Propagate the decoder frame premul info to SkBitmap

Propagate the decoder frame premul info to SkBitmap

Currently ImageFrame always sets up the resulting SkBitmap as
kPremul_SkAlphaType (or kOpaque_SkAlphaType in some cases), even
when AlphaOption == AlphaNotPremultiplied.

Update to set the correct bitmap alpha type.

R=noel@chromium.org,reed@google.com,xidachen@chromium.org
BUG=597035
Committed: https://crrev.com/18a861d65e82a16f8cfbc7d3f160cfb3746523e2
Cr-Commit-Position: refs/heads/master@{#382680}

[f(malita), 2016-03-21 21:42:28]

Description was changed from

==========
Propagate decoder frame premul info to SkBitmap

Currently ImageFrame always sets up the resulting SkBitmap as
kPremul_SkAlphaType (or kOpaque_SkAlphaType in some cases), even
when AlphaOption == AlphaNotPremultiplied.

Update to set the correct bitmap alpha type.

R=noel@chromium.org,reed@google.com,xidachen@chromium.org
==========

to

==========
Propagate the decoder frame premul info to SkBitmap

Currently ImageFrame always sets up the resulting SkBitmap as
kPremul_SkAlphaType (or kOpaque_SkAlphaType in some cases), even
when AlphaOption == AlphaNotPremultiplied.

Update to set the correct bitmap alpha type.

R=noel@chromium.org,reed@google.com,xidachen@chromium.org
==========

[Noel Gordon, 2016-03-21 22:37:16]

Thanks Florin, minor nits only ...

https://codereview.chromium.org/1820733004/diff/1/third_party/WebKit/Source/p...
File third_party/WebKit/Source/platform/image-decoders/ImageFrame.cpp (left):

https://codereview.chromium.org/1820733004/diff/1/third_party/WebKit/Source/p...
third_party/WebKit/Source/platform/image-decoders/ImageFrame.cpp:102: if
(!m_bitmap.tryAllocPixels(m_allocator, 0))
Maybe don't call computeAlphaType() here?  Instead use m_premultiplyAlpha ?
kPremul_SkAlphaType : kUnpremul_SkAlphaType;

https://codereview.chromium.org/1820733004/diff/1/third_party/WebKit/Source/p...
File
third_party/WebKit/Source/platform/image-decoders/gif/GIFImageDecoderTest.cpp
(right):

https://codereview.chromium.org/1820733004/diff/1/third_party/WebKit/Source/p...
third_party/WebKit/Source/platform/image-decoders/gif/GIFImageDecoderTest.cpp:524:
// Empirically chosen
// Empirically chosen truncation size:

https://codereview.chromium.org/1820733004/diff/1/third_party/WebKit/Source/p...
third_party/WebKit/Source/platform/image-decoders/gif/GIFImageDecoderTest.cpp:527:
static size_t kTruncateSize = 800;
static -> const ?

https://codereview.chromium.org/1820733004/diff/1/third_party/WebKit/Source/p...
third_party/WebKit/Source/platform/image-decoders/gif/GIFImageDecoderTest.cpp:540:
// Partial frame => alpha type reflects the requested format.
"Partial frame => alpha" -> "Partially decoded frame => the frame alpha type is
unknown and should reflect ..."

https://codereview.chromium.org/1820733004/diff/1/third_party/WebKit/Source/p...
third_party/WebKit/Source/platform/image-decoders/gif/GIFImageDecoderTest.cpp:552:
// Full frame => alpha type is known to be opaque
"Full frame =>" -> "Fully decoded frame => the frame alpha type is known
(opaque)."

https://codereview.chromium.org/1820733004/diff/1/third_party/WebKit/Source/p...
third_party/WebKit/Source/platform/image-decoders/gif/GIFImageDecoderTest.cpp:553:
premulDecoder->setData(fullData.get(), false);
false -> true

https://codereview.chromium.org/1820733004/diff/1/third_party/WebKit/Source/p...
third_party/WebKit/Source/platform/image-decoders/gif/GIFImageDecoderTest.cpp:555:
unpremulDecoder->setData(fullData.get(), false);
false -> true

[f(malita), 2016-03-21 22:46:24]

https://codereview.chromium.org/1820733004/diff/1/third_party/WebKit/Source/p...
File third_party/WebKit/Source/platform/image-decoders/ImageFrame.cpp (left):

https://codereview.chromium.org/1820733004/diff/1/third_party/WebKit/Source/p...
third_party/WebKit/Source/platform/image-decoders/ImageFrame.cpp:102: if
(!m_bitmap.tryAllocPixels(m_allocator, 0))
On 2016/03/21 22:37:16, noel gordon wrote:
> Maybe don't call computeAlphaType() here?  Instead use m_premultiplyAlpha ?
> kPremul_SkAlphaType : kUnpremul_SkAlphaType;

Done.

https://codereview.chromium.org/1820733004/diff/1/third_party/WebKit/Source/p...
File
third_party/WebKit/Source/platform/image-decoders/gif/GIFImageDecoderTest.cpp
(right):

https://codereview.chromium.org/1820733004/diff/1/third_party/WebKit/Source/p...
third_party/WebKit/Source/platform/image-decoders/gif/GIFImageDecoderTest.cpp:524:
// Empirically chosen
On 2016/03/21 22:37:16, noel gordon wrote:
> // Empirically chosen truncation size:

Done.

https://codereview.chromium.org/1820733004/diff/1/third_party/WebKit/Source/p...
third_party/WebKit/Source/platform/image-decoders/gif/GIFImageDecoderTest.cpp:527:
static size_t kTruncateSize = 800;
On 2016/03/21 22:37:16, noel gordon wrote:
> static -> const ?

doh, done.

https://codereview.chromium.org/1820733004/diff/1/third_party/WebKit/Source/p...
third_party/WebKit/Source/platform/image-decoders/gif/GIFImageDecoderTest.cpp:540:
// Partial frame => alpha type reflects the requested format.
On 2016/03/21 22:37:16, noel gordon wrote:
> "Partial frame => alpha" -> "Partially decoded frame => the frame alpha type
is
> unknown and should reflect ..."

Done.

https://codereview.chromium.org/1820733004/diff/1/third_party/WebKit/Source/p...
third_party/WebKit/Source/platform/image-decoders/gif/GIFImageDecoderTest.cpp:552:
// Full frame => alpha type is known to be opaque
On 2016/03/21 22:37:16, noel gordon wrote:
> "Full frame =>" -> "Fully decoded frame => the frame alpha type is known
> (opaque)."

Done.

https://codereview.chromium.org/1820733004/diff/1/third_party/WebKit/Source/p...
third_party/WebKit/Source/platform/image-decoders/gif/GIFImageDecoderTest.cpp:553:
premulDecoder->setData(fullData.get(), false);
On 2016/03/21 22:37:16, noel gordon wrote:
> false -> true

Done.

https://codereview.chromium.org/1820733004/diff/1/third_party/WebKit/Source/p...
third_party/WebKit/Source/platform/image-decoders/gif/GIFImageDecoderTest.cpp:555:
unpremulDecoder->setData(fullData.get(), false);
On 2016/03/21 22:37:16, noel gordon wrote:
> false -> true

Done.

[xidachen, 2016-03-21 23:13:07]

This is AWESOME! Thank you so much Florin.

[Noel Gordon, 2016-03-21 23:50:24]

On 2016/03/21 23:13:07, xidachen wrote:
> This is AWESOME! Thank you so much Florin.

@xidachen: could you please file a crbug.com about this issue, explain the
problem, and make the bug depend on your HTML5 image bitmap work/bugs? 

Once the bug is filed, we can then add a BUG=Number link to this CL's change
description, and go ahead and land it.

[Noel Gordon, 2016-03-21 23:55:26]

On 2016/03/21 22:46:24, f(malita) wrote:

Thanks Florin, and LGTM pending a BUG number, tests passing...

[f(malita), 2016-03-22 21:09:24]

Description was changed from

==========
Propagate the decoder frame premul info to SkBitmap

Currently ImageFrame always sets up the resulting SkBitmap as
kPremul_SkAlphaType (or kOpaque_SkAlphaType in some cases), even
when AlphaOption == AlphaNotPremultiplied.

Update to set the correct bitmap alpha type.

R=noel@chromium.org,reed@google.com,xidachen@chromium.org
==========

to

==========
Propagate the decoder frame premul info to SkBitmap

Currently ImageFrame always sets up the resulting SkBitmap as
kPremul_SkAlphaType (or kOpaque_SkAlphaType in some cases), even
when AlphaOption == AlphaNotPremultiplied.

Update to set the correct bitmap alpha type.

R=noel@chromium.org,reed@google.com,xidachen@chromium.org
BUG=597035
==========

[f(malita), 2016-03-22 21:09:50]

On 2016/03/21 23:55:26, noel gordon wrote:
> On 2016/03/21 22:46:24, f(malita) wrote:
> 
> Thanks Florin, and LGTM pending a BUG number, tests passing...

Bug added, thanks.

[f(malita), 2016-03-22 21:09:54]

The CQ bit was checked by fmalita@chromium.org

[commit-bot: I haz the power, 2016-03-22 21:10:19]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1820733004/20001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1820733004/20001

[commit-bot: I haz the power, 2016-03-22 21:16:51]

Description was changed from

==========
Propagate the decoder frame premul info to SkBitmap

Currently ImageFrame always sets up the resulting SkBitmap as
kPremul_SkAlphaType (or kOpaque_SkAlphaType in some cases), even
when AlphaOption == AlphaNotPremultiplied.

Update to set the correct bitmap alpha type.

R=noel@chromium.org,reed@google.com,xidachen@chromium.org
BUG=597035
==========

to

==========
Propagate the decoder frame premul info to SkBitmap

Currently ImageFrame always sets up the resulting SkBitmap as
kPremul_SkAlphaType (or kOpaque_SkAlphaType in some cases), even
when AlphaOption == AlphaNotPremultiplied.

Update to set the correct bitmap alpha type.

R=noel@chromium.org,reed@google.com,xidachen@chromium.org
BUG=597035
==========

[commit-bot: I haz the power, 2016-03-22 21:16:52]

Committed patchset #2 (id:20001)

[commit-bot: I haz the power, 2016-03-22 21:18:37]

Description was changed from

==========
Propagate the decoder frame premul info to SkBitmap

Currently ImageFrame always sets up the resulting SkBitmap as
kPremul_SkAlphaType (or kOpaque_SkAlphaType in some cases), even
when AlphaOption == AlphaNotPremultiplied.

Update to set the correct bitmap alpha type.

R=noel@chromium.org,reed@google.com,xidachen@chromium.org
BUG=597035
==========

to

==========
Propagate the decoder frame premul info to SkBitmap

Currently ImageFrame always sets up the resulting SkBitmap as
kPremul_SkAlphaType (or kOpaque_SkAlphaType in some cases), even
when AlphaOption == AlphaNotPremultiplied.

Update to set the correct bitmap alpha type.

R=noel@chromium.org,reed@google.com,xidachen@chromium.org
BUG=597035

Committed: https://crrev.com/18a861d65e82a16f8cfbc7d3f160cfb3746523e2
Cr-Commit-Position: refs/heads/master@{#382680}
==========

[commit-bot: I haz the power, 2016-03-22 21:18:38]

Patchset 2 (id:??) landed as
https://crrev.com/18a861d65e82a16f8cfbc7d3f160cfb3746523e2
Cr-Commit-Position: refs/heads/master@{#382680}