[04:10] <RAOF> Woot!  I think that's KDE fixed.
[04:11] <RAOF> At least as far as “not crashing the X server on logout with intel” can be regarded as “fixed” :P
[04:56] <Sarvatt> are you kidding me? LOL
[04:57] <Sarvatt> I had that patch backported in a bunch of the fixed ones I was testing, but didn't try that combination of adding the FreeScratchPixmapsForScreen with that patch at the same time, bah!
[04:59] <Sarvatt> checking if sandybridge works with that, I can dream :)
[05:06] <andrewaclt> The wiki says I should run a mainline kernel so I can get a batchbuffer dump. However, I'm having some driver issue it seems. Is there another way to get a batchbuffer dump?
[05:10] <RAOF> andrewaclt: It depends on which Ubuntu release you're running - Maverick has a 2.6.35-based kernel, so it's ready to go.  10.04 has a 2.6.33-based kernel (at least for its drm), so needs a newer kernel to get proper error_state.
[05:11] <andrewaclt> RAOF, Yeah, I installed the mainline 2.6.36 kernel, but it's very laggy
[05:12] <andrewaclt> the mouse pointer is.
[05:12] <RAOF> After a while, or from the get-go?
[05:12] <andrewaclt> get-go
[05:12] <RAOF> I haven't checked recently, but there was a serious memory leak in drm at the -rc1 stage or so.
[05:12] <andrewaclt> It I mean it's fine for 5-10 seconds, freezes, rinse and repeat
[05:13] <RAOF> Oh.
[05:13] <RAOF> That's probably the polling for output loop.
[05:14] <andrewaclt> Is it fixable? Kind of hard to debug my other issue :)
[05:14] <Sarvatt> i think theres a i915 module option to disable polling to fix that
[05:15] <Sarvatt> can change it at runtime, looking for where it is now
[05:16] <Sarvatt> http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=e58f637bb96d5a0ae0919b9998b891d1ba7e47c9
[05:17] <Sarvatt> echo 0 | sudo tee /sys/module/drm_kms_helper/parameters/poll
[05:17] <Sarvatt> err echo N I guess
[05:18] <Sarvatt> can boot with drm_kms_helper.poll=N added to grub to turn it off too
[05:18] <andrewaclt> that seems to have done it
[05:20] <andrewaclt> Sarvatt, Thanks.
[05:33] <andrewaclt> One last question, if I can't reproduce my gpu freeze on the vanilla kernel what do I do?
[06:42] <bryceh> andrewaclt, if you can find out an upstream kernel patch that fixes it, you can request that patch be added to the ubuntu kernel
[06:42] <bryceh> although honestly it's probably kinda too late to get anything more into the kernel
[06:42] <bryceh> at least, for this release
[14:08] <arie_> Hello, I am getting [ 7006.027503] fglrxinfo[1395]: segfault at 4 ip 00007f20bfdfd38e sp 00007fff0c55df40 error 4 in libGL.so.1.2[7f20bfda2000+ae000] when I run fglrxinfo
[14:08] <arie_> this is the line from dmesg btw
[14:09] <arie_> nothing in Xorg.0.log