Re: [PATCH 2/2] btrfs/163: replace sprout instead of seed

[Filipe Manana <fdmanana@gmail.com>]

On Sat, Sep 5, 2020 at 12:25 AM Anand Jain <anand.jain@oracle.com> wrote:
>
> Make this test case inline with the kernel patch [1] changes
> [1] btrfs: fix replace of seed device

Same comment as in the previous patch. Now that this is in Linus'
tree, it would be good to mention the commit id too.
>
> So use the sprout device as the replace target instead of the seed device.
> This change is compatible with the older kernels.
>
> While at this, this patch also fixes a typo fix as well.
>
> Signed-off-by: Anand Jain <anand.jain@oracle.com>
> ---
>  tests/btrfs/163     | 21 ++++++++++++++++-----
>  tests/btrfs/163.out |  5 ++++-
>  2 files changed, 20 insertions(+), 6 deletions(-)
>
> diff --git a/tests/btrfs/163 b/tests/btrfs/163
> index 24c725afb6b9..354d88502d47 100755
> --- a/tests/btrfs/163
> +++ b/tests/btrfs/163
> @@ -4,11 +4,15 @@
>  #
>  # FS QA Test 163
>  #
> -# Test case to verify that a seed device can be replaced
> +# Test case to verify that a sprouted device can be replaced
>  #  Create a seed device
>  #  Create a sprout device
>  #  Remount RW
> -#  Run device replace on the seed device
> +#  Run device replace on the sprout device
> +#
> +# Depends on the kernel patch
> +#   btrfs: fail replace of seed device
> +
>  seq=`basename $0`
>  seqres=$RESULT_DIR/$seq
>  echo "QA output created by $seq"
> @@ -39,6 +43,7 @@ _supported_fs btrfs
>  _supported_os Linux
>  _require_command "$BTRFS_TUNE_PROG" btrfstune
>  _require_scratch_dev_pool 3
> +_require_btrfs_forget_or_module_loadable
>
>  _scratch_dev_pool_get 3
>
> @@ -52,7 +57,7 @@ create_seed()
>         run_check _mount $dev_seed $SCRATCH_MNT
>         $XFS_IO_PROG -f -d -c "pwrite -S 0xab 0 4M" $SCRATCH_MNT/foobar >\
>                 /dev/null
> -       echo -- gloden --
> +       echo -- golden --
>         od -x $SCRATCH_MNT/foobar
>         _run_btrfs_util_prog filesystem show -m $SCRATCH_MNT
>         _scratch_unmount
> @@ -64,22 +69,28 @@ add_sprout()
>  {
>         _run_btrfs_util_prog device add -f $dev_sprout $SCRATCH_MNT
>         _run_btrfs_util_prog filesystem show -m $SCRATCH_MNT
> +       _mount -o remount,rw $dev_sprout $SCRATCH_MNT
> +       $XFS_IO_PROG -f -d -c "pwrite -S 0xcd 0 4M" $SCRATCH_MNT/foobar2 >\
> +               /dev/null

Same comment as for the other patch.
Why the direct IO? The test passes with buffered IO as well.
If there's a reason for direct IO, it should be mentioned in a comment
and _require_odirect added above.

>  }
>
>  replace_seed()
>  {
> -       _run_btrfs_util_prog replace start -fB $dev_seed $dev_replace_tgt $SCRATCH_MNT
> +       _run_btrfs_util_prog replace start -fB $dev_sprout $dev_replace_tgt $SCRATCH_MNT

So now the function should be renamed from replace_seed() to
replace_sprout() as well. Shouldn't it?

Other than that, it looks good and it works as expected.

Thanks.

>         _run_btrfs_util_prog filesystem show -m $SCRATCH_MNT
>         _scratch_unmount
> -       run_check _mount $dev_replace_tgt $SCRATCH_MNT
> +       _btrfs_forget_or_module_reload
> +       run_check _mount -o device=$dev_seed $dev_replace_tgt $SCRATCH_MNT
>         echo -- sprout --
>         od -x $SCRATCH_MNT/foobar
> +       od -x $SCRATCH_MNT/foobar2
>         _scratch_unmount
>
>  }
>
>  seed_is_mountable()
>  {
> +       _btrfs_forget_or_module_reload
>         run_check _mount $dev_seed $SCRATCH_MNT
>         _run_btrfs_util_prog filesystem show -m $SCRATCH_MNT
>         _scratch_unmount
> diff --git a/tests/btrfs/163.out b/tests/btrfs/163.out
> index 91f6f5b6f48a..351ef7b040b2 100644
> --- a/tests/btrfs/163.out
> +++ b/tests/btrfs/163.out
> @@ -1,5 +1,5 @@
>  QA output created by 163
> --- gloden --
> +-- golden --
>  0000000 abab abab abab abab abab abab abab abab
>  *
>  20000000
> @@ -7,3 +7,6 @@ QA output created by 163
>  0000000 abab abab abab abab abab abab abab abab
>  *
>  20000000
> +0000000 cdcd cdcd cdcd cdcd cdcd cdcd cdcd cdcd
> +*
> +20000000
> --
> 2.25.1
>


-- 
Filipe David Manana,

“Whether you think you can, or you think you can't — you're right.”