Commit 882ddbd9 authored by fbarchard@google.com's avatar fbarchard@google.com

wrapper for yuv to argb that handles destination misalignment and odd widths by using a memcpy

BUG=none
TEST=none
Review URL: https://webrtc-codereview.appspot.com/364001

git-svn-id: http://libyuv.googlecode.com/svn/trunk@145 16f28f9a-4ce2-e073-06de-1de4eb20be90
parent 0e6ce93c
Name: libyuv
URL: http://code.google.com/p/libyuv/
Version: 144
Version: 145
License: BSD
License File: LICENSE
......
......@@ -1512,12 +1512,16 @@ int I420ToARGB(const uint8* src_y, int src_stride_y,
#if defined(HAS_FASTCONVERTYUVTOARGBROW_NEON)
if (TestCpuFlag(kCpuHasNEON) && IS_ALIGNED(width, 16)) {
FastConvertYUVToARGBRow = FastConvertYUVToARGBRow_NEON;
} else if (TestCpuFlag(kCpuHasNEON)) {
FastConvertYUVToARGBRow = FastConvertYUVToARGBAnyRow_NEON;
} else
#elif defined(HAS_FASTCONVERTYUVTOARGBROW_SSSE3)
if (TestCpuFlag(kCpuHasSSSE3) &&
IS_ALIGNED(width, 8) &&
IS_ALIGNED(dst_argb, 16) && IS_ALIGNED(dst_stride_argb, 16)) {
FastConvertYUVToARGBRow = FastConvertYUVToARGBRow_SSSE3;
} else if (TestCpuFlag(kCpuHasSSSE3)) {
FastConvertYUVToARGBRow = FastConvertYUVToARGBAnyRow_SSSE3;
} else
#endif
{
......
......@@ -38,11 +38,6 @@
#define HAS_FASTCONVERTYUVTOARGBROW_SSSE3
#define HAS_FASTCONVERTYUVTOBGRAROW_SSSE3
#define HAS_FASTCONVERTYUVTOABGRROW_SSSE3
#define HAS_FASTCONVERTYUVTORGB565ROW_SSSE3
#define HAS_FASTCONVERTYUVTOARGB1555ROW_SSSE3
#define HAS_FASTCONVERTYUVTOARGB4444ROW_SSSE3
#define HAS_FASTCONVERTYUVTORGB24ROW_SSSE3
#define HAS_FASTCONVERTYUVTORAWROW_SSSE3
#define HAS_FASTCONVERTYUV444TOARGBROW_SSSE3
#define HAS_REVERSE_ROW_SSSE3
#define HAS_REVERSE_ROW_SSE2
......@@ -67,11 +62,6 @@
#define HAS_FASTCONVERTYUVTOARGBROW_NEON
#define HAS_FASTCONVERTYUVTOBGRAROW_NEON
#define HAS_FASTCONVERTYUVTOABGRROW_NEON
#define HAS_FASTCONVERTYUVTORGB565ROW_NEON
#define HAS_FASTCONVERTYUVTOARGB1555ROW_NEON
#define HAS_FASTCONVERTYUVTOARGB4444ROW_NEON
#define HAS_FASTCONVERTYUVTORGB24ROW_NEON
#define HAS_FASTCONVERTYUVTORAWROW_NEON
#endif
#ifdef __cplusplus
......@@ -91,7 +81,6 @@ typedef unsigned char __attribute__((vector_size(16))) uvec8;
typedef signed short __attribute__((vector_size(16))) vec16;
#endif
void FastConvertYUVToARGBRow_NEON(const uint8* y_buf,
const uint8* u_buf,
const uint8* v_buf,
......@@ -107,31 +96,6 @@ void FastConvertYUVToABGRRow_NEON(const uint8* y_buf,
const uint8* v_buf,
uint8* rgb_buf,
int width);
void FastConvertYUVToRGB565Row_NEON(const uint8* y_buf,
const uint8* u_buf,
const uint8* v_buf,
uint8* rgb_buf,
int width);
void FastConvertYUVToARGB1555Row_NEON(const uint8* y_buf,
const uint8* u_buf,
const uint8* v_buf,
uint8* rgb_buf,
int width);
void FastConvertYUVToARGB4444Row_NEON(const uint8* y_buf,
const uint8* u_buf,
const uint8* v_buf,
uint8* rgb_buf,
int width);
void FastConvertYUVToRGB24Row_NEON(const uint8* y_buf,
const uint8* u_buf,
const uint8* v_buf,
uint8* rgb_buf,
int width);
void FastConvertYUVToRAWRow_NEON(const uint8* y_buf,
const uint8* u_buf,
const uint8* v_buf,
uint8* rgb_buf,
int width);
void ARGBToYRow_SSSE3(const uint8* src_argb, uint8* dst_y, int pix);
void BGRAToYRow_SSSE3(const uint8* src_argb, uint8* dst_y, int pix);
......@@ -181,8 +145,6 @@ void ARGBToRGB565Row_SSE2(const uint8* src_argb, uint8* dst_rgb, int pix);
void ARGBToARGB1555Row_SSE2(const uint8* src_argb, uint8* dst_rgb, int pix);
void ARGBToARGB4444Row_SSE2(const uint8* src_argb, uint8* dst_rgb, int pix);
// ARGBToABGRRow_C is same as ABGRToARGB
// ARGBToBGRARow_C is same as BGRAToARGB
void ARGBToRGB24Row_C(const uint8* src_argb, uint8* dst_rgb, int pix);
void ARGBToRAWRow_C(const uint8* src_argb, uint8* dst_rgb, int pix);
void ARGBToRGB565Row_C(const uint8* src_argb, uint8* dst_rgb, int pix);
......@@ -220,34 +182,6 @@ void FastConvertYToARGBRow_C(const uint8* y_buf,
uint8* rgb_buf,
int width);
void FastConvertYUVToARGBRow_SSE2(const uint8* y_buf,
const uint8* u_buf,
const uint8* v_buf,
uint8* rgb_buf,
int width);
void FastConvertYUVToBGRARow_SSE2(const uint8* y_buf,
const uint8* u_buf,
const uint8* v_buf,
uint8* rgb_buf,
int width);
void FastConvertYUVToABGRRow_SSE2(const uint8* y_buf,
const uint8* u_buf,
const uint8* v_buf,
uint8* rgb_buf,
int width);
void FastConvertYUV444ToARGBRow_SSE2(const uint8* y_buf,
const uint8* u_buf,
const uint8* v_buf,
uint8* rgb_buf,
int width);
void FastConvertYToARGBRow_SSE2(const uint8* y_buf,
uint8* rgb_buf,
int width);
void FastConvertYUVToARGBRow_SSSE3(const uint8* y_buf,
const uint8* u_buf,
const uint8* v_buf,
......@@ -276,6 +210,19 @@ void FastConvertYToARGBRow_SSE2(const uint8* y_buf,
uint8* rgb_buf,
int width);
// 'Any' wrappers use memcpy()
void FastConvertYUVToARGBAnyRow_SSSE3(const uint8* y_buf,
const uint8* u_buf,
const uint8* v_buf,
uint8* rgb_buf,
int width);
void FastConvertYUVToARGBAnyRow_NEON(const uint8* y_buf,
const uint8* u_buf,
const uint8* v_buf,
uint8* rgb_buf,
int width);
#ifdef __cplusplus
} // extern "C"
} // namespace libyuv
......
......@@ -11,6 +11,7 @@
#include "row.h"
#include "libyuv/basic_types.h"
#include <string.h> // For memcpy
#ifdef __cplusplus
namespace libyuv {
......@@ -367,6 +368,39 @@ void ReverseRow_C(const uint8* src, uint8* dst, int width) {
}
}
// Wrappers to handle odd sizes/alignments
#if defined(HAS_FASTCONVERTYUVTOARGBROW_SSSE3)
void FastConvertYUVToARGBAnyRow_SSSE3(const uint8* y_buf,
const uint8* u_buf,
const uint8* v_buf,
uint8* rgb_buf,
int width) {
SIMD_ALIGNED(uint8 row[kMaxStride]);
FastConvertYUVToARGBRow_SSSE3(y_buf,
u_buf,
v_buf,
row,
width);
memcpy(rgb_buf, row, width << 2);
}
#endif
#if defined(HAS_FASTCONVERTYUVTOARGBROW_NEON)
void FastConvertYUVToARGBAnyRow_NEON(const uint8* y_buf,
const uint8* u_buf,
const uint8* v_buf,
uint8* rgb_buf,
int width) {
SIMD_ALIGNED(uint8 row[kMaxStride]);
FastConvertYUVToARGBRow_NEON(y_buf,
u_buf,
v_buf,
row,
width);
memcpy(rgb_buf, row, width << 2);
}
#endif
#ifdef __cplusplus
} // extern "C"
} // namespace libyuv
......
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