linux-ext4 - Re: data corruption with ext4 (from 2.6.27.4) exposed by rtorrent

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]

Date:	Thu, 6 Nov 2008 10:31:43 +0100
From:	Cryptooctoploid <cryptooctoploid@...il.com>
To:	"Theodore Tso" <tytso@....edu>
Cc:	"Roc Valles" <vallesroc@...il.com>, linux-ext4@...r.kernel.org,
	"Jindrich Makovicka" <makovick@...il.com>,
	Solofo.Ramangalahy@...l.net, torvalds@...ux-foundation.org
Subject: Re: data corruption with ext4 (from 2.6.27.4) exposed by rtorrent

On Wed, Nov 5, 2008 at 6:25 PM, Theodore Tso <tytso@....edu> wrote:
>
> Based on some reflection and further discussion on #ext4, I have a
> theory which we need to try to prove/disprove that it may be related
> to memory pressure combined with the delayed allocation code.

Jindrich Makovicka wrote:
>The following testcase was used to trigger the infamous MAP_SHARED
>dirty flag bug. Maybe it could be of some help here too:
>
>http://lkml.org/lkml/2006/12/27/180

Solofo.Ramangalahy wrote:
>The test case triggers corruption with 2.6.28-rc3 + ext4 patch queue:
. bunch of errors like
>  Chunk 71637 corrupted (0-1339)  (2756-4095)
>  Expected 213, got 0
>  with default mount.
>. nodelalloc is ok.

I also triggered the bug on my hardware (x86_64, with 2GB memory)
while running Linus' test program with TARGETSIZE (1600 << 20).
The symptoms are exactly the same as in the famous former bug,
that Linus discovered.

Here is an excerpt from the output:

Writing chunk 1149123/1149124 (99%)
Chunk 3 corrupted (0-1459)  (111c-16cf)
Expected 3, got 0
Written as (519444)1144595(687310)
Chunk 15 corrupted (0-1459)  (558c-5b3f)
Expected 15, got 0
Written as (99887)645566(1121294)
Chunk 16 corrupted (0-1215)  (5b40-5fff)
Expected 16, got 0
Written as (645566)1121294(920116)
Chunk 26 corrupted (0-1459)  (9448-99fb)
Expected 26, got 0
Written as (361425)1122832(756521)
Chunk 27 corrupted (0-1459)  (99fc-9faf)
Expected 27, got 0
Written as (1122832)756521(414420)
Chunk 34 corrupted (0-1459)  (c1e8-c79b)
Expected 34, got 0
Written as (204725)934398(543959)
Chunk 39 corrupted (404-1459)  (e000-e41f)
Expected 39, got 0
Written as (857308)1016367(413445)
Chunk 60 corrupted (0-1459)  (15630-15be3)
Expected 60, got 0
Written as (218658)603309(105651)
Chunk 84 corrupted (240-1459)  (1e000-1e4c3)
Expected 84, got 0
Written as (207103)982280(141982)
Chunk 109 corrupted (604-1459)  (27000-27357)
Expected 109, got 0
Written as (415001)965566(309924)
Chunk 114 corrupted (0-1459)  (28a28-28fdb)
Expected 114, got 0
Written as (420741)1107613(233665)
Chunk 119 corrupted (0-1459)  (2a6ac-2ac5f)
Expected 119, got 0
Written as (576520)1013770(680816)
Chunk 129 corrupted (0-1459)  (2dfb4-2e567)
Expected 129, got 0
Written as (272475)1136094(865523)
Chunk 130 corrupted (0-1459)  (2e568-2eb1b)
Expected 130, got 0
Written as (1136094)865523(71027)
Chunk 132 corrupted (0-1459)  (2f0d0-2f683)
Expected 132, got 0
Written as (71027)1087332(516188)
Chunk 139 corrupted (0-1459)  (318bc-31e6f)
Expected 139, got 0
Written as (31550)1035248(265709)
Chunk 141 corrupted (0-1459)  (32424-329d7)
Expected 141, got 0
Written as (265709)1039673(102047)
Chunk 150 corrupted (0-1459)  (35778-35d2b)
Expected 150, got 0
Written as (499376)970317(736517)
Chunk 168 corrupted (0-1459)  (3be20-3c3d3)
Expected 168, got 0
Written as (162733)1104408(467712)
Chunk 199 corrupted (0-275)  (46eec-46fff)
Expected 199, got 0
Written as (277601)946981(262145)
Chunk 212 corrupted (0-1459)  (4b910-4bec3)
Expected 212, got 0
Written as (72548)1090863(578882)
--
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