ARM Skia NEON patches - 15 - Preparation work for Blitmask optims

ARM Skia NEON patches - 15 - Preparation work for Blitmask optims





Blitmask: copy empty factory functions to a new file

Signed-off-by: Kévin PETIT <kevin.petit@arm.com>;

BUG=

Committed: http://code.google.com/p/skia/source/detail?r=10980

[djsollen, 2013-08-05 17:25:00]

see https://codereview.chromium.org/22229002/ for how we are doing this setup
elsewhere with other arm procs.

https://codereview.chromium.org/21120007/diff/1/src/opts/SkBlitRow_opts_arm.cpp
File src/opts/SkBlitRow_opts_arm.cpp (right):

https://codereview.chromium.org/21120007/diff/1/src/opts/SkBlitRow_opts_arm.c...
src/opts/SkBlitRow_opts_arm.cpp:375: SkBlitMask::ColorProc
SkBlitMask::PlatformColorProcs(SkBitmap::Config dstConfig,
I would either leave these functions here or move them into
SkBlitMask_opts_arm.cpp and have them import the neon header that exposes the
specific neon functions.

{
  if (sk_cpu_arm_has_neon()) {
    sk_blitmask_platform_color_procs_arm_neon(dstConfig, maskFormat, color);
  else {
    return NULL;

}

[kevin.petit.not.used.account, 2013-08-06 10:00:54]

https://codereview.chromium.org/21120007/diff/1/src/opts/SkBlitRow_opts_arm.cpp
File src/opts/SkBlitRow_opts_arm.cpp (right):

https://codereview.chromium.org/21120007/diff/1/src/opts/SkBlitRow_opts_arm.c...
src/opts/SkBlitRow_opts_arm.cpp:375: SkBlitMask::ColorProc
SkBlitMask::PlatformColorProcs(SkBitmap::Config dstConfig,
On 2013/08/05 17:25:00, djsollen wrote:
> I would either leave these functions here or move them into
> SkBlitMask_opts_arm.cpp and have them import the neon header that exposes the
> specific neon functions.

I agree that the current version duplicates the functions which can be unclear.

If you're not bothered by the number of files I'd prefer creating
SkBlitMask_opts_arm.cpp plus SkBlitMask_opts_none.cpp and
SkBlitMask_opts_SSE2.cpp for consistency.

If you are, I'll keep these functions in SkBlitRow_opts_arm.cpp.

The choice is yours.

> {
>   if (sk_cpu_arm_has_neon()) {
>     sk_blitmask_platform_color_procs_arm_neon(dstConfig, maskFormat, color);
>   else {
>     return NULL;
> 
> }

I'm using SK_ARM_NEON_WRAP that allows to deal with the no NEON and always NEON
cases at compile time.

[kevin.petit.not.used.account, 2013-08-08 13:19:41]

I've just uploaded a second revision of the patch that should be a reasonable
compromise.

[kevin.petit.not.used.account, 2013-08-19 09:41:04]

Ping. This one is preventing me from uploading the interesting bits that rely on
it.

[djsollen, 2013-08-19 13:36:22]

lgtm

[kevin.petit.not.used.account, 2013-08-22 13:09:37]

On 2013/08/19 13:36:22, djsollen wrote:
> lgtm

Are there any further actions required from me on this one?

[commit-bot: I haz the power, 2013-08-28 14:57:47]

CQ is trying da patch. Follow status at
https://skia-tree-status.appspot.com/cq/kevin.petit.arm@gmail.com/21120007/6001

[commit-bot: I haz the power, 2013-08-28 15:08:05]

Change committed as 10980