Bug 95140
| Summary: | [chromium] WebGL canvas cleared by printing | ||
|---|---|---|---|
| Product: | WebKit | Reporter: | James Robinson <jamesr> |
| Component: | WebGL | Assignee: | Nobody <webkit-unassigned> |
| Status: | RESOLVED WONTFIX | ||
| Severity: | Normal | CC: | gman, jbauman, kbr, schenney, twiz, zmo |
| Priority: | P2 | ||
| Version: | 528+ (Nightly build) | ||
| Hardware: | Unspecified | ||
| OS: | Unspecified | ||
James Robinson
Load up a page with WebGL that doesn't repaint on every frame (for instance http://trac.webkit.org/export/126806/trunk/LayoutTests/fast/canvas/webgl/webgl-composite-modes.html) and print. The webgl canvas content is gone after printing if the preserveDrawingBuffer attribute is false. It appears the readback is destructive on the canvas.
| Attachments | ||
|---|---|---|
| Add attachment proposed patch, testcase, etc. |
Kenneth Russell
Darn it -- this keeps regressing. We need a way to write printing layout tests.
James Robinson
Maybe we could add a testRunner hook that "prints" into a null WebCanvas so at least we'd hit all of the printing logic in WebKit and verify that it didn't hose any state, even though we couldn't test on what showed up in the printed output?
Kenneth Russell
That would be an improvement -- it would catch this particular error -- but it would be better if we could also (optionally) examine the printed results as a pixel test.
Stephen Chenney
https://code.google.com/p/chromium/issues/detail?id=231368