[Bug 95427] [IGT] gem_userptr_blits at mlocked-* fail or process killed

[bugzilla-daemon at freedesktop.org]

https://bugs.freedesktop.org/show_bug.cgi?id=95427

--- Comment #32 from Abdiel Janulgue <abdiel.janulgue at linux.intel.com> ---
Had a stab at this problem. Crashes actually doesn't come from the i915 driver
but rather core-mm triggering oom killer when the IGT process mlocks the whole
available system memory.

In other words the kernel killed it for mlocking too much memory causing the
test failure.

Now I understand that the purpose of test is to test i915's eviction mechanism.
by forcing the creation BOs through a narrow memory window. I tried various
memory locking size combos and it looks like we are still able to activate the
eviction mechanism if we mlock the free system memory subtracted by 128MB.

Any thoughts Chris?

-- 
You are receiving this mail because:
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/20180604/2529d940/attachment.html>