#include "libavutil/x86_cpu.h"
#include "libavcodec/dsputil.h"
#include "libavcodec/h264dsp.h"
#include "libavcodec/mpegvideo.h"
#include "libavcodec/simple_idct.h"
#include "dsputil_mmx.h"
#include "vp3dsp_mmx.h"
#include "vp3dsp_sse2.h"
#include "vp6dsp_mmx.h"
#include "vp6dsp_sse2.h"
#include "idct_xvid.h"
#include "dsputil_mmx_rnd_template.c"
#include "dsputil_mmx_avg_template.c"
#include "h264dsp_mmx.c"
#include "rv40dsp_mmx.c"
Go to the source code of this file.
Defines | |
#define | JUMPALIGN() __asm__ volatile (ASMALIGN(3)::) |
#define | MOVQ_ZERO(regd) __asm__ volatile ("pxor %%" #regd ", %%" #regd ::) |
#define | MOVQ_BFE(regd) |
#define | MOVQ_BONE(regd) __asm__ volatile ("movq %0, %%" #regd " \n\t" ::"m"(ff_bone)) |
#define | MOVQ_WTWO(regd) __asm__ volatile ("movq %0, %%" #regd " \n\t" ::"m"(ff_wtwo)) |
#define | PAVGB_MMX_NO_RND(rega, regb, regr, regfe) |
#define | PAVGB_MMX(rega, regb, regr, regfe) |
#define | PAVGBP_MMX_NO_RND(rega, regb, regr,regc, regd, regp) |
#define | PAVGBP_MMX(rega, regb, regr, regc, regd, regp) |
#define | DEF(x, y) x ## _no_rnd_ ## y ##_mmx |
#define | SET_RND MOVQ_WONE |
#define | PAVGBP(a, b, c, d, e, f) PAVGBP_MMX_NO_RND(a, b, c, d, e, f) |
#define | PAVGB(a, b, c, e) PAVGB_MMX_NO_RND(a, b, c, e) |
#define | OP_AVG(a, b, c, e) PAVGB_MMX(a, b, c, e) |
#define | DEF(x, y) x ## _ ## y ##_mmx |
#define | SET_RND MOVQ_WTWO |
#define | PAVGBP(a, b, c, d, e, f) PAVGBP_MMX(a, b, c, d, e, f) |
#define | PAVGB(a, b, c, e) PAVGB_MMX(a, b, c, e) |
#define | DEF(x) x ## _3dnow |
#define | PAVGB "pavgusb" |
#define | OP_AVG PAVGB |
#define | DEF(x) x ## _mmx2 |
#define | PAVGB "pavgb" |
#define | OP_AVG PAVGB |
#define | put_no_rnd_pixels16_mmx put_pixels16_mmx |
#define | put_no_rnd_pixels8_mmx put_pixels8_mmx |
#define | put_pixels16_mmx2 put_pixels16_mmx |
#define | put_pixels8_mmx2 put_pixels8_mmx |
#define | put_pixels4_mmx2 put_pixels4_mmx |
#define | put_no_rnd_pixels16_mmx2 put_no_rnd_pixels16_mmx |
#define | put_no_rnd_pixels8_mmx2 put_no_rnd_pixels8_mmx |
#define | put_pixels16_3dnow put_pixels16_mmx |
#define | put_pixels8_3dnow put_pixels8_mmx |
#define | put_pixels4_3dnow put_pixels4_mmx |
#define | put_no_rnd_pixels16_3dnow put_no_rnd_pixels16_mmx |
#define | put_no_rnd_pixels8_3dnow put_no_rnd_pixels8_mmx |
#define | put_signed_pixels_clamped_mmx_half(off) |
#define | CLEAR_BLOCKS(name, n) |
#define | H263_LOOP_FILTER |
#define | PAETH(cpu, abs3) |
#define | ABS3_MMX2 |
#define | ABS3_SSSE3 |
#define | QPEL_V_LOW(m3, m4, m5, m6, pw_20, pw_3, rnd, in0, in1, in2, in7, out, OP) |
#define | QPEL_BASE(OPNAME, ROUNDER, RND, OP_MMX2, OP_3DNOW) |
#define | QPEL_OP(OPNAME, ROUNDER, RND, OP, MMX) |
#define | PUT_OP(a, b, temp, size) "mov" #size " " #a ", " #b " \n\t" |
#define | AVG_3DNOW_OP(a, b, temp, size) |
#define | AVG_MMX2_OP(a, b, temp, size) |
#define | QPEL_2TAP_XY(OPNAME, SIZE, MMX, XY, HPEL) |
#define | QPEL_2TAP_L3(OPNAME, SIZE, MMX, XY, S0, S1, S2) |
#define | QPEL_2TAP(OPNAME, SIZE, MMX) |
#define | PREFETCH(name, op) |
#define | IF1(x) x |
#define | IF0(x) |
#define | MIX5(mono, stereo) |
#define | MIX_MISC(stereo) |
#define | ff_float_to_int16_interleave6_sse(a, b, c) float_to_int16_interleave_misc_sse(a,b,c,6) |
#define | ff_float_to_int16_interleave6_3dnow(a, b, c) float_to_int16_interleave_misc_3dnow(a,b,c,6) |
#define | ff_float_to_int16_interleave6_3dn2(a, b, c) float_to_int16_interleave_misc_3dnow(a,b,c,6) |
#define | ff_float_to_int16_interleave6_sse2 ff_float_to_int16_interleave6_sse |
#define | FLOAT_TO_INT16_INTERLEAVE(cpu, body) |
#define | SET_HPEL_FUNCS(PFX, IDX, SIZE, CPU) |
#define | SET_QPEL_FUNCS(PFX, IDX, SIZE, CPU) |
#define | H264_QPEL_FUNCS(x, y, CPU) |
Functions | |
void | put_pixels_clamped_mmx (const DCTELEM *block, uint8_t *pixels, int line_size) |
DECLARE_ASM_CONST (8, uint8_t, ff_vector128)[8] | |
void | put_signed_pixels_clamped_mmx (const DCTELEM *block, uint8_t *pixels, int line_size) |
void | add_pixels_clamped_mmx (const DCTELEM *block, uint8_t *pixels, int line_size) |
static void | put_pixels4_mmx (uint8_t *block, const uint8_t *pixels, int line_size, int h) |
static void | put_pixels8_mmx (uint8_t *block, const uint8_t *pixels, int line_size, int h) |
static void | put_pixels16_mmx (uint8_t *block, const uint8_t *pixels, int line_size, int h) |
static void | put_pixels16_sse2 (uint8_t *block, const uint8_t *pixels, int line_size, int h) |
static void | avg_pixels16_sse2 (uint8_t *block, const uint8_t *pixels, int line_size, int h) |
static void | clear_block_sse (DCTELEM *block) |
static void | clear_blocks_sse (DCTELEM *blocks) |
static void | add_bytes_mmx (uint8_t *dst, uint8_t *src, int w) |
static void | add_bytes_l2_mmx (uint8_t *dst, uint8_t *src1, uint8_t *src2, int w) |
static void | h263_v_loop_filter_mmx (uint8_t *src, int stride, int qscale) |
static void | transpose4x4 (uint8_t *dst, uint8_t *src, int dst_stride, int src_stride) |
static void | h263_h_loop_filter_mmx (uint8_t *src, int stride, int qscale) |
static void | draw_edges_mmx (uint8_t *buf, int wrap, int width, int height, int w) |
static void | gmc_mmx (uint8_t *dst, uint8_t *src, int stride, int h, int ox, int oy, int dxx, int dxy, int dyx, int dyy, int shift, int r, int width, int height) |
void | ff_put_cavs_qpel8_mc00_mmx2 (uint8_t *dst, uint8_t *src, int stride) |
void | ff_avg_cavs_qpel8_mc00_mmx2 (uint8_t *dst, uint8_t *src, int stride) |
void | ff_put_cavs_qpel16_mc00_mmx2 (uint8_t *dst, uint8_t *src, int stride) |
void | ff_avg_cavs_qpel16_mc00_mmx2 (uint8_t *dst, uint8_t *src, int stride) |
void | ff_put_vc1_mspel_mc00_mmx (uint8_t *dst, const uint8_t *src, int stride, int rnd) |
void | ff_avg_vc1_mspel_mc00_mmx2 (uint8_t *dst, const uint8_t *src, int stride, int rnd) |
static void | ff_idct_xvid_mmx_put (uint8_t *dest, int line_size, DCTELEM *block) |
static void | ff_idct_xvid_mmx_add (uint8_t *dest, int line_size, DCTELEM *block) |
static void | ff_idct_xvid_mmx2_put (uint8_t *dest, int line_size, DCTELEM *block) |
static void | ff_idct_xvid_mmx2_add (uint8_t *dest, int line_size, DCTELEM *block) |
static void | vorbis_inverse_coupling_3dnow (float *mag, float *ang, int blocksize) |
static void | vorbis_inverse_coupling_sse (float *mag, float *ang, int blocksize) |
static void | ac3_downmix_sse (float(*samples)[256], float(*matrix)[2], int out_ch, int in_ch, int len) |
static void | vector_fmul_3dnow (float *dst, const float *src, int len) |
static void | vector_fmul_sse (float *dst, const float *src, int len) |
static void | vector_fmul_reverse_3dnow2 (float *dst, const float *src0, const float *src1, int len) |
static void | vector_fmul_reverse_sse (float *dst, const float *src0, const float *src1, int len) |
static void | vector_fmul_add_3dnow (float *dst, const float *src0, const float *src1, const float *src2, int len) |
static void | vector_fmul_add_sse (float *dst, const float *src0, const float *src1, const float *src2, int len) |
static void | vector_fmul_window_3dnow2 (float *dst, const float *src0, const float *src1, const float *win, float add_bias, int len) |
static void | vector_fmul_window_sse (float *dst, const float *src0, const float *src1, const float *win, float add_bias, int len) |
static void | int32_to_float_fmul_scalar_sse (float *dst, const int *src, float mul, int len) |
static void | int32_to_float_fmul_scalar_sse2 (float *dst, const int *src, float mul, int len) |
static void | vector_clipf_sse (float *dst, const float *src, float min, float max, int len) |
static void | float_to_int16_3dnow (int16_t *dst, const float *src, long len) |
static void | float_to_int16_sse (int16_t *dst, const float *src, long len) |
static void | float_to_int16_sse2 (int16_t *dst, const float *src, long len) |
void | ff_float_to_int16_interleave6_sse (int16_t *dst, const float **src, int len) |
void | ff_float_to_int16_interleave6_3dnow (int16_t *dst, const float **src, int len) |
void | ff_float_to_int16_interleave6_3dn2 (int16_t *dst, const float **src, int len) |
int32_t | ff_scalarproduct_int16_mmx2 (int16_t *v1, int16_t *v2, int order, int shift) |
int32_t | ff_scalarproduct_int16_sse2 (int16_t *v1, int16_t *v2, int order, int shift) |
int32_t | ff_scalarproduct_and_madd_int16_mmx2 (int16_t *v1, int16_t *v2, int16_t *v3, int order, int mul) |
int32_t | ff_scalarproduct_and_madd_int16_sse2 (int16_t *v1, int16_t *v2, int16_t *v3, int order, int mul) |
int32_t | ff_scalarproduct_and_madd_int16_ssse3 (int16_t *v1, int16_t *v2, int16_t *v3, int order, int mul) |
void | ff_add_hfyu_median_prediction_mmx2 (uint8_t *dst, const uint8_t *top, const uint8_t *diff, int w, int *left, int *left_top) |
int | ff_add_hfyu_left_prediction_ssse3 (uint8_t *dst, const uint8_t *src, int w, int left) |
int | ff_add_hfyu_left_prediction_sse4 (uint8_t *dst, const uint8_t *src, int w, int left) |
void | ff_x264_deblock_v_luma_sse2 (uint8_t *pix, int stride, int alpha, int beta, int8_t *tc0) |
void | ff_x264_deblock_h_luma_sse2 (uint8_t *pix, int stride, int alpha, int beta, int8_t *tc0) |
void | ff_x264_deblock_h_luma_intra_mmxext (uint8_t *pix, int stride, int alpha, int beta) |
void | ff_x264_deblock_v_luma_intra_sse2 (uint8_t *pix, int stride, int alpha, int beta) |
void | ff_x264_deblock_h_luma_intra_sse2 (uint8_t *pix, int stride, int alpha, int beta) |
FLOAT_TO_INT16_INTERLEAVE (3dnow,"1: \n""pf2id (%2,%0), %%mm0 \n""pf2id 8(%2,%0), %%mm1 \n""pf2id (%3,%0), %%mm2 \n""pf2id 8(%3,%0), %%mm3 \n""packssdw %%mm1, %%mm0 \n""packssdw %%mm3, %%mm2 \n""movq %%mm0, %%mm1 \n""punpcklwd %%mm2, %%mm0 \n""punpckhwd %%mm2, %%mm1 \n""movq %%mm0, (%1,%0)\n""movq %%mm1, 8(%1,%0)\n""add $16, %0 \n""js 1b \n""femms \n") FLOAT_TO_INT16_INTERLEAVE(sse | |
mm0 n | cvtps2pi (%2,%0) |
mm0 n mm1 n mm2 n mm3 n packssdw mm0 n packssdw mm2 n movq mm1 n punpcklwd mm0 n punpckhwd mm1 n movq n movq n n js n emms n | FLOAT_TO_INT16_INTERLEAVE (sse2,"1: \n""cvtps2dq (%2,%0), %%xmm0 \n""cvtps2dq (%3,%0), %%xmm1 \n""packssdw %%xmm1, %%xmm0 \n""movhlps %%xmm0, %%xmm1 \n""punpcklwd %%xmm1, %%xmm0 \n""movdqa %%xmm0, (%1,%0) \n""add $16, %0 \n""js 1b \n") static void float_to_int16_interleave_3dn2(int16_t *dst |
void | dsputil_init_mmx (DSPContext *c, AVCodecContext *avctx) |
Variables | |
int | mm_flags |
const uint64_t | ff_bone = 0x0101010101010101ULL |
const uint64_t | ff_wtwo = 0x0002000200020002ULL |
const uint64_t | ff_pdw_80000000 [2] |
const uint64_t | ff_pw_3 = 0x0003000300030003ULL |
const uint64_t | ff_pw_4 = 0x0004000400040004ULL |
const xmm_reg | ff_pw_5 = {0x0005000500050005ULL, 0x0005000500050005ULL} |
const xmm_reg | ff_pw_8 = {0x0008000800080008ULL, 0x0008000800080008ULL} |
const uint64_t | ff_pw_15 = 0x000F000F000F000FULL |
const xmm_reg | ff_pw_16 = {0x0010001000100010ULL, 0x0010001000100010ULL} |
const uint64_t | ff_pw_20 = 0x0014001400140014ULL |
const xmm_reg | ff_pw_28 = {0x001C001C001C001CULL, 0x001C001C001C001CULL} |
const xmm_reg | ff_pw_32 = {0x0020002000200020ULL, 0x0020002000200020ULL} |
const uint64_t | ff_pw_42 = 0x002A002A002A002AULL |
const xmm_reg | ff_pw_64 = {0x0040004000400040ULL, 0x0040004000400040ULL} |
const uint64_t | ff_pw_96 = 0x0060006000600060ULL |
const uint64_t | ff_pw_128 = 0x0080008000800080ULL |
const uint64_t | ff_pw_255 = 0x00ff00ff00ff00ffULL |
const uint64_t | ff_pb_1 = 0x0101010101010101ULL |
const uint64_t | ff_pb_3 = 0x0303030303030303ULL |
const uint64_t | ff_pb_7 = 0x0707070707070707ULL |
const uint64_t | ff_pb_1F = 0x1F1F1F1F1F1F1F1FULL |
const uint64_t | ff_pb_3F = 0x3F3F3F3F3F3F3F3FULL |
const uint64_t | ff_pb_81 = 0x8181818181818181ULL |
const uint64_t | ff_pb_A1 = 0xA1A1A1A1A1A1A1A1ULL |
const uint64_t | ff_pb_FC = 0xFCFCFCFCFCFCFCFCULL |
const double | ff_pd_1 [2] = { 1.0, 1.0 } |
const double | ff_pd_2 [2] = { 2.0, 2.0 } |
__pad0__ | |
mm0 n mm1 n mm2 n mm3 n packssdw | mm1 |
mm0 n mm1 n mm2 n mm3 n packssdw mm0 n packssdw | mm3 |
mm0 n mm1 n mm2 n mm3 n packssdw mm0 n packssdw mm2 n movq | mm0 |
mm0 n mm1 n mm2 n mm3 n packssdw mm0 n packssdw mm2 n movq mm1 n punpcklwd | mm2 |
mm0 n mm1 n mm2 n mm3 n packssdw mm0 n packssdw mm2 n movq mm1 n punpcklwd mm0 n punpckhwd mm1 n movq n movq n | add |
mm0 n mm1 n mm2 n mm3 n packssdw mm0 n packssdw mm2 n movq mm1 n punpcklwd mm0 n punpckhwd mm1 n movq n movq n n js n emms n const float ** | src |
mm0 n mm1 n mm2 n mm3 n packssdw mm0 n packssdw mm2 n movq mm1 n punpcklwd mm0 n punpckhwd mm1 n movq n movq n n js n emms n const float long | len |
#define ABS3_MMX2 |
Value:
"psubw %%mm5, %%mm7 \n"\ "pmaxsw %%mm7, %%mm5 \n"\ "pxor %%mm6, %%mm6 \n"\ "pxor %%mm7, %%mm7 \n"\ "psubw %%mm3, %%mm6 \n"\ "psubw %%mm4, %%mm7 \n"\ "pmaxsw %%mm6, %%mm3 \n"\ "pmaxsw %%mm7, %%mm4 \n"\ "pxor %%mm7, %%mm7 \n"
Definition at line 949 of file dsputil_mmx.c.
#define ABS3_SSSE3 |
Value:
"pabsw %%mm3, %%mm3 \n"\ "pabsw %%mm4, %%mm4 \n"\ "pabsw %%mm5, %%mm5 \n"
Definition at line 960 of file dsputil_mmx.c.
#define AVG_3DNOW_OP | ( | a, | |||
b, | |||||
temp, | |||||
size | ) |
Value:
"mov" #size " " #b ", " #temp " \n\t"\ "pavgusb " #temp ", " #a " \n\t"\ "mov" #size " " #a ", " #b " \n\t"
Definition at line 1618 of file dsputil_mmx.c.
#define AVG_MMX2_OP | ( | a, | |||
b, | |||||
temp, | |||||
size | ) |
Value:
"mov" #size " " #b ", " #temp " \n\t"\ "pavgb " #temp ", " #a " \n\t"\ "mov" #size " " #a ", " #b " \n\t"
Definition at line 1622 of file dsputil_mmx.c.
#define CLEAR_BLOCKS | ( | name, | |||
n | ) |
Value:
static void name(DCTELEM *blocks)\ {\ __asm__ volatile(\ "pxor %%mm7, %%mm7 \n\t"\ "mov %1, %%"REG_a" \n\t"\ "1: \n\t"\ "movq %%mm7, (%0, %%"REG_a") \n\t"\ "movq %%mm7, 8(%0, %%"REG_a") \n\t"\ "movq %%mm7, 16(%0, %%"REG_a") \n\t"\ "movq %%mm7, 24(%0, %%"REG_a") \n\t"\ "add $32, %%"REG_a" \n\t"\ " js 1b \n\t"\ : : "r" (((uint8_t *)blocks)+128*n),\ "i" (-128*n)\ : "%"REG_a\ );\ }
Definition at line 493 of file dsputil_mmx.c.
#define DEF | ( | x | ) | x ## _mmx2 |
Definition at line 198 of file dsputil_mmx.c.
#define DEF | ( | x | ) | x ## _3dnow |
Definition at line 198 of file dsputil_mmx.c.
#define DEF | ( | x, | |||
y | ) | x ## _ ## y ##_mmx |
Definition at line 198 of file dsputil_mmx.c.
#define DEF | ( | x, | |||
y | ) | x ## _no_rnd_ ## y ##_mmx |
Definition at line 198 of file dsputil_mmx.c.
#define ff_float_to_int16_interleave6_3dn2 | ( | a, | |||
b, | |||||
c | ) | float_to_int16_interleave_misc_3dnow(a,b,c,6) |
Definition at line 2401 of file dsputil_mmx.c.
#define ff_float_to_int16_interleave6_3dnow | ( | a, | |||
b, | |||||
c | ) | float_to_int16_interleave_misc_3dnow(a,b,c,6) |
Definition at line 2400 of file dsputil_mmx.c.
#define ff_float_to_int16_interleave6_sse | ( | a, | |||
b, | |||||
c | ) | float_to_int16_interleave_misc_sse(a,b,c,6) |
Definition at line 2399 of file dsputil_mmx.c.
#define ff_float_to_int16_interleave6_sse2 ff_float_to_int16_interleave6_sse |
Definition at line 2403 of file dsputil_mmx.c.
#define FLOAT_TO_INT16_INTERLEAVE | ( | cpu, | |||
body | ) |
Definition at line 2405 of file dsputil_mmx.c.
#define H263_LOOP_FILTER |
Definition at line 634 of file dsputil_mmx.c.
Referenced by h263_h_loop_filter_mmx(), and h263_v_loop_filter_mmx().
#define H264_QPEL_FUNCS | ( | x, | |||
y, | |||||
CPU | ) |
Value:
c->put_h264_qpel_pixels_tab[0][x+y*4] = put_h264_qpel16_mc##x##y##_##CPU;\ c->put_h264_qpel_pixels_tab[1][x+y*4] = put_h264_qpel8_mc##x##y##_##CPU;\ c->avg_h264_qpel_pixels_tab[0][x+y*4] = avg_h264_qpel16_mc##x##y##_##CPU;\ c->avg_h264_qpel_pixels_tab[1][x+y*4] = avg_h264_qpel8_mc##x##y##_##CPU;
Referenced by dsputil_init_mmx().
#define IF0 | ( | x | ) |
#define IF1 | ( | x | ) | x |
#define JUMPALIGN | ( | ) | __asm__ volatile (ASMALIGN(3)::) |
#define MIX5 | ( | mono, | |||
stereo | ) |
#define MIX_MISC | ( | stereo | ) |
Value:
__asm__ volatile(\ "1: \n"\ "movaps (%3,%0), %%xmm0 \n"\ stereo("movaps %%xmm0, %%xmm1 \n")\ "mulps %%xmm6, %%xmm0 \n"\ stereo("mulps %%xmm7, %%xmm1 \n")\ "lea 1024(%3,%0), %1 \n"\ "mov %5, %2 \n"\ "2: \n"\ "movaps (%1), %%xmm2 \n"\ stereo("movaps %%xmm2, %%xmm3 \n")\ "mulps (%4,%2), %%xmm2 \n"\ stereo("mulps 16(%4,%2), %%xmm3 \n")\ "addps %%xmm2, %%xmm0 \n"\ stereo("addps %%xmm3, %%xmm1 \n")\ "add $1024, %1 \n"\ "add $32, %2 \n"\ "jl 2b \n"\ "movaps %%xmm0, (%3,%0) \n"\ stereo("movaps %%xmm1, 1024(%3,%0) \n")\ "add $16, %0 \n"\ "jl 1b \n"\ :"+&r"(i), "=&r"(j), "=&r"(k)\ :"r"(samples[0]+len), "r"(matrix_simd+in_ch), "g"((intptr_t)-32*(in_ch-1))\ :"memory"\ );
Definition at line 1982 of file dsputil_mmx.c.
Referenced by ac3_downmix_sse().
#define MOVQ_BFE | ( | regd | ) |
Value:
__asm__ volatile ( \ "pcmpeqd %%" #regd ", %%" #regd " \n\t"\ "paddb %%" #regd ", %%" #regd " \n\t" ::)
Definition at line 79 of file dsputil_mmx.c.
Referenced by pixels16(), pixels16_l2(), pixels16_x2(), pixels4(), pixels8(), pixels8_l2(), pixels8_x2(), and pixels8_y2().
#define MOVQ_BONE | ( | regd | ) | __asm__ volatile ("movq %0, %%" #regd " \n\t" ::"m"(ff_bone)) |
Definition at line 85 of file dsputil_mmx.c.
Referenced by avg_pixels8_xy2_TMPL(), put_no_rnd_pixels8_x2_TMPL(), and put_no_rnd_pixels8_y2_TMPL().
#define MOVQ_WTWO | ( | regd | ) | __asm__ volatile ("movq %0, %%" #regd " \n\t" ::"m"(ff_wtwo)) |
Definition at line 86 of file dsputil_mmx.c.
#define MOVQ_ZERO | ( | regd | ) | __asm__ volatile ("pxor %%" #regd ", %%" #regd ::) |
Definition at line 77 of file dsputil_mmx.c.
Referenced by add_pixels_clamped_mmx(), and pixels8_xy2().
#define OP_AVG PAVGB |
Definition at line 202 of file dsputil_mmx.c.
#define OP_AVG PAVGB |
Definition at line 202 of file dsputil_mmx.c.
#define OP_AVG | ( | a, | |||
b, | |||||
c, | |||||
e | ) | PAVGB_MMX(a, b, c, e) |
Definition at line 202 of file dsputil_mmx.c.
#define PAETH | ( | cpu, | |||
abs3 | ) |
Definition at line 896 of file dsputil_mmx.c.
#define PAVGB "pavgb" |
Definition at line 201 of file dsputil_mmx.c.
#define PAVGB "pavgusb" |
Definition at line 201 of file dsputil_mmx.c.
#define PAVGB | ( | a, | |||
b, | |||||
c, | |||||
e | ) | PAVGB_MMX(a, b, c, e) |
Definition at line 201 of file dsputil_mmx.c.
#define PAVGB | ( | a, | |||
b, | |||||
c, | |||||
e | ) | PAVGB_MMX_NO_RND(a, b, c, e) |
Definition at line 201 of file dsputil_mmx.c.
Referenced by avg_pixels16_l2_TMPL(), avg_pixels4_l2_TMPL(), avg_pixels4_TMPL(), avg_pixels8_l2_TMPL(), avg_pixels8_TMPL(), avg_pixels8_x2_TMPL(), avg_pixels8_xy2_TMPL(), avg_pixels8_y2_TMPL(), deInterlaceBlendLinear_TMPL(), deInterlaceInterpolateLinear_TMPL(), dering_TMPL(), doVertDefFilter_TMPL(), doVertLowPass_TMPL(), extract_even2avg_TMPL(), extract_odd2avg_TMPL(), pixels8_l2(), planar2x_TMPL(), put_no_rnd_pixels16_l2_TMPL(), put_no_rnd_pixels8_l2_TMPL(), put_no_rnd_pixels8_x2_TMPL(), put_no_rnd_pixels8_y2_TMPL(), put_pixels16_l2_TMPL(), put_pixels16_x2_TMPL(), put_pixels4_l2_TMPL(), put_pixels8_l2_TMPL(), put_pixels8_x2_TMPL(), put_pixels8_y2_TMPL(), rgb24toyv12_TMPL(), tempNoiseReducer_TMPL(), and vertX1Filter_TMPL().
#define PAVGB_MMX | ( | rega, | |||
regb, | |||||
regr, | |||||
regfe | ) |
Value:
"movq " #rega ", " #regr " \n\t"\ "por " #regb ", " #regr " \n\t"\ "pxor " #rega ", " #regb " \n\t"\ "pand " #regfe "," #regb " \n\t"\ "psrlq $1, " #regb " \n\t"\ "psubb " #regb ", " #regr " \n\t"
Definition at line 115 of file dsputil_mmx.c.
#define PAVGB_MMX_NO_RND | ( | rega, | |||
regb, | |||||
regr, | |||||
regfe | ) |
Value:
"movq " #rega ", " #regr " \n\t"\ "pand " #regb ", " #regr " \n\t"\ "pxor " #rega ", " #regb " \n\t"\ "pand " #regfe "," #regb " \n\t"\ "psrlq $1, " #regb " \n\t"\ "paddb " #regb ", " #regr " \n\t"
Definition at line 107 of file dsputil_mmx.c.
#define PAVGBP | ( | a, | |||
b, | |||||
c, | |||||
d, | |||||
e, | |||||
f | ) | PAVGBP_MMX(a, b, c, d, e, f) |
Definition at line 171 of file dsputil_mmx.c.
#define PAVGBP | ( | a, | |||
b, | |||||
c, | |||||
d, | |||||
e, | |||||
f | ) | PAVGBP_MMX_NO_RND(a, b, c, d, e, f) |
Definition at line 171 of file dsputil_mmx.c.
Referenced by pixels16_l2(), pixels16_x2(), pixels8_l2(), pixels8_x2(), and pixels8_y2().
#define PAVGBP_MMX | ( | rega, | |||
regb, | |||||
regr, | |||||
regc, | |||||
regd, | |||||
regp | ) |
Value:
"movq " #rega ", " #regr " \n\t"\ "movq " #regc ", " #regp " \n\t"\ "por " #regb ", " #regr " \n\t"\ "por " #regd ", " #regp " \n\t"\ "pxor " #rega ", " #regb " \n\t"\ "pxor " #regc ", " #regd " \n\t"\ "pand %%mm6, " #regb " \n\t"\ "pand %%mm6, " #regd " \n\t"\ "psrlq $1, " #regd " \n\t"\ "psrlq $1, " #regb " \n\t"\ "psubb " #regb ", " #regr " \n\t"\ "psubb " #regd ", " #regp " \n\t"
Definition at line 138 of file dsputil_mmx.c.
#define PAVGBP_MMX_NO_RND | ( | rega, | |||
regb, | |||||
regr, | |||||
regc, | |||||
regd, | |||||
regp | ) |
Value:
"movq " #rega ", " #regr " \n\t"\ "movq " #regc ", " #regp " \n\t"\ "pand " #regb ", " #regr " \n\t"\ "pand " #regd ", " #regp " \n\t"\ "pxor " #rega ", " #regb " \n\t"\ "pxor " #regc ", " #regd " \n\t"\ "pand %%mm6, " #regb " \n\t"\ "pand %%mm6, " #regd " \n\t"\ "psrlq $1, " #regb " \n\t"\ "psrlq $1, " #regd " \n\t"\ "paddb " #regb ", " #regr " \n\t"\ "paddb " #regd ", " #regp " \n\t"
Definition at line 124 of file dsputil_mmx.c.
#define PREFETCH | ( | name, | |||
op | ) |
Value:
static void name(void *mem, int stride, int h){\ const uint8_t *p= mem;\ do{\ __asm__ volatile(#op" %0" :: "m"(*p));\ p+= stride;\ }while(--h);\ }
Definition at line 1806 of file dsputil_mmx.c.
Referenced by hcscale_fast_TMPL(), hyscale_fast_TMPL(), interleaveBytes_TMPL(), rgb15to16_TMPL(), rgb15to32_TMPL(), rgb15tobgr24_TMPL(), rgb16to15_TMPL(), rgb16to32_TMPL(), rgb16tobgr24_TMPL(), rgb24to15_TMPL(), rgb24to16_TMPL(), rgb24tobgr15_TMPL(), rgb24tobgr16_TMPL(), rgb24tobgr24_TMPL(), rgb24tobgr32_TMPL(), rgb24toyv12_TMPL(), rgb32to15_TMPL(), rgb32to16_TMPL(), rgb32tobgr15_TMPL(), rgb32tobgr16_TMPL(), rgb32tobgr24_TMPL(), rgb32tobgr32_TMPL(), uyvytoyv12_TMPL(), vu9_to_vu12_TMPL(), yuvPlanartouyvy_TMPL(), yuvPlanartoyuy2_TMPL(), yuy2toyv12_TMPL(), and yvu9_to_yuy2_TMPL().
#define put_no_rnd_pixels16_3dnow put_no_rnd_pixels16_mmx |
Definition at line 220 of file dsputil_mmx.c.
#define put_no_rnd_pixels16_mmx put_pixels16_mmx |
Definition at line 210 of file dsputil_mmx.c.
#define put_no_rnd_pixels16_mmx2 put_no_rnd_pixels16_mmx |
Definition at line 215 of file dsputil_mmx.c.
#define put_no_rnd_pixels8_3dnow put_no_rnd_pixels8_mmx |
Definition at line 221 of file dsputil_mmx.c.
#define put_no_rnd_pixels8_mmx put_pixels8_mmx |
Definition at line 211 of file dsputil_mmx.c.
#define put_no_rnd_pixels8_mmx2 put_no_rnd_pixels8_mmx |
Definition at line 216 of file dsputil_mmx.c.
Definition at line 1617 of file dsputil_mmx.c.
#define put_pixels16_3dnow put_pixels16_mmx |
Definition at line 217 of file dsputil_mmx.c.
#define put_pixels16_mmx2 put_pixels16_mmx |
Definition at line 212 of file dsputil_mmx.c.
#define put_pixels4_3dnow put_pixels4_mmx |
Definition at line 219 of file dsputil_mmx.c.
#define put_pixels4_mmx2 put_pixels4_mmx |
Definition at line 214 of file dsputil_mmx.c.
#define put_pixels8_3dnow put_pixels8_mmx |
Definition at line 218 of file dsputil_mmx.c.
#define put_pixels8_mmx2 put_pixels8_mmx |
Definition at line 213 of file dsputil_mmx.c.
#define put_signed_pixels_clamped_mmx_half | ( | off | ) |
Value:
"movq "#off"(%2), %%mm1 \n\t"\ "movq 16+"#off"(%2), %%mm2 \n\t"\ "movq 32+"#off"(%2), %%mm3 \n\t"\ "movq 48+"#off"(%2), %%mm4 \n\t"\ "packsswb 8+"#off"(%2), %%mm1 \n\t"\ "packsswb 24+"#off"(%2), %%mm2 \n\t"\ "packsswb 40+"#off"(%2), %%mm3 \n\t"\ "packsswb 56+"#off"(%2), %%mm4 \n\t"\ "paddb %%mm0, %%mm1 \n\t"\ "paddb %%mm0, %%mm2 \n\t"\ "paddb %%mm0, %%mm3 \n\t"\ "paddb %%mm0, %%mm4 \n\t"\ "movq %%mm1, (%0) \n\t"\ "movq %%mm2, (%0, %3) \n\t"\ "movq %%mm3, (%0, %3, 2) \n\t"\ "movq %%mm4, (%0, %1) \n\t"
Definition at line 284 of file dsputil_mmx.c.
Referenced by put_signed_pixels_clamped_mmx().
#define QPEL_2TAP | ( | OPNAME, | |||
SIZE, | |||||
MMX | ) |
Value:
QPEL_2TAP_XY(OPNAME, SIZE, MMX, 20, _x2_ ## MMX)\ QPEL_2TAP_XY(OPNAME, SIZE, MMX, 02, _y2_ ## MMX)\ QPEL_2TAP_XY(OPNAME, SIZE, MMX, 22, _xy2_mmx)\ static const qpel_mc_func OPNAME ## 2tap_qpel ## SIZE ## _mc00_ ## MMX =\ OPNAME ## qpel ## SIZE ## _mc00_ ## MMX;\ static const qpel_mc_func OPNAME ## 2tap_qpel ## SIZE ## _mc21_ ## MMX =\ OPNAME ## 2tap_qpel ## SIZE ## _mc20_ ## MMX;\ static const qpel_mc_func OPNAME ## 2tap_qpel ## SIZE ## _mc12_ ## MMX =\ OPNAME ## 2tap_qpel ## SIZE ## _mc02_ ## MMX;\ static void OPNAME ## 2tap_qpel ## SIZE ## _mc32_ ## MMX(uint8_t *dst, uint8_t *src, int stride){\ OPNAME ## pixels ## SIZE ## _y2_ ## MMX(dst, src+1, stride, SIZE);\ }\ static void OPNAME ## 2tap_qpel ## SIZE ## _mc23_ ## MMX(uint8_t *dst, uint8_t *src, int stride){\ OPNAME ## pixels ## SIZE ## _x2_ ## MMX(dst, src+stride, stride, SIZE);\ }\ QPEL_2TAP_L3(OPNAME, SIZE, MMX, 10, 0, 1, 0)\ QPEL_2TAP_L3(OPNAME, SIZE, MMX, 30, 1, -1, 0)\ QPEL_2TAP_L3(OPNAME, SIZE, MMX, 01, 0, stride, 0)\ QPEL_2TAP_L3(OPNAME, SIZE, MMX, 03, stride, -stride, 0)\ QPEL_2TAP_L3(OPNAME, SIZE, MMX, 11, 0, stride, 1)\ QPEL_2TAP_L3(OPNAME, SIZE, MMX, 31, 1, stride, -1)\ QPEL_2TAP_L3(OPNAME, SIZE, MMX, 13, stride, -stride, 1)\ QPEL_2TAP_L3(OPNAME, SIZE, MMX, 33, stride+1, -stride, -1)\
Definition at line 1649 of file dsputil_mmx.c.
#define QPEL_2TAP_L3 | ( | OPNAME, | |||
SIZE, | |||||
MMX, | |||||
XY, | |||||
S0, | |||||
S1, | |||||
S2 | ) |
#define QPEL_2TAP_XY | ( | OPNAME, | |||
SIZE, | |||||
MMX, | |||||
XY, | |||||
HPEL | ) |
#define QPEL_BASE | ( | OPNAME, | |||
ROUNDER, | |||||
RND, | |||||
OP_MMX2, | |||||
OP_3DNOW | ) |
Definition at line 991 of file dsputil_mmx.c.
#define QPEL_OP | ( | OPNAME, | |||
ROUNDER, | |||||
RND, | |||||
OP, | |||||
MMX | ) |
Definition at line 1254 of file dsputil_mmx.c.
#define QPEL_V_LOW | ( | m3, | |||
m4, | |||||
m5, | |||||
m6, | |||||
pw_20, | |||||
pw_3, | |||||
rnd, | |||||
in0, | |||||
in1, | |||||
in2, | |||||
in7, | |||||
out, | |||||
OP | ) |
Value:
"paddw " #m4 ", " #m3 " \n\t" /* x1 */\ "movq "MANGLE(ff_pw_20)", %%mm4 \n\t" /* 20 */\ "pmullw " #m3 ", %%mm4 \n\t" /* 20x1 */\ "movq "#in7", " #m3 " \n\t" /* d */\ "movq "#in0", %%mm5 \n\t" /* D */\ "paddw " #m3 ", %%mm5 \n\t" /* x4 */\ "psubw %%mm5, %%mm4 \n\t" /* 20x1 - x4 */\ "movq "#in1", %%mm5 \n\t" /* C */\ "movq "#in2", %%mm6 \n\t" /* B */\ "paddw " #m6 ", %%mm5 \n\t" /* x3 */\ "paddw " #m5 ", %%mm6 \n\t" /* x2 */\ "paddw %%mm6, %%mm6 \n\t" /* 2x2 */\ "psubw %%mm6, %%mm5 \n\t" /* -2x2 + x3 */\ "pmullw "MANGLE(ff_pw_3)", %%mm5 \n\t" /* -6x2 + 3x3 */\ "paddw " #rnd ", %%mm4 \n\t" /* x2 */\ "paddw %%mm4, %%mm5 \n\t" /* 20x1 - 6x2 + 3x3 - x4 */\ "psraw $5, %%mm5 \n\t"\ "packuswb %%mm5, %%mm5 \n\t"\ OP(%%mm5, out, %%mm7, d)
Definition at line 970 of file dsputil_mmx.c.
#define SET_HPEL_FUNCS | ( | PFX, | |||
IDX, | |||||
SIZE, | |||||
CPU | ) |
Value:
c->PFX ## _pixels_tab[IDX][0] = PFX ## _pixels ## SIZE ## _ ## CPU; \ c->PFX ## _pixels_tab[IDX][1] = PFX ## _pixels ## SIZE ## _x2_ ## CPU; \ c->PFX ## _pixels_tab[IDX][2] = PFX ## _pixels ## SIZE ## _y2_ ## CPU; \ c->PFX ## _pixels_tab[IDX][3] = PFX ## _pixels ## SIZE ## _xy2_ ## CPU
Referenced by dsputil_init_mmx().
#define SET_QPEL_FUNCS | ( | PFX, | |||
IDX, | |||||
SIZE, | |||||
CPU | ) |
Value:
c->PFX ## _pixels_tab[IDX][ 0] = PFX ## SIZE ## _mc00_ ## CPU; \ c->PFX ## _pixels_tab[IDX][ 1] = PFX ## SIZE ## _mc10_ ## CPU; \ c->PFX ## _pixels_tab[IDX][ 2] = PFX ## SIZE ## _mc20_ ## CPU; \ c->PFX ## _pixels_tab[IDX][ 3] = PFX ## SIZE ## _mc30_ ## CPU; \ c->PFX ## _pixels_tab[IDX][ 4] = PFX ## SIZE ## _mc01_ ## CPU; \ c->PFX ## _pixels_tab[IDX][ 5] = PFX ## SIZE ## _mc11_ ## CPU; \ c->PFX ## _pixels_tab[IDX][ 6] = PFX ## SIZE ## _mc21_ ## CPU; \ c->PFX ## _pixels_tab[IDX][ 7] = PFX ## SIZE ## _mc31_ ## CPU; \ c->PFX ## _pixels_tab[IDX][ 8] = PFX ## SIZE ## _mc02_ ## CPU; \ c->PFX ## _pixels_tab[IDX][ 9] = PFX ## SIZE ## _mc12_ ## CPU; \ c->PFX ## _pixels_tab[IDX][10] = PFX ## SIZE ## _mc22_ ## CPU; \ c->PFX ## _pixels_tab[IDX][11] = PFX ## SIZE ## _mc32_ ## CPU; \ c->PFX ## _pixels_tab[IDX][12] = PFX ## SIZE ## _mc03_ ## CPU; \ c->PFX ## _pixels_tab[IDX][13] = PFX ## SIZE ## _mc13_ ## CPU; \ c->PFX ## _pixels_tab[IDX][14] = PFX ## SIZE ## _mc23_ ## CPU; \ c->PFX ## _pixels_tab[IDX][15] = PFX ## SIZE ## _mc33_ ## CPU
Referenced by dsputil_init_mmx().
#define SET_RND MOVQ_WTWO |
Definition at line 170 of file dsputil_mmx.c.
#define SET_RND MOVQ_WONE |
Definition at line 170 of file dsputil_mmx.c.
static void ac3_downmix_sse | ( | float(*) | samples[256], | |
float(*) | matrix[2], | |||
int | out_ch, | |||
int | in_ch, | |||
int | len | |||
) | [static] |
static void add_bytes_l2_mmx | ( | uint8_t * | dst, | |
uint8_t * | src1, | |||
uint8_t * | src2, | |||
int | w | |||
) | [static] |
static void add_bytes_mmx | ( | uint8_t * | dst, | |
uint8_t * | src, | |||
int | w | |||
) | [static] |
void add_pixels_clamped_mmx | ( | const DCTELEM * | block, | |
uint8_t * | pixels, | |||
int | line_size | |||
) |
Definition at line 318 of file dsputil_mmx.c.
Referenced by cavs_idct8_add_mmx(), dsputil_init_mmx(), ff_h264_idct8_add_mmx(), ff_idct_xvid_mmx2_add(), ff_idct_xvid_mmx_add(), ff_idct_xvid_sse2_add(), ff_simple_idct_add_mmx(), ff_vp3_idct_add_mmx(), and ff_vp3_idct_add_sse2().
static void avg_pixels16_sse2 | ( | uint8_t * | block, | |
const uint8_t * | pixels, | |||
int | line_size, | |||
int | h | |||
) | [static] |
static void clear_block_sse | ( | DCTELEM * | block | ) | [static] |
static void clear_blocks_sse | ( | DCTELEM * | blocks | ) | [static] |
DECLARE_ASM_CONST | ( | 8 | , | |
uint8_t | , | |||
ff_vector128 | ||||
) |
static void draw_edges_mmx | ( | uint8_t * | buf, | |
int | wrap, | |||
int | width, | |||
int | height, | |||
int | w | |||
) | [static] |
void dsputil_init_mmx | ( | DSPContext * | c, | |
AVCodecContext * | avctx | |||
) |
int ff_add_hfyu_left_prediction_sse4 | ( | uint8_t * | dst, | |
const uint8_t * | src, | |||
int | w, | |||
int | left | |||
) |
Referenced by dsputil_init_mmx().
int ff_add_hfyu_left_prediction_ssse3 | ( | uint8_t * | dst, | |
const uint8_t * | src, | |||
int | w, | |||
int | left | |||
) |
Referenced by dsputil_init_mmx().
void ff_add_hfyu_median_prediction_mmx2 | ( | uint8_t * | dst, | |
const uint8_t * | top, | |||
const uint8_t * | diff, | |||
int | w, | |||
int * | left, | |||
int * | left_top | |||
) |
Referenced by dsputil_init_mmx().
void ff_avg_cavs_qpel16_mc00_mmx2 | ( | uint8_t * | dst, | |
uint8_t * | src, | |||
int | stride | |||
) |
Definition at line 1831 of file dsputil_mmx.c.
void ff_avg_cavs_qpel8_mc00_mmx2 | ( | uint8_t * | dst, | |
uint8_t * | src, | |||
int | stride | |||
) |
Definition at line 1825 of file dsputil_mmx.c.
void ff_avg_vc1_mspel_mc00_mmx2 | ( | uint8_t * | dst, | |
const uint8_t * | src, | |||
int | stride, | |||
int | rnd | |||
) |
void ff_float_to_int16_interleave6_3dn2 | ( | int16_t * | dst, | |
const float ** | src, | |||
int | len | |||
) |
void ff_float_to_int16_interleave6_3dnow | ( | int16_t * | dst, | |
const float ** | src, | |||
int | len | |||
) |
void ff_float_to_int16_interleave6_sse | ( | int16_t * | dst, | |
const float ** | src, | |||
int | len | |||
) |
static void ff_idct_xvid_mmx2_add | ( | uint8_t * | dest, | |
int | line_size, | |||
DCTELEM * | block | |||
) | [static] |
static void ff_idct_xvid_mmx2_put | ( | uint8_t * | dest, | |
int | line_size, | |||
DCTELEM * | block | |||
) | [static] |
static void ff_idct_xvid_mmx_add | ( | uint8_t * | dest, | |
int | line_size, | |||
DCTELEM * | block | |||
) | [static] |
static void ff_idct_xvid_mmx_put | ( | uint8_t * | dest, | |
int | line_size, | |||
DCTELEM * | block | |||
) | [static] |
void ff_put_cavs_qpel16_mc00_mmx2 | ( | uint8_t * | dst, | |
uint8_t * | src, | |||
int | stride | |||
) |
Definition at line 1828 of file dsputil_mmx.c.
void ff_put_cavs_qpel8_mc00_mmx2 | ( | uint8_t * | dst, | |
uint8_t * | src, | |||
int | stride | |||
) |
Definition at line 1822 of file dsputil_mmx.c.
void ff_put_vc1_mspel_mc00_mmx | ( | uint8_t * | dst, | |
const uint8_t * | src, | |||
int | stride, | |||
int | rnd | |||
) |
int32_t ff_scalarproduct_and_madd_int16_mmx2 | ( | int16_t * | v1, | |
int16_t * | v2, | |||
int16_t * | v3, | |||
int | order, | |||
int | mul | |||
) |
Referenced by dsputil_init_mmx().
int32_t ff_scalarproduct_and_madd_int16_sse2 | ( | int16_t * | v1, | |
int16_t * | v2, | |||
int16_t * | v3, | |||
int | order, | |||
int | mul | |||
) |
Referenced by dsputil_init_mmx().
int32_t ff_scalarproduct_and_madd_int16_ssse3 | ( | int16_t * | v1, | |
int16_t * | v2, | |||
int16_t * | v3, | |||
int | order, | |||
int | mul | |||
) |
Referenced by dsputil_init_mmx().
int32_t ff_scalarproduct_int16_mmx2 | ( | int16_t * | v1, | |
int16_t * | v2, | |||
int | order, | |||
int | shift | |||
) |
Referenced by dsputil_init_mmx().
int32_t ff_scalarproduct_int16_sse2 | ( | int16_t * | v1, | |
int16_t * | v2, | |||
int | order, | |||
int | shift | |||
) |
Referenced by dsputil_init_mmx().
void ff_x264_deblock_h_luma_intra_mmxext | ( | uint8_t * | pix, | |
int | stride, | |||
int | alpha, | |||
int | beta | |||
) |
void ff_x264_deblock_h_luma_intra_sse2 | ( | uint8_t * | pix, | |
int | stride, | |||
int | alpha, | |||
int | beta | |||
) |
void ff_x264_deblock_h_luma_sse2 | ( | uint8_t * | pix, | |
int | stride, | |||
int | alpha, | |||
int | beta, | |||
int8_t * | tc0 | |||
) |
void ff_x264_deblock_v_luma_intra_sse2 | ( | uint8_t * | pix, | |
int | stride, | |||
int | alpha, | |||
int | beta | |||
) |
void ff_x264_deblock_v_luma_sse2 | ( | uint8_t * | pix, | |
int | stride, | |||
int | alpha, | |||
int | beta, | |||
int8_t * | tc0 | |||
) |
static void float_to_int16_3dnow | ( | int16_t * | dst, | |
const float * | src, | |||
long | len | |||
) | [static] |
mm0 n mm1 n mm2 n mm3 n packssdw mm0 n packssdw mm2 n movq mm1 n punpcklwd mm0 n punpckhwd mm1 n movq n movq n n js n emms n FLOAT_TO_INT16_INTERLEAVE | ( | sse2 | , | |
"1: \n""cvtps2dq | %2,%0, | |||
%%xmm0\n" " | cvtps2dq%3,%0, | |||
%%xmm1\n" "packssdw%% | xmm1, | |||
%%xmm0\n" "movhlps%% | xmm0, | |||
%%xmm1\n" "punpcklwd%% | xmm1, | |||
%%xmm0\n" "movdqa%% | xmm0, | |||
(%1,%0)\n" "add $ | 16, | |||
%0\n" "js 1b\n" | ||||
) |
FLOAT_TO_INT16_INTERLEAVE | ( | 3dnow | , | |
"1: \n""pf2id | %2,%0, | |||
%%mm0\n" "pf2id | 8%2,%0, | |||
%%mm1\n" " | pf2id%3,%0, | |||
%%mm2\n" "pf2id | 8%3,%0, | |||
%%mm3\n" "packssdw%% | mm1, | |||
%%mm0\n" "packssdw%% | mm3, | |||
%%mm2\n" "movq%% | mm0, | |||
%%mm1\n" "punpcklwd%% | mm2, | |||
%%mm0\n" "punpckhwd%% | mm2, | |||
%%mm1\n" "movq%% | mm0, | |||
(%1,%0)\n" "movq%% | mm1, | |||
8(%1,%0)\n" "add $ | 16, | |||
%0\n" "js 1b\n" "femms\n" | ||||
) |
static void float_to_int16_sse | ( | int16_t * | dst, | |
const float * | src, | |||
long | len | |||
) | [static] |
static void float_to_int16_sse2 | ( | int16_t * | dst, | |
const float * | src, | |||
long | len | |||
) | [static] |
static void gmc_mmx | ( | uint8_t * | dst, | |
uint8_t * | src, | |||
int | stride, | |||
int | h, | |||
int | ox, | |||
int | oy, | |||
int | dxx, | |||
int | dxy, | |||
int | dyx, | |||
int | dyy, | |||
int | shift, | |||
int | r, | |||
int | width, | |||
int | height | |||
) | [static] |
static void h263_h_loop_filter_mmx | ( | uint8_t * | src, | |
int | stride, | |||
int | qscale | |||
) | [static] |
static void h263_v_loop_filter_mmx | ( | uint8_t * | src, | |
int | stride, | |||
int | qscale | |||
) | [static] |
static void int32_to_float_fmul_scalar_sse | ( | float * | dst, | |
const int * | src, | |||
float | mul, | |||
int | len | |||
) | [static] |
static void int32_to_float_fmul_scalar_sse2 | ( | float * | dst, | |
const int * | src, | |||
float | mul, | |||
int | len | |||
) | [static] |
static void put_pixels16_mmx | ( | uint8_t * | block, | |
const uint8_t * | pixels, | |||
int | line_size, | |||
int | h | |||
) | [static] |
static void put_pixels16_sse2 | ( | uint8_t * | block, | |
const uint8_t * | pixels, | |||
int | line_size, | |||
int | h | |||
) | [static] |
static void put_pixels4_mmx | ( | uint8_t * | block, | |
const uint8_t * | pixels, | |||
int | line_size, | |||
int | h | |||
) | [static] |
Definition at line 359 of file dsputil_mmx.c.
static void put_pixels8_mmx | ( | uint8_t * | block, | |
const uint8_t * | pixels, | |||
int | line_size, | |||
int | h | |||
) | [static] |
Definition at line 385 of file dsputil_mmx.c.
Referenced by ff_put_cavs_qpel8_mc00_mmx2(), and ff_put_vc1_mspel_mc00_mmx().
void put_pixels_clamped_mmx | ( | const DCTELEM * | block, | |
uint8_t * | pixels, | |||
int | line_size | |||
) |
Definition at line 226 of file dsputil_mmx.c.
Referenced by dsputil_init_mmx(), ff_idct_xvid_mmx2_put(), ff_idct_xvid_mmx_put(), ff_idct_xvid_sse2_put(), and ff_simple_idct_put_mmx().
void put_signed_pixels_clamped_mmx | ( | const DCTELEM * | block, | |
uint8_t * | pixels, | |||
int | line_size | |||
) |
Definition at line 302 of file dsputil_mmx.c.
Referenced by dsputil_init_mmx(), ff_vp3_idct_put_mmx(), and ff_vp3_idct_put_sse2().
static void transpose4x4 | ( | uint8_t * | dst, | |
uint8_t * | src, | |||
int | dst_stride, | |||
int | src_stride | |||
) | [inline, static] |
Definition at line 726 of file dsputil_mmx.c.
Referenced by h263_h_loop_filter_mmx(), h264_h_loop_filter_chroma_intra_mmx2(), h264_h_loop_filter_chroma_mmx2(), and h264_h_loop_filter_luma_mmx2().
static void vector_clipf_sse | ( | float * | dst, | |
const float * | src, | |||
float | min, | |||
float | max, | |||
int | len | |||
) | [static] |
static void vector_fmul_3dnow | ( | float * | dst, | |
const float * | src, | |||
int | len | |||
) | [static] |
static void vector_fmul_add_3dnow | ( | float * | dst, | |
const float * | src0, | |||
const float * | src1, | |||
const float * | src2, | |||
int | len | |||
) | [static] |
static void vector_fmul_add_sse | ( | float * | dst, | |
const float * | src0, | |||
const float * | src1, | |||
const float * | src2, | |||
int | len | |||
) | [static] |
static void vector_fmul_reverse_3dnow2 | ( | float * | dst, | |
const float * | src0, | |||
const float * | src1, | |||
int | len | |||
) | [static] |
static void vector_fmul_reverse_sse | ( | float * | dst, | |
const float * | src0, | |||
const float * | src1, | |||
int | len | |||
) | [static] |
static void vector_fmul_sse | ( | float * | dst, | |
const float * | src, | |||
int | len | |||
) | [static] |
static void vector_fmul_window_3dnow2 | ( | float * | dst, | |
const float * | src0, | |||
const float * | src1, | |||
const float * | win, | |||
float | add_bias, | |||
int | len | |||
) | [static] |
static void vector_fmul_window_sse | ( | float * | dst, | |
const float * | src0, | |||
const float * | src1, | |||
const float * | win, | |||
float | add_bias, | |||
int | len | |||
) | [static] |
static void vorbis_inverse_coupling_3dnow | ( | float * | mag, | |
float * | ang, | |||
int | blocksize | |||
) | [static] |
static void vorbis_inverse_coupling_sse | ( | float * | mag, | |
float * | ang, | |||
int | blocksize | |||
) | [static] |
Definition at line 2459 of file dsputil_mmx.c.
mm0 n mm1 n mm2 n mm3 n packssdw mm0 n packssdw mm2 n movq mm1 n punpcklwd mm0 n punpckhwd mm1 n movq n movq n add |
Definition at line 2463 of file dsputil_mmx.c.
const uint64_t ff_bone = 0x0101010101010101ULL |
Definition at line 43 of file dsputil_mmx.c.
Referenced by h264_loop_filter_chroma_intra_mmx2(), h264_loop_filter_chroma_mmx2(), and h264_loop_filter_luma_mmx2().
const uint64_t ff_pb_1 = 0x0101010101010101ULL |
const uint64_t ff_pb_1F = 0x1F1F1F1F1F1F1F1FULL |
Definition at line 67 of file dsputil_mmx.c.
const uint64_t ff_pb_3 = 0x0303030303030303ULL |
const uint64_t ff_pb_3F = 0x3F3F3F3F3F3F3F3FULL |
const uint64_t ff_pb_7 = 0x0707070707070707ULL |
Definition at line 66 of file dsputil_mmx.c.
const uint64_t ff_pb_81 = 0x8181818181818181ULL |
Definition at line 69 of file dsputil_mmx.c.
const uint64_t ff_pb_A1 = 0xA1A1A1A1A1A1A1A1ULL |
Definition at line 70 of file dsputil_mmx.c.
const uint64_t ff_pb_FC = 0xFCFCFCFCFCFCFCFCULL |
Definition at line 71 of file dsputil_mmx.c.
Referenced by h263_h_loop_filter_mmx(), and h263_v_loop_filter_mmx().
const double ff_pd_1[2] = { 1.0, 1.0 } |
Definition at line 73 of file dsputil_mmx.c.
Referenced by apply_welch_window_sse2(), and ff_lpc_compute_autocorr_sse2().
const double ff_pd_2[2] = { 2.0, 2.0 } |
const uint64_t ff_pdw_80000000[2] |
Initial value:
{0x8000000080000000ULL, 0x8000000080000000ULL}
Definition at line 46 of file dsputil_mmx.c.
Referenced by vorbis_inverse_coupling_sse().
const uint64_t ff_pw_128 = 0x0080008000800080ULL |
Definition at line 61 of file dsputil_mmx.c.
const uint64_t ff_pw_15 = 0x000F000F000F000FULL |
Definition at line 53 of file dsputil_mmx.c.
const uint64_t ff_pw_20 = 0x0014001400140014ULL |
Definition at line 55 of file dsputil_mmx.c.
const uint64_t ff_pw_255 = 0x00ff00ff00ff00ffULL |
Definition at line 62 of file dsputil_mmx.c.
const uint64_t ff_pw_3 = 0x0003000300030003ULL |
Definition at line 57 of file dsputil_mmx.c.
Referenced by ff_h264_idct8_add_sse2(), ff_h264_idct_add_mmx(), H264_CHROMA_MC4_TMPL(), and H264_CHROMA_MC8_TMPL().
const uint64_t ff_pw_4 = 0x0004000400040004ULL |
Definition at line 50 of file dsputil_mmx.c.
Referenced by cavs_idct8_add_mmx(), deinterlace_line(), deinterlace_line_inplace(), and H264_CHROMA_MC8_TMPL().
const uint64_t ff_pw_42 = 0x002A002A002A002AULL |
Definition at line 58 of file dsputil_mmx.c.
Definition at line 59 of file dsputil_mmx.c.
Referenced by cavs_idct8_add_mmx(), ff_vp6_filter_diag4_mmx(), and H264_CHROMA_MC8_TMPL().
Definition at line 52 of file dsputil_mmx.c.
Referenced by ff_vp3_idct_mmx(), ff_vp3_idct_sse2(), H264_CHROMA_MC4_TMPL(), and H264_CHROMA_MC8_TMPL().
const uint64_t ff_pw_96 = 0x0060006000600060ULL |
Definition at line 60 of file dsputil_mmx.c.
const uint64_t ff_wtwo = 0x0002000200020002ULL |
Definition at line 44 of file dsputil_mmx.c.
mm0 n mm1 n mm2 n mm3 n packssdw mm0 n packssdw mm2 n movq mm1 n punpcklwd mm0 n punpckhwd mm1 n movq n movq n n js n emms n const float long len |
Definition at line 2487 of file dsputil_mmx.c.
mm0 n mm1 n mm2 n mm3 n packssdw mm0 n packssdw mm2 n movq mm1 n punpcklwd mm0 n punpckhwd mm1 n movq mm0 |
Definition at line 2463 of file dsputil_mmx.c.
Referenced by add_8x8basis_TMPL(), cavs_idct8_1d(), cavs_idct8_add_mmx(), deInterlaceBlendLinear_TMPL(), deInterlaceInterpolateLinear_TMPL(), deInterlaceL5_TMPL(), dering_TMPL(), doVertDefFilter_TMPL(), doVertLowPass_TMPL(), ff_h264_idct8_add_mmx(), ff_h264_idct_add_mmx(), ff_imdct_calc_3dn2(), ff_imdct_half_3dn2(), ff_vp3_h_loop_filter_mmx2(), H264_CHROMA_MC4_TMPL(), H264_CHROMA_MC8_TMPL(), h264_idct8_1d(), h264_loop_filter_chroma_intra_mmx2(), h264_loop_filter_luma_mmx2(), h264_loop_filter_strength_mmx2(), idct_col(), mmx_row(), mmx_row_head(), mmx_row_mid(), mmx_row_tail(), mmxext_row(), mmxext_row_head(), mmxext_row_mid(), mmxext_row_tail(), pixels16(), pixels16_l2(), pixels16_x2(), pixels4(), pixels8(), pixels8_l2(), pixels8_x2(), pixels8_y2(), tempNoiseReducer_TMPL(), try_8x8basis_TMPL(), vertX1Filter_TMPL(), vsad16_mmx(), vsad16_mmx2(), vsad_intra16_mmx(), vsad_intra16_mmx2(), yuv2packed2_TMPL(), and yuv2packedX_TMPL().
mm0 n mm1 n mm2 n mm3 n packssdw mm0 n packssdw mm2 n movq mm1 n punpcklwd mm0 n punpckhwd mm1 n movq n movq mm1 |
Definition at line 2463 of file dsputil_mmx.c.
Referenced by add_8x8basis_TMPL(), cavs_idct8_1d(), cavs_idct8_add_mmx(), deInterlaceBlendLinear_TMPL(), deInterlaceInterpolateLinear_TMPL(), deInterlaceL5_TMPL(), dering_TMPL(), do_a_deblock_TMPL(), doVertDefFilter_TMPL(), doVertLowPass_TMPL(), ff_h264_idct8_add_mmx(), ff_h264_idct_add_mmx(), ff_imdct_calc_3dn2(), ff_imdct_half_3dn2(), ff_vp3_h_loop_filter_mmx2(), H264_CHROMA_MC4_TMPL(), h264_idct8_1d(), h264_loop_filter_chroma_intra_mmx2(), h264_loop_filter_luma_mmx2(), h264_loop_filter_strength_mmx2(), idct_col(), mmx_row(), mmx_row_head(), mmx_row_mid(), mmx_row_tail(), mmxext_row(), mmxext_row_mid(), mmxext_row_tail(), pixels16(), pixels16_l2(), pixels16_x2(), pixels4(), pixels8(), pixels8_l2(), pixels8_x2(), pixels8_y2(), tempNoiseReducer_TMPL(), try_8x8basis_TMPL(), vertClassify_TMPL(), vertX1Filter_TMPL(), vsad16_mmx(), vsad16_mmx2(), vsad_intra16_mmx(), vsad_intra16_mmx2(), yuv2packed2_TMPL(), and yuv2packedX_TMPL().
Definition at line 2463 of file dsputil_mmx.c.
Referenced by cavs_idct8_1d(), cavs_idct8_add_mmx(), deInterlaceBlendLinear_TMPL(), dering_TMPL(), do_a_deblock_TMPL(), doVertDefFilter_TMPL(), doVertLowPass_TMPL(), ff_h264_idct8_add_mmx(), ff_h264_idct_add_mmx(), ff_imdct_half_3dn2(), ff_vp3_h_loop_filter_mmx2(), H264_CHROMA_MC8_TMPL(), h264_idct8_1d(), h264_loop_filter_chroma_intra_mmx2(), h264_loop_filter_luma_mmx2(), h264_loop_filter_strength_mmx2(), idct_col(), mmx_row(), mmx_row_head(), mmx_row_mid(), mmxext_row(), mmxext_row_head(), mmxext_row_mid(), pixels16(), pixels16_l2(), pixels16_x2(), pixels4(), pixels8(), pixels8_l2(), pixels8_x2(), pixels8_y2(), tempNoiseReducer_TMPL(), vertClassify_TMPL(), yuv2packed2_TMPL(), and yuv2packedX_TMPL().
Definition at line 2463 of file dsputil_mmx.c.
Referenced by cavs_idct8_1d(), cavs_idct8_add_mmx(), dering_TMPL(), do_a_deblock_TMPL(), doVertDefFilter_TMPL(), doVertLowPass_TMPL(), ff_h264_idct8_add_mmx(), ff_h264_idct_add_mmx(), ff_imdct_half_3dn2(), ff_vp3_h_loop_filter_mmx2(), h264_idct8_1d(), h264_loop_filter_chroma_intra_mmx2(), h264_loop_filter_luma_mmx2(), h264_loop_filter_strength_mmx2(), idct_col(), mmx_row(), mmx_row_head(), mmx_row_mid(), mmx_row_tail(), mmxext_row(), mmxext_row_head(), mmxext_row_mid(), mmxext_row_tail(), pixels16_l2(), pixels16_x2(), pixels8_l2(), pixels8_x2(), tempNoiseReducer_TMPL(), vertClassify_TMPL(), vertX1Filter_TMPL(), yuv2packed2_TMPL(), and yuv2packedX_TMPL().
int mm_flags |
Definition at line 40 of file dsputil_mmx.c.
mm0 n mm1 n mm2 n mm3 n packssdw mm0 n packssdw mm2 n movq mm1 n punpcklwd mm0 n punpckhwd mm1 n movq n movq n n js n emms n const float** src |
Definition at line 2487 of file dsputil_mmx.c.
Referenced by adpcm_decode_frame(), apply_dependent_coupling(), apply_independent_coupling(), avfilter_draw_slice(), bfi_decode_frame(), blockCopy_TMPL(), bmp_decode_frame(), bmp_encode_frame(), cavs_idct8_add_c(), cinaudio_decode_frame(), cmp_direct_inline(), cmp_inline(), copy(), copy_backptr(), decode(), decode_frame(), decode_init(), decode_p_frame(), deInterlaceBlendLinear_TMPL(), deInterlaceFF_TMPL(), deInterlaceInterpolateCubic_TMPL(), deInterlaceInterpolateLinear_TMPL(), deInterlaceL5_TMPL(), deInterlaceMedian_TMPL(), dering_TMPL(), do_a_deblock_TMPL(), doTest(), doVertDefFilter_TMPL(), doVertLowPass_TMPL(), duplicate_TMPL(), encode_frame(), extract_even2_TMPL(), extract_even_TMPL(), extract_odd2_TMPL(), ff_draw_horiz_band(), ff_ivi_inverse_slant_4x4(), ff_ivi_inverse_slant_8x8(), ff_ivi_process_empty_tile(), ff_jpegls_decode_picture(), ff_mjpeg_decode_frame(), ff_rtmp_packet_dump(), ff_snow_horizontal_compose97i_mmx(), ff_snow_horizontal_compose97i_sse2(), guess_mv(), h264_handle_packet(), halfpel_interpol(), http_prepare_data(), indeo3_decode_frame(), load_input_picture(), main(), mca(), memcpy_backptr(), msrle_decode_8_16_24_32(), pcm_bluray_decode_frame(), pcm_decode_frame(), pcx_encode_frame(), planar2x_TMPL(), postProcess_TMPL(), rgb15to16_TMPL(), rgb15to32_TMPL(), rgb15tobgr24_TMPL(), rgb16to15_TMPL(), rgb16to32_TMPL(), rgb16tobgr24_TMPL(), rgb24to15_TMPL(), rgb24to16_TMPL(), rgb24tobgr15_TMPL(), rgb24tobgr16_TMPL(), rgb24tobgr24_TMPL(), rgb24tobgr32_TMPL(), rgb24toyv12_TMPL(), rgb32to15_TMPL(), rgb32to16_TMPL(), rgb32tobgr15_TMPL(), rgb32tobgr16_TMPL(), rgb32tobgr24_TMPL(), rgb32tobgr32_TMPL(), svq1_decode_frame(), svq1_encode_plane(), svq1_motion_inter_4v_block(), svq1_motion_inter_block(), svq1_skip_block(), svq3_mc_dir_part(), tempNoiseReducer_TMPL(), tgv_decode_inter(), tmv_decode_frame(), transpose1_TMPL(), transpose2_TMPL(), uyvytoyuv420_TMPL(), uyvytoyuv422_TMPL(), uyvytoyv12_TMPL(), validate_acl_list(), vc1_interp_mc(), vc1_inv_trans_4x4_c(), vc1_inv_trans_4x8_c(), vc1_inv_trans_8x4_c(), vc1_inv_trans_8x8_c(), vc1_mc_1mv(), vc1_mc_4mv_chroma(), vc1_mc_4mv_luma(), vertClassify_TMPL(), vertX1Filter_TMPL(), wavpack_decode_frame(), yuv2yuv1_TMPL(), yuva420_bgr32_TMPL(), yuva420_rgb32_TMPL(), yuy2toyv12_TMPL(), yuyvtoyuv420_TMPL(), yuyvtoyuv422_TMPL(), zmbv_decode_xor_16(), and zmbv_decode_xor_8().