Bug 146561
| Summary: | REGRESSION (r139294): Images loaded via -webkit-mask-image now undergo same-origin checks | ||
|---|---|---|---|
| Product: | WebKit | Reporter: | Simon Fraser (smfr) <simon.fraser> |
| Component: | CSS | Assignee: | Nobody <webkit-unassigned> |
| Status: | RESOLVED FIXED | ||
| Severity: | Normal | CC: | ap, krit, simon.fraser |
| Priority: | P2 | ||
| Version: | 528+ (Nightly build) | ||
| Hardware: | Unspecified | ||
| OS: | Unspecified | ||
| See Also: | https://bugs.webkit.org/show_bug.cgi?id=139294 | ||
Simon Fraser (smfr)
After r139294, we're applying the same-origin policy to images loaded in -webkit-masks. Since this doesn't happen for CSS images or <img>, I don't think this is a progression.
| Attachments | ||
|---|---|---|
| Add attachment proposed patch, testcase, etc. |
Alexey Proskuryakov
> After r139294
I think that you meant bug 139294.
It's not obvious to me whether this is right or wrong. Does -webkit-mask ever affect content that would cause tainting if loaded cross-origin?
Dirk Schulze
(In reply to comment #1)
> > After r139294
>
> I think that you meant bug 139294.
>
> It's not obvious to me whether this is right or wrong. Does -webkit-mask
> ever affect content that would cause tainting if loaded cross-origin?
If webkit-mask does not load an image but references a mask element, we need a cross-origin check. However, at the time we had checked if the normal image loading works properly. If even images are cross-origin checked, then this might be a regression introduced at a later point.
Simon Fraser (smfr)
No longer an issue since the code was rolled out.