/external/XNNPACK/src/qs8-gemm/gen/ |
D | 3x8c2s4-minmax-rndnu-neon-mlal.c | 69 int8x8_t va2x0 = vld1_s8(a2); a2 += 8; in xnn_qs8_gemm_minmax_rndnu_ukernel_3x8c2s4__neon_mlal() local 185 int8x8_t va2x0 = vld1_s8(a2); a2 += 8; in xnn_qs8_gemm_minmax_rndnu_ukernel_3x8c2s4__neon_mlal() local
|
D | 4x8c2s4-minmax-rndnu-neon-mlal.c | 77 int8x8_t va2x0 = vld1_s8(a2); a2 += 8; in xnn_qs8_gemm_minmax_rndnu_ukernel_4x8c2s4__neon_mlal() local 225 int8x8_t va2x0 = vld1_s8(a2); a2 += 8; in xnn_qs8_gemm_minmax_rndnu_ukernel_4x8c2s4__neon_mlal() local
|
D | 3x8c4s2-minmax-rndnu-neon-mlal.c | 75 int8x8_t va2x0 = vld1_s8(a2); a2 += 8; in xnn_qs8_gemm_minmax_rndnu_ukernel_3x8c4s2__neon_mlal() local 179 int8x8_t va2x0 = vld1_s8(a2); a2 += 8; in xnn_qs8_gemm_minmax_rndnu_ukernel_3x8c4s2__neon_mlal() local
|
D | 3x8c2s4-minmax-rndnu-neon-mull.c | 67 int8x8_t va2x0 = vld1_s8(a2); a2 += 8; in xnn_qs8_gemm_minmax_rndnu_ukernel_3x8c2s4__neon_mull() local
|
D | 4x8c4s2-minmax-rndnu-neon-mlal.c | 85 int8x8_t va2x0 = vld1_s8(a2); a2 += 8; in xnn_qs8_gemm_minmax_rndnu_ukernel_4x8c4s2__neon_mlal() local 217 int8x8_t va2x0 = vld1_s8(a2); a2 += 8; in xnn_qs8_gemm_minmax_rndnu_ukernel_4x8c4s2__neon_mlal() local
|
D | 3x16c2s4-minmax-rndnu-neon-mlal.c | 75 int8x8_t va2x0 = vld1_s8(a2); a2 += 8; in xnn_qs8_gemm_minmax_rndnu_ukernel_3x16c2s4__neon_mlal() local 279 int8x8_t va2x0 = vld1_s8(a2); a2 += 8; in xnn_qs8_gemm_minmax_rndnu_ukernel_3x16c2s4__neon_mlal() local
|
D | 4x8c2s4-minmax-rndnu-neon-mull.c | 75 int8x8_t va2x0 = vld1_s8(a2); a2 += 8; in xnn_qs8_gemm_minmax_rndnu_ukernel_4x8c2s4__neon_mull() local
|
D | 3x8c4s2-minmax-rndnu-neon-mull.c | 73 int8x8_t va2x0 = vld1_s8(a2); a2 += 8; in xnn_qs8_gemm_minmax_rndnu_ukernel_3x8c4s2__neon_mull() local
|
D | 4x16c2s4-minmax-rndnu-neon-mlal.c | 85 int8x8_t va2x0 = vld1_s8(a2); a2 += 8; in xnn_qs8_gemm_minmax_rndnu_ukernel_4x16c2s4__neon_mlal() local 345 int8x8_t va2x0 = vld1_s8(a2); a2 += 8; in xnn_qs8_gemm_minmax_rndnu_ukernel_4x16c2s4__neon_mlal() local
|
D | 3x16c4s2-minmax-rndnu-neon-mlal.c | 87 int8x8_t va2x0 = vld1_s8(a2); a2 += 8; in xnn_qs8_gemm_minmax_rndnu_ukernel_3x16c4s2__neon_mlal() local 279 int8x8_t va2x0 = vld1_s8(a2); a2 += 8; in xnn_qs8_gemm_minmax_rndnu_ukernel_3x16c4s2__neon_mlal() local
|
D | 3x16c2s4-minmax-rndnu-neon-mull.c | 73 int8x8_t va2x0 = vld1_s8(a2); a2 += 8; in xnn_qs8_gemm_minmax_rndnu_ukernel_3x16c2s4__neon_mull() local
|
/external/XNNPACK/src/qs8-igemm/gen/ |
D | 3x8c2s4-minmax-rndnu-neon-mlal.c | 84 int8x8_t va2x0 = vld1_s8(a2); a2 += 8; in xnn_qs8_igemm_minmax_rndnu_ukernel_3x8c2s4__neon_mlal() local 200 int8x8_t va2x0 = vld1_s8(a2); a2 += 8; in xnn_qs8_igemm_minmax_rndnu_ukernel_3x8c2s4__neon_mlal() local
|
D | 4x8c2s4-minmax-rndnu-neon-mlal.c | 94 int8x8_t va2x0 = vld1_s8(a2); a2 += 8; in xnn_qs8_igemm_minmax_rndnu_ukernel_4x8c2s4__neon_mlal() local 242 int8x8_t va2x0 = vld1_s8(a2); a2 += 8; in xnn_qs8_igemm_minmax_rndnu_ukernel_4x8c2s4__neon_mlal() local
|
D | 3x8c4s2-minmax-rndnu-neon-mlal.c | 90 int8x8_t va2x0 = vld1_s8(a2); a2 += 8; in xnn_qs8_igemm_minmax_rndnu_ukernel_3x8c4s2__neon_mlal() local 194 int8x8_t va2x0 = vld1_s8(a2); a2 += 8; in xnn_qs8_igemm_minmax_rndnu_ukernel_3x8c4s2__neon_mlal() local
|
D | 4x8c4s2-minmax-rndnu-neon-mlal.c | 102 int8x8_t va2x0 = vld1_s8(a2); a2 += 8; in xnn_qs8_igemm_minmax_rndnu_ukernel_4x8c4s2__neon_mlal() local 234 int8x8_t va2x0 = vld1_s8(a2); a2 += 8; in xnn_qs8_igemm_minmax_rndnu_ukernel_4x8c4s2__neon_mlal() local
|
D | 3x8c2s4-minmax-rndnu-neon-mull.c | 82 int8x8_t va2x0 = vld1_s8(a2); a2 += 8; in xnn_qs8_igemm_minmax_rndnu_ukernel_3x8c2s4__neon_mull() local
|
D | 3x16c2s4-minmax-rndnu-neon-mlal.c | 90 int8x8_t va2x0 = vld1_s8(a2); a2 += 8; in xnn_qs8_igemm_minmax_rndnu_ukernel_3x16c2s4__neon_mlal() local 294 int8x8_t va2x0 = vld1_s8(a2); a2 += 8; in xnn_qs8_igemm_minmax_rndnu_ukernel_3x16c2s4__neon_mlal() local
|
D | 4x8c2s4-minmax-rndnu-neon-mull.c | 92 int8x8_t va2x0 = vld1_s8(a2); a2 += 8; in xnn_qs8_igemm_minmax_rndnu_ukernel_4x8c2s4__neon_mull() local
|
D | 3x8c4s2-minmax-rndnu-neon-mull.c | 88 int8x8_t va2x0 = vld1_s8(a2); a2 += 8; in xnn_qs8_igemm_minmax_rndnu_ukernel_3x8c4s2__neon_mull() local
|
D | 4x16c2s4-minmax-rndnu-neon-mlal.c | 102 int8x8_t va2x0 = vld1_s8(a2); a2 += 8; in xnn_qs8_igemm_minmax_rndnu_ukernel_4x16c2s4__neon_mlal() local 362 int8x8_t va2x0 = vld1_s8(a2); a2 += 8; in xnn_qs8_igemm_minmax_rndnu_ukernel_4x16c2s4__neon_mlal() local
|
D | 3x16c4s2-minmax-rndnu-neon-mlal.c | 102 int8x8_t va2x0 = vld1_s8(a2); a2 += 8; in xnn_qs8_igemm_minmax_rndnu_ukernel_3x16c4s2__neon_mlal() local 294 int8x8_t va2x0 = vld1_s8(a2); a2 += 8; in xnn_qs8_igemm_minmax_rndnu_ukernel_3x16c4s2__neon_mlal() local
|
/external/XNNPACK/src/bf16-gemm/gen/ |
D | 3x4c8-minmax-neonbf16-bfdot.c | 112 … const bfloat16x8_t va2x0 = vreinterpretq_bf16_u16(vbicq_u16(vreinterpretq_u16_bf16(va2), vm0)); in xnn_bf16_gemm_minmax_ukernel_3x4c8__neonbf16_bfdot() local
|
D | 3x4c8-minmax-neonbf16-bfmlal.c | 127 … const bfloat16x8_t va2x0 = vreinterpretq_bf16_u16(vbicq_u16(vreinterpretq_u16_bf16(va2), vm0)); in xnn_bf16_gemm_minmax_ukernel_3x4c8__neonbf16_bfmlal() local
|
D | 4x4c8-minmax-neonbf16-bfdot.c | 128 … const bfloat16x8_t va2x0 = vreinterpretq_bf16_u16(vbicq_u16(vreinterpretq_u16_bf16(va2), vm0)); in xnn_bf16_gemm_minmax_ukernel_4x4c8__neonbf16_bfdot() local
|
D | 4x4c8-minmax-neonbf16-bfmlal.c | 147 … const bfloat16x8_t va2x0 = vreinterpretq_bf16_u16(vbicq_u16(vreinterpretq_u16_bf16(va2), vm0)); in xnn_bf16_gemm_minmax_ukernel_4x4c8__neonbf16_bfmlal() local
|