https://intel-gfx-ci.01.org/tree/drm-tip/IGT_4602/fi-cfl-guc/igt@drv_selftest@live_hangcheck.html [ 387.078293] igt_reset_queue(vecs0): Failed to idle other inactive engines after device reset [ 387.078595] Dumping ftrace buffer: [ 387.078816] --------------------------------- [...] [ 387.179360] --------------------------------- [ 387.193367] Failed to switch back to kernel context; declaring wedged [ 387.193618] i915/intel_hangcheck_live_selftests: igt_reset_queue failed with error -5 [ 387.201181] Failed to switch back to kernel context; declaring wedged [ 387.239388] [drm:intel_guc_send_mmio [i915]] *ERROR* MMIO: GuC action 0x501 failed with error -110 0x501
That's not icl, that's just the guc being the guc -- there's a bug for the guc reset failure somewhere.
(In reply to Chris Wilson from comment #1) > That's not icl, that's just the guc being the guc -- there's a bug for the > guc reset failure somewhere. Yeah, that's why I put the issue on the GUC feature, not GPU hang ;)
But tagged icl. It's just that there is a bug in execlists reset that I was waiting for to be filed; and thought the guc bug was known...
(In reply to Chris Wilson from comment #3) > But tagged icl. It's just that there is a bug in execlists reset that I was > waiting for to be filed; and thought the guc bug was known... I'm done filing bugs for ICL (none in the backlog), so sorry if you did not get the bug you wanted :D
So, this is tagged as ICL, but the test link is for CFL - I'm assuming this should actually be tagged as CFL?
(In reply to James Ausmus from comment #5) > So, this is tagged as ICL, but the test link is for CFL - I'm assuming this > should actually be tagged as CFL? Oh my, thanks for pointing it out! That's what happens when you get tired I guess... Thanks again :)
*** This bug has been marked as a duplicate of bug 106685 ***
Closed as duplicate.
Re-opening to get rid of the duplicate given that this issue is fixed since we disabled GuC command submission, but the other bug is not fixed.
Now we can properly close it and archive it in CI Bug Log.
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.