Bug 83190 - Shadows clipped on position:fixed composited layers
Summary: Shadows clipped on position:fixed composited layers
Status: RESOLVED DUPLICATE of bug 63331
Alias: None
Product: WebKit
Classification: Unclassified
Component: Layout and Rendering (show other bugs)
Version: 528+ (Nightly build)
Hardware: Unspecified Unspecified
: P2 Normal
Assignee: Dean Jackson
URL:
Keywords: InRadar
: 83381 (view as bug list)
Depends on:
Blocks:
 
Reported: 2012-04-04 12:21 PDT by Vangelis Kokkevis
Modified: 2012-04-20 11:58 PDT (History)
5 users (show)

See Also:


Attachments
Reduced test case (643 bytes, text/html)
2012-04-20 00:41 PDT, Dean Jackson
no flags Details
Updated reduction (705 bytes, text/html)
2012-04-20 09:41 PDT, Dean Jackson
no flags Details

Note You need to log in before you can comment on or make changes to this bug.
Description Vangelis Kokkevis 2012-04-04 12:21:36 PDT
This bug was originally reported in:
http://code.google.com/p/chromium/issues/detail?id=121102

Composited layers seem to clip out shadow region corresponding to a -webkit-box-shadow property. Steps to reproduce:

(WebKit in 10.7 with accelerated canvas or Chrome)

1. Open http://www.tsetmc.com/Loader.aspx?ParTree=151311&i=23441366113375722
2. Click 's' to open the popup div with the box shadow 
3. Resize the main window until the popup overlaps the graph at the top left

At that point, the popup gets promoted to a composited layer and loses it shadow (and a chunk of its border). The composited layer borders reveal that the layer created is really too small.

An alternative way to trigger the bug on the same page is to simply add a -webkit-transform: translateZ(0) on the popup div's content.
Comment 1 Simon Fraser (smfr) 2012-04-04 13:04:03 PDT
When did this regress? r111456?
Comment 2 Vangelis Kokkevis 2012-04-04 13:15:33 PDT
(In reply to comment #1)
> When did this regress? r111456?

I'm not sure it's a regression, and if it is, it's probably not very recent. I see the same behavior in (official) Safari 5.1.3 (7534.53.10) .
Comment 3 Simon Fraser (smfr) 2012-04-06 17:11:12 PDT
*** Bug 83381 has been marked as a duplicate of this bug. ***
Comment 4 Simon Fraser (smfr) 2012-04-06 17:12:16 PDT
This needs to be looked soon.
Comment 5 Radar WebKit Bug Importer 2012-04-06 17:12:37 PDT
<rdar://problem/11204762>
Comment 6 Dean Jackson 2012-04-20 00:41:06 PDT
Created attachment 138056 [details]
Reduced test case

Here's a pretty simple reduction. Edit the marked point in the file to toggle the correct and incorrect results.
Comment 7 Simon Fraser (smfr) 2012-04-20 08:04:32 PDT
We're probably using the wrong clip rect for position:fixed things. Does it only repro with position:fixed?
Comment 8 Dean Jackson 2012-04-20 09:41:22 PDT
Created attachment 138107 [details]
Updated reduction
Comment 9 Simon Fraser (smfr) 2012-04-20 09:52:06 PDT
I think this is a dup of bug 63331
Comment 10 Dean Jackson 2012-04-20 11:58:29 PDT

*** This bug has been marked as a duplicate of bug 63331 ***