Closed
Bug 1504233
Opened 6 years ago
Closed 6 years ago
Web content is not clipped properly
Categories
(Core :: Web Painting, defect, P2)
Core
Web Painting
Tracking
()
RESOLVED
FIXED
mozilla65
Tracking | Status | |
---|---|---|
firefox-esr60 | --- | unaffected |
firefox63 | --- | unaffected |
firefox64 | --- | unaffected |
firefox65 | + | fixed |
People
(Reporter: soeren.hentzschel, Assigned: mikokm)
References
Details
(Keywords: regression, testcase-wanted)
Attachments
(7 files)
[Tracking Requested - why for this release]:
Please see the attached screenshot, it shows a rendering bug on Bugzilla. I don't know how to reproduce because it doesn't always happen but it already happened a few times today and never before.
As you can see there is content visible in the white area where no content should be visible. After scrolling the problem disappears.
Updated•6 years ago
|
Component: CSS Parsing and Computation → Web Painting
Comment 1•6 years ago
|
||
I'm guessing this is with current Nightly, correct?
Also: do you have WebRender enabled by any chance? (look for webrender in about:support)
Updated•6 years ago
|
Flags: needinfo?(soeren.hentzschel)
Reporter | ||
Comment 2•6 years ago
|
||
Oh, sorry, I forget to include some information. Yes, it was with today's Nightly and WebRender was disabled. It was on macOS 10.11.
Maybe a regression from bug 1503046? I don't know what exactly that ticket is about but it seems to have something to do with the stacking and it landed in current Nightly.
Flags: needinfo?(soeren.hentzschel)
Reporter | ||
Comment 3•6 years ago
|
||
I attached a video showing the regression in action in the newest Nightly (2018-11-03) on macOS Mojave.
Reporter | ||
Comment 4•6 years ago
|
||
I can confirm that's a regression from bug 1503046.
21:06.78 INFO: Last good revision: 806c9d31d1c6e94ca3e23a24e51753145e1079c5
21:06.78 INFO: First bad revision: 4408e5678aa6faaaebc4a69fa8bd6cbd161c30d2
21:06.78 INFO: Pushlog:
https://hg.mozilla.org/integration/autoland/pushloghtml?fromchange=806c9d31d1c6e94ca3e23a24e51753145e1079c5&tochange=4408e5678aa6faaaebc4a69fa8bd6cbd161c30d2
Assignee | ||
Comment 5•6 years ago
|
||
This looks like a clipping bug, it might be related how the new hit test info markers do not do clip chain fusing.
Flags: needinfo?(mikokm)
Keywords: testcase-wanted
Priority: -- → P2
Summary: CSS stacking (?) regression → Web content is not clipped properly
Updated•6 years ago
|
status-firefox63:
--- → unaffected
Assignee | ||
Comment 6•6 years ago
|
||
Assignee | ||
Comment 7•6 years ago
|
||
Unfortunately I haven't been able to reproduce this bug, so this is a speculative fix. This restores the previous behavior of fusing clip chains for hit test info items.
Assignee: nobody → mikokm
Status: NEW → ASSIGNED
Comment 9•6 years ago
|
||
I can also reproduce this if I go to https://bugzilla.mozilla.org/show_bug.cgi?id=1429522#c1 and double-click on the comment text to select it.
Comment 10•6 years ago
|
||
Comment 11•6 years ago
|
||
Pushed by mikokm@gmail.com:
https://hg.mozilla.org/integration/autoland/rev/f50639806f7f
Clip hit test info items properly r=mattwoodrow
Comment 12•6 years ago
|
||
bugherder |
Status: ASSIGNED → RESOLVED
Closed: 6 years ago
Resolution: --- → FIXED
Target Milestone: --- → mozilla65
Reporter | ||
Comment 13•6 years ago
|
||
(In reply to Miko Mynttinen [:miko] from comment #7)
> Unfortunately I haven't been able to reproduce this bug, so this is a
> speculative fix.
It's not yet fixed in today's Nightly (built from https://hg.mozilla.org/mozilla-central/rev/f6df375b86987b2772067a61873ebfe3a98c353a, so it should contain the fix).
Status: RESOLVED → REOPENED
Resolution: FIXED → ---
Assignee | ||
Comment 14•6 years ago
|
||
(In reply to Sören Hentzschel from comment #13)
> (In reply to Miko Mynttinen [:miko] from comment #7)
> > Unfortunately I haven't been able to reproduce this bug, so this is a
> > speculative fix.
>
> It's not yet fixed in today's Nightly (built from
> https://hg.mozilla.org/mozilla-central/rev/
> f6df375b86987b2772067a61873ebfe3a98c353a, so it should contain the fix).
Thank you for verifying this. Would you have some steps to reproduce so I can track this down further?
Flags: needinfo?(soeren.hentzschel)
Reporter | ||
Comment 15•6 years ago
|
||
1. open a bug on Bugzilla and press the vote button
2. wait until the page has been loaded
3. move the mouse
=> now there are row visible which should be hidden below the white header
4. scroll
=> the issue is gone
5. switch to another tab and swith back again
=> the issue is there again
Flags: needinfo?(soeren.hentzschel)
Assignee | ||
Comment 16•6 years ago
|
||
I managed to reproduce this on macOS 10.14 with the STR above. I think the problem here is very similar to bug 1475971, we need to copy the hit test information into new items during display list merging.
Assignee | ||
Comment 17•6 years ago
|
||
Reduced testcase.
Updated•6 years ago
|
Assignee | ||
Comment 20•6 years ago
|
||
Assignee | ||
Comment 21•6 years ago
|
||
Depends on D12612
Comment 24•6 years ago
|
||
Pushed by mikokm@gmail.com:
https://hg.mozilla.org/integration/autoland/rev/089d689d8eb5
Part 1: Update container item ASR properly when hit test information is present r=mattwoodrow
https://hg.mozilla.org/integration/autoland/rev/bb579a94d0f6
Part 2: Add a reftest r=mattwoodrow
Comment 26•6 years ago
|
||
bugherder |
https://hg.mozilla.org/mozilla-central/rev/089d689d8eb5
https://hg.mozilla.org/mozilla-central/rev/bb579a94d0f6
Status: REOPENED → RESOLVED
Closed: 6 years ago → 6 years ago
Resolution: --- → FIXED
Updated•6 years ago
|
QA Whiteboard: [good first verify]
You need to log in
before you can comment on or make changes to this bug.
Description
•