Re: [PATCH] Attempt to sync the fsstress writes to a frozen F.S

[Eric Sandeen <sandeen@redhat.com>]

On 5/11/11 2:10 AM, Surbhi Palande wrote:
> While the fsstress background writes are busy dirtying the page cache, if a
> fsfreeze happens then the background writes should stall. A sync should then
> not have any data to sync to the FS. If it does have any data to sync then
> sync will cause a deadlock by holding the s_umount write semaphore and waiting
> in the wait queue for the FS to thaw, whereas the F.S can never thaw without
> getting the s_umount write semaphore.
> 
> Signed-off-by: Surbhi Palande <surbhi.palande@canonical.com>

Seems ok to me.  In the future, when sending xfstests patches,
if you can add "xfstests" to the subject, and cc: the xfs list,
it'd be great.

I presume that this test does fail for you without your fixes?

I'll see if anyone on the xfs list has comments and if not, I can check this in.

Thanks,
-Eric

> ---
>  068 |    5 +++++
>  1 files changed, 5 insertions(+), 0 deletions(-)
> 
> diff --git a/068 b/068
> index 82c1a4e..b9ac58d 100755
> --- a/068
> +++ b/068
> @@ -101,6 +101,11 @@ do
>  	    tee -a $seq.full
>  	sleep 2
>  
> +	# there should be nothing to sync at this point. This may hang in case
> +	# of fsstress background writes dirtying the page cache while the F.S is frozen
> +	sync &
> +	sleep 2
> +
>  	echo "*** thawing  \$SCRATCH_MNT" | tee -a $seq.full
>  	xfs_freeze -u "$SCRATCH_MNT" | tee -a $seq.full
>  	[ $? != 0 ] && echo xfs_freeze -u "$SCRATCH_MNT" failed | \