Make SkSwizzler::Fill() support 565

Make SkSwizzler::Fill() support 565

BUG=skia:4135
Committed: https://skia.googlesource.com/skia/+/fdb788cbfcdccac65d8d1f535920c7f2fdba1ead

[msarett, 2015-07-29 17:10:41]

Patchset #1 (id:1) has been deleted

[msarett, 2015-07-29 17:11:31]

msarett@google.com changed reviewers:
+ emmaleer@google.com, scroggo@google.com

[scroggo, 2015-07-29 17:15:15]

lgtm

https://codereview.chromium.org/1267543002/diff/20001/src/codec/SkJpegCodec.cpp
File src/codec/SkJpegCodec.cpp (right):

https://codereview.chromium.org/1267543002/diff/20001/src/codec/SkJpegCodec.c...
src/codec/SkJpegCodec.cpp:358: // will be used.  Conveniently, these are zeros,
which is the
nit: if kYes_ZeroInitialized, we could skip this.

[msarett, 2015-07-29 17:29:39]

https://codereview.chromium.org/1267543002/diff/20001/src/codec/SkJpegCodec.cpp
File src/codec/SkJpegCodec.cpp (right):

https://codereview.chromium.org/1267543002/diff/20001/src/codec/SkJpegCodec.c...
src/codec/SkJpegCodec.cpp:358: // will be used.  Conveniently, these are zeros,
which is the
On 2015/07/29 17:15:14, scroggo wrote:
> nit: if kYes_ZeroInitialized, we could skip this.

Done.

[msarett, 2015-07-29 17:29:46]

The CQ bit was checked by msarett@google.com

[msarett, 2015-07-29 17:29:46]

The patchset sent to the CQ was uploaded after l-g-t-m from scroggo@google.com
Link to the patchset: https://codereview.chromium.org/1267543002/#ps40001
(title: "Check if pixel memory is zero initialized")

[commit-bot: I haz the power, 2015-07-29 17:30:03]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1267543002/40001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1267543002/40001

[commit-bot: I haz the power, 2015-07-29 17:37:32]

Committed patchset #2 (id:40001) as
https://skia.googlesource.com/skia/+/fdb788cbfcdccac65d8d1f535920c7f2fdba1ead

[scroggo, 2015-08-03 18:41:25]

https://codereview.chromium.org/1267543002/diff/40001/src/codec/SkJpegCodec.cpp
File src/codec/SkJpegCodec.cpp (right):

https://codereview.chromium.org/1267543002/diff/40001/src/codec/SkJpegCodec.c...
src/codec/SkJpegCodec.cpp:467: const SkCodec::Options&    fOpts;
I missed this when I first reviewed, but mention it in
https://codereview.chromium.org/1267583002/ (which also fixes it) - making this
a reference is not a good idea. We do not know the lifetime of the object passed
in, so we cannot be sure that the Options object we are referring to will still
exist when we need it.

We actually *do* know of an example when this does not work - if the client
called getScanlineDecoder with a NULL Options object, getScanlineDecoder puts
one on the stack and passes it to onGetScanlineDecoder. So as soon as the client
receives the JpegScanlineDecoder, the object they have is no longer valid.