New issue
Advanced search Search tips

Issue 919954 link

Starred by 1 user
Status: Assigned
Owner:
chrishtr@chromium.org
Cc:
vmp...@chromium.org
dtapu...@chromium.org
chrishtr@chromium.org
Components:
EstimatedDays: ----
NextAction: ----
OS: Mac
Pri: 3
Type: Bug



Sign in to add a comment

Image in a scrollable container fails to rerender after display:none is reset

Project Member Reported by dvoytenko@google.com, Jan 8 
Chrome Version: 71.0.3578.98 (Official Build) (64-bit)
OS: macOS High Siera 10.13.6

What steps will reproduce the problem?
(1) Open https://output.jsbin.com/hewokev/quiet. Bunch of images will be displayed. Each with the `display` defined as "block" in CSS.
(2) Click on "Toggle display". The images' `style.display` is set to "none". Observe that all images are now hidden.
(3) Click on "Toggle display" again. The images are now have their `style.display` reset and thus styled as `display: block`. Observe that images remain unpainted.
(4) Try to scroll the document or switch tabs - the images are repainted.

What is the expected result?

Images should be repainted once `display: none` is removed.

What happens instead?

Images remain unpainted. Important caveat: the container where images are hosted is `overflow: auto`. If it's instead `overflow: hidden` - images are repainted as a expected.

Comment 1 by dtapu...@chromium.org, Jan 8

Cc: vmp...@chromium.org
Components: -Blink>Compositing -Blink Blink>Paint>Invalidation
Labels: OS-Mac
Can only repro on Mac.

Happens on Stable but not on Canary.

vmpstr@ do you have any idea what change this was fixed by? I can run a bisect if you don't have an idea.

Comment 2 by dtapu...@chromium.org, Jan 8

Cc: chrishtr@chromium.org
Couldn't bisect on per revision build. But was able to bisect with release builds.

https://chromium.googlesource.com/chromium/src/+log/73.0.3652.0..73.0.3653.0?pretty=full

Only change I expect fixed it perhaps is: https://chromium.googlesource.com/chromium/src/+/395a93a5ea18f0890a237d5e687a80b69f39a473

+Chris perhaps he has some thoughts. I just want to be certain this is actually addressed and not just a race condition (since it always seemed to work with per revision bisect).

Comment 3 by schenney@chromium.org, Jan 9

Owner: chrishtr@chromium.org
Status: Assigned (was: Untriaged)
►

Sign in to add a comment