/external/XNNPACK/src/qs8-gemm/gen/ |
D | 1x16c8-minmax-rndnu-neon-mlal.c | 133 int16x8_t vprod0x12 = vmull_s8(vb12x0, va0x0); in xnn_qs8_gemm_minmax_rndnu_ukernel_1x16c8__neon_mlal() local 193 const int16x8_t vprod0x12 = vmull_s8(vb12, va0); in xnn_qs8_gemm_minmax_rndnu_ukernel_1x16c8__neon_mlal() local
|
D | 1x16c16-minmax-rndnu-neon-mlal.c | 119 int16x8_t vprod0x12 = vmull_s8(vget_low_s8(vb12), vget_low_s8(va0)); in xnn_qs8_gemm_minmax_rndnu_ukernel_1x16c16__neon_mlal() local
|
D | 1x16c8-minmax-rndnu-neon-mull.c | 104 const int16x8_t vprod0x12 = vmull_s8(vb12, va0); in xnn_qs8_gemm_minmax_rndnu_ukernel_1x16c8__neon_mull() local
|
D | 2x16c8-minmax-rndnu-neon-mlal.c | 193 int16x8_t vprod0x12 = vmull_s8(vb12x0, va0x0); in xnn_qs8_gemm_minmax_rndnu_ukernel_2x16c8__neon_mlal() local 290 const int16x8_t vprod0x12 = vmull_s8(vb12, va0); in xnn_qs8_gemm_minmax_rndnu_ukernel_2x16c8__neon_mlal() local
|
D | 3x16c8-minmax-rndnu-neon-mlal.c | 253 int16x8_t vprod0x12 = vmull_s8(vb12x0, va0x0); in xnn_qs8_gemm_minmax_rndnu_ukernel_3x16c8__neon_mlal() local 387 const int16x8_t vprod0x12 = vmull_s8(vb12, va0); in xnn_qs8_gemm_minmax_rndnu_ukernel_3x16c8__neon_mlal() local
|
D | 2x16c16-minmax-rndnu-neon-mlal.c | 178 int16x8_t vprod0x12 = vmull_s8(vget_low_s8(vb12), vget_low_s8(va0)); in xnn_qs8_gemm_minmax_rndnu_ukernel_2x16c16__neon_mlal() local
|
D | 2x16c8-minmax-rndnu-neon-mull.c | 151 const int16x8_t vprod0x12 = vmull_s8(vb12, va0); in xnn_qs8_gemm_minmax_rndnu_ukernel_2x16c8__neon_mull() local
|
D | 4x16c8-minmax-rndnu-neon-mlal.c | 313 int16x8_t vprod0x12 = vmull_s8(vb12x0, va0x0); in xnn_qs8_gemm_minmax_rndnu_ukernel_4x16c8__neon_mlal() local 484 const int16x8_t vprod0x12 = vmull_s8(vb12, va0); in xnn_qs8_gemm_minmax_rndnu_ukernel_4x16c8__neon_mlal() local
|
D | 3x16c8-minmax-rndnu-neon-mull.c | 198 const int16x8_t vprod0x12 = vmull_s8(vb12, va0); in xnn_qs8_gemm_minmax_rndnu_ukernel_3x16c8__neon_mull() local
|
D | 3x16c16-minmax-rndnu-neon-mlal.c | 237 int16x8_t vprod0x12 = vmull_s8(vget_low_s8(vb12), vget_low_s8(va0)); in xnn_qs8_gemm_minmax_rndnu_ukernel_3x16c16__neon_mlal() local
|
D | 4x16c8-minmax-rndnu-neon-mull.c | 245 const int16x8_t vprod0x12 = vmull_s8(vb12, va0); in xnn_qs8_gemm_minmax_rndnu_ukernel_4x16c8__neon_mull() local
|
D | 4x16c16-minmax-rndnu-neon-mlal.c | 296 int16x8_t vprod0x12 = vmull_s8(vget_low_s8(vb12), vget_low_s8(va0)); in xnn_qs8_gemm_minmax_rndnu_ukernel_4x16c16__neon_mlal() local
|
/external/XNNPACK/src/qs8-igemm/gen/ |
D | 1x16c8-minmax-rndnu-neon-mlal.c | 144 int16x8_t vprod0x12 = vmull_s8(vb12x0, va0x0); in xnn_qs8_igemm_minmax_rndnu_ukernel_1x16c8__neon_mlal() local 204 const int16x8_t vprod0x12 = vmull_s8(vb12, va0); in xnn_qs8_igemm_minmax_rndnu_ukernel_1x16c8__neon_mlal() local
|
D | 1x16c16-minmax-rndnu-neon-mlal.c | 130 int16x8_t vprod0x12 = vmull_s8(vget_low_s8(vb12), vget_low_s8(va0)); in xnn_qs8_igemm_minmax_rndnu_ukernel_1x16c16__neon_mlal() local
|
D | 1x16c8-minmax-rndnu-neon-mull.c | 115 const int16x8_t vprod0x12 = vmull_s8(vb12, va0); in xnn_qs8_igemm_minmax_rndnu_ukernel_1x16c8__neon_mull() local
|
D | 2x16c8-minmax-rndnu-neon-mlal.c | 206 int16x8_t vprod0x12 = vmull_s8(vb12x0, va0x0); in xnn_qs8_igemm_minmax_rndnu_ukernel_2x16c8__neon_mlal() local 303 const int16x8_t vprod0x12 = vmull_s8(vb12, va0); in xnn_qs8_igemm_minmax_rndnu_ukernel_2x16c8__neon_mlal() local
|
D | 3x16c8-minmax-rndnu-neon-mlal.c | 268 int16x8_t vprod0x12 = vmull_s8(vb12x0, va0x0); in xnn_qs8_igemm_minmax_rndnu_ukernel_3x16c8__neon_mlal() local 402 const int16x8_t vprod0x12 = vmull_s8(vb12, va0); in xnn_qs8_igemm_minmax_rndnu_ukernel_3x16c8__neon_mlal() local
|
D | 2x16c16-minmax-rndnu-neon-mlal.c | 191 int16x8_t vprod0x12 = vmull_s8(vget_low_s8(vb12), vget_low_s8(va0)); in xnn_qs8_igemm_minmax_rndnu_ukernel_2x16c16__neon_mlal() local
|
D | 2x16c8-minmax-rndnu-neon-mull.c | 164 const int16x8_t vprod0x12 = vmull_s8(vb12, va0); in xnn_qs8_igemm_minmax_rndnu_ukernel_2x16c8__neon_mull() local
|
D | 4x16c8-minmax-rndnu-neon-mlal.c | 330 int16x8_t vprod0x12 = vmull_s8(vb12x0, va0x0); in xnn_qs8_igemm_minmax_rndnu_ukernel_4x16c8__neon_mlal() local 501 const int16x8_t vprod0x12 = vmull_s8(vb12, va0); in xnn_qs8_igemm_minmax_rndnu_ukernel_4x16c8__neon_mlal() local
|
D | 3x16c8-minmax-rndnu-neon-mull.c | 213 const int16x8_t vprod0x12 = vmull_s8(vb12, va0); in xnn_qs8_igemm_minmax_rndnu_ukernel_3x16c8__neon_mull() local
|
D | 3x16c16-minmax-rndnu-neon-mlal.c | 252 int16x8_t vprod0x12 = vmull_s8(vget_low_s8(vb12), vget_low_s8(va0)); in xnn_qs8_igemm_minmax_rndnu_ukernel_3x16c16__neon_mlal() local
|
D | 4x16c16-minmax-rndnu-neon-mlal.c | 313 int16x8_t vprod0x12 = vmull_s8(vget_low_s8(vb12), vget_low_s8(va0)); in xnn_qs8_igemm_minmax_rndnu_ukernel_4x16c16__neon_mlal() local
|
D | 4x16c8-minmax-rndnu-neon-mull.c | 262 const int16x8_t vprod0x12 = vmull_s8(vb12, va0); in xnn_qs8_igemm_minmax_rndnu_ukernel_4x16c8__neon_mull() local
|