On Wed, Dec 25, 2024 at 08:51:20PM +0800, Zhang Yi wrote: > From: Zhang Yi <yi.zhang@xxxxxxxxxx> > > This addresses a data corruption issue encountered during partial page > zeroing in ext4 which the block size is smaller than the page size [1]. > Add a new test which is expanded upon generic/567, this test performs a > zeroing range test that spans two partial pages to cover this case, and > also generalize it to work for non-4k page sizes. > > Link: https://lore.kernel.org/linux-ext4/20241220011637.1157197-2-yi.zhang@xxxxxxxxxxxxxxx/ [1] > Signed-off-by: Zhang Yi <yi.zhang@xxxxxxxxxx> > --- > v1->v2: > - Add a new test instead of modifying generic/567. > - Generalize the test to work for non-4k page sizes. > v1: https://lore.kernel.org/fstests/20241223023930.2328634-1-yi.zhang@xxxxxxxxxxxxxxx/ > > tests/generic/758 | 76 +++++++++++++++++++++++++++++++++++++++++++ > tests/generic/758.out | 3 ++ > 2 files changed, 79 insertions(+) > create mode 100755 tests/generic/758 > create mode 100644 tests/generic/758.out > > diff --git a/tests/generic/758 b/tests/generic/758 > new file mode 100755 > index 00000000..e03b5e80 > --- /dev/null > +++ b/tests/generic/758 > @@ -0,0 +1,76 @@ > +#! /bin/bash > +# SPDX-License-Identifier: GPL-2.0 > +# Copyright (c) 2024 Huawei. All Rights Reserved. > +# > +# FS QA Test No. generic/758 "FS QA Test No. 758" is ok here, or whatever ./new spat out. > +# > +# Test mapped writes against zero-range to ensure we get the data > +# correctly written. This can expose data corruption bugs on filesystems > +# where the block size is smaller than the page size. > +# > +# (generic/567 is a similar test but for punch hole.) > +# > +. ./common/preamble > +_begin_fstest auto quick rw zero > + > +# Override the default cleanup function. > +_cleanup() > +{ > + cd / > + rm -r -f $verifyfile $testfile Don't bother deleting anything on $SCRATCH_MNT, it'll get mkfs'd out of existence soon enough. > +} > + > +# Import common functions. > +. ./common/filter > + > +_require_test > +_require_scratch > +_require_xfs_io_command "fzero" > + > +verifyfile=$TEST_DIR/verifyfile Also is there any harm in putting verifyfile on $SCRATCH_MNT and thereby not having to override _cleanup? --D > +testfile=$SCRATCH_MNT/testfile > + > +pagesz=$(getconf PAGE_SIZE) > + > +_scratch_mkfs > /dev/null 2>&1 > +_scratch_mount > + > +_dump_files() > +{ > + echo "---- testfile ----" > + _hexdump $testfile > + echo "---- verifyfile --" > + _hexdump $verifyfile > +} > + > +# Build verify file, the data in this file should be consistent with > +# that in the test file. > +$XFS_IO_PROG -f -c "pwrite -S 0x58 0 $((pagesz * 3))" \ > + -c "pwrite -S 0x59 $((pagesz / 2)) $((pagesz * 2))" \ > + $verifyfile | _filter_xfs_io >> /dev/null > + > +# Zero out straddling two pages to check that the mapped write after the > +# range-zeroing are correctly handled. > +$XFS_IO_PROG -t -f \ > + -c "pwrite -S 0x58 0 $((pagesz * 3))" \ > + -c "mmap -rw 0 $((pagesz * 3))" \ > + -c "mwrite -S 0x5a $((pagesz / 2)) $((pagesz * 2))" \ > + -c "fzero $((pagesz / 2)) $((pagesz * 2))" \ > + -c "mwrite -S 0x59 $((pagesz / 2)) $((pagesz * 2))" \ > + -c "close" \ > +$testfile | _filter_xfs_io > $seqres.full > + > +echo "==== Pre-Remount ===" > +if ! cmp -s $testfile $verifyfile; then > + echo "Data does not match pre-remount." > + _dump_files > +fi > +_scratch_cycle_mount > +echo "==== Post-Remount ==" > +if ! cmp -s $testfile $verifyfile; then > + echo "Data does not match post-remount." > + _dump_files > +fi > + > +status=0 > +exit > diff --git a/tests/generic/758.out b/tests/generic/758.out > new file mode 100644 > index 00000000..d01c1959 > --- /dev/null > +++ b/tests/generic/758.out > @@ -0,0 +1,3 @@ > +QA output created by 758 > +==== Pre-Remount === > +==== Post-Remount == > -- > 2.39.2 > >
