As detailed in the patch [1], kernel currently does not support resizes
with sparse_super2 enabled. Before the above patch, if we used the
EXT4_IOC_RESIZE_FS ioctl directly, wiht sparse_super2 enabled, the
kernel used to still try the resize and ultimatley leave the fs in an
inconsistent state. This also led to corruption and kernel BUGs.
This patchset adds a test for ext4 to ensure that the kernel handles
resizes with sparse_super2 correctly, and returns -EOPNOTSUPP.
Summary:
Patch 1: Fix the src/ext4_resize.c script to return accurate error
codes.
Patch 2: Add the ext4 test for checking resize functionality
Changes since v1 [2]
* In patch 2, don't iterate if the expected errno is returned
* Code cleanup and additional comments for clarity (also see extra
note in patch 2)
* No changes in patch 1
[1] https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=b1489186cc8391e0c1e342f9fbc3eedf6b944c61
[2] https://lore.kernel.org/fstests/[email protected]/
Ojaswin Mujoo (2):
src/ext4_resize.c: Refactor code and ensure accurate errno is returned
ext4: Test to ensure resize with sparse_super2 is handled correctly
src/ext4_resize.c | 46 +++++++++++++------
tests/ext4/056 | 108 +++++++++++++++++++++++++++++++++++++++++++++
tests/ext4/056.out | 2 +
3 files changed, 142 insertions(+), 14 deletions(-)
create mode 100755 tests/ext4/056
create mode 100644 tests/ext4/056.out
--
2.27.0
The current implementation of ext4_resize returned 1 whenever
there was an error. Modify this to return the correct error code.
This is important for tests that rely on correct error reporting, by
the kernel, for ext4 resize functionality.
Additionaly, perform some code cleanup.
Signed-off-by: Ojaswin Mujoo <[email protected]>
Reviewed-by: Ritesh Harjani <[email protected]>
---
src/ext4_resize.c | 46 ++++++++++++++++++++++++++++++++--------------
1 file changed, 32 insertions(+), 14 deletions(-)
diff --git a/src/ext4_resize.c b/src/ext4_resize.c
index 39e16529..78b66432 100644
--- a/src/ext4_resize.c
+++ b/src/ext4_resize.c
@@ -10,6 +10,7 @@
#include <unistd.h>
#include <stdint.h>
#include <stdlib.h>
+#include <string.h>
#include <sys/ioctl.h>
#include <sys/mount.h>
@@ -19,33 +20,50 @@ typedef unsigned long long __u64;
#define EXT4_IOC_RESIZE_FS _IOW('f', 16, __u64)
#endif
+#define pr_error(fmt, ...) do { \
+ fprintf (stderr, "ext4_resize.c: " fmt, ##__VA_ARGS__); \
+} while (0)
+
+static void usage(void)
+{
+ fprintf(stdout, "\nUsage: ext4_resize [mnt_path] [new_size(blocks)]\n");
+}
+
int main(int argc, char **argv)
{
__u64 new_size;
int error, fd;
- char *tmp = NULL;
+ char *mnt_dir = NULL, *tmp = NULL;
if (argc != 3) {
- fputs("insufficient arguments\n", stderr);
- return 1;
- }
- fd = open(argv[1], O_RDONLY);
- if (!fd) {
- perror(argv[1]);
- return 1;
+ error = EINVAL;
+ pr_error("insufficient arguments\n");
+ usage();
+ return error;
}
+ mnt_dir = argv[1];
+
errno = 0;
new_size = strtoull(argv[2], &tmp, 10);
if ((errno) || (*tmp != '\0')) {
- fprintf(stderr, "%s: invalid new size\n", argv[0]);
- return 1;
+ error = errno;
+ pr_error("invalid new size\n");
+ return error;
+ }
+
+ fd = open(mnt_dir, O_RDONLY);
+ if (fd < 0) {
+ error = errno;
+ pr_error("open() failed with error: %s\n", strerror(error));
+ return error;
}
- error = ioctl(fd, EXT4_IOC_RESIZE_FS, &new_size);
- if (error < 0) {
- perror("EXT4_IOC_RESIZE_FS");
- return 1;
+ if(ioctl(fd, EXT4_IOC_RESIZE_FS, &new_size) < 0) {
+ error = errno;
+ pr_error("EXT4_IOC_RESIZE_FS ioctl() failed with error: %s\n", strerror(error));
+ return error;
}
+
return 0;
}
--
2.27.0
Kernel currently doesn't support resize of EXT4 mounted
with sparse_super2 option enabled. Earlier, it used to leave the resize
incomplete and the fs would be left in an inconsistent state, however commit
b1489186cc83[1] fixed this to avoid the fs corruption by clearly returning
-EOPNOTSUPP.
Test to ensure that kernel handles resizing with sparse_super2 correctly. Run
resize for multiple iterations because this sometimes leads to kernel crash due
to fs corruption, which we want to detect.
Related commit in mainline:
[1] commit b1489186cc8391e0c1e342f9fbc3eedf6b944c61
ext4: add check to prevent attempting to resize an fs with sparse_super2
Signed-off-by: Ojaswin Mujoo <[email protected]>
---
I would like to add a few comments on the approach followed in this
test:
1. So although we check the return codes of the resize operation for
proper logging, the test is only considered to be passed if fsck
passes after the resize. This is because resizing a patched kernel
always keeps the fs consistent whereas resizing on unpatched kernel
always corrupts the fs.
2. I've noticed that running mkfs + resize multiple times on unpatched
kernel sometimes results in KASAN reporting use-after-free. Hence, if
we detect the kernel is unpatched (doesn't return EOPNOTSUPP on
resize) we continue iterating to capture this. In this case, we don't
run fsck in each iteration but run it only after all iterations are
complete because _check_scratch_fs exits the test if it fails.
3. In case we detect patched kernel, we stop iterating, and run fsck to
confirm success
tests/ext4/056 | 108 +++++++++++++++++++++++++++++++++++++++++++++
tests/ext4/056.out | 2 +
2 files changed, 110 insertions(+)
create mode 100755 tests/ext4/056
create mode 100644 tests/ext4/056.out
diff --git a/tests/ext4/056 b/tests/ext4/056
new file mode 100755
index 00000000..0f275dea
--- /dev/null
+++ b/tests/ext4/056
@@ -0,0 +1,108 @@
+#! /bin/bash
+# SPDX-License-Identifier: GPL-2.0
+# Copyright (c) 2022 IBM. All Rights Reserved.
+#
+# We don't currently support resize of EXT4 filesystems mounted
+# with sparse_super2 option enabled. Earlier, kernel used to leave the resize
+# incomplete and the fs would be left into an incomplete state, however commit
+# b1489186cc83[1] fixed this to avoid the fs corruption by clearly returning
+# -ENOTSUPP.
+#
+# This test ensures that kernel handles resizing with sparse_super2 correctly
+#
+# Related commit in mainline:
+#
+# [1] commit b1489186cc8391e0c1e342f9fbc3eedf6b944c61
+# ext4: add check to prevent attempting to resize an fs with sparse_super2
+#
+
+. ./common/preamble
+_begin_fstest auto ioctl resize quick
+
+# real QA test starts here
+
+INITIAL_FS_SIZE=1G
+RESIZED_FS_SIZE=$((2*1024*1024*1024)) # 2G
+ONLINE_RESIZE_BLOCK_LIMIT=$((256*1024*1024))
+
+STOP_ITER=255 # Arbitrary return code
+
+_supported_fs ext4
+_require_scratch_size $(($RESIZED_FS_SIZE/1024))
+_require_test_program "ext4_resize"
+
+log()
+{
+ echo "$seq: $*" >> $seqres.full 2>&1
+}
+
+do_resize()
+{
+ _mkfs_dev -E resize=$ONLINE_RESIZE_BLOCK_LIMIT -O sparse_super2 \
+ $SCRATCH_DEV $INITIAL_FS_SIZE >> $seqres.full 2>&1 \
+ || _fail "$MKFS_PROG failed. Exiting"
+
+ _scratch_mount || _fail "Failed to mount scratch partition. Exiting"
+
+ local BS=$(_get_block_size $SCRATCH_MNT)
+ local REQUIRED_BLOCKS=$(($RESIZED_FS_SIZE/$BS))
+
+ local RESIZE_RET
+ local EOPNOTSUPP=95
+
+ log "Resizing FS"
+ $here/src/ext4_resize $SCRATCH_MNT $REQUIRED_BLOCKS >> $seqres.full 2>&1
+ RESIZE_RET=$?
+
+ # Test appears to be successful. Stop iterating and confirm if FS is
+ # consistent.
+ if [ $RESIZE_RET = $EOPNOTSUPP ]
+ then
+ log "Resize operation not supported with sparse_super2"
+ log "Threw expected error $RESIZE_RET (EOPNOTSUPP)"
+ return $STOP_ITER
+ fi
+
+ # Test appears to be unsuccessful. Most probably, the fs is corrupt,
+ # iterate a few more times to further stress test this.
+ log "Something is wrong. Output of resize = $RESIZE_RET. \
+ Expected $EOPNOTSUPP (EOPNOTSUPP)"
+
+ # unmount after ioctl sometimes fails with "device busy" so add a small
+ # delay
+ sleep 0.2
+ _scratch_unmount >> $seqres.full 2>&1 \
+ || _fail "$UMOUNT_PROG failed. Exiting"
+}
+
+run_test()
+{
+ local FSCK_RET
+ local ITERS=8
+ local RET=0
+
+ for i in $(seq 1 $ITERS)
+ do
+ log "----------- Iteration: $i ------------"
+ do_resize
+ RET=$?
+
+ [ "$RET" = "$STOP_ITER" ] && break
+ done
+
+ log "-------- Iterations Complete ---------"
+ log "Checking if FS is in consistent state"
+ _check_scratch_fs
+ FSCK_RET=$?
+
+ [ "$FSCK_RET" -ne "0" ] && \
+ echo "fs corrupt. Check $seqres.full for more details"
+
+ return $FSCK_RET
+}
+
+echo "Silence is golden"
+run_test
+status=$?
+
+exit
diff --git a/tests/ext4/056.out b/tests/ext4/056.out
new file mode 100644
index 00000000..6142fcd2
--- /dev/null
+++ b/tests/ext4/056.out
@@ -0,0 +1,2 @@
+QA output created by 056
+Silence is golden
--
2.27.0
Greetings,
Just a gentle reminder to kindly let me know if any changes are
needed in this patchset.
Thank you,
Ojaswin
On Tue, Feb 22, 2022 at 11:20:51PM +0530, Ojaswin Mujoo wrote:
> As detailed in the patch [1], kernel currently does not support resizes
> with sparse_super2 enabled. Before the above patch, if we used the
> EXT4_IOC_RESIZE_FS ioctl directly, wiht sparse_super2 enabled, the
> kernel used to still try the resize and ultimatley leave the fs in an
> inconsistent state. This also led to corruption and kernel BUGs.
>
> This patchset adds a test for ext4 to ensure that the kernel handles
> resizes with sparse_super2 correctly, and returns -EOPNOTSUPP.
>
> Summary:
>
> Patch 1: Fix the src/ext4_resize.c script to return accurate error
> codes.
> Patch 2: Add the ext4 test for checking resize functionality
>
> Changes since v1 [2]
>
> * In patch 2, don't iterate if the expected errno is returned
> * Code cleanup and additional comments for clarity (also see extra
> note in patch 2)
> * No changes in patch 1
>
> [1] https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=b1489186cc8391e0c1e342f9fbc3eedf6b944c61
> [2] https://lore.kernel.org/fstests/[email protected]/
>
> Ojaswin Mujoo (2):
> src/ext4_resize.c: Refactor code and ensure accurate errno is returned
> ext4: Test to ensure resize with sparse_super2 is handled correctly
>
> src/ext4_resize.c | 46 +++++++++++++------
> tests/ext4/056 | 108 +++++++++++++++++++++++++++++++++++++++++++++
> tests/ext4/056.out | 2 +
> 3 files changed, 142 insertions(+), 14 deletions(-)
> create mode 100755 tests/ext4/056
> create mode 100644 tests/ext4/056.out
>
> --
> 2.27.0
>
On Tue, Feb 22, 2022 at 11:20:53PM +0530, Ojaswin Mujoo wrote:
> Kernel currently doesn't support resize of EXT4 mounted
> with sparse_super2 option enabled. Earlier, it used to leave the resize
> incomplete and the fs would be left in an inconsistent state, however commit
> b1489186cc83[1] fixed this to avoid the fs corruption by clearly returning
> -EOPNOTSUPP.
>
> Test to ensure that kernel handles resizing with sparse_super2 correctly. Run
> resize for multiple iterations because this sometimes leads to kernel crash due
> to fs corruption, which we want to detect.
>
> Related commit in mainline:
>
> [1] commit b1489186cc8391e0c1e342f9fbc3eedf6b944c61
>
> ext4: add check to prevent attempting to resize an fs with sparse_super2
>
> Signed-off-by: Ojaswin Mujoo <[email protected]>
> ---
>
> I would like to add a few comments on the approach followed in this
> test:
>
> 1. So although we check the return codes of the resize operation for
> proper logging, the test is only considered to be passed if fsck
> passes after the resize. This is because resizing a patched kernel
> always keeps the fs consistent whereas resizing on unpatched kernel
> always corrupts the fs.
>
> 2. I've noticed that running mkfs + resize multiple times on unpatched
> kernel sometimes results in KASAN reporting use-after-free. Hence, if
> we detect the kernel is unpatched (doesn't return EOPNOTSUPP on
> resize) we continue iterating to capture this. In this case, we don't
> run fsck in each iteration but run it only after all iterations are
> complete because _check_scratch_fs exits the test if it fails.
>
> 3. In case we detect patched kernel, we stop iterating, and run fsck to
> confirm success
>
> tests/ext4/056 | 108 +++++++++++++++++++++++++++++++++++++++++++++
> tests/ext4/056.out | 2 +
> 2 files changed, 110 insertions(+)
> create mode 100755 tests/ext4/056
> create mode 100644 tests/ext4/056.out
>
> diff --git a/tests/ext4/056 b/tests/ext4/056
> new file mode 100755
> index 00000000..0f275dea
> --- /dev/null
> +++ b/tests/ext4/056
> @@ -0,0 +1,108 @@
> +#! /bin/bash
> +# SPDX-License-Identifier: GPL-2.0
> +# Copyright (c) 2022 IBM. All Rights Reserved.
> +#
> +# We don't currently support resize of EXT4 filesystems mounted
> +# with sparse_super2 option enabled. Earlier, kernel used to leave the resize
> +# incomplete and the fs would be left into an incomplete state, however commit
> +# b1489186cc83[1] fixed this to avoid the fs corruption by clearly returning
> +# -ENOTSUPP.
> +#
> +# This test ensures that kernel handles resizing with sparse_super2 correctly
> +#
> +# Related commit in mainline:
> +#
> +# [1] commit b1489186cc8391e0c1e342f9fbc3eedf6b944c61
> +# ext4: add check to prevent attempting to resize an fs with sparse_super2
> +#
> +
> +. ./common/preamble
> +_begin_fstest auto ioctl resize quick
> +
> +# real QA test starts here
> +
> +INITIAL_FS_SIZE=1G
> +RESIZED_FS_SIZE=$((2*1024*1024*1024)) # 2G
> +ONLINE_RESIZE_BLOCK_LIMIT=$((256*1024*1024))
> +
> +STOP_ITER=255 # Arbitrary return code
> +
> +_supported_fs ext4
> +_require_scratch_size $(($RESIZED_FS_SIZE/1024))
> +_require_test_program "ext4_resize"
> +
> +log()
> +{
> + echo "$seq: $*" >> $seqres.full 2>&1
> +}
> +
> +do_resize()
> +{
> + _mkfs_dev -E resize=$ONLINE_RESIZE_BLOCK_LIMIT -O sparse_super2 \
> + $SCRATCH_DEV $INITIAL_FS_SIZE >> $seqres.full 2>&1 \
> + || _fail "$MKFS_PROG failed. Exiting"
> +
> + _scratch_mount || _fail "Failed to mount scratch partition. Exiting"
> +
> + local BS=$(_get_block_size $SCRATCH_MNT)
> + local REQUIRED_BLOCKS=$(($RESIZED_FS_SIZE/$BS))
> +
> + local RESIZE_RET
> + local EOPNOTSUPP=95
> +
> + log "Resizing FS"
> + $here/src/ext4_resize $SCRATCH_MNT $REQUIRED_BLOCKS >> $seqres.full 2>&1
> + RESIZE_RET=$?
> +
> + # Test appears to be successful. Stop iterating and confirm if FS is
> + # consistent.
> + if [ $RESIZE_RET = $EOPNOTSUPP ]
> + then
> + log "Resize operation not supported with sparse_super2"
> + log "Threw expected error $RESIZE_RET (EOPNOTSUPP)"
> + return $STOP_ITER
> + fi
> +
> + # Test appears to be unsuccessful. Most probably, the fs is corrupt,
> + # iterate a few more times to further stress test this.
> + log "Something is wrong. Output of resize = $RESIZE_RET. \
> + Expected $EOPNOTSUPP (EOPNOTSUPP)"
> +
> + # unmount after ioctl sometimes fails with "device busy" so add a small
> + # delay
> + sleep 0.2
> + _scratch_unmount >> $seqres.full 2>&1 \
> + || _fail "$UMOUNT_PROG failed. Exiting"
> +}
> +
> +run_test()
> +{
> + local FSCK_RET
> + local ITERS=8
> + local RET=0
> +
> + for i in $(seq 1 $ITERS)
> + do
> + log "----------- Iteration: $i ------------"
> + do_resize
> + RET=$?
> +
> + [ "$RET" = "$STOP_ITER" ] && break
> + done
> +
> + log "-------- Iterations Complete ---------"
> + log "Checking if FS is in consistent state"
> + _check_scratch_fs
_check_scratch_fs will exit the test on failure and print error message,
which will break the golden image, so there's no need to check fsck ret.
> + FSCK_RET=$?
> +
> + [ "$FSCK_RET" -ne "0" ] && \
> + echo "fs corrupt. Check $seqres.full for more details"
> +
> + return $FSCK_RET
So I removed above hunk on commit.
Thanks for the test! And my apology to the HUGE delay on review..
Thanks,
Eryu
> +}
> +
> +echo "Silence is golden"
> +run_test
> +status=$?
> +
> +exit
> diff --git a/tests/ext4/056.out b/tests/ext4/056.out
> new file mode 100644
> index 00000000..6142fcd2
> --- /dev/null
> +++ b/tests/ext4/056.out
> @@ -0,0 +1,2 @@
> +QA output created by 056
> +Silence is golden
> --
> 2.27.0
On Sun, Apr 10, 2022 at 06:36:14PM +0800, Eryu Guan wrote:
> On Tue, Feb 22, 2022 at 11:20:53PM +0530, Ojaswin Mujoo wrote:
> > Kernel currently doesn't support resize of EXT4 mounted
> > with sparse_super2 option enabled. Earlier, it used to leave the resize
> > incomplete and the fs would be left in an inconsistent state, however commit
> > b1489186cc83[1] fixed this to avoid the fs corruption by clearly returning
> > -EOPNOTSUPP.
> >
> > Test to ensure that kernel handles resizing with sparse_super2 correctly. Run
> > resize for multiple iterations because this sometimes leads to kernel crash due
> > to fs corruption, which we want to detect.
> >
> > Related commit in mainline:
> >
> > [1] commit b1489186cc8391e0c1e342f9fbc3eedf6b944c61
> >
> > ext4: add check to prevent attempting to resize an fs with sparse_super2
> >
> > Signed-off-by: Ojaswin Mujoo <[email protected]>
> > ---
> >
> > I would like to add a few comments on the approach followed in this
> > test:
> >
> > 1. So although we check the return codes of the resize operation for
> > proper logging, the test is only considered to be passed if fsck
> > passes after the resize. This is because resizing a patched kernel
> > always keeps the fs consistent whereas resizing on unpatched kernel
> > always corrupts the fs.
> >
> > 2. I've noticed that running mkfs + resize multiple times on unpatched
> > kernel sometimes results in KASAN reporting use-after-free. Hence, if
> > we detect the kernel is unpatched (doesn't return EOPNOTSUPP on
> > resize) we continue iterating to capture this. In this case, we don't
> > run fsck in each iteration but run it only after all iterations are
> > complete because _check_scratch_fs exits the test if it fails.
> >
> > 3. In case we detect patched kernel, we stop iterating, and run fsck to
> > confirm success
> >
> > tests/ext4/056 | 108 +++++++++++++++++++++++++++++++++++++++++++++
> > tests/ext4/056.out | 2 +
> > 2 files changed, 110 insertions(+)
> > create mode 100755 tests/ext4/056
> > create mode 100644 tests/ext4/056.out
> >
> > diff --git a/tests/ext4/056 b/tests/ext4/056
> > new file mode 100755
> > index 00000000..0f275dea
> > --- /dev/null
> > +++ b/tests/ext4/056
> > @@ -0,0 +1,108 @@
> > +#! /bin/bash
> > +# SPDX-License-Identifier: GPL-2.0
> > +# Copyright (c) 2022 IBM. All Rights Reserved.
> > +#
> > +# We don't currently support resize of EXT4 filesystems mounted
> > +# with sparse_super2 option enabled. Earlier, kernel used to leave the resize
> > +# incomplete and the fs would be left into an incomplete state, however commit
> > +# b1489186cc83[1] fixed this to avoid the fs corruption by clearly returning
> > +# -ENOTSUPP.
> > +#
> > +# This test ensures that kernel handles resizing with sparse_super2 correctly
> > +#
> > +# Related commit in mainline:
> > +#
> > +# [1] commit b1489186cc8391e0c1e342f9fbc3eedf6b944c61
> > +# ext4: add check to prevent attempting to resize an fs with sparse_super2
> > +#
> > +
> > +. ./common/preamble
> > +_begin_fstest auto ioctl resize quick
> > +
> > +# real QA test starts here
> > +
> > +INITIAL_FS_SIZE=1G
> > +RESIZED_FS_SIZE=$((2*1024*1024*1024)) # 2G
> > +ONLINE_RESIZE_BLOCK_LIMIT=$((256*1024*1024))
> > +
> > +STOP_ITER=255 # Arbitrary return code
> > +
> > +_supported_fs ext4
> > +_require_scratch_size $(($RESIZED_FS_SIZE/1024))
> > +_require_test_program "ext4_resize"
> > +
> > +log()
> > +{
> > + echo "$seq: $*" >> $seqres.full 2>&1
> > +}
> > +
> > +do_resize()
> > +{
> > + _mkfs_dev -E resize=$ONLINE_RESIZE_BLOCK_LIMIT -O sparse_super2 \
> > + $SCRATCH_DEV $INITIAL_FS_SIZE >> $seqres.full 2>&1 \
> > + || _fail "$MKFS_PROG failed. Exiting"
> > +
> > + _scratch_mount || _fail "Failed to mount scratch partition. Exiting"
> > +
> > + local BS=$(_get_block_size $SCRATCH_MNT)
> > + local REQUIRED_BLOCKS=$(($RESIZED_FS_SIZE/$BS))
> > +
> > + local RESIZE_RET
> > + local EOPNOTSUPP=95
> > +
> > + log "Resizing FS"
> > + $here/src/ext4_resize $SCRATCH_MNT $REQUIRED_BLOCKS >> $seqres.full 2>&1
> > + RESIZE_RET=$?
> > +
> > + # Test appears to be successful. Stop iterating and confirm if FS is
> > + # consistent.
> > + if [ $RESIZE_RET = $EOPNOTSUPP ]
> > + then
> > + log "Resize operation not supported with sparse_super2"
> > + log "Threw expected error $RESIZE_RET (EOPNOTSUPP)"
> > + return $STOP_ITER
> > + fi
> > +
> > + # Test appears to be unsuccessful. Most probably, the fs is corrupt,
> > + # iterate a few more times to further stress test this.
> > + log "Something is wrong. Output of resize = $RESIZE_RET. \
> > + Expected $EOPNOTSUPP (EOPNOTSUPP)"
> > +
> > + # unmount after ioctl sometimes fails with "device busy" so add a small
> > + # delay
> > + sleep 0.2
> > + _scratch_unmount >> $seqres.full 2>&1 \
> > + || _fail "$UMOUNT_PROG failed. Exiting"
> > +}
> > +
> > +run_test()
> > +{
> > + local FSCK_RET
> > + local ITERS=8
> > + local RET=0
> > +
> > + for i in $(seq 1 $ITERS)
> > + do
> > + log "----------- Iteration: $i ------------"
> > + do_resize
> > + RET=$?
> > +
> > + [ "$RET" = "$STOP_ITER" ] && break
> > + done
> > +
> > + log "-------- Iterations Complete ---------"
> > + log "Checking if FS is in consistent state"
> > + _check_scratch_fs
>
> _check_scratch_fs will exit the test on failure and print error message,
> which will break the golden image, so there's no need to check fsck ret.
>
> > + FSCK_RET=$?
> > +
> > + [ "$FSCK_RET" -ne "0" ] && \
> > + echo "fs corrupt. Check $seqres.full for more details"
> > +
> > + return $FSCK_RET
>
> So I removed above hunk on commit.
>
> Thanks for the test! And my apology to the HUGE delay on review..
No worries Eryu, thanks a lot for the review :)
Regards,
Ojaswin
>
> Thanks,
> Eryu
>
> > +}
> > +
> > +echo "Silence is golden"
> > +run_test
> > +status=$?
> > +
> > +exit
> > diff --git a/tests/ext4/056.out b/tests/ext4/056.out
> > new file mode 100644
> > index 00000000..6142fcd2
> > --- /dev/null
> > +++ b/tests/ext4/056.out
> > @@ -0,0 +1,2 @@
> > +QA output created by 056
> > +Silence is golden
> > --
> > 2.27.0