[PATCH] xfs/122: reduce xfs_dsb_t size

[PATCH] xfs/122: reduce xfs_dsb_t size

[Darrick J. Wong]

We're not going to do reverse-mapping on the realtime device for
some time, so let's fix xfs/122 in the meantime.

Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
---
 tests/xfs/122.out |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/tests/xfs/122.out b/tests/xfs/122.out
index 67f7698..168cfd5 100644
--- a/tests/xfs/122.out
+++ b/tests/xfs/122.out
@@ -138,7 +138,7 @@ sizeof(xfs_dir2_sf_off_t) = 2
 sizeof(xfs_disk_dquot_t) = 104
 sizeof(xfs_dq_logformat_t) = 24
 sizeof(xfs_dqblk_t) = 136
-sizeof(xfs_dsb_t) = 272
+sizeof(xfs_dsb_t) = 264
 sizeof(xfs_efd_log_format_32_t) = 28
 sizeof(xfs_efd_log_format_64_t) = 32
 sizeof(xfs_efi_log_format_32_t) = 28

Re: [PATCH] xfs/122: reduce xfs_dsb_t size

[Christoph Hellwig]

On Tue, Jan 24, 2017 at 09:08:19AM -0800, Darrick J. Wong wrote:
> We're not going to do reverse-mapping on the realtime device for
> some time, so let's fix xfs/122 in the meantime.
> 
> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>

Looks fine,

Reviewed-by: Christoph Hellwig <hch@lst.de>

Re: [PATCH] xfs/122: reduce xfs_dsb_t size

[Eric Sandeen]

On 1/24/17 11:09 AM, Christoph Hellwig wrote:
> On Tue, Jan 24, 2017 at 09:08:19AM -0800, Darrick J. Wong wrote:
>> We're not going to do reverse-mapping on the realtime device for
>> some time, so let's fix xfs/122 in the meantime.
>>
>> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
> 
> Looks fine,
> 
> Reviewed-by: Christoph Hellwig <hch@lst.de>

Philosophically (I'm good at that) how do we handle tests like this
that really only work on the latest version of the code, but don't
actually indicate a bug if it fails on something older?

-Eric

Re: [PATCH] xfs/122: reduce xfs_dsb_t size

[Christoph Hellwig]

On Tue, Jan 24, 2017 at 11:11:35AM -0600, Eric Sandeen wrote:
> Philosophically (I'm good at that) how do we handle tests like this
> that really only work on the latest version of the code, but don't
> actually indicate a bug if it fails on something older?

That might be another argument for doing the smaller than check that
Darrick suggested.  Or an argument for simply stopping to check xfs_dsb,
it's growing constantly.  Checking the offsets of known fields IFF
present wold seem more useful.