The screen corrupt and stayed still after suspend. You can move cursor around, but the server does not response. I can access to the machine through ssh, nothing show up in the log and dmesg. The cpu usage is normal too. I can't restart X, and have to restart the whole machine. After doing some bitsect with git, I found the bad commit is commit 8e942b70cb9a784b3f1311affd6fc74c4bcf68bb Author: Carl Worth <cworth@cworth.org> Date: Thu May 21 13:12:52 2009 -0700 Revert "Rely on BO pixmaps being present in acceleration paths." This reverts commit 4653a7db622ad54a3182d93c81331765d930db34. Eric was getting a little too ambitious about our brave, new world. We do still want the driver to work with old, non-GEM kernels after all. After revert this one, everything is fine. kernel: 2.6.30-rc7-git4 xserver: 1.6.1.901 xf86-video-intel: 8e942b70cb9a784b3f1311affd6fc74c4bcf68bb
Some updates on this bug. It seams that with kernel 2.6.29.3 everything works fine. Besides after revert 8e942b70cb9a784b3f1311affd6fc74c4bcf68bb I still hang for one time. So the bug might be lurk in somewhere else. I'll retest with latest kernel and report back.
OK, after retest with kernel 2.6.30-r8-git1, the hang still there. Maybe it is a regression of the kernel drm driver. I'll try to bisect the kernel instead.
Created attachment 26480 [details] intel_gpu_top
Created attachment 26481 [details] intel_gpu_dump This is what intel_gpu_dump shows after stripe a lot of MI_NOOP at the end of file.
It seams the same issue has also been reported as http://bugs.freedesktop.org/show_bug.cgi?id=22039
Marking as a DUP of an older 945GM hang... this is a high priority bug. *** This bug has been marked as a duplicate of bug 20520 ***
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.