xfs
[Top] [All Lists]

Re: [PATCH] xfs: fix uflags detection at xfs_fs_rm_xquota

To: Christoph Hellwig <hch@xxxxxxxxxxxxx>
Subject: Re: [PATCH] xfs: fix uflags detection at xfs_fs_rm_xquota
From: Jeff Liu <jeff.liu@xxxxxxxxxx>
Date: Wed, 16 Jul 2014 16:32:12 +0800
Cc: "xfs@xxxxxxxxxxx" <xfs@xxxxxxxxxxx>
Delivered-to: xfs@xxxxxxxxxxx
In-reply-to: <20140716071250.GA1461@xxxxxxxxxxxxx>
References: <53C4F5DB.2060108@xxxxxxxxxx> <20140715102520.GH30363@xxxxxxxxxxxxx> <53C62595.6010702@xxxxxxxxxx> <20140716071250.GA1461@xxxxxxxxxxxxx>
User-agent: Mozilla/5.0 (X11; Linux x86_64; rv:24.0) Gecko/20100101 Thunderbird/24.1.0
On 07/16/2014 15:12 PM, Christoph Hellwig wrote:
> On Wed, Jul 16, 2014 at 03:11:17PM +0800, Jeff Liu wrote:
>>> This code was introduced by commit
>>> 9da93f9b7cdf8ab28da6b364cdc1fafc8670b4dc which attempts to fix the
>>> broken quotarm code.  Seems like we need a testcase for this
>>> functionality or just remove it.  Physically removing the quota file
>>> never seems that useful to me to start with..
>>
>> Is there an approach to verify that a particular quota file inode has
>> been successfully removed?  It seems we have to do this in a test case
>> but am not yet figure out a way, or any thoughts for the test case?
> 
> umount the filesystem and check that the field in the superblock
> pointing to that kind of quota inode is zeroed out would be my
> suggestion.

Oh yes, after quota remove, the 'u' field of the quota inode will changed from
u.bmx[0] = [startoff,startblock,blockcount,extentflag] 0:[0,16,1,0] to u = 
(empty)
which can be used for the verification, thanks for your suggestion. :)


Cheers,
-Jeff

<Prev in Thread] Current Thread [Next in Thread>