Created attachment 22077 [details] syslog Hi! I've been testing latest git stuff (mesa, drm, video-intel, xserver-xorg, drm-intel-next w. 2.6.29) on my notebook (Latitude E5500) which features an Intel Mobile Intel GM45 Express Chipset and 4500MHD. The only error message I get in my log is "drm_mode_object_get called w/o mode_config lock" (which as I found out is not critical). Loading i915 with modeset=1 results in dark ttys, but X-Server (and compiz) works for at least a few minutes and will freeze eventually I (sysrq-keys still work). I can trigger that freeze for example by min/maximizing windows rapidly. I've attached a complete syslog and Xorg.0.log of a modeset=1 test. Please tell me if you need more info/debug output. Btw, I've found no doc of that so far: Has setting resolution of ttys at boot time already been implemented (like video=791 for framebuffers)?
Created attachment 22078 [details] Xorg.o.log
No the kernel just picks a default initial configuration w/o user input right now. We should probably add support for custom modes & configurations on the command line though.
Oh and you'll need the fbcon driver built in or loaded to see your console with the modesetting option enabled. So I think the main bug here is the hang... There have been some 2D fixes in this area recently (preventing rendering with bad state), can you try git master and see if you can still reproduce the hangs?
After adding i915 to my initram console is back with modesetting. There are problems with the correct resolution in X - but I think this is some issue with Gnome and randr. > There have been some 2D fixes in this area recently (preventing rendering with > bad state), can you try git master and see if you can still reproduce the > hangs? I'm now using the 2.6-rc of xorg in Debian/experimental and built git master of video-intel in top of that. Video-intel-2.6.1 (also in experimental) definitely freezes after some time (don't even need to stress the system), but so far git master proved quite stable. Again, I'm using kernel git and drm-intel-next. I'm closing this bug for now - it really seems fixed.
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.