linux-ext4 - Re: [PATCH v3] ext4: Don't set PageUptodate in ext4_end

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] [day] [month] [year] [list]

Date:	Tue, 10 May 2011 12:45:51 -0700
From:	Hugh Dickins <hughd@...gle.com>
To:	"Ted Ts'o" <tytso@....edu>
Cc:	linux-ext4@...r.kernel.org, Jim Meyering <jim@...ering.net>,
	Mingming Cao <cmm@...ibm.com>,
	Curt Wohlgemuth <curtw@...gle.com>
Subject: Re: [PATCH v3] ext4: Don't set PageUptodate in ext4_end_bio()

On Tue, May 10, 2011 at 12:17 PM, Ted Ts'o <tytso@....edu> wrote:
> On Tue, May 10, 2011 at 10:41:56AM -0700, Hugh Dickins wrote:
>>
>> I'm concerned that we've reached -rc7, with Linus planning on 2.6.39
>> release next week, but Curt's fix above to the mblk_io corruption bug
>> seems to have fallen through the cracks.
>
> It's in the ext4 master branch so it's queued to be pushed during the
> next merge window, and then it would go into the 2.6.39.x stable
> series.  So it didn't fall through the cracks; it was just a question
> of whether to push late in the rc series, or waiting until the merge
> window and then backporting to 2.6.39 stable.  Basically I got the
> patch in -rc5, and at that point, given that the vast majority of the
> file systems are 4k blocksize on x86, and most Power and Itanic users
> are using distribution kernels (and that's a very small number
> anyway), I decided not to push it to Linus at that point.
>
> It's a judgement call; and I could have gone the other way; it was
> very much a 49/51 sort of decision.

Hmm.  Well, thanks for the reply.  I do disagree with your decision,
and am surprised that you're still seeing it in the past tense.  This
is a corruption and a regression: prime material for a fix even at
this stage.

Admittedly, in my case the data was written correctly and only
temporarily presented incorrectly; but in other tests that could then
have got written back incorrectly.  And admittedly, not many people
outside of the large-page-size world are using blocksize less than
pagesize.  But even so....

Moral: cc Rafael's regression list next time?

Hugh
--
To unsubscribe from this list: send the line "unsubscribe linux-ext4" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html