[Bug 71391] GM45: Regression due to incorrect graphics base of stolen memory.

bugzilla-daemon at bugzilla.kernel.org bugzilla-daemon at bugzilla.kernel.org
Mon Mar 3 13:04:45 PST 2014


https://bugzilla.kernel.org/show_bug.cgi?id=71391

--- Comment #16 from Chris Wilson <chris at chris-wilson.co.uk> ---
(In reply to Pär Lindfors from comment #11)
> Chris: Don't know what tree your patch is against, it did not
> apply. So I added that same return 0 code manually to that
> function on d8efcf38b13df3e9e889cf7cc214cb85dc53600c (Linus
> master from 2014-02-28). This simple patch makes the problem go
> away.

Thanks. That just confirms using the stolen memory is conflicting with
something else.

> By "grab a video" do you mean with a camera pointed at the
> laptop? If so, then yes I can do that.

Yes. What we will be looking for here is whether there is any pattern to the
corruption. For example, if it is regular, it may be a scanout access
violation. If it is located to one area, we may only conflict over a few pages
(and we need to reserve a fraction of the stolen area for whomever). Or it may
be dirty cachelines, in which case we have angered the GPU gods.

-- 
You are receiving this mail because:
You are on the CC list for the bug.


More information about the intel-gfx-bugs mailing list