diff options
Diffstat (limited to 'files/include/libyuv')
| -rw-r--r-- | files/include/libyuv/convert.h | 123 | ||||
| -rw-r--r-- | files/include/libyuv/convert_argb.h | 126 | ||||
| -rw-r--r-- | files/include/libyuv/convert_from_argb.h | 51 | ||||
| -rw-r--r-- | files/include/libyuv/cpu_id.h | 7 | ||||
| -rw-r--r-- | files/include/libyuv/planar_functions.h | 56 | ||||
| -rw-r--r-- | files/include/libyuv/rotate.h | 64 | ||||
| -rw-r--r-- | files/include/libyuv/rotate_row.h | 45 | ||||
| -rw-r--r-- | files/include/libyuv/row.h | 1077 | ||||
| -rw-r--r-- | files/include/libyuv/scale_row.h | 43 | ||||
| -rw-r--r-- | files/include/libyuv/version.h | 2 |
10 files changed, 1394 insertions, 200 deletions
diff --git a/files/include/libyuv/convert.h b/files/include/libyuv/convert.h index 46d37159..88619a4f 100644 --- a/files/include/libyuv/convert.h +++ b/files/include/libyuv/convert.h @@ -151,6 +151,33 @@ int MM21ToI420(const uint8_t* src_y, int width, int height); +// Convert MM21 to YUY2 +LIBYUV_API +int MM21ToYUY2(const uint8_t* src_y, + int src_stride_y, + const uint8_t* src_uv, + int src_stride_uv, + uint8_t* dst_yuy2, + int dst_stride_yuy2, + int width, + int height); + +// Convert MT2T to P010 +// Note that src_y and src_uv point to packed 10-bit values, so the Y plane will +// be 10 / 8 times the dimensions of the image. Also for this reason, +// src_stride_y and src_stride_uv are given in bytes. +LIBYUV_API +int MT2TToP010(const uint8_t* src_y, + int src_stride_y, + const uint8_t* src_uv, + int src_stride_uv, + uint16_t* dst_y, + int dst_stride_y, + uint16_t* dst_uv, + int dst_stride_uv, + int width, + int height); + // Convert I422 to NV21. LIBYUV_API int I422ToNV21(const uint8_t* src_y, @@ -272,6 +299,23 @@ int I210ToI422(const uint16_t* src_y, int width, int height); +#define H410ToH420 I410ToI420 +LIBYUV_API +int I410ToI420(const uint16_t* src_y, + int src_stride_y, + const uint16_t* src_u, + int src_stride_u, + const uint16_t* src_v, + int src_stride_v, + uint8_t* dst_y, + int dst_stride_y, + uint8_t* dst_u, + int dst_stride_u, + uint8_t* dst_v, + int dst_stride_v, + int width, + int height); + #define H410ToH444 I410ToI444 LIBYUV_API int I410ToI444(const uint16_t* src_y, @@ -323,6 +367,23 @@ int I212ToI422(const uint16_t* src_y, int width, int height); +#define H212ToH420 I212ToI420 +LIBYUV_API +int I212ToI420(const uint16_t* src_y, + int src_stride_y, + const uint16_t* src_u, + int src_stride_u, + const uint16_t* src_v, + int src_stride_v, + uint8_t* dst_y, + int dst_stride_y, + uint8_t* dst_u, + int dst_stride_u, + uint8_t* dst_v, + int dst_stride_v, + int width, + int height); + #define H412ToH444 I412ToI444 LIBYUV_API int I412ToI444(const uint16_t* src_y, @@ -340,6 +401,23 @@ int I412ToI444(const uint16_t* src_y, int width, int height); +#define H412ToH420 I412ToI420 +LIBYUV_API +int I412ToI420(const uint16_t* src_y, + int src_stride_y, + const uint16_t* src_u, + int src_stride_u, + const uint16_t* src_v, + int src_stride_v, + uint8_t* dst_y, + int dst_stride_y, + uint8_t* dst_u, + int dst_stride_u, + uint8_t* dst_v, + int dst_stride_v, + int width, + int height); + #define I412ToI012 I410ToI010 #define H410ToH010 I410ToI010 #define H412ToH012 I410ToI010 @@ -560,6 +638,36 @@ int NV16ToNV24(const uint8_t* src_y, int width, int height); +// Convert P010 to I010. +LIBYUV_API +int P010ToI010(const uint16_t* src_y, + int src_stride_y, + const uint16_t* src_uv, + int src_stride_uv, + uint16_t* dst_y, + int dst_stride_y, + uint16_t* dst_u, + int dst_stride_u, + uint16_t* dst_v, + int dst_stride_v, + int width, + int height); + +// Convert P012 to I012. +LIBYUV_API +int P012ToI012(const uint16_t* src_y, + int src_stride_y, + const uint16_t* src_uv, + int src_stride_uv, + uint16_t* dst_y, + int dst_stride_y, + uint16_t* dst_u, + int dst_stride_u, + uint16_t* dst_v, + int dst_stride_v, + int width, + int height); + // Convert P010 to P410. LIBYUV_API int P010ToP410(const uint16_t* src_y, @@ -677,6 +785,21 @@ int ARGBToI420(const uint8_t* src_argb, int width, int height); +// Convert ARGB to I420 with Alpha +LIBYUV_API +int ARGBToI420Alpha(const uint8_t* src_argb, + int src_stride_argb, + uint8_t* dst_y, + int dst_stride_y, + uint8_t* dst_u, + int dst_stride_u, + uint8_t* dst_v, + int dst_stride_v, + uint8_t* dst_a, + int dst_stride_a, + int width, + int height); + // BGRA little endian (argb in memory) to I420. LIBYUV_API int BGRAToI420(const uint8_t* src_bgra, diff --git a/files/include/libyuv/convert_argb.h b/files/include/libyuv/convert_argb.h index f66d20ce..35eeac9b 100644 --- a/files/include/libyuv/convert_argb.h +++ b/files/include/libyuv/convert_argb.h @@ -67,6 +67,8 @@ LIBYUV_API extern const struct YuvConstants kYvuV2020Constants; // BT.2020 full I210ToAR30Matrix(a, b, e, f, c, d, g, h, i##VU, j, k) #define I410ToAB30Matrix(a, b, c, d, e, f, g, h, i, j, k) \ I410ToAR30Matrix(a, b, e, f, c, d, g, h, i##VU, j, k) +#define I012ToAB30Matrix(a, b, c, d, e, f, g, h, i, j, k) \ + I012ToAR30Matrix(a, b, e, f, c, d, g, h, i##VU, j, k) #define I420AlphaToABGRMatrix(a, b, c, d, e, f, g, h, i, j, k, l, m, n) \ I420AlphaToARGBMatrix(a, b, e, f, c, d, g, h, i, j, k##VU, l, m, n) #define I422AlphaToABGRMatrix(a, b, c, d, e, f, g, h, i, j, k, l, m, n) \ @@ -404,6 +406,32 @@ int U444ToABGR(const uint8_t* src_y, int width, int height); +// Convert I444 to RGB24. +LIBYUV_API +int I444ToRGB24(const uint8_t* src_y, + int src_stride_y, + const uint8_t* src_u, + int src_stride_u, + const uint8_t* src_v, + int src_stride_v, + uint8_t* dst_rgb24, + int dst_stride_rgb24, + int width, + int height); + +// Convert I444 to RAW. +LIBYUV_API +int I444ToRAW(const uint8_t* src_y, + int src_stride_y, + const uint8_t* src_u, + int src_stride_u, + const uint8_t* src_v, + int src_stride_v, + uint8_t* dst_raw, + int dst_stride_raw, + int width, + int height); + // Convert I010 to ARGB. LIBYUV_API int I010ToARGB(const uint16_t* src_y, @@ -1312,6 +1340,32 @@ int J420ToRAW(const uint8_t* src_y, int width, int height); +// Convert I422 to RGB24. +LIBYUV_API +int I422ToRGB24(const uint8_t* src_y, + int src_stride_y, + const uint8_t* src_u, + int src_stride_u, + const uint8_t* src_v, + int src_stride_v, + uint8_t* dst_rgb24, + int dst_stride_rgb24, + int width, + int height); + +// Convert I422 to RAW. +LIBYUV_API +int I422ToRAW(const uint8_t* src_y, + int src_stride_y, + const uint8_t* src_u, + int src_stride_u, + const uint8_t* src_v, + int src_stride_v, + uint8_t* dst_raw, + int dst_stride_raw, + int width, + int height); + LIBYUV_API int I420ToRGB565(const uint8_t* src_y, int src_stride_y, @@ -1495,6 +1549,20 @@ int I444ToARGBMatrix(const uint8_t* src_y, int width, int height); +// Convert I444 to RGB24 with matrix. +LIBYUV_API +int I444ToRGB24Matrix(const uint8_t* src_y, + int src_stride_y, + const uint8_t* src_u, + int src_stride_u, + const uint8_t* src_v, + int src_stride_v, + uint8_t* dst_rgb24, + int dst_stride_rgb24, + const struct YuvConstants* yuvconstants, + int width, + int height); + // Convert 10 bit 420 YUV to ARGB with matrix. LIBYUV_API int I010ToAR30Matrix(const uint16_t* src_y, @@ -1893,6 +1961,20 @@ int I420ToRGB24Matrix(const uint8_t* src_y, int width, int height); +// Convert I422 to RGB24 with matrix. +LIBYUV_API +int I422ToRGB24Matrix(const uint8_t* src_y, + int src_stride_y, + const uint8_t* src_u, + int src_stride_u, + const uint8_t* src_v, + int src_stride_v, + uint8_t* dst_rgb24, + int dst_stride_rgb24, + const struct YuvConstants* yuvconstants, + int width, + int height); + // Convert I420 to RGB565 with specified color matrix. LIBYUV_API int I420ToRGB565Matrix(const uint8_t* src_y, @@ -1907,6 +1989,20 @@ int I420ToRGB565Matrix(const uint8_t* src_y, int width, int height); +// Convert I422 to RGB565 with specified color matrix. +LIBYUV_API +int I422ToRGB565Matrix(const uint8_t* src_y, + int src_stride_y, + const uint8_t* src_u, + int src_stride_u, + const uint8_t* src_v, + int src_stride_v, + uint8_t* dst_rgb565, + int dst_stride_rgb565, + const struct YuvConstants* yuvconstants, + int width, + int height); + // Convert I420 to AR30 with matrix. LIBYUV_API int I420ToAR30Matrix(const uint8_t* src_y, @@ -1961,6 +2057,36 @@ int I422ToARGBMatrixFilter(const uint8_t* src_y, int height, enum FilterMode filter); +// Convert I422 to RGB24 with matrix and UV filter mode. +LIBYUV_API +int I422ToRGB24MatrixFilter(const uint8_t* src_y, + int src_stride_y, + const uint8_t* src_u, + int src_stride_u, + const uint8_t* src_v, + int src_stride_v, + uint8_t* dst_rgb24, + int dst_stride_rgb24, + const struct YuvConstants* yuvconstants, + int width, + int height, + enum FilterMode filter); + +// Convert I420 to RGB24 with matrix and UV filter mode. +LIBYUV_API +int I420ToRGB24MatrixFilter(const uint8_t* src_y, + int src_stride_y, + const uint8_t* src_u, + int src_stride_u, + const uint8_t* src_v, + int src_stride_v, + uint8_t* dst_rgb24, + int dst_stride_rgb24, + const struct YuvConstants* yuvconstants, + int width, + int height, + enum FilterMode filter); + // Convert I010 to AR30 with matrix and UV filter mode. LIBYUV_API int I010ToAR30MatrixFilter(const uint16_t* src_y, diff --git a/files/include/libyuv/convert_from_argb.h b/files/include/libyuv/convert_from_argb.h index 2a488838..ff2a581a 100644 --- a/files/include/libyuv/convert_from_argb.h +++ b/files/include/libyuv/convert_from_argb.h @@ -209,10 +209,10 @@ int ARGBToJ420(const uint8_t* src_argb, int src_stride_argb, uint8_t* dst_yj, int dst_stride_yj, - uint8_t* dst_u, - int dst_stride_u, - uint8_t* dst_v, - int dst_stride_v, + uint8_t* dst_uj, + int dst_stride_uj, + uint8_t* dst_vj, + int dst_stride_vj, int width, int height); @@ -222,10 +222,10 @@ int ARGBToJ422(const uint8_t* src_argb, int src_stride_argb, uint8_t* dst_yj, int dst_stride_yj, - uint8_t* dst_u, - int dst_stride_u, - uint8_t* dst_v, - int dst_stride_v, + uint8_t* dst_uj, + int dst_stride_uj, + uint8_t* dst_vj, + int dst_stride_vj, int width, int height); @@ -238,6 +238,41 @@ int ARGBToJ400(const uint8_t* src_argb, int width, int height); +// Convert ABGR to J420. (JPeg full range I420). +LIBYUV_API +int ABGRToJ420(const uint8_t* src_abgr, + int src_stride_abgr, + uint8_t* dst_yj, + int dst_stride_yj, + uint8_t* dst_uj, + int dst_stride_uj, + uint8_t* dst_vj, + int dst_stride_vj, + int width, + int height); + +// Convert ABGR to J422. +LIBYUV_API +int ABGRToJ422(const uint8_t* src_abgr, + int src_stride_abgr, + uint8_t* dst_yj, + int dst_stride_yj, + uint8_t* dst_uj, + int dst_stride_uj, + uint8_t* dst_vj, + int dst_stride_vj, + int width, + int height); + +// Convert ABGR to J400. (JPeg full range). +LIBYUV_API +int ABGRToJ400(const uint8_t* src_abgr, + int src_stride_abgr, + uint8_t* dst_yj, + int dst_stride_yj, + int width, + int height); + // Convert RGBA to J400. (JPeg full range). LIBYUV_API int RGBAToJ400(const uint8_t* src_rgba, diff --git a/files/include/libyuv/cpu_id.h b/files/include/libyuv/cpu_id.h index fb90c6c7..203f7e0d 100644 --- a/files/include/libyuv/cpu_id.h +++ b/files/include/libyuv/cpu_id.h @@ -55,6 +55,11 @@ static const int kCpuHasLOONGARCH = 0x2000000; static const int kCpuHasLSX = 0x4000000; static const int kCpuHasLASX = 0x8000000; +// These flags are only valid on RISCV processors. +static const int kCpuHasRISCV = 0x10000000; +static const int kCpuHasRVV = 0x20000000; +static const int kCpuHasRVVZVFH = 0x40000000; + // Optional init function. TestCpuFlag does an auto-init. // Returns cpu_info flags. LIBYUV_API @@ -78,6 +83,8 @@ LIBYUV_API int ArmCpuCaps(const char* cpuinfo_name); LIBYUV_API int MipsCpuCaps(const char* cpuinfo_name); +LIBYUV_API +int RiscvCpuCaps(const char* cpuinfo_name); // For testing, allow CPU flags to be disabled. // ie MaskCpuFlags(~kCpuHasSSSE3) to disable SSSE3. diff --git a/files/include/libyuv/planar_functions.h b/files/include/libyuv/planar_functions.h index 1ef2256b..154f2f21 100644 --- a/files/include/libyuv/planar_functions.h +++ b/files/include/libyuv/planar_functions.h @@ -85,13 +85,23 @@ void SetPlane(uint8_t* dst_y, // Convert a plane of tiles of 16 x H to linear. LIBYUV_API -void DetilePlane(const uint8_t* src_y, - int src_stride_y, - uint8_t* dst_y, - int dst_stride_y, - int width, - int height, - int tile_height); +int DetilePlane(const uint8_t* src_y, + int src_stride_y, + uint8_t* dst_y, + int dst_stride_y, + int width, + int height, + int tile_height); + +// Convert a plane of 16 bit tiles of 16 x H to linear. +LIBYUV_API +int DetilePlane_16(const uint16_t* src_y, + int src_stride_y, + uint16_t* dst_y, + int dst_stride_y, + int width, + int height, + int tile_height); // Convert a UV plane of tiles of 16 x H into linear U and V planes. LIBYUV_API @@ -105,6 +115,18 @@ void DetileSplitUVPlane(const uint8_t* src_uv, int height, int tile_height); +// Convert a Y and UV plane of tiles into interlaced YUY2. +LIBYUV_API +void DetileToYUY2(const uint8_t* src_y, + int src_stride_y, + const uint8_t* src_uv, + int src_stride_uv, + uint8_t* dst_yuy2, + int dst_stride_yuy2, + int width, + int height, + int tile_height); + // Split interleaved UV plane into separate U and V planes. LIBYUV_API void SplitUVPlane(const uint8_t* src_uv, @@ -370,7 +392,26 @@ int I210Copy(const uint16_t* src_y, int width, int height); +// Copy I410 to I410. +#define I410ToI410 I410Copy +LIBYUV_API +int I410Copy(const uint16_t* src_y, + int src_stride_y, + const uint16_t* src_u, + int src_stride_u, + const uint16_t* src_v, + int src_stride_v, + uint16_t* dst_y, + int dst_stride_y, + uint16_t* dst_u, + int dst_stride_u, + uint16_t* dst_v, + int dst_stride_v, + int width, + int height); + // Copy NV12. Supports inverting. +LIBYUV_API int NV12Copy(const uint8_t* src_y, int src_stride_y, const uint8_t* src_uv, @@ -383,6 +424,7 @@ int NV12Copy(const uint8_t* src_y, int height); // Copy NV21. Supports inverting. +LIBYUV_API int NV21Copy(const uint8_t* src_y, int src_stride_y, const uint8_t* src_vu, diff --git a/files/include/libyuv/rotate.h b/files/include/libyuv/rotate.h index 684ed5e6..37460c4a 100644 --- a/files/include/libyuv/rotate.h +++ b/files/include/libyuv/rotate.h @@ -85,6 +85,60 @@ int I444Rotate(const uint8_t* src_y, int height, enum RotationMode mode); +// Rotate I010 frame. +LIBYUV_API +int I010Rotate(const uint16_t* src_y, + int src_stride_y, + const uint16_t* src_u, + int src_stride_u, + const uint16_t* src_v, + int src_stride_v, + uint16_t* dst_y, + int dst_stride_y, + uint16_t* dst_u, + int dst_stride_u, + uint16_t* dst_v, + int dst_stride_v, + int width, + int height, + enum RotationMode mode); + +// Rotate I210 frame. +LIBYUV_API +int I210Rotate(const uint16_t* src_y, + int src_stride_y, + const uint16_t* src_u, + int src_stride_u, + const uint16_t* src_v, + int src_stride_v, + uint16_t* dst_y, + int dst_stride_y, + uint16_t* dst_u, + int dst_stride_u, + uint16_t* dst_v, + int dst_stride_v, + int width, + int height, + enum RotationMode mode); + +// Rotate I410 frame. +LIBYUV_API +int I410Rotate(const uint16_t* src_y, + int src_stride_y, + const uint16_t* src_u, + int src_stride_u, + const uint16_t* src_v, + int src_stride_v, + uint16_t* dst_y, + int dst_stride_y, + uint16_t* dst_u, + int dst_stride_u, + uint16_t* dst_v, + int dst_stride_v, + int width, + int height, + enum RotationMode mode); + // Rotate NV12 input and store in I420. LIBYUV_API int NV12ToI420Rotate(const uint8_t* src_y, @@ -156,6 +210,16 @@ void RotatePlane270(const uint8_t* src, int width, int height); +// Rotate a plane by 0, 90, 180, or 270. +LIBYUV_API +int RotatePlane_16(const uint16_t* src, + int src_stride, + uint16_t* dst, + int dst_stride, + int width, + int height, + enum RotationMode mode); + // Rotations for when U and V are interleaved. // These functions take one UV input pointer and // split the data into two buffers while diff --git a/files/include/libyuv/rotate_row.h b/files/include/libyuv/rotate_row.h index aa8528a9..2dd8c03d 100644 --- a/files/include/libyuv/rotate_row.h +++ b/files/include/libyuv/rotate_row.h @@ -42,6 +42,8 @@ extern "C" { // The following are available for GCC 32 or 64 bit: #if !defined(LIBYUV_DISABLE_X86) && (defined(__i386__) || defined(__x86_64__)) #define HAS_TRANSPOSEWX8_SSSE3 +#define HAS_TRANSPOSE4X4_32_SSE2 +#define HAS_TRANSPOSE4X4_32_AVX2 #endif // The following are available for 64 bit GCC: @@ -54,6 +56,7 @@ extern "C" { (defined(__ARM_NEON__) || defined(LIBYUV_NEON) || defined(__aarch64__)) #define HAS_TRANSPOSEWX8_NEON #define HAS_TRANSPOSEUVWX8_NEON +#define HAS_TRANSPOSE4X4_32_NEON #endif #if !defined(LIBYUV_DISABLE_MSA) && defined(__mips_msa) @@ -215,6 +218,48 @@ void TransposeUVWx16_Any_LSX(const uint8_t* src, uint8_t* dst_b, int dst_stride_b, int width); +void TransposeWxH_16_C(const uint16_t* src, + int src_stride, + uint16_t* dst, + int dst_stride, + int width, + int height); + +void TransposeWx8_16_C(const uint16_t* src, + int src_stride, + uint16_t* dst, + int dst_stride, + int width); +void TransposeWx1_16_C(const uint16_t* src, + int src_stride, + uint16_t* dst, + int dst_stride, + int width); + +// Transpose 32 bit values (ARGB) +void Transpose4x4_32_NEON(const uint8_t* src, + int src_stride, + uint8_t* dst, + int dst_stride, + int width); + +void Transpose4x4_32_SSE2(const uint8_t* src, + int src_stride, + uint8_t* dst, + int dst_stride, + int width); + +void Transpose4x4_32_AVX2(const uint8_t* src, + int src_stride, + uint8_t* dst, + int dst_stride, + int width); + +void Transpose4x4_32_C(const uint8_t* src, + int src_stride, + uint8_t* dst, + int dst_stride, + int width); #ifdef __cplusplus } // extern "C" diff --git a/files/include/libyuv/row.h b/files/include/libyuv/row.h index 1a1cf4b6..5b244d77 100644 --- a/files/include/libyuv/row.h +++ b/files/include/libyuv/row.h @@ -11,7 +11,8 @@ #ifndef INCLUDE_LIBYUV_ROW_H_ #define INCLUDE_LIBYUV_ROW_H_ -#include <stdlib.h> // For malloc. +#include <stddef.h> // For NULL +#include <stdlib.h> // For malloc #include "libyuv/basic_types.h" @@ -75,9 +76,6 @@ extern "C" { (defined(_M_IX86) || defined(__x86_64__) || defined(__i386__)) // Conversions: #define HAS_ABGRTOYROW_SSSE3 -#if !defined(LIBYUV_BIT_EXACT) -#define HAS_ABGRTOUVROW_SSSE3 -#endif #define HAS_ARGB1555TOARGBROW_SSE2 #define HAS_ARGB4444TOARGBROW_SSE2 #define HAS_ARGBEXTRACTALPHAROW_SSE2 @@ -92,12 +90,6 @@ extern "C" { #define HAS_ARGBTOYJROW_SSSE3 #define HAS_ARGBTOYROW_SSSE3 #define HAS_BGRATOYROW_SSSE3 -#if !defined(LIBYUV_BIT_EXACT) -#define HAS_ARGBTOUV444ROW_SSSE3 -#define HAS_ARGBTOUVJROW_SSSE3 -#define HAS_ARGBTOUVROW_SSSE3 -#define HAS_BGRATOUVROW_SSSE3 -#endif #define HAS_COPYROW_ERMS #define HAS_COPYROW_SSE2 #define HAS_H422TOARGBROW_SSSE3 @@ -111,6 +103,7 @@ extern "C" { #define HAS_I422TOUYVYROW_SSE2 #define HAS_I422TOYUY2ROW_SSE2 #define HAS_I444TOARGBROW_SSSE3 +#define HAS_I444TORGB24ROW_SSSE3 #define HAS_INTERPOLATEROW_SSSE3 #define HAS_J400TOARGBROW_SSE2 #define HAS_J422TOARGBROW_SSSE3 @@ -124,16 +117,13 @@ extern "C" { #define HAS_NV21TORGB24ROW_SSSE3 #define HAS_RAWTOARGBROW_SSSE3 #define HAS_RAWTORGB24ROW_SSSE3 +#define HAS_RAWTOYJROW_SSSE3 #define HAS_RAWTOYROW_SSSE3 #define HAS_RGB24TOARGBROW_SSSE3 +#define HAS_RGB24TOYJROW_SSSE3 #define HAS_RGB24TOYROW_SSSE3 #define HAS_RGB565TOARGBROW_SSE2 #define HAS_RGBATOYROW_SSSE3 -#if !defined(LIBYUV_BIT_EXACT) -#define HAS_RGB24TOYJROW_SSSE3 -#define HAS_RAWTOYJROW_SSSE3 -#define HAS_RGBATOUVROW_SSSE3 -#endif #define HAS_SETROW_ERMS #define HAS_SETROW_X86 #define HAS_SPLITUVROW_SSE2 @@ -145,13 +135,18 @@ extern "C" { #define HAS_YUY2TOUV422ROW_SSE2 #define HAS_YUY2TOUVROW_SSE2 #define HAS_YUY2TOYROW_SSE2 +#if !defined(LIBYUV_BIT_EXACT) +#define HAS_ABGRTOUVROW_SSSE3 +#define HAS_ARGBTOUV444ROW_SSSE3 +#define HAS_ARGBTOUVJROW_SSSE3 +#define HAS_ARGBTOUVROW_SSSE3 +#define HAS_BGRATOUVROW_SSSE3 +#define HAS_RGBATOUVROW_SSSE3 +#endif // Effects: #define HAS_ARGBADDROW_SSE2 #define HAS_ARGBAFFINEROW_SSE2 -#if !defined(LIBYUV_BIT_EXACT) -#define HAS_ARGBATTENUATEROW_SSSE3 -#endif #define HAS_ARGBBLENDROW_SSSE3 #define HAS_ARGBCOLORMATRIXROW_SSSE3 #define HAS_ARGBCOLORTABLEROW_X86 @@ -176,6 +171,9 @@ extern "C" { #define HAS_SOBELXROW_SSE2 #define HAS_SOBELXYROW_SSE2 #define HAS_SOBELYROW_SSE2 +#if !defined(LIBYUV_BIT_EXACT) +#define HAS_ARGBATTENUATEROW_SSSE3 +#endif // The following functions fail on gcc/clang 32 bit with fpic and framepointer. // caveat: clangcl uses row_win.cc which works. @@ -201,17 +199,10 @@ extern "C" { #define HAS_ARGBSHUFFLEROW_AVX2 #define HAS_ARGBTORGB565DITHERROW_AVX2 #define HAS_ARGBTOYJROW_AVX2 -#define HAS_RAWTOYJROW_AVX2 -#define HAS_RGB24TOYJROW_AVX2 #define HAS_ARGBTOYROW_AVX2 -#if !defined(LIBYUV_BIT_EXACT) -#define HAS_ARGBTOUVJROW_AVX2 -#define HAS_ARGBTOUVROW_AVX2 -#endif #define HAS_COPYROW_AVX #define HAS_H422TOARGBROW_AVX2 #define HAS_HALFFLOATROW_AVX2 -// #define HAS_HALFFLOATROW_F16C // Enable to test halffloat cast #define HAS_I422TOARGB1555ROW_AVX2 #define HAS_I422TOARGB4444ROW_AVX2 #define HAS_I422TOARGBROW_AVX2 @@ -219,6 +210,7 @@ extern "C" { #define HAS_I422TORGB565ROW_AVX2 #define HAS_I422TORGBAROW_AVX2 #define HAS_I444TOARGBROW_AVX2 +#define HAS_I444TORGB24ROW_AVX2 #define HAS_INTERPOLATEROW_AVX2 #define HAS_J422TOARGBROW_AVX2 #define HAS_MERGEUVROW_AVX2 @@ -228,6 +220,8 @@ extern "C" { #define HAS_NV12TORGB565ROW_AVX2 #define HAS_NV21TOARGBROW_AVX2 #define HAS_NV21TORGB24ROW_AVX2 +#define HAS_RAWTOYJROW_AVX2 +#define HAS_RGB24TOYJROW_AVX2 #define HAS_SPLITUVROW_AVX2 #define HAS_UYVYTOARGBROW_AVX2 #define HAS_UYVYTOUV422ROW_AVX2 @@ -237,16 +231,21 @@ extern "C" { #define HAS_YUY2TOUV422ROW_AVX2 #define HAS_YUY2TOUVROW_AVX2 #define HAS_YUY2TOYROW_AVX2 +// #define HAS_HALFFLOATROW_F16C // Enable to test half float cast +#if !defined(LIBYUV_BIT_EXACT) +#define HAS_ARGBTOUVJROW_AVX2 +#define HAS_ARGBTOUVROW_AVX2 +#endif // Effects: #define HAS_ARGBADDROW_AVX2 -#if !defined(LIBYUV_BIT_EXACT) -#define HAS_ARGBATTENUATEROW_AVX2 -#endif #define HAS_ARGBMULTIPLYROW_AVX2 #define HAS_ARGBSUBTRACTROW_AVX2 #define HAS_ARGBUNATTENUATEROW_AVX2 #define HAS_BLENDPLANEROW_AVX2 +#if !defined(LIBYUV_BIT_EXACT) +#define HAS_ARGBATTENUATEROW_AVX2 +#endif #if defined(__x86_64__) || !defined(__pic__) || defined(__clang__) || \ defined(_MSC_VER) @@ -282,28 +281,32 @@ extern "C" { // The following are available for gcc/clang x86 platforms: // TODO(fbarchard): Port to Visual C #if !defined(LIBYUV_DISABLE_X86) && (defined(__x86_64__) || defined(__i386__)) +#define HAS_AB64TOARGBROW_SSSE3 #define HAS_ABGRTOAR30ROW_SSSE3 +#define HAS_ABGRTOYJROW_SSSE3 +#define HAS_AR64TOARGBROW_SSSE3 +#define HAS_ARGBTOAB64ROW_SSSE3 #define HAS_ARGBTOAR30ROW_SSSE3 #define HAS_ARGBTOAR64ROW_SSSE3 -#define HAS_ARGBTOAB64ROW_SSSE3 -#define HAS_AR64TOARGBROW_SSSE3 -#define HAS_AB64TOARGBROW_SSSE3 #define HAS_CONVERT16TO8ROW_SSSE3 #define HAS_CONVERT8TO16ROW_SSE2 #define HAS_DETILEROW_SSE2 +#define HAS_DETILEROW_16_SSE2 +#define HAS_DETILEROW_16_AVX #define HAS_DETILESPLITUVROW_SSSE3 +#define HAS_DETILETOYUY2_SSE2 #define HAS_HALFMERGEUVROW_SSSE3 #define HAS_I210TOAR30ROW_SSSE3 #define HAS_I210TOARGBROW_SSSE3 #define HAS_I212TOAR30ROW_SSSE3 #define HAS_I212TOARGBROW_SSSE3 #define HAS_I400TOARGBROW_SSE2 -#define HAS_I422TOAR30ROW_SSSE3 #define HAS_I410TOAR30ROW_SSSE3 #define HAS_I410TOARGBROW_SSSE3 +#define HAS_I422TOAR30ROW_SSSE3 #define HAS_MERGEARGBROW_SSE2 -#define HAS_MERGEXRGBROW_SSE2 #define HAS_MERGERGBROW_SSSE3 +#define HAS_MERGEXRGBROW_SSE2 #define HAS_MIRRORUVROW_SSSE3 #define HAS_NV21TOYUV24ROW_SSSE3 #define HAS_P210TOAR30ROW_SSSE3 @@ -312,15 +315,17 @@ extern "C" { #define HAS_P410TOARGBROW_SSSE3 #define HAS_RAWTORGBAROW_SSSE3 #define HAS_RGB24MIRRORROW_SSSE3 -#if !defined(LIBYUV_BIT_EXACT) #define HAS_RGBATOYJROW_SSSE3 -#endif #define HAS_SPLITARGBROW_SSE2 #define HAS_SPLITARGBROW_SSSE3 +#define HAS_SPLITRGBROW_SSSE3 #define HAS_SPLITXRGBROW_SSE2 #define HAS_SPLITXRGBROW_SSSE3 -#define HAS_SPLITRGBROW_SSSE3 #define HAS_SWAPUVROW_SSSE3 +#define HAS_YUY2TONVUVROW_SSE2 +#if !defined(LIBYUV_BIT_EXACT) +#define HAS_ABGRTOUVJROW_SSSE3 +#endif #if defined(__x86_64__) || !defined(__pic__) // TODO(fbarchard): fix build error on android_full_debug=1 @@ -335,31 +340,20 @@ extern "C" { #if !defined(LIBYUV_DISABLE_X86) && \ (defined(__x86_64__) || defined(__i386__)) && \ (defined(CLANG_HAS_AVX2) || defined(GCC_HAS_AVX2)) +#define HAS_AB64TOARGBROW_AVX2 #define HAS_ABGRTOAR30ROW_AVX2 -#if !defined(LIBYUV_BIT_EXACT) -#define HAS_ABGRTOUVROW_AVX2 +#define HAS_ABGRTOYJROW_AVX2 #define HAS_ABGRTOYROW_AVX2 -#endif +#define HAS_AR64TOARGBROW_AVX2 +#define HAS_ARGBTOAB64ROW_AVX2 #define HAS_ARGBTOAR30ROW_AVX2 +#define HAS_ARGBTOAR64ROW_AVX2 #define HAS_ARGBTORAWROW_AVX2 #define HAS_ARGBTORGB24ROW_AVX2 -#define HAS_ARGBTOAR64ROW_AVX2 -#define HAS_ARGBTOAB64ROW_AVX2 -#define HAS_AR64TOARGBROW_AVX2 -#define HAS_AB64TOARGBROW_AVX2 #define HAS_CONVERT16TO8ROW_AVX2 -#define HAS_INTERPOLATEROW_16TO8_AVX2 #define HAS_CONVERT8TO16ROW_AVX2 #define HAS_DIVIDEROW_16_AVX2 #define HAS_HALFMERGEUVROW_AVX2 -#define HAS_MERGEAR64ROW_AVX2 -#define HAS_MERGEARGB16TO8ROW_AVX2 -#define HAS_MERGEARGBROW_AVX2 -#define HAS_MERGEXR30ROW_AVX2 -#define HAS_MERGEXR64ROW_AVX2 -#define HAS_MERGEXRGB16TO8ROW_AVX2 -#define HAS_MERGEXRGBROW_AVX2 -#define HAS_NV21TOYUV24ROW_AVX2 #define HAS_I210TOAR30ROW_AVX2 #define HAS_I210TOARGBROW_AVX2 #define HAS_I212TOAR30ROW_AVX2 @@ -367,23 +361,35 @@ extern "C" { #define HAS_I400TOARGBROW_AVX2 #define HAS_I410TOAR30ROW_AVX2 #define HAS_I410TOARGBROW_AVX2 -#define HAS_P210TOAR30ROW_AVX2 -#define HAS_P210TOARGBROW_AVX2 -#define HAS_P410TOAR30ROW_AVX2 -#define HAS_P410TOARGBROW_AVX2 #define HAS_I422TOAR30ROW_AVX2 #define HAS_I422TOUYVYROW_AVX2 #define HAS_I422TOYUY2ROW_AVX2 +#define HAS_INTERPOLATEROW_16TO8_AVX2 +#define HAS_MERGEAR64ROW_AVX2 +#define HAS_MERGEARGB16TO8ROW_AVX2 +#define HAS_MERGEARGBROW_AVX2 #define HAS_MERGEUVROW_16_AVX2 +#define HAS_MERGEXR30ROW_AVX2 +#define HAS_MERGEXR64ROW_AVX2 +#define HAS_MERGEXRGB16TO8ROW_AVX2 +#define HAS_MERGEXRGBROW_AVX2 #define HAS_MIRRORUVROW_AVX2 #define HAS_MULTIPLYROW_16_AVX2 -#if !defined(LIBYUV_BIT_EXACT) +#define HAS_NV21TOYUV24ROW_AVX2 +#define HAS_P210TOAR30ROW_AVX2 +#define HAS_P210TOARGBROW_AVX2 +#define HAS_P410TOAR30ROW_AVX2 +#define HAS_P410TOARGBROW_AVX2 #define HAS_RGBATOYJROW_AVX2 -#endif #define HAS_SPLITARGBROW_AVX2 -#define HAS_SPLITXRGBROW_AVX2 #define HAS_SPLITUVROW_16_AVX2 +#define HAS_SPLITXRGBROW_AVX2 #define HAS_SWAPUVROW_AVX2 +#define HAS_YUY2TONVUVROW_AVX2 +#if !defined(LIBYUV_BIT_EXACT) +#define HAS_ABGRTOUVJROW_AVX2 +#define HAS_ABGRTOUVROW_AVX2 +#endif #if defined(__x86_64__) || !defined(__pic__) // TODO(fbarchard): fix build error on android_full_debug=1 @@ -397,8 +403,9 @@ extern "C" { // TODO(fbarchard): Port to GCC and Visual C // TODO(fbarchard): re-enable HAS_ARGBTORGB24ROW_AVX512VBMI. Issue libyuv:789 #if !defined(LIBYUV_DISABLE_X86) && \ - (defined(__x86_64__) || defined(__i386__)) && (defined(CLANG_HAS_AVX512)) + (defined(__x86_64__) || defined(__i386__)) && defined(CLANG_HAS_AVX512) #define HAS_ARGBTORGB24ROW_AVX512VBMI +#define HAS_MERGEUVROW_AVX512BW #endif // The following are available for AVX512 clang x64 platforms: @@ -412,7 +419,9 @@ extern "C" { #if !defined(LIBYUV_DISABLE_NEON) && \ (defined(__aarch64__) || defined(__ARM_NEON__) || defined(LIBYUV_NEON)) #define HAS_AB64TOARGBROW_NEON +#define HAS_ABGRTOUVJROW_NEON #define HAS_ABGRTOUVROW_NEON +#define HAS_ABGRTOYJROW_NEON #define HAS_ABGRTOYROW_NEON #define HAS_AR64TOARGBROW_NEON #define HAS_ARGB1555TOARGBROW_NEON @@ -444,8 +453,11 @@ extern "C" { #define HAS_BYTETOFLOATROW_NEON #define HAS_CONVERT16TO8ROW_NEON #define HAS_COPYROW_NEON +#define HAS_DETILEROW_16_NEON #define HAS_DETILEROW_NEON #define HAS_DETILESPLITUVROW_NEON +#define HAS_DETILETOYUY2_NEON +#define HAS_UNPACKMT2T_NEON #define HAS_DIVIDEROW_16_NEON #define HAS_HALFFLOATROW_NEON #define HAS_HALFMERGEUVROW_NEON @@ -461,6 +473,7 @@ extern "C" { #define HAS_I422TOYUY2ROW_NEON #define HAS_I444ALPHATOARGBROW_NEON #define HAS_I444TOARGBROW_NEON +#define HAS_I444TORGB24ROW_NEON #define HAS_INTERPOLATEROW_16_NEON #define HAS_INTERPOLATEROW_NEON #define HAS_J400TOARGBROW_NEON @@ -513,6 +526,7 @@ extern "C" { #define HAS_UYVYTOUVROW_NEON #define HAS_UYVYTOYROW_NEON #define HAS_YUY2TOARGBROW_NEON +#define HAS_YUY2TONVUVROW_NEON #define HAS_YUY2TOUV422ROW_NEON #define HAS_YUY2TOUVROW_NEON #define HAS_YUY2TOYROW_NEON @@ -524,13 +538,13 @@ extern "C" { #define HAS_ARGBCOLORMATRIXROW_NEON #define HAS_ARGBGRAYROW_NEON #define HAS_ARGBMIRRORROW_NEON -#define HAS_RGB24MIRRORROW_NEON #define HAS_ARGBMULTIPLYROW_NEON #define HAS_ARGBQUANTIZEROW_NEON #define HAS_ARGBSEPIAROW_NEON #define HAS_ARGBSHADEROW_NEON #define HAS_ARGBSHUFFLEROW_NEON #define HAS_ARGBSUBTRACTROW_NEON +#define HAS_RGB24MIRRORROW_NEON #define HAS_SOBELROW_NEON #define HAS_SOBELTOPLANEROW_NEON #define HAS_SOBELXROW_NEON @@ -540,12 +554,13 @@ extern "C" { // The following are available on AArch64 platforms: #if !defined(LIBYUV_DISABLE_NEON) && defined(__aarch64__) +#define HAS_GAUSSCOL_F32_NEON +#define HAS_GAUSSROW_F32_NEON #define HAS_INTERPOLATEROW_16TO8_NEON #define HAS_SCALESUMSAMPLES_NEON -#define HAS_GAUSSROW_F32_NEON -#define HAS_GAUSSCOL_F32_NEON #endif #if !defined(LIBYUV_DISABLE_MSA) && defined(__mips_msa) +#define HAS_ABGRTOUVJROW_MSA #define HAS_ABGRTOUVROW_MSA #define HAS_ABGRTOYROW_MSA #define HAS_ARGB1555TOARGBROW_MSA @@ -581,27 +596,25 @@ extern "C" { #define HAS_BGRATOYROW_MSA #define HAS_HALFFLOATROW_MSA #define HAS_I400TOARGBROW_MSA -#define HAS_I422TOUYVYROW_MSA -#define HAS_I422TOYUY2ROW_MSA -#define HAS_I422TOARGBROW_MSA -#define HAS_I422TORGBAROW_MSA #define HAS_I422ALPHATOARGBROW_MSA +#define HAS_I422TOARGB1555ROW_MSA +#define HAS_I422TOARGB4444ROW_MSA +#define HAS_I422TOARGBROW_MSA #define HAS_I422TORGB24ROW_MSA #define HAS_I422TORGB565ROW_MSA -#define HAS_I422TOARGB4444ROW_MSA -#define HAS_I422TOARGB1555ROW_MSA -#define HAS_NV12TOARGBROW_MSA -#define HAS_NV12TORGB565ROW_MSA -#define HAS_NV21TOARGBROW_MSA -#define HAS_YUY2TOARGBROW_MSA -#define HAS_UYVYTOARGBROW_MSA +#define HAS_I422TORGBAROW_MSA +#define HAS_I422TOUYVYROW_MSA +#define HAS_I422TOYUY2ROW_MSA #define HAS_I444TOARGBROW_MSA #define HAS_INTERPOLATEROW_MSA #define HAS_J400TOARGBROW_MSA #define HAS_MERGEUVROW_MSA #define HAS_MIRRORROW_MSA -#define HAS_MIRRORUVROW_MSA #define HAS_MIRRORSPLITUVROW_MSA +#define HAS_MIRRORUVROW_MSA +#define HAS_NV12TOARGBROW_MSA +#define HAS_NV12TORGB565ROW_MSA +#define HAS_NV21TOARGBROW_MSA #define HAS_RAWTOARGBROW_MSA #define HAS_RAWTORGB24ROW_MSA #define HAS_RAWTOUVROW_MSA @@ -621,113 +634,208 @@ extern "C" { #define HAS_SOBELXYROW_MSA #define HAS_SOBELYROW_MSA #define HAS_SPLITUVROW_MSA +#define HAS_UYVYTOARGBROW_MSA #define HAS_UYVYTOUVROW_MSA #define HAS_UYVYTOYROW_MSA +#define HAS_YUY2TOARGBROW_MSA #define HAS_YUY2TOUV422ROW_MSA #define HAS_YUY2TOUVROW_MSA #define HAS_YUY2TOYROW_MSA #endif #if !defined(LIBYUV_DISABLE_LSX) && defined(__loongarch_sx) -#define HAS_ARGB4444TOARGBROW_LSX +#define HAS_ABGRTOUVROW_LSX +#define HAS_ABGRTOYROW_LSX #define HAS_ARGB1555TOARGBROW_LSX -#define HAS_RGB565TOARGBROW_LSX -#define HAS_RGB24TOARGBROW_LSX -#define HAS_RAWTOARGBROW_LSX -#define HAS_ARGB1555TOYROW_LSX #define HAS_ARGB1555TOUVROW_LSX -#define HAS_RGB565TOYROW_LSX -#define HAS_RGB565TOUVROW_LSX -#define HAS_RGB24TOYROW_LSX -#define HAS_RGB24TOUVROW_LSX -#define HAS_RAWTOYROW_LSX -#define HAS_RAWTOUVROW_LSX +#define HAS_ARGB1555TOYROW_LSX +#define HAS_ARGB4444TOARGBROW_LSX +#define HAS_ARGBADDROW_LSX +#define HAS_ARGBATTENUATEROW_LSX +#define HAS_ARGBBLENDROW_LSX +#define HAS_ARGBCOLORMATRIXROW_LSX +#define HAS_ARGBEXTRACTALPHAROW_LSX +#define HAS_ARGBGRAYROW_LSX +#define HAS_ARGBSEPIAROW_LSX +#define HAS_ARGBSHADEROW_LSX +#define HAS_ARGBSHUFFLEROW_LSX +#define HAS_ARGBSUBTRACTROW_LSX +#define HAS_ARGBQUANTIZEROW_LSX +#define HAS_ARGBSETROW_LSX +#define HAS_ARGBTOARGB1555ROW_LSX +#define HAS_ARGBTOARGB4444ROW_LSX +#define HAS_ARGBTORAWROW_LSX +#define HAS_ARGBTORGB24ROW_LSX +#define HAS_ARGBTORGB565ROW_LSX +#define HAS_ARGBTORGB565DITHERROW_LSX +#define HAS_ARGBTOUVJROW_LSX +#define HAS_ARGBTOUV444ROW_LSX +#define HAS_ARGBTOUVROW_LSX +#define HAS_ARGBTOYJROW_LSX +#define HAS_ARGBMIRRORROW_LSX +#define HAS_ARGBMULTIPLYROW_LSX +#define HAS_BGRATOUVROW_LSX +#define HAS_BGRATOYROW_LSX +#define HAS_I400TOARGBROW_LSX +#define HAS_I444TOARGBROW_LSX +#define HAS_INTERPOLATEROW_LSX +#define HAS_I422ALPHATOARGBROW_LSX +#define HAS_I422TOARGB1555ROW_LSX +#define HAS_I422TOARGB4444ROW_LSX +#define HAS_I422TORGB24ROW_LSX +#define HAS_I422TORGB565ROW_LSX +#define HAS_I422TORGBAROW_LSX +#define HAS_I422TOUYVYROW_LSX +#define HAS_I422TOYUY2ROW_LSX +#define HAS_J400TOARGBROW_LSX +#define HAS_MERGEUVROW_LSX +#define HAS_MIRRORROW_LSX +#define HAS_MIRRORUVROW_LSX +#define HAS_MIRRORSPLITUVROW_LSX #define HAS_NV12TOARGBROW_LSX #define HAS_NV12TORGB565ROW_LSX #define HAS_NV21TOARGBROW_LSX +#define HAS_RAWTOARGBROW_LSX +#define HAS_RAWTORGB24ROW_LSX +#define HAS_RAWTOUVROW_LSX +#define HAS_RAWTOYROW_LSX +#define HAS_RGB24TOARGBROW_LSX +#define HAS_RGB24TOUVROW_LSX +#define HAS_RGB24TOYROW_LSX +#define HAS_RGB565TOARGBROW_LSX +#define HAS_RGB565TOUVROW_LSX +#define HAS_RGB565TOYROW_LSX +#define HAS_RGBATOUVROW_LSX +#define HAS_RGBATOYROW_LSX +#define HAS_SETROW_LSX #define HAS_SOBELROW_LSX #define HAS_SOBELTOPLANEROW_LSX #define HAS_SOBELXYROW_LSX -#define HAS_ARGBTOYJROW_LSX -#define HAS_BGRATOYROW_LSX -#define HAS_BGRATOUVROW_LSX -#define HAS_ABGRTOYROW_LSX -#define HAS_ABGRTOUVROW_LSX -#define HAS_RGBATOYROW_LSX -#define HAS_RGBATOUVROW_LSX -#define HAS_ARGBTOUVJROW_LSX -#define HAS_I444TOARGBROW_LSX -#define HAS_I400TOARGBROW_LSX -#define HAS_J400TOARGBROW_LSX -#define HAS_YUY2TOARGBROW_LSX -#define HAS_UYVYTOARGBROW_LSX -#define HAS_INTERPOLATEROW_LSX -#define HAS_ARGBSETROW_LSX -#define HAS_RAWTORGB24ROW_LSX -#define HAS_MERGEUVROW_LSX -#define HAS_ARGBEXTRACTALPHAROW_LSX -#define HAS_ARGBBLENDROW_LSX -#define HAS_ARGBQUANTIZEROW_LSX -#define HAS_ARGBCOLORMATRIXROW_LSX #define HAS_SPLITUVROW_LSX -#define HAS_SETROW_LSX -#define HAS_MIRRORSPLITUVROW_LSX +#define HAS_UYVYTOARGBROW_LSX +#define HAS_UYVYTOUV422ROW_LSX +#define HAS_UYVYTOUVROW_LSX +#define HAS_UYVYTOYROW_LSX +#define HAS_YUY2TOARGBROW_LSX +#define HAS_YUY2TOUVROW_LSX +#define HAS_YUY2TOUV422ROW_LSX +#define HAS_YUY2TOYROW_LSX +#define HAS_ARGBTOYROW_LSX +#define HAS_ABGRTOYJROW_LSX +#define HAS_RGBATOYJROW_LSX +#define HAS_RGB24TOYJROW_LSX +#define HAS_RAWTOYJROW_LSX +#endif + +#if !defined(LIBYUV_DISABLE_LSX) && defined(__loongarch_sx) +#define HAS_I422TOARGBROW_LSX #endif #if !defined(LIBYUV_DISABLE_LASX) && defined(__loongarch_asx) +#define HAS_ARGB1555TOARGBROW_LASX +#define HAS_ARGB1555TOUVROW_LASX +#define HAS_ARGB1555TOYROW_LASX +#define HAS_ARGB4444TOARGBROW_LASX +#define HAS_ARGBADDROW_LASX +#define HAS_ARGBATTENUATEROW_LASX +#define HAS_ARGBGRAYROW_LASX +#define HAS_ARGBMIRRORROW_LASX +#define HAS_ARGBMULTIPLYROW_LASX +#define HAS_ARGBSEPIAROW_LASX +#define HAS_ARGBSHADEROW_LASX +#define HAS_ARGBSHUFFLEROW_LASX +#define HAS_ARGBSUBTRACTROW_LASX +#define HAS_ARGBTOARGB1555ROW_LASX +#define HAS_ARGBTOARGB4444ROW_LASX +#define HAS_ARGBTORAWROW_LASX +#define HAS_ARGBTORGB24ROW_LASX +#define HAS_ARGBTORGB565DITHERROW_LASX +#define HAS_ARGBTORGB565ROW_LASX +#define HAS_ARGBTOUV444ROW_LASX +#define HAS_ARGBTOUVJROW_LASX +#define HAS_ARGBTOUVROW_LASX +#define HAS_ARGBTOYJROW_LASX +#define HAS_ARGBTOYROW_LASX +#define HAS_ABGRTOYJROW_LASX +#define HAS_ABGRTOYROW_LASX +#define HAS_I422ALPHATOARGBROW_LASX +#define HAS_I422TOARGB1555ROW_LASX +#define HAS_I422TOARGB4444ROW_LASX #define HAS_I422TOARGBROW_LASX +#define HAS_I422TORGB24ROW_LASX +#define HAS_I422TORGB565ROW_LASX #define HAS_I422TORGBAROW_LASX -#define HAS_I422ALPHATOARGBROW_LASX -#define HAS_I422TOYUY2ROW_LASX #define HAS_I422TOUYVYROW_LASX +#define HAS_I422TOYUY2ROW_LASX #define HAS_MIRRORROW_LASX #define HAS_MIRRORUVROW_LASX -#define HAS_ARGBMIRRORROW_LASX -#define HAS_I422TORGB24ROW_LASX -#define HAS_I422TORGB565ROW_LASX -#define HAS_I422TOARGB4444ROW_LASX -#define HAS_I422TOARGB1555ROW_LASX -#define HAS_YUY2TOUVROW_LASX -#define HAS_YUY2TOYROW_LASX -#define HAS_YUY2TOUV422ROW_LASX -#define HAS_UYVYTOYROW_LASX -#define HAS_UYVYTOUVROW_LASX -#define HAS_UYVYTOUV422ROW_LASX -#define HAS_ARGBTOYROW_LASX -#define HAS_ARGBTOUVROW_LASX -#define HAS_ARGBTORGB24ROW_LASX -#define HAS_ARGBTORAWROW_LASX -#define HAS_ARGBTORGB565ROW_LASX -#define HAS_ARGBTOARGB1555ROW_LASX -#define HAS_ARGBTOARGB4444ROW_LASX -#define HAS_ARGBTOUV444ROW_LASX -#define HAS_ARGBMULTIPLYROW_LASX -#define HAS_ARGBADDROW_LASX -#define HAS_ARGBSUBTRACTROW_LASX -#define HAS_ARGBATTENUATEROW_LASX -#define HAS_ARGBTORGB565DITHERROW_LASX -#define HAS_ARGBSHUFFLEROW_LASX -#define HAS_ARGBSHADEROW_LASX -#define HAS_ARGBGRAYROW_LASX -#define HAS_ARGBSEPIAROW_LASX -#define HAS_ARGB4444TOARGBROW_LASX -#define HAS_ARGB1555TOARGBROW_LASX -#define HAS_RGB565TOARGBROW_LASX -#define HAS_RGB24TOARGBROW_LASX -#define HAS_RAWTOARGBROW_LASX -#define HAS_ARGB1555TOYROW_LASX -#define HAS_ARGB1555TOUVROW_LASX -#define HAS_RGB565TOYROW_LASX -#define HAS_RGB565TOUVROW_LASX -#define HAS_RGB24TOYROW_LASX -#define HAS_RGB24TOUVROW_LASX -#define HAS_RAWTOYROW_LASX -#define HAS_RAWTOUVROW_LASX #define HAS_NV12TOARGBROW_LASX #define HAS_NV12TORGB565ROW_LASX #define HAS_NV21TOARGBROW_LASX -#define HAS_ARGBTOYJROW_LASX -#define HAS_ARGBTOUVJROW_LASX +#define HAS_RAWTOARGBROW_LASX +#define HAS_RAWTOUVROW_LASX +#define HAS_RAWTOYROW_LASX +#define HAS_RGB24TOARGBROW_LASX +#define HAS_RGB24TOUVROW_LASX +#define HAS_RGB24TOYROW_LASX +#define HAS_RGB565TOARGBROW_LASX +#define HAS_RGB565TOUVROW_LASX +#define HAS_RGB565TOYROW_LASX +#define HAS_UYVYTOUV422ROW_LASX +#define HAS_UYVYTOUVROW_LASX +#define HAS_UYVYTOYROW_LASX +#define HAS_YUY2TOUV422ROW_LASX +#define HAS_YUY2TOUVROW_LASX +#define HAS_YUY2TOYROW_LASX +#define HAS_RGBATOYROW_LASX +#define HAS_RGBATOYJROW_LASX +#define HAS_BGRATOYROW_LASX +#define HAS_RGB24TOYJROW_LASX +#define HAS_RAWTOYJROW_LASX +#endif + +#if !defined(LIBYUV_DISABLE_RVV) && defined(__riscv_vector) +#define HAS_AB64TOARGBROW_RVV +#define HAS_AR64TOARGBROW_RVV +#define HAS_ARGBATTENUATEROW_RVV +#define HAS_ARGBTOAB64ROW_RVV +#define HAS_ARGBTOAR64ROW_RVV +#define HAS_ARGBTORAWROW_RVV +#define HAS_ARGBTORGB24ROW_RVV +#define HAS_ARGBTOYROW_RVV +#define HAS_ARGBTOYJROW_RVV +#define HAS_ABGRTOYROW_RVV +#define HAS_ABGRTOYJROW_RVV +#define HAS_BGRATOYROW_RVV +#define HAS_COPYROW_RVV +#define HAS_I400TOARGBROW_RVV +#define HAS_I422ALPHATOARGBROW_RVV +#define HAS_I422TOARGBROW_RVV +#define HAS_I422TORGB24ROW_RVV +#define HAS_I422TORGBAROW_RVV +#define HAS_I444ALPHATOARGBROW_RVV +#define HAS_I444TOARGBROW_RVV +#define HAS_I444TORGB24ROW_RVV +#define HAS_INTERPOLATEROW_RVV +#define HAS_J400TOARGBROW_RVV +#define HAS_MERGEARGBROW_RVV +#define HAS_MERGERGBROW_RVV +#define HAS_MERGEUVROW_RVV +#define HAS_MERGEXRGBROW_RVV +#define HAS_SPLITARGBROW_RVV +#define HAS_SPLITRGBROW_RVV +#define HAS_SPLITUVROW_RVV +#define HAS_SPLITXRGBROW_RVV +#define HAS_RAWTOARGBROW_RVV +#define HAS_RAWTORGB24ROW_RVV +#define HAS_RAWTORGBAROW_RVV +#define HAS_RAWTOYJROW_RVV +#define HAS_RAWTOYROW_RVV +#define HAS_RGB24TOARGBROW_RVV +#define HAS_RGB24TOYJROW_RVV +#define HAS_RGB24TOYROW_RVV +#define HAS_RGBATOYROW_RVV +#define HAS_RGBATOYJROW_RVV #endif #if defined(_MSC_VER) && !defined(__CLR_VER) && !defined(__clang__) @@ -789,8 +897,8 @@ typedef uint32_t ulvec32[8]; typedef uint8_t ulvec8[32]; #endif -#if defined(__aarch64__) || defined(__arm__) -// This struct is for ARM color conversion. +#if defined(__aarch64__) || defined(__arm__) || defined(__riscv) +// This struct is for ARM and RISC-V color conversion. struct YuvConstants { uvec8 kUVCoeff; vec16 kRGBCoeffBias; @@ -816,13 +924,21 @@ struct YuvConstants { #define IS_ALIGNED(p, a) (!((uintptr_t)(p) & ((a)-1))) -#define align_buffer_64(var, size) \ - uint8_t* var##_mem = (uint8_t*)(malloc((size) + 63)); /* NOLINT */ \ - uint8_t* var = (uint8_t*)(((intptr_t)(var##_mem) + 63) & ~63) /* NOLINT */ +#define align_buffer_64(var, size) \ + void* var##_mem = malloc((size) + 63); /* NOLINT */ \ + uint8_t* var = (uint8_t*)(((intptr_t)var##_mem + 63) & ~63) /* NOLINT */ #define free_aligned_buffer_64(var) \ free(var##_mem); \ - var = 0 + var = NULL + +#define align_buffer_64_16(var, size) \ + void* var##_mem = malloc((size)*2 + 63); /* NOLINT */ \ + uint16_t* var = (uint16_t*)(((intptr_t)var##_mem + 63) & ~63) /* NOLINT */ + +#define free_aligned_buffer_64_16(var) \ + free(var##_mem); \ + var = NULL #if defined(__APPLE__) || defined(__x86_64__) || defined(__llvm__) #define OMITFP @@ -894,6 +1010,12 @@ void I444ToARGBRow_NEON(const uint8_t* src_y, uint8_t* dst_argb, const struct YuvConstants* yuvconstants, int width); +void I444ToRGB24Row_NEON(const uint8_t* src_y, + const uint8_t* src_u, + const uint8_t* src_v, + uint8_t* dst_rgb24, + const struct YuvConstants* yuvconstants, + int width); void I422ToARGBRow_NEON(const uint8_t* src_y, const uint8_t* src_u, const uint8_t* src_v, @@ -981,6 +1103,50 @@ void UYVYToARGBRow_NEON(const uint8_t* src_uyvy, uint8_t* dst_argb, const struct YuvConstants* yuvconstants, int width); +void I444ToARGBRow_RVV(const uint8_t* src_y, + const uint8_t* src_u, + const uint8_t* src_v, + uint8_t* dst_argb, + const struct YuvConstants* yuvconstants, + int width); +void I444AlphaToARGBRow_RVV(const uint8_t* src_y, + const uint8_t* src_u, + const uint8_t* src_v, + const uint8_t* src_a, + uint8_t* dst_argb, + const struct YuvConstants* yuvconstants, + int width); +void I444ToRGB24Row_RVV(const uint8_t* src_y, + const uint8_t* src_u, + const uint8_t* src_v, + uint8_t* dst_rgb24, + const struct YuvConstants* yuvconstants, + int width); +void I422ToARGBRow_RVV(const uint8_t* src_y, + const uint8_t* src_u, + const uint8_t* src_v, + uint8_t* dst_argb, + const struct YuvConstants* yuvconstants, + int width); +void I422AlphaToARGBRow_RVV(const uint8_t* src_y, + const uint8_t* src_u, + const uint8_t* src_v, + const uint8_t* src_a, + uint8_t* dst_argb, + const struct YuvConstants* yuvconstants, + int width); +void I422ToRGBARow_RVV(const uint8_t* src_y, + const uint8_t* src_u, + const uint8_t* src_v, + uint8_t* dst_rgba, + const struct YuvConstants* yuvconstants, + int width); +void I422ToRGB24Row_RVV(const uint8_t* src_y, + const uint8_t* src_u, + const uint8_t* src_v, + uint8_t* dst_rgb24, + const struct YuvConstants* yuvconstants, + int width); void I444ToARGBRow_MSA(const uint8_t* src_y, const uint8_t* src_u, const uint8_t* src_v, @@ -1000,6 +1166,12 @@ void I422ToARGBRow_MSA(const uint8_t* src_y, uint8_t* dst_argb, const struct YuvConstants* yuvconstants, int width); +void I422ToARGBRow_LSX(const uint8_t* src_y, + const uint8_t* src_u, + const uint8_t* src_v, + uint8_t* dst_argb, + const struct YuvConstants* yuvconstants, + int width); void I422ToARGBRow_LASX(const uint8_t* src_y, const uint8_t* src_u, const uint8_t* src_v, @@ -1012,6 +1184,12 @@ void I422ToRGBARow_MSA(const uint8_t* src_y, uint8_t* dst_argb, const struct YuvConstants* yuvconstants, int width); +void I422ToRGBARow_LSX(const uint8_t* src_y, + const uint8_t* src_u, + const uint8_t* src_v, + uint8_t* dst_argb, + const struct YuvConstants* yuvconstants, + int width); void I422ToRGBARow_LASX(const uint8_t* src_y, const uint8_t* src_u, const uint8_t* src_v, @@ -1025,6 +1203,13 @@ void I422AlphaToARGBRow_MSA(const uint8_t* src_y, uint8_t* dst_argb, const struct YuvConstants* yuvconstants, int width); +void I422AlphaToARGBRow_LSX(const uint8_t* src_y, + const uint8_t* src_u, + const uint8_t* src_v, + const uint8_t* src_a, + uint8_t* dst_argb, + const struct YuvConstants* yuvconstants, + int width); void I422AlphaToARGBRow_LASX(const uint8_t* src_y, const uint8_t* src_u, const uint8_t* src_v, @@ -1038,6 +1223,12 @@ void I422ToRGB24Row_MSA(const uint8_t* src_y, uint8_t* dst_argb, const struct YuvConstants* yuvconstants, int width); +void I422ToRGB24Row_LSX(const uint8_t* src_y, + const uint8_t* src_u, + const uint8_t* src_v, + uint8_t* dst_argb, + const struct YuvConstants* yuvconstants, + int width); void I422ToRGB24Row_LASX(const uint8_t* src_y, const uint8_t* src_u, const uint8_t* src_v, @@ -1050,6 +1241,12 @@ void I422ToRGB565Row_MSA(const uint8_t* src_y, uint8_t* dst_rgb565, const struct YuvConstants* yuvconstants, int width); +void I422ToRGB565Row_LSX(const uint8_t* src_y, + const uint8_t* src_u, + const uint8_t* src_v, + uint8_t* dst_rgb565, + const struct YuvConstants* yuvconstants, + int width); void I422ToRGB565Row_LASX(const uint8_t* src_y, const uint8_t* src_u, const uint8_t* src_v, @@ -1062,6 +1259,12 @@ void I422ToARGB4444Row_MSA(const uint8_t* src_y, uint8_t* dst_argb4444, const struct YuvConstants* yuvconstants, int width); +void I422ToARGB4444Row_LSX(const uint8_t* src_y, + const uint8_t* src_u, + const uint8_t* src_v, + uint8_t* dst_argb4444, + const struct YuvConstants* yuvconstants, + int width); void I422ToARGB4444Row_LASX(const uint8_t* src_y, const uint8_t* src_u, const uint8_t* src_v, @@ -1074,6 +1277,12 @@ void I422ToARGB1555Row_MSA(const uint8_t* src_y, uint8_t* dst_argb1555, const struct YuvConstants* yuvconstants, int width); +void I422ToARGB1555Row_LSX(const uint8_t* src_y, + const uint8_t* src_u, + const uint8_t* src_v, + uint8_t* dst_argb1555, + const struct YuvConstants* yuvconstants, + int width); void I422ToARGB1555Row_LASX(const uint8_t* src_y, const uint8_t* src_u, const uint8_t* src_v, @@ -1148,9 +1357,13 @@ void ARGBToYRow_Any_AVX2(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); void ABGRToYRow_AVX2(const uint8_t* src_abgr, uint8_t* dst_y, int width); void ABGRToYRow_Any_AVX2(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); void ARGBToYRow_SSSE3(const uint8_t* src_argb, uint8_t* dst_y, int width); +void ARGBToYJRow_SSSE3(const uint8_t* src_argb, uint8_t* dst_y, int width); void ARGBToYJRow_AVX2(const uint8_t* src_argb, uint8_t* dst_y, int width); void ARGBToYJRow_Any_AVX2(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); -void ARGBToYJRow_SSSE3(const uint8_t* src_argb, uint8_t* dst_y, int width); +void ABGRToYRow_SSSE3(const uint8_t* src_abgr, uint8_t* dst_y, int width); +void ABGRToYJRow_SSSE3(const uint8_t* src_abgr, uint8_t* dst_y, int width); +void ABGRToYJRow_AVX2(const uint8_t* src_abgr, uint8_t* dst_y, int width); +void ABGRToYJRow_Any_AVX2(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); void RGBAToYJRow_AVX2(const uint8_t* src_rgba, uint8_t* dst_y, int width); void RGBAToYJRow_Any_AVX2(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); void RGBAToYJRow_SSSE3(const uint8_t* src_rgba, uint8_t* dst_y, int width); @@ -1164,13 +1377,23 @@ void RAWToYJRow_SSSE3(const uint8_t* src_raw, uint8_t* dst_yj, int width); void RGB24ToYJRow_AVX2(const uint8_t* src_rgb24, uint8_t* dst_yj, int width); void RAWToYJRow_AVX2(const uint8_t* src_raw, uint8_t* dst_yj, int width); void ARGBToYRow_NEON(const uint8_t* src_argb, uint8_t* dst_y, int width); -void ARGBToYJRow_NEON(const uint8_t* src_argb, uint8_t* dst_y, int width); -void RGBAToYJRow_NEON(const uint8_t* src_rgba, uint8_t* dst_y, int width); +void ARGBToYJRow_NEON(const uint8_t* src_argb, uint8_t* dst_yj, int width); +void ABGRToYJRow_NEON(const uint8_t* src_abgr, uint8_t* dst_yj, int width); +void RGBAToYJRow_NEON(const uint8_t* src_rgba, uint8_t* dst_yj, int width); +void ARGBToYRow_RVV(const uint8_t* src_argb, uint8_t* dst_y, int width); +void ARGBToYJRow_RVV(const uint8_t* src_argb, uint8_t* dst_yj, int width); +void ABGRToYJRow_RVV(const uint8_t* src_rgba, uint8_t* dst_yj, int width); +void RGBAToYJRow_RVV(const uint8_t* src_rgba, uint8_t* dst_yj, int width); void ARGBToYRow_MSA(const uint8_t* src_argb0, uint8_t* dst_y, int width); void ARGBToYJRow_MSA(const uint8_t* src_argb0, uint8_t* dst_y, int width); +void ARGBToYRow_LSX(const uint8_t* src_argb0, uint8_t* dst_y, int width); void ARGBToYRow_LASX(const uint8_t* src_argb0, uint8_t* dst_y, int width); void ARGBToYJRow_LSX(const uint8_t* src_argb0, uint8_t* dst_y, int width); +void ABGRToYJRow_LSX(const uint8_t* src_abgr, uint8_t* dst_yj, int width); +void RGBAToYJRow_LSX(const uint8_t* src_rgba, uint8_t* dst_yj, int width); void ARGBToYJRow_LASX(const uint8_t* src_argb0, uint8_t* dst_y, int width); +void ABGRToYJRow_LASX(const uint8_t* src_abgr, uint8_t* dst_yj, int width); +void RGBAToYJRow_LASX(const uint8_t* src_rgba, uint8_t* dst_yj, int width); void ARGBToUV444Row_NEON(const uint8_t* src_argb, uint8_t* dst_u, uint8_t* dst_v, @@ -1189,11 +1412,20 @@ void ARGBToUVRow_MSA(const uint8_t* src_argb, uint8_t* dst_u, uint8_t* dst_v, int width); +void ARGBToUVRow_LSX(const uint8_t* src_argb, + int src_stride_argb, + uint8_t* dst_u, + uint8_t* dst_v, + int width); void ARGBToUVRow_LASX(const uint8_t* src_argb, int src_stride_argb, uint8_t* dst_u, uint8_t* dst_v, int width); +void ARGBToUV444Row_LSX(const uint8_t* src_argb, + uint8_t* dst_u, + uint8_t* dst_v, + int width); void ARGBToUV444Row_LASX(const uint8_t* src_argb, uint8_t* dst_u, uint8_t* dst_v, @@ -1203,6 +1435,11 @@ void ARGBToUVJRow_NEON(const uint8_t* src_argb, uint8_t* dst_u, uint8_t* dst_v, int width); +void ABGRToUVJRow_NEON(const uint8_t* src_abgr, + int src_stride_abgr, + uint8_t* dst_uj, + uint8_t* dst_vj, + int width); void BGRAToUVRow_NEON(const uint8_t* src_bgra, int src_stride_bgra, uint8_t* dst_u, @@ -1258,6 +1495,11 @@ void ARGBToUVJRow_MSA(const uint8_t* src_rgb, uint8_t* dst_u, uint8_t* dst_v, int width); +void ABGRToUVJRow_MSA(const uint8_t* src_rgb, + int src_stride_rgb, + uint8_t* dst_u, + uint8_t* dst_v, + int width); void BGRAToUVRow_MSA(const uint8_t* src_rgb, int src_stride_rgb, uint8_t* dst_u, @@ -1372,6 +1614,13 @@ void ARGB1555ToYRow_NEON(const uint8_t* src_argb1555, void ARGB4444ToYRow_NEON(const uint8_t* src_argb4444, uint8_t* dst_y, int width); +void BGRAToYRow_RVV(const uint8_t* src_bgra, uint8_t* dst_y, int width); +void ABGRToYRow_RVV(const uint8_t* src_abgr, uint8_t* dst_y, int width); +void RGBAToYRow_RVV(const uint8_t* src_rgba, uint8_t* dst_y, int width); +void RGB24ToYRow_RVV(const uint8_t* src_rgb24, uint8_t* dst_y, int width); +void RGB24ToYJRow_RVV(const uint8_t* src_rgb24, uint8_t* dst_yj, int width); +void RAWToYRow_RVV(const uint8_t* src_raw, uint8_t* dst_y, int width); +void RAWToYJRow_RVV(const uint8_t* src_raw, uint8_t* dst_yj, int width); void BGRAToYRow_MSA(const uint8_t* src_argb, uint8_t* dst_y, int width); void ABGRToYRow_MSA(const uint8_t* src_argb, uint8_t* dst_y, int width); void RGBAToYRow_MSA(const uint8_t* src_argb, uint8_t* dst_y, int width); @@ -1384,6 +1633,8 @@ void BGRAToYRow_LSX(const uint8_t* src_bgra, uint8_t* dst_y, int width); void ABGRToYRow_LSX(const uint8_t* src_abgr, uint8_t* dst_y, int width); void RGBAToYRow_LSX(const uint8_t* src_rgba, uint8_t* dst_y, int width); void ARGB1555ToYRow_LSX(const uint8_t* src_argb1555, uint8_t* dst_y, int width); +void RGB24ToYJRow_LSX(const uint8_t* src_rgb24, uint8_t* dst_yj, int width); +void ABGRToYRow_LASX(const uint8_t* src_abgr, uint8_t* dst_y, int width); void ARGB1555ToYRow_LASX(const uint8_t* src_argb1555, uint8_t* dst_y, int width); @@ -1393,9 +1644,15 @@ void RGB24ToYRow_LSX(const uint8_t* src_rgb24, uint8_t* dst_y, int width); void RGB24ToYRow_LASX(const uint8_t* src_rgb24, uint8_t* dst_y, int width); void RAWToYRow_LSX(const uint8_t* src_raw, uint8_t* dst_y, int width); void RAWToYRow_LASX(const uint8_t* src_raw, uint8_t* dst_y, int width); +void RGBAToYRow_LASX(const uint8_t* src_rgba, uint8_t* dst_y, int width); +void BGRAToYRow_LASX(const uint8_t* src_bgra, uint8_t* dst_y, int width); +void RGB24ToYJRow_LASX(const uint8_t* src_rgb24, uint8_t* dst_yj, int width); +void RAWToYJRow_LSX(const uint8_t* src_raw, uint8_t* dst_yj, int width); +void RAWToYJRow_LASX(const uint8_t* src_raw, uint8_t* dst_yj, int width); void ARGBToYRow_C(const uint8_t* src_rgb, uint8_t* dst_y, int width); void ARGBToYJRow_C(const uint8_t* src_rgb, uint8_t* dst_y, int width); +void ABGRToYJRow_C(const uint8_t* src_rgb, uint8_t* dst_y, int width); void RGBAToYJRow_C(const uint8_t* src_rgb, uint8_t* dst_y, int width); void BGRAToYRow_C(const uint8_t* src_rgb, uint8_t* dst_y, int width); void ABGRToYRow_C(const uint8_t* src_rgb, uint8_t* dst_y, int width); @@ -1409,6 +1666,7 @@ void ARGB1555ToYRow_C(const uint8_t* src_argb1555, uint8_t* dst_y, int width); void ARGB4444ToYRow_C(const uint8_t* src_argb4444, uint8_t* dst_y, int width); void ARGBToYRow_Any_SSSE3(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); void ARGBToYJRow_Any_SSSE3(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); +void ABGRToYJRow_Any_SSSE3(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); void RGBAToYJRow_Any_SSSE3(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); void BGRAToYRow_Any_SSSE3(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); void ABGRToYRow_Any_SSSE3(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); @@ -1423,6 +1681,7 @@ void RGB24ToYJRow_Any_AVX2(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); void RAWToYJRow_Any_AVX2(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); void ARGBToYRow_Any_NEON(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); void ARGBToYJRow_Any_NEON(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); +void ABGRToYJRow_Any_NEON(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); void RGBAToYJRow_Any_NEON(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); void BGRAToYRow_Any_NEON(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); void ABGRToYRow_Any_NEON(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); @@ -1453,10 +1712,15 @@ void ARGB1555ToYRow_Any_MSA(const uint8_t* src_ptr, void BGRAToYRow_Any_LSX(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); void ABGRToYRow_Any_LSX(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); void RGBAToYRow_Any_LSX(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); +void ARGBToYRow_Any_LSX(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); void ARGBToYJRow_Any_LSX(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); void RGB24ToYRow_Any_LSX(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); void RGB565ToYRow_Any_LSX(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); +void ABGRToYJRow_Any_LSX(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); void RAWToYRow_Any_LSX(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); +void RGBAToYJRow_Any_LSX(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); +void RGB24ToYJRow_Any_LSX(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); +void RAWToYJRow_Any_LSX(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); void ARGB1555ToYRow_Any_LSX(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); @@ -1465,7 +1729,14 @@ void RGB565ToYRow_Any_LASX(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); void RGB24ToYRow_Any_LASX(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); void ARGBToYJRow_Any_LASX(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); void ARGBToYRow_Any_LASX(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); +void ABGRToYRow_Any_LASX(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); +void ABGRToYJRow_Any_LASX(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); void RAWToYRow_Any_LASX(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); +void RGBAToYRow_Any_LASX(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); +void RGBAToYJRow_Any_LASX(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); +void BGRAToYRow_Any_LASX(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); +void RGB24ToYJRow_Any_LASX(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); +void RAWToYJRow_Any_LASX(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); void ARGB1555ToYRow_Any_LASX(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); @@ -1485,6 +1756,11 @@ void ARGBToUVJRow_AVX2(const uint8_t* src_argb, uint8_t* dst_u, uint8_t* dst_v, int width); +void ABGRToUVJRow_AVX2(const uint8_t* src_abgr, + int src_stride_abgr, + uint8_t* dst_u, + uint8_t* dst_v, + int width); void ARGBToUVRow_SSSE3(const uint8_t* src_argb, int src_stride_argb, uint8_t* dst_u, @@ -1495,6 +1771,11 @@ void ARGBToUVJRow_SSSE3(const uint8_t* src_argb, uint8_t* dst_u, uint8_t* dst_v, int width); +void ABGRToUVJRow_SSSE3(const uint8_t* src_abgr, + int src_stride_abgr, + uint8_t* dst_u, + uint8_t* dst_v, + int width); void BGRAToUVRow_SSSE3(const uint8_t* src_bgra, int src_stride_bgra, uint8_t* dst_u, @@ -1525,6 +1806,11 @@ void ARGBToUVJRow_Any_AVX2(const uint8_t* src_ptr, uint8_t* dst_u, uint8_t* dst_v, int width); +void ABGRToUVJRow_Any_AVX2(const uint8_t* src_ptr, + int src_stride, + uint8_t* dst_u, + uint8_t* dst_v, + int width); void ARGBToUVRow_Any_SSSE3(const uint8_t* src_ptr, int src_stride, uint8_t* dst_u, @@ -1535,6 +1821,11 @@ void ARGBToUVJRow_Any_SSSE3(const uint8_t* src_ptr, uint8_t* dst_u, uint8_t* dst_v, int width); +void ABGRToUVJRow_Any_SSSE3(const uint8_t* src_ptr, + int src_stride, + uint8_t* dst_u, + uint8_t* dst_v, + int width); void BGRAToUVRow_Any_SSSE3(const uint8_t* src_ptr, int src_stride, uint8_t* dst_u, @@ -1568,11 +1859,20 @@ void ARGBToUVRow_Any_MSA(const uint8_t* src_ptr, uint8_t* dst_u, uint8_t* dst_v, int width); +void ARGBToUVRow_Any_LSX(const uint8_t* src_ptr, + int src_stride_ptr, + uint8_t* dst_u, + uint8_t* dst_v, + int width); void ARGBToUVRow_Any_LASX(const uint8_t* src_ptr, int src_stride_ptr, uint8_t* dst_u, uint8_t* dst_v, int width); +void ARGBToUV444Row_Any_LSX(const uint8_t* src_ptr, + uint8_t* dst_u, + uint8_t* dst_v, + int width); void ARGBToUV444Row_Any_LASX(const uint8_t* src_ptr, uint8_t* dst_u, uint8_t* dst_v, @@ -1582,6 +1882,11 @@ void ARGBToUVJRow_Any_NEON(const uint8_t* src_ptr, uint8_t* dst_u, uint8_t* dst_v, int width); +void ABGRToUVJRow_Any_NEON(const uint8_t* src_ptr, + int src_stride, + uint8_t* dst_u, + uint8_t* dst_v, + int width); void BGRAToUVRow_Any_NEON(const uint8_t* src_ptr, int src_stride, uint8_t* dst_u, @@ -1747,16 +2052,16 @@ void ARGBToUVJRow_C(const uint8_t* src_rgb, uint8_t* dst_u, uint8_t* dst_v, int width); +void ABGRToUVJRow_C(const uint8_t* src_rgb, + int src_stride_rgb, + uint8_t* dst_u, + uint8_t* dst_v, + int width); void ARGBToUVRow_C(const uint8_t* src_rgb, int src_stride_rgb, uint8_t* dst_u, uint8_t* dst_v, int width); -void ARGBToUVJRow_C(const uint8_t* src_rgb, - int src_stride_rgb, - uint8_t* dst_u, - uint8_t* dst_v, - int width); void BGRAToUVRow_C(const uint8_t* src_rgb, int src_stride_rgb, uint8_t* dst_u, @@ -1772,6 +2077,11 @@ void RGBAToUVRow_C(const uint8_t* src_rgb, uint8_t* dst_u, uint8_t* dst_v, int width); +void RGBAToUVJRow_C(const uint8_t* src_rgb, + int src_stride_rgb, + uint8_t* dst_u, + uint8_t* dst_v, + int width); void RGB24ToUVRow_C(const uint8_t* src_rgb, int src_stride_rgb, uint8_t* dst_u, @@ -1826,6 +2136,7 @@ void MirrorRow_AVX2(const uint8_t* src, uint8_t* dst, int width); void MirrorRow_SSSE3(const uint8_t* src, uint8_t* dst, int width); void MirrorRow_NEON(const uint8_t* src, uint8_t* dst, int width); void MirrorRow_MSA(const uint8_t* src, uint8_t* dst, int width); +void MirrorRow_LSX(const uint8_t* src, uint8_t* dst, int width); void MirrorRow_LASX(const uint8_t* src, uint8_t* dst, int width); void MirrorRow_C(const uint8_t* src, uint8_t* dst, int width); void MirrorRow_Any_AVX2(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); @@ -1833,17 +2144,20 @@ void MirrorRow_Any_SSSE3(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); void MirrorRow_Any_SSE2(const uint8_t* src, uint8_t* dst, int width); void MirrorRow_Any_NEON(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); void MirrorRow_Any_MSA(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); +void MirrorRow_Any_LSX(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); void MirrorRow_Any_LASX(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); void MirrorUVRow_AVX2(const uint8_t* src_uv, uint8_t* dst_uv, int width); void MirrorUVRow_SSSE3(const uint8_t* src_uv, uint8_t* dst_uv, int width); void MirrorUVRow_NEON(const uint8_t* src_uv, uint8_t* dst_uv, int width); void MirrorUVRow_MSA(const uint8_t* src_uv, uint8_t* dst_uv, int width); +void MirrorUVRow_LSX(const uint8_t* src_uv, uint8_t* dst_uv, int width); void MirrorUVRow_LASX(const uint8_t* src_uv, uint8_t* dst_uv, int width); void MirrorUVRow_C(const uint8_t* src_uv, uint8_t* dst_uv, int width); void MirrorUVRow_Any_AVX2(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); void MirrorUVRow_Any_SSSE3(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); void MirrorUVRow_Any_NEON(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); void MirrorUVRow_Any_MSA(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); +void MirrorUVRow_Any_LSX(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); void MirrorUVRow_Any_LASX(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); void MirrorSplitUVRow_SSSE3(const uint8_t* src, @@ -1867,10 +2181,13 @@ void MirrorSplitUVRow_C(const uint8_t* src_uv, uint8_t* dst_v, int width); +void MirrorRow_16_C(const uint16_t* src, uint16_t* dst, int width); + void ARGBMirrorRow_AVX2(const uint8_t* src, uint8_t* dst, int width); void ARGBMirrorRow_SSE2(const uint8_t* src, uint8_t* dst, int width); void ARGBMirrorRow_NEON(const uint8_t* src_argb, uint8_t* dst_argb, int width); void ARGBMirrorRow_MSA(const uint8_t* src, uint8_t* dst, int width); +void ARGBMirrorRow_LSX(const uint8_t* src, uint8_t* dst, int width); void ARGBMirrorRow_LASX(const uint8_t* src, uint8_t* dst, int width); void ARGBMirrorRow_C(const uint8_t* src, uint8_t* dst, int width); void ARGBMirrorRow_Any_AVX2(const uint8_t* src_ptr, @@ -1883,6 +2200,7 @@ void ARGBMirrorRow_Any_NEON(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); void ARGBMirrorRow_Any_MSA(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); +void ARGBMirrorRow_Any_LSX(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); void ARGBMirrorRow_Any_LASX(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); @@ -1925,6 +2243,10 @@ void SplitUVRow_LSX(const uint8_t* src_uv, uint8_t* dst_u, uint8_t* dst_v, int width); +void SplitUVRow_RVV(const uint8_t* src_uv, + uint8_t* dst_u, + uint8_t* dst_v, + int width); void SplitUVRow_Any_SSE2(const uint8_t* src_ptr, uint8_t* dst_u, uint8_t* dst_v, @@ -1949,7 +2271,6 @@ void DetileRow_C(const uint8_t* src, ptrdiff_t src_tile_stride, uint8_t* dst, int width); - void DetileRow_NEON(const uint8_t* src, ptrdiff_t src_tile_stride, uint8_t* dst, @@ -1966,6 +2287,42 @@ void DetileRow_Any_SSE2(const uint8_t* src, ptrdiff_t src_tile_stride, uint8_t* dst, int width); +void DetileRow_AVX(const uint8_t* src, + ptrdiff_t src_tile_stride, + uint8_t* dst, + int width); +void DetileRow_Any_AVX(const uint8_t* src, + ptrdiff_t src_tile_stride, + uint8_t* dst, + int width); +void DetileRow_16_C(const uint16_t* src, + ptrdiff_t src_tile_stride, + uint16_t* dst, + int width); +void DetileRow_16_NEON(const uint16_t* src, + ptrdiff_t src_tile_stride, + uint16_t* dst, + int width); +void DetileRow_16_Any_NEON(const uint16_t* src, + ptrdiff_t src_tile_stride, + uint16_t* dst, + int width); +void DetileRow_16_SSE2(const uint16_t* src, + ptrdiff_t src_tile_stride, + uint16_t* dst, + int width); +void DetileRow_16_Any_SSE2(const uint16_t* src, + ptrdiff_t src_tile_stride, + uint16_t* dst, + int width); +void DetileRow_16_AVX(const uint16_t* src, + ptrdiff_t src_tile_stride, + uint16_t* dst, + int width); +void DetileRow_16_Any_AVX(const uint16_t* src, + ptrdiff_t src_tile_stride, + uint16_t* dst, + int width); void DetileSplitUVRow_C(const uint8_t* src_uv, ptrdiff_t src_tile_stride, uint8_t* dst_u, @@ -1991,6 +2348,38 @@ void DetileSplitUVRow_Any_NEON(const uint8_t* src_uv, uint8_t* dst_u, uint8_t* dst_v, int width); +void DetileToYUY2_C(const uint8_t* src_y, + ptrdiff_t src_y_tile_stride, + const uint8_t* src_uv, + ptrdiff_t src_uv_tile_stride, + uint8_t* dst_yuy2, + int width); +void DetileToYUY2_SSE2(const uint8_t* src_y, + ptrdiff_t src_y_tile_stride, + const uint8_t* src_uv, + ptrdiff_t src_uv_tile_stride, + uint8_t* dst_yuy2, + int width); +void DetileToYUY2_Any_SSE2(const uint8_t* src_y, + ptrdiff_t src_y_tile_stride, + const uint8_t* src_uv, + ptrdiff_t src_uv_tile_stride, + uint8_t* dst_yuy2, + int width); +void DetileToYUY2_NEON(const uint8_t* src_y, + ptrdiff_t src_y_tile_stride, + const uint8_t* src_uv, + ptrdiff_t src_uv_tile_stride, + uint8_t* dst_yuy2, + int width); +void DetileToYUY2_Any_NEON(const uint8_t* src_y, + ptrdiff_t src_y_tile_stride, + const uint8_t* src_uv, + ptrdiff_t src_uv_tile_stride, + uint8_t* dst_yuy2, + int width); +void UnpackMT2T_C(const uint8_t* src, uint16_t* dst, size_t size); +void UnpackMT2T_NEON(const uint8_t* src, uint16_t* dst, size_t size); void MergeUVRow_C(const uint8_t* src_u, const uint8_t* src_v, uint8_t* dst_uv, @@ -2003,6 +2392,10 @@ void MergeUVRow_AVX2(const uint8_t* src_u, const uint8_t* src_v, uint8_t* dst_uv, int width); +void MergeUVRow_AVX512BW(const uint8_t* src_u, + const uint8_t* src_v, + uint8_t* dst_uv, + int width); void MergeUVRow_NEON(const uint8_t* src_u, const uint8_t* src_v, uint8_t* dst_uv, @@ -2015,6 +2408,10 @@ void MergeUVRow_LSX(const uint8_t* src_u, const uint8_t* src_v, uint8_t* dst_uv, int width); +void MergeUVRow_RVV(const uint8_t* src_u, + const uint8_t* src_v, + uint8_t* dst_uv, + int width); void MergeUVRow_Any_SSE2(const uint8_t* y_buf, const uint8_t* uv_buf, uint8_t* dst_ptr, @@ -2023,6 +2420,10 @@ void MergeUVRow_Any_AVX2(const uint8_t* y_buf, const uint8_t* uv_buf, uint8_t* dst_ptr, int width); +void MergeUVRow_Any_AVX512BW(const uint8_t* y_buf, + const uint8_t* uv_buf, + uint8_t* dst_ptr, + int width); void MergeUVRow_Any_NEON(const uint8_t* y_buf, const uint8_t* uv_buf, uint8_t* dst_ptr, @@ -2079,6 +2480,11 @@ void SplitRGBRow_NEON(const uint8_t* src_rgb, uint8_t* dst_g, uint8_t* dst_b, int width); +void SplitRGBRow_RVV(const uint8_t* src_rgb, + uint8_t* dst_r, + uint8_t* dst_g, + uint8_t* dst_b, + int width); void SplitRGBRow_Any_SSSE3(const uint8_t* src_ptr, uint8_t* dst_r, uint8_t* dst_g, @@ -2105,6 +2511,11 @@ void MergeRGBRow_NEON(const uint8_t* src_r, const uint8_t* src_b, uint8_t* dst_rgb, int width); +void MergeRGBRow_RVV(const uint8_t* src_r, + const uint8_t* src_g, + const uint8_t* src_b, + uint8_t* dst_rgb, + int width); void MergeRGBRow_Any_SSSE3(const uint8_t* y_buf, const uint8_t* u_buf, const uint8_t* v_buf, @@ -2139,6 +2550,12 @@ void MergeARGBRow_NEON(const uint8_t* src_r, const uint8_t* src_a, uint8_t* dst_argb, int width); +void MergeARGBRow_RVV(const uint8_t* src_r, + const uint8_t* src_g, + const uint8_t* src_b, + const uint8_t* src_a, + uint8_t* dst_argb, + int width); void MergeARGBRow_Any_SSE2(const uint8_t* y_buf, const uint8_t* u_buf, const uint8_t* v_buf, @@ -2187,6 +2604,12 @@ void SplitARGBRow_NEON(const uint8_t* src_rgba, uint8_t* dst_b, uint8_t* dst_a, int width); +void SplitARGBRow_RVV(const uint8_t* src_rgba, + uint8_t* dst_r, + uint8_t* dst_g, + uint8_t* dst_b, + uint8_t* dst_a, + int width); void SplitARGBRow_Any_SSE2(const uint8_t* src_ptr, uint8_t* dst_r, uint8_t* dst_g, @@ -2231,6 +2654,11 @@ void MergeXRGBRow_NEON(const uint8_t* src_r, const uint8_t* src_b, uint8_t* dst_argb, int width); +void MergeXRGBRow_RVV(const uint8_t* src_r, + const uint8_t* src_g, + const uint8_t* src_b, + uint8_t* dst_argb, + int width); void MergeXRGBRow_Any_SSE2(const uint8_t* y_buf, const uint8_t* u_buf, const uint8_t* v_buf, @@ -2271,6 +2699,11 @@ void SplitXRGBRow_NEON(const uint8_t* src_rgba, uint8_t* dst_g, uint8_t* dst_b, int width); +void SplitXRGBRow_RVV(const uint8_t* src_rgba, + uint8_t* dst_r, + uint8_t* dst_g, + uint8_t* dst_b, + int width); void SplitXRGBRow_Any_SSE2(const uint8_t* src_ptr, uint8_t* dst_r, uint8_t* dst_g, @@ -2604,8 +3037,8 @@ void Convert16To8Row_NEON(const uint16_t* src_y, uint8_t* dst_y, int scale, int width); -void Convert16To8Row_Any_NEON(const uint16_t* src_y, - uint8_t* dst_y, +void Convert16To8Row_Any_NEON(const uint16_t* src_ptr, + uint8_t* dst_ptr, int scale, int width); @@ -2614,6 +3047,7 @@ void CopyRow_AVX(const uint8_t* src, uint8_t* dst, int width); void CopyRow_ERMS(const uint8_t* src, uint8_t* dst, int width); void CopyRow_NEON(const uint8_t* src, uint8_t* dst, int width); void CopyRow_MIPS(const uint8_t* src, uint8_t* dst, int count); +void CopyRow_RVV(const uint8_t* src, uint8_t* dst, int count); void CopyRow_C(const uint8_t* src, uint8_t* dst, int count); void CopyRow_Any_SSE2(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); void CopyRow_Any_AVX(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); @@ -2713,6 +3147,10 @@ void ARGBShuffleRow_MSA(const uint8_t* src_argb, uint8_t* dst_argb, const uint8_t* shuffler, int width); +void ARGBShuffleRow_LSX(const uint8_t* src_argb, + uint8_t* dst_argb, + const uint8_t* shuffler, + int width); void ARGBShuffleRow_LASX(const uint8_t* src_argb, uint8_t* dst_argb, const uint8_t* shuffler, @@ -2733,6 +3171,10 @@ void ARGBShuffleRow_Any_MSA(const uint8_t* src_ptr, uint8_t* dst_ptr, const uint8_t* param, int width); +void ARGBShuffleRow_Any_LSX(const uint8_t* src_ptr, + uint8_t* dst_ptr, + const uint8_t* param, + int width); void ARGBShuffleRow_Any_LASX(const uint8_t* src_ptr, uint8_t* dst_ptr, const uint8_t* param, @@ -2765,14 +3207,18 @@ void RGB24ToARGBRow_LSX(const uint8_t* src_rgb24, uint8_t* dst_argb, int width); void RGB24ToARGBRow_LASX(const uint8_t* src_rgb24, uint8_t* dst_argb, int width); +void RGB24ToARGBRow_RVV(const uint8_t* src_rgb24, uint8_t* dst_argb, int width); void RAWToARGBRow_NEON(const uint8_t* src_raw, uint8_t* dst_argb, int width); void RAWToRGBARow_NEON(const uint8_t* src_raw, uint8_t* dst_rgba, int width); void RAWToARGBRow_MSA(const uint8_t* src_raw, uint8_t* dst_argb, int width); void RAWToARGBRow_LSX(const uint8_t* src_raw, uint8_t* dst_argb, int width); void RAWToARGBRow_LASX(const uint8_t* src_raw, uint8_t* dst_argb, int width); +void RAWToARGBRow_RVV(const uint8_t* src_raw, uint8_t* dst_argb, int width); +void RAWToRGBARow_RVV(const uint8_t* src_raw, uint8_t* dst_rgba, int width); void RAWToRGB24Row_NEON(const uint8_t* src_raw, uint8_t* dst_rgb24, int width); void RAWToRGB24Row_MSA(const uint8_t* src_raw, uint8_t* dst_rgb24, int width); void RAWToRGB24Row_LSX(const uint8_t* src_raw, uint8_t* dst_rgb24, int width); +void RAWToRGB24Row_RVV(const uint8_t* src_raw, uint8_t* dst_rgb24, int width); void RGB565ToARGBRow_NEON(const uint8_t* src_rgb565, uint8_t* dst_argb, int width); @@ -2932,15 +3378,15 @@ void ARGBToRGB24Row_AVX512VBMI(const uint8_t* src, uint8_t* dst, int width); void ARGBToRGB565DitherRow_C(const uint8_t* src_argb, uint8_t* dst_rgb, - const uint32_t dither4, + uint32_t dither4, int width); void ARGBToRGB565DitherRow_SSE2(const uint8_t* src, uint8_t* dst, - const uint32_t dither4, + uint32_t dither4, int width); void ARGBToRGB565DitherRow_AVX2(const uint8_t* src, uint8_t* dst, - const uint32_t dither4, + uint32_t dither4, int width); void ARGBToRGB565Row_AVX2(const uint8_t* src_argb, uint8_t* dst_rgb, int width); @@ -2968,7 +3414,7 @@ void ARGBToARGB4444Row_NEON(const uint8_t* src_argb, int width); void ARGBToRGB565DitherRow_NEON(const uint8_t* src_argb, uint8_t* dst_rgb, - const uint32_t dither4, + uint32_t dither4, int width); void ARGBToRGB24Row_MSA(const uint8_t* src_argb, uint8_t* dst_rgb, int width); void ARGBToRAWRow_MSA(const uint8_t* src_argb, uint8_t* dst_rgb, int width); @@ -2981,23 +3427,39 @@ void ARGBToARGB4444Row_MSA(const uint8_t* src_argb, int width); void ARGBToRGB565DitherRow_MSA(const uint8_t* src_argb, uint8_t* dst_rgb, - const uint32_t dither4, + uint32_t dither4, + int width); +void ARGBToRGB565DitherRow_LSX(const uint8_t* src_argb, + uint8_t* dst_rgb, + uint32_t dither4, int width); void ARGBToRGB565DitherRow_LASX(const uint8_t* src_argb, uint8_t* dst_rgb, - const uint32_t dither4, + uint32_t dither4, int width); +void ARGBToRGB24Row_LSX(const uint8_t* src_argb, uint8_t* dst_rgb, int width); void ARGBToRGB24Row_LASX(const uint8_t* src_argb, uint8_t* dst_rgb, int width); +void ARGBToRAWRow_LSX(const uint8_t* src_argb, uint8_t* dst_rgb, int width); void ARGBToRAWRow_LASX(const uint8_t* src_argb, uint8_t* dst_rgb, int width); +void ARGBToRGB565Row_LSX(const uint8_t* src_argb, uint8_t* dst_rgb, int width); void ARGBToRGB565Row_LASX(const uint8_t* src_argb, uint8_t* dst_rgb, int width); +void ARGBToARGB1555Row_LSX(const uint8_t* src_argb, + uint8_t* dst_rgb, + int width); void ARGBToARGB1555Row_LASX(const uint8_t* src_argb, uint8_t* dst_rgb, int width); +void ARGBToARGB4444Row_LSX(const uint8_t* src_argb, + uint8_t* dst_rgb, + int width); void ARGBToARGB4444Row_LASX(const uint8_t* src_argb, uint8_t* dst_rgb, int width); +void ARGBToRAWRow_RVV(const uint8_t* src_argb, uint8_t* dst_raw, int width); +void ARGBToRGB24Row_RVV(const uint8_t* src_argb, uint8_t* dst_rgb24, int width); + void ARGBToRGBARow_C(const uint8_t* src_argb, uint8_t* dst_rgb, int width); void ARGBToRGB24Row_C(const uint8_t* src_argb, uint8_t* dst_rgb, int width); void ARGBToRAWRow_C(const uint8_t* src_argb, uint8_t* dst_rgb, int width); @@ -3035,6 +3497,10 @@ void ARGBToAR64Row_NEON(const uint8_t* src_argb, uint16_t* dst_ar64, int width); void ARGBToAB64Row_NEON(const uint8_t* src_argb, uint16_t* dst_ab64, int width); void AR64ToARGBRow_NEON(const uint16_t* src_ar64, uint8_t* dst_argb, int width); void AB64ToARGBRow_NEON(const uint16_t* src_ab64, uint8_t* dst_argb, int width); +void ARGBToAR64Row_RVV(const uint8_t* src_argb, uint16_t* dst_ar64, int width); +void ARGBToAB64Row_RVV(const uint8_t* src_argb, uint16_t* dst_ab64, int width); +void AR64ToARGBRow_RVV(const uint16_t* src_ar64, uint8_t* dst_argb, int width); +void AB64ToARGBRow_RVV(const uint16_t* src_ab64, uint8_t* dst_argb, int width); void ARGBToAR64Row_Any_SSSE3(const uint8_t* src_ptr, uint16_t* dst_ptr, int width); @@ -3077,6 +3543,7 @@ void J400ToARGBRow_AVX2(const uint8_t* src_y, uint8_t* dst_argb, int width); void J400ToARGBRow_NEON(const uint8_t* src_y, uint8_t* dst_argb, int width); void J400ToARGBRow_MSA(const uint8_t* src_y, uint8_t* dst_argb, int width); void J400ToARGBRow_LSX(const uint8_t* src_y, uint8_t* dst_argb, int width); +void J400ToARGBRow_RVV(const uint8_t* src_y, uint8_t* dst_argb, int width); void J400ToARGBRow_C(const uint8_t* src_y, uint8_t* dst_argb, int width); void J400ToARGBRow_Any_SSE2(const uint8_t* src_ptr, uint8_t* dst_ptr, @@ -3096,6 +3563,12 @@ void I444ToARGBRow_C(const uint8_t* src_y, uint8_t* rgb_buf, const struct YuvConstants* yuvconstants, int width); +void I444ToRGB24Row_C(const uint8_t* src_y, + const uint8_t* src_u, + const uint8_t* src_v, + uint8_t* rgb_buf, + const struct YuvConstants* yuvconstants, + int width); void I422ToARGBRow_C(const uint8_t* src_y, const uint8_t* src_u, const uint8_t* src_v, @@ -3290,6 +3763,18 @@ void I444ToARGBRow_AVX2(const uint8_t* y_buf, uint8_t* dst_argb, const struct YuvConstants* yuvconstants, int width); +void I444ToRGB24Row_SSSE3(const uint8_t* y_buf, + const uint8_t* u_buf, + const uint8_t* v_buf, + uint8_t* dst_rgb24, + const struct YuvConstants* yuvconstants, + int width); +void I444ToRGB24Row_AVX2(const uint8_t* y_buf, + const uint8_t* u_buf, + const uint8_t* v_buf, + uint8_t* dst_rgb24, + const struct YuvConstants* yuvconstants, + int width); void I422ToARGBRow_SSSE3(const uint8_t* y_buf, const uint8_t* u_buf, const uint8_t* v_buf, @@ -3631,12 +4116,24 @@ void I444ToARGBRow_Any_SSSE3(const uint8_t* y_buf, uint8_t* dst_ptr, const struct YuvConstants* yuvconstants, int width); +void I444ToRGB24Row_Any_SSSE3(const uint8_t* y_buf, + const uint8_t* u_buf, + const uint8_t* v_buf, + uint8_t* dst_ptr, + const struct YuvConstants* yuvconstants, + int width); void I444ToARGBRow_Any_AVX2(const uint8_t* y_buf, const uint8_t* u_buf, const uint8_t* v_buf, uint8_t* dst_ptr, const struct YuvConstants* yuvconstants, int width); +void I444ToRGB24Row_Any_AVX2(const uint8_t* y_buf, + const uint8_t* u_buf, + const uint8_t* v_buf, + uint8_t* dst_ptr, + const struct YuvConstants* yuvconstants, + int width); void I422ToARGBRow_Any_SSSE3(const uint8_t* y_buf, const uint8_t* u_buf, const uint8_t* v_buf, @@ -3823,13 +4320,13 @@ void NV21ToRGB24Row_Any_AVX2(const uint8_t* y_buf, uint8_t* dst_ptr, const struct YuvConstants* yuvconstants, int width); -void NV21ToYUV24Row_Any_SSSE3(const uint8_t* src_y, - const uint8_t* src_vu, - uint8_t* dst_yuv24, +void NV21ToYUV24Row_Any_SSSE3(const uint8_t* y_buf, + const uint8_t* uv_buf, + uint8_t* dst_ptr, int width); -void NV21ToYUV24Row_Any_AVX2(const uint8_t* src_y, - const uint8_t* src_vu, - uint8_t* dst_yuv24, +void NV21ToYUV24Row_Any_AVX2(const uint8_t* y_buf, + const uint8_t* uv_buf, + uint8_t* dst_ptr, int width); void NV12ToRGB565Row_Any_SSSE3(const uint8_t* y_buf, const uint8_t* uv_buf, @@ -3976,6 +4473,10 @@ void I400ToARGBRow_LSX(const uint8_t* src_y, uint8_t* dst_argb, const struct YuvConstants* yuvconstants, int width); +void I400ToARGBRow_RVV(const uint8_t* src_y, + uint8_t* dst_argb, + const struct YuvConstants* yuvconstants, + int width); void I400ToARGBRow_Any_SSE2(const uint8_t* src_ptr, uint8_t* dst_ptr, const struct YuvConstants* param, @@ -4084,10 +4585,18 @@ void ARGBMultiplyRow_Any_MSA(const uint8_t* y_buf, const uint8_t* uv_buf, uint8_t* dst_ptr, int width); +void ARGBMultiplyRow_LSX(const uint8_t* src_argb0, + const uint8_t* src_argb1, + uint8_t* dst_argb, + int width); void ARGBMultiplyRow_LASX(const uint8_t* src_argb0, const uint8_t* src_argb1, uint8_t* dst_argb, int width); +void ARGBMultiplyRow_Any_LSX(const uint8_t* y_buf, + const uint8_t* uv_buf, + uint8_t* dst_ptr, + int width); void ARGBMultiplyRow_Any_LASX(const uint8_t* y_buf, const uint8_t* uv_buf, uint8_t* dst_ptr, @@ -4130,10 +4639,18 @@ void ARGBAddRow_Any_MSA(const uint8_t* y_buf, const uint8_t* uv_buf, uint8_t* dst_ptr, int width); +void ARGBAddRow_LSX(const uint8_t* src_argb0, + const uint8_t* src_argb1, + uint8_t* dst_argb, + int width); void ARGBAddRow_LASX(const uint8_t* src_argb0, const uint8_t* src_argb1, uint8_t* dst_argb, int width); +void ARGBAddRow_Any_LSX(const uint8_t* y_buf, + const uint8_t* uv_buf, + uint8_t* dst_ptr, + int width); void ARGBAddRow_Any_LASX(const uint8_t* y_buf, const uint8_t* uv_buf, uint8_t* dst_ptr, @@ -4177,10 +4694,18 @@ void ARGBSubtractRow_Any_MSA(const uint8_t* y_buf, const uint8_t* uv_buf, uint8_t* dst_ptr, int width); +void ARGBSubtractRow_LSX(const uint8_t* src_argb0, + const uint8_t* src_argb1, + uint8_t* dst_argb, + int width); void ARGBSubtractRow_LASX(const uint8_t* src_argb0, const uint8_t* src_argb1, uint8_t* dst_argb, int width); +void ARGBSubtractRow_Any_LSX(const uint8_t* y_buf, + const uint8_t* uv_buf, + uint8_t* dst_ptr, + int width); void ARGBSubtractRow_Any_LASX(const uint8_t* y_buf, const uint8_t* uv_buf, uint8_t* dst_ptr, @@ -4273,21 +4798,37 @@ void ARGBToRGB565DitherRow_Any_MSA(const uint8_t* src_ptr, uint8_t* dst_ptr, const uint32_t param, int width); +void ARGBToRGB565DitherRow_Any_LSX(const uint8_t* src_ptr, + uint8_t* dst_ptr, + const uint32_t param, + int width); void ARGBToRGB565DitherRow_Any_LASX(const uint8_t* src_ptr, uint8_t* dst_ptr, const uint32_t param, int width); - +void ARGBToRGB24Row_Any_LSX(const uint8_t* src_ptr, + uint8_t* dst_ptr, + int width); void ARGBToRGB24Row_Any_LASX(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); +void ARGBToRAWRow_Any_LSX(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); void ARGBToRAWRow_Any_LASX(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); +void ARGBToRGB565Row_Any_LSX(const uint8_t* src_ptr, + uint8_t* dst_ptr, + int width); void ARGBToRGB565Row_Any_LASX(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); +void ARGBToARGB1555Row_Any_LSX(const uint8_t* src_ptr, + uint8_t* dst_ptr, + int width); void ARGBToARGB1555Row_Any_LASX(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); +void ARGBToARGB4444Row_Any_LSX(const uint8_t* src_ptr, + uint8_t* dst_ptr, + int width); void ARGBToARGB4444Row_Any_LASX(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); @@ -4298,6 +4839,12 @@ void I444ToARGBRow_Any_NEON(const uint8_t* y_buf, uint8_t* dst_ptr, const struct YuvConstants* yuvconstants, int width); +void I444ToRGB24Row_Any_NEON(const uint8_t* y_buf, + const uint8_t* u_buf, + const uint8_t* v_buf, + uint8_t* dst_ptr, + const struct YuvConstants* yuvconstants, + int width); void I422ToARGBRow_Any_NEON(const uint8_t* y_buf, const uint8_t* u_buf, const uint8_t* v_buf, @@ -4443,6 +4990,12 @@ void I422ToARGBRow_Any_MSA(const uint8_t* y_buf, uint8_t* dst_ptr, const struct YuvConstants* yuvconstants, int width); +void I422ToARGBRow_Any_LSX(const uint8_t* y_buf, + const uint8_t* u_buf, + const uint8_t* v_buf, + uint8_t* dst_ptr, + const struct YuvConstants* yuvconstants, + int width); void I422ToARGBRow_Any_LASX(const uint8_t* y_buf, const uint8_t* u_buf, const uint8_t* v_buf, @@ -4455,6 +5008,12 @@ void I422ToRGBARow_Any_MSA(const uint8_t* y_buf, uint8_t* dst_ptr, const struct YuvConstants* yuvconstants, int width); +void I422ToRGBARow_Any_LSX(const uint8_t* y_buf, + const uint8_t* u_buf, + const uint8_t* v_buf, + uint8_t* dst_ptr, + const struct YuvConstants* yuvconstants, + int width); void I422ToRGBARow_Any_LASX(const uint8_t* y_buf, const uint8_t* u_buf, const uint8_t* v_buf, @@ -4468,6 +5027,13 @@ void I422AlphaToARGBRow_Any_MSA(const uint8_t* y_buf, uint8_t* dst_ptr, const struct YuvConstants* yuvconstants, int width); +void I422AlphaToARGBRow_Any_LSX(const uint8_t* y_buf, + const uint8_t* u_buf, + const uint8_t* v_buf, + const uint8_t* a_buf, + uint8_t* dst_ptr, + const struct YuvConstants* yuvconstants, + int width); void I422AlphaToARGBRow_Any_LASX(const uint8_t* y_buf, const uint8_t* u_buf, const uint8_t* v_buf, @@ -4481,6 +5047,12 @@ void I422ToRGB24Row_Any_MSA(const uint8_t* y_buf, uint8_t* dst_ptr, const struct YuvConstants* yuvconstants, int width); +void I422ToRGB24Row_Any_LSX(const uint8_t* y_buf, + const uint8_t* u_buf, + const uint8_t* v_buf, + uint8_t* dst_ptr, + const struct YuvConstants* yuvconstants, + int width); void I422ToRGB24Row_Any_LASX(const uint8_t* y_buf, const uint8_t* u_buf, const uint8_t* v_buf, @@ -4493,6 +5065,12 @@ void I422ToRGB565Row_Any_MSA(const uint8_t* y_buf, uint8_t* dst_ptr, const struct YuvConstants* yuvconstants, int width); +void I422ToRGB565Row_Any_LSX(const uint8_t* y_buf, + const uint8_t* u_buf, + const uint8_t* v_buf, + uint8_t* dst_ptr, + const struct YuvConstants* yuvconstants, + int width); void I422ToRGB565Row_Any_LASX(const uint8_t* y_buf, const uint8_t* u_buf, const uint8_t* v_buf, @@ -4505,6 +5083,12 @@ void I422ToARGB4444Row_Any_MSA(const uint8_t* y_buf, uint8_t* dst_ptr, const struct YuvConstants* yuvconstants, int width); +void I422ToARGB4444Row_Any_LSX(const uint8_t* y_buf, + const uint8_t* u_buf, + const uint8_t* v_buf, + uint8_t* dst_ptr, + const struct YuvConstants* yuvconstants, + int width); void I422ToARGB4444Row_Any_LASX(const uint8_t* y_buf, const uint8_t* u_buf, const uint8_t* v_buf, @@ -4517,6 +5101,12 @@ void I422ToARGB1555Row_Any_MSA(const uint8_t* y_buf, uint8_t* dst_ptr, const struct YuvConstants* yuvconstants, int width); +void I422ToARGB1555Row_Any_LSX(const uint8_t* y_buf, + const uint8_t* u_buf, + const uint8_t* v_buf, + uint8_t* dst_ptr, + const struct YuvConstants* yuvconstants, + int width); void I422ToARGB1555Row_Any_LASX(const uint8_t* y_buf, const uint8_t* u_buf, const uint8_t* v_buf, @@ -4592,6 +5182,10 @@ void YUY2ToUVRow_AVX2(const uint8_t* src_yuy2, uint8_t* dst_u, uint8_t* dst_v, int width); +void YUY2ToNVUVRow_AVX2(const uint8_t* src_yuy2, + int stride_yuy2, + uint8_t* dst_uv, + int width); void YUY2ToUV422Row_AVX2(const uint8_t* src_yuy2, uint8_t* dst_u, uint8_t* dst_v, @@ -4602,6 +5196,10 @@ void YUY2ToUVRow_SSE2(const uint8_t* src_yuy2, uint8_t* dst_u, uint8_t* dst_v, int width); +void YUY2ToNVUVRow_SSE2(const uint8_t* src_yuy2, + int stride_yuy2, + uint8_t* dst_uv, + int width); void YUY2ToUV422Row_SSE2(const uint8_t* src_yuy2, uint8_t* dst_u, uint8_t* dst_v, @@ -4612,17 +5210,27 @@ void YUY2ToUVRow_NEON(const uint8_t* src_yuy2, uint8_t* dst_u, uint8_t* dst_v, int width); +void YUY2ToNVUVRow_NEON(const uint8_t* src_yuy2, + int stride_yuy2, + uint8_t* dst_uv, + int width); void YUY2ToUV422Row_NEON(const uint8_t* src_yuy2, uint8_t* dst_u, uint8_t* dst_v, int width); void YUY2ToYRow_MSA(const uint8_t* src_yuy2, uint8_t* dst_y, int width); +void YUY2ToYRow_LSX(const uint8_t* src_yuy2, uint8_t* dst_y, int width); void YUY2ToYRow_LASX(const uint8_t* src_yuy2, uint8_t* dst_y, int width); void YUY2ToUVRow_MSA(const uint8_t* src_yuy2, int src_stride_yuy2, uint8_t* dst_u, uint8_t* dst_v, int width); +void YUY2ToUVRow_LSX(const uint8_t* src_yuy2, + int src_stride_yuy2, + uint8_t* dst_u, + uint8_t* dst_v, + int width); void YUY2ToUVRow_LASX(const uint8_t* src_yuy2, int src_stride_yuy2, uint8_t* dst_u, @@ -4632,6 +5240,10 @@ void YUY2ToUV422Row_MSA(const uint8_t* src_yuy2, uint8_t* dst_u, uint8_t* dst_v, int width); +void YUY2ToUV422Row_LSX(const uint8_t* src_yuy2, + uint8_t* dst_u, + uint8_t* dst_v, + int width); void YUY2ToUV422Row_LASX(const uint8_t* src_yuy2, uint8_t* dst_u, uint8_t* dst_v, @@ -4642,6 +5254,10 @@ void YUY2ToUVRow_C(const uint8_t* src_yuy2, uint8_t* dst_u, uint8_t* dst_v, int width); +void YUY2ToNVUVRow_C(const uint8_t* src_yuy2, + int src_stride_yuy2, + uint8_t* dst_uv, + int width); void YUY2ToUV422Row_C(const uint8_t* src_yuy2, uint8_t* dst_u, uint8_t* dst_v, @@ -4652,6 +5268,10 @@ void YUY2ToUVRow_Any_AVX2(const uint8_t* src_ptr, uint8_t* dst_u, uint8_t* dst_v, int width); +void YUY2ToNVUVRow_Any_AVX2(const uint8_t* src_yuy2, + int stride_yuy2, + uint8_t* dst_uv, + int width); void YUY2ToUV422Row_Any_AVX2(const uint8_t* src_ptr, uint8_t* dst_u, uint8_t* dst_v, @@ -4662,6 +5282,10 @@ void YUY2ToUVRow_Any_SSE2(const uint8_t* src_ptr, uint8_t* dst_u, uint8_t* dst_v, int width); +void YUY2ToNVUVRow_Any_SSE2(const uint8_t* src_yuy2, + int stride_yuy2, + uint8_t* dst_uv, + int width); void YUY2ToUV422Row_Any_SSE2(const uint8_t* src_ptr, uint8_t* dst_u, uint8_t* dst_v, @@ -4672,17 +5296,27 @@ void YUY2ToUVRow_Any_NEON(const uint8_t* src_ptr, uint8_t* dst_u, uint8_t* dst_v, int width); +void YUY2ToNVUVRow_Any_NEON(const uint8_t* src_yuy2, + int stride_yuy2, + uint8_t* dst_uv, + int width); void YUY2ToUV422Row_Any_NEON(const uint8_t* src_ptr, uint8_t* dst_u, uint8_t* dst_v, int width); void YUY2ToYRow_Any_MSA(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); +void YUY2ToYRow_Any_LSX(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); void YUY2ToYRow_Any_LASX(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); void YUY2ToUVRow_Any_MSA(const uint8_t* src_ptr, int src_stride_ptr, uint8_t* dst_u, uint8_t* dst_v, int width); +void YUY2ToUVRow_Any_LSX(const uint8_t* src_ptr, + int src_stride_ptr, + uint8_t* dst_u, + uint8_t* dst_v, + int width); void YUY2ToUVRow_Any_LASX(const uint8_t* src_ptr, int src_stride_ptr, uint8_t* dst_u, @@ -4692,6 +5326,10 @@ void YUY2ToUV422Row_Any_MSA(const uint8_t* src_ptr, uint8_t* dst_u, uint8_t* dst_v, int width); +void YUY2ToUV422Row_Any_LSX(const uint8_t* src_ptr, + uint8_t* dst_u, + uint8_t* dst_v, + int width); void YUY2ToUV422Row_Any_LASX(const uint8_t* src_ptr, uint8_t* dst_u, uint8_t* dst_v, @@ -4737,12 +5375,18 @@ void UYVYToUV422Row_NEON(const uint8_t* src_uyvy, uint8_t* dst_v, int width); void UYVYToYRow_MSA(const uint8_t* src_uyvy, uint8_t* dst_y, int width); +void UYVYToYRow_LSX(const uint8_t* src_uyvy, uint8_t* dst_y, int width); void UYVYToYRow_LASX(const uint8_t* src_uyvy, uint8_t* dst_y, int width); void UYVYToUVRow_MSA(const uint8_t* src_uyvy, int src_stride_uyvy, uint8_t* dst_u, uint8_t* dst_v, int width); +void UYVYToUVRow_LSX(const uint8_t* src_uyvy, + int src_stride_uyvy, + uint8_t* dst_u, + uint8_t* dst_v, + int width); void UYVYToUVRow_LASX(const uint8_t* src_uyvy, int src_stride_uyvy, uint8_t* dst_u, @@ -4752,6 +5396,10 @@ void UYVYToUV422Row_MSA(const uint8_t* src_uyvy, uint8_t* dst_u, uint8_t* dst_v, int width); +void UYVYToUV422Row_LSX(const uint8_t* src_uyvy, + uint8_t* dst_u, + uint8_t* dst_v, + int width); void UYVYToUV422Row_LASX(const uint8_t* src_uyvy, uint8_t* dst_u, uint8_t* dst_v, @@ -4798,12 +5446,18 @@ void UYVYToUV422Row_Any_NEON(const uint8_t* src_ptr, uint8_t* dst_v, int width); void UYVYToYRow_Any_MSA(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); +void UYVYToYRow_Any_LSX(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); void UYVYToYRow_Any_LASX(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); void UYVYToUVRow_Any_MSA(const uint8_t* src_ptr, int src_stride_ptr, uint8_t* dst_u, uint8_t* dst_v, int width); +void UYVYToUVRow_Any_LSX(const uint8_t* src_ptr, + int src_stride_ptr, + uint8_t* dst_u, + uint8_t* dst_v, + int width); void UYVYToUVRow_Any_LASX(const uint8_t* src_ptr, int src_stride_ptr, uint8_t* dst_u, @@ -4813,6 +5467,10 @@ void UYVYToUV422Row_Any_MSA(const uint8_t* src_ptr, uint8_t* dst_u, uint8_t* dst_v, int width); +void UYVYToUV422Row_Any_LSX(const uint8_t* src_ptr, + uint8_t* dst_u, + uint8_t* dst_v, + int width); void UYVYToUV422Row_Any_LASX(const uint8_t* src_ptr, uint8_t* dst_u, uint8_t* dst_v, @@ -4927,6 +5585,11 @@ void I422ToYUY2Row_MSA(const uint8_t* src_y, const uint8_t* src_v, uint8_t* dst_yuy2, int width); +void I422ToYUY2Row_LSX(const uint8_t* src_y, + const uint8_t* src_u, + const uint8_t* src_v, + uint8_t* dst_yuy2, + int width); void I422ToYUY2Row_LASX(const uint8_t* src_y, const uint8_t* src_u, const uint8_t* src_v, @@ -4937,6 +5600,11 @@ void I422ToUYVYRow_MSA(const uint8_t* src_y, const uint8_t* src_v, uint8_t* dst_uyvy, int width); +void I422ToUYVYRow_LSX(const uint8_t* src_y, + const uint8_t* src_u, + const uint8_t* src_v, + uint8_t* dst_uyvy, + int width); void I422ToUYVYRow_LASX(const uint8_t* src_y, const uint8_t* src_u, const uint8_t* src_v, @@ -4947,6 +5615,11 @@ void I422ToYUY2Row_Any_MSA(const uint8_t* y_buf, const uint8_t* v_buf, uint8_t* dst_ptr, int width); +void I422ToYUY2Row_Any_LSX(const uint8_t* y_buf, + const uint8_t* u_buf, + const uint8_t* v_buf, + uint8_t* dst_ptr, + int width); void I422ToYUY2Row_Any_LASX(const uint8_t* y_buf, const uint8_t* u_buf, const uint8_t* v_buf, @@ -4957,6 +5630,11 @@ void I422ToUYVYRow_Any_MSA(const uint8_t* y_buf, const uint8_t* v_buf, uint8_t* dst_ptr, int width); +void I422ToUYVYRow_Any_LSX(const uint8_t* y_buf, + const uint8_t* u_buf, + const uint8_t* v_buf, + uint8_t* dst_ptr, + int width); void I422ToUYVYRow_Any_LASX(const uint8_t* y_buf, const uint8_t* u_buf, const uint8_t* v_buf, @@ -4977,9 +5655,15 @@ void ARGBAttenuateRow_NEON(const uint8_t* src_argb, void ARGBAttenuateRow_MSA(const uint8_t* src_argb, uint8_t* dst_argb, int width); +void ARGBAttenuateRow_LSX(const uint8_t* src_argb, + uint8_t* dst_argb, + int width); void ARGBAttenuateRow_LASX(const uint8_t* src_argb, uint8_t* dst_argb, int width); +void ARGBAttenuateRow_RVV(const uint8_t* src_argb, + uint8_t* dst_argb, + int width); void ARGBAttenuateRow_Any_SSSE3(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); @@ -4992,6 +5676,9 @@ void ARGBAttenuateRow_Any_NEON(const uint8_t* src_ptr, void ARGBAttenuateRow_Any_MSA(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); +void ARGBAttenuateRow_Any_LSX(const uint8_t* src_ptr, + uint8_t* dst_ptr, + int width); void ARGBAttenuateRow_Any_LASX(const uint8_t* src_ptr, uint8_t* dst_ptr, int width); @@ -5018,12 +5705,14 @@ void ARGBGrayRow_C(const uint8_t* src_argb, uint8_t* dst_argb, int width); void ARGBGrayRow_SSSE3(const uint8_t* src_argb, uint8_t* dst_argb, int width); void ARGBGrayRow_NEON(const uint8_t* src_argb, uint8_t* dst_argb, int width); void ARGBGrayRow_MSA(const uint8_t* src_argb, uint8_t* dst_argb, int width); +void ARGBGrayRow_LSX(const uint8_t* src_argb, uint8_t* dst_argb, int width); void ARGBGrayRow_LASX(const uint8_t* src_argb, uint8_t* dst_argb, int width); void ARGBSepiaRow_C(uint8_t* dst_argb, int width); void ARGBSepiaRow_SSSE3(uint8_t* dst_argb, int width); void ARGBSepiaRow_NEON(uint8_t* dst_argb, int width); void ARGBSepiaRow_MSA(uint8_t* dst_argb, int width); +void ARGBSepiaRow_LSX(uint8_t* dst_argb, int width); void ARGBSepiaRow_LASX(uint8_t* dst_argb, int width); void ARGBColorMatrixRow_C(const uint8_t* src_argb, @@ -5103,6 +5792,10 @@ void ARGBShadeRow_MSA(const uint8_t* src_argb, uint8_t* dst_argb, int width, uint32_t value); +void ARGBShadeRow_LSX(const uint8_t* src_argb, + uint8_t* dst_argb, + int width, + uint32_t value); void ARGBShadeRow_LASX(const uint8_t* src_argb, uint8_t* dst_argb, int width, @@ -5175,6 +5868,11 @@ void InterpolateRow_LSX(uint8_t* dst_ptr, ptrdiff_t src_stride, int width, int source_y_fraction); +void InterpolateRow_RVV(uint8_t* dst_ptr, + const uint8_t* src_ptr, + ptrdiff_t src_stride, + int width, + int source_y_fraction); void InterpolateRow_Any_NEON(uint8_t* dst_ptr, const uint8_t* src_ptr, ptrdiff_t src_stride_ptr, @@ -5526,6 +6224,17 @@ void GaussCol_F32_C(const float* src0, float* dst, int width); +void GaussRow_C(const uint32_t* src, uint16_t* dst, int width); +void GaussCol_C(const uint16_t* src0, + const uint16_t* src1, + const uint16_t* src2, + const uint16_t* src3, + const uint16_t* src4, + uint32_t* dst, + int width); + +void ClampFloatToZero_SSE2(const float* src_x, float* dst_y, int width); + #ifdef __cplusplus } // extern "C" } // namespace libyuv diff --git a/files/include/libyuv/scale_row.h b/files/include/libyuv/scale_row.h index 6cb5e128..a7957c3f 100644 --- a/files/include/libyuv/scale_row.h +++ b/files/include/libyuv/scale_row.h @@ -133,6 +133,8 @@ extern "C" { #define HAS_SCALEROWDOWN34_NEON #define HAS_SCALEROWDOWN38_NEON #define HAS_SCALEROWDOWN4_NEON +#define HAS_SCALEUVROWDOWN2_NEON +#define HAS_SCALEUVROWDOWN2LINEAR_NEON #define HAS_SCALEUVROWDOWN2BOX_NEON #define HAS_SCALEUVROWDOWNEVEN_NEON #define HAS_SCALEROWUP2_LINEAR_NEON @@ -214,6 +216,17 @@ void ScalePlaneVertical_16To8(int src_height, int scale, enum FilterMode filtering); +void ScalePlaneDown2_16To8(int src_width, + int src_height, + int dst_width, + int dst_height, + int src_stride, + int dst_stride, + const uint16_t* src_ptr, + uint8_t* dst_ptr, + int scale, + enum FilterMode filtering); + // Simplify the filtering based on scale factors. enum FilterMode ScaleFilterReduce(int src_width, int src_height, @@ -259,6 +272,16 @@ void ScaleRowDown2_16_C(const uint16_t* src_ptr, ptrdiff_t src_stride, uint16_t* dst, int dst_width); +void ScaleRowDown2_16To8_C(const uint16_t* src_ptr, + ptrdiff_t src_stride, + uint8_t* dst, + int dst_width, + int scale); +void ScaleRowDown2_16To8_Odd_C(const uint16_t* src_ptr, + ptrdiff_t src_stride, + uint8_t* dst, + int dst_width, + int scale); void ScaleRowDown2Linear_C(const uint8_t* src_ptr, ptrdiff_t src_stride, uint8_t* dst, @@ -267,6 +290,16 @@ void ScaleRowDown2Linear_16_C(const uint16_t* src_ptr, ptrdiff_t src_stride, uint16_t* dst, int dst_width); +void ScaleRowDown2Linear_16To8_C(const uint16_t* src_ptr, + ptrdiff_t src_stride, + uint8_t* dst, + int dst_width, + int scale); +void ScaleRowDown2Linear_16To8_Odd_C(const uint16_t* src_ptr, + ptrdiff_t src_stride, + uint8_t* dst, + int dst_width, + int scale); void ScaleRowDown2Box_C(const uint8_t* src_ptr, ptrdiff_t src_stride, uint8_t* dst, @@ -279,6 +312,16 @@ void ScaleRowDown2Box_16_C(const uint16_t* src_ptr, ptrdiff_t src_stride, uint16_t* dst, int dst_width); +void ScaleRowDown2Box_16To8_C(const uint16_t* src_ptr, + ptrdiff_t src_stride, + uint8_t* dst, + int dst_width, + int scale); +void ScaleRowDown2Box_16To8_Odd_C(const uint16_t* src_ptr, + ptrdiff_t src_stride, + uint8_t* dst, + int dst_width, + int scale); void ScaleRowDown4_C(const uint8_t* src_ptr, ptrdiff_t src_stride, uint8_t* dst, diff --git a/files/include/libyuv/version.h b/files/include/libyuv/version.h index a85be048..b6623dbb 100644 --- a/files/include/libyuv/version.h +++ b/files/include/libyuv/version.h @@ -11,6 +11,6 @@ #ifndef INCLUDE_LIBYUV_VERSION_H_ #define INCLUDE_LIBYUV_VERSION_H_ -#define LIBYUV_VERSION 1837 +#define LIBYUV_VERSION 1871 #endif // INCLUDE_LIBYUV_VERSION_H_ |