Poclbm crashing with aticaldd.dll?

[Meatball]

Hey all,

I've been getting crashes the last day or so of the poclbm.exe process through GUIMiner.  Every so often the driver will crash with the following error in the event log.

Faulting application name: poclbm.exe, version: 0.0.0.0, time stamp: 0x4918019b
Faulting module name: aticaldd.dll, version: 6.14.10.1385, time stamp: 0x4dae39e9
Exception code: 0xc0000005
Fault offset: 0x0014c858
Faulting process id: 0xa0c
Faulting application start time: 0x01cc24519cb9cbbe
Faulting application path: C:\Program Files (x86)\Bitcoin\guiminer\poclbm.exe
Faulting module path: C:\Windows\system32\aticaldd.dll
Report Id: 42e1bd75-9049-11e0-8274-001b217ca570

I'm using Catalyst 11.5's with SDK 2.4 on Win7 x64.  Have a Radeon 6950 that is not overclocked, but does have the unlocked shaders bios.  I've tried yanking the drivers/SDK and reinstalling to no avail.

Anyone have any thoughts?

[1714775144]

1714775144

[1714775144]

1714775144

[1714775144]

1714775144

[1714775144]

1714775144

[1714775144]

1714775144

[Meatball]

Just an update.  Switching to phoenix miner and everything is stable.  Must be something messed up with poclbm on my machine.

[Phage]

Hi, I am a miner for third day today. The last two days there werent any errors whatsoever, while the last night and whole day today I am getting crashes on the poclbm.exe. The fault module name is aticaldd.dll. Also I am running 6950 too, but mine is running on default. It seems there is some problem on the miner side, not ours.

Btw, I was just about to start a new thread. 

[Meatball]

Hmm, that's strange considering it's the same Driver set/miner that I've been using for weeks without a problem.

[Phage]

I also see that our timestamps are the same. Here's my error log:

Problem signature:
  Problem Event Name:   APPCRASH
  Application Name:   poclbm.exe
  Application Version:   0.0.0.0
  Application Timestamp:   4918019b
  Fault Module Name:   aticaldd.dll
  Fault Module Version:   6.14.10.1385
  Fault Module Timestamp:   4dae39e9
  Exception Code:   c0000005
  Exception Offset:   0014c858
  OS Version:   6.1.7601.2.1.0.256.1
  Locale ID:   2057
  Additional Information 1:   0a9e
  Additional Information 2:   0a9e372d3b4ad19135b953a78882e789
  Additional Information 3:   0a9e
  Additional Information 4:   0a9e372d3b4ad19135b953a78882e789

Read our privacy statement online:
  http://go.microsoft.com/fwlink/?linkid=104288&clcid=0x0409

If the online privacy statement is not available, please read our privacy statement offline:
  C:\Windows\system32\en-US\erofflps.txt

[Meatball]

Hmm, interesting.  I was mining with Slush's fund and had the crashes occur at these times (All EST):

11:22:34 AM
11:16:21 AM
10:54:55 AM
10:50:01 AM
10:28:36 AM
09:56:37 AM
09:37:45 AM
09:21:27 AM
More might have occurred between these times, but I was AFK and wasn't around to restart the miner.
07:43:10 AM
07:23:17 AM

Did you have crashes at the same time?  I wonder if it has anything to do with the points where blocks were found...

[melco]

The same for me on one 6950 (I have 9 of them and only one card in separate PC crashes). Phoenix also crashes in the same way with poclbm core.
Any ideas?
Win7 32bit, Catalyst 11.5 latest

[Meatball]

I was never able to determine the problem other than the fact that I was running more than one copy of poclbm.  Had one running towards a pool and another running solo with a lower aggression in case the pool went down.  I can run for for days as long as I have just have one or the other going, but as soon as I launch the 2nd one, within 10 minutes or so, the driver crashes.

Not sure why it's doing.  I'm running multiples poclbm processes on other machines for the same card without a problem, it's just this machine, so I just run the pooled miner and not the secondary miner.

I've tried 11.5 and 11.6 and I'm running Win7 x64.

[Waltibaba]

My 6950 wasn't too stable either using guiminer, I switched to phoenix rising, and thats so much better. What clocks are you running on? maybe its just bad cards?
EDIT - just saw that its because of multiple poclbm threads processes; have you tried setting priority/affinity in task manager?

[Meatball]

I'm running 860/625 clocks, but have tried at stock clocks and get the same result.  I haven't really mucked around too much with process affinity, but I'll give it a whirl.

[frizzl]

The 69xx series has a known problem where anything using OpenCL doesn't share resources with other programs.

Those cards will all lock.

Read somewhere that youtube video would cause such a crash