Port morphology to SkOpts.

Port morphology to SkOpts.

Nothing too fancy.

Direction enums become enum classes so they don't get all confused.  An
alternative is to create one single Direction enum that both blur and
morphology opts use.

BUG=skia:4117
Committed: https://skia.googlesource.com/skia/+/d029ded92d409a004f2096c78f5a99b524206481

[mtklein_C, 2015-08-04 19:30:29]

mtklein@chromium.org changed reviewers:
+ djsollen@google.com

[djsollen, 2015-08-04 19:38:12]

https://codereview.chromium.org/1267343004/diff/1/src/core/SkOpts.cpp
File src/core/SkOpts.cpp (right):

https://codereview.chromium.org/1267343004/diff/1/src/core/SkOpts.cpp#newcode40
src/core/SkOpts.cpp:40: decltype(box_blur_xx) box_blur_xx =
portable::box_blur_xx;
why not

static const auto blur_x = portable::BlurDirection::kX;

and avoid having to declare those static functions.

https://codereview.chromium.org/1267343004/diff/1/src/core/SkOpts.cpp#newcode44
src/core/SkOpts.cpp:44: decltype(dilate_x) dilate_x = portable::dilate_x;
does decltype really help here?  It seems that Morph would be just fine since we
have a typedef.  Same for the box blurs.

[mtklein_C, 2015-08-04 19:41:48]

https://codereview.chromium.org/1267343004/diff/1/src/core/SkOpts.cpp
File src/core/SkOpts.cpp (right):

https://codereview.chromium.org/1267343004/diff/1/src/core/SkOpts.cpp#newcode40
src/core/SkOpts.cpp:40: decltype(box_blur_xx) box_blur_xx =
portable::box_blur_xx;
On 2015/08/04 19:38:12, djsollen wrote:
> why not
> 
> static const auto blur_x = portable::BlurDirection::kX;
> 
> and avoid having to declare those static functions.

This is mostly a style / layering thing.  I noticed that the implementation
details of those box blurs (that they're a single function template) were
leaking out.  Same deal of course with the morphs.

https://codereview.chromium.org/1267343004/diff/1/src/core/SkOpts.cpp#newcode44
src/core/SkOpts.cpp:44: decltype(dilate_x) dilate_x = portable::dilate_x;
On 2015/08/04 19:38:12, djsollen wrote:
> does decltype really help here?  It seems that Morph would be just fine since
we
> have a typedef.  Same for the box blurs.

Either would be fine.  On the other hand, decltype(x) x = ... works uniformly
for all these function pointers.

[djsollen, 2015-08-04 19:46:43]

lgtm

[djsollen, 2015-08-04 19:47:20]

https://codereview.chromium.org/1267343004/diff/1/gyp/opts.gypi
File gyp/opts.gypi (left):

https://codereview.chromium.org/1267343004/diff/1/gyp/opts.gypi#oldcode10
gyp/opts.gypi:10: '<(skia_src_path)/opts/SkMorphology_opts_none.cpp',
these lists are looking better with each checkin!

[mtklein_C, 2015-08-04 19:51:21]

On 2015/08/04 19:47:20, djsollen wrote:
> https://codereview.chromium.org/1267343004/diff/1/gyp/opts.gypi
> File gyp/opts.gypi (left):
> 
> https://codereview.chromium.org/1267343004/diff/1/gyp/opts.gypi#oldcode10
> gyp/opts.gypi:10: '<(skia_src_path)/opts/SkMorphology_opts_none.cpp',
> these lists are looking better with each checkin!

(I'm hoping we won't need lists by the time I'm done.)

[mtklein_C, 2015-08-04 20:53:37]

The CQ bit was checked by mtklein@chromium.org

[mtklein_C, 2015-08-04 20:53:37]

The patchset sent to the CQ was uploaded after l-g-t-m from djsollen@google.com
Link to the patchset: https://codereview.chromium.org/1267343004/#ps20001
(title: "rebase")

[commit-bot: I haz the power, 2015-08-04 20:53:48]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1267343004/20001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1267343004/20001

[commit-bot: I haz the power, 2015-08-04 21:09:11]

Committed patchset #2 (id:20001) as
https://skia.googlesource.com/skia/+/d029ded92d409a004f2096c78f5a99b524206481