Bug 447826
Summary: | Segfaults & Recursive Faults | ||||||||
---|---|---|---|---|---|---|---|---|---|
Product: | [Fedora] Fedora | Reporter: | Nigel Jones <dev> | ||||||
Component: | kernel | Assignee: | Kernel Maintainer List <kernel-maint> | ||||||
Status: | CLOSED NOTABUG | QA Contact: | Fedora Extras Quality Assurance <extras-qa> | ||||||
Severity: | low | Docs Contact: | |||||||
Priority: | low | ||||||||
Version: | 9 | CC: | nhorman | ||||||
Target Milestone: | --- | ||||||||
Target Release: | --- | ||||||||
Hardware: | All | ||||||||
OS: | Linux | ||||||||
Whiteboard: | |||||||||
Fixed In Version: | Doc Type: | Bug Fix | |||||||
Doc Text: | Story Points: | --- | |||||||
Clone Of: | Environment: | ||||||||
Last Closed: | 2008-05-23 13:33:46 UTC | Type: | --- | ||||||
Regression: | --- | Mount Type: | --- | ||||||
Documentation: | --- | CRM: | |||||||
Verified Versions: | Category: | --- | |||||||
oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | |||||||
Cloudforms Team: | --- | Target Upstream Version: | |||||||
Embargoed: | |||||||||
Attachments: |
|
Description
Nigel Jones
2008-05-21 23:13:20 UTC
Created attachment 306326 [details]
dmesg output after Recursive Fault
kernel BUG at mm/filemap.c:126! BUG_ON(page_mapped(page)); Your machine has taken a machine check error: "Tainted: G M" Can you run the mcelog program when that happens and see what the error is? it might be worth a run of memtest86 for a while too, just to rule out bad ram. These things are common indications of hardware problems of some kind (bad ram/insufficient power/cooling, or just general flakyness) (In reply to comment #2) > kernel BUG at mm/filemap.c:126! > BUG_ON(page_mapped(page)); > > > Your machine has taken a machine check error: > "Tainted: G M" > > Can you run the mcelog program when that happens and see what the error is? > I shall attempt this when I next time I see it, already had two segfaults today. I also just noticed another recursive fault (Kernel Oops spotted it in all fairness), so how am I meant to run mcelog? (In reply to comment #3) > it might be worth a run of memtest86 for a while too, just to rule out bad ram. > These things are common indications of hardware problems of some kind (bad > ram/insufficient power/cooling, or just general flakyness) I was considering bad RAM, I recently installed an extra 2 gig but dare I say it, it seems to run Vista okay, and I ran Fedora 8 quite happily until recently. I can understand your other points too although I'll give credit that it's worked pretty well for the last 8ish months w/ both Linux and Windows. I'll run memtest86 when I go to sleep tonight or have dinner though. Created attachment 306449 [details]
dmesg | mcelog --ascii
Okay scrap my last comment, google'd and got 'dmesg | mcelog --ascii' this is
the result.
"HARDWARE ERROR" seems to be the tell tale sign, I take it this is referring to
my nice dual core processor and in fact not a Kernel Bug? I'm a little
confused here, so a point in the right direction would be most appreciated.
Looking at the mcelog manpage, I think you just want to run mcelog without any arguments. The old method of writing events to the syslog is obsolete. (In reply to comment #6) > Looking at the mcelog manpage, I think you just want to run mcelog without any > arguments. The old method of writing events to the syslog is obsolete. That returned absolutely nothing (In reply to comment #3) > it might be worth a run of memtest86 for a while too, just to rule out bad ram. > These things are common indications of hardware problems of some kind (bad > ram/insufficient power/cooling, or just general flakyness) I think you might be right, I gave up memtest86+'ing it after 900 errors (spread over all 4 slots) in 40 minutes. Looks like I need to have a fiddle with the RAM config etc and work out whats going on. IMO it's a 'notabug' agree? yeah, sounds like a hardware fault of some sort, and given the recent addition of RAM, that's a likely suspect. We frequently see things like this where Windows runs just fine. It's purely by luck really. The access patterns of the two operating systems are completely different, and perhaps Linux employs more aggressive caching of data (or maybe we just read more of disk, or ...) So many variables, that it's not really a data point worth putting any faith in. |