/external/XNNPACK/src/qs8-gemm/gen/ |
D | 1x8c8-minmax-rndnu-neon-mlal.c | 81 int16x8_t vprod0x3 = vmull_s8(vb3x0, va0x0); in xnn_qs8_gemm_minmax_rndnu_ukernel_1x8c8__neon_mlal() local 118 const int16x8_t vprod0x3 = vmull_s8(vb3, va0); in xnn_qs8_gemm_minmax_rndnu_ukernel_1x8c8__neon_mlal() local
|
D | 1x8c8-minmax-fp32-neonv8-mlal.c | 82 int16x8_t vprod0x3 = vmull_s8(vb3x0, va0x0); in xnn_qs8_gemm_minmax_fp32_ukernel_1x8c8__neonv8_mlal() local 119 const int16x8_t vprod0x3 = vmull_s8(vb3, va0); in xnn_qs8_gemm_minmax_fp32_ukernel_1x8c8__neonv8_mlal() local
|
D | 1x8c8-minmax-fp32-neon-mlal.c | 81 int16x8_t vprod0x3 = vmull_s8(vb3x0, va0x0); in xnn_qs8_gemm_minmax_fp32_ukernel_1x8c8__neon_mlal() local 118 const int16x8_t vprod0x3 = vmull_s8(vb3, va0); in xnn_qs8_gemm_minmax_fp32_ukernel_1x8c8__neon_mlal() local
|
D | 2x8c8-minmax-fp32-neonv8-mlal.c | 107 int16x8_t vprod0x3 = vmull_s8(vb3x0, va0x0); in xnn_qs8_gemm_minmax_fp32_ukernel_2x8c8__neonv8_mlal() local 166 const int16x8_t vprod0x3 = vmull_s8(vb3, va0); in xnn_qs8_gemm_minmax_fp32_ukernel_2x8c8__neonv8_mlal() local
|
D | 2x8c8-minmax-fp32-neon-mlal.c | 106 int16x8_t vprod0x3 = vmull_s8(vb3x0, va0x0); in xnn_qs8_gemm_minmax_fp32_ukernel_2x8c8__neon_mlal() local 165 const int16x8_t vprod0x3 = vmull_s8(vb3, va0); in xnn_qs8_gemm_minmax_fp32_ukernel_2x8c8__neon_mlal() local
|
D | 2x8c8-minmax-rndnu-neon-mlal.c | 106 int16x8_t vprod0x3 = vmull_s8(vb3x0, va0x0); in xnn_qs8_gemm_minmax_rndnu_ukernel_2x8c8__neon_mlal() local 165 const int16x8_t vprod0x3 = vmull_s8(vb3, va0); in xnn_qs8_gemm_minmax_rndnu_ukernel_2x8c8__neon_mlal() local
|
D | 1x16c8-minmax-rndnu-neon-mlal.c | 97 int16x8_t vprod0x3 = vmull_s8(vb3x0, va0x0); in xnn_qs8_gemm_minmax_rndnu_ukernel_1x16c8__neon_mlal() local 166 const int16x8_t vprod0x3 = vmull_s8(vb3, va0); in xnn_qs8_gemm_minmax_rndnu_ukernel_1x16c8__neon_mlal() local
|
D | 1x8c8-minmax-rndnu-neon-mull.c | 69 const int16x8_t vprod0x3 = vmull_s8(vb3, va0); in xnn_qs8_gemm_minmax_rndnu_ukernel_1x8c8__neon_mull() local
|
D | 1x8c16-minmax-rndnu-neon-mlal.c | 76 int16x8_t vprod0x3 = vmull_s8(vget_low_s8(vb3), vget_low_s8(va0)); in xnn_qs8_gemm_minmax_rndnu_ukernel_1x8c16__neon_mlal() local
|
/external/XNNPACK/src/qc8-igemm/gen/ |
D | 1x8c8-minmax-fp32-neon-mlal.c | 92 int16x8_t vprod0x3 = vmull_s8(vb3x0, va0x0); in xnn_qc8_igemm_minmax_fp32_ukernel_1x8c8__neon_mlal() local 129 const int16x8_t vprod0x3 = vmull_s8(vb3, va0); in xnn_qc8_igemm_minmax_fp32_ukernel_1x8c8__neon_mlal() local
|
D | 1x8c8-minmax-fp32-neonv8-mlal.c | 93 int16x8_t vprod0x3 = vmull_s8(vb3x0, va0x0); in xnn_qc8_igemm_minmax_fp32_ukernel_1x8c8__neonv8_mlal() local 130 const int16x8_t vprod0x3 = vmull_s8(vb3, va0); in xnn_qc8_igemm_minmax_fp32_ukernel_1x8c8__neonv8_mlal() local
|
D | 2x8c8-minmax-fp32-neonv8-mlal.c | 120 int16x8_t vprod0x3 = vmull_s8(vb3x0, va0x0); in xnn_qc8_igemm_minmax_fp32_ukernel_2x8c8__neonv8_mlal() local 179 const int16x8_t vprod0x3 = vmull_s8(vb3, va0); in xnn_qc8_igemm_minmax_fp32_ukernel_2x8c8__neonv8_mlal() local
|
D | 2x8c8-minmax-fp32-neon-mlal.c | 119 int16x8_t vprod0x3 = vmull_s8(vb3x0, va0x0); in xnn_qc8_igemm_minmax_fp32_ukernel_2x8c8__neon_mlal() local 178 const int16x8_t vprod0x3 = vmull_s8(vb3, va0); in xnn_qc8_igemm_minmax_fp32_ukernel_2x8c8__neon_mlal() local
|
/external/XNNPACK/src/qc8-gemm/gen/ |
D | 1x8c8-minmax-fp32-neon-mlal.c | 81 int16x8_t vprod0x3 = vmull_s8(vb3x0, va0x0); in xnn_qc8_gemm_minmax_fp32_ukernel_1x8c8__neon_mlal() local 118 const int16x8_t vprod0x3 = vmull_s8(vb3, va0); in xnn_qc8_gemm_minmax_fp32_ukernel_1x8c8__neon_mlal() local
|
D | 1x8c8-minmax-fp32-neonv8-mlal.c | 82 int16x8_t vprod0x3 = vmull_s8(vb3x0, va0x0); in xnn_qc8_gemm_minmax_fp32_ukernel_1x8c8__neonv8_mlal() local 119 const int16x8_t vprod0x3 = vmull_s8(vb3, va0); in xnn_qc8_gemm_minmax_fp32_ukernel_1x8c8__neonv8_mlal() local
|
D | 2x8c8-minmax-fp32-neon-mlal.c | 106 int16x8_t vprod0x3 = vmull_s8(vb3x0, va0x0); in xnn_qc8_gemm_minmax_fp32_ukernel_2x8c8__neon_mlal() local 165 const int16x8_t vprod0x3 = vmull_s8(vb3, va0); in xnn_qc8_gemm_minmax_fp32_ukernel_2x8c8__neon_mlal() local
|
D | 2x8c8-minmax-fp32-neonv8-mlal.c | 107 int16x8_t vprod0x3 = vmull_s8(vb3x0, va0x0); in xnn_qc8_gemm_minmax_fp32_ukernel_2x8c8__neonv8_mlal() local 166 const int16x8_t vprod0x3 = vmull_s8(vb3, va0); in xnn_qc8_gemm_minmax_fp32_ukernel_2x8c8__neonv8_mlal() local
|
/external/XNNPACK/src/qs8-igemm/gen/ |
D | 1x8c8-minmax-fp32-neonv8-mlal.c | 93 int16x8_t vprod0x3 = vmull_s8(vb3x0, va0x0); in xnn_qs8_igemm_minmax_fp32_ukernel_1x8c8__neonv8_mlal() local 130 const int16x8_t vprod0x3 = vmull_s8(vb3, va0); in xnn_qs8_igemm_minmax_fp32_ukernel_1x8c8__neonv8_mlal() local
|
D | 1x8c8-minmax-rndnu-neon-mlal.c | 92 int16x8_t vprod0x3 = vmull_s8(vb3x0, va0x0); in xnn_qs8_igemm_minmax_rndnu_ukernel_1x8c8__neon_mlal() local 129 const int16x8_t vprod0x3 = vmull_s8(vb3, va0); in xnn_qs8_igemm_minmax_rndnu_ukernel_1x8c8__neon_mlal() local
|
D | 1x8c8-minmax-fp32-neon-mlal.c | 92 int16x8_t vprod0x3 = vmull_s8(vb3x0, va0x0); in xnn_qs8_igemm_minmax_fp32_ukernel_1x8c8__neon_mlal() local 129 const int16x8_t vprod0x3 = vmull_s8(vb3, va0); in xnn_qs8_igemm_minmax_fp32_ukernel_1x8c8__neon_mlal() local
|
D | 2x8c8-minmax-rndnu-neon-mlal.c | 119 int16x8_t vprod0x3 = vmull_s8(vb3x0, va0x0); in xnn_qs8_igemm_minmax_rndnu_ukernel_2x8c8__neon_mlal() local 178 const int16x8_t vprod0x3 = vmull_s8(vb3, va0); in xnn_qs8_igemm_minmax_rndnu_ukernel_2x8c8__neon_mlal() local
|
D | 2x8c8-minmax-fp32-neonv8-mlal.c | 120 int16x8_t vprod0x3 = vmull_s8(vb3x0, va0x0); in xnn_qs8_igemm_minmax_fp32_ukernel_2x8c8__neonv8_mlal() local 179 const int16x8_t vprod0x3 = vmull_s8(vb3, va0); in xnn_qs8_igemm_minmax_fp32_ukernel_2x8c8__neonv8_mlal() local
|
D | 2x8c8-minmax-fp32-neon-mlal.c | 119 int16x8_t vprod0x3 = vmull_s8(vb3x0, va0x0); in xnn_qs8_igemm_minmax_fp32_ukernel_2x8c8__neon_mlal() local 178 const int16x8_t vprod0x3 = vmull_s8(vb3, va0); in xnn_qs8_igemm_minmax_fp32_ukernel_2x8c8__neon_mlal() local
|
D | 1x16c8-minmax-rndnu-neon-mlal.c | 108 int16x8_t vprod0x3 = vmull_s8(vb3x0, va0x0); in xnn_qs8_igemm_minmax_rndnu_ukernel_1x16c8__neon_mlal() local 177 const int16x8_t vprod0x3 = vmull_s8(vb3, va0); in xnn_qs8_igemm_minmax_rndnu_ukernel_1x16c8__neon_mlal() local
|
D | 1x8c8-minmax-rndnu-neon-mull.c | 80 const int16x8_t vprod0x3 = vmull_s8(vb3, va0); in xnn_qs8_igemm_minmax_rndnu_ukernel_1x8c8__neon_mull() local
|