FC12 freeze/lockup + kernel panic

[cormanaz]

Greetings. I am having a problem with lockups on a new FC12 box (dual core 3 ghz, 4 gb memory,nvidia 8400gs). This happens reliably when replaying MythTV videos, but also randomly at other times using other apps (I suspect also boot but I can't be sure; it just occasionally stalls part way through the boot display).

Usually but not always this is accompanied by a kernel panic (caps lock + scroll lock lights flash on keyboard).

I did a core dump with kdump and it reports:

Thread 1 (<main task>):
Cannot access memory at address 0xffff880028025b70

I am in the process of running memtest86+ right now. It's been through several passes without errors. I'm going to let it run some more, but if that is dependable it's looking like the RAM sticks are not the problem.

Can anyone suggest what I should do next? Given the reliability of the problem with MythTV playback, which uses VDPAU, I'm thinking this problem might be with the graphics card. I can't find any diagnostic programs like memtest86+ for that, tho. Found this thing called Video Memory Stress Test, but it's for windoze.

Suggestions?

Steve

[Rubberman]

Which video driver are you using?

[cormanaz]

Which video driver are you using?

The latest nvidia drivers from RPM Fusion. rpm -qa | grep nvidia gives:

xorg-x11-drv-nvidia-libs-190.53-1.fc12.x86_64
nvidia-xconfig-1.0-1.fc12.x86_64
xorg-x11-drv-nvidia-190.53-1.fc12.x86_64
kmod-nvidia-2.6.31.9-174.fc12.x86_64-190.53-1.fc12.x86_64
nvidia-settings-1.0-3.2.fc12.x86_64
kmod-nvidia-2.6.31.12-174.2.3.fc12.x86_64-190.53-1.fc12.1.x86_64
kmod-nvidia-190.53-1.fc12.1.x86_64

Also FWIW I just finished running 10 passes of memtest86+ and it showed no problems.

[Rubberman]

I assume you are running the x86_64 (64-bit) version of FC12? At least the nVidia driver, the current release version, is the 64-bit one.

[Rubberman]

Anyway, you should report this to nvidia. Open a command-line window, su to root, and run the script: nvidia-bug-report.sh
Then, email a description of your problem with the file nvidia-bug-report.log.gz that was generated by the script to linux-bugs@nvidia.com

[cormanaz]

I assume you are running the x86_64 (64-bit) version of FC12? At least the nVidia driver, the current release version, is the 64-bit one.

Yep. It is the 64 bit version.

[cormanaz]

Anyway, you should report this to nvidia. Open a command-line window, su to root, and run the script: nvidia-bug-report.sh
Then, email a description of your problem with the file nvidia-bug-report.log.gz that was generated by the script to linux-bugs@nvidia.com

OK sounds like a plan. I sure don't know what else to do! Do you think it's possible I've just got a bad vid card, tho?

[Rubberman]

Possible. There should be some sort of reply pretty quickly.

[cormanaz]

I finally got crash working and got a proper crash report. It is attached for your reading enjoyment. Doesn't mean much to me, but the problem seems to be in the part that says:

NVRM: Xid (0001:00): 26, Ch 00000003 M 00000304 D 00000001 intr 00400000
NVRM: Xid (0001:00): 26, Ch 00000003 M 00000858 D 00000000 intr 04c00000
NVRM: Xid (0001:00): 26, Ch 00000003 M 00000858 D 00000000 intr 00400000
NVRM: Xid (0001:00): 26, Ch 00000003 M 00000858 D 00000000 intr 00c00000
NVRM: Xid (0001:00): 26, Ch 00000003 M 00000858 D 00000000 intr 00400000
NVRM: Xid (0001:00): 26, Ch 00000003 M 00000858 D 00000000 intr 00c00000
NVRM: Xid (0001:00): 26, Ch 00000003 M 00000858 D 00000000 intr 00400000
NVRM: Xid (0001:00): 26, Ch 00000003 M 00000858 D 00000000 intr 00c00000
BUG: unable to handle kernel paging request at ffffc9000a180000
IP: [<ffffffffa054e643>] os_mem_set+0x11/0x18 [nvidia]
PGD 12fc07067 PUD 12fc18067 PMD 12a4f2067 PTE 0
Thread overran stack, or stack corrupted
Oops: 0002 [#1] SMP

So your intuition that this is an nvidia problem seems to be correct. I sent the nvidia-bug-report.log.gz plus this info off to linux-bugs@nvidia.com so hopefully they will have some guideance.

[Rubberman]

Seems like the correct approach to me. At least nVidia might be able to make something of this, such as verifying that there is a hardware, or driver problem, or not...