[Bug 111732] [CI][SHARDS] igt at perf@gen8-unprivileged-single-ctx-counters|igt at gem_softpin@softpin| igt at gem_pipe_control_store_loop@reused-buffer - timeout - Received signal SIGQUIT.

[bugzilla-daemon at freedesktop.org]

https://bugs.freedesktop.org/show_bug.cgi?id=111732

--- Comment #5 from Chris Wilson <chris at chris-wilson.co.uk> ---
Showing all locks held in the system:
<4> [555.972811] 1 lock held by khungtaskd/34:
<4> [555.972830]  #0: ffffffff82247a60 (rcu_read_lock){....}, at:
debug_show_all_locks+0xe/0x1a0
<4> [555.972900] 3 locks held by kworker/3:2/160:
<4> [555.972919]  #0: ffff88817a08f528 ((wq_completion)events){+.+.}, at:
process_one_work+0x1de/0x620
<4> [555.972972]  #1: ffffc90000f17e78
((work_completion)(&(&engine->heartbeat.work)->work)){+.+.}, at:
process_one_work+0x1de/0x620
<4> [555.973024]  #2: ffff888172908e70 (&kernel#2){+.+.}, at:
heartbeat+0x3f7/0x5f0 [i915]
<4> [555.973427] 4 locks held by kworker/3:3/392:
<4> [555.973445]  #0: ffff88817a08f528 ((wq_completion)events){+.+.}, at:
process_one_work+0x1de/0x620
<4> [555.973497]  #1: ffffc9000034be78
((work_completion)(&(&gt->requests.retire_work)->work)){+.+.}, at:
process_one_work+0x1de/0x620
<4> [555.973547]  #2: ffff88817993ac70 (&timeline->mutex){+.+.}, at:
intel_gt_retire_requests_timeout+0xe2/0x540 [i915]
<4> [555.973857]  #3: ffff88816e17d928 (&ce->pin_mutex/1){+.+.}, at:
intel_context_unpin+0x54/0x100 [i915]
<4> [555.974242] 4 locks held by in:imklog/403:
<4> [555.974262] 4 locks held by rs:main Q:Reg/404:
<4> [555.974288] 4 locks held by dmesg/963:
<4> [555.974308] 3 locks held by kworker/1:1H/1084:
<4> [555.974324]  #0: ffff888179630f28 ((wq_completion)kblockd){+.+.}, at:
process_one_work+0x1de/0x620
<4> [555.974368]  #1: ffffc9000070be78
((work_completion)(&(&hctx->run_work)->work)){+.+.}, at:
process_one_work+0x1de/0x620
<4> [555.974415]  #2: ffff888176da6e58 (hctx->srcu){....}, at:
hctx_lock+0x5b/0xe0
<4> [555.974480] 5 locks held by perf/1180:
<4> [555.974499]  #0: ffff88815810ba80 (&perf->lock){+.+.}, at:
i915_perf_open_ioctl+0x242/0x6f0 [i915]
<4> [555.974873]  #1: ffff888170b93478 (&ctx->engines_mutex){+.+.}, at:
lrc_configure_all_contexts+0x26b/0x880 [i915]
<4> [555.975330]  #2: ffff88816e17d928 (&ce->pin_mutex){+.+.}, at:
lrc_configure_all_contexts+0x2d7/0x880 [i915]
<4> [555.975693]  #3: ffff888172908e70 (&kernel#2){+.+.}, at:
i915_request_create+0x16/0x1b0 [i915]
<4> [555.976136]  #4: ffff88817993ac70 (&timeline->mutex/1){+.+.}, at:
intel_context_prepare_remote_request+0x3f/0x180 [i915]
<4> [555.976485] 


Looks like lock inversion between the loop to update all contexts and timeline
retiring.

-- 
You are receiving this mail because:
You are on the CC list for the bug.
You are the QA Contact for the bug.
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.freedesktop.org/archives/intel-gfx-bugs/attachments/20191116/98dccde2/attachment-0001.html>