Errors from firegl/fglrx after install

[dmadls]

I've never posted a question before so I'll do my best.
I've been using pclos for several years now and absolutely love it.  I've been using UNIX systems for about 25 years.
I have a file server that I installed a 2010 era pclos release onto and never updated it.  It just ran great for 2 years doing it's file server stuff with samba and NFS and occasional usage for libre office, internet, etc.

A couple weeks ago I finally decided to update it and put 2012.9 on it.  I did all the updates (by updating respositories etc) and all updates are current on it.  But I did notice after the initial update that every 3 seconds, an error gets logged to the message, syslog, and user.log, and after about 3 days the root filesystem fills up and the system crashes...  It's the exact same message in all 3 logs. It is pasted below.

Seems to be graphics card related, I believe...
Graphics card is: ‎RV620 LE [Radeon HD 3450]
The graphics module that loaded is: ATI Radeon HD 2000 and later (radeon/fglrx)
CPU: Mobile AMD Athlon(tm) XP-M Processor 2800+
Flags: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 syscall nx mmxext 3dnowext 3dnow up extd_apicid

The system does have onboard video but I disabled it in the bios and run video from the ATI card.

From the message logs:
Feb  9 18:45:30 Hokie klogd: BUG: using smp_processor_id() in preemptible [00000000] code: firegl/1143
Feb  9 18:45:30 Hokie klogd: caller is kas_GetExecutionLevel+0x13/0x20 [fglrx]
Feb  9 18:45:30 Hokie klogd: Pid: 1143, comm: firegl Tainted: P           O 3.2.18-pclos2.bfs #1
Feb  9 18:45:30 Hokie klogd: Call Trace:
Feb  9 18:45:30 Hokie klogd:  [<c02af297>] debug_smp_processor_id+0xc7/0xe0
Feb  9 18:45:30 Hokie klogd:  [<efc6fd43>] kas_GetExecutionLevel+0x13/0x20 [fglrx]
Feb  9 18:45:30 Hokie klogd:  [<efc732fd>] KAS_GetExecutionLevel+0xd/0x10 [fglrx]
Feb  9 18:45:30 Hokie klogd:  [<efc9062b>] MCIL_GetExecutionLevel+0x4b/0x90 [fglrx]
Feb  9 18:45:30 Hokie klogd:  [<efd7ecc3>] AcquireWorkQueueLock+0x23/0xd0 [fglrx]
Feb  9 18:45:30 Hokie klogd:  [<efd6c69e>] PassiveRing_WorkerThreadRoutine+0x10e/0x300 [fglrx]
Feb  9 18:45:30 Hokie klogd:  [<c015fe53>] ? sched_clock_local+0xd3/0x1c0
Feb  9 18:45:30 Hokie klogd:  [<c0132e85>] ? pc_user_time+0x25/0x140
Feb  9 18:45:30 Hokie klogd:  [<c010b438>] ? __switch_to_xtra+0xf8/0x110
Feb  9 18:45:30 Hokie klogd:  [<c010202a>] ? __switch_to+0x1ca/0x1f0
Feb  9 18:45:30 Hokie klogd:  [<c0135afb>] ? get_parent_ip+0xb/0x40
Feb  9 18:45:30 Hokie klogd:  [<c0135bab>] ? sub_preempt_count+0x7b/0xb0
Feb  9 18:45:30 Hokie klogd:  [<c0439b6f>] ? schedule+0x59f/0xc40
Feb  9 18:45:30 Hokie klogd:  [<c0135afb>] ? get_parent_ip+0xb/0x40
Feb  9 18:45:30 Hokie klogd:  [<efc6efa0>] ? FGLDRM_name_info+0x130/0x130 [fglrx]
Feb  9 18:45:30 Hokie klogd:  [<efd80ff9>] IRQMGR_WorkerThreadRoutine+0x19/0x40 [fglrx]
Feb  9 18:45:30 Hokie klogd:  [<efc9225f>] IRQMGR_CallbackWrapper+0xf/0x20 [fglrx]
Feb  9 18:45:30 Hokie klogd:  [<efc6efa0>] ? FGLDRM_name_info+0x130/0x130 [fglrx]
Feb  9 18:45:30 Hokie klogd:  [<efc6f007>] kasThreadRoutineHelper+0x67/0xb0 [fglrx]
Feb  9 18:45:30 Hokie klogd:  [<efd80fe0>] ? IRQMGR_QueueDpc+0x30/0x30 [fglrx]
Feb  9 18:45:30 Hokie klogd:  [<efd80fe0>] ? IRQMGR_QueueDpc+0x30/0x30 [fglrx]
Feb  9 18:45:30 Hokie klogd:  [<c0159494>] kthread+0x74/0x80
Feb  9 18:45:30 Hokie klogd:  [<c0159420>] ? kthread_worker_fn+0x170/0x170
Feb  9 18:45:30 Hokie klogd:  [<c043d83e>] kernel_thread_helper+0x6/0x10

Thank you for any help.

[AndrzejL]

Try installing the non-bfs kernel. I had a trouble from the bfs kernel - it would run into segfault or something each time I plugged in my WD Mybook USB 3.0 2TB HDD. Something is fishy with the BFS kernel (PAE-BFS too) so try installing:

kernel-3.2.18-pclos2-1-1pclos2012

and reboot to it. Maybe it will help.

Regards.

Andrzej

[gseaman]

Try installing the non-bfs kernel. I had a trouble from the bfs kernel - it would run into segfault or something each time I plugged in my WD Mybook USB 3.0 2TB HDD. Something is fishy with the BFS kernel (PAE-BFS too) so try installing:

kernel-3.2.18-pclos2-1-1pclos2012

and reboot to it. Maybe it will help.

Regards.

Andrzej

Andrzej,
Have you reported problems to Con Kolivas? I watch his blog and he gets a very small number of bug reports, and he fixes them all, if they turn out to be bfs flaws.

Galen

[AndrzejL]

Try installing the non-bfs kernel. I had a trouble from the bfs kernel - it would run into segfault or something each time I plugged in my WD Mybook USB 3.0 2TB HDD. Something is fishy with the BFS kernel (PAE-BFS too) so try installing:

kernel-3.2.18-pclos2-1-1pclos2012

and reboot to it. Maybe it will help.

Regards.

Andrzej

Andrzej,
Have you reported problems to Con Kolivas? I watch his blog and he gets a very small number of bug reports, and he fixes them all, if they turn out to be bfs flaws.

Galen

I didn't. I might do that in the future but not the nearest future...

Cheers.

Andrzej

[gseaman]

I would report it, but I can't replicate the problem. 

Galen

[AndrzejL]

I would report it, but I can't replicate the problem. 

Galen

I would report it but I do not have any pclinuxos installs.

Regards.

Andrzej

[Just17]

I would report it, but I can't replicate the problem. 

Galen

I would report it but I do not have any pclinuxos installs.

Regards.

Andrzej

...........  and yet you are a tester ...........  seems rather odd ..........

[Texstar]

I would report it, but I can't replicate the problem.  

Galen

I cannot replicate the problem either because my ATI cards are the 4000 series.

[AndrzejL]

I would report it, but I can't replicate the problem.  

Galen

I would report it but I do not have any pclinuxos installs.

Regards.

Andrzej

...........  and yet you are a tester ...........  seems rather odd ..........

Indeed . I volunteered in the past when I still had it. Now all I have left is the title. It could be removed I guess. Never bothered to pursue that topic.

But let's get back to the original problem.

Regards.

Andrzej

[dmadls]

Thank you for the input.  I usually enjoy a good software / config challenge but I took the easy way out on this one.  I was in Best Buy today and they had all their Nvidia cards on sale.  I bought a GeForce 630 (PNY rebrand) and put it in my machine.  After booting, I logged in at the graphical login and was then put into text mode were it looked like an error was reported from harddrake during the "OK/FAILED" startups, the rest of the system was booted and it left me at a text login.  I'm not versed enough as to where to go from there... so, saved off some config files, fresh install of 2013.02, and 30 minutes later had a working system with all updates and my favorite apps loaded.  Maybe that was faster than troubleshooting kernels... not sure.