On Sat, Nov 05, 2022 at 11:23:29AM +0800, Zorro Lang wrote:
> Test nfs and its underlying fs, make sure file size as expected
> after writting a file, and the speculative allocation space can
> be shrunken.
>
> Signed-off-by: Zorro Lang <zlang@xxxxxxxxxx>
> ---
>
> Hi,
>
> The original bug reproducer is:
> 1. mount nfs3 backed by xfs
> 2. dd if=/dev/zero of=/nfs/10M bs=1M count=10
> 3. du -sh /nfs/10M
> 16M /nfs/10M
>
> As this was a xfs issue, so cc linux-xfs@ to get review.
>
> Thanks,
> Zorro
>
> tests/nfs/002 | 43 +++++++++++++++++++++++++++++++++++++++++++
> tests/nfs/002.out | 2 ++
> 2 files changed, 45 insertions(+)
> create mode 100755 tests/nfs/002
> create mode 100644 tests/nfs/002.out
>
> diff --git a/tests/nfs/002 b/tests/nfs/002
> new file mode 100755
> index 00000000..3d29958d
> --- /dev/null
> +++ b/tests/nfs/002
> @@ -0,0 +1,43 @@
> +#! /bin/bash
> +# SPDX-License-Identifier: GPL-2.0
> +# Copyright (c) 2022 Red Hat, Inc. All Rights Reserved.
> +#
> +# FS QA Test 002
> +#
> +# Make sure nfs gets expected file size after writting a big sized file. It's
> +# not only testing nfs, test its underlying fs too. For example a known old bug
> +# on xfs (underlying fs) caused nfs get larger file size (e.g. 16M) after
> +# writting 10M data to a file. It's fixed by a series of patches around
> +# 579b62faa5fb16 ("xfs: add background scanning to clear eofblocks inodes")
> +#
> +. ./common/preamble
> +_begin_fstest auto rw
> +
> +# real QA test starts here
> +_supported_fs nfs
> +_require_test
> +
> +localfile=$TEST_DIR/testfile.$seq
> +rm -rf $localfile
> +
> +$XFS_IO_PROG -f -t -c "pwrite 0 10m" -c "fsync" $localfile >>$seqres.full 2>&1
> +block_size=`stat -c '%B' $localfile`
> +iblocks_expected=$((10 * 1024 * 1024 / $block_size))
> +# Try several times for the speculative allocated file size can be shrunken
> +res=1
> +for ((i=0; i<10; i++));do
> + iblocks_real=`stat -c '%b' $localfile`
> + if [ "$iblocks_expected" = "$iblocks_real" ];then
> + res=0
> + break
> + fi
> + sleep 10
> +done
Hmm... this case sometimes fails on kernel 6.1.0-rc6 [1] (nfs4.2 base on xfs),
even I changed the sleep time to 20s * 10, it still fails. But I can't reproduce
this failure if the underlying fs is ext4... cc linux-xfs, to check if I miss
something for xfs? Or this's a xfs issue?
Thanks,
Zorro
[1]
# ./check nfs/002
FSTYP -- nfs
PLATFORM -- Linux/x86_64 dell-per640-04 6.1.0-rc6 #1 SMP PREEMPT_DYNAMIC Mon Nov 21 00:51:20 EST 2022
MKFS_OPTIONS -- dell-per640-04.dell2.lab.eng.bos.redhat.com:/mnt/xfstests/scratch/nfs-server
MOUNT_OPTIONS -- -o vers=4.2 -o context=system_u:object_r:root_t:s0 dell-per640-04.dell2.lab.eng.bos.redhat.com:/mnt/xfstests/scratch/nfs-server /mnt/xfstests/scratch/nfs-client
nfs/002 3s ... - output mismatch (see /var/lib/xfstests/results//nfs/002.out.bad)
--- tests/nfs/002.out 2022-11-21 01:29:33.861770474 -0500
+++ /var/lib/xfstests/results//nfs/002.out.bad 2022-11-21 13:27:37.424199056 -0500
@@ -1,2 +1,3 @@
QA output created by 002
+Write 20480 blocks, but get 32640 blocks
Silence is golden
...
(Run 'diff -u /var/lib/xfstests/tests/nfs/002.out /var/lib/xfstests/results//nfs/002.out.bad' to see the entire diff)
Ran: nfs/002
Failures: nfs/002
Failed 1 of 1 tests
> +if [ $res -ne 0 ];then
> + echo "Write $iblocks_expected blocks, but get $iblocks_real blocks"
> +fi
> +
> +echo "Silence is golden"
> +# success, all done
> +status=0
> +exit
> diff --git a/tests/nfs/002.out b/tests/nfs/002.out
> new file mode 100644
> index 00000000..61705c7c
> --- /dev/null
> +++ b/tests/nfs/002.out
> @@ -0,0 +1,2 @@
> +QA output created by 002
> +Silence is golden
> --
> 2.31.1
>
