[19:12] <stgraber> heya. A friend of mine just got a new DELL Latitude E6330 with an i7-3520M (ivy bridge). Running on 12.04 with the embedded HD4000 he'd get frequent full system locks (looks like bug 1030643), easily triggered when opening chrome. I know got him on the q-lts-backport PPA and that seem to have solved it. Anything I can try and get so we can have this fixed in 12.04's kernel/X?
[19:14] <stgraber> I already asked him to try without an external monitor (instead of LVDS + VGA) and to try booting the old kernel with the new X stack
[19:16] <tjaalton> stgraber: does he have all the updates?
[19:17] <stgraber> yep, it was on a fully up to date 12.04 system. I believe he also tried the kernel currently in -proposed.
[19:19] <tjaalton> then it would need bisecting the kernel. ivb is stable for me, but it's a desktop with just one monitor and no chrome
[19:20] <tjaalton> 3.4 would probably be stable like for others
[19:26] <stgraber> tjaalton: ok. I'll suggest he first checks that it's still failing when running the current precise 3.2 with the updated X stack (it really should), then go through the kernel mainline builds to confirm that a clean 3.2 has the same issue, then go up until we find a kernel that doesn't have the bug
[19:26] <stgraber> that should then be giving a start and end point to whoever wants to actually do the bisect to figure out the exact commit
[19:27] <tjaalton> stgraber: there are a number of commits from 3.3/3.4 backported to 3.2, so the mainlines can be buggy :)
[19:28] <tjaalton> i'd suggest to start from 3.4-rc1, and if that's stable, bisect comparing with 3.3
[19:29] <tjaalton> too bad that rc1 had around 150 commits to drm/i915..
[19:30] <stgraber> ok
[19:31] <bryceh> MIR for adding s2tc to main:  https://bugs.launchpad.net/ubuntu/+source/s2tc/+bug/1053065
[19:32] <bryceh> bug for adding s2tc recommends to mesa:  https://bugs.launchpad.net/ubuntu/+source/mesa/+bug/1053029
[19:37] <bryceh> Uploaded to quantal as mesa_9.0~git20120917.7cfd42ce-0ubuntu2; should show up on the component mismatches page shortly.
[19:38] <tjaalton> cool
[19:42] <bryceh> dunno if it can be sru'd