public inbox for linux-btrfs@vger.kernel.org
 help / color / mirror / Atom feed
From: Christoph Hellwig <hch@lst.de>
To: fstests@vger.kernel.org
Cc: linux-btrfs@vger.kernel.org
Subject: Re: [PATCH 4/4] btrfs: test read repair on a corrupted compressed extent
Date: Wed, 22 Jun 2022 11:21:40 +0200	[thread overview]
Message-ID: <20220622092140.GA26204@lst.de> (raw)
In-Reply-To: <20220622045844.3219390-5-hch@lst.de>

So while this test properly documents the current behavior, it failed
to grasp how broken that behavior ist: the current read repair code
writes back the uncompressed data to disk even for a compressed extent,
and this test verified the behavior.

Below is a correct test that fails on current mainline.  I'll send fixes
but right now they depend on a lot of prep work.

---
From 6b6c505f75c6c7cc15359f14053b1db43e3d3091 Mon Sep 17 00:00:00 2001
From: Christoph Hellwig <hch@lst.de>
Date: Wed, 22 Jun 2022 06:55:36 +0200
Subject: btrfs: test read repair on a corrupted compressed extent

Exercise read repair on a corrupted compressed sector.

Signed-off-by: Christoph Hellwig <hch@lst.de>
---
 tests/btrfs/270     | 82 +++++++++++++++++++++++++++++++++++++++++++++
 tests/btrfs/270.out |  7 ++++
 2 files changed, 89 insertions(+)
 create mode 100755 tests/btrfs/270
 create mode 100644 tests/btrfs/270.out

diff --git a/tests/btrfs/270 b/tests/btrfs/270
new file mode 100755
index 00000000..5b73fb15
--- /dev/null
+++ b/tests/btrfs/270
@@ -0,0 +1,82 @@
+#! /bin/bash
+# SPDX-License-Identifier: GPL-2.0
+# Copyright (c) 2017 Liu Bo.  All Rights Reserved.
+#
+# FS QA Test 270
+#
+# Regression test for btrfs buffered read repair of compressed data.
+#
+. ./common/preamble
+_begin_fstest auto quick read_repair compress
+
+. ./common/filter
+
+_supported_fs btrfs
+_require_btrfs_command inspect-internal dump-tree
+_require_non_zoned_device "${SCRATCH_DEV}" # no overwrites on zoned devices
+_require_scratch_dev_pool 2
+_scratch_dev_pool_get 2
+
+get_physical()
+{
+	local logical=$1
+	local stripe=$2
+	$BTRFS_UTIL_PROG inspect-internal dump-tree -t 3 $SCRATCH_DEV | \
+		grep $logical -A 6 | \
+		$AWK_PROG "(\$1 ~ /stripe/ && \$3 ~ /devid/ && \$2 ~ /$stripe/) { print \$6 }"
+}
+
+get_devid()
+{
+	local logical=$1
+	local stripe=$2
+	$BTRFS_UTIL_PROG inspect-internal dump-tree -t 3 $SCRATCH_DEV | \
+		grep $logical -A 6 | \
+		$AWK_PROG "(\$1 ~ /stripe/ && \$3 ~ /devid/ && \$2 ~ /$stripe/) { print \$4 }"
+}
+
+get_device_path()
+{
+	local devid=$1
+	echo "$SCRATCH_DEV_POOL" | $AWK_PROG "{print \$$devid}"
+}
+
+
+echo "step 1......mkfs.btrfs"
+_check_minimal_fs_size $(( 1024 * 1024 * 1024 ))
+_scratch_pool_mkfs "-d raid1 -b 1G" >>$seqres.full 2>&1
+_scratch_mount -ocompress
+
+# Create a file with all data being compressed
+$XFS_IO_PROG -f -c "pwrite -S 0xaa -W -b 128K 0 128K" \
+	"$SCRATCH_MNT/foobar" | _filter_xfs_io_offset
+
+logical_in_btrfs=$(_btrfs_get_first_logical $SCRATCH_MNT/foobar)
+physical=$(get_physical ${logical_in_btrfs} 1)
+devid=$(get_devid ${logical_in_btrfs} 1)
+devpath=$(get_device_path ${devid})
+
+_scratch_unmount
+echo "step 2......corrupt file extent"
+echo " corrupt stripe #1, devid $devid devpath $devpath physical $physical" \
+	>> $seqres.full
+dd if=$devpath of=$TEST_DIR/$seq.dump.good skip=$physical bs=1 count=4096 \
+	2>/dev/null
+$XFS_IO_PROG -c "pwrite -S 0xbb -b 4K $physical 4K" $devpath > /dev/null
+
+_scratch_mount
+
+echo "step 3......repair the bad copy"
+_btrfs_buffered_read_on_mirror 1 2 "$SCRATCH_MNT/foobar" 0 128K
+
+_scratch_unmount
+
+echo "step 4......check if the repair worked"
+dd if=$devpath of=$TEST_DIR/$seq.dump skip=$physical bs=1 count=4096 \
+	2>/dev/null
+cmp -bl $TEST_DIR/$seq.dump.good $TEST_DIR/$seq.dump
+
+_scratch_dev_pool_put
+# success, all done
+status=0
+exit
diff --git a/tests/btrfs/270.out b/tests/btrfs/270.out
new file mode 100644
index 00000000..6d744c02
--- /dev/null
+++ b/tests/btrfs/270.out
@@ -0,0 +1,7 @@
+QA output created by 270
+step 1......mkfs.btrfs
+wrote 131072/131072 bytes
+XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
+step 2......corrupt file extent
+step 3......repair the bad copy
+step 4......check if the repair worked
-- 
2.30.2


  reply	other threads:[~2022-06-22  9:21 UTC|newest]

Thread overview: 10+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-06-22  4:58 btrfs read repair: more tests Christoph Hellwig
2022-06-22  4:58 ` [PATCH 1/4] btrfs: fix the_btrfs_get_physical invocation in btrfs-map-logical Christoph Hellwig
2022-06-22  4:58 ` [PATCH 2/4] btrfs; add a test for impossible repair cases Christoph Hellwig
2022-06-22  4:58 ` [PATCH 3/4] btrfs: test checker pattern corruption on raid10 Christoph Hellwig
2022-06-22  4:58 ` [PATCH 4/4] btrfs: test read repair on a corrupted compressed extent Christoph Hellwig
2022-06-22  9:21   ` Christoph Hellwig [this message]
2022-06-22 12:41     ` Zorro Lang
2022-06-22 13:07       ` Christoph Hellwig
2022-06-24  2:25         ` Zorro Lang
2022-06-29 14:00 ` btrfs read repair: more tests Zorro Lang

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20220622092140.GA26204@lst.de \
    --to=hch@lst.de \
    --cc=fstests@vger.kernel.org \
    --cc=linux-btrfs@vger.kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox