Add MSA optimized I422AlphaToARGBRow_MSA and I422ToRGB24Row_MSA functions

Add MSA optimized I422AlphaToARGBRow_MSA and I422ToRGB24Row_MSA functions

R=fbarchard@google.com
BUG=libyuv:634
Performance Gain (vs C vectorized)
I422AlphaToARGBRow_MSA      : ~1.4x
I422AlphaToARGBRow_Any_MSA  : ~1.4x
I422ToRGB24Row_MSA          : ~4.8x
I422ToRGB24Row_Any_MSA      : ~4.8x

Performance Gain (vs C non-vectorized)
I422AlphaToARGBRow_MSA      : ~7.0x
I422AlphaToARGBRow_Any_MSA  : ~7.0x
I422ToRGB24Row_MSA          : ~7.9x
I422ToRGB24Row_Any_MSA      : ~7.7x

Committed: https://chromium.googlesource.com/libyuv/libyuv/+/532f5708a95b02185db463a481c0315524742dd8

[manojkumar.bhosale, 2016-10-26 06:46:01]

Description was changed from

==========
Add MSA optimized I422AlphaToARGBRow_MSA and I422ToRGB24Row_MSA functions

R=fbarchard@google.com
BUG=libyuv:634
==========

to

==========
Add MSA optimized I422AlphaToARGBRow_MSA and I422ToRGB24Row_MSA functions

R=fbarchard@google.com
BUG=libyuv:634

Performance Gain (vs C vectorized)
I422AlphaToARGBRow_MSA      : ~1.4x
I422AlphaToARGBRow_Any_MSA  : ~1.4x
I422ToRGB24Row_MSA          : ~4.8x
I422ToRGB24Row_Any_MSA      : ~4.8x

Performance Gain (vs C non-vectorized)
I422AlphaToARGBRow_MSA      : ~7.0x
I422AlphaToARGBRow_Any_MSA  : ~7.0x
I422ToRGB24Row_MSA          : ~7.9x
I422ToRGB24Row_Any_MSA      : ~7.7x
==========

[manojkumar.bhosale, 2016-10-26 06:46:01]

manojkumar.bhosale@imgtec.com changed reviewers:
+ kjellander@chromium.org

[manojkumar.bhosale, 2016-10-26 06:46:46]

Updated with reviewers, cc list and performance gain numbers

[fbarchard1, 2016-10-26 17:56:31]

lgtm

just nits about names and indent

https://codereview.chromium.org/2454433003/diff/1/source/convert_from.cc
File source/convert_from.cc (right):

https://codereview.chromium.org/2454433003/diff/1/source/convert_from.cc#newc...
source/convert_from.cc:563: I422ToRGB24Row = I422ToRGB24Row_Any_MSA;
when doing RGB24, its good to do RAW at the same time.  Its the same, but with
R,G,B memory ordering, vs RGB24 is B,G,R.
Neither are commonly used, but they come as a pair.

https://codereview.chromium.org/2454433003/diff/1/source/row_msa.cc
File source/row_msa.cc (right):

https://codereview.chromium.org/2454433003/diff/1/source/row_msa.cc#newcode23
source/row_msa.cc:23: #define LOAD_I422(psrc_y, psrc_u, psrc_v, out_y, out_u,
out_v) {   \
suggest READYUV422 as name, for consistency with row_gcc.cc macros.

https://codereview.chromium.org/2454433003/diff/1/source/row_msa.cc#newcode274
source/row_msa.cc:274: src3 = (v16u8) __msa_ilvr_b((v16i8) src3, (v16i8) src3);
consider a macro for STOREARGB since it will come up in all YUV (9+ formats) to
ARGB.

https://codereview.chromium.org/2454433003/diff/1/source/row_msa.cc#newcode301
source/row_msa.cc:301: { 26, 6, 7, 27, 8, 9, 28, 10, 11, 29, 12, 13, 30, 14, 15,
31 };
indent should be 4 from line above.

[fbarchard1, 2016-10-26 18:12:24]

Description was changed from

==========
Add MSA optimized I422AlphaToARGBRow_MSA and I422ToRGB24Row_MSA functions

R=fbarchard@google.com
BUG=libyuv:634

Performance Gain (vs C vectorized)
I422AlphaToARGBRow_MSA      : ~1.4x
I422AlphaToARGBRow_Any_MSA  : ~1.4x
I422ToRGB24Row_MSA          : ~4.8x
I422ToRGB24Row_Any_MSA      : ~4.8x

Performance Gain (vs C non-vectorized)
I422AlphaToARGBRow_MSA      : ~7.0x
I422AlphaToARGBRow_Any_MSA  : ~7.0x
I422ToRGB24Row_MSA          : ~7.9x
I422ToRGB24Row_Any_MSA      : ~7.7x
==========

to

==========
Add MSA optimized I422AlphaToARGBRow_MSA and I422ToRGB24Row_MSA functions

R=fbarchard@google.com
BUG=libyuv:634

Performance Gain (vs C vectorized)
I422AlphaToARGBRow_MSA      : ~1.4x
I422AlphaToARGBRow_Any_MSA  : ~1.4x
I422ToRGB24Row_MSA          : ~4.8x
I422ToRGB24Row_Any_MSA      : ~4.8x

Performance Gain (vs C non-vectorized)
I422AlphaToARGBRow_MSA      : ~7.0x
I422AlphaToARGBRow_Any_MSA  : ~7.0x
I422ToRGB24Row_MSA          : ~7.9x
I422ToRGB24Row_Any_MSA      : ~7.7x

Committed:
https://chromium.googlesource.com/libyuv/libyuv/+/532f5708a95b02185db463a481c...
==========

[fbarchard1, 2016-10-26 18:12:26]

Committed patchset #1 (id:1) manually as
532f5708a95b02185db463a481c0315524742dd8 (presubmit successful).