[Bugme-janitors] [Bug 9520] Quotas modification on a loaded system takes up to several minutes to complete

[bugme-daemon at bugzilla.kernel.org]

http://bugzilla.kernel.org/show_bug.cgi?id=9520





------- Comment #11 from oleg at tv-sign.ru  2007-12-09 08:58 -------
On 12/09, bugme-daemon at bugzilla.kernel.org wrote:
>
> ------- Comment #10 from gentuu at gmail.com  2007-12-09 08:25 -------
>
> - Ugh. this "endless" loop if allocation fails...
>
> I'll appreciate _any_ suggestions...  should we try several times and fail
> on permanent allocation failure then?

Heh. You need a more knowledgeable person in that area :) We have them on CC.
My opinion, we should break the loop, initialize() inodes we already found,
and repeat again.

Or return an error. Offtopic to this patch, but we don't check the return
value of ->initialize(), can't it fail?

> Yep, I know. But this doesn't really matter in the situation :)
>
>   IOW, list_for_each(pos, head) does nothing if there is no other
> > list_head's
> >   on list.
>
>
> That's why I'm explicitly freeing the last cache object.

Yep. But we don't do ->initialize() + dput() for that inode.

Oleg.


-- 
Configure bugmail: http://bugzilla.kernel.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.