Closed
Bug 980035
Opened 11 years ago
Closed 11 years ago
Support 16bit volatile buffer backed decoded images
Categories
(Core :: Graphics: ImageLib, defect)
Tracking
()
| Tracking | Status | |
|---|---|---|
| b2g-v1.3T | --- | fixed |
People
(Reporter: mwu, Assigned: mwu)
References
Details
Attachments
(2 files)
|
2.79 KB,
patch
|
seth
:
review+
|
Details | Diff | Splinter Review |
|
2.78 KB,
patch
|
Details | Diff | Splinter Review |
Right now, the volatile buffer is thrown away if it is "optimized". This avoids that was on android/gonk when we're using 16bpp images.
Attachment #8386311 -
Flags: review?(seth)
| Assignee | ||
Updated•11 years ago
|
blocking-b2g: --- → 1.3T?
Updated•11 years ago
|
blocking-b2g: 1.3T? → 1.3T+
Comment 1•11 years ago
|
||
Comment on attachment 8386311 [details] [diff] [review]
16bit-direct-vbuf
Review of attachment 8386311 [details] [diff] [review]:
-----------------------------------------------------------------
Looks good, Michael. Thanks.
Are there tests for this code?
Attachment #8386311 -
Flags: review?(seth) → review+
| Assignee | ||
Comment 2•11 years ago
|
||
(In reply to Seth Fowler [:seth] from comment #1)
> Are there tests for this code?
For the optimization/deoptimization path, I think any test which goes through the scaled image path in RasterImage.cpp should cover this. (assuming they exist - maybe a bad assumption)
For recovery from lost images, I was thinking of adding a code to purge every volatile buffer so those paths can be automatically tested on every platform. Right now I have platform specific test programs that I manually run to test the different redecoding paths.
| Assignee | ||
Comment 3•11 years ago
|
||
Status: NEW → RESOLVED
Closed: 11 years ago
Resolution: --- → FIXED
Target Milestone: --- → mozilla30
| Assignee | ||
Comment 5•11 years ago
|
||
Here's a version for tarako.
Comment 6•11 years ago
|
||
status-b2g-v1.3T:
--- → fixed
You need to log in
before you can comment on or make changes to this bug.
Description
•