Created attachment 78991 [details] Kernel Config My machine panics when it has a high workload - usually compiling I managed to photograph the panic and I'll attach it, the .config and info from my logs I'm not sure if I've assigned the bug to the right group This bug is reproducable - each time I've tried to recompile icedtea
Created attachment 79001 [details] Extract from /var/log/messages
Created attachment 79011 [details] Picture of Panic Sorry I don't have a text version of this
I've been noticing this since the start of the 3.6 serious but was unable to reliably bisect the issue.
Created attachment 79021 [details] Dmesg from SandyBridge Laptop
does the problem still exist in the latest upstream kernel?
I've just tried running 9924a1992a86ebdb7ca36ef790d2ba0da506296c and so far it hasn't failed (I tired compiling libreoffice and icedtea whilst watching a video) I've been mostly running 3.6.0-rc3 from git://people.freedesktop.org/~airlied/linux so my discreet card gets switched off when not in use I'll keep using the latest from master to see if it's truly resolved Would you like me to see figure out which rc's this bug was present in?
(In reply to comment #6) > I've just tried running 9924a1992a86ebdb7ca36ef790d2ba0da506296c and so far > it > hasn't failed (I tired compiling libreoffice and icedtea whilst watching a > video) > > I've been mostly running 3.6.0-rc3 from > git://people.freedesktop.org/~airlied/linux so my discreet card gets switched > off when not in use > > I'll keep using the latest from master to see if it's truly resolved > > Would you like me to see figure out which rc's this bug was present in? No, if the bug is fixed in the latest upstream kernel. Yes, if the problem still exists. :)
ping...
Damn I thought it was fixed but it happened again just now 3.7.0-rc6-tip+ 3587b1b097d70c2eb9fee95ea7995d13c05f66e5 Was the running kernel and the SHA of the first commit in the log I have pictures of the panic which I'll attach now - I'd like to point out this is the first panic I've had since switching to the newer kernels so something has changed to make it less frequent
Created attachment 87061 [details] Newest panic This is the latest panic - I have a less clearer one the goes right to the right edge of the screen if you need any of the longer lines
(In reply to comment #3) > I've been noticing this since the start of the 3.6 serious but was unable to > reliably bisect the issue. can you reproduce it with 3.6 kernel?
3.6 final or the latest point release?
(In reply to comment #3) > I've been noticing this since the start of the 3.6 serious but was unable to > reliably bisect the issue. I asked just to verify that which release that you did not notice the problem.
Then yes it was first noticed under a 3.6 rc
does the problem exists in 3.5?
No I'm quite sure the issue only appeared in the 3.6 merge window - unfortunately due to its infrequent nature I wasn't able to bisect it down to the problem commit. The idea of compiling icedtea between each bisect is a bit off putting even with an i7 - but I could give it a go this weekend if you think it'll help
*** Bug 50681 has been marked as a duplicate of this bug. ***
hi, Mike, can you help me check if the problem still exists in 3.9-rc1 please?
you can try to reproduce this by setting the cur_state of the processor cooling device to the max_state.
Hi Zhang I did encounter the pstates issue in rc1 (now fixed of course) but I've not had the original issue in a while I'm quite happy for this bug to be closed