I'm currently having some kind of weird issue in my rig. The rig itself consists of some cheap asus board, together with a 5850, 6850, 6950 and a 6770.
they all used to work just fine for the last months.
So the other night, I noticed on my mining tool, that my 6770 went offline, while all other cards continued to work. I thought, fine, I'll reboot the rig and it'll be all fine again.
after rebooting, all cards were still detected, X11 was starting up and not freezing. however when I started the second worker, I noticed that the system became laggy, so I tried to htop and found that all poclbm instances were using an entire core. So I did another reboot, noticed the problem still being there, used another pocl(newer version) - problem still there. I went to hashkill-gpu, which refused to start with no reason. fine again. I then tried CGMiner which told me on start that it couldnt fire the 5850, then started all other GPUs and also had the 100% issue. I thought, ok, maybe I raped the 5850 too hard over the past. but it wasnt that. after rebooting with the 5850 detached, I found the other gpus still using 100% each. So I reattached it, and tried mining anyways. despite all gpus coming online and having OK stale rates, my mining power was only about 50% of what it should be. after killing two pocl instances, the other 2 instances went to full capacity, however using both of my cores to the limit.
then I went installing a new linux stick. after everything was set up, I still had that same error.
finally, I tried each GPU single, and different combinations of them... all tries lead to each pocl instance having 100% CPU usage each.
So I'm out of elaborate guesses now, anyone else have any hints what's wrong with the rig?