[PATCH v2] fstests: generic/730: exclude btrfs for now

[PATCH v2] fstests: generic/730: exclude btrfs for now

[Qu Wenruo]

The test case always fail for btrfs:

  generic/730       - output mismatch (see /home/adam/xfstests-dev/results//generic/730.out.bad)
    --- tests/generic/730.out	2024-04-25 18:13:45.203549435 +0930
    +++ /home/adam/xfstests-dev/results//generic/730.out.bad	2025-06-04 15:10:39.062430952 +0930
    @@ -1,2 +1 @@
     QA output created by 730
    -cat: -: Input/output error
    ...

The root reason is that, btrfs doesn't implement its blk_holder_ops when
opening a block device.
Thus when the underlying block device is marked dead, btrfs is never
going to know thus no way to shutdown (nor btrfs has a way to shutdown
either).

I'm trying to improve the situation, but until then just exlucde btrfs
from the test case for now.

Signed-off-by: Qu Wenruo <wqu@suse.com>
---
Changelog:
v2:
- Update the root reason
  It's not the sb->s_op->shutdown, that is only for single device
  fs (through fs_bdev_mark_dead()).
  For a multi-devices fs, it should provide a blk_holder_ops when
  opening the block device.
---
 tests/generic/730 | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/tests/generic/730 b/tests/generic/730
index 6251980e..cae6489f 100755
--- a/tests/generic/730
+++ b/tests/generic/730
@@ -26,6 +26,10 @@ _require_test
 _require_block_device $TEST_DEV
 _require_scsi_debug
 
+if [ "$FSTYP" = "btrfs" ]; then
+	_notrun "btrfs doesn't support per-fs shutdown yet"
+fi
+
 size=$(_small_fs_size_mb 256)
 SCSI_DEBUG_DEV=`_get_scsi_debug_dev 512 512 0 $size`
 test -b "$SCSI_DEBUG_DEV" || _notrun "Failed to initialize scsi debug device"
-- 
2.49.0

Re: [PATCH v2] fstests: generic/730: exclude btrfs for now

[Anand Jain]

On 6/4/2025 3:10 PM, Qu Wenruo wrote:
> The test case always fail for btrfs:
> 
>    generic/730       - output mismatch (see /home/adam/xfstests-dev/results//generic/730.out.bad)
>      --- tests/generic/730.out	2024-04-25 18:13:45.203549435 +0930
>      +++ /home/adam/xfstests-dev/results//generic/730.out.bad	2025-06-04 15:10:39.062430952 +0930
>      @@ -1,2 +1 @@
>       QA output created by 730
>      -cat: -: Input/output error
>      ...



[PATCH v1 5/7] generic/730: add _require_scratch_shutdown

Fixed it.


> 
> The root reason is that, btrfs doesn't implement its blk_holder_ops when
> opening a block device.
> Thus when the underlying block device is marked dead, btrfs is never
> going to know thus no way to shutdown (nor btrfs has a way to shutdown
> either).
> 
> I'm trying to improve the situation, but until then just exlucde btrfs
> from the test case for now.
> 
> Signed-off-by: Qu Wenruo <wqu@suse.com>
> ---
> Changelog:
> v2:
> - Update the root reason
>    It's not the sb->s_op->shutdown, that is only for single device
>    fs (through fs_bdev_mark_dead()).
>    For a multi-devices fs, it should provide a blk_holder_ops when
>    opening the block device.

Along with that, I think it is a good idea to bring  XFS_IOC_GOINGDOWN, 
per fs (all devices in a multi-device fs) and vfs level so that we could 
support unmount --force.

>   
> +if [ "$FSTYP" = "btrfs" ]; then
> +	_notrun "btrfs doesn't support per-fs shutdown yet"
> +fi
> +

_require_scratch_shutdown() will take care.


Thanks, Anand

Re: [PATCH v2] fstests: generic/730: exclude btrfs for now

[Qu Wenruo]

在 2025/6/4 17:40, Anand Jain 写道:
> On 6/4/2025 3:10 PM, Qu Wenruo wrote:
>> The test case always fail for btrfs:
>>
>>    generic/730       - output mismatch (see /home/adam/xfstests-dev/ 
>> results//generic/730.out.bad)
>>      --- tests/generic/730.out    2024-04-25 18:13:45.203549435 +0930
>>      +++ /home/adam/xfstests-dev/results//generic/730.out.bad    
>> 2025-06-04 15:10:39.062430952 +0930
>>      @@ -1,2 +1 @@
>>       QA output created by 730
>>      -cat: -: Input/output error
>>      ...
> 
> 
> 
> [PATCH v1 5/7] generic/730: add _require_scratch_shutdown

That works fine for btrfs, but that's only for now.

The _require_scratch_shutdown is only going to check the shutdown ioctl, 
not the per-block-device shutdown.

I guess this means for fs supporting per-bdev shutdown, it must have a 
dedicated shutdown ioctl.

Thanks,
Qu>
> Fixed it.
> 
> 
>>
>> The root reason is that, btrfs doesn't implement its blk_holder_ops when
>> opening a block device.
>> Thus when the underlying block device is marked dead, btrfs is never
>> going to know thus no way to shutdown (nor btrfs has a way to shutdown
>> either).
>>
>> I'm trying to improve the situation, but until then just exlucde btrfs
>> from the test case for now.
>>
>> Signed-off-by: Qu Wenruo <wqu@suse.com>
>> ---
>> Changelog:
>> v2:
>> - Update the root reason
>>    It's not the sb->s_op->shutdown, that is only for single device
>>    fs (through fs_bdev_mark_dead()).
>>    For a multi-devices fs, it should provide a blk_holder_ops when
>>    opening the block device.
> 
> Along with that, I think it is a good idea to bring  XFS_IOC_GOINGDOWN, 
> per fs (all devices in a multi-device fs) and vfs level so that we could 
> support unmount --force.
> 
>> +if [ "$FSTYP" = "btrfs" ]; then
>> +    _notrun "btrfs doesn't support per-fs shutdown yet"
>> +fi
>> +
> 
> _require_scratch_shutdown() will take care.
> 
> 
> Thanks, Anand
>