[Bug 103707] [CI] igt at drv_selftest@live_coherency - dmesg-fail

[bugzilla-daemon at freedesktop.org]

https://bugs.freedesktop.org/show_bug.cgi?id=103707

--- Comment #1 from Chris Wilson <chris at chris-wilson.co.uk> ---
That's actually:

<7>[ 4062.736709] BUG: MAX_LOCKDEP_CHAINS too low!
<7>[ 4062.736710] turning off the locking correctness validator.

It's telling us that we've overflowed the static array lockdep uses for
tracking unique lockclasses -- we have too many locks! lockclasses are not
recycled when the module is unload, so every time we run a drv_selftest subtest
we load the module twice (the intention was to run drv_selftest in one batch so
the extra reload only happened once for the entire set). We can drop the
reload, so long as we teach igt to load i915.ko when required:
https://patchwork.freedesktop.org/series/33703/ but this still means that there
is an upper limit to the number of kselftests we run in this manner. Annoying.

I think the ultimate fix here is recycling the lockclasses on module unload.

-- 
You are receiving this mail because:
You are the assignee for the bug.
You are on the CC list for the bug.
You are the QA Contact for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.freedesktop.org/archives/intel-gfx-bugs/attachments/20171113/fcc340d4/attachment.html>