From: "Darrick J. Wong" <djwong@kernel.org>
To: Christoph Hellwig <hch@lst.de>
Cc: zlang@kernel.org, dchinner@redhat.com, linux-xfs@vger.kernel.org,
fstests@vger.kernel.org
Subject: Re: [PATCH] xfs/011: support byte-based grant heads are stored in bytes now
Date: Thu, 20 Jun 2024 12:56:06 -0700 [thread overview]
Message-ID: <20240620195606.GH103034@frogsfrogsfrogs> (raw)
In-Reply-To: <20240620072309.533010-1-hch@lst.de>
On Thu, Jun 20, 2024 at 09:23:09AM +0200, Christoph Hellwig wrote:
> New kernels where reservation grant track the actual reservation space
> consumed in bytes instead of LSNs in cycle/block tuples export different
> sysfs files for this information.
>
> Adapt the test to detect which version is exported, and simply check
> for a near-zero reservation space consumption for the byte based version.
>
> Based on work from Dave Chinner.
>
> Signed-off-by: Christoph Hellwig <hch@lst.de>
> ---
> tests/xfs/011 | 67 +++++++++++++++++++++++++++++++++------------------
> 1 file changed, 44 insertions(+), 23 deletions(-)
>
> diff --git a/tests/xfs/011 b/tests/xfs/011
> index ed44d074b..ef2366adf 100755
> --- a/tests/xfs/011
> +++ b/tests/xfs/011
> @@ -11,7 +11,18 @@
> . ./common/preamble
> _begin_fstest auto freeze log metadata quick
>
> -# Import common functions.
> +# real QA test starts here
> +_supported_fs xfs
> +
> +_require_scratch
> +_require_freeze
> +_require_xfs_sysfs $(_short_dev $TEST_DEV)/log
> +_require_command "$KILLALL_PROG" killall
> +
> +. ./common/filter
> +
> +devname=`_short_dev $SCRATCH_DEV`
> +attrprefix="/sys/fs/xfs/$devname/log"
>
> # Override the default cleanup function.
> _cleanup()
> @@ -24,27 +35,29 @@ _cleanup()
> rm -f $tmp.*
> }
>
> -# Use the information exported by XFS to sysfs to determine whether the log has
> -# active reservations after a filesystem freeze.
> -_check_scratch_log_state()
> +_check_scratch_log_state_new()
> {
> - devname=`_short_dev $SCRATCH_DEV`
> - attrpath="/sys/fs/xfs/$devname/log"
> -
> - # freeze the fs to ensure data is synced and the log is flushed. this
> - # means no outstanding transactions, and thus no outstanding log
> - # reservations, should exist
> - xfs_freeze -f $SCRATCH_MNT
> + # The grant heads record reservations in bytes. For complex reasons
> + # beyond the scope fo this test, these aren't going to be exactly zero
of
Why aren't they going to be exactly zero?
--D
> + # for frozen filesystems. Hence just check the value is between 0 and
> + # the maximum iclog size (256kB).
> + for attr in "reserve_grant_head_bytes" "write_grant_head_bytes"; do
> + space=`cat $attrprefix/$attr`
> + _within_tolerance $space 1024 0 $((256 * 1024))
> + done
> +}
>
> +_check_scratch_log_state_old()
> +{
> # the log head is exported in basic blocks and the log grant heads in
> # bytes. convert the log head to bytes for precise comparison
> - log_head_cycle=`awk -F : '{ print $1 }' $attrpath/log_head_lsn`
> - log_head_bytes=`awk -F : '{ print $2 }' $attrpath/log_head_lsn`
> + log_head_cycle=`awk -F : '{ print $1 }' $attrprefix/log_head_lsn`
> + log_head_bytes=`awk -F : '{ print $2 }' $attrprefix/log_head_lsn`
> log_head_bytes=$((log_head_bytes * 512))
>
> for attr in "reserve_grant_head" "write_grant_head"; do
> - cycle=`cat $attrpath/$attr | awk -F : '{ print $1 }'`
> - bytes=`cat $attrpath/$attr | awk -F : '{ print $2 }'`
> + cycle=`cat $attrprefix/$attr | awk -F : '{ print $1 }'`
> + bytes=`cat $attrprefix/$attr | awk -F : '{ print $2 }'`
>
> if [ $cycle != $log_head_cycle ] ||
> [ $bytes != $log_head_bytes ]
> @@ -54,17 +67,25 @@ _check_scratch_log_state()
> "possible leak detected."
> fi
> done
> -
> - xfs_freeze -u $SCRATCH_MNT
> }
>
> -# real QA test starts here
> -_supported_fs xfs
> +# Use the information exported by XFS to sysfs to determine whether the log has
> +# active reservations after a filesystem freeze.
> +_check_scratch_log_state()
> +{
> + # freeze the fs to ensure data is synced and the log is flushed. this
> + # means no outstanding transactions, and thus no outstanding log
> + # reservations, should exist
> + xfs_freeze -f $SCRATCH_MNT
>
> -_require_scratch
> -_require_freeze
> -_require_xfs_sysfs $(_short_dev $TEST_DEV)/log
> -_require_command "$KILLALL_PROG" killall
> + if [ -f "${attrprefix}/reserve_grant_head_bytes" ]; then
> + _check_scratch_log_state_new
> + else
> + _check_scratch_log_state_old
> + fi
> +
> + xfs_freeze -u $SCRATCH_MNT
> +}
>
> echo "Silence is golden."
>
> --
> 2.43.0
>
>
next prev parent reply other threads:[~2024-06-20 19:56 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-06-20 7:23 [PATCH] xfs/011: support byte-based grant heads are stored in bytes now Christoph Hellwig
2024-06-20 19:56 ` Darrick J. Wong [this message]
2024-06-21 6:29 ` Christoph Hellwig
2024-06-21 16:20 ` 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=20240620195606.GH103034@frogsfrogsfrogs \
--to=djwong@kernel.org \
--cc=dchinner@redhat.com \
--cc=fstests@vger.kernel.org \
--cc=hch@lst.de \
--cc=linux-xfs@vger.kernel.org \
--cc=zlang@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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox