Bug 682459

Summary: [abrt] kernel: kernel BUG at mm/slub.c:2834!: TAINTEDDie
Product: [Fedora] Fedora Reporter: Xavier Hourcade <public.oss>
Component: kernelAssignee: Kernel Maintainer List <kernel-maint>
Status: CLOSED NOTABUG QA Contact: Fedora Extras Quality Assurance <extras-qa>
Severity: unspecified Docs Contact:
Priority: unspecified    
Version: 14CC: gansalmon, itamar, jonathan, kernel-maint, madhu.chinakonda, public.oss
Target Milestone: ---   
Target Release: ---   
Hardware: x86_64   
OS: Unspecified   
Whiteboard: abrt_hash:814024449
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2011-06-27 16:57:25 UTC Type: ---
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Attachments:
Description Flags
File: backtrace none

Description Xavier Hourcade 2011-03-05 17:55:33 UTC
abrt version: 1.1.17
architecture: x86_64
Attached file: backtrace, 2789 bytes
cmdline: not_applicable
component: kernel
executable: kernel
kernel: 2.6.35.10-74.fc14.x86_64
package: kernel
reason: kernel BUG at mm/slub.c:2834!
release: Fedora release 14 (Laughlin)
tainted: 128
time: 1299345948
uid: 0

comment
-----
Please note that
- to date, more recent F14 stable kernels, don't permit building this kernel module.
- to date, more recent syntek trunk revs, don't build against any F14 stable kernel (eg. rev 100 or 101)

Whenever I reproduce this bug "on purpose" to document it (3rd time already), I seen to face two distinct levels of severity : once again I needed two attempts to capture the backtrace (both after fresh boot) :
- First crash was NOT handled, system just frooze, system logs just stopped.
- Second crash WAS handled, system recovered, oops was listed in ABRT.

Apart from the present bug, I can use this kernel module daily, just fine, with several videoconf apps.

Additional (extended) information about using this kernel module in Fedora, is available in another bug report (this is about the fact that this webcam is still not suported by default, so I provided here the steps I followed, so I could use my webcam, and links to related threads and bugs at sourceforge) :

https://bugzilla.redhat.com/show_bug.cgi?id=573241

Available if you wish any further info, please ask. Thanks,
Xavier

How to reproduce
-----
Hardware is ASUS V1S laptop with 174f:6a31 webcam

1. Boot kernel 2.6.35.10-74.fc14.x86_64
2. Build and install rev.99 of syntek kernel module @ sourceforge (see comments here after)
3. Start two concurrent processes initializing the webcam "almost at once"
    (eg. two distinct skype logins, I use panel shortcuts for each)

System ALWAYS crash ...if second processes tries accessing the webcam *while* first has not "released it".

Comment 1 Xavier Hourcade 2011-03-05 17:55:36 UTC
Created attachment 482464 [details]
File: backtrace

Comment 2 Xavier Hourcade 2011-03-25 16:09:51 UTC
Since I did patch the module as follow (2011-03-06, daily use), the kernel didn't crash.

(still using module source rev.99, and kernel 2.6.35.10-74.fc14.x86_64)

patch:
http://sourceforge.net/tracker/download.php?group_id=178178&atid=884191&file_id=403550&aid=3187007

thread:
http://sourceforge.net/tracker/?func=detail&aid=3187007&group_id=178178&atid=884191

Is the kernel source expected to protect against such module errors ?

Xavier

Comment 3 Dave Jones 2011-06-27 16:57:25 UTC
no, the kernel can't fix problems in third party code.

Comment 4 Xavier Hourcade 2012-02-24 21:16:18 UTC
Thanks for your answer ..last year :)