Bug 1381363

Summary: Multiple oops in 4.8.0
Product: [Fedora] Fedora Reporter: darrell pfeifer <darrellpf>
Component: kernelAssignee: Neil Horman <nhorman>
Status: CLOSED NOTABUG QA Contact: Fedora Extras Quality Assurance <extras-qa>
Severity: unspecified Docs Contact:
Priority: unspecified    
Version: rawhideCC: gansalmon, ichavero, itamar, jonathan, kernel-maint, labbott, madhu.chinakonda, mchehab, nhorman
Target Milestone: ---   
Target Release: ---   
Hardware: Unspecified   
OS: Unspecified   
Whiteboard:
Fixed In Version: Doc Type: If docs needed, set a value
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2016-10-18 13:31:49 UTC Type: Bug
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
journalctl output for the boot none

Description darrell pfeifer 2016-10-03 20:43:25 UTC
Created attachment 1206986 [details]
journalctl output for the boot

Description of problem

Multiple oops on boot.


Version-Release number of selected component (if applicable):

4.8.0-1.fc26.x86_64

Additional info:

Booted back to rc8, which was ok. Rebooted again into 4.8.0 final and there wasn't an oops.

Comment 1 Laura Abbott 2016-10-03 21:55:42 UTC
Just to clarify, this oops was only seen once?

Comment 2 darrell pfeifer 2016-10-03 21:57:11 UTC
Yes, just happened once. The next reboot it was gone. Not sure if it is a timing problem, so will just have to watch reboots to see if it happens again.

Comment 3 Laura Abbott 2016-10-03 22:03:32 UTC
I suspect it may be a random bit flip. The crash occurred in the memory allocator so everything that attempted to allocate crashed. We can monitor and see if it re-occurs.

Comment 4 Neil Horman 2016-10-18 13:31:24 UTC
I agree, this seems like an oddity.  Either it was a bit flip, or the slab allocator returned a pointer so large that when get_alloc_info computed the kasan offset, it overflowed to exactly zero/NULL.  Either way it needs to be reproducible to make progress.

Comment 5 Neil Horman 2016-10-18 13:31:24 UTC
I agree, this seems like an oddity.  Either it was a bit flip, or the slab allocator returned a pointer so large that when get_alloc_info computed the kasan offset, it overflowed to exactly zero/NULL.  Either way it needs to be reproducible to make progress.

Comment 6 Neil Horman 2016-10-18 13:31:49 UTC
please repoen if it occurs again