bisected to offending CL https://chromium-review.googlesource.com/c/chromium/src/+/900484. The part where left_rotated_view_ is changed from parent left_highlight_view_ to the widget content view is where this is introduced.

Issue 824852 has been merged into this issue.

+marcheu

These do not show up on dru or kevin, but they do on eve and caroline. Suspecting to be intel driver issue.

Deleted: dru.png 427 KB

	dru.png 427 KB View Download

Deleted: eve.png 592 KB

	eve.png 592 KB View Download

the artifact is a light box around the left "Drag here to use split screen" text on the eve.png

+reed

we have a workaround for splitview by changing [1] to use drawRect and drawCircle instead of drawRoundedRect, so suspect a skia issue.

reed@ - would you happen to have any ideas?

[1] https://cs.chromium.org/chromium/src/ash/wm/overview/rounded_rect_view.cc?rcl=9879f87f2d5c1eda59dbf3027ccd3417023d21a9&l=17

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/a9ce932269f8cf73d718047bedba984b770d23a6

commit a9ce932269f8cf73d718047bedba984b770d23a6
Author: Sammie Quon <sammiequon@google.com>
Date: Sat Mar 31 01:39:26 2018

splitview: Workaround for artifacts on split view indicators.

Some artifacts show up on intel devices on the splitview indicators
highlights, which are rounded_rect_views. Use addCircle and addRect
as a temporary workaround.

Test: manual
Bug:  824564 
Change-Id: Ic5bee0b6eaf412ea5fa569afefcde66843af4e4c
Reviewed-on: https://chromium-review.googlesource.com/986974
Reviewed-by: Mitsuru Oshima <oshima@chromium.org>
Commit-Queue: Sammie Quon <sammiequon@chromium.org>
Cr-Commit-Position: refs/heads/master@{#547354}
[modify] https://crrev.com/a9ce932269f8cf73d718047bedba984b770d23a6/ash/wm/overview/rounded_rect_view.cc

Note that this is probably a skia GPU rast bug. We only enable skia GPU rast on intel devices for now.

Over to reed@ for skia

I'm happy to take a look at this but no idea where to begin on debugging this. Can I build a version of ChromeOS for linux that would reproduce this?

re #11 - I tried undoing the workaround in #7 and it seems to be fixed, at least for our case.

Re #12, Great. Should we revert the workaround and call it a day or is there something else to do here?

RE #13, I'll double check on one more device and then revert.

This is reverted in https://chromium-review.googlesource.com/c/chromium/src/+/1097472, referenced the wrong bug in the cl.