| /external/libgav1/libgav1/src/dsp/arm/ |
| D | common_neon.h | 355 vzip1_u32(vreinterpret_u32_s8(a), vreinterpret_u32_s8(b))); in InterleaveLow32()
359 vzip_u32(vreinterpret_u32_s8(a), vreinterpret_u32_s8(b)).val[0]); in InterleaveLow32()
377 vzip2_u32(vreinterpret_u32_s8(a), vreinterpret_u32_s8(b))); in InterleaveHigh32()
381 vzip_u32(vreinterpret_u32_s8(a), vreinterpret_u32_s8(b)).val[1]); in InterleaveHigh32()
|
| /external/XNNPACK/src/qs8-vaddc/ |
| D | neon-ld64.c.in | 112 …vst1_lane_u32(__builtin_assume_aligned(output, 1), vreinterpret_u32_s8(vout${ABC[0:8]}), 0); outpu…
126 …vst1_lane_u32(__builtin_assume_aligned(output, 1), vreinterpret_u32_s8(vout${ABC[0:8]}), 0); outpu…
|
| /external/XNNPACK/src/qs8-vadd/ |
| D | neon-ld64.c.in | 119 …vst1_lane_u32(__builtin_assume_aligned(output, 1), vreinterpret_u32_s8(vout${ABC[0:8]}), 0); outpu…
133 …vst1_lane_u32(__builtin_assume_aligned(output, 1), vreinterpret_u32_s8(vout${ABC[0:8]}), 0); outpu…
|
| /external/XNNPACK/src/qs8-vaddc/gen/ |
| D | minmax-neon-ld64-x8.c | 85 …vst1_lane_u32(__builtin_assume_aligned(output, 1), vreinterpret_u32_s8(vout01234567), 0); output +… in xnn_qs8_vaddc_minmax_ukernel__neon_ld64_x8()
|
| D | minmax-neon-ld64-x16.c | 98 …vst1_lane_u32(__builtin_assume_aligned(output, 1), vreinterpret_u32_s8(vout01234567), 0); output +… in xnn_qs8_vaddc_minmax_ukernel__neon_ld64_x16()
|
| D | minmax-neon-ld64-x24.c | 111 …vst1_lane_u32(__builtin_assume_aligned(output, 1), vreinterpret_u32_s8(vout01234567), 0); output +… in xnn_qs8_vaddc_minmax_ukernel__neon_ld64_x24()
|
| D | minmax-neon-ld64-x32.c | 120 …vst1_lane_u32(__builtin_assume_aligned(output, 1), vreinterpret_u32_s8(vout01234567), 0); output +… in xnn_qs8_vaddc_minmax_ukernel__neon_ld64_x32()
|
| /external/XNNPACK/src/qs8-gemm/gen/ |
| D | 1x8c4-minmax-neondot.c | 127 … vst1_lane_u32(__builtin_assume_aligned(c0, 1), vreinterpret_u32_s8(vout0x01234567), 0); c0 += 4; in xnn_qs8_gemm_minmax_ukernel_1x8c4__neondot()
|
| D | 1x16c4-minmax-neondot.c | 154 … vst1_lane_u32(__builtin_assume_aligned(c0, 1), vreinterpret_u32_s8(vout0x01234567), 0); c0 += 4; in xnn_qs8_gemm_minmax_ukernel_1x16c4__neondot()
|
| D | 1x8c8-minmax-neon-mull-padal.c | 149 … vst1_lane_u32(__builtin_assume_aligned(c0, 1), vreinterpret_u32_s8(vout0x01234567), 0); c0 += 4; in xnn_qs8_gemm_minmax_ukernel_1x8c8__neon_mull_padal()
|
| D | 1x8c2-minmax-neon-mull-padal-dup.c | 152 … vst1_lane_u32(__builtin_assume_aligned(c0, 1), vreinterpret_u32_s8(vout0x01234567), 0); c0 += 4; in xnn_qs8_gemm_minmax_ukernel_1x8c2__neon_mull_padal_dup()
|
| D | 1x8-minmax-neon-mlal-lane.c | 193 … vst1_lane_u32(__builtin_assume_aligned(c0, 1), vreinterpret_u32_s8(vout0x01234567), 0); c0 += 4; in xnn_qs8_gemm_minmax_ukernel_1x8__neon_mlal_lane()
|
| D | 1x8c16-minmax-neon-mlal-padal.c | 157 … vst1_lane_u32(__builtin_assume_aligned(c0, 1), vreinterpret_u32_s8(vout0x01234567), 0); c0 += 4; in xnn_qs8_gemm_minmax_ukernel_1x8c16__neon_mlal_padal()
|
| /external/XNNPACK/src/qs8-vadd/gen/ |
| D | minmax-neon-ld64-x8.c | 90 …vst1_lane_u32(__builtin_assume_aligned(output, 1), vreinterpret_u32_s8(vout01234567), 0); output +… in xnn_qs8_vadd_minmax_ukernel__neon_ld64_x8()
|
| D | minmax-neon-ld64-x16.c | 107 …vst1_lane_u32(__builtin_assume_aligned(output, 1), vreinterpret_u32_s8(vout01234567), 0); output +… in xnn_qs8_vadd_minmax_ukernel__neon_ld64_x16()
|
| D | minmax-neon-ld64-x24.c | 124 …vst1_lane_u32(__builtin_assume_aligned(output, 1), vreinterpret_u32_s8(vout01234567), 0); output +… in xnn_qs8_vadd_minmax_ukernel__neon_ld64_x24()
|
| D | minmax-neon-ld64-x32.c | 137 …vst1_lane_u32(__builtin_assume_aligned(output, 1), vreinterpret_u32_s8(vout01234567), 0); output +… in xnn_qs8_vadd_minmax_ukernel__neon_ld64_x32()
|
| /external/XNNPACK/src/qs8-igemm/gen/ |
| D | 1x8c4-minmax-neondot.c | 134 … vst1_lane_u32(__builtin_assume_aligned(c0, 1), vreinterpret_u32_s8(vout0x01234567), 0); c0 += 4; in xnn_qs8_igemm_minmax_ukernel_1x8c4__neondot()
|
| D | 1x16c4-minmax-neondot.c | 161 … vst1_lane_u32(__builtin_assume_aligned(c0, 1), vreinterpret_u32_s8(vout0x01234567), 0); c0 += 4; in xnn_qs8_igemm_minmax_ukernel_1x16c4__neondot()
|
| D | 1x8c8-minmax-neon-mull-padal.c | 163 … vst1_lane_u32(__builtin_assume_aligned(c0, 1), vreinterpret_u32_s8(vout0x01234567), 0); c0 += 4; in xnn_qs8_igemm_minmax_ukernel_1x8c8__neon_mull_padal()
|
| D | 1x8c2-minmax-neon-mull-padal-dup.c | 166 … vst1_lane_u32(__builtin_assume_aligned(c0, 1), vreinterpret_u32_s8(vout0x01234567), 0); c0 += 4; in xnn_qs8_igemm_minmax_ukernel_1x8c2__neon_mull_padal_dup()
|
| D | 1x8-minmax-neon-mlal-lane.c | 206 … vst1_lane_u32(__builtin_assume_aligned(c0, 1), vreinterpret_u32_s8(vout0x01234567), 0); c0 += 4; in xnn_qs8_igemm_minmax_ukernel_1x8__neon_mlal_lane()
|
| D | 1x8-minmax-neon-mull-addw-dup.c | 203 … vst1_lane_u32(__builtin_assume_aligned(c0, 1), vreinterpret_u32_s8(vout0x01234567), 0); c0 += 4; in xnn_qs8_igemm_minmax_ukernel_1x8__neon_mull_addw_dup()
|
| /external/XNNPACK/src/qs8-dwconv/ |
| D | unipass-neon-mul16.c.in | 178 …vst1_lane_u32(__builtin_assume_aligned(output, 1), vreinterpret_u32_s8(vout${ABC[0:8]}), 0); outpu…
192 …vst1_lane_u32(__builtin_assume_aligned(output, 1), vreinterpret_u32_s8(vout${ABC[0:8]}), 0); outpu…
|
| /external/XNNPACK/src/qs8-gavgpool/ |
| D | unipass-neon.c.in | 241 …vst1_lane_u32(__builtin_assume_aligned(output, 1), vreinterpret_u32_s8(vout${ABC[0:8]}), 0); outpu…
255 …vst1_lane_u32(__builtin_assume_aligned(output, 1), vreinterpret_u32_s8(vout${ABC[0:8]}), 0); outpu…
|