[Bug 105581] [CI] igt at kms_plane_lowres@pipe-[a|b]-tiling-[none|x] - fail - Test assertion failure function igt_assert_plane_visible - Failed assertion: visible == visibility

[bugzilla-daemon at freedesktop.org]

https://bugs.freedesktop.org/show_bug.cgi?id=105581

Matt Roper <matthew.d.roper at intel.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |matthew.d.roper at intel.com

--- Comment #4 from Matt Roper <matthew.d.roper at intel.com> ---
This test puts a canary plane in the bottom left corner of the screen, then
switches to a smaller mode in which the plane is now fully offscreen; if the
driver still reports the plane as on-screen after the modeset then the test
fails.

I believe the failures here arise from test's fallback logic when it can't find
a suitable mode on the connector's modelist (e.g., because it's something like
an eDP panel that only has a single mode listed) --- the test uses a fixed
1024x768 mode in that case.  However the test doesn't actually check to ensure
1024x768 is truly small enough to ensure the plane is offscreen.  If the panel
was originally running with something like 1366x768 (which some of our CI
machines are), then a plane in the bottom left corner will still be visible
after switching to 1024x768.

I'm testing out a possible fix on trybot; will update tomorrow if it actually
solves the problem here.

-- 
You are receiving this mail because:
You are the assignee for the bug.
You are the QA Contact for the bug.
You are on the CC list for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.freedesktop.org/archives/intel-gfx-bugs/attachments/20190819/f973f209/attachment-0001.html>