Skip to contentSkip to conversation

f2fs/018: check data eof after partial truncation on compressed cluster

PATCH Superseded Chao Yu 2 participants 10 messages Aug 11, 2025 – Aug 14, 2025

v1 of 3 Superseded Latest is v3 →

lore
Share & tools
lore

Chao Yu

author
#0 lore raw

PATCH 1/3 f2fs/018: check data eof after partial truncation on compressed cluster

This is a regression testcase, it is added to check below case and its variants:
- write 16k data into compressed file (data will be compressed)
- truncate file to 12k (truncate partial data in compressed cluster)
- truncate file to 20k
- verify data in range of [12k, 16k] to see whether data is all zero or not

Cc Jaegeuk Kim <[email protected]>
---
 tests/f2fs/018     | 58 ++++++++++++++++++++++++++++++++++++++++++++++
 tests/f2fs/018.out |  2 ++
 2 files changed, 60 insertions(+)
 create mode 100755 tests/f2fs/018
 create mode 100644 tests/f2fs/018.out
tests/f2fs/018 new file +58 -0
new file mode 100755
index 00000000..ca727d24
@@ -0,0 +1,58 @@
+#! /bin/bash
+# SPDX-License-Identifier: GPL-2.0
+# Copyright (c) 2025 Chao Yu.  All Rights Reserved.
+#
+# FS QA Test No. f2fs/018
+#
+# This is a regression test to check whether page eof will be
+# zero or not after we truncate partial data in compressed
+# cluster.
+#
+. ./common/preamble
+_begin_fstest auto quick rw compress
+
+_fixed_by_kernel_commit ba8dac350faf \
+	"f2fs: fix to zero post-eof page"
+_fixed_by_kernel_commit xxxxxxxxxxxx \
+	"f2fs: fix to zero data after EOF for compressed file correctly"
+
+_require_scratch
+_require_fio
+
+testfile=$SCRATCH_MNT/testfile
+
+check_data_eof()
+{
+	local eof_start=$1
+	local eof_size=$2
+	local filesize=$3
+	local offset1=$4
+	local offset2=$5
+	local offset3=$6
+
+	_scratch_mkfs "-O extra_attr,compression" >> $seqres.full || _fail "mkfs failed"
+	_scratch_mount "-o compress_extension=*" >> $seqres.full
+
+	xfs_io -f -c "pwrite 0 $filesize" -c "fsync" $testfile >> $seqres.full
+	xfs_io -c "truncate $offset1" $testfile
+	xfs_io -c "truncate $offset2" $testfile
+
+	if [ "$offset3" ]; then
+		xfs_io -c "truncate $offset3" $testfile
+	fi
+
+	$FIO_PROG --name=verify-data --filename=$testfile --rw=read --verify=pattern \
+		--verify_pattern=0x00 --do_verify=1 --verify_only --offset=$eof_start \
+		--size=$eof_size >> $seqres.full 2>&1 || _fail "eof data is not zero"
+	_scratch_unmount
+}
+
+check_data_eof 12k 4k 16k 12k 20k
+check_data_eof 10k 6k 16k 10k 20k
+check_data_eof 12k 4k 16k 8k 12k 20k
+check_data_eof 10k 6k 16k 8k 10k 20k
+
+echo "Silence is golden"
+
+status=0
+exit
tests/f2fs/018.out new file +2 -0
new file mode 100644
index 00000000..8849e303
@@ -0,0 +1,2 @@
+QA output created by 018
+Silence is golden
Signature · 1 line
2.49.0
author
#1 lore raw

PATCH 2/3 f2fs/019: do sanity check on mapping table

This is a regression test:
1. create a file
2. write file to create a direct node at special offset
3. use inject.f2fs to inject nid of direct node w/ ino of the inode
4. check whether f2fs kernel module will detect and report such corruption in the fil

---
 tests/f2fs/019     | 45 +++++++++++++++++++++++++++++++++++++++++++++
 tests/f2fs/019.out |  2 ++
 2 files changed, 47 insertions(+)
 create mode 100755 tests/f2fs/019
 create mode 100644 tests/f2fs/019.out
tests/f2fs/019 new file +45 -0
new file mode 100755
index 00000000..0a02eb2f
@@ -0,0 +1,45 @@
+#! /bin/bash
+# SPDX-License-Identifier: GPL-2.0
+# Copyright (c) 2025 Chao Yu.  All Rights Reserved.
+#
+# FS QA Test No. f2fs/019
+#
+# This is a regression test:
+# 1. create a file
+# 2. write file to create a direct node at special offset
+# 3. use inject.f2fs to inject nid of direct node w/ ino of the inode
+# 4. check whether f2fs kernel module will detect and report such
+#    corruption in the file
+#
+. ./common/preamble
+_begin_fstest auto quick rw compress
+
+_cleanup()
+{
+	_scratch_mkfs >> $seqres.full
+}
+
+_fixed_by_kernel_commit 77de19b6867f \
+	"f2fs: fix to avoid out-of-boundary access in dnode page"
+
+export MKFS_OPTIONS=""
+_require_scratch
+_require_command "$F2FS_INJECT_PROG" inject.f2fs
+
+testfile=$SCRATCH_MNT/testfile
+
+_scratch_mkfs >> $seqres.full
+_scratch_mount
+
+xfs_io -f -c "pwrite 3738M 1M" -c "fsync" $testfile >> $seqres.full
+
+_scratch_unmount
+
+$F2FS_INJECT_PROG --node --mb addr --nid 5 --idx 937 --val 4 $SCRATCH_DEV >> $seqres.full
+
+_scratch_mount
+xfs_io -c "pread 3700M 40M" $testfile
+_scratch_unmount
+
+status=0
+exit
tests/f2fs/019.out new file +2 -0
new file mode 100644
index 00000000..2f7469e2
@@ -0,0 +1,2 @@
+QA output created by 019
+pread: Structure needs cleaning
Signature · 1 line
2.49.0
#2 lore raw

Re: [PATCH 2/3] f2fs/019: do sanity check on mapping table

Quoted 30 lines · Chao Yu

This is a regression test:
1. create a file
2. write file to create a direct node at special offset
3. use inject.f2fs to inject nid of direct node w/ ino of the inode
4. check whether f2fs kernel module will detect and report such corruption in the fil

---
 tests/f2fs/019     | 45 +++++++++++++++++++++++++++++++++++++++++++++
 tests/f2fs/019.out |  2 ++
 2 files changed, 47 insertions(+)
 create mode 100755 tests/f2fs/019
 create mode 100644 tests/f2fs/019.out
tests/f2fs/019 new file +9 -0
new file mode 100755
index 00000000..0a02eb2f
@@ -0,0 +1,45 @@
+#! /bin/bash
+# SPDX-License-Identifier: GPL-2.0
+# Copyright (c) 2025 Chao Yu.  All Rights Reserved.
+#
+# FS QA Test No. f2fs/019
+#
+# This is a regression test:
+# 1. create a file
+# 2. write file to create a direct node at special offset

tests/f2fs/019 line 19

Quoted 10 lines
tests/f2fs/019 new file +10 -0
+# 3. use inject.f2fs to inject nid of direct node w/ ino of the inode
+# 4. check whether f2fs kernel module will detect and report such
+#    corruption in the file
+#
+. ./common/preamble
+_begin_fstest auto quick rw compress
+
+_cleanup()
+{
+	_scratch_mkfs >> $seqres.full

Why mkfs SCRATCH_DEV in _cleanup ?

tests/f2fs/019 line 25

Quoted 6 lines
tests/f2fs/019 new file +6 -0
+}
+
+_fixed_by_kernel_commit 77de19b6867f \
+	"f2fs: fix to avoid out-of-boundary access in dnode page"
+
+export MKFS_OPTIONS=""

Can you give it a comment about why we need to clean mkfs option at here?

tests/f2fs/019 line 26

Quoted 1 line
tests/f2fs/019 new file +1 -0
+_require_scratch

Oh, if this case tends to have a corrupted SCRATCH_DEV, you can use _require_scratch_nocheck at here, don't need to mkfs in _cleanup.

tests/f2fs/019 line 34

Quoted 8 lines
tests/f2fs/019 new file +8 -0
+_require_command "$F2FS_INJECT_PROG" inject.f2fs
+
+testfile=$SCRATCH_MNT/testfile
+
+_scratch_mkfs >> $seqres.full
+_scratch_mount
+
+xfs_io -f -c "pwrite 3738M 1M" -c "fsync" $testfile >> $seqres.full

^^^^^^
$XFS_IO_PROG

tests/f2fs/019 line 41

Quoted 7 lines
tests/f2fs/019 new file +7 -0
+
+_scratch_unmount
+
+$F2FS_INJECT_PROG --node --mb addr --nid 5 --idx 937 --val 4 $SCRATCH_DEV >> $seqres.full
+
+_scratch_mount
+xfs_io -c "pread 3700M 40M" $testfile

^^^^^^
$XFS_IO_PROG

Quoted 15 lines
tests/f2fs/019 new file +4 -0
+_scratch_unmount
+
+status=0
+exit
tests/f2fs/019.out new file +2 -0
new file mode 100644
index 00000000..2f7469e2
@@ -0,0 +1,2 @@
+QA output created by 019
+pread: Structure needs cleaning
Signature · 1 line
2.49.0
author
#3 lore raw

Re: [PATCH 2/3] f2fs/019: do sanity check on mapping table

Quoted 35 lines · Zorro Lang
Quoted 34 lines · Chao Yu

This is a regression test:
1. create a file
2. write file to create a direct node at special offset
3. use inject.f2fs to inject nid of direct node w/ ino of the inode
4. check whether f2fs kernel module will detect and report such corruption in the fil

---
 tests/f2fs/019     | 45 +++++++++++++++++++++++++++++++++++++++++++++
 tests/f2fs/019.out |  2 ++
 2 files changed, 47 insertions(+)
 create mode 100755 tests/f2fs/019
 create mode 100644 tests/f2fs/019.out
tests/f2fs/019 new file +13 -0
new file mode 100755
index 00000000..0a02eb2f
@@ -0,0 +1,45 @@
+#! /bin/bash
+# SPDX-License-Identifier: GPL-2.0
+# Copyright (c) 2025 Chao Yu.  All Rights Reserved.
+#
+# FS QA Test No. f2fs/019
+#
+# This is a regression test:
+# 1. create a file
+# 2. write file to create a direct node at special offset
+# 3. use inject.f2fs to inject nid of direct node w/ ino of the inode
+# 4. check whether f2fs kernel module will detect and report such
+#    corruption in the file
+#
Quoted 8 lines

tests/f2fs/019 line 19

Quoted 6 lines
tests/f2fs/019 new file +6 -0
+. ./common/preamble
+_begin_fstest auto quick rw compress
+
+_cleanup()
+{
+	_scratch_mkfs >> $seqres.full

Why mkfs SCRATCH_DEV in _cleanup ?

Because SCRATCH_DEV will be corrupted after injection.

Quoted 9 lines

tests/f2fs/019 line 25

Quoted 7 lines
tests/f2fs/019 new file +6 -0
+}
+
+_fixed_by_kernel_commit 77de19b6867f \
+	"f2fs: fix to avoid out-of-boundary access in dnode page"
+
+export MKFS_OPTIONS=""

Can you give it a comment about why we need to clean mkfs option at here?

Sure.

Quoted 5 lines

tests/f2fs/019 line 26

Quoted 2 lines
tests/f2fs/019 new file +1 -0
+_require_scratch

Oh, if this case tends to have a corrupted SCRATCH_DEV, you can use _require_scratch_nocheck at here, don't need to mkfs in _cleanup.

Oh, my bad, you've mentioned the function on f2fs/012 once, but I forget to use it in this testcase.

Quoted 12 lines

tests/f2fs/019 line 34

Quoted 9 lines
tests/f2fs/019 new file +8 -0
+_require_command "$F2FS_INJECT_PROG" inject.f2fs
+
+testfile=$SCRATCH_MNT/testfile
+
+_scratch_mkfs >> $seqres.full
+_scratch_mount
+
+xfs_io -f -c "pwrite 3738M 1M" -c "fsync" $testfile >> $seqres.full

^^^^^^
$XFS_IO_PROG

Quoted 9 lines

tests/f2fs/019 line 41

Quoted 7 lines
tests/f2fs/019 new file +7 -0
+
+_scratch_unmount
+
+$F2FS_INJECT_PROG --node --mb addr --nid 5 --idx 937 --val 4 $SCRATCH_DEV >> $seqres.full
+
+_scratch_mount
+xfs_io -c "pread 3700M 40M" $testfile

^^^^^^
$XFS_IO_PROG

Will update them in v2.

Thanks,

Quoted 17 lines
Quoted 16 lines
tests/f2fs/019 new file +4 -0
+_scratch_unmount
+
+status=0
+exit
tests/f2fs/019.out new file +2 -0
new file mode 100644
index 00000000..2f7469e2
@@ -0,0 +1,2 @@
+QA output created by 019
+pread: Structure needs cleaning
Signature · 1 line
2.49.0
author
#4 lore raw

PATCH 3/3 f2fs/020: do sanity check on i_xattr_nid

In reply to Chao Yu

This is a regression test:
1. create directory
2. add a new xattr entry to create xattr node
3. use inject.f2fs to inject nid of xattr node w/ ino in a file
4. check whether f2fs kernel module will detect and report such corruption in the file

---
 tests/f2fs/020     | 50 ++++++++++++++++++++++++++++++++++++++++++++++
 tests/f2fs/020.out |  2 ++
 2 files changed, 52 insertions(+)
 create mode 100755 tests/f2fs/020
 create mode 100644 tests/f2fs/020.out
tests/f2fs/020 new file +50 -0
new file mode 100755
index 00000000..a6a08c8f
@@ -0,0 +1,50 @@
+#! /bin/bash
+# SPDX-License-Identifier: GPL-2.0
+# Copyright (c) 2025 Chao Yu.  All Rights Reserved.
+#
+# FS QA Test No. f2fs/020
+#
+# This is a regression test:
+# 1. create directory
+# 2. add a new xattr entry to create xattr node
+# 3. use inject.f2fs to inject nid of xattr node w/ ino in a file
+# 4. check whether f2fs kernel module will detect and report such
+#    corruption in the file
+#
+. ./common/preamble
+_begin_fstest auto quick rw
+
+_cleanup()
+{
+	_scratch_mkfs >> $seqres.full
+}
+
+_fixed_by_kernel_commit 061cf3a84bde \
+	"f2fs: fix to do sanity check on ino and xnid"
+
+export MKFS_OPTIONS=""
+_require_scratch
+_require_command "$F2FS_INJECT_PROG" inject.f2fs
+
+testdir=$SCRATCH_MNT/testdir
+
+_scratch_mkfs >> $seqres.full
+_scratch_mount "-o user_xattr,noinline_xattr"
+
+mkdir $testdir
+# add a new xattr entry to create xattr node
+setfattr -n user.abc -v 123 $testdir
+
+_scratch_unmount
+
+# inject i_xattr_nid w/ nid of inode node
+$F2FS_INJECT_PROG --node --mb i_xattr_nid --nid 4 --val 4 $SCRATCH_DEV >> $seqres.full
+
+_scratch_mount
+mkdir $testdir/dir >> $seqres.full 2>&1
+_scratch_unmount
+
+echo "Silence is golden"
+
+status=0
+exit
tests/f2fs/020.out new file +2 -0
new file mode 100644
index 00000000..20d7944e
@@ -0,0 +1,2 @@
+QA output created by 020
+Silence is golden
Signature · 1 line
2.49.0
#5 lore raw

Re: [PATCH 3/3] f2fs/020: do sanity check on i_xattr_nid

Quoted 37 lines · Chao Yu

This is a regression test:
1. create directory
2. add a new xattr entry to create xattr node
3. use inject.f2fs to inject nid of xattr node w/ ino in a file
4. check whether f2fs kernel module will detect and report such corruption in the file

---
 tests/f2fs/020     | 50 ++++++++++++++++++++++++++++++++++++++++++++++
 tests/f2fs/020.out |  2 ++
 2 files changed, 52 insertions(+)
 create mode 100755 tests/f2fs/020
 create mode 100644 tests/f2fs/020.out
tests/f2fs/020 new file +16 -0
new file mode 100755
index 00000000..a6a08c8f
@@ -0,0 +1,50 @@
+#! /bin/bash
+# SPDX-License-Identifier: GPL-2.0
+# Copyright (c) 2025 Chao Yu.  All Rights Reserved.
+#
+# FS QA Test No. f2fs/020
+#
+# This is a regression test:
+# 1. create directory
+# 2. add a new xattr entry to create xattr node
+# 3. use inject.f2fs to inject nid of xattr node w/ ino in a file
+# 4. check whether f2fs kernel module will detect and report such
+#    corruption in the file
+#
+. ./common/preamble
+_begin_fstest auto quick rw
+

tests/f2fs/020 line 26

Quoted 10 lines
tests/f2fs/020 new file +10 -0
+_cleanup()
+{
+	_scratch_mkfs >> $seqres.full
+}
+
+_fixed_by_kernel_commit 061cf3a84bde \
+	"f2fs: fix to do sanity check on ino and xnid"
+
+export MKFS_OPTIONS=""
+_require_scratch

Same review points with patch 2/3,

tests/f2fs/020 line 36

Quoted 10 lines
tests/f2fs/020 new file +10 -0
+_require_command "$F2FS_INJECT_PROG" inject.f2fs
+
+testdir=$SCRATCH_MNT/testdir
+
+_scratch_mkfs >> $seqres.full
+_scratch_mount "-o user_xattr,noinline_xattr"
+
+mkdir $testdir
+# add a new xattr entry to create xattr node
+setfattr -n user.abc -v 123 $testdir

You can call _require_attrs user at first, then use "$SETFATTR_PROG"

tests/f2fs/020 line 44

Quoted 8 lines
tests/f2fs/020 new file +8 -0
+
+_scratch_unmount
+
+# inject i_xattr_nid w/ nid of inode node
+$F2FS_INJECT_PROG --node --mb i_xattr_nid --nid 4 --val 4 $SCRATCH_DEV >> $seqres.full
+
+_scratch_mount
+mkdir $testdir/dir >> $seqres.full 2>&1

May I ask what kind of failure will this case hit, if there's the bug?

Thanks,
Zorro

Quoted 17 lines
tests/f2fs/020 new file +6 -0
+_scratch_unmount
+
+echo "Silence is golden"
+
+status=0
+exit
tests/f2fs/020.out new file +2 -0
new file mode 100644
index 00000000..20d7944e
@@ -0,0 +1,2 @@
+QA output created by 020
+Silence is golden
Signature · 1 line
2.49.0
author
#6 lore raw

Re: [PATCH 3/3] f2fs/020: do sanity check on i_xattr_nid

Quoted 51 lines · Zorro Lang
Quoted 37 lines · Chao Yu

This is a regression test:
1. create directory
2. add a new xattr entry to create xattr node
3. use inject.f2fs to inject nid of xattr node w/ ino in a file
4. check whether f2fs kernel module will detect and report such corruption in the file

---
 tests/f2fs/020     | 50 ++++++++++++++++++++++++++++++++++++++++++++++
 tests/f2fs/020.out |  2 ++
 2 files changed, 52 insertions(+)
 create mode 100755 tests/f2fs/020
 create mode 100644 tests/f2fs/020.out
tests/f2fs/020 new file +16 -0
new file mode 100755
index 00000000..a6a08c8f
@@ -0,0 +1,50 @@
+#! /bin/bash
+# SPDX-License-Identifier: GPL-2.0
+# Copyright (c) 2025 Chao Yu.  All Rights Reserved.
+#
+# FS QA Test No. f2fs/020
+#
+# This is a regression test:
+# 1. create directory
+# 2. add a new xattr entry to create xattr node
+# 3. use inject.f2fs to inject nid of xattr node w/ ino in a file
+# 4. check whether f2fs kernel module will detect and report such
+#    corruption in the file
+#
+. ./common/preamble
+_begin_fstest auto quick rw
+

tests/f2fs/020 line 26

Quoted 10 lines
tests/f2fs/020 new file +10 -0
+_cleanup()
+{
+	_scratch_mkfs >> $seqres.full
+}
+
+_fixed_by_kernel_commit 061cf3a84bde \
+	"f2fs: fix to do sanity check on ino and xnid"
+
+export MKFS_OPTIONS=""
+_require_scratch

Same review points with patch 2/3,

Quoted 12 lines

tests/f2fs/020 line 36

Quoted 10 lines
tests/f2fs/020 new file +10 -0
+_require_command "$F2FS_INJECT_PROG" inject.f2fs
+
+testdir=$SCRATCH_MNT/testdir
+
+_scratch_mkfs >> $seqres.full
+_scratch_mount "-o user_xattr,noinline_xattr"
+
+mkdir $testdir
+# add a new xattr entry to create xattr node
+setfattr -n user.abc -v 123 $testdir

You can call _require_attrs user at first, then use "$SETFATTR_PROG"

Okay, will update them in v2.

Quoted 11 lines

tests/f2fs/020 line 44

Quoted 9 lines
tests/f2fs/020 new file +8 -0
+
+_scratch_unmount
+
+# inject i_xattr_nid w/ nid of inode node
+$F2FS_INJECT_PROG --node --mb i_xattr_nid --nid 4 --val 4 $SCRATCH_DEV >> $seqres.full
+
+_scratch_mount
+mkdir $testdir/dir >> $seqres.full 2>&1

May I ask what kind of failure will this case hit, if there's the bug?

Bug like syzbot reported, you can check the details in below commit:

https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=061cf3a84bde

Thanks,

Quoted 22 lines

Thanks,
Zorro

Quoted 17 lines
tests/f2fs/020 new file +6 -0
+_scratch_unmount
+
+echo "Silence is golden"
+
+status=0
+exit
tests/f2fs/020.out new file +2 -0
new file mode 100644
index 00000000..20d7944e
@@ -0,0 +1,2 @@
+QA output created by 020
+Silence is golden
Signature · 1 line
2.49.0
#7 lore raw

Re: [PATCH 3/3] f2fs/020: do sanity check on i_xattr_nid

Quoted 79 lines · Chao Yu
Quoted 51 lines · Zorro Lang
Quoted 37 lines · Chao Yu

This is a regression test:
1. create directory
2. add a new xattr entry to create xattr node
3. use inject.f2fs to inject nid of xattr node w/ ino in a file
4. check whether f2fs kernel module will detect and report such corruption in the file

---
 tests/f2fs/020     | 50 ++++++++++++++++++++++++++++++++++++++++++++++
 tests/f2fs/020.out |  2 ++
 2 files changed, 52 insertions(+)
 create mode 100755 tests/f2fs/020
 create mode 100644 tests/f2fs/020.out
tests/f2fs/020 new file +16 -0
new file mode 100755
index 00000000..a6a08c8f
@@ -0,0 +1,50 @@
+#! /bin/bash
+# SPDX-License-Identifier: GPL-2.0
+# Copyright (c) 2025 Chao Yu.  All Rights Reserved.
+#
+# FS QA Test No. f2fs/020
+#
+# This is a regression test:
+# 1. create directory
+# 2. add a new xattr entry to create xattr node
+# 3. use inject.f2fs to inject nid of xattr node w/ ino in a file
+# 4. check whether f2fs kernel module will detect and report such
+#    corruption in the file
+#
+. ./common/preamble
+_begin_fstest auto quick rw
+

tests/f2fs/020 line 26

Quoted 10 lines
tests/f2fs/020 new file +10 -0
+_cleanup()
+{
+	_scratch_mkfs >> $seqres.full
+}
+
+_fixed_by_kernel_commit 061cf3a84bde \
+	"f2fs: fix to do sanity check on ino and xnid"
+
+export MKFS_OPTIONS=""
+_require_scratch

Same review points with patch 2/3,

Quoted 12 lines

tests/f2fs/020 line 36

Quoted 10 lines
tests/f2fs/020 new file +10 -0
+_require_command "$F2FS_INJECT_PROG" inject.f2fs
+
+testdir=$SCRATCH_MNT/testdir
+
+_scratch_mkfs >> $seqres.full
+_scratch_mount "-o user_xattr,noinline_xattr"
+
+mkdir $testdir
+# add a new xattr entry to create xattr node
+setfattr -n user.abc -v 123 $testdir

You can call _require_attrs user at first, then use "$SETFATTR_PROG"

Okay, will update them in v2.

Quoted 11 lines

tests/f2fs/020 line 44

Quoted 9 lines
tests/f2fs/020 new file +8 -0
+
+_scratch_unmount
+
+# inject i_xattr_nid w/ nid of inode node
+$F2FS_INJECT_PROG --node --mb i_xattr_nid --nid 4 --val 4 $SCRATCH_DEV >> $seqres.full
+
+_scratch_mount
+mkdir $testdir/dir >> $seqres.full 2>&1

May I ask what kind of failure will this case hit, if there's the bug?

Quoted 3 lines

Bug like syzbot reported, you can check the details in below commit:

https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=061cf3a84bde

So you hope to trigger a test failure by something likes:

"generic/361 _check_dmesg: something found in dmesg"

?

Did you try that, if it can fail this test on general kernel?

Thanks,
Zorro

Quoted 26 lines

Thanks,

Quoted 22 lines

Thanks,
Zorro

Quoted 17 lines
tests/f2fs/020 new file +6 -0
+_scratch_unmount
+
+echo "Silence is golden"
+
+status=0
+exit
tests/f2fs/020.out new file +2 -0
new file mode 100644
index 00000000..20d7944e
@@ -0,0 +1,2 @@
+QA output created by 020
+Silence is golden
Signature · 1 line
2.49.0
#8 lore raw
In reply to Chao Yu
Quoted 51 lines · Chao Yu

This is a regression testcase, it is added to check below case and its variants:
- write 16k data into compressed file (data will be compressed)
- truncate file to 12k (truncate partial data in compressed cluster)
- truncate file to 20k
- verify data in range of [12k, 16k] to see whether data is all zero or not

Cc Jaegeuk Kim <[email protected]>
---
 tests/f2fs/018     | 58 ++++++++++++++++++++++++++++++++++++++++++++++
 tests/f2fs/018.out |  2 ++
 2 files changed, 60 insertions(+)
 create mode 100755 tests/f2fs/018
 create mode 100644 tests/f2fs/018.out
tests/f2fs/018 new file +28 -0
new file mode 100755
index 00000000..ca727d24
@@ -0,0 +1,58 @@
+#! /bin/bash
+# SPDX-License-Identifier: GPL-2.0
+# Copyright (c) 2025 Chao Yu.  All Rights Reserved.
+#
+# FS QA Test No. f2fs/018
+#
+# This is a regression test to check whether page eof will be
+# zero or not after we truncate partial data in compressed
+# cluster.
+#
+. ./common/preamble
+_begin_fstest auto quick rw compress
+
+_fixed_by_kernel_commit ba8dac350faf \
+	"f2fs: fix to zero post-eof page"
+_fixed_by_kernel_commit xxxxxxxxxxxx \
+	"f2fs: fix to zero data after EOF for compressed file correctly"
+
+_require_scratch
+_require_fio
+
+testfile=$SCRATCH_MNT/testfile
+
+check_data_eof()
+{
+	local eof_start=$1
+	local eof_size=$2
+	local filesize=$3

tests/f2fs/018 line 38

Quoted 10 lines
tests/f2fs/018 new file +10 -0
+	local offset1=$4
+	local offset2=$5
+	local offset3=$6
+
+	_scratch_mkfs "-O extra_attr,compression" >> $seqres.full || _fail "mkfs failed"
+	_scratch_mount "-o compress_extension=*" >> $seqres.full
+
+	xfs_io -f -c "pwrite 0 $filesize" -c "fsync" $testfile >> $seqres.full
+	xfs_io -c "truncate $offset1" $testfile
+	xfs_io -c "truncate $offset2" $testfile

replace "xfs_io" with "$XFS_IO_PROG"

tests/f2fs/018 line 41

Quoted 3 lines
tests/f2fs/018 new file +3 -0
+
+	if [ "$offset3" ]; then
+		xfs_io -c "truncate $offset3" $testfile

Same as above.

And generally we need: _require_xfs_io_command "truncate"

tests/f2fs/018 line 46

Quoted 5 lines
tests/f2fs/018 new file +5 -0
+	fi
+
+	$FIO_PROG --name=verify-data --filename=$testfile --rw=read --verify=pattern \
+		--verify_pattern=0x00 --do_verify=1 --verify_only --offset=$eof_start \
+		--size=$eof_size >> $seqres.full 2>&1 || _fail "eof data is not zero"

Generally we do _require_fio $fio_config to make sure all arguments are supported by current fio version and FSTYP, refer to generic/366. But as this is a f2fs only test, and if you make sure you don't use any new/special fio arguments, I can accept this. Or you can refer to generic/366.

tests/f2fs/018 line 47

Quoted 1 line
tests/f2fs/018 new file +1 -0
+	_scratch_unmount

I'm not familar with f2fs much, may I ask why we need to mkfs&mount ... umount in each test loop?

Thanks,
Zorro

Quoted 22 lines
tests/f2fs/018 new file +11 -0
+}
+
+check_data_eof 12k 4k 16k 12k 20k
+check_data_eof 10k 6k 16k 10k 20k
+check_data_eof 12k 4k 16k 8k 12k 20k
+check_data_eof 10k 6k 16k 8k 10k 20k
+
+echo "Silence is golden"
+
+status=0
+exit
tests/f2fs/018.out new file +2 -0
new file mode 100644
index 00000000..8849e303
@@ -0,0 +1,2 @@
+QA output created by 018
+Silence is golden
Signature · 1 line
2.49.0
author
#9 lore raw
Quoted 56 lines · Zorro Lang
Quoted 55 lines · Chao Yu

This is a regression testcase, it is added to check below case and its variants:
- write 16k data into compressed file (data will be compressed)
- truncate file to 12k (truncate partial data in compressed cluster)
- truncate file to 20k
- verify data in range of [12k, 16k] to see whether data is all zero or not

Cc Jaegeuk Kim <[email protected]>
---
 tests/f2fs/018     | 58 ++++++++++++++++++++++++++++++++++++++++++++++
 tests/f2fs/018.out |  2 ++
 2 files changed, 60 insertions(+)
 create mode 100755 tests/f2fs/018
 create mode 100644 tests/f2fs/018.out
tests/f2fs/018 new file +32 -0
new file mode 100755
index 00000000..ca727d24
@@ -0,0 +1,58 @@
+#! /bin/bash
+# SPDX-License-Identifier: GPL-2.0
+# Copyright (c) 2025 Chao Yu.  All Rights Reserved.
+#
+# FS QA Test No. f2fs/018
+#
+# This is a regression test to check whether page eof will be
+# zero or not after we truncate partial data in compressed
+# cluster.
+#
+. ./common/preamble
+_begin_fstest auto quick rw compress
+
+_fixed_by_kernel_commit ba8dac350faf \
+	"f2fs: fix to zero post-eof page"
+_fixed_by_kernel_commit xxxxxxxxxxxx \
+	"f2fs: fix to zero data after EOF for compressed file correctly"
+
+_require_scratch
+_require_fio
+
+testfile=$SCRATCH_MNT/testfile
+
+check_data_eof()
+{
+	local eof_start=$1
+	local eof_size=$2
+	local filesize=$3
+	local offset1=$4
+	local offset2=$5
+	local offset3=$6
+
Quoted 8 lines

tests/f2fs/018 line 38

Quoted 6 lines
tests/f2fs/018 new file +6 -0
+	_scratch_mkfs "-O extra_attr,compression" >> $seqres.full || _fail "mkfs failed"
+	_scratch_mount "-o compress_extension=*" >> $seqres.full
+
+	xfs_io -f -c "pwrite 0 $filesize" -c "fsync" $testfile >> $seqres.full
+	xfs_io -c "truncate $offset1" $testfile
+	xfs_io -c "truncate $offset2" $testfile

replace "xfs_io" with "$XFS_IO_PROG"

Oops, yes, will update.

Quoted 8 lines

tests/f2fs/018 line 41

Quoted 4 lines
tests/f2fs/018 new file +3 -0
+
+	if [ "$offset3" ]; then
+		xfs_io -c "truncate $offset3" $testfile

Same as above.

And generally we need: _require_xfs_io_command "truncate"

Ditto,

Quoted 11 lines

tests/f2fs/018 line 46

Quoted 6 lines
tests/f2fs/018 new file +5 -0
+	fi
+
+	$FIO_PROG --name=verify-data --filename=$testfile --rw=read --verify=pattern \
+		--verify_pattern=0x00 --do_verify=1 --verify_only --offset=$eof_start \
+		--size=$eof_size >> $seqres.full 2>&1 || _fail "eof data is not zero"

Generally we do _require_fio $fio_config to make sure all arguments are supported by current fio version and FSTYP, refer to generic/366. But as this is a f2fs only test, and if you make sure you don't use any new/special fio arguments, I can accept this. Or you can refer to generic/366.

Okay, let me take a look.

Quoted 5 lines

tests/f2fs/018 line 47

Quoted 2 lines
tests/f2fs/018 new file +1 -0
+	_scratch_unmount

I'm not familar with f2fs much, may I ask why we need to mkfs&mount ... umount in each test loop?

Oh, seems it doesn't need to, let me check and clean up it.

Thanks,

Quoted 27 lines

Thanks,
Zorro

Quoted 22 lines
tests/f2fs/018 new file +11 -0
+}
+
+check_data_eof 12k 4k 16k 12k 20k
+check_data_eof 10k 6k 16k 10k 20k
+check_data_eof 12k 4k 16k 8k 12k 20k
+check_data_eof 10k 6k 16k 8k 10k 20k
+
+echo "Silence is golden"
+
+status=0
+exit
tests/f2fs/018.out new file +2 -0
new file mode 100644
index 00000000..8849e303
@@ -0,0 +1,2 @@
+QA output created by 018
+Silence is golden
Signature · 1 line
2.49.0