YUV to RGB for x64 use registers instead of memory.

YUV to RGB for x64 use registers instead of memory.

On Arm the YVU to RGB conversions move constants into registers.
This change does the same for 64 bit intel builds where additional
registers are available.
The AVX2 saves 3 instructions by because the 2nd argument needs to be a register, so a vmovdqu was avoided.

x64 builds using memory:
AVX2  I420ToARGB_Opt (3059 ms)
SSSE3 I420ToARGB_Opt (3959 ms)

Now using registers
AVX2  I420ToARGB_Opt (2906 ms)
SSSE3 I420ToARGB_Opt (3928 ms)

TBR=harryjin@google.com
BUG=libyuv:520
Committed: https://chromium.googlesource.com/libyuv/libyuv/+/431cb3667a346136189952ff93974f4ac9e79d78

[fbarchard, 2015-11-04 20:10:11]

Description was changed from

==========
YUV to RGB for x64 use registers instead of memory.

On Arm the YVU to RGB conversions move constants into registers.
This change does the same for 64 bit intel builds where additional
registers are available.

R=harryjin@google.com
BUG=none
==========

to

==========
YUV to RGB for x64 use registers instead of memory.

On Arm the YVU to RGB conversions move constants into registers.
This change does the same for 64 bit intel builds where additional
registers are available.

R=harryjin@google.com,xhwang@chromium.org
BUG=libyuv:520
==========

[fbarchard, 2015-11-04 20:32:47]

fbarchard@chromium.org changed reviewers:
+ fbarchard@chromium.org

[fbarchard, 2015-11-04 20:32:48]

no change in behavior for 32 bit. For 64 bit

Was using memory
AVX2  I420ToARGB_Opt (3059 ms)
SSSE3 I420ToARGB_Opt (3959 ms)

Now using registers
AVX2  I420ToARGB_Opt (2906 ms)
SSSE3 I420ToARGB_Opt (3928 ms)

[fbarchard, 2015-11-04 20:34:20]

Description was changed from

==========
YUV to RGB for x64 use registers instead of memory.

On Arm the YVU to RGB conversions move constants into registers.
This change does the same for 64 bit intel builds where additional
registers are available.

R=harryjin@google.com,xhwang@chromium.org
BUG=libyuv:520
==========

to

==========
YUV to RGB for x64 use registers instead of memory.

On Arm the YVU to RGB conversions move constants into registers.
This change does the same for 64 bit intel builds where additional
registers are available.
The AVX2 saves 3 instructions by because the 2nd argument needs to be a
register, so a vmovdqu was avoided.

x64 builds using memory:
AVX2  I420ToARGB_Opt (3059 ms)
SSSE3 I420ToARGB_Opt (3959 ms)

Now using registers
AVX2  I420ToARGB_Opt (2906 ms)
SSSE3 I420ToARGB_Opt (3928 ms)

R=harryjin@google.com,xhwang@chromium.org
BUG=libyuv:520
==========

[fbarchard, 2015-11-05 00:15:52]

Description was changed from

==========
YUV to RGB for x64 use registers instead of memory.

On Arm the YVU to RGB conversions move constants into registers.
This change does the same for 64 bit intel builds where additional
registers are available.
The AVX2 saves 3 instructions by because the 2nd argument needs to be a
register, so a vmovdqu was avoided.

x64 builds using memory:
AVX2  I420ToARGB_Opt (3059 ms)
SSSE3 I420ToARGB_Opt (3959 ms)

Now using registers
AVX2  I420ToARGB_Opt (2906 ms)
SSSE3 I420ToARGB_Opt (3928 ms)

R=harryjin@google.com,xhwang@chromium.org
BUG=libyuv:520
==========

to

==========
YUV to RGB for x64 use registers instead of memory.

On Arm the YVU to RGB conversions move constants into registers.
This change does the same for 64 bit intel builds where additional
registers are available.
The AVX2 saves 3 instructions by because the 2nd argument needs to be a
register, so a vmovdqu was avoided.

x64 builds using memory:
AVX2  I420ToARGB_Opt (3059 ms)
SSSE3 I420ToARGB_Opt (3959 ms)

Now using registers
AVX2  I420ToARGB_Opt (2906 ms)
SSSE3 I420ToARGB_Opt (3928 ms)

TBR=harryjin@google.com
BUG=libyuv:520
==========

[fbarchard1, 2015-11-05 00:16:22]

Committed patchset #2 (id:20001) manually as
431cb3667a346136189952ff93974f4ac9e79d78 (presubmit successful).