linux-ext4 - Re: [PATCH 4/4] ext4: fix suboptimal seek

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <87y4q6joy4.fsf@openvz.org>
Date:	Wed, 17 Dec 2014 18:06:27 +0300
From:	Dmitry Monakhov <dmonakhov@...nvz.org>
To:	Theodore Ts'o <tytso@....edu>
Cc:	linux-ext4@...r.kernel.org
Subject: Re: [PATCH 4/4] ext4: fix suboptimal seek_{data,hole} extents traversial

Theodore Ts'o <tytso@....edu> writes:

> On Fri, Dec 12, 2014 at 11:52:29AM +0300, Dmitry Monakhov wrote:
>> Oh. Indeed That is my mistake.
>> 
>> ret = ext4_fiemap(inode, &fie, offset, maxsize - offset);
>> /* No extents found, EOF */
>> if (!fie.fi_extents_mapped) {
>>      ret = -ENXIO;
>>      break;
>>   }                                                                           
>> Delalloc case handled incorrectly. Will fix that ASAP.
>
> Hi Dmitry, how are things going with a fix?  It would be great if I
> can get something which I can test and push to Linus before he
> releases -rc1.
I'm working on that. Sorry for delay. Delay is caused by confusion of
existing seek_xxx implementation (even w/o my changes)
For example: it is not obvious to believe that if bh is (bh_unwritten |
bh_uptodate) then it contains valid data. IMHO it is just a fallocated
area.  I already have simple fix for actual regression. But I try to
review,optimize and retest all related logic.

FYI: Currently test failed only on EXT4 w/o extents but w/ -odelalloc enabled.
This configuration is absent in xfstest-bld where conf/ext3 is EXT4 w/o
extents and w/ nodelalloc. That is why we oversee this from very beginning. 
>
> Thanks!!
>
> 					- Ted

Download attachment "signature.asc" of type "application/pgp-signature" (473 bytes)