[PATCH] xfs_db: check on-disk structure sizes

Eric Sandeen sandeen at sandeen.net
Wed Jan 13 00:02:36 CST 2016


LOn Jan 12, 2016, at 11:47 PM, Dave Chinner <david at fromorbit.com> wrote:
> 
>> On Tue, Jan 12, 2016 at 09:10:23PM -0600, Eric Sandeen wrote:
>> 
>> 
>>> On 1/12/16 7:29 PM, Darrick J. Wong wrote:
>>>> On Tue, Jan 12, 2016 at 09:01:22AM -0500, Brian Foster wrote:
>>>>> On Mon, Jan 11, 2016 at 03:46:44PM -0800, Darrick J. Wong wrote:
>>>>> Check on-disk structure sizes against known values.
>>>>> Use this to catch inadvertent changes in structure size due to padding
>>>>> and alignment issues, etc.
>>>>> 
>>>>> Signed-off-by: Darrick J. Wong <darrick.wong at oracle.com>
>>>>> ---
>>>> 
>>>> What's the need for this in userspace? Not a big deal really, but it
>>>> seems like it serves the fundamental purpose sufficiently in the kernel.
>>> 
>>> The primary point is to make sure that we didn't make any errors with the
>>> on-disk structures when porting libxfs changes.  The kernel build is the first
>>> line of defense since it tends to big get changes first, but I figure a
>>> defensive build check for xfsprogs won't harm anyone...
>> 
>> Does it need to actually be in the code?
>> 
>> $ pahole -s fs/xfs/xfs.ko | grep -w "xfs_dsb\|xfs_agf\|xfs_agi\|xfs_agfl"
>> xfs_agf    224    0
>> xfs_agfl    40    0
>> xfs_agi    336    0
>> xfs_dsb    264    0
>> 
>> pahole needs a binary w/ debuginfo, but maybe this could just be hooked up
>> in the Makefiles?
> 
> As I've pointed out previously to Darrick: xfstests:/tests/xfs/122
> 
> Make that build again, update it.
> 
Oh, I went looking for that and missed it somehow, thought it had been removed.   Ok then!

-Eric

> -Dave.
> -- 
> Dave Chinner
> david at fromorbit.com
> 



More information about the xfs mailing list