All of lore.kernel.org
 help / color / mirror / Atom feed
From: "Darrick J. Wong" <djwong@kernel.org>
To: djwong@kernel.org, guaneryu@gmail.com
Cc: linux-xfs@vger.kernel.org, fstests@vger.kernel.org, guan@eryu.me
Subject: [PATCH 2/3] xfs: test correct propagation of rt extent size hints on rtinherit dirs
Date: Tue, 31 Aug 2021 17:11:35 -0700	[thread overview]
Message-ID: <163045509592.769915.9044627867698975012.stgit@magnolia> (raw)
In-Reply-To: <163045508495.769915.4859353445119566326.stgit@magnolia>

From: Darrick J. Wong <djwong@kernel.org>

This is a regression test for the following fixes:

 xfs: standardize extent size hint validation
 xfs: validate extsz hints against rt extent size when rtinherit is set
 mkfs: validate rt extent size hint when rtinherit is set

These patches fix inadequate rtextsize alignment validation of extent
size hints on directories with the rtinherit and extszinherit flags set.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
 tests/xfs/774     |   80 +++++++++++++++++++++++++++++++++++++++++++++++++++++
 tests/xfs/774.out |    5 +++
 tests/xfs/776     |   57 ++++++++++++++++++++++++++++++++++++++
 tests/xfs/776.out |    5 +++
 4 files changed, 147 insertions(+)
 create mode 100755 tests/xfs/774
 create mode 100644 tests/xfs/774.out
 create mode 100755 tests/xfs/776
 create mode 100644 tests/xfs/776.out


diff --git a/tests/xfs/774 b/tests/xfs/774
new file mode 100755
index 00000000..4c6bc2c9
--- /dev/null
+++ b/tests/xfs/774
@@ -0,0 +1,80 @@
+#! /bin/bash
+# SPDX-License-Identifier: GPL-2.0
+# Copyright (c) 2021 Oracle.  All Rights Reserved.
+#
+# FS QA Test 774
+#
+# Regression test for kernel commits:
+#
+# 6b69e485894b ("xfs: standardize extent size hint validation")
+# 603f000b15f2 ("xfs: validate extsz hints against rt extent size when rtinherit is set")
+#
+# Regression test for xfsprogs commit:
+#
+# 1e8afffb ("mkfs: validate rt extent size hint when rtinherit is set")
+#
+# Collectively, these patches ensure that we cannot set the extent size hint on
+# a directory when the directory is configured to propagate its realtime and
+# extent size hint to newly created files when the hint size isn't aligned to
+# the size of a realtime extent.  If the patches aren't applied, the write will
+# fail and xfs_repair will say that the fs is corrupt.
+#
+. ./common/preamble
+_begin_fstest auto quick realtime
+
+# Import common functions.
+. ./common/filter
+
+# real QA test starts here
+_supported_fs generic
+_require_realtime
+_require_scratch
+
+# Check mkfs.xfs option parsing with regards to rtinherit.  XFS doesn't require
+# the realtime volume to be present to set rtinherit, so it's safe to call the
+# mkfs binary directly, in dry run mode, with exactly the parameters we want to
+# check.
+mkfs_args=(-f -N -r extsize=7b -d extszinherit=15 $SCRATCH_DEV)
+$MKFS_XFS_PROG -d rtinherit=1 "${mkfs_args[@]}" &>> $seqres.full && \
+	echo "mkfs should not succeed with heritable rtext-unaligned extent hint"
+$MKFS_XFS_PROG -d rtinherit=0 "${mkfs_args[@]}" &>> $seqres.full || \
+	echo "mkfs should succeed with uninheritable rtext-unaligned extent hint"
+
+# Move on to checking the kernel's behavior
+_scratch_mkfs -r extsize=7b | _filter_mkfs >> $seqres.full 2> $tmp.mkfs
+cat $tmp.mkfs >> $seqres.full
+. $tmp.mkfs
+_scratch_mount
+
+test $rtextsz -ne $dbsize || _notrun "failed to set large rt extent size"
+
+# Ensure there's no extent size hint set on the directory, then set the
+# rtinherit bit on the directory to test propagation.
+$XFS_IO_PROG -c 'extsize 0' -c 'chattr +t' $SCRATCH_MNT
+
+# Now try to set an extent size hint on the directory that isn't aligned to
+# the rt extent size.
+$XFS_IO_PROG -c "extsize $((rtextsz + dbsize))" $SCRATCH_MNT 2>&1 | _filter_scratch
+$XFS_IO_PROG -c 'stat -v' $SCRATCH_MNT > $tmp.stat
+cat $tmp.stat >> $seqres.full
+grep -q 'fsxattr.xflags.*rt-inherit' $tmp.stat || \
+	echo "rtinherit didn't get set on the directory?"
+grep 'fsxattr.extsize' $tmp.stat
+
+# Propagate the hint from directory to file
+echo moo > $SCRATCH_MNT/dummy
+$XFS_IO_PROG -c 'stat -v' $SCRATCH_MNT/dummy > $tmp.stat
+cat $tmp.stat >> $seqres.full
+grep -q 'fsxattr.xflags.*realtime' $tmp.stat || \
+	echo "realtime didnt' get set on the file?"
+grep 'fsxattr.extsize' $tmp.stat
+
+# Cycle the mount to force the inode verifier to run.
+_scratch_cycle_mount
+
+# Can we still access the dummy file?
+cat $SCRATCH_MNT/dummy
+
+# success, all done
+status=0
+exit
diff --git a/tests/xfs/774.out b/tests/xfs/774.out
new file mode 100644
index 00000000..767a504e
--- /dev/null
+++ b/tests/xfs/774.out
@@ -0,0 +1,5 @@
+QA output created by 774
+xfs_io: FS_IOC_FSSETXATTR SCRATCH_MNT: Invalid argument
+fsxattr.extsize = 0
+fsxattr.extsize = 0
+moo
diff --git a/tests/xfs/776 b/tests/xfs/776
new file mode 100755
index 00000000..a62da9a5
--- /dev/null
+++ b/tests/xfs/776
@@ -0,0 +1,57 @@
+#! /bin/bash
+# SPDX-License-Identifier: GPL-2.0
+# Copyright (c) 2021, Oracle.  All Rights Reserved.
+#
+# FS QA Test No. 776
+#
+# Functional test for xfsprogs commit:
+#
+# 5f062427 ("xfs_repair: validate alignment of inherited rt extent hints")
+#
+# This xfs_repair patch detects directories that are configured to propagate
+# their realtime and extent size hints to newly created realtime files when the
+# hint size isn't aligned to the size of a realtime extent.
+#
+# Since this is a test of userspace tool functionality, we don't need kernel
+# support, which in turn means that we omit _require_realtime.  Note that XFS
+# allows users to configure realtime extent size geometry and set RTINHERIT
+# flags even if the filesystem itself does not have a realtime volume attached.
+#
+. ./common/preamble
+_begin_fstest auto repair fuzzers
+
+# Import common functions.
+. ./common/filter
+
+# real QA test starts here
+_require_scratch
+
+echo "Format and mount"
+_scratch_mkfs -r extsize=7b | _filter_mkfs > $seqres.full 2>$tmp.mkfs
+cat $tmp.mkfs >> $seqres.full
+. $tmp.mkfs
+
+test $rtextsz -ne $dbsize || _notrun "failed to set large rt extent size"
+
+_scratch_mount >> $seqres.full 2>&1
+rootino=$(stat -c '%i' $SCRATCH_MNT)
+_scratch_unmount
+
+echo "Misconfigure the root directory"
+rtextsz_blks=$((rtextsz / dbsize))
+_scratch_xfs_db -x -c "inode $rootino" \
+	-c "write -d core.extsize $((rtextsz_blks + 1))" \
+	-c 'write -d core.rtinherit 1' \
+	-c 'write -d core.extszinherit 1' \
+	-c 'print' >> $seqres.full
+
+echo "Detect misconfigured directory"
+_scratch_xfs_repair -n >> $seqres.full 2>&1 && \
+	echo "repair did not catch error?"
+
+echo "Repair misconfigured directory"
+_scratch_xfs_repair >> $seqres.full 2>&1 || \
+	echo "repair did not fix error?"
+
+status=0
+exit
diff --git a/tests/xfs/776.out b/tests/xfs/776.out
new file mode 100644
index 00000000..05ea73b2
--- /dev/null
+++ b/tests/xfs/776.out
@@ -0,0 +1,5 @@
+QA output created by 776
+Format and mount
+Misconfigure the root directory
+Detect misconfigured directory
+Repair misconfigured directory


  parent reply	other threads:[~2021-09-01  0:11 UTC|newest]

Thread overview: 12+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-09-01  0:11 [PATCHSET 0/3] fstests: regression tests for 5.13/5.14 rt fixes Darrick J. Wong
2021-09-01  0:11 ` [PATCH 1/3] xfs: test fsx with extent size hints set on a realtime file Darrick J. Wong
2021-09-05 14:28   ` Eryu Guan
2021-09-13 17:37     ` Darrick J. Wong
2021-09-01  0:11 ` Darrick J. Wong [this message]
2021-09-05 14:35   ` [PATCH 2/3] xfs: test correct propagation of rt extent size hints on rtinherit dirs Eryu Guan
2021-09-13 17:57     ` Darrick J. Wong
2021-09-01  0:11 ` [PATCH 3/3] xfs: test adding realtime sections to filesystem Darrick J. Wong
2021-09-05 14:38   ` Eryu Guan
2021-09-13 18:02     ` Darrick J. Wong
  -- strict thread matches above, loose matches on Subject: below --
2021-09-15 23:42 [PATCHSET v2 0/3] fstests: regression tests for 5.13/5.14 rt fixes Darrick J. Wong
2021-09-15 23:42 ` [PATCH 2/3] xfs: test correct propagation of rt extent size hints on rtinherit dirs Darrick J. Wong
2021-07-28  0:10 [PATCHSET 0/3] fstests: regression tests for 5.13/5.14 rt fixes Darrick J. Wong
2021-07-28  0:10 ` [PATCH 2/3] xfs: test correct propagation of rt extent size hints on rtinherit dirs Darrick J. Wong

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=163045509592.769915.9044627867698975012.stgit@magnolia \
    --to=djwong@kernel.org \
    --cc=fstests@vger.kernel.org \
    --cc=guan@eryu.me \
    --cc=guaneryu@gmail.com \
    --cc=linux-xfs@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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.