGS_REPLACE 3 /* Replace the attribute */
#define XFS_ATTRI_OP_FLAGS_NVREMOVE 4 /* Remove attr w/ vlookup */
#define XFS_ATTRI_OP_FLAGS_NVSET 5 /* Set attr with w/ vlookup */
+#define XFS_ATTRI_OP_FLAGS_NVREPLACE 6 /* Replace attr name and val */
#define XFS_ATTRI_OP_FLAGS_TYPE_MASK 0xFF /* Flags type mask */
/*
@@ -1062,11 +1065,27 @@ struct xfs_icreate_log {
struct xfs_attri_log_format {
uint16_t alfi_type; /* attri log item type */
uint16_t alfi_size; /* size of this item */
- uint32_t __pad; /* pad to 64 bit aligned */
+
+ /*
+ * For NVREPLACE, this is the length of the new xattr value.
+ * alfi_value_len contains the length of the old xattr value.
+ */
+ uint32_t alfi_new_value_len;
+
uint64_t alfi_id; /* attri identifier */
uint64_t alfi_ino; /* the inode for this attr operation */
uint32_t alfi_op_flags; /* marks the op as a set or remove */
- uint32_t alfi_name_len; /* attr name length */
+ union {
+ uint32_t alfi_name_len; /* attr name length */
+ struct {
+ /*
+ * For NVREPLACE, these are the lengths of the old and
+ * new attr name.
+ */
+ uint16_t alfi_old_name_len;
+ uint16_t alfi_new_name_len;
+ };
+ };
uint32_t alfi_value_len; /* attr value length */
uint32_t alfi_attr_filter;/* attr filter flags */
};
diff --git a/fs/xfs/xfs_attr_item.c b/fs/xfs/xfs_attr_item.c
index 1ae2ecfe780c2..a1c8036310b21 100644
--- a/fs/xfs/xfs_attr_item.c
+++ b/fs/xfs/xfs_attr_item.c
@@ -73,8 +73,12 @@ static inline struct xfs_attri_log_nameval *
xfs_attri_log_nameval_alloc(
const void *name,
unsigned int name_len,
+ const void *new_name,
+ unsigned int new_name_len,
const void *value,
- unsigned int value_len)
+ unsigned int value_len,
+ const void *new_value,
+ unsigned int new_value_len)
{
struct xfs_attri_log_nameval *nv;
@@ -83,15 +87,26 @@ xfs_attri_log_nameval_alloc(
* this. But kvmalloc() utterly sucks, so we use our own version.
*/
nv = xlog_kvmalloc(sizeof(struct xfs_attri_log_nameval) +
- name_len + value_len);
+ name_len + new_name_len + value_len +
+ new_value_len);
nv->name.i_addr = nv + 1;
nv->name.i_len = name_len;
nv->name.i_type = XLOG_REG_TYPE_ATTR_NAME;
memcpy(nv->name.i_addr, name, name_len);
+ if (new_name_len) {
+ nv->new_name.i_addr = nv->name.i_addr + name_len;
+ nv->new_name.i_len = new_name_len;
+ memcpy(nv->new_name.i_addr, new_name, new_name_len);
+ } else {
+ nv->new_name.i_addr = NULL;
+ nv->new_name.i_len = 0;
+ }
+ nv->new_name.i_type = XLOG_REG_TYPE_ATTR_NEWNAME;
+
if (value_len) {
- nv->value.i_addr = nv->name.i_addr + name_len;
+ nv->value.i_addr = nv->name.i_addr + name_len + new_name_len;
nv->value.i_len = value_len;
memcpy(nv->value.i_addr, value, value_len);
} else {
@@ -100,6 +115,17 @@ xfs_attri_log_nameval_alloc(
}
nv->value.i_type = XLOG_REG_TYPE_ATTR_VALUE;
+ if (new_value_len) {
+ nv->new_value.i_addr = nv->name.i_addr + name_len +
+ new_name_len + value_len;
+ nv->new_value.i_len = new_value_len;
+ memcpy(nv->new_value.i_addr, new_value, new_value_len);
+ } else {
+ nv->new_value.i_addr = NULL;
+ nv->new_value.i_len = 0;
+ }
+ nv->new_value.i_type = XLOG_REG_TYPE_ATTR_NEWVALUE;
+
refcount_set(&nv->refcount, 1);
return nv;
}
@@ -145,11 +171,20 @@ xfs_attri_item_size(
*nbytes += sizeof(struct xfs_attri_log_format) +
xlog_calc_iovec_len(nv->name.i_len);
- if (!nv->value.i_len)
- return;
+ if (nv->new_name.i_len) {
+ *nvecs += 1;
+ *nbytes += xlog_calc_iovec_len(nv->new_name.i_len);
+ }
- *nvecs += 1;
- *nbytes += xlog_calc_iovec_len(nv->value.i_len);
+ if (nv->value.i_len) {
+ *nvecs += 1;
+ *nbytes += xlog_calc_iovec_len(nv->value.i_len);
+ }
+
+ if (nv->new_value.i_len) {
+ *nvecs += 1;
+ *nbytes += xlog_calc_iovec_len(nv->new_value.i_len);
+ }
}
/*
@@ -179,15 +214,28 @@ xfs_attri_item_format(
ASSERT(nv->name.i_len > 0);
attrip->attri_format.alfi_size++;
+ if (nv->new_name.i_len > 0)
+ attrip->attri_format.alfi_size++;
+
if (nv->value.i_len > 0)
attrip->attri_format.alfi_size++;
+ if (nv->new_value.i_len > 0)
+ attrip->attri_format.alfi_size++;
+
xlog_copy_iovec(lv, &vecp, XLOG_REG_TYPE_ATTRI_FORMAT,
&attrip->attri_format,
sizeof(struct xfs_attri_log_format));
xlog_copy_from_iovec(lv, &vecp, &nv->name);
+
+ if (nv->new_name.i_len > 0)
+ xlog_copy_from_iovec(lv, &vecp, &nv->new_name);
+
if (nv->value.i_len > 0)
xlog_copy_from_iovec(lv, &vecp, &nv->value);
+
+ if (nv->new_value.i_len > 0)
+ xlog_copy_from_iovec(lv, &vecp, &nv->new_value);
}
/*
@@ -333,7 +381,15 @@ xfs_attr_log_item(
ASSERT(!(attr->xattri_op_flags & ~XFS_ATTRI_OP_FLAGS_TYPE_MASK));
attrp->alfi_op_flags = attr->xattri_op_flags;
attrp->alfi_value_len = attr->xattri_nameval->value.i_len;
- attrp->alfi_name_len = attr->xattri_nameval->name.i_len;
+
+ if (xfs_attr_log_item_op(attrp) == XFS_ATTRI_OP_FLAGS_NVREPLACE) {
+ attrp->alfi_old_name_len = attr->xattri_nameval->name.i_len;
+ attrp->alfi_new_name_len = attr->xattri_nameval->new_name.i_len;
+ attrp->alfi_new_value_len = attr->xattri_nameval->new_value.i_len;
+ } else {
+ attrp->alfi_name_len = attr->xattri_nameval->name.i_len;
+ }
+
ASSERT(!(attr->xattri_da_args->attr_filter & ~XFS_ATTRI_FILTER_MASK));
attrp->alfi_attr_filter = attr->xattri_da_args->attr_filter;
}
@@ -374,8 +430,11 @@ xfs_attr_create_intent(
* Transfer our reference to the name/value buffer to the
* deferred work state structure.
*/
- attr->xattri_nameval = xfs_attri_log_nameval_alloc(args->name,
- args->namelen, args->value, args->valuelen);
+ attr->xattri_nameval = xfs_attri_log_nameval_alloc(
+ args->name, args->namelen,
+ args->new_name, args->new_namelen,
+ args->value, args->valuelen,
+ args->new_value, args->new_valuelen);
}
attrip = xfs_attri_init(mp, attr->xattri_nameval);
@@ -478,9 +537,6 @@ xfs_attri_validate(
!xfs_attri_can_use_without_log_assistance(mp))
return false;
- if (attrp->__pad != 0)
- return false;
-
if (attrp->alfi_op_flags & ~XFS_ATTRI_OP_FLAGS_TYPE_MASK)
return false;
@@ -489,23 +545,43 @@ xfs_attri_validate(
/* alfi_op_flags should be either a set or remove */
switch (op) {
+ case XFS_ATTRI_OP_FLAGS_REMOVE:
+ if (attrp->alfi_value_len != 0)
+ return false;
+ if (attrp->alfi_name_len == 0 ||
+ attrp->alfi_name_len > XATTR_NAME_MAX)
+ return false;
+ if (attrp->alfi_new_value_len != 0)
+ return false;
+ break;
case XFS_ATTRI_OP_FLAGS_SET:
case XFS_ATTRI_OP_FLAGS_REPLACE:
- case XFS_ATTRI_OP_FLAGS_REMOVE:
case XFS_ATTRI_OP_FLAGS_NVREMOVE:
case XFS_ATTRI_OP_FLAGS_NVSET:
+ if (attrp->alfi_name_len == 0 ||
+ attrp->alfi_name_len > XATTR_NAME_MAX)
+ return false;
+ if (attrp->alfi_value_len > XATTR_SIZE_MAX)
+ return false;
+ if (attrp->alfi_new_value_len != 0)
+ return false;
+ break;
+ case XFS_ATTRI_OP_FLAGS_NVREPLACE:
+ if (attrp->alfi_old_name_len == 0 ||
+ attrp->alfi_old_name_len > XATTR_NAME_MAX)
+ return false;
+ if (attrp->alfi_new_name_len == 0 ||
+ attrp->alfi_new_name_len > XATTR_NAME_MAX)
+ return false;
+ if (attrp->alfi_value_len > XATTR_SIZE_MAX)
+ return false;
+ if (attrp->alfi_new_value_len > XATTR_SIZE_MAX)
+ return false;
break;
default:
return false;
}
- if (attrp->alfi_value_len > XATTR_SIZE_MAX)
- return false;
-
- if ((attrp->alfi_name_len > XATTR_NAME_MAX) ||
- (attrp->alfi_name_len == 0))
- return false;
-
return xfs_verify_ino(mp, attrp->alfi_ino);
}
@@ -546,15 +622,20 @@ xfs_attri_recover_work(
args->whichfork = XFS_ATTR_FORK;
args->name = nv->name.i_addr;
args->namelen = nv->name.i_len;
+ args->new_name = nv->new_name.i_addr;
+ args->new_namelen = nv->new_name.i_len;
args->hashval = xfs_da_hashname(args->name, args->namelen);
args->value = nv->value.i_addr;
args->valuelen = nv->value.i_len;
+ args->new_value = nv->new_value.i_addr;
+ args->new_valuelen = nv->new_value.i_len;
args->attr_filter = attrp->alfi_attr_filter & XFS_ATTRI_FILTER_MASK;
args->op_flags = XFS_DA_OP_RECOVERY | XFS_DA_OP_OKNOENT |
XFS_DA_OP_LOGGED;
args->owner = args->dp->i_ino;
switch (xfs_attr_intent_op(attr)) {
+ case XFS_ATTRI_OP_FLAGS_NVREPLACE:
case XFS_ATTRI_OP_FLAGS_NVSET:
args->op_flags |= XFS_DA_OP_NVLOOKUP;
fallthrough;
@@ -666,7 +747,15 @@ xfs_attr_relog_intent(
new_attrp->alfi_ino = old_attrp->alfi_ino;
new_attrp->alfi_op_flags = old_attrp->alfi_op_flags;
new_attrp->alfi_value_len = old_attrp->alfi_value_len;
- new_attrp->alfi_name_len = old_attrp->alfi_name_len;
+
+ if (xfs_attr_log_item_op(old_attrp) == XFS_ATTRI_OP_FLAGS_NVREPLACE) {
+ new_attrp->alfi_new_name_len = old_attrp->alfi_new_name_len;
+ new_attrp->alfi_old_name_len = old_attrp->alfi_old_name_len;
+ new_attrp->alfi_new_value_len = old_attrp->alfi_new_value_len;
+ } else {
+ new_attrp->alfi_name_len = old_attrp->alfi_name_len;
+ }
+
new_attrp->alfi_attr_filter = old_attrp->alfi_attr_filter;
return &new_attrip->attri_item;
@@ -719,9 +808,13 @@ xlog_recover_attri_commit_pass2(
struct xfs_attri_log_nameval *nv;
const void *attr_name;
const void *attr_value = NULL;
+ const void *attr_new_name = NULL;
+ const void *attr_new_value = NULL;
size_t len;
unsigned int name_len = 0;
unsigned int value_len = 0;
+ unsigned int new_name_len = 0;
+ unsigned int new_value_len = 0;
unsigned int op, i = 0;
/* Validate xfs_attri_log_format before the large memory allocation */
@@ -773,6 +866,21 @@ xlog_recover_attri_commit_pass2(
}
name_len = attri_formatp->alfi_name_len;
break;
+ case XFS_ATTRI_OP_FLAGS_NVREPLACE:
+ /*
+ * Log item, attr name, new attr name, optional attr value,
+ * optional new attr value
+ */
+ if (item->ri_total < 3 || item->ri_total > 5) {
+ XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp,
+ attri_formatp, len);
+ return -EFSCORRUPTED;
+ }
+ name_len = attri_formatp->alfi_old_name_len;
+ new_name_len = attri_formatp->alfi_new_name_len;
+ value_len = attri_formatp->alfi_value_len;
+ new_value_len = attri_formatp->alfi_new_value_len;
+ break;
default:
XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp,
attri_formatp, len);
@@ -795,12 +903,30 @@ xlog_recover_attri_commit_pass2(
}
i++;
+ /* Validate the new attr name */
+ if (new_name_len > 0) {
+ if (item->ri_buf[i].i_len != xlog_calc_iovec_len(new_name_len)) {
+ XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp,
+ item->ri_buf[i].i_addr,
+ item->ri_buf[i].i_len);
+ return -EFSCORRUPTED;
+ }
+
+ attr_new_name = item->ri_buf[i].i_addr;
+ if (!xfs_attr_namecheck(attr_new_name, new_name_len)) {
+ XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp,
+ item->ri_buf[i].i_addr,
+ item->ri_buf[i].i_len);
+ return -EFSCORRUPTED;
+ }
+ i++;
+ }
+
/* Validate the attr value, if present */
if (value_len != 0) {
if (item->ri_buf[i].i_len != xlog_calc_iovec_len(value_len)) {
XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp,
- item->ri_buf[0].i_addr,
- item->ri_buf[0].i_len);
+ attri_formatp, len);
return -EFSCORRUPTED;
}
@@ -808,6 +934,18 @@ xlog_recover_attri_commit_pass2(
i++;
}
+ /* Validate the new attr value, if present */
+ if (new_value_len != 0) {
+ if (item->ri_buf[i].i_len != xlog_calc_iovec_len(new_value_len)) {
+ XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp,
+ attri_formatp, len);
+ return -EFSCORRUPTED;
+ }
+
+ attr_new_value = item->ri_buf[i].i_addr;
+ i++;
+ }
+
/*
* Make sure we got the correct number of buffers for the operation
* that we just loaded.
@@ -845,6 +983,23 @@ xlog_recover_attri_commit_pass2(
return -EFSCORRUPTED;
}
break;
+ case XFS_ATTRI_OP_FLAGS_NVREPLACE:
+ /*
+ * Name-value replace operations require the caller to
+ * specify the old and new names and values explicitly.
+ * Values are optional.
+ */
+ if (attr_name == NULL || name_len == 0) {
+ XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp,
+ attri_formatp, len);
+ return -EFSCORRUPTED;
+ }
+ if (attr_new_name == NULL || new_name_len == 0) {
+ XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp,
+ attri_formatp, len);
+ return -EFSCORRUPTED;
+ }
+ break;
}
/*
@@ -853,7 +1008,9 @@ xlog_recover_attri_commit_pass2(
* reference.
*/
nv = xfs_attri_log_nameval_alloc(attr_name, name_len,
- attr_value, value_len);
+ attr_new_name, new_name_len,
+ attr_value, value_len,
+ attr_new_value, new_value_len);
attrip = xfs_attri_init(mp, nv);
memcpy(&attrip->attri_format, attri_formatp, len);
diff --git a/fs/xfs/xfs_attr_item.h b/fs/xfs/xfs_attr_item.h
index 3280a79302876..9ae0b3696847b 100644
--- a/fs/xfs/xfs_attr_item.h
+++ b/fs/xfs/xfs_attr_item.h
@@ -13,7 +13,9 @@ struct kmem_zone;
struct xfs_attri_log_nameval {
struct xfs_log_iovec name;
+ struct xfs_log_iovec new_name; /* NVREPLACE only */
struct xfs_log_iovec value;
+ struct xfs_log_iovec new_value; /* NVREPLACE only */
refcount_t refcount;
/* name and value follow the end of this struct */
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 01/18] xfs: Expose init_xattrs in xfs_create_tmpfile
2023-12-31 19:33 ` [PATCHSET v13.0 4/7] xfs: Parent Pointers Darrick J. Wong
@ 2023-12-31 20:48 ` Darrick J. Wong
2023-12-31 20:49 ` [PATCH 02/18] xfs: add parent pointer support to attribute code Darrick J. Wong
` (16 subsequent siblings)
17 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:48 UTC (permalink / raw)
To: djwong; +Cc: Allison Henderson, catherine.hoang, allison.henderson, linux-xfs
From: Allison Henderson <allison.henderson@oracle.com>
Tmp files are used as part of rename operations and will need attr forks
initialized for parent pointers. Expose the init_xattrs parameter to
the calling function to initialize the fork.
Signed-off-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/xfs_inode.c | 5 +++--
fs/xfs/xfs_inode.h | 2 +-
fs/xfs/xfs_iops.c | 2 +-
3 files changed, 5 insertions(+), 4 deletions(-)
diff --git a/fs/xfs/xfs_inode.c b/fs/xfs/xfs_inode.c
index 88e0e93ded2e4..f1f8f85f941eb 100644
--- a/fs/xfs/xfs_inode.c
+++ b/fs/xfs/xfs_inode.c
@@ -1197,6 +1197,7 @@ xfs_create_tmpfile(
struct mnt_idmap *idmap,
struct xfs_inode *dp,
umode_t mode,
+ bool init_xattrs,
struct xfs_inode **ipp)
{
struct xfs_mount *mp = dp->i_mount;
@@ -1237,7 +1238,7 @@ xfs_create_tmpfile(
error = xfs_dialloc(&tp, dp->i_ino, mode, &ino);
if (!error)
error = xfs_init_new_inode(idmap, tp, dp, ino, mode,
- 0, 0, prid, false, &ip);
+ 0, 0, prid, init_xattrs, &ip);
if (error)
goto out_trans_cancel;
@@ -3038,7 +3039,7 @@ xfs_rename_alloc_whiteout(
int error;
error = xfs_create_tmpfile(idmap, dp, S_IFCHR | WHITEOUT_MODE,
- &tmpfile);
+ false, &tmpfile);
if (error)
return error;
diff --git a/fs/xfs/xfs_inode.h b/fs/xfs/xfs_inode.h
index 4826155ad9147..14c6d9e7ab975 100644
--- a/fs/xfs/xfs_inode.h
+++ b/fs/xfs/xfs_inode.h
@@ -518,7 +518,7 @@ int xfs_create(struct mnt_idmap *idmap,
umode_t mode, dev_t rdev, bool need_xattr,
struct xfs_inode **ipp);
int xfs_create_tmpfile(struct mnt_idmap *idmap,
- struct xfs_inode *dp, umode_t mode,
+ struct xfs_inode *dp, umode_t mode, bool init_xattrs,
struct xfs_inode **ipp);
int xfs_remove(struct xfs_inode *dp, struct xfs_name *name,
struct xfs_inode *ip);
diff --git a/fs/xfs/xfs_iops.c b/fs/xfs/xfs_iops.c
index 037606e5eee40..d9277f7faf534 100644
--- a/fs/xfs/xfs_iops.c
+++ b/fs/xfs/xfs_iops.c
@@ -202,7 +202,7 @@ xfs_generic_create(
xfs_create_need_xattr(dir, default_acl, acl),
&ip);
} else {
- error = xfs_create_tmpfile(idmap, XFS_I(dir), mode, &ip);
+ error = xfs_create_tmpfile(idmap, XFS_I(dir), mode, false, &ip);
}
if (unlikely(error))
goto out_free_acl;
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 02/18] xfs: add parent pointer support to attribute code
2023-12-31 19:33 ` [PATCHSET v13.0 4/7] xfs: Parent Pointers Darrick J. Wong
2023-12-31 20:48 ` [PATCH 01/18] xfs: Expose init_xattrs in xfs_create_tmpfile Darrick J. Wong
@ 2023-12-31 20:49 ` Darrick J. Wong
2023-12-31 20:49 ` [PATCH 03/18] xfs: define parent pointer ondisk extended attribute format Darrick J. Wong
` (15 subsequent siblings)
17 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:49 UTC (permalink / raw)
To: djwong
Cc: Mark Tinguely, Dave Chinner, Allison Henderson, catherine.hoang,
allison.henderson, linux-xfs
From: Allison Henderson <allison.henderson@oracle.com>
Add the new parent attribute type. XFS_ATTR_PARENT is used only for parent pointer
entries; it uses reserved blocks like XFS_ATTR_ROOT.
Signed-off-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/libxfs/xfs_attr.c | 3 ++-
fs/xfs/libxfs/xfs_da_format.h | 5 ++++-
fs/xfs/libxfs/xfs_log_format.h | 1 +
fs/xfs/scrub/attr.c | 2 +-
fs/xfs/xfs_trace.h | 3 ++-
5 files changed, 10 insertions(+), 4 deletions(-)
diff --git a/fs/xfs/libxfs/xfs_attr.c b/fs/xfs/libxfs/xfs_attr.c
index c6621af9554c2..579e63a1abf47 100644
--- a/fs/xfs/libxfs/xfs_attr.c
+++ b/fs/xfs/libxfs/xfs_attr.c
@@ -945,7 +945,8 @@ xfs_attr_set(
struct xfs_inode *dp = args->dp;
struct xfs_mount *mp = dp->i_mount;
struct xfs_trans_res tres;
- bool rsvd = (args->attr_filter & XFS_ATTR_ROOT);
+ bool rsvd = (args->attr_filter & (XFS_ATTR_ROOT |
+ XFS_ATTR_PARENT));
bool is_remove = args->op_flags & XFS_DA_OP_REMOVE;
int error, local;
int rmt_blks = 0;
diff --git a/fs/xfs/libxfs/xfs_da_format.h b/fs/xfs/libxfs/xfs_da_format.h
index 0e1ada44f21ba..6b5971ee6804c 100644
--- a/fs/xfs/libxfs/xfs_da_format.h
+++ b/fs/xfs/libxfs/xfs_da_format.h
@@ -709,12 +709,15 @@ struct xfs_attr3_leafblock {
#define XFS_ATTR_LOCAL_BIT 0 /* attr is stored locally */
#define XFS_ATTR_ROOT_BIT 1 /* limit access to trusted attrs */
#define XFS_ATTR_SECURE_BIT 2 /* limit access to secure attrs */
+#define XFS_ATTR_PARENT_BIT 3 /* parent pointer attrs */
#define XFS_ATTR_INCOMPLETE_BIT 7 /* attr in middle of create/delete */
#define XFS_ATTR_LOCAL (1u << XFS_ATTR_LOCAL_BIT)
#define XFS_ATTR_ROOT (1u << XFS_ATTR_ROOT_BIT)
#define XFS_ATTR_SECURE (1u << XFS_ATTR_SECURE_BIT)
+#define XFS_ATTR_PARENT (1u << XFS_ATTR_PARENT_BIT)
#define XFS_ATTR_INCOMPLETE (1u << XFS_ATTR_INCOMPLETE_BIT)
-#define XFS_ATTR_NSP_ONDISK_MASK (XFS_ATTR_ROOT | XFS_ATTR_SECURE)
+#define XFS_ATTR_NSP_ONDISK_MASK \
+ (XFS_ATTR_ROOT | XFS_ATTR_SECURE | XFS_ATTR_PARENT)
#define XFS_ATTR_NAMESPACE_STR \
{ XFS_ATTR_LOCAL, "local" }, \
diff --git a/fs/xfs/libxfs/xfs_log_format.h b/fs/xfs/libxfs/xfs_log_format.h
index 285a0a089df24..bded03634e53d 100644
--- a/fs/xfs/libxfs/xfs_log_format.h
+++ b/fs/xfs/libxfs/xfs_log_format.h
@@ -1056,6 +1056,7 @@ struct xfs_icreate_log {
*/
#define XFS_ATTRI_FILTER_MASK (XFS_ATTR_ROOT | \
XFS_ATTR_SECURE | \
+ XFS_ATTR_PARENT | \
XFS_ATTR_INCOMPLETE)
/*
diff --git a/fs/xfs/scrub/attr.c b/fs/xfs/scrub/attr.c
index ff83051c79818..247517f7e69bb 100644
--- a/fs/xfs/scrub/attr.c
+++ b/fs/xfs/scrub/attr.c
@@ -513,7 +513,7 @@ xchk_xattr_rec(
/* Retrieve the entry and check it. */
hash = be32_to_cpu(ent->hashval);
badflags = ~(XFS_ATTR_LOCAL | XFS_ATTR_ROOT | XFS_ATTR_SECURE |
- XFS_ATTR_INCOMPLETE);
+ XFS_ATTR_INCOMPLETE | XFS_ATTR_PARENT);
if ((ent->flags & badflags) != 0)
xchk_da_set_corrupt(ds, level);
if (ent->flags & XFS_ATTR_LOCAL) {
diff --git a/fs/xfs/xfs_trace.h b/fs/xfs/xfs_trace.h
index 2c838b7471191..1dac853aa659e 100644
--- a/fs/xfs/xfs_trace.h
+++ b/fs/xfs/xfs_trace.h
@@ -87,7 +87,8 @@ struct xfs_swapext_req;
#define XFS_ATTR_FILTER_FLAGS \
{ XFS_ATTR_ROOT, "ROOT" }, \
{ XFS_ATTR_SECURE, "SECURE" }, \
- { XFS_ATTR_INCOMPLETE, "INCOMPLETE" }
+ { XFS_ATTR_INCOMPLETE, "INCOMPLETE" }, \
+ { XFS_ATTR_PARENT, "PARENT" }
DECLARE_EVENT_CLASS(xfs_attr_list_class,
TP_PROTO(struct xfs_attr_list_context *ctx),
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 03/18] xfs: define parent pointer ondisk extended attribute format
2023-12-31 19:33 ` [PATCHSET v13.0 4/7] xfs: Parent Pointers Darrick J. Wong
2023-12-31 20:48 ` [PATCH 01/18] xfs: Expose init_xattrs in xfs_create_tmpfile Darrick J. Wong
2023-12-31 20:49 ` [PATCH 02/18] xfs: add parent pointer support to attribute code Darrick J. Wong
@ 2023-12-31 20:49 ` Darrick J. Wong
2023-12-31 20:49 ` [PATCH 04/18] xfs: add parent pointer validator functions Darrick J. Wong
` (14 subsequent siblings)
17 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:49 UTC (permalink / raw)
To: djwong
Cc: Dave Chinner, Allison Henderson, catherine.hoang,
allison.henderson, linux-xfs
From: Allison Henderson <allison.henderson@oracle.com>
We need to define the parent pointer attribute format before we start
adding support for it into all the code that needs to use it. The EA
format we will use encodes the following information:
name={parent inode #, parent inode generation, dirent namehash}
value={dirent name}
The inode/gen gives all the information we need to reliably identify the
parent without requiring child->parent lock ordering, and allows
userspace to do pathname component level reconstruction without the
kernel ever needing to verify the parent itself as part of ioctl calls.
Storing the dirent name hash in the key reduces hash collisions if a
file is hardlinked multiple times in the same directory.
By using the NVLOOKUP mode in the extended attribute code to match
parent pointers using both the xattr name and value, we can identify the
exact parent pointer EA we need to modify/remove in rename/unlink
operations without searching the entire EA space.
By storing the dirent name, we have enough information to be able to
validate and reconstruct damaged directory trees. Earlier iterations of
this patchset encoded the directory offset in the parent pointer key,
but this format required repair to keep that in sync across directory
rebuilds, which is unnecessary complexity.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
[djwong: replace diroffset with the namehash in the pptr key]
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/libxfs/xfs_da_format.h | 20 ++++++++++++++++++++
1 file changed, 20 insertions(+)
diff --git a/fs/xfs/libxfs/xfs_da_format.h b/fs/xfs/libxfs/xfs_da_format.h
index 6b5971ee6804c..edbd901d90659 100644
--- a/fs/xfs/libxfs/xfs_da_format.h
+++ b/fs/xfs/libxfs/xfs_da_format.h
@@ -878,4 +878,24 @@ static inline unsigned int xfs_dir2_dirblock_bytes(struct xfs_sb *sbp)
xfs_failaddr_t xfs_da3_blkinfo_verify(struct xfs_buf *bp,
struct xfs_da3_blkinfo *hdr3);
+/*
+ * Parent pointer attribute format definition
+ *
+ * The xattr name encodes the parent inode number, generation and the crc32c
+ * hash of the dirent name.
+ *
+ * The xattr value contains the dirent name.
+ */
+struct xfs_parent_name_rec {
+ __be64 p_ino;
+ __be32 p_gen;
+ __be32 p_namehash;
+};
+
+/*
+ * Maximum size of the dirent name that can be stored in a parent pointer.
+ * This matches the maximum dirent name length.
+ */
+#define XFS_PARENT_DIRENT_NAME_MAX_SIZE (MAXNAMELEN - 1)
+
#endif /* __XFS_DA_FORMAT_H__ */
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 04/18] xfs: add parent pointer validator functions
2023-12-31 19:33 ` [PATCHSET v13.0 4/7] xfs: Parent Pointers Darrick J. Wong
` (2 preceding siblings ...)
2023-12-31 20:49 ` [PATCH 03/18] xfs: define parent pointer ondisk extended attribute format Darrick J. Wong
@ 2023-12-31 20:49 ` Darrick J. Wong
2023-12-31 20:49 ` [PATCH 05/18] xfs: extend transaction reservations for parent attributes Darrick J. Wong
` (13 subsequent siblings)
17 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:49 UTC (permalink / raw)
To: djwong; +Cc: Allison Henderson, catherine.hoang, allison.henderson, linux-xfs
From: Allison Henderson <allison.henderson@oracle.com>
Attribute names of parent pointers are not strings. So we need to
modify attr_namecheck to verify parent pointer records when the
XFS_ATTR_PARENT flag is set. At the same time, we need to validate attr
values during log recovery if the xattr is really a parent pointer.
Signed-off-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
[djwong: move functions to xfs_parent.c, adjust for new disk format]
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/Makefile | 1
fs/xfs/libxfs/xfs_attr.c | 10 +++-
fs/xfs/libxfs/xfs_attr.h | 3 +
fs/xfs/libxfs/xfs_da_format.h | 8 +++
fs/xfs/libxfs/xfs_parent.c | 113 +++++++++++++++++++++++++++++++++++++++++
fs/xfs/libxfs/xfs_parent.h | 19 +++++++
fs/xfs/scrub/attr.c | 2 -
fs/xfs/xfs_attr_item.c | 42 ++++++++++++++-
fs/xfs/xfs_attr_list.c | 17 ++++--
9 files changed, 203 insertions(+), 12 deletions(-)
create mode 100644 fs/xfs/libxfs/xfs_parent.c
create mode 100644 fs/xfs/libxfs/xfs_parent.h
diff --git a/fs/xfs/Makefile b/fs/xfs/Makefile
index 09016465d4925..64ff5295d3fc9 100644
--- a/fs/xfs/Makefile
+++ b/fs/xfs/Makefile
@@ -41,6 +41,7 @@ xfs-y += $(addprefix libxfs/, \
xfs_inode_buf.o \
xfs_log_rlimit.o \
xfs_ag_resv.o \
+ xfs_parent.o \
xfs_rmap.o \
xfs_rmap_btree.o \
xfs_refcount.o \
diff --git a/fs/xfs/libxfs/xfs_attr.c b/fs/xfs/libxfs/xfs_attr.c
index 579e63a1abf47..d513ed57177e1 100644
--- a/fs/xfs/libxfs/xfs_attr.c
+++ b/fs/xfs/libxfs/xfs_attr.c
@@ -26,6 +26,7 @@
#include "xfs_trace.h"
#include "xfs_attr_item.h"
#include "xfs_xattr.h"
+#include "xfs_parent.h"
struct kmem_cache *xfs_attr_intent_cache;
@@ -1545,9 +1546,14 @@ xfs_attr_node_get(
/* Returns true if the attribute entry name is valid. */
bool
xfs_attr_namecheck(
- const void *name,
- size_t length)
+ struct xfs_mount *mp,
+ const void *name,
+ size_t length,
+ unsigned int flags)
{
+ if (flags & XFS_ATTR_PARENT)
+ return xfs_parent_namecheck(mp, name, length, flags);
+
/*
* MAXNAMELEN includes the trailing null, but (name/length) leave it
* out, so use >= for the length check.
diff --git a/fs/xfs/libxfs/xfs_attr.h b/fs/xfs/libxfs/xfs_attr.h
index b4e8ecee3e039..5b3a0d4b1583c 100644
--- a/fs/xfs/libxfs/xfs_attr.h
+++ b/fs/xfs/libxfs/xfs_attr.h
@@ -552,7 +552,8 @@ int xfs_attr_get(struct xfs_da_args *args);
int xfs_attr_set(struct xfs_da_args *args);
int xfs_attr_set_iter(struct xfs_attr_intent *attr);
int xfs_attr_remove_iter(struct xfs_attr_intent *attr);
-bool xfs_attr_namecheck(const void *name, size_t length);
+bool xfs_attr_namecheck(struct xfs_mount *mp, const void *name, size_t length,
+ unsigned int flags);
int xfs_attr_calc_size(struct xfs_da_args *args, int *local);
void xfs_init_attr_trans(struct xfs_da_args *args, struct xfs_trans_res *tres,
unsigned int *total);
diff --git a/fs/xfs/libxfs/xfs_da_format.h b/fs/xfs/libxfs/xfs_da_format.h
index edbd901d90659..a63387796a246 100644
--- a/fs/xfs/libxfs/xfs_da_format.h
+++ b/fs/xfs/libxfs/xfs_da_format.h
@@ -757,6 +757,14 @@ xfs_attr3_leaf_name(xfs_attr_leafblock_t *leafp, int idx)
return &((char *)leafp)[be16_to_cpu(entries[idx].nameidx)];
}
+static inline int
+xfs_attr3_leaf_flags(xfs_attr_leafblock_t *leafp, int idx)
+{
+ struct xfs_attr_leaf_entry *entries = xfs_attr3_leaf_entryp(leafp);
+
+ return entries[idx].flags;
+}
+
static inline xfs_attr_leaf_name_remote_t *
xfs_attr3_leaf_name_remote(xfs_attr_leafblock_t *leafp, int idx)
{
diff --git a/fs/xfs/libxfs/xfs_parent.c b/fs/xfs/libxfs/xfs_parent.c
new file mode 100644
index 0000000000000..1d45f926c13a6
--- /dev/null
+++ b/fs/xfs/libxfs/xfs_parent.c
@@ -0,0 +1,113 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Copyright (c) 2022-2024 Oracle.
+ * All rights reserved.
+ */
+#include "xfs.h"
+#include "xfs_fs.h"
+#include "xfs_format.h"
+#include "xfs_da_format.h"
+#include "xfs_log_format.h"
+#include "xfs_shared.h"
+#include "xfs_trans_resv.h"
+#include "xfs_mount.h"
+#include "xfs_bmap_btree.h"
+#include "xfs_inode.h"
+#include "xfs_error.h"
+#include "xfs_trace.h"
+#include "xfs_trans.h"
+#include "xfs_da_btree.h"
+#include "xfs_attr.h"
+#include "xfs_dir2.h"
+#include "xfs_dir2_priv.h"
+#include "xfs_attr_sf.h"
+#include "xfs_bmap.h"
+#include "xfs_defer.h"
+#include "xfs_log.h"
+#include "xfs_xattr.h"
+#include "xfs_parent.h"
+#include "xfs_trans_space.h"
+
+/*
+ * Parent pointer attribute handling.
+ *
+ * Because the attribute value is a filename component, it will never be longer
+ * than 255 bytes. This means the attribute will always be a local format
+ * attribute as it is xfs_attr_leaf_entsize_local_max() for v5 filesystems will
+ * always be larger than this (max is 75% of block size).
+ *
+ * Creating a new parent attribute will always create a new attribute - there
+ * should never, ever be an existing attribute in the tree for a new inode.
+ * ENOSPC behavior is problematic - creating the inode without the parent
+ * pointer is effectively a corruption, so we allow parent attribute creation
+ * to dip into the reserve block pool to avoid unexpected ENOSPC errors from
+ * occurring.
+ */
+
+/* Return true if parent pointer EA name is valid. */
+bool
+xfs_parent_namecheck(
+ struct xfs_mount *mp,
+ const struct xfs_parent_name_rec *rec,
+ size_t reclen,
+ unsigned int attr_flags)
+{
+ if (!(attr_flags & XFS_ATTR_PARENT))
+ return false;
+
+ /* pptr updates use logged xattrs, so we should never see this flag */
+ if (attr_flags & XFS_ATTR_INCOMPLETE)
+ return false;
+
+ if (reclen != sizeof(struct xfs_parent_name_rec))
+ return false;
+
+ /* Only one namespace bit allowed. */
+ if (hweight32(attr_flags & XFS_ATTR_NSP_ONDISK_MASK) > 1)
+ return false;
+
+ return true;
+}
+
+/* Return true if parent pointer EA value is valid. */
+bool
+xfs_parent_valuecheck(
+ struct xfs_mount *mp,
+ const void *value,
+ size_t valuelen)
+{
+ if (valuelen == 0 || valuelen > XFS_PARENT_DIRENT_NAME_MAX_SIZE)
+ return false;
+
+ if (value == NULL)
+ return false;
+
+ return true;
+}
+
+/* Return true if the ondisk parent pointer is consistent. */
+bool
+xfs_parent_hashcheck(
+ struct xfs_mount *mp,
+ const struct xfs_parent_name_rec *rec,
+ const void *value,
+ size_t valuelen)
+{
+ struct xfs_name dname = {
+ .name = value,
+ .len = valuelen,
+ };
+ xfs_ino_t p_ino;
+
+ /* Valid dirent name? */
+ if (!xfs_dir2_namecheck(value, valuelen))
+ return false;
+
+ /* Valid inode number? */
+ p_ino = be64_to_cpu(rec->p_ino);
+ if (!xfs_verify_dir_ino(mp, p_ino))
+ return false;
+
+ /* Namehash matches name? */
+ return be32_to_cpu(rec->p_namehash) == xfs_dir2_hashname(mp, &dname);
+}
diff --git a/fs/xfs/libxfs/xfs_parent.h b/fs/xfs/libxfs/xfs_parent.h
new file mode 100644
index 0000000000000..fcfeddb645f6d
--- /dev/null
+++ b/fs/xfs/libxfs/xfs_parent.h
@@ -0,0 +1,19 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Copyright (c) 2022-2024 Oracle.
+ * All Rights Reserved.
+ */
+#ifndef __XFS_PARENT_H__
+#define __XFS_PARENT_H__
+
+/* Metadata validators */
+bool xfs_parent_namecheck(struct xfs_mount *mp,
+ const struct xfs_parent_name_rec *rec, size_t reclen,
+ unsigned int attr_flags);
+bool xfs_parent_valuecheck(struct xfs_mount *mp, const void *value,
+ size_t valuelen);
+bool xfs_parent_hashcheck(struct xfs_mount *mp,
+ const struct xfs_parent_name_rec *rec, const void *value,
+ size_t valuelen);
+
+#endif /* __XFS_PARENT_H__ */
diff --git a/fs/xfs/scrub/attr.c b/fs/xfs/scrub/attr.c
index 247517f7e69bb..f213d745746fd 100644
--- a/fs/xfs/scrub/attr.c
+++ b/fs/xfs/scrub/attr.c
@@ -206,7 +206,7 @@ xchk_xattr_actor(
}
/* Does this name make sense? */
- if (!xfs_attr_namecheck(name, namelen)) {
+ if (!xfs_attr_namecheck(sc->mp, name, namelen, attr_flags)) {
xchk_fblock_set_corrupt(sc, XFS_ATTR_FORK, args.blkno);
return -ECANCELED;
}
diff --git a/fs/xfs/xfs_attr_item.c b/fs/xfs/xfs_attr_item.c
index a1c8036310b21..96b8f58bc5770 100644
--- a/fs/xfs/xfs_attr_item.c
+++ b/fs/xfs/xfs_attr_item.c
@@ -27,6 +27,7 @@
#include "xfs_error.h"
#include "xfs_log_priv.h"
#include "xfs_log_recover.h"
+#include "xfs_parent.h"
struct kmem_cache *xfs_attri_cache;
struct kmem_cache *xfs_attrd_cache;
@@ -687,7 +688,8 @@ xfs_attr_recover_work(
*/
attrp = &attrip->attri_format;
if (!xfs_attri_validate(mp, attrp) ||
- !xfs_attr_namecheck(nv->name.i_addr, nv->name.i_len))
+ !xfs_attr_namecheck(mp, nv->name.i_addr, nv->name.i_len,
+ attrp->alfi_attr_filter))
return -EFSCORRUPTED;
attr = xfs_attri_recover_work(mp, dfp, attrp, &ip, nv);
@@ -896,7 +898,8 @@ xlog_recover_attri_commit_pass2(
}
attr_name = item->ri_buf[i].i_addr;
- if (!xfs_attr_namecheck(attr_name, name_len)) {
+ if (!xfs_attr_namecheck(mp, attr_name, name_len,
+ attri_formatp->alfi_attr_filter)) {
XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp,
attri_formatp, len);
return -EFSCORRUPTED;
@@ -913,7 +916,8 @@ xlog_recover_attri_commit_pass2(
}
attr_new_name = item->ri_buf[i].i_addr;
- if (!xfs_attr_namecheck(attr_new_name, new_name_len)) {
+ if (!xfs_attr_namecheck(mp, attr_new_name, new_name_len,
+ attri_formatp->alfi_attr_filter)) {
XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp,
item->ri_buf[i].i_addr,
item->ri_buf[i].i_len);
@@ -931,6 +935,22 @@ xlog_recover_attri_commit_pass2(
}
attr_value = item->ri_buf[i].i_addr;
+ if ((attri_formatp->alfi_attr_filter & XFS_ATTR_PARENT) &&
+ !xfs_parent_valuecheck(mp, attr_value, value_len)) {
+ XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp,
+ item->ri_buf[i].i_addr,
+ item->ri_buf[i].i_len);
+ return -EFSCORRUPTED;
+ }
+ if ((attri_formatp->alfi_attr_filter & XFS_ATTR_PARENT) &&
+ (attr_name == NULL ||
+ !xfs_parent_hashcheck(mp, attr_name, attr_value,
+ value_len))) {
+ XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp,
+ item->ri_buf[i].i_addr,
+ item->ri_buf[i].i_len);
+ return -EFSCORRUPTED;
+ }
i++;
}
@@ -943,6 +963,22 @@ xlog_recover_attri_commit_pass2(
}
attr_new_value = item->ri_buf[i].i_addr;
+ if ((attri_formatp->alfi_attr_filter & XFS_ATTR_PARENT) &&
+ !xfs_parent_valuecheck(mp, attr_new_value, new_value_len)) {
+ XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp,
+ item->ri_buf[i].i_addr,
+ item->ri_buf[i].i_len);
+ return -EFSCORRUPTED;
+ }
+ if ((attri_formatp->alfi_attr_filter & XFS_ATTR_PARENT) &&
+ (attr_new_name == NULL ||
+ !xfs_parent_hashcheck(mp, attr_new_name, attr_new_value,
+ new_value_len))) {
+ XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp,
+ item->ri_buf[i].i_addr,
+ item->ri_buf[i].i_len);
+ return -EFSCORRUPTED;
+ }
i++;
}
diff --git a/fs/xfs/xfs_attr_list.c b/fs/xfs/xfs_attr_list.c
index 24516f3ff2df7..c86e5952c1378 100644
--- a/fs/xfs/xfs_attr_list.c
+++ b/fs/xfs/xfs_attr_list.c
@@ -59,9 +59,13 @@ xfs_attr_shortform_list(
struct xfs_attr_sf_sort *sbuf, *sbp;
struct xfs_attr_shortform *sf;
struct xfs_attr_sf_entry *sfe;
+ struct xfs_mount *mp;
int sbsize, nsbuf, count, i;
int error = 0;
+ ASSERT(context != NULL);
+ ASSERT(dp != NULL);
+ mp = dp->i_mount;
sf = (struct xfs_attr_shortform *)dp->i_af.if_u1.if_data;
ASSERT(sf != NULL);
if (!sf->hdr.count)
@@ -83,8 +87,9 @@ xfs_attr_shortform_list(
(dp->i_af.if_bytes + sf->hdr.count * 16) < context->bufsize)) {
for (i = 0, sfe = &sf->list[0]; i < sf->hdr.count; i++) {
if (XFS_IS_CORRUPT(context->dp->i_mount,
- !xfs_attr_namecheck(sfe->nameval,
- sfe->namelen))) {
+ !xfs_attr_namecheck(mp, sfe->nameval,
+ sfe->namelen,
+ sfe->flags))) {
xfs_dirattr_mark_sick(context->dp, XFS_ATTR_FORK);
return -EFSCORRUPTED;
}
@@ -178,8 +183,9 @@ xfs_attr_shortform_list(
cursor->offset = 0;
}
if (XFS_IS_CORRUPT(context->dp->i_mount,
- !xfs_attr_namecheck(sbp->name,
- sbp->namelen))) {
+ !xfs_attr_namecheck(mp, sbp->name,
+ sbp->namelen,
+ sbp->flags))) {
xfs_dirattr_mark_sick(context->dp, XFS_ATTR_FORK);
error = -EFSCORRUPTED;
goto out;
@@ -503,7 +509,8 @@ xfs_attr3_leaf_list_int(
}
if (XFS_IS_CORRUPT(context->dp->i_mount,
- !xfs_attr_namecheck(name, namelen))) {
+ !xfs_attr_namecheck(mp, name, namelen,
+ entry->flags))) {
xfs_dirattr_mark_sick(context->dp, XFS_ATTR_FORK);
return -EFSCORRUPTED;
}
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 05/18] xfs: extend transaction reservations for parent attributes
2023-12-31 19:33 ` [PATCHSET v13.0 4/7] xfs: Parent Pointers Darrick J. Wong
` (3 preceding siblings ...)
2023-12-31 20:49 ` [PATCH 04/18] xfs: add parent pointer validator functions Darrick J. Wong
@ 2023-12-31 20:49 ` Darrick J. Wong
2023-12-31 20:50 ` [PATCH 06/18] xfs: parent pointer attribute creation Darrick J. Wong
` (12 subsequent siblings)
17 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:49 UTC (permalink / raw)
To: djwong
Cc: Dave Chinner, Allison Henderson, catherine.hoang,
allison.henderson, linux-xfs
From: Allison Henderson <allison.henderson@oracle.com>
We need to add, remove or modify parent pointer attributes during
create/link/unlink/rename operations atomically with the dirents in the
parent directories being modified. This means they need to be modified
in the same transaction as the parent directories, and so we need to add
the required space for the attribute modifications to the transaction
reservations.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
[djwong: fix indenting errors, adjust for new log format]
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/libxfs/xfs_trans_resv.c | 326 ++++++++++++++++++++++++++++++++++------
1 file changed, 274 insertions(+), 52 deletions(-)
diff --git a/fs/xfs/libxfs/xfs_trans_resv.c b/fs/xfs/libxfs/xfs_trans_resv.c
index 6cd45e8c118da..b390d9aa02142 100644
--- a/fs/xfs/libxfs/xfs_trans_resv.c
+++ b/fs/xfs/libxfs/xfs_trans_resv.c
@@ -20,6 +20,9 @@
#include "xfs_qm.h"
#include "xfs_trans_space.h"
#include "xfs_rtbitmap.h"
+#include "xfs_attr_item.h"
+#include "xfs_log.h"
+#include "xfs_da_format.h"
#define _ALLOC true
#define _FREE false
@@ -422,29 +425,110 @@ xfs_calc_itruncate_reservation_minlogsize(
return xfs_calc_itruncate_reservation(mp, true);
}
+static inline unsigned int xfs_calc_pptr_link_overhead(void)
+{
+ return sizeof(struct xfs_attri_log_format) +
+ xlog_calc_iovec_len(sizeof(struct xfs_parent_name_rec)) +
+ xlog_calc_iovec_len(XFS_PARENT_DIRENT_NAME_MAX_SIZE);
+}
+static inline unsigned int xfs_calc_pptr_unlink_overhead(void)
+{
+ return sizeof(struct xfs_attri_log_format) +
+ xlog_calc_iovec_len(sizeof(struct xfs_parent_name_rec)) +
+ xlog_calc_iovec_len(XFS_PARENT_DIRENT_NAME_MAX_SIZE);
+}
+static inline unsigned int xfs_calc_pptr_replace_overhead(void)
+{
+ return sizeof(struct xfs_attri_log_format) +
+ xlog_calc_iovec_len(sizeof(struct xfs_parent_name_rec)) +
+ xlog_calc_iovec_len(XFS_PARENT_DIRENT_NAME_MAX_SIZE) +
+ xlog_calc_iovec_len(sizeof(struct xfs_parent_name_rec)) +
+ xlog_calc_iovec_len(XFS_PARENT_DIRENT_NAME_MAX_SIZE);
+}
+
/*
* In renaming a files we can modify:
* the five inodes involved: 5 * inode size
* the two directory btrees: 2 * (max depth + v2) * dir block size
* the two directory bmap btrees: 2 * max depth * block size
* And the bmap_finish transaction can free dir and bmap blocks (two sets
- * of bmap blocks) giving:
+ * of bmap blocks) giving (t2):
* the agf for the ags in which the blocks live: 3 * sector size
* the agfl for the ags in which the blocks live: 3 * sector size
* the superblock for the free block count: sector size
* the allocation btrees: 3 exts * 2 trees * (2 * max depth - 1) * block size
+ * If parent pointers are enabled (t3), then each transaction in the chain
+ * must be capable of setting or removing the extended attribute
+ * containing the parent information. It must also be able to handle
+ * the three xattr intent items that track the progress of the parent
+ * pointer update.
*/
STATIC uint
xfs_calc_rename_reservation(
struct xfs_mount *mp)
{
- return XFS_DQUOT_LOGRES(mp) +
- max((xfs_calc_inode_res(mp, 5) +
- xfs_calc_buf_res(2 * XFS_DIROP_LOG_COUNT(mp),
- XFS_FSB_TO_B(mp, 1))),
- (xfs_calc_buf_res(7, mp->m_sb.sb_sectsize) +
- xfs_calc_buf_res(xfs_allocfree_block_count(mp, 3),
- XFS_FSB_TO_B(mp, 1))));
+ unsigned int overhead = XFS_DQUOT_LOGRES(mp);
+ struct xfs_trans_resv *resp = M_RES(mp);
+ unsigned int t1, t2, t3 = 0;
+
+ t1 = xfs_calc_inode_res(mp, 5) +
+ xfs_calc_buf_res(2 * XFS_DIROP_LOG_COUNT(mp),
+ XFS_FSB_TO_B(mp, 1));
+
+ t2 = xfs_calc_buf_res(7, mp->m_sb.sb_sectsize) +
+ xfs_calc_buf_res(xfs_allocfree_block_count(mp, 3),
+ XFS_FSB_TO_B(mp, 1));
+
+ if (xfs_has_parent(mp)) {
+ unsigned int rename_overhead, exchange_overhead;
+
+ t3 = max(resp->tr_attrsetm.tr_logres,
+ resp->tr_attrrm.tr_logres);
+
+ /*
+ * For a standard rename, the three xattr intent log items
+ * are (1) replacing the pptr for the source file; (2)
+ * removing the pptr on the dest file; and (3) adding a
+ * pptr for the whiteout file in the src dir.
+ *
+ * For an RENAME_EXCHANGE, there are two xattr intent
+ * items to replace the pptr for both src and dest
+ * files. Link counts don't change and there is no
+ * whiteout.
+ *
+ * In the worst case we can end up relogging all log
+ * intent items to allow the log tail to move ahead, so
+ * they become overhead added to each transaction in a
+ * processing chain.
+ */
+ rename_overhead = xfs_calc_pptr_replace_overhead() +
+ xfs_calc_pptr_unlink_overhead() +
+ xfs_calc_pptr_link_overhead();
+ exchange_overhead = 2 * xfs_calc_pptr_replace_overhead();
+
+ overhead += max(rename_overhead, exchange_overhead);
+ }
+
+ return overhead + max3(t1, t2, t3);
+}
+
+static inline unsigned int
+xfs_rename_log_count(
+ struct xfs_mount *mp,
+ struct xfs_trans_resv *resp)
+{
+ /* One for the rename, one more for freeing blocks */
+ unsigned int ret = XFS_RENAME_LOG_COUNT;
+
+ /*
+ * Pre-reserve enough log reservation to handle the transaction
+ * rolling needed to remove or add one parent pointer.
+ */
+ if (xfs_has_parent(mp))
+ ret += max(resp->tr_attrsetm.tr_logcount,
+ resp->tr_attrrm.tr_logcount);
+
+ return ret;
}
/*
@@ -461,6 +545,23 @@ xfs_calc_iunlink_remove_reservation(
2 * M_IGEO(mp)->inode_cluster_size;
}
+static inline unsigned int
+xfs_link_log_count(
+ struct xfs_mount *mp,
+ struct xfs_trans_resv *resp)
+{
+ unsigned int ret = XFS_LINK_LOG_COUNT;
+
+ /*
+ * Pre-reserve enough log reservation to handle the transaction
+ * rolling needed to add one parent pointer.
+ */
+ if (xfs_has_parent(mp))
+ ret += resp->tr_attrsetm.tr_logcount;
+
+ return ret;
+}
+
/*
* For creating a link to an inode:
* the parent directory inode: inode size
@@ -477,14 +578,23 @@ STATIC uint
xfs_calc_link_reservation(
struct xfs_mount *mp)
{
- return XFS_DQUOT_LOGRES(mp) +
- xfs_calc_iunlink_remove_reservation(mp) +
- max((xfs_calc_inode_res(mp, 2) +
- xfs_calc_buf_res(XFS_DIROP_LOG_COUNT(mp),
- XFS_FSB_TO_B(mp, 1))),
- (xfs_calc_buf_res(3, mp->m_sb.sb_sectsize) +
- xfs_calc_buf_res(xfs_allocfree_block_count(mp, 1),
- XFS_FSB_TO_B(mp, 1))));
+ unsigned int overhead = XFS_DQUOT_LOGRES(mp);
+ struct xfs_trans_resv *resp = M_RES(mp);
+ unsigned int t1, t2, t3 = 0;
+
+ overhead += xfs_calc_iunlink_remove_reservation(mp);
+ t1 = xfs_calc_inode_res(mp, 2) +
+ xfs_calc_buf_res(XFS_DIROP_LOG_COUNT(mp), XFS_FSB_TO_B(mp, 1));
+ t2 = xfs_calc_buf_res(3, mp->m_sb.sb_sectsize) +
+ xfs_calc_buf_res(xfs_allocfree_block_count(mp, 1),
+ XFS_FSB_TO_B(mp, 1));
+
+ if (xfs_has_parent(mp)) {
+ t3 = resp->tr_attrsetm.tr_logres;
+ overhead += xfs_calc_pptr_link_overhead();
+ }
+
+ return overhead + max3(t1, t2, t3);
}
/*
@@ -499,6 +609,23 @@ xfs_calc_iunlink_add_reservation(xfs_mount_t *mp)
M_IGEO(mp)->inode_cluster_size;
}
+static inline unsigned int
+xfs_remove_log_count(
+ struct xfs_mount *mp,
+ struct xfs_trans_resv *resp)
+{
+ unsigned int ret = XFS_REMOVE_LOG_COUNT;
+
+ /*
+ * Pre-reserve enough log reservation to handle the transaction
+ * rolling needed to add one parent pointer.
+ */
+ if (xfs_has_parent(mp))
+ ret += resp->tr_attrrm.tr_logcount;
+
+ return ret;
+}
+
/*
* For removing a directory entry we can modify:
* the parent directory inode: inode size
@@ -515,14 +642,24 @@ STATIC uint
xfs_calc_remove_reservation(
struct xfs_mount *mp)
{
- return XFS_DQUOT_LOGRES(mp) +
- xfs_calc_iunlink_add_reservation(mp) +
- max((xfs_calc_inode_res(mp, 2) +
- xfs_calc_buf_res(XFS_DIROP_LOG_COUNT(mp),
- XFS_FSB_TO_B(mp, 1))),
- (xfs_calc_buf_res(4, mp->m_sb.sb_sectsize) +
- xfs_calc_buf_res(xfs_allocfree_block_count(mp, 2),
- XFS_FSB_TO_B(mp, 1))));
+ unsigned int overhead = XFS_DQUOT_LOGRES(mp);
+ struct xfs_trans_resv *resp = M_RES(mp);
+ unsigned int t1, t2, t3 = 0;
+
+ overhead += xfs_calc_iunlink_add_reservation(mp);
+
+ t1 = xfs_calc_inode_res(mp, 2) +
+ xfs_calc_buf_res(XFS_DIROP_LOG_COUNT(mp), XFS_FSB_TO_B(mp, 1));
+ t2 = xfs_calc_buf_res(4, mp->m_sb.sb_sectsize) +
+ xfs_calc_buf_res(xfs_allocfree_block_count(mp, 2),
+ XFS_FSB_TO_B(mp, 1));
+
+ if (xfs_has_parent(mp)) {
+ t3 = resp->tr_attrrm.tr_logres;
+ overhead += xfs_calc_pptr_unlink_overhead();
+ }
+
+ return overhead + max3(t1, t2, t3);
}
/*
@@ -571,12 +708,40 @@ xfs_calc_icreate_resv_alloc(
xfs_calc_finobt_res(mp);
}
+static inline unsigned int
+xfs_icreate_log_count(
+ struct xfs_mount *mp,
+ struct xfs_trans_resv *resp)
+{
+ unsigned int ret = XFS_CREATE_LOG_COUNT;
+
+ /*
+ * Pre-reserve enough log reservation to handle the transaction
+ * rolling needed to add one parent pointer.
+ */
+ if (xfs_has_parent(mp))
+ ret += resp->tr_attrsetm.tr_logcount;
+
+ return ret;
+}
+
STATIC uint
-xfs_calc_icreate_reservation(xfs_mount_t *mp)
+xfs_calc_icreate_reservation(
+ struct xfs_mount *mp)
{
- return XFS_DQUOT_LOGRES(mp) +
- max(xfs_calc_icreate_resv_alloc(mp),
- xfs_calc_create_resv_modify(mp));
+ struct xfs_trans_resv *resp = M_RES(mp);
+ unsigned int overhead = XFS_DQUOT_LOGRES(mp);
+ unsigned int t1, t2, t3 = 0;
+
+ t1 = xfs_calc_icreate_resv_alloc(mp);
+ t2 = xfs_calc_create_resv_modify(mp);
+
+ if (xfs_has_parent(mp)) {
+ t3 = resp->tr_attrsetm.tr_logres;
+ overhead += xfs_calc_pptr_link_overhead();
+ }
+
+ return overhead + max3(t1, t2, t3);
}
STATIC uint
@@ -589,6 +754,23 @@ xfs_calc_create_tmpfile_reservation(
return res + xfs_calc_iunlink_add_reservation(mp);
}
+static inline unsigned int
+xfs_mkdir_log_count(
+ struct xfs_mount *mp,
+ struct xfs_trans_resv *resp)
+{
+ unsigned int ret = XFS_MKDIR_LOG_COUNT;
+
+ /*
+ * Pre-reserve enough log reservation to handle the transaction
+ * rolling needed to add one parent pointer.
+ */
+ if (xfs_has_parent(mp))
+ ret += resp->tr_attrsetm.tr_logcount;
+
+ return ret;
+}
+
/*
* Making a new directory is the same as creating a new file.
*/
@@ -599,6 +781,22 @@ xfs_calc_mkdir_reservation(
return xfs_calc_icreate_reservation(mp);
}
+static inline unsigned int
+xfs_symlink_log_count(
+ struct xfs_mount *mp,
+ struct xfs_trans_resv *resp)
+{
+ unsigned int ret = XFS_SYMLINK_LOG_COUNT;
+
+ /*
+ * Pre-reserve enough log reservation to handle the transaction
+ * rolling needed to add one parent pointer.
+ */
+ if (xfs_has_parent(mp))
+ ret += resp->tr_attrsetm.tr_logcount;
+
+ return ret;
+}
/*
* Making a new symplink is the same as creating a new file, but
@@ -911,6 +1109,52 @@ xfs_calc_sb_reservation(
return xfs_calc_buf_res(1, mp->m_sb.sb_sectsize);
}
+/*
+ * Namespace reservations.
+ *
+ * These get tricky when parent pointers are enabled as we have attribute
+ * modifications occurring from within these transactions. Rather than confuse
+ * each of these reservation calculations with the conditional attribute
+ * reservations, add them here in a clear and concise manner. This requires that
+ * the attribute reservations have already been calculated.
+ *
+ * Note that we only include the static attribute reservation here; the runtime
+ * reservation will have to be modified by the size of the attributes being
+ * added/removed/modified. See the comments on the attribute reservation
+ * calculations for more details.
+ */
+STATIC void
+xfs_calc_namespace_reservations(
+ struct xfs_mount *mp,
+ struct xfs_trans_resv *resp)
+{
+ ASSERT(resp->tr_attrsetm.tr_logres > 0);
+
+ resp->tr_rename.tr_logres = xfs_calc_rename_reservation(mp);
+ resp->tr_rename.tr_logcount = xfs_rename_log_count(mp, resp);
+ resp->tr_rename.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
+
+ resp->tr_link.tr_logres = xfs_calc_link_reservation(mp);
+ resp->tr_link.tr_logcount = xfs_link_log_count(mp, resp);
+ resp->tr_link.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
+
+ resp->tr_remove.tr_logres = xfs_calc_remove_reservation(mp);
+ resp->tr_remove.tr_logcount = xfs_remove_log_count(mp, resp);
+ resp->tr_remove.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
+
+ resp->tr_symlink.tr_logres = xfs_calc_symlink_reservation(mp);
+ resp->tr_symlink.tr_logcount = xfs_symlink_log_count(mp, resp);
+ resp->tr_symlink.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
+
+ resp->tr_create.tr_logres = xfs_calc_icreate_reservation(mp);
+ resp->tr_create.tr_logcount = xfs_icreate_log_count(mp, resp);
+ resp->tr_create.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
+
+ resp->tr_mkdir.tr_logres = xfs_calc_mkdir_reservation(mp);
+ resp->tr_mkdir.tr_logcount = xfs_mkdir_log_count(mp, resp);
+ resp->tr_mkdir.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
+}
+
void
xfs_trans_resv_calc(
struct xfs_mount *mp,
@@ -930,35 +1174,11 @@ xfs_trans_resv_calc(
resp->tr_itruncate.tr_logcount = XFS_ITRUNCATE_LOG_COUNT;
resp->tr_itruncate.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
- resp->tr_rename.tr_logres = xfs_calc_rename_reservation(mp);
- resp->tr_rename.tr_logcount = XFS_RENAME_LOG_COUNT;
- resp->tr_rename.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
-
- resp->tr_link.tr_logres = xfs_calc_link_reservation(mp);
- resp->tr_link.tr_logcount = XFS_LINK_LOG_COUNT;
- resp->tr_link.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
-
- resp->tr_remove.tr_logres = xfs_calc_remove_reservation(mp);
- resp->tr_remove.tr_logcount = XFS_REMOVE_LOG_COUNT;
- resp->tr_remove.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
-
- resp->tr_symlink.tr_logres = xfs_calc_symlink_reservation(mp);
- resp->tr_symlink.tr_logcount = XFS_SYMLINK_LOG_COUNT;
- resp->tr_symlink.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
-
- resp->tr_create.tr_logres = xfs_calc_icreate_reservation(mp);
- resp->tr_create.tr_logcount = XFS_CREATE_LOG_COUNT;
- resp->tr_create.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
-
resp->tr_create_tmpfile.tr_logres =
xfs_calc_create_tmpfile_reservation(mp);
resp->tr_create_tmpfile.tr_logcount = XFS_CREATE_TMPFILE_LOG_COUNT;
resp->tr_create_tmpfile.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
- resp->tr_mkdir.tr_logres = xfs_calc_mkdir_reservation(mp);
- resp->tr_mkdir.tr_logcount = XFS_MKDIR_LOG_COUNT;
- resp->tr_mkdir.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
-
resp->tr_ifree.tr_logres = xfs_calc_ifree_reservation(mp);
resp->tr_ifree.tr_logcount = XFS_INACTIVE_LOG_COUNT;
resp->tr_ifree.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
@@ -988,6 +1208,8 @@ xfs_trans_resv_calc(
resp->tr_qm_dqalloc.tr_logcount = XFS_WRITE_LOG_COUNT;
resp->tr_qm_dqalloc.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
+ xfs_calc_namespace_reservations(mp, resp);
+
/*
* The following transactions are logged in logical format with
* a default log count.
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 06/18] xfs: parent pointer attribute creation
2023-12-31 19:33 ` [PATCHSET v13.0 4/7] xfs: Parent Pointers Darrick J. Wong
` (4 preceding siblings ...)
2023-12-31 20:49 ` [PATCH 05/18] xfs: extend transaction reservations for parent attributes Darrick J. Wong
@ 2023-12-31 20:50 ` Darrick J. Wong
2023-12-31 20:50 ` [PATCH 07/18] xfs: add parent attributes to link Darrick J. Wong
` (11 subsequent siblings)
17 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:50 UTC (permalink / raw)
To: djwong
Cc: Dave Chinner, Allison Henderson, catherine.hoang,
allison.henderson, linux-xfs
From: Allison Henderson <allison.henderson@oracle.com>
Add parent pointer attribute during xfs_create, and subroutines to
initialize attributes
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
[djwong: shorten names, adjust to new format, set init_xattrs for parent
pointers]
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/Makefile | 1
fs/xfs/libxfs/xfs_attr.c | 2 -
fs/xfs/libxfs/xfs_attr.h | 2 -
fs/xfs/libxfs/xfs_parent.c | 91 +++++++++++++++++++++++++++++++++++++++
fs/xfs/libxfs/xfs_parent.h | 77 +++++++++++++++++++++++++++++++++
fs/xfs/libxfs/xfs_trans_space.c | 52 ++++++++++++++++++++++
fs/xfs/libxfs/xfs_trans_space.h | 9 ++--
fs/xfs/scrub/tempfile.c | 2 -
fs/xfs/xfs_inode.c | 30 ++++++++++---
fs/xfs/xfs_iops.c | 15 ++++++
fs/xfs/xfs_super.c | 10 ++++
fs/xfs/xfs_xattr.c | 4 +-
fs/xfs/xfs_xattr.h | 2 +
13 files changed, 280 insertions(+), 17 deletions(-)
create mode 100644 fs/xfs/libxfs/xfs_trans_space.c
diff --git a/fs/xfs/Makefile b/fs/xfs/Makefile
index 64ff5295d3fc9..5a358113ad9d7 100644
--- a/fs/xfs/Makefile
+++ b/fs/xfs/Makefile
@@ -51,6 +51,7 @@ xfs-y += $(addprefix libxfs/, \
xfs_symlink_remote.o \
xfs_trans_inode.o \
xfs_trans_resv.o \
+ xfs_trans_space.o \
xfs_types.o \
)
# xfs_rtbitmap is shared with libxfs
diff --git a/fs/xfs/libxfs/xfs_attr.c b/fs/xfs/libxfs/xfs_attr.c
index d513ed57177e1..c13eb7b7b5b8f 100644
--- a/fs/xfs/libxfs/xfs_attr.c
+++ b/fs/xfs/libxfs/xfs_attr.c
@@ -895,7 +895,7 @@ xfs_attr_lookup(
return error;
}
-static void
+void
xfs_attr_defer_add(
struct xfs_da_args *args,
unsigned int op_flags)
diff --git a/fs/xfs/libxfs/xfs_attr.h b/fs/xfs/libxfs/xfs_attr.h
index 5b3a0d4b1583c..4a4d45a96dd6c 100644
--- a/fs/xfs/libxfs/xfs_attr.h
+++ b/fs/xfs/libxfs/xfs_attr.h
@@ -549,6 +549,7 @@ int xfs_inode_hasattr(struct xfs_inode *ip);
bool xfs_attr_is_leaf(struct xfs_inode *ip);
int xfs_attr_get_ilocked(struct xfs_da_args *args);
int xfs_attr_get(struct xfs_da_args *args);
+void xfs_attr_defer_add(struct xfs_da_args *args, unsigned int op_flags);
int xfs_attr_set(struct xfs_da_args *args);
int xfs_attr_set_iter(struct xfs_attr_intent *attr);
int xfs_attr_remove_iter(struct xfs_attr_intent *attr);
@@ -557,7 +558,6 @@ bool xfs_attr_namecheck(struct xfs_mount *mp, const void *name, size_t length,
int xfs_attr_calc_size(struct xfs_da_args *args, int *local);
void xfs_init_attr_trans(struct xfs_da_args *args, struct xfs_trans_res *tres,
unsigned int *total);
-
/*
* Check to see if the attr should be upgraded from non-existent or shortform to
* single-leaf-block attribute list.
diff --git a/fs/xfs/libxfs/xfs_parent.c b/fs/xfs/libxfs/xfs_parent.c
index 1d45f926c13a6..05ef155388a12 100644
--- a/fs/xfs/libxfs/xfs_parent.c
+++ b/fs/xfs/libxfs/xfs_parent.c
@@ -28,6 +28,8 @@
#include "xfs_parent.h"
#include "xfs_trans_space.h"
+struct kmem_cache *xfs_parent_args_cache;
+
/*
* Parent pointer attribute handling.
*
@@ -111,3 +113,92 @@ xfs_parent_hashcheck(
/* Namehash matches name? */
return be32_to_cpu(rec->p_namehash) == xfs_dir2_hashname(mp, &dname);
}
+
+/* Initializes a xfs_parent_name_rec to be stored as an attribute name. */
+static inline void
+xfs_init_parent_name_rec(
+ struct xfs_parent_name_rec *rec,
+ const struct xfs_inode *dp,
+ const struct xfs_name *name,
+ struct xfs_inode *ip)
+{
+ rec->p_ino = cpu_to_be64(dp->i_ino);
+ rec->p_gen = cpu_to_be32(VFS_IC(dp)->i_generation);
+ rec->p_namehash = cpu_to_be32(xfs_dir2_hashname(dp->i_mount, name));
+}
+
+/* Point the da args value fields at the non-key parts of a parent pointer. */
+static inline void
+xfs_init_parent_davalue(
+ struct xfs_da_args *args,
+ const struct xfs_name *name)
+{
+ args->valuelen = name->len;
+ args->value = (void *)name->name;
+}
+
+/*
+ * Allocate memory to control a logged parent pointer update as part of a
+ * dirent operation.
+ */
+int
+xfs_parent_args_alloc(
+ struct xfs_mount *mp,
+ struct xfs_parent_args **ppargsp)
+{
+ struct xfs_parent_args *ppargs;
+
+ ppargs = kmem_cache_zalloc(xfs_parent_args_cache, GFP_KERNEL);
+ if (!ppargs)
+ return -ENOMEM;
+
+ xfs_parent_args_init(mp, ppargs);
+ *ppargsp = ppargs;
+ return 0;
+}
+
+static inline xfs_dahash_t
+xfs_parent_hashname(
+ struct xfs_inode *ip,
+ const struct xfs_parent_args *ppargs)
+{
+ return xfs_da_hashname((const void *)&ppargs->rec,
+ sizeof(struct xfs_parent_name_rec));
+}
+
+/* Add a parent pointer to reflect a dirent addition. */
+int
+xfs_parent_addname(
+ struct xfs_trans *tp,
+ struct xfs_parent_args *ppargs,
+ struct xfs_inode *dp,
+ const struct xfs_name *parent_name,
+ struct xfs_inode *child)
+{
+ struct xfs_da_args *args = &ppargs->args;
+
+ if (XFS_IS_CORRUPT(tp->t_mountp,
+ !xfs_parent_valuecheck(tp->t_mountp, parent_name->name,
+ parent_name->len)))
+ return -EFSCORRUPTED;
+
+ xfs_init_parent_name_rec(&ppargs->rec, dp, parent_name, child);
+ args->hashval = xfs_parent_hashname(dp, ppargs);
+
+ args->trans = tp;
+ args->dp = child;
+
+ xfs_init_parent_davalue(&ppargs->args, parent_name);
+
+ xfs_attr_defer_add(args, XFS_ATTRI_OP_FLAGS_SET);
+ return 0;
+}
+
+/* Free a parent pointer context object. */
+void
+xfs_parent_args_free(
+ struct xfs_mount *mp,
+ struct xfs_parent_args *ppargs)
+{
+ kmem_cache_free(xfs_parent_args_cache, ppargs);
+}
diff --git a/fs/xfs/libxfs/xfs_parent.h b/fs/xfs/libxfs/xfs_parent.h
index fcfeddb645f6d..e2115a2b9648b 100644
--- a/fs/xfs/libxfs/xfs_parent.h
+++ b/fs/xfs/libxfs/xfs_parent.h
@@ -16,4 +16,79 @@ bool xfs_parent_hashcheck(struct xfs_mount *mp,
const struct xfs_parent_name_rec *rec, const void *value,
size_t valuelen);
-#endif /* __XFS_PARENT_H__ */
+extern struct kmem_cache *xfs_parent_args_cache;
+
+/*
+ * Dynamically allocd structure used to wrap the needed data to pass around
+ * the defer ops machinery
+ */
+struct xfs_parent_args {
+ struct xfs_parent_name_rec rec;
+ struct xfs_da_args args;
+};
+
+int xfs_parent_args_alloc(struct xfs_mount *mp,
+ struct xfs_parent_args **ppargsp);
+
+/*
+ * Initialize the parent pointer arguments structure. Caller must have zeroed
+ * the contents.
+ */
+static inline void
+xfs_parent_args_init(
+ struct xfs_mount *mp,
+ struct xfs_parent_args *ppargs)
+{
+ ppargs->args.geo = mp->m_attr_geo;
+ ppargs->args.whichfork = XFS_ATTR_FORK;
+ ppargs->args.attr_filter = XFS_ATTR_PARENT;
+ ppargs->args.op_flags = XFS_DA_OP_OKNOENT | XFS_DA_OP_LOGGED |
+ XFS_DA_OP_NVLOOKUP;
+ ppargs->args.name = (const uint8_t *)&ppargs->rec;
+ ppargs->args.namelen = sizeof(struct xfs_parent_name_rec);
+}
+
+/*
+ * Start a parent pointer update by allocating the context object we need to
+ * perform a parent pointer update.
+ */
+static inline int
+xfs_parent_start(
+ struct xfs_mount *mp,
+ struct xfs_parent_args **ppargsp)
+{
+ *ppargsp = NULL;
+
+ if (xfs_has_parent(mp))
+ return xfs_parent_args_alloc(mp, ppargsp);
+ return 0;
+}
+
+int xfs_parent_addname(struct xfs_trans *tp, struct xfs_parent_args *ppargs,
+ struct xfs_inode *dp, const struct xfs_name *parent_name,
+ struct xfs_inode *child);
+
+/* Schedule a parent pointer addition. */
+static inline int
+xfs_parent_add(struct xfs_trans *tp, struct xfs_parent_args *ppargs,
+ struct xfs_inode *dp, const struct xfs_name *parent_name,
+ struct xfs_inode *child)
+{
+ if (ppargs)
+ return xfs_parent_addname(tp, ppargs, dp, parent_name, child);
+ return 0;
+}
+
+void xfs_parent_args_free(struct xfs_mount *mp, struct xfs_parent_args *ppargs);
+
+/* Finish a parent pointer update by freeing the context object. */
+static inline void
+xfs_parent_finish(
+ struct xfs_mount *mp,
+ struct xfs_parent_args *ppargs)
+{
+ if (ppargs)
+ xfs_parent_args_free(mp, ppargs);
+}
+
+#endif /* __XFS_PARENT_H__ */
diff --git a/fs/xfs/libxfs/xfs_trans_space.c b/fs/xfs/libxfs/xfs_trans_space.c
new file mode 100644
index 0000000000000..90532c3fa2053
--- /dev/null
+++ b/fs/xfs/libxfs/xfs_trans_space.c
@@ -0,0 +1,52 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Copyright (c) 2000,2005 Silicon Graphics, Inc.
+ * All Rights Reserved.
+ */
+#include "xfs.h"
+#include "xfs_fs.h"
+#include "xfs_shared.h"
+#include "xfs_format.h"
+#include "xfs_da_format.h"
+#include "xfs_log_format.h"
+#include "xfs_trans_resv.h"
+#include "xfs_mount.h"
+#include "xfs_da_btree.h"
+#include "xfs_bmap_btree.h"
+#include "xfs_trans_space.h"
+
+/* Calculate the disk space required to add a parent pointer. */
+unsigned int
+xfs_parent_calc_space_res(
+ struct xfs_mount *mp,
+ unsigned int namelen)
+{
+ /*
+ * Parent pointers are always the first attr in an attr tree, and never
+ * larger than a block
+ */
+ return XFS_DAENTER_SPACE_RES(mp, XFS_ATTR_FORK) +
+ XFS_NEXTENTADD_SPACE_RES(mp, namelen, XFS_ATTR_FORK);
+}
+
+unsigned int
+xfs_create_space_res(
+ struct xfs_mount *mp,
+ unsigned int namelen)
+{
+ unsigned int ret;
+
+ ret = XFS_IALLOC_SPACE_RES(mp) + XFS_DIRENTER_SPACE_RES(mp, namelen);
+ if (xfs_has_parent(mp))
+ ret += xfs_parent_calc_space_res(mp, namelen);
+
+ return ret;
+}
+
+unsigned int
+xfs_mkdir_space_res(
+ struct xfs_mount *mp,
+ unsigned int namelen)
+{
+ return xfs_create_space_res(mp, namelen);
+}
diff --git a/fs/xfs/libxfs/xfs_trans_space.h b/fs/xfs/libxfs/xfs_trans_space.h
index 9640fc232c147..6cda87153b38c 100644
--- a/fs/xfs/libxfs/xfs_trans_space.h
+++ b/fs/xfs/libxfs/xfs_trans_space.h
@@ -80,8 +80,6 @@
/* This macro is not used - see inline code in xfs_attr_set */
#define XFS_ATTRSET_SPACE_RES(mp, v) \
(XFS_DAENTER_SPACE_RES(mp, XFS_ATTR_FORK) + XFS_B_TO_FSB(mp, v))
-#define XFS_CREATE_SPACE_RES(mp,nl) \
- (XFS_IALLOC_SPACE_RES(mp) + XFS_DIRENTER_SPACE_RES(mp,nl))
#define XFS_DIOSTRAT_SPACE_RES(mp, v) \
(XFS_EXTENTADD_SPACE_RES(mp, XFS_DATA_FORK) + (v))
#define XFS_GROWFS_SPACE_RES(mp) \
@@ -90,8 +88,6 @@
((b) + XFS_EXTENTADD_SPACE_RES(mp, XFS_DATA_FORK))
#define XFS_LINK_SPACE_RES(mp,nl) \
XFS_DIRENTER_SPACE_RES(mp,nl)
-#define XFS_MKDIR_SPACE_RES(mp,nl) \
- (XFS_IALLOC_SPACE_RES(mp) + XFS_DIRENTER_SPACE_RES(mp,nl))
#define XFS_QM_DQALLOC_SPACE_RES(mp) \
(XFS_EXTENTADD_SPACE_RES(mp, XFS_DATA_FORK) + \
XFS_DQUOT_CLUSTER_SIZE_FSB)
@@ -106,5 +102,10 @@
#define XFS_IFREE_SPACE_RES(mp) \
(xfs_has_finobt(mp) ? M_IGEO(mp)->inobt_maxlevels : 0)
+unsigned int xfs_parent_calc_space_res(struct xfs_mount *mp,
+ unsigned int namelen);
+
+unsigned int xfs_create_space_res(struct xfs_mount *mp, unsigned int namelen);
+unsigned int xfs_mkdir_space_res(struct xfs_mount *mp, unsigned int namelen);
#endif /* __XFS_TRANS_SPACE_H__ */
diff --git a/fs/xfs/scrub/tempfile.c b/fs/xfs/scrub/tempfile.c
index 9e0f1d311118b..43d48f1e331de 100644
--- a/fs/xfs/scrub/tempfile.c
+++ b/fs/xfs/scrub/tempfile.c
@@ -71,7 +71,7 @@ xrep_tempfile_create(
return error;
if (is_dir) {
- resblks = XFS_MKDIR_SPACE_RES(mp, 0);
+ resblks = xfs_mkdir_space_res(mp, 0);
tres = &M_RES(mp)->tr_mkdir;
} else {
resblks = XFS_IALLOC_SPACE_RES(mp);
diff --git a/fs/xfs/xfs_inode.c b/fs/xfs/xfs_inode.c
index f1f8f85f941eb..20aee9db1b43a 100644
--- a/fs/xfs/xfs_inode.c
+++ b/fs/xfs/xfs_inode.c
@@ -40,6 +40,8 @@
#include "xfs_log_priv.h"
#include "xfs_health.h"
#include "xfs_pnfs.h"
+#include "xfs_parent.h"
+#include "xfs_xattr.h"
struct kmem_cache *xfs_inode_cache;
@@ -1029,7 +1031,7 @@ xfs_dir_hook_del(
int
xfs_create(
struct mnt_idmap *idmap,
- xfs_inode_t *dp,
+ struct xfs_inode *dp,
struct xfs_name *name,
umode_t mode,
dev_t rdev,
@@ -1041,7 +1043,7 @@ xfs_create(
struct xfs_inode *ip = NULL;
struct xfs_trans *tp = NULL;
int error;
- bool unlock_dp_on_error = false;
+ bool unlock_dp_on_error = false;
prid_t prid;
struct xfs_dquot *udqp = NULL;
struct xfs_dquot *gdqp = NULL;
@@ -1049,6 +1051,7 @@ xfs_create(
struct xfs_trans_res *tres;
uint resblks;
xfs_ino_t ino;
+ struct xfs_parent_args *ppargs;
trace_xfs_create(dp, name);
@@ -1070,13 +1073,17 @@ xfs_create(
return error;
if (is_dir) {
- resblks = XFS_MKDIR_SPACE_RES(mp, name->len);
+ resblks = xfs_mkdir_space_res(mp, name->len);
tres = &M_RES(mp)->tr_mkdir;
} else {
- resblks = XFS_CREATE_SPACE_RES(mp, name->len);
+ resblks = xfs_create_space_res(mp, name->len);
tres = &M_RES(mp)->tr_create;
}
+ error = xfs_parent_start(mp, &ppargs);
+ if (error)
+ goto out_release_dquots;
+
/*
* Initially assume that the file does not exist and
* reserve the resources for that case. If that is not
@@ -1092,7 +1099,7 @@ xfs_create(
resblks, &tp);
}
if (error)
- goto out_release_dquots;
+ goto out_parent;
xfs_ilock(dp, XFS_ILOCK_EXCL | XFS_ILOCK_PARENT);
unlock_dp_on_error = true;
@@ -1135,6 +1142,14 @@ xfs_create(
xfs_bumplink(tp, dp);
}
+ /*
+ * If we have parent pointers, we need to add the attribute containing
+ * the parent information now.
+ */
+ error = xfs_parent_add(tp, ppargs, dp, name, ip);
+ if (error)
+ goto out_trans_cancel;
+
/*
* Create ip with a reference from dp, and add '.' and '..' references
* if it's a directory.
@@ -1167,6 +1182,7 @@ xfs_create(
*ipp = ip;
xfs_iunlock(ip, XFS_ILOCK_EXCL);
xfs_iunlock(dp, XFS_ILOCK_EXCL);
+ xfs_parent_finish(mp, ppargs);
return 0;
out_trans_cancel:
@@ -1182,6 +1198,8 @@ xfs_create(
xfs_finish_inode_setup(ip);
xfs_irele(ip);
}
+ out_parent:
+ xfs_parent_finish(mp, ppargs);
out_release_dquots:
xfs_qm_dqrele(udqp);
xfs_qm_dqrele(gdqp);
@@ -3039,7 +3057,7 @@ xfs_rename_alloc_whiteout(
int error;
error = xfs_create_tmpfile(idmap, dp, S_IFCHR | WHITEOUT_MODE,
- false, &tmpfile);
+ xfs_has_parent(dp->i_mount), &tmpfile);
if (error)
return error;
diff --git a/fs/xfs/xfs_iops.c b/fs/xfs/xfs_iops.c
index d9277f7faf534..62b425129d11c 100644
--- a/fs/xfs/xfs_iops.c
+++ b/fs/xfs/xfs_iops.c
@@ -158,6 +158,8 @@ xfs_create_need_xattr(
if (dir->i_sb->s_security)
return true;
#endif
+ if (xfs_has_parent(XFS_I(dir)->i_mount))
+ return true;
return false;
}
@@ -202,7 +204,18 @@ xfs_generic_create(
xfs_create_need_xattr(dir, default_acl, acl),
&ip);
} else {
- error = xfs_create_tmpfile(idmap, XFS_I(dir), mode, false, &ip);
+ bool init_xattrs = false;
+
+ /*
+ * If this temporary file will be linkable, set up the file
+ * with an attr fork to receive a parent pointer.
+ */
+ if (!(tmpfile->f_flags & O_EXCL) &&
+ xfs_has_parent(XFS_I(dir)->i_mount))
+ init_xattrs = true;
+
+ error = xfs_create_tmpfile(idmap, XFS_I(dir), mode,
+ init_xattrs, &ip);
}
if (unlikely(error))
goto out_free_acl;
diff --git a/fs/xfs/xfs_super.c b/fs/xfs/xfs_super.c
index e6e8e8fb17a19..e981c8b666a5d 100644
--- a/fs/xfs/xfs_super.c
+++ b/fs/xfs/xfs_super.c
@@ -44,6 +44,7 @@
#include "xfs_dahash_test.h"
#include "xfs_rtbitmap.h"
#include "xfs_swapext_item.h"
+#include "xfs_parent.h"
#include "scrub/stats.h"
#include "scrub/rcbag_btree.h"
@@ -2209,8 +2210,16 @@ xfs_init_caches(void)
if (!xfs_sxi_cache)
goto out_destroy_sxd_cache;
+ xfs_parent_args_cache = kmem_cache_create("xfs_parent_args",
+ sizeof(struct xfs_parent_args),
+ 0, 0, NULL);
+ if (!xfs_parent_args_cache)
+ goto out_destroy_sxi_cache;
+
return 0;
+ out_destroy_sxi_cache:
+ kmem_cache_destroy(xfs_sxi_cache);
out_destroy_sxd_cache:
kmem_cache_destroy(xfs_sxd_cache);
out_destroy_iul_cache:
@@ -2271,6 +2280,7 @@ xfs_destroy_caches(void)
* destroy caches.
*/
rcu_barrier();
+ kmem_cache_destroy(xfs_parent_args_cache);
kmem_cache_destroy(xfs_sxd_cache);
kmem_cache_destroy(xfs_sxi_cache);
kmem_cache_destroy(xfs_iunlink_cache);
diff --git a/fs/xfs/xfs_xattr.c b/fs/xfs/xfs_xattr.c
index 2339e3fcfb384..12405e4a70c1b 100644
--- a/fs/xfs/xfs_xattr.c
+++ b/fs/xfs/xfs_xattr.c
@@ -27,7 +27,7 @@
* they must release the permission by calling xlog_drop_incompat_feat
* when they're done.
*/
-static inline int
+int
xfs_attr_grab_log_assist(
struct xfs_mount *mp)
{
@@ -79,7 +79,7 @@ xfs_attr_grab_log_assist(
return error;
}
-static inline void
+void
xfs_attr_rele_log_assist(
struct xfs_mount *mp)
{
diff --git a/fs/xfs/xfs_xattr.h b/fs/xfs/xfs_xattr.h
index cec766cad26cd..f097002d06571 100644
--- a/fs/xfs/xfs_xattr.h
+++ b/fs/xfs/xfs_xattr.h
@@ -7,6 +7,8 @@
#define __XFS_XATTR_H__
int xfs_attr_change(struct xfs_da_args *args);
+int xfs_attr_grab_log_assist(struct xfs_mount *mp);
+void xfs_attr_rele_log_assist(struct xfs_mount *mp);
extern const struct xattr_handler * const xfs_xattr_handlers[];
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 07/18] xfs: add parent attributes to link
2023-12-31 19:33 ` [PATCHSET v13.0 4/7] xfs: Parent Pointers Darrick J. Wong
` (5 preceding siblings ...)
2023-12-31 20:50 ` [PATCH 06/18] xfs: parent pointer attribute creation Darrick J. Wong
@ 2023-12-31 20:50 ` Darrick J. Wong
2023-12-31 20:50 ` [PATCH 08/18] xfs: add parent attributes to symlink Darrick J. Wong
` (10 subsequent siblings)
17 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:50 UTC (permalink / raw)
To: djwong
Cc: Dave Chinner, Allison Henderson, catherine.hoang,
allison.henderson, linux-xfs
From: Allison Henderson <allison.henderson@oracle.com>
This patch modifies xfs_link to add a parent pointer to the inode.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
[djwong: minor rebase fixes]
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/libxfs/xfs_trans_space.c | 14 +++++++++++++
fs/xfs/libxfs/xfs_trans_space.h | 3 +--
fs/xfs/scrub/dir_repair.c | 2 +-
fs/xfs/scrub/orphanage.c | 2 +-
fs/xfs/xfs_inode.c | 41 +++++++++++++++++++++++++++++++++------
5 files changed, 52 insertions(+), 10 deletions(-)
diff --git a/fs/xfs/libxfs/xfs_trans_space.c b/fs/xfs/libxfs/xfs_trans_space.c
index 90532c3fa2053..cf775750120e8 100644
--- a/fs/xfs/libxfs/xfs_trans_space.c
+++ b/fs/xfs/libxfs/xfs_trans_space.c
@@ -50,3 +50,17 @@ xfs_mkdir_space_res(
{
return xfs_create_space_res(mp, namelen);
}
+
+unsigned int
+xfs_link_space_res(
+ struct xfs_mount *mp,
+ unsigned int namelen)
+{
+ unsigned int ret;
+
+ ret = XFS_DIRENTER_SPACE_RES(mp, namelen);
+ if (xfs_has_parent(mp))
+ ret += xfs_parent_calc_space_res(mp, namelen);
+
+ return ret;
+}
diff --git a/fs/xfs/libxfs/xfs_trans_space.h b/fs/xfs/libxfs/xfs_trans_space.h
index 6cda87153b38c..5539634009fb2 100644
--- a/fs/xfs/libxfs/xfs_trans_space.h
+++ b/fs/xfs/libxfs/xfs_trans_space.h
@@ -86,8 +86,6 @@
(2 * (mp)->m_alloc_maxlevels)
#define XFS_GROWFSRT_SPACE_RES(mp,b) \
((b) + XFS_EXTENTADD_SPACE_RES(mp, XFS_DATA_FORK))
-#define XFS_LINK_SPACE_RES(mp,nl) \
- XFS_DIRENTER_SPACE_RES(mp,nl)
#define XFS_QM_DQALLOC_SPACE_RES(mp) \
(XFS_EXTENTADD_SPACE_RES(mp, XFS_DATA_FORK) + \
XFS_DQUOT_CLUSTER_SIZE_FSB)
@@ -107,5 +105,6 @@ unsigned int xfs_parent_calc_space_res(struct xfs_mount *mp,
unsigned int xfs_create_space_res(struct xfs_mount *mp, unsigned int namelen);
unsigned int xfs_mkdir_space_res(struct xfs_mount *mp, unsigned int namelen);
+unsigned int xfs_link_space_res(struct xfs_mount *mp, unsigned int namelen);
#endif /* __XFS_TRANS_SPACE_H__ */
diff --git a/fs/xfs/scrub/dir_repair.c b/fs/xfs/scrub/dir_repair.c
index 48e30a9baeae0..e74f456c7b444 100644
--- a/fs/xfs/scrub/dir_repair.c
+++ b/fs/xfs/scrub/dir_repair.c
@@ -705,7 +705,7 @@ xrep_dir_replay_update(
uint resblks;
int error;
- resblks = XFS_LINK_SPACE_RES(mp, dirent->namelen);
+ resblks = xfs_link_space_res(mp, dirent->namelen);
error = xchk_trans_alloc(rd->sc, resblks);
if (error)
return error;
diff --git a/fs/xfs/scrub/orphanage.c b/fs/xfs/scrub/orphanage.c
index e1024a7bc9e96..84e6dcef067c1 100644
--- a/fs/xfs/scrub/orphanage.c
+++ b/fs/xfs/scrub/orphanage.c
@@ -326,7 +326,7 @@ xrep_adoption_trans_alloc(
/* Compute the worst case space reservation that we need. */
adopt->sc = sc;
- adopt->orphanage_blkres = XFS_LINK_SPACE_RES(mp, MAXNAMELEN);
+ adopt->orphanage_blkres = xfs_link_space_res(mp, MAXNAMELEN);
if (S_ISDIR(VFS_I(sc->ip)->i_mode))
child_blkres = XFS_RENAME_SPACE_RES(mp, xfs_name_dotdot.len);
adopt->child_blkres = child_blkres;
diff --git a/fs/xfs/xfs_inode.c b/fs/xfs/xfs_inode.c
index 20aee9db1b43a..4d7d923cf72ec 100644
--- a/fs/xfs/xfs_inode.c
+++ b/fs/xfs/xfs_inode.c
@@ -1309,14 +1309,15 @@ xfs_create_tmpfile(
int
xfs_link(
- xfs_inode_t *tdp,
- xfs_inode_t *sip,
+ struct xfs_inode *tdp,
+ struct xfs_inode *sip,
struct xfs_name *target_name)
{
- xfs_mount_t *mp = tdp->i_mount;
- xfs_trans_t *tp;
+ struct xfs_mount *mp = tdp->i_mount;
+ struct xfs_trans *tp;
int error, nospace_error = 0;
int resblks;
+ struct xfs_parent_args *ppargs;
trace_xfs_link(tdp, target_name);
@@ -1335,11 +1336,25 @@ xfs_link(
if (error)
goto std_return;
- resblks = XFS_LINK_SPACE_RES(mp, target_name->len);
+ error = xfs_parent_start(mp, &ppargs);
+ if (error)
+ goto std_return;
+
+ resblks = xfs_link_space_res(mp, target_name->len);
error = xfs_trans_alloc_dir(tdp, &M_RES(mp)->tr_link, sip, &resblks,
&tp, &nospace_error);
if (error)
- goto std_return;
+ goto out_parent;
+
+ /*
+ * We don't allow reservationless or quotaless hardlinking when parent
+ * pointers are enabled because we can't back out if the xattrs must
+ * grow.
+ */
+ if (ppargs && nospace_error) {
+ error = nospace_error;
+ goto error_return;
+ }
/*
* If we are using project inheritance, we only allow hard link
@@ -1379,6 +1394,17 @@ xfs_link(
xfs_trans_log_inode(tp, tdp, XFS_ILOG_CORE);
xfs_bumplink(tp, sip);
+
+ /*
+ * If we have parent pointers, we now need to add the parent record to
+ * the attribute fork of the inode. If this is the initial parent
+ * attribute, we need to create it correctly, otherwise we can just add
+ * the parent to the inode.
+ */
+ error = xfs_parent_add(tp, ppargs, tdp, target_name, sip);
+ if (error)
+ goto error_return;
+
xfs_dir_update_hook(tdp, sip, 1, target_name);
/*
@@ -1392,12 +1418,15 @@ xfs_link(
error = xfs_trans_commit(tp);
xfs_iunlock(tdp, XFS_ILOCK_EXCL);
xfs_iunlock(sip, XFS_ILOCK_EXCL);
+ xfs_parent_finish(mp, ppargs);
return error;
error_return:
xfs_trans_cancel(tp);
xfs_iunlock(tdp, XFS_ILOCK_EXCL);
xfs_iunlock(sip, XFS_ILOCK_EXCL);
+ out_parent:
+ xfs_parent_finish(mp, ppargs);
std_return:
if (error == -ENOSPC && nospace_error)
error = nospace_error;
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 08/18] xfs: add parent attributes to symlink
2023-12-31 19:33 ` [PATCHSET v13.0 4/7] xfs: Parent Pointers Darrick J. Wong
` (6 preceding siblings ...)
2023-12-31 20:50 ` [PATCH 07/18] xfs: add parent attributes to link Darrick J. Wong
@ 2023-12-31 20:50 ` Darrick J. Wong
2023-12-31 20:50 ` [PATCH 09/18] xfs: remove parent pointers in unlink Darrick J. Wong
` (9 subsequent siblings)
17 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:50 UTC (permalink / raw)
To: djwong; +Cc: Allison Henderson, catherine.hoang, allison.henderson, linux-xfs
From: Allison Henderson <allison.henderson@oracle.com>
This patch modifies xfs_symlink to add a parent pointer to the inode.
Signed-off-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
[djwong: minor rebase fixups]
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/libxfs/xfs_trans_space.c | 17 +++++++++++++++++
fs/xfs/libxfs/xfs_trans_space.h | 4 ++--
fs/xfs/scrub/symlink_repair.c | 2 +-
fs/xfs/xfs_symlink.c | 28 +++++++++++++++++++++++-----
4 files changed, 43 insertions(+), 8 deletions(-)
diff --git a/fs/xfs/libxfs/xfs_trans_space.c b/fs/xfs/libxfs/xfs_trans_space.c
index cf775750120e8..c8adda82debe0 100644
--- a/fs/xfs/libxfs/xfs_trans_space.c
+++ b/fs/xfs/libxfs/xfs_trans_space.c
@@ -64,3 +64,20 @@ xfs_link_space_res(
return ret;
}
+
+unsigned int
+xfs_symlink_space_res(
+ struct xfs_mount *mp,
+ unsigned int namelen,
+ unsigned int fsblocks)
+{
+ unsigned int ret;
+
+ ret = XFS_IALLOC_SPACE_RES(mp) + XFS_DIRENTER_SPACE_RES(mp, namelen) +
+ fsblocks;
+
+ if (xfs_has_parent(mp))
+ ret += xfs_parent_calc_space_res(mp, namelen);
+
+ return ret;
+}
diff --git a/fs/xfs/libxfs/xfs_trans_space.h b/fs/xfs/libxfs/xfs_trans_space.h
index 5539634009fb2..354ad1d6e18d6 100644
--- a/fs/xfs/libxfs/xfs_trans_space.h
+++ b/fs/xfs/libxfs/xfs_trans_space.h
@@ -95,8 +95,6 @@
XFS_DIRREMOVE_SPACE_RES(mp)
#define XFS_RENAME_SPACE_RES(mp,nl) \
(XFS_DIRREMOVE_SPACE_RES(mp) + XFS_DIRENTER_SPACE_RES(mp,nl))
-#define XFS_SYMLINK_SPACE_RES(mp,nl,b) \
- (XFS_IALLOC_SPACE_RES(mp) + XFS_DIRENTER_SPACE_RES(mp,nl) + (b))
#define XFS_IFREE_SPACE_RES(mp) \
(xfs_has_finobt(mp) ? M_IGEO(mp)->inobt_maxlevels : 0)
@@ -106,5 +104,7 @@ unsigned int xfs_parent_calc_space_res(struct xfs_mount *mp,
unsigned int xfs_create_space_res(struct xfs_mount *mp, unsigned int namelen);
unsigned int xfs_mkdir_space_res(struct xfs_mount *mp, unsigned int namelen);
unsigned int xfs_link_space_res(struct xfs_mount *mp, unsigned int namelen);
+unsigned int xfs_symlink_space_res(struct xfs_mount *mp, unsigned int namelen,
+ unsigned int fsblocks);
#endif /* __XFS_TRANS_SPACE_H__ */
diff --git a/fs/xfs/scrub/symlink_repair.c b/fs/xfs/scrub/symlink_repair.c
index 60246350ebfc9..62190fa738553 100644
--- a/fs/xfs/scrub/symlink_repair.c
+++ b/fs/xfs/scrub/symlink_repair.c
@@ -403,7 +403,7 @@ xrep_symlink_rebuild(
* unlikely.
*/
fs_blocks = xfs_symlink_blocks(sc->mp, target_len);
- resblks = XFS_SYMLINK_SPACE_RES(sc->mp, target_len, fs_blocks);
+ resblks = xfs_symlink_space_res(sc->mp, target_len, fs_blocks);
error = xfs_trans_reserve_quota_nblks(sc->tp, sc->tempip, resblks, 0,
true);
if (error)
diff --git a/fs/xfs/xfs_symlink.c b/fs/xfs/xfs_symlink.c
index 307a2b7e5426e..ced29d8c48c0a 100644
--- a/fs/xfs/xfs_symlink.c
+++ b/fs/xfs/xfs_symlink.c
@@ -25,6 +25,8 @@
#include "xfs_error.h"
#include "xfs_health.h"
#include "xfs_symlink_remote.h"
+#include "xfs_parent.h"
+#include "xfs_defer.h"
/* ----- Kernel only functions below ----- */
int
@@ -101,6 +103,7 @@ xfs_symlink(
struct xfs_dquot *pdqp = NULL;
uint resblks;
xfs_ino_t ino;
+ struct xfs_parent_args *ppargs;
*ipp = NULL;
@@ -131,18 +134,24 @@ xfs_symlink(
/*
* The symlink will fit into the inode data fork?
- * There can't be any attributes so we get the whole variable part.
+ * If there are no parent pointers, then there wont't be any attributes.
+ * So we get the whole variable part, and do not need to reserve extra
+ * blocks. Otherwise, we need to reserve the blocks.
*/
- if (pathlen <= XFS_LITINO(mp))
+ if (pathlen <= XFS_LITINO(mp) && !xfs_has_parent(mp))
fs_blocks = 0;
else
fs_blocks = xfs_symlink_blocks(mp, pathlen);
- resblks = XFS_SYMLINK_SPACE_RES(mp, link_name->len, fs_blocks);
+ resblks = xfs_symlink_space_res(mp, link_name->len, fs_blocks);
+
+ error = xfs_parent_start(mp, &ppargs);
+ if (error)
+ goto out_release_dquots;
error = xfs_trans_alloc_icreate(mp, &M_RES(mp)->tr_symlink, udqp, gdqp,
pdqp, resblks, &tp);
if (error)
- goto out_release_dquots;
+ goto out_parent;
xfs_ilock(dp, XFS_ILOCK_EXCL | XFS_ILOCK_PARENT);
unlock_dp_on_error = true;
@@ -162,7 +171,7 @@ xfs_symlink(
if (!error)
error = xfs_init_new_inode(idmap, tp, dp, ino,
S_IFLNK | (mode & ~S_IFMT), 1, 0, prid,
- false, &ip);
+ xfs_has_parent(mp), &ip);
if (error)
goto out_trans_cancel;
@@ -196,6 +205,12 @@ xfs_symlink(
goto out_trans_cancel;
xfs_trans_ichgtime(tp, dp, XFS_ICHGTIME_MOD | XFS_ICHGTIME_CHG);
xfs_trans_log_inode(tp, dp, XFS_ILOG_CORE);
+
+ /* Add parent pointer for the new symlink. */
+ error = xfs_parent_add(tp, ppargs, dp, link_name, ip);
+ if (error)
+ goto out_trans_cancel;
+
xfs_dir_update_hook(dp, ip, 1, link_name);
/*
@@ -217,6 +232,7 @@ xfs_symlink(
*ipp = ip;
xfs_iunlock(ip, XFS_ILOCK_EXCL);
xfs_iunlock(dp, XFS_ILOCK_EXCL);
+ xfs_parent_finish(mp, ppargs);
return 0;
out_trans_cancel:
@@ -232,6 +248,8 @@ xfs_symlink(
xfs_finish_inode_setup(ip);
xfs_irele(ip);
}
+out_parent:
+ xfs_parent_finish(mp, ppargs);
out_release_dquots:
xfs_qm_dqrele(udqp);
xfs_qm_dqrele(gdqp);
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 09/18] xfs: remove parent pointers in unlink
2023-12-31 19:33 ` [PATCHSET v13.0 4/7] xfs: Parent Pointers Darrick J. Wong
` (7 preceding siblings ...)
2023-12-31 20:50 ` [PATCH 08/18] xfs: add parent attributes to symlink Darrick J. Wong
@ 2023-12-31 20:50 ` Darrick J. Wong
2023-12-31 20:51 ` [PATCH 10/18] xfs: Add parent pointers to rename Darrick J. Wong
` (8 subsequent siblings)
17 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:50 UTC (permalink / raw)
To: djwong
Cc: Dave Chinner, Allison Henderson, catherine.hoang,
allison.henderson, linux-xfs
From: Allison Henderson <allison.henderson@oracle.com>
This patch removes the parent pointer attribute during unlink
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
[djwong: adjust to new ondisk format, minor rebase fixes]
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/libxfs/xfs_parent.c | 39 +++++++++++++++++++++++++++++++++++++++
fs/xfs/libxfs/xfs_parent.h | 16 ++++++++++++++++
fs/xfs/libxfs/xfs_trans_space.c | 13 +++++++++++++
fs/xfs/libxfs/xfs_trans_space.h | 3 +--
fs/xfs/xfs_inode.c | 25 +++++++++++++++++++------
5 files changed, 88 insertions(+), 8 deletions(-)
diff --git a/fs/xfs/libxfs/xfs_parent.c b/fs/xfs/libxfs/xfs_parent.c
index 05ef155388a12..1c3542d264618 100644
--- a/fs/xfs/libxfs/xfs_parent.c
+++ b/fs/xfs/libxfs/xfs_parent.c
@@ -27,6 +27,7 @@
#include "xfs_xattr.h"
#include "xfs_parent.h"
#include "xfs_trans_space.h"
+#include "xfs_health.h"
struct kmem_cache *xfs_parent_args_cache;
@@ -194,6 +195,44 @@ xfs_parent_addname(
return 0;
}
+/* Remove a parent pointer to reflect a dirent removal. */
+int
+xfs_parent_removename(
+ struct xfs_trans *tp,
+ struct xfs_parent_args *ppargs,
+ struct xfs_inode *dp,
+ const struct xfs_name *parent_name,
+ struct xfs_inode *child)
+{
+ struct xfs_da_args *args = &ppargs->args;
+
+ if (XFS_IS_CORRUPT(tp->t_mountp,
+ !xfs_parent_valuecheck(tp->t_mountp, parent_name->name,
+ parent_name->len)))
+ return -EFSCORRUPTED;
+
+ /*
+ * For regular attrs, removing an attr from a !hasattr inode is a nop.
+ * For parent pointers, we require that the pointer must exist if the
+ * caller wants us to remove the pointer.
+ */
+ if (XFS_IS_CORRUPT(child->i_mount, !xfs_inode_hasattr(child))) {
+ xfs_inode_mark_sick(child, XFS_SICK_INO_PARENT);
+ return -EFSCORRUPTED;
+ }
+
+ xfs_init_parent_name_rec(&ppargs->rec, dp, parent_name, child);
+ args->hashval = xfs_parent_hashname(dp, ppargs);
+
+ args->trans = tp;
+ args->dp = child;
+
+ xfs_init_parent_davalue(&ppargs->args, parent_name);
+
+ xfs_attr_defer_add(args, XFS_ATTRI_OP_FLAGS_REMOVE);
+ return 0;
+}
+
/* Free a parent pointer context object. */
void
xfs_parent_args_free(
diff --git a/fs/xfs/libxfs/xfs_parent.h b/fs/xfs/libxfs/xfs_parent.h
index e2115a2b9648b..31349130a330e 100644
--- a/fs/xfs/libxfs/xfs_parent.h
+++ b/fs/xfs/libxfs/xfs_parent.h
@@ -79,6 +79,22 @@ xfs_parent_add(struct xfs_trans *tp, struct xfs_parent_args *ppargs,
return 0;
}
+int xfs_parent_removename(struct xfs_trans *tp, struct xfs_parent_args *ppargs,
+ struct xfs_inode *dp, const struct xfs_name *parent_name,
+ struct xfs_inode *child);
+
+/* Schedule a parent pointer removal. */
+static inline int
+xfs_parent_remove(struct xfs_trans *tp, struct xfs_parent_args *ppargs,
+ struct xfs_inode *dp, const struct xfs_name *parent_name,
+ struct xfs_inode *child)
+{
+ if (ppargs)
+ return xfs_parent_removename(tp, ppargs, dp, parent_name,
+ child);
+ return 0;
+}
+
void xfs_parent_args_free(struct xfs_mount *mp, struct xfs_parent_args *ppargs);
/* Finish a parent pointer update by freeing the context object. */
diff --git a/fs/xfs/libxfs/xfs_trans_space.c b/fs/xfs/libxfs/xfs_trans_space.c
index c8adda82debe0..df729e4f1a4c9 100644
--- a/fs/xfs/libxfs/xfs_trans_space.c
+++ b/fs/xfs/libxfs/xfs_trans_space.c
@@ -81,3 +81,16 @@ xfs_symlink_space_res(
return ret;
}
+
+unsigned int
+xfs_remove_space_res(
+ struct xfs_mount *mp,
+ unsigned int namelen)
+{
+ unsigned int ret = XFS_DIRREMOVE_SPACE_RES(mp);
+
+ if (xfs_has_parent(mp))
+ ret += xfs_parent_calc_space_res(mp, namelen);
+
+ return ret;
+}
diff --git a/fs/xfs/libxfs/xfs_trans_space.h b/fs/xfs/libxfs/xfs_trans_space.h
index 354ad1d6e18d6..a4490813c56f1 100644
--- a/fs/xfs/libxfs/xfs_trans_space.h
+++ b/fs/xfs/libxfs/xfs_trans_space.h
@@ -91,8 +91,6 @@
XFS_DQUOT_CLUSTER_SIZE_FSB)
#define XFS_QM_QINOCREATE_SPACE_RES(mp) \
XFS_IALLOC_SPACE_RES(mp)
-#define XFS_REMOVE_SPACE_RES(mp) \
- XFS_DIRREMOVE_SPACE_RES(mp)
#define XFS_RENAME_SPACE_RES(mp,nl) \
(XFS_DIRREMOVE_SPACE_RES(mp) + XFS_DIRENTER_SPACE_RES(mp,nl))
#define XFS_IFREE_SPACE_RES(mp) \
@@ -106,5 +104,6 @@ unsigned int xfs_mkdir_space_res(struct xfs_mount *mp, unsigned int namelen);
unsigned int xfs_link_space_res(struct xfs_mount *mp, unsigned int namelen);
unsigned int xfs_symlink_space_res(struct xfs_mount *mp, unsigned int namelen,
unsigned int fsblocks);
+unsigned int xfs_remove_space_res(struct xfs_mount *mp, unsigned int namelen);
#endif /* __XFS_TRANS_SPACE_H__ */
diff --git a/fs/xfs/xfs_inode.c b/fs/xfs/xfs_inode.c
index 4d7d923cf72ec..85c9fa6bed2b9 100644
--- a/fs/xfs/xfs_inode.c
+++ b/fs/xfs/xfs_inode.c
@@ -2718,16 +2718,17 @@ xfs_iunpin_wait(
*/
int
xfs_remove(
- xfs_inode_t *dp,
+ struct xfs_inode *dp,
struct xfs_name *name,
- xfs_inode_t *ip)
+ struct xfs_inode *ip)
{
- xfs_mount_t *mp = dp->i_mount;
- xfs_trans_t *tp = NULL;
+ struct xfs_mount *mp = dp->i_mount;
+ struct xfs_trans *tp = NULL;
int is_dir = S_ISDIR(VFS_I(ip)->i_mode);
int dontcare;
int error = 0;
uint resblks;
+ struct xfs_parent_args *ppargs;
trace_xfs_remove(dp, name);
@@ -2744,6 +2745,10 @@ xfs_remove(
if (error)
goto std_return;
+ error = xfs_parent_start(mp, &ppargs);
+ if (error)
+ goto std_return;
+
/*
* We try to get the real space reservation first, allowing for
* directory btree deletion(s) implying possible bmap insert(s). If we
@@ -2755,12 +2760,12 @@ xfs_remove(
* the directory code can handle a reservationless update and we don't
* want to prevent a user from trying to free space by deleting things.
*/
- resblks = XFS_REMOVE_SPACE_RES(mp);
+ resblks = xfs_remove_space_res(mp, name->len);
error = xfs_trans_alloc_dir(dp, &M_RES(mp)->tr_remove, ip, &resblks,
&tp, &dontcare);
if (error) {
ASSERT(error != -ENOSPC);
- goto std_return;
+ goto out_parent;
}
/*
@@ -2820,6 +2825,11 @@ xfs_remove(
goto out_trans_cancel;
}
+ /* Remove parent pointer. */
+ error = xfs_parent_remove(tp, ppargs, dp, name, ip);
+ if (error)
+ goto out_trans_cancel;
+
/*
* Drop the link from dp to ip, and if ip was a directory, remove the
* '.' and '..' references since we freed the directory.
@@ -2843,6 +2853,7 @@ xfs_remove(
xfs_iunlock(ip, XFS_ILOCK_EXCL);
xfs_iunlock(dp, XFS_ILOCK_EXCL);
+ xfs_parent_finish(mp, ppargs);
return 0;
out_trans_cancel:
@@ -2850,6 +2861,8 @@ xfs_remove(
out_unlock:
xfs_iunlock(ip, XFS_ILOCK_EXCL);
xfs_iunlock(dp, XFS_ILOCK_EXCL);
+ out_parent:
+ xfs_parent_finish(mp, ppargs);
std_return:
return error;
}
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 10/18] xfs: Add parent pointers to rename
2023-12-31 19:33 ` [PATCHSET v13.0 4/7] xfs: Parent Pointers Darrick J. Wong
` (8 preceding siblings ...)
2023-12-31 20:50 ` [PATCH 09/18] xfs: remove parent pointers in unlink Darrick J. Wong
@ 2023-12-31 20:51 ` Darrick J. Wong
2023-12-31 20:51 ` [PATCH 11/18] xfs: Add parent pointers to xfs_cross_rename Darrick J. Wong
` (7 subsequent siblings)
17 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:51 UTC (permalink / raw)
To: djwong; +Cc: Allison Henderson, catherine.hoang, allison.henderson, linux-xfs
From: Allison Henderson <allison.henderson@oracle.com>
This patch removes the old parent pointer attribute during the rename
operation, and re-adds the updated parent pointer.
Signed-off-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
[djwong: adjust to new ondisk format]
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/libxfs/xfs_parent.c | 63 ++++++++++++++++++++++++++++++++++
fs/xfs/libxfs/xfs_parent.h | 20 +++++++++++
fs/xfs/libxfs/xfs_trans_space.c | 25 +++++++++++++
fs/xfs/libxfs/xfs_trans_space.h | 6 ++-
fs/xfs/scrub/orphanage.c | 3 +-
fs/xfs/scrub/parent_repair.c | 3 +-
fs/xfs/xfs_inode.c | 73 +++++++++++++++++++++++++++++++++++----
7 files changed, 182 insertions(+), 11 deletions(-)
diff --git a/fs/xfs/libxfs/xfs_parent.c b/fs/xfs/libxfs/xfs_parent.c
index 1c3542d264618..1bff67f8f1176 100644
--- a/fs/xfs/libxfs/xfs_parent.c
+++ b/fs/xfs/libxfs/xfs_parent.c
@@ -138,6 +138,19 @@ xfs_init_parent_davalue(
args->value = (void *)name->name;
}
+/*
+ * Point the da args new value fields at the non-key parts of a replacement
+ * parent pointer.
+ */
+static inline void
+xfs_init_parent_danewvalue(
+ struct xfs_da_args *args,
+ const struct xfs_name *name)
+{
+ args->new_valuelen = name->len;
+ args->new_value = (void *)name->name;
+}
+
/*
* Allocate memory to control a logged parent pointer update as part of a
* dirent operation.
@@ -233,6 +246,56 @@ xfs_parent_removename(
return 0;
}
+/* Replace one parent pointer with another to reflect a rename. */
+int
+xfs_parent_replacename(
+ struct xfs_trans *tp,
+ struct xfs_parent_args *ppargs,
+ struct xfs_inode *old_dp,
+ const struct xfs_name *old_name,
+ struct xfs_inode *new_dp,
+ const struct xfs_name *new_name,
+ struct xfs_inode *child)
+{
+ struct xfs_da_args *args = &ppargs->args;
+
+ if (XFS_IS_CORRUPT(tp->t_mountp,
+ !xfs_parent_valuecheck(tp->t_mountp, old_name->name,
+ old_name->len)))
+ return -EFSCORRUPTED;
+
+ if (XFS_IS_CORRUPT(tp->t_mountp,
+ !xfs_parent_valuecheck(tp->t_mountp, new_name->name,
+ new_name->len)))
+ return -EFSCORRUPTED;
+
+ /*
+ * For regular attrs, replacing an attr from a !hasattr inode becomes
+ * an attr-set operation. For replacing a parent pointer, however, we
+ * require that the old pointer must exist.
+ */
+ if (XFS_IS_CORRUPT(child->i_mount, !xfs_inode_hasattr(child))) {
+ xfs_inode_mark_sick(child, XFS_SICK_INO_PARENT);
+ return -EFSCORRUPTED;
+ }
+
+ xfs_init_parent_name_rec(&ppargs->rec, old_dp, old_name, child);
+ args->hashval = xfs_parent_hashname(old_dp, ppargs);
+
+ xfs_init_parent_name_rec(&ppargs->new_rec, new_dp, new_name, child);
+ args->new_name = (const uint8_t *)&ppargs->new_rec;
+ args->new_namelen = sizeof(struct xfs_parent_name_rec);
+
+ args->trans = tp;
+ args->dp = child;
+
+ xfs_init_parent_davalue(&ppargs->args, old_name);
+ xfs_init_parent_danewvalue(&ppargs->args, new_name);
+
+ xfs_attr_defer_add(args, XFS_ATTRI_OP_FLAGS_REPLACE);
+ return 0;
+}
+
/* Free a parent pointer context object. */
void
xfs_parent_args_free(
diff --git a/fs/xfs/libxfs/xfs_parent.h b/fs/xfs/libxfs/xfs_parent.h
index 31349130a330e..c68c501388e82 100644
--- a/fs/xfs/libxfs/xfs_parent.h
+++ b/fs/xfs/libxfs/xfs_parent.h
@@ -24,6 +24,7 @@ extern struct kmem_cache *xfs_parent_args_cache;
*/
struct xfs_parent_args {
struct xfs_parent_name_rec rec;
+ struct xfs_parent_name_rec new_rec;
struct xfs_da_args args;
};
@@ -95,6 +96,25 @@ xfs_parent_remove(struct xfs_trans *tp, struct xfs_parent_args *ppargs,
return 0;
}
+int xfs_parent_replacename(struct xfs_trans *tp,
+ struct xfs_parent_args *ppargs,
+ struct xfs_inode *old_dp, const struct xfs_name *old_name,
+ struct xfs_inode *new_dp, const struct xfs_name *new_name,
+ struct xfs_inode *child);
+
+/* Schedule a parent pointer replacement. */
+static inline int
+xfs_parent_replace(struct xfs_trans *tp, struct xfs_parent_args *ppargs,
+ struct xfs_inode *old_dp, const struct xfs_name *old_name,
+ struct xfs_inode *new_dp, const struct xfs_name *new_name,
+ struct xfs_inode *child)
+{
+ if (ppargs)
+ return xfs_parent_replacename(tp, ppargs, old_dp, old_name,
+ new_dp, new_name, child);
+ return 0;
+}
+
void xfs_parent_args_free(struct xfs_mount *mp, struct xfs_parent_args *ppargs);
/* Finish a parent pointer update by freeing the context object. */
diff --git a/fs/xfs/libxfs/xfs_trans_space.c b/fs/xfs/libxfs/xfs_trans_space.c
index df729e4f1a4c9..b9dc3752f702c 100644
--- a/fs/xfs/libxfs/xfs_trans_space.c
+++ b/fs/xfs/libxfs/xfs_trans_space.c
@@ -94,3 +94,28 @@ xfs_remove_space_res(
return ret;
}
+
+unsigned int
+xfs_rename_space_res(
+ struct xfs_mount *mp,
+ unsigned int src_namelen,
+ bool target_exists,
+ unsigned int target_namelen,
+ bool has_whiteout)
+{
+ unsigned int ret;
+
+ ret = XFS_DIRREMOVE_SPACE_RES(mp) +
+ XFS_DIRENTER_SPACE_RES(mp, target_namelen);
+
+ if (xfs_has_parent(mp)) {
+ if (has_whiteout)
+ ret += xfs_parent_calc_space_res(mp, src_namelen);
+ ret += 2 * xfs_parent_calc_space_res(mp, target_namelen);
+ }
+
+ if (target_exists)
+ ret += xfs_parent_calc_space_res(mp, target_namelen);
+
+ return ret;
+}
diff --git a/fs/xfs/libxfs/xfs_trans_space.h b/fs/xfs/libxfs/xfs_trans_space.h
index a4490813c56f1..1155ff2d37e29 100644
--- a/fs/xfs/libxfs/xfs_trans_space.h
+++ b/fs/xfs/libxfs/xfs_trans_space.h
@@ -91,8 +91,6 @@
XFS_DQUOT_CLUSTER_SIZE_FSB)
#define XFS_QM_QINOCREATE_SPACE_RES(mp) \
XFS_IALLOC_SPACE_RES(mp)
-#define XFS_RENAME_SPACE_RES(mp,nl) \
- (XFS_DIRREMOVE_SPACE_RES(mp) + XFS_DIRENTER_SPACE_RES(mp,nl))
#define XFS_IFREE_SPACE_RES(mp) \
(xfs_has_finobt(mp) ? M_IGEO(mp)->inobt_maxlevels : 0)
@@ -106,4 +104,8 @@ unsigned int xfs_symlink_space_res(struct xfs_mount *mp, unsigned int namelen,
unsigned int fsblocks);
unsigned int xfs_remove_space_res(struct xfs_mount *mp, unsigned int namelen);
+unsigned int xfs_rename_space_res(struct xfs_mount *mp,
+ unsigned int src_namelen, bool target_exists,
+ unsigned int target_namelen, bool has_whiteout);
+
#endif /* __XFS_TRANS_SPACE_H__ */
diff --git a/fs/xfs/scrub/orphanage.c b/fs/xfs/scrub/orphanage.c
index 84e6dcef067c1..ace7a0f23e474 100644
--- a/fs/xfs/scrub/orphanage.c
+++ b/fs/xfs/scrub/orphanage.c
@@ -328,7 +328,8 @@ xrep_adoption_trans_alloc(
adopt->sc = sc;
adopt->orphanage_blkres = xfs_link_space_res(mp, MAXNAMELEN);
if (S_ISDIR(VFS_I(sc->ip)->i_mode))
- child_blkres = XFS_RENAME_SPACE_RES(mp, xfs_name_dotdot.len);
+ child_blkres = xfs_rename_space_res(mp, 0, false,
+ xfs_name_dotdot.len, false);
adopt->child_blkres = child_blkres;
/*
diff --git a/fs/xfs/scrub/parent_repair.c b/fs/xfs/scrub/parent_repair.c
index 2eb0dbde9c459..099620fc119e9 100644
--- a/fs/xfs/scrub/parent_repair.c
+++ b/fs/xfs/scrub/parent_repair.c
@@ -169,7 +169,8 @@ xrep_parent_reset_dotdot(
* Reserve more space just in case we have to expand the dir. We're
* allowed to exceed quota to repair inconsistent metadata.
*/
- spaceres = XFS_RENAME_SPACE_RES(sc->mp, xfs_name_dotdot.len);
+ spaceres = xfs_rename_space_res(sc->mp, 0, false, xfs_name_dotdot.len,
+ false);
error = xfs_trans_reserve_more_inode(sc->tp, sc->ip, spaceres, 0,
true);
if (error)
diff --git a/fs/xfs/xfs_inode.c b/fs/xfs/xfs_inode.c
index 85c9fa6bed2b9..3723b4bdc47c7 100644
--- a/fs/xfs/xfs_inode.c
+++ b/fs/xfs/xfs_inode.c
@@ -3143,6 +3143,9 @@ xfs_rename(
struct xfs_trans *tp;
struct xfs_inode *wip = NULL; /* whiteout inode */
struct xfs_inode *inodes[__XFS_SORT_INODES];
+ struct xfs_parent_args *src_ppargs = NULL;
+ struct xfs_parent_args *tgt_ppargs = NULL;
+ struct xfs_parent_args *wip_ppargs = NULL;
int i;
int num_inodes = __XFS_SORT_INODES;
bool new_parent = (src_dp != target_dp);
@@ -3174,9 +3177,26 @@ xfs_rename(
xfs_sort_for_rename(src_dp, target_dp, src_ip, target_ip, wip,
inodes, &num_inodes);
+ error = xfs_parent_start(mp, &src_ppargs);
+ if (error)
+ goto out_release_wip;
+
+ if (wip) {
+ error = xfs_parent_start(mp, &wip_ppargs);
+ if (error)
+ goto out_src_ppargs;
+ }
+
+ if (target_ip) {
+ error = xfs_parent_start(mp, &tgt_ppargs);
+ if (error)
+ goto out_wip_ppargs;
+ }
+
retry:
nospace_error = 0;
- spaceres = XFS_RENAME_SPACE_RES(mp, target_name->len);
+ spaceres = xfs_rename_space_res(mp, src_name->len, target_ip != NULL,
+ target_name->len, wip != NULL);
error = xfs_trans_alloc(mp, &M_RES(mp)->tr_rename, spaceres, 0, 0, &tp);
if (error == -ENOSPC) {
nospace_error = error;
@@ -3185,7 +3205,17 @@ xfs_rename(
&tp);
}
if (error)
- goto out_release_wip;
+ goto out_tgt_ppargs;
+
+ /*
+ * We don't allow reservationless renaming when parent pointers are
+ * enabled because we can't back out if the xattrs must grow.
+ */
+ if (src_ppargs && nospace_error) {
+ error = nospace_error;
+ xfs_trans_cancel(tp);
+ goto out_tgt_ppargs;
+ }
/*
* Attach the dquots to the inodes
@@ -3193,7 +3223,7 @@ xfs_rename(
error = xfs_qm_vop_rename_dqattach(inodes);
if (error) {
xfs_trans_cancel(tp);
- goto out_release_wip;
+ goto out_tgt_ppargs;
}
/*
@@ -3262,6 +3292,15 @@ xfs_rename(
goto out_trans_cancel;
}
+ /*
+ * We don't allow quotaless renaming when parent pointers are enabled
+ * because we can't back out if the xattrs must grow.
+ */
+ if (src_ppargs && nospace_error) {
+ error = nospace_error;
+ goto out_trans_cancel;
+ }
+
/*
* Check for expected errors before we dirty the transaction
* so we can return an error without a transaction abort.
@@ -3454,6 +3493,21 @@ xfs_rename(
if (error)
goto out_trans_cancel;
+ /* Schedule parent pointer updates. */
+ error = xfs_parent_add(tp, wip_ppargs, src_dp, src_name, wip);
+ if (error)
+ goto out_trans_cancel;
+
+ error = xfs_parent_replace(tp, src_ppargs, src_dp, src_name, target_dp,
+ target_name, src_ip);
+ if (error)
+ goto out_trans_cancel;
+
+ error = xfs_parent_remove(tp, tgt_ppargs, target_dp, target_name,
+ target_ip);
+ if (error)
+ goto out_trans_cancel;
+
xfs_trans_ichgtime(tp, src_dp, XFS_ICHGTIME_MOD | XFS_ICHGTIME_CHG);
xfs_trans_log_inode(tp, src_dp, XFS_ILOG_CORE);
if (new_parent)
@@ -3475,14 +3529,19 @@ xfs_rename(
xfs_dir_update_hook(src_dp, wip, 1, src_name);
error = xfs_finish_rename(tp);
- xfs_iunlock_rename(inodes, num_inodes);
- if (wip)
- xfs_irele(wip);
- return error;
+ nospace_error = 0;
+ goto out_unlock;
out_trans_cancel:
xfs_trans_cancel(tp);
+out_unlock:
xfs_iunlock_rename(inodes, num_inodes);
+out_tgt_ppargs:
+ xfs_parent_finish(mp, tgt_ppargs);
+out_wip_ppargs:
+ xfs_parent_finish(mp, wip_ppargs);
+out_src_ppargs:
+ xfs_parent_finish(mp, src_ppargs);
out_release_wip:
if (wip)
xfs_irele(wip);
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 11/18] xfs: Add parent pointers to xfs_cross_rename
2023-12-31 19:33 ` [PATCHSET v13.0 4/7] xfs: Parent Pointers Darrick J. Wong
` (9 preceding siblings ...)
2023-12-31 20:51 ` [PATCH 10/18] xfs: Add parent pointers to rename Darrick J. Wong
@ 2023-12-31 20:51 ` Darrick J. Wong
2023-12-31 20:51 ` [PATCH 12/18] xfs: Filter XFS_ATTR_PARENT for getfattr Darrick J. Wong
` (6 subsequent siblings)
17 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:51 UTC (permalink / raw)
To: djwong; +Cc: Allison Henderson, catherine.hoang, allison.henderson, linux-xfs
From: Allison Henderson <allison.henderson@oracle.com>
Cross renames are handled separately from standard renames, and
need different handling to update the parent attributes correctly.
Signed-off-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/xfs_inode.c | 27 +++++++++++++++++++--------
1 file changed, 19 insertions(+), 8 deletions(-)
diff --git a/fs/xfs/xfs_inode.c b/fs/xfs/xfs_inode.c
index 3723b4bdc47c7..998df0d5dac3c 100644
--- a/fs/xfs/xfs_inode.c
+++ b/fs/xfs/xfs_inode.c
@@ -2967,15 +2967,17 @@ xfs_cross_rename(
struct xfs_inode *dp1,
struct xfs_name *name1,
struct xfs_inode *ip1,
+ struct xfs_parent_args *ip1_ppargs,
struct xfs_inode *dp2,
struct xfs_name *name2,
struct xfs_inode *ip2,
+ struct xfs_parent_args *ip2_ppargs,
int spaceres)
{
- int error = 0;
- int ip1_flags = 0;
- int ip2_flags = 0;
- int dp2_flags = 0;
+ int error = 0;
+ int ip1_flags = 0;
+ int ip2_flags = 0;
+ int dp2_flags = 0;
/* Swap inode number for dirent in first parent */
error = xfs_dir_replace(tp, dp1, name1, ip2->i_ino, spaceres);
@@ -3044,6 +3046,15 @@ xfs_cross_rename(
}
}
+ /* Schedule parent pointer replacements */
+ error = xfs_parent_replace(tp, ip1_ppargs, dp1, name1, dp2, name2, ip1);
+ if (error)
+ goto out_trans_abort;
+
+ error = xfs_parent_replace(tp, ip2_ppargs, dp2, name2, dp1, name1, ip2);
+ if (error)
+ goto out_trans_abort;
+
if (ip1_flags) {
xfs_trans_ichgtime(tp, ip1, ip1_flags);
xfs_trans_log_inode(tp, ip1, XFS_ILOG_CORE);
@@ -3260,10 +3271,10 @@ xfs_rename(
/* RENAME_EXCHANGE is unique from here on. */
if (flags & RENAME_EXCHANGE) {
error = xfs_cross_rename(tp, src_dp, src_name, src_ip,
- target_dp, target_name, target_ip,
- spaceres);
- xfs_iunlock_rename(inodes, num_inodes);
- return error;
+ src_ppargs, target_dp, target_name, target_ip,
+ tgt_ppargs, spaceres);
+ nospace_error = 0;
+ goto out_unlock;
}
/*
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 12/18] xfs: Filter XFS_ATTR_PARENT for getfattr
2023-12-31 19:33 ` [PATCHSET v13.0 4/7] xfs: Parent Pointers Darrick J. Wong
` (10 preceding siblings ...)
2023-12-31 20:51 ` [PATCH 11/18] xfs: Add parent pointers to xfs_cross_rename Darrick J. Wong
@ 2023-12-31 20:51 ` Darrick J. Wong
2023-12-31 20:51 ` [PATCH 13/18] xfs: pass the attr value to put_listent when possible Darrick J. Wong
` (5 subsequent siblings)
17 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:51 UTC (permalink / raw)
To: djwong; +Cc: Allison Henderson, catherine.hoang, allison.henderson, linux-xfs
From: Allison Henderson <allison.henderson@oracle.com>
Parent pointers returned to the get_fattr tool cause errors since
the tool cannot parse parent pointers. Fix this by filtering parent
parent pointers from xfs_xattr_put_listent.
Signed-off-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/xfs_xattr.c | 3 +++
1 file changed, 3 insertions(+)
diff --git a/fs/xfs/xfs_xattr.c b/fs/xfs/xfs_xattr.c
index 12405e4a70c1b..483685dbaaceb 100644
--- a/fs/xfs/xfs_xattr.c
+++ b/fs/xfs/xfs_xattr.c
@@ -259,6 +259,9 @@ xfs_xattr_put_listent(
ASSERT(context->count >= 0);
+ if (flags & XFS_ATTR_PARENT)
+ return;
+
if (flags & XFS_ATTR_ROOT) {
#ifdef CONFIG_XFS_POSIX_ACL
if (namelen == SGI_ACL_FILE_SIZE &&
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 13/18] xfs: pass the attr value to put_listent when possible
2023-12-31 19:33 ` [PATCHSET v13.0 4/7] xfs: Parent Pointers Darrick J. Wong
` (11 preceding siblings ...)
2023-12-31 20:51 ` [PATCH 12/18] xfs: Filter XFS_ATTR_PARENT for getfattr Darrick J. Wong
@ 2023-12-31 20:51 ` Darrick J. Wong
2023-12-31 20:52 ` [PATCH 14/18] xfs: Add parent pointer ioctl Darrick J. Wong
` (4 subsequent siblings)
17 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:51 UTC (permalink / raw)
To: djwong; +Cc: Allison Henderson, catherine.hoang, allison.henderson, linux-xfs
From: Allison Henderson <allison.henderson@oracle.com>
Pass the attr value to put_listent when we have local xattrs or
shortform xattrs.
Signed-off-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/libxfs/xfs_attr.h | 5 +++--
fs/xfs/libxfs/xfs_attr_sf.h | 1 +
fs/xfs/xfs_attr_list.c | 8 +++++++-
fs/xfs/xfs_ioctl.c | 1 +
fs/xfs/xfs_xattr.c | 1 +
5 files changed, 13 insertions(+), 3 deletions(-)
diff --git a/fs/xfs/libxfs/xfs_attr.h b/fs/xfs/libxfs/xfs_attr.h
index 4a4d45a96dd6c..0204f62298cb5 100644
--- a/fs/xfs/libxfs/xfs_attr.h
+++ b/fs/xfs/libxfs/xfs_attr.h
@@ -47,8 +47,9 @@ struct xfs_attrlist_cursor_kern {
/* void; state communicated via *context */
-typedef void (*put_listent_func_t)(struct xfs_attr_list_context *, int,
- unsigned char *, int, int);
+typedef void (*put_listent_func_t)(struct xfs_attr_list_context *context,
+ int flags, unsigned char *name, int namelen, void *value,
+ int valuelen);
struct xfs_attr_list_context {
struct xfs_trans *tp;
diff --git a/fs/xfs/libxfs/xfs_attr_sf.h b/fs/xfs/libxfs/xfs_attr_sf.h
index 37578b369d9b9..c6e259791bc37 100644
--- a/fs/xfs/libxfs/xfs_attr_sf.h
+++ b/fs/xfs/libxfs/xfs_attr_sf.h
@@ -24,6 +24,7 @@ typedef struct xfs_attr_sf_sort {
uint8_t flags; /* flags bits (see xfs_attr_leaf.h) */
xfs_dahash_t hash; /* this entry's hash value */
unsigned char *name; /* name value, pointer into buffer */
+ void *value;
} xfs_attr_sf_sort_t;
#define XFS_ATTR_SF_ENTSIZE_MAX /* max space for name&value */ \
diff --git a/fs/xfs/xfs_attr_list.c b/fs/xfs/xfs_attr_list.c
index c86e5952c1378..0eba5c9d21bb8 100644
--- a/fs/xfs/xfs_attr_list.c
+++ b/fs/xfs/xfs_attr_list.c
@@ -97,6 +97,7 @@ xfs_attr_shortform_list(
sfe->flags,
sfe->nameval,
(int)sfe->namelen,
+ &sfe->nameval[sfe->namelen],
(int)sfe->valuelen);
/*
* Either search callback finished early or
@@ -143,6 +144,7 @@ xfs_attr_shortform_list(
sbp->name = sfe->nameval;
sbp->namelen = sfe->namelen;
/* These are bytes, and both on-disk, don't endian-flip */
+ sbp->value = &sfe->nameval[sfe->namelen],
sbp->valuelen = sfe->valuelen;
sbp->flags = sfe->flags;
sfe = xfs_attr_sf_nextentry(sfe);
@@ -194,6 +196,7 @@ xfs_attr_shortform_list(
sbp->flags,
sbp->name,
sbp->namelen,
+ sbp->value,
sbp->valuelen);
if (context->seen_enough)
break;
@@ -481,6 +484,7 @@ xfs_attr3_leaf_list_int(
*/
for (; i < ichdr.count; entry++, i++) {
char *name;
+ void *value;
int namelen, valuelen;
if (be32_to_cpu(entry->hashval) != cursor->hashval) {
@@ -498,6 +502,7 @@ xfs_attr3_leaf_list_int(
name_loc = xfs_attr3_leaf_name_local(leaf, i);
name = name_loc->nameval;
namelen = name_loc->namelen;
+ value = &name_loc->nameval[name_loc->namelen];
valuelen = be16_to_cpu(name_loc->valuelen);
} else {
xfs_attr_leaf_name_remote_t *name_rmt;
@@ -505,6 +510,7 @@ xfs_attr3_leaf_list_int(
name_rmt = xfs_attr3_leaf_name_remote(leaf, i);
name = name_rmt->name;
namelen = name_rmt->namelen;
+ value = NULL;
valuelen = be32_to_cpu(name_rmt->valuelen);
}
@@ -515,7 +521,7 @@ xfs_attr3_leaf_list_int(
return -EFSCORRUPTED;
}
context->put_listent(context, entry->flags,
- name, namelen, valuelen);
+ name, namelen, value, valuelen);
if (context->seen_enough)
break;
cursor->offset++;
diff --git a/fs/xfs/xfs_ioctl.c b/fs/xfs/xfs_ioctl.c
index de16dbc9e7ded..968412c0ba59e 100644
--- a/fs/xfs/xfs_ioctl.c
+++ b/fs/xfs/xfs_ioctl.c
@@ -310,6 +310,7 @@ xfs_ioc_attr_put_listent(
int flags,
unsigned char *name,
int namelen,
+ void *value,
int valuelen)
{
struct xfs_attrlist *alist = context->buffer;
diff --git a/fs/xfs/xfs_xattr.c b/fs/xfs/xfs_xattr.c
index 483685dbaaceb..9ae2dbc7da6c2 100644
--- a/fs/xfs/xfs_xattr.c
+++ b/fs/xfs/xfs_xattr.c
@@ -252,6 +252,7 @@ xfs_xattr_put_listent(
int flags,
unsigned char *name,
int namelen,
+ void *value,
int valuelen)
{
char *prefix;
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 14/18] xfs: Add parent pointer ioctl
2023-12-31 19:33 ` [PATCHSET v13.0 4/7] xfs: Parent Pointers Darrick J. Wong
` (12 preceding siblings ...)
2023-12-31 20:51 ` [PATCH 13/18] xfs: pass the attr value to put_listent when possible Darrick J. Wong
@ 2023-12-31 20:52 ` Darrick J. Wong
2023-12-31 20:52 ` [PATCH 15/18] xfs: fix unit conversion error in xfs_log_calc_max_attrsetm_res Darrick J. Wong
` (3 subsequent siblings)
17 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:52 UTC (permalink / raw)
To: djwong; +Cc: Allison Henderson, catherine.hoang, allison.henderson, linux-xfs
From: Allison Henderson <allison.henderson@oracle.com>
This patch adds a new file ioctl to retrieve the parent pointer of a
given inode
Signed-off-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
[djwong: move new ioctl to xfs_fs_staging.h, adjust to new ondisk format]
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/Makefile | 1
fs/xfs/libxfs/xfs_fs.h | 1
fs/xfs/libxfs/xfs_fs_staging.h | 66 ++++++++++++++++
fs/xfs/libxfs/xfs_ondisk.h | 4 +
fs/xfs/libxfs/xfs_parent.c | 62 +++++++++++++++
fs/xfs/libxfs/xfs_parent.h | 25 ++++++
fs/xfs/xfs_ioctl.c | 145 ++++++++++++++++++++++++++++++++++++
fs/xfs/xfs_parent_utils.c | 161 ++++++++++++++++++++++++++++++++++++++++
fs/xfs/xfs_parent_utils.h | 20 +++++
fs/xfs/xfs_trace.c | 1
fs/xfs/xfs_trace.h | 73 ++++++++++++++++++
11 files changed, 558 insertions(+), 1 deletion(-)
create mode 100644 fs/xfs/xfs_parent_utils.c
create mode 100644 fs/xfs/xfs_parent_utils.h
diff --git a/fs/xfs/Makefile b/fs/xfs/Makefile
index 5a358113ad9d7..49480c81eaeab 100644
--- a/fs/xfs/Makefile
+++ b/fs/xfs/Makefile
@@ -89,6 +89,7 @@ xfs-y += xfs_aops.o \
xfs_mount.o \
xfs_mru_cache.o \
xfs_pwork.o \
+ xfs_parent_utils.o \
xfs_reflink.o \
xfs_stats.o \
xfs_super.o \
diff --git a/fs/xfs/libxfs/xfs_fs.h b/fs/xfs/libxfs/xfs_fs.h
index 63a145e50350b..e92b6a9612a95 100644
--- a/fs/xfs/libxfs/xfs_fs.h
+++ b/fs/xfs/libxfs/xfs_fs.h
@@ -811,6 +811,7 @@ struct xfs_scrub_metadata {
/* XFS_IOC_GETFSMAP ------ hoisted 59 */
#define XFS_IOC_SCRUB_METADATA _IOWR('X', 60, struct xfs_scrub_metadata)
#define XFS_IOC_AG_GEOMETRY _IOWR('X', 61, struct xfs_ag_geometry)
+/* XFS_IOC_GETPARENTS ---- staging 62 */
/*
* ioctl commands that replace IRIX syssgi()'s
diff --git a/fs/xfs/libxfs/xfs_fs_staging.h b/fs/xfs/libxfs/xfs_fs_staging.h
index e3d9f3b32b078..e0650af055895 100644
--- a/fs/xfs/libxfs/xfs_fs_staging.h
+++ b/fs/xfs/libxfs/xfs_fs_staging.h
@@ -104,4 +104,70 @@ struct xfs_exch_range {
#define XFS_IOC_EXCHANGE_RANGE _IOWR('X', 129, struct xfs_exch_range)
+/* Iterating parent pointers of files. */
+
+/* return parents of the handle, not the open fd */
+#define XFS_GETPARENTS_IFLAG_HANDLE (1U << 0)
+
+/* target was the root directory */
+#define XFS_GETPARENTS_OFLAG_ROOT (1U << 1)
+
+/* Cursor is done iterating pptrs */
+#define XFS_GETPARENTS_OFLAG_DONE (1U << 2)
+
+#define XFS_GETPARENTS_FLAG_ALL (XFS_GETPARENTS_IFLAG_HANDLE | \
+ XFS_GETPARENTS_OFLAG_ROOT | \
+ XFS_GETPARENTS_OFLAG_DONE)
+
+/* Get an inode parent pointer through ioctl */
+struct xfs_getparents_rec {
+ __u64 gpr_ino; /* Inode number */
+ __u32 gpr_gen; /* Inode generation */
+ __u32 gpr_pad; /* Reserved */
+ __u64 gpr_rsvd; /* Reserved */
+ __u8 gpr_name[]; /* File name and null terminator */
+};
+
+/* Iterate through an inodes parent pointers */
+struct xfs_getparents {
+ /* File handle, if XFS_GETPARENTS_IFLAG_HANDLE is set */
+ struct xfs_handle gp_handle;
+
+ /*
+ * Structure to track progress in iterating the parent pointers.
+ * Must be initialized to zeroes before the first ioctl call, and
+ * not touched by callers after that.
+ */
+ struct xfs_attrlist_cursor gp_cursor;
+
+ /* Operational flags: XFS_GETPARENTS_*FLAG* */
+ __u32 gp_flags;
+
+ /* Must be set to zero */
+ __u32 gp_reserved;
+
+ /* Size of the buffer in bytes, including this header */
+ __u32 gp_bufsize;
+
+ /* # of entries filled in (output) */
+ __u32 gp_count;
+
+ /* Must be set to zero */
+ __u64 gp_reserved2[5];
+
+ /* Byte offset of each record within the buffer */
+ __u32 gp_offsets[];
+};
+
+static inline struct xfs_getparents_rec*
+xfs_getparents_rec(
+ struct xfs_getparents *info,
+ unsigned int idx)
+{
+ return (struct xfs_getparents_rec *)((char *)info +
+ info->gp_offsets[idx]);
+}
+
+#define XFS_IOC_GETPARENTS _IOWR('X', 62, struct xfs_getparents)
+
#endif /* __XFS_FS_STAGING_H__ */
diff --git a/fs/xfs/libxfs/xfs_ondisk.h b/fs/xfs/libxfs/xfs_ondisk.h
index d9c988c5ad692..bffd39242d487 100644
--- a/fs/xfs/libxfs/xfs_ondisk.h
+++ b/fs/xfs/libxfs/xfs_ondisk.h
@@ -155,6 +155,10 @@ xfs_check_ondisk_structs(void)
XFS_CHECK_OFFSET(struct xfs_efi_log_format_32, efi_extents, 16);
XFS_CHECK_OFFSET(struct xfs_efi_log_format_64, efi_extents, 16);
+ /* parent pointer ioctls */
+ XFS_CHECK_STRUCT_SIZE(struct xfs_getparents_rec, 24);
+ XFS_CHECK_STRUCT_SIZE(struct xfs_getparents, 96);
+
/*
* The v5 superblock format extended several v4 header structures with
* additional data. While new fields are only accessible on v5
diff --git a/fs/xfs/libxfs/xfs_parent.c b/fs/xfs/libxfs/xfs_parent.c
index 1bff67f8f1176..48a2dfcc465fa 100644
--- a/fs/xfs/libxfs/xfs_parent.c
+++ b/fs/xfs/libxfs/xfs_parent.c
@@ -304,3 +304,65 @@ xfs_parent_args_free(
{
kmem_cache_free(xfs_parent_args_cache, ppargs);
}
+
+/* Convert an ondisk parent pointer to the incore format. */
+void
+xfs_parent_irec_from_disk(
+ struct xfs_parent_name_irec *irec,
+ const struct xfs_parent_name_rec *rec,
+ const void *value,
+ unsigned int valuelen)
+{
+ irec->p_ino = be64_to_cpu(rec->p_ino);
+ irec->p_gen = be32_to_cpu(rec->p_gen);
+ irec->p_namehash = be32_to_cpu(rec->p_namehash);
+ irec->p_namelen = valuelen;
+ memcpy(irec->p_name, value, valuelen);
+}
+
+/* Convert an incore parent pointer to the ondisk attr name format. */
+void
+xfs_parent_irec_to_disk(
+ struct xfs_parent_name_rec *rec,
+ const struct xfs_parent_name_irec *irec)
+{
+ rec->p_ino = cpu_to_be64(irec->p_ino);
+ rec->p_gen = cpu_to_be32(irec->p_gen);
+ rec->p_namehash = cpu_to_be32(irec->p_namehash);
+}
+
+/* Is this a valid incore parent pointer? */
+bool
+xfs_parent_verify_irec(
+ struct xfs_mount *mp,
+ const struct xfs_parent_name_irec *irec)
+{
+ struct xfs_name dname = {
+ .name = irec->p_name,
+ .len = irec->p_namelen,
+ };
+
+ if (!xfs_verify_dir_ino(mp, irec->p_ino))
+ return false;
+ if (!xfs_parent_valuecheck(mp, irec->p_name, irec->p_namelen))
+ return false;
+ if (!xfs_dir2_namecheck(irec->p_name, irec->p_namelen))
+ return false;
+ if (irec->p_namehash != xfs_dir2_hashname(mp, &dname))
+ return false;
+ return true;
+}
+
+/* Compute p_namehash for the this parent pointer. */
+void
+xfs_parent_irec_hashname(
+ struct xfs_mount *mp,
+ struct xfs_parent_name_irec *irec)
+{
+ struct xfs_name dname = {
+ .name = irec->p_name,
+ .len = irec->p_namelen,
+ };
+
+ irec->p_namehash = xfs_dir2_hashname(mp, &dname);
+}
diff --git a/fs/xfs/libxfs/xfs_parent.h b/fs/xfs/libxfs/xfs_parent.h
index c68c501388e82..e43ae5a7df826 100644
--- a/fs/xfs/libxfs/xfs_parent.h
+++ b/fs/xfs/libxfs/xfs_parent.h
@@ -127,4 +127,29 @@ xfs_parent_finish(
xfs_parent_args_free(mp, ppargs);
}
+/*
+ * Incore version of a parent pointer, also contains dirent name so callers
+ * can pass/obtain all the parent pointer information in a single structure
+ */
+struct xfs_parent_name_irec {
+ /* Parent pointer attribute name fields */
+ xfs_ino_t p_ino;
+ uint32_t p_gen;
+ xfs_dahash_t p_namehash;
+
+ /* Parent pointer attribute value fields */
+ uint8_t p_namelen;
+ unsigned char p_name[MAXNAMELEN];
+};
+
+void xfs_parent_irec_from_disk(struct xfs_parent_name_irec *irec,
+ const struct xfs_parent_name_rec *rec, const void *value,
+ unsigned int valuelen);
+void xfs_parent_irec_to_disk(struct xfs_parent_name_rec *rec,
+ const struct xfs_parent_name_irec *irec);
+void xfs_parent_irec_hashname(struct xfs_mount *mp,
+ struct xfs_parent_name_irec *irec);
+bool xfs_parent_verify_irec(struct xfs_mount *mp,
+ const struct xfs_parent_name_irec *irec);
+
#endif /* __XFS_PARENT_H__ */
diff --git a/fs/xfs/xfs_ioctl.c b/fs/xfs/xfs_ioctl.c
index 968412c0ba59e..5db70a11151dd 100644
--- a/fs/xfs/xfs_ioctl.c
+++ b/fs/xfs/xfs_ioctl.c
@@ -37,6 +37,7 @@
#include "xfs_health.h"
#include "xfs_reflink.h"
#include "xfs_ioctl.h"
+#include "xfs_parent_utils.h"
#include "xfs_xattr.h"
#include "xfs_rtbitmap.h"
#include "xfs_xchgrange.h"
@@ -1669,6 +1670,147 @@ xfs_ioc_scrub_metadata(
return 0;
}
+/*
+ * IOCTL routine to get the parent pointers of an inode and return it to user
+ * space. Caller must pass a buffer space containing a struct xfs_getparents,
+ * followed by a region large enough to contain an array of struct
+ * xfs_getparents_rec of a size specified in gp_bufsize. If the inode contains
+ * more parent pointers than can fit in the buffer space, caller may re-call
+ * the function using the returned gp_cursor to resume iteration. The
+ * number of xfs_getparents_rec returned will be stored in gp_count.
+ *
+ * Returns 0 on success or non-zero on failure
+ */
+STATIC int
+xfs_ioc_get_parent_pointer(
+ struct file *filp,
+ void __user *arg)
+{
+ struct xfs_getparents *ppi = NULL;
+ int error = 0;
+ struct xfs_inode *file_ip = XFS_I(file_inode(filp));
+ struct xfs_inode *call_ip = file_ip;
+ struct xfs_mount *mp = file_ip->i_mount;
+ void __user *o_pptr;
+ struct xfs_getparents_rec *i_pptr;
+ unsigned int bytes;
+
+ if (!capable(CAP_SYS_ADMIN))
+ return -EPERM;
+
+ /* Allocate an xfs_getparents to put the user data */
+ ppi = kvmalloc(sizeof(struct xfs_getparents), GFP_KERNEL);
+ if (!ppi)
+ return -ENOMEM;
+
+ /* Copy the data from the user */
+ error = copy_from_user(ppi, arg, sizeof(struct xfs_getparents));
+ if (error) {
+ error = -EFAULT;
+ goto out;
+ }
+
+ /* Check size of buffer requested by user */
+ if (ppi->gp_bufsize > XFS_XATTR_LIST_MAX) {
+ error = -ENOMEM;
+ goto out;
+ }
+ if (ppi->gp_bufsize < sizeof(struct xfs_getparents)) {
+ error = -EINVAL;
+ goto out;
+ }
+
+ if (ppi->gp_flags & ~XFS_GETPARENTS_FLAG_ALL) {
+ error = -EINVAL;
+ goto out;
+ }
+ ppi->gp_flags &= ~(XFS_GETPARENTS_OFLAG_ROOT | XFS_GETPARENTS_OFLAG_DONE);
+
+ /*
+ * Now that we know how big the trailing buffer is, expand
+ * our kernel xfs_getparents to be the same size
+ */
+ ppi = kvrealloc(ppi, sizeof(struct xfs_getparents), ppi->gp_bufsize,
+ GFP_KERNEL | __GFP_ZERO);
+ if (!ppi)
+ return -ENOMEM;
+
+ if (ppi->gp_flags & XFS_GETPARENTS_IFLAG_HANDLE) {
+ struct xfs_handle *hanp = &ppi->gp_handle;
+
+ if (memcmp(&hanp->ha_fsid, mp->m_fixedfsid,
+ sizeof(xfs_fsid_t))) {
+ error = -EINVAL;
+ goto out;
+ }
+
+ if (hanp->ha_fid.fid_ino != file_ip->i_ino) {
+ error = xfs_iget(mp, NULL, hanp->ha_fid.fid_ino,
+ XFS_IGET_UNTRUSTED, 0, &call_ip);
+ if (error)
+ goto out;
+
+ /*
+ * Reload the incore unlinked list to avoid failure in
+ * inodegc. Use an unlocked check here because
+ * unrecovered unlinked inodes should be somewhat rare.
+ */
+ if (xfs_inode_unlinked_incomplete(call_ip)) {
+ error = xfs_inode_reload_unlinked(call_ip);
+ if (error)
+ goto out;
+ }
+ }
+
+ if (VFS_I(call_ip)->i_generation != hanp->ha_fid.fid_gen) {
+ error = -EINVAL;
+ goto out;
+ }
+ }
+
+ /* Get the parent pointers */
+ error = xfs_getparent_pointers(call_ip, ppi);
+ if (error)
+ goto out;
+
+ /*
+ * If we ran out of buffer space before copying any parent pointers at
+ * all, the caller's buffer was too short. Tell userspace that, erm,
+ * the message is too long.
+ */
+ if (ppi->gp_count == 0 && !(ppi->gp_flags & XFS_GETPARENTS_OFLAG_DONE)) {
+ error = -EMSGSIZE;
+ goto out;
+ }
+
+ /* Copy the parent pointer head back to the user */
+ bytes = xfs_getparents_arraytop(ppi, ppi->gp_count);
+ error = copy_to_user(arg, ppi, bytes);
+ if (error) {
+ error = -EFAULT;
+ goto out;
+ }
+
+ if (ppi->gp_count == 0)
+ goto out;
+
+ /* Copy the parent pointer records back to the user. */
+ o_pptr = (__user char*)arg + ppi->gp_offsets[ppi->gp_count - 1];
+ i_pptr = xfs_getparents_rec(ppi, ppi->gp_count - 1);
+ bytes = ((char *)ppi + ppi->gp_bufsize) - (char *)i_pptr;
+ error = copy_to_user(o_pptr, i_pptr, bytes);
+ if (error) {
+ error = -EFAULT;
+ goto out;
+ }
+
+out:
+ if (call_ip != file_ip)
+ xfs_irele(call_ip);
+ kvfree(ppi);
+ return error;
+}
+
int
xfs_ioc_swapext(
struct xfs_swapext *sxp)
@@ -1947,7 +2089,8 @@ xfs_file_ioctl(
case XFS_IOC_FSGETXATTRA:
return xfs_ioc_fsgetxattra(ip, arg);
-
+ case XFS_IOC_GETPARENTS:
+ return xfs_ioc_get_parent_pointer(filp, arg);
case XFS_IOC_GETBMAP:
case XFS_IOC_GETBMAPA:
case XFS_IOC_GETBMAPX:
diff --git a/fs/xfs/xfs_parent_utils.c b/fs/xfs/xfs_parent_utils.c
new file mode 100644
index 0000000000000..95eeac1e79c21
--- /dev/null
+++ b/fs/xfs/xfs_parent_utils.c
@@ -0,0 +1,161 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Copyright (c) 2022-2024 Oracle.
+ * All rights reserved.
+ */
+#include "xfs.h"
+#include "xfs_fs.h"
+#include "xfs_format.h"
+#include "xfs_log_format.h"
+#include "xfs_shared.h"
+#include "xfs_trans_resv.h"
+#include "xfs_mount.h"
+#include "xfs_bmap_btree.h"
+#include "xfs_inode.h"
+#include "xfs_error.h"
+#include "xfs_trace.h"
+#include "xfs_trans.h"
+#include "xfs_da_format.h"
+#include "xfs_da_btree.h"
+#include "xfs_attr.h"
+#include "xfs_ioctl.h"
+#include "xfs_parent.h"
+#include "xfs_da_btree.h"
+#include "xfs_parent_utils.h"
+#include "xfs_health.h"
+
+struct xfs_getparent_ctx {
+ struct xfs_attr_list_context context;
+ struct xfs_parent_name_irec pptr_irec;
+ struct xfs_getparents *ppi;
+};
+
+static inline unsigned int
+xfs_getparents_rec_sizeof(
+ const struct xfs_parent_name_irec *irec)
+{
+ return round_up(sizeof(struct xfs_getparents_rec) + irec->p_namelen + 1,
+ sizeof(uint32_t));
+}
+
+static void
+xfs_getparent_listent(
+ struct xfs_attr_list_context *context,
+ int flags,
+ unsigned char *name,
+ int namelen,
+ void *value,
+ int valuelen)
+{
+ struct xfs_getparent_ctx *gp;
+ struct xfs_getparents *ppi;
+ struct xfs_getparents_rec *pptr;
+ struct xfs_parent_name_rec *rec = (void *)name;
+ struct xfs_parent_name_irec *irec;
+ struct xfs_mount *mp = context->dp->i_mount;
+ int arraytop;
+
+ gp = container_of(context, struct xfs_getparent_ctx, context);
+ ppi = gp->ppi;
+ irec = &gp->pptr_irec;
+
+ /* Ignore non-parent xattrs */
+ if (!(flags & XFS_ATTR_PARENT))
+ return;
+
+ /*
+ * Report corruption for anything that doesn't look like a parent
+ * pointer. The attr list functions filtered out INCOMPLETE attrs.
+ */
+ if (XFS_IS_CORRUPT(mp,
+ !xfs_parent_namecheck(mp, rec, namelen, flags)) ||
+ XFS_IS_CORRUPT(mp,
+ !xfs_parent_valuecheck(mp, value, valuelen)) ||
+ XFS_IS_CORRUPT(mp,
+ !xfs_parent_hashcheck(mp, rec, value, valuelen))) {
+ xfs_inode_mark_sick(context->dp, XFS_SICK_INO_PARENT);
+ context->seen_enough = -EFSCORRUPTED;
+ return;
+ }
+
+ xfs_parent_irec_from_disk(&gp->pptr_irec, rec, value, valuelen);
+
+ /*
+ * We found a parent pointer, but we've filled up the buffer. Signal
+ * to the caller that we did /not/ reach the end of the parent pointer
+ * recordset.
+ */
+ arraytop = xfs_getparents_arraytop(ppi, ppi->gp_count + 1);
+ context->firstu -= xfs_getparents_rec_sizeof(irec);
+ if (context->firstu < arraytop) {
+ context->seen_enough = 1;
+ return;
+ }
+
+ trace_xfs_getparent_listent(context->dp, ppi, irec);
+
+ /* Format the parent pointer directly into the caller buffer. */
+ ppi->gp_offsets[ppi->gp_count] = context->firstu;
+ pptr = xfs_getparents_rec(ppi, ppi->gp_count);
+ pptr->gpr_ino = irec->p_ino;
+ pptr->gpr_gen = irec->p_gen;
+ pptr->gpr_pad = 0;
+ pptr->gpr_rsvd = 0;
+
+ memcpy(pptr->gpr_name, irec->p_name, irec->p_namelen);
+ pptr->gpr_name[irec->p_namelen] = 0;
+ ppi->gp_count++;
+}
+
+/* Retrieve the parent pointers for a given inode. */
+int
+xfs_getparent_pointers(
+ struct xfs_inode *ip,
+ struct xfs_getparents *ppi)
+{
+ struct xfs_getparent_ctx *gp;
+ int error;
+
+ gp = kzalloc(sizeof(struct xfs_getparent_ctx), GFP_KERNEL);
+ if (!gp)
+ return -ENOMEM;
+ gp->ppi = ppi;
+ gp->context.dp = ip;
+ gp->context.resynch = 1;
+ gp->context.put_listent = xfs_getparent_listent;
+ gp->context.bufsize = round_down(ppi->gp_bufsize, sizeof(uint32_t));
+ gp->context.firstu = gp->context.bufsize;
+
+ /* Copy the cursor provided by caller */
+ memcpy(&gp->context.cursor, &ppi->gp_cursor,
+ sizeof(struct xfs_attrlist_cursor));
+ ppi->gp_count = 0;
+
+ trace_xfs_getparent_pointers(ip, ppi, &gp->context.cursor);
+
+ error = xfs_attr_list(&gp->context);
+ if (error)
+ goto out_free;
+ if (gp->context.seen_enough < 0) {
+ error = gp->context.seen_enough;
+ goto out_free;
+ }
+
+ /* Is this the root directory? */
+ if (ip->i_ino == ip->i_mount->m_sb.sb_rootino)
+ ppi->gp_flags |= XFS_GETPARENTS_OFLAG_ROOT;
+
+ /*
+ * If we did not run out of buffer space, then we reached the end of
+ * the pptr recordset, so set the DONE flag.
+ */
+ if (gp->context.seen_enough == 0)
+ ppi->gp_flags |= XFS_GETPARENTS_OFLAG_DONE;
+
+ /* Update the caller with the current cursor position */
+ memcpy(&ppi->gp_cursor, &gp->context.cursor,
+ sizeof(struct xfs_attrlist_cursor));
+out_free:
+ kfree(gp);
+ return error;
+}
diff --git a/fs/xfs/xfs_parent_utils.h b/fs/xfs/xfs_parent_utils.h
new file mode 100644
index 0000000000000..2a1d5306a02c2
--- /dev/null
+++ b/fs/xfs/xfs_parent_utils.h
@@ -0,0 +1,20 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Copyright (c) 2022-2024 Oracle.
+ * All rights reserved.
+ */
+#ifndef __XFS_PARENT_UTILS_H__
+#define __XFS_PARENT_UTILS_H__
+
+static inline unsigned int
+xfs_getparents_arraytop(
+ const struct xfs_getparents *ppi,
+ unsigned int nr)
+{
+ return sizeof(struct xfs_getparents) +
+ (nr * sizeof(ppi->gp_offsets[0]));
+}
+
+int xfs_getparent_pointers(struct xfs_inode *ip, struct xfs_getparents *ppi);
+
+#endif /* __XFS_PARENT_UTILS_H__ */
diff --git a/fs/xfs/xfs_trace.c b/fs/xfs/xfs_trace.c
index e38814f4380c8..7ccb7b3473943 100644
--- a/fs/xfs/xfs_trace.c
+++ b/fs/xfs/xfs_trace.c
@@ -42,6 +42,7 @@
#include "xfs_bmap.h"
#include "xfs_swapext.h"
#include "xfs_xchgrange.h"
+#include "xfs_parent.h"
/*
* We include this last to have the helpers above available for the trace
diff --git a/fs/xfs/xfs_trace.h b/fs/xfs/xfs_trace.h
index 1dac853aa659e..070c76f443737 100644
--- a/fs/xfs/xfs_trace.h
+++ b/fs/xfs/xfs_trace.h
@@ -83,6 +83,9 @@ struct xfs_perag;
struct xfs_bmap_intent;
struct xfs_swapext_intent;
struct xfs_swapext_req;
+struct xfs_getparents;
+struct xfs_parent_name_irec;
+struct xfs_attrlist_cursor_kern;
#define XFS_ATTR_FILTER_FLAGS \
{ XFS_ATTR_ROOT, "ROOT" }, \
@@ -4981,6 +4984,76 @@ TRACE_EVENT(xfs_swapext_delta_nextents,
__entry->d_nexts1, __entry->d_nexts2)
);
+TRACE_EVENT(xfs_getparent_listent,
+ TP_PROTO(struct xfs_inode *ip, const struct xfs_getparents *ppi,
+ const struct xfs_parent_name_irec *irec),
+ TP_ARGS(ip, ppi, irec),
+ TP_STRUCT__entry(
+ __field(dev_t, dev)
+ __field(xfs_ino_t, ino)
+ __field(unsigned int, count)
+ __field(unsigned int, bufsize)
+ __field(xfs_ino_t, parent_ino)
+ __field(unsigned int, parent_gen)
+ __field(unsigned int, namelen)
+ __dynamic_array(char, name, irec->p_namelen)
+ ),
+ TP_fast_assign(
+ __entry->dev = ip->i_mount->m_super->s_dev;
+ __entry->ino = ip->i_ino;
+ __entry->count = ppi->gp_count;
+ __entry->bufsize = ppi->gp_bufsize;
+ __entry->parent_ino = irec->p_ino;
+ __entry->parent_gen = irec->p_gen;
+ __entry->namelen = irec->p_namelen;
+ memcpy(__get_str(name), irec->p_name, irec->p_namelen);
+ ),
+ TP_printk("dev %d:%d ino 0x%llx bufsize %u count %u: parent_ino 0x%llx parent_gen 0x%x name '%.*s'",
+ MAJOR(__entry->dev), MINOR(__entry->dev),
+ __entry->ino,
+ __entry->bufsize,
+ __entry->count,
+ __entry->parent_ino,
+ __entry->parent_gen,
+ __entry->namelen,
+ __get_str(name))
+);
+
+TRACE_EVENT(xfs_getparent_pointers,
+ TP_PROTO(struct xfs_inode *ip, const struct xfs_getparents *ppi,
+ const struct xfs_attrlist_cursor_kern *cur),
+ TP_ARGS(ip, ppi, cur),
+ TP_STRUCT__entry(
+ __field(dev_t, dev)
+ __field(xfs_ino_t, ino)
+ __field(unsigned int, flags)
+ __field(unsigned int, bufsize)
+ __field(unsigned int, hashval)
+ __field(unsigned int, blkno)
+ __field(unsigned int, offset)
+ __field(int, initted)
+ ),
+ TP_fast_assign(
+ __entry->dev = ip->i_mount->m_super->s_dev;
+ __entry->ino = ip->i_ino;
+ __entry->flags = ppi->gp_flags;
+ __entry->bufsize = ppi->gp_bufsize;
+ __entry->hashval = cur->hashval;
+ __entry->blkno = cur->blkno;
+ __entry->offset = cur->offset;
+ __entry->initted = cur->initted;
+ ),
+ TP_printk("dev %d:%d ino 0x%llx flags 0x%x bufsize %u cur_init? %d hashval 0x%x blkno %u offset %u",
+ MAJOR(__entry->dev), MINOR(__entry->dev),
+ __entry->ino,
+ __entry->flags,
+ __entry->bufsize,
+ __entry->initted,
+ __entry->hashval,
+ __entry->blkno,
+ __entry->offset)
+);
+
#endif /* _TRACE_XFS_H */
#undef TRACE_INCLUDE_PATH
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 15/18] xfs: fix unit conversion error in xfs_log_calc_max_attrsetm_res
2023-12-31 19:33 ` [PATCHSET v13.0 4/7] xfs: Parent Pointers Darrick J. Wong
` (13 preceding siblings ...)
2023-12-31 20:52 ` [PATCH 14/18] xfs: Add parent pointer ioctl Darrick J. Wong
@ 2023-12-31 20:52 ` Darrick J. Wong
2023-12-31 20:52 ` [PATCH 16/18] xfs: drop compatibility minimum log size computations for reflink Darrick J. Wong
` (2 subsequent siblings)
17 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:52 UTC (permalink / raw)
To: djwong; +Cc: Allison Henderson, catherine.hoang, allison.henderson, linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
Dave and I were discussing some recent test regressions as a result of
me turning on nrext64=1 on realtime filesystems, when we noticed that
the minimum log size of a 32M filesystem jumped from 954 blocks to 4287
blocks.
Digging through xfs_log_calc_max_attrsetm_res, Dave noticed that @size
contains the maximum estimated amount of space needed for a local format
xattr, in bytes, but we feed this quantity to XFS_NEXTENTADD_SPACE_RES,
which requires units of blocks. This has resulted in an overestimation
of the minimum log size over the years.
We should nominally correct this, but there's a backwards compatibility
problem -- if we enable it now, the minimum log size will decrease. If
a corrected mkfs formats a filesystem with this new smaller log size, a
user will encounter mount failures on an uncorrected kernel due to the
larger minimum log size computations there.
However, the large extent counters feature is still EXPERIMENTAL, so we
can gate the correction on that feature (or any features that get added
after that) being enabled. Any filesystem with nrext64 or any of the
as-yet-undefined feature bits turned on will be rejected by old
uncorrected kernels, so this should be safe even in the upgrade case.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
---
fs/xfs/libxfs/xfs_log_rlimit.c | 33 +++++++++++++++++++++++++++++++++
1 file changed, 33 insertions(+)
diff --git a/fs/xfs/libxfs/xfs_log_rlimit.c b/fs/xfs/libxfs/xfs_log_rlimit.c
index 9975b93a7412d..b836de0de5b95 100644
--- a/fs/xfs/libxfs/xfs_log_rlimit.c
+++ b/fs/xfs/libxfs/xfs_log_rlimit.c
@@ -16,6 +16,29 @@
#include "xfs_bmap_btree.h"
#include "xfs_trace.h"
+/*
+ * Decide if the filesystem has the parent pointer feature or any feature
+ * added after that. If so, we can improve the accuracy of the transaction
+ * reservation computations that should lead to more efficient log grant use.
+ */
+static inline bool
+xfs_has_parent_or_newer_feature(
+ struct xfs_mount *mp)
+{
+ if (!xfs_sb_is_v5(&mp->m_sb))
+ return false;
+
+ if (xfs_sb_has_incompat_feature(&mp->m_sb,
+ ~(XFS_SB_FEAT_INCOMPAT_FTYPE |
+ XFS_SB_FEAT_INCOMPAT_SPINODES |
+ XFS_SB_FEAT_INCOMPAT_META_UUID |
+ XFS_SB_FEAT_INCOMPAT_BIGTIME |
+ XFS_SB_FEAT_INCOMPAT_NREXT64)))
+ return true;
+
+ return false;
+}
+
/*
* Calculate the maximum length in bytes that would be required for a local
* attribute value as large attributes out of line are not logged.
@@ -31,6 +54,16 @@ xfs_log_calc_max_attrsetm_res(
MAXNAMELEN - 1;
nblks = XFS_DAENTER_SPACE_RES(mp, XFS_ATTR_FORK);
nblks += XFS_B_TO_FSB(mp, size);
+
+ /*
+ * Starting with the parent pointer feature, every new fs feature
+ * corrects a unit conversion error in the xattr transaction
+ * reservation code that resulted in oversized minimum log size
+ * computations.
+ */
+ if (xfs_has_parent_or_newer_feature(mp))
+ size = XFS_B_TO_FSB(mp, size);
+
nblks += XFS_NEXTENTADD_SPACE_RES(mp, size, XFS_ATTR_FORK);
return M_RES(mp)->tr_attrsetm.tr_logres +
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 16/18] xfs: drop compatibility minimum log size computations for reflink
2023-12-31 19:33 ` [PATCHSET v13.0 4/7] xfs: Parent Pointers Darrick J. Wong
` (14 preceding siblings ...)
2023-12-31 20:52 ` [PATCH 15/18] xfs: fix unit conversion error in xfs_log_calc_max_attrsetm_res Darrick J. Wong
@ 2023-12-31 20:52 ` Darrick J. Wong
2023-12-31 20:52 ` [PATCH 17/18] xfs: don't remove the attr fork when parent pointers are enabled Darrick J. Wong
2023-12-31 20:53 ` [PATCH 18/18] xfs: Add the parent pointer support to the superblock version 5 Darrick J. Wong
17 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:52 UTC (permalink / raw)
To: djwong; +Cc: Allison Henderson, catherine.hoang, allison.henderson, linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
Having established that we can reduce the minimum log size computation
for filesystems with parent pointers or any newer feature, we should
also drop the compat minlogsize code that we added when we reduced the
transaction reservation size for rmap and reflink.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
---
fs/xfs/libxfs/xfs_log_rlimit.c | 10 ++++++++++
1 file changed, 10 insertions(+)
diff --git a/fs/xfs/libxfs/xfs_log_rlimit.c b/fs/xfs/libxfs/xfs_log_rlimit.c
index b836de0de5b95..1ccd4aa921756 100644
--- a/fs/xfs/libxfs/xfs_log_rlimit.c
+++ b/fs/xfs/libxfs/xfs_log_rlimit.c
@@ -81,6 +81,16 @@ xfs_log_calc_trans_resv_for_minlogblocks(
{
unsigned int rmap_maxlevels = mp->m_rmap_maxlevels;
+ /*
+ * Starting with the parent pointer feature, every new fs feature
+ * drops the oversized minimum log size computation introduced by the
+ * original reflink code.
+ */
+ if (xfs_has_parent_or_newer_feature(mp)) {
+ xfs_trans_resv_calc(mp, resv);
+ return;
+ }
+
/*
* In the early days of rmap+reflink, we always set the rmap maxlevels
* to 9 even if the AG was small enough that it would never grow to
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 17/18] xfs: don't remove the attr fork when parent pointers are enabled
2023-12-31 19:33 ` [PATCHSET v13.0 4/7] xfs: Parent Pointers Darrick J. Wong
` (15 preceding siblings ...)
2023-12-31 20:52 ` [PATCH 16/18] xfs: drop compatibility minimum log size computations for reflink Darrick J. Wong
@ 2023-12-31 20:52 ` Darrick J. Wong
2023-12-31 20:53 ` [PATCH 18/18] xfs: Add the parent pointer support to the superblock version 5 Darrick J. Wong
17 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:52 UTC (permalink / raw)
To: djwong; +Cc: Allison Henderson, catherine.hoang, allison.henderson, linux-xfs
From: Allison Henderson <allison.henderson@oracle.com>
When an inode is removed, it may also cause the attribute fork to be
removed if it is the last attribute. This transaction gets flushed to
the log, but if the system goes down before we could inactivate the symlink,
the log recovery tries to inactivate this inode (since it is on the unlinked
list) but the verifier trips over the remote value and leaks it.
Hence we ended up with a file in this odd state on a "clean" mount. The
"obvious" fix is to prohibit erasure of the attr fork to avoid tripping
over the verifiers when pptrs are enabled.
Signed-off-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/libxfs/xfs_attr_leaf.c | 6 ++++--
1 file changed, 4 insertions(+), 2 deletions(-)
diff --git a/fs/xfs/libxfs/xfs_attr_leaf.c b/fs/xfs/libxfs/xfs_attr_leaf.c
index f7a4839e93e3f..2168747aaa2dc 100644
--- a/fs/xfs/libxfs/xfs_attr_leaf.c
+++ b/fs/xfs/libxfs/xfs_attr_leaf.c
@@ -913,7 +913,8 @@ xfs_attr_sf_removename(
totsize -= size;
if (totsize == sizeof(xfs_attr_sf_hdr_t) && xfs_has_attr2(mp) &&
(dp->i_df.if_format != XFS_DINODE_FMT_BTREE) &&
- !(args->op_flags & (XFS_DA_OP_ADDNAME | XFS_DA_OP_REPLACE))) {
+ !(args->op_flags & (XFS_DA_OP_ADDNAME | XFS_DA_OP_REPLACE)) &&
+ !xfs_has_parent(mp)) {
xfs_attr_fork_remove(dp, args->trans);
} else {
xfs_idata_realloc(dp, -size, XFS_ATTR_FORK);
@@ -922,7 +923,8 @@ xfs_attr_sf_removename(
ASSERT(totsize > sizeof(xfs_attr_sf_hdr_t) ||
(args->op_flags & XFS_DA_OP_ADDNAME) ||
!xfs_has_attr2(mp) ||
- dp->i_df.if_format == XFS_DINODE_FMT_BTREE);
+ dp->i_df.if_format == XFS_DINODE_FMT_BTREE ||
+ xfs_has_parent(mp));
xfs_trans_log_inode(args->trans, dp,
XFS_ILOG_CORE | XFS_ILOG_ADATA);
}
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 18/18] xfs: Add the parent pointer support to the superblock version 5.
2023-12-31 19:33 ` [PATCHSET v13.0 4/7] xfs: Parent Pointers Darrick J. Wong
` (16 preceding siblings ...)
2023-12-31 20:52 ` [PATCH 17/18] xfs: don't remove the attr fork when parent pointers are enabled Darrick J. Wong
@ 2023-12-31 20:53 ` Darrick J. Wong
17 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:53 UTC (permalink / raw)
To: djwong
Cc: Mark Tinguely, Dave Chinner, Allison Henderson, Darrick J. Wong,
catherine.hoang, allison.henderson, linux-xfs
From: Allison Henderson <allison.henderson@oracle.com>
Add the parent pointer superblock flag so that we can actually mount
filesystems with this feature enabled.
Signed-off-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/libxfs/xfs_format.h | 4 +++-
fs/xfs/libxfs/xfs_fs.h | 1 +
fs/xfs/libxfs/xfs_sb.c | 4 ++++
fs/xfs/xfs_super.c | 4 ++++
4 files changed, 12 insertions(+), 1 deletion(-)
diff --git a/fs/xfs/libxfs/xfs_format.h b/fs/xfs/libxfs/xfs_format.h
index 8b952909ce1e2..b0aaa825539f6 100644
--- a/fs/xfs/libxfs/xfs_format.h
+++ b/fs/xfs/libxfs/xfs_format.h
@@ -373,13 +373,15 @@ xfs_sb_has_ro_compat_feature(
#define XFS_SB_FEAT_INCOMPAT_BIGTIME (1 << 3) /* large timestamps */
#define XFS_SB_FEAT_INCOMPAT_NEEDSREPAIR (1 << 4) /* needs xfs_repair */
#define XFS_SB_FEAT_INCOMPAT_NREXT64 (1 << 5) /* large extent counters */
+#define XFS_SB_FEAT_INCOMPAT_PARENT (1 << 6) /* parent pointers */
#define XFS_SB_FEAT_INCOMPAT_ALL \
(XFS_SB_FEAT_INCOMPAT_FTYPE| \
XFS_SB_FEAT_INCOMPAT_SPINODES| \
XFS_SB_FEAT_INCOMPAT_META_UUID| \
XFS_SB_FEAT_INCOMPAT_BIGTIME| \
XFS_SB_FEAT_INCOMPAT_NEEDSREPAIR| \
- XFS_SB_FEAT_INCOMPAT_NREXT64)
+ XFS_SB_FEAT_INCOMPAT_NREXT64| \
+ XFS_SB_FEAT_INCOMPAT_PARENT)
#define XFS_SB_FEAT_INCOMPAT_UNKNOWN ~XFS_SB_FEAT_INCOMPAT_ALL
static inline bool
diff --git a/fs/xfs/libxfs/xfs_fs.h b/fs/xfs/libxfs/xfs_fs.h
index e92b6a9612a95..efa68a2d82a1d 100644
--- a/fs/xfs/libxfs/xfs_fs.h
+++ b/fs/xfs/libxfs/xfs_fs.h
@@ -239,6 +239,7 @@ typedef struct xfs_fsop_resblks {
#define XFS_FSOP_GEOM_FLAGS_BIGTIME (1 << 21) /* 64-bit nsec timestamps */
#define XFS_FSOP_GEOM_FLAGS_INOBTCNT (1 << 22) /* inobt btree counter */
#define XFS_FSOP_GEOM_FLAGS_NREXT64 (1 << 23) /* large extent counters */
+#define XFS_FSOP_GEOM_FLAGS_PARENT (1U << 30) /* parent pointers */
/* atomic file extent swap available to userspace */
#define XFS_FSOP_GEOM_FLAGS_ATOMIC_SWAP (1U << 31)
diff --git a/fs/xfs/libxfs/xfs_sb.c b/fs/xfs/libxfs/xfs_sb.c
index 5de377c2b0fea..d4f72d4c85f83 100644
--- a/fs/xfs/libxfs/xfs_sb.c
+++ b/fs/xfs/libxfs/xfs_sb.c
@@ -176,6 +176,8 @@ xfs_sb_version_to_features(
features |= XFS_FEAT_NEEDSREPAIR;
if (sbp->sb_features_incompat & XFS_SB_FEAT_INCOMPAT_NREXT64)
features |= XFS_FEAT_NREXT64;
+ if (sbp->sb_features_incompat & XFS_SB_FEAT_INCOMPAT_PARENT)
+ features |= XFS_FEAT_PARENT;
return features;
}
@@ -1251,6 +1253,8 @@ xfs_fs_geometry(
geo->flags |= XFS_FSOP_GEOM_FLAGS_BIGTIME;
if (xfs_has_inobtcounts(mp))
geo->flags |= XFS_FSOP_GEOM_FLAGS_INOBTCNT;
+ if (xfs_has_parent(mp))
+ geo->flags |= XFS_FSOP_GEOM_FLAGS_PARENT;
if (xfs_has_sector(mp)) {
geo->flags |= XFS_FSOP_GEOM_FLAGS_SECTOR;
geo->logsectsize = sbp->sb_logsectsize;
diff --git a/fs/xfs/xfs_super.c b/fs/xfs/xfs_super.c
index e981c8b666a5d..8b4e5b8579a9b 100644
--- a/fs/xfs/xfs_super.c
+++ b/fs/xfs/xfs_super.c
@@ -1734,6 +1734,10 @@ xfs_fs_fill_super(
goto out_filestream_unmount;
}
+ if (xfs_has_parent(mp))
+ xfs_warn(mp,
+ "EXPERIMENTAL parent pointer feature enabled. Use at your own risk!");
+
error = xfs_mountfs(mp);
if (error)
goto out_filestream_unmount;
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 01/22] xfs: check dirents have parent pointers
2023-12-31 19:33 ` [PATCHSET v13.0 5/7] xfs: fsck for parent pointers Darrick J. Wong
@ 2023-12-31 20:53 ` Darrick J. Wong
2023-12-31 20:53 ` [PATCH 02/22] xfs: deferred scrub of dirents Darrick J. Wong
` (20 subsequent siblings)
21 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:53 UTC (permalink / raw)
To: djwong; +Cc: catherine.hoang, allison.henderson, linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
If the fs has parent pointers, we need to check that each child dirent
points to a file that has a parent pointer pointing back at us.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/libxfs/xfs_parent.c | 54 +++++++++++++++++++
fs/xfs/libxfs/xfs_parent.h | 10 ++++
fs/xfs/scrub/dir.c | 122 ++++++++++++++++++++++++++++++++++++++++++++
3 files changed, 185 insertions(+), 1 deletion(-)
diff --git a/fs/xfs/libxfs/xfs_parent.c b/fs/xfs/libxfs/xfs_parent.c
index 48a2dfcc465fa..09495eb368e2b 100644
--- a/fs/xfs/libxfs/xfs_parent.c
+++ b/fs/xfs/libxfs/xfs_parent.c
@@ -366,3 +366,57 @@ xfs_parent_irec_hashname(
irec->p_namehash = xfs_dir2_hashname(mp, &dname);
}
+
+static inline void
+xfs_parent_scratch_init(
+ struct xfs_trans *tp,
+ struct xfs_inode *ip,
+ const struct xfs_parent_name_irec *pptr,
+ struct xfs_parent_scratch *scr)
+{
+ memset(&scr->args, 0, sizeof(struct xfs_da_args));
+ scr->args.attr_filter = XFS_ATTR_PARENT;
+ scr->args.dp = ip;
+ scr->args.geo = ip->i_mount->m_attr_geo;
+ scr->args.name = (const unsigned char *)&scr->rec;
+ scr->args.namelen = sizeof(struct xfs_parent_name_rec);
+ scr->args.op_flags = XFS_DA_OP_NVLOOKUP;
+ scr->args.trans = tp;
+ scr->args.value = (void *)pptr->p_name;
+ scr->args.valuelen = pptr->p_namelen;
+ scr->args.whichfork = XFS_ATTR_FORK;
+ scr->args.hashval = xfs_da_hashname((const void *)&scr->rec,
+ sizeof(struct xfs_parent_name_rec));
+}
+
+/*
+ * Look up the @name associated with the parent pointer (@pptr) of @ip.
+ * Caller must hold at least ILOCK_SHARED. Returns 0 if the pointer is found,
+ * -ENOATTR if there is no match, or a negative errno. The scratchpad need not
+ * be initialized.
+ */
+int
+xfs_parent_lookup(
+ struct xfs_trans *tp,
+ struct xfs_inode *ip,
+ const struct xfs_parent_name_irec *pptr,
+ struct xfs_parent_scratch *scr)
+{
+ int error;
+
+ /*
+ * Make sure the attr fork iext tree is loaded in transaction context
+ * before we start down the rest of the call path.
+ */
+ if (xfs_inode_hasattr(ip)) {
+ error = xfs_iread_extents(tp, ip, XFS_ATTR_FORK);
+ if (error)
+ return error;
+ }
+
+ xfs_parent_irec_to_disk(&scr->rec, pptr);
+ xfs_parent_scratch_init(tp, ip, pptr, scr);
+ scr->args.op_flags |= XFS_DA_OP_OKNOENT;
+
+ return xfs_attr_get_ilocked(&scr->args);
+}
diff --git a/fs/xfs/libxfs/xfs_parent.h b/fs/xfs/libxfs/xfs_parent.h
index e43ae5a7df826..e4443da1d86f2 100644
--- a/fs/xfs/libxfs/xfs_parent.h
+++ b/fs/xfs/libxfs/xfs_parent.h
@@ -152,4 +152,14 @@ void xfs_parent_irec_hashname(struct xfs_mount *mp,
bool xfs_parent_verify_irec(struct xfs_mount *mp,
const struct xfs_parent_name_irec *irec);
+/* Scratchpad memory so that raw parent operations don't burn stack space. */
+struct xfs_parent_scratch {
+ struct xfs_parent_name_rec rec;
+ struct xfs_da_args args;
+};
+
+int xfs_parent_lookup(struct xfs_trans *tp, struct xfs_inode *ip,
+ const struct xfs_parent_name_irec *pptr,
+ struct xfs_parent_scratch *scratch);
+
#endif /* __XFS_PARENT_H__ */
diff --git a/fs/xfs/scrub/dir.c b/fs/xfs/scrub/dir.c
index 3fe6ffcf9c062..88370804025c4 100644
--- a/fs/xfs/scrub/dir.c
+++ b/fs/xfs/scrub/dir.c
@@ -16,6 +16,8 @@
#include "xfs_dir2.h"
#include "xfs_dir2_priv.h"
#include "xfs_health.h"
+#include "xfs_attr.h"
+#include "xfs_parent.h"
#include "scrub/scrub.h"
#include "scrub/common.h"
#include "scrub/dabtree.h"
@@ -41,6 +43,20 @@ xchk_setup_directory(
/* Directories */
+struct xchk_dir {
+ struct xfs_scrub *sc;
+
+ /* Scratch buffer for scanning pptr xattrs */
+ struct xfs_parent_name_irec pptr;
+
+ /* xattr key and da args for parent pointer validation. */
+ struct xfs_parent_scratch pptr_scratch;
+
+ /* Name buffer for dirent revalidation. */
+ uint8_t namebuf[MAXNAMELEN];
+
+};
+
/* Scrub a directory entry. */
/* Check that an inode's mode matches a given XFS_DIR3_FT_* type. */
@@ -63,6 +79,94 @@ xchk_dir_check_ftype(
xchk_fblock_set_corrupt(sc, XFS_DATA_FORK, offset);
}
+/*
+ * Try to lock a child file for checking parent pointers. Returns the inode
+ * flags for the locks we now hold, or zero if we failed.
+ */
+STATIC unsigned int
+xchk_dir_lock_child(
+ struct xfs_scrub *sc,
+ struct xfs_inode *ip)
+{
+ if (!xfs_ilock_nowait(ip, XFS_IOLOCK_SHARED))
+ return 0;
+
+ if (!xfs_ilock_nowait(ip, XFS_ILOCK_SHARED)) {
+ xfs_iunlock(ip, XFS_IOLOCK_SHARED);
+ return 0;
+ }
+
+ if (!xfs_inode_has_attr_fork(ip) || !xfs_need_iread_extents(&ip->i_af))
+ return XFS_IOLOCK_SHARED | XFS_ILOCK_SHARED;
+
+ xfs_iunlock(ip, XFS_ILOCK_SHARED);
+
+ if (!xfs_ilock_nowait(ip, XFS_ILOCK_EXCL)) {
+ xfs_iunlock(ip, XFS_IOLOCK_SHARED);
+ return 0;
+ }
+
+ return XFS_IOLOCK_SHARED | XFS_ILOCK_EXCL;
+}
+
+/* Check the backwards link (parent pointer) associated with this dirent. */
+STATIC int
+xchk_dir_parent_pointer(
+ struct xchk_dir *sd,
+ const struct xfs_name *name,
+ struct xfs_inode *ip)
+{
+ struct xfs_scrub *sc = sd->sc;
+ int error;
+
+ sd->pptr.p_ino = sc->ip->i_ino;
+ sd->pptr.p_gen = VFS_I(sc->ip)->i_generation;
+ sd->pptr.p_namelen = name->len;
+ memcpy(sd->pptr.p_name, name->name, name->len);
+ xfs_parent_irec_hashname(sc->mp, &sd->pptr);
+
+ error = xfs_parent_lookup(sc->tp, ip, &sd->pptr, &sd->pptr_scratch);
+ if (error == -ENOATTR)
+ xchk_fblock_xref_set_corrupt(sc, XFS_DATA_FORK, 0);
+
+ return 0;
+}
+
+/* Look for a parent pointer matching this dirent, if the child isn't busy. */
+STATIC int
+xchk_dir_check_pptr_fast(
+ struct xchk_dir *sd,
+ xfs_dir2_dataptr_t dapos,
+ const struct xfs_name *name,
+ struct xfs_inode *ip)
+{
+ struct xfs_scrub *sc = sd->sc;
+ unsigned int lockmode;
+ int error;
+
+ /* dot and dotdot entries do not have parent pointers */
+ if (xfs_dir2_samename(name, &xfs_name_dot) ||
+ xfs_dir2_samename(name, &xfs_name_dotdot))
+ return 0;
+
+ /* No self-referential non-dot or dotdot dirents. */
+ if (ip == sc->ip) {
+ xchk_fblock_set_corrupt(sc, XFS_DATA_FORK, 0);
+ return -ECANCELED;
+ }
+
+ /* Try to lock the inode. */
+ lockmode = xchk_dir_lock_child(sc, ip);
+ if (!lockmode) {
+ xchk_set_incomplete(sc);
+ return -ECANCELED;
+ }
+
+ error = xchk_dir_parent_pointer(sd, name, ip);
+ xfs_iunlock(ip, lockmode);
+ return error;
+}
+
/*
* Scrub a single directory entry.
*
@@ -80,6 +184,7 @@ xchk_dir_actor(
{
struct xfs_mount *mp = dp->i_mount;
struct xfs_inode *ip;
+ struct xchk_dir *sd = priv;
xfs_ino_t lookup_ino;
xfs_dablk_t offset;
int error = 0;
@@ -146,6 +251,14 @@ xchk_dir_actor(
goto out;
xchk_dir_check_ftype(sc, offset, ip, name->type);
+
+ if (xfs_has_parent(mp)) {
+ error = xchk_dir_check_pptr_fast(sd, dapos, name, ip);
+ if (error)
+ goto out_rele;
+ }
+
+out_rele:
xchk_irele(sc, ip);
out:
if (sc->sm->sm_flags & XFS_SCRUB_OFLAG_CORRUPT)
@@ -767,6 +880,7 @@ int
xchk_directory(
struct xfs_scrub *sc)
{
+ struct xchk_dir *sd;
int error;
if (!S_ISDIR(VFS_I(sc->ip)->i_mode))
@@ -799,8 +913,14 @@ xchk_directory(
if (sc->sm->sm_flags & XFS_SCRUB_OFLAG_CORRUPT)
return 0;
+ sd = kvzalloc(sizeof(struct xchk_dir), XCHK_GFP_FLAGS);
+ if (!sd)
+ return -ENOMEM;
+ sd->sc = sc;
+
/* Look up every name in this directory by hash. */
- error = xchk_dir_walk(sc, sc->ip, xchk_dir_actor, NULL);
+ error = xchk_dir_walk(sc, sc->ip, xchk_dir_actor, sd);
+ kvfree(sd);
if (error && error != -ECANCELED)
return error;
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 02/22] xfs: deferred scrub of dirents
2023-12-31 19:33 ` [PATCHSET v13.0 5/7] xfs: fsck for parent pointers Darrick J. Wong
2023-12-31 20:53 ` [PATCH 01/22] xfs: check dirents have " Darrick J. Wong
@ 2023-12-31 20:53 ` Darrick J. Wong
2023-12-31 20:53 ` [PATCH 03/22] xfs: create a parent pointer walk function for scrubbers Darrick J. Wong
` (19 subsequent siblings)
21 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:53 UTC (permalink / raw)
To: djwong; +Cc: catherine.hoang, allison.henderson, linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
If the trylock-based parent pointer check fails, retain those dirents
and check them at the end. This may involve dropping the locks on the
file being scanned, so yay.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/scrub/dir.c | 234 +++++++++++++++++++++++++++++++++++++++++++++++-
fs/xfs/scrub/readdir.c | 57 ++++++++++++
fs/xfs/scrub/readdir.h | 3 +
fs/xfs/scrub/trace.h | 34 +++++++
4 files changed, 325 insertions(+), 3 deletions(-)
diff --git a/fs/xfs/scrub/dir.c b/fs/xfs/scrub/dir.c
index 88370804025c4..cfaddde6a34d6 100644
--- a/fs/xfs/scrub/dir.c
+++ b/fs/xfs/scrub/dir.c
@@ -24,6 +24,10 @@
#include "scrub/readdir.h"
#include "scrub/health.h"
#include "scrub/repair.h"
+#include "scrub/trace.h"
+#include "scrub/xfile.h"
+#include "scrub/xfarray.h"
+#include "scrub/xfblob.h"
/* Set us up to scrub directories. */
int
@@ -43,6 +47,18 @@ xchk_setup_directory(
/* Directories */
+/* Deferred directory entry that we saved for later. */
+struct xchk_dirent {
+ /* Cookie for retrieval of the dirent name. */
+ xfblob_cookie name_cookie;
+
+ /* Child inode number. */
+ xfs_ino_t ino;
+
+ /* Length of the pptr name. */
+ uint8_t namelen;
+};
+
struct xchk_dir {
struct xfs_scrub *sc;
@@ -52,6 +68,15 @@ struct xchk_dir {
/* xattr key and da args for parent pointer validation. */
struct xfs_parent_scratch pptr_scratch;
+ /* Fixed-size array of xchk_dirent structures. */
+ struct xfarray *dir_entries;
+
+ /* Blobs containing dirent names. */
+ struct xfblob *dir_names;
+
+ /* If we've cycled the ILOCK, we must revalidate deferred dirents. */
+ bool need_revalidate;
+
/* Name buffer for dirent revalidation. */
uint8_t namebuf[MAXNAMELEN];
@@ -158,8 +183,26 @@ xchk_dir_check_pptr_fast(
/* Try to lock the inode. */
lockmode = xchk_dir_lock_child(sc, ip);
if (!lockmode) {
- xchk_set_incomplete(sc);
- return -ECANCELED;
+ struct xchk_dirent save_de = {
+ .namelen = name->len,
+ .ino = ip->i_ino,
+ };
+
+ /* Couldn't lock the inode, so save the dirent for later. */
+ trace_xchk_dir_defer(sc->ip, name->name, name->len, ip->i_ino);
+
+ error = xfblob_store(sd->dir_names, &save_de.name_cookie,
+ name->name, name->len);
+ if (!xchk_fblock_xref_process_error(sc, XFS_DATA_FORK, 0,
+ &error))
+ return error;
+
+ error = xfarray_append(sd->dir_entries, &save_de);
+ if (!xchk_fblock_xref_process_error(sc, XFS_DATA_FORK, 0,
+ &error))
+ return error;
+
+ return 0;
}
error = xchk_dir_parent_pointer(sd, name, ip);
@@ -875,6 +918,147 @@ xchk_directory_blocks(
return error;
}
+/*
+ * Revalidate a dirent that we collected in the past but couldn't check because
+ * of lock contention. Returns 0 if the dirent is still valid, -ENOENT if it
+ * has gone away on us, or a negative errno.
+ */
+STATIC int
+xchk_dir_revalidate_dirent(
+ struct xchk_dir *sd,
+ const struct xfs_name *xname,
+ xfs_ino_t ino)
+{
+ struct xfs_scrub *sc = sd->sc;
+ xfs_ino_t child_ino;
+ int error;
+
+ /*
+ * Look up the directory entry. If we get -ENOENT, the directory entry
+ * went away and there's nothing to revalidate. Return any other
+ * error.
+ */
+ error = xchk_dir_lookup(sc, sc->ip, xname, &child_ino);
+ if (error)
+ return error;
+
+ /* The inode number changed, nothing to revalidate. */
+ if (ino != child_ino)
+ return -ENOENT;
+
+ return 0;
+}
+
+/*
+ * Check a directory entry's parent pointers the slow way, which means we cycle
+ * locks a bunch and put up with revalidation until we get it done.
+ */
+STATIC int
+xchk_dir_slow_dirent(
+ struct xchk_dir *sd,
+ struct xchk_dirent *dirent)
+{
+ struct xfs_name xname = {
+ .name = sd->namebuf,
+ .len = dirent->namelen,
+ };
+ struct xfs_scrub *sc = sd->sc;
+ struct xfs_inode *ip;
+ unsigned int lockmode;
+ int error;
+
+ /* Check that the deferred dirent still exists. */
+ if (sd->need_revalidate) {
+ error = xchk_dir_revalidate_dirent(sd, &xname, dirent->ino);
+ if (error == -ENOENT)
+ return 0;
+ if (!xchk_fblock_xref_process_error(sc, XFS_DATA_FORK, 0,
+ &error))
+ return error;
+ }
+
+ error = xchk_iget(sc, dirent->ino, &ip);
+ if (error == -EINVAL || error == -ENOENT) {
+ xchk_fblock_set_corrupt(sc, XFS_DATA_FORK, 0);
+ return 0;
+ }
+ if (!xchk_fblock_xref_process_error(sc, XFS_DATA_FORK, 0, &error))
+ return error;
+
+ /*
+ * If we can grab both IOLOCK and ILOCK of the alleged child, we can
+ * proceed with the validation.
+ */
+ lockmode = xchk_dir_lock_child(sc, ip);
+ if (lockmode) {
+ trace_xchk_dir_slowpath(sc->ip, xname.name, xname.len,
+ ip->i_ino);
+ goto check_pptr;
+ }
+
+ /*
+ * We couldn't lock the child file. Drop all the locks and try to
+ * get them again, one at a time.
+ */
+ xchk_iunlock(sc, sc->ilock_flags);
+ sd->need_revalidate = true;
+
+ trace_xchk_dir_ultraslowpath(sc->ip, xname.name, xname.len, ip->i_ino);
+
+ error = xchk_dir_trylock_for_pptrs(sc, ip, &lockmode);
+ if (error)
+ goto out_rele;
+
+ /* Revalidate, since we just cycled the locks. */
+ error = xchk_dir_revalidate_dirent(sd, &xname, dirent->ino);
+ if (error == -ENOENT) {
+ error = 0;
+ goto out_unlock;
+ }
+ if (!xchk_fblock_xref_process_error(sc, XFS_DATA_FORK, 0, &error))
+ goto out_unlock;
+
+check_pptr:
+ error = xchk_dir_parent_pointer(sd, &xname, ip);
+out_unlock:
+ xfs_iunlock(ip, lockmode);
+out_rele:
+ xchk_irele(sc, ip);
+ return error;
+}
+
+/* Check all the dirents that we deferred the first time around. */
+STATIC int
+xchk_dir_finish_slow_dirents(
+ struct xchk_dir *sd)
+{
+ xfarray_idx_t array_cur;
+ int error;
+
+ foreach_xfarray_idx(sd->dir_entries, array_cur) {
+ struct xchk_dirent dirent;
+
+ if (sd->sc->sm->sm_flags & XFS_SCRUB_OFLAG_CORRUPT)
+ return 0;
+
+ error = xfarray_load(sd->dir_entries, array_cur, &dirent);
+ if (error)
+ return error;
+
+ error = xfblob_load(sd->dir_names, dirent.name_cookie,
+ sd->namebuf, dirent.namelen);
+ if (error)
+ return error;
+ sd->namebuf[MAXNAMELEN - 1] = 0;
+
+ error = xchk_dir_slow_dirent(sd, &dirent);
+ if (error)
+ return error;
+ }
+
+ return 0;
+}
+
/* Scrub a whole directory. */
int
xchk_directory(
@@ -918,10 +1102,54 @@ xchk_directory(
return -ENOMEM;
sd->sc = sc;
+ if (xfs_has_parent(sc->mp)) {
+ char *descr;
+
+ /*
+ * Set up some staging memory for dirents that we can't check
+ * due to locking contention.
+ */
+ descr = xchk_xfile_ino_descr(sc, "slow directory entries");
+ error = xfarray_create(descr, 0, sizeof(struct xchk_dirent),
+ &sd->dir_entries);
+ kfree(descr);
+ if (error)
+ goto out_sd;
+
+ descr = xchk_xfile_ino_descr(sc, "slow directory entry names");
+ error = xfblob_create(descr, &sd->dir_names);
+ kfree(descr);
+ if (error)
+ goto out_entries;
+ }
+
/* Look up every name in this directory by hash. */
error = xchk_dir_walk(sc, sc->ip, xchk_dir_actor, sd);
+ if (error == -ECANCELED)
+ error = 0;
+ if (error)
+ goto out_names;
+
+ if (xfs_has_parent(sc->mp)) {
+ error = xchk_dir_finish_slow_dirents(sd);
+ if (error == -ETIMEDOUT) {
+ /* Couldn't grab a lock, scrub was marked incomplete */
+ error = 0;
+ goto out_names;
+ }
+ if (error)
+ goto out_names;
+ }
+
+out_names:
+ if (sd->dir_names)
+ xfblob_destroy(sd->dir_names);
+out_entries:
+ if (sd->dir_entries)
+ xfarray_destroy(sd->dir_entries);
+out_sd:
kvfree(sd);
- if (error && error != -ECANCELED)
+ if (error)
return error;
/* If the dir is clean, it is clearly not zapped. */
diff --git a/fs/xfs/scrub/readdir.c b/fs/xfs/scrub/readdir.c
index d70dbbd4c9040..c9c875485b870 100644
--- a/fs/xfs/scrub/readdir.c
+++ b/fs/xfs/scrub/readdir.c
@@ -18,6 +18,7 @@
#include "xfs_trans.h"
#include "xfs_error.h"
#include "scrub/scrub.h"
+#include "scrub/common.h"
#include "scrub/readdir.h"
/* Call a function for every entry in a shortform directory. */
@@ -382,3 +383,59 @@ xchk_dir_lookup(
*ino = args.inumber;
return error;
}
+
+/*
+ * Try for a limited time to grab the IOLOCK and ILOCK of both the scrub target
+ * (@sc->ip) and the inode at the other end (@ip) of a directory or parent
+ * pointer link so that we can check that link.
+ *
+ * We do not know ahead of time that the directory tree is /not/ corrupt, so we
+ * cannot use the "lock two inode" functions because we do not know that there
+ * is not a racing thread trying to take the locks in opposite order. First
+ * take IOLOCK_EXCL of the scrub target, and then try to take IOLOCK_SHARED
+ * of @ip to synchronize with the VFS. Next, take ILOCK_EXCL of the scrub
+ * target and @ip to synchronize with XFS.
+ *
+ * If the trylocks succeed, *lockmode will be set to the locks held for @ip;
+ * @sc->ilock_flags will be set for the locks held for @sc->ip; and zero will
+ * be returned. If not, returns -EDEADLOCK to try again; or -ETIMEDOUT if
+ * XCHK_TRY_HARDER was set. Returns -EINTR if the process has been killed.
+ */
+int
+xchk_dir_trylock_for_pptrs(
+ struct xfs_scrub *sc,
+ struct xfs_inode *ip,
+ unsigned int *lockmode)
+{
+ unsigned int nr;
+ int error = 0;
+
+ ASSERT(sc->ilock_flags == 0);
+
+ *lockmode = 0;
+ for (nr = 0; nr < HZ; nr++) {
+ xchk_ilock(sc, XFS_IOLOCK_EXCL);
+ if (xfs_ilock_nowait(ip, XFS_IOLOCK_SHARED)) {
+ xchk_ilock(sc, XFS_ILOCK_EXCL);
+ if (xfs_ilock_nowait(ip, XFS_ILOCK_EXCL)) {
+ *lockmode = XFS_IOLOCK_SHARED | XFS_ILOCK_EXCL;
+ return 0;
+ }
+ xchk_iunlock(sc, XFS_ILOCK_EXCL);
+ xfs_iunlock(ip, XFS_IOLOCK_SHARED);
+ }
+ xchk_iunlock(sc, XFS_IOLOCK_EXCL);
+
+ if (xchk_should_terminate(sc, &error))
+ return error;
+
+ delay(1);
+ }
+
+ if (sc->flags & XCHK_TRY_HARDER) {
+ xchk_set_incomplete(sc);
+ return -ETIMEDOUT;
+ }
+
+ return -EDEADLOCK;
+}
diff --git a/fs/xfs/scrub/readdir.h b/fs/xfs/scrub/readdir.h
index 55787f4df123f..da501877a64dd 100644
--- a/fs/xfs/scrub/readdir.h
+++ b/fs/xfs/scrub/readdir.h
@@ -16,4 +16,7 @@ int xchk_dir_walk(struct xfs_scrub *sc, struct xfs_inode *dp,
int xchk_dir_lookup(struct xfs_scrub *sc, struct xfs_inode *dp,
const struct xfs_name *name, xfs_ino_t *ino);
+int xchk_dir_trylock_for_pptrs(struct xfs_scrub *sc, struct xfs_inode *ip,
+ unsigned int *lockmode);
+
#endif /* __XFS_SCRUB_READDIR_H__ */
diff --git a/fs/xfs/scrub/trace.h b/fs/xfs/scrub/trace.h
index 8d863f4737e90..651b73d33f2c4 100644
--- a/fs/xfs/scrub/trace.h
+++ b/fs/xfs/scrub/trace.h
@@ -1523,6 +1523,40 @@ DEFINE_EVENT(xchk_nlinks_diff_class, name, \
TP_ARGS(mp, ip, live))
DEFINE_SCRUB_NLINKS_DIFF_EVENT(xchk_nlinks_compare_inode);
+DECLARE_EVENT_CLASS(xchk_pptr_class,
+ TP_PROTO(struct xfs_inode *ip, const unsigned char *name,
+ unsigned int namelen, xfs_ino_t far_ino),
+ TP_ARGS(ip, name, namelen, far_ino),
+ TP_STRUCT__entry(
+ __field(dev_t, dev)
+ __field(xfs_ino_t, ino)
+ __field(unsigned int, namelen)
+ __dynamic_array(char, name, namelen)
+ __field(xfs_ino_t, far_ino)
+ ),
+ TP_fast_assign(
+ __entry->dev = ip->i_mount->m_super->s_dev;
+ __entry->ino = ip->i_ino;
+ __entry->namelen = namelen;
+ memcpy(__get_str(name), name, namelen);
+ __entry->far_ino = far_ino;
+ ),
+ TP_printk("dev %d:%d ino 0x%llx name '%.*s' far_ino 0x%llx",
+ MAJOR(__entry->dev), MINOR(__entry->dev),
+ __entry->ino,
+ __entry->namelen,
+ __get_str(name),
+ __entry->far_ino)
+)
+#define DEFINE_XCHK_PPTR_EVENT(name) \
+DEFINE_EVENT(xchk_pptr_class, name, \
+ TP_PROTO(struct xfs_inode *ip, const unsigned char *name, \
+ unsigned int namelen, xfs_ino_t far_ino), \
+ TP_ARGS(ip, name, namelen, far_ino))
+DEFINE_XCHK_PPTR_EVENT(xchk_dir_defer);
+DEFINE_XCHK_PPTR_EVENT(xchk_dir_slowpath);
+DEFINE_XCHK_PPTR_EVENT(xchk_dir_ultraslowpath);
+
/* repair tracepoints */
#if IS_ENABLED(CONFIG_XFS_ONLINE_REPAIR)
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 03/22] xfs: create a parent pointer walk function for scrubbers
2023-12-31 19:33 ` [PATCHSET v13.0 5/7] xfs: fsck for parent pointers Darrick J. Wong
2023-12-31 20:53 ` [PATCH 01/22] xfs: check dirents have " Darrick J. Wong
2023-12-31 20:53 ` [PATCH 02/22] xfs: deferred scrub of dirents Darrick J. Wong
@ 2023-12-31 20:53 ` Darrick J. Wong
2023-12-31 20:54 ` [PATCH 04/22] xfs: scrub parent pointers Darrick J. Wong
` (18 subsequent siblings)
21 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:53 UTC (permalink / raw)
To: djwong; +Cc: catherine.hoang, allison.henderson, linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
Build a parent pointer iteration function off of the existing xattr
walking code. This will be used by subsequent patches.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/scrub/listxattr.c | 84 +++++++++++++++++++++++++++++++++++++++++++---
fs/xfs/scrub/listxattr.h | 9 +++++
2 files changed, 88 insertions(+), 5 deletions(-)
diff --git a/fs/xfs/scrub/listxattr.c b/fs/xfs/scrub/listxattr.c
index c8d7d7d723177..dc893f2cdc1c3 100644
--- a/fs/xfs/scrub/listxattr.c
+++ b/fs/xfs/scrub/listxattr.c
@@ -17,11 +17,46 @@
#include "xfs_attr_leaf.h"
#include "xfs_attr_sf.h"
#include "xfs_trans.h"
+#include "xfs_parent.h"
#include "scrub/scrub.h"
#include "scrub/bitmap.h"
#include "scrub/dab_bitmap.h"
#include "scrub/listxattr.h"
+struct xchk_pptr_walk {
+ struct xfs_parent_name_irec *pptr_buf;
+ xchk_pptr_fn fn;
+ void *priv;
+};
+
+/* Call the parent pointer callback if this xattr is a valid parent pointer. */
+STATIC int
+xchk_pptr_walk_attr(
+ struct xfs_scrub *sc,
+ struct xfs_inode *ip,
+ unsigned int attr_flags,
+ const unsigned char *name,
+ unsigned int namelen,
+ const void *value,
+ unsigned int valuelen,
+ void *priv)
+{
+ struct xchk_pptr_walk *pw = priv;
+ const struct xfs_parent_name_rec *rec = (const void *)name;
+
+ /* Ignore anything that isn't a parent pointer. */
+ if (!(attr_flags & XFS_ATTR_PARENT))
+ return 0;
+
+ if (!xfs_parent_namecheck(sc->mp, rec, namelen, attr_flags))
+ return -EFSCORRUPTED;
+ if (!xfs_parent_valuecheck(sc->mp, value, valuelen))
+ return -EFSCORRUPTED;
+
+ xfs_parent_irec_from_disk(pw->pptr_buf, rec, value, valuelen);
+ return pw->fn(sc, ip, pw->pptr_buf, pw->priv);
+}
+
/* Call a function for every entry in a shortform xattr structure. */
STATIC int
xchk_xattr_walk_sf(
@@ -37,9 +72,16 @@ xchk_xattr_walk_sf(
sf = (struct xfs_attr_shortform *)ip->i_af.if_u1.if_data;
for (i = 0, sfe = &sf->list[0]; i < sf->hdr.count; i++) {
- error = attr_fn(sc, ip, sfe->flags, sfe->nameval, sfe->namelen,
- &sfe->nameval[sfe->namelen], sfe->valuelen,
- priv);
+ if (attr_fn == xchk_pptr_walk_attr)
+ error = xchk_pptr_walk_attr(sc, ip, sfe->flags,
+ sfe->nameval, sfe->namelen,
+ &sfe->nameval[sfe->namelen],
+ sfe->valuelen, priv);
+ else
+ error = attr_fn(sc, ip, sfe->flags,
+ sfe->nameval, sfe->namelen,
+ &sfe->nameval[sfe->namelen],
+ sfe->valuelen, priv);
if (error)
return error;
@@ -91,8 +133,12 @@ xchk_xattr_walk_leaf_entries(
valuelen = be32_to_cpu(name_rmt->valuelen);
}
- error = attr_fn(sc, ip, entry->flags, name, namelen, value,
- valuelen, priv);
+ if (attr_fn == xchk_pptr_walk_attr)
+ error = xchk_pptr_walk_attr(sc, ip, entry->flags, name,
+ namelen, value, valuelen, priv);
+ else
+ error = attr_fn(sc, ip, entry->flags, name, namelen,
+ value, valuelen, priv);
if (error)
return error;
@@ -308,3 +354,31 @@ xchk_xattr_walk(
return xchk_xattr_walk_node(sc, ip, attr_fn, priv);
}
+
+/*
+ * Walk every parent pointer of this file. The parent pointer will be
+ * formatted into the provided @pptr_buf, which is then passed to the callback
+ * function.
+ *
+ * The callback function must decide if an invalid parent_ino or invalid name
+ * should halt the parent pointer walk; the only validation done here is the
+ * structure of the xattrs themselves.
+ */
+int
+xchk_pptr_walk(
+ struct xfs_scrub *sc,
+ struct xfs_inode *ip,
+ xchk_pptr_fn pptr_fn,
+ struct xfs_parent_name_irec *pptr_buf,
+ void *priv)
+{
+ struct xchk_pptr_walk pw = {
+ .fn = pptr_fn,
+ .pptr_buf = pptr_buf,
+ .priv = priv,
+ };
+
+ ASSERT(xfs_has_parent(sc->mp));
+
+ return xchk_xattr_walk(sc, ip, xchk_pptr_walk_attr, &pw);
+}
diff --git a/fs/xfs/scrub/listxattr.h b/fs/xfs/scrub/listxattr.h
index 48fe89d05946b..7e4bd3ae75e15 100644
--- a/fs/xfs/scrub/listxattr.h
+++ b/fs/xfs/scrub/listxattr.h
@@ -14,4 +14,13 @@ typedef int (*xchk_xattr_fn)(struct xfs_scrub *sc, struct xfs_inode *ip,
int xchk_xattr_walk(struct xfs_scrub *sc, struct xfs_inode *ip,
xchk_xattr_fn attr_fn, void *priv);
+struct xfs_parent_name_irec;
+
+typedef int (*xchk_pptr_fn)(struct xfs_scrub *sc, struct xfs_inode *ip,
+ const struct xfs_parent_name_irec *pptr, void *priv);
+
+int xchk_pptr_walk(struct xfs_scrub *sc, struct xfs_inode *ip,
+ xchk_pptr_fn pptr_fn, struct xfs_parent_name_irec *pptr_buf,
+ void *priv);
+
#endif /* __XFS_SCRUB_LISTXATTR_H__ */
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 04/22] xfs: scrub parent pointers
2023-12-31 19:33 ` [PATCHSET v13.0 5/7] xfs: fsck for parent pointers Darrick J. Wong
` (2 preceding siblings ...)
2023-12-31 20:53 ` [PATCH 03/22] xfs: create a parent pointer walk function for scrubbers Darrick J. Wong
@ 2023-12-31 20:54 ` Darrick J. Wong
2023-12-31 20:54 ` [PATCH 05/22] xfs: deferred scrub of " Darrick J. Wong
` (17 subsequent siblings)
21 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:54 UTC (permalink / raw)
To: djwong; +Cc: catherine.hoang, allison.henderson, linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
Actually check parent pointers now.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/scrub/parent.c | 385 +++++++++++++++++++++++++++++++++++++++++++++++++
1 file changed, 385 insertions(+)
diff --git a/fs/xfs/scrub/parent.c b/fs/xfs/scrub/parent.c
index acb6282c3d148..7a5c57cdf93e4 100644
--- a/fs/xfs/scrub/parent.c
+++ b/fs/xfs/scrub/parent.c
@@ -15,11 +15,15 @@
#include "xfs_icache.h"
#include "xfs_dir2.h"
#include "xfs_dir2_priv.h"
+#include "xfs_attr.h"
+#include "xfs_parent.h"
#include "scrub/scrub.h"
#include "scrub/common.h"
#include "scrub/readdir.h"
#include "scrub/tempfile.h"
#include "scrub/repair.h"
+#include "scrub/listxattr.h"
+#include "scrub/trace.h"
/* Set us up to scrub parents. */
int
@@ -197,6 +201,384 @@ xchk_parent_validate(
return error;
}
+/*
+ * Checking of Parent Pointers
+ * ===========================
+ *
+ * On filesystems with directory parent pointers, we check the referential
+ * integrity by visiting each parent pointer of a child file and checking that
+ * the directory referenced by the pointer actually has a dirent pointing
+ * forward to the child file.
+ */
+
+struct xchk_pptrs {
+ struct xfs_scrub *sc;
+
+ /* Scratch buffer for scanning pptr xattrs */
+ struct xfs_parent_name_irec pptr;
+
+ /* How many parent pointers did we find at the end? */
+ unsigned long long pptrs_found;
+
+ /* Parent of this directory. */
+ xfs_ino_t parent_ino;
+};
+
+/* Does this parent pointer match the dotdot entry? */
+STATIC int
+xchk_parent_scan_dotdot(
+ struct xfs_scrub *sc,
+ struct xfs_inode *ip,
+ const struct xfs_parent_name_irec *pptr,
+ void *priv)
+{
+ struct xchk_pptrs *pp = priv;
+
+ if (pp->parent_ino == pptr->p_ino &&
+ xfs_parent_verify_irec(sc->mp, pptr))
+ return -ECANCELED;
+
+ return 0;
+}
+
+/* Look up the dotdot entry so that we can check it as we walk the pptrs. */
+STATIC int
+xchk_parent_pptr_and_dotdot(
+ struct xchk_pptrs *pp)
+{
+ struct xfs_scrub *sc = pp->sc;
+ int error;
+
+ /* Look up '..' */
+ error = xchk_dir_lookup(sc, sc->ip, &xfs_name_dotdot, &pp->parent_ino);
+ if (!xchk_fblock_process_error(sc, XFS_DATA_FORK, 0, &error))
+ return error;
+ if (!xfs_verify_dir_ino(sc->mp, pp->parent_ino)) {
+ xchk_fblock_set_corrupt(sc, XFS_DATA_FORK, 0);
+ return 0;
+ }
+
+ /* Is this the root dir? Then '..' must point to itself. */
+ if (sc->ip == sc->mp->m_rootip) {
+ if (sc->ip->i_ino != pp->parent_ino)
+ xchk_fblock_set_corrupt(sc, XFS_DATA_FORK, 0);
+ return 0;
+ }
+
+ /*
+ * If this is now an unlinked directory, the dotdot value is
+ * meaningless as long as it points to a valid inode.
+ */
+ if (VFS_I(sc->ip)->i_nlink == 0)
+ return 0;
+
+ if (pp->sc->sm->sm_flags & XFS_SCRUB_OFLAG_CORRUPT)
+ return 0;
+
+ /* Otherwise, walk the pptrs again, and check. */
+ error = xchk_pptr_walk(sc, sc->ip, xchk_parent_scan_dotdot, &pp->pptr,
+ pp);
+ if (error == -ECANCELED) {
+ /* Found a parent pointer that matches dotdot. */
+ return 0;
+ }
+ if (!error || error == -EFSCORRUPTED) {
+ /* Found a broken parent pointer or no match. */
+ xchk_fblock_set_corrupt(sc, XFS_ATTR_FORK, 0);
+ return 0;
+ }
+ return error;
+}
+
+/*
+ * Try to lock a parent directory for checking dirents. Returns the inode
+ * flags for the locks we now hold, or zero if we failed.
+ */
+STATIC unsigned int
+xchk_parent_lock_dir(
+ struct xfs_scrub *sc,
+ struct xfs_inode *dp)
+{
+ if (!xfs_ilock_nowait(dp, XFS_IOLOCK_SHARED))
+ return 0;
+
+ if (!xfs_ilock_nowait(dp, XFS_ILOCK_SHARED)) {
+ xfs_iunlock(dp, XFS_IOLOCK_SHARED);
+ return 0;
+ }
+
+ if (!xfs_need_iread_extents(&dp->i_df))
+ return XFS_IOLOCK_SHARED | XFS_ILOCK_SHARED;
+
+ xfs_iunlock(dp, XFS_ILOCK_SHARED);
+
+ if (!xfs_ilock_nowait(dp, XFS_ILOCK_EXCL)) {
+ xfs_iunlock(dp, XFS_IOLOCK_SHARED);
+ return 0;
+ }
+
+ return XFS_IOLOCK_SHARED | XFS_ILOCK_EXCL;
+}
+
+/* Check the forward link (dirent) associated with this parent pointer. */
+STATIC int
+xchk_parent_dirent(
+ struct xchk_pptrs *pp,
+ struct xfs_inode *dp)
+{
+ struct xfs_name xname = {
+ .name = pp->pptr.p_name,
+ .len = pp->pptr.p_namelen,
+ };
+ struct xfs_scrub *sc = pp->sc;
+ xfs_ino_t child_ino;
+ int error;
+
+ /*
+ * Use the name attached to this parent pointer to look up the
+ * directory entry in the alleged parent.
+ */
+ error = xchk_dir_lookup(sc, dp, &xname, &child_ino);
+ if (error == -ENOENT) {
+ xchk_fblock_xref_set_corrupt(sc, XFS_ATTR_FORK, 0);
+ return 0;
+ }
+ if (!xchk_fblock_xref_process_error(sc, XFS_ATTR_FORK, 0, &error))
+ return error;
+
+ /* Does the inode number match? */
+ if (child_ino != sc->ip->i_ino) {
+ xchk_fblock_xref_set_corrupt(sc, XFS_ATTR_FORK, 0);
+ return 0;
+ }
+
+ return 0;
+}
+
+/* Try to grab a parent directory. */
+STATIC int
+xchk_parent_iget(
+ struct xchk_pptrs *pp,
+ struct xfs_inode **dpp)
+{
+ struct xfs_scrub *sc = pp->sc;
+ struct xfs_inode *ip;
+ int error;
+
+ /* Validate inode number. */
+ error = xfs_dir_ino_validate(sc->mp, pp->pptr.p_ino);
+ if (error) {
+ xchk_fblock_set_corrupt(sc, XFS_ATTR_FORK, 0);
+ return -ECANCELED;
+ }
+
+ error = xchk_iget(sc, pp->pptr.p_ino, &ip);
+ if (error == -EINVAL || error == -ENOENT) {
+ xchk_fblock_set_corrupt(sc, XFS_ATTR_FORK, 0);
+ return -ECANCELED;
+ }
+ if (!xchk_fblock_xref_process_error(sc, XFS_ATTR_FORK, 0, &error))
+ return error;
+
+ /* The parent must be a directory. */
+ if (!S_ISDIR(VFS_I(ip)->i_mode)) {
+ xchk_fblock_xref_set_corrupt(sc, XFS_ATTR_FORK, 0);
+ goto out_rele;
+ }
+
+ /* Validate generation number. */
+ if (VFS_I(ip)->i_generation != pp->pptr.p_gen) {
+ xchk_fblock_xref_set_corrupt(sc, XFS_ATTR_FORK, 0);
+ goto out_rele;
+ }
+
+ *dpp = ip;
+ return 0;
+out_rele:
+ xchk_irele(sc, ip);
+ return 0;
+}
+
+/*
+ * Walk an xattr of a file. If this xattr is a parent pointer, follow it up
+ * to a parent directory and check that the parent has a dirent pointing back
+ * to us.
+ */
+STATIC int
+xchk_parent_scan_attr(
+ struct xfs_scrub *sc,
+ struct xfs_inode *ip,
+ unsigned int attr_flags,
+ const unsigned char *name,
+ unsigned int namelen,
+ const void *value,
+ unsigned int valuelen,
+ void *priv)
+{
+ struct xfs_name dname = {
+ .name = value,
+ .len = valuelen,
+ };
+ struct xchk_pptrs *pp = priv;
+ struct xfs_inode *dp = NULL;
+ const struct xfs_parent_name_rec *rec = (const void *)name;
+ unsigned int lockmode;
+ xfs_dahash_t computed_hash;
+ int error;
+
+ /* Ignore anything that isn't a parent pointer. */
+ if (!(attr_flags & XFS_ATTR_PARENT))
+ return 0;
+
+ /* Does the ondisk parent pointer structure make sense? */
+ if (!xfs_parent_namecheck(sc->mp, rec, namelen, attr_flags)) {
+ xchk_fblock_set_corrupt(sc, XFS_ATTR_FORK, 0);
+ return -ECANCELED;
+ }
+
+ if (!xfs_parent_valuecheck(sc->mp, value, valuelen)) {
+ xchk_fblock_set_corrupt(sc, XFS_ATTR_FORK, 0);
+ return -ECANCELED;
+ }
+
+ xfs_parent_irec_from_disk(&pp->pptr, rec, value, valuelen);
+
+ if (!xfs_parent_verify_irec(sc->mp, &pp->pptr)) {
+ xchk_fblock_set_corrupt(sc, XFS_ATTR_FORK, 0);
+ return -ECANCELED;
+ }
+
+ /* No self-referential parent pointers. */
+ if (pp->pptr.p_ino == sc->ip->i_ino) {
+ xchk_fblock_set_corrupt(sc, XFS_ATTR_FORK, 0);
+ return -ECANCELED;
+ }
+
+ /*
+ * If the namehash of the dirent name encoded in the parent pointer
+ * attr value doesn't match the namehash in the parent pointer key,
+ * the parent pointer is corrupt.
+ */
+ computed_hash = xfs_dir2_hashname(ip->i_mount, &dname);
+ if (pp->pptr.p_namehash != computed_hash) {
+ xchk_fblock_set_corrupt(sc, XFS_ATTR_FORK, 0);
+ return -ECANCELED;
+ }
+ pp->pptrs_found++;
+
+ error = xchk_parent_iget(pp, &dp);
+ if (error)
+ return error;
+ if (!dp)
+ return 0;
+
+ /* Try to lock the inode. */
+ lockmode = xchk_parent_lock_dir(sc, dp);
+ if (!lockmode) {
+ xchk_set_incomplete(sc);
+ error = -ECANCELED;
+ goto out_rele;
+ }
+
+ error = xchk_parent_dirent(pp, dp);
+ if (error)
+ goto out_unlock;
+
+out_unlock:
+ xfs_iunlock(dp, lockmode);
+out_rele:
+ xchk_irele(sc, dp);
+ return error;
+}
+
+/*
+ * Compare the number of parent pointers to the link count. For
+ * non-directories these should be the same. For unlinked directories the
+ * count should be zero; for linked directories, it should be nonzero.
+ */
+STATIC int
+xchk_parent_count_pptrs(
+ struct xchk_pptrs *pp)
+{
+ struct xfs_scrub *sc = pp->sc;
+
+ if (S_ISDIR(VFS_I(sc->ip)->i_mode)) {
+ if (sc->ip == sc->mp->m_rootip)
+ pp->pptrs_found++;
+
+ if (VFS_I(sc->ip)->i_nlink == 0 && pp->pptrs_found > 0)
+ xchk_ino_set_corrupt(sc, sc->ip->i_ino);
+ else if (VFS_I(sc->ip)->i_nlink > 0 &&
+ pp->pptrs_found == 0)
+ xchk_ino_set_corrupt(sc, sc->ip->i_ino);
+ } else {
+ if (VFS_I(sc->ip)->i_nlink != pp->pptrs_found)
+ xchk_ino_set_corrupt(sc, sc->ip->i_ino);
+ }
+
+ return 0;
+}
+
+/* Check parent pointers of a file. */
+STATIC int
+xchk_parent_pptr(
+ struct xfs_scrub *sc)
+{
+ struct xchk_pptrs *pp;
+ int error;
+
+ pp = kvzalloc(sizeof(struct xchk_pptrs), XCHK_GFP_FLAGS);
+ if (!pp)
+ return -ENOMEM;
+ pp->sc = sc;
+
+ error = xchk_xattr_walk(sc, sc->ip, xchk_parent_scan_attr, pp);
+ if (error == -ECANCELED) {
+ error = 0;
+ goto out_pp;
+ }
+ if (error)
+ goto out_pp;
+
+ if (pp->sc->sm->sm_flags & XFS_SCRUB_OFLAG_CORRUPT)
+ goto out_pp;
+
+ /*
+ * For subdirectories, make sure the dotdot entry references the same
+ * inode as the parent pointers.
+ *
+ * If we're scanning a /consistent/ directory, there should only be
+ * one parent pointer, and it should point to the same directory as
+ * the dotdot entry.
+ *
+ * However, a corrupt directory tree might feature a subdirectory with
+ * multiple parents. The directory loop scanner is responsible for
+ * correcting that kind of problem, so for now we only validate that
+ * the dotdot entry matches /one/ of the parents.
+ */
+ if (S_ISDIR(VFS_I(sc->ip)->i_mode)) {
+ error = xchk_parent_pptr_and_dotdot(pp);
+ if (error)
+ goto out_pp;
+ }
+
+ if (pp->sc->sm->sm_flags & XFS_SCRUB_OFLAG_CORRUPT)
+ goto out_pp;
+
+ /*
+ * Complain if the number of parent pointers doesn't match the link
+ * count. This could be a sign of missing parent pointers (or an
+ * incorrect link count).
+ */
+ error = xchk_parent_count_pptrs(pp);
+ if (error)
+ goto out_pp;
+
+out_pp:
+ kvfree(pp);
+ return error;
+}
+
/* Scrub a parent pointer. */
int
xchk_parent(
@@ -206,6 +588,9 @@ xchk_parent(
xfs_ino_t parent_ino;
int error = 0;
+ if (xfs_has_parent(mp))
+ return xchk_parent_pptr(sc);
+
/*
* If we're a directory, check that the '..' link points up to
* a directory that has one entry pointing to us.
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 05/22] xfs: deferred scrub of parent pointers
2023-12-31 19:33 ` [PATCHSET v13.0 5/7] xfs: fsck for parent pointers Darrick J. Wong
` (3 preceding siblings ...)
2023-12-31 20:54 ` [PATCH 04/22] xfs: scrub parent pointers Darrick J. Wong
@ 2023-12-31 20:54 ` Darrick J. Wong
2023-12-31 20:54 ` [PATCH 06/22] xfs: walk directory parent pointers to determine backref count Darrick J. Wong
` (16 subsequent siblings)
21 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:54 UTC (permalink / raw)
To: djwong; +Cc: catherine.hoang, allison.henderson, linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
If the trylock-based dirent check fails, retain those parent pointers
and check them at the end. This may involve dropping the locks on the
file being scanned, so yay.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/Makefile | 2
fs/xfs/scrub/parent.c | 263 ++++++++++++++++++++++++++++++++++++++++++++++++-
fs/xfs/scrub/trace.h | 3 +
3 files changed, 260 insertions(+), 8 deletions(-)
diff --git a/fs/xfs/Makefile b/fs/xfs/Makefile
index 49480c81eaeab..52ef808359966 100644
--- a/fs/xfs/Makefile
+++ b/fs/xfs/Makefile
@@ -177,6 +177,7 @@ xfs-y += $(addprefix scrub/, \
scrub.o \
symlink.o \
xfarray.o \
+ xfblob.o \
xfile.o \
)
@@ -218,7 +219,6 @@ xfs-y += $(addprefix scrub/, \
rmap_repair.o \
symlink_repair.o \
tempfile.o \
- xfblob.o \
xfbtree.o \
)
diff --git a/fs/xfs/scrub/parent.c b/fs/xfs/scrub/parent.c
index 7a5c57cdf93e4..3bacd3e14f5d3 100644
--- a/fs/xfs/scrub/parent.c
+++ b/fs/xfs/scrub/parent.c
@@ -23,6 +23,9 @@
#include "scrub/tempfile.h"
#include "scrub/repair.h"
#include "scrub/listxattr.h"
+#include "scrub/xfile.h"
+#include "scrub/xfarray.h"
+#include "scrub/xfblob.h"
#include "scrub/trace.h"
/* Set us up to scrub parents. */
@@ -211,17 +214,42 @@ xchk_parent_validate(
* forward to the child file.
*/
+/* Deferred parent pointer entry that we saved for later. */
+struct xchk_pptr {
+ /* Cookie for retrieval of the pptr name. */
+ xfblob_cookie name_cookie;
+
+ /* Parent pointer attr key. */
+ xfs_ino_t p_ino;
+ uint32_t p_gen;
+
+ /* Length of the pptr name. */
+ uint8_t namelen;
+};
+
struct xchk_pptrs {
struct xfs_scrub *sc;
/* Scratch buffer for scanning pptr xattrs */
struct xfs_parent_name_irec pptr;
+ /* Fixed-size array of xchk_pptr structures. */
+ struct xfarray *pptr_entries;
+
+ /* Blobs containing parent pointer names. */
+ struct xfblob *pptr_names;
+
/* How many parent pointers did we find at the end? */
unsigned long long pptrs_found;
/* Parent of this directory. */
xfs_ino_t parent_ino;
+
+ /* If we've cycled the ILOCK, we must revalidate all deferred pptrs. */
+ bool need_revalidate;
+
+ /* xattr key and da args for parent pointer revalidation. */
+ struct xfs_parent_scratch pptr_scratch;
};
/* Does this parent pointer match the dotdot entry? */
@@ -475,8 +503,27 @@ xchk_parent_scan_attr(
/* Try to lock the inode. */
lockmode = xchk_parent_lock_dir(sc, dp);
if (!lockmode) {
- xchk_set_incomplete(sc);
- error = -ECANCELED;
+ struct xchk_pptr save_pp = {
+ .p_ino = pp->pptr.p_ino,
+ .p_gen = pp->pptr.p_gen,
+ .namelen = pp->pptr.p_namelen,
+ };
+
+ /* Couldn't lock the inode, so save the pptr for later. */
+ trace_xchk_parent_defer(sc->ip, pp->pptr.p_name,
+ pp->pptr.p_namelen, dp->i_ino);
+
+ error = xfblob_store(pp->pptr_names, &save_pp.name_cookie,
+ pp->pptr.p_name, pp->pptr.p_namelen);
+ if (!xchk_fblock_xref_process_error(sc, XFS_ATTR_FORK, 0,
+ &error))
+ goto out_rele;
+
+ error = xfarray_append(pp->pptr_entries, &save_pp);
+ if (!xchk_fblock_xref_process_error(sc, XFS_ATTR_FORK, 0,
+ &error))
+ goto out_rele;
+
goto out_rele;
}
@@ -491,6 +538,159 @@ xchk_parent_scan_attr(
return error;
}
+/*
+ * Revalidate a parent pointer that we collected in the past but couldn't check
+ * because of lock contention. Returns 0 if the parent pointer is still valid,
+ * -ENOENT if it has gone away on us, or a negative errno.
+ */
+STATIC int
+xchk_parent_revalidate_pptr(
+ struct xchk_pptrs *pp)
+{
+ struct xfs_scrub *sc = pp->sc;
+ int error;
+
+ error = xfs_parent_lookup(sc->tp, sc->ip, &pp->pptr,
+ &pp->pptr_scratch);
+ if (error == -ENOATTR) {
+ /* Parent pointer went away, nothing to revalidate. */
+ return -ENOENT;
+ }
+
+ return error;
+}
+
+/*
+ * Check a parent pointer the slow way, which means we cycle locks a bunch
+ * and put up with revalidation until we get it done.
+ */
+STATIC int
+xchk_parent_slow_pptr(
+ struct xchk_pptrs *pp,
+ struct xchk_pptr *pptr)
+{
+ struct xfs_scrub *sc = pp->sc;
+ struct xfs_inode *dp = NULL;
+ unsigned int lockmode;
+ int error;
+
+ /* Restore the saved parent pointer into the irec. */
+ pp->pptr.p_ino = pptr->p_ino;
+ pp->pptr.p_gen = pptr->p_gen;
+
+ error = xfblob_load(pp->pptr_names, pptr->name_cookie, pp->pptr.p_name,
+ pptr->namelen);
+ if (error)
+ return error;
+ pp->pptr.p_name[MAXNAMELEN - 1] = 0;
+ pp->pptr.p_namelen = pptr->namelen;
+ xfs_parent_irec_hashname(sc->mp, &pp->pptr);
+
+ /* Check that the deferred parent pointer still exists. */
+ if (pp->need_revalidate) {
+ error = xchk_parent_revalidate_pptr(pp);
+ if (error == -ENOENT)
+ return 0;
+ if (!xchk_fblock_xref_process_error(sc, XFS_ATTR_FORK, 0,
+ &error))
+ return error;
+ }
+
+ error = xchk_parent_iget(pp, &dp);
+ if (error)
+ return error;
+ if (!dp)
+ return 0;
+
+ /*
+ * If we can grab both IOLOCK and ILOCK of the alleged parent, we
+ * can proceed with the validation.
+ */
+ lockmode = xchk_parent_lock_dir(sc, dp);
+ if (lockmode) {
+ trace_xchk_parent_slowpath(sc->ip, pp->pptr.p_name,
+ pptr->namelen, dp->i_ino);
+ goto check_dirent;
+ }
+
+ /*
+ * We couldn't lock the parent dir. Drop all the locks and try to
+ * get them again, one at a time.
+ */
+ xchk_iunlock(sc, sc->ilock_flags);
+ pp->need_revalidate = true;
+
+ trace_xchk_parent_ultraslowpath(sc->ip, pp->pptr.p_name, pptr->namelen,
+ dp->i_ino);
+
+ error = xchk_dir_trylock_for_pptrs(sc, dp, &lockmode);
+ if (error)
+ goto out_rele;
+
+ /* Revalidate the parent pointer now that we cycled locks. */
+ error = xchk_parent_revalidate_pptr(pp);
+ if (error == -ENOENT) {
+ error = 0;
+ goto out_unlock;
+ }
+ if (!xchk_fblock_xref_process_error(sc, XFS_ATTR_FORK, 0, &error))
+ goto out_unlock;
+
+check_dirent:
+ error = xchk_parent_dirent(pp, dp);
+out_unlock:
+ xfs_iunlock(dp, lockmode);
+out_rele:
+ xchk_irele(sc, dp);
+ return error;
+}
+
+/* Check all the parent pointers that we deferred the first time around. */
+STATIC int
+xchk_parent_finish_slow_pptrs(
+ struct xchk_pptrs *pp)
+{
+ xfarray_idx_t array_cur;
+ int error;
+
+ foreach_xfarray_idx(pp->pptr_entries, array_cur) {
+ struct xchk_pptr pptr;
+
+ if (pp->sc->sm->sm_flags & XFS_SCRUB_OFLAG_CORRUPT)
+ return 0;
+
+ error = xfarray_load(pp->pptr_entries, array_cur, &pptr);
+ if (error)
+ return error;
+
+ error = xchk_parent_slow_pptr(pp, &pptr);
+ if (error)
+ return error;
+ }
+
+ /* Empty out both xfiles now that we've checked everything. */
+ xfarray_truncate(pp->pptr_entries);
+ xfblob_truncate(pp->pptr_names);
+ return 0;
+}
+
+/* Count the number of parent pointers. */
+STATIC int
+xchk_parent_count_pptr(
+ struct xfs_scrub *sc,
+ struct xfs_inode *ip,
+ const struct xfs_parent_name_irec *pptr,
+ void *priv)
+{
+ struct xchk_pptrs *pp = priv;
+
+ if (!xfs_parent_verify_irec(sc->mp, pptr))
+ return -EFSCORRUPTED;
+
+ pp->pptrs_found++;
+ return 0;
+}
+
/*
* Compare the number of parent pointers to the link count. For
* non-directories these should be the same. For unlinked directories the
@@ -501,6 +701,24 @@ xchk_parent_count_pptrs(
struct xchk_pptrs *pp)
{
struct xfs_scrub *sc = pp->sc;
+ int error;
+
+ /*
+ * If we cycled the ILOCK while cross-checking parent pointers with
+ * dirents, then we need to recalculate the number of parent pointers.
+ */
+ if (pp->need_revalidate) {
+ pp->pptrs_found = 0;
+ error = xchk_pptr_walk(sc, sc->ip, xchk_parent_count_pptr,
+ &pp->pptr, pp);
+ if (error == -EFSCORRUPTED) {
+ /* Found a bad parent pointer */
+ xchk_fblock_set_corrupt(sc, XFS_ATTR_FORK, 0);
+ return 0;
+ }
+ if (error)
+ return error;
+ }
if (S_ISDIR(VFS_I(sc->ip)->i_mode)) {
if (sc->ip == sc->mp->m_rootip)
@@ -525,6 +743,7 @@ xchk_parent_pptr(
struct xfs_scrub *sc)
{
struct xchk_pptrs *pp;
+ char *descr;
int error;
pp = kvzalloc(sizeof(struct xchk_pptrs), XCHK_GFP_FLAGS);
@@ -532,16 +751,42 @@ xchk_parent_pptr(
return -ENOMEM;
pp->sc = sc;
+ /*
+ * Set up some staging memory for parent pointers that we can't check
+ * due to locking contention.
+ */
+ descr = xchk_xfile_ino_descr(sc, "slow parent pointer entries");
+ error = xfarray_create(descr, 0, sizeof(struct xchk_pptr),
+ &pp->pptr_entries);
+ kfree(descr);
+ if (error)
+ goto out_pp;
+
+ descr = xchk_xfile_ino_descr(sc, "slow parent pointer names");
+ error = xfblob_create(descr, &pp->pptr_names);
+ kfree(descr);
+ if (error)
+ goto out_entries;
+
error = xchk_xattr_walk(sc, sc->ip, xchk_parent_scan_attr, pp);
if (error == -ECANCELED) {
error = 0;
- goto out_pp;
+ goto out_names;
}
if (error)
- goto out_pp;
+ goto out_names;
+
+ error = xchk_parent_finish_slow_pptrs(pp);
+ if (error == -ETIMEDOUT) {
+ /* Couldn't grab a lock, scrub was marked incomplete */
+ error = 0;
+ goto out_names;
+ }
+ if (error)
+ goto out_names;
if (pp->sc->sm->sm_flags & XFS_SCRUB_OFLAG_CORRUPT)
- goto out_pp;
+ goto out_names;
/*
* For subdirectories, make sure the dotdot entry references the same
@@ -559,7 +804,7 @@ xchk_parent_pptr(
if (S_ISDIR(VFS_I(sc->ip)->i_mode)) {
error = xchk_parent_pptr_and_dotdot(pp);
if (error)
- goto out_pp;
+ goto out_names;
}
if (pp->sc->sm->sm_flags & XFS_SCRUB_OFLAG_CORRUPT)
@@ -572,8 +817,12 @@ xchk_parent_pptr(
*/
error = xchk_parent_count_pptrs(pp);
if (error)
- goto out_pp;
+ goto out_names;
+out_names:
+ xfblob_destroy(pp->pptr_names);
+out_entries:
+ xfarray_destroy(pp->pptr_entries);
out_pp:
kvfree(pp);
return error;
diff --git a/fs/xfs/scrub/trace.h b/fs/xfs/scrub/trace.h
index 651b73d33f2c4..f90743453cd22 100644
--- a/fs/xfs/scrub/trace.h
+++ b/fs/xfs/scrub/trace.h
@@ -1556,6 +1556,9 @@ DEFINE_EVENT(xchk_pptr_class, name, \
DEFINE_XCHK_PPTR_EVENT(xchk_dir_defer);
DEFINE_XCHK_PPTR_EVENT(xchk_dir_slowpath);
DEFINE_XCHK_PPTR_EVENT(xchk_dir_ultraslowpath);
+DEFINE_XCHK_PPTR_EVENT(xchk_parent_defer);
+DEFINE_XCHK_PPTR_EVENT(xchk_parent_slowpath);
+DEFINE_XCHK_PPTR_EVENT(xchk_parent_ultraslowpath);
/* repair tracepoints */
#if IS_ENABLED(CONFIG_XFS_ONLINE_REPAIR)
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 06/22] xfs: walk directory parent pointers to determine backref count
2023-12-31 19:33 ` [PATCHSET v13.0 5/7] xfs: fsck for parent pointers Darrick J. Wong
` (4 preceding siblings ...)
2023-12-31 20:54 ` [PATCH 05/22] xfs: deferred scrub of " Darrick J. Wong
@ 2023-12-31 20:54 ` Darrick J. Wong
2023-12-31 20:55 ` [PATCH 07/22] xfs: add raw parent pointer apis to support repair Darrick J. Wong
` (15 subsequent siblings)
21 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:54 UTC (permalink / raw)
To: djwong; +Cc: catherine.hoang, allison.henderson, linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
If the filesystem has parent pointers enabled, walk the parent pointers
of subdirectories to determine the true backref count. In theory each
subdir should have a single parent reachable via dotdot, but in the case
of (corrupt) subdirs with multiple parents, we need to keep the link
counts high enough that the directory loop detector will be able to
correct the multiple parents problems.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/scrub/common.h | 1 +
fs/xfs/scrub/nlinks.c | 71 +++++++++++++++++++++++++++++++++++++++++-
fs/xfs/scrub/nlinks.h | 3 ++
fs/xfs/scrub/nlinks_repair.c | 2 +
fs/xfs/scrub/parent.c | 61 ++++++++++++++++++++++++++++++++++++
fs/xfs/scrub/trace.c | 1 +
fs/xfs/scrub/trace.h | 27 ++++++++++++++++
7 files changed, 165 insertions(+), 1 deletion(-)
diff --git a/fs/xfs/scrub/common.h b/fs/xfs/scrub/common.h
index 2e6af46519b58..298669ca2eb92 100644
--- a/fs/xfs/scrub/common.h
+++ b/fs/xfs/scrub/common.h
@@ -212,6 +212,7 @@ static inline bool xchk_skip_xref(struct xfs_scrub_metadata *sm)
}
bool xchk_dir_looks_zapped(struct xfs_inode *dp);
+bool xchk_pptr_looks_zapped(struct xfs_inode *ip);
#ifdef CONFIG_XFS_ONLINE_REPAIR
/* Decide if a repair is required. */
diff --git a/fs/xfs/scrub/nlinks.c b/fs/xfs/scrub/nlinks.c
index 6f0b77da14dbb..4e62e287e1590 100644
--- a/fs/xfs/scrub/nlinks.c
+++ b/fs/xfs/scrub/nlinks.c
@@ -18,6 +18,7 @@
#include "xfs_dir2.h"
#include "xfs_dir2_priv.h"
#include "xfs_ag.h"
+#include "xfs_parent.h"
#include "scrub/scrub.h"
#include "scrub/common.h"
#include "scrub/repair.h"
@@ -29,6 +30,7 @@
#include "scrub/trace.h"
#include "scrub/readdir.h"
#include "scrub/tempfile.h"
+#include "scrub/listxattr.h"
/*
* Live Inode Link Count Checking
@@ -268,12 +270,17 @@ xchk_nlinks_collect_dirent(
* number of parents of the root directory.
*
* Otherwise, increment the number of backrefs pointing back to ino.
+ *
+ * If the filesystem has parent pointers, we walk the pptrs to
+ * determine the backref count.
*/
if (dotdot) {
if (dp == sc->mp->m_rootip)
error = xchk_nlinks_update_incore(xnc, ino, 1, 0, 0);
- else
+ else if (!xfs_has_parent(sc->mp))
error = xchk_nlinks_update_incore(xnc, ino, 0, 1, 0);
+ else
+ error = 0;
if (error)
goto out_unlock;
}
@@ -310,6 +317,46 @@ xchk_nlinks_collect_dirent(
return error;
}
+/* Bump the backref count for the inode referenced by this parent pointer. */
+STATIC int
+xchk_nlinks_collect_pptr(
+ struct xfs_scrub *sc,
+ struct xfs_inode *ip,
+ const struct xfs_parent_name_irec *pptr,
+ void *priv)
+{
+ struct xchk_nlink_ctrs *xnc = priv;
+ int error;
+
+ if (!xfs_parent_verify_irec(sc->mp, pptr))
+ return -EFSCORRUPTED;
+
+ /* Update the shadow link counts if we haven't already failed. */
+
+ if (xchk_iscan_aborted(&xnc->collect_iscan)) {
+ error = -ECANCELED;
+ goto out_incomplete;
+ }
+
+ trace_xchk_nlinks_collect_pptr(sc->mp, ip, pptr);
+
+ mutex_lock(&xnc->lock);
+
+ error = xchk_nlinks_update_incore(xnc, pptr->p_ino, 0, 1, 0);
+ if (error)
+ goto out_unlock;
+
+ mutex_unlock(&xnc->lock);
+ return 0;
+
+out_unlock:
+ mutex_unlock(&xnc->lock);
+ xchk_iscan_abort(&xnc->collect_iscan);
+out_incomplete:
+ xchk_set_incomplete(sc);
+ return error;
+}
+
/* Walk a directory to bump the observed link counts of the children. */
STATIC int
xchk_nlinks_collect_dir(
@@ -356,6 +403,28 @@ xchk_nlinks_collect_dir(
if (error)
goto out_abort;
+ /* Walk the parent pointers to get real backref counts. */
+ if (xfs_has_parent(sc->mp)) {
+ /*
+ * If the extended attributes look as though they has been
+ * zapped by the inode record repair code, we cannot scan for
+ * parent pointers.
+ */
+ if (xchk_pptr_looks_zapped(dp)) {
+ error = -EBUSY;
+ goto out_unlock;
+ }
+
+ error = xchk_pptr_walk(sc, dp, xchk_nlinks_collect_pptr,
+ &xnc->pptr, xnc);
+ if (error == -ECANCELED) {
+ error = 0;
+ goto out_unlock;
+ }
+ if (error)
+ goto out_abort;
+ }
+
xchk_iscan_mark_visited(&xnc->collect_iscan, dp);
goto out_unlock;
diff --git a/fs/xfs/scrub/nlinks.h b/fs/xfs/scrub/nlinks.h
index f4766e01b6469..2d63cb56b6a3c 100644
--- a/fs/xfs/scrub/nlinks.h
+++ b/fs/xfs/scrub/nlinks.h
@@ -23,6 +23,9 @@ struct xchk_nlink_ctrs {
struct xchk_iscan collect_iscan;
struct xchk_iscan compare_iscan;
+ /* Parent pointer for finding backrefs. */
+ struct xfs_parent_name_irec pptr;
+
/*
* Hook into directory updates so that we can receive live updates
* from other writer threads.
diff --git a/fs/xfs/scrub/nlinks_repair.c b/fs/xfs/scrub/nlinks_repair.c
index 87cb3400ff948..fb299b23d5f1d 100644
--- a/fs/xfs/scrub/nlinks_repair.c
+++ b/fs/xfs/scrub/nlinks_repair.c
@@ -18,6 +18,8 @@
#include "xfs_ialloc.h"
#include "xfs_sb.h"
#include "xfs_ag.h"
+#include "xfs_dir2.h"
+#include "xfs_parent.h"
#include "scrub/scrub.h"
#include "scrub/common.h"
#include "scrub/repair.h"
diff --git a/fs/xfs/scrub/parent.c b/fs/xfs/scrub/parent.c
index 3bacd3e14f5d3..555aee4b73b37 100644
--- a/fs/xfs/scrub/parent.c
+++ b/fs/xfs/scrub/parent.c
@@ -883,3 +883,64 @@ xchk_parent(
return error;
}
+
+/*
+ * Decide if this file's extended attributes (and therefore its parent
+ * pointers) have been zapped to satisfy the inode and ifork verifiers.
+ * Checking and repairing should be postponed until the extended attribute
+ * structure is fixed.
+ */
+bool
+xchk_pptr_looks_zapped(
+ struct xfs_inode *ip)
+{
+ struct xfs_mount *mp = ip->i_mount;
+ struct inode *inode = VFS_I(ip);
+
+ ASSERT(xfs_has_parent(mp));
+
+ /*
+ * Temporary files that cannot be linked into the directory tree do not
+ * have attr forks because they cannot ever have parents.
+ */
+ if (inode->i_nlink == 0 && !(inode->i_state & I_LINKABLE))
+ return false;
+
+ /*
+ * Directory tree roots do not have parents, so the expected outcome
+ * of a parent pointer scan is always the empty set. It's safe to scan
+ * them even if the attr fork was zapped.
+ */
+ if (ip == mp->m_rootip)
+ return false;
+
+ /*
+ * Metadata inodes are all rooted in the superblock and do not have
+ * any parents. Hence the attr fork will not be initialized, but
+ * there are no parent pointers that might have been zapped.
+ */
+ if (xfs_is_metadata_inode(ip))
+ return false;
+
+ /*
+ * Linked and linkable non-rootdir files should always have an
+ * attribute fork because that is where parent pointers are
+ * stored. If the fork is absent, something is amiss.
+ */
+ if (!xfs_inode_has_attr_fork(ip))
+ return true;
+
+ /* Repair zapped this file's attr fork a short time ago */
+ if (xfs_ifork_zapped(ip, XFS_ATTR_FORK))
+ return true;
+
+ /*
+ * If the dinode repair found a bad attr fork, it will reset the fork
+ * to extents format with zero records and wait for the bmapbta
+ * scrubber to reconstruct the block mappings. The extended attribute
+ * structure always contain some content when parent pointers are
+ * enabled, so this is a clear sign of a zapped attr fork.
+ */
+ return ip->i_af.if_format == XFS_DINODE_FMT_EXTENTS &&
+ ip->i_af.if_nextents == 0;
+}
diff --git a/fs/xfs/scrub/trace.c b/fs/xfs/scrub/trace.c
index e127f6d492c35..9fe1491adbb51 100644
--- a/fs/xfs/scrub/trace.c
+++ b/fs/xfs/scrub/trace.c
@@ -20,6 +20,7 @@
#include "xfs_da_format.h"
#include "xfs_dir2.h"
#include "xfs_rmap.h"
+#include "xfs_parent.h"
#include "scrub/scrub.h"
#include "scrub/xfile.h"
#include "scrub/xfarray.h"
diff --git a/fs/xfs/scrub/trace.h b/fs/xfs/scrub/trace.h
index f90743453cd22..d3a0cefea3684 100644
--- a/fs/xfs/scrub/trace.h
+++ b/fs/xfs/scrub/trace.h
@@ -28,6 +28,7 @@ struct xchk_fscounters;
struct xfbtree;
struct xfbtree_config;
struct xfs_rmap_update_params;
+struct xfs_parent_name_irec;
/*
* ftrace's __print_symbolic requires that all enum values be wrapped in the
@@ -1375,6 +1376,32 @@ TRACE_EVENT(xchk_nlinks_collect_dirent,
__get_str(name))
);
+TRACE_EVENT(xchk_nlinks_collect_pptr,
+ TP_PROTO(struct xfs_mount *mp, struct xfs_inode *dp,
+ const struct xfs_parent_name_irec *pptr),
+ TP_ARGS(mp, dp, pptr),
+ TP_STRUCT__entry(
+ __field(dev_t, dev)
+ __field(xfs_ino_t, dir)
+ __field(xfs_ino_t, ino)
+ __field(unsigned int, namelen)
+ __dynamic_array(char, name, pptr->p_namelen)
+ ),
+ TP_fast_assign(
+ __entry->dev = mp->m_super->s_dev;
+ __entry->dir = dp->i_ino;
+ __entry->ino = pptr->p_ino;
+ __entry->namelen = pptr->p_namelen;
+ memcpy(__get_str(name), pptr->p_name, pptr->p_namelen);
+ ),
+ TP_printk("dev %d:%d dir 0x%llx -> ino 0x%llx name '%.*s'",
+ MAJOR(__entry->dev), MINOR(__entry->dev),
+ __entry->dir,
+ __entry->ino,
+ __entry->namelen,
+ __get_str(name))
+);
+
TRACE_EVENT(xchk_nlinks_collect_metafile,
TP_PROTO(struct xfs_mount *mp, xfs_ino_t ino),
TP_ARGS(mp, ino),
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 07/22] xfs: add raw parent pointer apis to support repair
2023-12-31 19:33 ` [PATCHSET v13.0 5/7] xfs: fsck for parent pointers Darrick J. Wong
` (5 preceding siblings ...)
2023-12-31 20:54 ` [PATCH 06/22] xfs: walk directory parent pointers to determine backref count Darrick J. Wong
@ 2023-12-31 20:55 ` Darrick J. Wong
2023-12-31 20:55 ` [PATCH 08/22] xfs: set child file owner in xfs_da_args when changing parent pointers Darrick J. Wong
` (14 subsequent siblings)
21 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:55 UTC (permalink / raw)
To: djwong; +Cc: catherine.hoang, allison.henderson, linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
Add a couple of utility functions to set or remove parent pointers from
a file. These functions will be used by repair code, hence they skip
the xattr logging that regular parent pointer updates use.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/libxfs/xfs_dir2.c | 2 +-
fs/xfs/libxfs/xfs_dir2.h | 2 +-
fs/xfs/libxfs/xfs_parent.c | 46 ++++++++++++++++++++++++++++++++++++++++++++
fs/xfs/libxfs/xfs_parent.h | 8 ++++++++
4 files changed, 56 insertions(+), 2 deletions(-)
diff --git a/fs/xfs/libxfs/xfs_dir2.c b/fs/xfs/libxfs/xfs_dir2.c
index 51eed639f2dfe..525b23a3800b6 100644
--- a/fs/xfs/libxfs/xfs_dir2.c
+++ b/fs/xfs/libxfs/xfs_dir2.c
@@ -440,7 +440,7 @@ int
xfs_dir_removename(
struct xfs_trans *tp,
struct xfs_inode *dp,
- struct xfs_name *name,
+ const struct xfs_name *name,
xfs_ino_t ino,
xfs_extlen_t total) /* bmap's total block count */
{
diff --git a/fs/xfs/libxfs/xfs_dir2.h b/fs/xfs/libxfs/xfs_dir2.h
index f99788a1f3e63..ca1949ed4f5e8 100644
--- a/fs/xfs/libxfs/xfs_dir2.h
+++ b/fs/xfs/libxfs/xfs_dir2.h
@@ -55,7 +55,7 @@ extern int xfs_dir_lookup(struct xfs_trans *tp, struct xfs_inode *dp,
const struct xfs_name *name, xfs_ino_t *inum,
struct xfs_name *ci_name);
extern int xfs_dir_removename(struct xfs_trans *tp, struct xfs_inode *dp,
- struct xfs_name *name, xfs_ino_t ino,
+ const struct xfs_name *name, xfs_ino_t ino,
xfs_extlen_t tot);
extern int xfs_dir_replace(struct xfs_trans *tp, struct xfs_inode *dp,
const struct xfs_name *name, xfs_ino_t inum,
diff --git a/fs/xfs/libxfs/xfs_parent.c b/fs/xfs/libxfs/xfs_parent.c
index 09495eb368e2b..3c31c04dd9a20 100644
--- a/fs/xfs/libxfs/xfs_parent.c
+++ b/fs/xfs/libxfs/xfs_parent.c
@@ -420,3 +420,49 @@ xfs_parent_lookup(
return xfs_attr_get_ilocked(&scr->args);
}
+
+/*
+ * Attach the parent pointer (@pptr -> @name) to @ip immediately. Caller must
+ * not have a transaction or hold the ILOCK. This is for specialized repair
+ * functions only. The scratchpad need not be initialized.
+ */
+int
+xfs_parent_set(
+ struct xfs_inode *ip,
+ const struct xfs_parent_name_irec *pptr,
+ struct xfs_parent_scratch *scr)
+{
+ if (XFS_IS_CORRUPT(ip->i_mount,
+ !xfs_parent_verify_irec(ip->i_mount, pptr))) {
+ return -EFSCORRUPTED;
+ }
+
+ xfs_parent_irec_to_disk(&scr->rec, pptr);
+ xfs_parent_scratch_init(NULL, ip, pptr, scr);
+ scr->args.op_flags |= XFS_DA_OP_LOGGED;
+
+ return xfs_attr_set(&scr->args);
+}
+
+/*
+ * Remove the parent pointer (@rec -> @name) from @ip immediately. Caller must
+ * not have a transaction or hold the ILOCK. This is for specialized repair
+ * functions only. The scratchpad need not be initialized.
+ */
+int
+xfs_parent_unset(
+ struct xfs_inode *ip,
+ const struct xfs_parent_name_irec *pptr,
+ struct xfs_parent_scratch *scr)
+{
+ if (XFS_IS_CORRUPT(ip->i_mount,
+ !xfs_parent_verify_irec(ip->i_mount, pptr))) {
+ return -EFSCORRUPTED;
+ }
+
+ xfs_parent_irec_to_disk(&scr->rec, pptr);
+ xfs_parent_scratch_init(NULL, ip, pptr, scr);
+ scr->args.op_flags |= XFS_DA_OP_LOGGED | XFS_DA_OP_REMOVE;
+
+ return xfs_attr_set(&scr->args);
+}
diff --git a/fs/xfs/libxfs/xfs_parent.h b/fs/xfs/libxfs/xfs_parent.h
index e4443da1d86f2..58e59af818bd2 100644
--- a/fs/xfs/libxfs/xfs_parent.h
+++ b/fs/xfs/libxfs/xfs_parent.h
@@ -162,4 +162,12 @@ int xfs_parent_lookup(struct xfs_trans *tp, struct xfs_inode *ip,
const struct xfs_parent_name_irec *pptr,
struct xfs_parent_scratch *scratch);
+int xfs_parent_set(struct xfs_inode *ip,
+ const struct xfs_parent_name_irec *pptr,
+ struct xfs_parent_scratch *scratch);
+
+int xfs_parent_unset(struct xfs_inode *ip,
+ const struct xfs_parent_name_irec *rec,
+ struct xfs_parent_scratch *scratch);
+
#endif /* __XFS_PARENT_H__ */
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 08/22] xfs: set child file owner in xfs_da_args when changing parent pointers
2023-12-31 19:33 ` [PATCHSET v13.0 5/7] xfs: fsck for parent pointers Darrick J. Wong
` (6 preceding siblings ...)
2023-12-31 20:55 ` [PATCH 07/22] xfs: add raw parent pointer apis to support repair Darrick J. Wong
@ 2023-12-31 20:55 ` Darrick J. Wong
2023-12-31 20:55 ` [PATCH 09/22] xfs: salvage parent pointers when rebuilding xattr structures Darrick J. Wong
` (13 subsequent siblings)
21 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:55 UTC (permalink / raw)
To: djwong; +Cc: catherine.hoang, allison.henderson, linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
Now that struct xfs_da_args has an explicit file owner field, we must
set it when modifying parent pointers.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/libxfs/xfs_parent.c | 13 ++++++++++---
fs/xfs/libxfs/xfs_parent.h | 4 ++--
2 files changed, 12 insertions(+), 5 deletions(-)
diff --git a/fs/xfs/libxfs/xfs_parent.c b/fs/xfs/libxfs/xfs_parent.c
index 3c31c04dd9a20..3c3fcdf8b975b 100644
--- a/fs/xfs/libxfs/xfs_parent.c
+++ b/fs/xfs/libxfs/xfs_parent.c
@@ -201,6 +201,7 @@ xfs_parent_addname(
args->trans = tp;
args->dp = child;
+ args->owner = child->i_ino;
xfs_init_parent_davalue(&ppargs->args, parent_name);
@@ -239,6 +240,7 @@ xfs_parent_removename(
args->trans = tp;
args->dp = child;
+ args->owner = child->i_ino;
xfs_init_parent_davalue(&ppargs->args, parent_name);
@@ -288,6 +290,7 @@ xfs_parent_replacename(
args->trans = tp;
args->dp = child;
+ args->owner = child->i_ino;
xfs_init_parent_davalue(&ppargs->args, old_name);
xfs_init_parent_danewvalue(&ppargs->args, new_name);
@@ -371,6 +374,7 @@ static inline void
xfs_parent_scratch_init(
struct xfs_trans *tp,
struct xfs_inode *ip,
+ xfs_ino_t owner,
const struct xfs_parent_name_irec *pptr,
struct xfs_parent_scratch *scr)
{
@@ -387,6 +391,7 @@ xfs_parent_scratch_init(
scr->args.whichfork = XFS_ATTR_FORK;
scr->args.hashval = xfs_da_hashname((const void *)&scr->rec,
sizeof(struct xfs_parent_name_rec));
+ scr->args.owner = owner;
}
/*
@@ -415,7 +420,7 @@ xfs_parent_lookup(
}
xfs_parent_irec_to_disk(&scr->rec, pptr);
- xfs_parent_scratch_init(tp, ip, pptr, scr);
+ xfs_parent_scratch_init(tp, ip, ip->i_ino, pptr, scr);
scr->args.op_flags |= XFS_DA_OP_OKNOENT;
return xfs_attr_get_ilocked(&scr->args);
@@ -429,6 +434,7 @@ xfs_parent_lookup(
int
xfs_parent_set(
struct xfs_inode *ip,
+ xfs_ino_t owner,
const struct xfs_parent_name_irec *pptr,
struct xfs_parent_scratch *scr)
{
@@ -438,7 +444,7 @@ xfs_parent_set(
}
xfs_parent_irec_to_disk(&scr->rec, pptr);
- xfs_parent_scratch_init(NULL, ip, pptr, scr);
+ xfs_parent_scratch_init(NULL, ip, owner, pptr, scr);
scr->args.op_flags |= XFS_DA_OP_LOGGED;
return xfs_attr_set(&scr->args);
@@ -452,6 +458,7 @@ xfs_parent_set(
int
xfs_parent_unset(
struct xfs_inode *ip,
+ xfs_ino_t owner,
const struct xfs_parent_name_irec *pptr,
struct xfs_parent_scratch *scr)
{
@@ -461,7 +468,7 @@ xfs_parent_unset(
}
xfs_parent_irec_to_disk(&scr->rec, pptr);
- xfs_parent_scratch_init(NULL, ip, pptr, scr);
+ xfs_parent_scratch_init(NULL, ip, owner, pptr, scr);
scr->args.op_flags |= XFS_DA_OP_LOGGED | XFS_DA_OP_REMOVE;
return xfs_attr_set(&scr->args);
diff --git a/fs/xfs/libxfs/xfs_parent.h b/fs/xfs/libxfs/xfs_parent.h
index 58e59af818bd2..46bf96c7e3c92 100644
--- a/fs/xfs/libxfs/xfs_parent.h
+++ b/fs/xfs/libxfs/xfs_parent.h
@@ -162,11 +162,11 @@ int xfs_parent_lookup(struct xfs_trans *tp, struct xfs_inode *ip,
const struct xfs_parent_name_irec *pptr,
struct xfs_parent_scratch *scratch);
-int xfs_parent_set(struct xfs_inode *ip,
+int xfs_parent_set(struct xfs_inode *ip, xfs_ino_t owner,
const struct xfs_parent_name_irec *pptr,
struct xfs_parent_scratch *scratch);
-int xfs_parent_unset(struct xfs_inode *ip,
+int xfs_parent_unset(struct xfs_inode *ip, xfs_ino_t owner,
const struct xfs_parent_name_irec *rec,
struct xfs_parent_scratch *scratch);
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 09/22] xfs: salvage parent pointers when rebuilding xattr structures
2023-12-31 19:33 ` [PATCHSET v13.0 5/7] xfs: fsck for parent pointers Darrick J. Wong
` (7 preceding siblings ...)
2023-12-31 20:55 ` [PATCH 08/22] xfs: set child file owner in xfs_da_args when changing parent pointers Darrick J. Wong
@ 2023-12-31 20:55 ` Darrick J. Wong
2023-12-31 20:55 ` [PATCH 10/22] xfs: replace namebuf with parent pointer in directory repair Darrick J. Wong
` (12 subsequent siblings)
21 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:55 UTC (permalink / raw)
To: djwong; +Cc: catherine.hoang, allison.henderson, linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
When we're salvaging extended attributes, make sure we validate the ones
that claim to be parent pointers before adding them to the salvage pile.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/scrub/attr_repair.c | 41 ++++++++++++++++++++++++++++++++---------
fs/xfs/scrub/trace.h | 40 ++++++++++++++++++++++++++++++++++++++++
2 files changed, 72 insertions(+), 9 deletions(-)
diff --git a/fs/xfs/scrub/attr_repair.c b/fs/xfs/scrub/attr_repair.c
index 9a88d46392626..39e64d7559451 100644
--- a/fs/xfs/scrub/attr_repair.c
+++ b/fs/xfs/scrub/attr_repair.c
@@ -28,6 +28,7 @@
#include "xfs_swapext.h"
#include "xfs_xchgrange.h"
#include "xfs_acl.h"
+#include "xfs_parent.h"
#include "scrub/xfs_scrub.h"
#include "scrub/scrub.h"
#include "scrub/common.h"
@@ -124,6 +125,13 @@ xrep_xattr_want_salvage(
return false;
if (valuelen > XATTR_SIZE_MAX || valuelen < 0)
return false;
+ if (attr_flags & XFS_ATTR_PARENT) {
+ if (!xfs_parent_namecheck(rx->sc->mp, name, namelen,
+ attr_flags))
+ return false;
+ if (!xfs_parent_valuecheck(rx->sc->mp, value, valuelen))
+ return false;
+ }
return true;
}
@@ -151,14 +159,21 @@ xrep_xattr_salvage_key(
* Truncate the name to the first character that would trip namecheck.
* If we no longer have a name after that, ignore this attribute.
*/
- while (i < namelen && name[i] != 0)
- i++;
- if (i == 0)
- return 0;
- key.namelen = i;
+ if (flags & XFS_ATTR_PARENT) {
+ key.namelen = namelen;
- trace_xrep_xattr_salvage_rec(rx->sc->ip, flags, name, key.namelen,
- valuelen);
+ trace_xrep_xattr_salvage_pptr(rx->sc->ip, flags, name,
+ key.namelen, value, valuelen);
+ } else {
+ while (i < namelen && name[i] != 0)
+ i++;
+ if (i == 0)
+ return 0;
+ key.namelen = i;
+
+ trace_xrep_xattr_salvage_rec(rx->sc->ip, flags, name,
+ key.namelen, valuelen);
+ }
error = xfblob_store(rx->xattr_blobs, &key.name_cookie, name,
key.namelen);
@@ -562,6 +577,9 @@ xrep_xattr_insert_rec(
struct xchk_xattr_buf *ab = rx->sc->buf;
int error;
+ if (key->flags & XFS_ATTR_PARENT)
+ args.op_flags |= XFS_DA_OP_NVLOOKUP;
+
/*
* Grab pointers to the scrub buffer so that we can use them to insert
* attrs into the temp file.
@@ -595,8 +613,13 @@ xrep_xattr_insert_rec(
ab->name[key->namelen] = 0;
- trace_xrep_xattr_insert_rec(rx->sc->tempip, key->flags, ab->name,
- key->namelen, key->valuelen);
+ if (key->flags & XFS_ATTR_PARENT)
+ trace_xrep_xattr_insert_pptr(rx->sc->tempip, key->flags,
+ ab->name, key->namelen, ab->value,
+ key->valuelen);
+ else
+ trace_xrep_xattr_insert_rec(rx->sc->tempip, key->flags,
+ ab->name, key->namelen, key->valuelen);
/*
* xfs_attr_set creates and commits its own transaction. If the attr
diff --git a/fs/xfs/scrub/trace.h b/fs/xfs/scrub/trace.h
index d3a0cefea3684..8fa26a7811118 100644
--- a/fs/xfs/scrub/trace.h
+++ b/fs/xfs/scrub/trace.h
@@ -2662,6 +2662,46 @@ DEFINE_EVENT(xrep_xattr_salvage_class, name, \
DEFINE_XREP_XATTR_SALVAGE_EVENT(xrep_xattr_salvage_rec);
DEFINE_XREP_XATTR_SALVAGE_EVENT(xrep_xattr_insert_rec);
+DECLARE_EVENT_CLASS(xrep_pptr_salvage_class,
+ TP_PROTO(struct xfs_inode *ip, unsigned int flags, const void *name,
+ unsigned int namelen, const void *value, unsigned int valuelen),
+ TP_ARGS(ip, flags, name, namelen, value, valuelen),
+ TP_STRUCT__entry(
+ __field(dev_t, dev)
+ __field(xfs_ino_t, ino)
+ __field(xfs_ino_t, parent_ino)
+ __field(unsigned int, parent_gen)
+ __field(unsigned int, namelen)
+ __dynamic_array(char, name, valuelen)
+ ),
+ TP_fast_assign(
+ struct xfs_parent_name_irec pptr;
+
+ xfs_parent_irec_from_disk(&pptr, name, value, valuelen);
+
+ __entry->dev = ip->i_mount->m_super->s_dev;
+ __entry->ino = ip->i_ino;
+ __entry->parent_ino = pptr.p_ino;
+ __entry->parent_gen = pptr.p_gen;
+ __entry->namelen = pptr.p_namelen;
+ memcpy(__get_str(name), pptr.p_name, pptr.p_namelen);
+ ),
+ TP_printk("dev %d:%d ino 0x%llx parent_ino 0x%llx parent_gen 0x%x name '%.*s'",
+ MAJOR(__entry->dev), MINOR(__entry->dev),
+ __entry->ino,
+ __entry->parent_ino,
+ __entry->parent_gen,
+ __entry->namelen,
+ __get_str(name))
+)
+#define DEFINE_XREP_PPTR_SALVAGE_EVENT(name) \
+DEFINE_EVENT(xrep_pptr_salvage_class, name, \
+ TP_PROTO(struct xfs_inode *ip, unsigned int flags, const void *name, \
+ unsigned int namelen, const void *value, unsigned int valuelen), \
+ TP_ARGS(ip, flags, name, namelen, value, valuelen))
+DEFINE_XREP_PPTR_SALVAGE_EVENT(xrep_xattr_salvage_pptr);
+DEFINE_XREP_PPTR_SALVAGE_EVENT(xrep_xattr_insert_pptr);
+
TRACE_EVENT(xrep_xattr_class,
TP_PROTO(struct xfs_inode *ip, struct xfs_inode *arg_ip),
TP_ARGS(ip, arg_ip),
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 10/22] xfs: replace namebuf with parent pointer in directory repair
2023-12-31 19:33 ` [PATCHSET v13.0 5/7] xfs: fsck for parent pointers Darrick J. Wong
` (8 preceding siblings ...)
2023-12-31 20:55 ` [PATCH 09/22] xfs: salvage parent pointers when rebuilding xattr structures Darrick J. Wong
@ 2023-12-31 20:55 ` Darrick J. Wong
2023-12-31 20:56 ` [PATCH 11/22] xfs: repair directories by scanning directory parent pointers Darrick J. Wong
` (11 subsequent siblings)
21 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:55 UTC (permalink / raw)
To: djwong; +Cc: catherine.hoang, allison.henderson, linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
Replace the dirent name buffer at the end of struct xrep_dir with a
xfs_parent_name_irec object. The namebuf and p_name usage do not
overlap, so we can save 256 bytes of memory by allowing them to overlap.
Doing so makes the code a bit more complex, so this is called out
separately.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/scrub/dir_repair.c | 19 +++++++++++++------
1 file changed, 13 insertions(+), 6 deletions(-)
diff --git a/fs/xfs/scrub/dir_repair.c b/fs/xfs/scrub/dir_repair.c
index e74f456c7b444..13a1a3ef5e714 100644
--- a/fs/xfs/scrub/dir_repair.c
+++ b/fs/xfs/scrub/dir_repair.c
@@ -28,6 +28,7 @@
#include "xfs_swapext.h"
#include "xfs_xchgrange.h"
#include "xfs_ag.h"
+#include "xfs_parent.h"
#include "scrub/xfs_scrub.h"
#include "scrub/scrub.h"
#include "scrub/common.h"
@@ -131,8 +132,14 @@ struct xrep_dir {
/* Should we move this directory to the orphanage? */
bool needs_adoption;
- /* Directory entry name, plus the trailing null. */
- unsigned char namebuf[MAXNAMELEN];
+ /*
+ * Scratch buffer for reading parent pointers from child files. The
+ * p_name field is used to flush stashed dirents into the temporary
+ * directory in between parent pointers. At the very end of the
+ * repair, it can also be used to compute the lost+found filename
+ * if we need to reparent the directory.
+ */
+ struct xfs_parent_name_irec pptr;
};
/* Tear down all the incore stuff we created. */
@@ -696,7 +703,7 @@ xrep_dir_replay_update(
struct xfs_name name = {
.len = dirent->namelen,
.type = dirent->ftype,
- .name = rd->namebuf,
+ .name = rd->pptr.p_name,
};
struct xfs_mount *mp = rd->sc->mp;
#ifdef DEBUG
@@ -773,10 +780,10 @@ xrep_dir_replay_updates(
/* The dirent name is stored in the in-core buffer. */
error = xfblob_load(rd->dir_names, dirent.name_cookie,
- rd->namebuf, dirent.namelen);
+ rd->pptr.p_name, dirent.namelen);
if (error)
return error;
- rd->namebuf[MAXNAMELEN - 1] = 0;
+ rd->pptr.p_name[MAXNAMELEN - 1] = 0;
error = xrep_dir_replay_update(rd, &dirent);
if (error)
@@ -1416,7 +1423,7 @@ xrep_dir_move_to_orphanage(
if (error)
return error;
- error = xrep_adoption_compute_name(&rd->adoption, rd->namebuf);
+ error = xrep_adoption_compute_name(&rd->adoption, rd->pptr.p_name);
if (error)
return error;
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 11/22] xfs: repair directories by scanning directory parent pointers
2023-12-31 19:33 ` [PATCHSET v13.0 5/7] xfs: fsck for parent pointers Darrick J. Wong
` (9 preceding siblings ...)
2023-12-31 20:55 ` [PATCH 10/22] xfs: replace namebuf with parent pointer in directory repair Darrick J. Wong
@ 2023-12-31 20:56 ` Darrick J. Wong
2023-12-31 20:56 ` [PATCH 12/22] xfs: implement live updates for directory repairs Darrick J. Wong
` (10 subsequent siblings)
21 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:56 UTC (permalink / raw)
To: djwong; +Cc: catherine.hoang, allison.henderson, linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
For filesystems with parent pointers, scan the entire filesystem looking
for parent pointers that target the directory we're rebuilding instead
of trying to salvage whatever we can from the directory data blocks.
This will be more robust than salvaging, but there's more code to come.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/scrub/dir_repair.c | 327 ++++++++++++++++++++++++++++++++++++++++++++-
1 file changed, 323 insertions(+), 4 deletions(-)
diff --git a/fs/xfs/scrub/dir_repair.c b/fs/xfs/scrub/dir_repair.c
index 13a1a3ef5e714..cae22ad33bca3 100644
--- a/fs/xfs/scrub/dir_repair.c
+++ b/fs/xfs/scrub/dir_repair.c
@@ -44,6 +44,7 @@
#include "scrub/reap.h"
#include "scrub/findparent.h"
#include "scrub/orphanage.h"
+#include "scrub/listxattr.h"
/*
* Directory Repair
@@ -58,6 +59,15 @@
* being repaired and the temporary directory, and will later become important
* for parent pointer scanning.
*
+ * If parent pointers are enabled on this filesystem, we instead reconstruct
+ * the directory by visiting each parent pointer of each file in the filesystem
+ * and translating the relevant parent pointer records into dirents. In this
+ * case, it is advantageous to stash all directory entries created from parent
+ * pointers for a single child file before replaying them into the temporary
+ * directory. To save memory, the live filesystem scan reuses the findparent
+ * fields. Directory repair chooses either parent pointer scanning or
+ * directory entry salvaging, but not both.
+ *
* Directory entries added to the temporary directory do not elevate the link
* counts of the inodes found. When salvaging completes, the remaining stashed
* entries are replayed to the temporary directory. An atomic extent swap is
@@ -113,7 +123,15 @@ struct xrep_dir {
/*
* Information used to scan the filesystem to find the inumber of the
- * dotdot entry for this directory.
+ * dotdot entry for this directory. For directory salvaging when
+ * parent pointers are not enabled, we use the findparent_* functions
+ * on this object and access only the parent_ino field directly.
+ *
+ * When parent pointers are enabled, however, the pptr scanner uses the
+ * iscan, hooks, lock, and parent_ino fields of this object directly.
+ * @pscan.lock coordinates access to dir_entries, dir_names,
+ * parent_ino, subdirs, dirents, and args. This reduces the memory
+ * requirements of this structure.
*/
struct xrep_parent_scan_info pscan;
@@ -1003,6 +1021,261 @@ xrep_dir_salvage_entries(
}
+/*
+ * Examine a parent pointer of a file. If it leads us back to the directory
+ * that we're rebuilding, create an incore dirent from the parent pointer and
+ * stash it.
+ */
+STATIC int
+xrep_dir_scan_pptr(
+ struct xfs_scrub *sc,
+ struct xfs_inode *ip,
+ const struct xfs_parent_name_irec *pptr,
+ void *priv)
+{
+ struct xfs_name xname;
+ struct xrep_dir *rd = priv;
+ int error;
+
+ /*
+ * Ignore parent pointers that point back to a different dir, list the
+ * wrong generation number, or are invalid.
+ */
+ if (pptr->p_ino != sc->ip->i_ino ||
+ pptr->p_gen != VFS_I(sc->ip)->i_generation ||
+ !xfs_parent_verify_irec(sc->mp, pptr))
+ return 0;
+
+ /*
+ * Transform this parent pointer into a dirent and queue it for later
+ * addition to the temporary directory.
+ */
+ xname.name = pptr->p_name;
+ xname.len = pptr->p_namelen;
+ xname.type = xfs_mode_to_ftype(VFS_I(ip)->i_mode);
+
+ mutex_lock(&rd->pscan.lock);
+ error = xrep_dir_stash_createname(rd, &xname, ip->i_ino);
+ mutex_unlock(&rd->pscan.lock);
+ return error;
+}
+
+/*
+ * If this child dirent points to the directory being repaired, remember that
+ * fact so that we can reset the dotdot entry if necessary.
+ */
+STATIC int
+xrep_dir_scan_dirent(
+ struct xfs_scrub *sc,
+ struct xfs_inode *dp,
+ xfs_dir2_dataptr_t dapos,
+ const struct xfs_name *name,
+ xfs_ino_t ino,
+ void *priv)
+{
+ struct xrep_dir *rd = priv;
+
+ /* Dirent doesn't point to this directory. */
+ if (ino != rd->sc->ip->i_ino)
+ return 0;
+
+ /* Ignore garbage inum. */
+ if (!xfs_verify_dir_ino(rd->sc->mp, ino))
+ return 0;
+
+ /* No weird looking names. */
+ if (name->len >= MAXNAMELEN || name->len <= 0)
+ return 0;
+
+ /* Don't pick up dot or dotdot entries; we only want child dirents. */
+ if (xfs_dir2_samename(name, &xfs_name_dotdot) ||
+ xfs_dir2_samename(name, &xfs_name_dot))
+ return 0;
+
+ trace_xrep_dir_stash_createname(sc->tempip, &xfs_name_dotdot,
+ dp->i_ino);
+
+ xrep_findparent_scan_found(&rd->pscan, dp->i_ino);
+ return 0;
+}
+
+/*
+ * Decide if we want to look for child dirents or parent pointers in this file.
+ * Skip the dir being repaired and any files being used to stage repairs.
+ */
+static inline bool
+xrep_dir_want_scan(
+ struct xrep_dir *rd,
+ const struct xfs_inode *ip)
+{
+ return ip != rd->sc->ip && !xrep_is_tempfile(ip);
+}
+
+/*
+ * Take ILOCK on a file that we want to scan.
+ *
+ * Select ILOCK_EXCL if the file is a directory with an unloaded data bmbt or
+ * has an unloaded attr bmbt. Otherwise, take ILOCK_SHARED.
+ */
+static inline unsigned int
+xrep_dir_scan_ilock(
+ struct xrep_dir *rd,
+ struct xfs_inode *ip)
+{
+ uint lock_mode = XFS_ILOCK_SHARED;
+
+ /* Need to take the shared ILOCK to advance the iscan cursor. */
+ if (!xrep_dir_want_scan(rd, ip))
+ goto lock;
+
+ if (S_ISDIR(VFS_I(ip)->i_mode) && xfs_need_iread_extents(&ip->i_df)) {
+ lock_mode = XFS_ILOCK_EXCL;
+ goto lock;
+ }
+
+ if (xfs_inode_has_attr_fork(ip) && xfs_need_iread_extents(&ip->i_af))
+ lock_mode = XFS_ILOCK_EXCL;
+
+lock:
+ xfs_ilock(ip, lock_mode);
+ return lock_mode;
+}
+
+/*
+ * Scan this file for relevant child dirents or parent pointers that point to
+ * the directory we're rebuilding.
+ */
+STATIC int
+xrep_dir_scan_file(
+ struct xrep_dir *rd,
+ struct xfs_inode *ip)
+{
+ unsigned int lock_mode;
+ int error = 0;
+
+ lock_mode = xrep_dir_scan_ilock(rd, ip);
+
+ if (!xrep_dir_want_scan(rd, ip))
+ goto scan_done;
+
+ /*
+ * If the extended attributes look as though they has been zapped by
+ * the inode record repair code, we cannot scan for parent pointers.
+ */
+ if (xchk_pptr_looks_zapped(ip)) {
+ error = -EBUSY;
+ goto scan_done;
+ }
+
+ error = xchk_pptr_walk(rd->sc, ip, xrep_dir_scan_pptr, &rd->pptr, rd);
+ if (error)
+ goto scan_done;
+
+ if (S_ISDIR(VFS_I(ip)->i_mode)) {
+ /*
+ * If the directory looks as though it has been zapped by the
+ * inode record repair code, we cannot scan for child dirents.
+ */
+ if (xchk_dir_looks_zapped(ip)) {
+ error = -EBUSY;
+ goto scan_done;
+ }
+
+ error = xchk_dir_walk(rd->sc, ip, xrep_dir_scan_dirent, rd);
+ if (error)
+ goto scan_done;
+ }
+
+scan_done:
+ xchk_iscan_mark_visited(&rd->pscan.iscan, ip);
+ xfs_iunlock(ip, lock_mode);
+ return error;
+}
+
+/*
+ * Scan all files in the filesystem for parent pointers that we can turn into
+ * replacement dirents, and a dirent that we can use to set the dotdot pointer.
+ */
+STATIC int
+xrep_dir_scan_dirtree(
+ struct xrep_dir *rd)
+{
+ struct xfs_scrub *sc = rd->sc;
+ struct xfs_inode *ip;
+ int error;
+
+ /* Roots of directory trees are their own parents. */
+ if (sc->ip == sc->mp->m_rootip)
+ xrep_findparent_scan_found(&rd->pscan, sc->ip->i_ino);
+
+ /*
+ * Filesystem scans are time consuming. Drop the directory ILOCK and
+ * all other resources for the duration of the scan and hope for the
+ * best. The live update hooks will keep our scan information up to
+ * date even though we've dropped the locks.
+ */
+ xchk_trans_cancel(sc);
+ if (sc->ilock_flags & (XFS_ILOCK_SHARED | XFS_ILOCK_EXCL))
+ xchk_iunlock(sc, sc->ilock_flags & (XFS_ILOCK_SHARED |
+ XFS_ILOCK_EXCL));
+ error = xchk_trans_alloc_empty(sc);
+ if (error)
+ return error;
+
+ while ((error = xchk_iscan_iter(&rd->pscan.iscan, &ip)) == 1) {
+ bool flush;
+
+ error = xrep_dir_scan_file(rd, ip);
+ xchk_irele(sc, ip);
+ if (error)
+ break;
+
+ /* Flush stashed dirent updates to constrain memory usage. */
+ mutex_lock(&rd->pscan.lock);
+ flush = xrep_dir_want_flush_stashed(rd);
+ mutex_unlock(&rd->pscan.lock);
+ if (flush) {
+ xchk_trans_cancel(sc);
+
+ error = xrep_tempfile_iolock_polled(sc);
+ if (error)
+ break;
+
+ mutex_lock(&rd->pscan.lock);
+ error = xrep_dir_replay_updates(rd);
+ mutex_unlock(&rd->pscan.lock);
+ xrep_tempfile_iounlock(sc);
+ if (error)
+ break;
+
+ error = xchk_trans_alloc_empty(sc);
+ if (error)
+ break;
+ }
+
+ if (xchk_should_terminate(sc, &error))
+ break;
+ }
+ xchk_iscan_iter_finish(&rd->pscan.iscan);
+ if (error) {
+ /*
+ * If we couldn't grab an inode that was busy with a state
+ * change, change the error code so that we exit to userspace
+ * as quickly as possible.
+ */
+ if (error == -EBUSY)
+ return -ECANCELED;
+ return error;
+ }
+
+ /*
+ * Cancel the empty transaction so that we can (later) use the atomic
+ * extent swap helpers to lock files and commit the new directory.
+ */
+ xchk_trans_cancel(rd->sc);
+ return 0;
+}
+
/*
* Free all the directory blocks and reset the data fork. The caller must
* join the inode to the transaction. This function returns with the inode
@@ -1201,6 +1474,45 @@ xrep_dir_set_nlink(
return 0;
}
+/*
+ * Finish replaying stashed dirent updates, allocate a transaction for swapping
+ * extents, and take the ILOCKs of both directories before we commit the new
+ * directory structure.
+ */
+STATIC int
+xrep_dir_finalize_tempdir(
+ struct xrep_dir *rd)
+{
+ struct xfs_scrub *sc = rd->sc;
+ int error;
+
+ if (!xfs_has_parent(sc->mp))
+ return xrep_tempswap_trans_alloc(sc, XFS_DATA_FORK, &rd->tx);
+
+ /*
+ * Repair relies on the ILOCK to quiesce all possible dirent updates.
+ * Replay all queued dirent updates into the tempdir before swapping
+ * the contents, even if that means dropping the ILOCKs and the
+ * transaction.
+ */
+ do {
+ error = xrep_dir_replay_updates(rd);
+ if (error)
+ return error;
+
+ error = xrep_tempswap_trans_alloc(sc, XFS_DATA_FORK, &rd->tx);
+ if (error)
+ return error;
+
+ if (xfarray_length(rd->dir_entries) == 0)
+ break;
+
+ xchk_trans_cancel(sc);
+ xrep_tempfile_iunlock_both(sc);
+ } while (!xchk_should_terminate(sc, &error));
+ return error;
+}
+
/* Swap the temporary directory's data fork with the one being repaired. */
STATIC int
xrep_dir_swap(
@@ -1300,8 +1612,12 @@ xrep_dir_rebuild_tree(
if (error)
return error;
- /* Allocate transaction and ILOCK the scrub file and the temp file. */
- error = xrep_tempswap_trans_alloc(sc, XFS_DATA_FORK, &rd->tx);
+ /*
+ * Allocate transaction, lock inodes, and make sure that we've replayed
+ * all the stashed dirent updates to the tempdir. After this point,
+ * we're ready to swapext.
+ */
+ error = xrep_dir_finalize_tempdir(rd);
if (error)
return error;
@@ -1486,7 +1802,10 @@ xrep_directory(
if (error)
return error;
- error = xrep_dir_salvage_entries(rd);
+ if (xfs_has_parent(sc->mp))
+ error = xrep_dir_scan_dirtree(rd);
+ else
+ error = xrep_dir_salvage_entries(rd);
if (error)
goto out_teardown;
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 12/22] xfs: implement live updates for directory repairs
2023-12-31 19:33 ` [PATCHSET v13.0 5/7] xfs: fsck for parent pointers Darrick J. Wong
` (10 preceding siblings ...)
2023-12-31 20:56 ` [PATCH 11/22] xfs: repair directories by scanning directory parent pointers Darrick J. Wong
@ 2023-12-31 20:56 ` Darrick J. Wong
2023-12-31 20:56 ` [PATCH 13/22] xfs: replay unlocked parent pointer updates that accrue during xattr repair Darrick J. Wong
` (9 subsequent siblings)
21 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:56 UTC (permalink / raw)
To: djwong; +Cc: catherine.hoang, allison.henderson, linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
While we're scanning the filesystem for parent pointers that we can turn
into dirents, we cannot hold the IOLOCK or ILOCK of the directory being
repaired. Therefore, we need to set up a dirent hook so that we can
keep the temporary directory up to date with the rest of the filesystem.
Hence we add the ability to *remove* entries from the temporary dir.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/scrub/dir_repair.c | 221 +++++++++++++++++++++++++++++++++++++++++----
fs/xfs/scrub/findparent.c | 8 +-
fs/xfs/scrub/findparent.h | 10 ++
fs/xfs/scrub/trace.h | 2
4 files changed, 218 insertions(+), 23 deletions(-)
diff --git a/fs/xfs/scrub/dir_repair.c b/fs/xfs/scrub/dir_repair.c
index cae22ad33bca3..c66838fc144d5 100644
--- a/fs/xfs/scrub/dir_repair.c
+++ b/fs/xfs/scrub/dir_repair.c
@@ -85,6 +85,12 @@
* or to use dirent hooks to capture updates from other threads.
*/
+/* Create a dirent in the tempdir. */
+#define XREP_DIRENT_ADD (1)
+
+/* Remove a dirent from the tempdir. */
+#define XREP_DIRENT_REMOVE (2)
+
/* Directory entry to be restored in the new directory. */
struct xrep_dirent {
/* Cookie for retrieval of the dirent name. */
@@ -98,6 +104,9 @@ struct xrep_dirent {
/* File type of the dirent. */
uint8_t ftype;
+
+ /* XREP_DIRENT_{ADD,REMOVE} */
+ uint8_t action;
};
/*
@@ -341,6 +350,7 @@ xrep_dir_stash_createname(
xfs_ino_t ino)
{
struct xrep_dirent dirent = {
+ .action = XREP_DIRENT_ADD,
.ino = ino,
.namelen = name->len,
.ftype = name->type,
@@ -357,6 +367,34 @@ xrep_dir_stash_createname(
return xfarray_append(rd->dir_entries, &dirent);
}
+/*
+ * Remember that we want to remove a dirent from the tempdir. These stashed
+ * actions will be replayed later.
+ */
+STATIC int
+xrep_dir_stash_removename(
+ struct xrep_dir *rd,
+ const struct xfs_name *name,
+ xfs_ino_t ino)
+{
+ struct xrep_dirent dirent = {
+ .action = XREP_DIRENT_REMOVE,
+ .ino = ino,
+ .namelen = name->len,
+ .ftype = name->type,
+ };
+ int error;
+
+ trace_xrep_dir_stash_removename(rd->sc->tempip, name, ino);
+
+ error = xfblob_store(rd->dir_names, &dirent.name_cookie, name->name,
+ name->len);
+ if (error)
+ return error;
+
+ return xfarray_append(rd->dir_entries, &dirent);
+}
+
/* Allocate an in-core record to hold entries while we rebuild the dir data. */
STATIC int
xrep_dir_salvage_entry(
@@ -708,6 +746,43 @@ xrep_dir_replay_createname(
return xfs_dir2_node_addname(&rd->args);
}
+/* Replay a stashed removename onto the temporary directory. */
+STATIC int
+xrep_dir_replay_removename(
+ struct xrep_dir *rd,
+ const struct xfs_name *name,
+ xfs_extlen_t total)
+{
+ struct xfs_inode *dp = rd->args.dp;
+ bool is_block, is_leaf;
+ int error;
+
+ ASSERT(S_ISDIR(VFS_I(dp)->i_mode));
+
+ xrep_dir_init_args(rd, dp, name);
+ rd->args.op_flags = 0;
+ rd->args.total = total;
+
+ trace_xrep_dir_replay_removename(dp, name, 0);
+
+ if (dp->i_df.if_format == XFS_DINODE_FMT_LOCAL)
+ return xfs_dir2_sf_removename(&rd->args);
+
+ error = xfs_dir2_isblock(&rd->args, &is_block);
+ if (error)
+ return error;
+ if (is_block)
+ return xfs_dir2_block_removename(&rd->args);
+
+ error = xfs_dir2_isleaf(&rd->args, &is_leaf);
+ if (error)
+ return error;
+ if (is_leaf)
+ return xfs_dir2_leaf_removename(&rd->args);
+
+ return xfs_dir2_node_removename(&rd->args);
+}
+
/*
* Add this stashed incore directory entry to the temporary directory.
* The caller must hold the tempdir's IOLOCK, must not hold any ILOCKs, and
@@ -739,26 +814,64 @@ xrep_dir_replay_update(
xrep_tempfile_ilock(rd->sc);
xfs_trans_ijoin(rd->sc->tp, rd->sc->tempip, 0);
- /*
- * Create a replacement dirent in the temporary directory. Note that
- * _createname doesn't check for existing entries. There shouldn't be
- * any in the temporary dir, but we'll verify this in debug mode.
- */
+ switch (dirent->action) {
+ case XREP_DIRENT_ADD:
+ /*
+ * Create a replacement dirent in the temporary directory.
+ * Note that _createname doesn't check for existing entries.
+ * There shouldn't be any in the temporary dir, but we'll
+ * verify this in debug mode.
+ */
#ifdef DEBUG
- error = xchk_dir_lookup(rd->sc, rd->sc->tempip, &name, &ino);
- if (error != -ENOENT) {
- ASSERT(error != -ENOENT);
+ error = xchk_dir_lookup(rd->sc, rd->sc->tempip, &name, &ino);
+ if (error != -ENOENT) {
+ ASSERT(error != -ENOENT);
+ goto out_cancel;
+ }
+#endif
+
+ error = xrep_dir_replay_createname(rd, &name, dirent->ino,
+ resblks);
+ if (error)
+ goto out_cancel;
+
+ if (name.type == XFS_DIR3_FT_DIR)
+ rd->subdirs++;
+ rd->dirents++;
+ break;
+ case XREP_DIRENT_REMOVE:
+ /*
+ * Remove a dirent from the temporary directory. Note that
+ * _removename doesn't check the inode target of the exist
+ * entry. There should be a perfect match in the temporary
+ * dir, but we'll verify this in debug mode.
+ */
+#ifdef DEBUG
+ error = xchk_dir_lookup(rd->sc, rd->sc->tempip, &name, &ino);
+ if (error) {
+ ASSERT(error != 0);
+ goto out_cancel;
+ }
+ if (ino != dirent->ino) {
+ ASSERT(ino == dirent->ino);
+ error = -EIO;
+ goto out_cancel;
+ }
+#endif
+
+ error = xrep_dir_replay_removename(rd, &name, resblks);
+ if (error)
+ goto out_cancel;
+
+ if (name.type == XFS_DIR3_FT_DIR)
+ rd->subdirs--;
+ rd->dirents--;
+ break;
+ default:
+ ASSERT(0);
+ error = -EIO;
goto out_cancel;
}
-#endif
-
- error = xrep_dir_replay_createname(rd, &name, dirent->ino, resblks);
- if (error)
- goto out_cancel;
-
- if (name.type == XFS_DIR3_FT_DIR)
- rd->subdirs++;
- rd->dirents++;
/* Commit and unlock. */
error = xrep_trans_commit(rd->sc);
@@ -1276,6 +1389,71 @@ xrep_dir_scan_dirtree(
return 0;
}
+/*
+ * Capture dirent updates being made by other threads which are relevant to the
+ * directory being repaired.
+ */
+STATIC int
+xrep_dir_live_update(
+ struct notifier_block *nb,
+ unsigned long action,
+ void *data)
+{
+ struct xfs_dir_update_params *p = data;
+ struct xrep_dir *rd;
+ struct xfs_scrub *sc;
+ int error = 0;
+
+ rd = container_of(nb, struct xrep_dir, pscan.hooks.dirent_hook.nb);
+ sc = rd->sc;
+
+ /*
+ * This thread updated a child dirent in the directory that we're
+ * rebuilding. Stash the update for replay against the temporary
+ * directory.
+ */
+ if (p->dp->i_ino == sc->ip->i_ino &&
+ xchk_iscan_want_live_update(&rd->pscan.iscan, p->ip->i_ino)) {
+ mutex_lock(&rd->pscan.lock);
+ if (p->delta > 0)
+ error = xrep_dir_stash_createname(rd, p->name,
+ p->ip->i_ino);
+ else
+ error = xrep_dir_stash_removename(rd, p->name,
+ p->ip->i_ino);
+ mutex_unlock(&rd->pscan.lock);
+ if (error)
+ goto out_abort;
+ }
+
+ /*
+ * This thread updated another directory's child dirent that points to
+ * the directory that we're rebuilding, so remember the new dotdot
+ * target.
+ */
+ if (p->ip->i_ino == sc->ip->i_ino &&
+ xchk_iscan_want_live_update(&rd->pscan.iscan, p->dp->i_ino)) {
+ if (p->delta > 0) {
+ trace_xrep_dir_stash_createname(sc->tempip,
+ &xfs_name_dotdot,
+ p->dp->i_ino);
+
+ xrep_findparent_scan_found(&rd->pscan, p->dp->i_ino);
+ } else {
+ trace_xrep_dir_stash_removename(sc->tempip,
+ &xfs_name_dotdot,
+ rd->pscan.parent_ino);
+
+ xrep_findparent_scan_found(&rd->pscan, NULLFSINO);
+ }
+ }
+
+ return NOTIFY_DONE;
+out_abort:
+ xchk_iscan_abort(&rd->pscan.iscan);
+ return NOTIFY_DONE;
+}
+
/*
* Free all the directory blocks and reset the data fork. The caller must
* join the inode to the transaction. This function returns with the inode
@@ -1621,6 +1799,9 @@ xrep_dir_rebuild_tree(
if (error)
return error;
+ if (xchk_iscan_aborted(&rd->pscan.iscan))
+ return -ECANCELED;
+
/*
* Swap the tempdir's data fork with the file being repaired. This
* recreates the transaction and re-takes the ILOCK in the scrub
@@ -1676,7 +1857,11 @@ xrep_dir_setup_scan(
if (error)
goto out_xfarray;
- error = xrep_findparent_scan_start(sc, &rd->pscan);
+ if (xfs_has_parent(sc->mp))
+ error = __xrep_findparent_scan_start(sc, &rd->pscan,
+ xrep_dir_live_update);
+ else
+ error = xrep_findparent_scan_start(sc, &rd->pscan);
if (error)
goto out_xfblob;
diff --git a/fs/xfs/scrub/findparent.c b/fs/xfs/scrub/findparent.c
index 87047e9d49e47..9468029f73933 100644
--- a/fs/xfs/scrub/findparent.c
+++ b/fs/xfs/scrub/findparent.c
@@ -238,9 +238,10 @@ xrep_findparent_live_update(
* will be called when there is a dotdot update for the inode being repaired.
*/
int
-xrep_findparent_scan_start(
+__xrep_findparent_scan_start(
struct xfs_scrub *sc,
- struct xrep_parent_scan_info *pscan)
+ struct xrep_parent_scan_info *pscan,
+ notifier_fn_t custom_fn)
{
int error;
@@ -262,7 +263,8 @@ xrep_findparent_scan_start(
* ILOCK, which means that any in-progress inode updates will finish
* before we can scan the inode.
*/
- xfs_hook_setup(&pscan->hooks.dirent_hook, xrep_findparent_live_update);
+ xfs_hook_setup(&pscan->hooks.dirent_hook,
+ custom_fn ? custom_fn : xrep_findparent_live_update);
error = xfs_dir_hook_add(sc->mp, &pscan->hooks);
if (error)
goto out_iscan;
diff --git a/fs/xfs/scrub/findparent.h b/fs/xfs/scrub/findparent.h
index cb3a97f3fed48..29c6077df11e5 100644
--- a/fs/xfs/scrub/findparent.h
+++ b/fs/xfs/scrub/findparent.h
@@ -24,8 +24,14 @@ struct xrep_parent_scan_info {
bool lookup_parent;
};
-int xrep_findparent_scan_start(struct xfs_scrub *sc,
- struct xrep_parent_scan_info *pscan);
+int __xrep_findparent_scan_start(struct xfs_scrub *sc,
+ struct xrep_parent_scan_info *pscan,
+ notifier_fn_t custom_fn);
+static inline int xrep_findparent_scan_start(struct xfs_scrub *sc,
+ struct xrep_parent_scan_info *pscan)
+{
+ return __xrep_findparent_scan_start(sc, pscan, NULL);
+}
int xrep_findparent_scan(struct xrep_parent_scan_info *pscan);
void xrep_findparent_scan_teardown(struct xrep_parent_scan_info *pscan);
diff --git a/fs/xfs/scrub/trace.h b/fs/xfs/scrub/trace.h
index 8fa26a7811118..34e54ebf0daba 100644
--- a/fs/xfs/scrub/trace.h
+++ b/fs/xfs/scrub/trace.h
@@ -2816,6 +2816,8 @@ DEFINE_XREP_DIRENT_EVENT(xrep_dir_salvage_entry);
DEFINE_XREP_DIRENT_EVENT(xrep_dir_stash_createname);
DEFINE_XREP_DIRENT_EVENT(xrep_dir_replay_createname);
DEFINE_XREP_DIRENT_EVENT(xrep_adoption_reparent);
+DEFINE_XREP_DIRENT_EVENT(xrep_dir_stash_removename);
+DEFINE_XREP_DIRENT_EVENT(xrep_dir_replay_removename);
DECLARE_EVENT_CLASS(xrep_adoption_class,
TP_PROTO(struct xfs_inode *dp, struct xfs_inode *ip, bool moved),
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 13/22] xfs: replay unlocked parent pointer updates that accrue during xattr repair
2023-12-31 19:33 ` [PATCHSET v13.0 5/7] xfs: fsck for parent pointers Darrick J. Wong
` (11 preceding siblings ...)
2023-12-31 20:56 ` [PATCH 12/22] xfs: implement live updates for directory repairs Darrick J. Wong
@ 2023-12-31 20:56 ` Darrick J. Wong
2023-12-31 20:56 ` [PATCH 14/22] xfs: replace namebuf with parent pointer in parent pointer repair Darrick J. Wong
` (8 subsequent siblings)
21 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:56 UTC (permalink / raw)
To: djwong; +Cc: catherine.hoang, allison.henderson, linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
There are a few places where the extended attribute repair code drops
the ILOCK to apply stashed xattrs to the temporary file. Although
setxattr and removexattr are still locked out because we retain our hold
on the IOLOCK, this doesn't prevent renames from updating parent
pointers, because the VFS doesn't take i_rwsem on children that are
being moved.
Therefore, set up a dirent hook to capture parent pointer updates for
this file, and replay(?) the updates.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/scrub/attr_repair.c | 457 ++++++++++++++++++++++++++++++++++++++++++++
fs/xfs/scrub/trace.h | 71 +++++++
2 files changed, 526 insertions(+), 2 deletions(-)
diff --git a/fs/xfs/scrub/attr_repair.c b/fs/xfs/scrub/attr_repair.c
index 39e64d7559451..cc964dc427e23 100644
--- a/fs/xfs/scrub/attr_repair.c
+++ b/fs/xfs/scrub/attr_repair.c
@@ -95,6 +95,56 @@ struct xrep_xattr {
/* Number of attributes that we are salvaging. */
unsigned long long attrs_found;
+
+ /* Can we flush stashed attrs to the tempfile? */
+ bool can_flush;
+
+ /* Did the live update fail, and hence the repair is now out of date? */
+ bool live_update_aborted;
+
+ /* Lock protecting parent pointer updates */
+ struct mutex lock;
+
+ /* Fixed-size array of xrep_xattr_pptr structures. */
+ struct xfarray *pptr_recs;
+
+ /* Blobs containing parent pointer names. */
+ struct xfblob *pptr_names;
+
+ /* Hook to capture parent pointer updates. */
+ struct xfs_dir_hook hooks;
+
+ /* xattr key and da args for parent pointer replay. */
+ struct xfs_parent_scratch pptr_scratch;
+
+ /*
+ * Scratch buffer for scanning dirents to create pptr xattrs. At the
+ * very end of the repair, it can also be used to compute the
+ * lost+found filename if we need to reparent the file.
+ */
+ struct xfs_parent_name_irec pptr;
+};
+
+/* Create a parent pointer in the tempfile. */
+#define XREP_XATTR_PPTR_ADD (1)
+
+/* Remove a parent pointer from the tempfile. */
+#define XREP_XATTR_PPTR_REMOVE (2)
+
+/* A stashed parent pointer update. */
+struct xrep_xattr_pptr {
+ /* Cookie for retrieval of the pptr name. */
+ xfblob_cookie name_cookie;
+
+ /* Parent pointer attr key. */
+ xfs_ino_t p_ino;
+ uint32_t p_gen;
+
+ /* Length of the pptr name. */
+ uint8_t namelen;
+
+ /* XREP_XATTR_PPTR_{ADD,REMOVE} */
+ uint8_t action;
};
/* Set up to recreate the extended attributes. */
@@ -102,6 +152,9 @@ int
xrep_setup_xattr(
struct xfs_scrub *sc)
{
+ if (xfs_has_parent(sc->mp))
+ xchk_fsgates_enable(sc, XCHK_FSGATES_DIRENTS);
+
return xrep_tempfile_create(sc, S_IFREG);
}
@@ -712,11 +765,122 @@ xrep_xattr_want_flush_stashed(
{
unsigned long long bytes;
+ if (!rx->can_flush)
+ return false;
+
bytes = xfarray_bytes(rx->xattr_records) +
xfblob_bytes(rx->xattr_blobs);
return bytes > XREP_XATTR_MAX_STASH_BYTES;
}
+/*
+ * Did we observe rename changing parent pointer xattrs while we were flushing
+ * salvaged attrs?
+ */
+static inline bool
+xrep_xattr_saw_pptr_conflict(
+ struct xrep_xattr *rx)
+{
+ bool ret;
+
+ ASSERT(rx->can_flush);
+
+ if (!xfs_has_parent(rx->sc->mp))
+ return false;
+
+ ASSERT(xfs_isilocked(rx->sc->ip, XFS_ILOCK_EXCL));
+
+ mutex_lock(&rx->lock);
+ ret = xfarray_bytes(rx->pptr_recs) > 0;
+ mutex_unlock(&rx->lock);
+
+ return ret;
+}
+
+/*
+ * Reset the entire repair state back to initial conditions, now that we've
+ * detected a parent pointer update to the attr structure while we were
+ * flushing salvaged attrs. See the locking notes in dir_repair.c for more
+ * information on why this is all necessary.
+ */
+STATIC int
+xrep_xattr_full_reset(
+ struct xrep_xattr *rx)
+{
+ struct xfs_scrub *sc = rx->sc;
+ struct xfs_attr_sf_hdr *hdr;
+ struct xfs_ifork *ifp = &sc->tempip->i_af;
+ int error;
+
+ trace_xrep_xattr_full_reset(sc->ip, sc->tempip);
+
+ /* The temporary file's data fork had better not be in btree format. */
+ if (sc->tempip->i_df.if_format == XFS_DINODE_FMT_BTREE) {
+ ASSERT(0);
+ return -EIO;
+ }
+
+ /*
+ * We begin in transaction context with sc->ip ILOCKed but not joined
+ * to the transaction. To reset to the initial state, we must hold
+ * sc->ip's ILOCK to prevent rename from updating parent pointer
+ * information and the tempfile's ILOCK to clear its contents.
+ */
+ xchk_iunlock(rx->sc, XFS_ILOCK_EXCL);
+ xrep_tempfile_ilock_both(sc);
+ xfs_trans_ijoin(sc->tp, sc->ip, 0);
+ xfs_trans_ijoin(sc->tp, sc->tempip, 0);
+
+ /*
+ * Free all the blocks of the attr fork of the temp file, and reset
+ * it back to local format.
+ */
+ if (xfs_ifork_has_extents(&sc->tempip->i_af)) {
+ error = xrep_reap_ifork(sc, sc->tempip, XFS_ATTR_FORK);
+ if (error)
+ return error;
+
+ ASSERT(ifp->if_bytes == 0);
+ ifp->if_format = XFS_DINODE_FMT_LOCAL;
+ xfs_idata_realloc(sc->tempip, sizeof(*hdr), XFS_ATTR_FORK);
+ }
+
+ /* Reinitialize the attr fork to an empty shortform structure. */
+ hdr = (struct xfs_attr_sf_hdr *)ifp->if_u1.if_data;
+ memset(hdr, 0, sizeof(*hdr));
+ hdr->totsize = cpu_to_be16(sizeof(*hdr));
+ xfs_trans_log_inode(sc->tp, sc->tempip, XFS_ILOG_CORE | XFS_ILOG_ADATA);
+
+ /*
+ * Roll this transaction to commit our reset ondisk. The tempfile
+ * should no longer be joined to the transaction, so we drop its ILOCK.
+ * This should leave us in transaction context with sc->ip ILOCKed but
+ * not joined to the transaction.
+ */
+ error = xrep_roll_trans(sc);
+ if (error)
+ return error;
+ xrep_tempfile_iunlock(sc);
+
+ /*
+ * Erase any accumulated parent pointer updates now that we've erased
+ * the tempfile's attr fork. We're resetting the entire repair state
+ * back to where we were initially, except now we won't flush salvaged
+ * xattrs until the very end.
+ */
+ mutex_lock(&rx->lock);
+ xfarray_truncate(rx->pptr_recs);
+ xfblob_truncate(rx->pptr_names);
+ mutex_unlock(&rx->lock);
+
+ rx->can_flush = false;
+ rx->attrs_found = 0;
+
+ ASSERT(xfarray_bytes(rx->xattr_records) == 0);
+ ASSERT(xfblob_bytes(rx->xattr_blobs) == 0);
+ return 0;
+}
+
/* Extract as many attribute keys and values as we can. */
STATIC int
xrep_xattr_recover(
@@ -731,6 +895,7 @@ xrep_xattr_recover(
int nmap;
int error;
+restart:
/*
* Iterate each xattr leaf block in the attr fork to scan them for any
* attributes that we might salvage.
@@ -769,6 +934,14 @@ xrep_xattr_recover(
error = xrep_xattr_flush_stashed(rx);
if (error)
return error;
+
+ if (xrep_xattr_saw_pptr_conflict(rx)) {
+ error = xrep_xattr_full_reset(rx);
+ if (error)
+ return error;
+
+ goto restart;
+ }
}
}
}
@@ -929,6 +1102,195 @@ xrep_xattr_salvage_attributes(
return xrep_xattr_flush_stashed(rx);
}
+/*
+ * Add this stashed incore parent pointer to the temporary file.
+ * The caller must hold the tempdir's IOLOCK, must not hold any ILOCKs, and
+ * must not be in transaction context.
+ */
+STATIC int
+xrep_xattr_replay_pptr_update(
+ struct xrep_xattr *rx,
+ const struct xrep_xattr_pptr *pptr)
+{
+ struct xfs_scrub *sc = rx->sc;
+ int error;
+
+ rx->pptr.p_ino = pptr->p_ino;
+ rx->pptr.p_gen = pptr->p_gen;
+ rx->pptr.p_namelen = pptr->namelen;
+ xfs_parent_irec_hashname(sc->mp, &rx->pptr);
+
+ switch (pptr->action) {
+ case XREP_XATTR_PPTR_ADD:
+ /* Create parent pointer. */
+ trace_xrep_xattr_replay_parentadd(sc->tempip, &rx->pptr);
+
+ error = xfs_parent_set(sc->tempip, sc->ip->i_ino, &rx->pptr,
+ &rx->pptr_scratch);
+ if (error) {
+ ASSERT(error != -EEXIST);
+ return error;
+ }
+ break;
+ case XREP_XATTR_PPTR_REMOVE:
+ /* Remove parent pointer. */
+ trace_xrep_xattr_replay_parentremove(sc->tempip, &rx->pptr);
+
+ error = xfs_parent_unset(sc->tempip, sc->ip->i_ino, &rx->pptr,
+ &rx->pptr_scratch);
+ if (error) {
+ ASSERT(error != -ENOATTR);
+ return error;
+ }
+ break;
+ default:
+ ASSERT(0);
+ return -EIO;
+ }
+
+ return 0;
+}
+
+/*
+ * Flush stashed parent pointer updates that have been recorded by the scanner.
+ * This is done to reduce the memory requirements of the parent pointer
+ * rebuild, since files can have a lot of hardlinks and the fs can be busy.
+ *
+ * Caller must not hold transactions or ILOCKs. Caller must hold the tempfile
+ * IOLOCK.
+ */
+STATIC int
+xrep_xattr_replay_pptr_updates(
+ struct xrep_xattr *rx)
+{
+ xfarray_idx_t array_cur;
+ int error;
+
+ mutex_lock(&rx->lock);
+ foreach_xfarray_idx(rx->pptr_recs, array_cur) {
+ struct xrep_xattr_pptr pptr;
+
+ error = xfarray_load(rx->pptr_recs, array_cur, &pptr);
+ if (error)
+ goto out_unlock;
+
+ error = xfblob_load(rx->pptr_names, pptr.name_cookie,
+ rx->pptr.p_name, pptr.namelen);
+ if (error)
+ goto out_unlock;
+ rx->pptr.p_name[MAXNAMELEN - 1] = 0;
+ mutex_unlock(&rx->lock);
+
+ error = xrep_xattr_replay_pptr_update(rx, &pptr);
+ if (error)
+ return error;
+
+ mutex_lock(&rx->lock);
+ }
+
+ /* Empty out both arrays now that we've added the entries. */
+ xfarray_truncate(rx->pptr_recs);
+ xfblob_truncate(rx->pptr_names);
+ mutex_unlock(&rx->lock);
+ return 0;
+out_unlock:
+ mutex_unlock(&rx->lock);
+ return error;
+}
+
+/*
+ * Remember that we want to create a parent pointer in the tempfile. These
+ * stashed actions will be replayed later.
+ */
+STATIC int
+xrep_xattr_stash_parentadd(
+ struct xrep_xattr *rx,
+ const struct xfs_name *name,
+ const struct xfs_inode *dp)
+{
+ struct xrep_xattr_pptr pptr = {
+ .action = XREP_XATTR_PPTR_ADD,
+ .namelen = name->len,
+ .p_ino = dp->i_ino,
+ .p_gen = VFS_IC(dp)->i_generation,
+ };
+ int error;
+
+ trace_xrep_xattr_stash_parentadd(rx->sc->tempip, dp, name);
+
+ error = xfblob_store(rx->pptr_names, &pptr.name_cookie, name->name,
+ name->len);
+ if (error)
+ return error;
+
+ return xfarray_append(rx->pptr_recs, &pptr);
+}
+
+/*
+ * Remember that we want to remove a parent pointer from the tempfile. These
+ * stashed actions will be replayed later.
+ */
+STATIC int
+xrep_xattr_stash_parentremove(
+ struct xrep_xattr *rx,
+ const struct xfs_name *name,
+ const struct xfs_inode *dp)
+{
+ struct xrep_xattr_pptr pptr = {
+ .action = XREP_XATTR_PPTR_REMOVE,
+ .namelen = name->len,
+ .p_ino = dp->i_ino,
+ .p_gen = VFS_IC(dp)->i_generation,
+ };
+ int error;
+
+ trace_xrep_xattr_stash_parentremove(rx->sc->tempip, dp, name);
+
+ error = xfblob_store(rx->pptr_names, &pptr.name_cookie, name->name,
+ name->len);
+ if (error)
+ return error;
+
+ return xfarray_append(rx->pptr_recs, &pptr);
+}
+
+/*
+ * Capture dirent updates being made by other threads. We will have to replay
+ * the parent pointer updates before swapping attr forks.
+ */
+STATIC int
+xrep_xattr_live_dirent_update(
+ struct notifier_block *nb,
+ unsigned long action,
+ void *data)
+{
+ struct xfs_dir_update_params *p = data;
+ struct xrep_xattr *rx;
+ struct xfs_scrub *sc;
+ int error;
+
+ rx = container_of(nb, struct xrep_xattr, hooks.dirent_hook.nb);
+ sc = rx->sc;
+
+ /*
+ * This thread updated a dirent that points to the file that we're
+ * repairing, so stash the update for replay against the temporary
+ * file.
+ */
+ if (p->ip->i_ino != sc->ip->i_ino)
+ return NOTIFY_DONE;
+
+ mutex_lock(&rx->lock);
+ if (p->delta > 0)
+ error = xrep_xattr_stash_parentadd(rx, p->name, p->dp);
+ else
+ error = xrep_xattr_stash_parentremove(rx, p->name, p->dp);
+ if (error)
+ rx->live_update_aborted = true;
+ mutex_unlock(&rx->lock);
+ return NOTIFY_DONE;
+}
+
/*
* Prepare both inodes' attribute forks for extent swapping. Promote the
* tempfile from short format to leaf format, and if the file being repaired
@@ -1031,6 +1393,45 @@ xrep_xattr_swap(
return xrep_tempswap_contents(sc, tx);
}
+/*
+ * Finish replaying stashed parent pointer updates, allocate a transaction for
+ * swapping extents, and take the ILOCKs of both files before we commit the new
+ * extended attribute structure.
+ */
+STATIC int
+xrep_xattr_finalize_tempfile(
+ struct xrep_xattr *rx)
+{
+ struct xfs_scrub *sc = rx->sc;
+ int error;
+
+ if (!xfs_has_parent(sc->mp))
+ return xrep_tempswap_trans_alloc(sc, XFS_ATTR_FORK, &rx->tx);
+
+ /*
+ * Repair relies on the ILOCK to quiesce all possible xattr updates.
+ * Replay all queued parent pointer updates into the tempfile before
+ * swapping the contents, even if that means dropping the ILOCKs and
+ * the transaction.
+ */
+ do {
+ error = xrep_xattr_replay_pptr_updates(rx);
+ if (error)
+ return error;
+
+ error = xrep_tempswap_trans_alloc(sc, XFS_ATTR_FORK, &rx->tx);
+ if (error)
+ return error;
+
+ if (xfarray_length(rx->pptr_recs) == 0)
+ break;
+
+ xchk_trans_cancel(sc);
+ xrep_tempfile_iunlock_both(sc);
+ } while (!xchk_should_terminate(sc, &error));
+ return error;
+}
+
/*
* Swap the new extended attribute data (which we created in the tempfile) into
* the file being repaired.
@@ -1082,8 +1483,12 @@ xrep_xattr_rebuild_tree(
if (error)
return error;
- /* Allocate swapext transaction and lock both inodes. */
- error = xrep_tempswap_trans_alloc(rx->sc, XFS_ATTR_FORK, &rx->tx);
+ /*
+ * Allocate transaction, lock inodes, and make sure that we've replayed
+ * all the stashed parent pointer updates to the temp file. After this
+ * point, we're ready to swapext.
+ */
+ error = xrep_xattr_finalize_tempfile(rx);
if (error)
return error;
@@ -1124,8 +1529,15 @@ STATIC void
xrep_xattr_teardown(
struct xrep_xattr *rx)
{
+ if (xfs_has_parent(rx->sc->mp))
+ xfs_dir_hook_del(rx->sc->mp, &rx->hooks);
+ if (rx->pptr_names)
+ xfblob_destroy(rx->pptr_names);
+ if (rx->pptr_recs)
+ xfarray_destroy(rx->pptr_recs);
xfblob_destroy(rx->xattr_blobs);
xfarray_destroy(rx->xattr_records);
+ mutex_destroy(&rx->lock);
kfree(rx);
}
@@ -1144,6 +1556,9 @@ xrep_xattr_setup_scan(
if (!rx)
return -ENOMEM;
rx->sc = sc;
+ rx->can_flush = true;
+
+ mutex_init(&rx->lock);
/*
* Allocate enough memory to handle loading local attr values from the
@@ -1171,11 +1586,44 @@ xrep_xattr_setup_scan(
if (error)
goto out_keys;
+ if (xfs_has_parent(sc->mp)) {
+ ASSERT(sc->flags & XCHK_FSGATES_DIRENTS);
+
+ descr = xchk_xfile_ino_descr(sc,
+ "xattr retained parent pointer entries");
+ error = xfarray_create(descr, 0,
+ sizeof(struct xrep_xattr_pptr),
+ &rx->pptr_recs);
+ kfree(descr);
+ if (error)
+ goto out_values;
+
+ descr = xchk_xfile_ino_descr(sc,
+ "xattr retained parent pointer names");
+ error = xfblob_create(descr, &rx->pptr_names);
+ kfree(descr);
+ if (error)
+ goto out_pprecs;
+
+ xfs_hook_setup(&rx->hooks.dirent_hook,
+ xrep_xattr_live_dirent_update);
+ error = xfs_dir_hook_add(sc->mp, &rx->hooks);
+ if (error)
+ goto out_ppnames;
+ }
+
*rxp = rx;
return 0;
+out_ppnames:
+ xfblob_destroy(rx->pptr_names);
+out_pprecs:
+ xfarray_destroy(rx->pptr_recs);
+out_values:
+ xfblob_destroy(rx->xattr_blobs);
out_keys:
xfarray_destroy(rx->xattr_records);
out_rx:
+ mutex_destroy(&rx->lock);
kfree(rx);
return error;
}
@@ -1212,6 +1660,11 @@ xrep_xattr(
if (error)
goto out_scan;
+ if (rx->live_update_aborted) {
+ error = -EIO;
+ goto out_scan;
+ }
+
/* Last chance to abort before we start committing fixes. */
if (xchk_should_terminate(sc, &error))
goto out_scan;
diff --git a/fs/xfs/scrub/trace.h b/fs/xfs/scrub/trace.h
index 34e54ebf0daba..ebd5a91064281 100644
--- a/fs/xfs/scrub/trace.h
+++ b/fs/xfs/scrub/trace.h
@@ -2726,6 +2726,43 @@ DEFINE_EVENT(xrep_xattr_class, name, \
TP_ARGS(ip, arg_ip))
DEFINE_XREP_XATTR_EVENT(xrep_xattr_rebuild_tree);
DEFINE_XREP_XATTR_EVENT(xrep_xattr_reset_fork);
+DEFINE_XREP_XATTR_EVENT(xrep_xattr_full_reset);
+
+DECLARE_EVENT_CLASS(xrep_xattr_pptr_scan_class,
+ TP_PROTO(struct xfs_inode *ip, const struct xfs_inode *dp,
+ const struct xfs_name *name),
+ TP_ARGS(ip, dp, name),
+ TP_STRUCT__entry(
+ __field(dev_t, dev)
+ __field(xfs_ino_t, ino)
+ __field(xfs_ino_t, parent_ino)
+ __field(unsigned int, parent_gen)
+ __field(unsigned int, namelen)
+ __dynamic_array(char, name, name->len)
+ ),
+ TP_fast_assign(
+ __entry->dev = ip->i_mount->m_super->s_dev;
+ __entry->ino = ip->i_ino;
+ __entry->parent_ino = dp->i_ino;
+ __entry->parent_gen = VFS_IC(dp)->i_generation;
+ __entry->namelen = name->len;
+ memcpy(__get_str(name), name->name, name->len);
+ ),
+ TP_printk("dev %d:%d ino 0x%llx parent_ino 0x%llx parent_gen 0x%x name '%.*s'",
+ MAJOR(__entry->dev), MINOR(__entry->dev),
+ __entry->ino,
+ __entry->parent_ino,
+ __entry->parent_gen,
+ __entry->namelen,
+ __get_str(name))
+)
+#define DEFINE_XREP_XATTR_PPTR_SCAN_EVENT(name) \
+DEFINE_EVENT(xrep_xattr_pptr_scan_class, name, \
+ TP_PROTO(struct xfs_inode *ip, const struct xfs_inode *dp, \
+ const struct xfs_name *name), \
+ TP_ARGS(ip, dp, name))
+DEFINE_XREP_XATTR_PPTR_SCAN_EVENT(xrep_xattr_stash_parentadd);
+DEFINE_XREP_XATTR_PPTR_SCAN_EVENT(xrep_xattr_stash_parentremove);
TRACE_EVENT(xrep_dir_recover_dirblock,
TP_PROTO(struct xfs_inode *dp, xfs_dablk_t dabno, uint32_t magic,
@@ -2872,6 +2909,40 @@ DEFINE_XREP_PARENT_SALVAGE_EVENT(xrep_dir_salvaged_parent);
DEFINE_XREP_PARENT_SALVAGE_EVENT(xrep_findparent_dirent);
DEFINE_XREP_PARENT_SALVAGE_EVENT(xrep_findparent_from_dcache);
+DECLARE_EVENT_CLASS(xrep_pptr_class,
+ TP_PROTO(struct xfs_inode *ip, const struct xfs_parent_name_irec *pptr),
+ TP_ARGS(ip, pptr),
+ TP_STRUCT__entry(
+ __field(dev_t, dev)
+ __field(xfs_ino_t, ino)
+ __field(xfs_ino_t, parent_ino)
+ __field(unsigned int, parent_gen)
+ __field(unsigned int, namelen)
+ __dynamic_array(char, name, pptr->p_namelen)
+ ),
+ TP_fast_assign(
+ __entry->dev = ip->i_mount->m_super->s_dev;
+ __entry->ino = ip->i_ino;
+ __entry->parent_ino = pptr->p_ino;
+ __entry->parent_gen = pptr->p_gen;
+ __entry->namelen = pptr->p_namelen;
+ memcpy(__get_str(name), pptr->p_name, pptr->p_namelen);
+ ),
+ TP_printk("dev %d:%d ino 0x%llx parent_ino 0x%llx parent_gen 0x%x name '%.*s'",
+ MAJOR(__entry->dev), MINOR(__entry->dev),
+ __entry->ino,
+ __entry->parent_ino,
+ __entry->parent_gen,
+ __entry->namelen,
+ __get_str(name))
+)
+#define DEFINE_XREP_PPTR_EVENT(name) \
+DEFINE_EVENT(xrep_pptr_class, name, \
+ TP_PROTO(struct xfs_inode *ip, const struct xfs_parent_name_irec *pptr), \
+ TP_ARGS(ip, pptr))
+DEFINE_XREP_PPTR_EVENT(xrep_xattr_replay_parentadd);
+DEFINE_XREP_PPTR_EVENT(xrep_xattr_replay_parentremove);
+
TRACE_EVENT(xrep_nlinks_set_record,
TP_PROTO(struct xfs_mount *mp, xfs_ino_t ino,
const struct xchk_nlink *obs),
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 14/22] xfs: replace namebuf with parent pointer in parent pointer repair
2023-12-31 19:33 ` [PATCHSET v13.0 5/7] xfs: fsck for parent pointers Darrick J. Wong
` (12 preceding siblings ...)
2023-12-31 20:56 ` [PATCH 13/22] xfs: replay unlocked parent pointer updates that accrue during xattr repair Darrick J. Wong
@ 2023-12-31 20:56 ` Darrick J. Wong
2023-12-31 20:57 ` [PATCH 15/22] xfs: repair directory parent pointers by scanning for dirents Darrick J. Wong
` (7 subsequent siblings)
21 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:56 UTC (permalink / raw)
To: djwong; +Cc: catherine.hoang, allison.henderson, linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
Replace the dirent name buffer at the end of struct xrep_parent with a
xfs_parent_name_irec object. The namebuf and p_name usage do not
overlap, so we can save 256 bytes of memory by allowing them to overlap.
Doing so makes the code a bit more complex, so this is called out
separately.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/scrub/parent_repair.c | 11 ++++++++---
1 file changed, 8 insertions(+), 3 deletions(-)
diff --git a/fs/xfs/scrub/parent_repair.c b/fs/xfs/scrub/parent_repair.c
index 099620fc119e9..68cc3aee1d5c8 100644
--- a/fs/xfs/scrub/parent_repair.c
+++ b/fs/xfs/scrub/parent_repair.c
@@ -24,6 +24,7 @@
#include "xfs_trans_space.h"
#include "xfs_health.h"
#include "xfs_swapext.h"
+#include "xfs_parent.h"
#include "scrub/xfs_scrub.h"
#include "scrub/scrub.h"
#include "scrub/common.h"
@@ -63,8 +64,12 @@ struct xrep_parent {
/* Orphanage reparenting request. */
struct xrep_adoption adoption;
- /* Directory entry name, plus the trailing null. */
- unsigned char namebuf[MAXNAMELEN];
+ /*
+ * Scratch buffer for scanning dirents to create pptr xattrs. At the
+ * very end of the repair, it can also be used to compute the
+ * lost+found filename if we need to reparent the file.
+ */
+ struct xfs_parent_name_irec pptr;
};
/* Tear down all the incore stuff we created. */
@@ -236,7 +241,7 @@ xrep_parent_move_to_orphanage(
if (error)
return error;
- error = xrep_adoption_compute_name(&rp->adoption, rp->namebuf);
+ error = xrep_adoption_compute_name(&rp->adoption, rp->pptr.p_name);
if (error)
return error;
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 15/22] xfs: repair directory parent pointers by scanning for dirents
2023-12-31 19:33 ` [PATCHSET v13.0 5/7] xfs: fsck for parent pointers Darrick J. Wong
` (13 preceding siblings ...)
2023-12-31 20:56 ` [PATCH 14/22] xfs: replace namebuf with parent pointer in parent pointer repair Darrick J. Wong
@ 2023-12-31 20:57 ` Darrick J. Wong
2023-12-31 20:57 ` [PATCH 16/22] xfs: implement live updates for parent pointer repairs Darrick J. Wong
` (6 subsequent siblings)
21 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:57 UTC (permalink / raw)
To: djwong; +Cc: catherine.hoang, allison.henderson, linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
If parent pointers are enabled on the filesystem, we can repair the
entire dataset by walking the directories of the filesystem looking for
dirents that we can turn into parent pointers. Once we have a full
incore dataset, we'll figure out what to do with it, but that's for a
subsequent patch.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/scrub/parent_repair.c | 425 ++++++++++++++++++++++++++++++++++++++++++
fs/xfs/scrub/trace.h | 36 ++++
2 files changed, 458 insertions(+), 3 deletions(-)
diff --git a/fs/xfs/scrub/parent_repair.c b/fs/xfs/scrub/parent_repair.c
index 68cc3aee1d5c8..8d83aab8caa20 100644
--- a/fs/xfs/scrub/parent_repair.c
+++ b/fs/xfs/scrub/parent_repair.c
@@ -35,6 +35,9 @@
#include "scrub/readdir.h"
#include "scrub/tempfile.h"
#include "scrub/orphanage.h"
+#include "scrub/xfile.h"
+#include "scrub/xfarray.h"
+#include "scrub/xfblob.h"
/*
* Repairing The Directory Parent Pointer
@@ -50,20 +53,71 @@
* See the section on locking issues in dir_repair.c for more information about
* conflicts with the VFS. The findparent code wll keep our incore parent
* inode up to date.
+ *
+ * If parent pointers are enabled, we instead reconstruct the parent pointer
+ * information by visiting every directory entry of every directory in the
+ * system and translating the relevant dirents into parent pointers. In this
+ * case, it is advantageous to stash all parent pointers created from dirents
+ * from a single parent file before replaying them into the temporary file. To
+ * save memory, the live filesystem scan reuses the findparent object. Parent
+ * pointer repair chooses either directory scanning or findparent, but not
+ * both.
+ *
+ * When salvaging completes, the remaining stashed entries are replayed to the
+ * temporary file. All non-parent pointer extended attributes are copied to
+ * the temporary file's extended attributes. An atomic extent swap is used to
+ * commit the new directory blocks to the directory being repaired. This will
+ * disrupt attrmulti cursors.
*/
+/* A stashed parent pointer update. */
+struct xrep_pptr {
+ /* Cookie for retrieval of the pptr name. */
+ xfblob_cookie name_cookie;
+
+ /* Parent pointer attr key. */
+ xfs_ino_t p_ino;
+ uint32_t p_gen;
+
+ /* Length of the pptr name. */
+ uint8_t namelen;
+};
+
+/*
+ * Stash up to 8 pages of recovered parent pointers in pptr_recs and
+ * pptr_names before we write them to the temp file.
+ */
+#define XREP_PARENT_MAX_STASH_BYTES (PAGE_SIZE * 8)
+
struct xrep_parent {
struct xfs_scrub *sc;
+ /* Fixed-size array of xrep_pptr structures. */
+ struct xfarray *pptr_recs;
+
+ /* Blobs containing parent pointer names. */
+ struct xfblob *pptr_names;
+
/*
* Information used to scan the filesystem to find the inumber of the
- * dotdot entry for this directory.
+ * dotdot entry for this directory. On filesystems without parent
+ * pointers, we use the findparent_* functions on this object and
+ * access only the parent_ino field directly.
+ *
+ * When parent pointers are enabled, the directory entry scanner uses
+ * the iscan, hooks, and lock fields of this object directly.
+ * @pscan.lock coordinates access to pptr_recs, pptr_names, pptr, and
+ * pptr_scratch. This reduces the memory requirements of this
+ * structure.
*/
struct xrep_parent_scan_info pscan;
/* Orphanage reparenting request. */
struct xrep_adoption adoption;
+ /* xattr key and da args for parent pointer replay. */
+ struct xfs_parent_scratch pptr_scratch;
+
/*
* Scratch buffer for scanning dirents to create pptr xattrs. At the
* very end of the repair, it can also be used to compute the
@@ -78,6 +132,12 @@ xrep_parent_teardown(
struct xrep_parent *rp)
{
xrep_findparent_scan_teardown(&rp->pscan);
+ if (rp->pptr_names)
+ xfblob_destroy(rp->pptr_names);
+ rp->pptr_names = NULL;
+ if (rp->pptr_recs)
+ xfarray_destroy(rp->pptr_recs);
+ rp->pptr_recs = NULL;
}
/* Set up for a parent repair. */
@@ -86,6 +146,7 @@ xrep_setup_parent(
struct xfs_scrub *sc)
{
struct xrep_parent *rp;
+ int error;
xchk_fsgates_enable(sc, XCHK_FSGATES_DIRENTS);
@@ -95,6 +156,10 @@ xrep_setup_parent(
rp->sc = sc;
sc->buf = rp;
+ error = xrep_tempfile_create(sc, S_IFREG);
+ if (error)
+ return error;
+
return xrep_orphanage_try_create(sc);
}
@@ -150,6 +215,318 @@ xrep_parent_find_dotdot(
return error;
}
+/*
+ * Add this stashed incore parent pointer to the temporary file.
+ * The caller must hold the tempdir's IOLOCK, must not hold any ILOCKs, and
+ * must not be in transaction context.
+ */
+STATIC int
+xrep_parent_replay_update(
+ struct xrep_parent *rp,
+ const struct xrep_pptr *pptr)
+{
+ struct xfs_scrub *sc = rp->sc;
+ int error;
+
+ rp->pptr.p_ino = pptr->p_ino;
+ rp->pptr.p_gen = pptr->p_gen;
+ rp->pptr.p_namelen = pptr->namelen;
+ xfs_parent_irec_hashname(sc->mp, &rp->pptr);
+
+ /* Create parent pointer. */
+ trace_xrep_parent_replay_parentadd(sc->tempip, &rp->pptr);
+
+ error = xfs_parent_set(sc->tempip, sc->ip->i_ino, &rp->pptr,
+ &rp->pptr_scratch);
+ if (error)
+ return error;
+
+ return 0;
+}
+
+/*
+ * Flush stashed parent pointer updates that have been recorded by the scanner.
+ * This is done to reduce the memory requirements of the parent pointer
+ * rebuild, since files can have a lot of hardlinks and the fs can be busy.
+ *
+ * Caller must not hold transactions or ILOCKs. Caller must hold the tempfile
+ * IOLOCK.
+ */
+STATIC int
+xrep_parent_replay_updates(
+ struct xrep_parent *rp)
+{
+ xfarray_idx_t array_cur;
+ int error;
+
+ mutex_lock(&rp->pscan.lock);
+ foreach_xfarray_idx(rp->pptr_recs, array_cur) {
+ struct xrep_pptr pptr;
+
+ error = xfarray_load(rp->pptr_recs, array_cur, &pptr);
+ if (error)
+ goto out_unlock;
+
+ error = xfblob_load(rp->pptr_names, pptr.name_cookie,
+ rp->pptr.p_name, pptr.namelen);
+ if (error)
+ goto out_unlock;
+ rp->pptr.p_name[MAXNAMELEN - 1] = 0;
+ mutex_unlock(&rp->pscan.lock);
+
+ error = xrep_parent_replay_update(rp, &pptr);
+ if (error)
+ return error;
+
+ mutex_lock(&rp->pscan.lock);
+ }
+
+ /* Empty out both arrays now that we've added the entries. */
+ xfarray_truncate(rp->pptr_recs);
+ xfblob_truncate(rp->pptr_names);
+ mutex_unlock(&rp->pscan.lock);
+ return 0;
+out_unlock:
+ mutex_unlock(&rp->pscan.lock);
+ return error;
+}
+
+/*
+ * Remember that we want to create a parent pointer in the tempfile. These
+ * stashed actions will be replayed later.
+ */
+STATIC int
+xrep_parent_stash_parentadd(
+ struct xrep_parent *rp,
+ const struct xfs_name *name,
+ const struct xfs_inode *dp)
+{
+ struct xrep_pptr pptr = {
+ .namelen = name->len,
+ .p_ino = dp->i_ino,
+ .p_gen = VFS_IC(dp)->i_generation,
+ };
+ int error;
+
+ trace_xrep_parent_stash_parentadd(rp->sc->tempip, dp, name);
+
+ error = xfblob_store(rp->pptr_names, &pptr.name_cookie, name->name,
+ name->len);
+ if (error)
+ return error;
+
+ return xfarray_append(rp->pptr_recs, &pptr);
+}
+
+/*
+ * Examine an entry of a directory. If this dirent leads us back to the file
+ * whose parent pointers we're rebuilding, add a pptr to the temporary
+ * directory.
+ */
+STATIC int
+xrep_parent_scan_dirent(
+ struct xfs_scrub *sc,
+ struct xfs_inode *dp,
+ xfs_dir2_dataptr_t dapos,
+ const struct xfs_name *name,
+ xfs_ino_t ino,
+ void *priv)
+{
+ struct xrep_parent *rp = priv;
+ int error;
+
+ /* Dirent doesn't point to this directory. */
+ if (ino != rp->sc->ip->i_ino)
+ return 0;
+
+ /* No weird looking names. */
+ if (name->len == 0 || !xfs_dir2_namecheck(name->name, name->len))
+ return -EFSCORRUPTED;
+
+ /* No mismatching ftypes. */
+ if (name->type != xfs_mode_to_ftype(VFS_I(sc->ip)->i_mode))
+ return -EFSCORRUPTED;
+
+ /* Don't pick up dot or dotdot entries; we only want child dirents. */
+ if (xfs_dir2_samename(name, &xfs_name_dotdot) ||
+ xfs_dir2_samename(name, &xfs_name_dot))
+ return 0;
+
+ /*
+ * Transform this dirent into a parent pointer and queue it for later
+ * addition to the temporary file.
+ */
+ mutex_lock(&rp->pscan.lock);
+ error = xrep_parent_stash_parentadd(rp, name, dp);
+ mutex_unlock(&rp->pscan.lock);
+ return error;
+}
+
+/*
+ * Decide if we want to look for dirents in this directory. Skip the file
+ * being repaired and any files being used to stage repairs.
+ */
+static inline bool
+xrep_parent_want_scan(
+ struct xrep_parent *rp,
+ const struct xfs_inode *ip)
+{
+ return ip != rp->sc->ip && !xrep_is_tempfile(ip);
+}
+
+/*
+ * Take ILOCK on a file that we want to scan.
+ *
+ * Select ILOCK_EXCL if the file is a directory with an unloaded data bmbt.
+ * Otherwise, take ILOCK_SHARED.
+ */
+static inline unsigned int
+xrep_parent_scan_ilock(
+ struct xrep_parent *rp,
+ struct xfs_inode *ip)
+{
+ uint lock_mode = XFS_ILOCK_SHARED;
+
+ /* Still need to take the shared ILOCK to advance the iscan cursor. */
+ if (!xrep_parent_want_scan(rp, ip))
+ goto lock;
+
+ if (S_ISDIR(VFS_I(ip)->i_mode) && xfs_need_iread_extents(&ip->i_df)) {
+ lock_mode = XFS_ILOCK_EXCL;
+ goto lock;
+ }
+
+lock:
+ xfs_ilock(ip, lock_mode);
+ return lock_mode;
+}
+
+/*
+ * Scan this file for relevant child dirents that point to the file whose
+ * parent pointers we're rebuilding.
+ */
+STATIC int
+xrep_parent_scan_file(
+ struct xrep_parent *rp,
+ struct xfs_inode *ip)
+{
+ unsigned int lock_mode;
+ int error = 0;
+
+ lock_mode = xrep_parent_scan_ilock(rp, ip);
+
+ if (!xrep_parent_want_scan(rp, ip))
+ goto scan_done;
+
+ if (S_ISDIR(VFS_I(ip)->i_mode)) {
+ /*
+ * If the directory looks as though it has been zapped by the
+ * inode record repair code, we cannot scan for child dirents.
+ */
+ if (xchk_dir_looks_zapped(ip)) {
+ error = -EBUSY;
+ goto scan_done;
+ }
+
+ error = xchk_dir_walk(rp->sc, ip, xrep_parent_scan_dirent, rp);
+ if (error)
+ goto scan_done;
+ }
+
+scan_done:
+ xchk_iscan_mark_visited(&rp->pscan.iscan, ip);
+ xfs_iunlock(ip, lock_mode);
+ return error;
+}
+
+/* Decide if we've stashed too much pptr data in memory. */
+static inline bool
+xrep_parent_want_flush_stashed(
+ struct xrep_parent *rp)
+{
+ unsigned long long bytes;
+
+ bytes = xfarray_bytes(rp->pptr_recs) + xfblob_bytes(rp->pptr_names);
+ return bytes > XREP_PARENT_MAX_STASH_BYTES;
+}
+
+/*
+ * Scan all directories in the filesystem to look for dirents that we can turn
+ * into parent pointers.
+ */
+STATIC int
+xrep_parent_scan_dirtree(
+ struct xrep_parent *rp)
+{
+ struct xfs_scrub *sc = rp->sc;
+ struct xfs_inode *ip;
+ int error;
+
+ /*
+ * Filesystem scans are time consuming. Drop the file ILOCK and all
+ * other resources for the duration of the scan and hope for the best.
+ * The live update hooks will keep our scan information up to date.
+ */
+ xchk_trans_cancel(sc);
+ if (sc->ilock_flags & (XFS_ILOCK_SHARED | XFS_ILOCK_EXCL))
+ xchk_iunlock(sc, sc->ilock_flags & (XFS_ILOCK_SHARED |
+ XFS_ILOCK_EXCL));
+ error = xchk_trans_alloc_empty(sc);
+ if (error)
+ return error;
+
+ while ((error = xchk_iscan_iter(&rp->pscan.iscan, &ip)) == 1) {
+ bool flush;
+
+ error = xrep_parent_scan_file(rp, ip);
+ xchk_irele(sc, ip);
+ if (error)
+ break;
+
+ /* Flush stashed pptr updates to constrain memory usage. */
+ mutex_lock(&rp->pscan.lock);
+ flush = xrep_parent_want_flush_stashed(rp);
+ mutex_unlock(&rp->pscan.lock);
+ if (flush) {
+ xchk_trans_cancel(sc);
+
+ error = xrep_tempfile_iolock_polled(sc);
+ if (error)
+ break;
+
+ error = xrep_parent_replay_updates(rp);
+ xrep_tempfile_iounlock(sc);
+ if (error)
+ break;
+
+ error = xchk_trans_alloc_empty(sc);
+ if (error)
+ break;
+ }
+
+ if (xchk_should_terminate(sc, &error))
+ break;
+ }
+ xchk_iscan_iter_finish(&rp->pscan.iscan);
+ if (error) {
+ /*
+ * If we couldn't grab an inode that was busy with a state
+ * change, change the error code so that we exit to userspace
+ * as quickly as possible.
+ */
+ if (error == -EBUSY)
+ return -ECANCELED;
+ return error;
+ }
+
+ /*
+ * Cancel the empty transaction so that we can (later) use the atomic
+ * extent swap helpers to lock files and commit the new directory.
+ */
+ xchk_trans_cancel(rp->sc);
+ return 0;
+}
+
/* Reset a directory's dotdot entry, if needed. */
STATIC int
xrep_parent_reset_dotdot(
@@ -301,8 +678,39 @@ xrep_parent_setup_scan(
struct xrep_parent *rp)
{
struct xfs_scrub *sc = rp->sc;
+ char *descr;
+ int error;
- return xrep_findparent_scan_start(sc, &rp->pscan);
+ if (!xfs_has_parent(sc->mp))
+ return xrep_findparent_scan_start(sc, &rp->pscan);
+
+ /* Set up some staging memory for logging parent pointer updates. */
+ descr = xchk_xfile_ino_descr(sc, "parent pointer entries");
+ error = xfarray_create(descr, 0, sizeof(struct xrep_pptr),
+ &rp->pptr_recs);
+ kfree(descr);
+ if (error)
+ return error;
+
+ descr = xchk_xfile_ino_descr(sc, "parent pointer names");
+ error = xfblob_create(descr, &rp->pptr_names);
+ kfree(descr);
+ if (error)
+ goto out_recs;
+
+ error = xrep_findparent_scan_start(sc, &rp->pscan);
+ if (error)
+ goto out_names;
+
+ return 0;
+
+out_names:
+ xfblob_destroy(rp->pptr_names);
+ rp->pptr_names = NULL;
+out_recs:
+ xfarray_destroy(rp->pptr_recs);
+ rp->pptr_recs = NULL;
+ return error;
}
int
@@ -312,11 +720,22 @@ xrep_parent(
struct xrep_parent *rp = sc->buf;
int error;
+ /*
+ * When the parent pointers feature is enabled, repairs are committed
+ * by atomically committing a new xattr structure and reaping the old
+ * attr fork. Reaping requires rmap to be enabled.
+ */
+ if (xfs_has_parent(sc->mp) && !xfs_has_rmapbt(sc->mp))
+ return -EOPNOTSUPP;
+
error = xrep_parent_setup_scan(rp);
if (error)
return error;
- error = xrep_parent_find_dotdot(rp);
+ if (xfs_has_parent(sc->mp))
+ error = xrep_parent_scan_dirtree(rp);
+ else
+ error = xrep_parent_find_dotdot(rp);
if (error)
goto out_teardown;
diff --git a/fs/xfs/scrub/trace.h b/fs/xfs/scrub/trace.h
index ebd5a91064281..def5d72d3c55c 100644
--- a/fs/xfs/scrub/trace.h
+++ b/fs/xfs/scrub/trace.h
@@ -2942,6 +2942,42 @@ DEFINE_EVENT(xrep_pptr_class, name, \
TP_ARGS(ip, pptr))
DEFINE_XREP_PPTR_EVENT(xrep_xattr_replay_parentadd);
DEFINE_XREP_PPTR_EVENT(xrep_xattr_replay_parentremove);
+DEFINE_XREP_PPTR_EVENT(xrep_parent_replay_parentadd);
+
+DECLARE_EVENT_CLASS(xrep_pptr_scan_class,
+ TP_PROTO(struct xfs_inode *ip, const struct xfs_inode *dp,
+ const struct xfs_name *name),
+ TP_ARGS(ip, dp, name),
+ TP_STRUCT__entry(
+ __field(dev_t, dev)
+ __field(xfs_ino_t, ino)
+ __field(xfs_ino_t, parent_ino)
+ __field(unsigned int, parent_gen)
+ __field(unsigned int, namelen)
+ __dynamic_array(char, name, name->len)
+ ),
+ TP_fast_assign(
+ __entry->dev = ip->i_mount->m_super->s_dev;
+ __entry->ino = ip->i_ino;
+ __entry->parent_ino = dp->i_ino;
+ __entry->parent_gen = VFS_IC(dp)->i_generation;
+ __entry->namelen = name->len;
+ memcpy(__get_str(name), name->name, name->len);
+ ),
+ TP_printk("dev %d:%d ino 0x%llx parent_ino 0x%llx parent_gen 0x%x name '%.*s'",
+ MAJOR(__entry->dev), MINOR(__entry->dev),
+ __entry->ino,
+ __entry->parent_ino,
+ __entry->parent_gen,
+ __entry->namelen,
+ __get_str(name))
+)
+#define DEFINE_XREP_PPTR_SCAN_EVENT(name) \
+DEFINE_EVENT(xrep_pptr_scan_class, name, \
+ TP_PROTO(struct xfs_inode *ip, const struct xfs_inode *dp, \
+ const struct xfs_name *name), \
+ TP_ARGS(ip, dp, name))
+DEFINE_XREP_PPTR_SCAN_EVENT(xrep_parent_stash_parentadd);
TRACE_EVENT(xrep_nlinks_set_record,
TP_PROTO(struct xfs_mount *mp, xfs_ino_t ino,
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 16/22] xfs: implement live updates for parent pointer repairs
2023-12-31 19:33 ` [PATCHSET v13.0 5/7] xfs: fsck for parent pointers Darrick J. Wong
` (14 preceding siblings ...)
2023-12-31 20:57 ` [PATCH 15/22] xfs: repair directory parent pointers by scanning for dirents Darrick J. Wong
@ 2023-12-31 20:57 ` Darrick J. Wong
2023-12-31 20:57 ` [PATCH 17/22] xfs: remove pointless unlocked assertion Darrick J. Wong
` (5 subsequent siblings)
21 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:57 UTC (permalink / raw)
To: djwong; +Cc: catherine.hoang, allison.henderson, linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
While we're scanning the filesystem for dirents that we can turn into
parent pointers, we cannot hold the IOLOCK or ILOCK of the file being
repaired. Therefore, we need to set up a dirent hook so that we can
keep the temporary file's parent pionters up to date with the rest of
the filesystem. Hence we add the ability to *remove* pptrs from the
temporary file.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/scrub/parent_repair.c | 111 +++++++++++++++++++++++++++++++++++++++---
fs/xfs/scrub/trace.h | 2 +
2 files changed, 106 insertions(+), 7 deletions(-)
diff --git a/fs/xfs/scrub/parent_repair.c b/fs/xfs/scrub/parent_repair.c
index 8d83aab8caa20..b94eebec3cc79 100644
--- a/fs/xfs/scrub/parent_repair.c
+++ b/fs/xfs/scrub/parent_repair.c
@@ -70,6 +70,12 @@
* disrupt attrmulti cursors.
*/
+/* Create a parent pointer in the tempfile. */
+#define XREP_PPTR_ADD (1)
+
+/* Remove a parent pointer from the tempfile. */
+#define XREP_PPTR_REMOVE (2)
+
/* A stashed parent pointer update. */
struct xrep_pptr {
/* Cookie for retrieval of the pptr name. */
@@ -81,6 +87,9 @@ struct xrep_pptr {
/* Length of the pptr name. */
uint8_t namelen;
+
+ /* XREP_PPTR_{ADD,REMOVE} */
+ uint8_t action;
};
/*
@@ -233,13 +242,29 @@ xrep_parent_replay_update(
rp->pptr.p_namelen = pptr->namelen;
xfs_parent_irec_hashname(sc->mp, &rp->pptr);
- /* Create parent pointer. */
- trace_xrep_parent_replay_parentadd(sc->tempip, &rp->pptr);
+ switch (pptr->action) {
+ case XREP_PPTR_ADD:
+ /* Create parent pointer. */
+ trace_xrep_parent_replay_parentadd(sc->tempip, &rp->pptr);
- error = xfs_parent_set(sc->tempip, sc->ip->i_ino, &rp->pptr,
- &rp->pptr_scratch);
- if (error)
- return error;
+ error = xfs_parent_set(sc->tempip, sc->ip->i_ino, &rp->pptr,
+ &rp->pptr_scratch);
+ if (error)
+ return error;
+ break;
+ case XREP_PPTR_REMOVE:
+ /* Remove parent pointer. */
+ trace_xrep_parent_replay_parentremove(sc->tempip, &rp->pptr);
+
+ error = xfs_parent_unset(sc->tempip, sc->ip->i_ino, &rp->pptr,
+ &rp->pptr_scratch);
+ if (error)
+ return error;
+ break;
+ default:
+ ASSERT(0);
+ return -EIO;
+ }
return 0;
}
@@ -302,6 +327,7 @@ xrep_parent_stash_parentadd(
const struct xfs_inode *dp)
{
struct xrep_pptr pptr = {
+ .action = XREP_PPTR_ADD,
.namelen = name->len,
.p_ino = dp->i_ino,
.p_gen = VFS_IC(dp)->i_generation,
@@ -318,6 +344,34 @@ xrep_parent_stash_parentadd(
return xfarray_append(rp->pptr_recs, &pptr);
}
+/*
+ * Remember that we want to remove a parent pointer from the tempfile. These
+ * stashed actions will be replayed later.
+ */
+STATIC int
+xrep_parent_stash_parentremove(
+ struct xrep_parent *rp,
+ const struct xfs_name *name,
+ const struct xfs_inode *dp)
+{
+ struct xrep_pptr pptr = {
+ .action = XREP_PPTR_REMOVE,
+ .namelen = name->len,
+ .p_ino = dp->i_ino,
+ .p_gen = VFS_IC(dp)->i_generation,
+ };
+ int error;
+
+ trace_xrep_parent_stash_parentremove(rp->sc->tempip, dp, name);
+
+ error = xfblob_store(rp->pptr_names, &pptr.name_cookie, name->name,
+ name->len);
+ if (error)
+ return error;
+
+ return xfarray_append(rp->pptr_recs, &pptr);
+}
+
/*
* Examine an entry of a directory. If this dirent leads us back to the file
* whose parent pointers we're rebuilding, add a pptr to the temporary
@@ -527,6 +581,48 @@ xrep_parent_scan_dirtree(
return 0;
}
+/*
+ * Capture dirent updates being made by other threads which are relevant to the
+ * file being repaired.
+ */
+STATIC int
+xrep_parent_live_update(
+ struct notifier_block *nb,
+ unsigned long action,
+ void *data)
+{
+ struct xfs_dir_update_params *p = data;
+ struct xrep_parent *rp;
+ struct xfs_scrub *sc;
+ int error;
+
+ rp = container_of(nb, struct xrep_parent, pscan.hooks.dirent_hook.nb);
+ sc = rp->sc;
+
+ /*
+ * This thread updated a dirent that points to the file that we're
+ * repairing, so stash the update for replay against the temporary
+ * file.
+ */
+ if (p->ip->i_ino == sc->ip->i_ino &&
+ xchk_iscan_want_live_update(&rp->pscan.iscan, p->dp->i_ino)) {
+ mutex_lock(&rp->pscan.lock);
+ if (p->delta > 0)
+ error = xrep_parent_stash_parentadd(rp, p->name, p->dp);
+ else
+ error = xrep_parent_stash_parentremove(rp, p->name,
+ p->dp);
+ mutex_unlock(&rp->pscan.lock);
+ if (error)
+ goto out_abort;
+ }
+
+ return NOTIFY_DONE;
+out_abort:
+ xchk_iscan_abort(&rp->pscan.iscan);
+ return NOTIFY_DONE;
+}
+
/* Reset a directory's dotdot entry, if needed. */
STATIC int
xrep_parent_reset_dotdot(
@@ -698,7 +794,8 @@ xrep_parent_setup_scan(
if (error)
goto out_recs;
- error = xrep_findparent_scan_start(sc, &rp->pscan);
+ error = __xrep_findparent_scan_start(sc, &rp->pscan,
+ xrep_parent_live_update);
if (error)
goto out_names;
diff --git a/fs/xfs/scrub/trace.h b/fs/xfs/scrub/trace.h
index def5d72d3c55c..b03e1e69740fa 100644
--- a/fs/xfs/scrub/trace.h
+++ b/fs/xfs/scrub/trace.h
@@ -2943,6 +2943,7 @@ DEFINE_EVENT(xrep_pptr_class, name, \
DEFINE_XREP_PPTR_EVENT(xrep_xattr_replay_parentadd);
DEFINE_XREP_PPTR_EVENT(xrep_xattr_replay_parentremove);
DEFINE_XREP_PPTR_EVENT(xrep_parent_replay_parentadd);
+DEFINE_XREP_PPTR_EVENT(xrep_parent_replay_parentremove);
DECLARE_EVENT_CLASS(xrep_pptr_scan_class,
TP_PROTO(struct xfs_inode *ip, const struct xfs_inode *dp,
@@ -2978,6 +2979,7 @@ DEFINE_EVENT(xrep_pptr_scan_class, name, \
const struct xfs_name *name), \
TP_ARGS(ip, dp, name))
DEFINE_XREP_PPTR_SCAN_EVENT(xrep_parent_stash_parentadd);
+DEFINE_XREP_PPTR_SCAN_EVENT(xrep_parent_stash_parentremove);
TRACE_EVENT(xrep_nlinks_set_record,
TP_PROTO(struct xfs_mount *mp, xfs_ino_t ino,
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 17/22] xfs: remove pointless unlocked assertion
2023-12-31 19:33 ` [PATCHSET v13.0 5/7] xfs: fsck for parent pointers Darrick J. Wong
` (15 preceding siblings ...)
2023-12-31 20:57 ` [PATCH 16/22] xfs: implement live updates for parent pointer repairs Darrick J. Wong
@ 2023-12-31 20:57 ` Darrick J. Wong
2023-12-31 20:57 ` [PATCH 18/22] xfs: split xfs_bmap_add_attrfork into two pieces Darrick J. Wong
` (4 subsequent siblings)
21 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:57 UTC (permalink / raw)
To: djwong; +Cc: catherine.hoang, allison.henderson, linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
Remove this assertion about the inode not having an attr fork from
xfs_bmap_add_attrfork because the function handles that case just fine.
Weirder still, the function actually /requires/ the caller not to hold
the ILOCK, which means that its accesses are not stabilized.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/libxfs/xfs_bmap.c | 2 --
1 file changed, 2 deletions(-)
diff --git a/fs/xfs/libxfs/xfs_bmap.c b/fs/xfs/libxfs/xfs_bmap.c
index 44b8c315c5978..1dd1876a7b145 100644
--- a/fs/xfs/libxfs/xfs_bmap.c
+++ b/fs/xfs/libxfs/xfs_bmap.c
@@ -1023,8 +1023,6 @@ xfs_bmap_add_attrfork(
int logflags; /* logging flags */
int error; /* error return value */
- ASSERT(xfs_inode_has_attr_fork(ip) == 0);
-
mp = ip->i_mount;
ASSERT(!XFS_NOT_DQATTACHED(mp, ip));
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 18/22] xfs: split xfs_bmap_add_attrfork into two pieces
2023-12-31 19:33 ` [PATCHSET v13.0 5/7] xfs: fsck for parent pointers Darrick J. Wong
` (16 preceding siblings ...)
2023-12-31 20:57 ` [PATCH 17/22] xfs: remove pointless unlocked assertion Darrick J. Wong
@ 2023-12-31 20:57 ` Darrick J. Wong
2023-12-31 20:58 ` [PATCH 19/22] xfs: actually rebuild the parent pointer xattrs Darrick J. Wong
` (3 subsequent siblings)
21 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:57 UTC (permalink / raw)
To: djwong; +Cc: catherine.hoang, allison.henderson, linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
Split this function into two pieces -- one to make the actual changes to
the inode core to add the attr fork, and another one to deal with
getting the transaction and locking the inodes.
The next couple of patches will need this to be split into two. One
patch implements committing new parent pointer recordsets to damaged
files. If one file has an attr fork and the other does not, we have to
create the missing attr fork before the atomic swap transaction, and can
use the behavior encoded in the current xfs_bmap_add_attrfork.
The second patch adapts /lost+found adoptions to handle parent pointers
correctly. The adoption process will add a parent pointer to a child
that is being moved to /lost+found, but this requires that the attr fork
already exists. We don't know if we're actually going to commit the
adoption until we've already reserved a transaction and taken the
ILOCKs, which means that we must have a way to bypass the start of the
current xfs_bmap_add_attrfork.
Therefore, create xfs_attr_add_fork as the helper that creates a
transaction and takes locks; and make xfs_bmap_add_attrfork the function
that updates the inode core and allocates the incore attr fork.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/libxfs/xfs_attr.c | 39 ++++++++++++++++++++++++++++++++++++++-
fs/xfs/libxfs/xfs_bmap.c | 36 ++++++++++--------------------------
fs/xfs/libxfs/xfs_bmap.h | 3 ++-
3 files changed, 50 insertions(+), 28 deletions(-)
diff --git a/fs/xfs/libxfs/xfs_attr.c b/fs/xfs/libxfs/xfs_attr.c
index c13eb7b7b5b8f..ef32c2a22c617 100644
--- a/fs/xfs/libxfs/xfs_attr.c
+++ b/fs/xfs/libxfs/xfs_attr.c
@@ -935,6 +935,43 @@ xfs_attr_defer_add(
trace_xfs_attr_defer_add(new->xattri_dela_state, args->dp);
}
+STATIC int
+xfs_attr_add_fork(
+ struct xfs_inode *ip, /* incore inode pointer */
+ int size, /* space new attribute needs */
+ int rsvd) /* xact may use reserved blks */
+{
+ struct xfs_mount *mp = ip->i_mount;
+ struct xfs_trans *tp; /* transaction pointer */
+ unsigned int blks; /* space reservation */
+ int error; /* error return value */
+
+ ASSERT(!XFS_NOT_DQATTACHED(mp, ip));
+
+ blks = XFS_ADDAFORK_SPACE_RES(mp);
+
+ error = xfs_trans_alloc_inode(ip, &M_RES(mp)->tr_addafork, blks, 0,
+ rsvd, &tp);
+ if (error)
+ return error;
+
+ if (xfs_inode_has_attr_fork(ip))
+ goto trans_cancel;
+
+ error = xfs_bmap_add_attrfork(tp, ip, size, rsvd);
+ if (error)
+ goto trans_cancel;
+
+ error = xfs_trans_commit(tp);
+ xfs_iunlock(ip, XFS_ILOCK_EXCL);
+ return error;
+
+trans_cancel:
+ xfs_trans_cancel(tp);
+ xfs_iunlock(ip, XFS_ILOCK_EXCL);
+ return error;
+}
+
/*
* Note: If args->value is NULL the attribute will be removed, just like the
* Linux ->setattr API.
@@ -986,7 +1023,7 @@ xfs_attr_set(
xfs_attr_sf_entsize_byname(args->namelen,
args->valuelen);
- error = xfs_bmap_add_attrfork(dp, sf_size, rsvd);
+ error = xfs_attr_add_fork(dp, sf_size, rsvd);
if (error)
return error;
}
diff --git a/fs/xfs/libxfs/xfs_bmap.c b/fs/xfs/libxfs/xfs_bmap.c
index 1dd1876a7b145..d34354d2cdd49 100644
--- a/fs/xfs/libxfs/xfs_bmap.c
+++ b/fs/xfs/libxfs/xfs_bmap.c
@@ -1007,38 +1007,29 @@ xfs_bmap_set_attrforkoff(
}
/*
- * Convert inode from non-attributed to attributed.
- * Must not be in a transaction, ip must not be locked.
+ * Convert inode from non-attributed to attributed. Caller must hold the
+ * ILOCK_EXCL and the file cannot have an attr fork.
*/
int /* error code */
xfs_bmap_add_attrfork(
- xfs_inode_t *ip, /* incore inode pointer */
+ struct xfs_trans *tp,
+ struct xfs_inode *ip, /* incore inode pointer */
int size, /* space new attribute needs */
int rsvd) /* xact may use reserved blks */
{
- xfs_mount_t *mp; /* mount structure */
- xfs_trans_t *tp; /* transaction pointer */
- int blks; /* space reservation */
+ struct xfs_mount *mp = tp->t_mountp;
int version = 1; /* superblock attr version */
int logflags; /* logging flags */
int error; /* error return value */
- mp = ip->i_mount;
+ ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL));
ASSERT(!XFS_NOT_DQATTACHED(mp, ip));
-
- blks = XFS_ADDAFORK_SPACE_RES(mp);
-
- error = xfs_trans_alloc_inode(ip, &M_RES(mp)->tr_addafork, blks, 0,
- rsvd, &tp);
- if (error)
- return error;
- if (xfs_inode_has_attr_fork(ip))
- goto trans_cancel;
+ ASSERT(!xfs_inode_has_attr_fork(ip));
xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
error = xfs_bmap_set_attrforkoff(ip, size, &version);
if (error)
- goto trans_cancel;
+ return error;
xfs_ifork_init_attr(ip, XFS_DINODE_FMT_EXTENTS, 0);
logflags = 0;
@@ -1059,7 +1050,7 @@ xfs_bmap_add_attrfork(
if (logflags)
xfs_trans_log_inode(tp, ip, logflags);
if (error)
- goto trans_cancel;
+ return error;
if (!xfs_has_attr(mp) ||
(!xfs_has_attr2(mp) && version == 2)) {
bool log_sb = false;
@@ -1078,14 +1069,7 @@ xfs_bmap_add_attrfork(
xfs_log_sb(tp);
}
- error = xfs_trans_commit(tp);
- xfs_iunlock(ip, XFS_ILOCK_EXCL);
- return error;
-
-trans_cancel:
- xfs_trans_cancel(tp);
- xfs_iunlock(ip, XFS_ILOCK_EXCL);
- return error;
+ return 0;
}
/*
diff --git a/fs/xfs/libxfs/xfs_bmap.h b/fs/xfs/libxfs/xfs_bmap.h
index 87633449c379a..c9e297dba88d0 100644
--- a/fs/xfs/libxfs/xfs_bmap.h
+++ b/fs/xfs/libxfs/xfs_bmap.h
@@ -174,7 +174,8 @@ int xfs_bmap_longest_free_extent(struct xfs_perag *pag,
void xfs_trim_extent(struct xfs_bmbt_irec *irec, xfs_fileoff_t bno,
xfs_filblks_t len);
unsigned int xfs_bmap_compute_attr_offset(struct xfs_mount *mp);
-int xfs_bmap_add_attrfork(struct xfs_inode *ip, int size, int rsvd);
+int xfs_bmap_add_attrfork(struct xfs_trans *tp, struct xfs_inode *ip,
+ int size, int rsvd);
void xfs_bmap_local_to_extents_empty(struct xfs_trans *tp,
struct xfs_inode *ip, int whichfork);
int xfs_bmap_local_to_extents(struct xfs_trans *tp, struct xfs_inode *ip,
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 19/22] xfs: actually rebuild the parent pointer xattrs
2023-12-31 19:33 ` [PATCHSET v13.0 5/7] xfs: fsck for parent pointers Darrick J. Wong
` (17 preceding siblings ...)
2023-12-31 20:57 ` [PATCH 18/22] xfs: split xfs_bmap_add_attrfork into two pieces Darrick J. Wong
@ 2023-12-31 20:58 ` Darrick J. Wong
2023-12-31 20:58 ` [PATCH 20/22] xfs: adapt the orphanage code to handle parent pointers Darrick J. Wong
` (2 subsequent siblings)
21 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:58 UTC (permalink / raw)
To: djwong; +Cc: catherine.hoang, allison.henderson, linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
Once we've assembled all the parent pointers for a file, we need to
commit the new dataset atomically to that file. Parent pointer records
are embedded in the xattr structure, which means that we must write a
new extended attribute structure, again, atomically. Therefore, we must
copy the non-parent-pointer attributes from the file being repaired into
the temporary file's extended attributes and then call the atomic extent
swap mechanism to exchange the blocks.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/libxfs/xfs_attr.c | 2
fs/xfs/libxfs/xfs_attr.h | 2
fs/xfs/scrub/attr.c | 2
fs/xfs/scrub/attr_repair.c | 4
fs/xfs/scrub/attr_repair.h | 4
fs/xfs/scrub/findparent.c | 2
fs/xfs/scrub/listxattr.c | 12 +
fs/xfs/scrub/listxattr.h | 4
fs/xfs/scrub/parent.c | 2
fs/xfs/scrub/parent_repair.c | 696 +++++++++++++++++++++++++++++++++++++++++-
fs/xfs/scrub/trace.h | 2
11 files changed, 707 insertions(+), 25 deletions(-)
diff --git a/fs/xfs/libxfs/xfs_attr.c b/fs/xfs/libxfs/xfs_attr.c
index ef32c2a22c617..9cefaeca8f854 100644
--- a/fs/xfs/libxfs/xfs_attr.c
+++ b/fs/xfs/libxfs/xfs_attr.c
@@ -935,7 +935,7 @@ xfs_attr_defer_add(
trace_xfs_attr_defer_add(new->xattri_dela_state, args->dp);
}
-STATIC int
+int
xfs_attr_add_fork(
struct xfs_inode *ip, /* incore inode pointer */
int size, /* space new attribute needs */
diff --git a/fs/xfs/libxfs/xfs_attr.h b/fs/xfs/libxfs/xfs_attr.h
index 0204f62298cb5..a2cfe9e35fd43 100644
--- a/fs/xfs/libxfs/xfs_attr.h
+++ b/fs/xfs/libxfs/xfs_attr.h
@@ -650,4 +650,6 @@ xfs_attri_can_use_without_log_assistance(
return false;
}
+int xfs_attr_add_fork(struct xfs_inode *ip, int size, int rsvd);
+
#endif /* __XFS_ATTR_H__ */
diff --git a/fs/xfs/scrub/attr.c b/fs/xfs/scrub/attr.c
index f213d745746fd..555e1b65c78aa 100644
--- a/fs/xfs/scrub/attr.c
+++ b/fs/xfs/scrub/attr.c
@@ -647,7 +647,7 @@ xchk_xattr(
* iteration, which doesn't really follow the usual buffer
* locking order.
*/
- error = xchk_xattr_walk(sc, sc->ip, xchk_xattr_actor, NULL);
+ error = xchk_xattr_walk(sc, sc->ip, xchk_xattr_actor, NULL, NULL);
if (!xchk_fblock_process_error(sc, XFS_ATTR_FORK, 0, &error))
return error;
diff --git a/fs/xfs/scrub/attr_repair.c b/fs/xfs/scrub/attr_repair.c
index cc964dc427e23..0c369a243a635 100644
--- a/fs/xfs/scrub/attr_repair.c
+++ b/fs/xfs/scrub/attr_repair.c
@@ -1036,7 +1036,7 @@ xrep_xattr_reset_fork(
* fork. The caller must ILOCK the tempfile and join it to the transaction.
* This function returns with the inode joined to a clean scrub transaction.
*/
-STATIC int
+int
xrep_xattr_reset_tempfile_fork(
struct xfs_scrub *sc)
{
@@ -1356,7 +1356,7 @@ xrep_xattr_swap_prep(
}
/* Swap the temporary file's attribute fork with the one being repaired. */
-STATIC int
+int
xrep_xattr_swap(
struct xfs_scrub *sc,
struct xrep_tempswap *tx)
diff --git a/fs/xfs/scrub/attr_repair.h b/fs/xfs/scrub/attr_repair.h
index 0a9ffa7cfa906..ea73b8d498a24 100644
--- a/fs/xfs/scrub/attr_repair.h
+++ b/fs/xfs/scrub/attr_repair.h
@@ -6,6 +6,10 @@
#ifndef __XFS_SCRUB_ATTR_REPAIR_H__
#define __XFS_SCRUB_ATTR_REPAIR_H__
+struct xrep_tempswap;
+
+int xrep_xattr_swap(struct xfs_scrub *sc, struct xrep_tempswap *tx);
int xrep_xattr_reset_fork(struct xfs_scrub *sc);
+int xrep_xattr_reset_tempfile_fork(struct xfs_scrub *sc);
#endif /* __XFS_SCRUB_ATTR_REPAIR_H__ */
diff --git a/fs/xfs/scrub/findparent.c b/fs/xfs/scrub/findparent.c
index 9468029f73933..ceb76b26c6cd1 100644
--- a/fs/xfs/scrub/findparent.c
+++ b/fs/xfs/scrub/findparent.c
@@ -24,6 +24,7 @@
#include "xfs_trans_space.h"
#include "xfs_health.h"
#include "xfs_swapext.h"
+#include "xfs_parent.h"
#include "scrub/xfs_scrub.h"
#include "scrub/scrub.h"
#include "scrub/common.h"
@@ -33,6 +34,7 @@
#include "scrub/findparent.h"
#include "scrub/readdir.h"
#include "scrub/tempfile.h"
+#include "scrub/listxattr.h"
/*
* Finding the Parent of a Directory
diff --git a/fs/xfs/scrub/listxattr.c b/fs/xfs/scrub/listxattr.c
index dc893f2cdc1c3..44f9660ae544f 100644
--- a/fs/xfs/scrub/listxattr.c
+++ b/fs/xfs/scrub/listxattr.c
@@ -267,6 +267,7 @@ xchk_xattr_walk_node(
struct xfs_scrub *sc,
struct xfs_inode *ip,
xchk_xattr_fn attr_fn,
+ xchk_xattrleaf_fn leaf_fn,
void *priv)
{
struct xfs_attr3_icleaf_hdr leafhdr;
@@ -298,6 +299,12 @@ xchk_xattr_walk_node(
xfs_trans_brelse(sc->tp, leaf_bp);
+ if (leaf_fn) {
+ error = leaf_fn(sc, priv);
+ if (error)
+ goto out_bitmap;
+ }
+
/* Make sure we haven't seen this new leaf already. */
len = 1;
if (xdab_bitmap_test(&seen_dablks, leafhdr.forw, &len))
@@ -332,6 +339,7 @@ xchk_xattr_walk(
struct xfs_scrub *sc,
struct xfs_inode *ip,
xchk_xattr_fn attr_fn,
+ xchk_xattrleaf_fn leaf_fn,
void *priv)
{
int error;
@@ -352,7 +360,7 @@ xchk_xattr_walk(
if (xfs_attr_is_leaf(ip))
return xchk_xattr_walk_leaf(sc, ip, attr_fn, priv);
- return xchk_xattr_walk_node(sc, ip, attr_fn, priv);
+ return xchk_xattr_walk_node(sc, ip, attr_fn, leaf_fn, priv);
}
/*
@@ -380,5 +388,5 @@ xchk_pptr_walk(
ASSERT(xfs_has_parent(sc->mp));
- return xchk_xattr_walk(sc, ip, xchk_pptr_walk_attr, &pw);
+ return xchk_xattr_walk(sc, ip, xchk_pptr_walk_attr, NULL, &pw);
}
diff --git a/fs/xfs/scrub/listxattr.h b/fs/xfs/scrub/listxattr.h
index 7e4bd3ae75e15..6bd41443c5f44 100644
--- a/fs/xfs/scrub/listxattr.h
+++ b/fs/xfs/scrub/listxattr.h
@@ -11,8 +11,10 @@ typedef int (*xchk_xattr_fn)(struct xfs_scrub *sc, struct xfs_inode *ip,
unsigned int namelen, const void *value, unsigned int valuelen,
void *priv);
+typedef int (*xchk_xattrleaf_fn)(struct xfs_scrub *sc, void *priv);
+
int xchk_xattr_walk(struct xfs_scrub *sc, struct xfs_inode *ip,
- xchk_xattr_fn attr_fn, void *priv);
+ xchk_xattr_fn attr_fn, xchk_xattrleaf_fn leaf_fn, void *priv);
struct xfs_parent_name_irec;
diff --git a/fs/xfs/scrub/parent.c b/fs/xfs/scrub/parent.c
index 555aee4b73b37..6af212540c631 100644
--- a/fs/xfs/scrub/parent.c
+++ b/fs/xfs/scrub/parent.c
@@ -768,7 +768,7 @@ xchk_parent_pptr(
if (error)
goto out_entries;
- error = xchk_xattr_walk(sc, sc->ip, xchk_parent_scan_attr, pp);
+ error = xchk_xattr_walk(sc, sc->ip, xchk_parent_scan_attr, NULL, pp);
if (error == -ECANCELED) {
error = 0;
goto out_names;
diff --git a/fs/xfs/scrub/parent_repair.c b/fs/xfs/scrub/parent_repair.c
index b94eebec3cc79..00e6735717a37 100644
--- a/fs/xfs/scrub/parent_repair.c
+++ b/fs/xfs/scrub/parent_repair.c
@@ -25,6 +25,8 @@
#include "xfs_health.h"
#include "xfs_swapext.h"
#include "xfs_parent.h"
+#include "xfs_attr.h"
+#include "xfs_bmap.h"
#include "scrub/xfs_scrub.h"
#include "scrub/scrub.h"
#include "scrub/common.h"
@@ -34,10 +36,13 @@
#include "scrub/findparent.h"
#include "scrub/readdir.h"
#include "scrub/tempfile.h"
+#include "scrub/tempswap.h"
#include "scrub/orphanage.h"
#include "scrub/xfile.h"
#include "scrub/xfarray.h"
#include "scrub/xfblob.h"
+#include "scrub/attr_repair.h"
+#include "scrub/listxattr.h"
/*
* Repairing The Directory Parent Pointer
@@ -107,6 +112,23 @@ struct xrep_parent {
/* Blobs containing parent pointer names. */
struct xfblob *pptr_names;
+ /* xattr keys */
+ struct xfarray *xattr_records;
+
+ /* xattr values */
+ struct xfblob *xattr_blobs;
+
+ /* Scratch buffers for saving extended attributes */
+ unsigned char *xattr_name;
+ void *xattr_value;
+ unsigned int xattr_value_sz;
+
+ /*
+ * Information used to swap the attr fork, if the fs supports parent
+ * pointers.
+ */
+ struct xrep_tempswap tx;
+
/*
* Information used to scan the filesystem to find the inumber of the
* dotdot entry for this directory. On filesystems without parent
@@ -118,12 +140,17 @@ struct xrep_parent {
* @pscan.lock coordinates access to pptr_recs, pptr_names, pptr, and
* pptr_scratch. This reduces the memory requirements of this
* structure.
+ *
+ * The lock also controls access to xattr_records and xattr_blobs(?)
*/
struct xrep_parent_scan_info pscan;
/* Orphanage reparenting request. */
struct xrep_adoption adoption;
+ /* Have we seen any live updates of parent pointers recently? */
+ bool saw_pptr_updates;
+
/* xattr key and da args for parent pointer replay. */
struct xfs_parent_scratch pptr_scratch;
@@ -135,12 +162,43 @@ struct xrep_parent {
struct xfs_parent_name_irec pptr;
};
+struct xrep_parent_xattr {
+ /* Cookie for retrieval of the xattr name. */
+ xfblob_cookie name_cookie;
+
+ /* Cookie for retrieval of the xattr value. */
+ xfblob_cookie value_cookie;
+
+ /* XFS_ATTR_* flags */
+ int flags;
+
+ /* Length of the value and name. */
+ uint32_t valuelen;
+ uint16_t namelen;
+};
+
+/*
+ * Stash up to 8 pages of attrs in xattr_records/xattr_blobs before we write
+ * them to the temp file.
+ */
+#define XREP_PARENT_XATTR_MAX_STASH_BYTES (PAGE_SIZE * 8)
+
/* Tear down all the incore stuff we created. */
static void
xrep_parent_teardown(
struct xrep_parent *rp)
{
xrep_findparent_scan_teardown(&rp->pscan);
+ kvfree(rp->xattr_name);
+ rp->xattr_name = NULL;
+ kvfree(rp->xattr_value);
+ rp->xattr_value = NULL;
+ if (rp->xattr_blobs)
+ xfblob_destroy(rp->xattr_blobs);
+ rp->xattr_blobs = NULL;
+ if (rp->xattr_records)
+ xfarray_destroy(rp->xattr_records);
+ rp->xattr_records = NULL;
if (rp->pptr_names)
xfblob_destroy(rp->pptr_names);
rp->pptr_names = NULL;
@@ -574,10 +632,11 @@ xrep_parent_scan_dirtree(
}
/*
- * Cancel the empty transaction so that we can (later) use the atomic
- * extent swap helpers to lock files and commit the new directory.
+ * Retake sc->ip's ILOCK now that we're done flushing stashed parent
+ * pointers. We end this function with an empty transaction and the
+ * ILOCK.
*/
- xchk_trans_cancel(rp->sc);
+ xchk_ilock(rp->sc, XFS_ILOCK_EXCL);
return 0;
}
@@ -612,6 +671,8 @@ xrep_parent_live_update(
else
error = xrep_parent_stash_parentremove(rp, p->name,
p->dp);
+ if (!error)
+ rp->saw_pptr_updates = true;
mutex_unlock(&rp->pscan.lock);
if (error)
goto out_abort;
@@ -666,6 +727,45 @@ xrep_parent_reset_dotdot(
return xfs_trans_roll(&sc->tp);
}
+/* Pass back the parent inumber if this a parent pointer */
+STATIC int
+xrep_parent_lookup_pptr(
+ struct xfs_scrub *sc,
+ struct xfs_inode *ip,
+ const struct xfs_parent_name_irec *pptr,
+ void *priv)
+{
+ xfs_ino_t *inop = priv;
+
+ if (!xfs_parent_verify_irec(sc->mp, pptr))
+ return -EFSCORRUPTED;
+
+ *inop = pptr->p_ino;
+ return -ECANCELED;
+}
+
+/*
+ * Find the first parent of the scrub target by walking parent pointers for
+ * the purpose of deciding if we're going to move it to the orphanage.
+ * We don't care if the attr fork is zapped.
+ */
+STATIC int
+xrep_parent_lookup_pptrs(
+ struct xfs_scrub *sc,
+ xfs_ino_t *inop)
+{
+ struct xfs_parent_name_irec pptr;
+ int error;
+
+ *inop = NULLFSINO;
+
+ error = xchk_pptr_walk(sc, sc->ip, xrep_parent_lookup_pptr, &pptr,
+ inop);
+ if (error && error != -ECANCELED)
+ return error;
+ return 0;
+}
+
/*
* Move the current file to the orphanage.
*
@@ -682,14 +782,25 @@ xrep_parent_move_to_orphanage(
xfs_ino_t orig_parent, new_parent;
int error;
- /*
- * We are about to drop the ILOCK on sc->ip to lock the orphanage and
- * prepare for the adoption. Therefore, look up the old dotdot entry
- * for sc->ip so that we can compare it after we re-lock sc->ip.
- */
- error = xchk_dir_lookup(sc, sc->ip, &xfs_name_dotdot, &orig_parent);
- if (error)
- return error;
+ if (S_ISDIR(VFS_I(sc->ip)->i_mode)) {
+ /*
+ * We are about to drop the ILOCK on sc->ip to lock the
+ * orphanage and prepare for the adoption. Therefore, look up
+ * the old dotdot entry for sc->ip so that we can compare it
+ * after we re-lock sc->ip.
+ */
+ error = xchk_dir_lookup(sc, sc->ip, &xfs_name_dotdot,
+ &orig_parent);
+ if (error)
+ return error;
+ } else {
+ /*
+ * We haven't dropped the ILOCK since we swapped in the new
+ * parent pointers, which means that the file cannot have been
+ * moved in the directory tree, and there are no parents.
+ */
+ orig_parent = NULLFSINO;
+ }
/*
* Drop the ILOCK on the scrub target and commit the transaction.
@@ -722,9 +833,14 @@ xrep_parent_move_to_orphanage(
* Now that we've reacquired the ILOCK on sc->ip, look up the dotdot
* entry again. If the parent changed or the child was unlinked while
* the child directory was unlocked, we don't need to move the child to
- * the orphanage after all.
+ * the orphanage after all. For a non-directory, we have to scan for
+ * the first parent pointer to see if one has been added.
*/
- error = xchk_dir_lookup(sc, sc->ip, &xfs_name_dotdot, &new_parent);
+ if (S_ISDIR(VFS_I(sc->ip)->i_mode))
+ error = xchk_dir_lookup(sc, sc->ip, &xfs_name_dotdot,
+ &new_parent);
+ else
+ error = xrep_parent_lookup_pptrs(sc, &new_parent);
if (error)
return error;
@@ -751,6 +867,492 @@ xrep_parent_move_to_orphanage(
return 0;
}
+/* Ensure that the xattr value buffer is large enough. */
+STATIC int
+xrep_parent_alloc_xattr_value(
+ struct xrep_parent *rp,
+ size_t bufsize)
+{
+ void *new_val;
+
+ if (rp->xattr_value_sz >= bufsize)
+ return 0;
+
+ if (rp->xattr_value) {
+ kvfree(rp->xattr_value);
+ rp->xattr_value = NULL;
+ rp->xattr_value_sz = 0;
+ }
+
+ new_val = kvmalloc(bufsize, XCHK_GFP_FLAGS);
+ if (!new_val)
+ return -ENOMEM;
+
+ rp->xattr_value = new_val;
+ rp->xattr_value_sz = bufsize;
+ return 0;
+}
+
+/* Retrieve the (remote) value of a non-pptr xattr. */
+STATIC int
+xrep_parent_fetch_xattr_remote(
+ struct xrep_parent *rp,
+ struct xfs_inode *ip,
+ unsigned int attr_flags,
+ const unsigned char *name,
+ unsigned int namelen,
+ unsigned int valuelen)
+{
+ struct xfs_scrub *sc = rp->sc;
+ struct xfs_da_args args = {
+ .op_flags = XFS_DA_OP_NOTIME,
+ .attr_filter = attr_flags & XFS_ATTR_NSP_ONDISK_MASK,
+ .geo = sc->mp->m_attr_geo,
+ .whichfork = XFS_ATTR_FORK,
+ .dp = ip,
+ .name = name,
+ .namelen = namelen,
+ .hashval = xfs_da_hashname(name, namelen),
+ .trans = sc->tp,
+ .valuelen = valuelen,
+ .owner = ip->i_ino,
+ };
+ int error;
+
+ /*
+ * If we need a larger value buffer, try to allocate one. If that
+ * fails, return with -EDEADLOCK to try harder.
+ */
+ error = xrep_parent_alloc_xattr_value(rp, valuelen);
+ if (error == -ENOMEM)
+ return -EDEADLOCK;
+ if (error)
+ return error;
+
+ args.value = rp->xattr_value;
+ return xfs_attr_get_ilocked(&args);
+}
+
+/* Stash non-pptr attributes for later replay into the temporary file. */
+STATIC int
+xrep_parent_stash_xattr(
+ struct xfs_scrub *sc,
+ struct xfs_inode *ip,
+ unsigned int attr_flags,
+ const unsigned char *name,
+ unsigned int namelen,
+ const void *value,
+ unsigned int valuelen,
+ void *priv)
+{
+ struct xrep_parent_xattr key = {
+ .valuelen = valuelen,
+ .namelen = namelen,
+ .flags = attr_flags & XFS_ATTR_NSP_ONDISK_MASK,
+ };
+ struct xrep_parent *rp = priv;
+ int error;
+
+ if (attr_flags & (XFS_ATTR_INCOMPLETE | XFS_ATTR_PARENT))
+ return 0;
+
+ if (!value) {
+ error = xrep_parent_fetch_xattr_remote(rp, ip, attr_flags,
+ name, namelen, valuelen);
+ if (error)
+ return error;
+
+ value = rp->xattr_value;
+ }
+
+ trace_xrep_parent_stash_xattr(rp->sc->tempip, key.flags, (void *)name,
+ key.namelen, key.valuelen);
+
+ error = xfblob_store(rp->xattr_blobs, &key.name_cookie, name,
+ key.namelen);
+ if (error)
+ return error;
+
+ error = xfblob_store(rp->xattr_blobs, &key.value_cookie, value,
+ key.valuelen);
+ if (error)
+ return error;
+
+ return xfarray_append(rp->xattr_records, &key);
+}
+
+/* Insert one xattr key/value. */
+STATIC int
+xrep_parent_insert_xattr(
+ struct xrep_parent *rp,
+ const struct xrep_parent_xattr *key)
+{
+ struct xfs_da_args args = {
+ .dp = rp->sc->tempip,
+ .attr_filter = key->flags,
+ .namelen = key->namelen,
+ .valuelen = key->valuelen,
+ .op_flags = XFS_DA_OP_NOTIME,
+ .owner = rp->sc->ip->i_ino,
+ };
+ int error;
+
+ ASSERT(!(key->flags & XFS_ATTR_PARENT));
+
+ /*
+ * Grab pointers to the scrub buffer so that we can use them to insert
+ * attrs into the temp file.
+ */
+ args.name = rp->xattr_name;
+ args.value = rp->xattr_value;
+
+ /*
+ * The attribute name is stored near the end of the in-core buffer,
+ * though we reserve one more byte to ensure null termination.
+ */
+ rp->xattr_name[XATTR_NAME_MAX] = 0;
+
+ error = xfblob_load(rp->xattr_blobs, key->name_cookie, rp->xattr_name,
+ key->namelen);
+ if (error)
+ return error;
+
+ error = xfblob_free(rp->xattr_blobs, key->name_cookie);
+ if (error)
+ return error;
+
+ error = xfblob_load(rp->xattr_blobs, key->value_cookie, args.value,
+ key->valuelen);
+ if (error)
+ return error;
+
+ error = xfblob_free(rp->xattr_blobs, key->value_cookie);
+ if (error)
+ return error;
+
+ rp->xattr_name[key->namelen] = 0;
+
+ trace_xrep_parent_insert_xattr(rp->sc->tempip, key->flags,
+ rp->xattr_name, key->namelen, key->valuelen);
+
+ error = xfs_attr_set(&args);
+ if (error) {
+ ASSERT(error != -EEXIST);
+ return error;
+ }
+
+ return 0;
+}
+
+/*
+ * Periodically flush salvaged attributes to the temporary file. This is done
+ * to reduce the memory requirements of the xattr rebuild because files can
+ * contain millions of attributes.
+ */
+STATIC int
+xrep_parent_flush_xattrs(
+ struct xrep_parent *rp)
+{
+ xfarray_idx_t array_cur;
+ int error;
+
+ /*
+ * Entering this function, the scrub context has a reference to the
+ * inode being repaired, the temporary file, and the empty scrub
+ * transaction that we created for the xattr scan. We hold ILOCK_EXCL
+ * on the inode being repaired.
+ *
+ * To constrain kernel memory use, we occasionally flush salvaged
+ * xattrs from the xfarray and xfblob structures into the temporary
+ * file in preparation for swapping the xattr structures at the end.
+ * Updating the temporary file requires a transaction, so we commit the
+ * scrub transaction and drop the ILOCK so that xfs_attr_set can
+ * allocate whatever transaction it wants.
+ *
+ * We still hold IOLOCK_EXCL on the inode being repaired, which
+ * prevents anyone from adding xattrs (or parent pointers) while we're
+ * flushing.
+ */
+ xchk_trans_cancel(rp->sc);
+ xchk_iunlock(rp->sc, XFS_ILOCK_EXCL);
+
+ /*
+ * Take the IOLOCK of the temporary file while we modify xattrs. This
+ * isn't strictly required because the temporary file is never revealed
+ * to userspace, but we follow the same locking rules. We still hold
+ * sc->ip's IOLOCK.
+ */
+ error = xrep_tempfile_iolock_polled(rp->sc);
+ if (error)
+ return error;
+
+ /* Add all the salvaged attrs to the temporary file. */
+ foreach_xfarray_idx(rp->xattr_records, array_cur) {
+ struct xrep_parent_xattr key;
+
+ error = xfarray_load(rp->xattr_records, array_cur, &key);
+ if (error)
+ return error;
+
+ error = xrep_parent_insert_xattr(rp, &key);
+ if (error)
+ return error;
+ }
+
+ /* Empty out both arrays now that we've added the entries. */
+ xfarray_truncate(rp->xattr_records);
+ xfblob_truncate(rp->xattr_blobs);
+
+ xrep_tempfile_iounlock(rp->sc);
+
+ /* Recreate the empty transaction and relock the inode. */
+ error = xchk_trans_alloc_empty(rp->sc);
+ if (error)
+ return error;
+ xchk_ilock(rp->sc, XFS_ILOCK_EXCL);
+ return 0;
+}
+
+/* Decide if we've stashed too much xattr data in memory. */
+static inline bool
+xrep_parent_want_flush_xattrs(
+ struct xrep_parent *rp)
+{
+ unsigned long long bytes;
+
+ bytes = xfarray_bytes(rp->xattr_records) +
+ xfblob_bytes(rp->xattr_blobs);
+ return bytes > XREP_PARENT_XATTR_MAX_STASH_BYTES;
+}
+
+/* Flush staged attributes to the temporary file if we're over the limit. */
+STATIC int
+xrep_parent_try_flush_xattrs(
+ struct xfs_scrub *sc,
+ void *priv)
+{
+ struct xrep_parent *rp = priv;
+ int error;
+
+ if (!xrep_parent_want_flush_xattrs(rp))
+ return 0;
+
+ error = xrep_parent_flush_xattrs(rp);
+ if (error)
+ return error;
+
+ /*
+ * If there were any parent pointer updates to the xattr structure
+ * while we dropped the ILOCK, the xattr structure is now stale.
+ * Signal to the attr copy process that we need to start over, but
+ * this time without opportunistic attr flushing.
+ *
+ * This is unlikely to happen, so we're ok with restarting the copy.
+ */
+ mutex_lock(&rp->pscan.lock);
+ if (rp->saw_pptr_updates)
+ error = -ESTALE;
+ mutex_unlock(&rp->pscan.lock);
+ return error;
+}
+
+/* Copy all the non-pptr extended attributes into the temporary file. */
+STATIC int
+xrep_parent_copy_xattrs(
+ struct xrep_parent *rp)
+{
+ struct xfs_scrub *sc = rp->sc;
+ int error;
+
+ /*
+ * Clear the pptr updates flag. We hold sc->ip ILOCKed, so there
+ * can't be any parent pointer updates in progress.
+ */
+ mutex_lock(&rp->pscan.lock);
+ rp->saw_pptr_updates = false;
+ mutex_unlock(&rp->pscan.lock);
+
+ /* Copy xattrs, stopping periodically to flush the incore buffers. */
+ error = xchk_xattr_walk(sc, sc->ip, xrep_parent_stash_xattr,
+ xrep_parent_try_flush_xattrs, rp);
+ if (error && error != -ESTALE)
+ return error;
+
+ if (error == -ESTALE) {
+ /*
+ * The xattr copy collided with a parent pointer update.
+ * Restart the copy, but this time hold the ILOCK all the way
+ * to the end to lock out any directory parent pointer updates.
+ */
+ error = xchk_xattr_walk(sc, sc->ip, xrep_parent_stash_xattr,
+ NULL, rp);
+ if (error)
+ return error;
+ }
+
+ /* Flush any remaining stashed xattrs to the temporary file. */
+ if (xfarray_bytes(rp->xattr_records) == 0)
+ return 0;
+
+ return xrep_parent_flush_xattrs(rp);
+}
+
+/*
+ * Ensure that @sc->ip and @sc->tempip both have attribute forks before we
+ * head into the attr fork swap transaction. All files on a filesystem with
+ * parent pointers must have an attr fork because the parent pointer code
+ * does not itself add attribute forks.
+ *
+ * Note: Unlinkable unlinked files don't need one, but the overhead of having
+ * an unnecessary attr fork is not justified by the additional code complexity
+ * that would be needed to track that state correctly.
+ */
+STATIC int
+xrep_parent_ensure_attr_fork(
+ struct xrep_parent *rp)
+{
+ struct xfs_scrub *sc = rp->sc;
+ int error;
+
+ error = xfs_attr_add_fork(sc->tempip,
+ sizeof(struct xfs_attr_sf_hdr), 1);
+ if (error)
+ return error;
+ return xfs_attr_add_fork(sc->ip, sizeof(struct xfs_attr_sf_hdr), 1);
+}
+
+/*
+ * Finish replaying stashed parent pointer updates, allocate a transaction for
+ * swapping extents, and take the ILOCKs of both files before we commit the new
+ * attribute structure.
+ */
+STATIC int
+xrep_parent_finalize_tempfile(
+ struct xrep_parent *rp)
+{
+ struct xfs_scrub *sc = rp->sc;
+ int error;
+
+ /*
+ * Repair relies on the ILOCK to quiesce all possible xattr updates.
+ * Replay all queued parent pointer updates into the tempfile before
+ * swapping the contents, even if that means dropping the ILOCKs and
+ * the transaction.
+ */
+ do {
+ error = xrep_parent_replay_updates(rp);
+ if (error)
+ return error;
+
+ error = xrep_parent_ensure_attr_fork(rp);
+ if (error)
+ return error;
+
+ error = xrep_tempswap_trans_alloc(sc, XFS_ATTR_FORK, &rp->tx);
+ if (error)
+ return error;
+
+ if (xfarray_length(rp->pptr_recs) == 0)
+ break;
+
+ xchk_trans_cancel(sc);
+ xrep_tempfile_iunlock_both(sc);
+ } while (!xchk_should_terminate(sc, &error));
+ return error;
+}
+
+/*
+ * Replay all the stashed parent pointers into the temporary file, copy all
+ * the non-pptr xattrs from the file being repaired into the temporary file,
+ * and swap the extents atomically.
+ */
+STATIC int
+xrep_parent_rebuild_pptrs(
+ struct xrep_parent *rp)
+{
+ struct xfs_scrub *sc = rp->sc;
+ xfs_ino_t parent_ino = NULLFSINO;
+ int error;
+
+ /*
+ * Copy non-ppttr xattrs from the file being repaired into the
+ * temporary file's xattr structure. We hold sc->ip's IOLOCK, which
+ * prevents setxattr/removexattr calls from occurring, but renames
+ * update the parent pointers without holding IOLOCK. If we detect
+ * stale attr structures, we restart the scan but only flush at the
+ * end.
+ */
+ error = xrep_parent_copy_xattrs(rp);
+ if (error)
+ return error;
+
+ /*
+ * Cancel the empty transaction that we used to walk and copy attrs,
+ * and drop the ILOCK so that we can take the IOLOCK on the temporary
+ * file. We still hold sc->ip's IOLOCK.
+ */
+ xchk_trans_cancel(sc);
+ xchk_iunlock(sc, XFS_ILOCK_EXCL);
+
+ error = xrep_tempfile_iolock_polled(sc);
+ if (error)
+ return error;
+
+ /*
+ * Allocate transaction, lock inodes, and make sure that we've replayed
+ * all the stashed pptr updates to the tempdir. After this point,
+ * we're ready to swapext the attr fork.
+ */
+ error = xrep_parent_finalize_tempfile(rp);
+ if (error)
+ return error;
+
+ /* Last chance to abort before we start committing pptr fixes. */
+ if (xchk_should_terminate(sc, &error))
+ return error;
+
+ if (xchk_iscan_aborted(&rp->pscan.iscan))
+ return -ECANCELED;
+
+ /*
+ * Swap the attr fork and junk the old attr fork contents, which are
+ * now in the tempfile.
+ */
+ error = xrep_xattr_swap(sc, &rp->tx);
+ if (error)
+ return error;
+ error = xrep_xattr_reset_tempfile_fork(sc);
+ if (error)
+ return error;
+
+ /*
+ * Roll to get a transaction without any inodes joined to it. Then we
+ * can drop the tempfile's ILOCK and IOLOCK before doing more work on
+ * the scrub target file.
+ */
+ error = xfs_trans_roll(&sc->tp);
+ if (error)
+ return error;
+ xrep_tempfile_iunlock(sc);
+ xrep_tempfile_iounlock(sc);
+
+ /*
+ * We've committed the new parent pointers. Find at least one parent
+ * so that we can decide if we're moving this file to the orphanage.
+ * For this purpose, root directories are their own parents.
+ */
+ if (sc->ip == sc->mp->m_rootip) {
+ xrep_findparent_scan_found(&rp->pscan, sc->ip->i_ino);
+ } else {
+ error = xrep_parent_lookup_pptrs(sc, &parent_ino);
+ if (error)
+ return error;
+ if (parent_ino != NULLFSINO)
+ xrep_findparent_scan_found(&rp->pscan, parent_ino);
+ }
+ return 0;
+}
+
/*
* Commit the new parent pointer structure (currently only the dotdot entry) to
* the file that we're repairing.
@@ -759,13 +1361,24 @@ STATIC int
xrep_parent_rebuild_tree(
struct xrep_parent *rp)
{
+ int error;
+
+ if (xfs_has_parent(rp->sc->mp)) {
+ error = xrep_parent_rebuild_pptrs(rp);
+ if (error)
+ return error;
+ }
+
if (rp->pscan.parent_ino == NULLFSINO) {
if (xrep_orphanage_can_adopt(rp->sc))
return xrep_parent_move_to_orphanage(rp);
return -EFSCORRUPTED;
}
- return xrep_parent_reset_dotdot(rp);
+ if (S_ISDIR(VFS_I(rp->sc->ip)->i_mode))
+ return xrep_parent_reset_dotdot(rp);
+
+ return 0;
}
/* Set up the filesystem scan so we can look for parents. */
@@ -775,18 +1388,39 @@ xrep_parent_setup_scan(
{
struct xfs_scrub *sc = rp->sc;
char *descr;
+ struct xfs_da_geometry *geo = sc->mp->m_attr_geo;
+ int max_len;
int error;
if (!xfs_has_parent(sc->mp))
return xrep_findparent_scan_start(sc, &rp->pscan);
+ /* Buffers for copying non-pptr attrs to the tempfile */
+ rp->xattr_name = kvmalloc(XATTR_NAME_MAX + 1, XCHK_GFP_FLAGS);
+ if (!rp->xattr_name)
+ return -ENOMEM;
+
+ /*
+ * Allocate enough memory to handle loading local attr values from the
+ * xfblob data while flushing stashed attrs to the temporary file.
+ * We only realloc the buffer when salvaging remote attr values, so
+ * TRY_HARDER means we allocate the maximal attr value size.
+ */
+ if (sc->flags & XCHK_TRY_HARDER)
+ max_len = XATTR_SIZE_MAX;
+ else
+ max_len = xfs_attr_leaf_entsize_local_max(geo->blksize);
+ error = xrep_parent_alloc_xattr_value(rp, max_len);
+ if (error)
+ goto out_xattr_name;
+
/* Set up some staging memory for logging parent pointer updates. */
descr = xchk_xfile_ino_descr(sc, "parent pointer entries");
error = xfarray_create(descr, 0, sizeof(struct xrep_pptr),
&rp->pptr_recs);
kfree(descr);
if (error)
- return error;
+ goto out_xattr_value;
descr = xchk_xfile_ino_descr(sc, "parent pointer names");
error = xfblob_create(descr, &rp->pptr_names);
@@ -794,19 +1428,47 @@ xrep_parent_setup_scan(
if (error)
goto out_recs;
+ /* Set up some storage for copying attrs before the swap */
+ descr = xchk_xfile_ino_descr(sc,
+ "parent pointer retained xattr entries");
+ error = xfarray_create(descr, 0, sizeof(struct xrep_parent_xattr),
+ &rp->xattr_records);
+ kfree(descr);
+ if (error)
+ goto out_names;
+
+ descr = xchk_xfile_ino_descr(sc,
+ "parent pointer retained xattr values");
+ error = xfblob_create(descr, &rp->xattr_blobs);
+ kfree(descr);
+ if (error)
+ goto out_attr_keys;
+
error = __xrep_findparent_scan_start(sc, &rp->pscan,
xrep_parent_live_update);
if (error)
- goto out_names;
+ goto out_attr_values;
return 0;
+out_attr_values:
+ xfblob_destroy(rp->xattr_blobs);
+ rp->xattr_blobs = NULL;
+out_attr_keys:
+ xfarray_destroy(rp->xattr_records);
+ rp->xattr_records = NULL;
out_names:
xfblob_destroy(rp->pptr_names);
rp->pptr_names = NULL;
out_recs:
xfarray_destroy(rp->pptr_recs);
rp->pptr_recs = NULL;
+out_xattr_value:
+ kvfree(rp->xattr_value);
+ rp->xattr_value = NULL;
+out_xattr_name:
+ kvfree(rp->xattr_name);
+ rp->xattr_name = NULL;
return error;
}
@@ -836,7 +1498,7 @@ xrep_parent(
if (error)
goto out_teardown;
- /* Last chance to abort before we start committing fixes. */
+ /* Last chance to abort before we start committing dotdot fixes. */
if (xchk_should_terminate(sc, &error))
goto out_teardown;
diff --git a/fs/xfs/scrub/trace.h b/fs/xfs/scrub/trace.h
index b03e1e69740fa..10e2d6544c5ad 100644
--- a/fs/xfs/scrub/trace.h
+++ b/fs/xfs/scrub/trace.h
@@ -2661,6 +2661,8 @@ DEFINE_EVENT(xrep_xattr_salvage_class, name, \
TP_ARGS(ip, flags, name, namelen, valuelen))
DEFINE_XREP_XATTR_SALVAGE_EVENT(xrep_xattr_salvage_rec);
DEFINE_XREP_XATTR_SALVAGE_EVENT(xrep_xattr_insert_rec);
+DEFINE_XREP_XATTR_SALVAGE_EVENT(xrep_parent_stash_xattr);
+DEFINE_XREP_XATTR_SALVAGE_EVENT(xrep_parent_insert_xattr);
DECLARE_EVENT_CLASS(xrep_pptr_salvage_class,
TP_PROTO(struct xfs_inode *ip, unsigned int flags, const void *name,
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 20/22] xfs: adapt the orphanage code to handle parent pointers
2023-12-31 19:33 ` [PATCHSET v13.0 5/7] xfs: fsck for parent pointers Darrick J. Wong
` (18 preceding siblings ...)
2023-12-31 20:58 ` [PATCH 19/22] xfs: actually rebuild the parent pointer xattrs Darrick J. Wong
@ 2023-12-31 20:58 ` Darrick J. Wong
2023-12-31 20:58 ` [PATCH 21/22] xfs: repair link count of nondirectories after rebuilding " Darrick J. Wong
2023-12-31 20:58 ` [PATCH 22/22] xfs: inode repair should ensure there's an attr fork to store " Darrick J. Wong
21 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:58 UTC (permalink / raw)
To: djwong; +Cc: catherine.hoang, allison.henderson, linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
Adapt the orphanage's adoption code to update the child file's parent
pointers as part of the reparenting process. Also ensure that the child
has an attr fork to receive the parent pointer update, since the runtime
code assumes one exists.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/scrub/orphanage.c | 42 ++++++++++++++++++++++++++++++++++++++++++
fs/xfs/scrub/orphanage.h | 3 +++
fs/xfs/scrub/scrub.c | 2 ++
3 files changed, 47 insertions(+)
diff --git a/fs/xfs/scrub/orphanage.c b/fs/xfs/scrub/orphanage.c
index ace7a0f23e474..b894b807155a7 100644
--- a/fs/xfs/scrub/orphanage.c
+++ b/fs/xfs/scrub/orphanage.c
@@ -19,6 +19,8 @@
#include "xfs_icache.h"
#include "xfs_bmap.h"
#include "xfs_bmap_btree.h"
+#include "xfs_parent.h"
+#include "xfs_attr_sf.h"
#include "scrub/scrub.h"
#include "scrub/common.h"
#include "scrub/repair.h"
@@ -330,8 +332,12 @@ xrep_adoption_trans_alloc(
if (S_ISDIR(VFS_I(sc->ip)->i_mode))
child_blkres = xfs_rename_space_res(mp, 0, false,
xfs_name_dotdot.len, false);
+ if (xfs_has_parent(mp))
+ child_blkres += XFS_ADDAFORK_SPACE_RES(mp);
adopt->child_blkres = child_blkres;
+ xfs_parent_args_init(mp, &adopt->ppargs);
+
/*
* Allocate a transaction to link the child into the parent, along with
* enough disk space to handle expansion of both the orphanage and the
@@ -500,6 +506,21 @@ xrep_adoption_zap_dcache(
dput(d_orphanage);
}
+/*
+ * If we have to add an attr fork ahead of a parent pointer update, how much
+ * space should we ask for?
+ */
+static inline int
+xrep_adoption_attr_sizeof(
+ const struct xrep_adoption *adopt)
+{
+ size_t res = sizeof(struct xfs_attr_sf_hdr);
+
+ res += xfs_attr_sf_entsize_byname(sizeof(struct xfs_parent_name_rec),
+ adopt->xname.len);
+ return res;
+}
+
/*
* Move the current file to the orphanage under the computed name.
*
@@ -522,6 +543,19 @@ xrep_adoption_move(
if (error)
return error;
+ /*
+ * If this filesystem has parent pointers, ensure that the file being
+ * moved to the orphanage has an attribute fork. This is required
+ * because the parent pointer code does not itself add attr forks.
+ */
+ if (!xfs_inode_has_attr_fork(sc->ip) && xfs_has_parent(sc->mp)) {
+ int sf_size = xrep_adoption_attr_sizeof(adopt);
+
+ error = xfs_bmap_add_attrfork(sc->tp, sc->ip, sf_size, true);
+ if (error)
+ return error;
+ }
+
/* Create the new name in the orphanage. */
error = xfs_dir_createname(sc->tp, sc->orphanage, xname, sc->ip->i_ino,
adopt->orphanage_blkres);
@@ -546,6 +580,14 @@ xrep_adoption_move(
return error;
}
+ /* Add a parent pointer from the file back to the lost+found. */
+ if (xfs_has_parent(sc->mp)) {
+ error = xfs_parent_addname(sc->tp, &adopt->ppargs,
+ sc->orphanage, xname, sc->ip);
+ if (error)
+ return error;
+ }
+
/*
* Notify dirent hooks that we moved the file to /lost+found, and
* finish all the deferred work so that we know the adoption is fully
diff --git a/fs/xfs/scrub/orphanage.h b/fs/xfs/scrub/orphanage.h
index 9d40992583b24..74ce0bc05c6f1 100644
--- a/fs/xfs/scrub/orphanage.h
+++ b/fs/xfs/scrub/orphanage.h
@@ -54,6 +54,9 @@ struct xrep_adoption {
struct xfs_scrub *sc;
+ /* Parent pointer context tracking */
+ struct xfs_parent_args ppargs;
+
/* Block reservations for orphanage and child (if directory). */
unsigned int orphanage_blkres;
unsigned int child_blkres;
diff --git a/fs/xfs/scrub/scrub.c b/fs/xfs/scrub/scrub.c
index 440b8cb1957f4..d9c6d54ffad7f 100644
--- a/fs/xfs/scrub/scrub.c
+++ b/fs/xfs/scrub/scrub.c
@@ -19,6 +19,8 @@
#include "xfs_rmap.h"
#include "xfs_xchgrange.h"
#include "xfs_swapext.h"
+#include "xfs_dir2.h"
+#include "xfs_parent.h"
#include "scrub/scrub.h"
#include "scrub/common.h"
#include "scrub/trace.h"
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 21/22] xfs: repair link count of nondirectories after rebuilding parent pointers
2023-12-31 19:33 ` [PATCHSET v13.0 5/7] xfs: fsck for parent pointers Darrick J. Wong
` (19 preceding siblings ...)
2023-12-31 20:58 ` [PATCH 20/22] xfs: adapt the orphanage code to handle parent pointers Darrick J. Wong
@ 2023-12-31 20:58 ` Darrick J. Wong
2023-12-31 20:58 ` [PATCH 22/22] xfs: inode repair should ensure there's an attr fork to store " Darrick J. Wong
21 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:58 UTC (permalink / raw)
To: djwong; +Cc: catherine.hoang, allison.henderson, linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
Since the parent pointer scrubber does not exhaustively search the
filesystem for missing parent pointers, it doesn't have a good way to
determine that there are pointers missing from an otherwise uncorrupt
xattr structure. Instead, for nondirectories it employs a heuristic of
comparing the file link count to the number of parent pointers found.
However, we don't want this heuristic flagging a false corruption after
a repair has actually scanned the entire filesystem to rebuild the
parent pointers. Therefore, reset the file link count in this one case
because we actually know the correct link count.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/scrub/parent_repair.c | 97 ++++++++++++++++++++++++++++++++++++++++++
1 file changed, 97 insertions(+)
diff --git a/fs/xfs/scrub/parent_repair.c b/fs/xfs/scrub/parent_repair.c
index 00e6735717a37..04ea0b05ed088 100644
--- a/fs/xfs/scrub/parent_repair.c
+++ b/fs/xfs/scrub/parent_repair.c
@@ -27,6 +27,7 @@
#include "xfs_parent.h"
#include "xfs_attr.h"
#include "xfs_bmap.h"
+#include "xfs_ag.h"
#include "scrub/xfs_scrub.h"
#include "scrub/scrub.h"
#include "scrub/common.h"
@@ -160,6 +161,9 @@ struct xrep_parent {
* lost+found filename if we need to reparent the file.
*/
struct xfs_parent_name_irec pptr;
+
+ /* Number of parents we found after all other repairs */
+ unsigned long long parents;
};
struct xrep_parent_xattr {
@@ -1381,6 +1385,92 @@ xrep_parent_rebuild_tree(
return 0;
}
+/* Count the number of parent pointers. */
+STATIC int
+xrep_parent_count_pptr(
+ struct xfs_scrub *sc,
+ struct xfs_inode *ip,
+ const struct xfs_parent_name_irec *pptr,
+ void *priv)
+{
+ struct xrep_parent *rp = priv;
+
+ if (!xfs_parent_verify_irec(sc->mp, pptr))
+ return -EFSCORRUPTED;
+
+ rp->parents++;
+ return 0;
+}
+
+/*
+ * After all parent pointer rebuilding and adoption activity completes, reset
+ * the link count of this nondirectory, having scanned the fs to rebuild all
+ * parent pointers.
+ */
+STATIC int
+xrep_parent_set_nondir_nlink(
+ struct xrep_parent *rp)
+{
+ struct xfs_scrub *sc = rp->sc;
+ struct xfs_inode *ip = sc->ip;
+ struct xfs_perag *pag;
+ bool joined = false;
+ int error;
+
+ /* Count parent pointers so we can reset the file link count. */
+ rp->parents = 0;
+ error = xchk_pptr_walk(sc, ip, xrep_parent_count_pptr, &rp->pptr, rp);
+ if (error)
+ return error;
+
+ if (rp->parents > 0 && xfs_inode_on_unlinked_list(ip)) {
+ xfs_trans_ijoin(sc->tp, sc->ip, 0);
+ joined = true;
+
+ /*
+ * The file is on the unlinked list but we found parents.
+ * Remove the file from the unlinked list.
+ */
+ pag = xfs_perag_get(sc->mp, XFS_INO_TO_AGNO(sc->mp, ip->i_ino));
+ if (!pag) {
+ ASSERT(0);
+ return -EFSCORRUPTED;
+ }
+
+ error = xfs_iunlink_remove(sc->tp, pag, ip);
+ xfs_perag_put(pag);
+ if (error)
+ return error;
+ } else if (rp->parents == 0 && !xfs_inode_on_unlinked_list(ip)) {
+ xfs_trans_ijoin(sc->tp, sc->ip, 0);
+ joined = true;
+
+ /*
+ * The file is not on the unlinked list but we found no
+ * parents. Add the file to the unlinked list.
+ */
+ error = xfs_iunlink(sc->tp, ip);
+ if (error)
+ return error;
+ }
+
+ /* Set the correct link count. */
+ if (VFS_I(ip)->i_nlink != rp->parents) {
+ if (!joined) {
+ xfs_trans_ijoin(sc->tp, sc->ip, 0);
+ joined = true;
+ }
+
+ set_nlink(VFS_I(ip), min_t(unsigned long long, rp->parents,
+ XFS_NLINK_PINNED));
+ }
+
+ /* Log the inode to keep it moving forward if we dirtied anything. */
+ if (joined)
+ xfs_trans_log_inode(sc->tp, ip, XFS_ILOG_CORE);
+ return 0;
+}
+
/* Set up the filesystem scan so we can look for parents. */
STATIC int
xrep_parent_setup_scan(
@@ -1505,6 +1595,13 @@ xrep_parent(
error = xrep_parent_rebuild_tree(rp);
if (error)
goto out_teardown;
+ if (xfs_has_parent(sc->mp) && !S_ISDIR(VFS_I(sc->ip)->i_mode)) {
+ error = xrep_parent_set_nondir_nlink(rp);
+ if (error)
+ goto out_teardown;
+ }
+
+ error = xrep_defer_finish(sc);
out_teardown:
xrep_parent_teardown(rp);
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 22/22] xfs: inode repair should ensure there's an attr fork to store parent pointers
2023-12-31 19:33 ` [PATCHSET v13.0 5/7] xfs: fsck for parent pointers Darrick J. Wong
` (20 preceding siblings ...)
2023-12-31 20:58 ` [PATCH 21/22] xfs: repair link count of nondirectories after rebuilding " Darrick J. Wong
@ 2023-12-31 20:58 ` Darrick J. Wong
21 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:58 UTC (permalink / raw)
To: djwong; +Cc: catherine.hoang, allison.henderson, linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
The runtime parent pointer update code expects that any file being moved
around the directory tree already has an attr fork. However, if we had
to rebuild an inode core record, there's a chance that we zeroed forkoff
as part of the inode to pass the iget verifiers.
Therefore, if we performed any repairs on an inode core, ensure that the
inode has a nonzero forkoff before unlocking the inode.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/scrub/inode_repair.c | 41 +++++++++++++++++++++++++++++++++++++++++
1 file changed, 41 insertions(+)
diff --git a/fs/xfs/scrub/inode_repair.c b/fs/xfs/scrub/inode_repair.c
index 5867617c00cd8..187f782d51f22 100644
--- a/fs/xfs/scrub/inode_repair.c
+++ b/fs/xfs/scrub/inode_repair.c
@@ -1688,6 +1688,44 @@ xrep_inode_extsize(
}
}
+/* Ensure this file has an attr fork if it needs to hold a parent pointer. */
+STATIC int
+xrep_inode_pptr(
+ struct xfs_scrub *sc)
+{
+ struct xfs_mount *mp = sc->mp;
+ struct xfs_inode *ip = sc->ip;
+ struct inode *inode = VFS_I(ip);
+
+ if (!xfs_has_parent(mp))
+ return 0;
+
+ /*
+ * Unlinked inodes that cannot be added to the directory tree will not
+ * have a parent pointer.
+ */
+ if (inode->i_nlink == 0 && !(inode->i_state & I_LINKABLE))
+ return 0;
+
+ /* The root directory doesn't have a parent pointer. */
+ if (ip == mp->m_rootip)
+ return 0;
+
+ /*
+ * Metadata inodes are rooted in the superblock and do not have any
+ * parents.
+ */
+ if (xfs_is_metadata_inode(ip))
+ return 0;
+
+ /* Inode already has an attr fork; no further work possible here. */
+ if (xfs_inode_has_attr_fork(ip))
+ return 0;
+
+ return xfs_bmap_add_attrfork(sc->tp, ip,
+ sizeof(struct xfs_attr_sf_hdr), true);
+}
+
/* Fix any irregularities in an inode that the verifiers don't catch. */
STATIC int
xrep_inode_problems(
@@ -1696,6 +1734,9 @@ xrep_inode_problems(
int error;
error = xrep_inode_blockcounts(sc);
+ if (error)
+ return error;
+ error = xrep_inode_pptr(sc);
if (error)
return error;
xrep_inode_timestamps(sc->ip);
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 1/4] xfs: teach online scrub to find directory tree structure problems
2023-12-31 19:34 ` [PATCHSET v13.0 6/7] xfs: detect and correct directory tree problems Darrick J. Wong
@ 2023-12-31 20:59 ` Darrick J. Wong
2023-12-31 20:59 ` [PATCH 2/4] xfs: invalidate dirloop scrub path data when concurrent updates happen Darrick J. Wong
` (2 subsequent siblings)
3 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:59 UTC (permalink / raw)
To: djwong; +Cc: linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
Create a new scrubber that detects corruptions within the directory tree
structure itself. It can detect directories with multiple parents;
loops within the directory tree; and directory loops not accessible from
the root.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/Makefile | 1
fs/xfs/libxfs/xfs_fs.h | 3
fs/xfs/scrub/common.h | 1
fs/xfs/scrub/dirtree.c | 749 +++++++++++++++++++++++++++++++++++++++++++++
fs/xfs/scrub/dirtree.h | 123 +++++++
fs/xfs/scrub/ino_bitmap.h | 37 ++
fs/xfs/scrub/scrub.c | 7
fs/xfs/scrub/scrub.h | 1
fs/xfs/scrub/stats.c | 1
fs/xfs/scrub/trace.c | 4
fs/xfs/scrub/trace.h | 188 +++++++++++
fs/xfs/scrub/xfarray.h | 1
12 files changed, 1114 insertions(+), 2 deletions(-)
create mode 100644 fs/xfs/scrub/dirtree.c
create mode 100644 fs/xfs/scrub/dirtree.h
create mode 100644 fs/xfs/scrub/ino_bitmap.h
diff --git a/fs/xfs/Makefile b/fs/xfs/Makefile
index 52ef808359966..31cbc4fd88f99 100644
--- a/fs/xfs/Makefile
+++ b/fs/xfs/Makefile
@@ -163,6 +163,7 @@ xfs-y += $(addprefix scrub/, \
common.o \
dabtree.o \
dir.o \
+ dirtree.o \
fscounters.o \
health.o \
ialloc.o \
diff --git a/fs/xfs/libxfs/xfs_fs.h b/fs/xfs/libxfs/xfs_fs.h
index efa68a2d82a1d..48f38694f1232 100644
--- a/fs/xfs/libxfs/xfs_fs.h
+++ b/fs/xfs/libxfs/xfs_fs.h
@@ -719,9 +719,10 @@ struct xfs_scrub_metadata {
#define XFS_SCRUB_TYPE_QUOTACHECK 25 /* quota counters */
#define XFS_SCRUB_TYPE_NLINKS 26 /* inode link counts */
#define XFS_SCRUB_TYPE_HEALTHY 27 /* everything checked out ok */
+#define XFS_SCRUB_TYPE_DIRTREE 28 /* directory tree structure */
/* Number of scrub subcommands. */
-#define XFS_SCRUB_TYPE_NR 28
+#define XFS_SCRUB_TYPE_NR 29
/* i: Repair this metadata. */
#define XFS_SCRUB_IFLAG_REPAIR (1u << 0)
diff --git a/fs/xfs/scrub/common.h b/fs/xfs/scrub/common.h
index 298669ca2eb92..c3a580a5d7c7d 100644
--- a/fs/xfs/scrub/common.h
+++ b/fs/xfs/scrub/common.h
@@ -92,6 +92,7 @@ int xchk_setup_directory(struct xfs_scrub *sc);
int xchk_setup_xattr(struct xfs_scrub *sc);
int xchk_setup_symlink(struct xfs_scrub *sc);
int xchk_setup_parent(struct xfs_scrub *sc);
+int xchk_setup_dirtree(struct xfs_scrub *sc);
#ifdef CONFIG_XFS_RT
int xchk_setup_rtbitmap(struct xfs_scrub *sc);
int xchk_setup_rtsummary(struct xfs_scrub *sc);
diff --git a/fs/xfs/scrub/dirtree.c b/fs/xfs/scrub/dirtree.c
new file mode 100644
index 0000000000000..9edaf89f46fdf
--- /dev/null
+++ b/fs/xfs/scrub/dirtree.c
@@ -0,0 +1,749 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * Copyright (c) 2023-2024 Oracle. All Rights Reserved.
+ * Author: Darrick J. Wong <djwong@kernel.org>
+ */
+#include "xfs.h"
+#include "xfs_fs.h"
+#include "xfs_shared.h"
+#include "xfs_format.h"
+#include "xfs_trans_resv.h"
+#include "xfs_mount.h"
+#include "xfs_log_format.h"
+#include "xfs_trans.h"
+#include "xfs_inode.h"
+#include "xfs_icache.h"
+#include "xfs_dir2.h"
+#include "xfs_dir2_priv.h"
+#include "xfs_attr.h"
+#include "xfs_parent.h"
+#include "scrub/scrub.h"
+#include "scrub/common.h"
+#include "scrub/bitmap.h"
+#include "scrub/ino_bitmap.h"
+#include "scrub/xfile.h"
+#include "scrub/xfarray.h"
+#include "scrub/xfblob.h"
+#include "scrub/listxattr.h"
+#include "scrub/trace.h"
+#include "scrub/dirtree.h"
+
+/*
+ * Directory Tree Structure Validation
+ * ===================================
+ *
+ * Validating the tree qualities of the directory tree structure can be
+ * difficult. If the tree is frozen, running a depth (or breadth) first search
+ * and marking a bitmap suffices to determine if there is a cycle. XORing the
+ * mark bitmap with the inode bitmap afterwards tells us if there are
+ * disconnected cycles. If the tree is not frozen, directory updates can move
+ * subtrees across the scanner wavefront, which complicates the design greatly.
+ *
+ * Directory parent pointers change that by enabling an incremental approach to
+ * validation of the tree structure. Instead of using one thread to scan the
+ * entire filesystem, we instead can have multiple threads walking individual
+ * subdirectories upwards to the root. Each scanner thread must be able to
+ * take the IOLOCK of the directory that it is examining to prevent that
+ * directory from being moved within the tree. This was not possible prior to
+ * Linux 6.5 because the VFS did not take i_rwsem when moving subdirectories.
+ *
+ * If the walk terminates without reaching the root, we know the path is
+ * disconnected and ought to be attached to the lost and found. If on the walk
+ * we find the same subdir that we're scanning, we know this is a cycle and
+ * should delete an incoming edge. If we find multiple paths to the root, we
+ * know to delete an incoming edge.
+ *
+ * There are two big hitches with this approach: first, all file link counts
+ * must be correct to prevent other writers from doing the wrong thing with the
+ * directory tree structure. Second, because we're walking upwards in a tree
+ * of arbitrary depth, we cannot hold all the ILOCKs. Instead, we will use a
+ * directory update hook to invalidate the scan results if one of the paths
+ * we've scanned has changed.
+ */
+
+/* Clean up the dirtree checking resources. */
+STATIC void
+xchk_dirtree_buf_cleanup(
+ void *buf)
+{
+ struct xchk_dirtree *dl = buf;
+ struct xchk_dirpath *path, *n;
+
+ xchk_dirtree_for_each_path_safe(dl, path, n) {
+ list_del_init(&path->list);
+ xino_bitmap_destroy(&path->seen_inodes);
+ kfree(path);
+ }
+
+ xfblob_destroy(dl->path_names);
+ xfarray_destroy(dl->path_steps);
+ mutex_destroy(&dl->lock);
+}
+
+/* Set us up to look for directory loops. */
+int
+xchk_setup_dirtree(
+ struct xfs_scrub *sc)
+{
+ struct xchk_dirtree *dl;
+ char *descr;
+ int error;
+
+ dl = kvzalloc(sizeof(struct xchk_dirtree), XCHK_GFP_FLAGS);
+ if (!dl)
+ return -ENOMEM;
+ dl->sc = sc;
+ INIT_LIST_HEAD(&dl->path_list);
+ dl->root_ino = NULLFSINO;
+
+ mutex_init(&dl->lock);
+
+ descr = xchk_xfile_ino_descr(sc, "dirtree path steps");
+ error = xfarray_create(descr, 0, sizeof(struct xchk_dirpath_step),
+ &dl->path_steps);
+ kfree(descr);
+ if (error)
+ goto out_dl;
+
+ descr = xchk_xfile_ino_descr(sc, "dirtree path names");
+ error = xfblob_create(descr, &dl->path_names);
+ kfree(descr);
+ if (error)
+ goto out_steps;
+
+ error = xchk_setup_inode_contents(sc, 0);
+ if (error)
+ goto out_names;
+
+ sc->buf = dl;
+ sc->buf_cleanup = xchk_dirtree_buf_cleanup;
+ return 0;
+
+out_names:
+ xfblob_destroy(dl->path_names);
+out_steps:
+ xfarray_destroy(dl->path_steps);
+out_dl:
+ mutex_destroy(&dl->lock);
+ kvfree(dl);
+ return error;
+}
+
+/*
+ * Add the parent pointer described by @dl->pptr to the given path as a new
+ * step. Returns -ELNRNG if the path is too deep.
+ */
+STATIC int
+xchk_dirpath_append(
+ struct xchk_dirtree *dl,
+ struct xfs_inode *ip,
+ struct xchk_dirpath *path,
+ const struct xfs_parent_name_irec *pptr)
+{
+ struct xchk_dirpath_step step = {
+ .parent_ino = pptr->p_ino,
+ .parent_gen = pptr->p_gen,
+ .name_len = pptr->p_namelen,
+ };
+ int error;
+
+ /*
+ * If this path is more than 2 billion steps long, this directory tree
+ * is too far gone to fix.
+ */
+ if (path->nr_steps >= XFS_MAXLINK)
+ return -ELNRNG;
+
+ error = xfblob_store(dl->path_names, &step.name_cookie,
+ dl->pptr.p_name, dl->pptr.p_namelen);
+ if (error)
+ return error;
+
+ error = xino_bitmap_set(&path->seen_inodes, ip->i_ino);
+ if (error)
+ return error;
+
+ error = xfarray_append(dl->path_steps, &step);
+ if (error)
+ return error;
+
+ path->nr_steps++;
+ return 0;
+}
+
+/*
+ * Create an xchk_path for each parent pointer of the directory that we're
+ * scanning. For each path created, we will eventually try to walk towards the
+ * root with the goal of deleting all parents except for one that leads to the
+ * root.
+ *
+ * Returns -EFSCORRUPTED to signal that the inode being scanned has a corrupt
+ * parent pointer and hence there's no point in continuing; or -ENOSR if there
+ * are too many parent pointers for this directory.
+ */
+STATIC int
+xchk_dirtree_create_path(
+ struct xfs_scrub *sc,
+ struct xfs_inode *ip,
+ const struct xfs_parent_name_irec *pptr,
+ void *priv)
+{
+ struct xchk_dirtree *dl = priv;
+ struct xchk_dirpath *path;
+ int error;
+
+ if (!xfs_parent_verify_irec(sc->mp, pptr))
+ return -EFSCORRUPTED;
+
+ /*
+ * If there are more than 2 billion actual parent pointers for this
+ * subdirectory, this fs is too far gone to fix.
+ */
+ if (dl->nr_paths >= XFS_MAXLINK)
+ return -ENOSR;
+
+ trace_xchk_dirtree_create_path(sc, ip, dl->nr_paths, pptr);
+
+ /*
+ * Create a new xchk_path structure to remember this parent pointer
+ * and record the first name step.
+ */
+ path = kmalloc(sizeof(struct xchk_dirpath), XCHK_GFP_FLAGS);
+ if (!path)
+ return -ENOMEM;
+
+ INIT_LIST_HEAD(&path->list);
+ xino_bitmap_init(&path->seen_inodes);
+ path->nr_steps = 0;
+ path->outcome = XCHK_DIRPATH_SCANNING;
+
+ error = xchk_dirpath_append(dl, sc->ip, path, pptr);
+ if (error)
+ goto out_path;
+
+ path->first_step = xfarray_length(dl->path_steps) - 1;
+ path->second_step = XFARRAY_NULLIDX;
+ path->path_nr = dl->nr_paths;
+
+ list_add_tail(&path->list, &dl->path_list);
+ dl->nr_paths++;
+ return 0;
+out_path:
+ kfree(path);
+ return error;
+}
+
+/*
+ * Validate that the first step of this path still has a corresponding
+ * parent pointer in @sc->ip. We probably dropped @sc->ip's ILOCK while
+ * walking towards the roots, which is why this is necessary.
+ *
+ * This function has a side effect of loading the first parent pointer of this
+ * path into the parent pointer scratch pad. This prepares us to walk up the
+ * directory tree towards the root. Returns -ESTALE if the scan data is now
+ * out of date.
+ */
+STATIC int
+xchk_dirpath_revalidate(
+ struct xchk_dirtree *dl,
+ struct xchk_dirpath *path)
+{
+ struct xchk_dirpath_step step;
+ struct xfs_scrub *sc = dl->sc;
+ int error;
+
+ error = xfarray_load(dl->path_steps, path->first_step, &step);
+ if (error)
+ return error;
+
+ /*
+ * Check that this parent pointer is still attached to the inode that
+ * we're scanning/
+ */
+ dl->pptr.p_ino = step.parent_ino;
+ dl->pptr.p_gen = step.parent_gen;
+ dl->pptr.p_namelen = step.name_len;
+
+ error = xfblob_load(dl->path_names, step.name_cookie, dl->pptr.p_name,
+ step.name_len);
+ if (error)
+ return error;
+ xfs_parent_irec_hashname(sc->mp, &dl->pptr);
+
+ /*
+ * Look up the parent pointer that corresponds to the start of this
+ * path. If the parent pointer has disappeared on us, dump all the
+ * scan results and try again.
+ */
+ error = xfs_parent_lookup(sc->tp, sc->ip, &dl->pptr, &dl->scratch);
+ if (error == -ENOATTR) {
+ trace_xchk_dirpath_disappeared(dl->sc, sc->ip, path->path_nr,
+ path->first_step, &dl->pptr);
+ dl->stale = true;
+ return -ESTALE;
+ }
+
+ return error;
+}
+
+/*
+ * Walk the parent pointers of a directory at the end of a path and record
+ * the parent that we find.
+ */
+STATIC int
+xchk_dirpath_find_next_step(
+ struct xfs_scrub *sc,
+ struct xfs_inode *ip,
+ const struct xfs_parent_name_irec *pptr,
+ void *priv)
+{
+ unsigned int *parents_found = priv;
+
+ if (!xfs_parent_verify_irec(sc->mp, pptr))
+ return -EFSCORRUPTED;
+
+ /*
+ * If we've already set @dl->pptr.p_ino, then this directory has
+ * multiple parents. Signal this back to the caller via -EMLINK.
+ */
+ if (*parents_found > 0)
+ return -EMLINK;
+
+ (*parents_found)++;
+ return 0;
+}
+
+/* Set and log the outcome of a path walk. */
+static inline void
+xchk_dirpath_set_outcome(
+ struct xchk_dirtree *dl,
+ struct xchk_dirpath *path,
+ enum xchk_dirpath_outcome outcome)
+{
+ trace_xchk_dirpath_set_outcome(dl->sc, path->path_nr, path->nr_steps,
+ outcome);
+
+ path->outcome = outcome;
+}
+
+/*
+ * Scan the directory at the end of this path for its parent directory link.
+ * If we find one, extend the path. Returns -ESTALE if the scan data out of
+ * date. Returns -EFSCORRUPTED if the parent pointer is bad; or -ELNRNG if
+ * the path got too deep.
+ */
+STATIC int
+xchk_dirpath_step_up(
+ struct xchk_dirtree *dl,
+ struct xchk_dirpath *path)
+{
+ struct xfs_scrub *sc = dl->sc;
+ struct xfs_inode *dp;
+ unsigned int lock_mode;
+ unsigned int parents_found = 0;
+ int error;
+
+ /* Grab and lock the parent directory. */
+ error = xchk_iget(sc, dl->pptr.p_ino, &dp);
+ if (error)
+ return error;
+
+ lock_mode = xfs_ilock_attr_map_shared(dp);
+ mutex_lock(&dl->lock);
+
+ if (dl->stale) {
+ error = -ESTALE;
+ goto out_scanlock;
+ }
+
+ /* We've reached the root directory; the path is ok. */
+ if (dl->pptr.p_ino == dl->root_ino) {
+ xchk_dirpath_set_outcome(dl, path, XCHK_DIRPATH_OK);
+ error = 0;
+ goto out_scanlock;
+ }
+
+ /*
+ * The inode being scanned is its own distant ancestor! Get rid of
+ * this path.
+ */
+ if (dl->pptr.p_ino == sc->ip->i_ino) {
+ xchk_dirpath_set_outcome(dl, path, XCHK_DIRPATH_DELETE);
+ error = 0;
+ goto out_scanlock;
+ }
+
+ /*
+ * We've seen this inode before during the path walk. There's a loop
+ * above us in the directory tree. This probably means that we cannot
+ * continue, but let's keep walking paths to get a full picture.
+ */
+ if (xino_bitmap_test(&path->seen_inodes, dl->pptr.p_ino)) {
+ xchk_dirpath_set_outcome(dl, path, XCHK_DIRPATH_LOOP);
+ error = 0;
+ goto out_scanlock;
+ }
+
+ /* The handle encoded in the parent pointer must match. */
+ if (VFS_I(dp)->i_generation != dl->pptr.p_gen) {
+ trace_xchk_dirpath_badgen(dl->sc, dp, path->path_nr,
+ path->nr_steps, &dl->pptr);
+ error = -EFSCORRUPTED;
+ goto out_scanlock;
+ }
+
+ /* Parent pointer must point up to a directory. */
+ if (!S_ISDIR(VFS_I(dp)->i_mode)) {
+ trace_xchk_dirpath_nondir_parent(dl->sc, dp, path->path_nr,
+ path->nr_steps, &dl->pptr);
+ error = -EFSCORRUPTED;
+ goto out_scanlock;
+ }
+
+ /* Parent cannot be an unlinked directory. */
+ if (VFS_I(dp)->i_nlink == 0) {
+ trace_xchk_dirpath_unlinked_parent(dl->sc, dp, path->path_nr,
+ path->nr_steps, &dl->pptr);
+ error = -EFSCORRUPTED;
+ goto out_scanlock;
+ }
+
+ /*
+ * If the extended attributes look as though they has been zapped by
+ * the inode record repair code, we cannot scan for parent pointers.
+ */
+ if (xchk_pptr_looks_zapped(dp)) {
+ error = -EBUSY;
+ xchk_set_incomplete(sc);
+ goto out_scanlock;
+ }
+
+ /*
+ * Walk the parent pointers of @dp to find the parent of this directory
+ * to find the next step in our walk. If we find that @dp has exactly
+ * one parent, the parent pointer information will be in @dl->pptr.
+ */
+ mutex_unlock(&dl->lock);
+ error = xchk_pptr_walk(sc, dp, xchk_dirpath_find_next_step, &dl->pptr,
+ &parents_found);
+ mutex_lock(&dl->lock);
+ if (error == -EFSCORRUPTED || error == -EMLINK ||
+ (!error && parents_found == 0)) {
+ /*
+ * Further up the directory tree from @sc->ip, we found a
+ * corrupt parent pointer, multiple parent pointers while
+ * finding this directory's parent, or zero parents despite
+ * having a nonzero link count. Keep looking for other paths.
+ */
+ xchk_dirpath_set_outcome(dl, path, XCHK_DIRPATH_CORRUPT);
+ error = 0;
+ goto out_scanlock;
+ }
+ if (error)
+ goto out_scanlock;
+
+ if (dl->stale) {
+ error = -ESTALE;
+ goto out_scanlock;
+ }
+
+ trace_xchk_dirpath_found_next_step(sc, dp, path->path_nr,
+ path->nr_steps, &dl->pptr);
+
+ /* Append to the path steps */
+ error = xchk_dirpath_append(dl, dp, path, &dl->pptr);
+ if (error)
+ goto out_scanlock;
+
+ if (path->second_step == XFARRAY_NULLIDX)
+ path->second_step = xfarray_length(dl->path_steps) - 1;
+
+out_scanlock:
+ mutex_unlock(&dl->lock);
+ xfs_iunlock(dp, lock_mode);
+ xchk_irele(sc, dp);
+ return error;
+}
+
+/*
+ * Walk the directory tree upwards towards what is hopefully the root
+ * directory, recording path steps as we go. Returns -ESTALE if the scan data
+ * are out of date. Returns -EFSCORRUPTED only if the direct parent pointer of
+ * @sc->ip associated with this path is corrupt.
+ */
+STATIC int
+xchk_dirpath_walk_upwards(
+ struct xchk_dirtree *dl,
+ struct xchk_dirpath *path)
+{
+ struct xfs_scrub *sc = dl->sc;
+ int error;
+
+ ASSERT(sc->ilock_flags & XFS_ILOCK_EXCL);
+
+ /* Reload the start of this path and make sure it's still there. */
+ error = xchk_dirpath_revalidate(dl, path);
+ if (error)
+ return error;
+
+ trace_xchk_dirpath_walk_upwards(sc, sc->ip, path->path_nr, &dl->pptr);
+
+ /*
+ * The inode being scanned is its own direct ancestor!
+ * Get rid of this path.
+ */
+ if (dl->pptr.p_ino == sc->ip->i_ino) {
+ xchk_dirpath_set_outcome(dl, path, XCHK_DIRPATH_DELETE);
+ return 0;
+ }
+
+ /*
+ * Drop ILOCK_EXCL on the inode being scanned. We still hold
+ * IOLOCK_EXCL on it, so it cannot move around or be renamed.
+ *
+ * Beyond this point we're walking up the directory tree, which means
+ * that we can acquire and drop the ILOCK on an alias of sc->ip. The
+ * ILOCK state is no longer tracked in the scrub context. Hence we
+ * must drop @sc->ip's ILOCK during the walk.
+ */
+ mutex_unlock(&dl->lock);
+ xchk_iunlock(sc, XFS_ILOCK_EXCL);
+
+ /*
+ * Take the first step in the walk towards the root by checking the
+ * start of this path, which is a direct parent pointer of @sc->ip.
+ * If we see any kind of error here (including corruptions), the parent
+ * pointer of @sc->ip is corrupt. Stop the whole scan.
+ */
+ error = xchk_dirpath_step_up(dl, path);
+ if (error) {
+ xchk_ilock(sc, XFS_ILOCK_EXCL);
+ mutex_lock(&dl->lock);
+ return error;
+ }
+
+ /*
+ * Take steps upward from the second step in this path towards the
+ * root. If we hit corruption errors here, there's a problem
+ * *somewhere* in the path, but we don't need to stop scanning.
+ */
+ while (!error && path->outcome == XCHK_DIRPATH_SCANNING)
+ error = xchk_dirpath_step_up(dl, path);
+
+ /* Retake the locks we had, mark paths, etc. */
+ xchk_ilock(sc, XFS_ILOCK_EXCL);
+ mutex_lock(&dl->lock);
+ if (error == -EFSCORRUPTED) {
+ xchk_dirpath_set_outcome(dl, path, XCHK_DIRPATH_CORRUPT);
+ error = 0;
+ }
+ if (!error && dl->stale)
+ return -ESTALE;
+ return error;
+}
+
+/* Delete all the collected path information. */
+STATIC void
+xchk_dirtree_reset(
+ void *buf)
+{
+ struct xchk_dirtree *dl = buf;
+ struct xchk_dirpath *path, *n;
+
+ ASSERT(dl->sc->ilock_flags & XFS_ILOCK_EXCL);
+
+ xchk_dirtree_for_each_path_safe(dl, path, n) {
+ list_del_init(&path->list);
+ xino_bitmap_destroy(&path->seen_inodes);
+ kfree(path);
+ }
+ dl->nr_paths = 0;
+
+ xfarray_truncate(dl->path_steps);
+ xfblob_truncate(dl->path_names);
+
+ dl->stale = false;
+}
+
+/*
+ * For each parent pointer of this subdir, trace a path upwards towards the
+ * root directory and record what we find. Returns 0 for success;
+ * -EFSCORRUPTED if walking the parent pointers of @sc->ip failed, -ELNRNG if a
+ * path was too deep; -ENOSR if there were too many parent pointers; or
+ * a negative errno.
+ */
+STATIC int
+xchk_dirtree_find_paths_to_root(
+ struct xchk_dirtree *dl)
+{
+ struct xfs_scrub *sc = dl->sc;
+ struct xchk_dirpath *path;
+ int error = 0;
+
+ do {
+ if (xchk_should_terminate(sc, &error))
+ return error;
+
+ xchk_dirtree_reset(dl);
+
+ /*
+ * If the extended attributes look as though they has been
+ * zapped by the inode record repair code, we cannot scan for
+ * parent pointers.
+ */
+ if (xchk_pptr_looks_zapped(sc->ip)) {
+ xchk_set_incomplete(sc);
+ return -EBUSY;
+ }
+
+ /*
+ * Create path walk contexts for each parent of the directory
+ * that is being scanned. Directories are supposed to have
+ * only one parent, but this is how we detect multiple parents.
+ */
+ error = xchk_pptr_walk(sc, sc->ip, xchk_dirtree_create_path,
+ &dl->pptr, dl);
+ if (error)
+ return error;
+
+ xchk_dirtree_for_each_path(dl, path) {
+ /*
+ * Try to walk up each path to the root. This enables
+ * us to find directory loops in ancestors, and the
+ * like.
+ */
+ error = xchk_dirpath_walk_upwards(dl, path);
+ if (error == -EFSCORRUPTED) {
+ /*
+ * A parent pointer of @sc->ip is bad, don't
+ * bother continuing.
+ */
+ break;
+ }
+ if (error == -ESTALE) {
+ /* This had better be an invalidation. */
+ ASSERT(dl->stale);
+ break;
+ }
+ if (error)
+ return error;
+ }
+ } while (dl->stale);
+
+ return error;
+}
+
+/*
+ * Figure out what to do with the paths we tried to find. Do not call this
+ * if the scan results are stale.
+ */
+STATIC void
+xchk_dirtree_evaluate(
+ struct xchk_dirtree *dl,
+ struct xchk_dirtree_outcomes *oc)
+{
+ struct xchk_dirpath *path;
+
+ ASSERT(!dl->stale);
+
+ /* Scan the paths we have to decide what to do. */
+ memset(oc, 0, sizeof(struct xchk_dirtree_outcomes));
+ xchk_dirtree_for_each_path(dl, path) {
+ trace_xchk_dirpath_evaluate_path(dl->sc, path->path_nr,
+ path->nr_steps, path->outcome);
+
+ switch (path->outcome) {
+ case XCHK_DIRPATH_SCANNING:
+ /* shouldn't get here */
+ ASSERT(0);
+ break;
+ case XCHK_DIRPATH_DELETE:
+ /* This one is already going away. */
+ oc->bad++;
+ break;
+ case XCHK_DIRPATH_CORRUPT:
+ case XCHK_DIRPATH_LOOP:
+ /* Couldn't find the end of this path. */
+ oc->suspect++;
+ break;
+ case XCHK_DIRPATH_STALE:
+ /* shouldn't get here either */
+ ASSERT(0);
+ break;
+ case XCHK_DIRPATH_OK:
+ /* This path got all the way to the root. */
+ oc->good++;
+ break;
+ }
+ }
+
+ trace_xchk_dirtree_evaluate(dl, oc);
+}
+
+/* Look for directory loops. */
+int
+xchk_dirtree(
+ struct xfs_scrub *sc)
+{
+ struct xchk_dirtree_outcomes oc;
+ struct xchk_dirtree *dl = sc->buf;
+ int error;
+
+ /*
+ * Nondirectories do not point downwards to other files, so they cannot
+ * cause a cycle in the directory tree.
+ */
+ if (!S_ISDIR(VFS_I(sc->ip)->i_mode))
+ return -ENOENT;
+
+ ASSERT(xfs_has_parent(sc->mp));
+
+ /* Find the root of the directory tree. */
+ dl->root_ino = sc->mp->m_rootip->i_ino;
+
+ trace_xchk_dirtree_start(sc->ip, sc->sm, 0);
+
+ mutex_lock(&dl->lock);
+
+ /* Trace each parent pointer's path to the root. */
+ error = xchk_dirtree_find_paths_to_root(dl);
+ if (error == -EFSCORRUPTED || error == -ELNRNG || error == -ENOSR) {
+ /*
+ * Don't bother walking the paths if the xattr structure or the
+ * parent pointers are corrupt; this scan cannot be completed
+ * without full information.
+ */
+ xchk_ino_xref_set_corrupt(sc, sc->ip->i_ino);
+ error = 0;
+ goto out_scanlock;
+ }
+ if (error == -EBUSY) {
+ /*
+ * We couldn't scan some directory's parent pointers because
+ * the attr fork looked like it had been zapped. The
+ * scan was marked incomplete, so no further error code
+ * is necessary.
+ */
+ error = 0;
+ goto out_scanlock;
+ }
+ if (error)
+ goto out_scanlock;
+
+ /* Assess what we found in our path evaluation. */
+ xchk_dirtree_evaluate(dl, &oc);
+ if (xchk_dirtree_parentless(dl)) {
+ if (oc.good || oc.bad || oc.suspect)
+ xchk_ino_set_corrupt(sc, sc->ip->i_ino);
+ } else {
+ if (oc.bad || oc.good + oc.suspect != 1)
+ xchk_ino_set_corrupt(sc, sc->ip->i_ino);
+ if (oc.suspect)
+ xchk_ino_xref_set_corrupt(sc, sc->ip->i_ino);
+ }
+
+out_scanlock:
+ mutex_unlock(&dl->lock);
+ trace_xchk_dirtree_done(sc->ip, sc->sm, error);
+ return error;
+}
diff --git a/fs/xfs/scrub/dirtree.h b/fs/xfs/scrub/dirtree.h
new file mode 100644
index 0000000000000..a7797b618ec3f
--- /dev/null
+++ b/fs/xfs/scrub/dirtree.h
@@ -0,0 +1,123 @@
+/* SPDX-License-Identifier: GPL-2.0-or-later */
+/*
+ * Copyright (c) 2023-2024 Oracle. All Rights Reserved.
+ * Author: Darrick J. Wong <djwong@kernel.org>
+ */
+#ifndef __XFS_SCRUB_DIRTREE_H__
+#define __XFS_SCRUB_DIRTREE_H__
+
+/*
+ * Each of these represents one parent pointer path step in a chain going
+ * up towards the directory tree root. These are stored inside an xfarray.
+ */
+struct xchk_dirpath_step {
+ /* Directory entry name associated with this parent link. */
+ xfblob_cookie name_cookie;
+ unsigned int name_len;
+
+ /* Handle of the parent directory. */
+ unsigned int parent_gen;
+ xfs_ino_t parent_ino;
+};
+
+enum xchk_dirpath_outcome {
+ XCHK_DIRPATH_SCANNING = 0, /* still being put together */
+ XCHK_DIRPATH_DELETE, /* delete this path */
+ XCHK_DIRPATH_CORRUPT, /* corruption detected in path */
+ XCHK_DIRPATH_LOOP, /* cycle detected further up */
+ XCHK_DIRPATH_STALE, /* path is stale */
+ XCHK_DIRPATH_OK, /* path reaches the root */
+};
+
+/*
+ * Each of these represents one parent pointer path out of the directory being
+ * scanned. These exist in-core, and hopefully there aren't more than a
+ * handful of them.
+ */
+struct xchk_dirpath {
+ struct list_head list;
+
+ /* Index of the first step in this path. */
+ xfarray_idx_t first_step;
+
+ /* Index of the second step in this path. */
+ xfarray_idx_t second_step;
+
+ /* Inodes seen while walking this path. */
+ struct xino_bitmap seen_inodes;
+
+ /* Number of steps in this path. */
+ unsigned int nr_steps;
+
+ /* Which path is this? */
+ unsigned int path_nr;
+
+ /* What did we conclude from following this path? */
+ enum xchk_dirpath_outcome outcome;
+};
+
+struct xchk_dirtree_outcomes {
+ /* Number of XCHK_DIRPATH_DELETE */
+ unsigned int bad;
+
+ /* Number of XCHK_DIRPATH_CORRUPT or XCHK_DIRPATH_LOOP */
+ unsigned int suspect;
+
+ /* Number of XCHK_DIRPATH_OK */
+ unsigned int good;
+};
+
+struct xchk_dirtree {
+ struct xfs_scrub *sc;
+
+ /* Root inode that we're looking for. */
+ xfs_ino_t root_ino;
+
+ /* Scratch buffer for scanning pptr xattrs */
+ struct xfs_parent_scratch scratch;
+ struct xfs_parent_name_irec pptr;
+
+ /* lock for everything below here */
+ struct mutex lock;
+
+ /*
+ * All path steps observed during this scan. Each of the path
+ * steps for a particular pathwalk are recorded in sequential
+ * order in the xfarray. A pathwalk ends either with a step
+ * pointing to the root directory (success) or pointing to NULLFSINO
+ * (loop detected, empty dir detected, etc).
+ */
+ struct xfarray *path_steps;
+
+ /* All names observed during this scan. */
+ struct xfblob *path_names;
+
+ /* All paths being tracked by this scanner. */
+ struct list_head path_list;
+
+ /* Number of paths in path_list. */
+ unsigned int nr_paths;
+
+ /* Have the path data been invalidated by a concurrent update? */
+ bool stale:1;
+};
+
+#define xchk_dirtree_for_each_path_safe(dl, path, n) \
+ list_for_each_entry_safe((path), (n), &(dl)->path_list, list)
+
+#define xchk_dirtree_for_each_path(dl, path) \
+ list_for_each_entry((path), &(dl)->path_list, list)
+
+static inline bool
+xchk_dirtree_parentless(const struct xchk_dirtree *dl)
+{
+ struct xfs_scrub *sc = dl->sc;
+
+ if (sc->ip == sc->mp->m_rootip)
+ return true;
+ if (VFS_I(sc->ip)->i_nlink == 0)
+ return true;
+ return false;
+}
+
+#endif /* __XFS_SCRUB_DIRTREE_H__ */
diff --git a/fs/xfs/scrub/ino_bitmap.h b/fs/xfs/scrub/ino_bitmap.h
new file mode 100644
index 0000000000000..1300833679abf
--- /dev/null
+++ b/fs/xfs/scrub/ino_bitmap.h
@@ -0,0 +1,37 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * Copyright (c) 2023-2024 Oracle. All Rights Reserved.
+ * Author: Darrick J. Wong <djwong@kernel.org>
+ */
+#ifndef __XFS_SCRUB_INO_BITMAP_H__
+#define __XFS_SCRUB_INO_BITMAP_H__
+
+/* Bitmaps, but for type-checked for xfs_ino_t */
+
+struct xino_bitmap {
+ struct xbitmap64 inobitmap;
+};
+
+static inline void xino_bitmap_init(struct xino_bitmap *bitmap)
+{
+ xbitmap64_init(&bitmap->inobitmap);
+}
+
+static inline void xino_bitmap_destroy(struct xino_bitmap *bitmap)
+{
+ xbitmap64_destroy(&bitmap->inobitmap);
+}
+
+static inline int xino_bitmap_set(struct xino_bitmap *bitmap, xfs_ino_t ino)
+{
+ return xbitmap64_set(&bitmap->inobitmap, ino, 1);
+}
+
+static inline int xino_bitmap_test(struct xino_bitmap *bitmap, xfs_ino_t ino)
+{
+ uint64_t len = 1;
+
+ return xbitmap64_test(&bitmap->inobitmap, ino, &len);
+}
+
+#endif /* __XFS_SCRUB_INO_BITMAP_H__ */
diff --git a/fs/xfs/scrub/scrub.c b/fs/xfs/scrub/scrub.c
index d9c6d54ffad7f..b6f18d61d8816 100644
--- a/fs/xfs/scrub/scrub.c
+++ b/fs/xfs/scrub/scrub.c
@@ -439,6 +439,13 @@ static const struct xchk_meta_ops meta_scrub_ops[] = {
.scrub = xchk_health_record,
.repair = xrep_notsupported,
},
+ [XFS_SCRUB_TYPE_DIRTREE] = { /* directory tree structure */
+ .type = ST_INODE,
+ .setup = xchk_setup_dirtree,
+ .scrub = xchk_dirtree,
+ .has = xfs_has_parent,
+ .repair = xrep_notsupported,
+ },
};
static int
diff --git a/fs/xfs/scrub/scrub.h b/fs/xfs/scrub/scrub.h
index 665da3e3c1af1..ec685ef425778 100644
--- a/fs/xfs/scrub/scrub.h
+++ b/fs/xfs/scrub/scrub.h
@@ -185,6 +185,7 @@ int xchk_directory(struct xfs_scrub *sc);
int xchk_xattr(struct xfs_scrub *sc);
int xchk_symlink(struct xfs_scrub *sc);
int xchk_parent(struct xfs_scrub *sc);
+int xchk_dirtree(struct xfs_scrub *sc);
#ifdef CONFIG_XFS_RT
int xchk_rtbitmap(struct xfs_scrub *sc);
int xchk_rtsummary(struct xfs_scrub *sc);
diff --git a/fs/xfs/scrub/stats.c b/fs/xfs/scrub/stats.c
index b4ef1ebe28ab8..fd92df6389ac8 100644
--- a/fs/xfs/scrub/stats.c
+++ b/fs/xfs/scrub/stats.c
@@ -79,6 +79,7 @@ static const char *name_map[XFS_SCRUB_TYPE_NR] = {
[XFS_SCRUB_TYPE_FSCOUNTERS] = "fscounters",
[XFS_SCRUB_TYPE_QUOTACHECK] = "quotacheck",
[XFS_SCRUB_TYPE_NLINKS] = "nlinks",
+ [XFS_SCRUB_TYPE_DIRTREE] = "dirtree",
};
/* Format the scrub stats into a text buffer, similar to pcp style. */
diff --git a/fs/xfs/scrub/trace.c b/fs/xfs/scrub/trace.c
index 9fe1491adbb51..994a910eead80 100644
--- a/fs/xfs/scrub/trace.c
+++ b/fs/xfs/scrub/trace.c
@@ -30,6 +30,10 @@
#include "scrub/nlinks.h"
#include "scrub/fscounters.h"
#include "scrub/xfbtree.h"
+#include "scrub/bitmap.h"
+#include "scrub/ino_bitmap.h"
+#include "scrub/xfblob.h"
+#include "scrub/dirtree.h"
/* Figure out which block the btree cursor was pointing to. */
static inline xfs_fsblock_t
diff --git a/fs/xfs/scrub/trace.h b/fs/xfs/scrub/trace.h
index 10e2d6544c5ad..f5dddbc4d8594 100644
--- a/fs/xfs/scrub/trace.h
+++ b/fs/xfs/scrub/trace.h
@@ -29,6 +29,9 @@ struct xfbtree;
struct xfbtree_config;
struct xfs_rmap_update_params;
struct xfs_parent_name_irec;
+enum xchk_dirpath_outcome;
+struct xchk_dirtree;
+struct xchk_dirtree_outcomes;
/*
* ftrace's __print_symbolic requires that all enum values be wrapped in the
@@ -76,6 +79,7 @@ TRACE_DEFINE_ENUM(XFS_SCRUB_TYPE_FSCOUNTERS);
TRACE_DEFINE_ENUM(XFS_SCRUB_TYPE_QUOTACHECK);
TRACE_DEFINE_ENUM(XFS_SCRUB_TYPE_NLINKS);
TRACE_DEFINE_ENUM(XFS_SCRUB_TYPE_HEALTHY);
+TRACE_DEFINE_ENUM(XFS_SCRUB_TYPE_DIRTREE);
#define XFS_SCRUB_TYPE_STRINGS \
{ XFS_SCRUB_TYPE_PROBE, "probe" }, \
@@ -105,7 +109,8 @@ TRACE_DEFINE_ENUM(XFS_SCRUB_TYPE_HEALTHY);
{ XFS_SCRUB_TYPE_FSCOUNTERS, "fscounters" }, \
{ XFS_SCRUB_TYPE_QUOTACHECK, "quotacheck" }, \
{ XFS_SCRUB_TYPE_NLINKS, "nlinks" }, \
- { XFS_SCRUB_TYPE_HEALTHY, "healthy" }
+ { XFS_SCRUB_TYPE_HEALTHY, "healthy" }, \
+ { XFS_SCRUB_TYPE_DIRTREE, "dirtree" }
#define XFS_SCRUB_FLAG_STRINGS \
{ XFS_SCRUB_IFLAG_REPAIR, "repair" }, \
@@ -173,6 +178,8 @@ DEFINE_EVENT(xchk_class, name, \
DEFINE_SCRUB_EVENT(xchk_start);
DEFINE_SCRUB_EVENT(xchk_done);
DEFINE_SCRUB_EVENT(xchk_deadlock_retry);
+DEFINE_SCRUB_EVENT(xchk_dirtree_start);
+DEFINE_SCRUB_EVENT(xchk_dirtree_done);
DEFINE_SCRUB_EVENT(xrep_attempt);
DEFINE_SCRUB_EVENT(xrep_done);
@@ -1587,6 +1594,185 @@ DEFINE_XCHK_PPTR_EVENT(xchk_parent_defer);
DEFINE_XCHK_PPTR_EVENT(xchk_parent_slowpath);
DEFINE_XCHK_PPTR_EVENT(xchk_parent_ultraslowpath);
+DECLARE_EVENT_CLASS(xchk_dirtree_class,
+ TP_PROTO(struct xfs_scrub *sc, struct xfs_inode *ip,
+ unsigned int path_nr,
+ const struct xfs_parent_name_irec *pptr),
+ TP_ARGS(sc, ip, path_nr, pptr),
+ TP_STRUCT__entry(
+ __field(dev_t, dev)
+ __field(unsigned int, path_nr)
+ __field(xfs_ino_t, child_ino)
+ __field(unsigned int, child_gen)
+ __field(xfs_ino_t, parent_ino)
+ __field(unsigned int, parent_gen)
+ __field(unsigned int, namelen)
+ __dynamic_array(char, name, pptr->p_namelen)
+ ),
+ TP_fast_assign(
+ __entry->dev = sc->mp->m_super->s_dev;
+ __entry->path_nr = path_nr;
+ __entry->child_ino = ip->i_ino;
+ __entry->child_gen = VFS_I(ip)->i_generation;
+ __entry->parent_ino = pptr->p_ino;
+ __entry->parent_gen = pptr->p_gen;
+ __entry->namelen = pptr->p_namelen;
+ memcpy(__get_str(name), pptr->p_name, pptr->p_namelen);
+ ),
+ TP_printk("dev %d:%d path %u child_ino 0x%llx child_gen 0x%x parent_ino 0x%llx parent_gen 0x%x name '%.*s'",
+ MAJOR(__entry->dev), MINOR(__entry->dev),
+ __entry->path_nr,
+ __entry->child_ino,
+ __entry->child_gen,
+ __entry->parent_ino,
+ __entry->parent_gen,
+ __entry->namelen,
+ __get_str(name))
+);
+#define DEFINE_XCHK_DIRTREE_EVENT(name) \
+DEFINE_EVENT(xchk_dirtree_class, name, \
+ TP_PROTO(struct xfs_scrub *sc, struct xfs_inode *ip, \
+ unsigned int path_nr, \
+ const struct xfs_parent_name_irec *pptr), \
+ TP_ARGS(sc, ip, path_nr, pptr))
+DEFINE_XCHK_DIRTREE_EVENT(xchk_dirtree_create_path);
+DEFINE_XCHK_DIRTREE_EVENT(xchk_dirpath_walk_upwards);
+
+DECLARE_EVENT_CLASS(xchk_dirpath_class,
+ TP_PROTO(struct xfs_scrub *sc, struct xfs_inode *ip,
+ unsigned int path_nr, unsigned int step_nr,
+ const struct xfs_parent_name_irec *pptr),
+ TP_ARGS(sc, ip, path_nr, step_nr, pptr),
+ TP_STRUCT__entry(
+ __field(dev_t, dev)
+ __field(unsigned int, path_nr)
+ __field(unsigned int, step_nr)
+ __field(xfs_ino_t, child_ino)
+ __field(unsigned int, child_gen)
+ __field(xfs_ino_t, parent_ino)
+ __field(unsigned int, parent_gen)
+ __field(unsigned int, namelen)
+ __dynamic_array(char, name, pptr->p_namelen)
+ ),
+ TP_fast_assign(
+ __entry->dev = sc->mp->m_super->s_dev;
+ __entry->path_nr = path_nr;
+ __entry->step_nr = step_nr;
+ __entry->child_ino = ip->i_ino;
+ __entry->child_gen = VFS_I(ip)->i_generation;
+ __entry->parent_ino = pptr->p_ino;
+ __entry->parent_gen = pptr->p_gen;
+ __entry->namelen = pptr->p_namelen;
+ memcpy(__get_str(name), pptr->p_name, pptr->p_namelen);
+ ),
+ TP_printk("dev %d:%d path %u step %u child_ino 0x%llx child_gen 0x%x parent_ino 0x%llx parent_gen 0x%x name '%.*s'",
+ MAJOR(__entry->dev), MINOR(__entry->dev),
+ __entry->path_nr,
+ __entry->step_nr,
+ __entry->child_ino,
+ __entry->child_gen,
+ __entry->parent_ino,
+ __entry->parent_gen,
+ __entry->namelen,
+ __get_str(name))
+);
+#define DEFINE_XCHK_DIRPATH_EVENT(name) \
+DEFINE_EVENT(xchk_dirpath_class, name, \
+ TP_PROTO(struct xfs_scrub *sc, struct xfs_inode *ip, \
+ unsigned int path_nr, unsigned int step_nr, \
+ const struct xfs_parent_name_irec *pptr), \
+ TP_ARGS(sc, ip, path_nr, step_nr, pptr))
+DEFINE_XCHK_DIRPATH_EVENT(xchk_dirpath_disappeared);
+DEFINE_XCHK_DIRPATH_EVENT(xchk_dirpath_badgen);
+DEFINE_XCHK_DIRPATH_EVENT(xchk_dirpath_nondir_parent);
+DEFINE_XCHK_DIRPATH_EVENT(xchk_dirpath_unlinked_parent);
+DEFINE_XCHK_DIRPATH_EVENT(xchk_dirpath_found_next_step);
+
+TRACE_DEFINE_ENUM(XCHK_DIRPATH_SCANNING);
+TRACE_DEFINE_ENUM(XCHK_DIRPATH_DELETE);
+TRACE_DEFINE_ENUM(XCHK_DIRPATH_CORRUPT);
+TRACE_DEFINE_ENUM(XCHK_DIRPATH_LOOP);
+TRACE_DEFINE_ENUM(XCHK_DIRPATH_STALE);
+TRACE_DEFINE_ENUM(XCHK_DIRPATH_OK);
+
+#define XCHK_DIRPATH_OUTCOME_STRINGS \
+ { XCHK_DIRPATH_SCANNING, "scanning" }, \
+ { XCHK_DIRPATH_DELETE, "delete" }, \
+ { XCHK_DIRPATH_CORRUPT, "corrupt" }, \
+ { XCHK_DIRPATH_LOOP, "loop" }, \
+ { XCHK_DIRPATH_STALE, "stale" }, \
+ { XCHK_DIRPATH_OK, "ok" }
+
+DECLARE_EVENT_CLASS(xchk_dirpath_outcome_class,
+ TP_PROTO(struct xfs_scrub *sc, unsigned long long path_nr,
+ unsigned int nr_steps, \
+ unsigned int outcome),
+ TP_ARGS(sc, path_nr, nr_steps, outcome),
+ TP_STRUCT__entry(
+ __field(dev_t, dev)
+ __field(unsigned long long, path_nr)
+ __field(unsigned int, nr_steps)
+ __field(unsigned int, outcome)
+ ),
+ TP_fast_assign(
+ __entry->dev = sc->mp->m_super->s_dev;
+ __entry->path_nr = path_nr;
+ __entry->nr_steps = nr_steps;
+ __entry->outcome = outcome;
+ ),
+ TP_printk("dev %d:%d path %llu steps %u outcome %s",
+ MAJOR(__entry->dev), MINOR(__entry->dev),
+ __entry->path_nr,
+ __entry->nr_steps,
+ __print_symbolic(__entry->outcome, XCHK_DIRPATH_OUTCOME_STRINGS))
+);
+#define DEFINE_XCHK_DIRPATH_OUTCOME_EVENT(name) \
+DEFINE_EVENT(xchk_dirpath_outcome_class, name, \
+ TP_PROTO(struct xfs_scrub *sc, unsigned long long path_nr, \
+ unsigned int nr_steps, \
+ unsigned int outcome), \
+ TP_ARGS(sc, path_nr, nr_steps, outcome))
+DEFINE_XCHK_DIRPATH_OUTCOME_EVENT(xchk_dirpath_set_outcome);
+DEFINE_XCHK_DIRPATH_OUTCOME_EVENT(xchk_dirpath_evaluate_path);
+
+DECLARE_EVENT_CLASS(xchk_dirtree_evaluate_class,
+ TP_PROTO(const struct xchk_dirtree *dl,
+ const struct xchk_dirtree_outcomes *oc),
+ TP_ARGS(dl, oc),
+ TP_STRUCT__entry(
+ __field(dev_t, dev)
+ __field(xfs_ino_t, ino)
+ __field(xfs_ino_t, rootino)
+ __field(unsigned int, nr_paths)
+ __field(unsigned int, bad)
+ __field(unsigned int, suspect)
+ __field(unsigned int, good)
+ ),
+ TP_fast_assign(
+ __entry->dev = dl->sc->mp->m_super->s_dev;
+ __entry->ino = dl->sc->ip->i_ino;
+ __entry->rootino = dl->root_ino;
+ __entry->nr_paths = dl->nr_paths;
+ __entry->bad = oc->bad;
+ __entry->suspect = oc->suspect;
+ __entry->good = oc->good;
+ ),
+ TP_printk("dev %d:%d ino 0x%llx rootino 0x%llx nr_paths %u bad %u suspect %u good %u",
+ MAJOR(__entry->dev), MINOR(__entry->dev),
+ __entry->ino,
+ __entry->rootino,
+ __entry->nr_paths,
+ __entry->bad,
+ __entry->suspect,
+ __entry->good)
+);
+#define DEFINE_XCHK_DIRTREE_EVALUATE_EVENT(name) \
+DEFINE_EVENT(xchk_dirtree_evaluate_class, name, \
+ TP_PROTO(const struct xchk_dirtree *dl, \
+ const struct xchk_dirtree_outcomes *oc), \
+ TP_ARGS(dl, oc))
+DEFINE_XCHK_DIRTREE_EVALUATE_EVENT(xchk_dirtree_evaluate);
+
/* repair tracepoints */
#if IS_ENABLED(CONFIG_XFS_ONLINE_REPAIR)
diff --git a/fs/xfs/scrub/xfarray.h b/fs/xfs/scrub/xfarray.h
index f06af7eb484ec..6652070716095 100644
--- a/fs/xfs/scrub/xfarray.h
+++ b/fs/xfs/scrub/xfarray.h
@@ -8,6 +8,7 @@
/* xfile array index type, along with cursor initialization */
typedef uint64_t xfarray_idx_t;
+#define XFARRAY_NULLIDX ((__force xfarray_idx_t)-1ULL)
#define XFARRAY_CURSOR_INIT ((__force xfarray_idx_t)0)
/* Iterate each index of an xfile array. */
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 2/4] xfs: invalidate dirloop scrub path data when concurrent updates happen
2023-12-31 19:34 ` [PATCHSET v13.0 6/7] xfs: detect and correct directory tree problems Darrick J. Wong
2023-12-31 20:59 ` [PATCH 1/4] xfs: teach online scrub to find directory tree structure problems Darrick J. Wong
@ 2023-12-31 20:59 ` Darrick J. Wong
2023-12-31 20:59 ` [PATCH 3/4] xfs: report directory tree corruption in the health information Darrick J. Wong
2023-12-31 20:59 ` [PATCH 4/4] xfs: fix corruptions in the directory tree Darrick J. Wong
3 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:59 UTC (permalink / raw)
To: djwong; +Cc: linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
Add a dirent update hook so that we can detect directory tree updates
that affect any of the paths found by this scrubber and force it to
rescan.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/scrub/dirtree.c | 159 ++++++++++++++++++++++++++++++++++++++++++++++++
fs/xfs/scrub/dirtree.h | 19 ++++++
fs/xfs/scrub/trace.h | 65 ++++++++++++++++++++
3 files changed, 242 insertions(+), 1 deletion(-)
diff --git a/fs/xfs/scrub/dirtree.c b/fs/xfs/scrub/dirtree.c
index 9edaf89f46fdf..18005908434ed 100644
--- a/fs/xfs/scrub/dirtree.c
+++ b/fs/xfs/scrub/dirtree.c
@@ -69,6 +69,9 @@ xchk_dirtree_buf_cleanup(
struct xchk_dirtree *dl = buf;
struct xchk_dirpath *path, *n;
+ if (dl->scan_ino != NULLFSINO)
+ xfs_dir_hook_del(dl->sc->mp, &dl->hooks);
+
xchk_dirtree_for_each_path_safe(dl, path, n) {
list_del_init(&path->list);
xino_bitmap_destroy(&path->seen_inodes);
@@ -89,12 +92,15 @@ xchk_setup_dirtree(
char *descr;
int error;
+ xchk_fsgates_enable(sc, XCHK_FSGATES_DIRENTS);
+
dl = kvzalloc(sizeof(struct xchk_dirtree), XCHK_GFP_FLAGS);
if (!dl)
return -ENOMEM;
dl->sc = sc;
INIT_LIST_HEAD(&dl->path_list);
dl->root_ino = NULLFSINO;
+ dl->scan_ino = NULLFSINO;
mutex_init(&dl->lock);
@@ -542,6 +548,133 @@ xchk_dirpath_walk_upwards(
return error;
}
+/*
+ * Decide if this path step has been touched by this live update. Returns
+ * 1 for yes, 0 for no, or a negative errno.
+ */
+STATIC int
+xchk_dirpath_step_is_stale(
+ struct xchk_dirtree *dl,
+ struct xchk_dirpath *path,
+ unsigned int step_nr,
+ xfarray_idx_t step_idx,
+ struct xfs_dir_update_params *p,
+ struct xchk_dirpath_step *step)
+{
+ xfs_ino_t child_ino = step->parent_ino;
+ int error;
+
+ error = xfarray_load(dl->path_steps, step_idx, step);
+ if (error)
+ return error;
+
+ /*
+ * If the parent and child being updated are not the ones mentioned in
+ * this path step, the scan data is still ok.
+ */
+ if (p->ip->i_ino != child_ino || p->dp->i_ino != step->parent_ino)
+ return 0;
+
+ /*
+ * If the dirent name lengths or byte sequences are different, the scan
+ * data is still ok.
+ */
+ if (p->name->len != step->name_len)
+ return 0;
+
+ error = xfblob_load(dl->path_names, step->name_cookie,
+ dl->hook_namebuf, step->name_len);
+ if (error)
+ return error;
+
+ if (memcmp(dl->hook_namebuf, p->name->name, p->name->len) != 0)
+ return 0;
+
+ /* Exact match, scan data is out of date. */
+ trace_xchk_dirpath_changed(dl->sc, path->path_nr, step_nr, p->dp,
+ p->ip, p->name);
+ return 1;
+}
+
+/*
+ * Decide if this path has been touched by this live update. Returns 1 for
+ * yes, 0 for no, or a negative errno.
+ */
+STATIC int
+xchk_dirpath_is_stale(
+ struct xchk_dirtree *dl,
+ struct xchk_dirpath *path,
+ struct xfs_dir_update_params *p)
+{
+ struct xchk_dirpath_step step = {
+ .parent_ino = dl->scan_ino,
+ };
+ xfarray_idx_t idx = path->first_step;
+ unsigned int i;
+ int ret;
+
+ /*
+ * The child being updated has not been seen by this path at all; this
+ * path cannot be stale.
+ */
+ if (!xino_bitmap_test(&path->seen_inodes, p->ip->i_ino))
+ return 0;
+
+ ret = xchk_dirpath_step_is_stale(dl, path, 0, idx, p, &step);
+ if (ret != 0)
+ return ret;
+
+ for (i = 1, idx = path->second_step; i < path->nr_steps; i++, idx++) {
+ ret = xchk_dirpath_step_is_stale(dl, path, i, idx, p, &step);
+ if (ret != 0)
+ return ret;
+ }
+
+ return 0;
+}
+
+/*
+ * Decide if a directory update from the regular filesystem touches any of the
+ * paths we've scanned, and invalidate the scan data if true.
+ */
+STATIC int
+xchk_dirtree_live_update(
+ struct notifier_block *nb,
+ unsigned long action,
+ void *data)
+{
+ struct xfs_dir_update_params *p = data;
+ struct xchk_dirtree *dl;
+ struct xchk_dirpath *path;
+ int ret;
+
+ dl = container_of(nb, struct xchk_dirtree, hooks.dirent_hook.nb);
+
+ trace_xchk_dirtree_live_update(dl->sc, p->dp, action, p->ip, p->delta,
+ p->name);
+
+ mutex_lock(&dl->lock);
+
+ if (dl->stale || dl->aborted)
+ goto out_unlock;
+
+ xchk_dirtree_for_each_path(dl, path) {
+ ret = xchk_dirpath_is_stale(dl, path, p);
+ if (ret < 0) {
+ dl->aborted = true;
+ break;
+ }
+ if (ret == 1) {
+ dl->stale = true;
+ break;
+ }
+ }
+
+out_unlock:
+ mutex_unlock(&dl->lock);
+ return NOTIFY_DONE;
+}
+
/* Delete all the collected path information. */
STATIC void
xchk_dirtree_reset(
@@ -627,6 +760,8 @@ xchk_dirtree_find_paths_to_root(
}
if (error)
return error;
+ if (dl->aborted)
+ return 0;
}
} while (dl->stale);
@@ -698,11 +833,28 @@ xchk_dirtree(
ASSERT(xfs_has_parent(sc->mp));
- /* Find the root of the directory tree. */
+ /*
+ * Find the root of the directory tree. Remember which directory to
+ * scan, because the hook doesn't detach until after sc->ip gets
+ * released during teardown.
+ */
dl->root_ino = sc->mp->m_rootip->i_ino;
+ dl->scan_ino = sc->ip->i_ino;
trace_xchk_dirtree_start(sc->ip, sc->sm, 0);
+ /*
+ * Hook into the directory entry code so that we can capture updates to
+ * paths that we have already scanned. The scanner thread takes each
+ * directory's ILOCK, which means that any in-progress directory update
+ * will finish before we can scan the directory.
+ */
+ ASSERT(sc->flags & XCHK_FSGATES_DIRENTS);
+ xfs_hook_setup(&dl->hooks.dirent_hook, xchk_dirtree_live_update);
+ error = xfs_dir_hook_add(sc->mp, &dl->hooks);
+ if (error)
+ goto out;
+
mutex_lock(&dl->lock);
/* Trace each parent pointer's path to the root. */
@@ -729,6 +881,10 @@ xchk_dirtree(
}
if (error)
goto out_scanlock;
+ if (dl->aborted) {
+ xchk_set_incomplete(sc);
+ goto out_scanlock;
+ }
/* Assess what we found in our path evaluation. */
xchk_dirtree_evaluate(dl, &oc);
@@ -744,6 +900,7 @@ xchk_dirtree(
out_scanlock:
mutex_unlock(&dl->lock);
+out:
trace_xchk_dirtree_done(sc->ip, sc->sm, error);
return error;
}
diff --git a/fs/xfs/scrub/dirtree.h b/fs/xfs/scrub/dirtree.h
index a7797b618ec3f..ea374cf5f8362 100644
--- a/fs/xfs/scrub/dirtree.h
+++ b/fs/xfs/scrub/dirtree.h
@@ -73,13 +73,29 @@ struct xchk_dirtree {
/* Root inode that we're looking for. */
xfs_ino_t root_ino;
+ /*
+ * This is the inode that we're scanning. The live update hook can
+ * continue to be called after xchk_teardown drops sc->ip but before
+ * it calls buf_cleanup, so we keep a copy.
+ */
+ xfs_ino_t scan_ino;
+
/* Scratch buffer for scanning pptr xattrs */
struct xfs_parent_scratch scratch;
struct xfs_parent_name_irec pptr;
+ /*
+ * Hook into directory updates so that we can receive live updates
+ * from other writer threads.
+ */
+ struct xfs_dir_hook hooks;
+
/* lock for everything below here */
struct mutex lock;
+ /* buffer for the live update functions to use for dirent names */
+ unsigned char hook_namebuf[MAXNAMELEN];
+
/*
* All path steps observed during this scan. Each of the path
* steps for a particular pathwalk are recorded in sequential
@@ -100,6 +116,9 @@ struct xchk_dirtree {
/* Have the path data been invalidated by a concurrent update? */
bool stale:1;
+
+ /* Has the scan been aborted? */
+ bool aborted:1;
};
#define xchk_dirtree_for_each_path_safe(dl, path, n) \
diff --git a/fs/xfs/scrub/trace.h b/fs/xfs/scrub/trace.h
index f5dddbc4d8594..0d4c1580f61b7 100644
--- a/fs/xfs/scrub/trace.h
+++ b/fs/xfs/scrub/trace.h
@@ -1773,6 +1773,71 @@ DEFINE_EVENT(xchk_dirtree_evaluate_class, name, \
TP_ARGS(dl, oc))
DEFINE_XCHK_DIRTREE_EVALUATE_EVENT(xchk_dirtree_evaluate);
+TRACE_EVENT(xchk_dirpath_changed,
+ TP_PROTO(struct xfs_scrub *sc, unsigned int path_nr,
+ unsigned int step_nr, const struct xfs_inode *dp,
+ const struct xfs_inode *ip, const struct xfs_name *xname),
+ TP_ARGS(sc, path_nr, step_nr, dp, ip, xname),
+ TP_STRUCT__entry(
+ __field(dev_t, dev)
+ __field(unsigned int, path_nr)
+ __field(unsigned int, step_nr)
+ __field(xfs_ino_t, child_ino)
+ __field(xfs_ino_t, parent_ino)
+ __field(unsigned int, namelen)
+ __dynamic_array(char, name, xname->len)
+ ),
+ TP_fast_assign(
+ __entry->dev = sc->mp->m_super->s_dev;
+ __entry->path_nr = path_nr;
+ __entry->step_nr = step_nr;
+ __entry->child_ino = ip->i_ino;
+ __entry->parent_ino = dp->i_ino;
+ __entry->namelen = xname->len;
+ memcpy(__get_str(name), xname->name, xname->len);
+ ),
+ TP_printk("dev %d:%d path %u step %u child_ino 0x%llx parent_ino 0x%llx name '%.*s'",
+ MAJOR(__entry->dev), MINOR(__entry->dev),
+ __entry->path_nr,
+ __entry->step_nr,
+ __entry->child_ino,
+ __entry->parent_ino,
+ __entry->namelen,
+ __get_str(name))
+);
+
+TRACE_EVENT(xchk_dirtree_live_update,
+ TP_PROTO(struct xfs_scrub *sc, const struct xfs_inode *dp,
+ int action, const struct xfs_inode *ip, int delta,
+ const struct xfs_name *xname),
+ TP_ARGS(sc, dp, action, ip, delta, xname),
+ TP_STRUCT__entry(
+ __field(dev_t, dev)
+ __field(xfs_ino_t, parent_ino)
+ __field(int, action)
+ __field(xfs_ino_t, child_ino)
+ __field(int, delta)
+ __field(unsigned int, namelen)
+ __dynamic_array(char, name, xname->len)
+ ),
+ TP_fast_assign(
+ __entry->dev = sc->mp->m_super->s_dev;
+ __entry->parent_ino = dp->i_ino;
+ __entry->action = action;
+ __entry->child_ino = ip->i_ino;
+ __entry->delta = delta;
+ __entry->namelen = xname->len;
+ memcpy(__get_str(name), xname->name, xname->len);
+ ),
+ TP_printk("dev %d:%d parent_ino 0x%llx child_ino 0x%llx nlink_delta %d name '%.*s'",
+ MAJOR(__entry->dev), MINOR(__entry->dev),
+ __entry->parent_ino,
+ __entry->child_ino,
+ __entry->delta,
+ __entry->namelen,
+ __get_str(name))
+);
+
/* repair tracepoints */
#if IS_ENABLED(CONFIG_XFS_ONLINE_REPAIR)
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 3/4] xfs: report directory tree corruption in the health information
2023-12-31 19:34 ` [PATCHSET v13.0 6/7] xfs: detect and correct directory tree problems Darrick J. Wong
2023-12-31 20:59 ` [PATCH 1/4] xfs: teach online scrub to find directory tree structure problems Darrick J. Wong
2023-12-31 20:59 ` [PATCH 2/4] xfs: invalidate dirloop scrub path data when concurrent updates happen Darrick J. Wong
@ 2023-12-31 20:59 ` Darrick J. Wong
2023-12-31 20:59 ` [PATCH 4/4] xfs: fix corruptions in the directory tree Darrick J. Wong
3 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:59 UTC (permalink / raw)
To: djwong; +Cc: linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
Report directories that are the source of corruption in the directory
tree.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/libxfs/xfs_fs.h | 1 +
fs/xfs/libxfs/xfs_health.h | 4 +++-
fs/xfs/scrub/health.c | 1 +
fs/xfs/xfs_health.c | 1 +
4 files changed, 6 insertions(+), 1 deletion(-)
diff --git a/fs/xfs/libxfs/xfs_fs.h b/fs/xfs/libxfs/xfs_fs.h
index 48f38694f1232..2499a20f5f774 100644
--- a/fs/xfs/libxfs/xfs_fs.h
+++ b/fs/xfs/libxfs/xfs_fs.h
@@ -413,6 +413,7 @@ struct xfs_bulkstat {
#define XFS_BS_SICK_XATTR (1 << 5) /* extended attributes */
#define XFS_BS_SICK_SYMLINK (1 << 6) /* symbolic link remote target */
#define XFS_BS_SICK_PARENT (1 << 7) /* parent pointers */
+#define XFS_BS_SICK_DIRTREE (1 << 8) /* directory tree structure */
/*
* Project quota id helpers (previously projid was 16bit only
diff --git a/fs/xfs/libxfs/xfs_health.h b/fs/xfs/libxfs/xfs_health.h
index df07c5877ba44..bca1990f71da8 100644
--- a/fs/xfs/libxfs/xfs_health.h
+++ b/fs/xfs/libxfs/xfs_health.h
@@ -95,6 +95,7 @@ struct xfs_da_args;
/* Don't propagate sick status to ag health summary during inactivation */
#define XFS_SICK_INO_FORGET (1 << 12)
+#define XFS_SICK_INO_DIRTREE (1 << 13) /* directory tree structure */
/* Primary evidence of health problems in a given group. */
#define XFS_SICK_FS_PRIMARY (XFS_SICK_FS_COUNTERS | \
@@ -125,7 +126,8 @@ struct xfs_da_args;
XFS_SICK_INO_DIR | \
XFS_SICK_INO_XATTR | \
XFS_SICK_INO_SYMLINK | \
- XFS_SICK_INO_PARENT)
+ XFS_SICK_INO_PARENT | \
+ XFS_SICK_INO_DIRTREE)
#define XFS_SICK_INO_ZAPPED (XFS_SICK_INO_BMBTD_ZAPPED | \
XFS_SICK_INO_BMBTA_ZAPPED | \
diff --git a/fs/xfs/scrub/health.c b/fs/xfs/scrub/health.c
index 664d57247ddf5..12f3e9fca727f 100644
--- a/fs/xfs/scrub/health.c
+++ b/fs/xfs/scrub/health.c
@@ -110,6 +110,7 @@ static const struct xchk_health_map type_to_health_flag[XFS_SCRUB_TYPE_NR] = {
[XFS_SCRUB_TYPE_FSCOUNTERS] = { XHG_FS, XFS_SICK_FS_COUNTERS },
[XFS_SCRUB_TYPE_QUOTACHECK] = { XHG_FS, XFS_SICK_FS_QUOTACHECK },
[XFS_SCRUB_TYPE_NLINKS] = { XHG_FS, XFS_SICK_FS_NLINKS },
+ [XFS_SCRUB_TYPE_DIRTREE] = { XHG_INO, XFS_SICK_INO_DIRTREE },
};
/* Return the health status mask for this scrub type. */
diff --git a/fs/xfs/xfs_health.c b/fs/xfs/xfs_health.c
index bd884c154cf37..15ebfe331f277 100644
--- a/fs/xfs/xfs_health.c
+++ b/fs/xfs/xfs_health.c
@@ -454,6 +454,7 @@ static const struct ioctl_sick_map ino_map[] = {
{ XFS_SICK_INO_BMBTA_ZAPPED, XFS_BS_SICK_BMBTA },
{ XFS_SICK_INO_DIR_ZAPPED, XFS_BS_SICK_DIR },
{ XFS_SICK_INO_SYMLINK_ZAPPED, XFS_BS_SICK_SYMLINK },
+ { XFS_SICK_INO_DIRTREE, XFS_BS_SICK_DIRTREE },
{ 0, 0 },
};
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 4/4] xfs: fix corruptions in the directory tree
2023-12-31 19:34 ` [PATCHSET v13.0 6/7] xfs: detect and correct directory tree problems Darrick J. Wong
` (2 preceding siblings ...)
2023-12-31 20:59 ` [PATCH 3/4] xfs: report directory tree corruption in the health information Darrick J. Wong
@ 2023-12-31 20:59 ` Darrick J. Wong
3 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 20:59 UTC (permalink / raw)
To: djwong; +Cc: linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
Repair corruptions in the directory tree itself. Cycles are broken by
removing an incoming parent->child link. Multiply-owned directories are
fixed by pruning the extra parent -> child links Disconnected subtrees
are reconnected to the lost and found.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/Makefile | 1
fs/xfs/scrub/dirtree.c | 38 ++
fs/xfs/scrub/dirtree.h | 29 +
fs/xfs/scrub/dirtree_repair.c | 824 +++++++++++++++++++++++++++++++++++++++++
fs/xfs/scrub/orphanage.c | 6
fs/xfs/scrub/orphanage.h | 8
fs/xfs/scrub/repair.h | 4
fs/xfs/scrub/scrub.c | 2
fs/xfs/scrub/trace.h | 23 +
fs/xfs/xfs_inode.c | 2
fs/xfs/xfs_inode.h | 1
11 files changed, 930 insertions(+), 8 deletions(-)
create mode 100644 fs/xfs/scrub/dirtree_repair.c
diff --git a/fs/xfs/Makefile b/fs/xfs/Makefile
index 31cbc4fd88f99..c4a950ad517c9 100644
--- a/fs/xfs/Makefile
+++ b/fs/xfs/Makefile
@@ -204,6 +204,7 @@ xfs-y += $(addprefix scrub/, \
bmap_repair.o \
cow_repair.o \
dir_repair.o \
+ dirtree_repair.o \
findparent.o \
fscounters_repair.o \
ialloc_repair.o \
diff --git a/fs/xfs/scrub/dirtree.c b/fs/xfs/scrub/dirtree.c
index 18005908434ed..53fd89e48cfc6 100644
--- a/fs/xfs/scrub/dirtree.c
+++ b/fs/xfs/scrub/dirtree.c
@@ -26,6 +26,8 @@
#include "scrub/xfblob.h"
#include "scrub/listxattr.h"
#include "scrub/trace.h"
+#include "scrub/repair.h"
+#include "scrub/orphanage.h"
#include "scrub/dirtree.h"
/*
@@ -94,6 +96,12 @@ xchk_setup_dirtree(
xchk_fsgates_enable(sc, XCHK_FSGATES_DIRENTS);
+ if (xchk_could_repair(sc)) {
+ error = xrep_setup_dirtree(sc);
+ if (error)
+ return error;
+ }
+
dl = kvzalloc(sizeof(struct xchk_dirtree), XCHK_GFP_FLAGS);
if (!dl)
return -ENOMEM;
@@ -101,6 +109,7 @@ xchk_setup_dirtree(
INIT_LIST_HEAD(&dl->path_list);
dl->root_ino = NULLFSINO;
dl->scan_ino = NULLFSINO;
+ dl->parent_ino = NULLFSINO;
mutex_init(&dl->lock);
@@ -139,7 +148,7 @@ xchk_setup_dirtree(
* Add the parent pointer described by @dl->pptr to the given path as a new
* step. Returns -ELNRNG if the path is too deep.
*/
-STATIC int
+int
xchk_dirpath_append(
struct xchk_dirtree *dl,
struct xfs_inode *ip,
@@ -590,6 +599,22 @@ xchk_dirpath_step_is_stale(
if (memcmp(dl->hook_namebuf, p->name->name, p->name->len) != 0)
return 0;
+ /*
+ * If the update comes from the repair code itself, walk the state
+ * machine forward.
+ */
+ if (p->ip->i_ino == dl->scan_ino &&
+ path->outcome == XREP_DIRPATH_ADOPTING) {
+ xchk_dirpath_set_outcome(dl, path, XREP_DIRPATH_ADOPTED);
+ return 0;
+ }
+
+ if (p->ip->i_ino == dl->scan_ino &&
+ path->outcome == XREP_DIRPATH_DELETING) {
+ xchk_dirpath_set_outcome(dl, path, XREP_DIRPATH_DELETED);
+ return 0;
+ }
+
/* Exact match, scan data is out of date. */
trace_xchk_dirpath_changed(dl->sc, path->path_nr, step_nr, p->dp,
p->ip, p->name);
@@ -705,7 +730,7 @@ xchk_dirtree_reset(
* path was too deep; -ENOSR if there were too many parent pointers; or
* a negative errno.
*/
-STATIC int
+int
xchk_dirtree_find_paths_to_root(
struct xchk_dirtree *dl)
{
@@ -772,7 +797,7 @@ xchk_dirtree_find_paths_to_root(
* Figure out what to do with the paths we tried to find. Do not call this
* if the scan results are stale.
*/
-STATIC void
+void
xchk_dirtree_evaluate(
struct xchk_dirtree *dl,
struct xchk_dirtree_outcomes *oc)
@@ -809,6 +834,13 @@ xchk_dirtree_evaluate(
/* This path got all the way to the root. */
oc->good++;
break;
+ case XREP_DIRPATH_DELETING:
+ case XREP_DIRPATH_DELETED:
+ case XREP_DIRPATH_ADOPTING:
+ case XREP_DIRPATH_ADOPTED:
+ /* These should not be in progress! */
+ ASSERT(0);
+ break;
}
}
diff --git a/fs/xfs/scrub/dirtree.h b/fs/xfs/scrub/dirtree.h
index ea374cf5f8362..a5dca42906e0e 100644
--- a/fs/xfs/scrub/dirtree.h
+++ b/fs/xfs/scrub/dirtree.h
@@ -27,6 +27,11 @@ enum xchk_dirpath_outcome {
XCHK_DIRPATH_LOOP, /* cycle detected further up */
XCHK_DIRPATH_STALE, /* path is stale */
XCHK_DIRPATH_OK, /* path reaches the root */
+
+ XREP_DIRPATH_DELETING, /* path is being deleted */
+ XREP_DIRPATH_DELETED, /* path has been deleted */
+ XREP_DIRPATH_ADOPTING, /* path is being adopted */
+ XREP_DIRPATH_ADOPTED, /* path has been adopted */
};
/*
@@ -65,6 +70,9 @@ struct xchk_dirtree_outcomes {
/* Number of XCHK_DIRPATH_OK */
unsigned int good;
+
+ /* Directory needs to be added to lost+found */
+ bool needs_adoption;
};
struct xchk_dirtree {
@@ -80,16 +88,30 @@ struct xchk_dirtree {
*/
xfs_ino_t scan_ino;
+ /*
+ * If we start deleting redundant paths to this subdirectory, this is
+ * the inode number of the surviving parent and the dotdot entry will
+ * be set to this value. If the value is NULLFSINO, then use @root_ino
+ * as a stand-in until the orphanage can adopt the subdirectory.
+ */
+ xfs_ino_t parent_ino;
+
/* Scratch buffer for scanning pptr xattrs */
struct xfs_parent_scratch scratch;
struct xfs_parent_name_irec pptr;
+ /* Information for reparenting this directory. */
+ struct xrep_adoption adoption;
+
/*
* Hook into directory updates so that we can receive live updates
* from other writer threads.
*/
struct xfs_dir_hook hooks;
+ /* Parent pointer update arguments. */
+ struct xfs_parent_args ppargs;
+
/* lock for everything below here */
struct mutex lock;
@@ -139,4 +161,11 @@ xchk_dirtree_parentless(const struct xchk_dirtree *dl)
return false;
}
+int xchk_dirtree_find_paths_to_root(struct xchk_dirtree *dl);
+int xchk_dirpath_append(struct xchk_dirtree *dl, struct xfs_inode *ip,
+ struct xchk_dirpath *path,
+ const struct xfs_parent_name_irec *pptr);
+void xchk_dirtree_evaluate(struct xchk_dirtree *dl,
+ struct xchk_dirtree_outcomes *oc);
+
#endif /* __XFS_SCRUB_DIRTREE_H__ */
diff --git a/fs/xfs/scrub/dirtree_repair.c b/fs/xfs/scrub/dirtree_repair.c
new file mode 100644
index 0000000000000..ba35d574437d2
--- /dev/null
+++ b/fs/xfs/scrub/dirtree_repair.c
@@ -0,0 +1,824 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * Copyright (c) 2023-2024 Oracle. All Rights Reserved.
+ * Author: Darrick J. Wong <djwong@kernel.org>
+ */
+#include "xfs.h"
+#include "xfs_fs.h"
+#include "xfs_shared.h"
+#include "xfs_format.h"
+#include "xfs_trans_resv.h"
+#include "xfs_trans_space.h"
+#include "xfs_mount.h"
+#include "xfs_log_format.h"
+#include "xfs_trans.h"
+#include "xfs_inode.h"
+#include "xfs_icache.h"
+#include "xfs_dir2.h"
+#include "xfs_dir2_priv.h"
+#include "xfs_attr.h"
+#include "xfs_parent.h"
+#include "scrub/scrub.h"
+#include "scrub/common.h"
+#include "scrub/bitmap.h"
+#include "scrub/ino_bitmap.h"
+#include "scrub/xfile.h"
+#include "scrub/xfarray.h"
+#include "scrub/xfblob.h"
+#include "scrub/listxattr.h"
+#include "scrub/trace.h"
+#include "scrub/repair.h"
+#include "scrub/orphanage.h"
+#include "scrub/dirtree.h"
+#include "scrub/readdir.h"
+
+/*
+ * Directory Tree Structure Repairs
+ * ================================
+ *
+ * If we decide that the directory being scanned is participating in a
+ * directory loop, the only change we can make is to remove directory entries
+ * pointing down to @sc->ip. If that leaves it with no parents, the directory
+ * should be adopted by the orphanage.
+ */
+
+/* Set up to repair directory loops. */
+int
+xrep_setup_dirtree(
+ struct xfs_scrub *sc)
+{
+ return xrep_orphanage_try_create(sc);
+}
+
+/* Change the outcome of this path. */
+static inline void
+xrep_dirpath_set_outcome(
+ struct xchk_dirtree *dl,
+ struct xchk_dirpath *path,
+ enum xchk_dirpath_outcome outcome)
+{
+ trace_xrep_dirpath_set_outcome(dl->sc, path->path_nr, path->nr_steps,
+ outcome);
+
+ path->outcome = outcome;
+}
+
+/* Delete all paths. */
+STATIC void
+xrep_dirtree_delete_all_paths(
+ struct xchk_dirtree *dl,
+ struct xchk_dirtree_outcomes *oc)
+{
+ struct xchk_dirpath *path;
+
+ xchk_dirtree_for_each_path(dl, path) {
+ switch (path->outcome) {
+ case XCHK_DIRPATH_CORRUPT:
+ case XCHK_DIRPATH_LOOP:
+ oc->suspect--;
+ oc->bad++;
+ xrep_dirpath_set_outcome(dl, path, XCHK_DIRPATH_DELETE);
+ break;
+ case XCHK_DIRPATH_OK:
+ oc->good--;
+ oc->bad++;
+ xrep_dirpath_set_outcome(dl, path, XCHK_DIRPATH_DELETE);
+ break;
+ default:
+ break;
+ }
+ }
+
+ ASSERT(oc->suspect == 0);
+ ASSERT(oc->good == 0);
+}
+
+/* Since this is the surviving path, set the dotdot entry to this value. */
+STATIC void
+xrep_dirpath_retain_parent(
+ struct xchk_dirtree *dl,
+ struct xchk_dirpath *path)
+{
+ struct xchk_dirpath_step step;
+ int error;
+
+ error = xfarray_load(dl->path_steps, path->first_step, &step);
+ if (error)
+ return;
+
+ dl->parent_ino = step.parent_ino;
+}
+
+/* Find the one surviving path so we know how to set dotdot. */
+STATIC void
+xrep_dirtree_find_surviving_path(
+ struct xchk_dirtree *dl,
+ struct xchk_dirtree_outcomes *oc)
+{
+ struct xchk_dirpath *path;
+ bool foundit = false;
+
+ xchk_dirtree_for_each_path(dl, path) {
+ switch (path->outcome) {
+ case XCHK_DIRPATH_CORRUPT:
+ case XCHK_DIRPATH_LOOP:
+ case XCHK_DIRPATH_OK:
+ if (!foundit) {
+ xrep_dirpath_retain_parent(dl, path);
+ foundit = true;
+ continue;
+ }
+ ASSERT(foundit == false);
+ break;
+ default:
+ break;
+ }
+ }
+
+ ASSERT(oc->suspect + oc->good == 1);
+}
+
+/* Delete all paths except for the one good one. */
+STATIC void
+xrep_dirtree_keep_one_good_path(
+ struct xchk_dirtree *dl,
+ struct xchk_dirtree_outcomes *oc)
+{
+ struct xchk_dirpath *path;
+ bool foundit = false;
+
+ xchk_dirtree_for_each_path(dl, path) {
+ switch (path->outcome) {
+ case XCHK_DIRPATH_CORRUPT:
+ case XCHK_DIRPATH_LOOP:
+ oc->suspect--;
+ oc->bad++;
+ xrep_dirpath_set_outcome(dl, path, XCHK_DIRPATH_DELETE);
+ break;
+ case XCHK_DIRPATH_OK:
+ if (!foundit) {
+ xrep_dirpath_retain_parent(dl, path);
+ foundit = true;
+ continue;
+ }
+ oc->good--;
+ oc->bad++;
+ xrep_dirpath_set_outcome(dl, path, XCHK_DIRPATH_DELETE);
+ break;
+ default:
+ break;
+ }
+ }
+
+ ASSERT(oc->suspect == 0);
+ ASSERT(oc->good < 2);
+}
+
+/* Delete all paths except for one suspect one. */
+STATIC void
+xrep_dirtree_keep_one_suspect_path(
+ struct xchk_dirtree *dl,
+ struct xchk_dirtree_outcomes *oc)
+{
+ struct xchk_dirpath *path;
+ bool foundit = false;
+
+ xchk_dirtree_for_each_path(dl, path) {
+ switch (path->outcome) {
+ case XCHK_DIRPATH_CORRUPT:
+ case XCHK_DIRPATH_LOOP:
+ if (!foundit) {
+ xrep_dirpath_retain_parent(dl, path);
+ foundit = true;
+ continue;
+ }
+ oc->suspect--;
+ oc->bad++;
+ xrep_dirpath_set_outcome(dl, path, XCHK_DIRPATH_DELETE);
+ break;
+ case XCHK_DIRPATH_OK:
+ ASSERT(0);
+ break;
+ default:
+ break;
+ }
+ }
+
+ ASSERT(oc->suspect == 1);
+ ASSERT(oc->good == 0);
+}
+
+/*
+ * Figure out what to do with the paths we tried to find. Returns -EDEADLOCK
+ * if the scan results have become stale.
+ */
+STATIC void
+xrep_dirtree_decide_fate(
+ struct xchk_dirtree *dl,
+ struct xchk_dirtree_outcomes *oc)
+{
+ xchk_dirtree_evaluate(dl, oc);
+
+ /* Parentless directories should not have any paths at all. */
+ if (xchk_dirtree_parentless(dl)) {
+ xrep_dirtree_delete_all_paths(dl, oc);
+ return;
+ }
+
+ /* One path is exactly the number of paths we want. */
+ if (oc->good + oc->suspect == 1) {
+ xrep_dirtree_find_surviving_path(dl, oc);
+ return;
+ }
+
+ /* Zero paths means we should reattach the subdir to the orphanage. */
+ if (oc->good + oc->suspect == 0) {
+ if (dl->sc->orphanage)
+ oc->needs_adoption = true;
+ return;
+ }
+
+ /*
+ * Otherwise, this subdirectory has too many parents. If there's at
+ * least one good path, keep it and delete the others.
+ */
+ if (oc->good > 0) {
+ xrep_dirtree_keep_one_good_path(dl, oc);
+ return;
+ }
+
+ /*
+ * There are no good paths and there are too many suspect paths.
+ * Keep the first suspect path and delete the rest.
+ */
+ xrep_dirtree_keep_one_suspect_path(dl, oc);
+}
+
+/*
+ * Load the first step of this path into @step and @dl->pptr
+ * for later repair work.
+ */
+STATIC int
+xrep_dirtree_prep_path(
+ struct xchk_dirtree *dl,
+ struct xchk_dirpath *path,
+ struct xchk_dirpath_step *step)
+{
+ int error;
+
+ error = xfarray_load(dl->path_steps, path->first_step, step);
+ if (error)
+ return error;
+
+ dl->pptr.p_ino = step->parent_ino;
+ dl->pptr.p_gen = step->parent_gen;
+ dl->pptr.p_namelen = step->name_len;
+
+ return xfblob_load(dl->path_names, step->name_cookie, dl->pptr.p_name,
+ step->name_len);
+}
+
+/* Delete the VFS dentry for a removed child. */
+STATIC int
+xrep_dirtree_purge_dentry(
+ struct xchk_dirtree *dl,
+ struct xfs_inode *dp,
+ const struct xfs_name *name)
+{
+ struct qstr qname = QSTR_INIT(name->name, name->len);
+ struct dentry *parent_dentry, *child_dentry;
+ int error = 0;
+
+ /*
+ * Find the dentry for the parent directory. If there isn't one, we're
+ * done. Caller already holds i_rwsem for parent and child.
+ */
+ parent_dentry = d_find_alias(VFS_I(dp));
+ if (!parent_dentry)
+ return 0;
+
+ /* The VFS thinks the parent is a directory, right? */
+ if (!d_is_dir(parent_dentry)) {
+ ASSERT(d_is_dir(parent_dentry));
+ error = -EFSCORRUPTED;
+ goto out_dput_parent;
+ }
+
+ /*
+ * Try to find the dirent pointing to the child. If there isn't one,
+ * we're done.
+ */
+ qname.hash = full_name_hash(parent_dentry, name->name, name->len);
+ child_dentry = d_lookup(parent_dentry, &qname);
+ if (!child_dentry) {
+ error = 0;
+ goto out_dput_parent;
+ }
+
+ trace_xrep_dirtree_delete_child(dp->i_mount, child_dentry);
+
+ /* Child is not a directory? We're screwed. */
+ if (!d_is_dir(child_dentry)) {
+ ASSERT(d_is_dir(child_dentry));
+ error = -EFSCORRUPTED;
+ goto out_dput_child;
+ }
+
+ /* Replace the child dentry with a negative one. */
+ d_delete(child_dentry);
+
+out_dput_child:
+ dput(child_dentry);
+out_dput_parent:
+ dput(parent_dentry);
+ return error;
+}
+
+/*
+ * Prepare to delete a link by taking the IOLOCK of the parent and the child
+ * (scrub target). Caller must hold IOLOCK_EXCL on @sc->ip. Returns 0 if we
+ * took both locks, or a negative errno if we couldn't lock the parent in time.
+ */
+static inline int
+xrep_dirtree_unlink_iolock(
+ struct xfs_scrub *sc,
+ struct xfs_inode *dp)
+{
+ int error;
+
+ ASSERT(sc->ilock_flags & XFS_IOLOCK_EXCL);
+
+ if (xfs_ilock_nowait(dp, XFS_IOLOCK_EXCL))
+ return 0;
+
+ xchk_iunlock(sc, XFS_IOLOCK_EXCL);
+ do {
+ xfs_ilock(dp, XFS_IOLOCK_EXCL);
+ if (xchk_ilock_nowait(sc, XFS_IOLOCK_EXCL))
+ break;
+ xfs_iunlock(dp, XFS_IOLOCK_EXCL);
+
+ if (xchk_should_terminate(sc, &error)) {
+ xchk_ilock(sc, XFS_IOLOCK_EXCL);
+ return error;
+ }
+
+ delay(1);
+ } while (1);
+
+ return 0;
+}
+
+/*
+ * Remove a link from the directory tree and update the dcache. Returns
+ * -ESTALE if the scan data are now out of date.
+ */
+STATIC int
+xrep_dirtree_unlink(
+ struct xchk_dirtree *dl,
+ struct xfs_inode *dp,
+ struct xchk_dirpath *path,
+ struct xchk_dirpath_step *step)
+{
+ struct xfs_name name = {
+ .name = dl->pptr.p_name,
+ .len = step->name_len,
+ };
+ struct xfs_scrub *sc = dl->sc;
+ struct xfs_mount *mp = sc->mp;
+ xfs_ino_t dotdot_ino;
+ xfs_ino_t parent_ino = dl->parent_ino;
+ unsigned int resblks;
+ int dontcare;
+ int error;
+
+ /* Take IOLOCK_EXCL of the parent and child. */
+ error = xrep_dirtree_unlink_iolock(sc, dp);
+ if (error)
+ return error;
+
+ xfs_parent_args_init(mp, &dl->ppargs);
+
+ /*
+ * Create the transaction that we need to sever the path. Ignore
+ * EDQUOT and ENOSPC being returned via nospace_error because the
+ * directory code can handle a reservationless update.
+ */
+ resblks = xfs_remove_space_res(mp, step->name_len);
+ error = xfs_trans_alloc_dir(dp, &M_RES(mp)->tr_remove, sc->ip,
+ &resblks, &sc->tp, &dontcare);
+ if (error)
+ goto out_iolock;
+
+ /*
+ * Cancel if someone invalidate the paths while we were trying to get
+ * the ILOCK.
+ */
+ mutex_lock(&dl->lock);
+ if (dl->stale) {
+ mutex_unlock(&dl->lock);
+ error = -ESTALE;
+ goto out_trans_cancel;
+ }
+ xrep_dirpath_set_outcome(dl, path, XREP_DIRPATH_DELETING);
+ mutex_unlock(&dl->lock);
+
+ trace_xrep_dirtree_delete_path(dl->sc, sc->ip, path->path_nr,
+ &dl->pptr);
+
+ /*
+ * Decide if we need to reset the dotdot entry. Rules:
+ *
+ * - If there's a surviving parent, we want dotdot to point there.
+ * - If we don't have any surviving parents, then point dotdot at the
+ * root dir.
+ * - If dotdot is already set to the value we want, pass in NULLFSINO
+ * for no change necessary.
+ *
+ * Do this /before/ we dirty anything, in case the dotdot lookup
+ * fails.
+ */
+ error = xchk_dir_lookup(sc, sc->ip, &xfs_name_dotdot, &dotdot_ino);
+ if (error)
+ goto out_trans_cancel;
+ if (parent_ino == NULLFSINO)
+ parent_ino = dl->root_ino;
+ if (dotdot_ino == parent_ino)
+ parent_ino = NULLFSINO;
+
+ /* Drop the link from sc->ip's dotdot entry. */
+ error = xfs_droplink(sc->tp, dp);
+ if (error)
+ goto out_trans_cancel;
+
+ /* Reset the dotdot entry to a surviving parent. */
+ if (parent_ino != NULLFSINO) {
+ error = xfs_dir_replace(sc->tp, sc->ip, &xfs_name_dotdot,
+ parent_ino, 0);
+ if (error)
+ goto out_trans_cancel;
+ }
+
+ /* Drop the link from dp to sc->ip. */
+ error = xfs_droplink(sc->tp, sc->ip);
+ if (error)
+ goto out_trans_cancel;
+
+ error = xfs_dir_removename(sc->tp, dp, &name, sc->ip->i_ino, resblks);
+ if (error) {
+ ASSERT(error != -ENOENT);
+ goto out_trans_cancel;
+ }
+
+ if (xfs_has_parent(sc->mp)) {
+ error = xfs_parent_removename(sc->tp, &dl->ppargs, dp, &name,
+ sc->ip);
+ if (error)
+ goto out_trans_cancel;
+ }
+
+ /*
+ * Notify dirent hooks that we removed the bad link, invalidate the
+ * dcache, and commit the repair.
+ */
+ xfs_dir_update_hook(dp, sc->ip, -1, &name);
+ error = xrep_dirtree_purge_dentry(dl, dp, &name);
+ if (error)
+ goto out_trans_cancel;
+
+ error = xrep_trans_commit(sc);
+ goto out_ilock;
+
+out_trans_cancel:
+ xchk_trans_cancel(sc);
+out_ilock:
+ xfs_iunlock(sc->ip, XFS_ILOCK_EXCL);
+ xfs_iunlock(dp, XFS_ILOCK_EXCL);
+out_iolock:
+ xfs_iunlock(dp, XFS_IOLOCK_EXCL);
+ return error;
+}
+
+/*
+ * Delete a directory entry that points to this directory. Returns -ESTALE
+ * if the scan data are now out of date.
+ */
+STATIC int
+xrep_dirtree_delete_path(
+ struct xchk_dirtree *dl,
+ struct xchk_dirpath *path)
+{
+ struct xchk_dirpath_step step;
+ struct xfs_scrub *sc = dl->sc;
+ struct xfs_inode *dp;
+ int error;
+
+ /*
+ * Load the parent pointer and directory inode for this path, then
+ * drop the scan lock, the ILOCK, and the transaction so that
+ * _delete_path can reserve the proper transaction.
+ */
+ error = xrep_dirtree_prep_path(dl, path, &step);
+ if (error)
+ return error;
+
+ error = xchk_iget(sc, step.parent_ino, &dp);
+ if (error)
+ return error;
+
+ mutex_unlock(&dl->lock);
+ xchk_trans_cancel(sc);
+ xchk_iunlock(sc, XFS_ILOCK_EXCL);
+
+ /* Delete the directory link and release the parent. */
+ error = xrep_dirtree_unlink(dl, dp, path, &step);
+ xchk_irele(sc, dp);
+
+ /*
+ * Retake all the resources we had at the beginning even if the repair
+ * failed or the scan data are now stale. This keeps things simple for
+ * the caller.
+ */
+ xchk_trans_alloc_empty(sc);
+ xchk_ilock(sc, XFS_ILOCK_EXCL);
+ mutex_lock(&dl->lock);
+
+ if (!error && dl->stale)
+ error = -ESTALE;
+ return error;
+}
+
+/* Add a new path to represent our in-progress adoption. */
+STATIC int
+xrep_dirtree_create_adoption_path(
+ struct xchk_dirtree *dl)
+{
+ struct xfs_scrub *sc = dl->sc;
+ struct xchk_dirpath *path;
+ int error;
+
+ /*
+ * We should have capped the number of paths at XFS_MAXLINK-1 in the
+ * scanner.
+ */
+ if (dl->nr_paths > XFS_MAXLINK) {
+ ASSERT(dl->nr_paths <= XFS_MAXLINK);
+ return -EFSCORRUPTED;
+ }
+
+ /*
+ * Create a new xchk_path structure to remember this parent pointer
+ * and record the first name step.
+ */
+ path = kmalloc(sizeof(struct xchk_dirpath), XCHK_GFP_FLAGS);
+ if (!path)
+ return -ENOMEM;
+
+ INIT_LIST_HEAD(&path->list);
+ xino_bitmap_init(&path->seen_inodes);
+ path->nr_steps = 0;
+ path->outcome = XREP_DIRPATH_ADOPTING;
+
+ /*
+ * Record the new link that we just created in the orphanage. Because
+ * adoption is the last repair that we perform, we don't bother filling
+ * in the path all the way back to the root.
+ */
+ dl->pptr.p_ino = sc->orphanage->i_ino;
+ dl->pptr.p_gen = VFS_I(sc->orphanage)->i_generation;
+ dl->pptr.p_namelen = dl->adoption.xname.len;
+
+ error = xino_bitmap_set(&path->seen_inodes, sc->orphanage->i_ino);
+ if (error)
+ goto out_path;
+
+ trace_xrep_dirtree_create_adoption(sc, sc->ip, dl->nr_paths, &dl->pptr);
+
+ error = xchk_dirpath_append(dl, sc->ip, path, &dl->pptr);
+ if (error)
+ goto out_path;
+
+ path->first_step = xfarray_length(dl->path_steps) - 1;
+ path->second_step = XFARRAY_NULLIDX;
+ path->path_nr = dl->nr_paths;
+
+ list_add_tail(&path->list, &dl->path_list);
+ dl->nr_paths++;
+ return 0;
+
+out_path:
+ kfree(path);
+ return error;
+}
+
+/*
+ * Prepare to move a file to the orphanage by taking the IOLOCK of the
+ * orphanage and the child (scrub target). Caller must hold IOLOCK_EXCL on
+ * @sc->ip. Returns 0 if we took both locks, or a negative errno if we
+ * couldn't lock the orphanage in time.
+ */
+static inline int
+xrep_dirtree_adopt_iolock(
+ struct xfs_scrub *sc)
+{
+ int error;
+
+ ASSERT(sc->ilock_flags & XFS_IOLOCK_EXCL);
+
+ if (xrep_orphanage_ilock_nowait(sc, XFS_IOLOCK_EXCL))
+ return 0;
+
+ xchk_iunlock(sc, XFS_IOLOCK_EXCL);
+ do {
+ xrep_orphanage_ilock(sc, XFS_IOLOCK_EXCL);
+ if (xchk_ilock_nowait(sc, XFS_IOLOCK_EXCL))
+ break;
+ xrep_orphanage_iunlock(sc, XFS_IOLOCK_EXCL);
+
+ if (xchk_should_terminate(sc, &error)) {
+ xchk_ilock(sc, XFS_IOLOCK_EXCL);
+ return error;
+ }
+
+ delay(1);
+ } while (1);
+
+ return 0;
+}
+
+/*
+ * Reattach this orphaned directory to the orphanage. Do not call this with
+ * any resources held. Returns -ESTALE if the scan data have become out of
+ * date.
+ */
+STATIC int
+xrep_dirtree_adopt(
+ struct xchk_dirtree *dl)
+{
+ struct xfs_scrub *sc = dl->sc;
+ int error;
+
+ /* Take the IOLOCK of the orphanage and the scrub target. */
+ error = xrep_dirtree_adopt_iolock(sc);
+ if (error)
+ return error;
+
+ /*
+ * Set up for an adoption. The directory tree fixer runs after the
+ * link counts have been corrected. Therefore, we must bump the
+ * child's link count since there will be no further opportunity to fix
+ * errors.
+ */
+ error = xrep_adoption_trans_alloc(sc, &dl->adoption);
+ if (error)
+ goto out_iolock;
+ dl->adoption.bump_child_nlink = true;
+
+ /* Figure out what name we're going to use here. */
+ error = xrep_adoption_compute_name(&dl->adoption, dl->pptr.p_name);
+ if (error)
+ goto out_trans;
+
+ /*
+ * Now that we have a proposed name for the orphanage entry, create
+ * a faux path so that the live update hook will see it.
+ */
+ mutex_lock(&dl->lock);
+ if (dl->stale) {
+ mutex_unlock(&dl->lock);
+ error = -ESTALE;
+ goto out_trans;
+ }
+ error = xrep_dirtree_create_adoption_path(dl);
+ mutex_unlock(&dl->lock);
+ if (error)
+ goto out_trans;
+
+ /* Reparent the directory. */
+ error = xrep_adoption_move(&dl->adoption);
+ if (error)
+ goto out_trans;
+
+ /*
+ * Commit the name and release all inode locks except for the scrub
+ * target's IOLOCK.
+ */
+ error = xrep_trans_commit(sc);
+ goto out_ilock;
+
+out_trans:
+ xchk_trans_cancel(sc);
+out_ilock:
+ xchk_iunlock(sc, XFS_ILOCK_EXCL);
+ xrep_orphanage_iunlock(sc, XFS_ILOCK_EXCL);
+out_iolock:
+ xrep_orphanage_iunlock(sc, XFS_IOLOCK_EXCL);
+ return error;
+}
+
+/*
+ * This newly orphaned directory needs to be adopted by the orphanage.
+ * Make this happen.
+ */
+STATIC int
+xrep_dirtree_move_to_orphanage(
+ struct xchk_dirtree *dl)
+{
+ struct xfs_scrub *sc = dl->sc;
+ int error;
+
+ /*
+ * Start by dropping all the resources that we hold so that we can grab
+ * all the resources that we need for the adoption.
+ */
+ mutex_unlock(&dl->lock);
+ xchk_trans_cancel(sc);
+ xchk_iunlock(sc, XFS_ILOCK_EXCL);
+
+ /* Perform the adoption. */
+ error = xrep_dirtree_adopt(dl);
+
+ /*
+ * Retake all the resources we had at the beginning even if the repair
+ * failed or the scan data are now stale. This keeps things simple for
+ * the caller.
+ */
+ xchk_trans_alloc_empty(sc);
+ xchk_ilock(sc, XFS_ILOCK_EXCL);
+ mutex_lock(&dl->lock);
+
+ if (!error && dl->stale)
+ error = -ESTALE;
+ return error;
+}
+
+/*
+ * Try to fix all the problems. Returns -ESTALE if the scan data have become
+ * out of date.
+ */
+STATIC int
+xrep_dirtree_fix_problems(
+ struct xchk_dirtree *dl,
+ struct xchk_dirtree_outcomes *oc)
+{
+ struct xchk_dirpath *path;
+ int error;
+
+ /* Delete all the paths we don't want. */
+ xchk_dirtree_for_each_path(dl, path) {
+ if (path->outcome != XCHK_DIRPATH_DELETE)
+ continue;
+
+ error = xrep_dirtree_delete_path(dl, path);
+ if (error)
+ return error;
+ }
+
+ /* Reparent this directory to the orphanage. */
+ if (oc->needs_adoption) {
+ if (xrep_orphanage_can_adopt(dl->sc))
+ return xrep_dirtree_move_to_orphanage(dl);
+ return -EFSCORRUPTED;
+ }
+
+ return 0;
+}
+
+/* Fix directory loops involving this directory. */
+int
+xrep_dirtree(
+ struct xfs_scrub *sc)
+{
+ struct xchk_dirtree *dl = sc->buf;
+ struct xchk_dirtree_outcomes oc;
+ int error;
+
+ /*
+ * Prepare to fix the directory tree by retaking the scan lock. The
+ * order of resource acquisition is still IOLOCK -> transaction ->
+ * ILOCK -> scan lock.
+ */
+ mutex_lock(&dl->lock);
+ do {
+ /*
+ * Decide what we're going to do, then do it. An -ESTALE
+ * return here means the scan results are invalid and we have
+ * to walk again.
+ */
+ if (!dl->stale) {
+ xrep_dirtree_decide_fate(dl, &oc);
+
+ trace_xrep_dirtree_decided_fate(dl, &oc);
+
+ error = xrep_dirtree_fix_problems(dl, &oc);
+ if (!error || error != -ESTALE)
+ break;
+ }
+ error = xchk_dirtree_find_paths_to_root(dl);
+ if (error == -ELNRNG || error == -ENOSR)
+ error = -EFSCORRUPTED;
+ } while (!error);
+ mutex_unlock(&dl->lock);
+
+ return error;
+}
diff --git a/fs/xfs/scrub/orphanage.c b/fs/xfs/scrub/orphanage.c
index b894b807155a7..f1711d6031613 100644
--- a/fs/xfs/scrub/orphanage.c
+++ b/fs/xfs/scrub/orphanage.c
@@ -572,6 +572,12 @@ xrep_adoption_move(
xfs_bumplink(sc->tp, sc->orphanage);
xfs_trans_log_inode(sc->tp, sc->orphanage, XFS_ILOG_CORE);
+ /* Bump the link count of the child. */
+ if (adopt->bump_child_nlink) {
+ xfs_bumplink(sc->tp, sc->ip);
+ xfs_trans_log_inode(sc->tp, sc->ip, XFS_ILOG_CORE);
+ }
+
/* Replace the dotdot entry if the child is a subdirectory. */
if (isdir) {
error = xfs_dir_replace(sc->tp, sc->ip, &xfs_name_dotdot,
diff --git a/fs/xfs/scrub/orphanage.h b/fs/xfs/scrub/orphanage.h
index 74ce0bc05c6f1..2ee0e68226618 100644
--- a/fs/xfs/scrub/orphanage.h
+++ b/fs/xfs/scrub/orphanage.h
@@ -60,6 +60,14 @@ struct xrep_adoption {
/* Block reservations for orphanage and child (if directory). */
unsigned int orphanage_blkres;
unsigned int child_blkres;
+
+ /*
+ * Does the caller want us to bump the child link count? This is not
+ * needed when reattaching files that have become disconnected but have
+ * nlink > 1. It is necessary when changing the directory tree
+ * structure.
+ */
+ bool bump_child_nlink:1;
};
bool xrep_orphanage_can_adopt(struct xfs_scrub *sc);
diff --git a/fs/xfs/scrub/repair.h b/fs/xfs/scrub/repair.h
index 7ee5f7e5bffcb..22bce1def9393 100644
--- a/fs/xfs/scrub/repair.h
+++ b/fs/xfs/scrub/repair.h
@@ -95,6 +95,7 @@ int xrep_setup_directory(struct xfs_scrub *sc);
int xrep_setup_parent(struct xfs_scrub *sc);
int xrep_setup_nlinks(struct xfs_scrub *sc);
int xrep_setup_symlink(struct xfs_scrub *sc, unsigned int *resblks);
+int xrep_setup_dirtree(struct xfs_scrub *sc);
/* Repair setup functions */
int xrep_setup_ag_allocbt(struct xfs_scrub *sc);
@@ -132,6 +133,7 @@ int xrep_xattr(struct xfs_scrub *sc);
int xrep_directory(struct xfs_scrub *sc);
int xrep_parent(struct xfs_scrub *sc);
int xrep_symlink(struct xfs_scrub *sc);
+int xrep_dirtree(struct xfs_scrub *sc);
#ifdef CONFIG_XFS_RT
int xrep_rtbitmap(struct xfs_scrub *sc);
@@ -205,6 +207,7 @@ xrep_setup_nothing(
#define xrep_setup_directory xrep_setup_nothing
#define xrep_setup_parent xrep_setup_nothing
#define xrep_setup_nlinks xrep_setup_nothing
+#define xrep_setup_dirtree xrep_setup_nothing
#define xrep_setup_inode(sc, imap) ((void)0)
@@ -239,6 +242,7 @@ static inline int xrep_setup_symlink(struct xfs_scrub *sc, unsigned int *x)
#define xrep_directory xrep_notsupported
#define xrep_parent xrep_notsupported
#define xrep_symlink xrep_notsupported
+#define xrep_dirtree xrep_notsupported
#endif /* CONFIG_XFS_ONLINE_REPAIR */
diff --git a/fs/xfs/scrub/scrub.c b/fs/xfs/scrub/scrub.c
index b6f18d61d8816..4d7116250036f 100644
--- a/fs/xfs/scrub/scrub.c
+++ b/fs/xfs/scrub/scrub.c
@@ -444,7 +444,7 @@ static const struct xchk_meta_ops meta_scrub_ops[] = {
.setup = xchk_setup_dirtree,
.scrub = xchk_dirtree,
.has = xfs_has_parent,
- .repair = xrep_notsupported,
+ .repair = xrep_dirtree,
},
};
diff --git a/fs/xfs/scrub/trace.h b/fs/xfs/scrub/trace.h
index 0d4c1580f61b7..cf209ac9b5329 100644
--- a/fs/xfs/scrub/trace.h
+++ b/fs/xfs/scrub/trace.h
@@ -1694,6 +1694,10 @@ TRACE_DEFINE_ENUM(XCHK_DIRPATH_CORRUPT);
TRACE_DEFINE_ENUM(XCHK_DIRPATH_LOOP);
TRACE_DEFINE_ENUM(XCHK_DIRPATH_STALE);
TRACE_DEFINE_ENUM(XCHK_DIRPATH_OK);
+TRACE_DEFINE_ENUM(XREP_DIRPATH_DELETING);
+TRACE_DEFINE_ENUM(XREP_DIRPATH_DELETED);
+TRACE_DEFINE_ENUM(XREP_DIRPATH_ADOPTING);
+TRACE_DEFINE_ENUM(XREP_DIRPATH_ADOPTED);
#define XCHK_DIRPATH_OUTCOME_STRINGS \
{ XCHK_DIRPATH_SCANNING, "scanning" }, \
@@ -1701,7 +1705,11 @@ TRACE_DEFINE_ENUM(XCHK_DIRPATH_OK);
{ XCHK_DIRPATH_CORRUPT, "corrupt" }, \
{ XCHK_DIRPATH_LOOP, "loop" }, \
{ XCHK_DIRPATH_STALE, "stale" }, \
- { XCHK_DIRPATH_OK, "ok" }
+ { XCHK_DIRPATH_OK, "ok" }, \
+ { XREP_DIRPATH_DELETING, "deleting" }, \
+ { XREP_DIRPATH_DELETED, "deleted" }, \
+ { XREP_DIRPATH_ADOPTING, "adopting" }, \
+ { XREP_DIRPATH_ADOPTED, "adopted" }
DECLARE_EVENT_CLASS(xchk_dirpath_outcome_class,
TP_PROTO(struct xfs_scrub *sc, unsigned long long path_nr,
@@ -1747,6 +1755,7 @@ DECLARE_EVENT_CLASS(xchk_dirtree_evaluate_class,
__field(unsigned int, bad)
__field(unsigned int, suspect)
__field(unsigned int, good)
+ __field(bool, needs_adoption)
),
TP_fast_assign(
__entry->dev = dl->sc->mp->m_super->s_dev;
@@ -1756,15 +1765,17 @@ DECLARE_EVENT_CLASS(xchk_dirtree_evaluate_class,
__entry->bad = oc->bad;
__entry->suspect = oc->suspect;
__entry->good = oc->good;
+ __entry->needs_adoption = oc->needs_adoption ? 1 : 0;
),
- TP_printk("dev %d:%d ino 0x%llx rootino 0x%llx nr_paths %u bad %u suspect %u good %u",
+ TP_printk("dev %d:%d ino 0x%llx rootino 0x%llx nr_paths %u bad %u suspect %u good %u adopt? %d",
MAJOR(__entry->dev), MINOR(__entry->dev),
__entry->ino,
__entry->rootino,
__entry->nr_paths,
__entry->bad,
__entry->suspect,
- __entry->good)
+ __entry->good,
+ __entry->needs_adoption)
);
#define DEFINE_XCHK_DIRTREE_EVALUATE_EVENT(name) \
DEFINE_EVENT(xchk_dirtree_evaluate_class, name, \
@@ -3301,6 +3312,7 @@ DEFINE_REPAIR_DENTRY_EVENT(xrep_adoption_check_child);
DEFINE_REPAIR_DENTRY_EVENT(xrep_adoption_check_alias);
DEFINE_REPAIR_DENTRY_EVENT(xrep_adoption_check_dentry);
DEFINE_REPAIR_DENTRY_EVENT(xrep_adoption_invalidate_child);
+DEFINE_REPAIR_DENTRY_EVENT(xrep_dirtree_delete_child);
TRACE_EVENT(xrep_symlink_salvage_target,
TP_PROTO(struct xfs_inode *ip, char *target, unsigned int targetlen),
@@ -3603,6 +3615,11 @@ TRACE_EVENT(xrep_iunlink_commit_bucket,
__entry->agino)
);
+DEFINE_XCHK_DIRPATH_OUTCOME_EVENT(xrep_dirpath_set_outcome);
+DEFINE_XCHK_DIRTREE_EVENT(xrep_dirtree_delete_path);
+DEFINE_XCHK_DIRTREE_EVENT(xrep_dirtree_create_adoption);
+DEFINE_XCHK_DIRTREE_EVALUATE_EVENT(xrep_dirtree_decided_fate);
+
#endif /* IS_ENABLED(CONFIG_XFS_ONLINE_REPAIR) */
diff --git a/fs/xfs/xfs_inode.c b/fs/xfs/xfs_inode.c
index 998df0d5dac3c..01b100e28b541 100644
--- a/fs/xfs/xfs_inode.c
+++ b/fs/xfs/xfs_inode.c
@@ -914,7 +914,7 @@ xfs_init_new_inode(
* link count to go to zero, move the inode to AGI unlinked list so that it can
* be freed when the last active reference goes away via xfs_inactive().
*/
-static int /* error */
+int
xfs_droplink(
struct xfs_trans *tp,
struct xfs_inode *ip)
diff --git a/fs/xfs/xfs_inode.h b/fs/xfs/xfs_inode.h
index 14c6d9e7ab975..cbde77d711b49 100644
--- a/fs/xfs/xfs_inode.h
+++ b/fs/xfs/xfs_inode.h
@@ -622,6 +622,7 @@ void xfs_end_io(struct work_struct *work);
int xfs_ilock2_io_mmap(struct xfs_inode *ip1, struct xfs_inode *ip2);
void xfs_iunlock2_io_mmap(struct xfs_inode *ip1, struct xfs_inode *ip2);
void xfs_iunlock2_remapping(struct xfs_inode *ip1, struct xfs_inode *ip2);
+int xfs_droplink(struct xfs_trans *tp, struct xfs_inode *ip);
void xfs_bumplink(struct xfs_trans *tp, struct xfs_inode *ip);
void xfs_lock_inodes(struct xfs_inode **ips, int inodes, uint lock_mode);
void xfs_sort_inodes(struct xfs_inode **i_tab, unsigned int num_inodes);
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 1/3] xfs: reduce the rate of cond_resched calls inside scrub
2023-12-31 19:34 ` [PATCHSET v13.0 7/7] xfs: vectorize scrub kernel calls Darrick J. Wong
@ 2023-12-31 21:00 ` Darrick J. Wong
2023-12-31 21:00 ` [PATCH 2/3] xfs: introduce vectored scrub mode Darrick J. Wong
2023-12-31 21:00 ` [PATCH 3/3] xfs: only iget the file once when doing vectored scrub-by-handle Darrick J. Wong
2 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 21:00 UTC (permalink / raw)
To: djwong; +Cc: linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
We really don't want to call cond_resched every single time we go
through a loop in scrub -- there may be billions of records, and probing
into the scheduler itself has overhead. Reduce this overhead by only
calling cond_resched 10x per second; and add a counter so that we only
check jiffies once every 1000 records or so.
Surprisingly, this reduces scrub-only fstests runtime by about 2%. I
used the bmapinflate xfs_db command to produce a billion-extent file and
this stupid gadget reduced the scrub runtime by about 4%.
From a stupid microbenchmark of calling these things 1 billion times, I
estimate that cond_resched costs about 5.5ns per call; jiffes costs
about 0.3ns per read; and fatal_signal_pending costs about 0.4ns per
call.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/scrub/common.h | 25 -------------------
fs/xfs/scrub/scrub.c | 1 +
fs/xfs/scrub/scrub.h | 64 ++++++++++++++++++++++++++++++++++++++++++++++++
fs/xfs/scrub/xfarray.c | 10 ++++----
fs/xfs/scrub/xfarray.h | 3 ++
fs/xfs/scrub/xfile.c | 10 +++-----
6 files changed, 77 insertions(+), 36 deletions(-)
diff --git a/fs/xfs/scrub/common.h b/fs/xfs/scrub/common.h
index c3a580a5d7c7d..a90c82c18e3c9 100644
--- a/fs/xfs/scrub/common.h
+++ b/fs/xfs/scrub/common.h
@@ -6,31 +6,6 @@
#ifndef __XFS_SCRUB_COMMON_H__
#define __XFS_SCRUB_COMMON_H__
-/*
- * We /could/ terminate a scrub/repair operation early. If we're not
- * in a good place to continue (fatal signal, etc.) then bail out.
- * Note that we're careful not to make any judgements about *error.
- */
-static inline bool
-xchk_should_terminate(
- struct xfs_scrub *sc,
- int *error)
-{
- /*
- * If preemption is disabled, we need to yield to the scheduler every
- * few seconds so that we don't run afoul of the soft lockup watchdog
- * or RCU stall detector.
- */
- cond_resched();
-
- if (fatal_signal_pending(current)) {
- if (*error == 0)
- *error = -EINTR;
- return true;
- }
- return false;
-}
-
int xchk_trans_alloc(struct xfs_scrub *sc, uint resblks);
int xchk_trans_alloc_empty(struct xfs_scrub *sc);
void xchk_trans_cancel(struct xfs_scrub *sc);
diff --git a/fs/xfs/scrub/scrub.c b/fs/xfs/scrub/scrub.c
index 4d7116250036f..8874c28c2e7a8 100644
--- a/fs/xfs/scrub/scrub.c
+++ b/fs/xfs/scrub/scrub.c
@@ -623,6 +623,7 @@ xfs_scrub_metadata(
sc->sm = sm;
sc->ops = &meta_scrub_ops[sm->sm_type];
sc->sick_mask = xchk_health_mask_for_scrub_type(sm->sm_type);
+ sc->relax = INIT_XCHK_RELAX;
retry_op:
/*
* When repairs are allowed, prevent freezing or readonly remount while
diff --git a/fs/xfs/scrub/scrub.h b/fs/xfs/scrub/scrub.h
index ec685ef425778..ecd7aff7a48bf 100644
--- a/fs/xfs/scrub/scrub.h
+++ b/fs/xfs/scrub/scrub.h
@@ -8,6 +8,49 @@
struct xfs_scrub;
+struct xchk_relax {
+ unsigned long next_resched;
+ unsigned int resched_nr;
+ bool interruptible;
+};
+
+/* Yield to the scheduler at most 10x per second. */
+#define XCHK_RELAX_NEXT (jiffies + (HZ / 10))
+
+#define INIT_XCHK_RELAX \
+ (struct xchk_relax){ \
+ .next_resched = XCHK_RELAX_NEXT, \
+ .resched_nr = 0, \
+ .interruptible = true, \
+ }
+
+/*
+ * Relax during a scrub operation and exit if there's a fatal signal pending.
+ *
+ * If preemption is disabled, we need to yield to the scheduler every now and
+ * then so that we don't run afoul of the soft lockup watchdog or RCU stall
+ * detector. cond_resched calls are somewhat expensive (~5ns) so we want to
+ * ratelimit this to 10x per second. Amortize the cost of the other checks by
+ * only doing it once every 100 calls.
+ */
+static inline int xchk_maybe_relax(struct xchk_relax *widget)
+{
+ /* Amortize the cost of scheduling and checking signals. */
+ if (likely(++widget->resched_nr < 100))
+ return 0;
+ widget->resched_nr = 0;
+
+ if (unlikely(widget->next_resched <= jiffies)) {
+ cond_resched();
+ widget->next_resched = XCHK_RELAX_NEXT;
+ }
+
+ if (widget->interruptible && fatal_signal_pending(current))
+ return -EINTR;
+
+ return 0;
+}
+
/*
* Standard flags for allocating memory within scrub. NOFS context is
* configured by the process allocation scope. Scrub and repair must be able
@@ -123,6 +166,9 @@ struct xfs_scrub {
*/
unsigned int sick_mask;
+ /* next time we want to cond_resched() */
+ struct xchk_relax relax;
+
/* State tracking for single-AG operations. */
struct xchk_ag sa;
};
@@ -167,6 +213,24 @@ struct xfs_scrub_subord *xchk_scrub_create_subord(struct xfs_scrub *sc,
unsigned int subtype);
void xchk_scrub_free_subord(struct xfs_scrub_subord *sub);
+/*
+ * We /could/ terminate a scrub/repair operation early. If we're not
+ * in a good place to continue (fatal signal, etc.) then bail out.
+ * Note that we're careful not to make any judgements about *error.
+ */
+static inline bool
+xchk_should_terminate(
+ struct xfs_scrub *sc,
+ int *error)
+{
+ if (xchk_maybe_relax(&sc->relax)) {
+ if (*error == 0)
+ *error = -EINTR;
+ return true;
+ }
+ return false;
+}
+
/* Metadata scrubbers */
int xchk_tester(struct xfs_scrub *sc);
int xchk_superblock(struct xfs_scrub *sc);
diff --git a/fs/xfs/scrub/xfarray.c b/fs/xfs/scrub/xfarray.c
index 1bd690ac5d368..407e81a266a1c 100644
--- a/fs/xfs/scrub/xfarray.c
+++ b/fs/xfs/scrub/xfarray.c
@@ -7,9 +7,9 @@
#include "xfs_fs.h"
#include "xfs_shared.h"
#include "xfs_format.h"
+#include "scrub/scrub.h"
#include "scrub/xfile.h"
#include "scrub/xfarray.h"
-#include "scrub/scrub.h"
#include "scrub/trace.h"
/*
@@ -486,6 +486,9 @@ xfarray_sortinfo_alloc(
xfarray_sortinfo_lo(si)[0] = 0;
xfarray_sortinfo_hi(si)[0] = array->nr - 1;
+ si->relax = INIT_XCHK_RELAX;
+ if (flags & XFARRAY_SORT_KILLABLE)
+ si->relax.interruptible = false;
trace_xfarray_sort(si, nr_bytes);
*infop = si;
@@ -503,10 +506,7 @@ xfarray_sort_terminated(
* few seconds so that we don't run afoul of the soft lockup watchdog
* or RCU stall detector.
*/
- cond_resched();
-
- if ((si->flags & XFARRAY_SORT_KILLABLE) &&
- fatal_signal_pending(current)) {
+ if (xchk_maybe_relax(&si->relax)) {
if (*error == 0)
*error = -EINTR;
return true;
diff --git a/fs/xfs/scrub/xfarray.h b/fs/xfs/scrub/xfarray.h
index 6652070716095..2794dcc4cf955 100644
--- a/fs/xfs/scrub/xfarray.h
+++ b/fs/xfs/scrub/xfarray.h
@@ -127,6 +127,9 @@ struct xfarray_sortinfo {
/* XFARRAY_SORT_* flags; see below. */
unsigned int flags;
+ /* next time we want to cond_resched() */
+ struct xchk_relax relax;
+
/* Cache a page here for faster access. */
struct xfile_page xfpage;
void *page_kaddr;
diff --git a/fs/xfs/scrub/xfile.c b/fs/xfs/scrub/xfile.c
index ccef7fdcd7d9f..ffd35bce04d96 100644
--- a/fs/xfs/scrub/xfile.c
+++ b/fs/xfs/scrub/xfile.c
@@ -10,9 +10,9 @@
#include "xfs_log_format.h"
#include "xfs_trans_resv.h"
#include "xfs_mount.h"
+#include "scrub/scrub.h"
#include "scrub/xfile.h"
#include "scrub/xfarray.h"
-#include "scrub/scrub.h"
#include "scrub/trace.h"
#include <linux/shmem_fs.h>
@@ -607,6 +607,7 @@ xfile_dump(
loff_t holepos = 0;
loff_t datapos;
loff_t ret;
+ struct xchk_relax relax = INIT_XCHK_RELAX;
unsigned int pflags;
bool all_zeroes = true;
int error = 0;
@@ -633,12 +634,9 @@ xfile_dump(
unsigned int pagepos;
unsigned int pagelen;
- cond_resched();
-
- if (fatal_signal_pending(current)) {
- error = -EINTR;
+ error = xchk_maybe_relax(&relax);
+ if (error)
goto out_pflags;
- }
pagelen = min_t(u64, datalen, PAGE_SIZE);
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 2/3] xfs: introduce vectored scrub mode
2023-12-31 19:34 ` [PATCHSET v13.0 7/7] xfs: vectorize scrub kernel calls Darrick J. Wong
2023-12-31 21:00 ` [PATCH 1/3] xfs: reduce the rate of cond_resched calls inside scrub Darrick J. Wong
@ 2023-12-31 21:00 ` Darrick J. Wong
2023-12-31 21:00 ` [PATCH 3/3] xfs: only iget the file once when doing vectored scrub-by-handle Darrick J. Wong
2 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 21:00 UTC (permalink / raw)
To: djwong; +Cc: linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
Introduce a variant on XFS_SCRUB_METADATA that allows for a vectored
mode. The caller specifies the principal metadata object that they want
to scrub (allocation group, inode, etc.) once, followed by an array of
scrub types they want called on that object. The kernel runs the scrub
operations and writes the output flags and errno code to the
corresponding array element.
A new pseudo scrub type BARRIER is introduced to force the kernel to
return to userspace if any corruptions have been found when scrubbing
the previous scrub types in the array. This enables userspace to
schedule, for example, the sequence:
1. data fork
2. barrier
3. directory
If the data fork scrub is clean, then the kernel will perform the
directory scrub. If not, the barrier in 2 will exit back to userspace.
When running fstests in "rebuild all metadata after each test" mode, I
observed a 10% reduction in runtime due to fewer transitions across the
system call boundary.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/libxfs/xfs_fs.h | 10 ++++
fs/xfs/libxfs/xfs_fs_staging.h | 32 ++++++++++++
fs/xfs/scrub/scrub.c | 106 ++++++++++++++++++++++++++++++++++++++++
fs/xfs/scrub/trace.h | 78 +++++++++++++++++++++++++++++
fs/xfs/scrub/xfs_scrub.h | 2 +
fs/xfs/xfs_ioctl.c | 50 +++++++++++++++++++
6 files changed, 277 insertions(+), 1 deletion(-)
diff --git a/fs/xfs/libxfs/xfs_fs.h b/fs/xfs/libxfs/xfs_fs.h
index 2499a20f5f774..77fbca573e164 100644
--- a/fs/xfs/libxfs/xfs_fs.h
+++ b/fs/xfs/libxfs/xfs_fs.h
@@ -725,6 +725,15 @@ struct xfs_scrub_metadata {
/* Number of scrub subcommands. */
#define XFS_SCRUB_TYPE_NR 29
+/*
+ * This special type code only applies to the vectored scrub implementation.
+ *
+ * If any of the previous scrub vectors recorded runtime errors or have
+ * sv_flags bits set that match the OFLAG bits in the barrier vector's
+ * sv_flags, set the barrier's sv_ret to -ECANCELED and return to userspace.
+ */
+#define XFS_SCRUB_TYPE_BARRIER (-1U)
+
/* i: Repair this metadata. */
#define XFS_SCRUB_IFLAG_REPAIR (1u << 0)
@@ -813,6 +822,7 @@ struct xfs_scrub_metadata {
#define XFS_IOC_FREE_EOFBLOCKS _IOR ('X', 58, struct xfs_fs_eofblocks)
/* XFS_IOC_GETFSMAP ------ hoisted 59 */
#define XFS_IOC_SCRUB_METADATA _IOWR('X', 60, struct xfs_scrub_metadata)
+/* XFS_IOC_SCRUBV_METADATA -- staging 60 */
#define XFS_IOC_AG_GEOMETRY _IOWR('X', 61, struct xfs_ag_geometry)
/* XFS_IOC_GETPARENTS ---- staging 62 */
diff --git a/fs/xfs/libxfs/xfs_fs_staging.h b/fs/xfs/libxfs/xfs_fs_staging.h
index e0650af055895..69d29f213af3a 100644
--- a/fs/xfs/libxfs/xfs_fs_staging.h
+++ b/fs/xfs/libxfs/xfs_fs_staging.h
@@ -170,4 +170,36 @@ xfs_getparents_rec(
#define XFS_IOC_GETPARENTS _IOWR('X', 62, struct xfs_getparents)
+/* Vectored scrub calls to reduce the number of kernel transitions. */
+
+struct xfs_scrub_vec {
+ __u32 sv_type; /* XFS_SCRUB_TYPE_* */
+ __u32 sv_flags; /* XFS_SCRUB_FLAGS_* */
+ __s32 sv_ret; /* 0 or a negative error code */
+ __u32 sv_reserved; /* must be zero */
+};
+
+/* Vectored metadata scrub control structure. */
+struct xfs_scrub_vec_head {
+ __u64 svh_ino; /* inode number. */
+ __u32 svh_gen; /* inode generation. */
+ __u32 svh_agno; /* ag number. */
+ __u32 svh_flags; /* XFS_SCRUB_VEC_FLAGS_* */
+ __u16 svh_rest_us; /* wait this much time between vector items */
+ __u16 svh_nr; /* number of svh_vecs */
+ __u64 svh_reserved; /* must be zero */
+
+ struct xfs_scrub_vec svh_vecs[];
+};
+
+#define XFS_SCRUB_VEC_FLAGS_ALL (0)
+
+static inline size_t sizeof_xfs_scrub_vec(unsigned int nr)
+{
+ return sizeof(struct xfs_scrub_vec_head) +
+ nr * sizeof(struct xfs_scrub_vec);
+}
+
+#define XFS_IOC_SCRUBV_METADATA _IOWR('X', 60, struct xfs_scrub_vec_head)
+
#endif /* __XFS_FS_STAGING_H__ */
diff --git a/fs/xfs/scrub/scrub.c b/fs/xfs/scrub/scrub.c
index 8874c28c2e7a8..1a0018537b054 100644
--- a/fs/xfs/scrub/scrub.c
+++ b/fs/xfs/scrub/scrub.c
@@ -21,6 +21,7 @@
#include "xfs_swapext.h"
#include "xfs_dir2.h"
#include "xfs_parent.h"
+#include "xfs_icache.h"
#include "scrub/scrub.h"
#include "scrub/common.h"
#include "scrub/trace.h"
@@ -727,3 +728,108 @@ xfs_scrub_metadata(
run.retries++;
goto retry_op;
}
+
+/* Decide if there have been any scrub failures up to this point. */
+static inline bool
+xfs_scrubv_previous_failures(
+ struct xfs_mount *mp,
+ struct xfs_scrub_vec_head *vhead,
+ struct xfs_scrub_vec *barrier_vec)
+{
+ struct xfs_scrub_vec *v;
+ __u32 failmask;
+
+ failmask = barrier_vec->sv_flags & XFS_SCRUB_FLAGS_OUT;
+
+ for (v = vhead->svh_vecs; v < barrier_vec; v++) {
+ if (v->sv_type == XFS_SCRUB_TYPE_BARRIER)
+ continue;
+
+ /*
+ * Runtime errors count as a previous failure, except the ones
+ * used to ask userspace to retry.
+ */
+ if (v->sv_ret && v->sv_ret != -EBUSY && v->sv_ret != -ENOENT &&
+ v->sv_ret != -EUSERS)
+ return true;
+
+ /*
+ * If any of the out-flags on the scrub vector match the mask
+ * that was set on the barrier vector, that's a previous fail.
+ */
+ if (v->sv_flags & failmask)
+ return true;
+ }
+
+ return false;
+}
+
+/* Vectored scrub implementation to reduce ioctl calls. */
+int
+xfs_scrubv_metadata(
+ struct file *file,
+ struct xfs_scrub_vec_head *vhead)
+{
+ struct xfs_inode *ip_in = XFS_I(file_inode(file));
+ struct xfs_mount *mp = ip_in->i_mount;
+ struct xfs_scrub_vec *v;
+ unsigned int i;
+ int error = 0;
+
+ BUILD_BUG_ON(sizeof(struct xfs_scrub_vec_head) ==
+ sizeof(struct xfs_scrub_metadata));
+ BUILD_BUG_ON(XFS_IOC_SCRUB_METADATA == XFS_IOC_SCRUBV_METADATA);
+
+ trace_xchk_scrubv_start(ip_in, vhead);
+
+ if (vhead->svh_flags & ~XFS_SCRUB_VEC_FLAGS_ALL)
+ return -EINVAL;
+ for (i = 0, v = vhead->svh_vecs; i < vhead->svh_nr; i++, v++) {
+ if (v->sv_reserved)
+ return -EINVAL;
+ if (v->sv_type == XFS_SCRUB_TYPE_BARRIER &&
+ (v->sv_flags & ~XFS_SCRUB_FLAGS_OUT))
+ return -EINVAL;
+
+ trace_xchk_scrubv_item(mp, vhead, v);
+ }
+
+ /* Run all the scrubbers. */
+ for (i = 0, v = vhead->svh_vecs; i < vhead->svh_nr; i++, v++) {
+ struct xfs_scrub_metadata sm = {
+ .sm_type = v->sv_type,
+ .sm_flags = v->sv_flags,
+ .sm_ino = vhead->svh_ino,
+ .sm_gen = vhead->svh_gen,
+ .sm_agno = vhead->svh_agno,
+ };
+
+ if (v->sv_type == XFS_SCRUB_TYPE_BARRIER) {
+ if (xfs_scrubv_previous_failures(mp, vhead, v)) {
+ v->sv_ret = -ECANCELED;
+ trace_xchk_scrubv_barrier_fail(mp, vhead, v);
+ break;
+ }
+
+ continue;
+ }
+
+ v->sv_ret = xfs_scrub_metadata(file, &sm);
+ v->sv_flags = sm.sm_flags;
+
+ if (vhead->svh_rest_us) {
+ ktime_t expires;
+
+ expires = ktime_add_ns(ktime_get(),
+ vhead->svh_rest_us * 1000);
+ set_current_state(TASK_KILLABLE);
+ schedule_hrtimeout(&expires, HRTIMER_MODE_ABS);
+ }
+ if (fatal_signal_pending(current)) {
+ error = -EINTR;
+ break;
+ }
+ }
+
+ return error;
+}
diff --git a/fs/xfs/scrub/trace.h b/fs/xfs/scrub/trace.h
index cf209ac9b5329..cc2af405cd3a7 100644
--- a/fs/xfs/scrub/trace.h
+++ b/fs/xfs/scrub/trace.h
@@ -80,6 +80,7 @@ TRACE_DEFINE_ENUM(XFS_SCRUB_TYPE_QUOTACHECK);
TRACE_DEFINE_ENUM(XFS_SCRUB_TYPE_NLINKS);
TRACE_DEFINE_ENUM(XFS_SCRUB_TYPE_HEALTHY);
TRACE_DEFINE_ENUM(XFS_SCRUB_TYPE_DIRTREE);
+TRACE_DEFINE_ENUM(XFS_SCRUB_TYPE_BARRIER);
#define XFS_SCRUB_TYPE_STRINGS \
{ XFS_SCRUB_TYPE_PROBE, "probe" }, \
@@ -110,7 +111,8 @@ TRACE_DEFINE_ENUM(XFS_SCRUB_TYPE_DIRTREE);
{ XFS_SCRUB_TYPE_QUOTACHECK, "quotacheck" }, \
{ XFS_SCRUB_TYPE_NLINKS, "nlinks" }, \
{ XFS_SCRUB_TYPE_HEALTHY, "healthy" }, \
- { XFS_SCRUB_TYPE_DIRTREE, "dirtree" }
+ { XFS_SCRUB_TYPE_DIRTREE, "dirtree" }, \
+ { XFS_SCRUB_TYPE_BARRIER, "barrier" }
#define XFS_SCRUB_FLAG_STRINGS \
{ XFS_SCRUB_IFLAG_REPAIR, "repair" }, \
@@ -210,6 +212,80 @@ DEFINE_EVENT(xchk_fsgate_class, name, \
DEFINE_SCRUB_FSHOOK_EVENT(xchk_fsgates_enable);
DEFINE_SCRUB_FSHOOK_EVENT(xchk_fsgates_disable);
+DECLARE_EVENT_CLASS(xchk_vector_head_class,
+ TP_PROTO(struct xfs_inode *ip, struct xfs_scrub_vec_head *vhead),
+ TP_ARGS(ip, vhead),
+ TP_STRUCT__entry(
+ __field(dev_t, dev)
+ __field(xfs_ino_t, ino)
+ __field(xfs_agnumber_t, agno)
+ __field(xfs_ino_t, inum)
+ __field(unsigned int, gen)
+ __field(unsigned int, flags)
+ __field(unsigned short, rest_us)
+ __field(unsigned short, nr_vecs)
+ ),
+ TP_fast_assign(
+ __entry->dev = ip->i_mount->m_super->s_dev;
+ __entry->ino = ip->i_ino;
+ __entry->agno = vhead->svh_agno;
+ __entry->inum = vhead->svh_ino;
+ __entry->gen = vhead->svh_gen;
+ __entry->flags = vhead->svh_flags;
+ __entry->rest_us = vhead->svh_rest_us;
+ __entry->nr_vecs = vhead->svh_nr;
+ ),
+ TP_printk("dev %d:%d ino 0x%llx agno 0x%x inum 0x%llx gen 0x%x flags 0x%x rest_us %u nr_vecs %u",
+ MAJOR(__entry->dev), MINOR(__entry->dev),
+ __entry->ino,
+ __entry->agno,
+ __entry->inum,
+ __entry->gen,
+ __entry->flags,
+ __entry->rest_us,
+ __entry->nr_vecs)
+)
+#define DEFINE_SCRUBV_HEAD_EVENT(name) \
+DEFINE_EVENT(xchk_vector_head_class, name, \
+ TP_PROTO(struct xfs_inode *ip, struct xfs_scrub_vec_head *vhead), \
+ TP_ARGS(ip, vhead))
+
+DEFINE_SCRUBV_HEAD_EVENT(xchk_scrubv_start);
+
+DECLARE_EVENT_CLASS(xchk_vector_class,
+ TP_PROTO(struct xfs_mount *mp, struct xfs_scrub_vec_head *vhead,
+ struct xfs_scrub_vec *v),
+ TP_ARGS(mp, vhead, v),
+ TP_STRUCT__entry(
+ __field(dev_t, dev)
+ __field(unsigned int, vec_nr)
+ __field(unsigned int, vec_type)
+ __field(unsigned int, vec_flags)
+ __field(int, vec_ret)
+ ),
+ TP_fast_assign(
+ __entry->dev = mp->m_super->s_dev;
+ __entry->vec_nr = v - vhead->svh_vecs;
+ __entry->vec_type = v->sv_type;
+ __entry->vec_flags = v->sv_flags;
+ __entry->vec_ret = v->sv_ret;
+ ),
+ TP_printk("dev %d:%d vec[%u] type %s flags %s ret %d",
+ MAJOR(__entry->dev), MINOR(__entry->dev),
+ __entry->vec_nr,
+ __print_symbolic(__entry->vec_type, XFS_SCRUB_TYPE_STRINGS),
+ __print_flags(__entry->vec_flags, "|", XFS_SCRUB_FLAG_STRINGS),
+ __entry->vec_ret)
+)
+#define DEFINE_SCRUBV_EVENT(name) \
+DEFINE_EVENT(xchk_vector_class, name, \
+ TP_PROTO(struct xfs_mount *mp, struct xfs_scrub_vec_head *vhead, \
+ struct xfs_scrub_vec *v), \
+ TP_ARGS(mp, vhead, v))
+
+DEFINE_SCRUBV_EVENT(xchk_scrubv_barrier_fail);
+DEFINE_SCRUBV_EVENT(xchk_scrubv_item);
+
TRACE_EVENT(xchk_op_error,
TP_PROTO(struct xfs_scrub *sc, xfs_agnumber_t agno,
xfs_agblock_t bno, int error, void *ret_ip),
diff --git a/fs/xfs/scrub/xfs_scrub.h b/fs/xfs/scrub/xfs_scrub.h
index a39befa743ce0..61d010f19f003 100644
--- a/fs/xfs/scrub/xfs_scrub.h
+++ b/fs/xfs/scrub/xfs_scrub.h
@@ -8,8 +8,10 @@
#ifndef CONFIG_XFS_ONLINE_SCRUB
# define xfs_scrub_metadata(file, sm) (-ENOTTY)
+# define xfs_scrubv_metadata(file, vhead) (-ENOTTY)
#else
int xfs_scrub_metadata(struct file *file, struct xfs_scrub_metadata *sm);
+int xfs_scrubv_metadata(struct file *file, struct xfs_scrub_vec_head *vhead);
#endif /* CONFIG_XFS_ONLINE_SCRUB */
#endif /* __XFS_SCRUB_H__ */
diff --git a/fs/xfs/xfs_ioctl.c b/fs/xfs/xfs_ioctl.c
index 5db70a11151dd..a0dfefdf4c491 100644
--- a/fs/xfs/xfs_ioctl.c
+++ b/fs/xfs/xfs_ioctl.c
@@ -1936,6 +1936,54 @@ xfs_ioc_setlabel(
return error;
}
+STATIC int
+xfs_ioc_scrubv_metadata(
+ struct file *filp,
+ void __user *arg)
+{
+ struct xfs_scrub_vec_head __user *uhead = arg;
+ struct xfs_scrub_vec_head head;
+ struct xfs_scrub_vec_head *vhead;
+ size_t bytes;
+ int error;
+
+ if (!capable(CAP_SYS_ADMIN))
+ return -EPERM;
+
+ if (copy_from_user(&head, uhead, sizeof(head)))
+ return -EFAULT;
+
+ if (head.svh_reserved)
+ return -EINVAL;
+
+ bytes = sizeof_xfs_scrub_vec(head.svh_nr);
+ if (bytes > PAGE_SIZE)
+ return -ENOMEM;
+ vhead = kvmalloc(bytes, GFP_KERNEL | __GFP_RETRY_MAYFAIL);
+ if (!vhead)
+ return -ENOMEM;
+ memcpy(vhead, &head, sizeof(struct xfs_scrub_vec_head));
+
+ if (copy_from_user(&vhead->svh_vecs, &uhead->svh_vecs,
+ head.svh_nr * sizeof(struct xfs_scrub_vec))) {
+ error = -EFAULT;
+ goto err_free;
+ }
+
+ error = xfs_scrubv_metadata(filp, vhead);
+ if (error)
+ goto err_free;
+
+ if (copy_to_user(uhead, vhead, bytes)) {
+ error = -EFAULT;
+ goto err_free;
+ }
+
+err_free:
+ kvfree(vhead);
+ return error;
+}
+
static inline int
xfs_fs_eofblocks_from_user(
struct xfs_fs_eofblocks *src,
@@ -2099,6 +2147,8 @@ xfs_file_ioctl(
case FS_IOC_GETFSMAP:
return xfs_ioc_getfsmap(ip, arg);
+ case XFS_IOC_SCRUBV_METADATA:
+ return xfs_ioc_scrubv_metadata(filp, arg);
case XFS_IOC_SCRUB_METADATA:
return xfs_ioc_scrub_metadata(filp, arg);
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 3/3] xfs: only iget the file once when doing vectored scrub-by-handle
2023-12-31 19:34 ` [PATCHSET v13.0 7/7] xfs: vectorize scrub kernel calls Darrick J. Wong
2023-12-31 21:00 ` [PATCH 1/3] xfs: reduce the rate of cond_resched calls inside scrub Darrick J. Wong
2023-12-31 21:00 ` [PATCH 2/3] xfs: introduce vectored scrub mode Darrick J. Wong
@ 2023-12-31 21:00 ` Darrick J. Wong
2 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 21:00 UTC (permalink / raw)
To: djwong; +Cc: linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
If a program wants us to perform a scrub on a file handle and the fd
passed to ioctl() is not the file referenced in the handle, iget the
file once and pass it into the scrub code. This amortizes the untrusted
iget lookup over /all/ the scrubbers mentioned in the scrubv call.
When running fstests in "rebuild all metadata after each test" mode, I
observed a 10% reduction in runtime on account of avoiding repeated
inobt lookups.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
fs/xfs/scrub/scrub.c | 61 ++++++++++++++++++++++++++++++++++++++++++++++++++
1 file changed, 61 insertions(+)
diff --git a/fs/xfs/scrub/scrub.c b/fs/xfs/scrub/scrub.c
index 1a0018537b054..7b70dfb30287b 100644
--- a/fs/xfs/scrub/scrub.c
+++ b/fs/xfs/scrub/scrub.c
@@ -764,6 +764,31 @@ xfs_scrubv_previous_failures(
return false;
}
+/*
+ * If the caller provided us with a nonzero inode number that isn't the ioctl
+ * file, try to grab a reference to it to eliminate all further untrusted inode
+ * lookups. If we can't get the inode, let each scrub function try again.
+ */
+STATIC struct xfs_inode *
+xchk_scrubv_open_by_handle(
+ struct xfs_mount *mp,
+ const struct xfs_scrub_vec_head *vhead)
+{
+ struct xfs_inode *ip;
+ int error;
+
+ error = xfs_iget(mp, NULL, vhead->svh_ino, XFS_IGET_UNTRUSTED, 0, &ip);
+ if (error)
+ return NULL;
+
+ if (VFS_I(ip)->i_generation != vhead->svh_gen) {
+ xfs_irele(ip);
+ return NULL;
+ }
+
+ return ip;
+}
+
/* Vectored scrub implementation to reduce ioctl calls. */
int
xfs_scrubv_metadata(
@@ -772,7 +797,9 @@ xfs_scrubv_metadata(
{
struct xfs_inode *ip_in = XFS_I(file_inode(file));
struct xfs_mount *mp = ip_in->i_mount;
+ struct xfs_inode *handle_ip = NULL;
struct xfs_scrub_vec *v;
+ bool set_dontcache = false;
unsigned int i;
int error = 0;
@@ -791,9 +818,28 @@ xfs_scrubv_metadata(
(v->sv_flags & ~XFS_SCRUB_FLAGS_OUT))
return -EINVAL;
+ /*
+ * If we detect at least one inode-type scrub, we might
+ * consider setting dontcache at the end.
+ */
+ if (v->sv_type < XFS_SCRUB_TYPE_NR &&
+ meta_scrub_ops[v->sv_type].type == ST_INODE)
+ set_dontcache = true;
+
trace_xchk_scrubv_item(mp, vhead, v);
}
+ /*
+ * If the caller wants us to do a scrub-by-handle and the file used to
+ * call the ioctl is not the same file, load the incore inode and pin
+ * it across all the scrubv actions to avoid repeated UNTRUSTED
+ * lookups. The reference is not passed to deeper layers of scrub
+ * because each scrubber gets to decide its own strategy for getting an
+ * inode.
+ */
+ if (vhead->svh_ino && vhead->svh_ino != ip_in->i_ino)
+ handle_ip = xchk_scrubv_open_by_handle(mp, vhead);
+
/* Run all the scrubbers. */
for (i = 0, v = vhead->svh_vecs; i < vhead->svh_nr; i++, v++) {
struct xfs_scrub_metadata sm = {
@@ -817,6 +863,10 @@ xfs_scrubv_metadata(
v->sv_ret = xfs_scrub_metadata(file, &sm);
v->sv_flags = sm.sm_flags;
+ /* Leave the inode in memory if something's wrong with it. */
+ if (xchk_needs_repair(&sm))
+ set_dontcache = false;
+
if (vhead->svh_rest_us) {
ktime_t expires;
@@ -831,5 +881,16 @@ xfs_scrubv_metadata(
}
}
+ /*
+ * If we're holding the only reference to an inode opened via handle
+ * and the scan was clean, mark it dontcache so that we don't pollute
+ * the cache.
+ */
+ if (handle_ip) {
+ if (set_dontcache &&
+ atomic_read(&VFS_I(handle_ip)->i_count) == 1)
+ d_mark_dontcache(VFS_I(handle_ip));
+ xfs_irele(handle_ip);
+ }
return error;
}
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 1/1] xfs: Increase XFS_DEFER_OPS_NR_INODES to 5
2023-12-31 19:50 ` [PATCHSET v13.0 1/6] xfsprogs: retain ILOCK during directory updates Darrick J. Wong
@ 2023-12-31 23:02 ` Darrick J. Wong
0 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:02 UTC (permalink / raw)
To: djwong, cem
Cc: Allison Henderson, Catherine Hoang, catherine.hoang, linux-xfs,
allison.henderson
From: Allison Henderson <allison.henderson@oracle.com>
Renames that generate parent pointer updates can join up to 5
inodes locked in sorted order. So we need to increase the
number of defer ops inodes and relock them in the same way.
Signed-off-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Catherine Hoang <catherine.hoang@oracle.com>
[djwong: have one sorting function]
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
libxfs/libxfs_priv.h | 2 ++
libxfs/xfs_defer.c | 6 +++++-
libxfs/xfs_defer.h | 8 +++++++-
3 files changed, 14 insertions(+), 2 deletions(-)
diff --git a/libxfs/libxfs_priv.h b/libxfs/libxfs_priv.h
index ef29d7e5eb7..45cfe4408a9 100644
--- a/libxfs/libxfs_priv.h
+++ b/libxfs/libxfs_priv.h
@@ -437,6 +437,8 @@ void __xfs_buf_mark_corrupt(struct xfs_buf *bp, xfs_failaddr_t fa);
__mode = __mode; /* no set-but-unused warning */ \
})
#define xfs_lock_two_inodes(ip0,mode0,ip1,mode1) ((void) 0)
+#define xfs_lock_inodes(i_tab, nr, mode) ((void) 0)
+#define xfs_sort_inodes(i_tab, nr) ((void) 0)
/* space allocation */
#define XFS_EXTENT_BUSY_DISCARDED 0x01 /* undergoing a discard op. */
diff --git a/libxfs/xfs_defer.c b/libxfs/xfs_defer.c
index 7782eea458e..41e607d55f0 100644
--- a/libxfs/xfs_defer.c
+++ b/libxfs/xfs_defer.c
@@ -1090,7 +1090,11 @@ xfs_defer_ops_continue(
ASSERT(!(tp->t_flags & XFS_TRANS_DIRTY));
/* Lock the captured resources to the new transaction. */
- if (dfc->dfc_held.dr_inos == 2)
+ if (dfc->dfc_held.dr_inos > 2) {
+ xfs_sort_inodes(dfc->dfc_held.dr_ip, dfc->dfc_held.dr_inos);
+ xfs_lock_inodes(dfc->dfc_held.dr_ip, dfc->dfc_held.dr_inos,
+ XFS_ILOCK_EXCL);
+ } else if (dfc->dfc_held.dr_inos == 2)
xfs_lock_two_inodes(dfc->dfc_held.dr_ip[0], XFS_ILOCK_EXCL,
dfc->dfc_held.dr_ip[1], XFS_ILOCK_EXCL);
else if (dfc->dfc_held.dr_inos == 1)
diff --git a/libxfs/xfs_defer.h b/libxfs/xfs_defer.h
index e3cf81bafca..c9a1fe3fe36 100644
--- a/libxfs/xfs_defer.h
+++ b/libxfs/xfs_defer.h
@@ -77,7 +77,13 @@ extern const struct xfs_defer_op_type xfs_swapext_defer_type;
/*
* Deferred operation item relogging limits.
*/
-#define XFS_DEFER_OPS_NR_INODES 2 /* join up to two inodes */
+
+/*
+ * Rename w/ parent pointers can require up to 5 inodes with deferred ops to
+ * be joined to the transaction: src_dp, target_dp, src_ip, target_ip, and wip.
+ * These inodes are locked in sorted order by their inode numbers
+ */
+#define XFS_DEFER_OPS_NR_INODES 5
#define XFS_DEFER_OPS_NR_BUFS 2 /* join up to two buffers */
/* Resources that must be held across a transaction roll. */
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 01/11] xfs: allow newer INCOMPAT/RO_COMPAT feature bits to protect ATTRI log items
2023-12-31 19:50 ` [PATCHSET v13.0 2/6] xfsprogs: name-value xattr lookups Darrick J. Wong
@ 2023-12-31 23:02 ` Darrick J. Wong
2023-12-31 23:02 ` [PATCH 02/11] xfs: make xfs_attr_set require XFS_DA_OP_REMOVE Darrick J. Wong
` (9 subsequent siblings)
10 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:02 UTC (permalink / raw)
To: djwong, cem; +Cc: catherine.hoang, linux-xfs, allison.henderson
From: Darrick J. Wong <djwong@kernel.org>
Log recovery (which can include replaying ATTRI intent items) occurs on
rw and ro mounts. Dirty logs containing these log items must be
protected from being replayed by older kernels. The log incompat
feature XFS_SB_FEAT_INCOMPAT_LOG_XATTRS provides this protection.
However, adding this flag to the filesystem introduces performance
problems of its own -- each time we do, we must force the log and write
the primary superblock before writing any ATTRI log items. This was ok
when the only users were developers using the debug knob, but this sucks
for regular users. We'd like to avoid that.
If a filesystem has ro-compat or incompat feature bits set that weren't
defined at the time that ATTRI log items were defined, then any kernel
that doesn't know about ATTRI items will reject that filesystem. This
provides the same protection as the log-incompat feature, but at a much
lower cost because most ro-compat and incompat features are set on a
permanent basis.
Avoid the performance hit by detecting these feature bits and skipping
the xfs_add_incompat_log_feature calls.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
libxfs/xfs_attr.c | 6 +++++-
libxfs/xfs_attr.h | 23 +++++++++++++++++++++++
libxfs/xfs_format.h | 6 +++++-
3 files changed, 33 insertions(+), 2 deletions(-)
diff --git a/libxfs/xfs_attr.c b/libxfs/xfs_attr.c
index 8f527ac9292..8958434247f 100644
--- a/libxfs/xfs_attr.c
+++ b/libxfs/xfs_attr.c
@@ -883,9 +883,13 @@ xfs_attr_defer_add(
struct xfs_da_args *args,
unsigned int op_flags)
{
-
struct xfs_attr_intent *new;
+ /* ATTRI log items must be protected from older kernels */
+ if (args->op_flags & XFS_DA_OP_LOGGED)
+ ASSERT(xfs_attri_can_use_without_log_assistance(args->dp->i_mount) ||
+ xfs_sb_version_haslogxattrs(&args->dp->i_mount->m_sb));
+
new = kmem_cache_zalloc(xfs_attr_intent_cache, GFP_NOFS | __GFP_NOFAIL);
new->xattri_op_flags = op_flags;
new->xattri_da_args = args;
diff --git a/libxfs/xfs_attr.h b/libxfs/xfs_attr.h
index e4f55008552..273e8dff76c 100644
--- a/libxfs/xfs_attr.h
+++ b/libxfs/xfs_attr.h
@@ -620,4 +620,27 @@ void xfs_attr_intent_destroy_cache(void);
int xfs_attr_sf_totsize(struct xfs_inode *dp);
+/*
+ * Decide if this filesystem has a new enough permanent feature set to protect
+ * attri log items from being replayed on a kernel that does not have
+ * XFS_SB_FEAT_INCOMPAT_LOG_XATTRS set.
+ */
+static inline bool
+xfs_attri_can_use_without_log_assistance(
+ struct xfs_mount *mp)
+{
+ if (!xfs_sb_is_v5(&mp->m_sb))
+ return false;
+
+ if (xfs_sb_has_incompat_feature(&mp->m_sb,
+ ~(XFS_SB_FEAT_INCOMPAT_FTYPE |
+ XFS_SB_FEAT_INCOMPAT_SPINODES |
+ XFS_SB_FEAT_INCOMPAT_META_UUID |
+ XFS_SB_FEAT_INCOMPAT_BIGTIME |
+ XFS_SB_FEAT_INCOMPAT_NREXT64)))
+ return true;
+
+ return false;
+}
+
#endif /* __XFS_ATTR_H__ */
diff --git a/libxfs/xfs_format.h b/libxfs/xfs_format.h
index ec25010b577..8b952909ce1 100644
--- a/libxfs/xfs_format.h
+++ b/libxfs/xfs_format.h
@@ -390,7 +390,11 @@ xfs_sb_has_incompat_feature(
return (sbp->sb_features_incompat & feature) != 0;
}
-#define XFS_SB_FEAT_INCOMPAT_LOG_XATTRS (1 << 0) /* Delayed Attributes */
+/*
+ * Log contains ATTRI log intent items which are not otherwise protected by
+ * an INCOMPAT/RO_COMPAT feature flag.
+ */
+#define XFS_SB_FEAT_INCOMPAT_LOG_XATTRS (1 << 0)
/*
* Log contains SXI log intent items which are not otherwise protected by
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 02/11] xfs: make xfs_attr_set require XFS_DA_OP_REMOVE
2023-12-31 19:50 ` [PATCHSET v13.0 2/6] xfsprogs: name-value xattr lookups Darrick J. Wong
2023-12-31 23:02 ` [PATCH 01/11] xfs: allow newer INCOMPAT/RO_COMPAT feature bits to protect ATTRI log items Darrick J. Wong
@ 2023-12-31 23:02 ` Darrick J. Wong
2023-12-31 23:03 ` [PATCH 03/11] xfs: allow xattr matching on name and value for local/sf attrs Darrick J. Wong
` (8 subsequent siblings)
10 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:02 UTC (permalink / raw)
To: djwong, cem; +Cc: catherine.hoang, linux-xfs, allison.henderson
From: Darrick J. Wong <djwong@kernel.org>
Quite a few patches from now, we're going to change the parent pointer
xattr format to encode as much of the dirent name in the xattr name as
fits, and spill the rest of it to the xattr value. To make this work
correctly, we'll be adding the ability to look up xattrs based on name
/and/ value.
Internally, the xattr data structure supports attributes with a zero
length value, which is how we're going to store parent pointers for
short dirent names. The parent pointer repair code uses xfs_attr_set to
add missing and remove dangling parent pointers, so that interface must
be capable of setting an xattr with args->value == NULL.
The userspace API doesn't support this, so xfs_attr_set currently treats
a NULL args->value as a request to remove an attr. However, that's a
quirk of the existing callers and the interface. Make the callers of
xfs_attr_set to declare explicitly that they want to remove an xattr.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
db/attrset.c | 4 +++-
libxfs/xfs_attr.c | 10 +++++-----
2 files changed, 8 insertions(+), 6 deletions(-)
diff --git a/db/attrset.c b/db/attrset.c
index 2b6cdb5f5c3..123bdff1b62 100644
--- a/db/attrset.c
+++ b/db/attrset.c
@@ -185,7 +185,9 @@ attr_remove_f(
int argc,
char **argv)
{
- struct xfs_da_args args = { };
+ struct xfs_da_args args = {
+ .op_flags = XFS_DA_OP_REMOVE,
+ };
int c;
if (cur_typ == NULL) {
diff --git a/libxfs/xfs_attr.c b/libxfs/xfs_attr.c
index 8958434247f..ca6cfb1ee8a 100644
--- a/libxfs/xfs_attr.c
+++ b/libxfs/xfs_attr.c
@@ -924,6 +924,7 @@ xfs_attr_set(
struct xfs_mount *mp = dp->i_mount;
struct xfs_trans_res tres;
bool rsvd = (args->attr_filter & XFS_ATTR_ROOT);
+ bool is_remove = args->op_flags & XFS_DA_OP_REMOVE;
int error, local;
int rmt_blks = 0;
unsigned int total;
@@ -948,7 +949,7 @@ xfs_attr_set(
args->op_flags = XFS_DA_OP_OKNOENT |
(args->op_flags & XFS_DA_OP_LOGGED);
- if (args->value) {
+ if (!is_remove) {
XFS_STATS_INC(mp, xs_attr_set);
args->total = xfs_attr_calc_size(args, &local);
@@ -982,7 +983,7 @@ xfs_attr_set(
if (error)
return error;
- if (args->value || xfs_inode_hasattr(dp)) {
+ if (!is_remove || xfs_inode_hasattr(dp)) {
error = xfs_iext_count_may_overflow(dp, XFS_ATTR_FORK,
XFS_IEXT_ATTR_MANIP_CNT(rmt_blks));
if (error == -EFBIG)
@@ -995,8 +996,7 @@ xfs_attr_set(
error = xfs_attr_lookup(args);
switch (error) {
case -EEXIST:
- if (!args->value) {
- /* if no value, we are performing a remove operation */
+ if (is_remove) {
xfs_attr_defer_add(args, XFS_ATTRI_OP_FLAGS_REMOVE);
break;
}
@@ -1008,7 +1008,7 @@ xfs_attr_set(
break;
case -ENOATTR:
/* Can't remove what isn't there. */
- if (!args->value)
+ if (is_remove)
goto out_trans_cancel;
/* Pure replace fails if no existing attr to replace. */
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 03/11] xfs: allow xattr matching on name and value for local/sf attrs
2023-12-31 19:50 ` [PATCHSET v13.0 2/6] xfsprogs: name-value xattr lookups Darrick J. Wong
2023-12-31 23:02 ` [PATCH 01/11] xfs: allow newer INCOMPAT/RO_COMPAT feature bits to protect ATTRI log items Darrick J. Wong
2023-12-31 23:02 ` [PATCH 02/11] xfs: make xfs_attr_set require XFS_DA_OP_REMOVE Darrick J. Wong
@ 2023-12-31 23:03 ` Darrick J. Wong
2023-12-31 23:03 ` [PATCH 04/11] xfs: preserve NVLOOKUP in xfs_attr_set Darrick J. Wong
` (7 subsequent siblings)
10 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:03 UTC (permalink / raw)
To: djwong, cem; +Cc: catherine.hoang, linux-xfs, allison.henderson
From: Darrick J. Wong <djwong@kernel.org>
Add a new NVLOOKUP flag to signal that the caller wants to look up an
extended attribute by name and value. This only works with shortform
and local attributes. Only parent pointers need this functionality
and parent pointers cannot be remote xattrs, so this limitation is ok
for now.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
libxfs/xfs_attr_leaf.c | 45 +++++++++++++++++++++++++++++++++++++--------
libxfs/xfs_da_btree.h | 4 +++-
2 files changed, 40 insertions(+), 9 deletions(-)
diff --git a/libxfs/xfs_attr_leaf.c b/libxfs/xfs_attr_leaf.c
index 3d798828833..c0fc3c10dc4 100644
--- a/libxfs/xfs_attr_leaf.c
+++ b/libxfs/xfs_attr_leaf.c
@@ -505,10 +505,12 @@ xfs_attr3_leaf_read(
*/
static bool
xfs_attr_match(
- struct xfs_da_args *args,
- uint8_t namelen,
- unsigned char *name,
- int flags)
+ const struct xfs_da_args *args,
+ uint8_t namelen,
+ const unsigned char *name,
+ unsigned int valuelen,
+ const void *value,
+ int flags)
{
if (args->namelen != namelen)
@@ -516,6 +518,23 @@ xfs_attr_match(
if (memcmp(args->name, name, namelen) != 0)
return false;
+ if (args->op_flags & XFS_DA_OP_NVLOOKUP) {
+ if (args->valuelen != valuelen)
+ return false;
+ if (args->valuelen && !value) {
+ /* not implemented for remote values */
+ ASSERT(0);
+ return false;
+ }
+ if (valuelen && !args->value) {
+ /* caller gave us valuelen > 0 but no value?? */
+ ASSERT(0);
+ return false;
+ }
+ if (valuelen > 0 && memcmp(args->value, value, valuelen) != 0)
+ return false;
+ }
+
/* Recovery ignores the INCOMPLETE flag. */
if ((args->op_flags & XFS_DA_OP_RECOVERY) &&
args->attr_filter == (flags & XFS_ATTR_NSP_ONDISK_MASK))
@@ -534,6 +553,10 @@ xfs_attr_copy_value(
unsigned char *value,
int valuelen)
{
+ /* vlookups already supplied the attr value; don't copy anything */
+ if (args->op_flags & XFS_DA_OP_NVLOOKUP)
+ return 0;
+
/*
* No copy if all we have to do is get the length
*/
@@ -758,6 +781,7 @@ xfs_attr_sf_findname(
base += size, i++) {
size = xfs_attr_sf_entsize(sfe);
if (!xfs_attr_match(args, sfe->namelen, sfe->nameval,
+ sfe->valuelen, &sfe->nameval[sfe->namelen],
sfe->flags))
continue;
break;
@@ -926,6 +950,7 @@ xfs_attr_shortform_lookup(xfs_da_args_t *args)
for (i = 0; i < sf->hdr.count;
sfe = xfs_attr_sf_nextentry(sfe), i++) {
if (xfs_attr_match(args, sfe->namelen, sfe->nameval,
+ sfe->valuelen, &sfe->nameval[sfe->namelen],
sfe->flags))
return -EEXIST;
}
@@ -953,6 +978,7 @@ xfs_attr_shortform_getvalue(
for (i = 0; i < sf->hdr.count;
sfe = xfs_attr_sf_nextentry(sfe), i++) {
if (xfs_attr_match(args, sfe->namelen, sfe->nameval,
+ sfe->valuelen, &sfe->nameval[sfe->namelen],
sfe->flags))
return xfs_attr_copy_value(args,
&sfe->nameval[args->namelen], sfe->valuelen);
@@ -1005,7 +1031,7 @@ xfs_attr_shortform_to_leaf(
nargs.total = args->total;
nargs.whichfork = XFS_ATTR_FORK;
nargs.trans = args->trans;
- nargs.op_flags = XFS_DA_OP_OKNOENT;
+ nargs.op_flags = XFS_DA_OP_OKNOENT | XFS_DA_OP_NVLOOKUP;
nargs.owner = args->owner;
sfe = &sf->list[0];
@@ -1209,7 +1235,7 @@ xfs_attr3_leaf_to_shortform(
nargs.total = args->total;
nargs.whichfork = XFS_ATTR_FORK;
nargs.trans = args->trans;
- nargs.op_flags = XFS_DA_OP_OKNOENT;
+ nargs.op_flags = XFS_DA_OP_OKNOENT | XFS_DA_OP_NVLOOKUP;
nargs.owner = args->owner;
for (i = 0; i < ichdr.count; entry++, i++) {
@@ -2506,14 +2532,17 @@ xfs_attr3_leaf_lookup_int(
if (entry->flags & XFS_ATTR_LOCAL) {
name_loc = xfs_attr3_leaf_name_local(leaf, probe);
if (!xfs_attr_match(args, name_loc->namelen,
- name_loc->nameval, entry->flags))
+ name_loc->nameval,
+ be16_to_cpu(name_loc->valuelen),
+ &name_loc->nameval[name_loc->namelen],
+ entry->flags))
continue;
args->index = probe;
return -EEXIST;
} else {
name_rmt = xfs_attr3_leaf_name_remote(leaf, probe);
if (!xfs_attr_match(args, name_rmt->namelen,
- name_rmt->name, entry->flags))
+ name_rmt->name, 0, NULL, entry->flags))
continue;
args->index = probe;
args->rmtvaluelen = be32_to_cpu(name_rmt->valuelen);
diff --git a/libxfs/xfs_da_btree.h b/libxfs/xfs_da_btree.h
index 7a004786ee0..1bcb291150e 100644
--- a/libxfs/xfs_da_btree.h
+++ b/libxfs/xfs_da_btree.h
@@ -94,6 +94,7 @@ typedef struct xfs_da_args {
#define XFS_DA_OP_REMOVE (1u << 6) /* this is a remove operation */
#define XFS_DA_OP_RECOVERY (1u << 7) /* Log recovery operation */
#define XFS_DA_OP_LOGGED (1u << 8) /* Use intent items to track op */
+#define XFS_DA_OP_NVLOOKUP (1u << 9) /* Match local attr on name+value */
#define XFS_DA_OP_FLAGS \
{ XFS_DA_OP_JUSTCHECK, "JUSTCHECK" }, \
@@ -104,7 +105,8 @@ typedef struct xfs_da_args {
{ XFS_DA_OP_NOTIME, "NOTIME" }, \
{ XFS_DA_OP_REMOVE, "REMOVE" }, \
{ XFS_DA_OP_RECOVERY, "RECOVERY" }, \
- { XFS_DA_OP_LOGGED, "LOGGED" }
+ { XFS_DA_OP_LOGGED, "LOGGED" }, \
+ { XFS_DA_OP_NVLOOKUP, "NVLOOKUP" }
/*
* Storage for holding state during Btree searches and split/join ops.
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 04/11] xfs: preserve NVLOOKUP in xfs_attr_set
2023-12-31 19:50 ` [PATCHSET v13.0 2/6] xfsprogs: name-value xattr lookups Darrick J. Wong
` (2 preceding siblings ...)
2023-12-31 23:03 ` [PATCH 03/11] xfs: allow xattr matching on name and value for local/sf attrs Darrick J. Wong
@ 2023-12-31 23:03 ` Darrick J. Wong
2023-12-31 23:03 ` [PATCH 05/11] xfs: restructure xfs_attr_complete_op a bit Darrick J. Wong
` (6 subsequent siblings)
10 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:03 UTC (permalink / raw)
To: djwong, cem; +Cc: catherine.hoang, linux-xfs, allison.henderson
From: Darrick J. Wong <djwong@kernel.org>
Preserve the attr-value lookup flag when calling xfs_attr_set. Normal
xattr users will never use this, but parent pointer fsck will.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
libxfs/xfs_attr.c | 8 ++++----
1 file changed, 4 insertions(+), 4 deletions(-)
diff --git a/libxfs/xfs_attr.c b/libxfs/xfs_attr.c
index ca6cfb1ee8a..e714ea60319 100644
--- a/libxfs/xfs_attr.c
+++ b/libxfs/xfs_attr.c
@@ -943,11 +943,11 @@ xfs_attr_set(
/*
* We have no control over the attribute names that userspace passes us
* to remove, so we have to allow the name lookup prior to attribute
- * removal to fail as well. Preserve the logged flag, since we need
- * to pass that through to the logging code.
+ * removal to fail as well. Preserve the logged and vlookup flags,
+ * since we need to pass them through to the lower levels.
*/
- args->op_flags = XFS_DA_OP_OKNOENT |
- (args->op_flags & XFS_DA_OP_LOGGED);
+ args->op_flags &= (XFS_DA_OP_LOGGED | XFS_DA_OP_NVLOOKUP);
+ args->op_flags |= XFS_DA_OP_OKNOENT;
if (!is_remove) {
XFS_STATS_INC(mp, xs_attr_set);
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 05/11] xfs: restructure xfs_attr_complete_op a bit
2023-12-31 19:50 ` [PATCHSET v13.0 2/6] xfsprogs: name-value xattr lookups Darrick J. Wong
` (3 preceding siblings ...)
2023-12-31 23:03 ` [PATCH 04/11] xfs: preserve NVLOOKUP in xfs_attr_set Darrick J. Wong
@ 2023-12-31 23:03 ` Darrick J. Wong
2023-12-31 23:03 ` [PATCH 06/11] xfs: use helpers to extract xattr op from opflags Darrick J. Wong
` (5 subsequent siblings)
10 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:03 UTC (permalink / raw)
To: djwong, cem; +Cc: catherine.hoang, linux-xfs, allison.henderson
From: Darrick J. Wong <djwong@kernel.org>
Reduce the indentation in this function by flattening the nested if
statements. We're going to add more code later to this function later,
hence the early cleanup. No functional changes.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
libxfs/xfs_attr.c | 10 +++++-----
1 file changed, 5 insertions(+), 5 deletions(-)
diff --git a/libxfs/xfs_attr.c b/libxfs/xfs_attr.c
index e714ea60319..3f9c504e755 100644
--- a/libxfs/xfs_attr.c
+++ b/libxfs/xfs_attr.c
@@ -419,11 +419,11 @@ xfs_attr_complete_op(
bool do_replace = args->op_flags & XFS_DA_OP_REPLACE;
args->op_flags &= ~XFS_DA_OP_REPLACE;
- if (do_replace) {
- args->attr_filter &= ~XFS_ATTR_INCOMPLETE;
- return replace_state;
- }
- return XFS_DAS_DONE;
+ if (!do_replace)
+ return XFS_DAS_DONE;
+
+ args->attr_filter &= ~XFS_ATTR_INCOMPLETE;
+ return replace_state;
}
static int
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 06/11] xfs: use helpers to extract xattr op from opflags
2023-12-31 19:50 ` [PATCHSET v13.0 2/6] xfsprogs: name-value xattr lookups Darrick J. Wong
` (4 preceding siblings ...)
2023-12-31 23:03 ` [PATCH 05/11] xfs: restructure xfs_attr_complete_op a bit Darrick J. Wong
@ 2023-12-31 23:03 ` Darrick J. Wong
2023-12-31 23:04 ` [PATCH 07/11] xfs: log NVLOOKUP xattr removal operations Darrick J. Wong
` (4 subsequent siblings)
10 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:03 UTC (permalink / raw)
To: djwong, cem; +Cc: catherine.hoang, linux-xfs, allison.henderson
From: Darrick J. Wong <djwong@kernel.org>
Create helper functions to extract the xattr op from the ondisk xattri
log item and the incore attr intent item. These will get more use in
the patches that follow.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
libxfs/xfs_attr.h | 5 +++++
1 file changed, 5 insertions(+)
diff --git a/libxfs/xfs_attr.h b/libxfs/xfs_attr.h
index 273e8dff76c..ca51b93873b 100644
--- a/libxfs/xfs_attr.h
+++ b/libxfs/xfs_attr.h
@@ -529,6 +529,11 @@ struct xfs_attr_intent {
struct xfs_bmbt_irec xattri_map;
};
+static inline unsigned int
+xfs_attr_intent_op(const struct xfs_attr_intent *attr)
+{
+ return attr->xattri_op_flags & XFS_ATTRI_OP_FLAGS_TYPE_MASK;
+}
/*========================================================================
* Function prototypes for the kernel.
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 07/11] xfs: log NVLOOKUP xattr removal operations
2023-12-31 19:50 ` [PATCHSET v13.0 2/6] xfsprogs: name-value xattr lookups Darrick J. Wong
` (5 preceding siblings ...)
2023-12-31 23:03 ` [PATCH 06/11] xfs: use helpers to extract xattr op from opflags Darrick J. Wong
@ 2023-12-31 23:04 ` Darrick J. Wong
2023-12-31 23:04 ` [PATCH 08/11] xfs: log NVLOOKUP xattr setting operations Darrick J. Wong
` (3 subsequent siblings)
10 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:04 UTC (permalink / raw)
To: djwong, cem; +Cc: catherine.hoang, linux-xfs, allison.henderson
From: Darrick J. Wong <djwong@kernel.org>
If high level code wants to do a deferred xattr remove operation with
the NVLOOKUP flag set, we need to push this through the log.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
libxfs/xfs_attr.c | 2 ++
libxfs/xfs_log_format.h | 1 +
2 files changed, 3 insertions(+)
diff --git a/libxfs/xfs_attr.c b/libxfs/xfs_attr.c
index 3f9c504e755..c38048536af 100644
--- a/libxfs/xfs_attr.c
+++ b/libxfs/xfs_attr.c
@@ -902,6 +902,8 @@ xfs_attr_defer_add(
new->xattri_dela_state = xfs_attr_init_replace_state(args);
break;
case XFS_ATTRI_OP_FLAGS_REMOVE:
+ if (args->op_flags & XFS_DA_OP_NVLOOKUP)
+ new->xattri_op_flags = XFS_ATTRI_OP_FLAGS_NVREMOVE;
new->xattri_dela_state = xfs_attr_init_remove_state(args);
break;
default:
diff --git a/libxfs/xfs_log_format.h b/libxfs/xfs_log_format.h
index d4531060b6b..bf648b75194 100644
--- a/libxfs/xfs_log_format.h
+++ b/libxfs/xfs_log_format.h
@@ -1043,6 +1043,7 @@ struct xfs_icreate_log {
#define XFS_ATTRI_OP_FLAGS_SET 1 /* Set the attribute */
#define XFS_ATTRI_OP_FLAGS_REMOVE 2 /* Remove the attribute */
#define XFS_ATTRI_OP_FLAGS_REPLACE 3 /* Replace the attribute */
+#define XFS_ATTRI_OP_FLAGS_NVREMOVE 4 /* Remove attr w/ vlookup */
#define XFS_ATTRI_OP_FLAGS_TYPE_MASK 0xFF /* Flags type mask */
/*
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 08/11] xfs: log NVLOOKUP xattr setting operations
2023-12-31 19:50 ` [PATCHSET v13.0 2/6] xfsprogs: name-value xattr lookups Darrick J. Wong
` (6 preceding siblings ...)
2023-12-31 23:04 ` [PATCH 07/11] xfs: log NVLOOKUP xattr removal operations Darrick J. Wong
@ 2023-12-31 23:04 ` Darrick J. Wong
2023-12-31 23:04 ` [PATCH 09/11] xfs: log NVLOOKUP xattr nvreplace operations Darrick J. Wong
` (2 subsequent siblings)
10 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:04 UTC (permalink / raw)
To: djwong, cem; +Cc: catherine.hoang, linux-xfs, allison.henderson
From: Darrick J. Wong <djwong@kernel.org>
If high level code wants to do a deferred xattr set operation with the
NVLOOKUP flag set, we need to push this through the log.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
libxfs/xfs_attr.c | 2 ++
libxfs/xfs_log_format.h | 1 +
2 files changed, 3 insertions(+)
diff --git a/libxfs/xfs_attr.c b/libxfs/xfs_attr.c
index c38048536af..47684d07693 100644
--- a/libxfs/xfs_attr.c
+++ b/libxfs/xfs_attr.c
@@ -896,6 +896,8 @@ xfs_attr_defer_add(
switch (op_flags) {
case XFS_ATTRI_OP_FLAGS_SET:
+ if (args->op_flags & XFS_DA_OP_NVLOOKUP)
+ new->xattri_op_flags = XFS_ATTRI_OP_FLAGS_NVSET;
new->xattri_dela_state = xfs_attr_init_add_state(args);
break;
case XFS_ATTRI_OP_FLAGS_REPLACE:
diff --git a/libxfs/xfs_log_format.h b/libxfs/xfs_log_format.h
index bf648b75194..2ac520a18e9 100644
--- a/libxfs/xfs_log_format.h
+++ b/libxfs/xfs_log_format.h
@@ -1044,6 +1044,7 @@ struct xfs_icreate_log {
#define XFS_ATTRI_OP_FLAGS_REMOVE 2 /* Remove the attribute */
#define XFS_ATTRI_OP_FLAGS_REPLACE 3 /* Replace the attribute */
#define XFS_ATTRI_OP_FLAGS_NVREMOVE 4 /* Remove attr w/ vlookup */
+#define XFS_ATTRI_OP_FLAGS_NVSET 5 /* Set attr with w/ vlookup */
#define XFS_ATTRI_OP_FLAGS_TYPE_MASK 0xFF /* Flags type mask */
/*
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 09/11] xfs: log NVLOOKUP xattr nvreplace operations
2023-12-31 19:50 ` [PATCHSET v13.0 2/6] xfsprogs: name-value xattr lookups Darrick J. Wong
` (7 preceding siblings ...)
2023-12-31 23:04 ` [PATCH 08/11] xfs: log NVLOOKUP xattr setting operations Darrick J. Wong
@ 2023-12-31 23:04 ` Darrick J. Wong
2023-12-31 23:04 ` [PATCH 10/11] xfs_logprint: dump new attr log item fields Darrick J. Wong
2023-12-31 23:05 ` [PATCH 11/11] xfs_logprint: print missing attri header fields Darrick J. Wong
10 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:04 UTC (permalink / raw)
To: djwong, cem
Cc: Allison Henderson, catherine.hoang, linux-xfs, allison.henderson
From: Allison Henderson <allison.henderson@oracle.com>
(Formerly titled "xfs: Add new name to attri/d" and described as
follows:
This patch adds two new fields to the atti/d. They are nname and
nnamelen. This will be used for parent pointer updates since a
rename operation may cause the parent pointer to update both the
name and value. So we need to carry both the new name as well as
the target name in the attri/d.)
If high level code wants to do a deferred xattr nvreplace operation with
the NVLOOKUP flag set, we need to push this through the log. This log
item records the old name/value pair and the new name/value pair, and
completely replaces one with the other. Parent pointers will need this
ability to handle rename moving a child file between parents.
Signed-off-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
[djwong: reworked to handle new disk format]
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
libxfs/xfs_attr.c | 16 ++++++++++++++++
libxfs/xfs_attr.h | 4 ++--
libxfs/xfs_da_btree.h | 6 +++++-
libxfs/xfs_log_format.h | 27 +++++++++++++++++++++++----
4 files changed, 46 insertions(+), 7 deletions(-)
diff --git a/libxfs/xfs_attr.c b/libxfs/xfs_attr.c
index 47684d07693..3fe9041ae2c 100644
--- a/libxfs/xfs_attr.c
+++ b/libxfs/xfs_attr.c
@@ -423,6 +423,20 @@ xfs_attr_complete_op(
return XFS_DAS_DONE;
args->attr_filter &= ~XFS_ATTR_INCOMPLETE;
+ if (xfs_attr_intent_op(attr) != XFS_ATTRI_OP_FLAGS_NVREPLACE)
+ return replace_state;
+
+ /*
+ * NVREPLACE operations require the caller to set the old and new names
+ * and values explicitly.
+ */
+ ASSERT(args->new_namelen > 0);
+
+ args->name = args->new_name;
+ args->namelen = args->new_namelen;
+ args->hashval = xfs_da_hashname(args->name, args->namelen);
+ args->value = args->new_value;
+ args->valuelen = args->new_valuelen;
return replace_state;
}
@@ -901,6 +915,8 @@ xfs_attr_defer_add(
new->xattri_dela_state = xfs_attr_init_add_state(args);
break;
case XFS_ATTRI_OP_FLAGS_REPLACE:
+ if (args->op_flags & XFS_DA_OP_NVLOOKUP)
+ new->xattri_op_flags = XFS_ATTRI_OP_FLAGS_NVREPLACE;
new->xattri_dela_state = xfs_attr_init_replace_state(args);
break;
case XFS_ATTRI_OP_FLAGS_REMOVE:
diff --git a/libxfs/xfs_attr.h b/libxfs/xfs_attr.h
index ca51b93873b..b4e8ecee3e0 100644
--- a/libxfs/xfs_attr.h
+++ b/libxfs/xfs_attr.h
@@ -510,8 +510,8 @@ struct xfs_attr_intent {
struct xfs_da_args *xattri_da_args;
/*
- * Shared buffer containing the attr name and value so that the logging
- * code can share large memory buffers between log items.
+ * Shared buffer containing the attr name, new name, and value so that
+ * the logging code can share large memory buffers between log items.
*/
struct xfs_attri_log_nameval *xattri_nameval;
diff --git a/libxfs/xfs_da_btree.h b/libxfs/xfs_da_btree.h
index 1bcb291150e..93fcf49ab79 100644
--- a/libxfs/xfs_da_btree.h
+++ b/libxfs/xfs_da_btree.h
@@ -54,11 +54,15 @@ enum xfs_dacmp {
*/
typedef struct xfs_da_args {
struct xfs_da_geometry *geo; /* da block geometry */
- const uint8_t *name; /* string (maybe not NULL terminated) */
+ const uint8_t *name; /* string (maybe not NULL terminated) */
+ const uint8_t *new_name; /* new attr name */
int namelen; /* length of string (maybe no NULL) */
+ int new_namelen; /* new attr name len */
uint8_t filetype; /* filetype of inode for directories */
void *value; /* set of bytes (maybe contain NULLs) */
+ void *new_value; /* new xattr value (may contain NULLs) */
int valuelen; /* length of value */
+ int new_valuelen; /* length of new attr value */
unsigned int attr_filter; /* XFS_ATTR_{ROOT,SECURE,INCOMPLETE} */
unsigned int attr_flags; /* XATTR_{CREATE,REPLACE} */
xfs_dahash_t hashval; /* hash value of name */
diff --git a/libxfs/xfs_log_format.h b/libxfs/xfs_log_format.h
index 2ac520a18e9..285a0a089df 100644
--- a/libxfs/xfs_log_format.h
+++ b/libxfs/xfs_log_format.h
@@ -115,11 +115,13 @@ struct xfs_unmount_log_format {
#define XLOG_REG_TYPE_BUD_FORMAT 26
#define XLOG_REG_TYPE_ATTRI_FORMAT 27
#define XLOG_REG_TYPE_ATTRD_FORMAT 28
-#define XLOG_REG_TYPE_ATTR_NAME 29
+#define XLOG_REG_TYPE_ATTR_NAME 29
#define XLOG_REG_TYPE_ATTR_VALUE 30
#define XLOG_REG_TYPE_SXI_FORMAT 31
#define XLOG_REG_TYPE_SXD_FORMAT 32
-#define XLOG_REG_TYPE_MAX 32
+#define XLOG_REG_TYPE_ATTR_NEWNAME 33
+#define XLOG_REG_TYPE_ATTR_NEWVALUE 34
+#define XLOG_REG_TYPE_MAX 34
/*
* Flags to log operation header
@@ -1045,6 +1047,7 @@ struct xfs_icreate_log {
#define XFS_ATTRI_OP_FLAGS_REPLACE 3 /* Replace the attribute */
#define XFS_ATTRI_OP_FLAGS_NVREMOVE 4 /* Remove attr w/ vlookup */
#define XFS_ATTRI_OP_FLAGS_NVSET 5 /* Set attr with w/ vlookup */
+#define XFS_ATTRI_OP_FLAGS_NVREPLACE 6 /* Replace attr name and val */
#define XFS_ATTRI_OP_FLAGS_TYPE_MASK 0xFF /* Flags type mask */
/*
@@ -1062,11 +1065,27 @@ struct xfs_icreate_log {
struct xfs_attri_log_format {
uint16_t alfi_type; /* attri log item type */
uint16_t alfi_size; /* size of this item */
- uint32_t __pad; /* pad to 64 bit aligned */
+
+ /*
+ * For NVREPLACE, this is the length of the new xattr value.
+ * alfi_value_len contains the length of the old xattr value.
+ */
+ uint32_t alfi_new_value_len;
+
uint64_t alfi_id; /* attri identifier */
uint64_t alfi_ino; /* the inode for this attr operation */
uint32_t alfi_op_flags; /* marks the op as a set or remove */
- uint32_t alfi_name_len; /* attr name length */
+ union {
+ uint32_t alfi_name_len; /* attr name length */
+ struct {
+ /*
+ * For NVREPLACE, these are the lengths of the old and
+ * new attr name.
+ */
+ uint16_t alfi_old_name_len;
+ uint16_t alfi_new_name_len;
+ };
+ };
uint32_t alfi_value_len; /* attr value length */
uint32_t alfi_attr_filter;/* attr filter flags */
};
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 10/11] xfs_logprint: dump new attr log item fields
2023-12-31 19:50 ` [PATCHSET v13.0 2/6] xfsprogs: name-value xattr lookups Darrick J. Wong
` (8 preceding siblings ...)
2023-12-31 23:04 ` [PATCH 09/11] xfs: log NVLOOKUP xattr nvreplace operations Darrick J. Wong
@ 2023-12-31 23:04 ` Darrick J. Wong
2023-12-31 23:05 ` [PATCH 11/11] xfs_logprint: print missing attri header fields Darrick J. Wong
10 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:04 UTC (permalink / raw)
To: djwong, cem; +Cc: catherine.hoang, linux-xfs, allison.henderson
From: Darrick J. Wong <djwong@kernel.org>
Dump the new extended attribute log item fields. This was split out
from the previous patch to make libxfs resyncing easier. This code
needs more cleaning, which we'll do in the next few patches before
moving on to the parent pointer code.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
logprint/log_redo.c | 132 +++++++++++++++++++++++++++++++++++++++++----------
logprint/logprint.h | 6 ++
2 files changed, 111 insertions(+), 27 deletions(-)
diff --git a/logprint/log_redo.c b/logprint/log_redo.c
index 770485df75d..7531c6117bd 100644
--- a/logprint/log_redo.c
+++ b/logprint/log_redo.c
@@ -674,6 +674,12 @@ xfs_attri_copy_log_format(
return 1;
}
+static inline unsigned int
+xfs_attr_log_item_op(const struct xfs_attri_log_format *attrp)
+{
+ return attrp->alfi_op_flags & XFS_ATTRI_OP_FLAGS_TYPE_MASK;
+}
+
int
xlog_print_trans_attri(
char **ptr,
@@ -683,6 +689,10 @@ xlog_print_trans_attri(
struct xfs_attri_log_format *src_f = NULL;
xlog_op_header_t *head = NULL;
uint dst_len;
+ unsigned int name_len = 0;
+ unsigned int new_name_len = 0;
+ unsigned int value_len = 0;
+ unsigned int new_value_len = 0;
int error = 0;
dst_len = sizeof(struct xfs_attri_log_format);
@@ -705,27 +715,67 @@ xlog_print_trans_attri(
memmove((char*)src_f, *ptr, src_len);
*ptr += src_len;
- printf(_("ATTRI: #regs: %d name_len: %d, value_len: %d id: 0x%llx\n"),
- src_f->alfi_size, src_f->alfi_name_len, src_f->alfi_value_len,
- (unsigned long long)src_f->alfi_id);
+ if (xfs_attr_log_item_op(src_f) == XFS_ATTRI_OP_FLAGS_NVREPLACE) {
+ name_len = src_f->alfi_old_name_len;
+ new_name_len = src_f->alfi_new_name_len;
+ value_len = src_f->alfi_value_len;
+ new_value_len = src_f->alfi_new_value_len;
+ } else {
+ name_len = src_f->alfi_name_len;
+ value_len = src_f->alfi_value_len;
+ }
+
+ printf(_("ATTRI: #regs: %d name_len: %u, new_name_len: %u, value_len: %u, new_value_len: %u id: 0x%llx\n"),
+ src_f->alfi_size,
+ name_len,
+ new_name_len,
+ value_len,
+ new_value_len,
+ (unsigned long long)src_f->alfi_id);
+
+ if (name_len > 0) {
+ printf(_("\n"));
+ (*i)++;
+ head = (xlog_op_header_t *)*ptr;
+ xlog_print_op_header(head, *i, ptr);
+ error = xlog_print_trans_attri_name(ptr,
+ be32_to_cpu(head->oh_len), "name");
+ if (error)
+ goto error;
+ }
- if (src_f->alfi_name_len > 0) {
+ if (new_name_len > 0) {
printf(_("\n"));
(*i)++;
head = (xlog_op_header_t *)*ptr;
xlog_print_op_header(head, *i, ptr);
- error = xlog_print_trans_attri_name(ptr, be32_to_cpu(head->oh_len));
+ error = xlog_print_trans_attri_name(ptr,
+ be32_to_cpu(head->oh_len), "newname");
if (error)
goto error;
}
- if (src_f->alfi_value_len > 0) {
+ if (value_len > 0) {
printf(_("\n"));
(*i)++;
head = (xlog_op_header_t *)*ptr;
xlog_print_op_header(head, *i, ptr);
- error = xlog_print_trans_attri_value(ptr, be32_to_cpu(head->oh_len),
- src_f->alfi_value_len);
+ error = xlog_print_trans_attri_value(ptr,
+ be32_to_cpu(head->oh_len), value_len, "value");
+ if (error)
+ goto error;
+ }
+
+ if (new_value_len > 0) {
+ printf(_("\n"));
+ (*i)++;
+ head = (xlog_op_header_t *)*ptr;
+ xlog_print_op_header(head, *i, ptr);
+ error = xlog_print_trans_attri_value(ptr,
+ be32_to_cpu(head->oh_len), new_value_len,
+ "newvalue");
+ if (error)
+ goto error;
}
error:
free(src_f);
@@ -736,31 +786,33 @@ xlog_print_trans_attri(
int
xlog_print_trans_attri_name(
char **ptr,
- uint src_len)
+ uint src_len,
+ const char *tag)
{
- printf(_("ATTRI: name len:%u\n"), src_len);
+ printf(_("ATTRI: %s len:%u\n"), tag, src_len);
print_or_dump(*ptr, src_len);
*ptr += src_len;
return 0;
-} /* xlog_print_trans_attri */
+}
int
xlog_print_trans_attri_value(
char **ptr,
uint src_len,
- int value_len)
+ int value_len,
+ const char *tag)
{
int len = min(value_len, src_len);
- printf(_("ATTRI: value len:%u\n"), value_len);
+ printf(_("ATTRI: %s len:%u\n"), tag, value_len);
print_or_dump(*ptr, len);
*ptr += src_len;
return 0;
-} /* xlog_print_trans_attri_value */
+}
void
xlog_recover_print_attri(
@@ -768,7 +820,10 @@ xlog_recover_print_attri(
{
struct xfs_attri_log_format *f, *src_f = NULL;
uint src_len, dst_len;
-
+ unsigned int name_len = 0;
+ unsigned int new_name_len = 0;
+ unsigned int value_len = 0;
+ unsigned int new_value_len = 0;
int region = 0;
src_f = (struct xfs_attri_log_format *)item->ri_buf[0].i_addr;
@@ -788,24 +843,51 @@ xlog_recover_print_attri(
if (xfs_attri_copy_log_format((char*)src_f, src_len, f))
goto out;
- printf(_("ATTRI: #regs: %d name_len: %d, value_len: %d id: 0x%llx\n"),
- f->alfi_size, f->alfi_name_len, f->alfi_value_len, (unsigned long long)f->alfi_id);
+ if (xfs_attr_log_item_op(f) == XFS_ATTRI_OP_FLAGS_NVREPLACE) {
+ name_len = f->alfi_old_name_len;
+ new_name_len = f->alfi_new_name_len;
+ value_len = f->alfi_value_len;
+ new_value_len = f->alfi_new_value_len;
+ } else {
+ name_len = f->alfi_name_len;
+ value_len = f->alfi_value_len;
+ }
+
+ printf(_("ATTRI: #regs: %d name_len: %u, new_name_len: %u, value_len: %d, new_value_len: %u id: 0x%llx\n"),
+ f->alfi_size,
+ name_len,
+ new_name_len,
+ value_len,
+ new_value_len,
+ (unsigned long long)f->alfi_id);
- if (f->alfi_name_len > 0) {
+ if (name_len > 0) {
region++;
- printf(_("ATTRI: name len:%u\n"), f->alfi_name_len);
+ printf(_("ATTRI: name len:%u\n"), name_len);
print_or_dump((char *)item->ri_buf[region].i_addr,
- f->alfi_name_len);
+ name_len);
}
- if (f->alfi_value_len > 0) {
- int len = f->alfi_value_len;
+ if (new_name_len > 0) {
+ region++;
+ printf(_("ATTRI: newname len:%u\n"), new_name_len);
+ print_or_dump((char *)item->ri_buf[region].i_addr,
+ new_name_len);
+ }
+
+ if (value_len > 0) {
+ int len = min(MAX_ATTR_VAL_PRINT, value_len);
+
+ region++;
+ printf(_("ATTRI: value len:%u\n"), value_len);
+ print_or_dump((char *)item->ri_buf[region].i_addr, len);
+ }
- if (len > MAX_ATTR_VAL_PRINT)
- len = MAX_ATTR_VAL_PRINT;
+ if (new_value_len > 0) {
+ int len = min(MAX_ATTR_VAL_PRINT, new_value_len);
region++;
- printf(_("ATTRI: value len:%u\n"), f->alfi_value_len);
+ printf(_("ATTRI: newvalue len:%u\n"), new_value_len);
print_or_dump((char *)item->ri_buf[region].i_addr, len);
}
diff --git a/logprint/logprint.h b/logprint/logprint.h
index 892b280b548..8742b98a9d1 100644
--- a/logprint/logprint.h
+++ b/logprint/logprint.h
@@ -59,8 +59,10 @@ extern void xlog_recover_print_bud(struct xlog_recover_item *item);
#define MAX_ATTR_VAL_PRINT 128
extern int xlog_print_trans_attri(char **ptr, uint src_len, int *i);
-extern int xlog_print_trans_attri_name(char **ptr, uint src_len);
-extern int xlog_print_trans_attri_value(char **ptr, uint src_len, int value_len);
+extern int xlog_print_trans_attri_name(char **ptr, uint src_len,
+ const char *tag);
+extern int xlog_print_trans_attri_value(char **ptr, uint src_len, int value_len,
+ const char *tag);
extern void xlog_recover_print_attri(struct xlog_recover_item *item);
extern int xlog_print_trans_attrd(char **ptr, uint len);
extern void xlog_recover_print_attrd(struct xlog_recover_item *item);
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 11/11] xfs_logprint: print missing attri header fields
2023-12-31 19:50 ` [PATCHSET v13.0 2/6] xfsprogs: name-value xattr lookups Darrick J. Wong
` (9 preceding siblings ...)
2023-12-31 23:04 ` [PATCH 10/11] xfs_logprint: dump new attr log item fields Darrick J. Wong
@ 2023-12-31 23:05 ` Darrick J. Wong
10 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:05 UTC (permalink / raw)
To: djwong, cem; +Cc: catherine.hoang, linux-xfs, allison.henderson
From: Darrick J. Wong <djwong@kernel.org>
Not sure why logprint doesn't print the op flags, inode, or attr filter
fields. Make it do that.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
logprint/log_redo.c | 10 ++++++++--
1 file changed, 8 insertions(+), 2 deletions(-)
diff --git a/logprint/log_redo.c b/logprint/log_redo.c
index 7531c6117bd..e6401bb293e 100644
--- a/logprint/log_redo.c
+++ b/logprint/log_redo.c
@@ -725,8 +725,11 @@ xlog_print_trans_attri(
value_len = src_f->alfi_value_len;
}
- printf(_("ATTRI: #regs: %d name_len: %u, new_name_len: %u, value_len: %u, new_value_len: %u id: 0x%llx\n"),
+ printf(_("ATTRI: #regs: %d f: 0x%x, ino: 0x%llx, attr_filter: 0x%x, name_len: %u, new_name_len: %u, value_len: %u, new_value_len: %u id: 0x%llx\n"),
src_f->alfi_size,
+ src_f->alfi_op_flags,
+ (unsigned long long)src_f->alfi_ino,
+ src_f->alfi_attr_filter,
name_len,
new_name_len,
value_len,
@@ -853,8 +856,11 @@ xlog_recover_print_attri(
value_len = f->alfi_value_len;
}
- printf(_("ATTRI: #regs: %d name_len: %u, new_name_len: %u, value_len: %d, new_value_len: %u id: 0x%llx\n"),
+ printf(_("ATTRI: #regs: %d f: 0x%x, ino: 0x%llx, attr_filter: 0x%x, name_len: %u, new_name_len: %u, value_len: %u, new_value_len: %u id: 0x%llx\n"),
f->alfi_size,
+ f->alfi_op_flags,
+ (unsigned long long)f->alfi_ino,
+ f->alfi_attr_filter,
name_len,
new_name_len,
value_len,
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 01/32] xfs: add parent pointer support to attribute code
2023-12-31 19:50 ` [PATCHSET v13.0 3/6] xfsprogs: Parent Pointers Darrick J. Wong
@ 2023-12-31 23:05 ` Darrick J. Wong
2023-12-31 23:05 ` [PATCH 02/32] xfs: define parent pointer ondisk extended attribute format Darrick J. Wong
` (30 subsequent siblings)
31 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:05 UTC (permalink / raw)
To: djwong, cem
Cc: Mark Tinguely, Dave Chinner, Allison Henderson, catherine.hoang,
linux-xfs, allison.henderson
From: Allison Henderson <allison.henderson@oracle.com>
Add the new parent attribute type. XFS_ATTR_PARENT is used only for
parent pointer entries; it uses reserved blocks like XFS_ATTR_ROOT.
Signed-off-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
libxfs/xfs_attr.c | 3 ++-
libxfs/xfs_da_format.h | 5 ++++-
libxfs/xfs_log_format.h | 1 +
3 files changed, 7 insertions(+), 2 deletions(-)
diff --git a/libxfs/xfs_attr.c b/libxfs/xfs_attr.c
index 3fe9041ae2c..78884485c08 100644
--- a/libxfs/xfs_attr.c
+++ b/libxfs/xfs_attr.c
@@ -943,7 +943,8 @@ xfs_attr_set(
struct xfs_inode *dp = args->dp;
struct xfs_mount *mp = dp->i_mount;
struct xfs_trans_res tres;
- bool rsvd = (args->attr_filter & XFS_ATTR_ROOT);
+ bool rsvd = (args->attr_filter & (XFS_ATTR_ROOT |
+ XFS_ATTR_PARENT));
bool is_remove = args->op_flags & XFS_DA_OP_REMOVE;
int error, local;
int rmt_blks = 0;
diff --git a/libxfs/xfs_da_format.h b/libxfs/xfs_da_format.h
index 0e1ada44f21..6b5971ee680 100644
--- a/libxfs/xfs_da_format.h
+++ b/libxfs/xfs_da_format.h
@@ -709,12 +709,15 @@ struct xfs_attr3_leafblock {
#define XFS_ATTR_LOCAL_BIT 0 /* attr is stored locally */
#define XFS_ATTR_ROOT_BIT 1 /* limit access to trusted attrs */
#define XFS_ATTR_SECURE_BIT 2 /* limit access to secure attrs */
+#define XFS_ATTR_PARENT_BIT 3 /* parent pointer attrs */
#define XFS_ATTR_INCOMPLETE_BIT 7 /* attr in middle of create/delete */
#define XFS_ATTR_LOCAL (1u << XFS_ATTR_LOCAL_BIT)
#define XFS_ATTR_ROOT (1u << XFS_ATTR_ROOT_BIT)
#define XFS_ATTR_SECURE (1u << XFS_ATTR_SECURE_BIT)
+#define XFS_ATTR_PARENT (1u << XFS_ATTR_PARENT_BIT)
#define XFS_ATTR_INCOMPLETE (1u << XFS_ATTR_INCOMPLETE_BIT)
-#define XFS_ATTR_NSP_ONDISK_MASK (XFS_ATTR_ROOT | XFS_ATTR_SECURE)
+#define XFS_ATTR_NSP_ONDISK_MASK \
+ (XFS_ATTR_ROOT | XFS_ATTR_SECURE | XFS_ATTR_PARENT)
#define XFS_ATTR_NAMESPACE_STR \
{ XFS_ATTR_LOCAL, "local" }, \
diff --git a/libxfs/xfs_log_format.h b/libxfs/xfs_log_format.h
index 285a0a089df..bded03634e5 100644
--- a/libxfs/xfs_log_format.h
+++ b/libxfs/xfs_log_format.h
@@ -1056,6 +1056,7 @@ struct xfs_icreate_log {
*/
#define XFS_ATTRI_FILTER_MASK (XFS_ATTR_ROOT | \
XFS_ATTR_SECURE | \
+ XFS_ATTR_PARENT | \
XFS_ATTR_INCOMPLETE)
/*
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 02/32] xfs: define parent pointer ondisk extended attribute format
2023-12-31 19:50 ` [PATCHSET v13.0 3/6] xfsprogs: Parent Pointers Darrick J. Wong
2023-12-31 23:05 ` [PATCH 01/32] xfs: add parent pointer support to attribute code Darrick J. Wong
@ 2023-12-31 23:05 ` Darrick J. Wong
2023-12-31 23:05 ` [PATCH 03/32] xfs: add parent pointer validator functions Darrick J. Wong
` (29 subsequent siblings)
31 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:05 UTC (permalink / raw)
To: djwong, cem
Cc: Dave Chinner, Allison Henderson, catherine.hoang, linux-xfs,
allison.henderson
From: Allison Henderson <allison.henderson@oracle.com>
We need to define the parent pointer attribute format before we start
adding support for it into all the code that needs to use it. The EA
format we will use encodes the following information:
name={parent inode #, parent inode generation, dirent namehash}
value={dirent name}
The inode/gen gives all the information we need to reliably identify the
parent without requiring child->parent lock ordering, and allows
userspace to do pathname component level reconstruction without the
kernel ever needing to verify the parent itself as part of ioctl calls.
Storing the dirent name hash in the key reduces hash collisions if a
file is hardlinked multiple times in the same directory.
By using the NVLOOKUP mode in the extended attribute code to match
parent pointers using both the xattr name and value, we can identify the
exact parent pointer EA we need to modify/remove in rename/unlink
operations without searching the entire EA space.
By storing the dirent name, we have enough information to be able to
validate and reconstruct damaged directory trees. Earlier iterations of
this patchset encoded the directory offset in the parent pointer key,
but this format required repair to keep that in sync across directory
rebuilds, which is unnecessary complexity.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
[djwong: replace diroffset with the namehash in the pptr key]
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
libxfs/xfs_da_format.h | 20 ++++++++++++++++++++
1 file changed, 20 insertions(+)
diff --git a/libxfs/xfs_da_format.h b/libxfs/xfs_da_format.h
index 6b5971ee680..edbd901d906 100644
--- a/libxfs/xfs_da_format.h
+++ b/libxfs/xfs_da_format.h
@@ -878,4 +878,24 @@ static inline unsigned int xfs_dir2_dirblock_bytes(struct xfs_sb *sbp)
xfs_failaddr_t xfs_da3_blkinfo_verify(struct xfs_buf *bp,
struct xfs_da3_blkinfo *hdr3);
+/*
+ * Parent pointer attribute format definition
+ *
+ * The xattr name encodes the parent inode number, generation and the crc32c
+ * hash of the dirent name.
+ *
+ * The xattr value contains the dirent name.
+ */
+struct xfs_parent_name_rec {
+ __be64 p_ino;
+ __be32 p_gen;
+ __be32 p_namehash;
+};
+
+/*
+ * Maximum size of the dirent name that can be stored in a parent pointer.
+ * This matches the maximum dirent name length.
+ */
+#define XFS_PARENT_DIRENT_NAME_MAX_SIZE (MAXNAMELEN - 1)
+
#endif /* __XFS_DA_FORMAT_H__ */
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 03/32] xfs: add parent pointer validator functions
2023-12-31 19:50 ` [PATCHSET v13.0 3/6] xfsprogs: Parent Pointers Darrick J. Wong
2023-12-31 23:05 ` [PATCH 01/32] xfs: add parent pointer support to attribute code Darrick J. Wong
2023-12-31 23:05 ` [PATCH 02/32] xfs: define parent pointer ondisk extended attribute format Darrick J. Wong
@ 2023-12-31 23:05 ` Darrick J. Wong
2023-12-31 23:06 ` [PATCH 04/32] xfs: extend transaction reservations for parent attributes Darrick J. Wong
` (28 subsequent siblings)
31 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:05 UTC (permalink / raw)
To: djwong, cem
Cc: Allison Henderson, catherine.hoang, linux-xfs, allison.henderson
From: Allison Henderson <allison.henderson@oracle.com>
Attribute names of parent pointers are not strings. So we need to
modify attr_namecheck to verify parent pointer records when the
XFS_ATTR_PARENT flag is set. At the same time, we need to validate attr
values during log recovery if the xattr is really a parent pointer.
Signed-off-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
[djwong: move functions to xfs_parent.c, adjust for new disk format]
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
libxfs/Makefile | 2 +
libxfs/xfs_attr.c | 10 +++-
libxfs/xfs_attr.h | 3 +
libxfs/xfs_da_format.h | 8 +++
libxfs/xfs_parent.c | 114 ++++++++++++++++++++++++++++++++++++++++++++++++
libxfs/xfs_parent.h | 19 ++++++++
repair/attr_repair.c | 19 +++++---
7 files changed, 165 insertions(+), 10 deletions(-)
create mode 100644 libxfs/xfs_parent.c
create mode 100644 libxfs/xfs_parent.h
diff --git a/libxfs/Makefile b/libxfs/Makefile
index 0fb8f7b39bc..38594965882 100644
--- a/libxfs/Makefile
+++ b/libxfs/Makefile
@@ -50,6 +50,7 @@ HFILES = \
xfs_ialloc_btree.h \
xfs_inode_buf.h \
xfs_inode_fork.h \
+ xfs_parent.h \
xfs_quota_defs.h \
xfs_refcount.h \
xfs_refcount_btree.h \
@@ -101,6 +102,7 @@ CFILES = cache.c \
xfs_inode_fork.c \
xfs_ialloc_btree.c \
xfs_log_rlimit.c \
+ xfs_parent.c \
xfs_refcount.c \
xfs_refcount_btree.c \
xfs_rmap.c \
diff --git a/libxfs/xfs_attr.c b/libxfs/xfs_attr.c
index 78884485c08..04561f0318a 100644
--- a/libxfs/xfs_attr.c
+++ b/libxfs/xfs_attr.c
@@ -24,6 +24,7 @@
#include "xfs_quota_defs.h"
#include "xfs_trans_space.h"
#include "xfs_trace.h"
+#include "xfs_parent.h"
struct kmem_cache *xfs_attr_intent_cache;
@@ -1543,9 +1544,14 @@ xfs_attr_node_get(
/* Returns true if the attribute entry name is valid. */
bool
xfs_attr_namecheck(
- const void *name,
- size_t length)
+ struct xfs_mount *mp,
+ const void *name,
+ size_t length,
+ unsigned int flags)
{
+ if (flags & XFS_ATTR_PARENT)
+ return xfs_parent_namecheck(mp, name, length, flags);
+
/*
* MAXNAMELEN includes the trailing null, but (name/length) leave it
* out, so use >= for the length check.
diff --git a/libxfs/xfs_attr.h b/libxfs/xfs_attr.h
index b4e8ecee3e0..5b3a0d4b158 100644
--- a/libxfs/xfs_attr.h
+++ b/libxfs/xfs_attr.h
@@ -552,7 +552,8 @@ int xfs_attr_get(struct xfs_da_args *args);
int xfs_attr_set(struct xfs_da_args *args);
int xfs_attr_set_iter(struct xfs_attr_intent *attr);
int xfs_attr_remove_iter(struct xfs_attr_intent *attr);
-bool xfs_attr_namecheck(const void *name, size_t length);
+bool xfs_attr_namecheck(struct xfs_mount *mp, const void *name, size_t length,
+ unsigned int flags);
int xfs_attr_calc_size(struct xfs_da_args *args, int *local);
void xfs_init_attr_trans(struct xfs_da_args *args, struct xfs_trans_res *tres,
unsigned int *total);
diff --git a/libxfs/xfs_da_format.h b/libxfs/xfs_da_format.h
index edbd901d906..a63387796a2 100644
--- a/libxfs/xfs_da_format.h
+++ b/libxfs/xfs_da_format.h
@@ -757,6 +757,14 @@ xfs_attr3_leaf_name(xfs_attr_leafblock_t *leafp, int idx)
return &((char *)leafp)[be16_to_cpu(entries[idx].nameidx)];
}
+static inline int
+xfs_attr3_leaf_flags(xfs_attr_leafblock_t *leafp, int idx)
+{
+ struct xfs_attr_leaf_entry *entries = xfs_attr3_leaf_entryp(leafp);
+
+ return entries[idx].flags;
+}
+
static inline xfs_attr_leaf_name_remote_t *
xfs_attr3_leaf_name_remote(xfs_attr_leafblock_t *leafp, int idx)
{
diff --git a/libxfs/xfs_parent.c b/libxfs/xfs_parent.c
new file mode 100644
index 00000000000..6874d8afe23
--- /dev/null
+++ b/libxfs/xfs_parent.c
@@ -0,0 +1,114 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Copyright (c) 2022-2024 Oracle.
+ * All rights reserved.
+ */
+#include "libxfs_priv.h"
+#include "xfs_shared.h"
+#include "xfs_format.h"
+#include "xfs_log_format.h"
+#include "xfs_trans_resv.h"
+#include "xfs_mount.h"
+#include "xfs_inode.h"
+#include "xfs_trans_resv.h"
+#include "xfs_mount.h"
+#include "xfs_trace.h"
+#include "xfs.h"
+#include "xfs_fs.h"
+#include "xfs_da_format.h"
+#include "xfs_bmap_btree.h"
+#include "xfs_trans.h"
+#include "xfs_da_btree.h"
+#include "xfs_attr.h"
+#include "xfs_dir2.h"
+#include "xfs_dir2_priv.h"
+#include "xfs_attr_sf.h"
+#include "xfs_bmap.h"
+#include "xfs_parent.h"
+#include "xfs_da_format.h"
+#include "xfs_format.h"
+#include "xfs_trans_space.h"
+
+/*
+ * Parent pointer attribute handling.
+ *
+ * Because the attribute value is a filename component, it will never be longer
+ * than 255 bytes. This means the attribute will always be a local format
+ * attribute as it is xfs_attr_leaf_entsize_local_max() for v5 filesystems will
+ * always be larger than this (max is 75% of block size).
+ *
+ * Creating a new parent attribute will always create a new attribute - there
+ * should never, ever be an existing attribute in the tree for a new inode.
+ * ENOSPC behavior is problematic - creating the inode without the parent
+ * pointer is effectively a corruption, so we allow parent attribute creation
+ * to dip into the reserve block pool to avoid unexpected ENOSPC errors from
+ * occurring.
+ */
+
+/* Return true if parent pointer EA name is valid. */
+bool
+xfs_parent_namecheck(
+ struct xfs_mount *mp,
+ const struct xfs_parent_name_rec *rec,
+ size_t reclen,
+ unsigned int attr_flags)
+{
+ if (!(attr_flags & XFS_ATTR_PARENT))
+ return false;
+
+ /* pptr updates use logged xattrs, so we should never see this flag */
+ if (attr_flags & XFS_ATTR_INCOMPLETE)
+ return false;
+
+ if (reclen != sizeof(struct xfs_parent_name_rec))
+ return false;
+
+ /* Only one namespace bit allowed. */
+ if (hweight32(attr_flags & XFS_ATTR_NSP_ONDISK_MASK) > 1)
+ return false;
+
+ return true;
+}
+
+/* Return true if parent pointer EA value is valid. */
+bool
+xfs_parent_valuecheck(
+ struct xfs_mount *mp,
+ const void *value,
+ size_t valuelen)
+{
+ if (valuelen == 0 || valuelen > XFS_PARENT_DIRENT_NAME_MAX_SIZE)
+ return false;
+
+ if (value == NULL)
+ return false;
+
+ return true;
+}
+
+/* Return true if the ondisk parent pointer is consistent. */
+bool
+xfs_parent_hashcheck(
+ struct xfs_mount *mp,
+ const struct xfs_parent_name_rec *rec,
+ const void *value,
+ size_t valuelen)
+{
+ struct xfs_name dname = {
+ .name = value,
+ .len = valuelen,
+ };
+ xfs_ino_t p_ino;
+
+ /* Valid dirent name? */
+ if (!xfs_dir2_namecheck(value, valuelen))
+ return false;
+
+ /* Valid inode number? */
+ p_ino = be64_to_cpu(rec->p_ino);
+ if (!xfs_verify_dir_ino(mp, p_ino))
+ return false;
+
+ /* Namehash matches name? */
+ return be32_to_cpu(rec->p_namehash) == xfs_dir2_hashname(mp, &dname);
+}
diff --git a/libxfs/xfs_parent.h b/libxfs/xfs_parent.h
new file mode 100644
index 00000000000..fcfeddb645f
--- /dev/null
+++ b/libxfs/xfs_parent.h
@@ -0,0 +1,19 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Copyright (c) 2022-2024 Oracle.
+ * All Rights Reserved.
+ */
+#ifndef __XFS_PARENT_H__
+#define __XFS_PARENT_H__
+
+/* Metadata validators */
+bool xfs_parent_namecheck(struct xfs_mount *mp,
+ const struct xfs_parent_name_rec *rec, size_t reclen,
+ unsigned int attr_flags);
+bool xfs_parent_valuecheck(struct xfs_mount *mp, const void *value,
+ size_t valuelen);
+bool xfs_parent_hashcheck(struct xfs_mount *mp,
+ const struct xfs_parent_name_rec *rec, const void *value,
+ size_t valuelen);
+
+#endif /* __XFS_PARENT_H__ */
diff --git a/repair/attr_repair.c b/repair/attr_repair.c
index b7ebf103dd6..314a9cb4d6e 100644
--- a/repair/attr_repair.c
+++ b/repair/attr_repair.c
@@ -294,8 +294,9 @@ process_shortform_attr(
}
/* namecheck checks for null chars in attr names. */
- if (!libxfs_attr_namecheck(currententry->nameval,
- currententry->namelen)) {
+ if (!libxfs_attr_namecheck(mp, currententry->nameval,
+ currententry->namelen,
+ currententry->flags)) {
do_warn(
_("entry contains illegal character in shortform attribute name\n"));
junkit = 1;
@@ -471,12 +472,14 @@ process_leaf_attr_local(
xfs_dablk_t da_bno,
xfs_ino_t ino)
{
- xfs_attr_leaf_name_local_t *local;
+ xfs_attr_leaf_name_local_t *local;
+ int flags;
local = xfs_attr3_leaf_name_local(leaf, i);
+ flags = xfs_attr3_leaf_flags(leaf, i);
if (local->namelen == 0 ||
- !libxfs_attr_namecheck(local->nameval,
- local->namelen)) {
+ !libxfs_attr_namecheck(mp, local->nameval,
+ local->namelen, flags)) {
do_warn(
_("attribute entry %d in attr block %u, inode %" PRIu64 " has bad name (namelen = %d)\n"),
i, da_bno, ino, local->namelen);
@@ -527,12 +530,14 @@ process_leaf_attr_remote(
{
xfs_attr_leaf_name_remote_t *remotep;
char* value;
+ int flags;
remotep = xfs_attr3_leaf_name_remote(leaf, i);
+ flags = xfs_attr3_leaf_flags(leaf, i);
if (remotep->namelen == 0 ||
- !libxfs_attr_namecheck(remotep->name,
- remotep->namelen) ||
+ !libxfs_attr_namecheck(mp, remotep->name,
+ remotep->namelen, flags) ||
be32_to_cpu(entry->hashval) !=
libxfs_da_hashname((unsigned char *)&remotep->name[0],
remotep->namelen) ||
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 04/32] xfs: extend transaction reservations for parent attributes
2023-12-31 19:50 ` [PATCHSET v13.0 3/6] xfsprogs: Parent Pointers Darrick J. Wong
` (2 preceding siblings ...)
2023-12-31 23:05 ` [PATCH 03/32] xfs: add parent pointer validator functions Darrick J. Wong
@ 2023-12-31 23:06 ` Darrick J. Wong
2023-12-31 23:06 ` [PATCH 05/32] xfs: parent pointer attribute creation Darrick J. Wong
` (27 subsequent siblings)
31 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:06 UTC (permalink / raw)
To: djwong, cem
Cc: Dave Chinner, Allison Henderson, catherine.hoang, linux-xfs,
allison.henderson
From: Allison Henderson <allison.henderson@oracle.com>
We need to add, remove or modify parent pointer attributes during
create/link/unlink/rename operations atomically with the dirents in the
parent directories being modified. This means they need to be modified
in the same transaction as the parent directories, and so we need to add
the required space for the attribute modifications to the transaction
reservations.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
[djwong: fix indenting errors, adjust for new log format]
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
libxfs/libxfs_priv.h | 1
libxfs/xfs_trans_resv.c | 324 +++++++++++++++++++++++++++++++++++++++--------
2 files changed, 273 insertions(+), 52 deletions(-)
diff --git a/libxfs/libxfs_priv.h b/libxfs/libxfs_priv.h
index 45cfe4408a9..411c33b3956 100644
--- a/libxfs/libxfs_priv.h
+++ b/libxfs/libxfs_priv.h
@@ -483,6 +483,7 @@ static inline int retzero(void) { return 0; }
#define xfs_icreate_log(tp, agno, agbno, cnt, isize, len, gen) ((void) 0)
#define xfs_sb_validate_fsb_count(sbp, nblks) (0)
+#define xlog_calc_iovec_len(len) roundup(len, sizeof(uint32_t))
/*
* Prototypes for kernel static functions that are aren't in their
diff --git a/libxfs/xfs_trans_resv.c b/libxfs/xfs_trans_resv.c
index 82b3d1522b6..78e7a575baa 100644
--- a/libxfs/xfs_trans_resv.c
+++ b/libxfs/xfs_trans_resv.c
@@ -19,6 +19,7 @@
#include "xfs_trans_space.h"
#include "xfs_quota_defs.h"
#include "xfs_rtbitmap.h"
+#include "xfs_da_format.h"
#define _ALLOC true
#define _FREE false
@@ -421,29 +422,110 @@ xfs_calc_itruncate_reservation_minlogsize(
return xfs_calc_itruncate_reservation(mp, true);
}
+static inline unsigned int xfs_calc_pptr_link_overhead(void)
+{
+ return sizeof(struct xfs_attri_log_format) +
+ xlog_calc_iovec_len(sizeof(struct xfs_parent_name_rec)) +
+ xlog_calc_iovec_len(XFS_PARENT_DIRENT_NAME_MAX_SIZE);
+}
+static inline unsigned int xfs_calc_pptr_unlink_overhead(void)
+{
+ return sizeof(struct xfs_attri_log_format) +
+ xlog_calc_iovec_len(sizeof(struct xfs_parent_name_rec)) +
+ xlog_calc_iovec_len(XFS_PARENT_DIRENT_NAME_MAX_SIZE);
+}
+static inline unsigned int xfs_calc_pptr_replace_overhead(void)
+{
+ return sizeof(struct xfs_attri_log_format) +
+ xlog_calc_iovec_len(sizeof(struct xfs_parent_name_rec)) +
+ xlog_calc_iovec_len(XFS_PARENT_DIRENT_NAME_MAX_SIZE) +
+ xlog_calc_iovec_len(sizeof(struct xfs_parent_name_rec)) +
+ xlog_calc_iovec_len(XFS_PARENT_DIRENT_NAME_MAX_SIZE);
+}
+
/*
* In renaming a files we can modify:
* the five inodes involved: 5 * inode size
* the two directory btrees: 2 * (max depth + v2) * dir block size
* the two directory bmap btrees: 2 * max depth * block size
* And the bmap_finish transaction can free dir and bmap blocks (two sets
- * of bmap blocks) giving:
+ * of bmap blocks) giving (t2):
* the agf for the ags in which the blocks live: 3 * sector size
* the agfl for the ags in which the blocks live: 3 * sector size
* the superblock for the free block count: sector size
* the allocation btrees: 3 exts * 2 trees * (2 * max depth - 1) * block size
+ * If parent pointers are enabled (t3), then each transaction in the chain
+ * must be capable of setting or removing the extended attribute
+ * containing the parent information. It must also be able to handle
+ * the three xattr intent items that track the progress of the parent
+ * pointer update.
*/
STATIC uint
xfs_calc_rename_reservation(
struct xfs_mount *mp)
{
- return XFS_DQUOT_LOGRES(mp) +
- max((xfs_calc_inode_res(mp, 5) +
- xfs_calc_buf_res(2 * XFS_DIROP_LOG_COUNT(mp),
- XFS_FSB_TO_B(mp, 1))),
- (xfs_calc_buf_res(7, mp->m_sb.sb_sectsize) +
- xfs_calc_buf_res(xfs_allocfree_block_count(mp, 3),
- XFS_FSB_TO_B(mp, 1))));
+ unsigned int overhead = XFS_DQUOT_LOGRES(mp);
+ struct xfs_trans_resv *resp = M_RES(mp);
+ unsigned int t1, t2, t3 = 0;
+
+ t1 = xfs_calc_inode_res(mp, 5) +
+ xfs_calc_buf_res(2 * XFS_DIROP_LOG_COUNT(mp),
+ XFS_FSB_TO_B(mp, 1));
+
+ t2 = xfs_calc_buf_res(7, mp->m_sb.sb_sectsize) +
+ xfs_calc_buf_res(xfs_allocfree_block_count(mp, 3),
+ XFS_FSB_TO_B(mp, 1));
+
+ if (xfs_has_parent(mp)) {
+ unsigned int rename_overhead, exchange_overhead;
+
+ t3 = max(resp->tr_attrsetm.tr_logres,
+ resp->tr_attrrm.tr_logres);
+
+ /*
+ * For a standard rename, the three xattr intent log items
+ * are (1) replacing the pptr for the source file; (2)
+ * removing the pptr on the dest file; and (3) adding a
+ * pptr for the whiteout file in the src dir.
+ *
+ * For an RENAME_EXCHANGE, there are two xattr intent
+ * items to replace the pptr for both src and dest
+ * files. Link counts don't change and there is no
+ * whiteout.
+ *
+ * In the worst case we can end up relogging all log
+ * intent items to allow the log tail to move ahead, so
+ * they become overhead added to each transaction in a
+ * processing chain.
+ */
+ rename_overhead = xfs_calc_pptr_replace_overhead() +
+ xfs_calc_pptr_unlink_overhead() +
+ xfs_calc_pptr_link_overhead();
+ exchange_overhead = 2 * xfs_calc_pptr_replace_overhead();
+
+ overhead += max(rename_overhead, exchange_overhead);
+ }
+
+ return overhead + max3(t1, t2, t3);
+}
+
+static inline unsigned int
+xfs_rename_log_count(
+ struct xfs_mount *mp,
+ struct xfs_trans_resv *resp)
+{
+ /* One for the rename, one more for freeing blocks */
+ unsigned int ret = XFS_RENAME_LOG_COUNT;
+
+ /*
+ * Pre-reserve enough log reservation to handle the transaction
+ * rolling needed to remove or add one parent pointer.
+ */
+ if (xfs_has_parent(mp))
+ ret += max(resp->tr_attrsetm.tr_logcount,
+ resp->tr_attrrm.tr_logcount);
+
+ return ret;
}
/*
@@ -460,6 +542,23 @@ xfs_calc_iunlink_remove_reservation(
2 * M_IGEO(mp)->inode_cluster_size;
}
+static inline unsigned int
+xfs_link_log_count(
+ struct xfs_mount *mp,
+ struct xfs_trans_resv *resp)
+{
+ unsigned int ret = XFS_LINK_LOG_COUNT;
+
+ /*
+ * Pre-reserve enough log reservation to handle the transaction
+ * rolling needed to add one parent pointer.
+ */
+ if (xfs_has_parent(mp))
+ ret += resp->tr_attrsetm.tr_logcount;
+
+ return ret;
+}
+
/*
* For creating a link to an inode:
* the parent directory inode: inode size
@@ -476,14 +575,23 @@ STATIC uint
xfs_calc_link_reservation(
struct xfs_mount *mp)
{
- return XFS_DQUOT_LOGRES(mp) +
- xfs_calc_iunlink_remove_reservation(mp) +
- max((xfs_calc_inode_res(mp, 2) +
- xfs_calc_buf_res(XFS_DIROP_LOG_COUNT(mp),
- XFS_FSB_TO_B(mp, 1))),
- (xfs_calc_buf_res(3, mp->m_sb.sb_sectsize) +
- xfs_calc_buf_res(xfs_allocfree_block_count(mp, 1),
- XFS_FSB_TO_B(mp, 1))));
+ unsigned int overhead = XFS_DQUOT_LOGRES(mp);
+ struct xfs_trans_resv *resp = M_RES(mp);
+ unsigned int t1, t2, t3 = 0;
+
+ overhead += xfs_calc_iunlink_remove_reservation(mp);
+ t1 = xfs_calc_inode_res(mp, 2) +
+ xfs_calc_buf_res(XFS_DIROP_LOG_COUNT(mp), XFS_FSB_TO_B(mp, 1));
+ t2 = xfs_calc_buf_res(3, mp->m_sb.sb_sectsize) +
+ xfs_calc_buf_res(xfs_allocfree_block_count(mp, 1),
+ XFS_FSB_TO_B(mp, 1));
+
+ if (xfs_has_parent(mp)) {
+ t3 = resp->tr_attrsetm.tr_logres;
+ overhead += xfs_calc_pptr_link_overhead();
+ }
+
+ return overhead + max3(t1, t2, t3);
}
/*
@@ -498,6 +606,23 @@ xfs_calc_iunlink_add_reservation(xfs_mount_t *mp)
M_IGEO(mp)->inode_cluster_size;
}
+static inline unsigned int
+xfs_remove_log_count(
+ struct xfs_mount *mp,
+ struct xfs_trans_resv *resp)
+{
+ unsigned int ret = XFS_REMOVE_LOG_COUNT;
+
+ /*
+ * Pre-reserve enough log reservation to handle the transaction
+ * rolling needed to add one parent pointer.
+ */
+ if (xfs_has_parent(mp))
+ ret += resp->tr_attrrm.tr_logcount;
+
+ return ret;
+}
+
/*
* For removing a directory entry we can modify:
* the parent directory inode: inode size
@@ -514,14 +639,24 @@ STATIC uint
xfs_calc_remove_reservation(
struct xfs_mount *mp)
{
- return XFS_DQUOT_LOGRES(mp) +
- xfs_calc_iunlink_add_reservation(mp) +
- max((xfs_calc_inode_res(mp, 2) +
- xfs_calc_buf_res(XFS_DIROP_LOG_COUNT(mp),
- XFS_FSB_TO_B(mp, 1))),
- (xfs_calc_buf_res(4, mp->m_sb.sb_sectsize) +
- xfs_calc_buf_res(xfs_allocfree_block_count(mp, 2),
- XFS_FSB_TO_B(mp, 1))));
+ unsigned int overhead = XFS_DQUOT_LOGRES(mp);
+ struct xfs_trans_resv *resp = M_RES(mp);
+ unsigned int t1, t2, t3 = 0;
+
+ overhead += xfs_calc_iunlink_add_reservation(mp);
+
+ t1 = xfs_calc_inode_res(mp, 2) +
+ xfs_calc_buf_res(XFS_DIROP_LOG_COUNT(mp), XFS_FSB_TO_B(mp, 1));
+ t2 = xfs_calc_buf_res(4, mp->m_sb.sb_sectsize) +
+ xfs_calc_buf_res(xfs_allocfree_block_count(mp, 2),
+ XFS_FSB_TO_B(mp, 1));
+
+ if (xfs_has_parent(mp)) {
+ t3 = resp->tr_attrrm.tr_logres;
+ overhead += xfs_calc_pptr_unlink_overhead();
+ }
+
+ return overhead + max3(t1, t2, t3);
}
/*
@@ -570,12 +705,40 @@ xfs_calc_icreate_resv_alloc(
xfs_calc_finobt_res(mp);
}
+static inline unsigned int
+xfs_icreate_log_count(
+ struct xfs_mount *mp,
+ struct xfs_trans_resv *resp)
+{
+ unsigned int ret = XFS_CREATE_LOG_COUNT;
+
+ /*
+ * Pre-reserve enough log reservation to handle the transaction
+ * rolling needed to add one parent pointer.
+ */
+ if (xfs_has_parent(mp))
+ ret += resp->tr_attrsetm.tr_logcount;
+
+ return ret;
+}
+
STATIC uint
-xfs_calc_icreate_reservation(xfs_mount_t *mp)
+xfs_calc_icreate_reservation(
+ struct xfs_mount *mp)
{
- return XFS_DQUOT_LOGRES(mp) +
- max(xfs_calc_icreate_resv_alloc(mp),
- xfs_calc_create_resv_modify(mp));
+ struct xfs_trans_resv *resp = M_RES(mp);
+ unsigned int overhead = XFS_DQUOT_LOGRES(mp);
+ unsigned int t1, t2, t3 = 0;
+
+ t1 = xfs_calc_icreate_resv_alloc(mp);
+ t2 = xfs_calc_create_resv_modify(mp);
+
+ if (xfs_has_parent(mp)) {
+ t3 = resp->tr_attrsetm.tr_logres;
+ overhead += xfs_calc_pptr_link_overhead();
+ }
+
+ return overhead + max3(t1, t2, t3);
}
STATIC uint
@@ -588,6 +751,23 @@ xfs_calc_create_tmpfile_reservation(
return res + xfs_calc_iunlink_add_reservation(mp);
}
+static inline unsigned int
+xfs_mkdir_log_count(
+ struct xfs_mount *mp,
+ struct xfs_trans_resv *resp)
+{
+ unsigned int ret = XFS_MKDIR_LOG_COUNT;
+
+ /*
+ * Pre-reserve enough log reservation to handle the transaction
+ * rolling needed to add one parent pointer.
+ */
+ if (xfs_has_parent(mp))
+ ret += resp->tr_attrsetm.tr_logcount;
+
+ return ret;
+}
+
/*
* Making a new directory is the same as creating a new file.
*/
@@ -598,6 +778,22 @@ xfs_calc_mkdir_reservation(
return xfs_calc_icreate_reservation(mp);
}
+static inline unsigned int
+xfs_symlink_log_count(
+ struct xfs_mount *mp,
+ struct xfs_trans_resv *resp)
+{
+ unsigned int ret = XFS_SYMLINK_LOG_COUNT;
+
+ /*
+ * Pre-reserve enough log reservation to handle the transaction
+ * rolling needed to add one parent pointer.
+ */
+ if (xfs_has_parent(mp))
+ ret += resp->tr_attrsetm.tr_logcount;
+
+ return ret;
+}
/*
* Making a new symplink is the same as creating a new file, but
@@ -910,6 +1106,52 @@ xfs_calc_sb_reservation(
return xfs_calc_buf_res(1, mp->m_sb.sb_sectsize);
}
+/*
+ * Namespace reservations.
+ *
+ * These get tricky when parent pointers are enabled as we have attribute
+ * modifications occurring from within these transactions. Rather than confuse
+ * each of these reservation calculations with the conditional attribute
+ * reservations, add them here in a clear and concise manner. This requires that
+ * the attribute reservations have already been calculated.
+ *
+ * Note that we only include the static attribute reservation here; the runtime
+ * reservation will have to be modified by the size of the attributes being
+ * added/removed/modified. See the comments on the attribute reservation
+ * calculations for more details.
+ */
+STATIC void
+xfs_calc_namespace_reservations(
+ struct xfs_mount *mp,
+ struct xfs_trans_resv *resp)
+{
+ ASSERT(resp->tr_attrsetm.tr_logres > 0);
+
+ resp->tr_rename.tr_logres = xfs_calc_rename_reservation(mp);
+ resp->tr_rename.tr_logcount = xfs_rename_log_count(mp, resp);
+ resp->tr_rename.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
+
+ resp->tr_link.tr_logres = xfs_calc_link_reservation(mp);
+ resp->tr_link.tr_logcount = xfs_link_log_count(mp, resp);
+ resp->tr_link.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
+
+ resp->tr_remove.tr_logres = xfs_calc_remove_reservation(mp);
+ resp->tr_remove.tr_logcount = xfs_remove_log_count(mp, resp);
+ resp->tr_remove.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
+
+ resp->tr_symlink.tr_logres = xfs_calc_symlink_reservation(mp);
+ resp->tr_symlink.tr_logcount = xfs_symlink_log_count(mp, resp);
+ resp->tr_symlink.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
+
+ resp->tr_create.tr_logres = xfs_calc_icreate_reservation(mp);
+ resp->tr_create.tr_logcount = xfs_icreate_log_count(mp, resp);
+ resp->tr_create.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
+
+ resp->tr_mkdir.tr_logres = xfs_calc_mkdir_reservation(mp);
+ resp->tr_mkdir.tr_logcount = xfs_mkdir_log_count(mp, resp);
+ resp->tr_mkdir.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
+}
+
void
xfs_trans_resv_calc(
struct xfs_mount *mp,
@@ -929,35 +1171,11 @@ xfs_trans_resv_calc(
resp->tr_itruncate.tr_logcount = XFS_ITRUNCATE_LOG_COUNT;
resp->tr_itruncate.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
- resp->tr_rename.tr_logres = xfs_calc_rename_reservation(mp);
- resp->tr_rename.tr_logcount = XFS_RENAME_LOG_COUNT;
- resp->tr_rename.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
-
- resp->tr_link.tr_logres = xfs_calc_link_reservation(mp);
- resp->tr_link.tr_logcount = XFS_LINK_LOG_COUNT;
- resp->tr_link.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
-
- resp->tr_remove.tr_logres = xfs_calc_remove_reservation(mp);
- resp->tr_remove.tr_logcount = XFS_REMOVE_LOG_COUNT;
- resp->tr_remove.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
-
- resp->tr_symlink.tr_logres = xfs_calc_symlink_reservation(mp);
- resp->tr_symlink.tr_logcount = XFS_SYMLINK_LOG_COUNT;
- resp->tr_symlink.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
-
- resp->tr_create.tr_logres = xfs_calc_icreate_reservation(mp);
- resp->tr_create.tr_logcount = XFS_CREATE_LOG_COUNT;
- resp->tr_create.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
-
resp->tr_create_tmpfile.tr_logres =
xfs_calc_create_tmpfile_reservation(mp);
resp->tr_create_tmpfile.tr_logcount = XFS_CREATE_TMPFILE_LOG_COUNT;
resp->tr_create_tmpfile.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
- resp->tr_mkdir.tr_logres = xfs_calc_mkdir_reservation(mp);
- resp->tr_mkdir.tr_logcount = XFS_MKDIR_LOG_COUNT;
- resp->tr_mkdir.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
-
resp->tr_ifree.tr_logres = xfs_calc_ifree_reservation(mp);
resp->tr_ifree.tr_logcount = XFS_INACTIVE_LOG_COUNT;
resp->tr_ifree.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
@@ -987,6 +1205,8 @@ xfs_trans_resv_calc(
resp->tr_qm_dqalloc.tr_logcount = XFS_WRITE_LOG_COUNT;
resp->tr_qm_dqalloc.tr_logflags |= XFS_TRANS_PERM_LOG_RES;
+ xfs_calc_namespace_reservations(mp, resp);
+
/*
* The following transactions are logged in logical format with
* a default log count.
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 05/32] xfs: parent pointer attribute creation
2023-12-31 19:50 ` [PATCHSET v13.0 3/6] xfsprogs: Parent Pointers Darrick J. Wong
` (3 preceding siblings ...)
2023-12-31 23:06 ` [PATCH 04/32] xfs: extend transaction reservations for parent attributes Darrick J. Wong
@ 2023-12-31 23:06 ` Darrick J. Wong
2023-12-31 23:06 ` [PATCH 06/32] xfs: add parent attributes to link Darrick J. Wong
` (26 subsequent siblings)
31 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:06 UTC (permalink / raw)
To: djwong, cem
Cc: Dave Chinner, Allison Henderson, catherine.hoang, linux-xfs,
allison.henderson
From: Allison Henderson <allison.henderson@oracle.com>
Add parent pointer attribute during xfs_create, and subroutines to
initialize attributes
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
[djwong: shorten names, adjust to new format, set init_xattrs for parent
pointers]
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
include/libxfs.h | 1 +
include/xfs_inode.h | 6 +++
libxfs/Makefile | 1 +
libxfs/init.c | 3 ++
libxfs/libxfs_api_defs.h | 2 +
libxfs/libxfs_priv.h | 5 ++-
libxfs/xfs_attr.c | 2 +
libxfs/xfs_attr.h | 2 +
libxfs/xfs_parent.c | 91 ++++++++++++++++++++++++++++++++++++++++++++++
libxfs/xfs_parent.h | 77 ++++++++++++++++++++++++++++++++++++++-
libxfs/xfs_trans_space.c | 52 ++++++++++++++++++++++++++
libxfs/xfs_trans_space.h | 9 +++--
repair/phase6.c | 8 ++--
13 files changed, 246 insertions(+), 13 deletions(-)
create mode 100644 libxfs/xfs_trans_space.c
diff --git a/include/libxfs.h b/include/libxfs.h
index 77ecfda4bc7..425112b0693 100644
--- a/include/libxfs.h
+++ b/include/libxfs.h
@@ -87,6 +87,7 @@ struct iomap;
#include "xfs_rtbitmap.h"
#include "xfs_symlink_remote.h"
#include "xfs_ag_resv.h"
+#include "xfs_parent.h"
#ifndef ARRAY_SIZE
#define ARRAY_SIZE(x) (sizeof(x) / sizeof((x)[0]))
diff --git a/include/xfs_inode.h b/include/xfs_inode.h
index 47959314811..088a6b34f04 100644
--- a/include/xfs_inode.h
+++ b/include/xfs_inode.h
@@ -268,6 +268,12 @@ static inline struct inode *VFS_I(struct xfs_inode *ip)
return &ip->i_vnode;
}
+/* convert from const xfs inode to const vfs inode */
+static inline const struct inode *VFS_IC(const struct xfs_inode *ip)
+{
+ return &ip->i_vnode;
+}
+
/* We only have i_size in the xfs inode in userspace */
static inline loff_t i_size_read(struct inode *inode)
{
diff --git a/libxfs/Makefile b/libxfs/Makefile
index 38594965882..e0bdaefb209 100644
--- a/libxfs/Makefile
+++ b/libxfs/Makefile
@@ -113,6 +113,7 @@ CFILES = cache.c \
xfs_symlink_remote.c \
xfs_trans_inode.c \
xfs_trans_resv.c \
+ xfs_trans_space.c \
xfs_types.c
#
diff --git a/libxfs/init.c b/libxfs/init.c
index 2e59ba0d0a2..b6b1282201c 100644
--- a/libxfs/init.c
+++ b/libxfs/init.c
@@ -213,6 +213,8 @@ init_caches(void)
"xfs_extfree_item");
xfs_trans_cache = kmem_cache_init(
sizeof(struct xfs_trans), "xfs_trans");
+ xfs_parent_args_cache = kmem_cache_init(
+ sizeof(struct xfs_parent_args), "xfs_parent_args");
}
static int
@@ -230,6 +232,7 @@ destroy_caches(void)
xfs_btree_destroy_cur_caches();
leaked += kmem_cache_destroy(xfs_extfree_item_cache);
leaked += kmem_cache_destroy(xfs_trans_cache);
+ leaked += kmem_cache_destroy(xfs_parent_args_cache);
return leaked;
}
diff --git a/libxfs/libxfs_api_defs.h b/libxfs/libxfs_api_defs.h
index eba9a8386d2..6ab10be3ad6 100644
--- a/libxfs/libxfs_api_defs.h
+++ b/libxfs/libxfs_api_defs.h
@@ -89,6 +89,7 @@
#define xfs_bwrite libxfs_bwrite
#define xfs_calc_dquots_per_chunk libxfs_calc_dquots_per_chunk
#define xfs_compute_rextslog libxfs_compute_rextslog
+#define xfs_create_space_res libxfs_create_space_res
#define xfs_da3_node_hdr_from_disk libxfs_da3_node_hdr_from_disk
#define xfs_da_get_buf libxfs_da_get_buf
#define xfs_da_hashname libxfs_da_hashname
@@ -175,6 +176,7 @@
#define xfs_log_get_max_trans_res libxfs_log_get_max_trans_res
#define xfs_log_sb libxfs_log_sb
#define xfs_mode_to_ftype libxfs_mode_to_ftype
+#define xfs_mkdir_space_res libxfs_mkdir_space_res
#define xfs_perag_get libxfs_perag_get
#define xfs_perag_hold libxfs_perag_hold
#define xfs_perag_put libxfs_perag_put
diff --git a/libxfs/libxfs_priv.h b/libxfs/libxfs_priv.h
index 411c33b3956..123e25d2f5e 100644
--- a/libxfs/libxfs_priv.h
+++ b/libxfs/libxfs_priv.h
@@ -562,8 +562,9 @@ struct xfs_item_ops;
bool xfs_log_check_lsn(struct xfs_mount *, xfs_lsn_t);
void xfs_log_item_init(struct xfs_mount *mp, struct xfs_log_item *lip, int type,
const struct xfs_item_ops *ops);
-#define xfs_attr_use_log_assist(mp) (0)
-#define xlog_drop_incompat_feat(log) do { } while (0)
+#define xfs_attr_grab_log_assist(mp) (0)
+#define xfs_attr_rele_log_assist(mp) ((void) 0)
+#define xlog_drop_incompat_feat(log,w) do { } while (0)
#define xfs_log_in_recovery(mp) (false)
/* xfs_icache.c */
diff --git a/libxfs/xfs_attr.c b/libxfs/xfs_attr.c
index 04561f0318a..c4f543db474 100644
--- a/libxfs/xfs_attr.c
+++ b/libxfs/xfs_attr.c
@@ -893,7 +893,7 @@ xfs_attr_lookup(
return error;
}
-static void
+void
xfs_attr_defer_add(
struct xfs_da_args *args,
unsigned int op_flags)
diff --git a/libxfs/xfs_attr.h b/libxfs/xfs_attr.h
index 5b3a0d4b158..4a4d45a96dd 100644
--- a/libxfs/xfs_attr.h
+++ b/libxfs/xfs_attr.h
@@ -549,6 +549,7 @@ int xfs_inode_hasattr(struct xfs_inode *ip);
bool xfs_attr_is_leaf(struct xfs_inode *ip);
int xfs_attr_get_ilocked(struct xfs_da_args *args);
int xfs_attr_get(struct xfs_da_args *args);
+void xfs_attr_defer_add(struct xfs_da_args *args, unsigned int op_flags);
int xfs_attr_set(struct xfs_da_args *args);
int xfs_attr_set_iter(struct xfs_attr_intent *attr);
int xfs_attr_remove_iter(struct xfs_attr_intent *attr);
@@ -557,7 +558,6 @@ bool xfs_attr_namecheck(struct xfs_mount *mp, const void *name, size_t length,
int xfs_attr_calc_size(struct xfs_da_args *args, int *local);
void xfs_init_attr_trans(struct xfs_da_args *args, struct xfs_trans_res *tres,
unsigned int *total);
-
/*
* Check to see if the attr should be upgraded from non-existent or shortform to
* single-leaf-block attribute list.
diff --git a/libxfs/xfs_parent.c b/libxfs/xfs_parent.c
index 6874d8afe23..32e1d1f62ec 100644
--- a/libxfs/xfs_parent.c
+++ b/libxfs/xfs_parent.c
@@ -29,6 +29,8 @@
#include "xfs_format.h"
#include "xfs_trans_space.h"
+struct kmem_cache *xfs_parent_args_cache;
+
/*
* Parent pointer attribute handling.
*
@@ -112,3 +114,92 @@ xfs_parent_hashcheck(
/* Namehash matches name? */
return be32_to_cpu(rec->p_namehash) == xfs_dir2_hashname(mp, &dname);
}
+
+/* Initializes a xfs_parent_name_rec to be stored as an attribute name. */
+static inline void
+xfs_init_parent_name_rec(
+ struct xfs_parent_name_rec *rec,
+ const struct xfs_inode *dp,
+ const struct xfs_name *name,
+ struct xfs_inode *ip)
+{
+ rec->p_ino = cpu_to_be64(dp->i_ino);
+ rec->p_gen = cpu_to_be32(VFS_IC(dp)->i_generation);
+ rec->p_namehash = cpu_to_be32(xfs_dir2_hashname(dp->i_mount, name));
+}
+
+/* Point the da args value fields at the non-key parts of a parent pointer. */
+static inline void
+xfs_init_parent_davalue(
+ struct xfs_da_args *args,
+ const struct xfs_name *name)
+{
+ args->valuelen = name->len;
+ args->value = (void *)name->name;
+}
+
+/*
+ * Allocate memory to control a logged parent pointer update as part of a
+ * dirent operation.
+ */
+int
+xfs_parent_args_alloc(
+ struct xfs_mount *mp,
+ struct xfs_parent_args **ppargsp)
+{
+ struct xfs_parent_args *ppargs;
+
+ ppargs = kmem_cache_zalloc(xfs_parent_args_cache, GFP_KERNEL);
+ if (!ppargs)
+ return -ENOMEM;
+
+ xfs_parent_args_init(mp, ppargs);
+ *ppargsp = ppargs;
+ return 0;
+}
+
+static inline xfs_dahash_t
+xfs_parent_hashname(
+ struct xfs_inode *ip,
+ const struct xfs_parent_args *ppargs)
+{
+ return xfs_da_hashname((const void *)&ppargs->rec,
+ sizeof(struct xfs_parent_name_rec));
+}
+
+/* Add a parent pointer to reflect a dirent addition. */
+int
+xfs_parent_addname(
+ struct xfs_trans *tp,
+ struct xfs_parent_args *ppargs,
+ struct xfs_inode *dp,
+ const struct xfs_name *parent_name,
+ struct xfs_inode *child)
+{
+ struct xfs_da_args *args = &ppargs->args;
+
+ if (XFS_IS_CORRUPT(tp->t_mountp,
+ !xfs_parent_valuecheck(tp->t_mountp, parent_name->name,
+ parent_name->len)))
+ return -EFSCORRUPTED;
+
+ xfs_init_parent_name_rec(&ppargs->rec, dp, parent_name, child);
+ args->hashval = xfs_parent_hashname(dp, ppargs);
+
+ args->trans = tp;
+ args->dp = child;
+
+ xfs_init_parent_davalue(&ppargs->args, parent_name);
+
+ xfs_attr_defer_add(args, XFS_ATTRI_OP_FLAGS_SET);
+ return 0;
+}
+
+/* Free a parent pointer context object. */
+void
+xfs_parent_args_free(
+ struct xfs_mount *mp,
+ struct xfs_parent_args *ppargs)
+{
+ kmem_cache_free(xfs_parent_args_cache, ppargs);
+}
diff --git a/libxfs/xfs_parent.h b/libxfs/xfs_parent.h
index fcfeddb645f..e2115a2b964 100644
--- a/libxfs/xfs_parent.h
+++ b/libxfs/xfs_parent.h
@@ -16,4 +16,79 @@ bool xfs_parent_hashcheck(struct xfs_mount *mp,
const struct xfs_parent_name_rec *rec, const void *value,
size_t valuelen);
-#endif /* __XFS_PARENT_H__ */
+extern struct kmem_cache *xfs_parent_args_cache;
+
+/*
+ * Dynamically allocd structure used to wrap the needed data to pass around
+ * the defer ops machinery
+ */
+struct xfs_parent_args {
+ struct xfs_parent_name_rec rec;
+ struct xfs_da_args args;
+};
+
+int xfs_parent_args_alloc(struct xfs_mount *mp,
+ struct xfs_parent_args **ppargsp);
+
+/*
+ * Initialize the parent pointer arguments structure. Caller must have zeroed
+ * the contents.
+ */
+static inline void
+xfs_parent_args_init(
+ struct xfs_mount *mp,
+ struct xfs_parent_args *ppargs)
+{
+ ppargs->args.geo = mp->m_attr_geo;
+ ppargs->args.whichfork = XFS_ATTR_FORK;
+ ppargs->args.attr_filter = XFS_ATTR_PARENT;
+ ppargs->args.op_flags = XFS_DA_OP_OKNOENT | XFS_DA_OP_LOGGED |
+ XFS_DA_OP_NVLOOKUP;
+ ppargs->args.name = (const uint8_t *)&ppargs->rec;
+ ppargs->args.namelen = sizeof(struct xfs_parent_name_rec);
+}
+
+/*
+ * Start a parent pointer update by allocating the context object we need to
+ * perform a parent pointer update.
+ */
+static inline int
+xfs_parent_start(
+ struct xfs_mount *mp,
+ struct xfs_parent_args **ppargsp)
+{
+ *ppargsp = NULL;
+
+ if (xfs_has_parent(mp))
+ return xfs_parent_args_alloc(mp, ppargsp);
+ return 0;
+}
+
+int xfs_parent_addname(struct xfs_trans *tp, struct xfs_parent_args *ppargs,
+ struct xfs_inode *dp, const struct xfs_name *parent_name,
+ struct xfs_inode *child);
+
+/* Schedule a parent pointer addition. */
+static inline int
+xfs_parent_add(struct xfs_trans *tp, struct xfs_parent_args *ppargs,
+ struct xfs_inode *dp, const struct xfs_name *parent_name,
+ struct xfs_inode *child)
+{
+ if (ppargs)
+ return xfs_parent_addname(tp, ppargs, dp, parent_name, child);
+ return 0;
+}
+
+void xfs_parent_args_free(struct xfs_mount *mp, struct xfs_parent_args *ppargs);
+
+/* Finish a parent pointer update by freeing the context object. */
+static inline void
+xfs_parent_finish(
+ struct xfs_mount *mp,
+ struct xfs_parent_args *ppargs)
+{
+ if (ppargs)
+ xfs_parent_args_free(mp, ppargs);
+}
+
+#endif /* __XFS_PARENT_H__ */
diff --git a/libxfs/xfs_trans_space.c b/libxfs/xfs_trans_space.c
new file mode 100644
index 00000000000..3408e700f01
--- /dev/null
+++ b/libxfs/xfs_trans_space.c
@@ -0,0 +1,52 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Copyright (c) 2000,2005 Silicon Graphics, Inc.
+ * All Rights Reserved.
+ */
+#include "libxfs_priv.h"
+#include "xfs_fs.h"
+#include "xfs_shared.h"
+#include "xfs_format.h"
+#include "xfs_da_format.h"
+#include "xfs_log_format.h"
+#include "xfs_trans_resv.h"
+#include "xfs_mount.h"
+#include "xfs_da_btree.h"
+#include "xfs_bmap_btree.h"
+#include "xfs_trans_space.h"
+
+/* Calculate the disk space required to add a parent pointer. */
+unsigned int
+xfs_parent_calc_space_res(
+ struct xfs_mount *mp,
+ unsigned int namelen)
+{
+ /*
+ * Parent pointers are always the first attr in an attr tree, and never
+ * larger than a block
+ */
+ return XFS_DAENTER_SPACE_RES(mp, XFS_ATTR_FORK) +
+ XFS_NEXTENTADD_SPACE_RES(mp, namelen, XFS_ATTR_FORK);
+}
+
+unsigned int
+xfs_create_space_res(
+ struct xfs_mount *mp,
+ unsigned int namelen)
+{
+ unsigned int ret;
+
+ ret = XFS_IALLOC_SPACE_RES(mp) + XFS_DIRENTER_SPACE_RES(mp, namelen);
+ if (xfs_has_parent(mp))
+ ret += xfs_parent_calc_space_res(mp, namelen);
+
+ return ret;
+}
+
+unsigned int
+xfs_mkdir_space_res(
+ struct xfs_mount *mp,
+ unsigned int namelen)
+{
+ return xfs_create_space_res(mp, namelen);
+}
diff --git a/libxfs/xfs_trans_space.h b/libxfs/xfs_trans_space.h
index 9640fc232c1..6cda87153b3 100644
--- a/libxfs/xfs_trans_space.h
+++ b/libxfs/xfs_trans_space.h
@@ -80,8 +80,6 @@
/* This macro is not used - see inline code in xfs_attr_set */
#define XFS_ATTRSET_SPACE_RES(mp, v) \
(XFS_DAENTER_SPACE_RES(mp, XFS_ATTR_FORK) + XFS_B_TO_FSB(mp, v))
-#define XFS_CREATE_SPACE_RES(mp,nl) \
- (XFS_IALLOC_SPACE_RES(mp) + XFS_DIRENTER_SPACE_RES(mp,nl))
#define XFS_DIOSTRAT_SPACE_RES(mp, v) \
(XFS_EXTENTADD_SPACE_RES(mp, XFS_DATA_FORK) + (v))
#define XFS_GROWFS_SPACE_RES(mp) \
@@ -90,8 +88,6 @@
((b) + XFS_EXTENTADD_SPACE_RES(mp, XFS_DATA_FORK))
#define XFS_LINK_SPACE_RES(mp,nl) \
XFS_DIRENTER_SPACE_RES(mp,nl)
-#define XFS_MKDIR_SPACE_RES(mp,nl) \
- (XFS_IALLOC_SPACE_RES(mp) + XFS_DIRENTER_SPACE_RES(mp,nl))
#define XFS_QM_DQALLOC_SPACE_RES(mp) \
(XFS_EXTENTADD_SPACE_RES(mp, XFS_DATA_FORK) + \
XFS_DQUOT_CLUSTER_SIZE_FSB)
@@ -106,5 +102,10 @@
#define XFS_IFREE_SPACE_RES(mp) \
(xfs_has_finobt(mp) ? M_IGEO(mp)->inobt_maxlevels : 0)
+unsigned int xfs_parent_calc_space_res(struct xfs_mount *mp,
+ unsigned int namelen);
+
+unsigned int xfs_create_space_res(struct xfs_mount *mp, unsigned int namelen);
+unsigned int xfs_mkdir_space_res(struct xfs_mount *mp, unsigned int namelen);
#endif /* __XFS_TRANS_SPACE_H__ */
diff --git a/repair/phase6.c b/repair/phase6.c
index 75391378291..825f0cf3956 100644
--- a/repair/phase6.c
+++ b/repair/phase6.c
@@ -924,7 +924,7 @@ mk_orphanage(xfs_mount_t *mp)
/*
* could not be found, create it
*/
- nres = XFS_MKDIR_SPACE_RES(mp, xname.len);
+ nres = libxfs_mkdir_space_res(mp, xname.len);
i = -libxfs_trans_alloc(mp, &M_RES(mp)->tr_mkdir, nres, 0, 0, &tp);
if (i)
res_failed(i);
@@ -1335,7 +1335,7 @@ longform_dir2_rebuild(
p->name.name[1] == '.'))))
continue;
- nres = XFS_CREATE_SPACE_RES(mp, p->name.len);
+ nres = libxfs_create_space_res(mp, p->name.len);
error = -libxfs_trans_alloc(mp, &M_RES(mp)->tr_create,
nres, 0, 0, &tp);
if (error)
@@ -2947,7 +2947,7 @@ _("error %d fixing shortform directory %llu\n"),
do_warn(_("recreating root directory .. entry\n"));
- nres = XFS_MKDIR_SPACE_RES(mp, 2);
+ nres = libxfs_mkdir_space_res(mp, 2);
error = -libxfs_trans_alloc(mp, &M_RES(mp)->tr_mkdir,
nres, 0, 0, &tp);
if (error)
@@ -3002,7 +3002,7 @@ _("error %d fixing shortform directory %llu\n"),
do_warn(
_("creating missing \".\" entry in dir ino %" PRIu64 "\n"), ino);
- nres = XFS_MKDIR_SPACE_RES(mp, 1);
+ nres = libxfs_mkdir_space_res(mp, 1);
error = -libxfs_trans_alloc(mp, &M_RES(mp)->tr_mkdir,
nres, 0, 0, &tp);
if (error)
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 06/32] xfs: add parent attributes to link
2023-12-31 19:50 ` [PATCHSET v13.0 3/6] xfsprogs: Parent Pointers Darrick J. Wong
` (4 preceding siblings ...)
2023-12-31 23:06 ` [PATCH 05/32] xfs: parent pointer attribute creation Darrick J. Wong
@ 2023-12-31 23:06 ` Darrick J. Wong
2023-12-31 23:06 ` [PATCH 07/32] xfs: add parent attributes to symlink Darrick J. Wong
` (25 subsequent siblings)
31 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:06 UTC (permalink / raw)
To: djwong, cem
Cc: Dave Chinner, Allison Henderson, catherine.hoang, linux-xfs,
allison.henderson
From: Allison Henderson <allison.henderson@oracle.com>
This patch modifies xfs_link to add a parent pointer to the inode.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
[djwong: minor rebase fixes]
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
libxfs/xfs_trans_space.c | 14 ++++++++++++++
libxfs/xfs_trans_space.h | 3 +--
2 files changed, 15 insertions(+), 2 deletions(-)
diff --git a/libxfs/xfs_trans_space.c b/libxfs/xfs_trans_space.c
index 3408e700f01..039bbd91e87 100644
--- a/libxfs/xfs_trans_space.c
+++ b/libxfs/xfs_trans_space.c
@@ -50,3 +50,17 @@ xfs_mkdir_space_res(
{
return xfs_create_space_res(mp, namelen);
}
+
+unsigned int
+xfs_link_space_res(
+ struct xfs_mount *mp,
+ unsigned int namelen)
+{
+ unsigned int ret;
+
+ ret = XFS_DIRENTER_SPACE_RES(mp, namelen);
+ if (xfs_has_parent(mp))
+ ret += xfs_parent_calc_space_res(mp, namelen);
+
+ return ret;
+}
diff --git a/libxfs/xfs_trans_space.h b/libxfs/xfs_trans_space.h
index 6cda87153b3..5539634009f 100644
--- a/libxfs/xfs_trans_space.h
+++ b/libxfs/xfs_trans_space.h
@@ -86,8 +86,6 @@
(2 * (mp)->m_alloc_maxlevels)
#define XFS_GROWFSRT_SPACE_RES(mp,b) \
((b) + XFS_EXTENTADD_SPACE_RES(mp, XFS_DATA_FORK))
-#define XFS_LINK_SPACE_RES(mp,nl) \
- XFS_DIRENTER_SPACE_RES(mp,nl)
#define XFS_QM_DQALLOC_SPACE_RES(mp) \
(XFS_EXTENTADD_SPACE_RES(mp, XFS_DATA_FORK) + \
XFS_DQUOT_CLUSTER_SIZE_FSB)
@@ -107,5 +105,6 @@ unsigned int xfs_parent_calc_space_res(struct xfs_mount *mp,
unsigned int xfs_create_space_res(struct xfs_mount *mp, unsigned int namelen);
unsigned int xfs_mkdir_space_res(struct xfs_mount *mp, unsigned int namelen);
+unsigned int xfs_link_space_res(struct xfs_mount *mp, unsigned int namelen);
#endif /* __XFS_TRANS_SPACE_H__ */
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 07/32] xfs: add parent attributes to symlink
2023-12-31 19:50 ` [PATCHSET v13.0 3/6] xfsprogs: Parent Pointers Darrick J. Wong
` (5 preceding siblings ...)
2023-12-31 23:06 ` [PATCH 06/32] xfs: add parent attributes to link Darrick J. Wong
@ 2023-12-31 23:06 ` Darrick J. Wong
2023-12-31 23:07 ` [PATCH 08/32] xfs: remove parent pointers in unlink Darrick J. Wong
` (24 subsequent siblings)
31 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:06 UTC (permalink / raw)
To: djwong, cem
Cc: Allison Henderson, catherine.hoang, linux-xfs, allison.henderson
From: Allison Henderson <allison.henderson@oracle.com>
This patch modifies xfs_symlink to add a parent pointer to the inode.
Signed-off-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
[djwong: minor rebase fixups]
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
libxfs/xfs_trans_space.c | 17 +++++++++++++++++
libxfs/xfs_trans_space.h | 4 ++--
2 files changed, 19 insertions(+), 2 deletions(-)
diff --git a/libxfs/xfs_trans_space.c b/libxfs/xfs_trans_space.c
index 039bbd91e87..bf4a41492c2 100644
--- a/libxfs/xfs_trans_space.c
+++ b/libxfs/xfs_trans_space.c
@@ -64,3 +64,20 @@ xfs_link_space_res(
return ret;
}
+
+unsigned int
+xfs_symlink_space_res(
+ struct xfs_mount *mp,
+ unsigned int namelen,
+ unsigned int fsblocks)
+{
+ unsigned int ret;
+
+ ret = XFS_IALLOC_SPACE_RES(mp) + XFS_DIRENTER_SPACE_RES(mp, namelen) +
+ fsblocks;
+
+ if (xfs_has_parent(mp))
+ ret += xfs_parent_calc_space_res(mp, namelen);
+
+ return ret;
+}
diff --git a/libxfs/xfs_trans_space.h b/libxfs/xfs_trans_space.h
index 5539634009f..354ad1d6e18 100644
--- a/libxfs/xfs_trans_space.h
+++ b/libxfs/xfs_trans_space.h
@@ -95,8 +95,6 @@
XFS_DIRREMOVE_SPACE_RES(mp)
#define XFS_RENAME_SPACE_RES(mp,nl) \
(XFS_DIRREMOVE_SPACE_RES(mp) + XFS_DIRENTER_SPACE_RES(mp,nl))
-#define XFS_SYMLINK_SPACE_RES(mp,nl,b) \
- (XFS_IALLOC_SPACE_RES(mp) + XFS_DIRENTER_SPACE_RES(mp,nl) + (b))
#define XFS_IFREE_SPACE_RES(mp) \
(xfs_has_finobt(mp) ? M_IGEO(mp)->inobt_maxlevels : 0)
@@ -106,5 +104,7 @@ unsigned int xfs_parent_calc_space_res(struct xfs_mount *mp,
unsigned int xfs_create_space_res(struct xfs_mount *mp, unsigned int namelen);
unsigned int xfs_mkdir_space_res(struct xfs_mount *mp, unsigned int namelen);
unsigned int xfs_link_space_res(struct xfs_mount *mp, unsigned int namelen);
+unsigned int xfs_symlink_space_res(struct xfs_mount *mp, unsigned int namelen,
+ unsigned int fsblocks);
#endif /* __XFS_TRANS_SPACE_H__ */
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 08/32] xfs: remove parent pointers in unlink
2023-12-31 19:50 ` [PATCHSET v13.0 3/6] xfsprogs: Parent Pointers Darrick J. Wong
` (6 preceding siblings ...)
2023-12-31 23:06 ` [PATCH 07/32] xfs: add parent attributes to symlink Darrick J. Wong
@ 2023-12-31 23:07 ` Darrick J. Wong
2023-12-31 23:07 ` [PATCH 09/32] xfs: Add parent pointers to rename Darrick J. Wong
` (23 subsequent siblings)
31 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:07 UTC (permalink / raw)
To: djwong, cem
Cc: Dave Chinner, Allison Henderson, catherine.hoang, linux-xfs,
allison.henderson
From: Allison Henderson <allison.henderson@oracle.com>
This patch removes the parent pointer attribute during unlink
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
[djwong: adjust to new ondisk format, minor rebase fixes]
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
libxfs/libxfs_api_defs.h | 1 +
libxfs/xfs_parent.c | 39 +++++++++++++++++++++++++++++++++++++++
libxfs/xfs_parent.h | 16 ++++++++++++++++
libxfs/xfs_trans_space.c | 13 +++++++++++++
libxfs/xfs_trans_space.h | 3 +--
repair/phase6.c | 6 +++---
6 files changed, 73 insertions(+), 5 deletions(-)
diff --git a/libxfs/libxfs_api_defs.h b/libxfs/libxfs_api_defs.h
index 6ab10be3ad6..1b69124767c 100644
--- a/libxfs/libxfs_api_defs.h
+++ b/libxfs/libxfs_api_defs.h
@@ -193,6 +193,7 @@
#define xfs_refcountbt_stage_cursor libxfs_refcountbt_stage_cursor
#define xfs_refcount_get_rec libxfs_refcount_get_rec
#define xfs_refcount_lookup_le libxfs_refcount_lookup_le
+#define xfs_remove_space_res libxfs_remove_space_res
#define xfs_rmap_alloc libxfs_rmap_alloc
#define xfs_rmapbt_calc_reserves libxfs_rmapbt_calc_reserves
diff --git a/libxfs/xfs_parent.c b/libxfs/xfs_parent.c
index 32e1d1f62ec..b3004af9161 100644
--- a/libxfs/xfs_parent.c
+++ b/libxfs/xfs_parent.c
@@ -28,6 +28,7 @@
#include "xfs_da_format.h"
#include "xfs_format.h"
#include "xfs_trans_space.h"
+#include "xfs_health.h"
struct kmem_cache *xfs_parent_args_cache;
@@ -195,6 +196,44 @@ xfs_parent_addname(
return 0;
}
+/* Remove a parent pointer to reflect a dirent removal. */
+int
+xfs_parent_removename(
+ struct xfs_trans *tp,
+ struct xfs_parent_args *ppargs,
+ struct xfs_inode *dp,
+ const struct xfs_name *parent_name,
+ struct xfs_inode *child)
+{
+ struct xfs_da_args *args = &ppargs->args;
+
+ if (XFS_IS_CORRUPT(tp->t_mountp,
+ !xfs_parent_valuecheck(tp->t_mountp, parent_name->name,
+ parent_name->len)))
+ return -EFSCORRUPTED;
+
+ /*
+ * For regular attrs, removing an attr from a !hasattr inode is a nop.
+ * For parent pointers, we require that the pointer must exist if the
+ * caller wants us to remove the pointer.
+ */
+ if (XFS_IS_CORRUPT(child->i_mount, !xfs_inode_hasattr(child))) {
+ xfs_inode_mark_sick(child, XFS_SICK_INO_PARENT);
+ return -EFSCORRUPTED;
+ }
+
+ xfs_init_parent_name_rec(&ppargs->rec, dp, parent_name, child);
+ args->hashval = xfs_parent_hashname(dp, ppargs);
+
+ args->trans = tp;
+ args->dp = child;
+
+ xfs_init_parent_davalue(&ppargs->args, parent_name);
+
+ xfs_attr_defer_add(args, XFS_ATTRI_OP_FLAGS_REMOVE);
+ return 0;
+}
+
/* Free a parent pointer context object. */
void
xfs_parent_args_free(
diff --git a/libxfs/xfs_parent.h b/libxfs/xfs_parent.h
index e2115a2b964..31349130a33 100644
--- a/libxfs/xfs_parent.h
+++ b/libxfs/xfs_parent.h
@@ -79,6 +79,22 @@ xfs_parent_add(struct xfs_trans *tp, struct xfs_parent_args *ppargs,
return 0;
}
+int xfs_parent_removename(struct xfs_trans *tp, struct xfs_parent_args *ppargs,
+ struct xfs_inode *dp, const struct xfs_name *parent_name,
+ struct xfs_inode *child);
+
+/* Schedule a parent pointer removal. */
+static inline int
+xfs_parent_remove(struct xfs_trans *tp, struct xfs_parent_args *ppargs,
+ struct xfs_inode *dp, const struct xfs_name *parent_name,
+ struct xfs_inode *child)
+{
+ if (ppargs)
+ return xfs_parent_removename(tp, ppargs, dp, parent_name,
+ child);
+ return 0;
+}
+
void xfs_parent_args_free(struct xfs_mount *mp, struct xfs_parent_args *ppargs);
/* Finish a parent pointer update by freeing the context object. */
diff --git a/libxfs/xfs_trans_space.c b/libxfs/xfs_trans_space.c
index bf4a41492c2..86a91a3a844 100644
--- a/libxfs/xfs_trans_space.c
+++ b/libxfs/xfs_trans_space.c
@@ -81,3 +81,16 @@ xfs_symlink_space_res(
return ret;
}
+
+unsigned int
+xfs_remove_space_res(
+ struct xfs_mount *mp,
+ unsigned int namelen)
+{
+ unsigned int ret = XFS_DIRREMOVE_SPACE_RES(mp);
+
+ if (xfs_has_parent(mp))
+ ret += xfs_parent_calc_space_res(mp, namelen);
+
+ return ret;
+}
diff --git a/libxfs/xfs_trans_space.h b/libxfs/xfs_trans_space.h
index 354ad1d6e18..a4490813c56 100644
--- a/libxfs/xfs_trans_space.h
+++ b/libxfs/xfs_trans_space.h
@@ -91,8 +91,6 @@
XFS_DQUOT_CLUSTER_SIZE_FSB)
#define XFS_QM_QINOCREATE_SPACE_RES(mp) \
XFS_IALLOC_SPACE_RES(mp)
-#define XFS_REMOVE_SPACE_RES(mp) \
- XFS_DIRREMOVE_SPACE_RES(mp)
#define XFS_RENAME_SPACE_RES(mp,nl) \
(XFS_DIRREMOVE_SPACE_RES(mp) + XFS_DIRENTER_SPACE_RES(mp,nl))
#define XFS_IFREE_SPACE_RES(mp) \
@@ -106,5 +104,6 @@ unsigned int xfs_mkdir_space_res(struct xfs_mount *mp, unsigned int namelen);
unsigned int xfs_link_space_res(struct xfs_mount *mp, unsigned int namelen);
unsigned int xfs_symlink_space_res(struct xfs_mount *mp, unsigned int namelen,
unsigned int fsblocks);
+unsigned int xfs_remove_space_res(struct xfs_mount *mp, unsigned int namelen);
#endif /* __XFS_TRANS_SPACE_H__ */
diff --git a/repair/phase6.c b/repair/phase6.c
index 825f0cf3956..9b43e58b3d3 100644
--- a/repair/phase6.c
+++ b/repair/phase6.c
@@ -1275,7 +1275,7 @@ longform_dir2_rebuild(
libxfs_dir_ino_validate(mp, pip.i_ino))
pip.i_ino = mp->m_sb.sb_rootino;
- nres = XFS_REMOVE_SPACE_RES(mp);
+ nres = libxfs_remove_space_res(mp, 0);
error = -libxfs_trans_alloc(mp, &M_RES(mp)->tr_remove, nres, 0, 0, &tp);
if (error)
res_failed(error);
@@ -1381,7 +1381,7 @@ dir2_kill_block(
int nres;
xfs_trans_t *tp;
- nres = XFS_REMOVE_SPACE_RES(mp);
+ nres = libxfs_remove_space_res(mp, 0);
error = -libxfs_trans_alloc(mp, &M_RES(mp)->tr_remove, nres, 0, 0, &tp);
if (error)
res_failed(error);
@@ -2900,7 +2900,7 @@ process_dir_inode(
* inode but it's easier than wedging a
* new define in ourselves.
*/
- nres = no_modify ? 0 : XFS_REMOVE_SPACE_RES(mp);
+ nres = no_modify ? 0 : libxfs_remove_space_res(mp, 0);
error = -libxfs_trans_alloc(mp, &M_RES(mp)->tr_remove,
nres, 0, 0, &tp);
if (error)
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 09/32] xfs: Add parent pointers to rename
2023-12-31 19:50 ` [PATCHSET v13.0 3/6] xfsprogs: Parent Pointers Darrick J. Wong
` (7 preceding siblings ...)
2023-12-31 23:07 ` [PATCH 08/32] xfs: remove parent pointers in unlink Darrick J. Wong
@ 2023-12-31 23:07 ` Darrick J. Wong
2023-12-31 23:07 ` [PATCH 10/32] xfs: pass the attr value to put_listent when possible Darrick J. Wong
` (22 subsequent siblings)
31 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:07 UTC (permalink / raw)
To: djwong, cem
Cc: Allison Henderson, catherine.hoang, linux-xfs, allison.henderson
From: Allison Henderson <allison.henderson@oracle.com>
This patch removes the old parent pointer attribute during the rename
operation, and re-adds the updated parent pointer.
Signed-off-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
[djwong: adjust to new ondisk format]
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
libxfs/xfs_parent.c | 63 ++++++++++++++++++++++++++++++++++++++++++++++
libxfs/xfs_parent.h | 20 +++++++++++++++
libxfs/xfs_trans_space.c | 25 ++++++++++++++++++
libxfs/xfs_trans_space.h | 6 +++-
4 files changed, 112 insertions(+), 2 deletions(-)
diff --git a/libxfs/xfs_parent.c b/libxfs/xfs_parent.c
index b3004af9161..6c98f95f274 100644
--- a/libxfs/xfs_parent.c
+++ b/libxfs/xfs_parent.c
@@ -139,6 +139,19 @@ xfs_init_parent_davalue(
args->value = (void *)name->name;
}
+/*
+ * Point the da args new value fields at the non-key parts of a replacement
+ * parent pointer.
+ */
+static inline void
+xfs_init_parent_danewvalue(
+ struct xfs_da_args *args,
+ const struct xfs_name *name)
+{
+ args->new_valuelen = name->len;
+ args->new_value = (void *)name->name;
+}
+
/*
* Allocate memory to control a logged parent pointer update as part of a
* dirent operation.
@@ -234,6 +247,56 @@ xfs_parent_removename(
return 0;
}
+/* Replace one parent pointer with another to reflect a rename. */
+int
+xfs_parent_replacename(
+ struct xfs_trans *tp,
+ struct xfs_parent_args *ppargs,
+ struct xfs_inode *old_dp,
+ const struct xfs_name *old_name,
+ struct xfs_inode *new_dp,
+ const struct xfs_name *new_name,
+ struct xfs_inode *child)
+{
+ struct xfs_da_args *args = &ppargs->args;
+
+ if (XFS_IS_CORRUPT(tp->t_mountp,
+ !xfs_parent_valuecheck(tp->t_mountp, old_name->name,
+ old_name->len)))
+ return -EFSCORRUPTED;
+
+ if (XFS_IS_CORRUPT(tp->t_mountp,
+ !xfs_parent_valuecheck(tp->t_mountp, new_name->name,
+ new_name->len)))
+ return -EFSCORRUPTED;
+
+ /*
+ * For regular attrs, replacing an attr from a !hasattr inode becomes
+ * an attr-set operation. For replacing a parent pointer, however, we
+ * require that the old pointer must exist.
+ */
+ if (XFS_IS_CORRUPT(child->i_mount, !xfs_inode_hasattr(child))) {
+ xfs_inode_mark_sick(child, XFS_SICK_INO_PARENT);
+ return -EFSCORRUPTED;
+ }
+
+ xfs_init_parent_name_rec(&ppargs->rec, old_dp, old_name, child);
+ args->hashval = xfs_parent_hashname(old_dp, ppargs);
+
+ xfs_init_parent_name_rec(&ppargs->new_rec, new_dp, new_name, child);
+ args->new_name = (const uint8_t *)&ppargs->new_rec;
+ args->new_namelen = sizeof(struct xfs_parent_name_rec);
+
+ args->trans = tp;
+ args->dp = child;
+
+ xfs_init_parent_davalue(&ppargs->args, old_name);
+ xfs_init_parent_danewvalue(&ppargs->args, new_name);
+
+ xfs_attr_defer_add(args, XFS_ATTRI_OP_FLAGS_REPLACE);
+ return 0;
+}
+
/* Free a parent pointer context object. */
void
xfs_parent_args_free(
diff --git a/libxfs/xfs_parent.h b/libxfs/xfs_parent.h
index 31349130a33..c68c501388e 100644
--- a/libxfs/xfs_parent.h
+++ b/libxfs/xfs_parent.h
@@ -24,6 +24,7 @@ extern struct kmem_cache *xfs_parent_args_cache;
*/
struct xfs_parent_args {
struct xfs_parent_name_rec rec;
+ struct xfs_parent_name_rec new_rec;
struct xfs_da_args args;
};
@@ -95,6 +96,25 @@ xfs_parent_remove(struct xfs_trans *tp, struct xfs_parent_args *ppargs,
return 0;
}
+int xfs_parent_replacename(struct xfs_trans *tp,
+ struct xfs_parent_args *ppargs,
+ struct xfs_inode *old_dp, const struct xfs_name *old_name,
+ struct xfs_inode *new_dp, const struct xfs_name *new_name,
+ struct xfs_inode *child);
+
+/* Schedule a parent pointer replacement. */
+static inline int
+xfs_parent_replace(struct xfs_trans *tp, struct xfs_parent_args *ppargs,
+ struct xfs_inode *old_dp, const struct xfs_name *old_name,
+ struct xfs_inode *new_dp, const struct xfs_name *new_name,
+ struct xfs_inode *child)
+{
+ if (ppargs)
+ return xfs_parent_replacename(tp, ppargs, old_dp, old_name,
+ new_dp, new_name, child);
+ return 0;
+}
+
void xfs_parent_args_free(struct xfs_mount *mp, struct xfs_parent_args *ppargs);
/* Finish a parent pointer update by freeing the context object. */
diff --git a/libxfs/xfs_trans_space.c b/libxfs/xfs_trans_space.c
index 86a91a3a844..373f5cc2497 100644
--- a/libxfs/xfs_trans_space.c
+++ b/libxfs/xfs_trans_space.c
@@ -94,3 +94,28 @@ xfs_remove_space_res(
return ret;
}
+
+unsigned int
+xfs_rename_space_res(
+ struct xfs_mount *mp,
+ unsigned int src_namelen,
+ bool target_exists,
+ unsigned int target_namelen,
+ bool has_whiteout)
+{
+ unsigned int ret;
+
+ ret = XFS_DIRREMOVE_SPACE_RES(mp) +
+ XFS_DIRENTER_SPACE_RES(mp, target_namelen);
+
+ if (xfs_has_parent(mp)) {
+ if (has_whiteout)
+ ret += xfs_parent_calc_space_res(mp, src_namelen);
+ ret += 2 * xfs_parent_calc_space_res(mp, target_namelen);
+ }
+
+ if (target_exists)
+ ret += xfs_parent_calc_space_res(mp, target_namelen);
+
+ return ret;
+}
diff --git a/libxfs/xfs_trans_space.h b/libxfs/xfs_trans_space.h
index a4490813c56..1155ff2d37e 100644
--- a/libxfs/xfs_trans_space.h
+++ b/libxfs/xfs_trans_space.h
@@ -91,8 +91,6 @@
XFS_DQUOT_CLUSTER_SIZE_FSB)
#define XFS_QM_QINOCREATE_SPACE_RES(mp) \
XFS_IALLOC_SPACE_RES(mp)
-#define XFS_RENAME_SPACE_RES(mp,nl) \
- (XFS_DIRREMOVE_SPACE_RES(mp) + XFS_DIRENTER_SPACE_RES(mp,nl))
#define XFS_IFREE_SPACE_RES(mp) \
(xfs_has_finobt(mp) ? M_IGEO(mp)->inobt_maxlevels : 0)
@@ -106,4 +104,8 @@ unsigned int xfs_symlink_space_res(struct xfs_mount *mp, unsigned int namelen,
unsigned int fsblocks);
unsigned int xfs_remove_space_res(struct xfs_mount *mp, unsigned int namelen);
+unsigned int xfs_rename_space_res(struct xfs_mount *mp,
+ unsigned int src_namelen, bool target_exists,
+ unsigned int target_namelen, bool has_whiteout);
+
#endif /* __XFS_TRANS_SPACE_H__ */
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 10/32] xfs: pass the attr value to put_listent when possible
2023-12-31 19:50 ` [PATCHSET v13.0 3/6] xfsprogs: Parent Pointers Darrick J. Wong
` (8 preceding siblings ...)
2023-12-31 23:07 ` [PATCH 09/32] xfs: Add parent pointers to rename Darrick J. Wong
@ 2023-12-31 23:07 ` Darrick J. Wong
2023-12-31 23:07 ` [PATCH 11/32] xfs: Add parent pointer ioctl Darrick J. Wong
` (21 subsequent siblings)
31 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:07 UTC (permalink / raw)
To: djwong, cem
Cc: Allison Henderson, catherine.hoang, linux-xfs, allison.henderson
From: Allison Henderson <allison.henderson@oracle.com>
Pass the attr value to put_listent when we have local xattrs or
shortform xattrs.
Signed-off-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
libxfs/xfs_attr.h | 5 +++--
libxfs/xfs_attr_sf.h | 1 +
2 files changed, 4 insertions(+), 2 deletions(-)
diff --git a/libxfs/xfs_attr.h b/libxfs/xfs_attr.h
index 4a4d45a96dd..0204f62298c 100644
--- a/libxfs/xfs_attr.h
+++ b/libxfs/xfs_attr.h
@@ -47,8 +47,9 @@ struct xfs_attrlist_cursor_kern {
/* void; state communicated via *context */
-typedef void (*put_listent_func_t)(struct xfs_attr_list_context *, int,
- unsigned char *, int, int);
+typedef void (*put_listent_func_t)(struct xfs_attr_list_context *context,
+ int flags, unsigned char *name, int namelen, void *value,
+ int valuelen);
struct xfs_attr_list_context {
struct xfs_trans *tp;
diff --git a/libxfs/xfs_attr_sf.h b/libxfs/xfs_attr_sf.h
index 37578b369d9..c6e259791bc 100644
--- a/libxfs/xfs_attr_sf.h
+++ b/libxfs/xfs_attr_sf.h
@@ -24,6 +24,7 @@ typedef struct xfs_attr_sf_sort {
uint8_t flags; /* flags bits (see xfs_attr_leaf.h) */
xfs_dahash_t hash; /* this entry's hash value */
unsigned char *name; /* name value, pointer into buffer */
+ void *value;
} xfs_attr_sf_sort_t;
#define XFS_ATTR_SF_ENTSIZE_MAX /* max space for name&value */ \
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 11/32] xfs: Add parent pointer ioctl
2023-12-31 19:50 ` [PATCHSET v13.0 3/6] xfsprogs: Parent Pointers Darrick J. Wong
` (9 preceding siblings ...)
2023-12-31 23:07 ` [PATCH 10/32] xfs: pass the attr value to put_listent when possible Darrick J. Wong
@ 2023-12-31 23:07 ` Darrick J. Wong
2023-12-31 23:08 ` [PATCH 12/32] xfs: fix unit conversion error in xfs_log_calc_max_attrsetm_res Darrick J. Wong
` (20 subsequent siblings)
31 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:07 UTC (permalink / raw)
To: djwong, cem
Cc: Allison Henderson, catherine.hoang, linux-xfs, allison.henderson
From: Allison Henderson <allison.henderson@oracle.com>
This patch adds a new file ioctl to retrieve the parent pointer of a
given inode
Signed-off-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
[djwong: move new ioctl to xfs_fs_staging.h, adjust to new ondisk format]
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
libxfs/xfs_fs.h | 1
libxfs/xfs_fs_staging.h | 66 +++++++++++
libxfs/xfs_ondisk.h | 4 +
libxfs/xfs_parent.c | 62 +++++++++++
libxfs/xfs_parent.h | 25 ++++
man/man2/ioctl_xfs_getparents.2 | 227 +++++++++++++++++++++++++++++++++++++++
6 files changed, 385 insertions(+)
create mode 100644 man/man2/ioctl_xfs_getparents.2
diff --git a/libxfs/xfs_fs.h b/libxfs/xfs_fs.h
index 63a145e5035..e92b6a9612a 100644
--- a/libxfs/xfs_fs.h
+++ b/libxfs/xfs_fs.h
@@ -811,6 +811,7 @@ struct xfs_scrub_metadata {
/* XFS_IOC_GETFSMAP ------ hoisted 59 */
#define XFS_IOC_SCRUB_METADATA _IOWR('X', 60, struct xfs_scrub_metadata)
#define XFS_IOC_AG_GEOMETRY _IOWR('X', 61, struct xfs_ag_geometry)
+/* XFS_IOC_GETPARENTS ---- staging 62 */
/*
* ioctl commands that replace IRIX syssgi()'s
diff --git a/libxfs/xfs_fs_staging.h b/libxfs/xfs_fs_staging.h
index e3d9f3b32b0..e0650af0558 100644
--- a/libxfs/xfs_fs_staging.h
+++ b/libxfs/xfs_fs_staging.h
@@ -104,4 +104,70 @@ struct xfs_exch_range {
#define XFS_IOC_EXCHANGE_RANGE _IOWR('X', 129, struct xfs_exch_range)
+/* Iterating parent pointers of files. */
+
+/* return parents of the handle, not the open fd */
+#define XFS_GETPARENTS_IFLAG_HANDLE (1U << 0)
+
+/* target was the root directory */
+#define XFS_GETPARENTS_OFLAG_ROOT (1U << 1)
+
+/* Cursor is done iterating pptrs */
+#define XFS_GETPARENTS_OFLAG_DONE (1U << 2)
+
+#define XFS_GETPARENTS_FLAG_ALL (XFS_GETPARENTS_IFLAG_HANDLE | \
+ XFS_GETPARENTS_OFLAG_ROOT | \
+ XFS_GETPARENTS_OFLAG_DONE)
+
+/* Get an inode parent pointer through ioctl */
+struct xfs_getparents_rec {
+ __u64 gpr_ino; /* Inode number */
+ __u32 gpr_gen; /* Inode generation */
+ __u32 gpr_pad; /* Reserved */
+ __u64 gpr_rsvd; /* Reserved */
+ __u8 gpr_name[]; /* File name and null terminator */
+};
+
+/* Iterate through an inodes parent pointers */
+struct xfs_getparents {
+ /* File handle, if XFS_GETPARENTS_IFLAG_HANDLE is set */
+ struct xfs_handle gp_handle;
+
+ /*
+ * Structure to track progress in iterating the parent pointers.
+ * Must be initialized to zeroes before the first ioctl call, and
+ * not touched by callers after that.
+ */
+ struct xfs_attrlist_cursor gp_cursor;
+
+ /* Operational flags: XFS_GETPARENTS_*FLAG* */
+ __u32 gp_flags;
+
+ /* Must be set to zero */
+ __u32 gp_reserved;
+
+ /* Size of the buffer in bytes, including this header */
+ __u32 gp_bufsize;
+
+ /* # of entries filled in (output) */
+ __u32 gp_count;
+
+ /* Must be set to zero */
+ __u64 gp_reserved2[5];
+
+ /* Byte offset of each record within the buffer */
+ __u32 gp_offsets[];
+};
+
+static inline struct xfs_getparents_rec*
+xfs_getparents_rec(
+ struct xfs_getparents *info,
+ unsigned int idx)
+{
+ return (struct xfs_getparents_rec *)((char *)info +
+ info->gp_offsets[idx]);
+}
+
+#define XFS_IOC_GETPARENTS _IOWR('X', 62, struct xfs_getparents)
+
#endif /* __XFS_FS_STAGING_H__ */
diff --git a/libxfs/xfs_ondisk.h b/libxfs/xfs_ondisk.h
index d9c988c5ad6..bffd39242d4 100644
--- a/libxfs/xfs_ondisk.h
+++ b/libxfs/xfs_ondisk.h
@@ -155,6 +155,10 @@ xfs_check_ondisk_structs(void)
XFS_CHECK_OFFSET(struct xfs_efi_log_format_32, efi_extents, 16);
XFS_CHECK_OFFSET(struct xfs_efi_log_format_64, efi_extents, 16);
+ /* parent pointer ioctls */
+ XFS_CHECK_STRUCT_SIZE(struct xfs_getparents_rec, 24);
+ XFS_CHECK_STRUCT_SIZE(struct xfs_getparents, 96);
+
/*
* The v5 superblock format extended several v4 header structures with
* additional data. While new fields are only accessible on v5
diff --git a/libxfs/xfs_parent.c b/libxfs/xfs_parent.c
index 6c98f95f274..92b541737cb 100644
--- a/libxfs/xfs_parent.c
+++ b/libxfs/xfs_parent.c
@@ -305,3 +305,65 @@ xfs_parent_args_free(
{
kmem_cache_free(xfs_parent_args_cache, ppargs);
}
+
+/* Convert an ondisk parent pointer to the incore format. */
+void
+xfs_parent_irec_from_disk(
+ struct xfs_parent_name_irec *irec,
+ const struct xfs_parent_name_rec *rec,
+ const void *value,
+ unsigned int valuelen)
+{
+ irec->p_ino = be64_to_cpu(rec->p_ino);
+ irec->p_gen = be32_to_cpu(rec->p_gen);
+ irec->p_namehash = be32_to_cpu(rec->p_namehash);
+ irec->p_namelen = valuelen;
+ memcpy(irec->p_name, value, valuelen);
+}
+
+/* Convert an incore parent pointer to the ondisk attr name format. */
+void
+xfs_parent_irec_to_disk(
+ struct xfs_parent_name_rec *rec,
+ const struct xfs_parent_name_irec *irec)
+{
+ rec->p_ino = cpu_to_be64(irec->p_ino);
+ rec->p_gen = cpu_to_be32(irec->p_gen);
+ rec->p_namehash = cpu_to_be32(irec->p_namehash);
+}
+
+/* Is this a valid incore parent pointer? */
+bool
+xfs_parent_verify_irec(
+ struct xfs_mount *mp,
+ const struct xfs_parent_name_irec *irec)
+{
+ struct xfs_name dname = {
+ .name = irec->p_name,
+ .len = irec->p_namelen,
+ };
+
+ if (!xfs_verify_dir_ino(mp, irec->p_ino))
+ return false;
+ if (!xfs_parent_valuecheck(mp, irec->p_name, irec->p_namelen))
+ return false;
+ if (!xfs_dir2_namecheck(irec->p_name, irec->p_namelen))
+ return false;
+ if (irec->p_namehash != xfs_dir2_hashname(mp, &dname))
+ return false;
+ return true;
+}
+
+/* Compute p_namehash for the this parent pointer. */
+void
+xfs_parent_irec_hashname(
+ struct xfs_mount *mp,
+ struct xfs_parent_name_irec *irec)
+{
+ struct xfs_name dname = {
+ .name = irec->p_name,
+ .len = irec->p_namelen,
+ };
+
+ irec->p_namehash = xfs_dir2_hashname(mp, &dname);
+}
diff --git a/libxfs/xfs_parent.h b/libxfs/xfs_parent.h
index c68c501388e..e43ae5a7df8 100644
--- a/libxfs/xfs_parent.h
+++ b/libxfs/xfs_parent.h
@@ -127,4 +127,29 @@ xfs_parent_finish(
xfs_parent_args_free(mp, ppargs);
}
+/*
+ * Incore version of a parent pointer, also contains dirent name so callers
+ * can pass/obtain all the parent pointer information in a single structure
+ */
+struct xfs_parent_name_irec {
+ /* Parent pointer attribute name fields */
+ xfs_ino_t p_ino;
+ uint32_t p_gen;
+ xfs_dahash_t p_namehash;
+
+ /* Parent pointer attribute value fields */
+ uint8_t p_namelen;
+ unsigned char p_name[MAXNAMELEN];
+};
+
+void xfs_parent_irec_from_disk(struct xfs_parent_name_irec *irec,
+ const struct xfs_parent_name_rec *rec, const void *value,
+ unsigned int valuelen);
+void xfs_parent_irec_to_disk(struct xfs_parent_name_rec *rec,
+ const struct xfs_parent_name_irec *irec);
+void xfs_parent_irec_hashname(struct xfs_mount *mp,
+ struct xfs_parent_name_irec *irec);
+bool xfs_parent_verify_irec(struct xfs_mount *mp,
+ const struct xfs_parent_name_irec *irec);
+
#endif /* __XFS_PARENT_H__ */
diff --git a/man/man2/ioctl_xfs_getparents.2 b/man/man2/ioctl_xfs_getparents.2
new file mode 100644
index 00000000000..d6987979e85
--- /dev/null
+++ b/man/man2/ioctl_xfs_getparents.2
@@ -0,0 +1,227 @@
+.\" Copyright (c) 2019-2024 Oracle. All rights reserved.
+.\"
+.\" %%%LICENSE_START(GPLv2+_DOC_FULL)
+.\" SPDX-License-Identifier: GPL-2.0-or-later
+.\" %%%LICENSE_END
+.TH IOCTL-XFS-GETPARENTS 2 2023-08-18 "XFS"
+.SH NAME
+ioctl_xfs_getparents \- query XFS directory parent information
+.SH SYNOPSIS
+.br
+.B #include <xfs/xfs_fs.h>
+.br
+.B #include <xfs/xfs_fs_staging.h>
+.PP
+.BI "int ioctl(int " fd ", XFS_IOC_GETPARENTS, struct xfs_getparents *" arg );
+.SH DESCRIPTION
+This command is used to get a file's parent pointers.
+Parent pointers point upwards in the directory tree from a child file towards a
+parent directories.
+Each entry in a parent directory must have a corresponding parent pointer in
+the child.
+
+Calling programs should allocate a large memory buffer and initialize the
+beginning of the buffer to a header of the following form:
+.PP
+.in +4n
+.nf
+struct xfs_getparents {
+ struct xfs_handle gp_handle;
+ struct xfs_attrlist_cursor gp_cursor;
+ __u32 gp_flags;
+ __u32 gp_reserved;
+ __u32 gp_bufsize;
+ __u32 gp_count;
+ __u64 gp_reserved2[5];
+ __u32 gp_offsets[];
+};
+.fi
+.in
+
+.PP
+If the flag
+.B XFS_GETPARENTS_IFLAG_HANDLE
+is set,
+the field
+.I gp_handle
+will be interpreted as if it contains file handle information.
+If the file handle is not stale, the file represented by the handle will be the
+target of the query.
+If the flag is not set, the file represented by
+.I fd
+will be queried instead.
+
+.PP
+The field
+.I gp_cursor
+tracks the progress of iterating through the parent pointers.
+Calling programs must initialize this to zero before the first system call
+and must not touch it after that.
+
+.PP
+The field
+.I gp_flags
+control the behavior of the query operation and provide more information
+about the outcome of the operation.
+.RS 0.4i
+.TP
+.B XFS_GETPARENTS_IFLAG_HANDLE
+If the caller sets this flag, the kernel driver will interpret the
+.I gp_handle
+field as if it were a file handle.
+If the handle maps to an allocated file, that file will be queried for
+parents instead of the open file descriptor.
+.TP
+.B XFS_GETPARENTS_OFLAG_ROOT
+The file queried was the root directory.
+.TP
+.B XFS_GETPARENTS_OFLAG_DONE
+There are no more parent pointers to query.
+.RE
+
+.PP
+The fields
+.I gp_reserved
+and
+.I gp_reserved2
+must be zero.
+
+.PP
+The field
+.I gp_bufsize
+should be set to the size of the buffer, in bytes.
+
+.PP
+The field
+.I gp_count
+will be set to the number of parent pointer records returned.
+
+.PP
+Each element of the array
+.I gp_offsets
+will be set to the byte offset within the buffer of each parent record.
+
+Parent pointer records are returned in the following form:
+.PP
+.in +4n
+.nf
+
+struct xfs_getparents_rec {
+ __u64 gpr_ino;
+ __u32 gpr_gen;
+ __u32 gpr_pad;
+ __u64 gpr_rsvd;
+ __u8 gpr_name[];
+};
+.fi
+.in
+
+.PP
+The field
+.I gpr_ino
+and
+.I gpr_gen
+will be set to the inode number and generation number of the parent.
+
+.PP
+The fields
+.I gpr_pad
+and
+.I gpr_rsvd
+will be set to zero.
+
+.PP
+The array
+.I gpr_name
+will be set to a NULL-terminated byte sequence representing the filename
+stored in the parent pointer.
+
+.SH SAMPLE PROGRAM
+Calling programs should allocate a large memory buffer, initialize the head
+structure to zeroes, set gp_bufsize to the size of the buffer, and call the
+ioctl.
+The kernel will fill out the gp_offsets array with integer offsets to
+struct xfs_getparents_rec objects that are written within the provided memory
+buffer.
+The size of the gp_offsets array is given by gp_count.
+The XFS_GETPARENTS_OFLAG_DONE flag will be set in gp_flags when there are no
+more parent pointers to be read.
+The below code is an example of XFS_IOC_GETPARENTS usage:
+
+.nf
+#include <stdio.h>
+#include <string.h>
+#include <errno.h>
+#include <xfs/linux.h>
+#include <xfs/xfs.h>
+#include <xfs/xfs_types.h>
+#include <xfs/xfs_fs.h>
+#include <xfs/xfs_fs_staging.h>
+
+int main() {
+ struct xfs_getparents *pi;
+ struct xfs_getparents_rec *p;
+ int i, error, fd, nr_ptrs = 4;
+
+ error = malloc(65536);
+ if (!error) {
+ perror("malloc");
+ return 1;
+ }
+
+ memset(pi, 0, sizeof(*pi));
+ pi->gp_bufsize = 65536;
+
+ fd = open("/mnt/test/foo.txt", O_RDONLY | O_CREAT);
+ if (fd == -1)
+ return errno;
+
+ do {
+ error = ioctl(fd, XFS_IOC_GETPARENTS, pi);
+ if (error)
+ return error;
+
+ for (i = 0; i < pi->gp_count; i++) {
+ p = xfs_getparents_rec(pi, i);
+ printf("inode = %llu\\n", (unsigned long long)p->gpr_ino);
+ printf("generation = %u\\n", (unsigned int)p->gpr_gen);
+ printf("name = \\"%s\\"\\n\\n", (char *)p->gpr_name);
+ }
+ } while (!(pi->gp_flags & XFS_GETPARENTS_OFLAG_DONE));
+
+ return 0;
+}
+.fi
+
+.SH RETURN VALUE
+On error, \-1 is returned, and
+.I errno
+is set to indicate the error.
+.PP
+.SH ERRORS
+Error codes can be one of, but are not limited to, the following:
+.TP
+.B EFSBADCRC
+Metadata checksum validation failed while performing the query.
+.TP
+.B EFSCORRUPTED
+Metadata corruption was encountered while performing the query.
+.TP
+.B EINVAL
+One or more of the arguments specified is invalid.
+.TP
+.B EOPNOTSUPP
+Repairs of the requested metadata object are not supported.
+.TP
+.B EROFS
+Filesystem is read-only and a repair was requested.
+.TP
+.B ESHUTDOWN
+Filesystem is shut down due to previous errors.
+.TP
+.B EIO
+An I/O error was encountered while performing the query.
+.SH CONFORMING TO
+This API is specific to XFS filesystem on the Linux kernel.
+.SH SEE ALSO
+.BR ioctl (2)
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 12/32] xfs: fix unit conversion error in xfs_log_calc_max_attrsetm_res
2023-12-31 19:50 ` [PATCHSET v13.0 3/6] xfsprogs: Parent Pointers Darrick J. Wong
` (10 preceding siblings ...)
2023-12-31 23:07 ` [PATCH 11/32] xfs: Add parent pointer ioctl Darrick J. Wong
@ 2023-12-31 23:08 ` Darrick J. Wong
2023-12-31 23:08 ` [PATCH 13/32] xfs: drop compatibility minimum log size computations for reflink Darrick J. Wong
` (19 subsequent siblings)
31 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:08 UTC (permalink / raw)
To: djwong, cem
Cc: Allison Henderson, catherine.hoang, linux-xfs, allison.henderson
From: Darrick J. Wong <djwong@kernel.org>
Dave and I were discussing some recent test regressions as a result of
me turning on nrext64=1 on realtime filesystems, when we noticed that
the minimum log size of a 32M filesystem jumped from 954 blocks to 4287
blocks.
Digging through xfs_log_calc_max_attrsetm_res, Dave noticed that @size
contains the maximum estimated amount of space needed for a local format
xattr, in bytes, but we feed this quantity to XFS_NEXTENTADD_SPACE_RES,
which requires units of blocks. This has resulted in an overestimation
of the minimum log size over the years.
We should nominally correct this, but there's a backwards compatibility
problem -- if we enable it now, the minimum log size will decrease. If
a corrected mkfs formats a filesystem with this new smaller log size, a
user will encounter mount failures on an uncorrected kernel due to the
larger minimum log size computations there.
However, the large extent counters feature is still EXPERIMENTAL, so we
can gate the correction on that feature (or any features that get added
after that) being enabled. Any filesystem with nrext64 or any of the
as-yet-undefined feature bits turned on will be rejected by old
uncorrected kernels, so this should be safe even in the upgrade case.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
---
libxfs/xfs_log_rlimit.c | 33 +++++++++++++++++++++++++++++++++
1 file changed, 33 insertions(+)
diff --git a/libxfs/xfs_log_rlimit.c b/libxfs/xfs_log_rlimit.c
index cba24493f86..096e3ee4b6a 100644
--- a/libxfs/xfs_log_rlimit.c
+++ b/libxfs/xfs_log_rlimit.c
@@ -16,6 +16,29 @@
#include "xfs_bmap_btree.h"
#include "xfs_trace.h"
+/*
+ * Decide if the filesystem has the parent pointer feature or any feature
+ * added after that. If so, we can improve the accuracy of the transaction
+ * reservation computations that should lead to more efficient log grant use.
+ */
+static inline bool
+xfs_has_parent_or_newer_feature(
+ struct xfs_mount *mp)
+{
+ if (!xfs_sb_is_v5(&mp->m_sb))
+ return false;
+
+ if (xfs_sb_has_incompat_feature(&mp->m_sb,
+ ~(XFS_SB_FEAT_INCOMPAT_FTYPE |
+ XFS_SB_FEAT_INCOMPAT_SPINODES |
+ XFS_SB_FEAT_INCOMPAT_META_UUID |
+ XFS_SB_FEAT_INCOMPAT_BIGTIME |
+ XFS_SB_FEAT_INCOMPAT_NREXT64)))
+ return true;
+
+ return false;
+}
+
/*
* Calculate the maximum length in bytes that would be required for a local
* attribute value as large attributes out of line are not logged.
@@ -31,6 +54,16 @@ xfs_log_calc_max_attrsetm_res(
MAXNAMELEN - 1;
nblks = XFS_DAENTER_SPACE_RES(mp, XFS_ATTR_FORK);
nblks += XFS_B_TO_FSB(mp, size);
+
+ /*
+ * Starting with the parent pointer feature, every new fs feature
+ * corrects a unit conversion error in the xattr transaction
+ * reservation code that resulted in oversized minimum log size
+ * computations.
+ */
+ if (xfs_has_parent_or_newer_feature(mp))
+ size = XFS_B_TO_FSB(mp, size);
+
nblks += XFS_NEXTENTADD_SPACE_RES(mp, size, XFS_ATTR_FORK);
return M_RES(mp)->tr_attrsetm.tr_logres +
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 13/32] xfs: drop compatibility minimum log size computations for reflink
2023-12-31 19:50 ` [PATCHSET v13.0 3/6] xfsprogs: Parent Pointers Darrick J. Wong
` (11 preceding siblings ...)
2023-12-31 23:08 ` [PATCH 12/32] xfs: fix unit conversion error in xfs_log_calc_max_attrsetm_res Darrick J. Wong
@ 2023-12-31 23:08 ` Darrick J. Wong
2023-12-31 23:08 ` [PATCH 14/32] xfs: don't remove the attr fork when parent pointers are enabled Darrick J. Wong
` (18 subsequent siblings)
31 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:08 UTC (permalink / raw)
To: djwong, cem
Cc: Allison Henderson, catherine.hoang, linux-xfs, allison.henderson
From: Darrick J. Wong <djwong@kernel.org>
Having established that we can reduce the minimum log size computation
for filesystems with parent pointers or any newer feature, we should
also drop the compat minlogsize code that we added when we reduced the
transaction reservation size for rmap and reflink.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
---
libxfs/xfs_log_rlimit.c | 10 ++++++++++
1 file changed, 10 insertions(+)
diff --git a/libxfs/xfs_log_rlimit.c b/libxfs/xfs_log_rlimit.c
index 096e3ee4b6a..2a5fa3334c8 100644
--- a/libxfs/xfs_log_rlimit.c
+++ b/libxfs/xfs_log_rlimit.c
@@ -81,6 +81,16 @@ xfs_log_calc_trans_resv_for_minlogblocks(
{
unsigned int rmap_maxlevels = mp->m_rmap_maxlevels;
+ /*
+ * Starting with the parent pointer feature, every new fs feature
+ * drops the oversized minimum log size computation introduced by the
+ * original reflink code.
+ */
+ if (xfs_has_parent_or_newer_feature(mp)) {
+ xfs_trans_resv_calc(mp, resv);
+ return;
+ }
+
/*
* In the early days of rmap+reflink, we always set the rmap maxlevels
* to 9 even if the AG was small enough that it would never grow to
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 14/32] xfs: don't remove the attr fork when parent pointers are enabled
2023-12-31 19:50 ` [PATCHSET v13.0 3/6] xfsprogs: Parent Pointers Darrick J. Wong
` (12 preceding siblings ...)
2023-12-31 23:08 ` [PATCH 13/32] xfs: drop compatibility minimum log size computations for reflink Darrick J. Wong
@ 2023-12-31 23:08 ` Darrick J. Wong
2023-12-31 23:09 ` [PATCH 15/32] xfs: Add the parent pointer support to the superblock version 5 Darrick J. Wong
` (17 subsequent siblings)
31 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:08 UTC (permalink / raw)
To: djwong, cem
Cc: Allison Henderson, catherine.hoang, linux-xfs, allison.henderson
From: Allison Henderson <allison.henderson@oracle.com>
When an inode is removed, it may also cause the attribute fork to be
removed if it is the last attribute. This transaction gets flushed to
the log, but if the system goes down before we could inactivate the symlink,
the log recovery tries to inactivate this inode (since it is on the unlinked
list) but the verifier trips over the remote value and leaks it.
Hence we ended up with a file in this odd state on a "clean" mount. The
"obvious" fix is to prohibit erasure of the attr fork to avoid tripping
over the verifiers when pptrs are enabled.
Signed-off-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
libxfs/xfs_attr_leaf.c | 6 ++++--
1 file changed, 4 insertions(+), 2 deletions(-)
diff --git a/libxfs/xfs_attr_leaf.c b/libxfs/xfs_attr_leaf.c
index c0fc3c10dc4..14020c09146 100644
--- a/libxfs/xfs_attr_leaf.c
+++ b/libxfs/xfs_attr_leaf.c
@@ -910,7 +910,8 @@ xfs_attr_sf_removename(
totsize -= size;
if (totsize == sizeof(xfs_attr_sf_hdr_t) && xfs_has_attr2(mp) &&
(dp->i_df.if_format != XFS_DINODE_FMT_BTREE) &&
- !(args->op_flags & (XFS_DA_OP_ADDNAME | XFS_DA_OP_REPLACE))) {
+ !(args->op_flags & (XFS_DA_OP_ADDNAME | XFS_DA_OP_REPLACE)) &&
+ !xfs_has_parent(mp)) {
xfs_attr_fork_remove(dp, args->trans);
} else {
xfs_idata_realloc(dp, -size, XFS_ATTR_FORK);
@@ -919,7 +920,8 @@ xfs_attr_sf_removename(
ASSERT(totsize > sizeof(xfs_attr_sf_hdr_t) ||
(args->op_flags & XFS_DA_OP_ADDNAME) ||
!xfs_has_attr2(mp) ||
- dp->i_df.if_format == XFS_DINODE_FMT_BTREE);
+ dp->i_df.if_format == XFS_DINODE_FMT_BTREE ||
+ xfs_has_parent(mp));
xfs_trans_log_inode(args->trans, dp,
XFS_ILOG_CORE | XFS_ILOG_ADATA);
}
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 15/32] xfs: Add the parent pointer support to the superblock version 5.
2023-12-31 19:50 ` [PATCHSET v13.0 3/6] xfsprogs: Parent Pointers Darrick J. Wong
` (13 preceding siblings ...)
2023-12-31 23:08 ` [PATCH 14/32] xfs: don't remove the attr fork when parent pointers are enabled Darrick J. Wong
@ 2023-12-31 23:09 ` Darrick J. Wong
2023-12-31 23:09 ` [PATCH 16/32] libfrog: add parent pointer support code Darrick J. Wong
` (16 subsequent siblings)
31 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:09 UTC (permalink / raw)
To: djwong, cem
Cc: Mark Tinguely, Dave Chinner, Allison Henderson, Darrick J. Wong,
catherine.hoang, linux-xfs, allison.henderson
From: Allison Henderson <allison.henderson@oracle.com>
Add the parent pointer superblock flag so that we can actually mount
filesystems with this feature enabled.
Signed-off-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
libfrog/fsgeom.c | 6 ++++--
libxfs/xfs_format.h | 4 +++-
libxfs/xfs_fs.h | 1 +
libxfs/xfs_sb.c | 4 ++++
4 files changed, 12 insertions(+), 3 deletions(-)
diff --git a/libfrog/fsgeom.c b/libfrog/fsgeom.c
index 6980d3ffab6..061995fa2c7 100644
--- a/libfrog/fsgeom.c
+++ b/libfrog/fsgeom.c
@@ -31,6 +31,7 @@ xfs_report_geom(
int bigtime_enabled;
int inobtcount;
int nrext64;
+ int parent;
isint = geo->logstart > 0;
lazycount = geo->flags & XFS_FSOP_GEOM_FLAGS_LAZYSB ? 1 : 0;
@@ -49,6 +50,7 @@ xfs_report_geom(
bigtime_enabled = geo->flags & XFS_FSOP_GEOM_FLAGS_BIGTIME ? 1 : 0;
inobtcount = geo->flags & XFS_FSOP_GEOM_FLAGS_INOBTCNT ? 1 : 0;
nrext64 = geo->flags & XFS_FSOP_GEOM_FLAGS_NREXT64 ? 1 : 0;
+ parent = geo->flags & XFS_FSOP_GEOM_FLAGS_PARENT ? 1 : 0;
printf(_(
"meta-data=%-22s isize=%-6d agcount=%u, agsize=%u blks\n"
@@ -57,7 +59,7 @@ xfs_report_geom(
" =%-22s reflink=%-4u bigtime=%u inobtcount=%u nrext64=%u\n"
"data =%-22s bsize=%-6u blocks=%llu, imaxpct=%u\n"
" =%-22s sunit=%-6u swidth=%u blks\n"
-"naming =version %-14u bsize=%-6u ascii-ci=%d, ftype=%d\n"
+"naming =version %-14u bsize=%-6u ascii-ci=%d, ftype=%d, parent=%d\n"
"log =%-22s bsize=%-6d blocks=%u, version=%d\n"
" =%-22s sectsz=%-5u sunit=%d blks, lazy-count=%d\n"
"realtime =%-22s extsz=%-6d blocks=%lld, rtextents=%lld\n"),
@@ -68,7 +70,7 @@ xfs_report_geom(
"", geo->blocksize, (unsigned long long)geo->datablocks,
geo->imaxpct,
"", geo->sunit, geo->swidth,
- dirversion, geo->dirblocksize, cimode, ftype_enabled,
+ dirversion, geo->dirblocksize, cimode, ftype_enabled, parent,
isint ? _("internal log") : logname ? logname : _("external"),
geo->blocksize, geo->logblocks, logversion,
"", geo->logsectsize, geo->logsunit / geo->blocksize, lazycount,
diff --git a/libxfs/xfs_format.h b/libxfs/xfs_format.h
index 8b952909ce1..b0aaa825539 100644
--- a/libxfs/xfs_format.h
+++ b/libxfs/xfs_format.h
@@ -373,13 +373,15 @@ xfs_sb_has_ro_compat_feature(
#define XFS_SB_FEAT_INCOMPAT_BIGTIME (1 << 3) /* large timestamps */
#define XFS_SB_FEAT_INCOMPAT_NEEDSREPAIR (1 << 4) /* needs xfs_repair */
#define XFS_SB_FEAT_INCOMPAT_NREXT64 (1 << 5) /* large extent counters */
+#define XFS_SB_FEAT_INCOMPAT_PARENT (1 << 6) /* parent pointers */
#define XFS_SB_FEAT_INCOMPAT_ALL \
(XFS_SB_FEAT_INCOMPAT_FTYPE| \
XFS_SB_FEAT_INCOMPAT_SPINODES| \
XFS_SB_FEAT_INCOMPAT_META_UUID| \
XFS_SB_FEAT_INCOMPAT_BIGTIME| \
XFS_SB_FEAT_INCOMPAT_NEEDSREPAIR| \
- XFS_SB_FEAT_INCOMPAT_NREXT64)
+ XFS_SB_FEAT_INCOMPAT_NREXT64| \
+ XFS_SB_FEAT_INCOMPAT_PARENT)
#define XFS_SB_FEAT_INCOMPAT_UNKNOWN ~XFS_SB_FEAT_INCOMPAT_ALL
static inline bool
diff --git a/libxfs/xfs_fs.h b/libxfs/xfs_fs.h
index e92b6a9612a..efa68a2d82a 100644
--- a/libxfs/xfs_fs.h
+++ b/libxfs/xfs_fs.h
@@ -239,6 +239,7 @@ typedef struct xfs_fsop_resblks {
#define XFS_FSOP_GEOM_FLAGS_BIGTIME (1 << 21) /* 64-bit nsec timestamps */
#define XFS_FSOP_GEOM_FLAGS_INOBTCNT (1 << 22) /* inobt btree counter */
#define XFS_FSOP_GEOM_FLAGS_NREXT64 (1 << 23) /* large extent counters */
+#define XFS_FSOP_GEOM_FLAGS_PARENT (1U << 30) /* parent pointers */
/* atomic file extent swap available to userspace */
#define XFS_FSOP_GEOM_FLAGS_ATOMIC_SWAP (1U << 31)
diff --git a/libxfs/xfs_sb.c b/libxfs/xfs_sb.c
index fd017d18cda..d150170d87b 100644
--- a/libxfs/xfs_sb.c
+++ b/libxfs/xfs_sb.c
@@ -174,6 +174,8 @@ xfs_sb_version_to_features(
features |= XFS_FEAT_NEEDSREPAIR;
if (sbp->sb_features_incompat & XFS_SB_FEAT_INCOMPAT_NREXT64)
features |= XFS_FEAT_NREXT64;
+ if (sbp->sb_features_incompat & XFS_SB_FEAT_INCOMPAT_PARENT)
+ features |= XFS_FEAT_PARENT;
return features;
}
@@ -1249,6 +1251,8 @@ xfs_fs_geometry(
geo->flags |= XFS_FSOP_GEOM_FLAGS_BIGTIME;
if (xfs_has_inobtcounts(mp))
geo->flags |= XFS_FSOP_GEOM_FLAGS_INOBTCNT;
+ if (xfs_has_parent(mp))
+ geo->flags |= XFS_FSOP_GEOM_FLAGS_PARENT;
if (xfs_has_sector(mp)) {
geo->flags |= XFS_FSOP_GEOM_FLAGS_SECTOR;
geo->logsectsize = sbp->sb_logsectsize;
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 16/32] libfrog: add parent pointer support code
2023-12-31 19:50 ` [PATCHSET v13.0 3/6] xfsprogs: Parent Pointers Darrick J. Wong
` (14 preceding siblings ...)
2023-12-31 23:09 ` [PATCH 15/32] xfs: Add the parent pointer support to the superblock version 5 Darrick J. Wong
@ 2023-12-31 23:09 ` Darrick J. Wong
2023-12-31 23:09 ` [PATCH 17/32] libfrog: detect looping paths when walking directory parent pointers Darrick J. Wong
` (15 subsequent siblings)
31 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:09 UTC (permalink / raw)
To: djwong, cem
Cc: Allison Henderson, catherine.hoang, linux-xfs, allison.henderson
From: Darrick J. Wong <djwong@kernel.org>
Add some support code to libfrog so that client programs can walk file
descriptors and handles upwards through the directory tree; and obtain a
reasonable file path from a file descriptor/handle. This code will be
used in xfsprogs utilities.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
---
include/handle.h | 1
libfrog/Makefile | 2
libfrog/getparents.c | 345 ++++++++++++++++++++++++++++++++++++++++++++++++++
libfrog/getparents.h | 36 +++++
libfrog/paths.c | 167 ++++++++++++++++++++++++
libfrog/paths.h | 25 ++++
libhandle/handle.c | 7 +
7 files changed, 580 insertions(+), 3 deletions(-)
create mode 100644 libfrog/getparents.c
create mode 100644 libfrog/getparents.h
diff --git a/include/handle.h b/include/handle.h
index 34246f3854d..ba06500516c 100644
--- a/include/handle.h
+++ b/include/handle.h
@@ -17,6 +17,7 @@ struct parent;
extern int path_to_handle (char *__path, void **__hanp, size_t *__hlen);
extern int path_to_fshandle (char *__path, void **__fshanp, size_t *__fshlen);
extern int fd_to_handle (int fd, void **hanp, size_t *hlen);
+extern int handle_to_fsfd(void *, char **);
extern int handle_to_fshandle (void *__hanp, size_t __hlen, void **__fshanp,
size_t *__fshlen);
extern void free_handle (void *__hanp, size_t __hlen);
diff --git a/libfrog/Makefile b/libfrog/Makefile
index bbc5b887cd3..8c2d040bc29 100644
--- a/libfrog/Makefile
+++ b/libfrog/Makefile
@@ -20,6 +20,7 @@ convert.c \
crc32.c \
file_exchange.c \
fsgeom.c \
+getparents.c \
histogram.c \
list_sort.c \
linux.c \
@@ -46,6 +47,7 @@ dahashselftest.h \
div64.h \
file_exchange.h \
fsgeom.h \
+getparents.h \
histogram.h \
logging.h \
paths.h \
diff --git a/libfrog/getparents.c b/libfrog/getparents.c
new file mode 100644
index 00000000000..016fe3f026d
--- /dev/null
+++ b/libfrog/getparents.c
@@ -0,0 +1,345 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * Copyright (c) 2017-2024 Oracle. All Rights Reserved.
+ * Author: Darrick J. Wong <djwong@kernel.org>
+ */
+#include "platform_defs.h"
+#include "xfs.h"
+#include "xfs_arch.h"
+#include "list.h"
+#include "libfrog/paths.h"
+#include "handle.h"
+#include "libfrog/getparents.h"
+
+/* Allocate a buffer large enough for some parent pointer records. */
+static inline struct xfs_getparents *
+alloc_pptr_buf(
+ size_t bufsize)
+{
+ struct xfs_getparents *pi;
+
+ pi = calloc(bufsize, 1);
+ if (!pi)
+ return NULL;
+ pi->gp_bufsize = bufsize;
+ return pi;
+}
+
+/*
+ * Walk all parents of the given file handle. Returns 0 on success or positive
+ * errno.
+ */
+static int
+call_getparents(
+ int fd,
+ struct xfs_handle *handle,
+ walk_parent_fn fn,
+ void *arg)
+{
+ struct xfs_getparents *pi;
+ struct xfs_getparents_rec *p;
+ unsigned int i;
+ ssize_t ret = -1;
+
+ pi = alloc_pptr_buf(XFS_XATTR_LIST_MAX);
+ if (!pi)
+ return errno;
+
+ if (handle) {
+ memcpy(&pi->gp_handle, handle, sizeof(struct xfs_handle));
+ pi->gp_flags = XFS_GETPARENTS_IFLAG_HANDLE;
+ }
+
+ ret = ioctl(fd, XFS_IOC_GETPARENTS, pi);
+ while (!ret) {
+ if (pi->gp_flags & XFS_GETPARENTS_OFLAG_ROOT) {
+ struct parent_rec rec = {
+ .p_flags = PARENT_IS_ROOT,
+ };
+
+ ret = fn(&rec, arg);
+ goto out_pi;
+ }
+
+ for (i = 0; i < pi->gp_count; i++) {
+ struct parent_rec rec = { 0 };
+
+ p = xfs_getparents_rec(pi, i);
+ rec.p_ino = p->gpr_ino;
+ rec.p_gen = p->gpr_gen;
+ strncpy((char *)rec.p_name, (char *)p->gpr_name,
+ MAXNAMELEN - 1);
+
+ ret = fn(&rec, arg);
+ if (ret)
+ goto out_pi;
+ }
+
+ if (pi->gp_flags & XFS_GETPARENTS_OFLAG_DONE)
+ break;
+
+ ret = ioctl(fd, XFS_IOC_GETPARENTS, pi);
+ }
+ if (ret)
+ ret = errno;
+
+out_pi:
+ free(pi);
+ return ret;
+}
+
+/* Walk all parent pointers of this handle. Returns 0 or positive errno. */
+int
+handle_walk_parents(
+ void *hanp,
+ size_t hlen,
+ walk_parent_fn fn,
+ void *arg)
+{
+ char *mntpt;
+ int fd;
+
+ if (hlen != sizeof(struct xfs_handle))
+ return EINVAL;
+
+ fd = handle_to_fsfd(hanp, &mntpt);
+ if (fd < 0)
+ return errno;
+
+ return call_getparents(fd, hanp, fn, arg);
+}
+
+/* Walk all parent pointers of this fd. Returns 0 or positive errno. */
+int
+fd_walk_parents(
+ int fd,
+ walk_parent_fn fn,
+ void *arg)
+{
+ return call_getparents(fd, NULL, fn, arg);
+}
+
+struct walk_ppaths_info {
+ walk_path_fn fn;
+ void *arg;
+ char *mntpt;
+ struct path_list *path;
+ int fd;
+};
+
+struct walk_ppath_level_info {
+ struct xfs_handle newhandle;
+ struct path_component *pc;
+ struct walk_ppaths_info *wpi;
+};
+
+static int handle_walk_ppath(struct walk_ppaths_info *wpi,
+ struct xfs_handle *handle);
+
+static int
+handle_walk_ppath_rec(
+ const struct parent_rec *rec,
+ void *arg)
+{
+ struct walk_ppath_level_info *wpli = arg;
+ struct walk_ppaths_info *wpi = wpli->wpi;
+ int ret = 0;
+
+ if (rec->p_flags & PARENT_IS_ROOT)
+ return wpi->fn(wpi->mntpt, wpi->path, wpi->arg);
+
+ ret = path_component_change(wpli->pc, rec->p_name,
+ strlen((char *)rec->p_name), rec->p_ino);
+ if (ret)
+ return ret;
+
+ wpli->newhandle.ha_fid.fid_ino = rec->p_ino;
+ wpli->newhandle.ha_fid.fid_gen = rec->p_gen;
+
+ path_list_add_parent_component(wpi->path, wpli->pc);
+ ret = handle_walk_ppath(wpi, &wpli->newhandle);
+ path_list_del_component(wpi->path, wpli->pc);
+
+ return ret;
+}
+
+/*
+ * Recursively walk all parents of the given file handle; if we hit the
+ * fs root then we call the associated function with the constructed path.
+ * Returns 0 for success or positive errno.
+ */
+static int
+handle_walk_ppath(
+ struct walk_ppaths_info *wpi,
+ struct xfs_handle *handle)
+{
+ struct walk_ppath_level_info *wpli;
+ int ret;
+
+ wpli = malloc(sizeof(struct walk_ppath_level_info));
+ if (!wpli)
+ return errno;
+ wpli->pc = path_component_init("", 0);
+ if (!wpli->pc) {
+ ret = errno;
+ free(wpli);
+ return ret;
+ }
+ wpli->wpi = wpi;
+ memcpy(&wpli->newhandle, handle, sizeof(struct xfs_handle));
+
+ ret = call_getparents(wpi->fd, handle, handle_walk_ppath_rec, wpli);
+
+ path_component_free(wpli->pc);
+ free(wpli);
+ return ret;
+}
+
+/*
+ * Call the given function on all known paths from the vfs root to the inode
+ * described in the handle. Returns 0 for success or positive errno.
+ */
+int
+handle_walk_parent_paths(
+ void *hanp,
+ size_t hlen,
+ walk_path_fn fn,
+ void *arg)
+{
+ struct walk_ppaths_info wpi;
+ ssize_t ret;
+
+ if (hlen != sizeof(struct xfs_handle))
+ return EINVAL;
+
+ wpi.fd = handle_to_fsfd(hanp, &wpi.mntpt);
+ if (wpi.fd < 0)
+ return errno;
+ wpi.path = path_list_init();
+ if (!wpi.path)
+ return errno;
+ wpi.fn = fn;
+ wpi.arg = arg;
+
+ ret = handle_walk_ppath(&wpi, hanp);
+ path_list_free(wpi.path);
+
+ return ret;
+}
+
+/*
+ * Call the given function on all known paths from the vfs root to the inode
+ * referred to by the file description. Returns 0 or positive errno.
+ */
+int
+fd_walk_parent_paths(
+ int fd,
+ walk_path_fn fn,
+ void *arg)
+{
+ struct walk_ppaths_info wpi;
+ void *hanp;
+ size_t hlen;
+ int fsfd;
+ int ret;
+
+ ret = fd_to_handle(fd, &hanp, &hlen);
+ if (ret)
+ return errno;
+
+ fsfd = handle_to_fsfd(hanp, &wpi.mntpt);
+ if (fsfd < 0)
+ return errno;
+ wpi.fd = fd;
+ wpi.path = path_list_init();
+ if (!wpi.path)
+ return errno;
+ wpi.fn = fn;
+ wpi.arg = arg;
+
+ ret = handle_walk_ppath(&wpi, hanp);
+ path_list_free(wpi.path);
+
+ return ret;
+}
+
+struct path_walk_info {
+ char *buf;
+ size_t len;
+ size_t written;
+};
+
+/* Helper that stringifies the first full path that we find. */
+static int
+handle_to_path_walk(
+ const char *mntpt,
+ const struct path_list *path,
+ void *arg)
+{
+ struct path_walk_info *pwi = arg;
+ int mntpt_len = strlen(mntpt);
+ int ret;
+
+ /* Trim trailing slashes from the mountpoint */
+ while (mntpt_len > 0 && mntpt[mntpt_len - 1] == '/')
+ mntpt_len--;
+
+ ret = snprintf(pwi->buf, pwi->len, "%.*s", mntpt_len, mntpt);
+ if (ret != mntpt_len)
+ return ENAMETOOLONG;
+ pwi->written += ret;
+
+ ret = path_list_to_string(path, pwi->buf + ret, pwi->len - ret);
+ if (ret < 0)
+ return ENAMETOOLONG;
+
+ pwi->written += ret;
+ return ECANCELED;
+}
+
+/*
+ * Return any eligible path to this file handle. Returns 0 for success or
+ * positive errno.
+ */
+int
+handle_to_path(
+ void *hanp,
+ size_t hlen,
+ char *path,
+ size_t pathlen)
+{
+ struct path_walk_info pwi = { .buf = path, .len = pathlen };
+ int ret;
+
+ ret = handle_walk_parent_paths(hanp, hlen, handle_to_path_walk, &pwi);
+ if (ret && ret != ECANCELED)
+ return ret;
+ if (!pwi.written)
+ return ENODATA;
+
+ path[pwi.written] = 0;
+ return 0;
+}
+
+/*
+ * Return any eligible path to this file description. Returns 0 for success
+ * or positive errno.
+ */
+int
+fd_to_path(
+ int fd,
+ char *path,
+ size_t pathlen)
+{
+ struct path_walk_info pwi = { .buf = path, .len = pathlen };
+ int ret;
+
+ ret = fd_walk_parent_paths(fd, handle_to_path_walk, &pwi);
+ if (ret && ret != ECANCELED)
+ return ret;
+ if (!pwi.written)
+ return ENODATA;
+
+ path[pwi.written] = 0;
+ return ret;
+}
diff --git a/libfrog/getparents.h b/libfrog/getparents.h
new file mode 100644
index 00000000000..41beb4ce3be
--- /dev/null
+++ b/libfrog/getparents.h
@@ -0,0 +1,36 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * Copyright (c) 2023-2024 Oracle. All Rights Reserved.
+ * Author: Darrick J. Wong <djwong@kernel.org>
+ */
+#ifndef __LIBFROG_GETPARENTS_H_
+#define __LIBFROG_GETPARENTS_H_
+
+struct path_list;
+
+struct parent_rec {
+ uint64_t p_ino;
+ uint32_t p_gen;
+ uint32_t p_flags;
+ unsigned char p_name[MAXNAMELEN];
+};
+
+/* This is the root directory. */
+#define PARENT_IS_ROOT (1U << 0)
+
+typedef int (*walk_parent_fn)(const struct parent_rec *rec, void *arg);
+typedef int (*walk_path_fn)(const char *mntpt, const struct path_list *path,
+ void *arg);
+
+int fd_walk_parents(int fd, walk_parent_fn fn, void *arg);
+int handle_walk_parents(void *hanp, size_t hanlen, walk_parent_fn fn,
+ void *arg);
+
+int fd_walk_parent_paths(int fd, walk_path_fn fn, void *arg);
+int handle_walk_parent_paths(void *hanp, size_t hanlen, walk_path_fn fn,
+ void *arg);
+
+int fd_to_path(int fd, char *path, size_t pathlen);
+int handle_to_path(void *hanp, size_t hlen, char *path, size_t pathlen);
+
+#endif /* __LIBFROG_GETPARENTS_H_ */
diff --git a/libfrog/paths.c b/libfrog/paths.c
index d8c42163a7b..b3c5236990e 100644
--- a/libfrog/paths.c
+++ b/libfrog/paths.c
@@ -15,6 +15,7 @@
#include "paths.h"
#include "input.h"
#include "projects.h"
+#include "list.h"
#include <limits.h>
extern char *progname;
@@ -567,3 +568,169 @@ fs_table_insert_project_path(
return error;
}
+
+/* Structured path components. */
+
+struct path_list {
+ struct list_head p_head;
+};
+
+struct path_component {
+ struct list_head pc_list;
+ uint64_t pc_ino;
+ char *pc_fname;
+};
+
+/* Initialize a path component with a given name. */
+struct path_component *
+path_component_init(
+ const char *name,
+ uint64_t ino)
+{
+ struct path_component *pc;
+
+ pc = malloc(sizeof(struct path_component));
+ if (!pc)
+ return NULL;
+ INIT_LIST_HEAD(&pc->pc_list);
+ pc->pc_fname = strdup(name);
+ if (!pc->pc_fname) {
+ free(pc);
+ return NULL;
+ }
+ pc->pc_ino = ino;
+ return pc;
+}
+
+/* Free a path component. */
+void
+path_component_free(
+ struct path_component *pc)
+{
+ free(pc->pc_fname);
+ free(pc);
+}
+
+/* Change a path component's filename or returns positive errno. */
+int
+path_component_change(
+ struct path_component *pc,
+ const void *name,
+ size_t namelen,
+ uint64_t ino)
+{
+ void *p;
+
+ p = realloc(pc->pc_fname, namelen + 1);
+ if (!p)
+ return errno;
+ pc->pc_fname = p;
+ memcpy(pc->pc_fname, name, namelen);
+ pc->pc_fname[namelen] = 0;
+ pc->pc_ino = ino;
+ return 0;
+}
+
+/* Initialize a pathname or returns positive errno. */
+struct path_list *
+path_list_init(void)
+{
+ struct path_list *path;
+
+ path = malloc(sizeof(struct path_list));
+ if (!path)
+ return NULL;
+ INIT_LIST_HEAD(&path->p_head);
+ return path;
+}
+
+/* Empty out a pathname. */
+void
+path_list_free(
+ struct path_list *path)
+{
+ struct path_component *pos;
+ struct path_component *n;
+
+ list_for_each_entry_safe(pos, n, &path->p_head, pc_list) {
+ path_list_del_component(path, pos);
+ path_component_free(pos);
+ }
+ free(path);
+}
+
+/* Add a parent component to a pathname. */
+void
+path_list_add_parent_component(
+ struct path_list *path,
+ struct path_component *pc)
+{
+ list_add(&pc->pc_list, &path->p_head);
+}
+
+/* Add a component to a pathname. */
+void
+path_list_add_component(
+ struct path_list *path,
+ struct path_component *pc)
+{
+ list_add_tail(&pc->pc_list, &path->p_head);
+}
+
+/* Remove a component from a pathname. */
+void
+path_list_del_component(
+ struct path_list *path,
+ struct path_component *pc)
+{
+ list_del_init(&pc->pc_list);
+}
+
+/*
+ * Convert a pathname into a string or returns -1 if the buffer isn't long
+ * enough.
+ */
+ssize_t
+path_list_to_string(
+ const struct path_list *path,
+ char *buf,
+ size_t buflen)
+{
+ struct path_component *pos;
+ char *buf_end = buf + buflen;
+ ssize_t bytes = 0;
+ int ret;
+
+ list_for_each_entry(pos, &path->p_head, pc_list) {
+ if (buf >= buf_end)
+ return -1;
+
+ ret = snprintf(buf, buflen, "/%s", pos->pc_fname);
+ if (ret < 0 || ret >= buflen)
+ return -1;
+
+ bytes += ret;
+ buf += ret;
+ buflen -= ret;
+ }
+ return bytes;
+}
+
+/* Walk each component of a path. */
+int
+path_walk_components(
+ const struct path_list *path,
+ path_walk_fn_t fn,
+ void *arg)
+{
+ struct path_component *pos;
+ int ret;
+
+ list_for_each_entry(pos, &path->p_head, pc_list) {
+ ret = fn(pos->pc_fname, pos->pc_ino, arg);
+ if (ret)
+ return ret;
+ }
+
+ return 0;
+}
diff --git a/libfrog/paths.h b/libfrog/paths.h
index f20a2c3ef58..6be74c42b07 100644
--- a/libfrog/paths.h
+++ b/libfrog/paths.h
@@ -58,4 +58,29 @@ typedef struct fs_cursor {
extern void fs_cursor_initialise(char *__dir, uint __flags, fs_cursor_t *__cp);
extern fs_path_t *fs_cursor_next_entry(fs_cursor_t *__cp);
+/* Path information. */
+
+struct path_list;
+struct path_component;
+
+struct path_component *path_component_init(const char *name, uint64_t ino);
+void path_component_free(struct path_component *pc);
+int path_component_change(struct path_component *pc, const void *name,
+ size_t namelen, uint64_t ino);
+
+struct path_list *path_list_init(void);
+void path_list_free(struct path_list *path);
+void path_list_add_parent_component(struct path_list *path,
+ struct path_component *pc);
+void path_list_add_component(struct path_list *path, struct path_component *pc);
+void path_list_del_component(struct path_list *path, struct path_component *pc);
+
+ssize_t path_list_to_string(const struct path_list *path, char *buf,
+ size_t buflen);
+
+typedef int (*path_walk_fn_t)(const char *name, uint64_t ino, void *arg);
+
+int path_walk_components(const struct path_list *path, path_walk_fn_t fn,
+ void *arg);
+
#endif /* __LIBFROG_PATH_H__ */
diff --git a/libhandle/handle.c b/libhandle/handle.c
index 333c2190900..1e8fe9ac5f1 100644
--- a/libhandle/handle.c
+++ b/libhandle/handle.c
@@ -29,7 +29,6 @@ typedef union {
} comarg_t;
static int obj_to_handle(char *, int, unsigned int, comarg_t, void**, size_t*);
-static int handle_to_fsfd(void *, char **);
static char *path_to_fspath(char *path);
@@ -203,8 +202,10 @@ handle_to_fshandle(
return 0;
}
-static int
-handle_to_fsfd(void *hanp, char **path)
+int
+handle_to_fsfd(
+ void *hanp,
+ char **path)
{
struct fdhash *fdhp;
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 17/32] libfrog: detect looping paths when walking directory parent pointers
2023-12-31 19:50 ` [PATCHSET v13.0 3/6] xfsprogs: Parent Pointers Darrick J. Wong
` (15 preceding siblings ...)
2023-12-31 23:09 ` [PATCH 16/32] libfrog: add parent pointer support code Darrick J. Wong
@ 2023-12-31 23:09 ` Darrick J. Wong
2023-12-31 23:09 ` [PATCH 18/32] xfs_io: adapt parent command to new parent pointer ioctls Darrick J. Wong
` (14 subsequent siblings)
31 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:09 UTC (permalink / raw)
To: djwong, cem; +Cc: catherine.hoang, linux-xfs, allison.henderson
From: Darrick J. Wong <djwong@kernel.org>
Detect loops when we're walking directory parent pointers so that we
don't loop infinitely.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
libfrog/getparents.c | 3 +++
libfrog/paths.c | 16 ++++++++++++++++
libfrog/paths.h | 2 ++
3 files changed, 21 insertions(+)
diff --git a/libfrog/getparents.c b/libfrog/getparents.c
index 016fe3f026d..fa4e4a1c9c0 100644
--- a/libfrog/getparents.c
+++ b/libfrog/getparents.c
@@ -148,6 +148,9 @@ handle_walk_ppath_rec(
if (rec->p_flags & PARENT_IS_ROOT)
return wpi->fn(wpi->mntpt, wpi->path, wpi->arg);
+ if (path_will_loop(wpi->path, rec->p_ino))
+ return 0;
+
ret = path_component_change(wpli->pc, rec->p_name,
strlen((char *)rec->p_name), rec->p_ino);
if (ret)
diff --git a/libfrog/paths.c b/libfrog/paths.c
index b3c5236990e..9ba2a2f313b 100644
--- a/libfrog/paths.c
+++ b/libfrog/paths.c
@@ -734,3 +734,19 @@ path_walk_components(
return 0;
}
+
+/* Will this path contain a loop if we add this inode? */
+bool
+path_will_loop(
+ const struct path_list *path_list,
+ uint64_t ino)
+{
+ struct path_component *pc;
+
+ list_for_each_entry(pc, &path_list->p_head, pc_list) {
+ if (pc->pc_ino == ino)
+ return true;
+ }
+
+ return false;
+}
diff --git a/libfrog/paths.h b/libfrog/paths.h
index 6be74c42b07..895171aa342 100644
--- a/libfrog/paths.h
+++ b/libfrog/paths.h
@@ -83,4 +83,6 @@ typedef int (*path_walk_fn_t)(const char *name, uint64_t ino, void *arg);
int path_walk_components(const struct path_list *path, path_walk_fn_t fn,
void *arg);
+bool path_will_loop(const struct path_list *path, uint64_t ino);
+
#endif /* __LIBFROG_PATH_H__ */
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 18/32] xfs_io: adapt parent command to new parent pointer ioctls
2023-12-31 19:50 ` [PATCHSET v13.0 3/6] xfsprogs: Parent Pointers Darrick J. Wong
` (16 preceding siblings ...)
2023-12-31 23:09 ` [PATCH 17/32] libfrog: detect looping paths when walking directory parent pointers Darrick J. Wong
@ 2023-12-31 23:09 ` Darrick J. Wong
2023-12-31 23:10 ` [PATCH 19/32] xfs_io: Add i, n and f flags to parent command Darrick J. Wong
` (13 subsequent siblings)
31 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:09 UTC (permalink / raw)
To: djwong, cem
Cc: Allison Henderson, catherine.hoang, linux-xfs, allison.henderson
From: Darrick J. Wong <djwong@kernel.org>
For ages, xfs_io has had a totally useless 'parent' command that enabled
callers to walk the parents or print the directory tree path of an open
file. This code used the ioctl interface presented by SGI's version of
parent pointers that was never merged. Rework the code in here to use
the new ioctl interfaces that we've settled upon. Get rid of the old
parent pointer checking code since xfs_repair/xfs_scrub will take care
of that.
(This originally was in the "xfsprogs: implement the upper half of
parent pointers" megapatch.)
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
---
io/parent.c | 470 ++++++++++++++---------------------------------------
man/man8/xfs_io.8 | 21 +-
2 files changed, 133 insertions(+), 358 deletions(-)
diff --git a/io/parent.c b/io/parent.c
index 8f63607ffec..65fd892bffc 100644
--- a/io/parent.c
+++ b/io/parent.c
@@ -7,365 +7,112 @@
#include "command.h"
#include "input.h"
#include "libfrog/paths.h"
-#include "parent.h"
+#include "libfrog/getparents.h"
#include "handle.h"
-#include "jdm.h"
#include "init.h"
#include "io.h"
-#define PARENTBUF_SZ 16384
-#define BSTATBUF_SZ 16384
-
static cmdinfo_t parent_cmd;
-static int verbose_flag;
-static int err_status;
-static __u64 inodes_checked;
static char *mntpt;
-/*
- * check out a parent entry to see if the values seem valid
- */
-static void
-check_parent_entry(struct xfs_bstat *bstatp, parent_t *parent)
+static int
+pptr_print(
+ const struct parent_rec *rec,
+ void *arg)
{
- int sts;
- char fullpath[PATH_MAX];
- struct stat statbuf;
- char *str;
-
- sprintf(fullpath, _("%s%s"), mntpt, parent->p_name);
-
- sts = lstat(fullpath, &statbuf);
- if (sts != 0) {
- fprintf(stderr,
- _("inode-path for inode: %llu is incorrect - path \"%s\" non-existent\n"),
- (unsigned long long) bstatp->bs_ino, fullpath);
- if (verbose_flag) {
- fprintf(stderr,
- _("path \"%s\" does not stat for inode: %llu; err = %s\n"),
- fullpath,
- (unsigned long long) bstatp->bs_ino,
- strerror(errno));
- }
- err_status++;
- return;
- } else {
- if (verbose_flag > 1) {
- printf(_("path \"%s\" found\n"), fullpath);
- }
- }
-
- if (statbuf.st_ino != bstatp->bs_ino) {
- fprintf(stderr,
- _("inode-path for inode: %llu is incorrect - wrong inode#\n"),
- (unsigned long long) bstatp->bs_ino);
- if (verbose_flag) {
- fprintf(stderr,
- _("ino mismatch for path \"%s\" %llu vs %llu\n"),
- fullpath,
- (unsigned long long)statbuf.st_ino,
- (unsigned long long)bstatp->bs_ino);
- }
- err_status++;
- return;
- } else if (verbose_flag > 1) {
- printf(_("inode number match: %llu\n"),
- (unsigned long long)statbuf.st_ino);
- }
+ const char *name = (char *)rec->p_name;
+ unsigned int namelen;
- /* get parent path */
- str = strrchr(fullpath, '/');
- *str = '\0';
- sts = stat(fullpath, &statbuf);
- if (sts != 0) {
- fprintf(stderr,
- _("parent path \"%s\" does not stat: %s\n"),
- fullpath,
- strerror(errno));
- err_status++;
- return;
- } else {
- if (parent->p_ino != statbuf.st_ino) {
- fprintf(stderr,
- _("inode-path for inode: %llu is incorrect - wrong parent inode#\n"),
- (unsigned long long) bstatp->bs_ino);
- if (verbose_flag) {
- fprintf(stderr,
- _("ino mismatch for path \"%s\" %llu vs %llu\n"),
- fullpath,
- (unsigned long long)parent->p_ino,
- (unsigned long long)statbuf.st_ino);
- }
- err_status++;
- return;
- } else {
- if (verbose_flag > 1) {
- printf(_("parent ino match for %llu\n"),
- (unsigned long long) parent->p_ino);
- }
- }
+ if (rec->p_flags & PARENT_IS_ROOT) {
+ printf(_("Root directory.\n"));
+ return 0;
}
-}
-
-static void
-check_parents(parent_t *parentbuf, size_t *parentbuf_size,
- jdm_fshandle_t *fshandlep, struct xfs_bstat *statp)
-{
- int error, i;
- __u32 count;
- parent_t *entryp;
-
- do {
- error = jdm_parentpaths(fshandlep, statp, parentbuf, *parentbuf_size, &count);
-
- if (error == ERANGE) {
- *parentbuf_size *= 2;
- parentbuf = (parent_t *)realloc(parentbuf, *parentbuf_size);
- } else if (error) {
- fprintf(stderr, _("parentpaths failed for ino %llu: %s\n"),
- (unsigned long long) statp->bs_ino,
- strerror(errno));
- err_status++;
- break;
- }
- } while (error == ERANGE);
-
- if (count == 0) {
- /* no links for inode - something wrong here */
- fprintf(stderr, _("inode-path for inode: %llu is missing\n"),
- (unsigned long long) statp->bs_ino);
- err_status++;
- }
+ namelen = strlen(name);
+ printf(_("p_ino = %llu\n"), (unsigned long long)rec->p_ino);
+ printf(_("p_gen = %u\n"), (unsigned int)rec->p_gen);
+ printf(_("p_namelen = %u\n"), namelen);
+ printf(_("p_name = \"%s\"\n\n"), rec->p_name);
- entryp = parentbuf;
- for (i = 0; i < count; i++) {
- check_parent_entry(statp, entryp);
- entryp = (parent_t*) (((char*)entryp) + entryp->p_reclen);
- }
+ return 0;
}
static int
-do_bulkstat(parent_t *parentbuf, size_t *parentbuf_size,
- struct xfs_bstat *bstatbuf, int fsfd, jdm_fshandle_t *fshandlep)
+print_parents(
+ struct xfs_handle *handle)
{
- __s32 buflenout;
- __u64 lastino = 0;
- struct xfs_bstat *p;
- struct xfs_bstat *endp;
- struct xfs_fsop_bulkreq bulkreq;
- struct stat mntstat;
-
- if (stat(mntpt, &mntstat)) {
- fprintf(stderr, _("can't stat mount point \"%s\": %s\n"),
- mntpt, strerror(errno));
- return 1;
- }
-
- bulkreq.lastip = &lastino;
- bulkreq.icount = BSTATBUF_SZ;
- bulkreq.ubuffer = (void *)bstatbuf;
- bulkreq.ocount = &buflenout;
-
- while (xfsctl(mntpt, fsfd, XFS_IOC_FSBULKSTAT, &bulkreq) == 0) {
- if (*(bulkreq.ocount) == 0) {
- return 0;
- }
- for (p = bstatbuf, endp = bstatbuf + *bulkreq.ocount; p < endp; p++) {
-
- /* inode being modified, get synced data with iget */
- if ( (!p->bs_nlink || !p->bs_mode) && p->bs_ino != 0 ) {
-
- if (xfsctl(mntpt, fsfd, XFS_IOC_FSBULKSTAT_SINGLE, &bulkreq) < 0) {
- fprintf(stderr,
- _("failed to get bulkstat information for inode %llu\n"),
- (unsigned long long) p->bs_ino);
- continue;
- }
- if (!p->bs_nlink || !p->bs_mode || !p->bs_ino) {
- fprintf(stderr,
- _("failed to get valid bulkstat information for inode %llu\n"),
- (unsigned long long) p->bs_ino);
- continue;
- }
- }
-
- /* skip root */
- if (p->bs_ino == mntstat.st_ino) {
- continue;
- }
-
- if (verbose_flag > 1) {
- printf(_("checking inode %llu\n"),
- (unsigned long long) p->bs_ino);
- }
-
- /* print dotted progress */
- if ((inodes_checked % 100) == 0 && verbose_flag == 1) {
- printf("."); fflush(stdout);
- }
- inodes_checked++;
-
- check_parents(parentbuf, parentbuf_size, fshandlep, p);
- }
+ int ret;
- }/*while*/
+ if (handle)
+ ret = handle_walk_parents(handle, sizeof(*handle), pptr_print,
+ NULL);
+ else
+ ret = fd_walk_parents(file->fd, pptr_print, NULL);
+ if (ret)
+ fprintf(stderr, "%s: %s\n", file->name, strerror(ret));
- fprintf(stderr, _("syssgi bulkstat failed: %s\n"), strerror(errno));
- return 1;
+ return 0;
}
static int
-parent_check(void)
+path_print(
+ const char *mntpt,
+ const struct path_list *path,
+ void *arg)
{
- int fsfd;
- jdm_fshandle_t *fshandlep;
- parent_t *parentbuf;
- size_t parentbuf_size = PARENTBUF_SZ;
- struct xfs_bstat *bstatbuf;
-
- err_status = 0;
- inodes_checked = 0;
-
- sync();
-
- fsfd = file->fd;
-
- fshandlep = jdm_getfshandle(mntpt);
- if (fshandlep == NULL) {
- fprintf(stderr, _("unable to open \"%s\" for jdm: %s\n"),
- mntpt,
- strerror(errno));
- return 1;
- }
+ char buf[PATH_MAX];
+ size_t len = PATH_MAX;
+ int mntpt_len = strlen(mntpt);
+ int ret;
- /* allocate buffers */
- bstatbuf = (struct xfs_bstat *)calloc(BSTATBUF_SZ, sizeof(struct xfs_bstat));
- parentbuf = (parent_t *)malloc(parentbuf_size);
- if (!bstatbuf || !parentbuf) {
- fprintf(stderr, _("unable to allocate buffers: %s\n"),
- strerror(errno));
- err_status = 1;
- goto out;
- }
-
- if (do_bulkstat(parentbuf, &parentbuf_size, bstatbuf, fsfd, fshandlep) != 0)
- err_status++;
+ /* Trim trailing slashes from the mountpoint */
+ while (mntpt_len > 0 && mntpt[mntpt_len - 1] == '/')
+ mntpt_len--;
- if (err_status > 0)
- fprintf(stderr, _("num errors: %d\n"), err_status);
- else
- printf(_("succeeded checking %llu inodes\n"),
- (unsigned long long) inodes_checked);
+ ret = snprintf(buf, len, "%.*s", mntpt_len, mntpt);
+ if (ret != mntpt_len)
+ return ENAMETOOLONG;
-out:
- free(bstatbuf);
- free(parentbuf);
- free(fshandlep);
- return err_status;
-}
+ ret = path_list_to_string(path, buf + ret, len - ret);
+ if (ret < 0)
+ return ENAMETOOLONG;
-static void
-print_parent_entry(parent_t *parent, int fullpath)
-{
- printf(_("p_ino = %llu\n"), (unsigned long long) parent->p_ino);
- printf(_("p_gen = %u\n"), parent->p_gen);
- printf(_("p_reclen = %u\n"), parent->p_reclen);
- if (fullpath)
- printf(_("p_name = \"%s%s\"\n"), mntpt, parent->p_name);
- else
- printf(_("p_name = \"%s\"\n"), parent->p_name);
+ printf("%s\n", buf);
+ return 0;
}
static int
-parent_list(int fullpath)
+print_paths(
+ struct xfs_handle *handle)
{
- void *handlep = NULL;
- size_t handlen;
- int error, i;
- int retval = 1;
- __u32 count;
- parent_t *entryp;
- parent_t *parentbuf = NULL;
- char *path = file->name;
- int pb_size = PARENTBUF_SZ;
+ int ret;
- /* XXXX for linux libhandle version - to set libhandle fsfd cache */
- {
- void *fshandle;
- size_t fshlen;
-
- if (path_to_fshandle(mntpt, &fshandle, &fshlen) != 0) {
- fprintf(stderr, _("%s: failed path_to_fshandle \"%s\": %s\n"),
- progname, path, strerror(errno));
- goto error;
- }
- free_handle(fshandle, fshlen);
- }
-
- if (path_to_handle(path, &handlep, &handlen) != 0) {
- fprintf(stderr, _("%s: path_to_handle failed for \"%s\"\n"), progname, path);
- goto error;
- }
-
- do {
- parentbuf = (parent_t *)realloc(parentbuf, pb_size);
- if (!parentbuf) {
- fprintf(stderr, _("%s: unable to allocate parent buffer: %s\n"),
- progname, strerror(errno));
- goto error;
- }
-
- if (fullpath) {
- error = parentpaths_by_handle(handlep,
- handlen,
- parentbuf,
- pb_size,
- &count);
- } else {
- error = parents_by_handle(handlep,
- handlen,
- parentbuf,
- pb_size,
- &count);
- }
- if (error == ERANGE) {
- pb_size *= 2;
- } else if (error) {
- fprintf(stderr, _("%s: %s call failed for \"%s\": %s\n"),
- progname, fullpath ? "parentpaths" : "parents",
- path, strerror(errno));
- goto error;
- }
- } while (error == ERANGE);
-
- if (count == 0) {
- /* no links for inode - something wrong here */
- fprintf(stderr, _("%s: inode-path is missing\n"), progname);
- goto error;
- }
-
- entryp = parentbuf;
- for (i = 0; i < count; i++) {
- print_parent_entry(entryp, fullpath);
- entryp = (parent_t*) (((char*)entryp) + entryp->p_reclen);
- }
-
- retval = 0;
-error:
- free(handlep);
- free(parentbuf);
- return retval;
+ if (handle)
+ ret = handle_walk_parent_paths(handle, sizeof(*handle),
+ path_print, NULL);
+ else
+ ret = fd_walk_parent_paths(file->fd, path_print, NULL);
+ if (ret)
+ fprintf(stderr, "%s: %s\n", file->name, strerror(ret));
+ return 0;
}
static int
-parent_f(int argc, char **argv)
+parent_f(
+ int argc,
+ char **argv)
{
- int c;
- int listpath_flag = 0;
- int check_flag = 0;
- fs_path_t *fs;
- static int tab_init;
+ struct xfs_handle handle;
+ void *hanp = NULL;
+ size_t hlen;
+ struct fs_path *fs;
+ char *p;
+ uint64_t ino = 0;
+ uint32_t gen = 0;
+ int c;
+ int listpath_flag = 0;
+ int ret;
+ static int tab_init;
if (!tab_init) {
tab_init = 1;
@@ -380,46 +127,74 @@ parent_f(int argc, char **argv)
}
mntpt = fs->fs_dir;
- verbose_flag = 0;
-
- while ((c = getopt(argc, argv, "cpv")) != EOF) {
+ while ((c = getopt(argc, argv, "p")) != EOF) {
switch (c) {
- case 'c':
- check_flag = 1;
- break;
case 'p':
listpath_flag = 1;
break;
- case 'v':
- verbose_flag++;
- break;
default:
return command_usage(&parent_cmd);
}
}
- if (!check_flag && !listpath_flag) /* default case */
- exitcode = parent_list(listpath_flag);
- else {
- if (listpath_flag)
- exitcode = parent_list(listpath_flag);
- if (check_flag)
- exitcode = parent_check();
+ /*
+ * Always initialize the fshandle table because we need it for
+ * the ppaths functions to work.
+ */
+ ret = path_to_fshandle((char *)mntpt, &hanp, &hlen);
+ if (ret) {
+ perror(mntpt);
+ return 0;
}
+ if (optind + 2 == argc) {
+ ino = strtoull(argv[optind], &p, 0);
+ if (*p != '\0' || ino == 0) {
+ fprintf(stderr,
+ _("Bad inode number '%s'.\n"),
+ argv[optind]);
+ return 0;
+ }
+ gen = strtoul(argv[optind + 1], &p, 0);
+ if (*p != '\0') {
+ fprintf(stderr,
+ _("Bad generation number '%s'.\n"),
+ argv[optind + 1]);
+ return 0;
+ }
+
+ memcpy(&handle, hanp, sizeof(handle));
+ handle.ha_fid.fid_len = sizeof(xfs_fid_t) -
+ sizeof(handle.ha_fid.fid_len);
+ handle.ha_fid.fid_pad = 0;
+ handle.ha_fid.fid_ino = ino;
+ handle.ha_fid.fid_gen = gen;
+
+ } else if (optind != argc) {
+ return command_usage(&parent_cmd);
+ }
+
+ if (listpath_flag)
+ exitcode = print_paths(ino ? &handle : NULL);
+ else
+ exitcode = print_parents(ino ? &handle : NULL);
+
+ if (hanp)
+ free_handle(hanp, hlen);
+
return 0;
}
static void
parent_help(void)
{
- printf(_(
+printf(_(
"\n"
" list the current file's parents and their filenames\n"
"\n"
-" -c -- check the current file's file system for parent consistency\n"
-" -p -- list the current file's parents and their full paths\n"
-" -v -- verbose mode\n"
+" -p -- list the current file's paths up to the root\n"
+"\n"
+"If ino and gen are supplied, use them instead.\n"
"\n"));
}
@@ -430,11 +205,10 @@ parent_init(void)
parent_cmd.cfunc = parent_f;
parent_cmd.argmin = 0;
parent_cmd.argmax = -1;
- parent_cmd.args = _("[-cpv]");
+ parent_cmd.args = _("[-p] [ino gen]");
parent_cmd.flags = CMD_NOMAP_OK;
- parent_cmd.oneline = _("print or check parent inodes");
+ parent_cmd.oneline = _("print parent inodes");
parent_cmd.help = parent_help;
- if (expert)
- add_command(&parent_cmd);
+ add_command(&parent_cmd);
}
diff --git a/man/man8/xfs_io.8 b/man/man8/xfs_io.8
index 6ebb479a344..4eda47c2772 100644
--- a/man/man8/xfs_io.8
+++ b/man/man8/xfs_io.8
@@ -1016,25 +1016,26 @@ and
options behave as described above, in
.B chproj.
.TP
-.BR parent " [ " \-cpv " ]"
+.BR parent " [ " \-p " ] [" " ino gen " "]"
By default this command prints out the parent inode numbers,
inode generation numbers and basenames of all the hardlinks which
point to the inode of the current file.
+
+If the optional
+.B ino
+and
+.B gen
+parameters are provided, they will be used to create a file handle on the same
+filesystem as the open file.
+The parents of the file represented by the handle will be reported instead of
+the open file.
+
.RS 1.0i
.PD 0
.TP 0.4i
.B \-p
the output is similar to the default output except pathnames up to
the mount-point are printed out instead of the component name.
-.TP
-.B \-c
-the file's filesystem will check all the parent attributes for consistency.
-.TP
-.B \-v
-verbose output will be printed.
-.RE
-.IP
-.B [NOTE: Not currently operational on Linux.]
.RE
.PD
.TP
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 19/32] xfs_io: Add i, n and f flags to parent command
2023-12-31 19:50 ` [PATCHSET v13.0 3/6] xfsprogs: Parent Pointers Darrick J. Wong
` (17 preceding siblings ...)
2023-12-31 23:09 ` [PATCH 18/32] xfs_io: adapt parent command to new parent pointer ioctls Darrick J. Wong
@ 2023-12-31 23:10 ` Darrick J. Wong
2023-12-31 23:10 ` [PATCH 20/32] xfs_logprint: decode parent pointers in ATTRI items fully Darrick J. Wong
` (12 subsequent siblings)
31 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:10 UTC (permalink / raw)
To: djwong, cem
Cc: Allison Henderson, catherine.hoang, linux-xfs, allison.henderson
From: Allison Henderson <allison.henderson@oracle.com>
This patch adds the flags i, n, and f to the parent command. These flags add
filtering options that are used by the new parent pointer tests in xfstests, and
help to improve the test run time. The flags are:
-i: Only show parent pointer records containing the given inode
-n: Only show parent pointer records containing the given filename
-f: Print records in short format: ino/gen/namelen/name
Signed-off-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
[djwong: adapt to new getparents ioctl]
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
io/parent.c | 89 +++++++++++++++++++++++++++++++++++++++++++++++------
man/man8/xfs_io.8 | 11 ++++++-
2 files changed, 89 insertions(+), 11 deletions(-)
diff --git a/io/parent.c b/io/parent.c
index 65fd892bffc..6bb7571e1bd 100644
--- a/io/parent.c
+++ b/io/parent.c
@@ -15,11 +15,18 @@
static cmdinfo_t parent_cmd;
static char *mntpt;
+struct pptr_args {
+ uint64_t filter_ino;
+ char *filter_name;
+ bool shortformat;
+};
+
static int
pptr_print(
const struct parent_rec *rec,
void *arg)
{
+ struct pptr_args *args = arg;
const char *name = (char *)rec->p_name;
unsigned int namelen;
@@ -28,7 +35,22 @@ pptr_print(
return 0;
}
+ if (args->filter_ino && rec->p_ino != args->filter_ino)
+ return 0;
+ if (args->filter_name && strcmp(args->filter_name, name))
+ return 0;
+
namelen = strlen(name);
+
+ if (args->shortformat) {
+ printf("%llu/%u/%u/%s\n",
+ (unsigned long long)rec->p_ino,
+ (unsigned int)rec->p_gen,
+ namelen,
+ rec->p_name);
+ return 0;
+ }
+
printf(_("p_ino = %llu\n"), (unsigned long long)rec->p_ino);
printf(_("p_gen = %u\n"), (unsigned int)rec->p_gen);
printf(_("p_namelen = %u\n"), namelen);
@@ -39,32 +61,55 @@ pptr_print(
static int
print_parents(
- struct xfs_handle *handle)
+ struct xfs_handle *handle,
+ struct pptr_args *args)
{
int ret;
if (handle)
ret = handle_walk_parents(handle, sizeof(*handle), pptr_print,
- NULL);
+ args);
else
- ret = fd_walk_parents(file->fd, pptr_print, NULL);
+ ret = fd_walk_parents(file->fd, pptr_print, args);
if (ret)
fprintf(stderr, "%s: %s\n", file->name, strerror(ret));
return 0;
}
+static int
+filter_path_components(
+ const char *name,
+ uint64_t ino,
+ void *arg)
+{
+ struct pptr_args *args = arg;
+
+ if (args->filter_ino && ino == args->filter_ino)
+ return ECANCELED;
+ if (args->filter_name && !strcmp(args->filter_name, name))
+ return ECANCELED;
+ return 0;
+}
+
static int
path_print(
const char *mntpt,
const struct path_list *path,
void *arg)
{
+ struct pptr_args *args = arg;
char buf[PATH_MAX];
size_t len = PATH_MAX;
int mntpt_len = strlen(mntpt);
int ret;
+ if (args->filter_ino || args->filter_name) {
+ ret = path_walk_components(path, filter_path_components, args);
+ if (ret != ECANCELED)
+ return 0;
+ }
+
/* Trim trailing slashes from the mountpoint */
while (mntpt_len > 0 && mntpt[mntpt_len - 1] == '/')
mntpt_len--;
@@ -83,15 +128,16 @@ path_print(
static int
print_paths(
- struct xfs_handle *handle)
+ struct xfs_handle *handle,
+ struct pptr_args *args)
{
int ret;
if (handle)
ret = handle_walk_parent_paths(handle, sizeof(*handle),
- path_print, NULL);
+ path_print, args);
else
- ret = fd_walk_parent_paths(file->fd, path_print, NULL);
+ ret = fd_walk_parent_paths(file->fd, path_print, args);
if (ret)
fprintf(stderr, "%s: %s\n", file->name, strerror(ret));
return 0;
@@ -103,6 +149,7 @@ parent_f(
char **argv)
{
struct xfs_handle handle;
+ struct pptr_args args = { 0 };
void *hanp = NULL;
size_t hlen;
struct fs_path *fs;
@@ -127,11 +174,27 @@ parent_f(
}
mntpt = fs->fs_dir;
- while ((c = getopt(argc, argv, "p")) != EOF) {
+ while ((c = getopt(argc, argv, "pfi:n:")) != EOF) {
switch (c) {
case 'p':
listpath_flag = 1;
break;
+ case 'i':
+ args.filter_ino = strtoull(optarg, &p, 0);
+ if (*p != '\0' || args.filter_ino == 0) {
+ fprintf(stderr,
+ _("Bad inode number '%s'.\n"),
+ optarg);
+ return 0;
+ }
+
+ break;
+ case 'n':
+ args.filter_name = optarg;
+ break;
+ case 'f':
+ args.shortformat = true;
+ break;
default:
return command_usage(&parent_cmd);
}
@@ -175,9 +238,9 @@ parent_f(
}
if (listpath_flag)
- exitcode = print_paths(ino ? &handle : NULL);
+ exitcode = print_paths(ino ? &handle : NULL, &args);
else
- exitcode = print_parents(ino ? &handle : NULL);
+ exitcode = print_parents(ino ? &handle : NULL, &args);
if (hanp)
free_handle(hanp, hlen);
@@ -195,6 +258,12 @@ printf(_(
" -p -- list the current file's paths up to the root\n"
"\n"
"If ino and gen are supplied, use them instead.\n"
+"\n"
+" -i -- Only show parent pointer records containing the given inode\n"
+"\n"
+" -n -- Only show parent pointer records containing the given filename\n"
+"\n"
+" -f -- Print records in short format: ino/gen/namelen/filename\n"
"\n"));
}
@@ -205,7 +274,7 @@ parent_init(void)
parent_cmd.cfunc = parent_f;
parent_cmd.argmin = 0;
parent_cmd.argmax = -1;
- parent_cmd.args = _("[-p] [ino gen]");
+ parent_cmd.args = _("[-p] [ino gen] [-i ino] [-n name] [-f]");
parent_cmd.flags = CMD_NOMAP_OK;
parent_cmd.oneline = _("print parent inodes");
parent_cmd.help = parent_help;
diff --git a/man/man8/xfs_io.8 b/man/man8/xfs_io.8
index 4eda47c2772..aa9907c1e5e 100644
--- a/man/man8/xfs_io.8
+++ b/man/man8/xfs_io.8
@@ -1016,7 +1016,7 @@ and
options behave as described above, in
.B chproj.
.TP
-.BR parent " [ " \-p " ] [" " ino gen " "]"
+.BR parent " [ " \-fp " ] [-i " ino "] [-n " name "] [" " ino gen " "]"
By default this command prints out the parent inode numbers,
inode generation numbers and basenames of all the hardlinks which
point to the inode of the current file.
@@ -1033,6 +1033,15 @@ the open file.
.RS 1.0i
.PD 0
.TP 0.4i
+.B \-f
+Print records in short format: ino/gen/namelen/name
+.TP 0.4i
+.B \-i
+Only show parent pointer records containing this inode number.
+.TP 0.4i
+.B \-n
+Only show parent pointer records containing this directory entry name.
+.TP 0.4i
.B \-p
the output is similar to the default output except pathnames up to
the mount-point are printed out instead of the component name.
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 20/32] xfs_logprint: decode parent pointers in ATTRI items fully
2023-12-31 19:50 ` [PATCHSET v13.0 3/6] xfsprogs: Parent Pointers Darrick J. Wong
` (18 preceding siblings ...)
2023-12-31 23:10 ` [PATCH 19/32] xfs_io: Add i, n and f flags to parent command Darrick J. Wong
@ 2023-12-31 23:10 ` Darrick J. Wong
2023-12-31 23:10 ` [PATCH 21/32] xfs_spaceman: report file paths Darrick J. Wong
` (11 subsequent siblings)
31 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:10 UTC (permalink / raw)
To: djwong, cem
Cc: Allison Henderson, catherine.hoang, linux-xfs, allison.henderson
From: Allison Henderson <allison.henderson@oracle.com>
This patch modifies the ATTRI print routines to look for the parent
pointer flag, and decode logged parent pointers fully when dumping log
contents. Between the existing ATTRI: printouts and the new ones
introduced here, we can figure out what was stored in each log iovec,
as well as the higher level parent pointer that was logged.
Signed-off-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
[djwong: adjust to new ondisk format]
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
libxfs/libxfs_api_defs.h | 4 ++
logprint/log_redo.c | 81 ++++++++++++++++++++++++++++++++++++++++++++++
2 files changed, 85 insertions(+)
diff --git a/libxfs/libxfs_api_defs.h b/libxfs/libxfs_api_defs.h
index 1b69124767c..c94972fb84b 100644
--- a/libxfs/libxfs_api_defs.h
+++ b/libxfs/libxfs_api_defs.h
@@ -177,6 +177,10 @@
#define xfs_log_sb libxfs_log_sb
#define xfs_mode_to_ftype libxfs_mode_to_ftype
#define xfs_mkdir_space_res libxfs_mkdir_space_res
+#define xfs_parent_add libxfs_parent_add
+#define xfs_parent_finish libxfs_parent_finish
+#define xfs_parent_irec_from_disk libxfs_parent_irec_from_disk
+#define xfs_parent_start libxfs_parent_start
#define xfs_perag_get libxfs_perag_get
#define xfs_perag_hold libxfs_perag_hold
#define xfs_perag_put libxfs_perag_put
diff --git a/logprint/log_redo.c b/logprint/log_redo.c
index e6401bb293e..948924d5bcb 100644
--- a/logprint/log_redo.c
+++ b/logprint/log_redo.c
@@ -674,6 +674,59 @@ xfs_attri_copy_log_format(
return 1;
}
+static void
+dump_pptr(
+ const char *tag,
+ const void *name_ptr,
+ unsigned int name_len,
+ const void *value_ptr,
+ unsigned int value_len)
+{
+ struct xfs_parent_name_irec irec;
+
+ if (name_len < sizeof(struct xfs_parent_name_rec)) {
+ printf("PPTR: %s CORRUPT\n", tag);
+ return;
+ }
+
+ libxfs_parent_irec_from_disk(&irec, name_ptr, value_ptr, value_len);
+
+ printf("PPTR: %s attr_namelen %u attr_valuelen %u\n", tag, name_len, value_len);
+ printf("PPTR: %s parent_ino %llu parent_gen %u namehash 0x%x namelen %u name '%.*s'\n",
+ tag,
+ (unsigned long long)irec.p_ino,
+ irec.p_gen,
+ irec.p_namehash,
+ irec.p_namelen,
+ irec.p_namelen,
+ irec.p_name);
+}
+
+static void
+dump_pptr_update(
+ const void *name_ptr,
+ unsigned int name_len,
+ const void *new_name_ptr,
+ unsigned int new_name_len,
+ const void *value_ptr,
+ unsigned int value_len,
+ const void *new_value_ptr,
+ unsigned int new_value_len)
+{
+ if (new_name_ptr && name_ptr) {
+ dump_pptr("OLDNAME", name_ptr, name_len, value_ptr, value_len);
+ dump_pptr("NEWNAME", new_name_ptr, new_name_len, new_value_ptr,
+ new_value_len);
+ return;
+ }
+
+ if (name_ptr)
+ dump_pptr("NAME", name_ptr, name_len, value_ptr, value_len);
+ if (new_name_ptr)
+ dump_pptr("NEWNAME", new_name_ptr, new_name_len, new_value_ptr,
+ new_value_len);
+}
+
static inline unsigned int
xfs_attr_log_item_op(const struct xfs_attri_log_format *attrp)
{
@@ -688,6 +741,10 @@ xlog_print_trans_attri(
{
struct xfs_attri_log_format *src_f = NULL;
xlog_op_header_t *head = NULL;
+ void *name_ptr = NULL;
+ void *new_name_ptr = NULL;
+ void *value_ptr = NULL;
+ void *new_value_ptr = NULL;
uint dst_len;
unsigned int name_len = 0;
unsigned int new_name_len = 0;
@@ -741,6 +798,7 @@ xlog_print_trans_attri(
(*i)++;
head = (xlog_op_header_t *)*ptr;
xlog_print_op_header(head, *i, ptr);
+ name_ptr = *ptr;
error = xlog_print_trans_attri_name(ptr,
be32_to_cpu(head->oh_len), "name");
if (error)
@@ -752,6 +810,7 @@ xlog_print_trans_attri(
(*i)++;
head = (xlog_op_header_t *)*ptr;
xlog_print_op_header(head, *i, ptr);
+ new_name_ptr = *ptr;
error = xlog_print_trans_attri_name(ptr,
be32_to_cpu(head->oh_len), "newname");
if (error)
@@ -763,6 +822,7 @@ xlog_print_trans_attri(
(*i)++;
head = (xlog_op_header_t *)*ptr;
xlog_print_op_header(head, *i, ptr);
+ value_ptr = *ptr;
error = xlog_print_trans_attri_value(ptr,
be32_to_cpu(head->oh_len), value_len, "value");
if (error)
@@ -774,12 +834,19 @@ xlog_print_trans_attri(
(*i)++;
head = (xlog_op_header_t *)*ptr;
xlog_print_op_header(head, *i, ptr);
+ new_value_ptr = *ptr;
error = xlog_print_trans_attri_value(ptr,
be32_to_cpu(head->oh_len), new_value_len,
"newvalue");
if (error)
goto error;
}
+
+ if (src_f->alfi_attr_filter & XFS_ATTR_PARENT)
+ dump_pptr_update(name_ptr, name_len,
+ new_name_ptr, new_name_len,
+ value_ptr, value_len,
+ new_value_ptr, new_value_len);
error:
free(src_f);
@@ -822,6 +889,10 @@ xlog_recover_print_attri(
struct xlog_recover_item *item)
{
struct xfs_attri_log_format *f, *src_f = NULL;
+ void *name_ptr = NULL;
+ void *new_name_ptr = NULL;
+ void *value_ptr = NULL;
+ void *new_value_ptr = NULL;
uint src_len, dst_len;
unsigned int name_len = 0;
unsigned int new_name_len = 0;
@@ -872,6 +943,7 @@ xlog_recover_print_attri(
printf(_("ATTRI: name len:%u\n"), name_len);
print_or_dump((char *)item->ri_buf[region].i_addr,
name_len);
+ name_ptr = item->ri_buf[region].i_addr;
}
if (new_name_len > 0) {
@@ -879,6 +951,7 @@ xlog_recover_print_attri(
printf(_("ATTRI: newname len:%u\n"), new_name_len);
print_or_dump((char *)item->ri_buf[region].i_addr,
new_name_len);
+ new_name_ptr = item->ri_buf[region].i_addr;
}
if (value_len > 0) {
@@ -887,6 +960,7 @@ xlog_recover_print_attri(
region++;
printf(_("ATTRI: value len:%u\n"), value_len);
print_or_dump((char *)item->ri_buf[region].i_addr, len);
+ value_ptr = item->ri_buf[region].i_addr;
}
if (new_value_len > 0) {
@@ -895,8 +969,15 @@ xlog_recover_print_attri(
region++;
printf(_("ATTRI: newvalue len:%u\n"), new_value_len);
print_or_dump((char *)item->ri_buf[region].i_addr, len);
+ new_value_ptr = item->ri_buf[region].i_addr;
}
+ if (src_f->alfi_attr_filter & XFS_ATTR_PARENT)
+ dump_pptr_update(name_ptr, name_len,
+ new_name_ptr, new_name_len,
+ value_ptr, value_len,
+ new_value_ptr, new_value_len);
+
out:
free(f);
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 21/32] xfs_spaceman: report file paths
2023-12-31 19:50 ` [PATCHSET v13.0 3/6] xfsprogs: Parent Pointers Darrick J. Wong
` (19 preceding siblings ...)
2023-12-31 23:10 ` [PATCH 20/32] xfs_logprint: decode parent pointers in ATTRI items fully Darrick J. Wong
@ 2023-12-31 23:10 ` Darrick J. Wong
2023-12-31 23:10 ` [PATCH 22/32] xfs_scrub: use parent pointers when possible to report file operations Darrick J. Wong
` (10 subsequent siblings)
31 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:10 UTC (permalink / raw)
To: djwong, cem; +Cc: catherine.hoang, linux-xfs, allison.henderson
From: Darrick J. Wong <djwong@kernel.org>
Teach the health command to report file paths when possible.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
man/man8/xfs_spaceman.8 | 7 +++++-
spaceman/Makefile | 4 ++--
spaceman/file.c | 7 ++++++
spaceman/health.c | 53 ++++++++++++++++++++++++++++++++++++++---------
spaceman/space.h | 3 +++
5 files changed, 61 insertions(+), 13 deletions(-)
diff --git a/man/man8/xfs_spaceman.8 b/man/man8/xfs_spaceman.8
index ece840d7300..0d299132a78 100644
--- a/man/man8/xfs_spaceman.8
+++ b/man/man8/xfs_spaceman.8
@@ -91,7 +91,7 @@ The output will have the same format that
.BR "xfs_info" "(8)"
prints when querying a filesystem.
.TP
-.BI "health [ \-a agno] [ \-c ] [ \-f ] [ \-i inum ] [ \-q ] [ paths ]"
+.BI "health [ \-a agno] [ \-c ] [ \-f ] [ \-i inum ] [ \-n ] [ \-q ] [ paths ]"
Reports the health of the given group of filesystem metadata.
.RS 1.0i
.PD 0
@@ -111,6 +111,11 @@ Report on the health of metadata that affect the entire filesystem.
.B \-i inum
Report on the health of a specific inode.
.TP
+.B \-n
+When reporting on the health of a file, try to report the full file path,
+if possible.
+This option is disabled by default to minimize runtime.
+.TP
.B \-q
Report only unhealthy metadata.
.TP
diff --git a/spaceman/Makefile b/spaceman/Makefile
index 1f048d54a4d..d6fccc361cf 100644
--- a/spaceman/Makefile
+++ b/spaceman/Makefile
@@ -10,8 +10,8 @@ HFILES = init.h space.h
CFILES = info.c init.c file.c health.c prealloc.c trim.c
LSRCFILES = xfs_info.sh
-LLDLIBS = $(LIBXCMD) $(LIBFROG)
-LTDEPENDENCIES = $(LIBXCMD) $(LIBFROG)
+LLDLIBS = $(LIBHANDLE) $(LIBXCMD) $(LIBFROG)
+LTDEPENDENCIES = $(LIBHANDLE) $(LIBXCMD) $(LIBFROG)
LLDFLAGS = -static
ifeq ($(ENABLE_EDITLINE),yes)
diff --git a/spaceman/file.c b/spaceman/file.c
index eec7ee9f4ba..850688ace15 100644
--- a/spaceman/file.c
+++ b/spaceman/file.c
@@ -14,6 +14,7 @@
#include "libfrog/paths.h"
#include "libfrog/fsgeom.h"
#include "space.h"
+#include "handle.h"
static cmdinfo_t print_cmd;
@@ -106,6 +107,12 @@ addfile(
file->name = filename;
memcpy(&file->xfd, xfd, sizeof(struct xfs_fd));
memcpy(&file->fs_path, fs_path, sizeof(file->fs_path));
+
+ /* Try to capture a fs handle for reporting paths. */
+ file->fshandle = NULL;
+ file->fshandle_len = 0;
+ path_to_fshandle(filename, &file->fshandle, &file->fshandle_len);
+
return 0;
}
diff --git a/spaceman/health.c b/spaceman/health.c
index 12fb67bab28..ab5bc074988 100644
--- a/spaceman/health.c
+++ b/spaceman/health.c
@@ -13,11 +13,13 @@
#include "libfrog/fsgeom.h"
#include "libfrog/bulkstat.h"
#include "space.h"
+#include "libfrog/getparents.h"
static cmdinfo_t health_cmd;
static unsigned long long reported;
static bool comprehensive;
static bool quiet;
+static bool report_paths;
static bool has_realtime(const struct xfs_fsop_geom *g)
{
@@ -269,6 +271,38 @@ report_file_health(
#define BULKSTAT_NR (128)
+static void
+report_inode(
+ const struct xfs_bulkstat *bs)
+{
+ char descr[PATH_MAX];
+ int ret;
+
+ if (report_paths && file->fshandle &&
+ (file->xfd.fsgeom.flags & XFS_FSOP_GEOM_FLAGS_PARENT)) {
+ struct xfs_handle handle;
+
+ memcpy(&handle.ha_fsid, file->fshandle, sizeof(handle.ha_fsid));
+ handle.ha_fid.fid_len = sizeof(xfs_fid_t) -
+ sizeof(handle.ha_fid.fid_len);
+ handle.ha_fid.fid_pad = 0;
+ handle.ha_fid.fid_ino = bs->bs_ino;
+ handle.ha_fid.fid_gen = bs->bs_gen;
+
+ ret = handle_to_path(&handle, sizeof(struct xfs_handle), descr,
+ sizeof(descr) - 1);
+ if (ret)
+ goto report_inum;
+
+ goto report_status;
+ }
+
+report_inum:
+ snprintf(descr, sizeof(descr) - 1, _("inode %"PRIu64), bs->bs_ino);
+report_status:
+ report_sick(descr, inode_flags, bs->bs_sick, bs->bs_checked);
+}
+
/*
* Report on all files' health for a given @agno. If @agno is NULLAGNUMBER,
* report on all files in the filesystem.
@@ -278,7 +312,6 @@ report_bulkstat_health(
xfs_agnumber_t agno)
{
struct xfs_bulkstat_req *breq;
- char descr[256];
uint32_t i;
int error;
@@ -296,13 +329,8 @@ report_bulkstat_health(
error = -xfrog_bulkstat(&file->xfd, breq);
if (error)
break;
- for (i = 0; i < breq->hdr.ocount; i++) {
- snprintf(descr, sizeof(descr) - 1, _("inode %"PRIu64),
- breq->bulkstat[i].bs_ino);
- report_sick(descr, inode_flags,
- breq->bulkstat[i].bs_sick,
- breq->bulkstat[i].bs_checked);
- }
+ for (i = 0; i < breq->hdr.ocount; i++)
+ report_inode(&breq->bulkstat[i]);
} while (breq->hdr.ocount > 0);
if (error)
@@ -312,7 +340,7 @@ report_bulkstat_health(
return error;
}
-#define OPT_STRING ("a:cfi:q")
+#define OPT_STRING ("a:cfi:nq")
/* Report on health problems in XFS filesystem. */
static int
@@ -327,6 +355,7 @@ health_f(
int ret;
reported = 0;
+ report_paths = false;
if (file->xfd.fsgeom.version != XFS_FSOP_GEOM_VERSION_V5) {
perror("health");
@@ -362,6 +391,9 @@ health_f(
return 1;
}
break;
+ case 'n':
+ report_paths = true;
+ break;
case 'q':
quiet = true;
break;
@@ -449,6 +481,7 @@ health_help(void)
" -c -- Report on the health of all inodes.\n"
" -f -- Report health of the overall filesystem.\n"
" -i inum -- Report health of a given inode number.\n"
+" -n -- Try to report file names.\n"
" -q -- Only report unhealthy metadata.\n"
" paths -- Report health of the given file path.\n"
"\n"));
@@ -460,7 +493,7 @@ static cmdinfo_t health_cmd = {
.cfunc = health_f,
.argmin = 0,
.argmax = -1,
- .args = "[-a agno] [-c] [-f] [-i inum] [-q] [paths]",
+ .args = "[-a agno] [-c] [-f] [-i inum] [-n] [-q] [paths]",
.flags = CMD_FLAG_ONESHOT,
.help = health_help,
};
diff --git a/spaceman/space.h b/spaceman/space.h
index 723209edd99..28fa35a3047 100644
--- a/spaceman/space.h
+++ b/spaceman/space.h
@@ -10,6 +10,9 @@ struct fileio {
struct xfs_fd xfd; /* XFS runtime support context */
struct fs_path fs_path; /* XFS path information */
char *name; /* file name at time of open */
+
+ void *fshandle;
+ size_t fshandle_len;
};
extern struct fileio *filetable; /* open file table */
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 22/32] xfs_scrub: use parent pointers when possible to report file operations
2023-12-31 19:50 ` [PATCHSET v13.0 3/6] xfsprogs: Parent Pointers Darrick J. Wong
` (20 preceding siblings ...)
2023-12-31 23:10 ` [PATCH 21/32] xfs_spaceman: report file paths Darrick J. Wong
@ 2023-12-31 23:10 ` Darrick J. Wong
2023-12-31 23:11 ` [PATCH 23/32] xfs_db: report parent pointers in version command Darrick J. Wong
` (9 subsequent siblings)
31 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:10 UTC (permalink / raw)
To: djwong, cem; +Cc: catherine.hoang, linux-xfs, allison.henderson
From: Darrick J. Wong <djwong@kernel.org>
If parent pointers are available, use them to supply file paths when
doing things to files, instead of merely printing the inode number.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
scrub/common.c | 41 +++++++++++++++++++++++++++++++++++++++--
1 file changed, 39 insertions(+), 2 deletions(-)
diff --git a/scrub/common.c b/scrub/common.c
index aca59648711..248a33ef324 100644
--- a/scrub/common.c
+++ b/scrub/common.c
@@ -9,6 +9,7 @@
#include <syslog.h>
#include "platform_defs.h"
#include "libfrog/paths.h"
+#include "libfrog/getparents.h"
#include "xfs_scrub.h"
#include "common.h"
#include "progress.h"
@@ -405,19 +406,55 @@ scrub_render_ino_descr(
...)
{
va_list args;
+ size_t pathlen = 0;
uint32_t agno;
uint32_t agino;
int ret;
+ if (ctx->mnt.fsgeom.flags & XFS_FSOP_GEOM_FLAGS_PARENT) {
+ struct xfs_handle handle;
+
+ memcpy(&handle.ha_fsid, ctx->fshandle, sizeof(handle.ha_fsid));
+ handle.ha_fid.fid_len = sizeof(xfs_fid_t) -
+ sizeof(handle.ha_fid.fid_len);
+ handle.ha_fid.fid_pad = 0;
+ handle.ha_fid.fid_ino = ino;
+ handle.ha_fid.fid_gen = gen;
+
+ ret = handle_to_path(&handle, sizeof(struct xfs_handle), buf,
+ buflen);
+ if (ret)
+ goto report_inum;
+
+ /*
+ * Leave at least 16 bytes for the description of what went
+ * wrong. If we can't do that, we'll use the inode number.
+ */
+ pathlen = strlen(buf);
+ if (pathlen >= buflen - 16)
+ goto report_inum;
+
+ if (format) {
+ buf[pathlen] = ' ';
+ buf[pathlen + 1] = 0;
+ pathlen++;
+ }
+
+ goto report_format;
+ }
+
+report_inum:
agno = cvt_ino_to_agno(&ctx->mnt, ino);
agino = cvt_ino_to_agino(&ctx->mnt, ino);
ret = snprintf(buf, buflen, _("inode %"PRIu64" (%"PRIu32"/%"PRIu32")%s"),
ino, agno, agino, format ? " " : "");
if (ret < 0 || ret >= buflen || format == NULL)
return ret;
+ pathlen = ret;
+report_format:
va_start(args, format);
- ret += vsnprintf(buf + ret, buflen - ret, format, args);
+ pathlen += vsnprintf(buf + pathlen, buflen - pathlen, format, args);
va_end(args);
- return ret;
+ return pathlen;
}
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 23/32] xfs_db: report parent pointers in version command
2023-12-31 19:50 ` [PATCHSET v13.0 3/6] xfsprogs: Parent Pointers Darrick J. Wong
` (21 preceding siblings ...)
2023-12-31 23:10 ` [PATCH 22/32] xfs_scrub: use parent pointers when possible to report file operations Darrick J. Wong
@ 2023-12-31 23:11 ` Darrick J. Wong
2023-12-31 23:11 ` [PATCH 24/32] xfs_db: report parent bit on xattrs Darrick J. Wong
` (8 subsequent siblings)
31 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:11 UTC (permalink / raw)
To: djwong, cem; +Cc: catherine.hoang, linux-xfs, allison.henderson
From: Darrick J. Wong <djwong@kernel.org>
Report the presents of PARENT pointers from the version subcommand.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
db/sb.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/db/sb.c b/db/sb.c
index b48767f47fe..9a5d665dfbd 100644
--- a/db/sb.c
+++ b/db/sb.c
@@ -706,6 +706,8 @@ version_string(
strcat(s, ",NEEDSREPAIR");
if (xfs_has_large_extent_counts(mp))
strcat(s, ",NREXT64");
+ if (xfs_has_parent(mp))
+ strcat(s, ",PARENT");
return s;
}
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 24/32] xfs_db: report parent bit on xattrs
2023-12-31 19:50 ` [PATCHSET v13.0 3/6] xfsprogs: Parent Pointers Darrick J. Wong
` (22 preceding siblings ...)
2023-12-31 23:11 ` [PATCH 23/32] xfs_db: report parent pointers in version command Darrick J. Wong
@ 2023-12-31 23:11 ` Darrick J. Wong
2023-12-31 23:11 ` [PATCH 25/32] xfs_db: report parent pointers embedded in xattrs Darrick J. Wong
` (7 subsequent siblings)
31 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:11 UTC (permalink / raw)
To: djwong, cem
Cc: Allison Henderson, catherine.hoang, linux-xfs, allison.henderson
From: Darrick J. Wong <djwong@kernel.org>
Display the parent bit on xattr keys
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
---
db/attr.c | 3 +++
db/attrshort.c | 3 +++
2 files changed, 6 insertions(+)
diff --git a/db/attr.c b/db/attr.c
index ba722e146e2..f29e4a54454 100644
--- a/db/attr.c
+++ b/db/attr.c
@@ -82,6 +82,9 @@ const field_t attr_leaf_entry_flds[] = {
{ "local", FLDT_UINT1,
OI(LEOFF(flags) + bitsz(uint8_t) - XFS_ATTR_LOCAL_BIT - 1), C1, 0,
TYP_NONE },
+ { "parent", FLDT_UINT1,
+ OI(LEOFF(flags) + bitsz(uint8_t) - XFS_ATTR_PARENT_BIT - 1), C1, 0,
+ TYP_NONE },
{ "pad2", FLDT_UINT8X, OI(LEOFF(pad2)), C1, FLD_SKIPALL, TYP_NONE },
{ NULL }
};
diff --git a/db/attrshort.c b/db/attrshort.c
index e234fbd8365..872d771d5ed 100644
--- a/db/attrshort.c
+++ b/db/attrshort.c
@@ -44,6 +44,9 @@ const field_t attr_sf_entry_flds[] = {
{ "secure", FLDT_UINT1,
OI(EOFF(flags) + bitsz(uint8_t) - XFS_ATTR_SECURE_BIT - 1), C1, 0,
TYP_NONE },
+ { "parent", FLDT_UINT1,
+ OI(EOFF(flags) + bitsz(uint8_t) - XFS_ATTR_PARENT_BIT - 1), C1, 0,
+ TYP_NONE },
{ "name", FLDT_CHARNS, OI(EOFF(nameval)), attr_sf_entry_name_count,
FLD_COUNT, TYP_NONE },
{ "value", FLDT_CHARNS, attr_sf_entry_value_offset,
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 25/32] xfs_db: report parent pointers embedded in xattrs
2023-12-31 19:50 ` [PATCHSET v13.0 3/6] xfsprogs: Parent Pointers Darrick J. Wong
` (23 preceding siblings ...)
2023-12-31 23:11 ` [PATCH 24/32] xfs_db: report parent bit on xattrs Darrick J. Wong
@ 2023-12-31 23:11 ` Darrick J. Wong
2023-12-31 23:11 ` [PATCH 26/32] xfs_db: obfuscate dirent and parent pointer names consistently Darrick J. Wong
` (6 subsequent siblings)
31 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:11 UTC (permalink / raw)
To: djwong, cem; +Cc: catherine.hoang, linux-xfs, allison.henderson
From: Darrick J. Wong <djwong@kernel.org>
Decode the parent pointer inode, generation, namehash, and name fields
if the parent pointer passes basic validation checks.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
db/attr.c | 62 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++
db/attrshort.c | 48 +++++++++++++++++++++++++++++++++++++++++++
2 files changed, 110 insertions(+)
diff --git a/db/attr.c b/db/attr.c
index f29e4a54454..9e7bbd164df 100644
--- a/db/attr.c
+++ b/db/attr.c
@@ -19,6 +19,8 @@ static int attr_leaf_entries_count(void *obj, int startoff);
static int attr_leaf_hdr_count(void *obj, int startoff);
static int attr_leaf_name_local_count(void *obj, int startoff);
static int attr_leaf_name_local_name_count(void *obj, int startoff);
+static int attr_leaf_name_pptr_count(void *obj, int startoff);
+static int attr_leaf_name_pptr_namelen(void *obj, int startoff);
static int attr_leaf_name_local_value_count(void *obj, int startoff);
static int attr_leaf_name_local_value_offset(void *obj, int startoff,
int idx);
@@ -111,6 +113,8 @@ const field_t attr_leaf_map_flds[] = {
#define LNOFF(f) bitize(offsetof(xfs_attr_leaf_name_local_t, f))
#define LVOFF(f) bitize(offsetof(xfs_attr_leaf_name_remote_t, f))
+#define PPOFF(f) bitize(offsetof(xfs_attr_leaf_name_local_t, nameval) + \
+ offsetof(struct xfs_parent_name_rec, f))
const field_t attr_leaf_name_flds[] = {
{ "valuelen", FLDT_UINT16D, OI(LNOFF(valuelen)),
attr_leaf_name_local_count, FLD_COUNT, TYP_NONE },
@@ -118,6 +122,14 @@ const field_t attr_leaf_name_flds[] = {
attr_leaf_name_local_count, FLD_COUNT, TYP_NONE },
{ "name", FLDT_CHARNS, OI(LNOFF(nameval)),
attr_leaf_name_local_name_count, FLD_COUNT, TYP_NONE },
+ { "parent_ino", FLDT_INO, OI(PPOFF(p_ino)),
+ attr_leaf_name_pptr_count, FLD_COUNT, TYP_INODE },
+ { "parent_gen", FLDT_UINT32D, OI(PPOFF(p_gen)),
+ attr_leaf_name_pptr_count, FLD_COUNT, TYP_NONE },
+ { "parent_namehash", FLDT_UINT32X, OI(PPOFF(p_namehash)),
+ attr_leaf_name_pptr_count, FLD_COUNT, TYP_NONE },
+ { "parent_name", FLDT_CHARNS, attr_leaf_name_local_value_offset,
+ attr_leaf_name_pptr_namelen, FLD_COUNT|FLD_OFFSET, TYP_NONE },
{ "value", FLDT_CHARNS, attr_leaf_name_local_value_offset,
attr_leaf_name_local_value_count, FLD_COUNT|FLD_OFFSET, TYP_NONE },
{ "valueblk", FLDT_UINT32X, OI(LVOFF(valueblk)),
@@ -273,6 +285,26 @@ attr_leaf_name_local_count(
__attr_leaf_name_local_count);
}
+static int
+__attr_leaf_name_pptr_count(
+ struct xfs_attr_leafblock *leaf,
+ struct xfs_attr_leaf_entry *e,
+ int i)
+{
+ if ((e->flags & XFS_ATTR_NSP_ONDISK_MASK) == XFS_ATTR_PARENT)
+ return 1;
+ return 0;
+}
+
+static int
+attr_leaf_name_pptr_count(
+ void *obj,
+ int startoff)
+{
+ return attr_leaf_entry_walk(obj, startoff,
+ __attr_leaf_name_pptr_count);
+}
+
static int
__attr_leaf_name_local_name_count(
struct xfs_attr_leafblock *leaf,
@@ -283,6 +315,8 @@ __attr_leaf_name_local_name_count(
if (!(e->flags & XFS_ATTR_LOCAL))
return 0;
+ if ((e->flags & XFS_ATTR_NSP_ONDISK_MASK) == XFS_ATTR_PARENT)
+ return 0;
l = xfs_attr3_leaf_name_local(leaf, i);
return l->namelen;
@@ -297,6 +331,32 @@ attr_leaf_name_local_name_count(
__attr_leaf_name_local_name_count);
}
+static int
+__attr_leaf_name_pptr_namelen(
+ struct xfs_attr_leafblock *leaf,
+ struct xfs_attr_leaf_entry *e,
+ int i)
+{
+ struct xfs_attr_leaf_name_local *l;
+
+ if (!(e->flags & XFS_ATTR_LOCAL))
+ return 0;
+ if ((e->flags & XFS_ATTR_NSP_ONDISK_MASK) != XFS_ATTR_PARENT)
+ return 0;
+
+ l = xfs_attr3_leaf_name_local(leaf, i);
+ return be16_to_cpu(l->valuelen);
+}
+
+static int
+attr_leaf_name_pptr_namelen(
+ void *obj,
+ int startoff)
+{
+ return attr_leaf_entry_walk(obj, startoff,
+ __attr_leaf_name_pptr_namelen);
+}
+
static int
__attr_leaf_name_local_value_count(
struct xfs_attr_leafblock *leaf,
@@ -307,6 +367,8 @@ __attr_leaf_name_local_value_count(
if (!(e->flags & XFS_ATTR_LOCAL))
return 0;
+ if ((e->flags & XFS_ATTR_NSP_ONDISK_MASK) == XFS_ATTR_PARENT)
+ return 0;
l = xfs_attr3_leaf_name_local(leaf, i);
return be16_to_cpu(l->valuelen);
diff --git a/db/attrshort.c b/db/attrshort.c
index 872d771d5ed..9cd3411bee4 100644
--- a/db/attrshort.c
+++ b/db/attrshort.c
@@ -13,6 +13,8 @@
#include "attrshort.h"
static int attr_sf_entry_name_count(void *obj, int startoff);
+static int attr_sf_entry_pptr_count(void *obj, int startoff);
+static int attr_sf_entry_pptr_namelen(void *obj, int startoff);
static int attr_sf_entry_value_count(void *obj, int startoff);
static int attr_sf_entry_value_offset(void *obj, int startoff, int idx);
static int attr_shortform_list_count(void *obj, int startoff);
@@ -34,6 +36,8 @@ const field_t attr_sf_hdr_flds[] = {
};
#define EOFF(f) bitize(offsetof(struct xfs_attr_sf_entry, f))
+#define PPOFF(f) bitize(offsetof(struct xfs_attr_sf_entry, nameval) + \
+ offsetof(struct xfs_parent_name_rec, f))
const field_t attr_sf_entry_flds[] = {
{ "namelen", FLDT_UINT8D, OI(EOFF(namelen)), C1, 0, TYP_NONE },
{ "valuelen", FLDT_UINT8D, OI(EOFF(valuelen)), C1, 0, TYP_NONE },
@@ -49,11 +53,33 @@ const field_t attr_sf_entry_flds[] = {
TYP_NONE },
{ "name", FLDT_CHARNS, OI(EOFF(nameval)), attr_sf_entry_name_count,
FLD_COUNT, TYP_NONE },
+ { "parent_ino", FLDT_INO, OI(PPOFF(p_ino)), attr_sf_entry_pptr_count,
+ FLD_COUNT, TYP_INODE },
+ { "parent_gen", FLDT_UINT32D, OI(PPOFF(p_gen)), attr_sf_entry_pptr_count,
+ FLD_COUNT, TYP_NONE },
+ { "parent_namehash", FLDT_UINT32X, OI(PPOFF(p_namehash)),
+ attr_sf_entry_pptr_count, FLD_COUNT, TYP_NONE },
+ { "parent_name", FLDT_CHARNS, attr_sf_entry_value_offset,
+ attr_sf_entry_pptr_namelen, FLD_COUNT|FLD_OFFSET, TYP_NONE },
{ "value", FLDT_CHARNS, attr_sf_entry_value_offset,
attr_sf_entry_value_count, FLD_COUNT|FLD_OFFSET, TYP_NONE },
{ NULL }
};
+static int
+attr_sf_entry_pptr_count(
+ void *obj,
+ int startoff)
+{
+ struct xfs_attr_sf_entry *e;
+
+ ASSERT(bitoffs(startoff) == 0);
+ e = (struct xfs_attr_sf_entry *)((char *)obj + byteize(startoff));
+ if ((e->flags & XFS_ATTR_NSP_ONDISK_MASK) == XFS_ATTR_PARENT)
+ return 1;
+ return 0;
+}
+
static int
attr_sf_entry_name_count(
void *obj,
@@ -63,6 +89,8 @@ attr_sf_entry_name_count(
ASSERT(bitoffs(startoff) == 0);
e = (struct xfs_attr_sf_entry *)((char *)obj + byteize(startoff));
+ if ((e->flags & XFS_ATTR_NSP_ONDISK_MASK) == XFS_ATTR_PARENT)
+ return 0;
return e->namelen;
}
@@ -84,6 +112,22 @@ attr_sf_entry_size(
return bitize((int)xfs_attr_sf_entsize(e));
}
+static int
+attr_sf_entry_pptr_namelen(
+ void *obj,
+ int startoff)
+{
+ struct xfs_attr_sf_entry *e;
+
+ ASSERT(bitoffs(startoff) == 0);
+ e = (struct xfs_attr_sf_entry *)((char *)obj + byteize(startoff));
+
+ if ((e->flags & XFS_ATTR_NSP_ONDISK_MASK) != XFS_ATTR_PARENT)
+ return 0;
+
+ return e->valuelen;
+}
+
static int
attr_sf_entry_value_count(
void *obj,
@@ -93,6 +137,10 @@ attr_sf_entry_value_count(
ASSERT(bitoffs(startoff) == 0);
e = (struct xfs_attr_sf_entry *)((char *)obj + byteize(startoff));
+
+ if ((e->flags & XFS_ATTR_NSP_ONDISK_MASK) == XFS_ATTR_PARENT)
+ return 0;
+
return e->valuelen;
}
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 26/32] xfs_db: obfuscate dirent and parent pointer names consistently
2023-12-31 19:50 ` [PATCHSET v13.0 3/6] xfsprogs: Parent Pointers Darrick J. Wong
` (24 preceding siblings ...)
2023-12-31 23:11 ` [PATCH 25/32] xfs_db: report parent pointers embedded in xattrs Darrick J. Wong
@ 2023-12-31 23:11 ` Darrick J. Wong
2023-12-31 23:12 ` [PATCH 27/32] libxfs: export attr3_leaf_hdr_from_disk via libxfs_api_defs.h Darrick J. Wong
` (5 subsequent siblings)
31 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:11 UTC (permalink / raw)
To: djwong, cem; +Cc: catherine.hoang, linux-xfs, allison.henderson
From: Darrick J. Wong <djwong@kernel.org>
When someone wants to perform an obfuscated metadump of a filesystem
where parent pointers are enabled, we have to use the *exact* same
obfuscated name for both the directory entry and the parent pointer.
Create a name remapping table so that when we obfuscate a dirent name or
a parent pointer name, we can apply the same obfuscation when we find
the corresponding parent pointer or dirent.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
db/metadump.c | 323 ++++++++++++++++++++++++++++++++++++++++++++--
libxfs/libxfs_api_defs.h | 3
2 files changed, 315 insertions(+), 11 deletions(-)
diff --git a/db/metadump.c b/db/metadump.c
index bac35b9cc68..5f5a33335b0 100644
--- a/db/metadump.c
+++ b/db/metadump.c
@@ -21,6 +21,14 @@
#include "dir2.h"
#include "obfuscate.h"
+#undef REMAP_DEBUG
+
+#ifdef REMAP_DEBUG
+# define remap_debug printf
+#else
+# define remap_debug(...) ((void)0)
+#endif
+
#define DEFAULT_MAX_EXT_SIZE XFS_MAX_BMBT_EXTLEN
/* copy all metadata structures to/from a file */
@@ -719,6 +727,111 @@ nametable_add(xfs_dahash_t hash, int namelen, unsigned char *name)
return ent;
}
+/*
+ * Obfuscated name remapping table for parent pointer-enabled filesystems.
+ * When this feature is enabled, we have to maintain consistency between the
+ * names that appears in the dirent and the corresponding parent pointer.
+ */
+
+struct remap_ent {
+ struct remap_ent *next;
+ xfs_ino_t dir_ino;
+ xfs_dahash_t namehash;
+ uint8_t namelen;
+
+ uint8_t names[];
+};
+
+static inline uint8_t *remap_ent_before(struct remap_ent *ent)
+{
+ return &ent->names[0];
+}
+
+static inline uint8_t *remap_ent_after(struct remap_ent *ent)
+{
+ return &ent->names[ent->namelen];
+}
+
+#define REMAP_TABLE_SIZE 4096
+
+static struct remap_ent *remaptable[REMAP_TABLE_SIZE];
+
+static void
+remaptable_clear(void)
+{
+ int i;
+ struct remap_ent *ent, *next;
+
+ for (i = 0; i < REMAP_TABLE_SIZE; i++) {
+ ent = remaptable[i];
+
+ while (ent) {
+ next = ent->next;
+ free(ent);
+ ent = next;
+ }
+ }
+}
+
+/* Try to find a remapping table entry. */
+static struct remap_ent *
+remaptable_find(
+ xfs_ino_t dir_ino,
+ xfs_dahash_t namehash,
+ const unsigned char *name,
+ unsigned int namelen)
+{
+ struct remap_ent *ent = remaptable[namehash % REMAP_TABLE_SIZE];
+
+ remap_debug("REMAP FIND: 0x%lx hash 0x%x '%.*s'\n",
+ dir_ino, namehash, namelen, name);
+
+ while (ent) {
+ remap_debug("REMAP ENT: 0x%lx hash 0x%x '%.*s'\n",
+ ent->dir_ino, ent->namehash, ent->namelen,
+ remap_ent_before(ent));
+
+ if (ent->dir_ino == dir_ino &&
+ ent->namehash == namehash &&
+ ent->namelen == namelen &&
+ !memcmp(remap_ent_before(ent), name, namelen))
+ return ent;
+ ent = ent->next;
+ }
+
+ return NULL;
+}
+
+/* Remember the remapping for a particular dirent that we obfuscated. */
+static struct remap_ent *
+remaptable_add(
+ xfs_ino_t dir_ino,
+ xfs_dahash_t namehash,
+ const unsigned char *old_name,
+ unsigned int namelen,
+ const unsigned char *new_name)
+{
+ struct remap_ent *ent;
+
+ ent = malloc(sizeof(struct remap_ent) + (namelen * 2));
+ if (!ent)
+ return NULL;
+
+ ent->dir_ino = dir_ino;
+ ent->namehash = namehash;
+ ent->namelen = namelen;
+ memcpy(remap_ent_before(ent), old_name, namelen);
+ memcpy(remap_ent_after(ent), new_name, namelen);
+ ent->next = remaptable[namehash % REMAP_TABLE_SIZE];
+
+ remaptable[namehash % REMAP_TABLE_SIZE] = ent;
+
+ remap_debug("REMAP ADD: 0x%lx hash 0x%x '%.*s' -> '%.*s'\n",
+ dir_ino, namehash, namelen, old_name, namelen,
+ new_name);
+ return ent;
+}
+
#define ORPHANAGE "lost+found"
#define ORPHANAGE_LEN (sizeof (ORPHANAGE) - 1)
@@ -844,6 +957,7 @@ generate_obfuscated_name(
int namelen,
unsigned char *name)
{
+ unsigned char *orig_name = NULL;
xfs_dahash_t hash;
/*
@@ -865,8 +979,37 @@ generate_obfuscated_name(
name++;
/* Obfuscate the name (if possible) */
-
hash = dirattr_hashname(ino != 0, name, namelen);
+
+ /*
+ * If we're obfuscating a dirent name on a pptrs filesystem, see if we
+ * already processed the parent pointer and use the same name.
+ */
+ if (xfs_has_parent(mp) && ino) {
+ struct remap_ent *remap;
+
+ remap = remaptable_find(metadump.cur_ino, hash, name, namelen);
+ if (remap) {
+ remap_debug("found obfuscated dir 0x%lx '%.*s' -> 0x%lx -> '%.*s' \n",
+ cur_ino, namelen,
+ remap_ent_before(remap), ino, namelen,
+ remap_ent_after(remap));
+ memcpy(name, remap_ent_after(remap), namelen);
+ return;
+ }
+
+ /*
+ * If we haven't procesed this dirent name before, save the
+ * old name for a remap table entry. Obfuscate the name.
+ */
+ orig_name = malloc(namelen);
+ if (!orig_name) {
+ orig_name = name;
+ goto add_remap;
+ }
+ memcpy(orig_name, name, namelen);
+ }
+
obfuscate_name(hash, namelen, name, ino != 0);
ASSERT(hash == dirattr_hashname(ino != 0, name, namelen));
@@ -891,6 +1034,26 @@ generate_obfuscated_name(
"in dir inode %llu\n",
(unsigned long long) ino,
(unsigned long long) metadump.cur_ino);
+
+ /*
+ * We've obfuscated a name in the directory entry. Remember this
+ * remapping for when we come across the parent pointer later.
+ */
+ if (!orig_name)
+ return;
+
+add_remap:
+ remap_debug("obfuscating dir 0x%lx '%.*s' -> 0x%lx -> '%.*s' \n",
+ metadump.cur_ino, namelen, orig_name, ino, namelen,
+ name);
+
+ if (!remaptable_add(metadump.cur_ino, hash, orig_name, namelen, name))
+ print_warning("unable to record remapped dirent name for inode %llu "
+ "in dir inode %llu\n",
+ (unsigned long long) ino,
+ (unsigned long long) metadump.cur_ino);
+ if (orig_name && orig_name != name)
+ free(orig_name);
}
static void
@@ -1026,6 +1189,125 @@ process_sf_symlink(
memset(&buf[len], 0, XFS_DFORK_DSIZE(dip, mp) - len);
}
+static inline bool
+want_obfuscate_pptr(
+ unsigned int nsp_flags,
+ const void *name,
+ unsigned int namelen,
+ const void *value,
+ unsigned int valuelen)
+{
+ if (!metadump.obfuscate)
+ return false;
+
+ /* Ignore if parent pointers aren't enabled. */
+ if (!xfs_has_parent(mp))
+ return false;
+
+ /* Ignore anything not claiming to be a parent pointer. */
+ if (!(nsp_flags & XFS_ATTR_PARENT))
+ return false;
+
+ /* Obfuscate this parent pointer if it passes basic checks. */
+ if (libxfs_parent_namecheck(mp, name, namelen, nsp_flags) &&
+ libxfs_parent_valuecheck(mp, value, valuelen) &&
+ libxfs_parent_hashcheck(mp, name, value, valuelen))
+ return true;
+
+ /* Ignore otherwise. */
+ return false;
+}
+
+static void
+obfuscate_parent_pointer(
+ const struct xfs_parent_name_rec *rec,
+ unsigned char *value,
+ unsigned int valuelen)
+{
+ struct xfs_parent_name_irec irec;
+ struct remap_ent *remap;
+ char *old_name = irec.p_name;
+ xfs_dahash_t hash;
+ xfs_ino_t child_ino = metadump.cur_ino;
+
+ libxfs_parent_irec_from_disk(&irec, rec, value, valuelen);
+
+ /*
+ * We don't obfuscate "lost+found" or any orphan files
+ * therein. If When the name table is used for extended
+ * attributes, the inode number provided is 0, in which
+ * case we don't need to make this check.
+ */
+ metadump.cur_ino = irec.p_ino;
+ if (in_lost_found(child_ino, valuelen, value)) {
+ metadump.cur_ino = child_ino;
+ return;
+ }
+ metadump.cur_ino = child_ino;
+
+ /*
+ * If the name starts with a slash, just skip over it. It isn't
+ * included in the hash and we don't record it in the name table.
+ */
+ if (*value == '/') {
+ old_name++;
+ value++;
+ valuelen--;
+ }
+
+ hash = libxfs_da_hashname(value, valuelen);
+
+ /*
+ * If we already processed the dirent, use the same name for the parent
+ * pointer.
+ */
+ remap = remaptable_find(irec.p_ino, hash, value, valuelen);
+ if (remap) {
+ remap_debug("found obfuscated pptr 0x%lx '%.*s' -> 0x%lx -> '%.*s' \n",
+ irec.p_ino, valuelen, remap_ent_before(remap),
+ metadump.cur_ino, valuelen,
+ remap_ent_after(remap));
+ memcpy(value, remap_ent_after(remap), valuelen);
+ return;
+ }
+
+ /*
+ * Obfuscate the parent pointer name and remember this for later
+ * in case we encounter the dirent and need to reuse the name there.
+ */
+ obfuscate_name(hash, valuelen, value, true);
+
+ remap_debug("obfuscated pptr 0x%lx '%.*s' -> 0x%lx -> '%.*s'\n",
+ irec.p_ino, valuelen, old_name, metadump.cur_ino,
+ valuelen, value);
+ if (!remaptable_add(irec.p_ino, hash, old_name, valuelen, value))
+ print_warning("unable to record remapped pptr name for inode %llu "
+ "in dir inode %llu\n",
+ (unsigned long long) metadump.cur_ino,
+ (unsigned long long) irec.p_ino);
+}
+
+static inline bool
+want_obfuscate_attr(
+ unsigned int nsp_flags,
+ const void *name,
+ unsigned int namelen,
+ const void *value,
+ unsigned int valuelen)
+{
+ if (!metadump.obfuscate)
+ return false;
+
+ /*
+ * If we didn't already obfuscate the parent pointer, it's probably
+ * corrupt. Leave it intact for analysis.
+ */
+ if (nsp_flags & XFS_ATTR_PARENT)
+ return false;
+
+ return true;
+}
+
static void
process_sf_attr(
struct xfs_dinode *dip)
@@ -1055,7 +1337,7 @@ process_sf_attr(
asfep = &asfp->list[0];
for (i = 0; (i < asfp->hdr.count) &&
((char *)asfep - (char *)asfp < ino_attr_size); i++) {
-
+ void *name, *value;
int namelen = asfep->namelen;
if (namelen == 0) {
@@ -1072,11 +1354,16 @@ process_sf_attr(
break;
}
- if (metadump.obfuscate) {
- generate_obfuscated_name(0, asfep->namelen,
- &asfep->nameval[0]);
- memset(&asfep->nameval[asfep->namelen], 'v',
- asfep->valuelen);
+ name = &asfep->nameval[0];
+ value = &asfep->nameval[asfep->namelen];
+
+ if (want_obfuscate_pptr(asfep->flags, name, namelen, value,
+ asfep->valuelen)) {
+ obfuscate_parent_pointer(name, value, asfep->valuelen);
+ } else if (want_obfuscate_attr(asfep->flags, name, namelen,
+ value, asfep->valuelen)) {
+ generate_obfuscated_name(0, asfep->namelen, name);
+ memset(value, 'v', asfep->valuelen);
}
asfep = (struct xfs_attr_sf_entry *)((char *)asfep +
@@ -1445,6 +1732,9 @@ process_attr_block(
break;
}
if (entry->flags & XFS_ATTR_LOCAL) {
+ void *name, *value;
+ unsigned int valuelen;
+
local = xfs_attr3_leaf_name_local(leaf, i);
if (local->namelen == 0) {
if (metadump.show_warnings)
@@ -1453,11 +1743,21 @@ process_attr_block(
(long long)metadump.cur_ino);
break;
}
- if (metadump.obfuscate) {
+
+ name = &local->nameval[0];
+ value = &local->nameval[local->namelen];
+ valuelen = be16_to_cpu(local->valuelen);
+
+ if (want_obfuscate_pptr(entry->flags, name,
+ local->namelen, value,
+ valuelen)) {
+ obfuscate_parent_pointer(name, value, valuelen);
+ } else if (want_obfuscate_attr(entry->flags, name,
+ local->namelen, value,
+ valuelen)) {
generate_obfuscated_name(0, local->namelen,
- &local->nameval[0]);
- memset(&local->nameval[local->namelen], 'v',
- be16_to_cpu(local->valuelen));
+ name);
+ memset(value, 'v', valuelen);
}
/* zero from end of nameval[] to next name start */
nlen = local->namelen;
@@ -3046,5 +3346,6 @@ metadump_f(
metadump.mdops->release();
out:
+ remaptable_clear();
return 0;
}
diff --git a/libxfs/libxfs_api_defs.h b/libxfs/libxfs_api_defs.h
index c94972fb84b..a6b561b5b40 100644
--- a/libxfs/libxfs_api_defs.h
+++ b/libxfs/libxfs_api_defs.h
@@ -181,6 +181,9 @@
#define xfs_parent_finish libxfs_parent_finish
#define xfs_parent_irec_from_disk libxfs_parent_irec_from_disk
#define xfs_parent_start libxfs_parent_start
+#define xfs_parent_hashcheck libxfs_parent_hashcheck
+#define xfs_parent_namecheck libxfs_parent_namecheck
+#define xfs_parent_valuecheck libxfs_parent_valuecheck
#define xfs_perag_get libxfs_perag_get
#define xfs_perag_hold libxfs_perag_hold
#define xfs_perag_put libxfs_perag_put
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 27/32] libxfs: export attr3_leaf_hdr_from_disk via libxfs_api_defs.h
2023-12-31 19:50 ` [PATCHSET v13.0 3/6] xfsprogs: Parent Pointers Darrick J. Wong
` (25 preceding siblings ...)
2023-12-31 23:11 ` [PATCH 26/32] xfs_db: obfuscate dirent and parent pointer names consistently Darrick J. Wong
@ 2023-12-31 23:12 ` Darrick J. Wong
2023-12-31 23:12 ` [PATCH 28/32] xfs_db: add a parents command to list the parents of a file Darrick J. Wong
` (4 subsequent siblings)
31 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:12 UTC (permalink / raw)
To: djwong, cem; +Cc: catherine.hoang, linux-xfs, allison.henderson
From: Darrick J. Wong <djwong@kernel.org>
Do the xfs -> libxfs switcheroo and cleanups separately so the next
patch doesn't become an even larger mess.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
db/attr.c | 2 +-
db/metadump.c | 2 +-
libxfs/libxfs_api_defs.h | 5 +++++
repair/attr_repair.c | 6 +++---
4 files changed, 10 insertions(+), 5 deletions(-)
diff --git a/db/attr.c b/db/attr.c
index 9e7bbd164df..95969d115d4 100644
--- a/db/attr.c
+++ b/db/attr.c
@@ -256,7 +256,7 @@ attr_leaf_entry_walk(
return 0;
off = byteize(startoff);
- xfs_attr3_leaf_hdr_from_disk(mp->m_attr_geo, &leafhdr, leaf);
+ libxfs_attr3_leaf_hdr_from_disk(mp->m_attr_geo, &leafhdr, leaf);
entries = xfs_attr3_leaf_entryp(leaf);
for (i = 0; i < leafhdr.count; i++) {
diff --git a/db/metadump.c b/db/metadump.c
index 5f5a33335b0..f5b930d51d2 100644
--- a/db/metadump.c
+++ b/db/metadump.c
@@ -1701,7 +1701,7 @@ process_attr_block(
}
/* Ok, it's a leaf - get header; accounts for crc & non-crc */
- xfs_attr3_leaf_hdr_from_disk(mp->m_attr_geo, &hdr, leaf);
+ libxfs_attr3_leaf_hdr_from_disk(mp->m_attr_geo, &hdr, leaf);
nentries = hdr.count;
if (nentries == 0 ||
diff --git a/libxfs/libxfs_api_defs.h b/libxfs/libxfs_api_defs.h
index a6b561b5b40..22e4c569170 100644
--- a/libxfs/libxfs_api_defs.h
+++ b/libxfs/libxfs_api_defs.h
@@ -36,7 +36,10 @@
#define xfs_ascii_ci_hashname libxfs_ascii_ci_hashname
+#define xfs_attr3_leaf_hdr_from_disk libxfs_attr3_leaf_hdr_from_disk
+#define xfs_attr3_leaf_read libxfs_attr3_leaf_read
#define xfs_attr_get libxfs_attr_get
+#define xfs_attr_is_leaf libxfs_attr_is_leaf
#define xfs_attr_leaf_newentsize libxfs_attr_leaf_newentsize
#define xfs_attr_namecheck libxfs_attr_namecheck
#define xfs_attr_set libxfs_attr_set
@@ -91,6 +94,7 @@
#define xfs_compute_rextslog libxfs_compute_rextslog
#define xfs_create_space_res libxfs_create_space_res
#define xfs_da3_node_hdr_from_disk libxfs_da3_node_hdr_from_disk
+#define xfs_da3_node_read libxfs_da3_node_read
#define xfs_da_get_buf libxfs_da_get_buf
#define xfs_da_hashname libxfs_da_hashname
#define xfs_da_read_buf libxfs_da_read_buf
@@ -164,6 +168,7 @@
#define xfs_inobt_stage_cursor libxfs_inobt_stage_cursor
#define xfs_inode_from_disk libxfs_inode_from_disk
#define xfs_inode_from_disk_ts libxfs_inode_from_disk_ts
+#define xfs_inode_hasattr libxfs_inode_hasattr
#define xfs_inode_to_disk libxfs_inode_to_disk
#define xfs_inode_validate_cowextsize libxfs_inode_validate_cowextsize
#define xfs_inode_validate_extsize libxfs_inode_validate_extsize
diff --git a/repair/attr_repair.c b/repair/attr_repair.c
index 314a9cb4d6e..b0f6ee11ae4 100644
--- a/repair/attr_repair.c
+++ b/repair/attr_repair.c
@@ -596,7 +596,7 @@ process_leaf_attr_block(
da_freemap_t *attr_freemap;
struct xfs_attr3_icleaf_hdr leafhdr;
- xfs_attr3_leaf_hdr_from_disk(mp->m_attr_geo, &leafhdr, leaf);
+ libxfs_attr3_leaf_hdr_from_disk(mp->m_attr_geo, &leafhdr, leaf);
clearit = usedbs = 0;
firstb = mp->m_sb.sb_blocksize;
stop = xfs_attr3_leaf_hdr_size(leaf);
@@ -819,7 +819,7 @@ process_leaf_attr_level(xfs_mount_t *mp,
}
leaf = bp->b_addr;
- xfs_attr3_leaf_hdr_from_disk(mp->m_attr_geo, &leafhdr, leaf);
+ libxfs_attr3_leaf_hdr_from_disk(mp->m_attr_geo, &leafhdr, leaf);
/* check magic number for leaf directory btree block */
if (!(leafhdr.magic == XFS_ATTR_LEAF_MAGIC ||
@@ -1017,7 +1017,7 @@ process_longform_leaf_root(
* check sibling pointers in leaf block or root block 0 before
* we have to release the btree block
*/
- xfs_attr3_leaf_hdr_from_disk(mp->m_attr_geo, &leafhdr, bp->b_addr);
+ libxfs_attr3_leaf_hdr_from_disk(mp->m_attr_geo, &leafhdr, bp->b_addr);
if (leafhdr.forw != 0 || leafhdr.back != 0) {
if (!no_modify) {
do_warn(
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 28/32] xfs_db: add a parents command to list the parents of a file
2023-12-31 19:50 ` [PATCHSET v13.0 3/6] xfsprogs: Parent Pointers Darrick J. Wong
` (26 preceding siblings ...)
2023-12-31 23:12 ` [PATCH 27/32] libxfs: export attr3_leaf_hdr_from_disk via libxfs_api_defs.h Darrick J. Wong
@ 2023-12-31 23:12 ` Darrick J. Wong
2023-12-31 23:12 ` [PATCH 29/32] libxfs: create new files with attr forks if necessary Darrick J. Wong
` (3 subsequent siblings)
31 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:12 UTC (permalink / raw)
To: djwong, cem; +Cc: catherine.hoang, linux-xfs, allison.henderson
From: Darrick J. Wong <djwong@kernel.org>
Create a command to dump the parents of a file.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
db/namei.c | 335 ++++++++++++++++++++++++++++++++++++++++++++++
libxfs/libxfs_api_defs.h | 1
man/man8/xfs_db.8 | 9 +
3 files changed, 345 insertions(+)
diff --git a/db/namei.c b/db/namei.c
index a8577b97222..fb7f63fda07 100644
--- a/db/namei.c
+++ b/db/namei.c
@@ -598,6 +598,338 @@ static struct cmdinfo ls_cmd = {
.help = ls_help,
};
+static void
+pptr_emit(
+ struct xfs_mount *mp,
+ const struct xfs_parent_name_irec *irec)
+{
+ struct xfs_name xname = {
+ .name = irec->p_name,
+ .len = irec->p_namelen,
+ };
+ xfs_dahash_t hash;
+ bool good;
+
+ hash = libxfs_dir2_hashname(mp, &xname);
+ good = libxfs_parent_verify_irec(mp, irec);
+
+ dbprintf("%18llu:0x%08x 0x%08x:0x%08x %3d %.*s %s\n",
+ irec->p_ino, irec->p_gen, irec->p_namehash, hash,
+ xname.len, xname.len, xname.name,
+ good ? _("(good)") : _("(corrupt)"));
+}
+
+static int
+list_sf_pptrs(
+ struct xfs_inode *ip)
+{
+ struct xfs_parent_name_irec irec;
+ struct xfs_attr_shortform *sf;
+ struct xfs_attr_sf_entry *sfe;
+ unsigned int i;
+
+ sf = (struct xfs_attr_shortform *)ip->i_af.if_u1.if_data;
+ for (i = 0, sfe = &sf->list[0]; i < sf->hdr.count; i++) {
+ void *name = sfe->nameval;
+ void *value = &sfe->nameval[sfe->namelen];
+
+ if ((sfe->flags & XFS_ATTR_PARENT) &&
+ libxfs_parent_namecheck(mp, name, sfe->namelen, sfe->flags) &&
+ libxfs_parent_valuecheck(mp, value, sfe->valuelen)) {
+ libxfs_parent_irec_from_disk(&irec, name, value,
+ sfe->valuelen);
+ pptr_emit(mp, &irec);
+ }
+
+ sfe = xfs_attr_sf_nextentry(sfe);
+ }
+
+ return 0;
+}
+
+static void
+list_leaf_pptr_entries(
+ struct xfs_inode *ip,
+ struct xfs_buf *bp)
+{
+ struct xfs_parent_name_irec irec;
+ struct xfs_attr3_icleaf_hdr ichdr;
+ struct xfs_mount *mp = ip->i_mount;
+ struct xfs_attr_leafblock *leaf = bp->b_addr;
+ struct xfs_attr_leaf_entry *entry;
+ unsigned int i;
+
+ libxfs_attr3_leaf_hdr_from_disk(mp->m_attr_geo, &ichdr, leaf);
+ entry = xfs_attr3_leaf_entryp(leaf);
+
+ for (i = 0; i < ichdr.count; entry++, i++) {
+ struct xfs_attr_leaf_name_local *name_loc;
+ void *value;
+ void *name;
+ unsigned int namelen, valuelen;
+
+ if (!(entry->flags & XFS_ATTR_LOCAL) ||
+ !(entry->flags & XFS_ATTR_PARENT))
+ continue;
+
+ name_loc = xfs_attr3_leaf_name_local(leaf, i);
+ name = name_loc->nameval;
+ namelen = name_loc->namelen;
+ value = &name_loc->nameval[name_loc->namelen];
+ valuelen = be16_to_cpu(name_loc->valuelen);
+
+ if (libxfs_parent_namecheck(mp, name, namelen, entry->flags) &&
+ libxfs_parent_valuecheck(mp, value, valuelen)) {
+ libxfs_parent_irec_from_disk(&irec, name, value,
+ valuelen);
+ pptr_emit(mp, &irec);
+ }
+ }
+}
+
+static int
+list_leaf_pptrs(
+ struct xfs_inode *ip)
+{
+ struct xfs_buf *leaf_bp;
+ int error;
+
+ error = -libxfs_attr3_leaf_read(NULL, ip, ip->i_ino, 0, &leaf_bp);
+ if (error)
+ return error;
+
+ list_leaf_pptr_entries(ip, leaf_bp);
+ libxfs_trans_brelse(NULL, leaf_bp);
+ return 0;
+}
+
+static int
+find_leftmost_attr_leaf(
+ struct xfs_inode *ip,
+ struct xfs_buf **leaf_bpp)
+{
+ struct xfs_da3_icnode_hdr nodehdr;
+ struct xfs_mount *mp = ip->i_mount;
+ struct xfs_da_intnode *node;
+ struct xfs_da_node_entry *btree;
+ struct xfs_buf *bp;
+ xfs_dablk_t blkno = 0;
+ unsigned int expected_level = 0;
+ int error;
+
+ for (;;) {
+ uint16_t magic;
+
+ error = -libxfs_da3_node_read(NULL, ip, blkno, &bp,
+ XFS_ATTR_FORK);
+ if (error)
+ return error;
+
+ node = bp->b_addr;
+ magic = be16_to_cpu(node->hdr.info.magic);
+ if (magic == XFS_ATTR_LEAF_MAGIC ||
+ magic == XFS_ATTR3_LEAF_MAGIC)
+ break;
+
+ error = EFSCORRUPTED;
+ if (magic != XFS_DA_NODE_MAGIC &&
+ magic != XFS_DA3_NODE_MAGIC)
+ goto out_buf;
+
+ libxfs_da3_node_hdr_from_disk(mp, &nodehdr, node);
+
+ if (nodehdr.count == 0 || nodehdr.level >= XFS_DA_NODE_MAXDEPTH)
+ goto out_buf;
+
+ /* Check the level from the root node. */
+ if (blkno == 0)
+ expected_level = nodehdr.level - 1;
+ else if (expected_level != nodehdr.level)
+ goto out_buf;
+ else
+ expected_level--;
+
+ /* Find the next level towards the leaves of the dabtree. */
+ btree = nodehdr.btree;
+ blkno = be32_to_cpu(btree->before);
+ libxfs_trans_brelse(NULL, bp);
+ }
+
+ error = EFSCORRUPTED;
+ if (expected_level != 0)
+ goto out_buf;
+
+ *leaf_bpp = bp;
+ return 0;
+
+out_buf:
+ libxfs_trans_brelse(NULL, bp);
+ return error;
+}
+
+static int
+list_node_pptrs(
+ struct xfs_inode *ip)
+{
+ struct xfs_attr3_icleaf_hdr leafhdr;
+ struct xfs_mount *mp = ip->i_mount;
+ struct xfs_attr_leafblock *leaf;
+ struct xfs_buf *leaf_bp;
+ int error;
+
+ error = find_leftmost_attr_leaf(ip, &leaf_bp);
+ if (error)
+ return error;
+
+ for (;;) {
+ list_leaf_pptr_entries(ip, leaf_bp);
+
+ /* Find the right sibling of this leaf block. */
+ leaf = leaf_bp->b_addr;
+ libxfs_attr3_leaf_hdr_from_disk(mp->m_attr_geo, &leafhdr, leaf);
+ if (leafhdr.forw == 0)
+ goto out_leaf;
+
+ libxfs_trans_brelse(NULL, leaf_bp);
+
+ error = -libxfs_attr3_leaf_read(NULL, ip, ip->i_ino,
+ leafhdr.forw, &leaf_bp);
+ if (error)
+ return error;
+ }
+
+out_leaf:
+ libxfs_trans_brelse(NULL, leaf_bp);
+ return error;
+}
+
+static int
+list_pptrs(
+ struct xfs_inode *ip)
+{
+ int error;
+
+ if (!libxfs_inode_hasattr(ip))
+ return 0;
+
+ if (ip->i_af.if_format == XFS_DINODE_FMT_LOCAL)
+ return list_sf_pptrs(ip);
+
+ /* attr functions require that the attr fork is loaded */
+ error = -libxfs_iread_extents(NULL, ip, XFS_ATTR_FORK);
+ if (error)
+ return error;
+
+ if (libxfs_attr_is_leaf(ip))
+ return list_leaf_pptrs(ip);
+
+ return list_node_pptrs(ip);
+}
+
+/* If the io cursor points to a file, list its parents. */
+static int
+parent_cur(
+ char *tag)
+{
+ struct xfs_inode *ip;
+ int error = 0;
+
+ if (!xfs_has_parent(mp))
+ return 0;
+
+ if (iocur_top->typ != &typtab[TYP_INODE])
+ return ENOTDIR;
+
+ error = -libxfs_iget(mp, NULL, iocur_top->ino, 0, &ip);
+ if (error)
+ return error;
+
+ /* List the parents of a file. */
+ if (tag)
+ dbprintf(_("%s:\n"), tag);
+
+ error = list_pptrs(ip);
+ if (error)
+ goto rele;
+
+rele:
+ libxfs_irele(ip);
+ return error;
+}
+
+static void
+parent_help(void)
+{
+ dbprintf(_(
+"\n"
+" List the parents of the currently selected file.\n"
+"\n"
+" Parent pointers will be listed in the format:\n"
+" inode_number:inode_gen ondisk_namehash:namehash name_length name\n"
+ ));
+}
+
+static int
+parent_f(
+ int argc,
+ char **argv)
+{
+ int c;
+ int error = 0;
+
+ while ((c = getopt(argc, argv, "")) != -1) {
+ switch (c) {
+ default:
+ ls_help();
+ return 0;
+ }
+ }
+
+ if (optind == argc) {
+ error = parent_cur(NULL);
+ if (error) {
+ dbprintf("%s\n", strerror(error));
+ exitcode = 1;
+ }
+
+ return 0;
+ }
+
+ for (c = optind; c < argc; c++) {
+ push_cur();
+
+ error = path_walk(argv[c]);
+ if (error)
+ goto err_cur;
+
+ error = parent_cur(argv[c]);
+ if (error)
+ goto err_cur;
+
+ pop_cur();
+ }
+
+ return 0;
+err_cur:
+ pop_cur();
+ if (error) {
+ dbprintf("%s: %s\n", argv[c], strerror(error));
+ exitcode = 1;
+ }
+ return 0;
+}
+
+static struct cmdinfo parent_cmd = {
+ .name = "parent",
+ .altname = "pptr",
+ .cfunc = parent_f,
+ .argmin = 0,
+ .argmax = -1,
+ .canpush = 0,
+ .args = "[paths...]",
+ .help = parent_help,
+};
+
void
namei_init(void)
{
@@ -606,4 +938,7 @@ namei_init(void)
ls_cmd.oneline = _("list directory contents");
add_command(&ls_cmd);
+
+ parent_cmd.oneline = _("list parent pointers");
+ add_command(&parent_cmd);
}
diff --git a/libxfs/libxfs_api_defs.h b/libxfs/libxfs_api_defs.h
index 22e4c569170..7ea7eebfbca 100644
--- a/libxfs/libxfs_api_defs.h
+++ b/libxfs/libxfs_api_defs.h
@@ -189,6 +189,7 @@
#define xfs_parent_hashcheck libxfs_parent_hashcheck
#define xfs_parent_namecheck libxfs_parent_namecheck
#define xfs_parent_valuecheck libxfs_parent_valuecheck
+#define xfs_parent_verify_irec libxfs_parent_verify_irec
#define xfs_perag_get libxfs_perag_get
#define xfs_perag_hold libxfs_perag_hold
#define xfs_perag_put libxfs_perag_put
diff --git a/man/man8/xfs_db.8 b/man/man8/xfs_db.8
index a7f6d55ed8b..937b17e79a3 100644
--- a/man/man8/xfs_db.8
+++ b/man/man8/xfs_db.8
@@ -943,6 +943,15 @@ See the
.B print
command.
.TP
+.BI "parent [" paths "]..."
+List the parents of a file.
+If a path resolves to a file, the parents of that file will be listed.
+If no paths are supplied and the IO cursor points at an inode, the parents of
+that file will be listed.
+
+The output format is:
+inode number, inode generation, ondisk namehash, namehash, name length, name.
+.TP
.BI "path " dir_path
Walk the directory tree to an inode using the supplied path.
Absolute and relative paths are supported.
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 29/32] libxfs: create new files with attr forks if necessary
2023-12-31 19:50 ` [PATCHSET v13.0 3/6] xfsprogs: Parent Pointers Darrick J. Wong
` (27 preceding siblings ...)
2023-12-31 23:12 ` [PATCH 28/32] xfs_db: add a parents command to list the parents of a file Darrick J. Wong
@ 2023-12-31 23:12 ` Darrick J. Wong
2023-12-31 23:12 ` [PATCH 30/32] xfsprogs: Fix default superblock attr bits Darrick J. Wong
` (2 subsequent siblings)
31 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:12 UTC (permalink / raw)
To: djwong, cem; +Cc: catherine.hoang, linux-xfs, allison.henderson
From: Darrick J. Wong <djwong@kernel.org>
Create new files with attr forks if they're going to have parent
pointers. In the next patch we'll fix mkfs to use the same parent
creation functions as the kernel, so we're going to need this.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
libxfs/init.c | 4 ++++
libxfs/util.c | 14 ++++++++++++++
2 files changed, 18 insertions(+)
diff --git a/libxfs/init.c b/libxfs/init.c
index b6b1282201c..397ce088d3a 100644
--- a/libxfs/init.c
+++ b/libxfs/init.c
@@ -655,14 +655,18 @@ void
libxfs_compute_all_maxlevels(
struct xfs_mount *mp)
{
+ struct xfs_ino_geometry *igeo = M_IGEO(mp);
+
xfs_alloc_compute_maxlevels(mp);
xfs_bmap_compute_maxlevels(mp, XFS_DATA_FORK);
xfs_bmap_compute_maxlevels(mp, XFS_ATTR_FORK);
+ igeo->attr_fork_offset = xfs_bmap_compute_attr_offset(mp);
xfs_ialloc_setup_geometry(mp);
xfs_rmapbt_compute_maxlevels(mp);
xfs_refcountbt_compute_maxlevels(mp);
xfs_agbtree_compute_maxlevels(mp);
+
}
/*
diff --git a/libxfs/util.c b/libxfs/util.c
index 03191ebcd08..5106a6433da 100644
--- a/libxfs/util.c
+++ b/libxfs/util.c
@@ -340,6 +340,20 @@ libxfs_init_new_inode(
ASSERT(0);
}
+ /*
+ * If we need to create attributes immediately after allocating the
+ * inode, initialise an empty attribute fork right now. We use the
+ * default fork offset for attributes here as we don't know exactly what
+ * size or how many attributes we might be adding. We can do this
+ * safely here because we know the data fork is completely empty and
+ * this saves us from needing to run a separate transaction to set the
+ * fork offset in the immediate future.
+ */
+ if (xfs_has_parent(tp->t_mountp) && xfs_has_attr(tp->t_mountp)) {
+ ip->i_forkoff = xfs_default_attroffset(ip) >> 3;
+ xfs_ifork_init_attr(ip, XFS_DINODE_FMT_EXTENTS, 0);
+ }
+
/*
* Log the new values stuffed into the inode.
*/
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 30/32] xfsprogs: Fix default superblock attr bits
2023-12-31 19:50 ` [PATCHSET v13.0 3/6] xfsprogs: Parent Pointers Darrick J. Wong
` (28 preceding siblings ...)
2023-12-31 23:12 ` [PATCH 29/32] libxfs: create new files with attr forks if necessary Darrick J. Wong
@ 2023-12-31 23:12 ` Darrick J. Wong
2023-12-31 23:13 ` [PATCH 31/32] mkfs: Add parent pointers during protofile creation Darrick J. Wong
2023-12-31 23:13 ` [PATCH 32/32] mkfs: enable formatting with parent pointers Darrick J. Wong
31 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:12 UTC (permalink / raw)
To: djwong, cem
Cc: Allison Henderson, catherine.hoang, linux-xfs, allison.henderson
From: Allison Henderson <allison.henderson@oracle.com>
Recent parent pointer testing discovered that the default attr
configuration has XFS_SB_VERSION2_ATTR2BIT enabled but
XFS_SB_VERSION_ATTRBIT disabled. This is incorrect since
XFS_SB_VERSION2_ATTR2BIT describes the format of the attr where
as XFS_SB_VERSION_ATTRBIT enables or disables attrs. Fix this
by enableing XFS_SB_VERSION_ATTRBIT for either attr version 1 or 2
Signed-off-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
mkfs/xfs_mkfs.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/mkfs/xfs_mkfs.c b/mkfs/xfs_mkfs.c
index 8b0fbe97ddc..cbfb89b6795 100644
--- a/mkfs/xfs_mkfs.c
+++ b/mkfs/xfs_mkfs.c
@@ -3424,7 +3424,7 @@ sb_set_features(
sbp->sb_versionnum |= XFS_SB_VERSION_DALIGNBIT;
if (fp->log_version == 2)
sbp->sb_versionnum |= XFS_SB_VERSION_LOGV2BIT;
- if (fp->attr_version == 1)
+ if (fp->attr_version >= 1)
sbp->sb_versionnum |= XFS_SB_VERSION_ATTRBIT;
if (fp->nci)
sbp->sb_versionnum |= XFS_SB_VERSION_BORGBIT;
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 31/32] mkfs: Add parent pointers during protofile creation
2023-12-31 19:50 ` [PATCHSET v13.0 3/6] xfsprogs: Parent Pointers Darrick J. Wong
` (29 preceding siblings ...)
2023-12-31 23:12 ` [PATCH 30/32] xfsprogs: Fix default superblock attr bits Darrick J. Wong
@ 2023-12-31 23:13 ` Darrick J. Wong
2023-12-31 23:13 ` [PATCH 32/32] mkfs: enable formatting with parent pointers Darrick J. Wong
31 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:13 UTC (permalink / raw)
To: djwong, cem
Cc: Allison Henderson, catherine.hoang, linux-xfs, allison.henderson
From: Allison Henderson <allison.henderson@oracle.com>
Inodes created from protofile parsing will also need to add the
appropriate parent pointers.
Signed-off-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
[djwong: use xfs_parent_add from libxfs instead of open-coding xfs_attr_set]
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
mkfs/proto.c | 60 ++++++++++++++++++++++++++++++++++++++++++++--------------
1 file changed, 46 insertions(+), 14 deletions(-)
diff --git a/mkfs/proto.c b/mkfs/proto.c
index 457899ac178..cc06bdfaf57 100644
--- a/mkfs/proto.c
+++ b/mkfs/proto.c
@@ -347,11 +347,12 @@ newregfile(
static void
newdirent(
- xfs_mount_t *mp,
- xfs_trans_t *tp,
- xfs_inode_t *pip,
- struct xfs_name *name,
- xfs_ino_t inum)
+ struct xfs_mount *mp,
+ struct xfs_trans *tp,
+ struct xfs_inode *pip,
+ struct xfs_name *name,
+ struct xfs_inode *ip,
+ struct xfs_parent_args *ppargs)
{
int error;
int rsv;
@@ -364,9 +365,13 @@ newdirent(
rsv = XFS_DIRENTER_SPACE_RES(mp, name->len);
- error = -libxfs_dir_createname(tp, pip, name, inum, rsv);
+ error = -libxfs_dir_createname(tp, pip, name, ip->i_ino, rsv);
if (error)
fail(_("directory createname error"), error);
+
+ error = -libxfs_parent_add(tp, ppargs, pip, name, ip);
+ if (error)
+ fail(_("committing parent pointers failed."), error);
}
static void
@@ -383,6 +388,20 @@ newdirectory(
fail(_("directory create error"), error);
}
+static struct xfs_parent_args *
+newpptr(
+ struct xfs_mount *mp)
+{
+ struct xfs_parent_args *ret;
+ int error;
+
+ error = -libxfs_parent_start(mp, &ret);
+ if (error)
+ fail(_("initializing parent pointer"), error);
+
+ return ret;
+}
+
static void
parseproto(
xfs_mount_t *mp,
@@ -417,6 +436,7 @@ parseproto(
struct cred creds;
char *value;
struct xfs_name xname;
+ struct xfs_parent_args *ppargs = NULL;
memset(&creds, 0, sizeof(creds));
mstr = getstr(pp);
@@ -491,6 +511,7 @@ parseproto(
case IF_REGULAR:
buf = newregfile(pp, &len);
tp = getres(mp, XFS_B_TO_FSB(mp, len));
+ ppargs = newpptr(mp);
error = -libxfs_dir_ialloc(&tp, pip, mode|S_IFREG, 1, 0,
&creds, fsxp, &ip);
if (error)
@@ -500,7 +521,7 @@ parseproto(
free(buf);
libxfs_trans_ijoin(tp, pip, 0);
xname.type = XFS_DIR3_FT_REG_FILE;
- newdirent(mp, tp, pip, &xname, ip->i_ino);
+ newdirent(mp, tp, pip, &xname, ip, ppargs);
break;
case IF_RESERVED: /* pre-allocated space only */
@@ -514,7 +535,7 @@ parseproto(
exit(1);
}
tp = getres(mp, XFS_B_TO_FSB(mp, llen));
-
+ ppargs = newpptr(mp);
error = -libxfs_dir_ialloc(&tp, pip, mode|S_IFREG, 1, 0,
&creds, fsxp, &ip);
if (error)
@@ -523,17 +544,19 @@ parseproto(
libxfs_trans_ijoin(tp, pip, 0);
xname.type = XFS_DIR3_FT_REG_FILE;
- newdirent(mp, tp, pip, &xname, ip->i_ino);
+ newdirent(mp, tp, pip, &xname, ip, ppargs);
libxfs_trans_log_inode(tp, ip, flags);
error = -libxfs_trans_commit(tp);
if (error)
fail(_("Space preallocation failed."), error);
+ libxfs_parent_finish(mp, ppargs);
rsvfile(mp, ip, llen);
libxfs_irele(ip);
return;
case IF_BLOCK:
tp = getres(mp, 0);
+ ppargs = newpptr(mp);
majdev = getnum(getstr(pp), 0, 0, false);
mindev = getnum(getstr(pp), 0, 0, false);
error = -libxfs_dir_ialloc(&tp, pip, mode|S_IFBLK, 1,
@@ -543,12 +566,13 @@ parseproto(
}
libxfs_trans_ijoin(tp, pip, 0);
xname.type = XFS_DIR3_FT_BLKDEV;
- newdirent(mp, tp, pip, &xname, ip->i_ino);
+ newdirent(mp, tp, pip, &xname, ip, ppargs);
flags |= XFS_ILOG_DEV;
break;
case IF_CHAR:
tp = getres(mp, 0);
+ ppargs = newpptr(mp);
majdev = getnum(getstr(pp), 0, 0, false);
mindev = getnum(getstr(pp), 0, 0, false);
error = -libxfs_dir_ialloc(&tp, pip, mode|S_IFCHR, 1,
@@ -557,24 +581,26 @@ parseproto(
fail(_("Inode allocation failed"), error);
libxfs_trans_ijoin(tp, pip, 0);
xname.type = XFS_DIR3_FT_CHRDEV;
- newdirent(mp, tp, pip, &xname, ip->i_ino);
+ newdirent(mp, tp, pip, &xname, ip, ppargs);
flags |= XFS_ILOG_DEV;
break;
case IF_FIFO:
tp = getres(mp, 0);
+ ppargs = newpptr(mp);
error = -libxfs_dir_ialloc(&tp, pip, mode|S_IFIFO, 1, 0,
&creds, fsxp, &ip);
if (error)
fail(_("Inode allocation failed"), error);
libxfs_trans_ijoin(tp, pip, 0);
xname.type = XFS_DIR3_FT_FIFO;
- newdirent(mp, tp, pip, &xname, ip->i_ino);
+ newdirent(mp, tp, pip, &xname, ip, ppargs);
break;
case IF_SYMLINK:
buf = getstr(pp);
len = (int)strlen(buf);
tp = getres(mp, XFS_B_TO_FSB(mp, len));
+ ppargs = newpptr(mp);
error = -libxfs_dir_ialloc(&tp, pip, mode|S_IFLNK, 1, 0,
&creds, fsxp, &ip);
if (error)
@@ -582,7 +608,7 @@ parseproto(
writesymlink(tp, ip, buf, len);
libxfs_trans_ijoin(tp, pip, 0);
xname.type = XFS_DIR3_FT_SYMLINK;
- newdirent(mp, tp, pip, &xname, ip->i_ino);
+ newdirent(mp, tp, pip, &xname, ip, ppargs);
break;
case IF_DIRECTORY:
tp = getres(mp, 0);
@@ -597,9 +623,10 @@ parseproto(
libxfs_log_sb(tp);
isroot = 1;
} else {
+ ppargs = newpptr(mp);
libxfs_trans_ijoin(tp, pip, 0);
xname.type = XFS_DIR3_FT_DIR;
- newdirent(mp, tp, pip, &xname, ip->i_ino);
+ newdirent(mp, tp, pip, &xname, ip, ppargs);
libxfs_bumplink(tp, pip);
libxfs_trans_log_inode(tp, pip, XFS_ILOG_CORE);
}
@@ -608,6 +635,9 @@ parseproto(
error = -libxfs_trans_commit(tp);
if (error)
fail(_("Directory inode allocation failed."), error);
+
+ libxfs_parent_finish(mp, ppargs);
+
/*
* RT initialization. Do this here to ensure that
* the RT inodes get placed after the root inode.
@@ -635,6 +665,8 @@ parseproto(
fail(_("Error encountered creating file from prototype file"),
error);
}
+
+ libxfs_parent_finish(mp, ppargs);
libxfs_irele(ip);
}
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 32/32] mkfs: enable formatting with parent pointers
2023-12-31 19:50 ` [PATCHSET v13.0 3/6] xfsprogs: Parent Pointers Darrick J. Wong
` (30 preceding siblings ...)
2023-12-31 23:13 ` [PATCH 31/32] mkfs: Add parent pointers during protofile creation Darrick J. Wong
@ 2023-12-31 23:13 ` Darrick J. Wong
31 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:13 UTC (permalink / raw)
To: djwong, cem
Cc: Allison Henderson, catherine.hoang, linux-xfs, allison.henderson
From: Allison Henderson <allison.henderson@oracle.com>
Enable parent pointer support in mkfs via the '-n parent' parameter.
Signed-off-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
mkfs/xfs_mkfs.c | 29 ++++++++++++++++++++++++++---
1 file changed, 26 insertions(+), 3 deletions(-)
diff --git a/mkfs/xfs_mkfs.c b/mkfs/xfs_mkfs.c
index cbfb89b6795..482275e0e0d 100644
--- a/mkfs/xfs_mkfs.c
+++ b/mkfs/xfs_mkfs.c
@@ -113,6 +113,7 @@ enum {
N_SIZE = 0,
N_VERSION,
N_FTYPE,
+ N_PARENT,
N_MAX_OPTS,
};
@@ -648,6 +649,7 @@ static struct opt_params nopts = {
[N_SIZE] = "size",
[N_VERSION] = "version",
[N_FTYPE] = "ftype",
+ [N_PARENT] = "parent",
[N_MAX_OPTS] = NULL,
},
.subopt_params = {
@@ -671,6 +673,14 @@ static struct opt_params nopts = {
.maxval = 1,
.defaultval = 1,
},
+ { .index = N_PARENT,
+ .conflicts = { { NULL, LAST_CONFLICT } },
+ .minval = 0,
+ .maxval = 1,
+ .defaultval = 1,
+ },
+
+
},
};
@@ -1030,7 +1040,7 @@ usage( void )
sunit=value|su=num,sectsize=num,lazy-count=0|1,\n\
concurrency=num]\n\
/* label */ [-L label (maximum 12 characters)]\n\
-/* naming */ [-n size=num,version=2|ci,ftype=0|1]\n\
+/* naming */ [-n size=num,version=2|ci,ftype=0|1,parent=0|1]]\n\
/* no-op info only */ [-N]\n\
/* prototype file */ [-p fname]\n\
/* quiet */ [-q]\n\
@@ -1865,6 +1875,9 @@ naming_opts_parser(
case N_FTYPE:
cli->sb_feat.dirftype = getnum(value, opts, subopt);
break;
+ case N_PARENT:
+ cli->sb_feat.parent_pointers = getnum(value, &nopts, N_PARENT);
+ break;
default:
return -EINVAL;
}
@@ -2382,6 +2395,14 @@ _("inode btree counters not supported without finobt support\n"));
cli->sb_feat.inobtcnt = false;
}
+ if ((cli->sb_feat.parent_pointers) &&
+ cli->sb_feat.dir_version == 4) {
+ fprintf(stderr,
+_("parent pointers not supported on v4 filesystems\n"));
+ usage();
+ cli->sb_feat.parent_pointers = false;
+ }
+
if (cli->xi->rt.name) {
if (cli->sb_feat.reflink && cli_opt_set(&mopts, M_REFLINK)) {
fprintf(stderr,
@@ -3443,8 +3464,6 @@ sb_set_features(
sbp->sb_features2 |= XFS_SB_VERSION2_LAZYSBCOUNTBIT;
if (fp->projid32bit)
sbp->sb_features2 |= XFS_SB_VERSION2_PROJID32BIT;
- if (fp->parent_pointers)
- sbp->sb_features2 |= XFS_SB_VERSION2_PARENTBIT;
if (fp->crcs_enabled)
sbp->sb_features2 |= XFS_SB_VERSION2_CRCBIT;
if (fp->attr_version == 2)
@@ -3485,6 +3504,10 @@ sb_set_features(
sbp->sb_features_ro_compat |= XFS_SB_FEAT_RO_COMPAT_INOBTCNT;
if (fp->bigtime)
sbp->sb_features_incompat |= XFS_SB_FEAT_INCOMPAT_BIGTIME;
+ if (fp->parent_pointers) {
+ sbp->sb_features_incompat |= XFS_SB_FEAT_INCOMPAT_PARENT;
+ sbp->sb_versionnum |= XFS_SB_VERSION_ATTRBIT;
+ }
/*
* Sparse inode chunk support has two main inode alignment requirements.
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 01/18] xfs: create a blob array data structure
2023-12-31 19:51 ` [PATCHSET v13.0 4/6] xfsprogs: fsck for " Darrick J. Wong
@ 2023-12-31 23:13 ` Darrick J. Wong
2023-12-31 23:13 ` [PATCH 02/18] xfs: check dirents have parent pointers Darrick J. Wong
` (16 subsequent siblings)
17 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:13 UTC (permalink / raw)
To: djwong, cem; +Cc: catherine.hoang, linux-xfs, allison.henderson
From: Darrick J. Wong <djwong@kernel.org>
Create a simple 'blob array' data structure for storage of arbitrarily
sized metadata objects that will be used to reconstruct metadata. For
the intended usage (temporarily storing extended attribute names and
values) we only have to support storing objects and retrieving them.
Use the xfile abstraction to store the attribute information in memory
that can be swapped out.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
libxfs/Makefile | 2 +
libxfs/xfblob.c | 147 +++++++++++++++++++++++++++++++++++++++++++++++++++++++
libxfs/xfblob.h | 24 +++++++++
libxfs/xfile.c | 11 ++++
libxfs/xfile.h | 1
5 files changed, 185 insertions(+)
create mode 100644 libxfs/xfblob.c
create mode 100644 libxfs/xfblob.h
diff --git a/libxfs/Makefile b/libxfs/Makefile
index e0bdaefb209..42dce62ecf9 100644
--- a/libxfs/Makefile
+++ b/libxfs/Makefile
@@ -28,6 +28,7 @@ HFILES = \
libxfs_priv.h \
linux-err.h \
topology.h \
+ xfblob.h \
xfbtree.h \
xfile.h \
xfs_ag_resv.h \
@@ -73,6 +74,7 @@ CFILES = cache.c \
topology.c \
trans.c \
util.c \
+ xfblob.c \
xfbtree.c \
xfile.c \
xfs_ag.c \
diff --git a/libxfs/xfblob.c b/libxfs/xfblob.c
new file mode 100644
index 00000000000..d826e5f3cb0
--- /dev/null
+++ b/libxfs/xfblob.c
@@ -0,0 +1,147 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * Copyright (c) 2022-2024 Oracle. All Rights Reserved.
+ * Author: Darrick J. Wong <djwong@kernel.org>
+ */
+#include "libxfs_priv.h"
+#include "libxfs.h"
+#include "libxfs/xfile.h"
+#include "libxfs/xfblob.h"
+
+/*
+ * XFS Blob Storage
+ * ================
+ * Stores and retrieves blobs using an xfile. Objects are appended to the file
+ * and the offset is returned as a magic cookie for retrieval.
+ */
+
+#define XB_KEY_MAGIC 0xABAADDAD
+struct xb_key {
+ uint32_t xb_magic; /* XB_KEY_MAGIC */
+ uint32_t xb_size; /* size of the blob, in bytes */
+ loff_t xb_offset; /* byte offset of this key */
+ /* blob comes after here */
+} __packed;
+
+/* Initialize a blob storage object. */
+int
+xfblob_create(
+ const char *description,
+ struct xfblob **blobp)
+{
+ struct xfblob *blob;
+ struct xfile *xfile;
+ int error;
+
+ error = xfile_create(description, 0, &xfile);
+ if (error)
+ return error;
+
+ blob = malloc(sizeof(struct xfblob));
+ if (!blob) {
+ error = -ENOMEM;
+ goto out_xfile;
+ }
+
+ blob->xfile = xfile;
+ blob->last_offset = PAGE_SIZE;
+
+ *blobp = blob;
+ return 0;
+
+out_xfile:
+ xfile_destroy(xfile);
+ return error;
+}
+
+/* Destroy a blob storage object. */
+void
+xfblob_destroy(
+ struct xfblob *blob)
+{
+ xfile_destroy(blob->xfile);
+ kfree(blob);
+}
+
+/* Retrieve a blob. */
+int
+xfblob_load(
+ struct xfblob *blob,
+ xfblob_cookie cookie,
+ void *ptr,
+ uint32_t size)
+{
+ struct xb_key key;
+ int error;
+
+ error = xfile_obj_load(blob->xfile, &key, sizeof(key), cookie);
+ if (error)
+ return error;
+
+ if (key.xb_magic != XB_KEY_MAGIC || key.xb_offset != cookie) {
+ ASSERT(0);
+ return -ENODATA;
+ }
+ if (size < key.xb_size) {
+ ASSERT(0);
+ return -EFBIG;
+ }
+
+ return xfile_obj_load(blob->xfile, ptr, key.xb_size,
+ cookie + sizeof(key));
+}
+
+/* Store a blob. */
+int
+xfblob_store(
+ struct xfblob *blob,
+ xfblob_cookie *cookie,
+ const void *ptr,
+ uint32_t size)
+{
+ struct xb_key key = {
+ .xb_offset = blob->last_offset,
+ .xb_magic = XB_KEY_MAGIC,
+ .xb_size = size,
+ };
+ loff_t pos = blob->last_offset;
+ int error;
+
+ error = xfile_obj_store(blob->xfile, &key, sizeof(key), pos);
+ if (error)
+ return error;
+
+ pos += sizeof(key);
+ error = xfile_obj_store(blob->xfile, ptr, size, pos);
+ if (error)
+ goto out_err;
+
+ *cookie = blob->last_offset;
+ blob->last_offset += sizeof(key) + size;
+ return 0;
+out_err:
+ xfile_discard(blob->xfile, blob->last_offset, sizeof(key));
+ return error;
+}
+
+/* Free a blob. */
+int
+xfblob_free(
+ struct xfblob *blob,
+ xfblob_cookie cookie)
+{
+ struct xb_key key;
+ int error;
+
+ error = xfile_obj_load(blob->xfile, &key, sizeof(key), cookie);
+ if (error)
+ return error;
+
+ if (key.xb_magic != XB_KEY_MAGIC || key.xb_offset != cookie) {
+ ASSERT(0);
+ return -ENODATA;
+ }
+
+ xfile_discard(blob->xfile, cookie, sizeof(key) + key.xb_size);
+ return 0;
+}
diff --git a/libxfs/xfblob.h b/libxfs/xfblob.h
new file mode 100644
index 00000000000..28bf4ab2898
--- /dev/null
+++ b/libxfs/xfblob.h
@@ -0,0 +1,24 @@
+/* SPDX-License-Identifier: GPL-2.0-or-later */
+/*
+ * Copyright (c) 2022-2024 Oracle. All Rights Reserved.
+ * Author: Darrick J. Wong <djwong@kernel.org>
+ */
+#ifndef __XFS_SCRUB_XFBLOB_H__
+#define __XFS_SCRUB_XFBLOB_H__
+
+struct xfblob {
+ struct xfile *xfile;
+ loff_t last_offset;
+};
+
+typedef loff_t xfblob_cookie;
+
+int xfblob_create(const char *descr, struct xfblob **blobp);
+void xfblob_destroy(struct xfblob *blob);
+int xfblob_load(struct xfblob *blob, xfblob_cookie cookie, void *ptr,
+ uint32_t size);
+int xfblob_store(struct xfblob *blob, xfblob_cookie *cookie, const void *ptr,
+ uint32_t size);
+int xfblob_free(struct xfblob *blob, xfblob_cookie cookie);
+
+#endif /* __XFS_SCRUB_XFBLOB_H__ */
diff --git a/libxfs/xfile.c b/libxfs/xfile.c
index 7f785feb125..87362f96c43 100644
--- a/libxfs/xfile.c
+++ b/libxfs/xfile.c
@@ -479,3 +479,14 @@ xfile_prealloc(
return -errno;
return 0;
}
+
+/* Discard pages backing a range of the xfile. */
+void
+xfile_discard(
+ struct xfile *xf,
+ loff_t pos,
+ unsigned long long count)
+{
+ fallocate(xf->fcb->fd, FALLOC_FL_PUNCH_HOLE | FALLOC_FL_KEEP_SIZE,
+ pos, count);
+}
diff --git a/libxfs/xfile.h b/libxfs/xfile.h
index ac368432382..595a369c89e 100644
--- a/libxfs/xfile.h
+++ b/libxfs/xfile.h
@@ -61,6 +61,7 @@ struct xfile_stat {
int xfile_stat(struct xfile *xf, struct xfile_stat *statbuf);
unsigned long long xfile_bytes(struct xfile *xf);
int xfile_dump(struct xfile *xf);
+void xfile_discard(struct xfile *xf, loff_t pos, unsigned long long count);
static inline loff_t xfile_size(struct xfile *xf)
{
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 02/18] xfs: check dirents have parent pointers
2023-12-31 19:51 ` [PATCHSET v13.0 4/6] xfsprogs: fsck for " Darrick J. Wong
2023-12-31 23:13 ` [PATCH 01/18] xfs: create a blob array data structure Darrick J. Wong
@ 2023-12-31 23:13 ` Darrick J. Wong
2023-12-31 23:14 ` [PATCH 03/18] man2: update ioctl_xfs_scrub_metadata.2 for " Darrick J. Wong
` (15 subsequent siblings)
17 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:13 UTC (permalink / raw)
To: djwong, cem; +Cc: catherine.hoang, linux-xfs, allison.henderson
From: Darrick J. Wong <djwong@kernel.org>
If the fs has parent pointers, we need to check that each child dirent
points to a file that has a parent pointer pointing back at us.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
libxfs/xfs_parent.c | 54 +++++++++++++++++++++++++++++++++++++++++++++++++++
libxfs/xfs_parent.h | 10 +++++++++
2 files changed, 64 insertions(+)
diff --git a/libxfs/xfs_parent.c b/libxfs/xfs_parent.c
index 92b541737cb..024e89756e6 100644
--- a/libxfs/xfs_parent.c
+++ b/libxfs/xfs_parent.c
@@ -367,3 +367,57 @@ xfs_parent_irec_hashname(
irec->p_namehash = xfs_dir2_hashname(mp, &dname);
}
+
+static inline void
+xfs_parent_scratch_init(
+ struct xfs_trans *tp,
+ struct xfs_inode *ip,
+ const struct xfs_parent_name_irec *pptr,
+ struct xfs_parent_scratch *scr)
+{
+ memset(&scr->args, 0, sizeof(struct xfs_da_args));
+ scr->args.attr_filter = XFS_ATTR_PARENT;
+ scr->args.dp = ip;
+ scr->args.geo = ip->i_mount->m_attr_geo;
+ scr->args.name = (const unsigned char *)&scr->rec;
+ scr->args.namelen = sizeof(struct xfs_parent_name_rec);
+ scr->args.op_flags = XFS_DA_OP_NVLOOKUP;
+ scr->args.trans = tp;
+ scr->args.value = (void *)pptr->p_name;
+ scr->args.valuelen = pptr->p_namelen;
+ scr->args.whichfork = XFS_ATTR_FORK;
+ scr->args.hashval = xfs_da_hashname((const void *)&scr->rec,
+ sizeof(struct xfs_parent_name_rec));
+}
+
+/*
+ * Look up the @name associated with the parent pointer (@pptr) of @ip.
+ * Caller must hold at least ILOCK_SHARED. Returns 0 if the pointer is found,
+ * -ENOATTR if there is no match, or a negative errno. The scratchpad need not
+ * be initialized.
+ */
+int
+xfs_parent_lookup(
+ struct xfs_trans *tp,
+ struct xfs_inode *ip,
+ const struct xfs_parent_name_irec *pptr,
+ struct xfs_parent_scratch *scr)
+{
+ int error;
+
+ /*
+ * Make sure the attr fork iext tree is loaded in transaction context
+ * before we start down the rest of the call path.
+ */
+ if (xfs_inode_hasattr(ip)) {
+ error = xfs_iread_extents(tp, ip, XFS_ATTR_FORK);
+ if (error)
+ return error;
+ }
+
+ xfs_parent_irec_to_disk(&scr->rec, pptr);
+ xfs_parent_scratch_init(tp, ip, pptr, scr);
+ scr->args.op_flags |= XFS_DA_OP_OKNOENT;
+
+ return xfs_attr_get_ilocked(&scr->args);
+}
diff --git a/libxfs/xfs_parent.h b/libxfs/xfs_parent.h
index e43ae5a7df8..e4443da1d86 100644
--- a/libxfs/xfs_parent.h
+++ b/libxfs/xfs_parent.h
@@ -152,4 +152,14 @@ void xfs_parent_irec_hashname(struct xfs_mount *mp,
bool xfs_parent_verify_irec(struct xfs_mount *mp,
const struct xfs_parent_name_irec *irec);
+/* Scratchpad memory so that raw parent operations don't burn stack space. */
+struct xfs_parent_scratch {
+ struct xfs_parent_name_rec rec;
+ struct xfs_da_args args;
+};
+
+int xfs_parent_lookup(struct xfs_trans *tp, struct xfs_inode *ip,
+ const struct xfs_parent_name_irec *pptr,
+ struct xfs_parent_scratch *scratch);
+
#endif /* __XFS_PARENT_H__ */
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 03/18] man2: update ioctl_xfs_scrub_metadata.2 for parent pointers
2023-12-31 19:51 ` [PATCHSET v13.0 4/6] xfsprogs: fsck for " Darrick J. Wong
2023-12-31 23:13 ` [PATCH 01/18] xfs: create a blob array data structure Darrick J. Wong
2023-12-31 23:13 ` [PATCH 02/18] xfs: check dirents have parent pointers Darrick J. Wong
@ 2023-12-31 23:14 ` Darrick J. Wong
2023-12-31 23:14 ` [PATCH 04/18] xfs: add raw parent pointer apis to support repair Darrick J. Wong
` (14 subsequent siblings)
17 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:14 UTC (permalink / raw)
To: djwong, cem; +Cc: catherine.hoang, linux-xfs, allison.henderson
From: Darrick J. Wong <djwong@kernel.org>
Update the man page for the scrub ioctl to reflect the new scrubbing
abilities when parent pointers are enabled.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
man/man2/ioctl_xfs_scrub_metadata.2 | 20 ++++++++++++++++----
1 file changed, 16 insertions(+), 4 deletions(-)
diff --git a/man/man2/ioctl_xfs_scrub_metadata.2 b/man/man2/ioctl_xfs_scrub_metadata.2
index 9963f1913e6..75ae52bb584 100644
--- a/man/man2/ioctl_xfs_scrub_metadata.2
+++ b/man/man2/ioctl_xfs_scrub_metadata.2
@@ -109,12 +109,11 @@ must be zero.
.nf
.B XFS_SCRUB_TYPE_BMBTD
.B XFS_SCRUB_TYPE_BMBTA
+.fi
+.TP
.B XFS_SCRUB_TYPE_BMBTC
-.fi
-.TP
-.B XFS_SCRUB_TYPE_PARENT
Examine a given inode's data block map, extended attribute block map,
-copy on write block map, or parent inode pointer.
+or copy on write block map.
Inode records are examined for obviously incorrect values and
discrepancies with the three block map types.
The block maps are checked for obviously wrong values and
@@ -133,9 +132,22 @@ The inode to examine can be specified in the same manner as
.TP
.B XFS_SCRUB_TYPE_DIR
Examine the entries in a given directory for invalid data or dangling pointers.
+If the filesystem supports directory parent pointers, each entry will be
+checked to confirm that the child file has a matching parent pointer.
The directory to examine can be specified in the same manner as
.BR XFS_SCRUB_TYPE_INODE "."
+.TP
+.B XFS_SCRUB_TYPE_PARENT
+For filesystems that support directory parent pointers, this scrubber
+examines all the parent pointers attached to a file and confirms that the
+parent directory has an entry matching the parent pointer.
+For filesystems that do not support directory parent pointers, this scrubber
+checks that a subdirectory's dotdot entry points to a directory with an entry
+that points back to the subdirectory.
+The inode to examine can be specified in the same manner as
+.BR XFS_SCRUB_TYPE_INODE "."
+
.TP
.B XFS_SCRUB_TYPE_SYMLINK
Examine the target of a symbolic link for obvious pathname problems.
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 04/18] xfs: add raw parent pointer apis to support repair
2023-12-31 19:51 ` [PATCHSET v13.0 4/6] xfsprogs: fsck for " Darrick J. Wong
` (2 preceding siblings ...)
2023-12-31 23:14 ` [PATCH 03/18] man2: update ioctl_xfs_scrub_metadata.2 for " Darrick J. Wong
@ 2023-12-31 23:14 ` Darrick J. Wong
2023-12-31 23:14 ` [PATCH 05/18] xfs: set child file owner in xfs_da_args when changing parent pointers Darrick J. Wong
` (13 subsequent siblings)
17 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:14 UTC (permalink / raw)
To: djwong, cem; +Cc: catherine.hoang, linux-xfs, allison.henderson
From: Darrick J. Wong <djwong@kernel.org>
Add a couple of utility functions to set or remove parent pointers from
a file. These functions will be used by repair code, hence they skip
the xattr logging that regular parent pointer updates use.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
libxfs/xfs_dir2.c | 2 +-
libxfs/xfs_dir2.h | 2 +-
libxfs/xfs_parent.c | 46 ++++++++++++++++++++++++++++++++++++++++++++++
libxfs/xfs_parent.h | 8 ++++++++
4 files changed, 56 insertions(+), 2 deletions(-)
diff --git a/libxfs/xfs_dir2.c b/libxfs/xfs_dir2.c
index 79b6ec893fd..b906f39e0fe 100644
--- a/libxfs/xfs_dir2.c
+++ b/libxfs/xfs_dir2.c
@@ -439,7 +439,7 @@ int
xfs_dir_removename(
struct xfs_trans *tp,
struct xfs_inode *dp,
- struct xfs_name *name,
+ const struct xfs_name *name,
xfs_ino_t ino,
xfs_extlen_t total) /* bmap's total block count */
{
diff --git a/libxfs/xfs_dir2.h b/libxfs/xfs_dir2.h
index f99788a1f3e..ca1949ed4f5 100644
--- a/libxfs/xfs_dir2.h
+++ b/libxfs/xfs_dir2.h
@@ -55,7 +55,7 @@ extern int xfs_dir_lookup(struct xfs_trans *tp, struct xfs_inode *dp,
const struct xfs_name *name, xfs_ino_t *inum,
struct xfs_name *ci_name);
extern int xfs_dir_removename(struct xfs_trans *tp, struct xfs_inode *dp,
- struct xfs_name *name, xfs_ino_t ino,
+ const struct xfs_name *name, xfs_ino_t ino,
xfs_extlen_t tot);
extern int xfs_dir_replace(struct xfs_trans *tp, struct xfs_inode *dp,
const struct xfs_name *name, xfs_ino_t inum,
diff --git a/libxfs/xfs_parent.c b/libxfs/xfs_parent.c
index 024e89756e6..f7cef51e1ec 100644
--- a/libxfs/xfs_parent.c
+++ b/libxfs/xfs_parent.c
@@ -421,3 +421,49 @@ xfs_parent_lookup(
return xfs_attr_get_ilocked(&scr->args);
}
+
+/*
+ * Attach the parent pointer (@pptr -> @name) to @ip immediately. Caller must
+ * not have a transaction or hold the ILOCK. This is for specialized repair
+ * functions only. The scratchpad need not be initialized.
+ */
+int
+xfs_parent_set(
+ struct xfs_inode *ip,
+ const struct xfs_parent_name_irec *pptr,
+ struct xfs_parent_scratch *scr)
+{
+ if (XFS_IS_CORRUPT(ip->i_mount,
+ !xfs_parent_verify_irec(ip->i_mount, pptr))) {
+ return -EFSCORRUPTED;
+ }
+
+ xfs_parent_irec_to_disk(&scr->rec, pptr);
+ xfs_parent_scratch_init(NULL, ip, pptr, scr);
+ scr->args.op_flags |= XFS_DA_OP_LOGGED;
+
+ return xfs_attr_set(&scr->args);
+}
+
+/*
+ * Remove the parent pointer (@rec -> @name) from @ip immediately. Caller must
+ * not have a transaction or hold the ILOCK. This is for specialized repair
+ * functions only. The scratchpad need not be initialized.
+ */
+int
+xfs_parent_unset(
+ struct xfs_inode *ip,
+ const struct xfs_parent_name_irec *pptr,
+ struct xfs_parent_scratch *scr)
+{
+ if (XFS_IS_CORRUPT(ip->i_mount,
+ !xfs_parent_verify_irec(ip->i_mount, pptr))) {
+ return -EFSCORRUPTED;
+ }
+
+ xfs_parent_irec_to_disk(&scr->rec, pptr);
+ xfs_parent_scratch_init(NULL, ip, pptr, scr);
+ scr->args.op_flags |= XFS_DA_OP_LOGGED | XFS_DA_OP_REMOVE;
+
+ return xfs_attr_set(&scr->args);
+}
diff --git a/libxfs/xfs_parent.h b/libxfs/xfs_parent.h
index e4443da1d86..58e59af818b 100644
--- a/libxfs/xfs_parent.h
+++ b/libxfs/xfs_parent.h
@@ -162,4 +162,12 @@ int xfs_parent_lookup(struct xfs_trans *tp, struct xfs_inode *ip,
const struct xfs_parent_name_irec *pptr,
struct xfs_parent_scratch *scratch);
+int xfs_parent_set(struct xfs_inode *ip,
+ const struct xfs_parent_name_irec *pptr,
+ struct xfs_parent_scratch *scratch);
+
+int xfs_parent_unset(struct xfs_inode *ip,
+ const struct xfs_parent_name_irec *rec,
+ struct xfs_parent_scratch *scratch);
+
#endif /* __XFS_PARENT_H__ */
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 05/18] xfs: set child file owner in xfs_da_args when changing parent pointers
2023-12-31 19:51 ` [PATCHSET v13.0 4/6] xfsprogs: fsck for " Darrick J. Wong
` (3 preceding siblings ...)
2023-12-31 23:14 ` [PATCH 04/18] xfs: add raw parent pointer apis to support repair Darrick J. Wong
@ 2023-12-31 23:14 ` Darrick J. Wong
2023-12-31 23:15 ` [PATCH 06/18] xfs: remove pointless unlocked assertion Darrick J. Wong
` (12 subsequent siblings)
17 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:14 UTC (permalink / raw)
To: djwong, cem; +Cc: catherine.hoang, linux-xfs, allison.henderson
From: Darrick J. Wong <djwong@kernel.org>
Now that struct xfs_da_args has an explicit file owner field, we must
set it when modifying parent pointers.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
libxfs/xfs_parent.c | 13 ++++++++++---
libxfs/xfs_parent.h | 4 ++--
2 files changed, 12 insertions(+), 5 deletions(-)
diff --git a/libxfs/xfs_parent.c b/libxfs/xfs_parent.c
index f7cef51e1ec..8f4196c20f1 100644
--- a/libxfs/xfs_parent.c
+++ b/libxfs/xfs_parent.c
@@ -202,6 +202,7 @@ xfs_parent_addname(
args->trans = tp;
args->dp = child;
+ args->owner = child->i_ino;
xfs_init_parent_davalue(&ppargs->args, parent_name);
@@ -240,6 +241,7 @@ xfs_parent_removename(
args->trans = tp;
args->dp = child;
+ args->owner = child->i_ino;
xfs_init_parent_davalue(&ppargs->args, parent_name);
@@ -289,6 +291,7 @@ xfs_parent_replacename(
args->trans = tp;
args->dp = child;
+ args->owner = child->i_ino;
xfs_init_parent_davalue(&ppargs->args, old_name);
xfs_init_parent_danewvalue(&ppargs->args, new_name);
@@ -372,6 +375,7 @@ static inline void
xfs_parent_scratch_init(
struct xfs_trans *tp,
struct xfs_inode *ip,
+ xfs_ino_t owner,
const struct xfs_parent_name_irec *pptr,
struct xfs_parent_scratch *scr)
{
@@ -388,6 +392,7 @@ xfs_parent_scratch_init(
scr->args.whichfork = XFS_ATTR_FORK;
scr->args.hashval = xfs_da_hashname((const void *)&scr->rec,
sizeof(struct xfs_parent_name_rec));
+ scr->args.owner = owner;
}
/*
@@ -416,7 +421,7 @@ xfs_parent_lookup(
}
xfs_parent_irec_to_disk(&scr->rec, pptr);
- xfs_parent_scratch_init(tp, ip, pptr, scr);
+ xfs_parent_scratch_init(tp, ip, ip->i_ino, pptr, scr);
scr->args.op_flags |= XFS_DA_OP_OKNOENT;
return xfs_attr_get_ilocked(&scr->args);
@@ -430,6 +435,7 @@ xfs_parent_lookup(
int
xfs_parent_set(
struct xfs_inode *ip,
+ xfs_ino_t owner,
const struct xfs_parent_name_irec *pptr,
struct xfs_parent_scratch *scr)
{
@@ -439,7 +445,7 @@ xfs_parent_set(
}
xfs_parent_irec_to_disk(&scr->rec, pptr);
- xfs_parent_scratch_init(NULL, ip, pptr, scr);
+ xfs_parent_scratch_init(NULL, ip, owner, pptr, scr);
scr->args.op_flags |= XFS_DA_OP_LOGGED;
return xfs_attr_set(&scr->args);
@@ -453,6 +459,7 @@ xfs_parent_set(
int
xfs_parent_unset(
struct xfs_inode *ip,
+ xfs_ino_t owner,
const struct xfs_parent_name_irec *pptr,
struct xfs_parent_scratch *scr)
{
@@ -462,7 +469,7 @@ xfs_parent_unset(
}
xfs_parent_irec_to_disk(&scr->rec, pptr);
- xfs_parent_scratch_init(NULL, ip, pptr, scr);
+ xfs_parent_scratch_init(NULL, ip, owner, pptr, scr);
scr->args.op_flags |= XFS_DA_OP_LOGGED | XFS_DA_OP_REMOVE;
return xfs_attr_set(&scr->args);
diff --git a/libxfs/xfs_parent.h b/libxfs/xfs_parent.h
index 58e59af818b..46bf96c7e3c 100644
--- a/libxfs/xfs_parent.h
+++ b/libxfs/xfs_parent.h
@@ -162,11 +162,11 @@ int xfs_parent_lookup(struct xfs_trans *tp, struct xfs_inode *ip,
const struct xfs_parent_name_irec *pptr,
struct xfs_parent_scratch *scratch);
-int xfs_parent_set(struct xfs_inode *ip,
+int xfs_parent_set(struct xfs_inode *ip, xfs_ino_t owner,
const struct xfs_parent_name_irec *pptr,
struct xfs_parent_scratch *scratch);
-int xfs_parent_unset(struct xfs_inode *ip,
+int xfs_parent_unset(struct xfs_inode *ip, xfs_ino_t owner,
const struct xfs_parent_name_irec *rec,
struct xfs_parent_scratch *scratch);
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 06/18] xfs: remove pointless unlocked assertion
2023-12-31 19:51 ` [PATCHSET v13.0 4/6] xfsprogs: fsck for " Darrick J. Wong
` (4 preceding siblings ...)
2023-12-31 23:14 ` [PATCH 05/18] xfs: set child file owner in xfs_da_args when changing parent pointers Darrick J. Wong
@ 2023-12-31 23:15 ` Darrick J. Wong
2023-12-31 23:15 ` [PATCH 07/18] xfs: split xfs_bmap_add_attrfork into two pieces Darrick J. Wong
` (11 subsequent siblings)
17 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:15 UTC (permalink / raw)
To: djwong, cem; +Cc: catherine.hoang, linux-xfs, allison.henderson
From: Darrick J. Wong <djwong@kernel.org>
Remove this assertion about the inode not having an attr fork from
xfs_bmap_add_attrfork because the function handles that case just fine.
Weirder still, the function actually /requires/ the caller not to hold
the ILOCK, which means that its accesses are not stabilized.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
libxfs/xfs_bmap.c | 2 --
1 file changed, 2 deletions(-)
diff --git a/libxfs/xfs_bmap.c b/libxfs/xfs_bmap.c
index c6f2f4ace53..f16e4369306 100644
--- a/libxfs/xfs_bmap.c
+++ b/libxfs/xfs_bmap.c
@@ -1017,8 +1017,6 @@ xfs_bmap_add_attrfork(
int logflags; /* logging flags */
int error; /* error return value */
- ASSERT(xfs_inode_has_attr_fork(ip) == 0);
-
mp = ip->i_mount;
ASSERT(!XFS_NOT_DQATTACHED(mp, ip));
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 07/18] xfs: split xfs_bmap_add_attrfork into two pieces
2023-12-31 19:51 ` [PATCHSET v13.0 4/6] xfsprogs: fsck for " Darrick J. Wong
` (5 preceding siblings ...)
2023-12-31 23:15 ` [PATCH 06/18] xfs: remove pointless unlocked assertion Darrick J. Wong
@ 2023-12-31 23:15 ` Darrick J. Wong
2023-12-31 23:15 ` [PATCH 08/18] xfs: actually rebuild the parent pointer xattrs Darrick J. Wong
` (10 subsequent siblings)
17 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:15 UTC (permalink / raw)
To: djwong, cem; +Cc: catherine.hoang, linux-xfs, allison.henderson
From: Darrick J. Wong <djwong@kernel.org>
Split this function into two pieces -- one to make the actual changes to
the inode core to add the attr fork, and another one to deal with
getting the transaction and locking the inodes.
The next couple of patches will need this to be split into two. One
patch implements committing new parent pointer recordsets to damaged
files. If one file has an attr fork and the other does not, we have to
create the missing attr fork before the atomic swap transaction, and can
use the behavior encoded in the current xfs_bmap_add_attrfork.
The second patch adapts /lost+found adoptions to handle parent pointers
correctly. The adoption process will add a parent pointer to a child
that is being moved to /lost+found, but this requires that the attr fork
already exists. We don't know if we're actually going to commit the
adoption until we've already reserved a transaction and taken the
ILOCKs, which means that we must have a way to bypass the start of the
current xfs_bmap_add_attrfork.
Therefore, create xfs_attr_add_fork as the helper that creates a
transaction and takes locks; and make xfs_bmap_add_attrfork the function
that updates the inode core and allocates the incore attr fork.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
libxfs/xfs_attr.c | 39 ++++++++++++++++++++++++++++++++++++++-
libxfs/xfs_bmap.c | 36 ++++++++++--------------------------
libxfs/xfs_bmap.h | 3 ++-
3 files changed, 50 insertions(+), 28 deletions(-)
diff --git a/libxfs/xfs_attr.c b/libxfs/xfs_attr.c
index c4f543db474..cb4c2726fd7 100644
--- a/libxfs/xfs_attr.c
+++ b/libxfs/xfs_attr.c
@@ -933,6 +933,43 @@ xfs_attr_defer_add(
trace_xfs_attr_defer_add(new->xattri_dela_state, args->dp);
}
+STATIC int
+xfs_attr_add_fork(
+ struct xfs_inode *ip, /* incore inode pointer */
+ int size, /* space new attribute needs */
+ int rsvd) /* xact may use reserved blks */
+{
+ struct xfs_mount *mp = ip->i_mount;
+ struct xfs_trans *tp; /* transaction pointer */
+ unsigned int blks; /* space reservation */
+ int error; /* error return value */
+
+ ASSERT(!XFS_NOT_DQATTACHED(mp, ip));
+
+ blks = XFS_ADDAFORK_SPACE_RES(mp);
+
+ error = xfs_trans_alloc_inode(ip, &M_RES(mp)->tr_addafork, blks, 0,
+ rsvd, &tp);
+ if (error)
+ return error;
+
+ if (xfs_inode_has_attr_fork(ip))
+ goto trans_cancel;
+
+ error = xfs_bmap_add_attrfork(tp, ip, size, rsvd);
+ if (error)
+ goto trans_cancel;
+
+ error = xfs_trans_commit(tp);
+ xfs_iunlock(ip, XFS_ILOCK_EXCL);
+ return error;
+
+trans_cancel:
+ xfs_trans_cancel(tp);
+ xfs_iunlock(ip, XFS_ILOCK_EXCL);
+ return error;
+}
+
/*
* Note: If args->value is NULL the attribute will be removed, just like the
* Linux ->setattr API.
@@ -984,7 +1021,7 @@ xfs_attr_set(
xfs_attr_sf_entsize_byname(args->namelen,
args->valuelen);
- error = xfs_bmap_add_attrfork(dp, sf_size, rsvd);
+ error = xfs_attr_add_fork(dp, sf_size, rsvd);
if (error)
return error;
}
diff --git a/libxfs/xfs_bmap.c b/libxfs/xfs_bmap.c
index f16e4369306..5c69720e19e 100644
--- a/libxfs/xfs_bmap.c
+++ b/libxfs/xfs_bmap.c
@@ -1001,38 +1001,29 @@ xfs_bmap_set_attrforkoff(
}
/*
- * Convert inode from non-attributed to attributed.
- * Must not be in a transaction, ip must not be locked.
+ * Convert inode from non-attributed to attributed. Caller must hold the
+ * ILOCK_EXCL and the file cannot have an attr fork.
*/
int /* error code */
xfs_bmap_add_attrfork(
- xfs_inode_t *ip, /* incore inode pointer */
+ struct xfs_trans *tp,
+ struct xfs_inode *ip, /* incore inode pointer */
int size, /* space new attribute needs */
int rsvd) /* xact may use reserved blks */
{
- xfs_mount_t *mp; /* mount structure */
- xfs_trans_t *tp; /* transaction pointer */
- int blks; /* space reservation */
+ struct xfs_mount *mp = tp->t_mountp;
int version = 1; /* superblock attr version */
int logflags; /* logging flags */
int error; /* error return value */
- mp = ip->i_mount;
+ ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL));
ASSERT(!XFS_NOT_DQATTACHED(mp, ip));
-
- blks = XFS_ADDAFORK_SPACE_RES(mp);
-
- error = xfs_trans_alloc_inode(ip, &M_RES(mp)->tr_addafork, blks, 0,
- rsvd, &tp);
- if (error)
- return error;
- if (xfs_inode_has_attr_fork(ip))
- goto trans_cancel;
+ ASSERT(!xfs_inode_has_attr_fork(ip));
xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
error = xfs_bmap_set_attrforkoff(ip, size, &version);
if (error)
- goto trans_cancel;
+ return error;
xfs_ifork_init_attr(ip, XFS_DINODE_FMT_EXTENTS, 0);
logflags = 0;
@@ -1053,7 +1044,7 @@ xfs_bmap_add_attrfork(
if (logflags)
xfs_trans_log_inode(tp, ip, logflags);
if (error)
- goto trans_cancel;
+ return error;
if (!xfs_has_attr(mp) ||
(!xfs_has_attr2(mp) && version == 2)) {
bool log_sb = false;
@@ -1072,14 +1063,7 @@ xfs_bmap_add_attrfork(
xfs_log_sb(tp);
}
- error = xfs_trans_commit(tp);
- xfs_iunlock(ip, XFS_ILOCK_EXCL);
- return error;
-
-trans_cancel:
- xfs_trans_cancel(tp);
- xfs_iunlock(ip, XFS_ILOCK_EXCL);
- return error;
+ return 0;
}
/*
diff --git a/libxfs/xfs_bmap.h b/libxfs/xfs_bmap.h
index 87633449c37..c9e297dba88 100644
--- a/libxfs/xfs_bmap.h
+++ b/libxfs/xfs_bmap.h
@@ -174,7 +174,8 @@ int xfs_bmap_longest_free_extent(struct xfs_perag *pag,
void xfs_trim_extent(struct xfs_bmbt_irec *irec, xfs_fileoff_t bno,
xfs_filblks_t len);
unsigned int xfs_bmap_compute_attr_offset(struct xfs_mount *mp);
-int xfs_bmap_add_attrfork(struct xfs_inode *ip, int size, int rsvd);
+int xfs_bmap_add_attrfork(struct xfs_trans *tp, struct xfs_inode *ip,
+ int size, int rsvd);
void xfs_bmap_local_to_extents_empty(struct xfs_trans *tp,
struct xfs_inode *ip, int whichfork);
int xfs_bmap_local_to_extents(struct xfs_trans *tp, struct xfs_inode *ip,
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 08/18] xfs: actually rebuild the parent pointer xattrs
2023-12-31 19:51 ` [PATCHSET v13.0 4/6] xfsprogs: fsck for " Darrick J. Wong
` (6 preceding siblings ...)
2023-12-31 23:15 ` [PATCH 07/18] xfs: split xfs_bmap_add_attrfork into two pieces Darrick J. Wong
@ 2023-12-31 23:15 ` Darrick J. Wong
2023-12-31 23:15 ` [PATCH 09/18] xfs_repair: add parent pointers when messing with /lost+found Darrick J. Wong
` (9 subsequent siblings)
17 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:15 UTC (permalink / raw)
To: djwong, cem; +Cc: catherine.hoang, linux-xfs, allison.henderson
From: Darrick J. Wong <djwong@kernel.org>
Once we've assembled all the parent pointers for a file, we need to
commit the new dataset atomically to that file. Parent pointer records
are embedded in the xattr structure, which means that we must write a
new extended attribute structure, again, atomically. Therefore, we must
copy the non-parent-pointer attributes from the file being repaired into
the temporary file's extended attributes and then call the atomic extent
swap mechanism to exchange the blocks.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
libxfs/xfs_attr.c | 2 +-
libxfs/xfs_attr.h | 2 ++
2 files changed, 3 insertions(+), 1 deletion(-)
diff --git a/libxfs/xfs_attr.c b/libxfs/xfs_attr.c
index cb4c2726fd7..5da0ac9f706 100644
--- a/libxfs/xfs_attr.c
+++ b/libxfs/xfs_attr.c
@@ -933,7 +933,7 @@ xfs_attr_defer_add(
trace_xfs_attr_defer_add(new->xattri_dela_state, args->dp);
}
-STATIC int
+int
xfs_attr_add_fork(
struct xfs_inode *ip, /* incore inode pointer */
int size, /* space new attribute needs */
diff --git a/libxfs/xfs_attr.h b/libxfs/xfs_attr.h
index 0204f62298c..a2cfe9e35fd 100644
--- a/libxfs/xfs_attr.h
+++ b/libxfs/xfs_attr.h
@@ -650,4 +650,6 @@ xfs_attri_can_use_without_log_assistance(
return false;
}
+int xfs_attr_add_fork(struct xfs_inode *ip, int size, int rsvd);
+
#endif /* __XFS_ATTR_H__ */
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 09/18] xfs_repair: add parent pointers when messing with /lost+found
2023-12-31 19:51 ` [PATCHSET v13.0 4/6] xfsprogs: fsck for " Darrick J. Wong
` (7 preceding siblings ...)
2023-12-31 23:15 ` [PATCH 08/18] xfs: actually rebuild the parent pointer xattrs Darrick J. Wong
@ 2023-12-31 23:15 ` Darrick J. Wong
2023-12-31 23:16 ` [PATCH 10/18] xfs_repair: build a parent pointer index Darrick J. Wong
` (8 subsequent siblings)
17 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:15 UTC (permalink / raw)
To: djwong, cem; +Cc: catherine.hoang, linux-xfs, allison.henderson
From: Darrick J. Wong <djwong@kernel.org>
Make sure that the /lost+found gets created with parent pointers, and
that lost children being put in there get new parent pointers.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
libxfs/libxfs_api_defs.h | 2 +
repair/phase6.c | 73 ++++++++++++++++++++++++++++++++++++++++++++++
2 files changed, 75 insertions(+)
diff --git a/libxfs/libxfs_api_defs.h b/libxfs/libxfs_api_defs.h
index 7ea7eebfbca..1aa2d9f0679 100644
--- a/libxfs/libxfs_api_defs.h
+++ b/libxfs/libxfs_api_defs.h
@@ -185,6 +185,8 @@
#define xfs_parent_add libxfs_parent_add
#define xfs_parent_finish libxfs_parent_finish
#define xfs_parent_irec_from_disk libxfs_parent_irec_from_disk
+#define xfs_parent_irec_hashname libxfs_parent_irec_hashname
+#define xfs_parent_lookup libxfs_parent_lookup
#define xfs_parent_start libxfs_parent_start
#define xfs_parent_hashcheck libxfs_parent_hashcheck
#define xfs_parent_namecheck libxfs_parent_namecheck
diff --git a/repair/phase6.c b/repair/phase6.c
index 9b43e58b3d3..2bdddafe213 100644
--- a/repair/phase6.c
+++ b/repair/phase6.c
@@ -902,6 +902,12 @@ mk_orphanage(xfs_mount_t *mp)
const int mode = 0755;
int nres;
struct xfs_name xname;
+ struct xfs_parent_args *ppargs;
+
+ i = -libxfs_parent_start(mp, &ppargs);
+ if (i)
+ do_error(_("%d - couldn't allocate parent pointer for %s\n"),
+ i, ORPHANAGE);
/*
* check for an existing lost+found first, if it exists, return
@@ -991,6 +997,12 @@ mk_orphanage(xfs_mount_t *mp)
_("can't make %s, createname error %d\n"),
ORPHANAGE, error);
+ error = -libxfs_parent_add(tp, ppargs, pip, &xname, ip);
+ if (error)
+ do_error(
+ _("committing %s parent pointer failed, error %d.\n"),
+ ORPHANAGE, error);
+
/*
* bump up the link count in the root directory to account
* for .. in the new directory, and update the irec copy of the
@@ -1012,10 +1024,51 @@ mk_orphanage(xfs_mount_t *mp)
}
libxfs_irele(ip);
libxfs_irele(pip);
+ libxfs_parent_finish(mp, ppargs);
return(ino);
}
+/*
+ * Add a parent pointer back to the orphanage for any file we're moving into
+ * the orphanage, being careful not to trip over any existing parent pointer.
+ * You never know when the orphanage might get corrupted.
+ */
+static void
+add_orphan_pptr(
+ struct xfs_trans *tp,
+ struct xfs_inode *orphanage_ip,
+ const struct xfs_name *xname,
+ struct xfs_inode *ip,
+ struct xfs_parent_args *ppargs)
+{
+ struct xfs_parent_name_irec pptr = {
+ .p_ino = orphanage_ip->i_ino,
+ .p_gen = VFS_I(orphanage_ip)->i_generation,
+ .p_namelen = xname->len,
+ };
+ struct xfs_parent_scratch scr = { };
+ struct xfs_mount *mp = tp->t_mountp;
+ int error;
+
+ memcpy(pptr.p_name, xname->name, xname->len);
+ libxfs_parent_irec_hashname(mp, &pptr);
+
+ error = -libxfs_parent_lookup(tp, ip, &pptr, &scr);
+ if (!error)
+ return;
+ if (error != ENOATTR)
+ do_log(
+ _("cannot look up parent pointer for '%.*s', err %d\n"),
+ xname->len, xname->name, error);
+
+ error = -libxfs_parent_add(tp, ppargs, orphanage_ip, xname, ip);
+ if (error)
+ do_error(
+ _("adding '%.*s' parent pointer failed, error %d.\n"),
+ xname->len, xname->name, error);
+}
+
/*
* move a file to the orphange.
*/
@@ -1036,6 +1089,13 @@ mv_orphanage(
ino_tree_node_t *irec;
int ino_offset = 0;
struct xfs_name xname;
+ struct xfs_parent_args *ppargs;
+
+ err = -libxfs_parent_start(mp, &ppargs);
+ if (err)
+ do_error(
+ _("%d - couldn't allocate parent pointer for lost inode\n"),
+ err);
xname.name = fname;
xname.len = snprintf((char *)fname, sizeof(fname), "%llu",
@@ -1087,6 +1147,10 @@ mv_orphanage(
do_error(
_("name create failed in %s (%d)\n"), ORPHANAGE, err);
+ if (ppargs)
+ add_orphan_pptr(tp, orphanage_ip, &xname,
+ ino_p, ppargs);
+
if (irec)
add_inode_ref(irec, ino_offset);
else
@@ -1121,6 +1185,10 @@ mv_orphanage(
do_error(
_("name create failed in %s (%d)\n"), ORPHANAGE, err);
+ if (ppargs)
+ add_orphan_pptr(tp, orphanage_ip, &xname,
+ ino_p, ppargs);
+
if (irec)
add_inode_ref(irec, ino_offset);
else
@@ -1169,6 +1237,10 @@ mv_orphanage(
_("name create failed in %s (%d)\n"), ORPHANAGE, err);
ASSERT(err == 0);
+ if (ppargs)
+ add_orphan_pptr(tp, orphanage_ip, &xname, ino_p,
+ ppargs);
+
set_nlink(VFS_I(ino_p), 1);
libxfs_trans_log_inode(tp, ino_p, XFS_ILOG_CORE);
err = -libxfs_trans_commit(tp);
@@ -1178,6 +1250,7 @@ mv_orphanage(
}
libxfs_irele(ino_p);
libxfs_irele(orphanage_ip);
+ libxfs_parent_finish(mp, ppargs);
}
static int
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 10/18] xfs_repair: build a parent pointer index
2023-12-31 19:51 ` [PATCHSET v13.0 4/6] xfsprogs: fsck for " Darrick J. Wong
` (8 preceding siblings ...)
2023-12-31 23:15 ` [PATCH 09/18] xfs_repair: add parent pointers when messing with /lost+found Darrick J. Wong
@ 2023-12-31 23:16 ` Darrick J. Wong
2023-12-31 23:16 ` [PATCH 11/18] xfs_repair: move the global dirent name store to a separate object Darrick J. Wong
` (7 subsequent siblings)
17 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:16 UTC (permalink / raw)
To: djwong, cem; +Cc: catherine.hoang, linux-xfs, allison.henderson
From: Darrick J. Wong <djwong@kernel.org>
When we're walking directories during phase 6, build an index of parent
pointers that we expect to find.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
repair/Makefile | 2 +
repair/phase6.c | 43 +++++++++++-
repair/pptr.c | 204 +++++++++++++++++++++++++++++++++++++++++++++++++++++++
repair/pptr.h | 15 ++++
4 files changed, 262 insertions(+), 2 deletions(-)
create mode 100644 repair/pptr.c
create mode 100644 repair/pptr.h
diff --git a/repair/Makefile b/repair/Makefile
index 250c86cca2d..a5102015651 100644
--- a/repair/Makefile
+++ b/repair/Makefile
@@ -24,6 +24,7 @@ HFILES = \
err_protos.h \
globals.h \
incore.h \
+ pptr.h \
prefetch.h \
progress.h \
protos.h \
@@ -63,6 +64,7 @@ CFILES = \
phase5.c \
phase6.c \
phase7.c \
+ pptr.c \
prefetch.c \
progress.c \
quotacheck.c \
diff --git a/repair/phase6.c b/repair/phase6.c
index 2bdddafe213..e1cd1984986 100644
--- a/repair/phase6.c
+++ b/repair/phase6.c
@@ -18,6 +18,7 @@
#include "dinode.h"
#include "progress.h"
#include "versions.h"
+#include "repair/pptr.h"
static struct cred zerocr;
static struct fsxattr zerofsx;
@@ -996,6 +997,7 @@ mk_orphanage(xfs_mount_t *mp)
do_error(
_("can't make %s, createname error %d\n"),
ORPHANAGE, error);
+ add_parent_ptr(ip->i_ino, ORPHANAGE, pip);
error = -libxfs_parent_add(tp, ppargs, pip, &xname, ip);
if (error)
@@ -1248,6 +1250,10 @@ mv_orphanage(
do_error(
_("orphanage name create failed (%d)\n"), err);
}
+
+ if (xfs_has_parent(mp))
+ add_parent_ptr(ino_p->i_ino, xname.name, orphanage_ip);
+
libxfs_irele(ino_p);
libxfs_irele(orphanage_ip);
libxfs_parent_finish(mp, ppargs);
@@ -2549,6 +2555,7 @@ shortform_dir2_entry_check(
struct xfs_dir2_sf_entry *next_sfep;
struct xfs_ifork *ifp;
struct ino_tree_node *irec;
+ xfs_dir2_dataptr_t diroffset;
int max_size;
int ino_offset;
int i;
@@ -2727,8 +2734,9 @@ shortform_dir2_entry_check(
/*
* check for duplicate names in directory.
*/
- if (!dir_hash_add(mp, hashtab, (xfs_dir2_dataptr_t)
- (sfep - xfs_dir2_sf_firstentry(sfp)),
+ diroffset = xfs_dir2_byte_to_dataptr(
+ xfs_dir2_sf_get_offset(sfep));
+ if (!dir_hash_add(mp, hashtab, diroffset,
lino, sfep->namelen, sfep->name,
libxfs_dir2_sf_get_ftype(mp, sfep))) {
do_warn(
@@ -2762,6 +2770,7 @@ _("entry \"%s\" (ino %" PRIu64 ") in dir %" PRIu64 " is a duplicate name, "),
next_sfep = shortform_dir2_junk(mp, sfp, sfep,
lino, &max_size, &i,
&bytes_deleted, ino_dirty);
+ dir_hash_junkit(hashtab, diroffset);
continue;
} else if (parent == ino) {
add_inode_reached(irec, ino_offset);
@@ -2786,6 +2795,7 @@ _("entry \"%s\" (ino %" PRIu64 ") in dir %" PRIu64 " is a duplicate name, "),
next_sfep = shortform_dir2_junk(mp, sfp, sfep,
lino, &max_size, &i,
&bytes_deleted, ino_dirty);
+ dir_hash_junkit(hashtab, diroffset);
continue;
}
}
@@ -2877,6 +2887,30 @@ _("entry \"%s\" (ino %" PRIu64 ") in dir %" PRIu64 " is a duplicate name, "),
}
}
+static void
+dir_hash_add_parent_ptrs(
+ struct xfs_inode *dp,
+ struct dir_hash_tab *hashtab)
+{
+ struct dir_hash_ent *p;
+
+ if (!xfs_has_parent(dp->i_mount))
+ return;
+
+ for (p = hashtab->first; p; p = p->nextbyorder) {
+ if (p->junkit)
+ continue;
+ if (p->name.name[0] == '/')
+ continue;
+ if (p->name.name[0] == '.' &&
+ (p->name.len == 1 ||
+ (p->name.len == 2 && p->name.name[1] == '.')))
+ continue;
+
+ add_parent_ptr(p->inum, p->name.name, dp);
+ }
+}
+
/*
* processes all reachable inodes in directories
*/
@@ -3003,6 +3037,7 @@ _("error %d fixing shortform directory %llu\n"),
default:
break;
}
+ dir_hash_add_parent_ptrs(ip, hashtab);
dir_hash_done(hashtab);
/*
@@ -3294,6 +3329,8 @@ phase6(xfs_mount_t *mp)
ino_tree_node_t *irec;
int i;
+ parent_ptr_init(mp);
+
memset(&zerocr, 0, sizeof(struct cred));
memset(&zerofsx, 0, sizeof(struct fsxattr));
orphanage_ino = 0;
@@ -3394,4 +3431,6 @@ _(" - resetting contents of realtime bitmap and summary inodes\n"));
irec = next_ino_rec(irec);
}
}
+
+ parent_ptr_free(mp);
}
diff --git a/repair/pptr.c b/repair/pptr.c
new file mode 100644
index 00000000000..68cdd0ae424
--- /dev/null
+++ b/repair/pptr.c
@@ -0,0 +1,204 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * Copyright (c) 2023-2024 Oracle. All Rights Reserved.
+ * Author: Darrick J. Wong <djwong@kernel.org>
+ */
+#include "libxfs.h"
+#include "libxfs/xfile.h"
+#include "libxfs/xfblob.h"
+#include "libfrog/platform.h"
+#include "repair/err_protos.h"
+#include "repair/slab.h"
+#include "repair/pptr.h"
+
+#undef PPTR_DEBUG
+
+#ifdef PPTR_DEBUG
+# define dbg_printf(f, a...) do {printf(f, ## a); fflush(stdout); } while (0)
+#else
+# define dbg_printf(f, a...)
+#endif
+
+/*
+ * Parent Pointer Validation
+ * =========================
+ *
+ * Phase 6 validates the connectivity of the directory tree after validating
+ * that all the space metadata are correct, and confirming all the inodes that
+ * we intend to keep. The first part of phase 6 walks the directories of the
+ * filesystem to ensure that every file that isn't the root directory has a
+ * parent. Unconnected files are attached to the orphanage. Filesystems with
+ * the directory parent pointer feature enabled must also ensure that for every
+ * directory entry that points to a child file, that child has a matching
+ * parent pointer.
+ *
+ * There are many ways that we could check the parent pointers, but the means
+ * that we have chosen is to build a per-AG master index of all parent pointers
+ * of all inodes stored in that AG, and use that as the basis for comparison.
+ * This consumes a lot of memory, but performing both a forward scan to check
+ * dirent -> parent pointer and a backwards scan of parent pointer -> dirent
+ * takes longer than the simple method presented here. Userspace adds the
+ * additional twist that inodes are not cached (and there are no ILOCKs), which
+ * makes that approach even less attractive.
+ *
+ * During the directory walk at the start of phase 6, we transform each child
+ * directory entry found into its parent pointer equivalent. In other words,
+ * the forward information:
+ *
+ * (dir_ino, name, child_ino)
+ *
+ * becomes this backwards information:
+ *
+ * (child_agino*, dir_ino*, dir_gen, name*)
+ *
+ * Key fields are starred.
+ *
+ * This tuple is recorded in the per-AG master parent pointer index. Note
+ * that names are stored separately in an xfblob data structure so that the
+ * rest of the information can be sorted and processed as fixed-size records;
+ * the incore parent pointer record contains a pointer to the xfblob data.
+ */
+
+struct ag_pptr {
+ /* parent directory handle */
+ xfs_ino_t parent_ino;
+ unsigned int parent_gen;
+
+ /* dirent name length */
+ unsigned int namelen;
+
+ /* cookie for the actual dirent name */
+ xfblob_cookie name_cookie;
+
+ /* agino of the child file */
+ xfs_agino_t child_agino;
+
+ /* hash of the dirent name */
+ xfs_dahash_t namehash;
+};
+
+struct ag_pptrs {
+ /* Lock to protect pptr_recs during the dirent scan. */
+ pthread_mutex_t lock;
+
+ /* Parent pointer records for files in this AG. */
+ struct xfs_slab *pptr_recs;
+};
+
+/* Global names storage file. */
+static struct xfblob *names;
+static pthread_mutex_t names_mutex = PTHREAD_MUTEX_INITIALIZER;
+static struct ag_pptrs *fs_pptrs;
+
+void
+parent_ptr_free(
+ struct xfs_mount *mp)
+{
+ xfs_agnumber_t agno;
+
+ if (!xfs_has_parent(mp))
+ return;
+
+ for (agno = 0; agno < mp->m_sb.sb_agcount; agno++) {
+ free_slab(&fs_pptrs[agno].pptr_recs);
+ pthread_mutex_destroy(&fs_pptrs[agno].lock);
+ }
+ free(fs_pptrs);
+ fs_pptrs = NULL;
+
+ xfblob_destroy(names);
+}
+
+void
+parent_ptr_init(
+ struct xfs_mount *mp)
+{
+ char *descr;
+ xfs_agnumber_t agno;
+ int error;
+
+ if (!xfs_has_parent(mp))
+ return;
+
+ descr = kasprintf("xfs_repair (%s): parent pointer names",
+ mp->m_fsname);
+ error = -xfblob_create(descr, &names);
+ kfree(descr);
+ if (error)
+ do_error(_("init parent pointer names failed: %s\n"),
+ strerror(error));
+
+ fs_pptrs = calloc(mp->m_sb.sb_agcount, sizeof(struct ag_pptrs));
+ if (!fs_pptrs)
+ do_error(
+ _("init parent pointer per-AG record array failed: %s\n"),
+ strerror(errno));
+
+ for (agno = 0; agno < mp->m_sb.sb_agcount; agno++) {
+ error = pthread_mutex_init(&fs_pptrs[agno].lock, NULL);
+ if (error)
+ do_error(
+ _("init agno %u parent pointer lock failed: %s\n"),
+ agno, strerror(error));
+
+ error = -init_slab(&fs_pptrs[agno].pptr_recs,
+ sizeof(struct ag_pptr));
+ if (error)
+ do_error(
+ _("init agno %u parent pointer recs failed: %s\n"),
+ agno, strerror(error));
+ }
+}
+
+/* Remember that @dp has a dirent (@fname, @ino). */
+void
+add_parent_ptr(
+ xfs_ino_t ino,
+ const unsigned char *fname,
+ struct xfs_inode *dp)
+{
+ struct xfs_mount *mp = dp->i_mount;
+ struct xfs_name dname = {
+ .name = fname,
+ .len = strlen(fname),
+ };
+ struct ag_pptr ag_pptr = {
+ .child_agino = XFS_INO_TO_AGINO(mp, ino),
+ .parent_ino = dp->i_ino,
+ .parent_gen = VFS_I(dp)->i_generation,
+ .namelen = dname.len,
+ };
+ struct ag_pptrs *ag_pptrs;
+ xfs_agnumber_t agno = XFS_INO_TO_AGNO(mp, ino);
+ int error;
+
+ if (!xfs_has_parent(mp))
+ return;
+
+ ag_pptr.namehash = libxfs_dir2_hashname(mp, &dname);
+
+ pthread_mutex_lock(&names_mutex);
+ error = -xfblob_store(names, &ag_pptr.name_cookie, fname,
+ ag_pptr.namelen);
+ pthread_mutex_unlock(&names_mutex);
+ if (error)
+ do_error(_("storing name '%s' failed: %s\n"),
+ fname, strerror(error));
+
+ ag_pptrs = &fs_pptrs[agno];
+ pthread_mutex_lock(&ag_pptrs->lock);
+ error = -slab_add(ag_pptrs->pptr_recs, &ag_pptr);
+ pthread_mutex_unlock(&ag_pptrs->lock);
+ if (error)
+ do_error(_("storing name '%s' key failed: %s\n"),
+ fname, strerror(error));
+
+ dbg_printf(
+ _("%s: dp %llu gen 0x%x fname '%s' ino %llu namecookie 0x%llx\n"),
+ __func__,
+ (unsigned long long)dp->i_ino,
+ VFS_I(dp)->i_generation,
+ fname,
+ (unsigned long long)ino,
+ (unsigned long long)ag_pptr.name_cookie);
+}
diff --git a/repair/pptr.h b/repair/pptr.h
new file mode 100644
index 00000000000..ef85807b117
--- /dev/null
+++ b/repair/pptr.h
@@ -0,0 +1,15 @@
+/* SPDX-License-Identifier: GPL-2.0-or-later */
+/*
+ * Copyright (c) 2023-2024 Oracle. All Rights Reserved.
+ * Author: Darrick J. Wong <djwong@kernel.org>
+ */
+#ifndef __REPAIR_PPTR_H__
+#define __REPAIR_PPTR_H__
+
+void parent_ptr_free(struct xfs_mount *mp);
+void parent_ptr_init(struct xfs_mount *mp);
+
+void add_parent_ptr(xfs_ino_t ino, const unsigned char *fname,
+ struct xfs_inode *dp);
+
+#endif /* __REPAIR_PPTR_H__ */
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 11/18] xfs_repair: move the global dirent name store to a separate object
2023-12-31 19:51 ` [PATCHSET v13.0 4/6] xfsprogs: fsck for " Darrick J. Wong
` (9 preceding siblings ...)
2023-12-31 23:16 ` [PATCH 10/18] xfs_repair: build a parent pointer index Darrick J. Wong
@ 2023-12-31 23:16 ` Darrick J. Wong
2023-12-31 23:16 ` [PATCH 12/18] xfs_repair: deduplicate strings stored in string blob Darrick J. Wong
` (6 subsequent siblings)
17 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:16 UTC (permalink / raw)
To: djwong, cem; +Cc: catherine.hoang, linux-xfs, allison.henderson
From: Darrick J. Wong <djwong@kernel.org>
Abstract the main parent pointer dirent names xfblob object into a
separate data structure to hide implementation details.
The goals here are (a) reduce memory usage when we can by deduplicating
dirent names that exist in multiple directories; and (b) provide a
unique id for each name in the system so that sorting incore parent
pointer records can be done in a stable manner. Fast stable sorting of
records is required for the dirent <-> pptr matching algorithm.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
repair/Makefile | 2 +
repair/pptr.c | 11 ++++---
repair/strblobs.c | 79 +++++++++++++++++++++++++++++++++++++++++++++++++++++
repair/strblobs.h | 19 +++++++++++++
4 files changed, 106 insertions(+), 5 deletions(-)
create mode 100644 repair/strblobs.c
create mode 100644 repair/strblobs.h
diff --git a/repair/Makefile b/repair/Makefile
index a5102015651..320f2f9a21d 100644
--- a/repair/Makefile
+++ b/repair/Makefile
@@ -35,6 +35,7 @@ HFILES = \
rt.h \
scan.h \
slab.h \
+ strblobs.h \
threads.h \
versions.h
@@ -75,6 +76,7 @@ CFILES = \
sb.c \
scan.c \
slab.c \
+ strblobs.c \
threads.c \
versions.c \
xfs_repair.c
diff --git a/repair/pptr.c b/repair/pptr.c
index 68cdd0ae424..88970f81348 100644
--- a/repair/pptr.c
+++ b/repair/pptr.c
@@ -10,6 +10,7 @@
#include "repair/err_protos.h"
#include "repair/slab.h"
#include "repair/pptr.h"
+#include "repair/strblobs.h"
#undef PPTR_DEBUG
@@ -56,7 +57,7 @@
* This tuple is recorded in the per-AG master parent pointer index. Note
* that names are stored separately in an xfblob data structure so that the
* rest of the information can be sorted and processed as fixed-size records;
- * the incore parent pointer record contains a pointer to the xfblob data.
+ * the incore parent pointer record contains a pointer to the strblob data.
*/
struct ag_pptr {
@@ -86,7 +87,7 @@ struct ag_pptrs {
};
/* Global names storage file. */
-static struct xfblob *names;
+static struct strblobs *nameblobs;
static pthread_mutex_t names_mutex = PTHREAD_MUTEX_INITIALIZER;
static struct ag_pptrs *fs_pptrs;
@@ -106,7 +107,7 @@ parent_ptr_free(
free(fs_pptrs);
fs_pptrs = NULL;
- xfblob_destroy(names);
+ strblobs_destroy(&nameblobs);
}
void
@@ -122,7 +123,7 @@ parent_ptr_init(
descr = kasprintf("xfs_repair (%s): parent pointer names",
mp->m_fsname);
- error = -xfblob_create(descr, &names);
+ error = strblobs_init(descr, &nameblobs);
kfree(descr);
if (error)
do_error(_("init parent pointer names failed: %s\n"),
@@ -178,7 +179,7 @@ add_parent_ptr(
ag_pptr.namehash = libxfs_dir2_hashname(mp, &dname);
pthread_mutex_lock(&names_mutex);
- error = -xfblob_store(names, &ag_pptr.name_cookie, fname,
+ error = strblobs_store(nameblobs, &ag_pptr.name_cookie, fname,
ag_pptr.namelen);
pthread_mutex_unlock(&names_mutex);
if (error)
diff --git a/repair/strblobs.c b/repair/strblobs.c
new file mode 100644
index 00000000000..45d2559c722
--- /dev/null
+++ b/repair/strblobs.c
@@ -0,0 +1,79 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * Copyright (c) 2023-2024 Oracle. All Rights Reserved.
+ * Author: Darrick J. Wong <djwong@kernel.org>
+ */
+#include "libxfs.h"
+#include "libxfs/xfile.h"
+#include "libxfs/xfblob.h"
+#include "repair/strblobs.h"
+
+/*
+ * String Blob Structure
+ * =====================
+ *
+ * This data structure wraps the storage of strings with explicit length in an
+ * xfblob structure.
+ */
+struct strblobs {
+ struct xfblob *strings;
+};
+
+/* Initialize a string blob structure. */
+int
+strblobs_init(
+ const char *descr,
+ struct strblobs **sblobs)
+{
+ struct strblobs *sb;
+ int error;
+
+ sb = malloc(sizeof(struct strblobs));
+ if (!sb)
+ return ENOMEM;
+
+ error = -xfblob_create(descr, &sb->strings);
+ if (error)
+ goto out_free;
+
+ *sblobs = sb;
+ return 0;
+
+out_free:
+ free(sb);
+ return error;
+}
+
+/* Deconstruct a string blob structure. */
+void
+strblobs_destroy(
+ struct strblobs **sblobs)
+{
+ struct strblobs *sb = *sblobs;
+
+ xfblob_destroy(sb->strings);
+ free(sb);
+ *sblobs = NULL;
+}
+
+/* Store a string and return a cookie for its retrieval. */
+int
+strblobs_store(
+ struct strblobs *sblobs,
+ xfblob_cookie *str_cookie,
+ const unsigned char *str,
+ unsigned int str_len)
+{
+ return -xfblob_store(sblobs->strings, str_cookie, str, str_len);
+}
+
+/* Retrieve a previously stored string. */
+int
+strblobs_load(
+ struct strblobs *sblobs,
+ xfblob_cookie str_cookie,
+ unsigned char *str,
+ unsigned int str_len)
+{
+ return -xfblob_load(sblobs->strings, str_cookie, str, str_len);
+}
diff --git a/repair/strblobs.h b/repair/strblobs.h
new file mode 100644
index 00000000000..27e98eee208
--- /dev/null
+++ b/repair/strblobs.h
@@ -0,0 +1,19 @@
+/* SPDX-License-Identifier: GPL-2.0-or-later */
+/*
+ * Copyright (c) 2023-2024 Oracle. All Rights Reserved.
+ * Author: Darrick J. Wong <djwong@kernel.org>
+ */
+#ifndef __REPAIR_STRBLOBS_H__
+#define __REPAIR_STRBLOBS_H__
+
+struct strblobs;
+
+int strblobs_init(const char *descr, struct strblobs **sblobs);
+void strblobs_destroy(struct strblobs **sblobs);
+
+int strblobs_store(struct strblobs *sblobs, xfblob_cookie *str_cookie,
+ const unsigned char *str, unsigned int str_len);
+int strblobs_load(struct strblobs *sblobs, xfblob_cookie str_cookie,
+ unsigned char *str, unsigned int str_len);
+
+#endif /* __REPAIR_STRBLOBS_H__ */
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 12/18] xfs_repair: deduplicate strings stored in string blob
2023-12-31 19:51 ` [PATCHSET v13.0 4/6] xfsprogs: fsck for " Darrick J. Wong
` (10 preceding siblings ...)
2023-12-31 23:16 ` [PATCH 11/18] xfs_repair: move the global dirent name store to a separate object Darrick J. Wong
@ 2023-12-31 23:16 ` Darrick J. Wong
2023-12-31 23:16 ` [PATCH 13/18] xfs_repair: check parent pointers Darrick J. Wong
` (5 subsequent siblings)
17 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:16 UTC (permalink / raw)
To: djwong, cem; +Cc: catherine.hoang, linux-xfs, allison.henderson
From: Darrick J. Wong <djwong@kernel.org>
Reduce the memory requirements of the string blob structure by
deduplicating the strings stored within.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
repair/pptr.c | 13 ++++-
repair/strblobs.c | 140 +++++++++++++++++++++++++++++++++++++++++++++++++++--
repair/strblobs.h | 9 +++
3 files changed, 153 insertions(+), 9 deletions(-)
diff --git a/repair/pptr.c b/repair/pptr.c
index 88970f81348..3ea5514531c 100644
--- a/repair/pptr.c
+++ b/repair/pptr.c
@@ -50,7 +50,7 @@
*
* becomes this backwards information:
*
- * (child_agino*, dir_ino*, dir_gen, name*)
+ * (child_agino*, dir_ino*, dir_gen, name_cookie*)
*
* Key fields are starred.
*
@@ -58,6 +58,10 @@
* that names are stored separately in an xfblob data structure so that the
* rest of the information can be sorted and processed as fixed-size records;
* the incore parent pointer record contains a pointer to the strblob data.
+ * Because string blobs are deduplicated, there's a 1:1 mapping of name cookies
+ * to strings, which means that we can use the name cookie as a comparison key
+ * instead of loading the full dentry name every time we want to perform a
+ * comparison.
*/
struct ag_pptr {
@@ -115,15 +119,18 @@ parent_ptr_init(
struct xfs_mount *mp)
{
char *descr;
+ uint64_t iused;
xfs_agnumber_t agno;
int error;
if (!xfs_has_parent(mp))
return;
+ /* One hash bucket per inode, up to about 8M of memory on 64-bit. */
+ iused = min(mp->m_sb.sb_icount - mp->m_sb.sb_ifree, 1048573);
descr = kasprintf("xfs_repair (%s): parent pointer names",
mp->m_fsname);
- error = strblobs_init(descr, &nameblobs);
+ error = strblobs_init(descr, iused, &nameblobs);
kfree(descr);
if (error)
do_error(_("init parent pointer names failed: %s\n"),
@@ -180,7 +187,7 @@ add_parent_ptr(
pthread_mutex_lock(&names_mutex);
error = strblobs_store(nameblobs, &ag_pptr.name_cookie, fname,
- ag_pptr.namelen);
+ ag_pptr.namelen, ag_pptr.namehash);
pthread_mutex_unlock(&names_mutex);
if (error)
do_error(_("storing name '%s' failed: %s\n"),
diff --git a/repair/strblobs.c b/repair/strblobs.c
index 45d2559c722..4ed21e9536d 100644
--- a/repair/strblobs.c
+++ b/repair/strblobs.c
@@ -13,22 +13,42 @@
* =====================
*
* This data structure wraps the storage of strings with explicit length in an
- * xfblob structure.
+ * xfblob structure. It stores a hashtable of string checksums to provide
+ * fast(ish) lookups of existing strings to enable deduplication of the strings
+ * contained within.
*/
+struct strblob_hashent {
+ struct strblob_hashent *next;
+
+ xfblob_cookie str_cookie;
+ unsigned int str_len;
+ xfs_dahash_t str_hash;
+};
+
struct strblobs {
struct xfblob *strings;
+ unsigned int nr_buckets;
+
+ struct strblob_hashent *buckets[];
};
+static inline size_t strblobs_sizeof(unsigned int nr_buckets)
+{
+ return sizeof(struct strblobs) +
+ (nr_buckets * sizeof(struct strblobs_hashent *));
+}
+
/* Initialize a string blob structure. */
int
strblobs_init(
const char *descr,
+ unsigned int hash_buckets,
struct strblobs **sblobs)
{
struct strblobs *sb;
int error;
- sb = malloc(sizeof(struct strblobs));
+ sb = calloc(strblobs_sizeof(hash_buckets), 1);
if (!sb)
return ENOMEM;
@@ -36,6 +56,7 @@ strblobs_init(
if (error)
goto out_free;
+ sb->nr_buckets = hash_buckets;
*sblobs = sb;
return 0;
@@ -50,21 +71,132 @@ strblobs_destroy(
struct strblobs **sblobs)
{
struct strblobs *sb = *sblobs;
+ struct strblob_hashent *ent, *ent_next;
+ unsigned int bucket;
+
+ for (bucket = 0; bucket < sb->nr_buckets; bucket++) {
+ ent = sb->buckets[bucket];
+ while (ent != NULL) {
+ ent_next = ent->next;
+ free(ent);
+ ent = ent_next;
+ }
+ }
xfblob_destroy(sb->strings);
free(sb);
*sblobs = NULL;
}
+/*
+ * Search the string hashtable for a matching entry. Sets sets the cookie and
+ * returns 0 if one is found; ENOENT if there is no match; or a positive errno.
+ */
+static int
+__strblobs_lookup(
+ struct strblobs *sblobs,
+ xfblob_cookie *str_cookie,
+ const unsigned char *str,
+ unsigned int str_len,
+ xfs_dahash_t str_hash)
+{
+ struct strblob_hashent *ent;
+ char *buf = NULL;
+ unsigned int bucket;
+ int error;
+
+ bucket = str_hash % sblobs->nr_buckets;
+ ent = sblobs->buckets[bucket];
+
+ for (ent = sblobs->buckets[bucket]; ent != NULL; ent = ent->next) {
+ if (ent->str_len != str_len || ent->str_hash != str_hash)
+ continue;
+
+ if (!buf) {
+ buf = malloc(str_len);
+ if (!buf)
+ return ENOMEM;
+ }
+
+ error = strblobs_load(sblobs, ent->str_cookie, buf, str_len);
+ if (error)
+ goto out;
+
+ if (memcmp(str, buf, str_len))
+ continue;
+
+ *str_cookie = ent->str_cookie;
+ goto out;
+ }
+ error = ENOENT;
+
+out:
+ free(buf);
+ return error;
+}
+
+/*
+ * Search the string hashtable for a matching entry. Sets sets the cookie and
+ * returns 0 if one is found; ENOENT if there is no match; or a positive errno.
+ */
+int
+strblobs_lookup(
+ struct strblobs *sblobs,
+ xfblob_cookie *str_cookie,
+ const unsigned char *str,
+ unsigned int str_len,
+ xfs_dahash_t str_hash)
+{
+ return __strblobs_lookup(sblobs, str_cookie, str, str_len, str_hash);
+}
+
+/* Remember a string in the hashtable. */
+static int
+strblobs_hash(
+ struct strblobs *sblobs,
+ xfblob_cookie str_cookie,
+ const unsigned char *str,
+ unsigned int str_len,
+ xfs_dahash_t str_hash)
+{
+ struct strblob_hashent *ent;
+ unsigned int bucket;
+
+ bucket = str_hash % sblobs->nr_buckets;
+
+ ent = malloc(sizeof(struct strblob_hashent));
+ if (!ent)
+ return ENOMEM;
+
+ ent->str_cookie = str_cookie;
+ ent->str_len = str_len;
+ ent->str_hash = str_hash;
+ ent->next = sblobs->buckets[bucket];
+
+ sblobs->buckets[bucket] = ent;
+ return 0;
+}
+
/* Store a string and return a cookie for its retrieval. */
int
strblobs_store(
struct strblobs *sblobs,
xfblob_cookie *str_cookie,
const unsigned char *str,
- unsigned int str_len)
+ unsigned int str_len,
+ xfs_dahash_t str_hash)
{
- return -xfblob_store(sblobs->strings, str_cookie, str, str_len);
+ int error;
+
+ error = __strblobs_lookup(sblobs, str_cookie, str, str_len, str_hash);
+ if (error != ENOENT)
+ return error;
+
+ error = -xfblob_store(sblobs->strings, str_cookie, str, str_len);
+ if (error)
+ return error;
+
+ return strblobs_hash(sblobs, *str_cookie, str, str_len, str_hash);
}
/* Retrieve a previously stored string. */
diff --git a/repair/strblobs.h b/repair/strblobs.h
index 27e98eee208..40cd6d8e91c 100644
--- a/repair/strblobs.h
+++ b/repair/strblobs.h
@@ -8,12 +8,17 @@
struct strblobs;
-int strblobs_init(const char *descr, struct strblobs **sblobs);
+int strblobs_init(const char *descr, unsigned int hash_buckets,
+ struct strblobs **sblobs);
void strblobs_destroy(struct strblobs **sblobs);
int strblobs_store(struct strblobs *sblobs, xfblob_cookie *str_cookie,
- const unsigned char *str, unsigned int str_len);
+ const unsigned char *str, unsigned int str_len,
+ xfs_dahash_t hash);
int strblobs_load(struct strblobs *sblobs, xfblob_cookie str_cookie,
unsigned char *str, unsigned int str_len);
+int strblobs_lookup(struct strblobs *sblobs, xfblob_cookie *str_cookie,
+ const unsigned char *str, unsigned int str_len,
+ xfs_dahash_t hash);
#endif /* __REPAIR_STRBLOBS_H__ */
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 13/18] xfs_repair: check parent pointers
2023-12-31 19:51 ` [PATCHSET v13.0 4/6] xfsprogs: fsck for " Darrick J. Wong
` (11 preceding siblings ...)
2023-12-31 23:16 ` [PATCH 12/18] xfs_repair: deduplicate strings stored in string blob Darrick J. Wong
@ 2023-12-31 23:16 ` Darrick J. Wong
2023-12-31 23:17 ` [PATCH 14/18] xfs_repair: dump garbage parent pointer attributes Darrick J. Wong
` (4 subsequent siblings)
17 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:16 UTC (permalink / raw)
To: djwong, cem; +Cc: catherine.hoang, linux-xfs, allison.henderson
From: Darrick J. Wong <djwong@kernel.org>
Use the parent pointer index that we constructed in the previous patch
to check that each file's parent pointer records exactly match the
directory entries that we recorded while walking directory entries.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
libxfs/xfblob.c | 9 +
libxfs/xfblob.h | 2
repair/Makefile | 2
repair/listxattr.c | 271 +++++++++++++++++
repair/listxattr.h | 15 +
repair/phase6.c | 2
repair/pptr.c | 819 ++++++++++++++++++++++++++++++++++++++++++++++++++++
repair/pptr.h | 2
8 files changed, 1122 insertions(+)
create mode 100644 repair/listxattr.c
create mode 100644 repair/listxattr.h
diff --git a/libxfs/xfblob.c b/libxfs/xfblob.c
index d826e5f3cb0..acaa2120a90 100644
--- a/libxfs/xfblob.c
+++ b/libxfs/xfblob.c
@@ -145,3 +145,12 @@ xfblob_free(
xfile_discard(blob->xfile, cookie, sizeof(key) + key.xb_size);
return 0;
}
+
+/* Drop all the blobs. */
+void
+xfblob_truncate(
+ struct xfblob *blob)
+{
+ xfile_discard(blob->xfile, PAGE_SIZE, blob->last_offset - PAGE_SIZE);
+ blob->last_offset = PAGE_SIZE;
+}
diff --git a/libxfs/xfblob.h b/libxfs/xfblob.h
index 28bf4ab2898..1939202e12d 100644
--- a/libxfs/xfblob.h
+++ b/libxfs/xfblob.h
@@ -21,4 +21,6 @@ int xfblob_store(struct xfblob *blob, xfblob_cookie *cookie, const void *ptr,
uint32_t size);
int xfblob_free(struct xfblob *blob, xfblob_cookie cookie);
+void xfblob_truncate(struct xfblob *blob);
+
#endif /* __XFS_SCRUB_XFBLOB_H__ */
diff --git a/repair/Makefile b/repair/Makefile
index 320f2f9a21d..1f72c811056 100644
--- a/repair/Makefile
+++ b/repair/Makefile
@@ -24,6 +24,7 @@ HFILES = \
err_protos.h \
globals.h \
incore.h \
+ listxattr.h \
pptr.h \
prefetch.h \
progress.h \
@@ -58,6 +59,7 @@ CFILES = \
incore_ext.c \
incore_ino.c \
init.c \
+ listxattr.c \
phase1.c \
phase2.c \
phase3.c \
diff --git a/repair/listxattr.c b/repair/listxattr.c
new file mode 100644
index 00000000000..52630987f83
--- /dev/null
+++ b/repair/listxattr.c
@@ -0,0 +1,271 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * Copyright (c) 2022-2024 Oracle. All Rights Reserved.
+ * Author: Darrick J. Wong <djwong@kernel.org>
+ */
+#include "libxfs.h"
+#include "libxlog.h"
+#include "libfrog/bitmap.h"
+#include "repair/listxattr.h"
+
+/* Call a function for every entry in a shortform xattr structure. */
+STATIC int
+xattr_walk_sf(
+ struct xfs_inode *ip,
+ xattr_walk_fn attr_fn,
+ void *priv)
+{
+ struct xfs_attr_shortform *sf;
+ struct xfs_attr_sf_entry *sfe;
+ unsigned int i;
+ int error;
+
+ sf = (struct xfs_attr_shortform *)ip->i_af.if_u1.if_data;
+ for (i = 0, sfe = &sf->list[0]; i < sf->hdr.count; i++) {
+ error = attr_fn(ip, sfe->flags, sfe->nameval, sfe->namelen,
+ &sfe->nameval[sfe->namelen], sfe->valuelen,
+ priv);
+ if (error)
+ return error;
+
+ sfe = xfs_attr_sf_nextentry(sfe);
+ }
+
+ return 0;
+}
+
+/* Call a function for every entry in this xattr leaf block. */
+STATIC int
+xattr_walk_leaf_entries(
+ struct xfs_inode *ip,
+ xattr_walk_fn attr_fn,
+ struct xfs_buf *bp,
+ void *priv)
+{
+ struct xfs_attr3_icleaf_hdr ichdr;
+ struct xfs_mount *mp = ip->i_mount;
+ struct xfs_attr_leafblock *leaf = bp->b_addr;
+ struct xfs_attr_leaf_entry *entry;
+ unsigned int i;
+ int error;
+
+ libxfs_attr3_leaf_hdr_from_disk(mp->m_attr_geo, &ichdr, leaf);
+ entry = xfs_attr3_leaf_entryp(leaf);
+
+ for (i = 0; i < ichdr.count; entry++, i++) {
+ void *value;
+ char *name;
+ unsigned int namelen, valuelen;
+
+ if (entry->flags & XFS_ATTR_LOCAL) {
+ struct xfs_attr_leaf_name_local *name_loc;
+
+ name_loc = xfs_attr3_leaf_name_local(leaf, i);
+ name = name_loc->nameval;
+ namelen = name_loc->namelen;
+ value = &name_loc->nameval[name_loc->namelen];
+ valuelen = be16_to_cpu(name_loc->valuelen);
+ } else {
+ struct xfs_attr_leaf_name_remote *name_rmt;
+
+ name_rmt = xfs_attr3_leaf_name_remote(leaf, i);
+ name = name_rmt->name;
+ namelen = name_rmt->namelen;
+ value = NULL;
+ valuelen = be32_to_cpu(name_rmt->valuelen);
+ }
+
+ error = attr_fn(ip, entry->flags, name, namelen, value,
+ valuelen, priv);
+ if (error)
+ return error;
+
+ }
+
+ return 0;
+}
+
+/*
+ * Call a function for every entry in a leaf-format xattr structure. Avoid
+ * memory allocations for the loop detector since there's only one block.
+ */
+STATIC int
+xattr_walk_leaf(
+ struct xfs_inode *ip,
+ xattr_walk_fn attr_fn,
+ void *priv)
+{
+ struct xfs_buf *leaf_bp;
+ int error;
+
+ error = -libxfs_attr3_leaf_read(NULL, ip, ip->i_ino, 0, &leaf_bp);
+ if (error)
+ return error;
+
+ error = xattr_walk_leaf_entries(ip, attr_fn, leaf_bp, priv);
+ libxfs_trans_brelse(NULL, leaf_bp);
+ return error;
+}
+
+/* Find the leftmost leaf in the xattr dabtree. */
+STATIC int
+xattr_walk_find_leftmost_leaf(
+ struct xfs_inode *ip,
+ struct bitmap *seen_blocks,
+ struct xfs_buf **leaf_bpp)
+{
+ struct xfs_da3_icnode_hdr nodehdr;
+ struct xfs_mount *mp = ip->i_mount;
+ struct xfs_da_intnode *node;
+ struct xfs_da_node_entry *btree;
+ struct xfs_buf *bp;
+ //xfs_failaddr_t fa;
+ xfs_dablk_t blkno = 0;
+ unsigned int expected_level = 0;
+ int error;
+
+ for (;;) {
+ uint16_t magic;
+
+ error = -libxfs_da3_node_read(NULL, ip, blkno, &bp,
+ XFS_ATTR_FORK);
+ if (error)
+ return error;
+
+ node = bp->b_addr;
+ magic = be16_to_cpu(node->hdr.info.magic);
+ if (magic == XFS_ATTR_LEAF_MAGIC ||
+ magic == XFS_ATTR3_LEAF_MAGIC)
+ break;
+
+ error = EFSCORRUPTED;
+ if (magic != XFS_DA_NODE_MAGIC &&
+ magic != XFS_DA3_NODE_MAGIC)
+ goto out_buf;
+
+ libxfs_da3_node_hdr_from_disk(mp, &nodehdr, node);
+
+ if (nodehdr.count == 0 || nodehdr.level >= XFS_DA_NODE_MAXDEPTH)
+ goto out_buf;
+
+ /* Check the level from the root node. */
+ if (blkno == 0)
+ expected_level = nodehdr.level - 1;
+ else if (expected_level != nodehdr.level)
+ goto out_buf;
+ else
+ expected_level--;
+
+ /* Remember that we've seen this node. */
+ error = -bitmap_set(seen_blocks, blkno, 1);
+ if (error)
+ goto out_buf;
+
+ /* Find the next level towards the leaves of the dabtree. */
+ btree = nodehdr.btree;
+ blkno = be32_to_cpu(btree->before);
+ libxfs_trans_brelse(NULL, bp);
+
+ /* Make sure we haven't seen this new block already. */
+ if (bitmap_test(seen_blocks, blkno, 1))
+ return EFSCORRUPTED;
+ }
+
+ error = EFSCORRUPTED;
+ if (expected_level != 0)
+ goto out_buf;
+
+ /* Remember that we've seen this leaf. */
+ error = -bitmap_set(seen_blocks, blkno, 1);
+ if (error)
+ goto out_buf;
+
+ *leaf_bpp = bp;
+ return 0;
+
+out_buf:
+ libxfs_trans_brelse(NULL, bp);
+ return error;
+}
+
+/* Call a function for every entry in a node-format xattr structure. */
+STATIC int
+xattr_walk_node(
+ struct xfs_inode *ip,
+ xattr_walk_fn attr_fn,
+ void *priv)
+{
+ struct xfs_attr3_icleaf_hdr leafhdr;
+ struct bitmap *seen_blocks;
+ struct xfs_mount *mp = ip->i_mount;
+ struct xfs_attr_leafblock *leaf;
+ struct xfs_buf *leaf_bp;
+ int error;
+
+ bitmap_alloc(&seen_blocks);
+
+ error = xattr_walk_find_leftmost_leaf(ip, seen_blocks, &leaf_bp);
+ if (error)
+ goto out_bitmap;
+
+ for (;;) {
+ error = xattr_walk_leaf_entries(ip, attr_fn, leaf_bp,
+ priv);
+ if (error)
+ goto out_leaf;
+
+ /* Find the right sibling of this leaf block. */
+ leaf = leaf_bp->b_addr;
+ libxfs_attr3_leaf_hdr_from_disk(mp->m_attr_geo, &leafhdr, leaf);
+ if (leafhdr.forw == 0)
+ goto out_leaf;
+
+ libxfs_trans_brelse(NULL, leaf_bp);
+
+ /* Make sure we haven't seen this new leaf already. */
+ if (bitmap_test(seen_blocks, leafhdr.forw, 1))
+ goto out_bitmap;
+
+ error = -libxfs_attr3_leaf_read(NULL, ip, ip->i_ino,
+ leafhdr.forw, &leaf_bp);
+ if (error)
+ goto out_bitmap;
+
+ /* Remember that we've seen this new leaf. */
+ error = -bitmap_set(seen_blocks, leafhdr.forw, 1);
+ if (error)
+ goto out_leaf;
+ }
+
+out_leaf:
+ libxfs_trans_brelse(NULL, leaf_bp);
+out_bitmap:
+ bitmap_free(&seen_blocks);
+ return error;
+}
+
+/* Call a function for every extended attribute in a file. */
+int
+xattr_walk(
+ struct xfs_inode *ip,
+ xattr_walk_fn attr_fn,
+ void *priv)
+{
+ int error;
+
+ if (!libxfs_inode_hasattr(ip))
+ return 0;
+
+ if (ip->i_af.if_format == XFS_DINODE_FMT_LOCAL)
+ return xattr_walk_sf(ip, attr_fn, priv);
+
+ /* attr functions require that the attr fork is loaded */
+ error = -libxfs_iread_extents(NULL, ip, XFS_ATTR_FORK);
+ if (error)
+ return error;
+
+ if (libxfs_attr_is_leaf(ip))
+ return xattr_walk_leaf(ip, attr_fn, priv);
+
+ return xattr_walk_node(ip, attr_fn, priv);
+}
diff --git a/repair/listxattr.h b/repair/listxattr.h
new file mode 100644
index 00000000000..2d26fce0f32
--- /dev/null
+++ b/repair/listxattr.h
@@ -0,0 +1,15 @@
+/* SPDX-License-Identifier: GPL-2.0-or-later */
+/*
+ * Copyright (c) 2022-2024 Oracle. All Rights Reserved.
+ * Author: Darrick J. Wong <djwong@kernel.org>
+ */
+#ifndef __REPAIR_LISTXATTR_H__
+#define __REPAIR_LISTXATTR_H__
+
+typedef int (*xattr_walk_fn)(struct xfs_inode *ip, unsigned int attr_flags,
+ const unsigned char *name, unsigned int namelen,
+ const void *value, unsigned int valuelen, void *priv);
+
+int xattr_walk(struct xfs_inode *ip, xattr_walk_fn attr_fn, void *priv);
+
+#endif /* __REPAIR_LISTXATTR_H__ */
diff --git a/repair/phase6.c b/repair/phase6.c
index e1cd1984986..5e95dabbe09 100644
--- a/repair/phase6.c
+++ b/repair/phase6.c
@@ -3432,5 +3432,7 @@ _(" - resetting contents of realtime bitmap and summary inodes\n"));
}
}
+ /* Check and repair directory parent pointers, if enabled. */
+ check_parent_ptrs(mp);
parent_ptr_free(mp);
}
diff --git a/repair/pptr.c b/repair/pptr.c
index 3ea5514531c..11aa8d4e322 100644
--- a/repair/pptr.c
+++ b/repair/pptr.c
@@ -7,8 +7,13 @@
#include "libxfs/xfile.h"
#include "libxfs/xfblob.h"
#include "libfrog/platform.h"
+#include "libfrog/workqueue.h"
+#include "repair/globals.h"
#include "repair/err_protos.h"
#include "repair/slab.h"
+#include "repair/listxattr.h"
+#include "repair/threads.h"
+#include "repair/incore.h"
#include "repair/pptr.h"
#include "repair/strblobs.h"
@@ -62,6 +67,65 @@
* to strings, which means that we can use the name cookie as a comparison key
* instead of loading the full dentry name every time we want to perform a
* comparison.
+ *
+ * Once we've finished with the forward scan, we get to work on the backwards
+ * scan. Each AG is processed independently. First, we sort the per-AG master
+ * records in order of child_agino, dir_ino, and name_cookie. Each inode in
+ * the AG is then processed in numerical order.
+ *
+ * The first thing that happens to the file is that we read all the extended
+ * attributes to look for parent pointers. Attributes that claim to be parent
+ * pointers but are obviously garbage are thrown away. The rest of the ondisk
+ * parent pointers for that file are stored in memory like this:
+ *
+ * (dir_ino*, dir_gen, name_cookie*)
+ *
+ * After loading the ondisk parent pointer name, we search the strblobs
+ * structure to see if it has already recorded the name. If so, this value is
+ * used as the name cookie. If the name has not yet been recorded, we flag the
+ * incore record for later deletion.
+ *
+ * When we've concluded the xattr scan, the per-file records are sorted in
+ * order of dir_ino and name_cookie.
+ *
+ * There are three possibilities here:
+ *
+ * A. The first record in the per-AG master index is an exact match for the
+ * first record in the per-file index. Everything is consistent, and we can
+ * proceed with the lockstep scan detailed below.
+ *
+ * B. The per-AG master index cursor points to a higher inode number than the
+ * first inode we are scanning. Delete the ondisk parent pointers
+ * corresponding to the per-file records until condition (B) is no longer true.
+ *
+ * C. The per-AG master index cursor instead points to a lower inode number
+ * than the one we are scanning. This means that there exists a directory
+ * entry pointing at an inode that is free. We supposedly already settled
+ * which inodes are free and which aren't, which means in-memory information is
+ * inconsistent. Abort.
+ *
+ * Otherwise, we are ready to check the file parent pointers against the
+ * master. If the ondisk directory metadata are all consistent, this recordset
+ * should correspond exactly to the subset of the master records with a
+ * child_agino matching the file that we're scanning. We should be able to
+ * walk both sets in lockstep, and find one of the following outcomes:
+ *
+ * 1) The master index cursor is ahead of the ondisk index cursor. This means
+ * that the inode has parent pointers that were not found during the dirent
+ * scan. These should be deleted.
+ *
+ * 2) The ondisk index gets ahead of the master index. This means that the
+ * dirent scan found parent pointers that are not attached to the inode.
+ * These should be added.
+ *
+ * 3) The parent_gen or (dirent) name are not consistent. Update the parent
+ * pointer to the values that we found during the dirent scan.
+ *
+ * 4) Everything matches. Move on to the next parent pointer.
+ *
+ * The current implementation does not try to rebuild directories from parent
+ * pointer information, as this requires a lengthy scan of the filesystem for
+ * each broken directory.
*/
struct ag_pptr {
@@ -82,6 +146,24 @@ struct ag_pptr {
xfs_dahash_t namehash;
};
+struct file_pptr {
+ /* parent directory handle */
+ unsigned long long parent_ino;
+ unsigned int parent_gen;
+
+ /* Is the name stored in the global nameblobs structure? */
+ unsigned int name_in_nameblobs;
+
+ /* hash of the dirent name */
+ xfs_dahash_t namehash;
+
+ /* parent pointer name length */
+ unsigned int namelen;
+
+ /* cookie for the file dirent name */
+ xfblob_cookie name_cookie;
+};
+
struct ag_pptrs {
/* Lock to protect pptr_recs during the dirent scan. */
pthread_mutex_t lock;
@@ -90,11 +172,99 @@ struct ag_pptrs {
struct xfs_slab *pptr_recs;
};
+struct file_scan {
+ struct ag_pptrs *ag_pptrs;
+
+ /* cursor for comparing ag_pptrs.pptr_recs against file_pptrs_recs */
+ struct xfs_slab_cursor *ag_pptr_recs_cur;
+
+ /* xfs_parent_name_rec records for a file that we're checking */
+ struct xfs_slab *file_pptr_recs;
+
+ /* cursor for comparing file_pptr_recs against pptrs_recs */
+ struct xfs_slab_cursor *file_pptr_recs_cur;
+
+ /* names associated with file_pptr_recs */
+ struct xfblob *file_pptr_names;
+
+ /* Number of parent pointers recorded for this file. */
+ unsigned int nr_file_pptrs;
+
+ /* Does this file have garbage xattrs with ATTR_PARENT set? */
+ bool have_garbage;
+};
+
/* Global names storage file. */
static struct strblobs *nameblobs;
static pthread_mutex_t names_mutex = PTHREAD_MUTEX_INITIALIZER;
static struct ag_pptrs *fs_pptrs;
+static int
+cmp_ag_pptr(
+ const void *a,
+ const void *b)
+{
+ const struct ag_pptr *pa = a;
+ const struct ag_pptr *pb = b;
+
+ if (pa->child_agino < pb->child_agino)
+ return -1;
+ if (pa->child_agino > pb->child_agino)
+ return 1;
+
+ if (pa->parent_ino < pb->parent_ino)
+ return -1;
+ if (pa->parent_ino > pb->parent_ino)
+ return 1;
+
+ if (pa->namehash < pb->namehash)
+ return -1;
+ if (pa->namehash > pb->namehash)
+ return 1;
+
+ if (pa->name_cookie < pb->name_cookie)
+ return -1;
+ if (pa->name_cookie > pb->name_cookie)
+ return 1;
+
+ return 0;
+}
+
+static int
+cmp_file_pptr(
+ const void *a,
+ const void *b)
+{
+ const struct file_pptr *pa = a;
+ const struct file_pptr *pb = b;
+
+ if (pa->parent_ino < pb->parent_ino)
+ return -1;
+ if (pa->parent_ino > pb->parent_ino)
+ return 1;
+
+ /*
+ * Push the parent pointer names that we didn't find in the dirent scan
+ * towards the end of the list so that we delete them as excess.
+ */
+ if (!pa->name_in_nameblobs && pb->name_in_nameblobs)
+ return 1;
+ if (pa->name_in_nameblobs && !pb->name_in_nameblobs)
+ return -1;
+
+ if (pa->namehash < pb->namehash)
+ return -1;
+ if (pa->namehash > pb->namehash)
+ return 1;
+
+ if (pa->name_cookie < pb->name_cookie)
+ return -1;
+ if (pa->name_cookie > pb->name_cookie)
+ return 1;
+
+ return 0;
+}
+
void
parent_ptr_free(
struct xfs_mount *mp)
@@ -210,3 +380,652 @@ add_parent_ptr(
(unsigned long long)ino,
(unsigned long long)ag_pptr.name_cookie);
}
+
+/* Schedule this ATTR_PARENT extended attribute for deletion. */
+static void
+record_garbage_xattr(
+ struct xfs_inode *ip,
+ struct file_scan *fscan,
+ unsigned int attr_filter,
+ const char *name,
+ unsigned int namelen,
+ const void *value,
+ unsigned int valuelen)
+{
+ if (no_modify) {
+ if (!fscan->have_garbage)
+ do_warn(
+ _("would delete garbage parent pointer extended attributes in ino %llu\n"),
+ (unsigned long long)ip->i_ino);
+ fscan->have_garbage = true;
+ return;
+ }
+
+ if (fscan->have_garbage)
+ return;
+ fscan->have_garbage = true;
+
+ do_warn(
+ _("deleting garbage parent pointer extended attributes in ino %llu\n"),
+ (unsigned long long)ip->i_ino);
+ /* XXX do the work */
+}
+
+/*
+ * Store this file parent pointer's name in the file scan namelist unless it's
+ * already in the global list.
+ */
+static int
+store_file_pptr_name(
+ struct file_scan *fscan,
+ struct file_pptr *file_pptr,
+ const struct xfs_parent_name_irec *irec)
+{
+ int error;
+
+ error = strblobs_lookup(nameblobs, &file_pptr->name_cookie,
+ irec->p_name, irec->p_namelen, file_pptr->namehash);
+ if (!error) {
+ file_pptr->name_in_nameblobs = true;
+ return 0;
+ }
+ if (error != ENOENT)
+ return error;
+
+ file_pptr->name_in_nameblobs = false;
+ return -xfblob_store(fscan->file_pptr_names, &file_pptr->name_cookie,
+ irec->p_name, irec->p_namelen);
+}
+
+/* Decide if this is a directory parent pointer and stash it if so. */
+static int
+examine_xattr(
+ struct xfs_inode *ip,
+ unsigned int attr_flags,
+ const unsigned char *name,
+ unsigned int namelen,
+ const void *value,
+ unsigned int valuelen,
+ void *priv)
+{
+ struct file_pptr file_pptr = { };
+ struct xfs_name dname = {
+ .name = value,
+ .len = valuelen,
+ };
+ struct xfs_parent_name_irec irec;
+ struct xfs_mount *mp = ip->i_mount;
+ struct file_scan *fscan = priv;
+ const struct xfs_parent_name_rec *rec = (const void *)name;
+ int error;
+
+ /* Ignore anything that isn't a parent pointer. */
+ if (!(attr_flags & XFS_ATTR_PARENT))
+ return 0;
+
+ /* No incomplete parent pointers. */
+ if (attr_flags & XFS_ATTR_INCOMPLETE)
+ goto corrupt;
+
+ /* Does the ondisk parent pointer structure make sense? */
+ if (!xfs_parent_namecheck(mp, rec, namelen, attr_flags) ||
+ !xfs_parent_valuecheck(mp, value, valuelen))
+ goto corrupt;
+
+ /*
+ * If the namehash of the dirent name encoded in the parent pointer
+ * attr value doesn't match the namehash in the parent pointer key,
+ * delete this attribute.
+ */
+ if (!xfs_parent_hashcheck(mp, rec, value, valuelen)) {
+ xfs_dahash_t computed_hash;
+
+ computed_hash = libxfs_dir2_hashname(ip->i_mount, &dname);
+ do_warn(
+ _("bad hash 0x%x for ino %llu parent pointer '%.*s', expected 0x%x\n"),
+ irec.p_namehash,
+ (unsigned long long)ip->i_ino,
+ irec.p_namelen,
+ (const char *)irec.p_name,
+ computed_hash);
+ goto corrupt;
+ }
+
+ libxfs_parent_irec_from_disk(&irec, rec, value, valuelen);
+
+ file_pptr.parent_ino = irec.p_ino;
+ file_pptr.parent_gen = irec.p_gen;
+ file_pptr.namelen = irec.p_namelen;
+ file_pptr.namehash = irec.p_namehash;
+
+ error = store_file_pptr_name(fscan, &file_pptr, &irec);
+ if (error)
+ do_error(
+ _("storing ino %llu parent pointer '%.*s' failed: %s\n"),
+ (unsigned long long)ip->i_ino,
+ irec.p_namelen,
+ (const char *)irec.p_name,
+ strerror(error));
+
+ error = -slab_add(fscan->file_pptr_recs, &file_pptr);
+ if (error)
+ do_error(_("storing ino %llu parent pointer rec failed: %s\n"),
+ (unsigned long long)ip->i_ino,
+ strerror(error));
+
+ dbg_printf(
+ _("%s: dp %llu gen 0x%x fname '%.*s' namelen %u ino %llu namecookie 0x%llx global? %d\n"),
+ __func__,
+ (unsigned long long)irec.p_ino,
+ irec.p_gen,
+ irec.p_namelen,
+ (const char *)irec.p_name,
+ irec.p_namelen,
+ (unsigned long long)ip->i_ino,
+ (unsigned long long)file_pptr.name_cookie,
+ file_pptr.name_in_nameblobs);
+
+ fscan->nr_file_pptrs++;
+ return 0;
+corrupt:
+ record_garbage_xattr(ip, fscan, attr_flags, name, namelen, value,
+ valuelen);
+ return 0;
+}
+
+/* Load a file parent pointer name from wherever we stored it. */
+static int
+load_file_pptr_name(
+ struct file_scan *fscan,
+ const struct file_pptr *file_pptr,
+ unsigned char *name)
+{
+ if (file_pptr->name_in_nameblobs)
+ return strblobs_load(nameblobs, file_pptr->name_cookie,
+ name, file_pptr->namelen);
+
+ return -xfblob_load(fscan->file_pptr_names, file_pptr->name_cookie,
+ name, file_pptr->namelen);
+}
+
+/* Remove all pptrs from @ip. */
+static void
+clear_all_pptrs(
+ struct xfs_inode *ip)
+{
+ if (no_modify) {
+ do_warn(_("would delete unlinked ino %llu parent pointers\n"),
+ (unsigned long long)ip->i_ino);
+ return;
+ }
+
+ do_warn(_("deleting unlinked ino %llu parent pointers\n"),
+ (unsigned long long)ip->i_ino);
+ /* XXX actually do the work */
+}
+
+/* Add @ag_pptr to @ip. */
+static void
+add_missing_parent_ptr(
+ struct xfs_inode *ip,
+ struct file_scan *fscan,
+ const struct ag_pptr *ag_pptr)
+{
+ unsigned char name[MAXNAMELEN];
+ int error;
+
+ error = strblobs_load(nameblobs, ag_pptr->name_cookie, name,
+ ag_pptr->namelen);
+ if (error)
+ do_error(
+ _("loading missing name for ino %llu parent pointer (ino %llu gen 0x%x namecookie 0x%llx) failed: %s\n"),
+ (unsigned long long)ip->i_ino,
+ (unsigned long long)ag_pptr->parent_ino,
+ ag_pptr->parent_gen,
+ (unsigned long long)ag_pptr->name_cookie,
+ strerror(error));
+
+ if (no_modify) {
+ do_warn(
+ _("would add missing ino %llu parent pointer (ino %llu gen 0x%x name '%.*s')\n"),
+ (unsigned long long)ip->i_ino,
+ (unsigned long long)ag_pptr->parent_ino,
+ ag_pptr->parent_gen,
+ ag_pptr->namelen,
+ name);
+ return;
+ } else {
+ do_warn(
+ _("adding missing ino %llu parent pointer (ino %llu gen 0x%x name '%.*s')\n"),
+ (unsigned long long)ip->i_ino,
+ (unsigned long long)ag_pptr->parent_ino,
+ ag_pptr->parent_gen,
+ ag_pptr->namelen,
+ name);
+ }
+
+ /* XXX actually do the work */
+}
+
+/* Remove @file_pptr from @ip. */
+static void
+remove_incorrect_parent_ptr(
+ struct xfs_inode *ip,
+ struct file_scan *fscan,
+ const struct file_pptr *file_pptr)
+{
+ unsigned char name[MAXNAMELEN] = { };
+ int error;
+
+ error = load_file_pptr_name(fscan, file_pptr, name);
+ if (error)
+ do_error(
+ _("loading incorrect name for ino %llu parent pointer (ino %llu gen 0x%x namecookie 0x%llx) failed: %s\n"),
+ (unsigned long long)ip->i_ino,
+ (unsigned long long)file_pptr->parent_ino,
+ file_pptr->parent_gen,
+ (unsigned long long)file_pptr->name_cookie,
+ strerror(error));
+
+ if (no_modify) {
+ do_warn(
+ _("would remove bad ino %llu parent pointer (ino %llu gen 0x%x name '%.*s')\n"),
+ (unsigned long long)ip->i_ino,
+ (unsigned long long)file_pptr->parent_ino,
+ file_pptr->parent_gen,
+ file_pptr->namelen,
+ name);
+ return;
+ }
+
+ do_warn(
+ _("removing bad ino %llu parent pointer (ino %llu gen 0x%x name '%.*s')\n"),
+ (unsigned long long)ip->i_ino,
+ (unsigned long long)file_pptr->parent_ino,
+ file_pptr->parent_gen,
+ file_pptr->namelen,
+ name);
+
+ /* XXX actually do the work */
+}
+
+/*
+ * We found parent pointers that point to the same inode and directory offset.
+ * Make sure they have the same generation number and dirent name.
+ */
+static void
+compare_parent_ptrs(
+ struct xfs_inode *ip,
+ struct file_scan *fscan,
+ const struct ag_pptr *ag_pptr,
+ const struct file_pptr *file_pptr)
+{
+ unsigned char name1[MAXNAMELEN] = { };
+ unsigned char name2[MAXNAMELEN] = { };
+ int error;
+
+ error = strblobs_load(nameblobs, ag_pptr->name_cookie, name1,
+ ag_pptr->namelen);
+ if (error)
+ do_error(
+ _("loading master-list name for ino %llu parent pointer (ino %llu gen 0x%x namecookie 0x%llx namelen %u) failed: %s\n"),
+ (unsigned long long)ip->i_ino,
+ (unsigned long long)ag_pptr->parent_ino,
+ ag_pptr->parent_gen,
+ (unsigned long long)ag_pptr->name_cookie,
+ ag_pptr->namelen,
+ strerror(error));
+
+ error = load_file_pptr_name(fscan, file_pptr, name2);
+ if (error)
+ do_error(
+ _("loading file-list name for ino %llu parent pointer (ino %llu gen 0x%x namecookie 0x%llx namelen %u) failed: %s\n"),
+ (unsigned long long)ip->i_ino,
+ (unsigned long long)file_pptr->parent_ino,
+ file_pptr->parent_gen,
+ (unsigned long long)file_pptr->name_cookie,
+ ag_pptr->namelen,
+ strerror(error));
+
+ if (ag_pptr->parent_gen != file_pptr->parent_gen)
+ goto reset;
+ if (ag_pptr->namelen != file_pptr->namelen)
+ goto reset;
+ if (ag_pptr->namehash != file_pptr->namehash)
+ goto reset;
+ if (memcmp(name1, name2, ag_pptr->namelen))
+ goto reset;
+
+ return;
+
+reset:
+ if (no_modify) {
+ do_warn(
+ _("would update ino %llu parent pointer (ino %llu gen 0x%x name '%.*s')\n"),
+ (unsigned long long)ip->i_ino,
+ (unsigned long long)ag_pptr->parent_ino,
+ ag_pptr->parent_gen,
+ ag_pptr->namelen,
+ name1);
+ return;
+ }
+
+ do_warn(
+ _("updating ino %llu parent pointer (ino %llu gen 0x%x name '%.*s')\n"),
+ (unsigned long long)ip->i_ino,
+ (unsigned long long)ag_pptr->parent_ino,
+ ag_pptr->parent_gen,
+ ag_pptr->namelen,
+ name1);
+
+ /* XXX do the work */
+}
+
+static int
+cmp_file_to_ag_pptr(
+ const struct file_pptr *fp,
+ const struct ag_pptr *ap)
+{
+ /*
+ * We finished iterating all the pptrs attached to the file before we
+ * ran out of pptrs that we found in the directory scan. Return 1 so
+ * the caller adds the pptr from the dir scan.
+ */
+ if (!fp)
+ return 1;
+
+ if (fp->parent_ino > ap->parent_ino)
+ return 1;
+ if (fp->parent_ino < ap->parent_ino)
+ return -1;
+
+ if (fp->namehash < ap->namehash)
+ return -1;
+ if (fp->namehash > ap->namehash)
+ return 1;
+
+ /*
+ * If this parent pointer wasn't found in the dirent scan, we know it
+ * should be removed.
+ */
+ if (!fp->name_in_nameblobs)
+ return -1;
+
+ if (fp->name_cookie < ap->name_cookie)
+ return -1;
+ if (fp->name_cookie > ap->name_cookie)
+ return 1;
+
+ return 0;
+}
+
+/*
+ * Make sure that the parent pointers we observed match the ones ondisk.
+ *
+ * Earlier, we generated a master list of parent pointers for files in this AG
+ * based on what we saw during the directory walk at the start of phase 6.
+ * Now that we've read in all of this file's parent pointers, make sure the
+ * lists match.
+ */
+static void
+crosscheck_file_parent_ptrs(
+ struct xfs_inode *ip,
+ struct file_scan *fscan)
+{
+ struct ag_pptr *ag_pptr;
+ struct file_pptr *file_pptr;
+ struct xfs_mount *mp = ip->i_mount;
+ xfs_agnumber_t agno = XFS_INO_TO_AGNO(mp, ip->i_ino);
+ xfs_agino_t agino = XFS_INO_TO_AGINO(mp, ip->i_ino);
+ int error;
+
+ ag_pptr = peek_slab_cursor(fscan->ag_pptr_recs_cur);
+
+ if (!ag_pptr || ag_pptr->child_agino > agino) {
+ /*
+ * The cursor for the master pptr list has gone beyond this
+ * file that we're scanning. Evidently it has no parents at
+ * all, so we better not have found any pptrs attached to the
+ * file.
+ */
+ if (fscan->nr_file_pptrs > 0)
+ clear_all_pptrs(ip);
+
+ return;
+ }
+
+ if (ag_pptr->child_agino < agino) {
+ /*
+ * The cursor for the master pptr list is behind the file that
+ * we're scanning. This suggests that the incore inode tree
+ * doesn't know about a file that is mentioned by a dirent.
+ * At this point the inode liveness is supposed to be settled,
+ * which means our incore information is inconsistent.
+ */
+ do_error(
+ _("found dirent referring to ino %llu even though inobt scan moved on to ino %llu?!\n"),
+ (unsigned long long)XFS_AGINO_TO_INO(mp, agno,
+ ag_pptr->child_agino),
+ (unsigned long long)ip->i_ino);
+ /* does not return */
+ }
+
+ /*
+ * The master pptr list cursor is pointing to the inode that we want
+ * to check. Sort the pptr records that we recorded from the ondisk
+ * pptrs for this file, then set up for the comparison.
+ */
+ qsort_slab(fscan->file_pptr_recs, cmp_file_pptr);
+
+ error = -init_slab_cursor(fscan->file_pptr_recs, cmp_file_pptr,
+ &fscan->file_pptr_recs_cur);
+ if (error)
+ do_error(_("init ino %llu parent pointer cursor failed: %s\n"),
+ (unsigned long long)ip->i_ino, strerror(error));
+
+ do {
+ int cmp_result;
+
+ file_pptr = peek_slab_cursor(fscan->file_pptr_recs_cur);
+
+ dbg_printf(
+ _("%s: dp %llu dp_gen 0x%x namelen %u ino %llu namecookie 0x%llx (master)\n"),
+ __func__,
+ (unsigned long long)ag_pptr->parent_ino,
+ ag_pptr->parent_gen,
+ ag_pptr->namelen,
+ (unsigned long long)ip->i_ino,
+ (unsigned long long)ag_pptr->name_cookie);
+
+ if (file_pptr) {
+ dbg_printf(
+ _("%s: dp %llu dp_gen 0x%x namelen %u ino %llu namecookie 0x%llx (file)\n"),
+ __func__,
+ (unsigned long long)file_pptr->parent_ino,
+ file_pptr->parent_gen,
+ file_pptr->namelen,
+ (unsigned long long)ip->i_ino,
+ (unsigned long long)file_pptr->name_cookie);
+ } else {
+ dbg_printf(
+ _("%s: ran out of parent pointers for ino %llu (file)\n"),
+ __func__,
+ (unsigned long long)ip->i_ino);
+ }
+
+ cmp_result = cmp_file_to_ag_pptr(file_pptr, ag_pptr);
+ if (cmp_result > 0) {
+ /*
+ * The master pptr list knows about pptrs that are not
+ * in the ondisk metadata. Add the missing pptr and
+ * advance only the master pptr cursor.
+ */
+ add_missing_parent_ptr(ip, fscan, ag_pptr);
+ advance_slab_cursor(fscan->ag_pptr_recs_cur);
+ } else if (cmp_result < 0) {
+ /*
+ * The ondisk pptrs mention a link that is not in the
+ * master list. Delete the extra pptr and advance only
+ * the file pptr cursor.
+ */
+ remove_incorrect_parent_ptr(ip, fscan, file_pptr);
+ advance_slab_cursor(fscan->file_pptr_recs_cur);
+ } else {
+ /*
+ * Exact match, make sure the parent_gen and dirent
+ * name parts of the parent pointer match. Move both
+ * cursors forward.
+ */
+ compare_parent_ptrs(ip, fscan, ag_pptr, file_pptr);
+ advance_slab_cursor(fscan->ag_pptr_recs_cur);
+ advance_slab_cursor(fscan->file_pptr_recs_cur);
+ }
+
+ ag_pptr = peek_slab_cursor(fscan->ag_pptr_recs_cur);
+ } while (ag_pptr && ag_pptr->child_agino == agino);
+
+ while ((file_pptr = pop_slab_cursor(fscan->file_pptr_recs_cur))) {
+ dbg_printf(
+ _("%s: dp %llu dp_gen 0x%x namelen %u ino %llu namecookie 0x%llx (excess)\n"),
+ __func__,
+ (unsigned long long)file_pptr->parent_ino,
+ file_pptr->parent_gen,
+ file_pptr->namelen,
+ (unsigned long long)ip->i_ino,
+ (unsigned long long)file_pptr->name_cookie);
+
+ /*
+ * The master pptr list does not have any more pptrs for this
+ * file, but we still have unprocessed ondisk pptrs. Delete
+ * all these ondisk pptrs.
+ */
+ remove_incorrect_parent_ptr(ip, fscan, file_pptr);
+ }
+}
+
+/* Ensure this file's parent pointers match what we found in the dirent scan. */
+static void
+check_file_parent_ptrs(
+ struct xfs_inode *ip,
+ struct file_scan *fscan)
+{
+ int error;
+
+ error = -init_slab(&fscan->file_pptr_recs, sizeof(struct file_pptr));
+ if (error)
+ do_error(_("init file parent pointer recs failed: %s\n"),
+ strerror(error));
+
+ fscan->have_garbage = false;
+ fscan->nr_file_pptrs = 0;
+
+ error = xattr_walk(ip, examine_xattr, fscan);
+ if (error && !no_modify)
+ do_error(_("ino %llu parent pointer scan failed: %s\n"),
+ (unsigned long long)ip->i_ino,
+ strerror(error));
+ if (error) {
+ do_warn(_("ino %llu parent pointer scan failed: %s\n"),
+ (unsigned long long)ip->i_ino,
+ strerror(error));
+ goto out_free;
+ }
+
+ crosscheck_file_parent_ptrs(ip, fscan);
+
+out_free:
+ free_slab(&fscan->file_pptr_recs);
+ xfblob_truncate(fscan->file_pptr_names);
+}
+
+/* Check all the parent pointers of files in this AG. */
+static void
+check_ag_parent_ptrs(
+ struct workqueue *wq,
+ uint32_t agno,
+ void *arg)
+{
+ struct xfs_mount *mp = wq->wq_ctx;
+ struct file_scan fscan = {
+ .ag_pptrs = &fs_pptrs[agno],
+ };
+ struct ag_pptrs *ag_pptrs = &fs_pptrs[agno];
+ struct ino_tree_node *irec;
+ char *descr;
+ int error;
+
+ qsort_slab(ag_pptrs->pptr_recs, cmp_ag_pptr);
+
+ error = -init_slab_cursor(ag_pptrs->pptr_recs, cmp_ag_pptr,
+ &fscan.ag_pptr_recs_cur);
+ if (error)
+ do_error(
+ _("init agno %u parent pointer slab cursor failed: %s\n"),
+ agno, strerror(error));
+
+ descr = kasprintf("xfs_repair (%s): file parent pointer names",
+ mp->m_fsname);
+ error = -xfblob_create(descr, &fscan.file_pptr_names);
+ kfree(descr);
+ if (error)
+ do_error(
+ _("init agno %u file parent pointer names failed: %s\n"),
+ agno, strerror(error));
+
+ for (irec = findfirst_inode_rec(agno);
+ irec != NULL;
+ irec = next_ino_rec(irec)) {
+ unsigned int ino_offset;
+
+ for (ino_offset = 0;
+ ino_offset < XFS_INODES_PER_CHUNK;
+ ino_offset++) {
+ struct xfs_inode *ip;
+ xfs_ino_t ino;
+
+ if (is_inode_free(irec, ino_offset))
+ continue;
+
+ ino = XFS_AGINO_TO_INO(mp, agno,
+ irec->ino_startnum + ino_offset);
+ error = -libxfs_iget(mp, NULL, ino, 0, &ip);
+ if (error && !no_modify)
+ do_error(
+ _("loading ino %llu for parent pointer check failed: %s\n"),
+ (unsigned long long)ino,
+ strerror(error));
+ if (error) {
+ do_warn(
+ _("loading ino %llu for parent pointer check failed: %s\n"),
+ (unsigned long long)ino,
+ strerror(error));
+ continue;
+ }
+
+ check_file_parent_ptrs(ip, &fscan);
+ libxfs_irele(ip);
+ }
+ }
+
+ xfblob_destroy(fscan.file_pptr_names);
+ free_slab_cursor(&fscan.ag_pptr_recs_cur);
+}
+
+/* Check all the parent pointers of all files in this filesystem. */
+void
+check_parent_ptrs(
+ struct xfs_mount *mp)
+{
+ struct workqueue wq;
+ xfs_agnumber_t agno;
+
+ if (!xfs_has_parent(mp))
+ return;
+
+ create_work_queue(&wq, mp, ag_stride);
+
+ for (agno = 0; agno < mp->m_sb.sb_agcount; agno++)
+ queue_work(&wq, check_ag_parent_ptrs, agno, NULL);
+
+ destroy_work_queue(&wq);
+}
diff --git a/repair/pptr.h b/repair/pptr.h
index ef85807b117..f5ffcc137e3 100644
--- a/repair/pptr.h
+++ b/repair/pptr.h
@@ -12,4 +12,6 @@ void parent_ptr_init(struct xfs_mount *mp);
void add_parent_ptr(xfs_ino_t ino, const unsigned char *fname,
struct xfs_inode *dp);
+void check_parent_ptrs(struct xfs_mount *mp);
+
#endif /* __REPAIR_PPTR_H__ */
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 14/18] xfs_repair: dump garbage parent pointer attributes
2023-12-31 19:51 ` [PATCHSET v13.0 4/6] xfsprogs: fsck for " Darrick J. Wong
` (12 preceding siblings ...)
2023-12-31 23:16 ` [PATCH 13/18] xfs_repair: check parent pointers Darrick J. Wong
@ 2023-12-31 23:17 ` Darrick J. Wong
2023-12-31 23:17 ` [PATCH 15/18] xfs_repair: update ondisk parent pointer records Darrick J. Wong
` (3 subsequent siblings)
17 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:17 UTC (permalink / raw)
To: djwong, cem; +Cc: catherine.hoang, linux-xfs, allison.henderson
From: Darrick J. Wong <djwong@kernel.org>
Delete xattrs that have ATTR_PARENT set but are so garbage that they
clearly aren't parent pointers.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
repair/pptr.c | 145 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++-
1 file changed, 143 insertions(+), 2 deletions(-)
diff --git a/repair/pptr.c b/repair/pptr.c
index 11aa8d4e322..21b15ab80ea 100644
--- a/repair/pptr.c
+++ b/repair/pptr.c
@@ -192,6 +192,29 @@ struct file_scan {
/* Does this file have garbage xattrs with ATTR_PARENT set? */
bool have_garbage;
+
+ /* xattrs that we have to remove from this file */
+ struct xfs_slab *garbage_xattr_recs;
+
+ /* attr names associated with garbage_xattr_recs */
+ struct xfblob *garbage_xattr_names;
+};
+
+struct garbage_xattr {
+ /* xfs_da_args.attr_filter for the attribute being removed */
+ unsigned int attr_filter;
+
+ /* attribute name length */
+ unsigned int attrnamelen;
+
+ /* attribute value length */
+ unsigned int attrvaluelen;
+
+ /* cookie for the attribute name */
+ xfblob_cookie attrname_cookie;
+
+ /* cookie for the attribute value */
+ xfblob_cookie attrvalue_cookie;
};
/* Global names storage file. */
@@ -381,6 +404,78 @@ add_parent_ptr(
(unsigned long long)ag_pptr.name_cookie);
}
+/* Remove garbage extended attributes that have ATTR_PARENT set. */
+static void
+remove_garbage_xattrs(
+ struct xfs_inode *ip,
+ struct file_scan *fscan)
+{
+ struct xfs_slab_cursor *cur;
+ struct garbage_xattr *ga;
+ void *buf = NULL;
+ size_t bufsize = 0;
+ int error;
+
+ error = -init_slab_cursor(fscan->garbage_xattr_recs, NULL, &cur);
+ if (error)
+ do_error(_("init garbage xattr cursor failed: %s\n"),
+ strerror(error));
+
+ while ((ga = pop_slab_cursor(cur)) != NULL) {
+ struct xfs_da_args args = {
+ .dp = ip,
+ .attr_filter = ga->attr_filter,
+ .namelen = ga->attrnamelen,
+ .valuelen = ga->attrvaluelen,
+ .op_flags = XFS_DA_OP_REMOVE | XFS_DA_OP_NVLOOKUP,
+ };
+ size_t desired = ga->attrnamelen + ga->attrvaluelen;
+
+ if (desired > bufsize) {
+ free(buf);
+ buf = malloc(desired);
+ if (!buf)
+ do_error(
+ _("allocating %zu bytes to remove ino %llu garbage xattr failed: %s\n"),
+ desired,
+ (unsigned long long)ip->i_ino,
+ strerror(errno));
+ bufsize = desired;
+ }
+
+ args.name = buf;
+ args.value = buf + ga->attrnamelen;
+
+ error = -xfblob_load(fscan->garbage_xattr_names,
+ ga->attrname_cookie, buf, ga->attrnamelen);
+ if (error)
+ do_error(
+ _("loading garbage xattr name failed: %s\n"),
+ strerror(error));
+
+ error = -xfblob_load(fscan->garbage_xattr_names,
+ ga->attrvalue_cookie, args.value,
+ ga->attrvaluelen);
+ if (error)
+ do_error(
+ _("loading garbage xattr value failed: %s\n"),
+ strerror(error));
+
+ error = -libxfs_attr_set(&args);
+ if (error)
+ do_error(
+ _("removing ino %llu garbage xattr failed: %s\n"),
+ (unsigned long long)ip->i_ino,
+ strerror(error));
+ }
+
+ free(buf);
+ free_slab_cursor(&cur);
+ free_slab(&fscan->garbage_xattr_recs);
+ xfblob_destroy(fscan->garbage_xattr_names);
+ fscan->garbage_xattr_names = NULL;
+}
+
/* Schedule this ATTR_PARENT extended attribute for deletion. */
static void
record_garbage_xattr(
@@ -392,6 +487,15 @@ record_garbage_xattr(
const void *value,
unsigned int valuelen)
{
+ struct garbage_xattr garbage_xattr = {
+ .attr_filter = attr_filter,
+ .attrnamelen = namelen,
+ .attrvaluelen = valuelen,
+ };
+ struct xfs_mount *mp = ip->i_mount;
+ char *descr;
+ int error;
+
if (no_modify) {
if (!fscan->have_garbage)
do_warn(
@@ -402,13 +506,47 @@ record_garbage_xattr(
}
if (fscan->have_garbage)
- return;
+ goto stuffit;
fscan->have_garbage = true;
do_warn(
_("deleting garbage parent pointer extended attributes in ino %llu\n"),
(unsigned long long)ip->i_ino);
- /* XXX do the work */
+
+ error = -init_slab(&fscan->garbage_xattr_recs,
+ sizeof(struct garbage_xattr));
+ if (error)
+ do_error(_("init garbage xattr recs failed: %s\n"),
+ strerror(error));
+
+ descr = kasprintf("xfs_repair (%s): garbage xattr names",
+ mp->m_fsname);
+ error = -xfblob_create(descr, &fscan->garbage_xattr_names);
+ kfree(descr);
+ if (error)
+ do_error("init garbage xattr names failed: %s\n",
+ strerror(error));
+
+stuffit:
+ error = -xfblob_store(fscan->garbage_xattr_names,
+ &garbage_xattr.attrname_cookie, name, namelen);
+ if (error)
+ do_error(_("storing ino %llu garbage xattr failed: %s\n"),
+ (unsigned long long)ip->i_ino,
+ strerror(error));
+
+ error = -xfblob_store(fscan->garbage_xattr_names,
+ &garbage_xattr.attrvalue_cookie, value, valuelen);
+ if (error)
+ do_error(_("storing ino %llu garbage xattr failed: %s\n"),
+ (unsigned long long)ip->i_ino,
+ strerror(error));
+
+ error = -slab_add(fscan->garbage_xattr_recs, &garbage_xattr);
+ if (error)
+ do_error(_("storing ino %llu garbage xattr rec failed: %s\n"),
+ (unsigned long long)ip->i_ino,
+ strerror(error));
}
/*
@@ -931,6 +1069,9 @@ check_file_parent_ptrs(
goto out_free;
}
+ if (!no_modify && fscan->have_garbage)
+ remove_garbage_xattrs(ip, fscan);
+
crosscheck_file_parent_ptrs(ip, fscan);
out_free:
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 15/18] xfs_repair: update ondisk parent pointer records
2023-12-31 19:51 ` [PATCHSET v13.0 4/6] xfsprogs: fsck for " Darrick J. Wong
` (13 preceding siblings ...)
2023-12-31 23:17 ` [PATCH 14/18] xfs_repair: dump garbage parent pointer attributes Darrick J. Wong
@ 2023-12-31 23:17 ` Darrick J. Wong
2023-12-31 23:17 ` [PATCH 16/18] xfs_repair: wipe ondisk parent pointers when there are none Darrick J. Wong
` (2 subsequent siblings)
17 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:17 UTC (permalink / raw)
To: djwong, cem; +Cc: catherine.hoang, linux-xfs, allison.henderson
From: Darrick J. Wong <djwong@kernel.org>
Update the ondisk parent pointer records as necessary.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
libxfs/libxfs_api_defs.h | 2 +
repair/pptr.c | 89 ++++++++++++++++++++++++++++++++++++++++++++--
2 files changed, 88 insertions(+), 3 deletions(-)
diff --git a/libxfs/libxfs_api_defs.h b/libxfs/libxfs_api_defs.h
index 1aa2d9f0679..d5662c42f2b 100644
--- a/libxfs/libxfs_api_defs.h
+++ b/libxfs/libxfs_api_defs.h
@@ -187,9 +187,11 @@
#define xfs_parent_irec_from_disk libxfs_parent_irec_from_disk
#define xfs_parent_irec_hashname libxfs_parent_irec_hashname
#define xfs_parent_lookup libxfs_parent_lookup
+#define xfs_parent_set libxfs_parent_set
#define xfs_parent_start libxfs_parent_start
#define xfs_parent_hashcheck libxfs_parent_hashcheck
#define xfs_parent_namecheck libxfs_parent_namecheck
+#define xfs_parent_unset libxfs_parent_unset
#define xfs_parent_valuecheck libxfs_parent_valuecheck
#define xfs_parent_verify_irec libxfs_parent_verify_irec
#define xfs_perag_get libxfs_perag_get
diff --git a/repair/pptr.c b/repair/pptr.c
index 21b15ab80ea..7f65ae1aac3 100644
--- a/repair/pptr.c
+++ b/repair/pptr.c
@@ -686,6 +686,48 @@ load_file_pptr_name(
name, file_pptr->namelen);
}
+/* Add an on disk parent pointer to a file. */
+static int
+add_file_pptr(
+ struct xfs_inode *ip,
+ const struct ag_pptr *ag_pptr,
+ const unsigned char *name)
+{
+ struct xfs_parent_name_irec pptr_rec = {
+ .p_ino = ag_pptr->parent_ino,
+ .p_gen = ag_pptr->parent_gen,
+ .p_namelen = ag_pptr->namelen,
+ };
+ struct xfs_parent_scratch scratch;
+ int error;
+
+ memcpy(pptr_rec.p_name, name, ag_pptr->namelen);
+ libxfs_parent_irec_hashname(ip->i_mount, &pptr_rec);
+ error = -libxfs_parent_lookup(NULL, ip, &pptr_rec, &scratch);
+ if (!error || error != ENOATTR)
+ return error;
+ return -libxfs_parent_set(ip, ip->i_ino, &pptr_rec, &scratch);
+}
+
+/* Remove an on disk parent pointer from a file. */
+static int
+remove_file_pptr(
+ struct xfs_inode *ip,
+ const struct file_pptr *file_pptr,
+ const unsigned char *name)
+{
+ struct xfs_parent_name_irec pptr_rec = {
+ .p_ino = file_pptr->parent_ino,
+ .p_gen = file_pptr->parent_gen,
+ .p_namelen = file_pptr->namelen,
+ };
+ struct xfs_parent_scratch scratch;
+
+ memcpy(pptr_rec.p_name, name, file_pptr->namelen);
+ libxfs_parent_irec_hashname(ip->i_mount, &pptr_rec);
+ return -libxfs_parent_unset(ip, ip->i_ino, &pptr_rec, &scratch);
+}
+
/* Remove all pptrs from @ip. */
static void
clear_all_pptrs(
@@ -742,7 +784,16 @@ add_missing_parent_ptr(
name);
}
- /* XXX actually do the work */
+ error = add_file_pptr(ip, ag_pptr, name);
+ if (error)
+ do_error(
+ _("adding ino %llu pptr (ino %llu gen 0x%x name '%.*s') failed: %s\n"),
+ (unsigned long long)ip->i_ino,
+ (unsigned long long)ag_pptr->parent_ino,
+ ag_pptr->parent_gen,
+ ag_pptr->namelen,
+ name,
+ strerror(error));
}
/* Remove @file_pptr from @ip. */
@@ -784,7 +835,16 @@ remove_incorrect_parent_ptr(
file_pptr->namelen,
name);
- /* XXX actually do the work */
+ error = remove_file_pptr(ip, file_pptr, name);
+ if (error)
+ do_error(
+ _("removing ino %llu pptr (ino %llu gen 0x%x name '%.*s') failed: %s\n"),
+ (unsigned long long)ip->i_ino,
+ (unsigned long long)file_pptr->parent_ino,
+ file_pptr->parent_gen,
+ file_pptr->namelen,
+ name,
+ strerror(error));
}
/*
@@ -856,7 +916,30 @@ compare_parent_ptrs(
ag_pptr->namelen,
name1);
- /* XXX do the work */
+ if (ag_pptr->parent_gen != file_pptr->parent_gen ||
+ ag_pptr->namehash != file_pptr->namehash) {
+ error = remove_file_pptr(ip, file_pptr, name2);
+ if (error)
+ do_error(
+ _("erasing ino %llu pptr (ino %llu gen 0x%x name '%.*s') failed: %s\n"),
+ (unsigned long long)ip->i_ino,
+ (unsigned long long)file_pptr->parent_ino,
+ file_pptr->parent_gen,
+ file_pptr->namelen,
+ name2,
+ strerror(error));
+ }
+
+ error = add_file_pptr(ip, ag_pptr, name1);
+ if (error)
+ do_error(
+ _("updating ino %llu pptr (ino %llu gen 0x%x name '%.*s') failed: %s\n"),
+ (unsigned long long)ip->i_ino,
+ (unsigned long long)ag_pptr->parent_ino,
+ ag_pptr->parent_gen,
+ ag_pptr->namelen,
+ name1,
+ strerror(error));
}
static int
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 16/18] xfs_repair: wipe ondisk parent pointers when there are none
2023-12-31 19:51 ` [PATCHSET v13.0 4/6] xfsprogs: fsck for " Darrick J. Wong
` (14 preceding siblings ...)
2023-12-31 23:17 ` [PATCH 15/18] xfs_repair: update ondisk parent pointer records Darrick J. Wong
@ 2023-12-31 23:17 ` Darrick J. Wong
2023-12-31 23:17 ` [PATCH 17/18] xfs_repair: upgrade an existing filesystem to have parent pointers Darrick J. Wong
2023-12-31 23:18 ` [PATCH 18/18] xfs_scrub: use parent pointers to report lost file data Darrick J. Wong
17 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:17 UTC (permalink / raw)
To: djwong, cem; +Cc: catherine.hoang, linux-xfs, allison.henderson
From: Darrick J. Wong <djwong@kernel.org>
Erase all the parent pointers when there aren't any found by the
directory entry scan.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
repair/pptr.c | 41 ++++++++++++++++++++++++++++++++++++++---
1 file changed, 38 insertions(+), 3 deletions(-)
diff --git a/repair/pptr.c b/repair/pptr.c
index 7f65ae1aac3..46a214f205e 100644
--- a/repair/pptr.c
+++ b/repair/pptr.c
@@ -731,8 +731,13 @@ remove_file_pptr(
/* Remove all pptrs from @ip. */
static void
clear_all_pptrs(
- struct xfs_inode *ip)
+ struct xfs_inode *ip,
+ struct file_scan *fscan)
{
+ struct xfs_slab_cursor *cur;
+ struct file_pptr *file_pptr;
+ int error;
+
if (no_modify) {
do_warn(_("would delete unlinked ino %llu parent pointers\n"),
(unsigned long long)ip->i_ino);
@@ -741,7 +746,37 @@ clear_all_pptrs(
do_warn(_("deleting unlinked ino %llu parent pointers\n"),
(unsigned long long)ip->i_ino);
- /* XXX actually do the work */
+
+ error = -init_slab_cursor(fscan->file_pptr_recs, NULL, &cur);
+ if (error)
+ do_error(_("init ino %llu pptr cursor failed: %s\n"),
+ (unsigned long long)ip->i_ino,
+ strerror(error));
+
+ while ((file_pptr = pop_slab_cursor(cur)) != NULL) {
+ unsigned char name[MAXNAMELEN];
+
+ error = load_file_pptr_name(fscan, file_pptr, name);
+ if (error)
+ do_error(
+ _("loading incorrect name for ino %llu parent pointer (ino %llu gen 0x%x namecookie 0x%llx) failed: %s\n"),
+ (unsigned long long)ip->i_ino,
+ (unsigned long long)file_pptr->parent_ino,
+ file_pptr->parent_gen,
+ (unsigned long long)file_pptr->name_cookie,
+ strerror(error));
+
+ error = remove_file_pptr(ip, file_pptr, name);
+ if (error)
+ do_error(
+ _("wiping ino %llu pptr (ino %llu gen 0x%x) failed: %s\n"),
+ (unsigned long long)ip->i_ino,
+ (unsigned long long)file_pptr->parent_ino,
+ file_pptr->parent_gen,
+ strerror(error));
+ }
+
+ free_slab_cursor(&cur);
}
/* Add @ag_pptr to @ip. */
@@ -1010,7 +1045,7 @@ crosscheck_file_parent_ptrs(
* file.
*/
if (fscan->nr_file_pptrs > 0)
- clear_all_pptrs(ip);
+ clear_all_pptrs(ip, fscan);
return;
}
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 17/18] xfs_repair: upgrade an existing filesystem to have parent pointers
2023-12-31 19:51 ` [PATCHSET v13.0 4/6] xfsprogs: fsck for " Darrick J. Wong
` (15 preceding siblings ...)
2023-12-31 23:17 ` [PATCH 16/18] xfs_repair: wipe ondisk parent pointers when there are none Darrick J. Wong
@ 2023-12-31 23:17 ` Darrick J. Wong
2023-12-31 23:18 ` [PATCH 18/18] xfs_scrub: use parent pointers to report lost file data Darrick J. Wong
17 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:17 UTC (permalink / raw)
To: djwong, cem; +Cc: catherine.hoang, linux-xfs, allison.henderson
From: Darrick J. Wong <djwong@kernel.org>
Upgrade an existing filesystem to have parent pointers.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
man/man8/xfs_admin.8 | 8 ++++++++
repair/globals.c | 1 +
repair/globals.h | 1 +
repair/phase2.c | 39 +++++++++++++++++++++++++++++++++++++++
repair/pptr.c | 15 ++++++++++++++-
repair/xfs_repair.c | 11 +++++++++++
6 files changed, 74 insertions(+), 1 deletion(-)
diff --git a/man/man8/xfs_admin.8 b/man/man8/xfs_admin.8
index 467fb2dfd0a..28f28b6dd8f 100644
--- a/man/man8/xfs_admin.8
+++ b/man/man8/xfs_admin.8
@@ -177,6 +177,14 @@ and online repairs to space usage metadata.
The filesystem cannot be downgraded after this feature is enabled.
This upgrade can fail if any AG has less than 5% free space remaining.
This feature was added to Linux 4.8.
+.TP 0.4i
+.B parent
+Store in each child file a mirror a pointing back to the parent directory.
+This enables much stronger cross-referencing and online repairs of the
+directory tree.
+The filesystem cannot be downgraded after this feature is enabled.
+This upgrade can fail if the filesystem has less than 25% free space remaining.
+This feature is not upstream yet.
.RE
.TP
.BI \-U " uuid"
diff --git a/repair/globals.c b/repair/globals.c
index cff620e8f0e..7d95e210e8e 100644
--- a/repair/globals.c
+++ b/repair/globals.c
@@ -55,6 +55,7 @@ bool add_nrext64;
bool add_finobt; /* add free inode btrees */
bool add_reflink; /* add reference count btrees */
bool add_rmapbt; /* add reverse mapping btrees */
+bool add_parent; /* add parent pointers */
/* misc status variables */
diff --git a/repair/globals.h b/repair/globals.h
index 76d22fd3b2c..71a64b94365 100644
--- a/repair/globals.h
+++ b/repair/globals.h
@@ -96,6 +96,7 @@ extern bool add_nrext64;
extern bool add_finobt; /* add free inode btrees */
extern bool add_reflink; /* add reference count btrees */
extern bool add_rmapbt; /* add reverse mapping btrees */
+extern bool add_parent; /* add parent pointers */
/* misc status variables */
diff --git a/repair/phase2.c b/repair/phase2.c
index be0d791a8b5..a58fa7d8a7b 100644
--- a/repair/phase2.c
+++ b/repair/phase2.c
@@ -265,6 +265,28 @@ set_rmapbt(
return true;
}
+static bool
+set_parent(
+ struct xfs_mount *mp,
+ struct xfs_sb *new_sb)
+{
+ if (xfs_has_parent(mp)) {
+ printf(_("Filesystem already supports parent pointers.\n"));
+ exit(0);
+ }
+
+ if (!xfs_has_crc(mp)) {
+ printf(
+ _("Parent pointer feature only supported on V5 filesystems.\n"));
+ exit(0);
+ }
+
+ printf(_("Adding parent pointers to filesystem.\n"));
+ new_sb->sb_features_incompat |= XFS_SB_FEAT_INCOMPAT_PARENT;
+ new_sb->sb_features_incompat |= XFS_SB_FEAT_INCOMPAT_NEEDSREPAIR;
+ return true;
+}
+
struct check_state {
struct xfs_sb sb;
uint64_t features;
@@ -406,6 +428,19 @@ check_fs_free_space(
libxfs_trans_cancel(tp);
}
+ /*
+ * If we're adding parent pointers, we need at least 25% free since
+ * scanning the entire filesystem to guesstimate the overhead is
+ * prohibitively expensive.
+ */
+ if (xfs_has_parent(mp) && !(old->features & XFS_FEAT_PARENT)) {
+ if (mp->m_sb.sb_fdblocks < mp->m_sb.sb_dblocks / 4) {
+ printf(
+ _("Filesystem does not have enough space to add parent pointers.\n"));
+ exit(1);
+ }
+ }
+
/*
* Would the post-upgrade filesystem have enough free space on the data
* device after making per-AG reservations?
@@ -438,6 +473,8 @@ need_check_fs_free_space(
return true;
if (xfs_has_rmapbt(mp) && !(old->features & XFS_FEAT_RMAPBT))
return true;
+ if (xfs_has_parent(mp) && !(old->features & XFS_FEAT_PARENT))
+ return true;
return false;
}
@@ -519,6 +556,8 @@ upgrade_filesystem(
dirty |= set_reflink(mp, &new_sb);
if (add_rmapbt)
dirty |= set_rmapbt(mp, &new_sb);
+ if (add_parent)
+ dirty |= set_parent(mp, &new_sb);
if (!dirty)
return;
diff --git a/repair/pptr.c b/repair/pptr.c
index 46a214f205e..77f49dbcb84 100644
--- a/repair/pptr.c
+++ b/repair/pptr.c
@@ -809,7 +809,7 @@ add_missing_parent_ptr(
ag_pptr->namelen,
name);
return;
- } else {
+ } else if (!add_parent) {
do_warn(
_("adding missing ino %llu parent pointer (ino %llu gen 0x%x name '%.*s')\n"),
(unsigned long long)ip->i_ino,
@@ -817,6 +817,19 @@ add_missing_parent_ptr(
ag_pptr->parent_gen,
ag_pptr->namelen,
name);
+ } else {
+ static bool warned = false;
+ static pthread_mutex_t lock = PTHREAD_MUTEX_INITIALIZER;
+
+ if (!warned) {
+ pthread_mutex_lock(&lock);
+ if (!warned) {
+ do_warn(
+ _("setting parent pointers to upgrade filesystem\n"));
+ warned = true;
+ }
+ pthread_mutex_unlock(&lock);
+ }
}
error = add_file_pptr(ip, ag_pptr, name);
diff --git a/repair/xfs_repair.c b/repair/xfs_repair.c
index e94b0a79378..32c28a980ff 100644
--- a/repair/xfs_repair.c
+++ b/repair/xfs_repair.c
@@ -72,6 +72,7 @@ enum c_opt_nums {
CONVERT_FINOBT,
CONVERT_REFLINK,
CONVERT_RMAPBT,
+ CONVERT_PARENT,
C_MAX_OPTS,
};
@@ -83,6 +84,7 @@ static char *c_opts[] = {
[CONVERT_FINOBT] = "finobt",
[CONVERT_REFLINK] = "reflink",
[CONVERT_RMAPBT] = "rmapbt",
+ [CONVERT_PARENT] = "parent",
[C_MAX_OPTS] = NULL,
};
@@ -369,6 +371,15 @@ process_args(int argc, char **argv)
_("-c rmapbt only supports upgrades\n"));
add_rmapbt = true;
break;
+ case CONVERT_PARENT:
+ if (!val)
+ do_abort(
+ _("-c parent requires a parameter\n"));
+ if (strtol(val, NULL, 0) != 1)
+ do_abort(
+ _("-c parent only supports upgrades\n"));
+ add_parent = true;
+ break;
default:
unknown('c', val);
break;
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 18/18] xfs_scrub: use parent pointers to report lost file data
2023-12-31 19:51 ` [PATCHSET v13.0 4/6] xfsprogs: fsck for " Darrick J. Wong
` (16 preceding siblings ...)
2023-12-31 23:17 ` [PATCH 17/18] xfs_repair: upgrade an existing filesystem to have parent pointers Darrick J. Wong
@ 2023-12-31 23:18 ` Darrick J. Wong
17 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:18 UTC (permalink / raw)
To: djwong, cem; +Cc: catherine.hoang, linux-xfs, allison.henderson
From: Darrick J. Wong <djwong@kernel.org>
If parent pointers are enabled, compute the path to the file while we're
doing the fsmap scan and report that, instead of walking the entire
directory tree to print the paths of the (hopefully few) files that lost
data.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
scrub/phase6.c | 75 +++++++++++++++++++++++++++++++++++++++++++++++---------
1 file changed, 63 insertions(+), 12 deletions(-)
diff --git a/scrub/phase6.c b/scrub/phase6.c
index 99a32bc7962..66ca57507e9 100644
--- a/scrub/phase6.c
+++ b/scrub/phase6.c
@@ -21,6 +21,7 @@
#include "read_verify.h"
#include "spacemap.h"
#include "vfs.h"
+#include "libfrog/bulkstat.h"
/*
* Phase 6: Verify data file integrity.
@@ -371,6 +372,24 @@ report_dirent_loss(
return error;
}
+struct ioerr_filerange {
+ uint64_t physical;
+ uint64_t length;
+};
+
+/*
+ * If reverse mapping and parent pointers are enabled, we can map media errors
+ * directly back to a filename and a file position without needing to walk the
+ * directory tree.
+ */
+static inline bool
+can_use_pptrs(
+ const struct scrub_ctx *ctx)
+{
+ return (ctx->mnt.fsgeom.flags & XFS_FSOP_GEOM_FLAGS_PARENT) &&
+ (ctx->mnt.fsgeom.flags & XFS_FSOP_GEOM_FLAGS_RMAPBT);
+}
+
/* Use a fsmap to report metadata lost to a media error. */
static int
report_ioerr_fsmap(
@@ -379,16 +398,18 @@ report_ioerr_fsmap(
void *arg)
{
const char *type;
+ struct xfs_bulkstat bs = { };
char buf[DESCR_BUFSZ];
- uint64_t err_physical = *(uint64_t *)arg;
+ struct ioerr_filerange *fr = arg;
uint64_t err_off;
+ int ret;
/* Don't care about unwritten extents. */
if (map->fmr_flags & FMR_OF_PREALLOC)
return 0;
- if (err_physical > map->fmr_physical)
- err_off = err_physical - map->fmr_physical;
+ if (fr->physical > map->fmr_physical)
+ err_off = fr->physical - map->fmr_physical;
else
err_off = 0;
@@ -411,23 +432,43 @@ report_ioerr_fsmap(
}
}
+ if (can_use_pptrs(ctx)) {
+ ret = -xfrog_bulkstat_single(&ctx->mnt, map->fmr_owner, 0, &bs);
+ if (ret)
+ str_liberror(ctx, ret,
+ _("bulkstat for media error report"));
+ }
+
/* Report extent maps */
if (map->fmr_flags & FMR_OF_EXTENT_MAP) {
bool attr = (map->fmr_flags & FMR_OF_ATTR_FORK);
scrub_render_ino_descr(ctx, buf, DESCR_BUFSZ,
- map->fmr_owner, 0, " %s",
+ map->fmr_owner, bs.bs_gen, " %s",
attr ? _("extended attribute") :
_("file data"));
str_corrupt(ctx, buf, _("media error in extent map"));
}
/*
- * XXX: If we had a getparent() call we could report IO errors
- * efficiently. Until then, we'll have to scan the dir tree
- * to find the bad file's pathname.
+ * If directory parent pointers are available, use that to find the
+ * pathname to a file, and report that path as having lost its
+ * extended attributes, or the precise offset of the lost file data.
*/
+ if (!can_use_pptrs(ctx))
+ return 0;
+ scrub_render_ino_descr(ctx, buf, DESCR_BUFSZ, map->fmr_owner,
+ bs.bs_gen, NULL);
+
+ if (map->fmr_flags & FMR_OF_ATTR_FORK) {
+ str_corrupt(ctx, buf, _("media error in extended attributes"));
+ return 0;
+ }
+
+ str_unfixable_error(ctx, buf,
+ _("media error at data offset %llu length %llu."),
+ err_off, fr->length);
return 0;
}
@@ -442,6 +483,10 @@ report_ioerr(
void *arg)
{
struct fsmap keys[2];
+ struct ioerr_filerange fr = {
+ .physical = start,
+ .length = length,
+ };
struct disk_ioerr_report *dioerr = arg;
dev_t dev;
@@ -457,7 +502,7 @@ report_ioerr(
(keys + 1)->fmr_offset = ULLONG_MAX;
(keys + 1)->fmr_flags = UINT_MAX;
return -scrub_iterate_fsmap(dioerr->ctx, keys, report_ioerr_fsmap,
- &start);
+ &fr);
}
/* Report all the media errors found on a disk. */
@@ -501,10 +546,16 @@ report_all_media_errors(
return ret;
}
- /* Scan the directory tree to get file paths. */
- ret = scan_fs_tree(ctx, report_dir_loss, report_dirent_loss, vs);
- if (ret)
- return ret;
+ /*
+ * Scan the directory tree to get file paths if we didn't already use
+ * directory parent pointers to report the loss.
+ */
+ if (!can_use_pptrs(ctx)) {
+ ret = scan_fs_tree(ctx, report_dir_loss, report_dirent_loss,
+ vs);
+ if (ret)
+ return ret;
+ }
/* Scan for unlinked files. */
return scrub_scan_all_inodes(ctx, report_inode_loss, vs);
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 1/6] xfs_db: add link and unlink expert commands
2023-12-31 19:51 ` [PATCHSET v13.0 5/6] xfs: detect and correct directory tree problems Darrick J. Wong
@ 2023-12-31 23:18 ` Darrick J. Wong
2023-12-31 23:18 ` [PATCH 2/6] xfs: teach online scrub to find directory tree structure problems Darrick J. Wong
` (4 subsequent siblings)
5 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:18 UTC (permalink / raw)
To: djwong, cem; +Cc: linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
Create a pair of commands to create and remove directory entries to
support functional testing of directory tree corruption.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
db/namei.c | 374 ++++++++++++++++++++++++++++++++++++++++++++++
include/xfs_inode.h | 4
libxfs/libxfs_api_defs.h | 6 +
man/man8/xfs_db.8 | 20 ++
4 files changed, 404 insertions(+)
diff --git a/db/namei.c b/db/namei.c
index fb7f63fda07..196da7f90e9 100644
--- a/db/namei.c
+++ b/db/namei.c
@@ -930,6 +930,372 @@ static struct cmdinfo parent_cmd = {
.help = parent_help,
};
+static void
+link_help(void)
+{
+ dbprintf(_(
+"\n"
+" Create a directory entry in the current directory that points to the\n"
+" specified file.\n"
+"\n"
+" Options:\n"
+" -i -- Point to this specific inode number.\n"
+" -p -- Point to the inode given by this path.\n"
+" -t -- Set the file type to this value.\n"
+" name -- Create this directory entry with this name.\n"
+ ));
+}
+
+static int
+create_child(
+ struct xfs_mount *mp,
+ xfs_ino_t parent_ino,
+ const char *name,
+ unsigned int ftype,
+ xfs_ino_t child_ino)
+{
+ struct xfs_name xname = {
+ .name = name,
+ .len = strlen(name),
+ .type = ftype,
+ };
+ struct xfs_parent_args *ppargs;
+ struct xfs_trans *tp;
+ struct xfs_inode *dp, *ip;
+ unsigned int resblks;
+ bool isdir;
+ int error;
+
+ error = -libxfs_iget(mp, NULL, parent_ino, 0, &dp);
+ if (error)
+ return error;
+
+ if (!S_ISDIR(VFS_I(dp)->i_mode)) {
+ error = -ENOTDIR;
+ goto out_dp;
+ }
+
+ error = -libxfs_iget(mp, NULL, child_ino, 0, &ip);
+ if (error)
+ goto out_dp;
+ isdir = S_ISDIR(VFS_I(ip)->i_mode);
+
+ if (xname.type == XFS_DIR3_FT_UNKNOWN)
+ xname.type = libxfs_mode_to_ftype(VFS_I(ip)->i_mode);
+
+ error = -libxfs_parent_start(mp, &ppargs);
+ if (error)
+ goto out_ip;
+
+ resblks = libxfs_link_space_res(mp, MAXNAMELEN);
+ error = -libxfs_trans_alloc(mp, &M_RES(mp)->tr_link, resblks, 0, 0,
+ &tp);
+ if (error)
+ goto out_parent;
+
+ libxfs_trans_ijoin(tp, dp, 0);
+ libxfs_trans_ijoin(tp, ip, 0);
+
+ error = -libxfs_dir_createname(tp, dp, &xname, ip->i_ino, resblks);
+ if (error)
+ goto out_trans;
+
+ /* bump dp's link to ip */
+ libxfs_bumplink(tp, ip);
+
+ /* bump ip's dotdot link to dp */
+ if (isdir)
+ libxfs_bumplink(tp, dp);
+
+ /* Replace the dotdot entry in the child directory. */
+ if (isdir) {
+ error = -libxfs_dir_replace(tp, ip, &xfs_name_dotdot,
+ dp->i_ino, resblks);
+ if (error)
+ goto out_trans;
+ }
+
+ error = -libxfs_parent_add(tp, ppargs, dp, &xname, ip);
+ if (error)
+ goto out_trans;
+
+ error = -libxfs_trans_commit(tp);
+ goto out_parent;
+
+out_trans:
+ libxfs_trans_cancel(tp);
+out_parent:
+ libxfs_parent_finish(mp, ppargs);
+out_ip:
+ libxfs_irele(ip);
+out_dp:
+ libxfs_irele(dp);
+ return error;
+}
+
+static const char *ftype_map[] = {
+ [XFS_DIR3_FT_REG_FILE] = "reg",
+ [XFS_DIR3_FT_DIR] = "dir",
+ [XFS_DIR3_FT_CHRDEV] = "cdev",
+ [XFS_DIR3_FT_BLKDEV] = "bdev",
+ [XFS_DIR3_FT_FIFO] = "fifo",
+ [XFS_DIR3_FT_SOCK] = "sock",
+ [XFS_DIR3_FT_SYMLINK] = "symlink",
+ [XFS_DIR3_FT_WHT] = "whiteout",
+};
+
+static int
+link_f(
+ int argc,
+ char **argv)
+{
+ xfs_ino_t child_ino = NULLFSINO;
+ int ftype = XFS_DIR3_FT_UNKNOWN;
+ unsigned int i;
+ int c;
+ int error = 0;
+
+ while ((c = getopt(argc, argv, "i:p:t:")) != -1) {
+ switch (c) {
+ case 'i':
+ errno = 0;
+ child_ino = strtoull(optarg, NULL, 0);
+ if (errno == ERANGE) {
+ printf("%s: unknown inode number\n", optarg);
+ exitcode = 1;
+ return 0;
+ }
+ break;
+ case 'p':
+ push_cur();
+ error = path_walk(optarg);
+ if (error) {
+ printf("%s: %s\n", optarg, strerror(error));
+ exitcode = 1;
+ return 0;
+ } else if (iocur_top->typ != &typtab[TYP_INODE]) {
+ printf("%s: does not point to an inode\n",
+ optarg);
+ exitcode = 1;
+ return 0;
+ } else {
+ child_ino = iocur_top->ino;
+ }
+ pop_cur();
+ break;
+ case 't':
+ for (i = 0; i < ARRAY_SIZE(ftype_map); i++) {
+ if (ftype_map[i] &&
+ !strcmp(ftype_map[i], optarg)) {
+ ftype = i;
+ break;
+ }
+ }
+ if (i == ARRAY_SIZE(ftype_map)) {
+ printf("%s: unknown file type\n", optarg);
+ exitcode = 1;
+ return 0;
+ }
+ break;
+ default:
+ link_help();
+ return 0;
+ }
+ }
+
+ if (child_ino == NULLFSINO) {
+ printf("link: need to specify child via -i or -p\n");
+ exitcode = 1;
+ return 0;
+ }
+
+ if (iocur_top->typ != &typtab[TYP_INODE]) {
+ printf("io cursor does not point to an inode.\n");
+ exitcode = 1;
+ return 0;
+ }
+
+ if (optind + 1 != argc) {
+ printf("link: need directory entry name");
+ exitcode = 1;
+ return 0;
+ }
+
+ error = create_child(mp, iocur_top->ino, argv[optind], ftype,
+ child_ino);
+ if (error) {
+ printf("link failed: %s\n", strerror(error));
+ exitcode = 1;
+ return 0;
+ }
+
+ return 0;
+}
+
+static struct cmdinfo link_cmd = {
+ .name = "link",
+ .cfunc = link_f,
+ .argmin = 0,
+ .argmax = -1,
+ .canpush = 0,
+ .args = "[-i ino] [-p path] [-t ftype] name",
+ .help = link_help,
+};
+
+static void
+unlink_help(void)
+{
+ dbprintf(_(
+"\n"
+" Remove a directory entry from the current directory.\n"
+"\n"
+" Options:\n"
+" name -- Remove the directory entry with this name.\n"
+ ));
+}
+
+static void
+droplink(
+ struct xfs_trans *tp,
+ struct xfs_inode *ip)
+{
+ struct inode *inode = VFS_I(ip);
+
+ libxfs_trans_ichgtime(tp, ip, XFS_ICHGTIME_CHG);
+
+ if (inode->i_nlink != XFS_NLINK_PINNED)
+ drop_nlink(VFS_I(ip));
+
+ libxfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
+}
+
+static int
+remove_child(
+ struct xfs_mount *mp,
+ xfs_ino_t parent_ino,
+ const char *name)
+{
+ struct xfs_name xname = {
+ .name = name,
+ .len = strlen(name),
+ };
+ struct xfs_parent_args *ppargs;
+ struct xfs_trans *tp;
+ struct xfs_inode *dp, *ip;
+ xfs_ino_t child_ino;
+ unsigned int resblks;
+ int error;
+
+ error = -libxfs_iget(mp, NULL, parent_ino, 0, &dp);
+ if (error)
+ return error;
+
+ if (!S_ISDIR(VFS_I(dp)->i_mode)) {
+ error = -ENOTDIR;
+ goto out_dp;
+ }
+
+ error = -libxfs_dir_lookup(NULL, dp, &xname, &child_ino, NULL);
+ if (error)
+ goto out_dp;
+
+ error = -libxfs_iget(mp, NULL, child_ino, 0, &ip);
+ if (error)
+ goto out_dp;
+
+ error = -libxfs_parent_start(mp, &ppargs);
+ if (error)
+ goto out_ip;
+
+ resblks = libxfs_remove_space_res(mp, MAXNAMELEN);
+ error = -libxfs_trans_alloc(mp, &M_RES(mp)->tr_remove, resblks, 0, 0,
+ &tp);
+ if (error)
+ goto out_parent;
+
+ libxfs_trans_ijoin(tp, dp, 0);
+ libxfs_trans_ijoin(tp, ip, 0);
+
+ if (S_ISDIR(VFS_I(ip)->i_mode)) {
+ /* drop ip's dotdot link to dp */
+ droplink(tp, dp);
+ } else {
+ libxfs_trans_log_inode(tp, dp, XFS_ILOG_CORE);
+ }
+
+ /* drop dp's link to ip */
+ droplink(tp, ip);
+
+ error = -libxfs_dir_removename(tp, dp, &xname, ip->i_ino, resblks);
+ if (error)
+ goto out_trans;
+
+ error = -libxfs_parent_remove(tp, ppargs, dp, &xname, ip);
+ if (error)
+ goto out_trans;
+
+ error = -libxfs_trans_commit(tp);
+ goto out_parent;
+
+out_trans:
+ libxfs_trans_cancel(tp);
+out_parent:
+ libxfs_parent_finish(mp, ppargs);
+out_ip:
+ libxfs_irele(ip);
+out_dp:
+ libxfs_irele(dp);
+ return error;
+}
+
+static int
+unlink_f(
+ int argc,
+ char **argv)
+{
+ int c;
+ int error = 0;
+
+ while ((c = getopt(argc, argv, "")) != -1) {
+ switch (c) {
+ default:
+ unlink_help();
+ return 0;
+ }
+ }
+
+ if (iocur_top->typ != &typtab[TYP_INODE]) {
+ printf("io cursor does not point to an inode.\n");
+ exitcode = 1;
+ return 0;
+ }
+
+ if (optind + 1 != argc) {
+ printf("unlink: need directory entry name");
+ exitcode = 1;
+ return 0;
+ }
+
+ error = remove_child(mp, iocur_top->ino, argv[optind]);
+ if (error) {
+ printf("unlink failed: %s\n", strerror(error));
+ exitcode = 1;
+ return 0;
+ }
+
+ return 0;
+}
+
+static struct cmdinfo unlink_cmd = {
+ .name = "unlink",
+ .cfunc = unlink_f,
+ .argmin = 0,
+ .argmax = -1,
+ .canpush = 0,
+ .args = "name",
+ .help = unlink_help,
+};
+
void
namei_init(void)
{
@@ -941,4 +1307,12 @@ namei_init(void)
parent_cmd.oneline = _("list parent pointers");
add_command(&parent_cmd);
+
+ if (expert_mode) {
+ link_cmd.oneline = _("create directory link");
+ add_command(&link_cmd);
+
+ unlink_cmd.oneline = _("remove directory link");
+ add_command(&unlink_cmd);
+ }
}
diff --git a/include/xfs_inode.h b/include/xfs_inode.h
index 088a6b34f04..496d504747c 100644
--- a/include/xfs_inode.h
+++ b/include/xfs_inode.h
@@ -315,6 +315,10 @@ static inline void inc_nlink(struct inode *inode)
{
inode->i_nlink++;
}
+static inline void drop_nlink(struct inode *inode)
+{
+ inode->i_nlink--;
+}
static inline bool xfs_is_reflink_inode(struct xfs_inode *ip)
{
diff --git a/libxfs/libxfs_api_defs.h b/libxfs/libxfs_api_defs.h
index d5662c42f2b..49372a44029 100644
--- a/libxfs/libxfs_api_defs.h
+++ b/libxfs/libxfs_api_defs.h
@@ -135,12 +135,16 @@
#define xfs_dir_init libxfs_dir_init
#define xfs_dir_ino_validate libxfs_dir_ino_validate
#define xfs_dir_lookup libxfs_dir_lookup
+#define xfs_dir_removename libxfs_dir_removename
#define xfs_dir_replace libxfs_dir_replace
#define xfs_dqblk_repair libxfs_dqblk_repair
#define xfs_dquot_from_disk_ts libxfs_dquot_from_disk_ts
#define xfs_dquot_verify libxfs_dquot_verify
+#define xfs_bumplink libxfs_bumplink
+#define xfs_droplink libxfs_droplink
+
#define xfs_finobt_calc_reserves libxfs_finobt_calc_reserves
#define xfs_free_extent libxfs_free_extent
#define xfs_free_extent_later libxfs_free_extent_later
@@ -177,6 +181,7 @@
#define xfs_iread_extents libxfs_iread_extents
#define xfs_irele libxfs_irele
+#define xfs_link_space_res libxfs_link_space_res
#define xfs_log_calc_minimum_size libxfs_log_calc_minimum_size
#define xfs_log_get_max_trans_res libxfs_log_get_max_trans_res
#define xfs_log_sb libxfs_log_sb
@@ -187,6 +192,7 @@
#define xfs_parent_irec_from_disk libxfs_parent_irec_from_disk
#define xfs_parent_irec_hashname libxfs_parent_irec_hashname
#define xfs_parent_lookup libxfs_parent_lookup
+#define xfs_parent_remove libxfs_parent_remove
#define xfs_parent_set libxfs_parent_set
#define xfs_parent_start libxfs_parent_start
#define xfs_parent_hashcheck libxfs_parent_hashcheck
diff --git a/man/man8/xfs_db.8 b/man/man8/xfs_db.8
index 937b17e79a3..638a8dc9352 100644
--- a/man/man8/xfs_db.8
+++ b/man/man8/xfs_db.8
@@ -884,6 +884,21 @@ will result in truncation and a warning will be issued. If no
.I label
is given, the current filesystem label is printed.
.TP
+.BI "link [-i " ino "] [-p " path "] [-t " ftype "] name"
+In the current directory, create a directory entry with the given
+.I name
+pointing to a file.
+The file must be specified either as a directory tree path as given by the
+.I path
+option; or directly as an inode number as given by the
+.I ino
+option.
+The file type in the directory entry will be determined from the mode of the
+child file unless the
+.I ftype
+option is given.
+The file being targetted must not be on the iunlink list.
+.TP
.BI "log [stop | start " filename ]
Start logging output to
.IR filename ,
@@ -1035,6 +1050,11 @@ Print the timestamps in the current locale's date and time format instead of
raw seconds since the Unix epoch.
.RE
.TP
+.BI "unlink name"
+In the current directory, remove a directory entry with the given
+.IR name .
+The file being targetted will not be put on the iunlink list.
+.TP
.BI "uuid [" uuid " | " generate " | " rewrite " | " restore ]
Set the filesystem universally unique identifier (UUID).
The filesystem UUID can be used by
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 2/6] xfs: teach online scrub to find directory tree structure problems
2023-12-31 19:51 ` [PATCHSET v13.0 5/6] xfs: detect and correct directory tree problems Darrick J. Wong
2023-12-31 23:18 ` [PATCH 1/6] xfs_db: add link and unlink expert commands Darrick J. Wong
@ 2023-12-31 23:18 ` Darrick J. Wong
2023-12-31 23:18 ` [PATCH 3/6] xfs: report directory tree corruption in the health information Darrick J. Wong
` (3 subsequent siblings)
5 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:18 UTC (permalink / raw)
To: djwong, cem; +Cc: linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
Create a new scrubber that detects corruptions within the directory tree
structure itself. It can detect directories with multiple parents;
loops within the directory tree; and directory loops not accessible from
the root.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
libfrog/scrub.c | 5 +++++
libxfs/xfs_fs.h | 3 ++-
man/man2/ioctl_xfs_scrub_metadata.2 | 14 ++++++++++++++
3 files changed, 21 insertions(+), 1 deletion(-)
diff --git a/libfrog/scrub.c b/libfrog/scrub.c
index baaa4b4d940..a2146e228f5 100644
--- a/libfrog/scrub.c
+++ b/libfrog/scrub.c
@@ -149,6 +149,11 @@ const struct xfrog_scrub_descr xfrog_scrubbers[XFS_SCRUB_TYPE_NR] = {
.descr = "retained health records",
.group = XFROG_SCRUB_GROUP_NONE,
},
+ [XFS_SCRUB_TYPE_DIRTREE] = {
+ .name = "dirtree",
+ .descr = "directory tree structure",
+ .group = XFROG_SCRUB_GROUP_INODE,
+ },
};
#undef DEP
diff --git a/libxfs/xfs_fs.h b/libxfs/xfs_fs.h
index efa68a2d82a..48f38694f12 100644
--- a/libxfs/xfs_fs.h
+++ b/libxfs/xfs_fs.h
@@ -719,9 +719,10 @@ struct xfs_scrub_metadata {
#define XFS_SCRUB_TYPE_QUOTACHECK 25 /* quota counters */
#define XFS_SCRUB_TYPE_NLINKS 26 /* inode link counts */
#define XFS_SCRUB_TYPE_HEALTHY 27 /* everything checked out ok */
+#define XFS_SCRUB_TYPE_DIRTREE 28 /* directory tree structure */
/* Number of scrub subcommands. */
-#define XFS_SCRUB_TYPE_NR 28
+#define XFS_SCRUB_TYPE_NR 29
/* i: Repair this metadata. */
#define XFS_SCRUB_IFLAG_REPAIR (1u << 0)
diff --git a/man/man2/ioctl_xfs_scrub_metadata.2 b/man/man2/ioctl_xfs_scrub_metadata.2
index 75ae52bb584..44aa139b297 100644
--- a/man/man2/ioctl_xfs_scrub_metadata.2
+++ b/man/man2/ioctl_xfs_scrub_metadata.2
@@ -148,6 +148,20 @@ that points back to the subdirectory.
The inode to examine can be specified in the same manner as
.BR XFS_SCRUB_TYPE_INODE "."
+.TP
+.B XFS_SCRUB_TYPE_DIRTREE
+This scrubber looks for problems in the directory tree structure such as loops
+and directories accessible through more than one path.
+Problems are detected by walking parent pointers upwards towards the root.
+Loops are detected by comparing the parent directory at each step against the
+directories already examined.
+Directories with multiple paths are detected by counting the parent pointers
+attached to a directory.
+Non-directories do not have links pointing away from the directory tree root
+and can be skipped.
+The directory to examine can be specified in the same manner as
+.BR XFS_SCRUB_TYPE_INODE "."
+
.TP
.B XFS_SCRUB_TYPE_SYMLINK
Examine the target of a symbolic link for obvious pathname problems.
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 3/6] xfs: report directory tree corruption in the health information
2023-12-31 19:51 ` [PATCHSET v13.0 5/6] xfs: detect and correct directory tree problems Darrick J. Wong
2023-12-31 23:18 ` [PATCH 1/6] xfs_db: add link and unlink expert commands Darrick J. Wong
2023-12-31 23:18 ` [PATCH 2/6] xfs: teach online scrub to find directory tree structure problems Darrick J. Wong
@ 2023-12-31 23:18 ` Darrick J. Wong
2023-12-31 23:19 ` [PATCH 4/6] xfs_scrub: fix erroring out of check_inode_names Darrick J. Wong
` (2 subsequent siblings)
5 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:18 UTC (permalink / raw)
To: djwong, cem; +Cc: linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
Report directories that are the source of corruption in the directory
tree.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
libxfs/xfs_fs.h | 1 +
libxfs/xfs_health.h | 4 +++-
man/man2/ioctl_xfs_bulkstat.2 | 3 +++
man/man2/ioctl_xfs_fsbulkstat.2 | 3 +++
spaceman/health.c | 4 ++++
5 files changed, 14 insertions(+), 1 deletion(-)
diff --git a/libxfs/xfs_fs.h b/libxfs/xfs_fs.h
index 48f38694f12..2499a20f5f7 100644
--- a/libxfs/xfs_fs.h
+++ b/libxfs/xfs_fs.h
@@ -413,6 +413,7 @@ struct xfs_bulkstat {
#define XFS_BS_SICK_XATTR (1 << 5) /* extended attributes */
#define XFS_BS_SICK_SYMLINK (1 << 6) /* symbolic link remote target */
#define XFS_BS_SICK_PARENT (1 << 7) /* parent pointers */
+#define XFS_BS_SICK_DIRTREE (1 << 8) /* directory tree structure */
/*
* Project quota id helpers (previously projid was 16bit only
diff --git a/libxfs/xfs_health.h b/libxfs/xfs_health.h
index df07c5877ba..bca1990f71d 100644
--- a/libxfs/xfs_health.h
+++ b/libxfs/xfs_health.h
@@ -95,6 +95,7 @@ struct xfs_da_args;
/* Don't propagate sick status to ag health summary during inactivation */
#define XFS_SICK_INO_FORGET (1 << 12)
+#define XFS_SICK_INO_DIRTREE (1 << 13) /* directory tree structure */
/* Primary evidence of health problems in a given group. */
#define XFS_SICK_FS_PRIMARY (XFS_SICK_FS_COUNTERS | \
@@ -125,7 +126,8 @@ struct xfs_da_args;
XFS_SICK_INO_DIR | \
XFS_SICK_INO_XATTR | \
XFS_SICK_INO_SYMLINK | \
- XFS_SICK_INO_PARENT)
+ XFS_SICK_INO_PARENT | \
+ XFS_SICK_INO_DIRTREE)
#define XFS_SICK_INO_ZAPPED (XFS_SICK_INO_BMBTD_ZAPPED | \
XFS_SICK_INO_BMBTA_ZAPPED | \
diff --git a/man/man2/ioctl_xfs_bulkstat.2 b/man/man2/ioctl_xfs_bulkstat.2
index 3203ca0c5d2..b6d51aa4381 100644
--- a/man/man2/ioctl_xfs_bulkstat.2
+++ b/man/man2/ioctl_xfs_bulkstat.2
@@ -326,6 +326,9 @@ Symbolic link target.
.TP
.B XFS_BS_SICK_PARENT
Parent pointers.
+.TP
+.B XFS_BS_SICK_DIRTREE
+Directory is the source of corruption in the directory tree.
.RE
.SH ERRORS
Error codes can be one of, but are not limited to, the following:
diff --git a/man/man2/ioctl_xfs_fsbulkstat.2 b/man/man2/ioctl_xfs_fsbulkstat.2
index 3f059942a21..cd38d2fd6f2 100644
--- a/man/man2/ioctl_xfs_fsbulkstat.2
+++ b/man/man2/ioctl_xfs_fsbulkstat.2
@@ -239,6 +239,9 @@ Symbolic link target.
.TP
.B XFS_BS_SICK_PARENT
Parent pointers.
+.TP
+.B XFS_BS_SICK_DIRTREE
+Directory is the source of corruption in the directory tree.
.RE
.SH RETURN VALUE
On error, \-1 is returned, and
diff --git a/spaceman/health.c b/spaceman/health.c
index ab5bc074988..8ba78152cb6 100644
--- a/spaceman/health.c
+++ b/spaceman/health.c
@@ -169,6 +169,10 @@ static const struct flag_map inode_flags[] = {
.mask = XFS_BS_SICK_PARENT,
.descr = "parent pointers",
},
+ {
+ .mask = XFS_BS_SICK_DIRTREE,
+ .descr = "directory tree structure",
+ },
{0},
};
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 4/6] xfs_scrub: fix erroring out of check_inode_names
2023-12-31 19:51 ` [PATCHSET v13.0 5/6] xfs: detect and correct directory tree problems Darrick J. Wong
` (2 preceding siblings ...)
2023-12-31 23:18 ` [PATCH 3/6] xfs: report directory tree corruption in the health information Darrick J. Wong
@ 2023-12-31 23:19 ` Darrick J. Wong
2023-12-31 23:19 ` [PATCH 5/6] xfs_scrub: detect and repair directory tree corruptions Darrick J. Wong
2023-12-31 23:19 ` [PATCH 6/6] xfs_scrub: defer phase5 file scans if dirloop fails Darrick J. Wong
5 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:19 UTC (permalink / raw)
To: djwong, cem; +Cc: linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
The early exit logic in this function is a bit suboptimal -- we don't
need to close the @fd if we haven't even opened it, and since all errors
are fatal, we don't need to bump the progress counter. The logic in
this function is about to get more involved due to the addition of the
directory tree structure checker, so clean up these warts.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
scrub/phase5.c | 10 +++++-----
1 file changed, 5 insertions(+), 5 deletions(-)
diff --git a/scrub/phase5.c b/scrub/phase5.c
index 0df8c46e9f5..b3719627755 100644
--- a/scrub/phase5.c
+++ b/scrub/phase5.c
@@ -279,7 +279,7 @@ check_inode_names(
if (bstat->bs_xflags & FS_XFLAG_HASATTR) {
error = check_xattr_names(ctx, &dsc, handle, bstat);
if (error)
- goto out;
+ goto err;
}
/*
@@ -295,16 +295,16 @@ check_inode_names(
if (error == ESTALE)
return ESTALE;
str_errno(ctx, descr_render(&dsc));
- goto out;
+ goto err;
}
error = check_dirent_names(ctx, &dsc, &fd, bstat);
if (error)
- goto out;
+ goto err_fd;
}
-out:
progress_add(1);
+err_fd:
if (fd >= 0) {
err2 = close(fd);
if (err2)
@@ -312,7 +312,7 @@ check_inode_names(
if (!error && err2)
error = err2;
}
-
+err:
if (error)
*aborted = true;
if (!error && *aborted)
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 5/6] xfs_scrub: detect and repair directory tree corruptions
2023-12-31 19:51 ` [PATCHSET v13.0 5/6] xfs: detect and correct directory tree problems Darrick J. Wong
` (3 preceding siblings ...)
2023-12-31 23:19 ` [PATCH 4/6] xfs_scrub: fix erroring out of check_inode_names Darrick J. Wong
@ 2023-12-31 23:19 ` Darrick J. Wong
2023-12-31 23:19 ` [PATCH 6/6] xfs_scrub: defer phase5 file scans if dirloop fails Darrick J. Wong
5 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:19 UTC (permalink / raw)
To: djwong, cem; +Cc: linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
Now that we have online fsck for directory tree structure problems, we
need to find a place to call it. The scanner requires that parent
pointers are enabled, that directory link counts are correct, and that
every directory entry has a corresponding parent pointer. Therefore, we
can only run it after phase 4 fixes every file, and phase 5 resets the
link counts.
In other words, we call it as part of the phase 5 file scan that we do
to warn about weird looking file names. This has the added benefit that
opening the directory by handle is less likely to fail if there are
loops in the directory structure. For now, only plumb in enough to try
to fix directory tree problems right away; the next patch will make
phase 5 retry the dirloop scanner until the problems are fixed or we
stop making forward progress.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
scrub/phase5.c | 56 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++
1 file changed, 56 insertions(+)
diff --git a/scrub/phase5.c b/scrub/phase5.c
index b3719627755..6c8dee66e6e 100644
--- a/scrub/phase5.c
+++ b/scrub/phase5.c
@@ -252,6 +252,47 @@ render_ino_from_handle(
bstat->bs_gen, NULL);
}
+/*
+ * Check the directory structure for problems that could cause open_by_handle
+ * not to work. Returns 0 for no problems; EADDRNOTAVAIL if the there are
+ * problems that would prevent name checking.
+ */
+static int
+check_dir_connection(
+ struct scrub_ctx *ctx,
+ struct descr *dsc,
+ const struct xfs_bulkstat *bstat)
+{
+ struct scrub_item sri = { };
+ int error;
+
+ /* The dirtree scrubber only works when parent pointers are enabled */
+ if (!(ctx->mnt.fsgeom.flags & XFS_FSOP_GEOM_FLAGS_PARENT))
+ return 0;
+
+ scrub_item_init_file(&sri, bstat);
+ scrub_item_schedule(&sri, XFS_SCRUB_TYPE_DIRTREE);
+
+ error = scrub_item_check_file(ctx, &sri, -1);
+ if (error) {
+ str_liberror(ctx, error, _("checking directory loops"));
+ return error;
+ }
+
+ error = repair_file_corruption(ctx, &sri, -1);
+ if (error) {
+ str_liberror(ctx, error, _("repairing directory loops"));
+ return error;
+ }
+
+ /* No directory tree problems? Clear this inode if it was deferred. */
+ if (repair_item_count_needsrepair(&sri) == 0)
+ return 0;
+
+ str_corrupt(ctx, descr_render(dsc), _("directory loop uncorrected!"));
+ return EADDRNOTAVAIL;
+}
+
/*
* Verify the connectivity of the directory tree.
* We know that the kernel's open-by-handle function will try to reconnect
@@ -275,6 +316,20 @@ check_inode_names(
descr_set(&dsc, bstat);
background_sleep();
+ /*
+ * Try to fix directory loops before we have problems opening files by
+ * handle.
+ */
+ if (S_ISDIR(bstat->bs_mode)) {
+ error = check_dir_connection(ctx, &dsc, bstat);
+ if (error == EADDRNOTAVAIL) {
+ error = 0;
+ goto out;
+ }
+ if (error)
+ goto err;
+ }
+
/* Warn about naming problems in xattrs. */
if (bstat->bs_xflags & FS_XFLAG_HASATTR) {
error = check_xattr_names(ctx, &dsc, handle, bstat);
@@ -315,6 +370,7 @@ check_inode_names(
err:
if (error)
*aborted = true;
+out:
if (!error && *aborted)
error = ECANCELED;
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 6/6] xfs_scrub: defer phase5 file scans if dirloop fails
2023-12-31 19:51 ` [PATCHSET v13.0 5/6] xfs: detect and correct directory tree problems Darrick J. Wong
` (4 preceding siblings ...)
2023-12-31 23:19 ` [PATCH 5/6] xfs_scrub: detect and repair directory tree corruptions Darrick J. Wong
@ 2023-12-31 23:19 ` Darrick J. Wong
5 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:19 UTC (permalink / raw)
To: djwong, cem; +Cc: linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
If we cannot fix dirloop problems during the initial phase 5 inode scan,
defer them until later.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
scrub/phase5.c | 215 ++++++++++++++++++++++++++++++++++++++++++++++++++++----
scrub/repair.c | 13 +++
scrub/repair.h | 2 +
3 files changed, 216 insertions(+), 14 deletions(-)
diff --git a/scrub/phase5.c b/scrub/phase5.c
index 6c8dee66e6e..f6c295c64ad 100644
--- a/scrub/phase5.c
+++ b/scrub/phase5.c
@@ -18,6 +18,8 @@
#include "libfrog/workqueue.h"
#include "libfrog/fsgeom.h"
#include "libfrog/scrub.h"
+#include "libfrog/bitmap.h"
+#include "libfrog/bulkstat.h"
#include "xfs_scrub.h"
#include "common.h"
#include "inodes.h"
@@ -29,6 +31,36 @@
/* Phase 5: Full inode scans and check directory connectivity. */
+struct ncheck_state {
+ struct scrub_ctx *ctx;
+
+ /* Have we aborted this scan? */
+ bool aborted;
+
+ /* Is this the last time we're going to process deferred inodes? */
+ bool last_call;
+
+ /* Did we fix at least one thing while walking @cur->deferred? */
+ bool fixed_something;
+
+ /* Lock for this structure */
+ pthread_mutex_t lock;
+
+ /*
+ * Inodes that are involved with directory tree structure corruptions
+ * are marked here. This will be NULL until the first corruption is
+ * noted.
+ */
+ struct bitmap *new_deferred;
+
+ /*
+ * Inodes that we're reprocessing due to earlier directory tree
+ * structure corruption problems are marked here. This will be NULL
+ * during the first (parallel) inode scan.
+ */
+ struct bitmap *cur_deferred;
+};
+
/*
* Warn about problematic bytes in a directory/attribute name. That means
* terminal control characters and escape sequences, since that could be used
@@ -252,6 +284,26 @@ render_ino_from_handle(
bstat->bs_gen, NULL);
}
+/* Defer this inode until later. */
+static inline int
+defer_inode(
+ struct ncheck_state *ncs,
+ uint64_t ino)
+{
+ int error;
+
+ pthread_mutex_lock(&ncs->lock);
+ if (!ncs->new_deferred) {
+ error = -bitmap_alloc(&ncs->new_deferred);
+ if (error)
+ goto unlock;
+ }
+ error = -bitmap_set(ncs->new_deferred, ino, 1);
+unlock:
+ pthread_mutex_unlock(&ncs->lock);
+ return error;
+}
+
/*
* Check the directory structure for problems that could cause open_by_handle
* not to work. Returns 0 for no problems; EADDRNOTAVAIL if the there are
@@ -260,7 +312,7 @@ render_ino_from_handle(
static int
check_dir_connection(
struct scrub_ctx *ctx,
- struct descr *dsc,
+ struct ncheck_state *ncs,
const struct xfs_bulkstat *bstat)
{
struct scrub_item sri = { };
@@ -279,17 +331,31 @@ check_dir_connection(
return error;
}
- error = repair_file_corruption(ctx, &sri, -1);
+ if (ncs->last_call)
+ error = repair_file_corruption_now(ctx, &sri, -1);
+ else
+ error = repair_file_corruption(ctx, &sri, -1);
if (error) {
str_liberror(ctx, error, _("repairing directory loops"));
return error;
}
/* No directory tree problems? Clear this inode if it was deferred. */
- if (repair_item_count_needsrepair(&sri) == 0)
+ if (repair_item_count_needsrepair(&sri) == 0) {
+ if (ncs->cur_deferred)
+ ncs->fixed_something = true;
return 0;
+ }
+
+ /* Don't defer anything during last call. */
+ if (ncs->last_call)
+ return 0;
+
+ /* Directory tree structure problems exist; do not check names yet. */
+ error = defer_inode(ncs, bstat->bs_ino);
+ if (error)
+ return error;
- str_corrupt(ctx, descr_render(dsc), _("directory loop uncorrected!"));
return EADDRNOTAVAIL;
}
@@ -308,7 +374,7 @@ check_inode_names(
void *arg)
{
DEFINE_DESCR(dsc, ctx, render_ino_from_handle);
- bool *aborted = arg;
+ struct ncheck_state *ncs = arg;
int fd = -1;
int error = 0;
int err2;
@@ -321,7 +387,7 @@ check_inode_names(
* handle.
*/
if (S_ISDIR(bstat->bs_mode)) {
- error = check_dir_connection(ctx, &dsc, bstat);
+ error = check_dir_connection(ctx, ncs, bstat);
if (error == EADDRNOTAVAIL) {
error = 0;
goto out;
@@ -369,14 +435,120 @@ check_inode_names(
}
err:
if (error)
- *aborted = true;
+ ncs->aborted = true;
out:
- if (!error && *aborted)
+ if (!error && ncs->aborted)
error = ECANCELED;
return error;
}
+/* Try to check_inode_names on a specific inode. */
+static int
+retry_deferred_inode(
+ struct ncheck_state *ncs,
+ struct xfs_handle *handle,
+ uint64_t ino)
+{
+ struct xfs_bulkstat bstat;
+ struct scrub_ctx *ctx = ncs->ctx;
+ unsigned int flags = 0;
+ int error;
+
+ error = -xfrog_bulkstat_single(&ctx->mnt, ino, flags, &bstat);
+ if (error == ENOENT) {
+ /* Directory is gone, mark it clear. */
+ ncs->fixed_something = true;
+ return 0;
+ }
+ if (error)
+ return error;
+
+ handle->ha_fid.fid_ino = bstat.bs_ino;
+ handle->ha_fid.fid_gen = bstat.bs_gen;
+
+ return check_inode_names(ncs->ctx, handle, &bstat, ncs);
+}
+
+/* Try to check_inode_names on a range of inodes from the bitmap. */
+static int
+retry_deferred_inode_range(
+ uint64_t ino,
+ uint64_t len,
+ void *arg)
+{
+ struct xfs_handle handle = { };
+ struct ncheck_state *ncs = arg;
+ struct scrub_ctx *ctx = ncs->ctx;
+ uint64_t i;
+ int error;
+
+ memcpy(&handle.ha_fsid, ctx->fshandle, sizeof(handle.ha_fsid));
+ handle.ha_fid.fid_len = sizeof(xfs_fid_t) -
+ sizeof(handle.ha_fid.fid_len);
+ handle.ha_fid.fid_pad = 0;
+
+ for (i = 0; i < len; i++) {
+ error = retry_deferred_inode(ncs, &handle, ino + i);
+ if (error)
+ return error;
+ }
+
+ return 0;
+}
+
+/*
+ * Try to check_inode_names on inodes that were deferred due to directory tree
+ * problems until we stop making progress.
+ */
+static int
+retry_deferred_inodes(
+ struct scrub_ctx *ctx,
+ struct ncheck_state *ncs)
+{
+ int error;
+
+ if (!ncs->new_deferred)
+ return 0;
+
+ /*
+ * Try to repair things until we stop making forward progress or we
+ * don't observe any new corruptions. During the loop, we do not
+ * complain about the corruptions that do not get fixed.
+ */
+ do {
+ ncs->cur_deferred = ncs->new_deferred;
+ ncs->new_deferred = NULL;
+ ncs->fixed_something = false;
+
+ error = -bitmap_iterate(ncs->cur_deferred,
+ retry_deferred_inode_range, ncs);
+ if (error)
+ return error;
+
+ bitmap_free(&ncs->cur_deferred);
+ } while (ncs->fixed_something && ncs->new_deferred);
+
+ /*
+ * Try one last time to fix things, and complain about any problems
+ * that remain.
+ */
+ if (!ncs->new_deferred)
+ return 0;
+
+ ncs->cur_deferred = ncs->new_deferred;
+ ncs->new_deferred = NULL;
+ ncs->last_call = true;
+
+ error = -bitmap_iterate(ncs->cur_deferred,
+ retry_deferred_inode_range, ncs);
+ if (error)
+ return error;
+
+ bitmap_free(&ncs->cur_deferred);
+ return 0;
+}
+
#ifndef FS_IOC_GETFSLABEL
# define FSLABEL_MAX 256
# define FS_IOC_GETFSLABEL _IOR(0x94, 49, char[FSLABEL_MAX])
@@ -568,9 +740,10 @@ int
phase5_func(
struct scrub_ctx *ctx)
{
- bool aborted = false;
+ struct ncheck_state ncs = { .ctx = ctx };
int ret;
+
/*
* Check and fix anything that requires a full filesystem scan. We do
* this after we've checked all inodes and repaired anything that could
@@ -590,14 +763,28 @@ _("Filesystem has errors, skipping connectivity checks."));
if (ret)
return ret;
- ret = scrub_scan_all_inodes(ctx, check_inode_names, &aborted);
+ pthread_mutex_init(&ncs.lock, NULL);
+
+ ret = scrub_scan_all_inodes(ctx, check_inode_names, &ncs);
if (ret)
- return ret;
- if (aborted)
- return ECANCELED;
+ goto out_lock;
+ if (ncs.aborted) {
+ ret = ECANCELED;
+ goto out_lock;
+ }
+
+ ret = retry_deferred_inodes(ctx, &ncs);
+ if (ret)
+ goto out_lock;
scrub_report_preen_triggers(ctx);
- return 0;
+out_lock:
+ pthread_mutex_destroy(&ncs.lock);
+ if (ncs.new_deferred)
+ bitmap_free(&ncs.new_deferred);
+ if (ncs.cur_deferred)
+ bitmap_free(&ncs.cur_deferred);
+ return ret;
}
/* Estimate how much work we're going to do. */
diff --git a/scrub/repair.c b/scrub/repair.c
index 0258210722b..4fed86134ed 100644
--- a/scrub/repair.c
+++ b/scrub/repair.c
@@ -732,6 +732,19 @@ repair_file_corruption(
XRM_REPAIR_ONLY | XRM_NOPROGRESS);
}
+/* Repair all parts of this file or complain if we cannot. */
+int
+repair_file_corruption_now(
+ struct scrub_ctx *ctx,
+ struct scrub_item *sri,
+ int override_fd)
+{
+ repair_item_boost_priorities(sri);
+
+ return repair_item_class(ctx, sri, override_fd, SCRUB_ITEM_CORRUPT,
+ XRM_REPAIR_ONLY | XRM_NOPROGRESS | XRM_FINAL_WARNING);
+}
+
/*
* Repair everything in this filesystem object that needs it. This includes
* cross-referencing and preening.
diff --git a/scrub/repair.h b/scrub/repair.h
index 411a379f6fa..ec4aa381a82 100644
--- a/scrub/repair.h
+++ b/scrub/repair.h
@@ -76,6 +76,8 @@ int action_list_process(struct scrub_ctx *ctx, struct action_list *alist,
int repair_item_corruption(struct scrub_ctx *ctx, struct scrub_item *sri);
int repair_file_corruption(struct scrub_ctx *ctx, struct scrub_item *sri,
int override_fd);
+int repair_file_corruption_now(struct scrub_ctx *ctx, struct scrub_item *sri,
+ int override_fd);
int repair_item(struct scrub_ctx *ctx, struct scrub_item *sri,
unsigned int repair_flags);
int repair_item_to_action_item(struct scrub_ctx *ctx,
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 01/10] xfs: introduce vectored scrub mode
2023-12-31 19:51 ` [PATCHSET 6/6] xfs_scrub: vectorize kernel calls Darrick J. Wong
@ 2023-12-31 23:19 ` Darrick J. Wong
2023-12-31 23:20 ` [PATCH 02/10] libfrog: support vectored scrub Darrick J. Wong
` (8 subsequent siblings)
9 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:19 UTC (permalink / raw)
To: djwong, cem; +Cc: linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
Introduce a variant on XFS_SCRUB_METADATA that allows for vectored mode.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
libxfs/xfs_fs.h | 10 ++
libxfs/xfs_fs_staging.h | 32 ++++++
man/man2/ioctl_xfs_scrubv_metadata.2 | 168 ++++++++++++++++++++++++++++++++++
3 files changed, 210 insertions(+)
create mode 100644 man/man2/ioctl_xfs_scrubv_metadata.2
diff --git a/libxfs/xfs_fs.h b/libxfs/xfs_fs.h
index 2499a20f5f7..77fbca573e1 100644
--- a/libxfs/xfs_fs.h
+++ b/libxfs/xfs_fs.h
@@ -725,6 +725,15 @@ struct xfs_scrub_metadata {
/* Number of scrub subcommands. */
#define XFS_SCRUB_TYPE_NR 29
+/*
+ * This special type code only applies to the vectored scrub implementation.
+ *
+ * If any of the previous scrub vectors recorded runtime errors or have
+ * sv_flags bits set that match the OFLAG bits in the barrier vector's
+ * sv_flags, set the barrier's sv_ret to -ECANCELED and return to userspace.
+ */
+#define XFS_SCRUB_TYPE_BARRIER (-1U)
+
/* i: Repair this metadata. */
#define XFS_SCRUB_IFLAG_REPAIR (1u << 0)
@@ -813,6 +822,7 @@ struct xfs_scrub_metadata {
#define XFS_IOC_FREE_EOFBLOCKS _IOR ('X', 58, struct xfs_fs_eofblocks)
/* XFS_IOC_GETFSMAP ------ hoisted 59 */
#define XFS_IOC_SCRUB_METADATA _IOWR('X', 60, struct xfs_scrub_metadata)
+/* XFS_IOC_SCRUBV_METADATA -- staging 60 */
#define XFS_IOC_AG_GEOMETRY _IOWR('X', 61, struct xfs_ag_geometry)
/* XFS_IOC_GETPARENTS ---- staging 62 */
diff --git a/libxfs/xfs_fs_staging.h b/libxfs/xfs_fs_staging.h
index e0650af0558..69d29f213af 100644
--- a/libxfs/xfs_fs_staging.h
+++ b/libxfs/xfs_fs_staging.h
@@ -170,4 +170,36 @@ xfs_getparents_rec(
#define XFS_IOC_GETPARENTS _IOWR('X', 62, struct xfs_getparents)
+/* Vectored scrub calls to reduce the number of kernel transitions. */
+
+struct xfs_scrub_vec {
+ __u32 sv_type; /* XFS_SCRUB_TYPE_* */
+ __u32 sv_flags; /* XFS_SCRUB_FLAGS_* */
+ __s32 sv_ret; /* 0 or a negative error code */
+ __u32 sv_reserved; /* must be zero */
+};
+
+/* Vectored metadata scrub control structure. */
+struct xfs_scrub_vec_head {
+ __u64 svh_ino; /* inode number. */
+ __u32 svh_gen; /* inode generation. */
+ __u32 svh_agno; /* ag number. */
+ __u32 svh_flags; /* XFS_SCRUB_VEC_FLAGS_* */
+ __u16 svh_rest_us; /* wait this much time between vector items */
+ __u16 svh_nr; /* number of svh_vecs */
+ __u64 svh_reserved; /* must be zero */
+
+ struct xfs_scrub_vec svh_vecs[];
+};
+
+#define XFS_SCRUB_VEC_FLAGS_ALL (0)
+
+static inline size_t sizeof_xfs_scrub_vec(unsigned int nr)
+{
+ return sizeof(struct xfs_scrub_vec_head) +
+ nr * sizeof(struct xfs_scrub_vec);
+}
+
+#define XFS_IOC_SCRUBV_METADATA _IOWR('X', 60, struct xfs_scrub_vec_head)
+
#endif /* __XFS_FS_STAGING_H__ */
diff --git a/man/man2/ioctl_xfs_scrubv_metadata.2 b/man/man2/ioctl_xfs_scrubv_metadata.2
new file mode 100644
index 00000000000..05a4adaba48
--- /dev/null
+++ b/man/man2/ioctl_xfs_scrubv_metadata.2
@@ -0,0 +1,168 @@
+.\" Copyright (c) 2023-2024 Oracle. All rights reserved.
+.\"
+.\" %%%LICENSE_START(GPLv2+_DOC_FULL)
+.\" SPDX-License-Identifier: GPL-2.0-or-later
+.\" %%%LICENSE_END
+.TH IOCTL-XFS-SCRUBV-METADATA 2 2023-08-18 "XFS"
+.SH NAME
+ioctl_xfs_scrubv_metadata \- check a lot of XFS filesystem metadata
+.SH SYNOPSIS
+.br
+.B #include <xfs/xfs_fs.h>
+.br
+.B #include <xfs/xfs_fs_staging.h>
+.PP
+.BI "int ioctl(int " dest_fd ", XFS_IOC_SCRUBV_METADATA, struct xfs_scrub_vec_head *" arg );
+.SH DESCRIPTION
+This XFS ioctl asks the kernel driver to examine several pieces of filesystem
+metadata for errors or suboptimal metadata.
+Multiple scrub types can be invoked to target a single filesystem object.
+See
+.BR ioctl_xfs_scrub_metadata (2)
+for a discussion of metadata validation, and documentation of the various
+.B XFS_SCRUB_TYPE
+and
+.B XFS_SCRUB_FLAGS
+values referenced below.
+
+The types and location of the metadata to scrub are conveyed as a vector with
+a header of the following form:
+.PP
+.in +4n
+.nf
+
+struct xfs_scrub_vec_head {
+ __u64 svh_ino;
+ __u32 svh_gen;
+ __u32 svh_agno;
+ __u32 svh_flags;
+ __u16 svh_rest_us;
+ __u16 svh_nr;
+ __u64 svh_reserved;
+
+ struct xfs_scrub_vec svh_vecs[];
+};
+.fi
+.in
+.PP
+The field
+.IR svh_ino ,
+.IR svh_gen ,
+and
+.IR svh_agno
+correspond to the
+.IR sm_ino ,
+.IR sm_gen ,
+and
+.IR sm_agno
+fields of the regular scrub ioctl.
+Exactly one filesystem object can be specified in a single call.
+The kernel will proceed with each vector in
+.I svh_vecs
+until progress is no longer possible.
+
+The field
+.I svh_rest_us
+specifies an amount of time to pause between each scrub invocation to give
+the system a chance to process other requests.
+
+The field
+.I svh_nr
+specifies the number of vectors in the
+.I svh_recs
+flex array.
+
+.PP
+The field
+.I svh_reserved
+must be zero.
+
+Each vector has the following form:
+.PP
+.in +4n
+.nf
+
+struct xfs_scrub_vec {
+ __u32 sv_type;
+ __u32 sv_flags;
+ __s32 sv_ret;
+ __u32 sv_reserved;
+};
+.fi
+.in
+
+.PP
+The fields
+.I sv_type
+and
+.I sv_flags
+indicate the type of metadata to check and the behavioral changes that
+userspace will permit of the kernel.
+The
+.I sv_flags
+field will be updated upon completion of the scrub call.
+See the documentation of
+.B XFS_SCRUB_TYPE_*
+and
+.B XFS_SCRUB_[IO]FLAG_*
+values in
+.BR ioctl_xfs_scrub_metadata (2)
+for a detailed description of their purpose.
+
+.PP
+If a vector's
+.I sv_type
+field is set to the value
+.BR XFS_SCRUB_TYPE_BARRIER ,
+the kernel will stop processing vectors and return to userspace if a scrubber
+flags corruption by setting one of the
+.B XFS_SCRUB_OFLAG_*
+values in
+.I sv_flags
+or
+returns an operation error in
+.IR sv_ret .
+Otherwise, the kernel returns only after processing all vectors.
+
+The
+.I sv_ret
+field is set to the return value of the scrub function.
+See the RETURN VALUE
+section of the
+.BR ioctl_xfs_scrub_metadata (2)
+manual page for more information.
+
+The
+.B sv_reserved
+field must be zero.
+
+.SH RETURN VALUE
+On error, \-1 is returned, and
+.I errno
+is set to indicate the error.
+.PP
+.SH ERRORS
+Error codes can be one of, but are not limited to, the following:
+.TP
+.B EINVAL
+One or more of the arguments specified is invalid.
+.TP
+.B EINTR
+The operation was interrupted.
+.TP
+.B ENOMEM
+There was not sufficient memory to perform the scrub or repair operation.
+.TP
+.B EFAULT
+A memory fault was encountered while reading or writing the vector.
+.SH CONFORMING TO
+This API is specific to XFS filesystem on the Linux kernel.
+.SH NOTES
+These operations may block other filesystem operations for a long time.
+A calling process can stop the operation by being sent a fatal
+signal, but non-fatal signals are blocked.
+.SH SEE ALSO
+.BR ioctl (2)
+.BR ioctl_xfs_scrub_metadata (2)
+.BR xfs_scrub (8)
+.BR xfs_repair (8)
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 02/10] libfrog: support vectored scrub
2023-12-31 19:51 ` [PATCHSET 6/6] xfs_scrub: vectorize kernel calls Darrick J. Wong
2023-12-31 23:19 ` [PATCH 01/10] xfs: introduce vectored scrub mode Darrick J. Wong
@ 2023-12-31 23:20 ` Darrick J. Wong
2023-12-31 23:20 ` [PATCH 03/10] xfs_io: " Darrick J. Wong
` (7 subsequent siblings)
9 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:20 UTC (permalink / raw)
To: djwong, cem; +Cc: linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
Enhance libfrog to support performing vectored metadata scrub.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
libfrog/fsgeom.h | 6 +++
libfrog/scrub.c | 124 ++++++++++++++++++++++++++++++++++++++++++++++++++++++
libfrog/scrub.h | 4 ++
3 files changed, 134 insertions(+)
diff --git a/libfrog/fsgeom.h b/libfrog/fsgeom.h
index 7e002c5137a..4f3542eafec 100644
--- a/libfrog/fsgeom.h
+++ b/libfrog/fsgeom.h
@@ -56,6 +56,12 @@ struct xfs_fd {
/* Only use XFS_IOC_EXCHANGE_RANGE for file data exchanges. */
#define XFROG_FLAG_FORCE_EXCH_RANGE (1 << 3)
+/* Only use the older one-at-a-time scrub ioctl. */
+#define XFROG_FLAG_SCRUB_FORCE_SINGLE (1 << 4)
+
+/* Only use the vectored scrub ioctl. */
+#define XFROG_FLAG_SCRUB_FORCE_VECTOR (1 << 5)
+
/* Static initializers */
#define XFS_FD_INIT(_fd) { .fd = (_fd), }
#define XFS_FD_INIT_EMPTY XFS_FD_INIT(-1)
diff --git a/libfrog/scrub.c b/libfrog/scrub.c
index a2146e228f5..8264aab00ef 100644
--- a/libfrog/scrub.c
+++ b/libfrog/scrub.c
@@ -171,3 +171,127 @@ xfrog_scrub_metadata(
return 0;
}
+
+/* Decide if there have been any scrub failures up to this point. */
+static inline int
+xfrog_scrubv_previous_failures(
+ struct xfs_scrub_vec_head *vhead,
+ struct xfs_scrub_vec *barrier_vec)
+{
+ struct xfs_scrub_vec *v;
+ __u32 failmask;
+
+ failmask = barrier_vec->sv_flags & XFS_SCRUB_FLAGS_OUT;
+ for (v = vhead->svh_vecs; v < barrier_vec; v++) {
+ if (v->sv_type == XFS_SCRUB_TYPE_BARRIER)
+ continue;
+
+ /*
+ * Runtime errors count as a previous failure, except the ones
+ * used to ask userspace to retry.
+ */
+ if (v->sv_ret && v->sv_ret != -EBUSY && v->sv_ret != -ENOENT &&
+ v->sv_ret != -EUSERS)
+ return -ECANCELED;
+
+ /*
+ * If any of the out-flags on the scrub vector match the mask
+ * that was set on the barrier vector, that's a previous fail.
+ */
+ if (v->sv_flags & failmask)
+ return -ECANCELED;
+ }
+
+ return 0;
+}
+
+static int
+xfrog_scrubv_fallback(
+ struct xfs_fd *xfd,
+ struct xfs_scrub_vec_head *vhead)
+{
+ struct xfs_scrub_vec *v;
+ unsigned int i;
+
+ if (vhead->svh_flags & ~XFS_SCRUB_VEC_FLAGS_ALL)
+ return -EINVAL;
+ for (i = 0, v = vhead->svh_vecs; i < vhead->svh_nr; i++, v++) {
+ if (v->sv_reserved)
+ return -EINVAL;
+ if (v->sv_type == XFS_SCRUB_TYPE_BARRIER &&
+ (v->sv_flags & ~XFS_SCRUB_FLAGS_OUT))
+ return -EINVAL;
+ }
+
+ /* Run all the scrubbers. */
+ for (i = 0, v = vhead->svh_vecs; i < vhead->svh_nr; i++, v++) {
+ struct xfs_scrub_metadata sm = {
+ .sm_type = v->sv_type,
+ .sm_flags = v->sv_flags,
+ .sm_ino = vhead->svh_ino,
+ .sm_gen = vhead->svh_gen,
+ .sm_agno = vhead->svh_agno,
+ };
+ struct timespec tv;
+
+ if (v->sv_type == XFS_SCRUB_TYPE_BARRIER) {
+ v->sv_ret = xfrog_scrubv_previous_failures(vhead, v);
+ if (v->sv_ret)
+ break;
+ continue;
+ }
+
+ v->sv_ret = xfrog_scrub_metadata(xfd, &sm);
+ v->sv_flags = sm.sm_flags;
+
+ if (vhead->svh_rest_us) {
+ tv.tv_sec = 0;
+ tv.tv_nsec = vhead->svh_rest_us * 1000;
+ nanosleep(&tv, NULL);
+ }
+ }
+
+ return 0;
+}
+
+/* Invoke the vectored scrub ioctl. */
+static int
+xfrog_scrubv_call(
+ struct xfs_fd *xfd,
+ struct xfs_scrub_vec_head *vhead)
+{
+ int ret;
+
+ ret = ioctl(xfd->fd, XFS_IOC_SCRUBV_METADATA, vhead);
+ if (ret)
+ return -errno;
+
+ return 0;
+}
+
+/* Invoke the vectored scrub ioctl. Returns zero or negative error code. */
+int
+xfrog_scrubv_metadata(
+ struct xfs_fd *xfd,
+ struct xfs_scrub_vec_head *vhead)
+{
+ int error = 0;
+
+ if (xfd->flags & XFROG_FLAG_SCRUB_FORCE_SINGLE)
+ goto try_single;
+
+ error = xfrog_scrubv_call(xfd, vhead);
+ if (error == 0 || (xfd->flags & XFROG_FLAG_SCRUB_FORCE_VECTOR))
+ return error;
+
+ /* If the vectored scrub ioctl wasn't found, force single mode. */
+ switch (error) {
+ case -EOPNOTSUPP:
+ case -ENOTTY:
+ xfd->flags |= XFROG_FLAG_SCRUB_FORCE_SINGLE;
+ break;
+ }
+
+try_single:
+ return xfrog_scrubv_fallback(xfd, vhead);
+}
diff --git a/libfrog/scrub.h b/libfrog/scrub.h
index 27230c62f71..43456230479 100644
--- a/libfrog/scrub.h
+++ b/libfrog/scrub.h
@@ -28,4 +28,8 @@ extern const struct xfrog_scrub_descr xfrog_scrubbers[XFS_SCRUB_TYPE_NR];
int xfrog_scrub_metadata(struct xfs_fd *xfd, struct xfs_scrub_metadata *meta);
+struct xfs_scrub_vec_head;
+
+int xfrog_scrubv_metadata(struct xfs_fd *xfd, struct xfs_scrub_vec_head *vhead);
+
#endif /* __LIBFROG_SCRUB_H__ */
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 03/10] xfs_io: support vectored scrub
2023-12-31 19:51 ` [PATCHSET 6/6] xfs_scrub: vectorize kernel calls Darrick J. Wong
2023-12-31 23:19 ` [PATCH 01/10] xfs: introduce vectored scrub mode Darrick J. Wong
2023-12-31 23:20 ` [PATCH 02/10] libfrog: support vectored scrub Darrick J. Wong
@ 2023-12-31 23:20 ` Darrick J. Wong
2023-12-31 23:20 ` [PATCH 04/10] xfs_scrub: split the scrub epilogue code into a separate function Darrick J. Wong
` (6 subsequent siblings)
9 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:20 UTC (permalink / raw)
To: djwong, cem; +Cc: linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
Create a new scrubv command to xfs_io to support the vectored scrub
ioctl.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
io/scrub.c | 371 +++++++++++++++++++++++++++++++++++++++++++++++------
man/man8/xfs_io.8 | 51 +++++++
2 files changed, 382 insertions(+), 40 deletions(-)
diff --git a/io/scrub.c b/io/scrub.c
index a77cd872fed..456d1594f22 100644
--- a/io/scrub.c
+++ b/io/scrub.c
@@ -12,10 +12,13 @@
#include "libfrog/paths.h"
#include "libfrog/fsgeom.h"
#include "libfrog/scrub.h"
+#include "libfrog/logging.h"
#include "io.h"
+#include "list.h"
static struct cmdinfo scrub_cmd;
static struct cmdinfo repair_cmd;
+static const struct cmdinfo scrubv_cmd;
static void
scrub_help(void)
@@ -197,31 +200,38 @@ parse_args(
return 0;
}
-static int
-scrub_f(
- int argc,
- char **argv)
+static void
+report_scrub_outcome(
+ uint32_t flags)
{
- struct xfs_scrub_metadata meta;
- int error;
-
- error = parse_args(argc, argv, &scrub_cmd, &meta);
- if (error)
- return error;
-
- error = ioctl(file->fd, XFS_IOC_SCRUB_METADATA, &meta);
- if (error)
- perror("scrub");
- if (meta.sm_flags & XFS_SCRUB_OFLAG_CORRUPT)
+ if (flags & XFS_SCRUB_OFLAG_CORRUPT)
printf(_("Corruption detected.\n"));
- if (meta.sm_flags & XFS_SCRUB_OFLAG_PREEN)
+ if (flags & XFS_SCRUB_OFLAG_PREEN)
printf(_("Optimization possible.\n"));
- if (meta.sm_flags & XFS_SCRUB_OFLAG_XFAIL)
+ if (flags & XFS_SCRUB_OFLAG_XFAIL)
printf(_("Cross-referencing failed.\n"));
- if (meta.sm_flags & XFS_SCRUB_OFLAG_XCORRUPT)
+ if (flags & XFS_SCRUB_OFLAG_XCORRUPT)
printf(_("Corruption detected during cross-referencing.\n"));
- if (meta.sm_flags & XFS_SCRUB_OFLAG_INCOMPLETE)
+ if (flags & XFS_SCRUB_OFLAG_INCOMPLETE)
printf(_("Scan was not complete.\n"));
+}
+
+static int
+scrub_f(
+ int argc,
+ char **argv)
+{
+ struct xfs_scrub_metadata meta;
+ int error;
+
+ error = parse_args(argc, argv, &scrub_cmd, &meta);
+ if (error)
+ return error;
+
+ error = ioctl(file->fd, XFS_IOC_SCRUB_METADATA, &meta);
+ if (error)
+ perror("scrub");
+ report_scrub_outcome(meta.sm_flags);
return 0;
}
@@ -239,6 +249,7 @@ scrub_init(void)
scrub_cmd.help = scrub_help;
add_command(&scrub_cmd);
+ add_command(&scrubv_cmd);
}
static void
@@ -267,34 +278,41 @@ repair_help(void)
printf("\n");
}
-static int
-repair_f(
- int argc,
- char **argv)
+static void
+report_repair_outcome(
+ uint32_t flags)
{
- struct xfs_scrub_metadata meta;
- int error;
-
- error = parse_args(argc, argv, &repair_cmd, &meta);
- if (error)
- return error;
- meta.sm_flags |= XFS_SCRUB_IFLAG_REPAIR;
-
- error = ioctl(file->fd, XFS_IOC_SCRUB_METADATA, &meta);
- if (error)
- perror("repair");
- if (meta.sm_flags & XFS_SCRUB_OFLAG_CORRUPT)
+ if (flags & XFS_SCRUB_OFLAG_CORRUPT)
printf(_("Corruption remains.\n"));
- if (meta.sm_flags & XFS_SCRUB_OFLAG_PREEN)
+ if (flags & XFS_SCRUB_OFLAG_PREEN)
printf(_("Optimization possible.\n"));
- if (meta.sm_flags & XFS_SCRUB_OFLAG_XFAIL)
+ if (flags & XFS_SCRUB_OFLAG_XFAIL)
printf(_("Cross-referencing failed.\n"));
- if (meta.sm_flags & XFS_SCRUB_OFLAG_XCORRUPT)
+ if (flags & XFS_SCRUB_OFLAG_XCORRUPT)
printf(_("Corruption still detected during cross-referencing.\n"));
- if (meta.sm_flags & XFS_SCRUB_OFLAG_INCOMPLETE)
+ if (flags & XFS_SCRUB_OFLAG_INCOMPLETE)
printf(_("Repair was not complete.\n"));
- if (meta.sm_flags & XFS_SCRUB_OFLAG_NO_REPAIR_NEEDED)
+ if (flags & XFS_SCRUB_OFLAG_NO_REPAIR_NEEDED)
printf(_("Metadata did not need repair or optimization.\n"));
+}
+
+static int
+repair_f(
+ int argc,
+ char **argv)
+{
+ struct xfs_scrub_metadata meta;
+ int error;
+
+ error = parse_args(argc, argv, &repair_cmd, &meta);
+ if (error)
+ return error;
+ meta.sm_flags |= XFS_SCRUB_IFLAG_REPAIR;
+
+ error = ioctl(file->fd, XFS_IOC_SCRUB_METADATA, &meta);
+ if (error)
+ perror("repair");
+ report_repair_outcome(meta.sm_flags);
return 0;
}
@@ -315,3 +333,276 @@ repair_init(void)
add_command(&repair_cmd);
}
+
+static void
+scrubv_help(void)
+{
+ printf(_(
+"\n"
+" Scrubs pieces of XFS filesystem metadata. The first argument is the group\n"
+" of metadata to examine. If the group is 'ag', the second parameter should\n"
+" be the AG number. If the group is 'inode', the second and third parameters\n"
+" should be the inode number and generation number to act upon; if these are\n"
+" omitted, the scrub is performed on the open file. If the group is 'fs',\n"
+" 'summary', or 'probe', there are no other parameters.\n"
+"\n"
+" Flags are -d for debug, and -r to allow repairs.\n"
+" -b NN will insert a scrub barrier after every NN scrubs, and -m sets the\n"
+" desired corruption mask in all barriers. -w pauses for some microseconds\n"
+" after each scrub call.\n"
+"\n"
+" Example:\n"
+" 'scrubv ag 3' - scrub all metadata in AG 3.\n"
+" 'scrubv ag 3 -b 2 -m 0x4' - scrub all metadata in AG 3, and use barriers\n"
+" every third scrub to exit early if there are optimizations.\n"
+" 'scrubv fs' - scrub all non-AG non-file metadata.\n"
+" 'scrubv inode' - scrub all metadata for the open file.\n"
+" 'scrubv inode 128 13525' - scrub all metadata for inode 128 gen 13525.\n"
+" 'scrubv probe' - check for presence of online scrub.\n"
+" 'scrubv summary' - scrub all summary metadata.\n"));
+}
+
+/* Fill out the scrub vectors for a group of scrubber (ag, ino, fs, summary) */
+static void
+scrubv_fill_group(
+ struct xfs_scrub_vec_head *vhead,
+ int barrier_interval,
+ __u32 barrier_mask,
+ enum xfrog_scrub_group group)
+{
+ const struct xfrog_scrub_descr *d;
+ unsigned int i;
+
+ for (i = 0, d = xfrog_scrubbers; i < XFS_SCRUB_TYPE_NR; i++, d++) {
+ if (d->group != group)
+ continue;
+ vhead->svh_vecs[vhead->svh_nr++].sv_type = i;
+
+ if (barrier_interval &&
+ vhead->svh_nr % (barrier_interval + 1) == 0) {
+ struct xfs_scrub_vec *v;
+
+ v = &vhead->svh_vecs[vhead->svh_nr++];
+ v->sv_flags = barrier_mask;
+ v->sv_type = XFS_SCRUB_TYPE_BARRIER;
+ }
+ }
+}
+
+/* Declare a structure big enough to handle all scrub types + barriers */
+struct scrubv_head {
+ struct xfs_scrub_vec_head head;
+ struct xfs_scrub_vec __vecs[XFS_SCRUB_TYPE_NR * 2];
+};
+
+
+static int
+scrubv_f(
+ int argc,
+ char **argv)
+{
+ struct scrubv_head bighead = { };
+ struct xfs_fd xfd = XFS_FD_INIT(file->fd);
+ struct xfs_scrub_vec_head *vhead = &bighead.head;
+ struct xfs_scrub_vec *v;
+ uint32_t flags = 0;
+ __u32 barrier_mask = XFS_SCRUB_OFLAG_CORRUPT;
+ enum xfrog_scrub_group group;
+ bool debug = false;
+ int version = -1;
+ int barrier_interval = 0;
+ int rest_us = 0;
+ int c;
+ int error;
+
+ while ((c = getopt(argc, argv, "b:dm:rv:w:")) != EOF) {
+ switch (c) {
+ case 'b':
+ barrier_interval = atoi(optarg);
+ if (barrier_interval < 0) {
+ fprintf(stderr,
+ _("Negative barrier interval makes no sense.\n"));
+ exitcode = 1;
+ return command_usage(&scrubv_cmd);
+ }
+ break;
+ case 'd':
+ debug = true;
+ break;
+ case 'm':
+ barrier_mask = strtoul(optarg, NULL, 0);
+ break;
+ case 'r':
+ flags |= XFS_SCRUB_IFLAG_REPAIR;
+ break;
+ case 'v':
+ if (!strcmp("single", optarg)) {
+ version = 0;
+ } else if (!strcmp("vector", optarg)) {
+ version = 1;
+ } else {
+ fprintf(stderr,
+ _("API version must be 'single' or 'vector'.\n"));
+ exitcode = 1;
+ return command_usage(&scrubv_cmd);
+ }
+ break;
+ case 'w':
+ rest_us = atoi(optarg);
+ if (rest_us < 0) {
+ fprintf(stderr,
+ _("Rest time must be positive.\n"));
+ exitcode = 1;
+ return command_usage(&scrubv_cmd);
+ }
+ break;
+ default:
+ exitcode = 1;
+ return command_usage(&scrubv_cmd);
+ }
+ }
+ if (optind > argc - 1) {
+ fprintf(stderr,
+ _("Must have at least one positional argument.\n"));
+ exitcode = 1;
+ return command_usage(&scrubv_cmd);
+ }
+
+ if ((flags & XFS_SCRUB_IFLAG_REPAIR) && !expert) {
+ printf(_("Repair flag requires expert mode.\n"));
+ return 1;
+ }
+
+ vhead->svh_rest_us = rest_us;
+ for (c = 0, v = vhead->svh_vecs; c < vhead->svh_nr; c++, v++)
+ v->sv_flags = flags;
+
+ /* Extract group and domain information from cmdline. */
+ if (!strcmp(argv[optind], "probe"))
+ group = XFROG_SCRUB_GROUP_NONE;
+ else if (!strcmp(argv[optind], "agheader"))
+ group = XFROG_SCRUB_GROUP_AGHEADER;
+ else if (!strcmp(argv[optind], "ag"))
+ group = XFROG_SCRUB_GROUP_PERAG;
+ else if (!strcmp(argv[optind], "fs"))
+ group = XFROG_SCRUB_GROUP_FS;
+ else if (!strcmp(argv[optind], "inode"))
+ group = XFROG_SCRUB_GROUP_INODE;
+ else if (!strcmp(argv[optind], "iscan"))
+ group = XFROG_SCRUB_GROUP_ISCAN;
+ else if (!strcmp(argv[optind], "summary"))
+ group = XFROG_SCRUB_GROUP_SUMMARY;
+ else {
+ printf(_("Unknown group '%s'.\n"), argv[optind]);
+ exitcode = 1;
+ return command_usage(&scrubv_cmd);
+ }
+ optind++;
+
+ switch (group) {
+ case XFROG_SCRUB_GROUP_INODE:
+ if (!parse_inode(argc, argv, optind, &vhead->svh_ino,
+ &vhead->svh_gen)) {
+ exitcode = 1;
+ return command_usage(&scrubv_cmd);
+ }
+ break;
+ case XFROG_SCRUB_GROUP_AGHEADER:
+ case XFROG_SCRUB_GROUP_PERAG:
+ if (!parse_agno(argc, argv, optind, &vhead->svh_agno)) {
+ exitcode = 1;
+ return command_usage(&scrubv_cmd);
+ }
+ break;
+ case XFROG_SCRUB_GROUP_FS:
+ case XFROG_SCRUB_GROUP_SUMMARY:
+ case XFROG_SCRUB_GROUP_ISCAN:
+ case XFROG_SCRUB_GROUP_NONE:
+ if (!parse_none(argc, optind)) {
+ exitcode = 1;
+ return command_usage(&scrubv_cmd);
+ }
+ break;
+ default:
+ ASSERT(0);
+ break;
+ }
+ scrubv_fill_group(vhead, barrier_interval, barrier_mask, group);
+ assert(vhead->svh_nr < ARRAY_SIZE(bighead.__vecs));
+
+ error = -xfd_prepare_geometry(&xfd);
+ if (error) {
+ xfrog_perror(error, "xfd_prepare_geometry");
+ exitcode = 1;
+ return 0;
+ }
+
+ switch (version) {
+ case 0:
+ xfd.flags |= XFROG_FLAG_SCRUB_FORCE_SINGLE;
+ break;
+ case 1:
+ xfd.flags |= XFROG_FLAG_SCRUB_FORCE_VECTOR;
+ break;
+ default:
+ break;
+ }
+
+ error = -xfrog_scrubv_metadata(&xfd, vhead);
+ if (error) {
+ xfrog_perror(error, "xfrog_scrub_many");
+ exitcode = 1;
+ return 0;
+ }
+
+ /* Figure out what happened. */
+ for (c = 0, v = vhead->svh_vecs; debug && c < vhead->svh_nr; c++, v++) {
+ const char *type;
+
+ if (v->sv_type == XFS_SCRUB_TYPE_BARRIER)
+ type = _("barrier");
+ else
+ type = _(xfrog_scrubbers[v->sv_type].descr);
+ printf(_("[%02u] %-25s: flags 0x%x ret %d\n"), c, type,
+ v->sv_flags, v->sv_ret);
+ }
+
+ /* Figure out what happened. */
+ for (c = 0, v = vhead->svh_vecs; c < vhead->svh_nr; c++, v++) {
+ /* Report barrier failures. */
+ if (v->sv_type == XFS_SCRUB_TYPE_BARRIER) {
+ if (v->sv_ret) {
+ printf(_("barrier: FAILED\n"));
+ break;
+ }
+ continue;
+ }
+
+ printf("%s: ", _(xfrog_scrubbers[v->sv_type].descr));
+ switch (v->sv_ret) {
+ case 0:
+ break;
+ default:
+ printf("%s\n", strerror(-v->sv_ret));
+ continue;
+ }
+ if (!(v->sv_flags & XFS_SCRUB_FLAGS_OUT))
+ printf(_("OK.\n"));
+ else if (v->sv_flags & XFS_SCRUB_IFLAG_REPAIR)
+ report_repair_outcome(v->sv_flags);
+ else
+ report_scrub_outcome(v->sv_flags);
+ }
+
+ return 0;
+}
+
+static const struct cmdinfo scrubv_cmd = {
+ .name = "scrubv",
+ .cfunc = scrubv_f,
+ .argmin = 1,
+ .argmax = -1,
+ .flags = CMD_NOMAP_OK,
+ .oneline = N_("vectored metadata scrub"),
+ .help = scrubv_help,
+};
diff --git a/man/man8/xfs_io.8 b/man/man8/xfs_io.8
index aa9907c1e5e..5a6b2724504 100644
--- a/man/man8/xfs_io.8
+++ b/man/man8/xfs_io.8
@@ -1428,6 +1428,57 @@ inode number and generation number are specified.
.RE
.PD
.TP
+.BI "scrubv [ \-b NN ] [ \-d ] [ \-f ] [ \-r ] [ \-v NN ] [ \-w ms ] " group " [ " agnumber " | " "ino" " " "gen" " ]"
+Scrub a bunch of internal XFS filesystem metadata.
+The
+.BI group
+parameter specifies which group of metadata to scrub.
+Valid groups are
+.IR ag ", " agheader ", " inode ", " iscan ", " fs ", " probe ", " rtgroup ", or " summary .
+
+For
+.BR ag " and " agheader
+metadata, one AG number must be specified.
+For
+.B inode
+metadata, the scrub is applied to the open file unless the
+inode number and generation number are specified.
+For
+.B rtgroup
+metadata, one rt group number must be specified.
+
+.RS 1.0i
+.PD 0
+.TP
+.BI "\-b " NN
+Inject scrub barriers into the vector stream at the given interval.
+Barriers abort vector processing if any previous scrub function found
+corruption.
+.TP
+.BI \-d
+Enables debug mode.
+.TP
+.BI \-f
+Permit the kernel to freeze the filesystem in order to scrub or repair.
+.TP
+.BI \-r
+Repair metadata if corruptions are found.
+This option requires expert mode.
+.TP
+.BI "\-v " NN
+Force a particular API version.
+.B single
+selects XFS_SCRUB_METADATA (one-by-one).
+.B vector
+selects XFS_SCRUBV_METADATA (vectored).
+If no option is specified, vector mode will be used, with a fallback to single
+mode if the kernel doesn't recognize the vector mode ioctl.
+.TP
+.BI "\-w " us
+Wait the given number of microseconds between each scrub function.
+.RE
+.PD
+.TP
.BI "repair " type " [ " agnumber " | " "ino" " " "gen" " ]"
Repair internal XFS filesystem metadata. The
.BI type
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 04/10] xfs_scrub: split the scrub epilogue code into a separate function
2023-12-31 19:51 ` [PATCHSET 6/6] xfs_scrub: vectorize kernel calls Darrick J. Wong
` (2 preceding siblings ...)
2023-12-31 23:20 ` [PATCH 03/10] xfs_io: " Darrick J. Wong
@ 2023-12-31 23:20 ` Darrick J. Wong
2023-12-31 23:21 ` [PATCH 05/10] xfs_scrub: split the repair " Darrick J. Wong
` (5 subsequent siblings)
9 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:20 UTC (permalink / raw)
To: djwong, cem; +Cc: linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
Move all the code that updates the internal state in response to a scrub
ioctl() call completion into a separate function. This will help with
vectorizing scrub calls later on.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
scrub/scrub.c | 52 ++++++++++++++++++++++++++++++++++++++--------------
1 file changed, 38 insertions(+), 14 deletions(-)
diff --git a/scrub/scrub.c b/scrub/scrub.c
index 1b0609e7418..c4b4367e458 100644
--- a/scrub/scrub.c
+++ b/scrub/scrub.c
@@ -22,6 +22,10 @@
#include "descr.h"
#include "scrub_private.h"
+static int scrub_epilogue(struct scrub_ctx *ctx, struct descr *dsc,
+ struct scrub_item *sri, struct xfs_scrub_metadata *meta,
+ int error);
+
/* Online scrub and repair wrappers. */
/* Format a scrub description. */
@@ -117,12 +121,32 @@ xfs_check_metadata(
dbg_printf("check %s flags %xh\n", descr_render(&dsc), meta.sm_flags);
error = -xfrog_scrub_metadata(xfdp, &meta);
+ return scrub_epilogue(ctx, &dsc, sri, &meta, error);
+}
+
+/*
+ * Update all internal state after a scrub ioctl call.
+ * Returns 0 for success, or ECANCELED to abort the program.
+ */
+static int
+scrub_epilogue(
+ struct scrub_ctx *ctx,
+ struct descr *dsc,
+ struct scrub_item *sri,
+ struct xfs_scrub_metadata *meta,
+ int error)
+{
+ unsigned int scrub_type = meta->sm_type;
+ enum xfrog_scrub_group group;
+
+ group = xfrog_scrubbers[scrub_type].group;
+
switch (error) {
case 0:
/* No operational errors encountered. */
if (!sri->sri_revalidate &&
debug_tweak_on("XFS_SCRUB_FORCE_REPAIR"))
- meta.sm_flags |= XFS_SCRUB_OFLAG_CORRUPT;
+ meta->sm_flags |= XFS_SCRUB_OFLAG_CORRUPT;
break;
case ENOENT:
/* Metadata not present, just skip it. */
@@ -130,13 +154,13 @@ xfs_check_metadata(
return 0;
case ESHUTDOWN:
/* FS already crashed, give up. */
- str_error(ctx, descr_render(&dsc),
+ str_error(ctx, descr_render(dsc),
_("Filesystem is shut down, aborting."));
return ECANCELED;
case EIO:
case ENOMEM:
/* Abort on I/O errors or insufficient memory. */
- str_liberror(ctx, error, descr_render(&dsc));
+ str_liberror(ctx, error, descr_render(dsc));
return ECANCELED;
case EDEADLOCK:
case EBUSY:
@@ -152,7 +176,7 @@ _("Filesystem is shut down, aborting."));
return 0;
default:
/* Operational error. Log it and move on. */
- str_liberror(ctx, error, descr_render(&dsc));
+ str_liberror(ctx, error, descr_render(dsc));
scrub_item_clean_state(sri, scrub_type);
return 0;
}
@@ -163,27 +187,27 @@ _("Filesystem is shut down, aborting."));
* we'll try the scan again, just in case the fs was busy.
* Only retry so many times.
*/
- if (want_retry(&meta) && scrub_item_schedule_retry(sri, scrub_type))
+ if (want_retry(meta) && scrub_item_schedule_retry(sri, scrub_type))
return 0;
/* Complain about incomplete or suspicious metadata. */
- scrub_warn_incomplete_scrub(ctx, &dsc, &meta);
+ scrub_warn_incomplete_scrub(ctx, dsc, meta);
/*
* If we need repairs or there were discrepancies, schedule a
* repair if desired, otherwise complain.
*/
- if (is_corrupt(&meta) || xref_disagrees(&meta)) {
+ if (is_corrupt(meta) || xref_disagrees(meta)) {
if (ctx->mode != SCRUB_MODE_REPAIR) {
/* Dry-run mode, so log an error and forget it. */
- str_corrupt(ctx, descr_render(&dsc),
+ str_corrupt(ctx, descr_render(dsc),
_("Repairs are required."));
scrub_item_clean_state(sri, scrub_type);
return 0;
}
/* Schedule repairs. */
- scrub_item_save_state(sri, scrub_type, meta.sm_flags);
+ scrub_item_save_state(sri, scrub_type, meta->sm_flags);
return 0;
}
@@ -191,12 +215,12 @@ _("Repairs are required."));
* If we could optimize, schedule a repair if desired,
* otherwise complain.
*/
- if (is_unoptimized(&meta)) {
+ if (is_unoptimized(meta)) {
if (ctx->mode == SCRUB_MODE_DRY_RUN) {
/* Dry-run mode, so log an error and forget it. */
if (group != XFROG_SCRUB_GROUP_INODE) {
/* AG or FS metadata, always warn. */
- str_info(ctx, descr_render(&dsc),
+ str_info(ctx, descr_render(dsc),
_("Optimization is possible."));
} else if (!ctx->preen_triggers[scrub_type]) {
/* File metadata, only warn once per type. */
@@ -210,7 +234,7 @@ _("Optimization is possible."));
}
/* Schedule optimizations. */
- scrub_item_save_state(sri, scrub_type, meta.sm_flags);
+ scrub_item_save_state(sri, scrub_type, meta->sm_flags);
return 0;
}
@@ -221,8 +245,8 @@ _("Optimization is possible."));
* re-examine the object as repairs progress to see if the kernel will
* deem it completely consistent at some point.
*/
- if (xref_failed(&meta) && ctx->mode == SCRUB_MODE_REPAIR) {
- scrub_item_save_state(sri, scrub_type, meta.sm_flags);
+ if (xref_failed(meta) && ctx->mode == SCRUB_MODE_REPAIR) {
+ scrub_item_save_state(sri, scrub_type, meta->sm_flags);
return 0;
}
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 05/10] xfs_scrub: split the repair epilogue code into a separate function
2023-12-31 19:51 ` [PATCHSET 6/6] xfs_scrub: vectorize kernel calls Darrick J. Wong
` (3 preceding siblings ...)
2023-12-31 23:20 ` [PATCH 04/10] xfs_scrub: split the scrub epilogue code into a separate function Darrick J. Wong
@ 2023-12-31 23:21 ` Darrick J. Wong
2023-12-31 23:21 ` [PATCH 06/10] xfs_scrub: convert scrub and repair epilogues to use xfs_scrub_vec Darrick J. Wong
` (4 subsequent siblings)
9 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:21 UTC (permalink / raw)
To: djwong, cem; +Cc: linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
Move all the code that updates the internal state in response to a
repair ioctl() call completion into a separate function. This will help
with vectorizing repair calls later on.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
scrub/repair.c | 69 +++++++++++++++++++++++++++++++++++++-------------------
1 file changed, 45 insertions(+), 24 deletions(-)
diff --git a/scrub/repair.c b/scrub/repair.c
index 4fed86134ed..0b99e335191 100644
--- a/scrub/repair.c
+++ b/scrub/repair.c
@@ -20,6 +20,11 @@
#include "descr.h"
#include "scrub_private.h"
+static int repair_epilogue(struct scrub_ctx *ctx, struct descr *dsc,
+ struct scrub_item *sri, unsigned int repair_flags,
+ struct xfs_scrub_metadata *oldm,
+ struct xfs_scrub_metadata *meta, int error);
+
/* General repair routines. */
/*
@@ -133,6 +138,22 @@ xfs_repair_metadata(
_("Attempting optimization."));
error = -xfrog_scrub_metadata(xfdp, &meta);
+ return repair_epilogue(ctx, &dsc, sri, repair_flags, &oldm, &meta,
+ error);
+}
+
+static int
+repair_epilogue(
+ struct scrub_ctx *ctx,
+ struct descr *dsc,
+ struct scrub_item *sri,
+ unsigned int repair_flags,
+ struct xfs_scrub_metadata *oldm,
+ struct xfs_scrub_metadata *meta,
+ int error)
+{
+ unsigned int scrub_type = meta->sm_type;
+
switch (error) {
case 0:
/* No operational errors encountered. */
@@ -141,12 +162,12 @@ xfs_repair_metadata(
case EBUSY:
/* Filesystem is busy, try again later. */
if (debug || verbose)
- str_info(ctx, descr_render(&dsc),
+ str_info(ctx, descr_render(dsc),
_("Filesystem is busy, deferring repair."));
return 0;
case ESHUTDOWN:
/* Filesystem is already shut down, abort. */
- str_error(ctx, descr_render(&dsc),
+ str_error(ctx, descr_render(dsc),
_("Filesystem is shut down, aborting."));
return ECANCELED;
case ENOTTY:
@@ -157,7 +178,7 @@ _("Filesystem is shut down, aborting."));
* how to perform the repair, don't requeue the request. Mark
* it done and move on.
*/
- if (is_unoptimized(&oldm) ||
+ if (is_unoptimized(oldm) ||
debug_tweak_on("XFS_SCRUB_FORCE_REPAIR")) {
scrub_item_clean_state(sri, scrub_type);
return 0;
@@ -175,14 +196,14 @@ _("Filesystem is shut down, aborting."));
fallthrough;
case EINVAL:
/* Kernel doesn't know how to repair this? */
- str_corrupt(ctx, descr_render(&dsc),
+ str_corrupt(ctx, descr_render(dsc),
_("Don't know how to fix; offline repair required."));
scrub_item_clean_state(sri, scrub_type);
return 0;
case EROFS:
/* Read-only filesystem, can't fix. */
- if (verbose || debug || needs_repair(&oldm))
- str_error(ctx, descr_render(&dsc),
+ if (verbose || debug || needs_repair(oldm))
+ str_error(ctx, descr_render(dsc),
_("Read-only filesystem; cannot make changes."));
return ECANCELED;
case ENOENT:
@@ -192,7 +213,7 @@ _("Read-only filesystem; cannot make changes."));
case ENOMEM:
case ENOSPC:
/* Don't care if preen fails due to low resources. */
- if (is_unoptimized(&oldm) && !needs_repair(&oldm)) {
+ if (is_unoptimized(oldm) && !needs_repair(oldm)) {
scrub_item_clean_state(sri, scrub_type);
return 0;
}
@@ -207,7 +228,7 @@ _("Read-only filesystem; cannot make changes."));
*/
if (!(repair_flags & XRM_FINAL_WARNING))
return 0;
- str_liberror(ctx, error, descr_render(&dsc));
+ str_liberror(ctx, error, descr_render(dsc));
scrub_item_clean_state(sri, scrub_type);
return 0;
}
@@ -218,12 +239,12 @@ _("Read-only filesystem; cannot make changes."));
* the repair again, just in case the fs was busy. Only retry so many
* times.
*/
- if (want_retry(&meta) && scrub_item_schedule_retry(sri, scrub_type))
+ if (want_retry(meta) && scrub_item_schedule_retry(sri, scrub_type))
return 0;
if (repair_flags & XRM_FINAL_WARNING)
- scrub_warn_incomplete_scrub(ctx, &dsc, &meta);
- if (needs_repair(&meta) || is_incomplete(&meta)) {
+ scrub_warn_incomplete_scrub(ctx, dsc, meta);
+ if (needs_repair(meta) || is_incomplete(meta)) {
/*
* Still broken; if we've been told not to complain then we
* just requeue this and try again later. Otherwise we
@@ -231,9 +252,9 @@ _("Read-only filesystem; cannot make changes."));
*/
if (!(repair_flags & XRM_FINAL_WARNING))
return 0;
- str_corrupt(ctx, descr_render(&dsc),
+ str_corrupt(ctx, descr_render(dsc),
_("Repair unsuccessful; offline repair required."));
- } else if (xref_failed(&meta)) {
+ } else if (xref_failed(meta)) {
/*
* This metadata object itself looks ok, but we still noticed
* inconsistencies when comparing it with the other filesystem
@@ -242,31 +263,31 @@ _("Repair unsuccessful; offline repair required."));
* reverify the cross-referencing as repairs progress.
*/
if (repair_flags & XRM_FINAL_WARNING) {
- str_info(ctx, descr_render(&dsc),
+ str_info(ctx, descr_render(dsc),
_("Seems correct but cross-referencing failed; offline repair recommended."));
} else {
if (verbose)
- str_info(ctx, descr_render(&dsc),
+ str_info(ctx, descr_render(dsc),
_("Seems correct but cross-referencing failed; will keep checking."));
return 0;
}
- } else if (meta.sm_flags & XFS_SCRUB_OFLAG_NO_REPAIR_NEEDED) {
+ } else if (meta->sm_flags & XFS_SCRUB_OFLAG_NO_REPAIR_NEEDED) {
if (verbose)
- str_info(ctx, descr_render(&dsc),
+ str_info(ctx, descr_render(dsc),
_("No modification needed."));
} else {
/* Clean operation, no corruption detected. */
- if (is_corrupt(&oldm))
- record_repair(ctx, descr_render(&dsc),
+ if (is_corrupt(oldm))
+ record_repair(ctx, descr_render(dsc),
_("Repairs successful."));
- else if (xref_disagrees(&oldm))
- record_repair(ctx, descr_render(&dsc),
+ else if (xref_disagrees(oldm))
+ record_repair(ctx, descr_render(dsc),
_("Repairs successful after discrepancy in cross-referencing."));
- else if (xref_failed(&oldm))
- record_repair(ctx, descr_render(&dsc),
+ else if (xref_failed(oldm))
+ record_repair(ctx, descr_render(dsc),
_("Repairs successful after cross-referencing failure."));
else
- record_preen(ctx, descr_render(&dsc),
+ record_preen(ctx, descr_render(dsc),
_("Optimization successful."));
}
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 06/10] xfs_scrub: convert scrub and repair epilogues to use xfs_scrub_vec
2023-12-31 19:51 ` [PATCHSET 6/6] xfs_scrub: vectorize kernel calls Darrick J. Wong
` (4 preceding siblings ...)
2023-12-31 23:21 ` [PATCH 05/10] xfs_scrub: split the repair " Darrick J. Wong
@ 2023-12-31 23:21 ` Darrick J. Wong
2023-12-31 23:21 ` [PATCH 07/10] xfs_scrub: vectorize scrub calls Darrick J. Wong
` (3 subsequent siblings)
9 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:21 UTC (permalink / raw)
To: djwong, cem; +Cc: linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
Convert the scrub and repair epilogue code to pass around xfs_scrub_vecs
as we prepare for vectorized operation.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
scrub/repair.c | 35 ++++++++++++++++++-----------------
scrub/scrub.c | 27 ++++++++++++++-------------
scrub/scrub_private.h | 34 +++++++++++++++++-----------------
3 files changed, 49 insertions(+), 47 deletions(-)
diff --git a/scrub/repair.c b/scrub/repair.c
index 0b99e335191..7a710a159e6 100644
--- a/scrub/repair.c
+++ b/scrub/repair.c
@@ -22,8 +22,8 @@
static int repair_epilogue(struct scrub_ctx *ctx, struct descr *dsc,
struct scrub_item *sri, unsigned int repair_flags,
- struct xfs_scrub_metadata *oldm,
- struct xfs_scrub_metadata *meta, int error);
+ const struct xfs_scrub_vec *oldm,
+ const struct xfs_scrub_vec *meta);
/* General repair routines. */
@@ -93,10 +93,9 @@ xfs_repair_metadata(
unsigned int repair_flags)
{
struct xfs_scrub_metadata meta = { 0 };
- struct xfs_scrub_metadata oldm;
+ struct xfs_scrub_vec oldm, vec;
DEFINE_DESCR(dsc, ctx, format_scrub_descr);
bool repair_only;
- int error;
/*
* If the caller boosted the priority of this scrub type on behalf of a
@@ -124,22 +123,24 @@ xfs_repair_metadata(
break;
}
- if (!is_corrupt(&meta) && repair_only)
+ vec.sv_type = scrub_type;
+ vec.sv_flags = sri->sri_state[scrub_type] & SCRUB_ITEM_REPAIR_ANY;
+ memcpy(&oldm, &vec, sizeof(struct xfs_scrub_vec));
+ if (!is_corrupt(&vec) && repair_only)
return 0;
- memcpy(&oldm, &meta, sizeof(oldm));
- oldm.sm_flags = sri->sri_state[scrub_type] & SCRUB_ITEM_REPAIR_ANY;
- descr_set(&dsc, &oldm);
+ descr_set(&dsc, &meta);
- if (needs_repair(&oldm))
+ if (needs_repair(&vec))
str_info(ctx, descr_render(&dsc), _("Attempting repair."));
else if (debug || verbose)
str_info(ctx, descr_render(&dsc),
_("Attempting optimization."));
- error = -xfrog_scrub_metadata(xfdp, &meta);
- return repair_epilogue(ctx, &dsc, sri, repair_flags, &oldm, &meta,
- error);
+ vec.sv_ret = xfrog_scrub_metadata(xfdp, &meta);
+ vec.sv_flags = meta.sm_flags;
+
+ return repair_epilogue(ctx, &dsc, sri, repair_flags, &oldm, &vec);
}
static int
@@ -148,11 +149,11 @@ repair_epilogue(
struct descr *dsc,
struct scrub_item *sri,
unsigned int repair_flags,
- struct xfs_scrub_metadata *oldm,
- struct xfs_scrub_metadata *meta,
- int error)
+ const struct xfs_scrub_vec *oldm,
+ const struct xfs_scrub_vec *meta)
{
- unsigned int scrub_type = meta->sm_type;
+ unsigned int scrub_type = meta->sv_type;
+ int error = -meta->sv_ret;
switch (error) {
case 0:
@@ -271,7 +272,7 @@ _("Repair unsuccessful; offline repair required."));
_("Seems correct but cross-referencing failed; will keep checking."));
return 0;
}
- } else if (meta->sm_flags & XFS_SCRUB_OFLAG_NO_REPAIR_NEEDED) {
+ } else if (meta->sv_flags & XFS_SCRUB_OFLAG_NO_REPAIR_NEEDED) {
if (verbose)
str_info(ctx, descr_render(dsc),
_("No modification needed."));
diff --git a/scrub/scrub.c b/scrub/scrub.c
index c4b4367e458..2fb2293558e 100644
--- a/scrub/scrub.c
+++ b/scrub/scrub.c
@@ -23,8 +23,7 @@
#include "scrub_private.h"
static int scrub_epilogue(struct scrub_ctx *ctx, struct descr *dsc,
- struct scrub_item *sri, struct xfs_scrub_metadata *meta,
- int error);
+ struct scrub_item *sri, struct xfs_scrub_vec *vec);
/* Online scrub and repair wrappers. */
@@ -62,7 +61,7 @@ void
scrub_warn_incomplete_scrub(
struct scrub_ctx *ctx,
struct descr *dsc,
- struct xfs_scrub_metadata *meta)
+ const struct xfs_scrub_vec *meta)
{
if (is_incomplete(meta))
str_info(ctx, descr_render(dsc), _("Check incomplete."));
@@ -91,8 +90,8 @@ xfs_check_metadata(
{
DEFINE_DESCR(dsc, ctx, format_scrub_descr);
struct xfs_scrub_metadata meta = { };
+ struct xfs_scrub_vec vec;
enum xfrog_scrub_group group;
- int error;
background_sleep();
@@ -120,8 +119,10 @@ xfs_check_metadata(
dbg_printf("check %s flags %xh\n", descr_render(&dsc), meta.sm_flags);
- error = -xfrog_scrub_metadata(xfdp, &meta);
- return scrub_epilogue(ctx, &dsc, sri, &meta, error);
+ vec.sv_ret = xfrog_scrub_metadata(xfdp, &meta);
+ vec.sv_type = scrub_type;
+ vec.sv_flags = meta.sm_flags;
+ return scrub_epilogue(ctx, &dsc, sri, &vec);
}
/*
@@ -133,11 +134,11 @@ scrub_epilogue(
struct scrub_ctx *ctx,
struct descr *dsc,
struct scrub_item *sri,
- struct xfs_scrub_metadata *meta,
- int error)
+ struct xfs_scrub_vec *meta)
{
- unsigned int scrub_type = meta->sm_type;
+ unsigned int scrub_type = meta->sv_type;
enum xfrog_scrub_group group;
+ int error = -meta->sv_ret;
group = xfrog_scrubbers[scrub_type].group;
@@ -146,7 +147,7 @@ scrub_epilogue(
/* No operational errors encountered. */
if (!sri->sri_revalidate &&
debug_tweak_on("XFS_SCRUB_FORCE_REPAIR"))
- meta->sm_flags |= XFS_SCRUB_OFLAG_CORRUPT;
+ meta->sv_flags |= XFS_SCRUB_OFLAG_CORRUPT;
break;
case ENOENT:
/* Metadata not present, just skip it. */
@@ -207,7 +208,7 @@ _("Repairs are required."));
}
/* Schedule repairs. */
- scrub_item_save_state(sri, scrub_type, meta->sm_flags);
+ scrub_item_save_state(sri, scrub_type, meta->sv_flags);
return 0;
}
@@ -234,7 +235,7 @@ _("Optimization is possible."));
}
/* Schedule optimizations. */
- scrub_item_save_state(sri, scrub_type, meta->sm_flags);
+ scrub_item_save_state(sri, scrub_type, meta->sv_flags);
return 0;
}
@@ -246,7 +247,7 @@ _("Optimization is possible."));
* deem it completely consistent at some point.
*/
if (xref_failed(meta) && ctx->mode == SCRUB_MODE_REPAIR) {
- scrub_item_save_state(sri, scrub_type, meta->sm_flags);
+ scrub_item_save_state(sri, scrub_type, meta->sv_flags);
return 0;
}
diff --git a/scrub/scrub_private.h b/scrub/scrub_private.h
index bcfabda16be..98a9238f2aa 100644
--- a/scrub/scrub_private.h
+++ b/scrub/scrub_private.h
@@ -13,40 +13,40 @@ int format_scrub_descr(struct scrub_ctx *ctx, char *buf, size_t buflen,
/* Predicates for scrub flag state. */
-static inline bool is_corrupt(struct xfs_scrub_metadata *sm)
+static inline bool is_corrupt(const struct xfs_scrub_vec *sv)
{
- return sm->sm_flags & XFS_SCRUB_OFLAG_CORRUPT;
+ return sv->sv_flags & XFS_SCRUB_OFLAG_CORRUPT;
}
-static inline bool is_unoptimized(struct xfs_scrub_metadata *sm)
+static inline bool is_unoptimized(const struct xfs_scrub_vec *sv)
{
- return sm->sm_flags & XFS_SCRUB_OFLAG_PREEN;
+ return sv->sv_flags & XFS_SCRUB_OFLAG_PREEN;
}
-static inline bool xref_failed(struct xfs_scrub_metadata *sm)
+static inline bool xref_failed(const struct xfs_scrub_vec *sv)
{
- return sm->sm_flags & XFS_SCRUB_OFLAG_XFAIL;
+ return sv->sv_flags & XFS_SCRUB_OFLAG_XFAIL;
}
-static inline bool xref_disagrees(struct xfs_scrub_metadata *sm)
+static inline bool xref_disagrees(const struct xfs_scrub_vec *sv)
{
- return sm->sm_flags & XFS_SCRUB_OFLAG_XCORRUPT;
+ return sv->sv_flags & XFS_SCRUB_OFLAG_XCORRUPT;
}
-static inline bool is_incomplete(struct xfs_scrub_metadata *sm)
+static inline bool is_incomplete(const struct xfs_scrub_vec *sv)
{
- return sm->sm_flags & XFS_SCRUB_OFLAG_INCOMPLETE;
+ return sv->sv_flags & XFS_SCRUB_OFLAG_INCOMPLETE;
}
-static inline bool is_suspicious(struct xfs_scrub_metadata *sm)
+static inline bool is_suspicious(const struct xfs_scrub_vec *sv)
{
- return sm->sm_flags & XFS_SCRUB_OFLAG_WARNING;
+ return sv->sv_flags & XFS_SCRUB_OFLAG_WARNING;
}
/* Should we fix it? */
-static inline bool needs_repair(struct xfs_scrub_metadata *sm)
+static inline bool needs_repair(const struct xfs_scrub_vec *sv)
{
- return is_corrupt(sm) || xref_disagrees(sm);
+ return is_corrupt(sv) || xref_disagrees(sv);
}
/*
@@ -54,13 +54,13 @@ static inline bool needs_repair(struct xfs_scrub_metadata *sm)
* scan/repair; or if there were cross-referencing problems but the object was
* not obviously corrupt.
*/
-static inline bool want_retry(struct xfs_scrub_metadata *sm)
+static inline bool want_retry(const struct xfs_scrub_vec *sv)
{
- return is_incomplete(sm) || (xref_disagrees(sm) && !is_corrupt(sm));
+ return is_incomplete(sv) || (xref_disagrees(sv) && !is_corrupt(sv));
}
void scrub_warn_incomplete_scrub(struct scrub_ctx *ctx, struct descr *dsc,
- struct xfs_scrub_metadata *meta);
+ const struct xfs_scrub_vec *meta);
/* Scrub item functions */
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 07/10] xfs_scrub: vectorize scrub calls
2023-12-31 19:51 ` [PATCHSET 6/6] xfs_scrub: vectorize kernel calls Darrick J. Wong
` (5 preceding siblings ...)
2023-12-31 23:21 ` [PATCH 06/10] xfs_scrub: convert scrub and repair epilogues to use xfs_scrub_vec Darrick J. Wong
@ 2023-12-31 23:21 ` Darrick J. Wong
2023-12-31 23:21 ` [PATCH 08/10] xfs_scrub: vectorize repair calls Darrick J. Wong
` (2 subsequent siblings)
9 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:21 UTC (permalink / raw)
To: djwong, cem; +Cc: linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
Use the new vectorized kernel scrub calls to reduce the overhead of
checking metadata.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
scrub/phase1.c | 2
scrub/scrub.c | 261 +++++++++++++++++++++++++++++++++++--------------
scrub/scrub.h | 2
scrub/scrub_private.h | 21 ++++
4 files changed, 213 insertions(+), 73 deletions(-)
diff --git a/scrub/phase1.c b/scrub/phase1.c
index 095c045915a..091b59e57e7 100644
--- a/scrub/phase1.c
+++ b/scrub/phase1.c
@@ -216,6 +216,8 @@ _("Kernel metadata scrubbing facility is not available."));
return ECANCELED;
}
+ check_scrubv(ctx);
+
/*
* Normally, callers are required to pass -n if the provided path is a
* readonly filesystem or the kernel wasn't built with online repair
diff --git a/scrub/scrub.c b/scrub/scrub.c
index 2fb2293558e..ec8d5e92cea 100644
--- a/scrub/scrub.c
+++ b/scrub/scrub.c
@@ -22,11 +22,39 @@
#include "descr.h"
#include "scrub_private.h"
-static int scrub_epilogue(struct scrub_ctx *ctx, struct descr *dsc,
- struct scrub_item *sri, struct xfs_scrub_vec *vec);
-
/* Online scrub and repair wrappers. */
+/* Describe the current state of a vectored scrub. */
+static int
+format_scrubv_descr(
+ struct scrub_ctx *ctx,
+ char *buf,
+ size_t buflen,
+ void *where)
+{
+ struct scrubv_head *bh = where;
+ struct xfs_scrub_vec_head *vhead = &bh->head;
+ struct xfs_scrub_vec *v = bh->head.svh_vecs + bh->i;
+ const struct xfrog_scrub_descr *sc = &xfrog_scrubbers[v->sv_type];
+
+ switch (sc->group) {
+ case XFROG_SCRUB_GROUP_AGHEADER:
+ case XFROG_SCRUB_GROUP_PERAG:
+ return snprintf(buf, buflen, _("AG %u %s"), vhead->svh_agno,
+ _(sc->descr));
+ case XFROG_SCRUB_GROUP_INODE:
+ return scrub_render_ino_descr(ctx, buf, buflen,
+ vhead->svh_ino, vhead->svh_gen, "%s",
+ _(sc->descr));
+ case XFROG_SCRUB_GROUP_FS:
+ case XFROG_SCRUB_GROUP_SUMMARY:
+ case XFROG_SCRUB_GROUP_ISCAN:
+ case XFROG_SCRUB_GROUP_NONE:
+ return snprintf(buf, buflen, _("%s"), _(sc->descr));
+ }
+ return -1;
+}
+
/* Format a scrub description. */
int
format_scrub_descr(
@@ -80,51 +108,6 @@ scrub_warn_incomplete_scrub(
_("Cross-referencing failed."));
}
-/* Do a read-only check of some metadata. */
-static int
-xfs_check_metadata(
- struct scrub_ctx *ctx,
- struct xfs_fd *xfdp,
- unsigned int scrub_type,
- struct scrub_item *sri)
-{
- DEFINE_DESCR(dsc, ctx, format_scrub_descr);
- struct xfs_scrub_metadata meta = { };
- struct xfs_scrub_vec vec;
- enum xfrog_scrub_group group;
-
- background_sleep();
-
- group = xfrog_scrubbers[scrub_type].group;
- meta.sm_type = scrub_type;
- switch (group) {
- case XFROG_SCRUB_GROUP_AGHEADER:
- case XFROG_SCRUB_GROUP_PERAG:
- meta.sm_agno = sri->sri_agno;
- break;
- case XFROG_SCRUB_GROUP_FS:
- case XFROG_SCRUB_GROUP_SUMMARY:
- case XFROG_SCRUB_GROUP_ISCAN:
- case XFROG_SCRUB_GROUP_NONE:
- break;
- case XFROG_SCRUB_GROUP_INODE:
- meta.sm_ino = sri->sri_ino;
- meta.sm_gen = sri->sri_gen;
- break;
- }
-
- assert(!debug_tweak_on("XFS_SCRUB_NO_KERNEL"));
- assert(scrub_type < XFS_SCRUB_TYPE_NR);
- descr_set(&dsc, &meta);
-
- dbg_printf("check %s flags %xh\n", descr_render(&dsc), meta.sm_flags);
-
- vec.sv_ret = xfrog_scrub_metadata(xfdp, &meta);
- vec.sv_type = scrub_type;
- vec.sv_flags = meta.sm_flags;
- return scrub_epilogue(ctx, &dsc, sri, &vec);
-}
-
/*
* Update all internal state after a scrub ioctl call.
* Returns 0 for success, or ECANCELED to abort the program.
@@ -256,6 +239,88 @@ _("Optimization is possible."));
return 0;
}
+/* Fill out the scrub vector header. */
+void
+scrub_item_to_vhead(
+ struct scrubv_head *bighead,
+ const struct scrub_item *sri)
+{
+ struct xfs_scrub_vec_head *vhead = &bighead->head;
+
+ if (bg_mode > 1)
+ vhead->svh_rest_us = bg_mode - 1;
+ if (sri->sri_agno != -1)
+ vhead->svh_agno = sri->sri_agno;
+ if (sri->sri_ino != -1ULL) {
+ vhead->svh_ino = sri->sri_ino;
+ vhead->svh_gen = sri->sri_gen;
+ }
+}
+
+/* Add a scrubber to the scrub vector. */
+void
+scrub_vhead_add(
+ struct scrubv_head *bighead,
+ const struct scrub_item *sri,
+ unsigned int scrub_type)
+{
+ struct xfs_scrub_vec_head *vhead = &bighead->head;
+ struct xfs_scrub_vec *v;
+
+ v = &vhead->svh_vecs[vhead->svh_nr++];
+ v->sv_type = scrub_type;
+ bighead->i = v - vhead->svh_vecs;
+}
+
+/* Do a read-only check of some metadata. */
+static int
+scrub_call_kernel(
+ struct scrub_ctx *ctx,
+ struct xfs_fd *xfdp,
+ struct scrub_item *sri)
+{
+ DEFINE_DESCR(dsc, ctx, format_scrubv_descr);
+ struct scrubv_head bh = { };
+ struct xfs_scrub_vec *v;
+ unsigned int scrub_type;
+ int error;
+
+ assert(!debug_tweak_on("XFS_SCRUB_NO_KERNEL"));
+
+ scrub_item_to_vhead(&bh, sri);
+ descr_set(&dsc, &bh);
+
+ foreach_scrub_type(scrub_type) {
+ if (!(sri->sri_state[scrub_type] & SCRUB_ITEM_NEEDSCHECK))
+ continue;
+ scrub_vhead_add(&bh, sri, scrub_type);
+
+ dbg_printf("check %s flags %xh tries %u\n", descr_render(&dsc),
+ sri->sri_state[scrub_type],
+ sri->sri_tries[scrub_type]);
+ }
+
+ error = -xfrog_scrubv_metadata(xfdp, &bh.head);
+ if (error)
+ return error;
+
+ foreach_bighead_vec(&bh, v) {
+ error = scrub_epilogue(ctx, &dsc, sri, v);
+ if (error)
+ return error;
+
+ /*
+ * Progress is counted by the inode for inode metadata; for
+ * everything else, it's counted for each scrub call.
+ */
+ if (!(sri->sri_state[v->sv_type] & SCRUB_ITEM_NEEDSCHECK) &&
+ sri->sri_ino == -1ULL)
+ progress_add(1);
+ }
+
+ return 0;
+}
+
/* Bulk-notify user about things that could be optimized. */
void
scrub_report_preen_triggers(
@@ -291,6 +356,37 @@ scrub_item_schedule_group(
}
}
+/* Decide if we call the kernel again to finish scrub/repair activity. */
+static inline bool
+scrub_item_call_kernel_again_future(
+ struct scrub_item *sri,
+ uint8_t work_mask,
+ const struct scrub_item *old)
+{
+ unsigned int scrub_type;
+ unsigned int nr = 0;
+
+ /* If there's nothing to do, we're done. */
+ foreach_scrub_type(scrub_type) {
+ if (sri->sri_state[scrub_type] & work_mask)
+ nr++;
+ }
+ if (!nr)
+ return false;
+
+ foreach_scrub_type(scrub_type) {
+ uint8_t statex = sri->sri_state[scrub_type] ^
+ old->sri_state[scrub_type];
+
+ if (statex & work_mask)
+ return true;
+ if (sri->sri_tries[scrub_type] != old->sri_tries[scrub_type])
+ return true;
+ }
+
+ return false;
+}
+
/* Decide if we call the kernel again to finish scrub/repair activity. */
bool
scrub_item_call_kernel_again(
@@ -319,6 +415,29 @@ scrub_item_call_kernel_again(
return false;
}
+/*
+ * For each scrub item whose state matches the state_flags, set up the item
+ * state for a kernel call. Returns true if any work was scheduled.
+ */
+bool
+scrub_item_schedule_work(
+ struct scrub_item *sri,
+ uint8_t state_flags)
+{
+ unsigned int scrub_type;
+ unsigned int nr = 0;
+
+ foreach_scrub_type(scrub_type) {
+ if (!(sri->sri_state[scrub_type] & state_flags))
+ continue;
+
+ sri->sri_tries[scrub_type] = SCRUB_ITEM_MAX_RETRIES;
+ nr++;
+ }
+
+ return nr > 0;
+}
+
/* Run all the incomplete scans on this scrub principal. */
int
scrub_item_check_file(
@@ -329,8 +448,10 @@ scrub_item_check_file(
struct xfs_fd xfd;
struct scrub_item old_sri;
struct xfs_fd *xfdp = &ctx->mnt;
- unsigned int scrub_type;
- int error;
+ int error = 0;
+
+ if (!scrub_item_schedule_work(sri, SCRUB_ITEM_NEEDSCHECK))
+ return 0;
/*
* If the caller passed us a file descriptor for a scrub, use it
@@ -343,31 +464,15 @@ scrub_item_check_file(
xfdp = &xfd;
}
- foreach_scrub_type(scrub_type) {
- if (!(sri->sri_state[scrub_type] & SCRUB_ITEM_NEEDSCHECK))
- continue;
-
- sri->sri_tries[scrub_type] = SCRUB_ITEM_MAX_RETRIES;
- do {
- memcpy(&old_sri, sri, sizeof(old_sri));
- error = xfs_check_metadata(ctx, xfdp, scrub_type, sri);
- if (error)
- return error;
- } while (scrub_item_call_kernel_again(sri, scrub_type,
- SCRUB_ITEM_NEEDSCHECK, &old_sri));
-
- /*
- * Progress is counted by the inode for inode metadata; for
- * everything else, it's counted for each scrub call.
- */
- if (sri->sri_ino == -1ULL)
- progress_add(1);
-
+ do {
+ memcpy(&old_sri, sri, sizeof(old_sri));
+ error = scrub_call_kernel(ctx, xfdp, sri);
if (error)
- break;
- }
+ return error;
+ } while (scrub_item_call_kernel_again_future(sri, SCRUB_ITEM_NEEDSCHECK,
+ &old_sri));
- return error;
+ return 0;
}
/* How many items do we have to check? */
@@ -562,3 +667,13 @@ can_force_rebuild(
return __scrub_test(ctx, XFS_SCRUB_TYPE_PROBE,
XFS_SCRUB_IFLAG_REPAIR | XFS_SCRUB_IFLAG_FORCE_REBUILD);
}
+
+void
+check_scrubv(
+ struct scrub_ctx *ctx)
+{
+ struct xfs_scrub_vec_head head = { };
+
+ /* We set the fallback flag if this doesn't work. */
+ xfrog_scrubv_metadata(&ctx->mnt, &head);
+}
diff --git a/scrub/scrub.h b/scrub/scrub.h
index 90578108a1c..183b89379cb 100644
--- a/scrub/scrub.h
+++ b/scrub/scrub.h
@@ -138,6 +138,8 @@ bool can_scrub_parent(struct scrub_ctx *ctx);
bool can_repair(struct scrub_ctx *ctx);
bool can_force_rebuild(struct scrub_ctx *ctx);
+void check_scrubv(struct scrub_ctx *ctx);
+
int scrub_file(struct scrub_ctx *ctx, int fd, const struct xfs_bulkstat *bstat,
unsigned int type, struct scrub_item *sri);
diff --git a/scrub/scrub_private.h b/scrub/scrub_private.h
index 98a9238f2aa..ecdce680d81 100644
--- a/scrub/scrub_private.h
+++ b/scrub/scrub_private.h
@@ -8,6 +8,26 @@
/* Shared code between scrub.c and repair.c. */
+/*
+ * Declare a structure big enough to handle all scrub types + barriers, and
+ * an iteration pointer. So far we only need two barriers.
+ */
+struct scrubv_head {
+ struct xfs_scrub_vec_head head;
+ struct xfs_scrub_vec __vecs[XFS_SCRUB_TYPE_NR + 2];
+ unsigned int i;
+};
+
+#define foreach_bighead_vec(bh, v) \
+ for ((bh)->i = 0, (v) = (bh)->head.svh_vecs; \
+ (bh)->i < (bh)->head.svh_nr; \
+ (bh)->i++, (v)++)
+
+void scrub_item_to_vhead(struct scrubv_head *bighead,
+ const struct scrub_item *sri);
+void scrub_vhead_add(struct scrubv_head *bighead, const struct scrub_item *sri,
+ unsigned int scrub_type);
+
int format_scrub_descr(struct scrub_ctx *ctx, char *buf, size_t buflen,
void *where);
@@ -104,5 +124,6 @@ scrub_item_schedule_retry(struct scrub_item *sri, unsigned int scrub_type)
bool scrub_item_call_kernel_again(struct scrub_item *sri,
unsigned int scrub_type, uint8_t work_mask,
const struct scrub_item *old);
+bool scrub_item_schedule_work(struct scrub_item *sri, uint8_t state_flags);
#endif /* XFS_SCRUB_SCRUB_PRIVATE_H_ */
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 08/10] xfs_scrub: vectorize repair calls
2023-12-31 19:51 ` [PATCHSET 6/6] xfs_scrub: vectorize kernel calls Darrick J. Wong
` (6 preceding siblings ...)
2023-12-31 23:21 ` [PATCH 07/10] xfs_scrub: vectorize scrub calls Darrick J. Wong
@ 2023-12-31 23:21 ` Darrick J. Wong
2023-12-31 23:22 ` [PATCH 09/10] xfs_scrub: use scrub barriers to reduce kernel calls Darrick J. Wong
2023-12-31 23:22 ` [PATCH 10/10] xfs_scrub: try spot repairs of metadata items to make scrub progress Darrick J. Wong
9 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:21 UTC (permalink / raw)
To: djwong, cem; +Cc: linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
Use the new vectorized scrub kernel calls to reduce the overhead of
performing repairs.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
scrub/repair.c | 267 +++++++++++++++++++++++++++----------------------
scrub/scrub.c | 79 +++-----------
scrub/scrub_private.h | 7 +
3 files changed, 166 insertions(+), 187 deletions(-)
diff --git a/scrub/repair.c b/scrub/repair.c
index 7a710a159e6..a8e61255c5f 100644
--- a/scrub/repair.c
+++ b/scrub/repair.c
@@ -20,11 +20,6 @@
#include "descr.h"
#include "scrub_private.h"
-static int repair_epilogue(struct scrub_ctx *ctx, struct descr *dsc,
- struct scrub_item *sri, unsigned int repair_flags,
- const struct xfs_scrub_vec *oldm,
- const struct xfs_scrub_vec *meta);
-
/* General repair routines. */
/*
@@ -83,64 +78,14 @@ repair_want_service_downgrade(
return false;
}
-/* Repair some metadata. */
-static int
-xfs_repair_metadata(
- struct scrub_ctx *ctx,
- struct xfs_fd *xfdp,
- unsigned int scrub_type,
- struct scrub_item *sri,
- unsigned int repair_flags)
+static inline void
+restore_oldvec(
+ struct xfs_scrub_vec *oldvec,
+ const struct scrub_item *sri,
+ unsigned int scrub_type)
{
- struct xfs_scrub_metadata meta = { 0 };
- struct xfs_scrub_vec oldm, vec;
- DEFINE_DESCR(dsc, ctx, format_scrub_descr);
- bool repair_only;
-
- /*
- * If the caller boosted the priority of this scrub type on behalf of a
- * higher level repair by setting IFLAG_REPAIR, turn off REPAIR_ONLY.
- */
- repair_only = (repair_flags & XRM_REPAIR_ONLY) &&
- scrub_item_type_boosted(sri, scrub_type);
-
- assert(scrub_type < XFS_SCRUB_TYPE_NR);
- assert(!debug_tweak_on("XFS_SCRUB_NO_KERNEL"));
- meta.sm_type = scrub_type;
- meta.sm_flags = XFS_SCRUB_IFLAG_REPAIR;
- if (use_force_rebuild)
- meta.sm_flags |= XFS_SCRUB_IFLAG_FORCE_REBUILD;
- switch (xfrog_scrubbers[scrub_type].group) {
- case XFROG_SCRUB_GROUP_AGHEADER:
- case XFROG_SCRUB_GROUP_PERAG:
- meta.sm_agno = sri->sri_agno;
- break;
- case XFROG_SCRUB_GROUP_INODE:
- meta.sm_ino = sri->sri_ino;
- meta.sm_gen = sri->sri_gen;
- break;
- default:
- break;
- }
-
- vec.sv_type = scrub_type;
- vec.sv_flags = sri->sri_state[scrub_type] & SCRUB_ITEM_REPAIR_ANY;
- memcpy(&oldm, &vec, sizeof(struct xfs_scrub_vec));
- if (!is_corrupt(&vec) && repair_only)
- return 0;
-
- descr_set(&dsc, &meta);
-
- if (needs_repair(&vec))
- str_info(ctx, descr_render(&dsc), _("Attempting repair."));
- else if (debug || verbose)
- str_info(ctx, descr_render(&dsc),
- _("Attempting optimization."));
-
- vec.sv_ret = xfrog_scrub_metadata(xfdp, &meta);
- vec.sv_flags = meta.sm_flags;
-
- return repair_epilogue(ctx, &dsc, sri, repair_flags, &oldm, &vec);
+ oldvec->sv_type = scrub_type;
+ oldvec->sv_flags = sri->sri_state[scrub_type] & SCRUB_ITEM_REPAIR_ANY;
}
static int
@@ -149,12 +94,15 @@ repair_epilogue(
struct descr *dsc,
struct scrub_item *sri,
unsigned int repair_flags,
- const struct xfs_scrub_vec *oldm,
const struct xfs_scrub_vec *meta)
{
+ struct xfs_scrub_vec oldv;
+ struct xfs_scrub_vec *oldm = &oldv;
unsigned int scrub_type = meta->sv_type;
int error = -meta->sv_ret;
+ restore_oldvec(oldm, sri, meta->sv_type);
+
switch (error) {
case 0:
/* No operational errors encountered. */
@@ -296,6 +244,132 @@ _("Repair unsuccessful; offline repair required."));
return 0;
}
+/* Decide if the dependent scrub types of the given scrub type are ok. */
+static bool
+repair_item_dependencies_ok(
+ const struct scrub_item *sri,
+ unsigned int scrub_type)
+{
+ unsigned int dep_mask = repair_deps[scrub_type];
+ unsigned int b;
+
+ for (b = 0; dep_mask && b < XFS_SCRUB_TYPE_NR; b++, dep_mask >>= 1) {
+ if (!(dep_mask & 1))
+ continue;
+ /*
+ * If this lower level object also needs repair, we can't fix
+ * the higher level item.
+ */
+ if (sri->sri_state[b] & SCRUB_ITEM_NEEDSREPAIR)
+ return false;
+ }
+
+ return true;
+}
+
+/* Decide if we want to repair a particular type of metadata. */
+static bool
+can_repair_now(
+ const struct scrub_item *sri,
+ unsigned int scrub_type,
+ __u32 repair_mask,
+ unsigned int repair_flags)
+{
+ struct xfs_scrub_vec oldvec;
+ bool repair_only;
+
+ /* Do we even need to repair this thing? */
+ if (!(sri->sri_state[scrub_type] & repair_mask))
+ return false;
+
+ restore_oldvec(&oldvec, sri, scrub_type);
+
+ /*
+ * If the caller boosted the priority of this scrub type on behalf of a
+ * higher level repair by setting IFLAG_REPAIR, ignore REPAIR_ONLY.
+ */
+ repair_only = (repair_flags & XRM_REPAIR_ONLY) &&
+ !(sri->sri_state[scrub_type] & SCRUB_ITEM_BOOST_REPAIR);
+ if (!is_corrupt(&oldvec) && repair_only)
+ return false;
+
+ /*
+ * Don't try to repair higher level items if their lower-level
+ * dependencies haven't been verified, unless this is our last chance
+ * to fix things without complaint.
+ */
+ if (!(repair_flags & XRM_FINAL_WARNING) &&
+ !repair_item_dependencies_ok(sri, scrub_type))
+ return false;
+
+ return true;
+}
+
+/*
+ * Repair some metadata.
+ *
+ * Returns 0 for success (or repair item deferral), or ECANCELED to abort the
+ * program.
+ */
+static int
+repair_call_kernel(
+ struct scrub_ctx *ctx,
+ struct xfs_fd *xfdp,
+ struct scrub_item *sri,
+ __u32 repair_mask,
+ unsigned int repair_flags)
+{
+ DEFINE_DESCR(dsc, ctx, format_scrubv_descr);
+ struct scrubv_head bh = { };
+ struct xfs_scrub_vec *v;
+ unsigned int scrub_type;
+ int error;
+
+ assert(!debug_tweak_on("XFS_SCRUB_NO_KERNEL"));
+
+ scrub_item_to_vhead(&bh, sri);
+ descr_set(&dsc, &bh);
+
+ foreach_scrub_type(scrub_type) {
+ if (scrub_excessive_errors(ctx))
+ return ECANCELED;
+
+ if (!can_repair_now(sri, scrub_type, repair_mask,
+ repair_flags))
+ continue;
+
+ scrub_vhead_add(&bh, sri, scrub_type, true);
+
+ if (sri->sri_state[scrub_type] & SCRUB_ITEM_NEEDSREPAIR)
+ str_info(ctx, descr_render(&dsc),
+ _("Attempting repair."));
+ else if (debug || verbose)
+ str_info(ctx, descr_render(&dsc),
+ _("Attempting optimization."));
+
+ dbg_printf("repair %s flags %xh tries %u\n", descr_render(&dsc),
+ sri->sri_state[scrub_type],
+ sri->sri_tries[scrub_type]);
+ }
+
+ error = -xfrog_scrubv_metadata(xfdp, &bh.head);
+ if (error)
+ return error;
+
+ foreach_bighead_vec(&bh, v) {
+ error = repair_epilogue(ctx, &dsc, sri, repair_flags, v);
+ if (error)
+ return error;
+
+ /* Maybe update progress if we fixed the problem. */
+ if (!(repair_flags & XRM_NOPROGRESS) &&
+ !(sri->sri_state[v->sv_type] & SCRUB_ITEM_REPAIR_ANY))
+ progress_add(1);
+ }
+
+ return 0;
+}
+
/*
* Prioritize action items in order of how long we can wait.
*
@@ -632,29 +706,6 @@ action_list_process(
return ret;
}
-/* Decide if the dependent scrub types of the given scrub type are ok. */
-static bool
-repair_item_dependencies_ok(
- const struct scrub_item *sri,
- unsigned int scrub_type)
-{
- unsigned int dep_mask = repair_deps[scrub_type];
- unsigned int b;
-
- for (b = 0; dep_mask && b < XFS_SCRUB_TYPE_NR; b++, dep_mask >>= 1) {
- if (!(dep_mask & 1))
- continue;
- /*
- * If this lower level object also needs repair, we can't fix
- * the higher level item.
- */
- if (sri->sri_state[b] & SCRUB_ITEM_NEEDSREPAIR)
- return false;
- }
-
- return true;
-}
-
/*
* For a given filesystem object, perform all repairs of a given class
* (corrupt, xcorrupt, xfail, preen) if the repair item says it's needed.
@@ -670,13 +721,14 @@ repair_item_class(
struct xfs_fd xfd;
struct scrub_item old_sri;
struct xfs_fd *xfdp = &ctx->mnt;
- unsigned int scrub_type;
int error = 0;
if (ctx->mode == SCRUB_MODE_DRY_RUN)
return 0;
if (ctx->mode == SCRUB_MODE_PREEN && !(repair_mask & SCRUB_ITEM_PREEN))
return 0;
+ if (!scrub_item_schedule_work(sri, repair_mask))
+ return 0;
/*
* If the caller passed us a file descriptor for a scrub, use it
@@ -689,39 +741,14 @@ repair_item_class(
xfdp = &xfd;
}
- foreach_scrub_type(scrub_type) {
- if (scrub_excessive_errors(ctx))
- return ECANCELED;
-
- if (!(sri->sri_state[scrub_type] & repair_mask))
- continue;
-
- /*
- * Don't try to repair higher level items if their lower-level
- * dependencies haven't been verified, unless this is our last
- * chance to fix things without complaint.
- */
- if (!(flags & XRM_FINAL_WARNING) &&
- !repair_item_dependencies_ok(sri, scrub_type))
- continue;
-
- sri->sri_tries[scrub_type] = SCRUB_ITEM_MAX_RETRIES;
- do {
- memcpy(&old_sri, sri, sizeof(old_sri));
- error = xfs_repair_metadata(ctx, xfdp, scrub_type, sri,
- flags);
- if (error)
- return error;
- } while (scrub_item_call_kernel_again(sri, scrub_type,
- repair_mask, &old_sri));
-
- /* Maybe update progress if we fixed the problem. */
- if (!(flags & XRM_NOPROGRESS) &&
- !(sri->sri_state[scrub_type] & SCRUB_ITEM_REPAIR_ANY))
- progress_add(1);
- }
-
- return error;
+ do {
+ memcpy(&old_sri, sri, sizeof(struct scrub_item));
+ error = repair_call_kernel(ctx, xfdp, sri, repair_mask, flags);
+ if (error)
+ return error;
+ } while (scrub_item_call_kernel_again(sri, repair_mask, &old_sri));
+
+ return 0;
}
/*
diff --git a/scrub/scrub.c b/scrub/scrub.c
index ec8d5e92cea..9f982eee701 100644
--- a/scrub/scrub.c
+++ b/scrub/scrub.c
@@ -25,7 +25,7 @@
/* Online scrub and repair wrappers. */
/* Describe the current state of a vectored scrub. */
-static int
+int
format_scrubv_descr(
struct scrub_ctx *ctx,
char *buf,
@@ -55,35 +55,6 @@ format_scrubv_descr(
return -1;
}
-/* Format a scrub description. */
-int
-format_scrub_descr(
- struct scrub_ctx *ctx,
- char *buf,
- size_t buflen,
- void *where)
-{
- struct xfs_scrub_metadata *meta = where;
- const struct xfrog_scrub_descr *sc = &xfrog_scrubbers[meta->sm_type];
-
- switch (sc->group) {
- case XFROG_SCRUB_GROUP_AGHEADER:
- case XFROG_SCRUB_GROUP_PERAG:
- return snprintf(buf, buflen, _("AG %u %s"), meta->sm_agno,
- _(sc->descr));
- case XFROG_SCRUB_GROUP_INODE:
- return scrub_render_ino_descr(ctx, buf, buflen,
- meta->sm_ino, meta->sm_gen, "%s",
- _(sc->descr));
- case XFROG_SCRUB_GROUP_FS:
- case XFROG_SCRUB_GROUP_SUMMARY:
- case XFROG_SCRUB_GROUP_ISCAN:
- case XFROG_SCRUB_GROUP_NONE:
- return snprintf(buf, buflen, _("%s"), _(sc->descr));
- }
- return -1;
-}
-
/* Warn about strange circumstances after scrub. */
void
scrub_warn_incomplete_scrub(
@@ -262,13 +233,18 @@ void
scrub_vhead_add(
struct scrubv_head *bighead,
const struct scrub_item *sri,
- unsigned int scrub_type)
+ unsigned int scrub_type,
+ bool repair)
{
struct xfs_scrub_vec_head *vhead = &bighead->head;
struct xfs_scrub_vec *v;
v = &vhead->svh_vecs[vhead->svh_nr++];
v->sv_type = scrub_type;
+ if (repair)
+ v->sv_flags |= XFS_SCRUB_IFLAG_REPAIR;
+ if (repair && use_force_rebuild)
+ v->sv_flags |= XFS_SCRUB_IFLAG_FORCE_REBUILD;
bighead->i = v - vhead->svh_vecs;
}
@@ -293,7 +269,7 @@ scrub_call_kernel(
foreach_scrub_type(scrub_type) {
if (!(sri->sri_state[scrub_type] & SCRUB_ITEM_NEEDSCHECK))
continue;
- scrub_vhead_add(&bh, sri, scrub_type);
+ scrub_vhead_add(&bh, sri, scrub_type, false);
dbg_printf("check %s flags %xh tries %u\n", descr_render(&dsc),
sri->sri_state[scrub_type],
@@ -357,8 +333,8 @@ scrub_item_schedule_group(
}
/* Decide if we call the kernel again to finish scrub/repair activity. */
-static inline bool
-scrub_item_call_kernel_again_future(
+bool
+scrub_item_call_kernel_again(
struct scrub_item *sri,
uint8_t work_mask,
const struct scrub_item *old)
@@ -374,6 +350,11 @@ scrub_item_call_kernel_again_future(
if (!nr)
return false;
+ /*
+ * We are willing to go again if the last call had any effect on the
+ * state of the scrub item that the caller cares about or if the kernel
+ * asked us to try again.
+ */
foreach_scrub_type(scrub_type) {
uint8_t statex = sri->sri_state[scrub_type] ^
old->sri_state[scrub_type];
@@ -387,34 +368,6 @@ scrub_item_call_kernel_again_future(
return false;
}
-/* Decide if we call the kernel again to finish scrub/repair activity. */
-bool
-scrub_item_call_kernel_again(
- struct scrub_item *sri,
- unsigned int scrub_type,
- uint8_t work_mask,
- const struct scrub_item *old)
-{
- uint8_t statex;
-
- /* If there's nothing to do, we're done. */
- if (!(sri->sri_state[scrub_type] & work_mask))
- return false;
-
- /*
- * We are willing to go again if the last call had any effect on the
- * state of the scrub item that the caller cares about, if the freeze
- * flag got set, or if the kernel asked us to try again...
- */
- statex = sri->sri_state[scrub_type] ^ old->sri_state[scrub_type];
- if (statex & work_mask)
- return true;
- if (sri->sri_tries[scrub_type] != old->sri_tries[scrub_type])
- return true;
-
- return false;
-}
-
/*
* For each scrub item whose state matches the state_flags, set up the item
* state for a kernel call. Returns true if any work was scheduled.
@@ -469,7 +422,7 @@ scrub_item_check_file(
error = scrub_call_kernel(ctx, xfdp, sri);
if (error)
return error;
- } while (scrub_item_call_kernel_again_future(sri, SCRUB_ITEM_NEEDSCHECK,
+ } while (scrub_item_call_kernel_again(sri, SCRUB_ITEM_NEEDSCHECK,
&old_sri));
return 0;
diff --git a/scrub/scrub_private.h b/scrub/scrub_private.h
index ecdce680d81..ceda8ea1505 100644
--- a/scrub/scrub_private.h
+++ b/scrub/scrub_private.h
@@ -26,9 +26,9 @@ struct scrubv_head {
void scrub_item_to_vhead(struct scrubv_head *bighead,
const struct scrub_item *sri);
void scrub_vhead_add(struct scrubv_head *bighead, const struct scrub_item *sri,
- unsigned int scrub_type);
+ unsigned int scrub_type, bool repair);
-int format_scrub_descr(struct scrub_ctx *ctx, char *buf, size_t buflen,
+int format_scrubv_descr(struct scrub_ctx *ctx, char *buf, size_t buflen,
void *where);
/* Predicates for scrub flag state. */
@@ -121,8 +121,7 @@ scrub_item_schedule_retry(struct scrub_item *sri, unsigned int scrub_type)
return true;
}
-bool scrub_item_call_kernel_again(struct scrub_item *sri,
- unsigned int scrub_type, uint8_t work_mask,
+bool scrub_item_call_kernel_again(struct scrub_item *sri, uint8_t work_mask,
const struct scrub_item *old);
bool scrub_item_schedule_work(struct scrub_item *sri, uint8_t state_flags);
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 09/10] xfs_scrub: use scrub barriers to reduce kernel calls
2023-12-31 19:51 ` [PATCHSET 6/6] xfs_scrub: vectorize kernel calls Darrick J. Wong
` (7 preceding siblings ...)
2023-12-31 23:21 ` [PATCH 08/10] xfs_scrub: vectorize repair calls Darrick J. Wong
@ 2023-12-31 23:22 ` Darrick J. Wong
2023-12-31 23:22 ` [PATCH 10/10] xfs_scrub: try spot repairs of metadata items to make scrub progress Darrick J. Wong
9 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:22 UTC (permalink / raw)
To: djwong, cem; +Cc: linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
Use scrub barriers so that we can submit a single scrub request for a
bunch of things, and have the kernel stop midway through if it finds
anything broken.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
scrub/phase2.c | 15 ++-------
scrub/phase3.c | 17 +---------
scrub/repair.c | 32 ++++++++++++++++++-
scrub/scrub.c | 81 ++++++++++++++++++++++++++++++++++++++++++++++++-
scrub/scrub.h | 17 ++++++++++
scrub/scrub_private.h | 4 ++
6 files changed, 134 insertions(+), 32 deletions(-)
diff --git a/scrub/phase2.c b/scrub/phase2.c
index 57c6d0ef213..d435da07125 100644
--- a/scrub/phase2.c
+++ b/scrub/phase2.c
@@ -91,21 +91,12 @@ scan_ag_metadata(
snprintf(descr, DESCR_BUFSZ, _("AG %u"), agno);
/*
- * First we scrub and fix the AG headers, because we need
- * them to work well enough to check the AG btrees.
+ * First we scrub and fix the AG headers, because we need them to work
+ * well enough to check the AG btrees. Then scrub the AG btrees.
*/
scrub_item_schedule_group(&sri, XFROG_SCRUB_GROUP_AGHEADER);
- ret = scrub_item_check(ctx, &sri);
- if (ret)
- goto err;
-
- /* Repair header damage. */
- ret = repair_item_corruption(ctx, &sri);
- if (ret)
- goto err;
-
- /* Now scrub the AG btrees. */
scrub_item_schedule_group(&sri, XFROG_SCRUB_GROUP_PERAG);
+
ret = scrub_item_check(ctx, &sri);
if (ret)
goto err;
diff --git a/scrub/phase3.c b/scrub/phase3.c
index 98e5c5a1f9f..09a1ea452bb 100644
--- a/scrub/phase3.c
+++ b/scrub/phase3.c
@@ -145,25 +145,11 @@ scrub_inode(
/* Scrub the inode. */
scrub_item_schedule(&sri, XFS_SCRUB_TYPE_INODE);
- error = scrub_item_check_file(ctx, &sri, fd);
- if (error)
- goto out;
-
- error = try_inode_repair(ictx, &sri, fd);
- if (error)
- goto out;
/* Scrub all block mappings. */
scrub_item_schedule(&sri, XFS_SCRUB_TYPE_BMBTD);
scrub_item_schedule(&sri, XFS_SCRUB_TYPE_BMBTA);
scrub_item_schedule(&sri, XFS_SCRUB_TYPE_BMBTC);
- error = scrub_item_check_file(ctx, &sri, fd);
- if (error)
- goto out;
-
- error = try_inode_repair(ictx, &sri, fd);
- if (error)
- goto out;
/*
* Check file data contents, e.g. symlink and directory entries.
@@ -182,11 +168,12 @@ scrub_inode(
scrub_item_schedule(&sri, XFS_SCRUB_TYPE_XATTR);
scrub_item_schedule(&sri, XFS_SCRUB_TYPE_PARENT);
+
+ /* Try to check and repair the file while it's open. */
error = scrub_item_check_file(ctx, &sri, fd);
if (error)
goto out;
- /* Try to repair the file while it's open. */
error = try_inode_repair(ictx, &sri, fd);
if (error)
goto out;
diff --git a/scrub/repair.c b/scrub/repair.c
index a8e61255c5f..1ce0283cb7f 100644
--- a/scrub/repair.c
+++ b/scrub/repair.c
@@ -323,6 +323,7 @@ repair_call_kernel(
struct scrubv_head bh = { };
struct xfs_scrub_vec *v;
unsigned int scrub_type;
+ bool need_barrier = false;
int error;
assert(!debug_tweak_on("XFS_SCRUB_NO_KERNEL"));
@@ -338,6 +339,11 @@ repair_call_kernel(
repair_flags))
continue;
+ if (need_barrier) {
+ scrub_vhead_add_barrier(&bh);
+ need_barrier = false;
+ }
+
scrub_vhead_add(&bh, sri, scrub_type, true);
if (sri->sri_state[scrub_type] & SCRUB_ITEM_NEEDSREPAIR)
@@ -350,6 +356,17 @@ repair_call_kernel(
dbg_printf("repair %s flags %xh tries %u\n", descr_render(&dsc),
sri->sri_state[scrub_type],
sri->sri_tries[scrub_type]);
+
+ /*
+ * One of the other scrub types depends on this one. Set us up
+ * to add a repair barrier if we decide to schedule a repair
+ * after this one. If the UNFIXED flag is set, that means this
+ * is our last chance to fix things, so we skip the barriers
+ * just let everything run.
+ */
+ if (!(repair_flags & XRM_FINAL_WARNING) &&
+ (sri->sri_state[scrub_type] & SCRUB_ITEM_BARRIER))
+ need_barrier = true;
}
error = -xfrog_scrubv_metadata(xfdp, &bh.head);
@@ -357,6 +374,16 @@ repair_call_kernel(
return error;
foreach_bighead_vec(&bh, v) {
+ /* Deal with barriers separately. */
+ if (v->sv_type == XFS_SCRUB_TYPE_BARRIER) {
+ /* -ECANCELED means the kernel stopped here. */
+ if (v->sv_ret == -ECANCELED)
+ return 0;
+ if (v->sv_ret)
+ return -v->sv_ret;
+ continue;
+ }
+
error = repair_epilogue(ctx, &dsc, sri, repair_flags, v);
if (error)
return error;
@@ -445,7 +472,8 @@ repair_item_boost_priorities(
* bits are left untouched to force a rescan in phase 4.
*/
#define MUSTFIX_STATES (SCRUB_ITEM_CORRUPT | \
- SCRUB_ITEM_BOOST_REPAIR)
+ SCRUB_ITEM_BOOST_REPAIR | \
+ SCRUB_ITEM_BARRIER)
/*
* Figure out which AG metadata must be fixed before we can move on
* to the inode scan.
@@ -727,7 +755,7 @@ repair_item_class(
return 0;
if (ctx->mode == SCRUB_MODE_PREEN && !(repair_mask & SCRUB_ITEM_PREEN))
return 0;
- if (!scrub_item_schedule_work(sri, repair_mask))
+ if (!scrub_item_schedule_work(sri, repair_mask, repair_deps))
return 0;
/*
diff --git a/scrub/scrub.c b/scrub/scrub.c
index 9f982eee701..2ec3cbc9aac 100644
--- a/scrub/scrub.c
+++ b/scrub/scrub.c
@@ -24,6 +24,35 @@
/* Online scrub and repair wrappers. */
+/*
+ * Bitmap showing the correctness dependencies between scrub types for scrubs.
+ * Dependencies cannot cross scrub groups.
+ */
+#define DEP(x) (1U << (x))
+static const unsigned int scrub_deps[XFS_SCRUB_TYPE_NR] = {
+ [XFS_SCRUB_TYPE_AGF] = DEP(XFS_SCRUB_TYPE_SB),
+ [XFS_SCRUB_TYPE_AGFL] = DEP(XFS_SCRUB_TYPE_SB) |
+ DEP(XFS_SCRUB_TYPE_AGF),
+ [XFS_SCRUB_TYPE_AGI] = DEP(XFS_SCRUB_TYPE_SB),
+ [XFS_SCRUB_TYPE_BNOBT] = DEP(XFS_SCRUB_TYPE_AGF),
+ [XFS_SCRUB_TYPE_CNTBT] = DEP(XFS_SCRUB_TYPE_AGF),
+ [XFS_SCRUB_TYPE_INOBT] = DEP(XFS_SCRUB_TYPE_AGI),
+ [XFS_SCRUB_TYPE_FINOBT] = DEP(XFS_SCRUB_TYPE_AGI),
+ [XFS_SCRUB_TYPE_RMAPBT] = DEP(XFS_SCRUB_TYPE_AGF),
+ [XFS_SCRUB_TYPE_REFCNTBT] = DEP(XFS_SCRUB_TYPE_AGF),
+ [XFS_SCRUB_TYPE_BMBTD] = DEP(XFS_SCRUB_TYPE_INODE),
+ [XFS_SCRUB_TYPE_BMBTA] = DEP(XFS_SCRUB_TYPE_INODE),
+ [XFS_SCRUB_TYPE_BMBTC] = DEP(XFS_SCRUB_TYPE_INODE),
+ [XFS_SCRUB_TYPE_DIR] = DEP(XFS_SCRUB_TYPE_BMBTD),
+ [XFS_SCRUB_TYPE_XATTR] = DEP(XFS_SCRUB_TYPE_BMBTA),
+ [XFS_SCRUB_TYPE_SYMLINK] = DEP(XFS_SCRUB_TYPE_BMBTD),
+ [XFS_SCRUB_TYPE_PARENT] = DEP(XFS_SCRUB_TYPE_BMBTD),
+ [XFS_SCRUB_TYPE_QUOTACHECK] = DEP(XFS_SCRUB_TYPE_UQUOTA) |
+ DEP(XFS_SCRUB_TYPE_GQUOTA) |
+ DEP(XFS_SCRUB_TYPE_PQUOTA),
+};
+#undef DEP
+
/* Describe the current state of a vectored scrub. */
int
format_scrubv_descr(
@@ -248,6 +277,21 @@ scrub_vhead_add(
bighead->i = v - vhead->svh_vecs;
}
+/* Add a barrier to the scrub vector. */
+void
+scrub_vhead_add_barrier(
+ struct scrubv_head *bighead)
+{
+ struct xfs_scrub_vec_head *vhead = &bighead->head;
+ struct xfs_scrub_vec *v;
+
+ v = &vhead->svh_vecs[vhead->svh_nr++];
+ v->sv_type = XFS_SCRUB_TYPE_BARRIER;
+ v->sv_flags = XFS_SCRUB_OFLAG_CORRUPT | XFS_SCRUB_OFLAG_XFAIL |
+ XFS_SCRUB_OFLAG_XCORRUPT | XFS_SCRUB_OFLAG_INCOMPLETE;
+ bighead->i = v - vhead->svh_vecs;
+}
+
/* Do a read-only check of some metadata. */
static int
scrub_call_kernel(
@@ -259,6 +303,7 @@ scrub_call_kernel(
struct scrubv_head bh = { };
struct xfs_scrub_vec *v;
unsigned int scrub_type;
+ bool need_barrier = false;
int error;
assert(!debug_tweak_on("XFS_SCRUB_NO_KERNEL"));
@@ -269,8 +314,17 @@ scrub_call_kernel(
foreach_scrub_type(scrub_type) {
if (!(sri->sri_state[scrub_type] & SCRUB_ITEM_NEEDSCHECK))
continue;
+
+ if (need_barrier) {
+ scrub_vhead_add_barrier(&bh);
+ need_barrier = false;
+ }
+
scrub_vhead_add(&bh, sri, scrub_type, false);
+ if (sri->sri_state[scrub_type] & SCRUB_ITEM_BARRIER)
+ need_barrier = true;
+
dbg_printf("check %s flags %xh tries %u\n", descr_render(&dsc),
sri->sri_state[scrub_type],
sri->sri_tries[scrub_type]);
@@ -281,6 +335,16 @@ scrub_call_kernel(
return error;
foreach_bighead_vec(&bh, v) {
+ /* Deal with barriers separately. */
+ if (v->sv_type == XFS_SCRUB_TYPE_BARRIER) {
+ /* -ECANCELED means the kernel stopped here. */
+ if (v->sv_ret == -ECANCELED)
+ return 0;
+ if (v->sv_ret)
+ return -v->sv_ret;
+ continue;
+ }
+
error = scrub_epilogue(ctx, &dsc, sri, v);
if (error)
return error;
@@ -375,15 +439,25 @@ scrub_item_call_kernel_again(
bool
scrub_item_schedule_work(
struct scrub_item *sri,
- uint8_t state_flags)
+ uint8_t state_flags,
+ const unsigned int *schedule_deps)
{
unsigned int scrub_type;
unsigned int nr = 0;
foreach_scrub_type(scrub_type) {
+ unsigned int j;
+
+ sri->sri_state[scrub_type] &= ~SCRUB_ITEM_BARRIER;
+
if (!(sri->sri_state[scrub_type] & state_flags))
continue;
+ foreach_scrub_type(j) {
+ if (schedule_deps[scrub_type] & (1U << j))
+ sri->sri_state[j] |= SCRUB_ITEM_BARRIER;
+ }
+
sri->sri_tries[scrub_type] = SCRUB_ITEM_MAX_RETRIES;
nr++;
}
@@ -403,7 +477,7 @@ scrub_item_check_file(
struct xfs_fd *xfdp = &ctx->mnt;
int error = 0;
- if (!scrub_item_schedule_work(sri, SCRUB_ITEM_NEEDSCHECK))
+ if (!scrub_item_schedule_work(sri, SCRUB_ITEM_NEEDSCHECK, scrub_deps))
return 0;
/*
@@ -627,6 +701,9 @@ check_scrubv(
{
struct xfs_scrub_vec_head head = { };
+ if (debug_tweak_on("XFS_SCRUB_FORCE_SINGLE"))
+ ctx->mnt.flags |= XFROG_FLAG_SCRUB_FORCE_SINGLE;
+
/* We set the fallback flag if this doesn't work. */
xfrog_scrubv_metadata(&ctx->mnt, &head);
}
diff --git a/scrub/scrub.h b/scrub/scrub.h
index 183b89379cb..c3eed1b261d 100644
--- a/scrub/scrub.h
+++ b/scrub/scrub.h
@@ -30,6 +30,9 @@ enum xfrog_scrub_group;
/* This scrub type needs to be checked. */
#define SCRUB_ITEM_NEEDSCHECK (1 << 5)
+/* Scrub barrier. */
+#define SCRUB_ITEM_BARRIER (1 << 6)
+
/* All of the state flags that we need to prioritize repair work. */
#define SCRUB_ITEM_REPAIR_ANY (SCRUB_ITEM_CORRUPT | \
SCRUB_ITEM_PREEN | \
@@ -126,6 +129,20 @@ scrub_item_check(struct scrub_ctx *ctx, struct scrub_item *sri)
return scrub_item_check_file(ctx, sri, -1);
}
+/* Count the number of metadata objects still needing a scrub. */
+static inline unsigned int
+scrub_item_count_needscheck(
+ const struct scrub_item *sri)
+{
+ unsigned int ret = 0;
+ unsigned int i;
+
+ foreach_scrub_type(i)
+ if (sri->sri_state[i] & SCRUB_ITEM_NEEDSCHECK)
+ ret++;
+ return ret;
+}
+
void scrub_report_preen_triggers(struct scrub_ctx *ctx);
bool can_scrub_fs_metadata(struct scrub_ctx *ctx);
diff --git a/scrub/scrub_private.h b/scrub/scrub_private.h
index ceda8ea1505..5014feee515 100644
--- a/scrub/scrub_private.h
+++ b/scrub/scrub_private.h
@@ -27,6 +27,7 @@ void scrub_item_to_vhead(struct scrubv_head *bighead,
const struct scrub_item *sri);
void scrub_vhead_add(struct scrubv_head *bighead, const struct scrub_item *sri,
unsigned int scrub_type, bool repair);
+void scrub_vhead_add_barrier(struct scrubv_head *bighead);
int format_scrubv_descr(struct scrub_ctx *ctx, char *buf, size_t buflen,
void *where);
@@ -123,6 +124,7 @@ scrub_item_schedule_retry(struct scrub_item *sri, unsigned int scrub_type)
bool scrub_item_call_kernel_again(struct scrub_item *sri, uint8_t work_mask,
const struct scrub_item *old);
-bool scrub_item_schedule_work(struct scrub_item *sri, uint8_t state_flags);
+bool scrub_item_schedule_work(struct scrub_item *sri, uint8_t state_flags,
+ const unsigned int *schedule_deps);
#endif /* XFS_SCRUB_SCRUB_PRIVATE_H_ */
^ permalink raw reply related [flat|nested] 185+ messages in thread
* [PATCH 10/10] xfs_scrub: try spot repairs of metadata items to make scrub progress
2023-12-31 19:51 ` [PATCHSET 6/6] xfs_scrub: vectorize kernel calls Darrick J. Wong
` (8 preceding siblings ...)
2023-12-31 23:22 ` [PATCH 09/10] xfs_scrub: use scrub barriers to reduce kernel calls Darrick J. Wong
@ 2023-12-31 23:22 ` Darrick J. Wong
9 siblings, 0 replies; 185+ messages in thread
From: Darrick J. Wong @ 2023-12-31 23:22 UTC (permalink / raw)
To: djwong, cem; +Cc: linux-xfs
From: Darrick J. Wong <djwong@kernel.org>
Now that we've enabled scrub dependency barriers, it's possible that a
scrub_item_check call will return with some of the scrub items still in
NEEDSCHECK state. If, for example, scrub type B depends on scrub type
A being clean and A is not clean, B will still be in NEEDSCHECK state.
In order to make as much scanning progress as possible during phase 2
and phase 3, allow ourselves to try some spot repairs in the hopes that
it will enable us to make progress towards at least scanning the whole
metadata item. If we can't make any forward progress, we'll queue the
scrub item for repair in phase 4, which means that anything still in in
NEEDSCHECK state becomes CORRUPT state. (At worst, the NEEDSCHECK item
will actually be clean by phase 4, and xfs_scrub will report that it
didn't need any work after all.)
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
scrub/phase2.c | 78 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++
scrub/phase3.c | 71 ++++++++++++++++++++++++++++++++++++++++++++++++++-
scrub/repair.c | 15 +++++++++++
3 files changed, 163 insertions(+), 1 deletion(-)
diff --git a/scrub/phase2.c b/scrub/phase2.c
index d435da07125..c24d137358c 100644
--- a/scrub/phase2.c
+++ b/scrub/phase2.c
@@ -69,6 +69,53 @@ defer_fs_repair(
return 0;
}
+/*
+ * If we couldn't check all the scheduled metadata items, try performing spot
+ * repairs until we check everything or stop making forward progress.
+ */
+static int
+repair_and_scrub_loop(
+ struct scrub_ctx *ctx,
+ struct scrub_item *sri,
+ const char *descr,
+ bool *defer)
+{
+ unsigned int to_check;
+ int ret;
+
+ *defer = false;
+ if (ctx->mode != SCRUB_MODE_REPAIR)
+ return 0;
+
+ to_check = scrub_item_count_needscheck(sri);
+ while (to_check > 0) {
+ unsigned int nr;
+
+ ret = repair_item_corruption(ctx, sri);
+ if (ret)
+ return ret;
+
+ ret = scrub_item_check(ctx, sri);
+ if (ret)
+ return ret;
+
+ nr = scrub_item_count_needscheck(sri);
+ if (nr == to_check) {
+ /*
+ * We cannot make forward scanning progress with this
+ * metadata, so defer the rest until phase 4.
+ */
+ str_info(ctx, descr,
+ _("Unable to make forward checking progress; will try again in phase 4."));
+ *defer = true;
+ return 0;
+ }
+ to_check = nr;
+ }
+
+ return 0;
+}
+
/* Scrub each AG's metadata btrees. */
static void
scan_ag_metadata(
@@ -82,6 +129,7 @@ scan_ag_metadata(
struct scan_ctl *sctl = arg;
char descr[DESCR_BUFSZ];
unsigned int difficulty;
+ bool defer_repairs;
int ret;
if (sctl->aborted)
@@ -97,10 +145,22 @@ scan_ag_metadata(
scrub_item_schedule_group(&sri, XFROG_SCRUB_GROUP_AGHEADER);
scrub_item_schedule_group(&sri, XFROG_SCRUB_GROUP_PERAG);
+ /*
+ * Try to check all of the AG metadata items that we just scheduled.
+ * If we return with some types still needing a check, try repairing
+ * any damaged metadata that we've found so far, and try again. Abort
+ * if we stop making forward progress.
+ */
ret = scrub_item_check(ctx, &sri);
if (ret)
goto err;
+ ret = repair_and_scrub_loop(ctx, &sri, descr, &defer_repairs);
+ if (ret)
+ goto err;
+ if (defer_repairs)
+ goto defer;
+
/*
* Figure out if we need to perform early fixing. The only
* reason we need to do this is if the inobt is broken, which
@@ -117,6 +177,7 @@ scan_ag_metadata(
if (ret)
goto err;
+defer:
/* Everything else gets fixed during phase 4. */
ret = defer_fs_repair(ctx, &sri);
if (ret)
@@ -137,11 +198,18 @@ scan_fs_metadata(
struct scrub_ctx *ctx = (struct scrub_ctx *)wq->wq_ctx;
struct scan_ctl *sctl = arg;
unsigned int difficulty;
+ bool defer_repairs;
int ret;
if (sctl->aborted)
goto out;
+ /*
+ * Try to check all of the metadata files that we just scheduled. If
+ * we return with some types still needing a check, try repairing any
+ * damaged metadata that we've found so far, and try again. Abort if
+ * we stop making forward progress.
+ */
scrub_item_init_fs(&sri);
scrub_item_schedule(&sri, type);
ret = scrub_item_check(ctx, &sri);
@@ -150,10 +218,20 @@ scan_fs_metadata(
goto out;
}
+ ret = repair_and_scrub_loop(ctx, &sri, xfrog_scrubbers[type].descr,
+ &defer_repairs);
+ if (ret) {
+ sctl->aborted = true;
+ goto out;
+ }
+ if (defer_repairs)
+ goto defer;
+
/* Complain about metadata corruptions that might not be fixable. */
difficulty = repair_item_difficulty(&sri);
warn_repair_difficulties(ctx, difficulty, xfrog_scrubbers[type].descr);
+defer:
ret = defer_fs_repair(ctx, &sri);
if (ret) {
sctl->aborted = true;
diff --git a/scrub/phase3.c b/scrub/phase3.c
index 09a1ea452bb..046a42c1da8 100644
--- a/scrub/phase3.c
+++ b/scrub/phase3.c
@@ -99,6 +99,58 @@ try_inode_repair(
return repair_file_corruption(ictx->ctx, sri, fd);
}
+/*
+ * If we couldn't check all the scheduled file metadata items, try performing
+ * spot repairs until we check everything or stop making forward progress.
+ */
+static int
+repair_and_scrub_inode_loop(
+ struct scrub_ctx *ctx,
+ struct xfs_bulkstat *bstat,
+ int fd,
+ struct scrub_item *sri,
+ bool *defer)
+{
+ unsigned int to_check;
+ int error;
+
+ *defer = false;
+ if (ctx->mode != SCRUB_MODE_REPAIR)
+ return 0;
+
+ to_check = scrub_item_count_needscheck(sri);
+ while (to_check > 0) {
+ unsigned int nr;
+
+ error = repair_file_corruption(ctx, sri, fd);
+ if (error)
+ return error;
+
+ error = scrub_item_check_file(ctx, sri, fd);
+ if (error)
+ return error;
+
+ nr = scrub_item_count_needscheck(sri);
+ if (nr == to_check) {
+ char descr[DESCR_BUFSZ];
+
+ /*
+ * We cannot make forward scanning progress with this
+ * inode, so defer the rest until phase 4.
+ */
+ scrub_render_ino_descr(ctx, descr, DESCR_BUFSZ,
+ bstat->bs_ino, bstat->bs_gen, NULL);
+ str_info(ctx, descr,
+ _("Unable to make forward checking progress; will try again in phase 4."));
+ *defer = true;
+ return 0;
+ }
+ to_check = nr;
+ }
+
+ return 0;
+}
+
/* Verify the contents, xattrs, and extent maps of an inode. */
static int
scrub_inode(
@@ -169,11 +221,28 @@ scrub_inode(
scrub_item_schedule(&sri, XFS_SCRUB_TYPE_XATTR);
scrub_item_schedule(&sri, XFS_SCRUB_TYPE_PARENT);
- /* Try to check and repair the file while it's open. */
+ /*
+ * Try to check all of the metadata items that we just scheduled. If
+ * we return with some types still needing a check and the space
+ * metadata isn't also in need of repairs, try repairing any damaged
+ * file metadata that we've found so far, and try checking the file
+ * again. Worst case, defer the repairs and the checks to phase 4 if
+ * we can't make any progress on anything.
+ */
error = scrub_item_check_file(ctx, &sri, fd);
if (error)
goto out;
+ if (!ictx->always_defer_repairs) {
+ bool defer_repairs;
+
+ error = repair_and_scrub_inode_loop(ctx, bstat, fd, &sri,
+ &defer_repairs);
+ if (error || defer_repairs)
+ goto out;
+ }
+
+ /* Try to repair the file while it's open. */
error = try_inode_repair(ictx, &sri, fd);
if (error)
goto out;
diff --git a/scrub/repair.c b/scrub/repair.c
index 1ce0283cb7f..7e131001e13 100644
--- a/scrub/repair.c
+++ b/scrub/repair.c
@@ -859,6 +859,7 @@ repair_item_to_action_item(
struct action_item **aitemp)
{
struct action_item *aitem;
+ unsigned int scrub_type;
if (repair_item_count_needsrepair(sri) == 0)
return 0;
@@ -874,6 +875,20 @@ repair_item_to_action_item(
INIT_LIST_HEAD(&aitem->list);
memcpy(&aitem->sri, sri, sizeof(struct scrub_item));
+ /*
+ * If the scrub item indicates that there is unchecked metadata, assume
+ * that the scrub type checker depends on something that couldn't be
+ * fixed. Mark that type as corrupt so that phase 4 will try it again.
+ */
+ foreach_scrub_type(scrub_type) {
+ __u8 *state = aitem->sri.sri_state;
+
+ if (state[scrub_type] & SCRUB_ITEM_NEEDSCHECK) {
+ state[scrub_type] &= ~SCRUB_ITEM_NEEDSCHECK;
+ state[scrub_type] |= SCRUB_ITEM_CORRUPT;
+ }
+ }
+
*aitemp = aitem;
return 0;
}
^ permalink raw reply related [flat|nested] 185+ messages in thread
end of thread, other threads:[~2024-01-01 1:07 UTC | newest]
Thread overview: 185+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2023-12-31 18:18 [NYE PATCHRIVER 2/4] xfs: online repair part 2 Darrick J. Wong
2023-12-31 19:32 ` [PATCHSET v13.0 1/7] xfs: design documentation for online fsck, " Darrick J. Wong
2023-12-31 20:42 ` [PATCH 1/4] docs: update the parent pointers documentation to the final version Darrick J. Wong
2023-12-31 20:42 ` [PATCH 2/4] docs: update online directory and parent pointer repair sections Darrick J. Wong
2023-12-31 20:42 ` [PATCH 3/4] docs: update offline parent pointer repair strategy Darrick J. Wong
2023-12-31 20:43 ` [PATCH 4/4] docs: describe xfs directory tree online fsck Darrick J. Wong
2023-12-31 19:33 ` [PATCHSET v13.0 2/7] xfs: retain ILOCK during directory updates Darrick J. Wong
2023-12-31 20:43 ` [PATCH 1/7] xfs: Increase XFS_DEFER_OPS_NR_INODES to 5 Darrick J. Wong
2023-12-31 20:43 ` [PATCH 2/7] xfs: Increase XFS_QM_TRANS_MAXDQS " Darrick J. Wong
2023-12-31 20:43 ` [PATCH 3/7] xfs: Hold inode locks in xfs_ialloc Darrick J. Wong
2023-12-31 20:44 ` [PATCH 4/7] xfs: Hold inode locks in xfs_trans_alloc_dir Darrick J. Wong
2023-12-31 20:44 ` [PATCH 5/7] xfs: Hold inode locks in xfs_rename Darrick J. Wong
2023-12-31 20:44 ` [PATCH 6/7] xfs: don't pick up IOLOCK during rmapbt repair scan Darrick J. Wong
2023-12-31 20:44 ` [PATCH 7/7] xfs: unlock new repair tempfiles after creation Darrick J. Wong
2023-12-31 19:33 ` [PATCHSET v13.0 3/7] xfs: name-value xattr lookups Darrick J. Wong
2023-12-31 20:45 ` [PATCH 01/14] xfs: require XFS_SB_FEAT_INCOMPAT_LOG_XATTRS for attr log intent item recovery Darrick J. Wong
2023-12-31 20:45 ` [PATCH 02/14] xfs: allow newer INCOMPAT/RO_COMPAT feature bits to protect ATTRI log items Darrick J. Wong
2023-12-31 20:45 ` [PATCH 03/14] xfs: check opcode and iovec count match in xlog_recover_attri_commit_pass2 Darrick J. Wong
2023-12-31 20:45 ` [PATCH 04/14] xfs: make xfs_attr_set require XFS_DA_OP_REMOVE Darrick J. Wong
2023-12-31 20:46 ` [PATCH 05/14] xfs: allow xattr matching on name and value for local/sf attrs Darrick J. Wong
2023-12-31 20:46 ` [PATCH 06/14] xfs: preserve NVLOOKUP in xfs_attr_set Darrick J. Wong
2023-12-31 20:46 ` [PATCH 07/14] xfs: restructure xfs_attr_complete_op a bit Darrick J. Wong
2023-12-31 20:46 ` [PATCH 08/14] xfs: use helpers to extract xattr op from opflags Darrick J. Wong
2023-12-31 20:47 ` [PATCH 09/14] xfs: validate recovered name buffers when recovering xattr items Darrick J. Wong
2023-12-31 20:47 ` [PATCH 10/14] xfs: always set args->value in xfs_attri_item_recover Darrick J. Wong
2023-12-31 20:47 ` [PATCH 11/14] xfs: use local variables for name and value length in _attri_commit_pass2 Darrick J. Wong
2023-12-31 20:47 ` [PATCH 12/14] xfs: log NVLOOKUP xattr removal operations Darrick J. Wong
2023-12-31 20:48 ` [PATCH 13/14] xfs: log NVLOOKUP xattr setting operations Darrick J. Wong
2023-12-31 20:48 ` [PATCH 14/14] xfs: log NVLOOKUP xattr nvreplace operations Darrick J. Wong
2023-12-31 19:33 ` [PATCHSET v13.0 4/7] xfs: Parent Pointers Darrick J. Wong
2023-12-31 20:48 ` [PATCH 01/18] xfs: Expose init_xattrs in xfs_create_tmpfile Darrick J. Wong
2023-12-31 20:49 ` [PATCH 02/18] xfs: add parent pointer support to attribute code Darrick J. Wong
2023-12-31 20:49 ` [PATCH 03/18] xfs: define parent pointer ondisk extended attribute format Darrick J. Wong
2023-12-31 20:49 ` [PATCH 04/18] xfs: add parent pointer validator functions Darrick J. Wong
2023-12-31 20:49 ` [PATCH 05/18] xfs: extend transaction reservations for parent attributes Darrick J. Wong
2023-12-31 20:50 ` [PATCH 06/18] xfs: parent pointer attribute creation Darrick J. Wong
2023-12-31 20:50 ` [PATCH 07/18] xfs: add parent attributes to link Darrick J. Wong
2023-12-31 20:50 ` [PATCH 08/18] xfs: add parent attributes to symlink Darrick J. Wong
2023-12-31 20:50 ` [PATCH 09/18] xfs: remove parent pointers in unlink Darrick J. Wong
2023-12-31 20:51 ` [PATCH 10/18] xfs: Add parent pointers to rename Darrick J. Wong
2023-12-31 20:51 ` [PATCH 11/18] xfs: Add parent pointers to xfs_cross_rename Darrick J. Wong
2023-12-31 20:51 ` [PATCH 12/18] xfs: Filter XFS_ATTR_PARENT for getfattr Darrick J. Wong
2023-12-31 20:51 ` [PATCH 13/18] xfs: pass the attr value to put_listent when possible Darrick J. Wong
2023-12-31 20:52 ` [PATCH 14/18] xfs: Add parent pointer ioctl Darrick J. Wong
2023-12-31 20:52 ` [PATCH 15/18] xfs: fix unit conversion error in xfs_log_calc_max_attrsetm_res Darrick J. Wong
2023-12-31 20:52 ` [PATCH 16/18] xfs: drop compatibility minimum log size computations for reflink Darrick J. Wong
2023-12-31 20:52 ` [PATCH 17/18] xfs: don't remove the attr fork when parent pointers are enabled Darrick J. Wong
2023-12-31 20:53 ` [PATCH 18/18] xfs: Add the parent pointer support to the superblock version 5 Darrick J. Wong
2023-12-31 19:33 ` [PATCHSET v13.0 5/7] xfs: fsck for parent pointers Darrick J. Wong
2023-12-31 20:53 ` [PATCH 01/22] xfs: check dirents have " Darrick J. Wong
2023-12-31 20:53 ` [PATCH 02/22] xfs: deferred scrub of dirents Darrick J. Wong
2023-12-31 20:53 ` [PATCH 03/22] xfs: create a parent pointer walk function for scrubbers Darrick J. Wong
2023-12-31 20:54 ` [PATCH 04/22] xfs: scrub parent pointers Darrick J. Wong
2023-12-31 20:54 ` [PATCH 05/22] xfs: deferred scrub of " Darrick J. Wong
2023-12-31 20:54 ` [PATCH 06/22] xfs: walk directory parent pointers to determine backref count Darrick J. Wong
2023-12-31 20:55 ` [PATCH 07/22] xfs: add raw parent pointer apis to support repair Darrick J. Wong
2023-12-31 20:55 ` [PATCH 08/22] xfs: set child file owner in xfs_da_args when changing parent pointers Darrick J. Wong
2023-12-31 20:55 ` [PATCH 09/22] xfs: salvage parent pointers when rebuilding xattr structures Darrick J. Wong
2023-12-31 20:55 ` [PATCH 10/22] xfs: replace namebuf with parent pointer in directory repair Darrick J. Wong
2023-12-31 20:56 ` [PATCH 11/22] xfs: repair directories by scanning directory parent pointers Darrick J. Wong
2023-12-31 20:56 ` [PATCH 12/22] xfs: implement live updates for directory repairs Darrick J. Wong
2023-12-31 20:56 ` [PATCH 13/22] xfs: replay unlocked parent pointer updates that accrue during xattr repair Darrick J. Wong
2023-12-31 20:56 ` [PATCH 14/22] xfs: replace namebuf with parent pointer in parent pointer repair Darrick J. Wong
2023-12-31 20:57 ` [PATCH 15/22] xfs: repair directory parent pointers by scanning for dirents Darrick J. Wong
2023-12-31 20:57 ` [PATCH 16/22] xfs: implement live updates for parent pointer repairs Darrick J. Wong
2023-12-31 20:57 ` [PATCH 17/22] xfs: remove pointless unlocked assertion Darrick J. Wong
2023-12-31 20:57 ` [PATCH 18/22] xfs: split xfs_bmap_add_attrfork into two pieces Darrick J. Wong
2023-12-31 20:58 ` [PATCH 19/22] xfs: actually rebuild the parent pointer xattrs Darrick J. Wong
2023-12-31 20:58 ` [PATCH 20/22] xfs: adapt the orphanage code to handle parent pointers Darrick J. Wong
2023-12-31 20:58 ` [PATCH 21/22] xfs: repair link count of nondirectories after rebuilding " Darrick J. Wong
2023-12-31 20:58 ` [PATCH 22/22] xfs: inode repair should ensure there's an attr fork to store " Darrick J. Wong
2023-12-31 19:34 ` [PATCHSET v13.0 6/7] xfs: detect and correct directory tree problems Darrick J. Wong
2023-12-31 20:59 ` [PATCH 1/4] xfs: teach online scrub to find directory tree structure problems Darrick J. Wong
2023-12-31 20:59 ` [PATCH 2/4] xfs: invalidate dirloop scrub path data when concurrent updates happen Darrick J. Wong
2023-12-31 20:59 ` [PATCH 3/4] xfs: report directory tree corruption in the health information Darrick J. Wong
2023-12-31 20:59 ` [PATCH 4/4] xfs: fix corruptions in the directory tree Darrick J. Wong
2023-12-31 19:34 ` [PATCHSET v13.0 7/7] xfs: vectorize scrub kernel calls Darrick J. Wong
2023-12-31 21:00 ` [PATCH 1/3] xfs: reduce the rate of cond_resched calls inside scrub Darrick J. Wong
2023-12-31 21:00 ` [PATCH 2/3] xfs: introduce vectored scrub mode Darrick J. Wong
2023-12-31 21:00 ` [PATCH 3/3] xfs: only iget the file once when doing vectored scrub-by-handle Darrick J. Wong
2023-12-31 19:50 ` [PATCHSET v13.0 1/6] xfsprogs: retain ILOCK during directory updates Darrick J. Wong
2023-12-31 23:02 ` [PATCH 1/1] xfs: Increase XFS_DEFER_OPS_NR_INODES to 5 Darrick J. Wong
2023-12-31 19:50 ` [PATCHSET v13.0 2/6] xfsprogs: name-value xattr lookups Darrick J. Wong
2023-12-31 23:02 ` [PATCH 01/11] xfs: allow newer INCOMPAT/RO_COMPAT feature bits to protect ATTRI log items Darrick J. Wong
2023-12-31 23:02 ` [PATCH 02/11] xfs: make xfs_attr_set require XFS_DA_OP_REMOVE Darrick J. Wong
2023-12-31 23:03 ` [PATCH 03/11] xfs: allow xattr matching on name and value for local/sf attrs Darrick J. Wong
2023-12-31 23:03 ` [PATCH 04/11] xfs: preserve NVLOOKUP in xfs_attr_set Darrick J. Wong
2023-12-31 23:03 ` [PATCH 05/11] xfs: restructure xfs_attr_complete_op a bit Darrick J. Wong
2023-12-31 23:03 ` [PATCH 06/11] xfs: use helpers to extract xattr op from opflags Darrick J. Wong
2023-12-31 23:04 ` [PATCH 07/11] xfs: log NVLOOKUP xattr removal operations Darrick J. Wong
2023-12-31 23:04 ` [PATCH 08/11] xfs: log NVLOOKUP xattr setting operations Darrick J. Wong
2023-12-31 23:04 ` [PATCH 09/11] xfs: log NVLOOKUP xattr nvreplace operations Darrick J. Wong
2023-12-31 23:04 ` [PATCH 10/11] xfs_logprint: dump new attr log item fields Darrick J. Wong
2023-12-31 23:05 ` [PATCH 11/11] xfs_logprint: print missing attri header fields Darrick J. Wong
2023-12-31 19:50 ` [PATCHSET v13.0 3/6] xfsprogs: Parent Pointers Darrick J. Wong
2023-12-31 23:05 ` [PATCH 01/32] xfs: add parent pointer support to attribute code Darrick J. Wong
2023-12-31 23:05 ` [PATCH 02/32] xfs: define parent pointer ondisk extended attribute format Darrick J. Wong
2023-12-31 23:05 ` [PATCH 03/32] xfs: add parent pointer validator functions Darrick J. Wong
2023-12-31 23:06 ` [PATCH 04/32] xfs: extend transaction reservations for parent attributes Darrick J. Wong
2023-12-31 23:06 ` [PATCH 05/32] xfs: parent pointer attribute creation Darrick J. Wong
2023-12-31 23:06 ` [PATCH 06/32] xfs: add parent attributes to link Darrick J. Wong
2023-12-31 23:06 ` [PATCH 07/32] xfs: add parent attributes to symlink Darrick J. Wong
2023-12-31 23:07 ` [PATCH 08/32] xfs: remove parent pointers in unlink Darrick J. Wong
2023-12-31 23:07 ` [PATCH 09/32] xfs: Add parent pointers to rename Darrick J. Wong
2023-12-31 23:07 ` [PATCH 10/32] xfs: pass the attr value to put_listent when possible Darrick J. Wong
2023-12-31 23:07 ` [PATCH 11/32] xfs: Add parent pointer ioctl Darrick J. Wong
2023-12-31 23:08 ` [PATCH 12/32] xfs: fix unit conversion error in xfs_log_calc_max_attrsetm_res Darrick J. Wong
2023-12-31 23:08 ` [PATCH 13/32] xfs: drop compatibility minimum log size computations for reflink Darrick J. Wong
2023-12-31 23:08 ` [PATCH 14/32] xfs: don't remove the attr fork when parent pointers are enabled Darrick J. Wong
2023-12-31 23:09 ` [PATCH 15/32] xfs: Add the parent pointer support to the superblock version 5 Darrick J. Wong
2023-12-31 23:09 ` [PATCH 16/32] libfrog: add parent pointer support code Darrick J. Wong
2023-12-31 23:09 ` [PATCH 17/32] libfrog: detect looping paths when walking directory parent pointers Darrick J. Wong
2023-12-31 23:09 ` [PATCH 18/32] xfs_io: adapt parent command to new parent pointer ioctls Darrick J. Wong
2023-12-31 23:10 ` [PATCH 19/32] xfs_io: Add i, n and f flags to parent command Darrick J. Wong
2023-12-31 23:10 ` [PATCH 20/32] xfs_logprint: decode parent pointers in ATTRI items fully Darrick J. Wong
2023-12-31 23:10 ` [PATCH 21/32] xfs_spaceman: report file paths Darrick J. Wong
2023-12-31 23:10 ` [PATCH 22/32] xfs_scrub: use parent pointers when possible to report file operations Darrick J. Wong
2023-12-31 23:11 ` [PATCH 23/32] xfs_db: report parent pointers in version command Darrick J. Wong
2023-12-31 23:11 ` [PATCH 24/32] xfs_db: report parent bit on xattrs Darrick J. Wong
2023-12-31 23:11 ` [PATCH 25/32] xfs_db: report parent pointers embedded in xattrs Darrick J. Wong
2023-12-31 23:11 ` [PATCH 26/32] xfs_db: obfuscate dirent and parent pointer names consistently Darrick J. Wong
2023-12-31 23:12 ` [PATCH 27/32] libxfs: export attr3_leaf_hdr_from_disk via libxfs_api_defs.h Darrick J. Wong
2023-12-31 23:12 ` [PATCH 28/32] xfs_db: add a parents command to list the parents of a file Darrick J. Wong
2023-12-31 23:12 ` [PATCH 29/32] libxfs: create new files with attr forks if necessary Darrick J. Wong
2023-12-31 23:12 ` [PATCH 30/32] xfsprogs: Fix default superblock attr bits Darrick J. Wong
2023-12-31 23:13 ` [PATCH 31/32] mkfs: Add parent pointers during protofile creation Darrick J. Wong
2023-12-31 23:13 ` [PATCH 32/32] mkfs: enable formatting with parent pointers Darrick J. Wong
2023-12-31 19:51 ` [PATCHSET v13.0 4/6] xfsprogs: fsck for " Darrick J. Wong
2023-12-31 23:13 ` [PATCH 01/18] xfs: create a blob array data structure Darrick J. Wong
2023-12-31 23:13 ` [PATCH 02/18] xfs: check dirents have parent pointers Darrick J. Wong
2023-12-31 23:14 ` [PATCH 03/18] man2: update ioctl_xfs_scrub_metadata.2 for " Darrick J. Wong
2023-12-31 23:14 ` [PATCH 04/18] xfs: add raw parent pointer apis to support repair Darrick J. Wong
2023-12-31 23:14 ` [PATCH 05/18] xfs: set child file owner in xfs_da_args when changing parent pointers Darrick J. Wong
2023-12-31 23:15 ` [PATCH 06/18] xfs: remove pointless unlocked assertion Darrick J. Wong
2023-12-31 23:15 ` [PATCH 07/18] xfs: split xfs_bmap_add_attrfork into two pieces Darrick J. Wong
2023-12-31 23:15 ` [PATCH 08/18] xfs: actually rebuild the parent pointer xattrs Darrick J. Wong
2023-12-31 23:15 ` [PATCH 09/18] xfs_repair: add parent pointers when messing with /lost+found Darrick J. Wong
2023-12-31 23:16 ` [PATCH 10/18] xfs_repair: build a parent pointer index Darrick J. Wong
2023-12-31 23:16 ` [PATCH 11/18] xfs_repair: move the global dirent name store to a separate object Darrick J. Wong
2023-12-31 23:16 ` [PATCH 12/18] xfs_repair: deduplicate strings stored in string blob Darrick J. Wong
2023-12-31 23:16 ` [PATCH 13/18] xfs_repair: check parent pointers Darrick J. Wong
2023-12-31 23:17 ` [PATCH 14/18] xfs_repair: dump garbage parent pointer attributes Darrick J. Wong
2023-12-31 23:17 ` [PATCH 15/18] xfs_repair: update ondisk parent pointer records Darrick J. Wong
2023-12-31 23:17 ` [PATCH 16/18] xfs_repair: wipe ondisk parent pointers when there are none Darrick J. Wong
2023-12-31 23:17 ` [PATCH 17/18] xfs_repair: upgrade an existing filesystem to have parent pointers Darrick J. Wong
2023-12-31 23:18 ` [PATCH 18/18] xfs_scrub: use parent pointers to report lost file data Darrick J. Wong
2023-12-31 19:51 ` [PATCHSET v13.0 5/6] xfs: detect and correct directory tree problems Darrick J. Wong
2023-12-31 23:18 ` [PATCH 1/6] xfs_db: add link and unlink expert commands Darrick J. Wong
2023-12-31 23:18 ` [PATCH 2/6] xfs: teach online scrub to find directory tree structure problems Darrick J. Wong
2023-12-31 23:18 ` [PATCH 3/6] xfs: report directory tree corruption in the health information Darrick J. Wong
2023-12-31 23:19 ` [PATCH 4/6] xfs_scrub: fix erroring out of check_inode_names Darrick J. Wong
2023-12-31 23:19 ` [PATCH 5/6] xfs_scrub: detect and repair directory tree corruptions Darrick J. Wong
2023-12-31 23:19 ` [PATCH 6/6] xfs_scrub: defer phase5 file scans if dirloop fails Darrick J. Wong
2023-12-31 19:51 ` [PATCHSET 6/6] xfs_scrub: vectorize kernel calls Darrick J. Wong
2023-12-31 23:19 ` [PATCH 01/10] xfs: introduce vectored scrub mode Darrick J. Wong
2023-12-31 23:20 ` [PATCH 02/10] libfrog: support vectored scrub Darrick J. Wong
2023-12-31 23:20 ` [PATCH 03/10] xfs_io: " Darrick J. Wong
2023-12-31 23:20 ` [PATCH 04/10] xfs_scrub: split the scrub epilogue code into a separate function Darrick J. Wong
2023-12-31 23:21 ` [PATCH 05/10] xfs_scrub: split the repair " Darrick J. Wong
2023-12-31 23:21 ` [PATCH 06/10] xfs_scrub: convert scrub and repair epilogues to use xfs_scrub_vec Darrick J. Wong
2023-12-31 23:21 ` [PATCH 07/10] xfs_scrub: vectorize scrub calls Darrick J. Wong
2023-12-31 23:21 ` [PATCH 08/10] xfs_scrub: vectorize repair calls Darrick J. Wong
2023-12-31 23:22 ` [PATCH 09/10] xfs_scrub: use scrub barriers to reduce kernel calls Darrick J. Wong
2023-12-31 23:22 ` [PATCH 10/10] xfs_scrub: try spot repairs of metadata items to make scrub progress Darrick J. Wong
2023-12-31 19:59 ` [PATCHSET v13.0 1/3] fstests: adjust tests for xfs parent pointers Darrick J. Wong
2023-12-27 13:46 ` [PATCH 01/11] generic: test recovery of extended attribute updates Darrick J. Wong
2023-12-27 13:46 ` [PATCH 02/11] xfs/206: filter out the parent= status from mkfs Darrick J. Wong
2023-12-27 13:47 ` [PATCH 03/11] xfs/122: update for parent pointers Darrick J. Wong
2023-12-27 13:47 ` [PATCH 04/11] populate: create hardlinks " Darrick J. Wong
2023-12-27 13:47 ` [PATCH 05/11] xfs/021: adapt golden output files " Darrick J. Wong
2023-12-27 13:47 ` [PATCH 06/11] xfs/{018,191,288}: disable parent pointers for this test Darrick J. Wong
2023-12-27 13:48 ` [PATCH 07/11] xfs/306: fix formatting failures with parent pointers Darrick J. Wong
2023-12-27 13:48 ` [PATCH 08/11] common: add helpers for parent pointer tests Darrick J. Wong
2023-12-27 13:48 ` [PATCH 09/11] xfs: add parent pointer test Darrick J. Wong
2023-12-27 13:48 ` [PATCH 10/11] xfs: add multi link " Darrick J. Wong
2023-12-27 13:49 ` [PATCH 11/11] xfs: add parent pointer inject test Darrick J. Wong
2023-12-31 19:59 ` [PATCHSET v13.0 2/3] xfs: detect and correct directory tree structures Darrick J. Wong
2023-12-27 13:49 ` [PATCH 1/2] common/fuzzy: stress directory tree modifications with the dirtree tester Darrick J. Wong
2023-12-27 13:49 ` [PATCH 2/2] scrub: test correction of directory tree corruptions Darrick J. Wong
2023-12-31 19:59 ` [PATCHSET 3/3] xfs_scrub: vectorize kernel calls Darrick J. Wong
2023-12-27 13:49 ` [PATCH 1/1] xfs/122: update for vectored scrub Darrick J. Wong
2023-12-31 20:03 ` [PATCHSET v13.0 1/2] xfs-documentation: document attr log item changes for parent pointers Darrick J. Wong
2023-12-27 14:07 ` [PATCH 1/1] design: document new name-value logged attribute variants Darrick J. Wong
2023-12-31 20:03 ` [PATCHSET v13.0 2/2] xfs-documentation: document parent pointers Darrick J. Wong
2023-12-27 14:07 ` [PATCH 1/1] design: document the parent pointer ondisk format Darrick J. Wong
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).