I frequently play Dota 2 ( http://dota2.com/ ) on my machine. Recently Dota 2 had a significant update, which included some graphics improvements. Since then, I experience frequent freezes ingame. Every 5 to 10 minutes, the game will freeze up, for about 5 seconds, and then most of the time continue on as before. Looking at my logs, I found messages like this, associated with the freezes: kernel: [ 976.678551] [drm] GPU HANG: ecode 8:0:0x85dfffdb, in GLRenderThread [5482], reason: Hang on render ring, action: reset kernel: [ 976.678552] [drm] GPU hangs can indicate a bug anywhere in the entire gfx stack, including userspace. kernel: [ 976.678553] [drm] Please file a _new_ bug report on bugs.freedesktop.org against DRI -> DRM/Intel kernel: [ 976.678553] [drm] drm/i915 developers can then reassign to the right component if it's not a kernel issue. kernel: [ 976.678553] [drm] The gpu crash dump is required to analyze gpu hangs, so please always attach it. kernel: [ 976.678553] [drm] GPU crash dump saved to /sys/class/drm/card0/error kernel: [ 976.693084] drm/i915: Resetting chip after gpu hang kernel: [ 984.688951] drm/i915: Resetting chip after gpu hang kernel: [ 994.739240] kworker/dying (4978) used greatest stack depth: 12344 bytes left kernel: [ 996.707763] drm/i915: Resetting chip after gpu hang kernel: [ 1010.655648] drm/i915: Resetting chip after gpu hang kernel: [ 1022.686415] drm/i915: Resetting chip after gpu hang I was then at kernel 4.4.19. I've since tried 4.4.39, 4.9.0 and currently I'm running drm-tip 8c3608fc86d03a4ed5082734f751c2dc85a13ecb. Neither kernel upgrade helped. I have a $ cat /proc/cpuinfo | grep name model name : Intel(R) Core(TM) i5-5675C CPU @ 3.10GHz so the GPU is a Iris 6200. I've upgraded my userspace driver to 169c74fa6c2cd9c28dd7bfacd9639cd245b8c8a8 that also didn't help. Motherboard: Z97M-G43 Display connector: DVI Distribution: Gentoo Reverting to mesa 11.2.2 resolves the issue, I've tried several versions after that, including a recent commit, ce7a045feeef8cad155f1c9aa07f166e146e3d00.
Created attachment 129240 [details] dmesg from start to crash
Created attachment 129241 [details] GPU error from /sys/class/drm/card0/error
I'm not sure what standard practice is, the api trace is to big to upload here, so I put it on google drive, this link should allow it to be downloaded: https://drive.google.com/open?id=0B48wafqQit54YVhEcWozTUdKWjg
The earliest version that repros the issue, as far as I'm aware, is 12.0.1.
Dylan, can you investigate?
Issue it not reproducible. OS: Ubuntu 16.04 Mesa 18.1.0-devel (git-f746aa3) CPU: Intel® Core™ i5-4310M CPU @ 2.70GHz × 4 GPU: Intel® Haswell Mobile But apitrace is crashing and the is following API error: Mesa: User error: GL_INVALID_VALUE in glBindBufferRange(offset misaligned 369968/32) Mesa: User error: GL_INVALID_VALUE in glBindBufferRange(offset misaligned 370192/32) apitrace: warning: caught signal 11 29828287: error: caught an unhandled exception /home/vadym/mesa/apitrace/build/glretrace+0x21802c /lib/x86_64-linux-gnu/libpthread.so.0+0x1138f /home/vadym/mesa/apitrace/build/glretrace+0x1956db /home/vadym/mesa/apitrace/build/glretrace+0x17463 /home/vadym/mesa/apitrace/build/glretrace+0x17b1b /usr/lib/x86_64-linux-gnu/libstdc++.so.6+0xb8c7f /lib/x86_64-linux-gnu/libpthread.so.0+0x76b9 /lib/x86_64-linux-gnu/libc.so.6: clone+0x6c ? apitrace: info: taking default action for signal 11
This is for a really old version of mesa, if you can reproduce please reopen.
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.