Bug 134916 – Clean up image subsampling code, make it less iOS-specific

• Status: RESOLVED
• Resolution: FIXED
• Component: New Bugs

[Simon Fraser (smfr)]

Clean up image subsampling code, make it less iOS-specific

[Simon Fraser (smfr)]

Created attachment 234904 [details]
Patch

[WebKit Commit Bot]

Attachment 234904 [details] did not pass style-queue:


ERROR: Source/WebCore/platform/graphics/BitmapImage.h:323:  Please declare enum bitfields as unsigned integral types.  [runtime/enum_bitfields] [5]
Total errors found: 1 in 20 files


If any of these errors are false positives, please file a bug against check-webkit-style.

[Simon Fraser (smfr)]

I tried to make some ref tests, and failed (bug 134918, and slight ref mismatches).

[Simon Fraser (smfr)]

Created attachment 234906 [details]
Patch

[WebKit Commit Bot]

Attachment 234906 [details] did not pass style-queue:


ERROR: Source/WebCore/platform/graphics/BitmapImage.h:323:  Please declare enum bitfields as unsigned integral types.  [runtime/enum_bitfields] [5]
Total errors found: 1 in 20 files


If any of these errors are false positives, please file a bug against check-webkit-style.

[Build Bot]

Comment on attachment 234906 [details]
Patch

Attachment 234906 [details] did not pass mac-ews (mac):
Output: http://webkit-queues.appspot.com/results/6534446131970048

New failing tests:
media/video-ended-event-negative-playback.html

[Build Bot]

Created attachment 234909 [details]
Archive of layout-test-results from webkit-ews-07 for mac-mountainlion

The attached test failures were seen while running run-webkit-tests on the mac-ews.
Bot: webkit-ews-07  Port: mac-mountainlion  Platform: Mac OS X 10.8.5

[Dean Jackson]

Comment on attachment 234906 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=234906&action=review

> Source/WebCore/platform/graphics/BitmapImage.cpp:390
> +    // We may have cached a frame with a higher subsampling level, in which case we need to
> +    // re-decode with a lower level.
> +    if (index < m_frames.size() && m_frames[index].m_frame && subsamplingLevel < m_frames[index].m_subsamplingLevel) {
>          // If the image is already cached, but at too small a size, re-decode a larger version.

The wording around higher and lower, smaller and larger is often confusing. Maybe the comment should be "higher subsampling level (smaller image)"?

And, the second comment here mostly duplicates the comment you added.

> Source/WebCore/platform/graphics/BitmapImage.cpp:395
> +        WTFLogAlways("BitmapImage %p frameAtIndex recaching, m_decodedSize now %u (size change %d)", this, m_decodedSize, sizeChange);

Always? Why not add a new logging channel?

[Simon Fraser (smfr)]

http://trac.webkit.org/changeset/171957

[WebKit Commit Bot]

Re-opened since this is blocked by bug 135538

[Carlos Alberto Lopez Perez]

(In reply to comment #9)
> http://trac.webkit.org/changeset/171957

This has caused lot of regressions on EFL and GTK.

Reported on bug 135537 and bug 135532

[Simon Fraser (smfr)]

That's why I rolled it out.

[Simon Fraser (smfr)]

Relanded with fixes:
http://trac.webkit.org/changeset/172348