Commit 304a611d authored by fbarchard@google.com's avatar fbarchard@google.com

Unaligned version of ABGRToARGB for SSSE3

BUG=196
TESTED=libyuv_unittest --gtest_filter=*ABGRToARGB*
Review URL: https://webrtc-codereview.appspot.com/1166006

git-svn-id: http://libyuv.googlecode.com/svn/trunk@595 16f28f9a-4ce2-e073-06de-1de4eb20be90
parent 51398e0b
Name: libyuv
URL: http://code.google.com/p/libyuv/
Version: 594
Version: 595
License: BSD
License File: LICENSE
......
......@@ -635,6 +635,8 @@ void ARGBSetRows_C(uint8* dst, uint32 v32, int width, int dst_stride,
void BGRAToARGBRow_SSSE3(const uint8* src_bgra, uint8* dst_argb, int pix);
void ABGRToARGBRow_SSSE3(const uint8* src_abgr, uint8* dst_argb, int pix);
void ABGRToARGBRow_Unaligned_SSSE3(const uint8* src_abgr, uint8* dst_argb,
int pix);
void RGBAToARGBRow_SSSE3(const uint8* src_rgba, uint8* dst_argb, int pix);
void RGB24ToARGBRow_SSSE3(const uint8* src_rgb24, uint8* dst_argb, int pix);
void RAWToARGBRow_SSSE3(const uint8* src_raw, uint8* dst_argb, int pix);
......
......@@ -11,6 +11,6 @@
#ifndef INCLUDE_LIBYUV_VERSION_H_ // NOLINT
#define INCLUDE_LIBYUV_VERSION_H_
#define LIBYUV_VERSION 594
#define LIBYUV_VERSION 595
#endif // INCLUDE_LIBYUV_VERSION_H_ NOLINT
......@@ -350,9 +350,13 @@ int ABGRToARGB(const uint8* src_abgr, int src_stride_abgr,
void (*ABGRToARGBRow)(const uint8* src_abgr, uint8* dst_argb, int pix) =
ABGRToARGBRow_C;
#if defined(HAS_ABGRTOARGBROW_SSSE3)
if (TestCpuFlag(kCpuHasSSSE3) && IS_ALIGNED(width, 4) &&
IS_ALIGNED(src_abgr, 16) && IS_ALIGNED(src_stride_abgr, 16) &&
IS_ALIGNED(dst_argb, 16) && IS_ALIGNED(dst_stride_argb, 16)) {
if (TestCpuFlag(kCpuHasSSSE3) && IS_ALIGNED(width, 4)) {
// TODO(fbarchard): Port to posix.
#if defined(_M_IX86)
ABGRToARGBRow = ABGRToARGBRow_Unaligned_SSSE3;
#endif
if (IS_ALIGNED(src_abgr, 16) && IS_ALIGNED(src_stride_abgr, 16) &&
IS_ALIGNED(dst_argb, 16) && IS_ALIGNED(dst_stride_argb, 16))
ABGRToARGBRow = ABGRToARGBRow_SSSE3;
}
#elif defined(HAS_ABGRTOARGBROW_NEON)
......
......@@ -251,6 +251,28 @@ void BGRAToARGBRow_SSSE3(const uint8* src_bgra, uint8* dst_argb, int pix) {
}
}
__declspec(naked) __declspec(align(16))
void ABGRToARGBRow_Unaligned_SSSE3(const uint8* src_abgr, uint8* dst_argb,
int pix) {
__asm {
mov eax, [esp + 4] // src_abgr
mov edx, [esp + 8] // dst_argb
mov ecx, [esp + 12] // pix
movdqa xmm5, kShuffleMaskABGRToARGB
sub edx, eax
align 16
convertloop:
movdqu xmm0, [eax]
pshufb xmm0, xmm5
sub ecx, 4
movdqu [eax + edx], xmm0
lea eax, [eax + 16]
jg convertloop
ret
}
}
__declspec(naked) __declspec(align(16))
void ABGRToARGBRow_SSSE3(const uint8* src_abgr, uint8* dst_argb, int pix) {
__asm {
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment