make info real in SkPixelRef, and add bitmap::asImageInfo

make info real in SkPixelRef, and add bitmap::asImageInfo

BUG=
R=scroggo@google.com

Committed: https://code.google.com/p/skia/source/detail?r=12586

[scroggo, 2013-12-09 21:33:41]

It appears this is mostly a subset of https://codereview.chromium.org/68973005/,
which was reverted an odd number of times?

https://codereview.chromium.org/108663004/diff/20001/src/image/SkImagePriv.h
File src/image/SkImagePriv.h (right):

https://codereview.chromium.org/108663004/diff/20001/src/image/SkImagePriv.h#...
src/image/SkImagePriv.h:47: return SkAlign4(info.minRowBytes());
Why is this function different (from minRowBytes()) that it needs to be 4 byte
aligned? Can you document that difference?

https://codereview.chromium.org/108663004/diff/20001/tests/PixelRefTest.cpp
File tests/PixelRefTest.cpp (right):

https://codereview.chromium.org/108663004/diff/20001/tests/PixelRefTest.cpp#n...
tests/PixelRefTest.cpp:14: { SkBitmap::kNo_Config,         kPremul_SkAlphaType, 
  kPMColor_SkColorType,   false },
Why isn't kIndex8_Config tested here?

[reed1, 2013-12-09 21:46:13]

Yes, this is a subset, hopeful containing non-chrome-breaking aspects, but
useful things as asImageInfo. It was getting just to hard to land the other CL
as is, since it was changing several things/apis at once. Now that CL will be
rebased to just focus on a new signature for lockPixels (returning rowbytes).

https://codereview.chromium.org/108663004/diff/20001/src/image/SkImagePriv.h
File src/image/SkImagePriv.h (right):

https://codereview.chromium.org/108663004/diff/20001/src/image/SkImagePriv.h#...
src/image/SkImagePriv.h:47: return SkAlign4(info.minRowBytes());
On 2013/12/09 21:33:41, scroggo wrote:
> Why is this function different (from minRowBytes()) that it needs to be 4 byte
> aligned? Can you document that difference?

Don't know. I was just keeping the old behavior.

https://codereview.chromium.org/108663004/diff/20001/tests/PixelRefTest.cpp
File tests/PixelRefTest.cpp (right):

https://codereview.chromium.org/108663004/diff/20001/tests/PixelRefTest.cpp#n...
tests/PixelRefTest.cpp:14: { SkBitmap::kNo_Config,         kPremul_SkAlphaType, 
  kPMColor_SkColorType,   false },
On 2013/12/09 21:33:41, scroggo wrote:
> Why isn't kIndex8_Config tested here?

Done.

[reed1, 2013-12-09 21:53:50]

ptal

[scroggo, 2013-12-09 21:57:06]

lgtm

[reed1, 2013-12-09 22:01:12]

Committed patchset #5 manually as r12586 (presubmit successful).