/external/libvpx/libvpx/vpx_dsp/arm/ |
D | highbd_intrapred_neon.c | 20 static INLINE uint16x4_t dc_sum_4(const uint16_t *ref) { in dc_sum_4() 21 const uint16x4_t ref_u16 = vld1_u16(ref); in dc_sum_4() 22 const uint16x4_t p0 = vpadd_u16(ref_u16, ref_u16); in dc_sum_4() 27 const uint16x4_t dc) { in dc_store_4x4() 28 const uint16x4_t dc_dup = vdup_lane_u16(dc, 0); in dc_store_4x4() 38 const uint16x4_t a = vld1_u16(above); in vpx_highbd_dc_predictor_4x4_neon() 39 const uint16x4_t l = vld1_u16(left); in vpx_highbd_dc_predictor_4x4_neon() 40 uint16x4_t sum; in vpx_highbd_dc_predictor_4x4_neon() 41 uint16x4_t dc; in vpx_highbd_dc_predictor_4x4_neon() 53 const uint16x4_t sum = dc_sum_4(left); in vpx_highbd_dc_left_predictor_4x4_neon() [all …]
|
D | sad4d_neon.c | 37 uint16x4_t a[2]; in sad4x_4d() 77 const uint16x4_t a0 = vadd_u16(vget_low_u16(sum[0]), vget_high_u16(sum[0])); in sad_512_pel_final_neon() 78 const uint16x4_t a1 = vadd_u16(vget_low_u16(sum[1]), vget_high_u16(sum[1])); in sad_512_pel_final_neon() 79 const uint16x4_t a2 = vadd_u16(vget_low_u16(sum[2]), vget_high_u16(sum[2])); in sad_512_pel_final_neon() 80 const uint16x4_t a3 = vadd_u16(vget_low_u16(sum[3]), vget_high_u16(sum[3])); in sad_512_pel_final_neon() 81 const uint16x4_t b0 = vpadd_u16(a0, a1); in sad_512_pel_final_neon() 82 const uint16x4_t b1 = vpadd_u16(a2, a3); in sad_512_pel_final_neon() 90 const uint16x4_t a0 = vpadd_u16(vget_low_u16(sum[0]), vget_high_u16(sum[0])); in sad_1024_pel_final_neon() 91 const uint16x4_t a1 = vpadd_u16(vget_low_u16(sum[1]), vget_high_u16(sum[1])); in sad_1024_pel_final_neon() 92 const uint16x4_t a2 = vpadd_u16(vget_low_u16(sum[2]), vget_high_u16(sum[2])); in sad_1024_pel_final_neon() [all …]
|
D | quantize_neon.c | 154 const uint16x4_t eob_max_0 = in vpx_quantize_b_neon() 156 const uint16x4_t eob_max_1 = vpmax_u16(eob_max_0, eob_max_0); in vpx_quantize_b_neon() 157 const uint16x4_t eob_max_2 = vpmax_u16(eob_max_1, eob_max_1); in vpx_quantize_b_neon() 310 const uint16x4_t eob_max_0 = in vpx_quantize_b_32x32_neon() 312 const uint16x4_t eob_max_1 = vpmax_u16(eob_max_0, eob_max_0); in vpx_quantize_b_32x32_neon() 313 const uint16x4_t eob_max_2 = vpmax_u16(eob_max_1, eob_max_1); in vpx_quantize_b_32x32_neon()
|
D | intrapred_neon.c | 20 static INLINE uint16x4_t dc_sum_4(const uint8_t *ref) { in dc_sum_4() 22 const uint16x4_t p0 = vpaddl_u8(ref_u8); in dc_sum_4() 40 uint16x4_t sum; in vpx_dc_predictor_4x4_neon() 50 const uint16x4_t sum = dc_sum_4(left); in vpx_dc_left_predictor_4x4_neon() 58 const uint16x4_t sum = dc_sum_4(above); in vpx_dc_top_predictor_4x4_neon() 75 static INLINE uint16x4_t dc_sum_8(const uint8_t *ref) { in dc_sum_8() 77 uint16x4_t sum = vpaddl_u8(ref_u8); in dc_sum_8() 97 uint16x4_t sum = vadd_u16(vget_low_u16(p0), vget_high_u16(p0)); in vpx_dc_predictor_8x8_neon() 107 const uint16x4_t sum = dc_sum_8(left); in vpx_dc_left_predictor_8x8_neon() 115 const uint16x4_t sum = dc_sum_8(above); in vpx_dc_top_predictor_8x8_neon() [all …]
|
/external/clang/test/CodeGen/ |
D | aarch64-neon-3v.c | 80 uint16x4_t test_vand_u16(uint16x4_t a, uint16x4_t b) { in test_vand_u16() 192 uint16x4_t test_vorr_u16(uint16x4_t a, uint16x4_t b) { in test_vorr_u16() 304 uint16x4_t test_veor_u16(uint16x4_t a, uint16x4_t b) { in test_veor_u16() 427 uint16x4_t test_vbic_u16(uint16x4_t a, uint16x4_t b) { in test_vbic_u16() 555 uint16x4_t test_vorn_u16(uint16x4_t a, uint16x4_t b) { in test_vorn_u16()
|
D | arm_neon_intrinsics.c | 63 uint16x4_t test_vaba_u16(uint16x4_t a, uint16x4_t b, uint16x4_t c) { in test_vaba_u16() 219 uint32x4_t test_vabal_u16(uint32x4_t a, uint16x4_t b, uint16x4_t c) { in test_vabal_u16() 290 uint16x4_t test_vabd_u16(uint16x4_t a, uint16x4_t b) { in test_vabd_u16() 460 uint32x4_t test_vabdl_u16(uint16x4_t a, uint16x4_t b) { in test_vabdl_u16() 595 uint16x4_t test_vadd_u16(uint16x4_t a, uint16x4_t b) { in test_vadd_u16() 738 uint16x4_t test_vaddhn_u32(uint32x4_t a, uint32x4_t b) { in test_vaddhn_u32() 809 uint32x4_t test_vaddl_u16(uint16x4_t a, uint16x4_t b) { in test_vaddl_u16() 869 uint32x4_t test_vaddw_u16(uint32x4_t a, uint16x4_t b) { in test_vaddw_u16() 922 uint16x4_t test_vand_u16(uint16x4_t a, uint16x4_t b) { in test_vand_u16() 1041 uint16x4_t test_vbic_u16(uint16x4_t a, uint16x4_t b) { in test_vbic_u16() [all …]
|
D | aarch64-neon-2velem.c | 187 uint16x4_t test_vmul_lane_u16(uint16x4_t a, uint16x4_t v) { in test_vmul_lane_u16() 195 uint16x8_t test_vmulq_lane_u16(uint16x8_t a, uint16x4_t v) { in test_vmulq_lane_u16() 251 uint16x4_t test_vmul_laneq_u16(uint16x4_t a, uint16x8_t v) { in test_vmul_laneq_u16() 960 uint32x4_t test_vmull_lane_u16(uint16x4_t a, uint16x4_t v) { in test_vmull_lane_u16() 1011 uint32x4_t test_vmull_high_lane_u16(uint16x8_t a, uint16x4_t v) { in test_vmull_high_lane_u16() 1060 uint32x4_t test_vmull_laneq_u16(uint16x4_t a, uint16x8_t v) { in test_vmull_laneq_u16() 1809 uint16x4_t test_vmul_lane_u16_0(uint16x4_t a, uint16x4_t v) { in test_vmul_lane_u16_0() 1817 uint16x8_t test_vmulq_lane_u16_0(uint16x8_t a, uint16x4_t v) { in test_vmulq_lane_u16_0() 1873 uint16x4_t test_vmul_laneq_u16_0(uint16x4_t a, uint16x8_t v) { in test_vmul_laneq_u16_0() 2510 uint32x4_t test_vmull_lane_u16_0(uint16x4_t a, uint16x4_t v) { in test_vmull_lane_u16_0() [all …]
|
D | aarch64-neon-perm.c | 73 uint16x4_t test_vuzp1_u16(uint16x4_t a, uint16x4_t b) { in test_vuzp1_u16() 220 uint16x4_t test_vuzp2_u16(uint16x4_t a, uint16x4_t b) { in test_vuzp2_u16() 367 uint16x4_t test_vzip1_u16(uint16x4_t a, uint16x4_t b) { in test_vzip1_u16() 514 uint16x4_t test_vzip2_u16(uint16x4_t a, uint16x4_t b) { in test_vzip2_u16() 661 uint16x4_t test_vtrn1_u16(uint16x4_t a, uint16x4_t b) { in test_vtrn1_u16() 808 uint16x4_t test_vtrn2_u16(uint16x4_t a, uint16x4_t b) { in test_vtrn2_u16() 1014 uint16x4x2_t test_vuzp_u16(uint16x4_t a, uint16x4_t b) { in test_vuzp_u16() 1478 uint16x4x2_t test_vzip_u16(uint16x4_t a, uint16x4_t b) { in test_vzip_u16() 1942 uint16x4x2_t test_vtrn_u16(uint16x4_t a, uint16x4_t b) { in test_vtrn_u16()
|
D | aarch64-neon-misc.c | 23 uint16x4_t test_vceqz_s16(int16x4_t a) { in test_vceqz_s16() 119 uint16x4_t test_vceqz_u16(uint16x4_t a) { in test_vceqz_u16() 223 uint16x4_t test_vceqz_p16(poly16x4_t a) { in test_vceqz_p16() 271 uint16x4_t test_vcgez_s16(int16x4_t a) { in test_vcgez_s16() 387 uint16x4_t test_vclez_s16(int16x4_t a) { in test_vclez_s16() 503 uint16x4_t test_vcgtz_s16(int16x4_t a) { in test_vcgtz_s16() 619 uint16x4_t test_vcltz_s16(int16x4_t a) { in test_vcltz_s16() 787 uint16x4_t test_vrev32_u16(uint16x4_t a) { in test_vrev32_u16() 878 uint16x4_t test_vrev64_u16(uint16x4_t a) { in test_vrev64_u16() 1001 uint16x4_t test_vpaddl_u8(uint8x8_t a) { in test_vpaddl_u8() [all …]
|
D | aarch64-neon-intrinsics.c | 55 uint16x4_t test_vadd_u16(uint16x4_t v1, uint16x4_t v2) { in test_vadd_u16() 186 uint16x4_t test_vsub_u16(uint16x4_t v1, uint16x4_t v2) { in test_vsub_u16() 313 uint16x4_t test_vmul_u16(uint16x4_t v1, uint16x4_t v2) { in test_vmul_u16() 444 uint16x4_t test_vmla_u16(uint16x4_t v1, uint16x4_t v2, uint16x4_t v3) { in test_vmla_u16() 565 uint16x4_t test_vmls_u16(uint16x4_t v1, uint16x4_t v2, uint16x4_t v3) { in test_vmls_u16() 787 uint16x4_t test_vaba_u16(uint16x4_t v1, uint16x4_t v2, uint16x4_t v3) { in test_vaba_u16() 910 uint16x4_t test_vabd_u16(uint16x4_t v1, uint16x4_t v2) { in test_vabd_u16() 1040 int8x8_t test_vbsl_s16(uint16x4_t v1, int16x4_t v2, int16x4_t v3) { in test_vbsl_s16() 1098 uint16x4_t test_vbsl_u16(uint16x4_t v1, uint16x4_t v2, uint16x4_t v3) { in test_vbsl_u16() 1191 poly16x4_t test_vbsl_p16(uint16x4_t v1, poly16x4_t v2, poly16x4_t v3) { in test_vbsl_p16() [all …]
|
D | aarch64-neon-across.c | 38 uint32_t test_vaddlv_u16(uint16x4_t a) { in test_vaddlv_u16() 126 uint16_t test_vmaxv_u16(uint16x4_t a) { in test_vmaxv_u16() 216 uint16_t test_vminv_u16(uint16x4_t a) { in test_vminv_u16() 306 uint16_t test_vaddv_u16(uint16x4_t a) { in test_vaddv_u16()
|
D | aarch64-neon-extract.c | 103 uint16x4_t test_vext_u16(uint16x4_t a, uint16x4_t b) { in test_vext_u16()
|
/external/libaom/libaom/aom_dsp/arm/ |
D | intrapred_neon.c | 31 const uint16x4_t p0 = vpaddl_u8(A); // cascading summation of the top in dc_4x4() 32 const uint16x4_t p1 = vpadd_u16(p0, p0); in dc_4x4() 38 const uint16x4_t p0 = vpaddl_u8(L); // cascading summation of the left in dc_4x4() 39 const uint16x4_t p1 = vpadd_u16(p0, p0); in dc_4x4() 99 const uint16x4_t p0 = vpaddl_u8(A); // cascading summation of the top in dc_8x8() 100 const uint16x4_t p1 = vpadd_u16(p0, p0); in dc_8x8() 101 const uint16x4_t p2 = vpadd_u16(p1, p1); in dc_8x8() 107 const uint16x4_t p0 = vpaddl_u8(L); // cascading summation of the left in dc_8x8() 108 const uint16x4_t p1 = vpadd_u16(p0, p0); in dc_8x8() 109 const uint16x4_t p2 = vpadd_u16(p1, p1); in dc_8x8() [all …]
|
/external/neon_2_sse/ |
D | NEON_2_SSE.h | 146 typedef __m64_128 uint16x4_t; typedef 451 _NEON2SSESTORAGE uint16x4_t vadd_u16(uint16x4_t a, uint16x4_t b); // VADD.I16 d0,d0,d0 468 _NEON2SSESTORAGE uint32x4_t vaddl_u16(uint16x4_t a, uint16x4_t b); // VADDL.U16 q0,d0,d0 475 _NEON2SSESTORAGE uint32x4_t vaddw_u16(uint32x4_t a, uint16x4_t b); // VADDW.U16 q0,q0,d0 482 _NEON2SSESTORAGE uint16x4_t vhadd_u16(uint16x4_t a, uint16x4_t b); // VHADD.U16 d0,d0,d0 495 _NEON2SSESTORAGE uint16x4_t vrhadd_u16(uint16x4_t a, uint16x4_t b); // VRHADD.U16 d0,d0,d0 509 _NEON2SSESTORAGE uint16x4_t vqadd_u16(uint16x4_t a, uint16x4_t b); // VQADD.U16 d0,d0,d0 525 _NEON2SSESTORAGE uint16x4_t vaddhn_u32(uint32x4_t a, uint32x4_t b); // VADDHN.I32 d0,q0,q0 532 _NEON2SSESTORAGE uint16x4_t vraddhn_u32(uint32x4_t a, uint32x4_t b); // VRADDHN.I32 d0,q0,q0 541 _NEON2SSESTORAGE uint16x4_t vmul_u16(uint16x4_t a, uint16x4_t b); // VMUL.I16 d0,d0,d0 [all …]
|
/external/libaom/libaom/av1/common/arm/ |
D | cfl_neon.c | 33 static INLINE void vsth_u16(uint16_t *ptr, uint16x4_t val) { in vsth_u16() 50 const uint16x4_t top = vpaddl_u8(vldh_dup_u8(input)); in cfl_luma_subsampling_420_lbd_neon() 51 const uint16x4_t sum = vpadal_u8(top, vldh_dup_u8(input + input_stride)); in cfl_luma_subsampling_420_lbd_neon() 54 const uint16x4_t top = vpaddl_u8(vld1_u8(input)); in cfl_luma_subsampling_420_lbd_neon() 55 const uint16x4_t sum = vpadal_u8(top, vld1_u8(input + input_stride)); in cfl_luma_subsampling_420_lbd_neon() 88 const uint16x4_t top = vpaddl_u8(vldh_dup_u8(input)); in cfl_luma_subsampling_422_lbd_neon() 91 const uint16x4_t top = vpaddl_u8(vld1_u8(input)); in cfl_luma_subsampling_422_lbd_neon() 150 const uint16x4_t top = vld1_u16(input); in cfl_luma_subsampling_420_hbd_neon() 151 const uint16x4_t bot = vld1_u16(input + input_stride); in cfl_luma_subsampling_420_hbd_neon() 152 const uint16x4_t sum = vadd_u16(top, bot); in cfl_luma_subsampling_420_hbd_neon() [all …]
|
D | blend_a64_vmask_neon.c | 31 uint16x4_t tmp0_16 = vdup_n_u16(0), tmp1_16 = vdup_n_u16(0); in aom_blend_a64_vmask_neon() 85 const uint16x4_t m1 = vdup_n_u16((uint16_t)mask[i]); in aom_blend_a64_vmask_neon() 86 const uint16x4_t m2 = vdup_n_u16((uint16_t)mask[i + 1]); in aom_blend_a64_vmask_neon() 88 const uint16x4_t max_minus_m1 = vdup_n_u16(64 - (uint16_t)mask[i]); in aom_blend_a64_vmask_neon() 89 const uint16x4_t max_minus_m2 = vdup_n_u16(64 - (uint16_t)mask[i + 1]); in aom_blend_a64_vmask_neon()
|
D | transpose_neon.h | 188 static INLINE void transpose_u16_4x8(uint16x4_t *a0, uint16x4_t *a1, in transpose_u16_4x8() 189 uint16x4_t *a2, uint16x4_t *a3, in transpose_u16_4x8() 190 uint16x4_t *a4, uint16x4_t *a5, in transpose_u16_4x8() 191 uint16x4_t *a6, uint16x4_t *a7, in transpose_u16_4x8()
|
D | convolve_neon.h | 25 uint16x4_t tmp0, tmp1; in wiener_convolve8_vert_4x8() 117 static INLINE uint16x4_t wiener_convolve8_horiz_4x8( in wiener_convolve8_horiz_4x8() 122 uint16x4_t res; in wiener_convolve8_horiz_4x8() 202 static INLINE uint16x4_t convolve8_4x4_s32( in convolve8_4x4_s32() 208 uint16x4_t res; in convolve8_4x4_s32()
|
/external/zlib/ |
D | adler32_simd.c | 290 (uint16x4_t) { 32, 31, 30, 29 }); in adler32_simd_() 292 (uint16x4_t) { 28, 27, 26, 25 }); in adler32_simd_() 294 (uint16x4_t) { 24, 23, 22, 21 }); in adler32_simd_() 296 (uint16x4_t) { 20, 19, 18, 17 }); in adler32_simd_() 298 (uint16x4_t) { 16, 15, 14, 13 }); in adler32_simd_() 300 (uint16x4_t) { 12, 11, 10, 9 }); in adler32_simd_() 302 (uint16x4_t) { 8, 7, 6, 5 }); in adler32_simd_() 304 (uint16x4_t) { 4, 3, 2, 1 }); in adler32_simd_()
|
/external/v8/third_party/zlib/ |
D | adler32_simd.c | 290 (uint16x4_t) { 32, 31, 30, 29 }); in adler32_simd_() 292 (uint16x4_t) { 28, 27, 26, 25 }); in adler32_simd_() 294 (uint16x4_t) { 24, 23, 22, 21 }); in adler32_simd_() 296 (uint16x4_t) { 20, 19, 18, 17 }); in adler32_simd_() 298 (uint16x4_t) { 16, 15, 14, 13 }); in adler32_simd_() 300 (uint16x4_t) { 12, 11, 10, 9 }); in adler32_simd_() 302 (uint16x4_t) { 8, 7, 6, 5 }); in adler32_simd_() 304 (uint16x4_t) { 4, 3, 2, 1 }); in adler32_simd_()
|
/external/webp/src/dsp/ |
D | rescaler_neon.c | 82 const uint16x4_t C0 = vmovn_u32(B0); in RescalerExportRowExpand_NEON() 83 const uint16x4_t C1 = vmovn_u32(B1); in RescalerExportRowExpand_NEON() 102 const uint16x4_t E0 = vmovn_u32(D0); in RescalerExportRowExpand_NEON() 103 const uint16x4_t E1 = vmovn_u32(D1); in RescalerExportRowExpand_NEON() 142 const uint16x4_t D0 = vmovn_u32(C0); in RescalerExportRowShrink_NEON() 143 const uint16x4_t D1 = vmovn_u32(C1); in RescalerExportRowShrink_NEON() 159 const uint16x4_t B0 = vmovn_u32(A0); in RescalerExportRowShrink_NEON() 160 const uint16x4_t B1 = vmovn_u32(A1); in RescalerExportRowShrink_NEON()
|
/external/libhevc/encoder/arm/ |
D | ihevce_coarse_layer_sad_neon.c | 192 uint16x4_t tmp_a0, tmp_a1; in hme_store_4x4_sads_high_speed_neon() 212 uint16x4_t tmp_a; in hme_store_4x4_sads_high_speed_neon() 345 uint16x4_t tmp_b0, tmp_b1; in hme_store_4x4_sads_high_quality_neon() 374 uint16x4_t tmp_a, tmp_b; in hme_store_4x4_sads_high_quality_neon() 545 uint16x4_t curr = vld1_u16((U16 *)pi2_sads_4x4_current + sad_pos); in hme_combine_4x4_sads_and_compute_cost_high_speed_neon() 546 uint16x4_t south = vld1_u16((U16 *)pi2_sads_4x4_south + sad_pos); in hme_combine_4x4_sads_and_compute_cost_high_speed_neon() 547 uint16x4_t east = vld1_u16((U16 *)pi2_sads_4x4_east + sad_pos); in hme_combine_4x4_sads_and_compute_cost_high_speed_neon() 548 uint16x4_t sad_4x8 = vadd_u16(curr, south); in hme_combine_4x4_sads_and_compute_cost_high_speed_neon() 549 uint16x4_t sad_8x4 = vadd_u16(curr, east); in hme_combine_4x4_sads_and_compute_cost_high_speed_neon() 551 uint16x4_t mv_wt = in hme_combine_4x4_sads_and_compute_cost_high_speed_neon() [all …]
|
/external/libgav1/libgav1/src/dsp/arm/ |
D | intrapred_smooth_neon.cc | 58 inline uint16x4_t CalculatePred(const uint16x4_t weighted_top, in CalculatePred() 59 const uint16x4_t weighted_left, in CalculatePred() 60 const uint16x4_t weighted_bl, in CalculatePred() 61 const uint16x4_t weighted_tr) { in CalculatePred() 102 const uint16x4_t dest_0 = in Smooth4Or8xN_NEON() 109 const uint16x4_t dest_1 = CalculatePred( in Smooth4Or8xN_NEON() 126 const uint16x4_t dest_0 = CalculatePred( in CalculateWeightsAndPred() 129 const uint16x4_t dest_1 = CalculatePred( in CalculateWeightsAndPred() 138 const uint16x4_t dest_2 = CalculatePred( in CalculateWeightsAndPred() 141 const uint16x4_t dest_3 = CalculatePred( in CalculateWeightsAndPred()
|
/external/libgav1/libgav1/src/utils/ |
D | entropy_decoder.cc | 170 uint16x4_t cdf_vec = vld1_u16(cdf); in UpdateCdf5() 173 const uint16x4_t zero = vdup_n_u16(0); in UpdateCdf5() 174 const uint16x4_t cdf_max_probability = in UpdateCdf5() 176 const uint16x4_t index = vcreate_u16(0x0003000200010000); in UpdateCdf5() 177 const uint16x4_t symbol_vec = vdup_n_u16(symbol); in UpdateCdf5() 178 const uint16x4_t mask = vclt_u16(index, symbol_vec); in UpdateCdf5() 179 const uint16x4_t a = vbsl_u16(mask, cdf_max_probability, zero); in UpdateCdf5() 182 const uint16x4_t delta = vreinterpret_u16_s16(vshl_s16(diff, negative_rate)); in UpdateCdf5() 782 uint16x4_t cdf_vec = vld1_u16(cdf); in ReadSymbol4() 784 const uint16x4_t delta = vshl_u16(cdf_vec, negative_rate); in ReadSymbol4() [all …]
|
/external/skqp/third_party/skcms/src/ |
D | Transform_inl.h | 1046 (uint16x4_t)cast<U16>(to_fixed(r * 65535)), in exec_ops() 1047 (uint16x4_t)cast<U16>(to_fixed(g * 65535)), in exec_ops() 1048 (uint16x4_t)cast<U16>(to_fixed(b * 65535)), in exec_ops() 1065 (uint16x4_t)cast<U16>(to_fixed(r * 65535)), in exec_ops() 1066 (uint16x4_t)cast<U16>(to_fixed(g * 65535)), in exec_ops() 1067 (uint16x4_t)cast<U16>(to_fixed(b * 65535)), in exec_ops() 1068 (uint16x4_t)cast<U16>(to_fixed(a * 65535)), in exec_ops() 1086 (uint16x4_t)swap_endian_16(cast<U16>(to_fixed(r * 65535))), in exec_ops() 1087 (uint16x4_t)swap_endian_16(cast<U16>(to_fixed(g * 65535))), in exec_ops() 1088 (uint16x4_t)swap_endian_16(cast<U16>(to_fixed(b * 65535))), in exec_ops() [all …]
|