Deferred GIF image decodes should report decode failures

Deferred GIF image decodes should report decode failures

For fast/images/size-failure.html fails due to slimming paint,
attempt to help resolve that issue. The image in the test is a
single-frame GIF which fails to decode due to corrupt LZW data
but the decoding failure is not recorded in the deferred image
decode system, causing continual re-decode requests while this
image is painted. All these extra decode requests fail.

When the GIF decoder landed in deferred, it no longer recorded
decode failures for GIF, and that was a regression in behavior
compared to the main-thread GIF decoder [1].

DecodingImageGenerator.cpp: include the frame index in the API
used for decoding (used for ImageFrameGenerator trace).

ImageFrameGenerator.cpp: removed m_decodeCount (was always set
to 0, never updated, and only used for trace). Change the name
of the decode failure member (AndEmpty suffix adds nothing) to
|m_decodeFailed|. Use it to record decoding failures, even for
multi-frame images like GIF to deny further requests to decode
the image (and undo the GIF behavioral regression [1]).

Add a new API to retrieve the decoding failure state. No mutex
is used when reading the state (not needed).

The comments, "This method is called to populate a discardable
memory owned by Skia" are removed. The deferred decoder writes
to the memory it is told to write to; it is not discardable in
all cases, other parts of the code even say that. GIF does not
currently decode to discardable memory, for example.

DeferredImageDecoder.cpp: detect decoding failures when frames
are requested and early-return on all decoding paths.

ImageFrameGeneratorTest.cpp: update the tests (the frame index
is the first argument of m_generator->decodeAndScale now).

[1] https://codereview.chromium.org/19838002

BUG=501543
Committed: https://crrev.com/edb61309dc83fb68645fbf712a885c5ad5d977b1
Cr-Commit-Position: refs/heads/master@{#366329}

[Noel Gordon, 2015-12-14 03:47:02]

Patchset #1 (id:1) has been deleted

[Noel Gordon, 2015-12-14 03:54:17]

Description was changed from

==========
fast/images/size-failure.html fails due to slimming paint

BUG=501543
==========

to

==========
fast/images/size-failure.html fails due to slimming paint

Attempt to help resolve that issue. The image in the test is a
single frame GIF which fails to decode due to corrupt LZW data
but the decoding failure is not recorded in the deferred image
decode system, causing continual re-decode requests while this
image is painted. All these extra decode requests fail.

When the GIF decoder landed in deferred, it no longer recorded
decode failures for GIF, and that was a regression in behavior
compared to the main-thread GIF decoder.

DecodingImageGenerator.cpp: include the frame index in the API
decoding calls (used for ImageFrameGenerator trace).

ImageFrameGenerator.cpp: removed m_decodeCount (was always set
to 0, never updated, and only used for trace). Change the name
of the decode failure member (AndEmpty suffix adds nothing) to
|m_decodeFailed|. Use it to record decoding failures, even for
multi-frame images like GIF to deny further requests to decode
the image (and undo the GIF behavioral regression).

Add a new API to retrieve the decoding failure state. No mutex
is used when reading the state (not needed).

The comments, "This method is called to populate a discardable
memory owned by Skia" are removed. The deferred decoder writes
to the memory it is told to write to; it is not discardable in
all cases: other parts of the code even say that. GIF does not
decode to discardable at all.

DeferredImageDecoder.cpp: detect decoding failures when frames
are reqeusted and early return on all decoding paths.

ImageFrameGeneratorTest.cpp: updates the test (the frame index
is the first argument of m_generator->decodeAndScale now).

BUG=501543
==========

[Noel Gordon, 2015-12-14 03:58:08]

Description was changed from

==========
fast/images/size-failure.html fails due to slimming paint

Attempt to help resolve that issue. The image in the test is a
single frame GIF which fails to decode due to corrupt LZW data
but the decoding failure is not recorded in the deferred image
decode system, causing continual re-decode requests while this
image is painted. All these extra decode requests fail.

When the GIF decoder landed in deferred, it no longer recorded
decode failures for GIF, and that was a regression in behavior
compared to the main-thread GIF decoder.

DecodingImageGenerator.cpp: include the frame index in the API
decoding calls (used for ImageFrameGenerator trace).

ImageFrameGenerator.cpp: removed m_decodeCount (was always set
to 0, never updated, and only used for trace). Change the name
of the decode failure member (AndEmpty suffix adds nothing) to
|m_decodeFailed|. Use it to record decoding failures, even for
multi-frame images like GIF to deny further requests to decode
the image (and undo the GIF behavioral regression).

Add a new API to retrieve the decoding failure state. No mutex
is used when reading the state (not needed).

The comments, "This method is called to populate a discardable
memory owned by Skia" are removed. The deferred decoder writes
to the memory it is told to write to; it is not discardable in
all cases: other parts of the code even say that. GIF does not
decode to discardable at all.

DeferredImageDecoder.cpp: detect decoding failures when frames
are reqeusted and early return on all decoding paths.

ImageFrameGeneratorTest.cpp: updates the test (the frame index
is the first argument of m_generator->decodeAndScale now).

BUG=501543
==========

to

==========
fast/images/size-failure.html fails due to slimming paint

Attempt to help resolve that issue. The image in the test is a
single frame GIF which fails to decode due to corrupt LZW data
but the decoding failure is not recorded in the deferred image
decode system, causing continual re-decode requests while this
image is painted. All these extra decode requests fail.

When the GIF decoder landed in deferred, it no longer recorded
decode failures for GIF, and that was a regression in behavior
compared to the main-thread GIF decoder.

DecodingImageGenerator.cpp: include the frame index in the API
decoding calls (used for ImageFrameGenerator trace).

ImageFrameGenerator.cpp: removed m_decodeCount (was always set
to 0, never updated, and only used for trace). Change the name
of the decode failure member (AndEmpty suffix adds nothing) to
|m_decodeFailed|. Use it to record decoding failures, even for
multi-frame images like GIF to deny further requests to decode
the image (and undo the GIF behavioral regression).

Add a new API to retrieve the decoding failure state. No mutex
is used when reading the state (not needed).

The comments, "This method is called to populate a discardable
memory owned by Skia" are removed. The deferred decoder writes
to the memory it is told to write to; it is not discardable in
all cases: other parts of the code even say that. GIF does not
currently decode to discardable memory.

DeferredImageDecoder.cpp: detect decoding failures when frames
are requested and early return on all decoding paths.

ImageFrameGeneratorTest.cpp: updates the test (the frame index
is the first argument of m_generator->decodeAndScale now).

BUG=501543
==========

[Noel Gordon, 2015-12-14 04:07:12]

Description was changed from

==========
fast/images/size-failure.html fails due to slimming paint

Attempt to help resolve that issue. The image in the test is a
single frame GIF which fails to decode due to corrupt LZW data
but the decoding failure is not recorded in the deferred image
decode system, causing continual re-decode requests while this
image is painted. All these extra decode requests fail.

When the GIF decoder landed in deferred, it no longer recorded
decode failures for GIF, and that was a regression in behavior
compared to the main-thread GIF decoder.

DecodingImageGenerator.cpp: include the frame index in the API
decoding calls (used for ImageFrameGenerator trace).

ImageFrameGenerator.cpp: removed m_decodeCount (was always set
to 0, never updated, and only used for trace). Change the name
of the decode failure member (AndEmpty suffix adds nothing) to
|m_decodeFailed|. Use it to record decoding failures, even for
multi-frame images like GIF to deny further requests to decode
the image (and undo the GIF behavioral regression).

Add a new API to retrieve the decoding failure state. No mutex
is used when reading the state (not needed).

The comments, "This method is called to populate a discardable
memory owned by Skia" are removed. The deferred decoder writes
to the memory it is told to write to; it is not discardable in
all cases: other parts of the code even say that. GIF does not
currently decode to discardable memory.

DeferredImageDecoder.cpp: detect decoding failures when frames
are requested and early return on all decoding paths.

ImageFrameGeneratorTest.cpp: updates the test (the frame index
is the first argument of m_generator->decodeAndScale now).

BUG=501543
==========

to

==========
fast/images/size-failure.html fails due to slimming paint

Attempt to help resolve that issue. The image in the test is a
single frame GIF which fails to decode due to corrupt LZW data
but the decoding failure is not recorded in the deferred image
decode system, causing continual re-decode requests while this
image is painted. All these extra decode requests fail.

When the GIF decoder landed in deferred, it no longer recorded
decode failures for GIF, and that was a regression in behavior
compared to the main-thread GIF decoder.

DecodingImageGenerator.cpp: include the frame index in the API
decoding calls (used for ImageFrameGenerator trace).

ImageFrameGenerator.cpp: removed m_decodeCount (was always set
to 0, never updated, and only used for trace). Change the name
of the decode failure member (AndEmpty suffix adds nothing) to
|m_decodeFailed|. Use it to record decoding failures, even for
multi-frame images like GIF to deny further requests to decode
the image (and undo the GIF behavioral regression).

Add a new API to retrieve the decoding failure state. No mutex
is used when reading the state (not needed).

The comments, "This method is called to populate a discardable
memory owned by Skia" are removed. The deferred decoder writes
to the memory it is told to write to; it is not discardable in
all cases, other parts of the code even say that. GIF does not
currently decode to discardable memory, for example.

DeferredImageDecoder.cpp: detect decoding failures when frames
are requested and early return on all decoding paths.

ImageFrameGeneratorTest.cpp: updates the test (the frame index
is the first argument of m_generator->decodeAndScale now).

BUG=501543
==========

[Noel Gordon, 2015-12-14 04:21:55]

noel@chromium.org changed reviewers:
+ enne@chromium.org

[Noel Gordon, 2015-12-14 04:48:01]

The CQ bit was checked by noel@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2015-12-14 04:48:16]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1527433002/20001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1527433002/20001

[commit-bot: I haz the power, 2015-12-14 05:52:55]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2015-12-14 05:52:56]

Dry run: This issue passed the CQ dry run.

[Noel Gordon, 2015-12-14 11:13:41]

Description was changed from

==========
fast/images/size-failure.html fails due to slimming paint

Attempt to help resolve that issue. The image in the test is a
single frame GIF which fails to decode due to corrupt LZW data
but the decoding failure is not recorded in the deferred image
decode system, causing continual re-decode requests while this
image is painted. All these extra decode requests fail.

When the GIF decoder landed in deferred, it no longer recorded
decode failures for GIF, and that was a regression in behavior
compared to the main-thread GIF decoder.

DecodingImageGenerator.cpp: include the frame index in the API
decoding calls (used for ImageFrameGenerator trace).

ImageFrameGenerator.cpp: removed m_decodeCount (was always set
to 0, never updated, and only used for trace). Change the name
of the decode failure member (AndEmpty suffix adds nothing) to
|m_decodeFailed|. Use it to record decoding failures, even for
multi-frame images like GIF to deny further requests to decode
the image (and undo the GIF behavioral regression).

Add a new API to retrieve the decoding failure state. No mutex
is used when reading the state (not needed).

The comments, "This method is called to populate a discardable
memory owned by Skia" are removed. The deferred decoder writes
to the memory it is told to write to; it is not discardable in
all cases, other parts of the code even say that. GIF does not
currently decode to discardable memory, for example.

DeferredImageDecoder.cpp: detect decoding failures when frames
are requested and early return on all decoding paths.

ImageFrameGeneratorTest.cpp: updates the test (the frame index
is the first argument of m_generator->decodeAndScale now).

BUG=501543
==========

to

==========
fast/images/size-failure.html fails due to slimming paint

Attempt to help resolve that issue. The image in the test is a
single frame GIF which fails to decode due to corrupt LZW data
but the decoding failure is not recorded in the deferred image
decode system, causing continual re-decode requests while this
image is painted. All these extra decode requests fail.

When the GIF decoder landed in deferred, it no longer recorded
decode failures for GIF, and that was a regression in behavior
compared to the main-thread GIF decoder.

DecodingImageGenerator.cpp: include the frame index in the API
decoding calls (used for ImageFrameGenerator trace).

ImageFrameGenerator.cpp: removed m_decodeCount (was always set
to 0, never updated, and only used for trace). Change the name
of the decode failure member (AndEmpty suffix adds nothing) to
|m_decodeFailed|. Use it to record decoding failures, even for
multi-frame images like GIF to deny further requests to decode
the image (and undo the GIF behavioral regression).

Add a new API to retrieve the decoding failure state. No mutex
is used when reading the state (not needed).

The comments, "This method is called to populate a discardable
memory owned by Skia" are removed. The deferred decoder writes
to the memory it is told to write to; it is not discardable in
all cases, other parts of the code even say that. GIF does not
currently decode to discardable memory, for example.

DeferredImageDecoder.cpp: detect decoding failures when frames
are requested and early return on all decoding paths.

ImageFrameGeneratorTest.cpp: update the tests (the frame index
is the first argument of m_generator->decodeAndScale now).

BUG=501543
==========

[Noel Gordon, 2015-12-14 11:15:05]

Description was changed from

==========
fast/images/size-failure.html fails due to slimming paint

Attempt to help resolve that issue. The image in the test is a
single frame GIF which fails to decode due to corrupt LZW data
but the decoding failure is not recorded in the deferred image
decode system, causing continual re-decode requests while this
image is painted. All these extra decode requests fail.

When the GIF decoder landed in deferred, it no longer recorded
decode failures for GIF, and that was a regression in behavior
compared to the main-thread GIF decoder.

DecodingImageGenerator.cpp: include the frame index in the API
decoding calls (used for ImageFrameGenerator trace).

ImageFrameGenerator.cpp: removed m_decodeCount (was always set
to 0, never updated, and only used for trace). Change the name
of the decode failure member (AndEmpty suffix adds nothing) to
|m_decodeFailed|. Use it to record decoding failures, even for
multi-frame images like GIF to deny further requests to decode
the image (and undo the GIF behavioral regression).

Add a new API to retrieve the decoding failure state. No mutex
is used when reading the state (not needed).

The comments, "This method is called to populate a discardable
memory owned by Skia" are removed. The deferred decoder writes
to the memory it is told to write to; it is not discardable in
all cases, other parts of the code even say that. GIF does not
currently decode to discardable memory, for example.

DeferredImageDecoder.cpp: detect decoding failures when frames
are requested and early return on all decoding paths.

ImageFrameGeneratorTest.cpp: update the tests (the frame index
is the first argument of m_generator->decodeAndScale now).

BUG=501543
==========

to

==========
fast/images/size-failure.html fails due to slimming paint

Attempt to help resolve that issue. The image in the test is a
single frame GIF which fails to decode due to corrupt LZW data
but the decoding failure is not recorded in the deferred image
decode system, causing continual re-decode requests while this
image is painted. All these extra decode requests fail.

When the GIF decoder landed in deferred, it no longer recorded
decode failures for GIF, and that was a regression in behavior
compared to the main-thread GIF decoder.

DecodingImageGenerator.cpp: include the frame index in the API
decoding calls (used for ImageFrameGenerator trace).

ImageFrameGenerator.cpp: removed m_decodeCount (was always set
to 0, never updated, and only used for trace). Change the name
of the decode failure member (AndEmpty suffix adds nothing) to
|m_decodeFailed|. Use it to record decoding failures, even for
multi-frame images like GIF to deny further requests to decode
the image (and undo the GIF behavioral regression).

Add a new API to retrieve the decoding failure state. No mutex
is used when reading the state (not needed).

The comments, "This method is called to populate a discardable
memory owned by Skia" are removed. The deferred decoder writes
to the memory it is told to write to; it is not discardable in
all cases, other parts of the code even say that. GIF does not
currently decode to discardable memory, for example.

DeferredImageDecoder.cpp: detect decoding failures when frames
are requested and early-return on all decoding paths.

ImageFrameGeneratorTest.cpp: update the tests (the frame index
is the first argument of m_generator->decodeAndScale now).

BUG=501543
==========

[Noel Gordon, 2015-12-14 15:31:23]

Description was changed from

==========
fast/images/size-failure.html fails due to slimming paint

Attempt to help resolve that issue. The image in the test is a
single frame GIF which fails to decode due to corrupt LZW data
but the decoding failure is not recorded in the deferred image
decode system, causing continual re-decode requests while this
image is painted. All these extra decode requests fail.

When the GIF decoder landed in deferred, it no longer recorded
decode failures for GIF, and that was a regression in behavior
compared to the main-thread GIF decoder.

DecodingImageGenerator.cpp: include the frame index in the API
decoding calls (used for ImageFrameGenerator trace).

ImageFrameGenerator.cpp: removed m_decodeCount (was always set
to 0, never updated, and only used for trace). Change the name
of the decode failure member (AndEmpty suffix adds nothing) to
|m_decodeFailed|. Use it to record decoding failures, even for
multi-frame images like GIF to deny further requests to decode
the image (and undo the GIF behavioral regression).

Add a new API to retrieve the decoding failure state. No mutex
is used when reading the state (not needed).

The comments, "This method is called to populate a discardable
memory owned by Skia" are removed. The deferred decoder writes
to the memory it is told to write to; it is not discardable in
all cases, other parts of the code even say that. GIF does not
currently decode to discardable memory, for example.

DeferredImageDecoder.cpp: detect decoding failures when frames
are requested and early-return on all decoding paths.

ImageFrameGeneratorTest.cpp: update the tests (the frame index
is the first argument of m_generator->decodeAndScale now).

BUG=501543
==========

to

==========
fast/images/size-failure.html fails due to slimming paint

Attempt to help resolve that issue. The image in the test is a
single frame GIF which fails to decode due to corrupt LZW data
but the decoding failure is not recorded in the deferred image
decode system, causing continual re-decode requests while this
image is painted. All these extra decode requests fail.

When the GIF decoder landed in deferred, it no longer recorded
decode failures for GIF, and that was a regression in behavior
compared to the main-thread GIF decoder.

DecodingImageGenerator.cpp: include the frame index in the API
decoding calls (used for ImageFrameGenerator trace).

ImageFrameGenerator.cpp: removed m_decodeCount (was always set
to 0, never updated, and only used for trace). Change the name
of the decode failure member (AndEmpty suffix adds nothing) to
|m_decodeFailed|. Use it to record decoding failures, even for
multi-frame images like GIF to deny further requests to decode
the image (and undo the GIF behavioral regression [1]).

Add a new API to retrieve the decoding failure state. No mutex
is used when reading the state (not needed).

The comments, "This method is called to populate a discardable
memory owned by Skia" are removed. The deferred decoder writes
to the memory it is told to write to; it is not discardable in
all cases, other parts of the code even say that. GIF does not
currently decode to discardable memory, for example.

DeferredImageDecoder.cpp: detect decoding failures when frames
are requested and early-return on all decoding paths.

ImageFrameGeneratorTest.cpp: update the tests (the frame index
is the first argument of m_generator->decodeAndScale now).

[1] https://codereview.chromium.org/19838002

BUG=501543
==========

[Noel Gordon, 2015-12-14 15:32:13]

Description was changed from

==========
fast/images/size-failure.html fails due to slimming paint

Attempt to help resolve that issue. The image in the test is a
single frame GIF which fails to decode due to corrupt LZW data
but the decoding failure is not recorded in the deferred image
decode system, causing continual re-decode requests while this
image is painted. All these extra decode requests fail.

When the GIF decoder landed in deferred, it no longer recorded
decode failures for GIF, and that was a regression in behavior
compared to the main-thread GIF decoder.

DecodingImageGenerator.cpp: include the frame index in the API
decoding calls (used for ImageFrameGenerator trace).

ImageFrameGenerator.cpp: removed m_decodeCount (was always set
to 0, never updated, and only used for trace). Change the name
of the decode failure member (AndEmpty suffix adds nothing) to
|m_decodeFailed|. Use it to record decoding failures, even for
multi-frame images like GIF to deny further requests to decode
the image (and undo the GIF behavioral regression [1]).

Add a new API to retrieve the decoding failure state. No mutex
is used when reading the state (not needed).

The comments, "This method is called to populate a discardable
memory owned by Skia" are removed. The deferred decoder writes
to the memory it is told to write to; it is not discardable in
all cases, other parts of the code even say that. GIF does not
currently decode to discardable memory, for example.

DeferredImageDecoder.cpp: detect decoding failures when frames
are requested and early-return on all decoding paths.

ImageFrameGeneratorTest.cpp: update the tests (the frame index
is the first argument of m_generator->decodeAndScale now).

[1] https://codereview.chromium.org/19838002

BUG=501543
==========

to

==========
fast/images/size-failure.html fails due to slimming paint

Attempt to help resolve that issue. The image in the test is a
single frame GIF which fails to decode due to corrupt LZW data
but the decoding failure is not recorded in the deferred image
decode system, causing continual re-decode requests while this
image is painted. All these extra decode requests fail.

When the GIF decoder landed in deferred, it no longer recorded
decode failures for GIF, and that was a regression in behavior
compared to the main-thread GIF decoder [1].

DecodingImageGenerator.cpp: include the frame index in the API
decoding calls (used for ImageFrameGenerator trace).

ImageFrameGenerator.cpp: removed m_decodeCount (was always set
to 0, never updated, and only used for trace). Change the name
of the decode failure member (AndEmpty suffix adds nothing) to
|m_decodeFailed|. Use it to record decoding failures, even for
multi-frame images like GIF to deny further requests to decode
the image (and undo the GIF behavioral regression [1]).

Add a new API to retrieve the decoding failure state. No mutex
is used when reading the state (not needed).

The comments, "This method is called to populate a discardable
memory owned by Skia" are removed. The deferred decoder writes
to the memory it is told to write to; it is not discardable in
all cases, other parts of the code even say that. GIF does not
currently decode to discardable memory, for example.

DeferredImageDecoder.cpp: detect decoding failures when frames
are requested and early-return on all decoding paths.

ImageFrameGeneratorTest.cpp: update the tests (the frame index
is the first argument of m_generator->decodeAndScale now).

[1] https://codereview.chromium.org/19838002

BUG=501543
==========

[Noel Gordon, 2015-12-16 01:14:49]

PTAL.

[scroggo_chromium, 2015-12-18 18:12:05]

scroggo@chromium.org changed reviewers:
+ scroggo@chromium.org

[scroggo_chromium, 2015-12-18 18:12:06]

lgtm

https://codereview.chromium.org/1527433002/diff/20001/third_party/WebKit/Sour...
File third_party/WebKit/Source/platform/graphics/ImageFrameGenerator.cpp
(right):

https://codereview.chromium.org/1527433002/diff/20001/third_party/WebKit/Sour...
third_party/WebKit/Source/platform/graphics/ImageFrameGenerator.cpp:202: // by
Skia. If not, make a copy.
I support removing the comment above ("This method is called to populate a
discardable memory owned by Skia") for the exact reason you mentioned - this
method populates whatever memory was handed to it at |pixels|. I think this
comment should be modified for the same reason. You can just remove "by Skia".

[Noel Gordon, 2015-12-21 03:23:23]

https://codereview.chromium.org/1527433002/diff/20001/third_party/WebKit/Sour...
File third_party/WebKit/Source/platform/graphics/ImageFrameGenerator.cpp
(right):

https://codereview.chromium.org/1527433002/diff/20001/third_party/WebKit/Sour...
third_party/WebKit/Source/platform/graphics/ImageFrameGenerator.cpp:202: // by
Skia. If not, make a copy.
On 2015/12/18 18:12:06, scroggo_chromium wrote:
> I support removing the comment above ("This method is called to populate a
> discardable memory owned by Skia") for the exact reason you mentioned - this
> method populates whatever memory was handed to it at |pixels|. I think this
> comment should be modified for the same reason. You can just remove "by Skia".

Thanks Leon, done.  Removed the "by Skia" bit.

[Noel Gordon, 2015-12-21 03:23:32]

The CQ bit was checked by noel@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2015-12-21 03:23:39]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1527433002/40001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1527433002/40001

[Noel Gordon, 2015-12-21 04:39:12]

Description was changed from

==========
fast/images/size-failure.html fails due to slimming paint

Attempt to help resolve that issue. The image in the test is a
single frame GIF which fails to decode due to corrupt LZW data
but the decoding failure is not recorded in the deferred image
decode system, causing continual re-decode requests while this
image is painted. All these extra decode requests fail.

When the GIF decoder landed in deferred, it no longer recorded
decode failures for GIF, and that was a regression in behavior
compared to the main-thread GIF decoder [1].

DecodingImageGenerator.cpp: include the frame index in the API
decoding calls (used for ImageFrameGenerator trace).

ImageFrameGenerator.cpp: removed m_decodeCount (was always set
to 0, never updated, and only used for trace). Change the name
of the decode failure member (AndEmpty suffix adds nothing) to
|m_decodeFailed|. Use it to record decoding failures, even for
multi-frame images like GIF to deny further requests to decode
the image (and undo the GIF behavioral regression [1]).

Add a new API to retrieve the decoding failure state. No mutex
is used when reading the state (not needed).

The comments, "This method is called to populate a discardable
memory owned by Skia" are removed. The deferred decoder writes
to the memory it is told to write to; it is not discardable in
all cases, other parts of the code even say that. GIF does not
currently decode to discardable memory, for example.

DeferredImageDecoder.cpp: detect decoding failures when frames
are requested and early-return on all decoding paths.

ImageFrameGeneratorTest.cpp: update the tests (the frame index
is the first argument of m_generator->decodeAndScale now).

[1] https://codereview.chromium.org/19838002

BUG=501543
==========

to

==========
Deferred GIF image decodes should report decode failures

For fast/images/size-failure.html fails due to slimming paint,
attempt to help resolve that issue. The image in the test is a
single frame GIF which fails to decode due to corrupt LZW data
but the decoding failure is not recorded in the deferred image
decode system, causing continual re-decode requests while this
image is painted. All these extra decode requests fail.

When the GIF decoder landed in deferred, it no longer recorded
decode failures for GIF, and that was a regression in behavior
compared to the main-thread GIF decoder [1].

DecodingImageGenerator.cpp: include the frame index in the API
decoding calls (used for ImageFrameGenerator trace).

ImageFrameGenerator.cpp: removed m_decodeCount (was always set
to 0, never updated, and only used for trace). Change the name
of the decode failure member (AndEmpty suffix adds nothing) to
|m_decodeFailed|. Use it to record decoding failures, even for
multi-frame images like GIF to deny further requests to decode
the image (and undo the GIF behavioral regression [1]).

Add a new API to retrieve the decoding failure state. No mutex
is used when reading the state (not needed).

The comments, "This method is called to populate a discardable
memory owned by Skia" are removed. The deferred decoder writes
to the memory it is told to write to; it is not discardable in
all cases, other parts of the code even say that. GIF does not
currently decode to discardable memory, for example.

DeferredImageDecoder.cpp: detect decoding failures when frames
are requested and early-return on all decoding paths.

ImageFrameGeneratorTest.cpp: update the tests (the frame index
is the first argument of m_generator->decodeAndScale now).

[1] https://codereview.chromium.org/19838002

BUG=501543
==========

[Noel Gordon, 2015-12-21 04:43:29]

Description was changed from

==========
Deferred GIF image decodes should report decode failures

For fast/images/size-failure.html fails due to slimming paint,
attempt to help resolve that issue. The image in the test is a
single frame GIF which fails to decode due to corrupt LZW data
but the decoding failure is not recorded in the deferred image
decode system, causing continual re-decode requests while this
image is painted. All these extra decode requests fail.

When the GIF decoder landed in deferred, it no longer recorded
decode failures for GIF, and that was a regression in behavior
compared to the main-thread GIF decoder [1].

DecodingImageGenerator.cpp: include the frame index in the API
decoding calls (used for ImageFrameGenerator trace).

ImageFrameGenerator.cpp: removed m_decodeCount (was always set
to 0, never updated, and only used for trace). Change the name
of the decode failure member (AndEmpty suffix adds nothing) to
|m_decodeFailed|. Use it to record decoding failures, even for
multi-frame images like GIF to deny further requests to decode
the image (and undo the GIF behavioral regression [1]).

Add a new API to retrieve the decoding failure state. No mutex
is used when reading the state (not needed).

The comments, "This method is called to populate a discardable
memory owned by Skia" are removed. The deferred decoder writes
to the memory it is told to write to; it is not discardable in
all cases, other parts of the code even say that. GIF does not
currently decode to discardable memory, for example.

DeferredImageDecoder.cpp: detect decoding failures when frames
are requested and early-return on all decoding paths.

ImageFrameGeneratorTest.cpp: update the tests (the frame index
is the first argument of m_generator->decodeAndScale now).

[1] https://codereview.chromium.org/19838002

BUG=501543
==========

to

==========
Deferred GIF image decodes should report decode failures

For fast/images/size-failure.html fails due to slimming paint,
attempt to help resolve that issue. The image in the test is a
single-frame GIF which fails to decode due to corrupt LZW data
but the decoding failure is not recorded in the deferred image
decode system, causing continual re-decode requests while this
image is painted. All the extra decode requests fail.

When the GIF decoder landed in deferred, it no longer recorded
decode failures for GIF, and that was a regression in behavior
compared to the main-thread GIF decoder [1].

DecodingImageGenerator.cpp: include the frame index in the API
decoding calls (used for ImageFrameGenerator trace).

ImageFrameGenerator.cpp: removed m_decodeCount (was always set
to 0, never updated, and only used for trace). Change the name
of the decode failure member (AndEmpty suffix adds nothing) to
|m_decodeFailed|. Use it to record decoding failures, even for
multi-frame images like GIF to deny further requests to decode
the image (and undo the GIF behavioral regression [1]).

Add a new API to retrieve the decoding failure state. No mutex
is used when reading the state (not needed).

The comments, "This method is called to populate a discardable
memory owned by Skia" are removed. The deferred decoder writes
to the memory it is told to write to; it is not discardable in
all cases, other parts of the code even say that. GIF does not
currently decode to discardable memory, for example.

DeferredImageDecoder.cpp: detect decoding failures when frames
are requested and early-return on all decoding paths.

ImageFrameGeneratorTest.cpp: update the tests (the frame index
is the first argument of m_generator->decodeAndScale now).

[1] https://codereview.chromium.org/19838002

BUG=501543
==========

[Noel Gordon, 2015-12-21 04:55:49]

Description was changed from

==========
Deferred GIF image decodes should report decode failures

For fast/images/size-failure.html fails due to slimming paint,
attempt to help resolve that issue. The image in the test is a
single-frame GIF which fails to decode due to corrupt LZW data
but the decoding failure is not recorded in the deferred image
decode system, causing continual re-decode requests while this
image is painted. All the extra decode requests fail.

When the GIF decoder landed in deferred, it no longer recorded
decode failures for GIF, and that was a regression in behavior
compared to the main-thread GIF decoder [1].

DecodingImageGenerator.cpp: include the frame index in the API
decoding calls (used for ImageFrameGenerator trace).

ImageFrameGenerator.cpp: removed m_decodeCount (was always set
to 0, never updated, and only used for trace). Change the name
of the decode failure member (AndEmpty suffix adds nothing) to
|m_decodeFailed|. Use it to record decoding failures, even for
multi-frame images like GIF to deny further requests to decode
the image (and undo the GIF behavioral regression [1]).

Add a new API to retrieve the decoding failure state. No mutex
is used when reading the state (not needed).

The comments, "This method is called to populate a discardable
memory owned by Skia" are removed. The deferred decoder writes
to the memory it is told to write to; it is not discardable in
all cases, other parts of the code even say that. GIF does not
currently decode to discardable memory, for example.

DeferredImageDecoder.cpp: detect decoding failures when frames
are requested and early-return on all decoding paths.

ImageFrameGeneratorTest.cpp: update the tests (the frame index
is the first argument of m_generator->decodeAndScale now).

[1] https://codereview.chromium.org/19838002

BUG=501543
==========

to

==========
Deferred GIF image decodes should report decode failures

For fast/images/size-failure.html fails due to slimming paint,
attempt to help resolve that issue. The image in the test is a
single-frame GIF which fails to decode due to corrupt LZW data
but the decoding failure is not recorded in the deferred image
decode system, causing continual re-decode requests while this
image is painted. All the extra decode requests fail.

When the GIF decoder landed in deferred, it no longer recorded
decode failures for GIF, and that was a regression in behavior
compared to the main-thread GIF decoder [1].

DecodingImageGenerator.cpp: include the frame index in the API
used for decoding (used for ImageFrameGenerator trace).

ImageFrameGenerator.cpp: removed m_decodeCount (was always set
to 0, never updated, and only used for trace). Change the name
of the decode failure member (AndEmpty suffix adds nothing) to
|m_decodeFailed|. Use it to record decoding failures, even for
multi-frame images like GIF to deny further requests to decode
the image (and undo the GIF behavioral regression [1]).

Add a new API to retrieve the decoding failure state. No mutex
is used when reading the state (not needed).

The comments, "This method is called to populate a discardable
memory owned by Skia" are removed. The deferred decoder writes
to the memory it is told to write to; it is not discardable in
all cases, other parts of the code even say that. GIF does not
currently decode to discardable memory, for example.

DeferredImageDecoder.cpp: detect decoding failures when frames
are requested and early-return on all decoding paths.

ImageFrameGeneratorTest.cpp: update the tests (the frame index
is the first argument of m_generator->decodeAndScale now).

[1] https://codereview.chromium.org/19838002

BUG=501543
==========

[Noel Gordon, 2015-12-21 04:57:12]

Description was changed from

==========
Deferred GIF image decodes should report decode failures

For fast/images/size-failure.html fails due to slimming paint,
attempt to help resolve that issue. The image in the test is a
single-frame GIF which fails to decode due to corrupt LZW data
but the decoding failure is not recorded in the deferred image
decode system, causing continual re-decode requests while this
image is painted. All the extra decode requests fail.

When the GIF decoder landed in deferred, it no longer recorded
decode failures for GIF, and that was a regression in behavior
compared to the main-thread GIF decoder [1].

DecodingImageGenerator.cpp: include the frame index in the API
used for decoding (used for ImageFrameGenerator trace).

ImageFrameGenerator.cpp: removed m_decodeCount (was always set
to 0, never updated, and only used for trace). Change the name
of the decode failure member (AndEmpty suffix adds nothing) to
|m_decodeFailed|. Use it to record decoding failures, even for
multi-frame images like GIF to deny further requests to decode
the image (and undo the GIF behavioral regression [1]).

Add a new API to retrieve the decoding failure state. No mutex
is used when reading the state (not needed).

The comments, "This method is called to populate a discardable
memory owned by Skia" are removed. The deferred decoder writes
to the memory it is told to write to; it is not discardable in
all cases, other parts of the code even say that. GIF does not
currently decode to discardable memory, for example.

DeferredImageDecoder.cpp: detect decoding failures when frames
are requested and early-return on all decoding paths.

ImageFrameGeneratorTest.cpp: update the tests (the frame index
is the first argument of m_generator->decodeAndScale now).

[1] https://codereview.chromium.org/19838002

BUG=501543
==========

to

==========
Deferred GIF image decodes should report decode failures

For fast/images/size-failure.html fails due to slimming paint,
attempt to help resolve that issue. The image in the test is a
single-frame GIF which fails to decode due to corrupt LZW data
but the decoding failure is not recorded in the deferred image
decode system, causing continual re-decode requests while this
image is painted. All these extra decode requests fail.

When the GIF decoder landed in deferred, it no longer recorded
decode failures for GIF, and that was a regression in behavior
compared to the main-thread GIF decoder [1].

DecodingImageGenerator.cpp: include the frame index in the API
used for decoding (used for ImageFrameGenerator trace).

ImageFrameGenerator.cpp: removed m_decodeCount (was always set
to 0, never updated, and only used for trace). Change the name
of the decode failure member (AndEmpty suffix adds nothing) to
|m_decodeFailed|. Use it to record decoding failures, even for
multi-frame images like GIF to deny further requests to decode
the image (and undo the GIF behavioral regression [1]).

Add a new API to retrieve the decoding failure state. No mutex
is used when reading the state (not needed).

The comments, "This method is called to populate a discardable
memory owned by Skia" are removed. The deferred decoder writes
to the memory it is told to write to; it is not discardable in
all cases, other parts of the code even say that. GIF does not
currently decode to discardable memory, for example.

DeferredImageDecoder.cpp: detect decoding failures when frames
are requested and early-return on all decoding paths.

ImageFrameGeneratorTest.cpp: update the tests (the frame index
is the first argument of m_generator->decodeAndScale now).

[1] https://codereview.chromium.org/19838002

BUG=501543
==========

[commit-bot: I haz the power, 2015-12-21 05:05:49]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2015-12-21 05:05:50]

Dry run: This issue passed the CQ dry run.

[Noel Gordon, 2015-12-21 05:08:13]

The CQ bit was checked by noel@chromium.org

[Noel Gordon, 2015-12-21 05:08:13]

The patchset sent to the CQ was uploaded after l-g-t-m from scroggo@chromium.org
Link to the patchset: https://codereview.chromium.org/1527433002/#ps40001
(title: "Patch for landing")

[commit-bot: I haz the power, 2015-12-21 05:08:48]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1527433002/40001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1527433002/40001

[commit-bot: I haz the power, 2015-12-21 05:12:58]

Description was changed from

==========
Deferred GIF image decodes should report decode failures

For fast/images/size-failure.html fails due to slimming paint,
attempt to help resolve that issue. The image in the test is a
single-frame GIF which fails to decode due to corrupt LZW data
but the decoding failure is not recorded in the deferred image
decode system, causing continual re-decode requests while this
image is painted. All these extra decode requests fail.

When the GIF decoder landed in deferred, it no longer recorded
decode failures for GIF, and that was a regression in behavior
compared to the main-thread GIF decoder [1].

DecodingImageGenerator.cpp: include the frame index in the API
used for decoding (used for ImageFrameGenerator trace).

ImageFrameGenerator.cpp: removed m_decodeCount (was always set
to 0, never updated, and only used for trace). Change the name
of the decode failure member (AndEmpty suffix adds nothing) to
|m_decodeFailed|. Use it to record decoding failures, even for
multi-frame images like GIF to deny further requests to decode
the image (and undo the GIF behavioral regression [1]).

Add a new API to retrieve the decoding failure state. No mutex
is used when reading the state (not needed).

The comments, "This method is called to populate a discardable
memory owned by Skia" are removed. The deferred decoder writes
to the memory it is told to write to; it is not discardable in
all cases, other parts of the code even say that. GIF does not
currently decode to discardable memory, for example.

DeferredImageDecoder.cpp: detect decoding failures when frames
are requested and early-return on all decoding paths.

ImageFrameGeneratorTest.cpp: update the tests (the frame index
is the first argument of m_generator->decodeAndScale now).

[1] https://codereview.chromium.org/19838002

BUG=501543
==========

to

==========
Deferred GIF image decodes should report decode failures

For fast/images/size-failure.html fails due to slimming paint,
attempt to help resolve that issue. The image in the test is a
single-frame GIF which fails to decode due to corrupt LZW data
but the decoding failure is not recorded in the deferred image
decode system, causing continual re-decode requests while this
image is painted. All these extra decode requests fail.

When the GIF decoder landed in deferred, it no longer recorded
decode failures for GIF, and that was a regression in behavior
compared to the main-thread GIF decoder [1].

DecodingImageGenerator.cpp: include the frame index in the API
used for decoding (used for ImageFrameGenerator trace).

ImageFrameGenerator.cpp: removed m_decodeCount (was always set
to 0, never updated, and only used for trace). Change the name
of the decode failure member (AndEmpty suffix adds nothing) to
|m_decodeFailed|. Use it to record decoding failures, even for
multi-frame images like GIF to deny further requests to decode
the image (and undo the GIF behavioral regression [1]).

Add a new API to retrieve the decoding failure state. No mutex
is used when reading the state (not needed).

The comments, "This method is called to populate a discardable
memory owned by Skia" are removed. The deferred decoder writes
to the memory it is told to write to; it is not discardable in
all cases, other parts of the code even say that. GIF does not
currently decode to discardable memory, for example.

DeferredImageDecoder.cpp: detect decoding failures when frames
are requested and early-return on all decoding paths.

ImageFrameGeneratorTest.cpp: update the tests (the frame index
is the first argument of m_generator->decodeAndScale now).

[1] https://codereview.chromium.org/19838002

BUG=501543
==========

[commit-bot: I haz the power, 2015-12-21 05:13:02]

Committed patchset #2 (id:40001)

[commit-bot: I haz the power, 2015-12-21 05:13:51]

Description was changed from

==========
Deferred GIF image decodes should report decode failures

For fast/images/size-failure.html fails due to slimming paint,
attempt to help resolve that issue. The image in the test is a
single-frame GIF which fails to decode due to corrupt LZW data
but the decoding failure is not recorded in the deferred image
decode system, causing continual re-decode requests while this
image is painted. All these extra decode requests fail.

When the GIF decoder landed in deferred, it no longer recorded
decode failures for GIF, and that was a regression in behavior
compared to the main-thread GIF decoder [1].

DecodingImageGenerator.cpp: include the frame index in the API
used for decoding (used for ImageFrameGenerator trace).

ImageFrameGenerator.cpp: removed m_decodeCount (was always set
to 0, never updated, and only used for trace). Change the name
of the decode failure member (AndEmpty suffix adds nothing) to
|m_decodeFailed|. Use it to record decoding failures, even for
multi-frame images like GIF to deny further requests to decode
the image (and undo the GIF behavioral regression [1]).

Add a new API to retrieve the decoding failure state. No mutex
is used when reading the state (not needed).

The comments, "This method is called to populate a discardable
memory owned by Skia" are removed. The deferred decoder writes
to the memory it is told to write to; it is not discardable in
all cases, other parts of the code even say that. GIF does not
currently decode to discardable memory, for example.

DeferredImageDecoder.cpp: detect decoding failures when frames
are requested and early-return on all decoding paths.

ImageFrameGeneratorTest.cpp: update the tests (the frame index
is the first argument of m_generator->decodeAndScale now).

[1] https://codereview.chromium.org/19838002

BUG=501543
==========

to

==========
Deferred GIF image decodes should report decode failures

For fast/images/size-failure.html fails due to slimming paint,
attempt to help resolve that issue. The image in the test is a
single-frame GIF which fails to decode due to corrupt LZW data
but the decoding failure is not recorded in the deferred image
decode system, causing continual re-decode requests while this
image is painted. All these extra decode requests fail.

When the GIF decoder landed in deferred, it no longer recorded
decode failures for GIF, and that was a regression in behavior
compared to the main-thread GIF decoder [1].

DecodingImageGenerator.cpp: include the frame index in the API
used for decoding (used for ImageFrameGenerator trace).

ImageFrameGenerator.cpp: removed m_decodeCount (was always set
to 0, never updated, and only used for trace). Change the name
of the decode failure member (AndEmpty suffix adds nothing) to
|m_decodeFailed|. Use it to record decoding failures, even for
multi-frame images like GIF to deny further requests to decode
the image (and undo the GIF behavioral regression [1]).

Add a new API to retrieve the decoding failure state. No mutex
is used when reading the state (not needed).

The comments, "This method is called to populate a discardable
memory owned by Skia" are removed. The deferred decoder writes
to the memory it is told to write to; it is not discardable in
all cases, other parts of the code even say that. GIF does not
currently decode to discardable memory, for example.

DeferredImageDecoder.cpp: detect decoding failures when frames
are requested and early-return on all decoding paths.

ImageFrameGeneratorTest.cpp: update the tests (the frame index
is the first argument of m_generator->decodeAndScale now).

[1] https://codereview.chromium.org/19838002

BUG=501543

Committed: https://crrev.com/edb61309dc83fb68645fbf712a885c5ad5d977b1
Cr-Commit-Position: refs/heads/master@{#366329}
==========

[commit-bot: I haz the power, 2015-12-21 05:13:52]

Patchset 2 (id:??) landed as
https://crrev.com/edb61309dc83fb68645fbf712a885c5ad5d977b1
Cr-Commit-Position: refs/heads/master@{#366329}