/third_party/libjpeg-turbo/simd/loongson/ |
D | jquanti-mmi.c | 40 mm0 = mm2; \ 47 mm0 = _mm_xor_si64(mm0, mm2); /* val = -val */ \ 49 mm0 = _mm_sub_pi16(mm0, mm2); \ 55 mm0 = _mm_add_pi16(mm0, corr0); /* correction + roundfactor */ \ 58 mm4 = mm0; \ 64 mm0 = _mm_mulhi_pi16(mm0, recip0); \ 67 mm0 = _mm_add_pi16(mm0, mm4); /* reciprocal is always negative */ \ 78 mm4 = mm0; \ 81 mm0 = _mm_mulhi_pi16(mm0, mm6); \ 90 mm0 = _mm_add_pi16(mm0, mm6); \ [all …]
|
D | jdsample-mmi.c | 57 mm0 = mm7; \ 59 mm0 = _mm_srli_si64(mm0, 2 * BYTE_BIT); /* mm0=( 1 2 3 -) */ \ 66 mm0 = _mm_or_si64(mm0, mm4); /* mm0=( 1 2 3 4) */ \ 85 mm0 = _mm_add_pi16(mm0, PW_SEVEN); \ 92 mm0 = _mm_add_pi16(mm0, mm7); \ 94 mm0 = _mm_srli_pi16(mm0, 4); /* mm0=OutrLO=( 1 3 5 7) */ \ 97 mm0 = _mm_slli_pi16(mm0, BYTE_BIT); \ 99 mm1 = _mm_or_si64(mm1, mm0); /* mm1=OutrL=( 0 1 2 3 4 5 6 7) */ \ 114 __m64 mm0, mm1, mm2, mm3 = 0.0, mm4, mm5, mm6, mm7 = 0.0; in jsimd_h2v2_fancy_upsample_mmi() local 147 mm0 = _mm_load_si64((__m64 *)inptr0); /* mm0 = row[ 0][0] */ in jsimd_h2v2_fancy_upsample_mmi() [all …]
|
D | jcsample-mmi.c | 45 __m64 mm0, mm1, mm2, mm3, mm4, mm5, mm6 = 0.0, mm7; in jsimd_h2v2_downsample_mmi() local 65 mm0 = _mm_load_si64((__m64 *)&inptr0[0]); in jsimd_h2v2_downsample_mmi() 70 mm4 = mm0; in jsimd_h2v2_downsample_mmi() 72 mm0 = _mm_and_si64(mm0, mm6); in jsimd_h2v2_downsample_mmi() 76 mm0 = _mm_add_pi16(mm0, mm4); in jsimd_h2v2_downsample_mmi() 88 mm0 = _mm_add_pi16(mm0, mm1); in jsimd_h2v2_downsample_mmi() 90 mm0 = _mm_add_pi16(mm0, mm7); in jsimd_h2v2_downsample_mmi() 92 mm0 = _mm_srli_pi16(mm0, 2); in jsimd_h2v2_downsample_mmi() 95 mm0 = _mm_packs_pu16(mm0, mm2); in jsimd_h2v2_downsample_mmi() 97 _mm_store_si64((__m64 *)&outptr[0], mm0); in jsimd_h2v2_downsample_mmi()
|
D | jccolext-mmi.c | 37 #define mmA mm0 51 #define mmC mm0 65 #define mmE mm0 79 #define mmG mm0 99 __m64 mm0, mm1, mm2, mm3, mm4, mm5, mm6, mm7; in jsimd_rgb_ycc_convert_mmi() local 327 wk[0] = mm0; in jsimd_rgb_ycc_convert_mmi() 360 mm6 = mm0; in jsimd_rgb_ycc_convert_mmi() 361 mm0 = _mm_unpacklo_pi16(mm0, mm2); in jsimd_rgb_ycc_convert_mmi() 363 mm5 = mm0; in jsimd_rgb_ycc_convert_mmi() 365 mm0 = _mm_madd_pi16(mm0, PW_F0299_F0337); in jsimd_rgb_ycc_convert_mmi() [all …]
|
D | jdcolext-mmi.c | 36 #define mmA mm0 50 #define mmC mm0 64 #define mmE mm0 78 #define mmG mm0 98 __m64 mm0, mm1, mm2, mm3, mm4, mm5, mm6, mm7; in jsimd_ycc_rgb_convert_mmi() local 120 mm0 = mm4; /* mm0=mm4={0xFF 0x00 0xFF 0x00 ..} */ in jsimd_ycc_rgb_convert_mmi() 124 mm0 = _mm_and_si64(mm0, mm1); /* mm0=Cr(0246)=CrE */ in jsimd_ycc_rgb_convert_mmi() 128 mm0 = _mm_add_pi16(mm0, mm7); in jsimd_ycc_rgb_convert_mmi() 146 mm6 = mm0; /* mm6 = CrE */ in jsimd_ycc_rgb_convert_mmi() 148 mm0 = _mm_add_pi16(mm0, mm0); /* mm0 = 2*CrE */ in jsimd_ycc_rgb_convert_mmi() [all …]
|
/third_party/flutter/skia/third_party/externals/libjpeg-turbo/simd/loongson/ |
D | jquanti-mmi.c | 40 mm0 = mm2; \ 47 mm0 = _mm_xor_si64(mm0, mm2); /* val = -val */ \ 49 mm0 = _mm_sub_pi16(mm0, mm2); \ 55 mm0 = _mm_add_pi16(mm0, corr0); /* correction + roundfactor */ \ 58 mm4 = mm0; \ 64 mm0 = _mm_mulhi_pi16(mm0, recip0); \ 67 mm0 = _mm_add_pi16(mm0, mm4); /* reciprocal is always negative */ \ 78 mm4 = mm0; \ 81 mm0 = _mm_mulhi_pi16(mm0, mm6); \ 90 mm0 = _mm_add_pi16(mm0, mm6); \ [all …]
|
D | jdsample-mmi.c | 57 mm0 = mm7; \ 59 mm0 = _mm_srli_si64(mm0, 2 * BYTE_BIT); /* mm0=( 1 2 3 -) */ \ 66 mm0 = _mm_or_si64(mm0, mm4); /* mm0=( 1 2 3 4) */ \ 85 mm0 = _mm_add_pi16(mm0, PW_SEVEN); \ 92 mm0 = _mm_add_pi16(mm0, mm7); \ 94 mm0 = _mm_srli_pi16(mm0, 4); /* mm0=OutrLO=( 1 3 5 7) */ \ 97 mm0 = _mm_slli_pi16(mm0, BYTE_BIT); \ 99 mm1 = _mm_or_si64(mm1, mm0); /* mm1=OutrL=( 0 1 2 3 4 5 6 7) */ \ 114 __m64 mm0, mm1, mm2, mm3 = 0.0, mm4, mm5, mm6, mm7 = 0.0; in jsimd_h2v2_fancy_upsample_mmi() local 147 mm0 = _mm_load_si64((__m64 *)inptr0); /* mm0 = row[ 0][0] */ in jsimd_h2v2_fancy_upsample_mmi() [all …]
|
D | jccolext-mmi.c | 36 #define mmA mm0 50 #define mmC mm0 64 #define mmE mm0 78 #define mmG mm0 98 __m64 mm0, mm1, mm2, mm3, mm4, mm5, mm6, mm7; in jsimd_rgb_ycc_convert_mmi() local 313 wk[0] = mm0; in jsimd_rgb_ycc_convert_mmi() 346 mm6 = mm0; in jsimd_rgb_ycc_convert_mmi() 347 mm0 = _mm_unpacklo_pi16(mm0, mm2); in jsimd_rgb_ycc_convert_mmi() 349 mm5 = mm0; in jsimd_rgb_ycc_convert_mmi() 351 mm0 = _mm_madd_pi16(mm0, PW_F0299_F0337); in jsimd_rgb_ycc_convert_mmi() [all …]
|
D | jcsample-mmi.c | 45 __m64 mm0, mm1, mm2, mm3, mm4, mm5, mm6 = 0.0, mm7; in jsimd_h2v2_downsample_mmi() local 65 mm0 = _mm_load_si64((__m64 *)&inptr0[0]); in jsimd_h2v2_downsample_mmi() 70 mm4 = mm0; in jsimd_h2v2_downsample_mmi() 72 mm0 = _mm_and_si64(mm0, mm6); in jsimd_h2v2_downsample_mmi() 76 mm0 = _mm_add_pi16(mm0, mm4); in jsimd_h2v2_downsample_mmi() 88 mm0 = _mm_add_pi16(mm0, mm1); in jsimd_h2v2_downsample_mmi() 90 mm0 = _mm_add_pi16(mm0, mm7); in jsimd_h2v2_downsample_mmi() 92 mm0 = _mm_srli_pi16(mm0, 2); in jsimd_h2v2_downsample_mmi() 95 mm0 = _mm_packs_pu16(mm0, mm2); in jsimd_h2v2_downsample_mmi() 97 _mm_store_si64((__m64 *)&outptr[0], mm0); in jsimd_h2v2_downsample_mmi()
|
D | jdcolext-mmi.c | 36 #define mmA mm0 50 #define mmC mm0 64 #define mmE mm0 78 #define mmG mm0 98 __m64 mm0, mm1, mm2, mm3, mm4, mm5, mm6, mm7; in jsimd_ycc_rgb_convert_mmi() local 120 mm0 = mm4; /* mm0=mm4={0xFF 0x00 0xFF 0x00 ..} */ in jsimd_ycc_rgb_convert_mmi() 124 mm0 = _mm_and_si64(mm0, mm1); /* mm0=Cr(0246)=CrE */ in jsimd_ycc_rgb_convert_mmi() 128 mm0 = _mm_add_pi16(mm0, mm7); in jsimd_ycc_rgb_convert_mmi() 146 mm6 = mm0; /* mm6 = CrE */ in jsimd_ycc_rgb_convert_mmi() 148 mm0 = _mm_add_pi16(mm0, mm0); /* mm0 = 2*CrE */ in jsimd_ycc_rgb_convert_mmi() [all …]
|
/third_party/mesa3d/src/mesa/x86/ |
D | read_rgba_span_x86.S | 123 movq (%ebx), %mm0 130 movq %mm0, %mm3 131 movq %mm0, %mm4 138 pand %mm1, %mm0 140 por %mm3, %mm0 142 movq %mm0, (%ecx) 227 movq (%ebx), %mm0 230 movq %mm0, %mm3 231 movq %mm0, %mm4 238 pand %mm1, %mm0 [all …]
|
/third_party/ffmpeg/libavcodec/ |
D | xvididct.c | 160 int mm0, mm1, mm2, mm3, mm4, mm5, mm6, mm7, spill; in idct_col_8() local 169 mm0 = MULT(TAN1, mm4, 16) + mm7; in idct_col_8() 174 mm7 = mm0 + mm2; in idct_col_8() 176 mm0 = mm0 - mm2; in idct_col_8() 178 mm6 = mm0 + mm1; in idct_col_8() 179 mm5 = mm0 - mm1; in idct_col_8() 191 LOAD_BUTTERFLY(mm0, mm1, 0 * 8, 4 * 8, spill, in); in idct_col_8() 193 BUTTERFLY(mm0, mm3, spill); in idct_col_8() 194 BUTTERFLY(mm0, mm7, spill); in idct_col_8() 195 in[8 * 0] = (int16_t) (mm0 >> COL_SHIFT); in idct_col_8() [all …]
|
/third_party/ffmpeg/libavcodec/x86/ |
D | h264_intrapred.asm | 55 movq mm0, [r0+0] 58 movq [r0+r1*1+0], mm0 60 movq [r0+r1*2+0], mm0 132 pxor mm0, mm0 134 psadbw mm0, [r0+0] 138 paddw mm0, mm1 139 movd r6d, mm0 194 movq mm0, [r0+0] 196 movq mm1, mm0 198 punpcklbw mm0, mm7 [all …]
|
D | simple_idct.asm | 83 movq mm0, [blockq + %1] ; R4 R0 r4 r0 88 pand mm4, mm0 97 pmaddwd mm4, mm0 ; C4R4+C4R0 C4r4+C4r0 99 pmaddwd mm0, mm5 ; -C4R4+C4R0 -C4r4+C4r0 112 paddd mm0, [coeffs + 8] 113 paddd mm1, mm0 ; A1 a1 114 paddd mm0, mm0 115 psubd mm0, mm1 ; A2 a2 141 movq mm2, mm0 ; A2 a2 145 psubd mm0, mm4 ; a2-B2 a2-b2 [all …]
|
D | hpeldsp_rnd_template.c | 42 PAVGBP(%%mm0, %%mm1, %%mm4, %%mm2, %%mm3, %%mm5) in DEF() 51 PAVGBP(%%mm0, %%mm1, %%mm4, %%mm2, %%mm3, %%mm5) in DEF() 74 PAVGBP(%%mm0, %%mm1, %%mm4, %%mm2, %%mm3, %%mm5) in DEF() 81 PAVGBP(%%mm0, %%mm1, %%mm4, %%mm2, %%mm3, %%mm5) in DEF() 90 PAVGBP(%%mm0, %%mm1, %%mm4, %%mm2, %%mm3, %%mm5) in DEF() 97 PAVGBP(%%mm0, %%mm1, %%mm4, %%mm2, %%mm3, %%mm5) in DEF() 119 PAVGBP(%%mm1, %%mm0, %%mm4, %%mm2, %%mm1, %%mm5) in DEF() 126 PAVGBP(%%mm1, %%mm2, %%mm4, %%mm0, %%mm1, %%mm5) in DEF() 147 PAVGB(%%mm0, %%mm1, %%mm2, %%mm6) in DEF() 148 PAVGB_MMX(%%mm3, %%mm2, %%mm0, %%mm6) in DEF() [all …]
|
/third_party/gstreamer/gstplugins_good/gst/deinterlace/tvtime/tomsmocomp/ |
D | SearchLoopBottom.inc | 8 "movq %%mm6, %%mm0\n\t" // just use the results of our weird bob 20 "pxor %%mm0, %%mm0\n\t" 21 "pcmpeqb %%mm0, %%mm7\n\t" // all ff where weave better, else 00 22 "pcmpeqb %%mm7, %%mm0\n\t" // all ff where bob better, else 00 23 "pand %%mm6, %%mm0\n\t" // use bob for these pixel values 25 "por %%mm7, %%mm0\n\t" // combine both 29 V_PMINUB ("%%mm4", _TENS, "%%mm0") // the most we care about 33 "pxor %%mm0, %%mm0\n\t" 34 "pcmpeqb %%mm0, %%mm7\n\t" // all ff where weave better, else 00 35 "pcmpeqb %%mm7, %%mm0\n\t" // all ff where bob better, else 00 [all …]
|
D | StrangeBob.inc | 36 "movq -2(%%"XBX"), %%mm0\n\t" // value a from top left 39 "movq %%mm0, %%mm3\n\t" 41 "psubusb %%mm0, %%mm1\n\t" 50 "movq -4(%%"XBX"), %%mm0\n\t" // value j 52 "movq %%mm0, %%mm2\n\t" 54 "movq %%mm0, %%mm3\n\t" 55 "psubusb %%mm1, %%mm0\n\t" 57 "por %%mm1, %%mm0\n\t" // abs(j,n) 59 "movq %%mm0, %%mm1\n\t" 66 "pand %%mm1, %%mm0\n\t" [all …]
|
/third_party/libjpeg-turbo/simd/i386/ |
D | jquant-3dn.asm | 60 movq mm0, MMWORD [ebx+eax*SIZEOF_JSAMPLE] 63 psubb mm0, mm7 ; mm0=(01234567) 66 punpcklbw mm2, mm0 ; mm2=(*0*1*2*3) 67 punpckhbw mm0, mm0 ; mm0=(*4*5*6*7) 73 punpcklwd mm5, mm0 ; mm5=(***4***5) 74 punpckhwd mm0, mm0 ; mm0=(***6***7) 81 psrad mm0, (DWORD_BIT-BYTE_BIT) ; mm0=(67) 83 pi2fd mm0, mm0 88 movq MMWORD [MMBLOCK(0,3,edi,SIZEOF_FAST_FLOAT)], mm0 159 movq mm0, MMWORD [MMBLOCK(0,0,esi,SIZEOF_FAST_FLOAT)] [all …]
|
/third_party/flutter/skia/third_party/externals/libjpeg-turbo/simd/i386/ |
D | jquant-3dn.asm | 62 movq mm0, MMWORD [ebx+eax*SIZEOF_JSAMPLE] 65 psubb mm0, mm7 ; mm0=(01234567) 68 punpcklbw mm2, mm0 ; mm2=(*0*1*2*3) 69 punpckhbw mm0, mm0 ; mm0=(*4*5*6*7) 75 punpcklwd mm5, mm0 ; mm5=(***4***5) 76 punpckhwd mm0, mm0 ; mm0=(***6***7) 83 psrad mm0, (DWORD_BIT-BYTE_BIT) ; mm0=(67) 85 pi2fd mm0, mm0 90 movq MMWORD [MMBLOCK(0,3,edi,SIZEOF_FAST_FLOAT)], mm0 161 movq mm0, MMWORD [MMBLOCK(0,0,esi,SIZEOF_FAST_FLOAT)] [all …]
|
/third_party/skia/third_party/externals/libjpeg-turbo/simd/i386/ |
D | jquant-3dn.asm | 60 movq mm0, MMWORD [ebx+eax*SIZEOF_JSAMPLE] 63 psubb mm0, mm7 ; mm0=(01234567) 66 punpcklbw mm2, mm0 ; mm2=(*0*1*2*3) 67 punpckhbw mm0, mm0 ; mm0=(*4*5*6*7) 73 punpcklwd mm5, mm0 ; mm5=(***4***5) 74 punpckhwd mm0, mm0 ; mm0=(***6***7) 81 psrad mm0, (DWORD_BIT-BYTE_BIT) ; mm0=(67) 83 pi2fd mm0, mm0 88 movq MMWORD [MMBLOCK(0,3,edi,SIZEOF_FAST_FLOAT)], mm0 159 movq mm0, MMWORD [MMBLOCK(0,0,esi,SIZEOF_FAST_FLOAT)] [all …]
|
/third_party/openh264/codec/encoder/core/x86/ |
D | intra_pred.asm | 332 movq mm0, [r1] 334 punpcklbw mm0, mm7 335 pmullw mm0, mm5 340 psubw mm1, mm0 354 LOAD_COLUMN_C mm0, mm2, mm3, mm4, r1, r2 365 punpckhbw mm0, mm4 366 pmullw mm0, mm5 369 psubw mm7, mm0 538 movq mm0, [r1-8] 539 psrlq mm0, 38h [all …]
|
D | memzero.asm | 93 pxor mm0, mm0 95 movq [r0], mm0 96 movq [r0+8], mm0 97 movq [r0+16], mm0 98 movq [r0+24], mm0 99 movq [r0+32], mm0 100 movq [r0+40], mm0 101 movq [r0+48], mm0 102 movq [r0+56], mm0 120 pxor mm0, mm0 [all …]
|
/third_party/openh264/codec/decoder/core/x86/ |
D | intra_pred.asm | 408 movq mm0, [r0] 410 punpcklbw mm0, mm7 411 pmullw mm0, mm5 416 psubw mm1, mm0 430 LOAD_COLUMN_C mm0, mm2, mm3, mm4, r0, r1 441 punpckhbw mm0, mm4 442 pmullw mm0, mm5 445 psubw mm7, mm0 576 movq mm0, [r2-8] 577 psrlq mm0, 38h [all …]
|
/third_party/gstreamer/gstplugins_good/gst/goom2k1/ |
D | filters_mmx.s | 56 movq (%esi), %mm0 /* b1-v1-r1-a1-b2-v2-r2-a2 */ 57 movq %mm0, %mm1 /* b1-v1-r1-a1-b2-v2-r2-a2 */ 62 punpcklbw %mm7, %mm0 /* 00-b2-00-v2-00-r2-00-a2 */ 81 pmullw %mm6, %mm0 /* c1*b2-c1*v2-c1*r2-c1*a2 */ 83 paddw %mm1, %mm0 103 paddw %mm1, %mm0 104 paddw %mm2, %mm0 107 psrlw $8, %mm0 108 packuswb %mm7, %mm0 115 movd %mm0, (%ebx)
|
/third_party/pulseaudio/src/pulsecore/ |
D | svolume_mmx.c | 111 VOLUME_32x16 (%%mm1, %%mm0) in pa_volume_s16ne_mmx() 125 VOLUME_32x16 (%%mm1, %%mm0) in pa_volume_s16ne_mmx() 140 VOLUME_32x16 (%%mm1, %%mm0) in pa_volume_s16ne_mmx() 185 VOLUME_32x16 (%%mm1, %%mm0) in pa_volume_s16re_mmx() 201 VOLUME_32x16 (%%mm1, %%mm0) in pa_volume_s16re_mmx() 202 SWAP_16 (%%mm0) in pa_volume_s16re_mmx() 218 VOLUME_32x16 (%%mm1, %%mm0) in pa_volume_s16re_mmx() 220 SWAP_16_2 (%%mm0, %%mm2) in pa_volume_s16re_mmx()
|