Created attachment 70932 [details] i915_error_state_decoded While either cycling through internal/external TFTs or switching back from the console, GNOME3 crashed in the sense that the Shell and Mutter went away and the keyboard no longer worked (probably none of the windows had focus). I did not see the Shell and Mutter disappear so it happened either while I was looking at the console or when switching. I was still able to use the pointer and e.g. switch tabs in the gnome-terminal, as well as open a new terminal from the menu, but could not switch focus between them by clicking. This is from dmesg: [54282.564118] [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck timer elapsed... GPU hung [54282.564134] [drm] capturing error event; look for more information in /debug/dri/0/i915_error_state 54282.570984] [drm:i915_wait_request] *ERROR* i915_wait_request returns -11 (awaiting 116142 at 116139, next 116143) [54283.072037] [drm:i915_reset] *ERROR* Failed to reset chip. The batchbuffer was: batchbuffer (render ring) at 0x085bb000: 0x085bb000: 0x09000000: MI_LOAD_SCAN_LINES_INCL 0x085bb004: 0x00000383: dword 1 0x085bb008: 0x09000000: MI_LOAD_SCAN_LINES_INCL 0x085bb00c: 0x00000383: dword 1 0x085bb010: 0x01820000: MI_WAIT_FOR_EVENT 0x085bb014: HEAD 0x02000006: MI_FLUSH 0x085bb018: 0x54f08806: XY_SRC_COPY_BLT (rgb enabled, alpha enabled, src tile 1, dst tile 1) 0x085bb01c: 0x03cc0600: format 8888, pitch 1536, rop 0xcc, clipping disabled, 0x085bb020: 0x00000000: dst (0,0) 0x085bb024: 0x038405a0: dst (1440,900) 0x085bb028: 0x078c4000: dst offset 0x078c4000 0x085bb02c: 0x00000000: src (0,0) 0x085bb030: 0x00000600: src pitch 1536 0x085bb034: 0x068a8000: src offset 0x068a8000 0x085bb038: 0x05000000: MI_BATCH_BUFFER_END (the full i915_error_state is attached). Please note that shortly before that, the internal TFT of the ThinkPad had stopped working after a suspend to memory resume - backlight was still on, but it was just blank. I filed this as Debug Bug #694946: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=694946 The other times I had this problem (happens 10-30% of the time) I rebooted rather quickly in order to get the notebook back to work. lspci Output: 00:02.0 VGA compatible controller [0300]: Intel Corporation Mobile 4 Series Chipset Integrated Graphics Controller [8086:2a42] (rev 07)
Death in MI_WAIT - retesting on the very latest versions of ddx and kernel (preferrably drm-intel-next or 3.7) would be interesting - we've fixed too many bugs in this area recently. Can you readily reproduce this?
Thanks for the reply > Can you readily reproduce this? Happened only this time so far. The TFT failures are somewhat reproducable and I will see whether it happens afterwards next time as well - but I doubt it is readily reproducable. I switched around TFTs and VTs quite a bit in order to diagnoe the other problem. I'm running the 3.6.8 kernel from experimental now and might upgrade the Intel driver from wheezy's 2.19 to experimental's 2.20.5, but this a company notebook I need for my day-job, so need it working primarily, sorry.
The bug fixes are available in stable releases upstream.
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.