[Qemu-devel] [PATCH v2 0/2] block/file-posix: Fix unaligned O_DIRECT block status

[Qemu-devel] [PATCH v2 0/2] block/file-posix: Fix unaligned O_DIRECT block status

[Max Reitz]

The user-visible problem:
$ echo > foo
$ qemu-img map --image-opts driver=file,filename=foo,cache.direct=on
Offset          Length          Mapped to       File
qemu-img: block/io.c:2093: bdrv_co_block_status: Assertion `*pnum &&
QEMU_IS_ALIGNED(*pnum, align) && align > offset - aligned_offset'
failed.

The internal problem: file-posix truncates status requests to the EOF.
If the EOF is not aligned at the request_alignment,
bdrv_co_block_status() won't like that.

See patch 1 for a deeper discussion (including two possible alternatives
how we could address the problem).
(As I note there, I’ve looked through all block drivers, and I didn’t
find any other which could have the same problem.  gluster uses the same
block-status code, but it doesn’t set a request_alignment.  NBD
force-aligns the server response in nbd_parse_blockstatus_payload().
qcow2... Should be fine as long as no crypto driver has a block limit
exceeding the qcow2 cluster size.  And so on.)

Patch 2 adds a test.  After writing that test, I noticed that we already
had one: 109 fails with -c none before patch 1.  Er, well, at least the
new test is more succinct and has the correct default cache mode, so it
will actually do the test if you run ./check without enforcing any cache
on a filesystem that supports O_DIRECT.


v2:
- Shortened the assert() in patch 1 [Eric]


git backport-diff against v1:

Key:
[----] : patches are identical
[####] : number of functional differences between upstream/downstream patch
[down] : patch is downstream-only
The flags [FC] indicate (F)unctional and (C)ontextual differences, respectively

001/2:[0003] [FC] 'block/file-posix: Unaligned O_DIRECT block-status'
002/2:[----] [--] 'iotests: Test unaligned raw images with O_DIRECT'


Max Reitz (2):
  block/file-posix: Unaligned O_DIRECT block-status
  iotests: Test unaligned raw images with O_DIRECT

 block/file-posix.c         | 16 ++++++++
 tests/qemu-iotests/221     |  4 ++
 tests/qemu-iotests/253     | 84 ++++++++++++++++++++++++++++++++++++++
 tests/qemu-iotests/253.out | 14 +++++++
 tests/qemu-iotests/group   |  1 +
 5 files changed, 119 insertions(+)
 create mode 100755 tests/qemu-iotests/253
 create mode 100644 tests/qemu-iotests/253.out

-- 
2.21.0

[Qemu-devel] [PATCH v2 1/2] block/file-posix: Unaligned O_DIRECT block-status

[Max Reitz]

Currently, qemu crashes whenever someone queries the block status of an
unaligned image tail of an O_DIRECT image:
$ echo > foo
$ qemu-img map --image-opts driver=file,filename=foo,cache.direct=on
Offset          Length          Mapped to       File
qemu-img: block/io.c:2093: bdrv_co_block_status: Assertion `*pnum &&
QEMU_IS_ALIGNED(*pnum, align) && align > offset - aligned_offset'
failed.

This is because bdrv_co_block_status() checks that the result returned
by the driver's implementation is aligned to the request_alignment, but
file-posix can fail to do so, which is actually mentioned in a comment
there: "[...] possibly including a partial sector at EOF".

Fix this by rounding up those partial sectors.

There are two possible alternative fixes:
(1) We could refuse to open unaligned image files with O_DIRECT
    altogether.  That sounds reasonable until you realize that qcow2
    does necessarily not fill up its metadata clusters, and that nobody
    runs qemu-img create with O_DIRECT.  Therefore, unpreallocated qcow2
    files usually have an unaligned image tail.

(2) bdrv_co_block_status() could ignore unaligned tails.  It actually
    throws away everything past the EOF already, so that sounds
    reasonable.
    Unfortunately, the block layer knows file lengths only with a
    granularity of BDRV_SECTOR_SIZE, so bdrv_co_block_status() usually
    would have to guess whether its file length information is inexact
    or whether the driver is broken.

Fixing what raw_co_block_status() returns is the safest thing to do.

There seems to be no other block driver that sets request_alignment and
does not make sure that it always returns aligned values.

Cc: qemu-stable@nongnu.org
Signed-off-by: Max Reitz <mreitz@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
---
 block/file-posix.c | 16 ++++++++++++++++
 1 file changed, 16 insertions(+)

diff --git a/block/file-posix.c b/block/file-posix.c
index e09e15bbf8..d018429672 100644
--- a/block/file-posix.c
+++ b/block/file-posix.c
@@ -2488,6 +2488,8 @@ static int coroutine_fn raw_co_block_status(BlockDriverState *bs,
     off_t data = 0, hole = 0;
     int ret;
 
+    assert(QEMU_IS_ALIGNED(offset | bytes, bs->bl.request_alignment));
+
     ret = fd_open(bs);
     if (ret < 0) {
         return ret;
@@ -2513,6 +2515,20 @@ static int coroutine_fn raw_co_block_status(BlockDriverState *bs,
         /* On a data extent, compute bytes to the end of the extent,
          * possibly including a partial sector at EOF. */
         *pnum = MIN(bytes, hole - offset);
+
+        /*
+         * We are not allowed to return partial sectors, though, so
+         * round up if necessary.
+         */
+        if (!QEMU_IS_ALIGNED(*pnum, bs->bl.request_alignment)) {
+            int64_t file_length = raw_getlength(bs);
+            if (file_length > 0) {
+                /* Ignore errors, this is just a safeguard */
+                assert(hole == file_length);
+            }
+            *pnum = ROUND_UP(*pnum, bs->bl.request_alignment);
+        }
+
         ret = BDRV_BLOCK_DATA;
     } else {
         /* On a hole, compute bytes to the beginning of the next extent.  */
-- 
2.21.0

[Qemu-devel] [PATCH v2 2/2] iotests: Test unaligned raw images with O_DIRECT

[Max Reitz]

We already have 221 for accesses through the page cache, but it is
better to create a new file for O_DIRECT instead of integrating those
test cases into 221.  This way, we can make use of
_supported_cache_modes (and _default_cache_mode) so the test is
automatically skipped on filesystems that do not support O_DIRECT.

As part of the split, add _supported_cache_modes to 221.  With that, it
no longer fails when run with -c none or -c directsync.

Signed-off-by: Max Reitz <mreitz@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
---
 tests/qemu-iotests/221     |  4 ++
 tests/qemu-iotests/253     | 84 ++++++++++++++++++++++++++++++++++++++
 tests/qemu-iotests/253.out | 14 +++++++
 tests/qemu-iotests/group   |  1 +
 4 files changed, 103 insertions(+)
 create mode 100755 tests/qemu-iotests/253
 create mode 100644 tests/qemu-iotests/253.out

diff --git a/tests/qemu-iotests/221 b/tests/qemu-iotests/221
index 25dd47bcfe..0e9096fec7 100755
--- a/tests/qemu-iotests/221
+++ b/tests/qemu-iotests/221
@@ -1,6 +1,7 @@
 #!/usr/bin/env bash
 #
 # Test qemu-img vs. unaligned images
+# (See also 253, which is the O_DIRECT version)
 #
 # Copyright (C) 2018-2019 Red Hat, Inc.
 #
@@ -37,6 +38,9 @@ _supported_fmt raw
 _supported_proto file
 _supported_os Linux
 
+_default_cache_mode writeback
+_supported_cache_modes writeback writethrough unsafe
+
 echo
 echo "=== Check mapping of unaligned raw image ==="
 echo
diff --git a/tests/qemu-iotests/253 b/tests/qemu-iotests/253
new file mode 100755
index 0000000000..d88d5afa45
--- /dev/null
+++ b/tests/qemu-iotests/253
@@ -0,0 +1,84 @@
+#!/usr/bin/env bash
+#
+# Test qemu-img vs. unaligned images; O_DIRECT version
+# (Originates from 221)
+#
+# Copyright (C) 2019 Red Hat, Inc.
+#
+# This program is free software; you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation; either version 2 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program.  If not, see <http://www.gnu.org/licenses/>.
+#
+
+seq="$(basename $0)"
+echo "QA output created by $seq"
+
+status=1 # failure is the default!
+
+_cleanup()
+{
+    _cleanup_test_img
+}
+trap "_cleanup; exit \$status" 0 1 2 3 15
+
+# get standard environment, filters and checks
+. ./common.rc
+. ./common.filter
+
+_supported_fmt raw
+_supported_proto file
+_supported_os Linux
+
+_default_cache_mode none
+_supported_cache_modes none directsync
+
+echo
+echo "=== Check mapping of unaligned raw image ==="
+echo
+
+# We do not know how large a physical sector is, but it is certainly
+# going to be a factor of 1 MB
+size=$((1 * 1024 * 1024 - 1))
+
+# qemu-img create rounds size up to BDRV_SECTOR_SIZE
+_make_test_img $size
+$QEMU_IMG map --output=json --image-opts \
+    "driver=$IMGFMT,file.driver=file,file.filename=$TEST_IMG,cache.direct=on" \
+    | _filter_qemu_img_map
+
+# so we resize it and check again
+truncate --size=$size "$TEST_IMG"
+$QEMU_IMG map --output=json --image-opts \
+    "driver=$IMGFMT,file.driver=file,file.filename=$TEST_IMG,cache.direct=on" \
+    | _filter_qemu_img_map
+
+# qemu-io with O_DIRECT always writes whole physical sectors.  Again,
+# we do not know how large a physical sector is, so we just start
+# writing from a 64 kB boundary, which should always be aligned.
+offset=$((1 * 1024 * 1024 - 64 * 1024))
+$QEMU_IO -c "w $offset $((size - offset))" "$TEST_IMG" | _filter_qemu_io
+$QEMU_IMG map --output=json --image-opts \
+    "driver=$IMGFMT,file.driver=file,file.filename=$TEST_IMG,cache.direct=on" \
+    | _filter_qemu_img_map
+
+# Resize it and check again -- contrary to 221, we may not get partial
+# sectors here, so there should be only two areas (one zero, one
+# data).
+truncate --size=$size "$TEST_IMG"
+$QEMU_IMG map --output=json --image-opts \
+    "driver=$IMGFMT,file.driver=file,file.filename=$TEST_IMG,cache.direct=on" \
+    | _filter_qemu_img_map
+
+# success, all done
+echo '*** done'
+rm -f $seq.full
+status=0
diff --git a/tests/qemu-iotests/253.out b/tests/qemu-iotests/253.out
new file mode 100644
index 0000000000..607c0baa0b
--- /dev/null
+++ b/tests/qemu-iotests/253.out
@@ -0,0 +1,14 @@
+QA output created by 253
+
+=== Check mapping of unaligned raw image ===
+
+Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1048575
+[{ "start": 0, "length": 1048576, "depth": 0, "zero": true, "data": false, "offset": OFFSET}]
+[{ "start": 0, "length": 1048576, "depth": 0, "zero": true, "data": false, "offset": OFFSET}]
+wrote 65535/65535 bytes at offset 983040
+63.999 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
+[{ "start": 0, "length": 983040, "depth": 0, "zero": true, "data": false, "offset": OFFSET},
+{ "start": 983040, "length": 65536, "depth": 0, "zero": false, "data": true, "offset": OFFSET}]
+[{ "start": 0, "length": 983040, "depth": 0, "zero": true, "data": false, "offset": OFFSET},
+{ "start": 983040, "length": 65536, "depth": 0, "zero": false, "data": true, "offset": OFFSET}]
+*** done
diff --git a/tests/qemu-iotests/group b/tests/qemu-iotests/group
index 00e474ab0a..52b7c16e15 100644
--- a/tests/qemu-iotests/group
+++ b/tests/qemu-iotests/group
@@ -250,3 +250,4 @@
 248 rw auto quick
 249 rw auto quick
 252 rw auto backing quick
+253 rw auto quick
-- 
2.21.0

Re: [Qemu-devel] [PATCH v2 0/2] block/file-posix: Fix unaligned O_DIRECT block status

[Kevin Wolf]

Am 15.05.2019 um 06:15 hat Max Reitz geschrieben:
> The user-visible problem:
> $ echo > foo
> $ qemu-img map --image-opts driver=file,filename=foo,cache.direct=on
> Offset          Length          Mapped to       File
> qemu-img: block/io.c:2093: bdrv_co_block_status: Assertion `*pnum &&
> QEMU_IS_ALIGNED(*pnum, align) && align > offset - aligned_offset'
> failed.
> 
> The internal problem: file-posix truncates status requests to the EOF.
> If the EOF is not aligned at the request_alignment,
> bdrv_co_block_status() won't like that.
> 
> See patch 1 for a deeper discussion (including two possible alternatives
> how we could address the problem).
> (As I note there, I’ve looked through all block drivers, and I didn’t
> find any other which could have the same problem.  gluster uses the same
> block-status code, but it doesn’t set a request_alignment.  NBD
> force-aligns the server response in nbd_parse_blockstatus_payload().
> qcow2... Should be fine as long as no crypto driver has a block limit
> exceeding the qcow2 cluster size.  And so on.)
> 
> Patch 2 adds a test.  After writing that test, I noticed that we already
> had one: 109 fails with -c none before patch 1.  Er, well, at least the
> new test is more succinct and has the correct default cache mode, so it
> will actually do the test if you run ./check without enforcing any cache
> on a filesystem that supports O_DIRECT.

Thanks, applied to the block branch.

Kevin