/external/XNNPACK/src/qs8-gemm/gen/ |
D | 1x16c8-minmax-neon-mlal-padal.c | 125 int16x8_t vprod0x10 = vmull_s8(vb10x0, va0x0); in xnn_qs8_gemm_minmax_ukernel_1x16c8__neon_mlal_padal() local 187 const int16x8_t vprod0x10 = vmull_s8(vb10, va0); in xnn_qs8_gemm_minmax_ukernel_1x16c8__neon_mlal_padal() local
|
D | 1x16c16-minmax-neon-mlal-padal.c | 113 int16x8_t vprod0x10 = vmull_s8(vget_low_s8(vb10), vget_low_s8(va0)); in xnn_qs8_gemm_minmax_ukernel_1x16c16__neon_mlal_padal() local
|
D | 1x16c8-minmax-neon-mull-padal.c | 98 const int16x8_t vprod0x10 = vmull_s8(vb10, va0); in xnn_qs8_gemm_minmax_ukernel_1x16c8__neon_mull_padal() local
|
D | 2x16c8-minmax-neon-mlal-padal.c | 179 int16x8_t vprod0x10 = vmull_s8(vb10x0, va0x0); in xnn_qs8_gemm_minmax_ukernel_2x16c8__neon_mlal_padal() local 280 const int16x8_t vprod0x10 = vmull_s8(vb10, va0); in xnn_qs8_gemm_minmax_ukernel_2x16c8__neon_mlal_padal() local
|
D | 3x16c8-minmax-neon-mlal-padal.c | 233 int16x8_t vprod0x10 = vmull_s8(vb10x0, va0x0); in xnn_qs8_gemm_minmax_ukernel_3x16c8__neon_mlal_padal() local 373 const int16x8_t vprod0x10 = vmull_s8(vb10, va0); in xnn_qs8_gemm_minmax_ukernel_3x16c8__neon_mlal_padal() local
|
D | 2x16c16-minmax-neon-mlal-padal.c | 166 int16x8_t vprod0x10 = vmull_s8(vget_low_s8(vb10), vget_low_s8(va0)); in xnn_qs8_gemm_minmax_ukernel_2x16c16__neon_mlal_padal() local
|
D | 2x16c8-minmax-neon-mull-padal.c | 141 const int16x8_t vprod0x10 = vmull_s8(vb10, va0); in xnn_qs8_gemm_minmax_ukernel_2x16c8__neon_mull_padal() local
|
D | 4x16c8-minmax-neon-mlal-padal.c | 287 int16x8_t vprod0x10 = vmull_s8(vb10x0, va0x0); in xnn_qs8_gemm_minmax_ukernel_4x16c8__neon_mlal_padal() local 466 const int16x8_t vprod0x10 = vmull_s8(vb10, va0); in xnn_qs8_gemm_minmax_ukernel_4x16c8__neon_mlal_padal() local
|
D | 3x16c8-minmax-neon-mull-padal.c | 184 const int16x8_t vprod0x10 = vmull_s8(vb10, va0); in xnn_qs8_gemm_minmax_ukernel_3x16c8__neon_mull_padal() local
|
D | 3x16c16-minmax-neon-mlal-padal.c | 219 int16x8_t vprod0x10 = vmull_s8(vget_low_s8(vb10), vget_low_s8(va0)); in xnn_qs8_gemm_minmax_ukernel_3x16c16__neon_mlal_padal() local
|
D | 4x16c16-minmax-neon-mlal-padal.c | 272 int16x8_t vprod0x10 = vmull_s8(vget_low_s8(vb10), vget_low_s8(va0)); in xnn_qs8_gemm_minmax_ukernel_4x16c16__neon_mlal_padal() local
|
D | 4x16c8-minmax-neon-mull-padal.c | 227 const int16x8_t vprod0x10 = vmull_s8(vb10, va0); in xnn_qs8_gemm_minmax_ukernel_4x16c8__neon_mull_padal() local
|
/external/XNNPACK/src/qs8-igemm/gen/ |
D | 1x16c8-minmax-neon-mlal-padal.c | 136 int16x8_t vprod0x10 = vmull_s8(vb10x0, va0x0); in xnn_qs8_igemm_minmax_ukernel_1x16c8__neon_mlal_padal() local 198 const int16x8_t vprod0x10 = vmull_s8(vb10, va0); in xnn_qs8_igemm_minmax_ukernel_1x16c8__neon_mlal_padal() local
|
D | 1x16c8-minmax-neon-mull-padal.c | 109 const int16x8_t vprod0x10 = vmull_s8(vb10, va0); in xnn_qs8_igemm_minmax_ukernel_1x16c8__neon_mull_padal() local
|
D | 1x16c16-minmax-neon-mlal-padal.c | 124 int16x8_t vprod0x10 = vmull_s8(vget_low_s8(vb10), vget_low_s8(va0)); in xnn_qs8_igemm_minmax_ukernel_1x16c16__neon_mlal_padal() local
|
D | 2x16c8-minmax-neon-mlal-padal.c | 192 int16x8_t vprod0x10 = vmull_s8(vb10x0, va0x0); in xnn_qs8_igemm_minmax_ukernel_2x16c8__neon_mlal_padal() local 293 const int16x8_t vprod0x10 = vmull_s8(vb10, va0); in xnn_qs8_igemm_minmax_ukernel_2x16c8__neon_mlal_padal() local
|
D | 3x16c8-minmax-neon-mlal-padal.c | 248 int16x8_t vprod0x10 = vmull_s8(vb10x0, va0x0); in xnn_qs8_igemm_minmax_ukernel_3x16c8__neon_mlal_padal() local 388 const int16x8_t vprod0x10 = vmull_s8(vb10, va0); in xnn_qs8_igemm_minmax_ukernel_3x16c8__neon_mlal_padal() local
|
D | 2x16c8-minmax-neon-mull-padal.c | 154 const int16x8_t vprod0x10 = vmull_s8(vb10, va0); in xnn_qs8_igemm_minmax_ukernel_2x16c8__neon_mull_padal() local
|
D | 2x16c16-minmax-neon-mlal-padal.c | 179 int16x8_t vprod0x10 = vmull_s8(vget_low_s8(vb10), vget_low_s8(va0)); in xnn_qs8_igemm_minmax_ukernel_2x16c16__neon_mlal_padal() local
|
D | 4x16c8-minmax-neon-mlal-padal.c | 304 int16x8_t vprod0x10 = vmull_s8(vb10x0, va0x0); in xnn_qs8_igemm_minmax_ukernel_4x16c8__neon_mlal_padal() local 483 const int16x8_t vprod0x10 = vmull_s8(vb10, va0); in xnn_qs8_igemm_minmax_ukernel_4x16c8__neon_mlal_padal() local
|
D | 3x16c8-minmax-neon-mull-padal.c | 199 const int16x8_t vprod0x10 = vmull_s8(vb10, va0); in xnn_qs8_igemm_minmax_ukernel_3x16c8__neon_mull_padal() local
|
D | 3x16c16-minmax-neon-mlal-padal.c | 234 int16x8_t vprod0x10 = vmull_s8(vget_low_s8(vb10), vget_low_s8(va0)); in xnn_qs8_igemm_minmax_ukernel_3x16c16__neon_mlal_padal() local
|
D | 4x16c8-minmax-neon-mull-padal.c | 244 const int16x8_t vprod0x10 = vmull_s8(vb10, va0); in xnn_qs8_igemm_minmax_ukernel_4x16c8__neon_mull_padal() local
|
D | 4x16c16-minmax-neon-mlal-padal.c | 289 int16x8_t vprod0x10 = vmull_s8(vget_low_s8(vb10), vget_low_s8(va0)); in xnn_qs8_igemm_minmax_ukernel_4x16c16__neon_mlal_padal() local
|