/external/XNNPACK/src/qs8-gemm/gen/ |
D | 2x8c8-minmax-neon-mlal-padal.c | 93 int16x8_t vprod1x1 = vmull_s8(vb1x0, va1x0); in xnn_qs8_gemm_minmax_ukernel_2x8c8__neon_mlal_padal() local 156 const int16x8_t vprod1x1 = vmull_s8(vb1, va1); in xnn_qs8_gemm_minmax_ukernel_2x8c8__neon_mlal_padal() local
|
D | 2x4c8-xw-minmax-wasmsimd.c | 81 const v128_t vprod1x1 = wasm_i16x8_mul(vxa1, vxb1); in xnn_qs8_gemm_xw_minmax_ukernel_2x4c8__wasmsimd() local
|
D | 2x4c8-minmax-wasmsimd-ld128.c | 80 const v128_t vprod1x1 = wasm_i16x8_mul(vxb1, vxa1); in xnn_qs8_gemm_minmax_ukernel_2x4c8__wasmsimd_ld128() local
|
D | 2x4c8-minmax-wasmsimd-ld64.c | 81 const v128_t vprod1x1 = wasm_i16x8_mul(vxa1, vxb1); in xnn_qs8_gemm_minmax_ukernel_2x4c8__wasmsimd_ld64() local
|
D | 3x8c8-minmax-neon-mlal-padal.c | 112 int16x8_t vprod1x1 = vmull_s8(vb1x0, va1x0); in xnn_qs8_gemm_minmax_ukernel_3x8c8__neon_mlal_padal() local 199 const int16x8_t vprod1x1 = vmull_s8(vb1, va1); in xnn_qs8_gemm_minmax_ukernel_3x8c8__neon_mlal_padal() local
|
D | 3x4c8-xw-minmax-wasmsimd.c | 96 const v128_t vprod1x1 = wasm_i16x8_mul(vxa1, vxb1); in xnn_qs8_gemm_xw_minmax_ukernel_3x4c8__wasmsimd() local
|
D | 3x4c8-minmax-wasmsimd-ld128.c | 94 const v128_t vprod1x1 = wasm_i16x8_mul(vxb1, vxa1); in xnn_qs8_gemm_minmax_ukernel_3x4c8__wasmsimd_ld128() local
|
D | 3x4c8-minmax-wasmsimd-ld64.c | 96 const v128_t vprod1x1 = wasm_i16x8_mul(vxa1, vxb1); in xnn_qs8_gemm_minmax_ukernel_3x4c8__wasmsimd_ld64() local
|
D | 2x8c8-minmax-neon-mull-padal.c | 81 const int16x8_t vprod1x1 = vmull_s8(vb1, va1); in xnn_qs8_gemm_minmax_ukernel_2x8c8__neon_mull_padal() local
|
D | 2x8c16-minmax-neon-mlal-padal.c | 89 int16x8_t vprod1x1 = vmull_s8(vget_low_s8(vb1), vget_low_s8(va1)); in xnn_qs8_gemm_minmax_ukernel_2x8c16__neon_mlal_padal() local
|
D | 4x8c8-minmax-neon-mlal-padal.c | 131 int16x8_t vprod1x1 = vmull_s8(vb1x0, va1x0); in xnn_qs8_gemm_minmax_ukernel_4x8c8__neon_mlal_padal() local 242 const int16x8_t vprod1x1 = vmull_s8(vb1, va1); in xnn_qs8_gemm_minmax_ukernel_4x8c8__neon_mlal_padal() local
|
D | 2x16c8-minmax-neon-mlal-padal.c | 117 int16x8_t vprod1x1 = vmull_s8(vb1x0, va1x0); in xnn_qs8_gemm_minmax_ukernel_2x16c8__neon_mlal_padal() local 236 const int16x8_t vprod1x1 = vmull_s8(vb1, va1); in xnn_qs8_gemm_minmax_ukernel_2x16c8__neon_mlal_padal() local
|
D | 3x8c16-minmax-neon-mlal-padal.c | 107 int16x8_t vprod1x1 = vmull_s8(vget_low_s8(vb1), vget_low_s8(va1)); in xnn_qs8_gemm_minmax_ukernel_3x8c16__neon_mlal_padal() local
|
/external/XNNPACK/src/qs8-igemm/gen/ |
D | 2x8c8-minmax-neon-mlal-padal.c | 106 int16x8_t vprod1x1 = vmull_s8(vb1x0, va1x0); in xnn_qs8_igemm_minmax_ukernel_2x8c8__neon_mlal_padal() local 169 const int16x8_t vprod1x1 = vmull_s8(vb1, va1); in xnn_qs8_igemm_minmax_ukernel_2x8c8__neon_mlal_padal() local
|
D | 2x4c8-minmax-wasmsimd-ld128.c | 93 const v128_t vprod1x1 = wasm_i16x8_mul(vxb1, vxa1); in xnn_qs8_igemm_minmax_ukernel_2x4c8__wasmsimd_ld128() local
|
D | 2x4c8-minmax-wasmsimd-ld64.c | 94 const v128_t vprod1x1 = wasm_i16x8_mul(vxa1, vxb1); in xnn_qs8_igemm_minmax_ukernel_2x4c8__wasmsimd_ld64() local
|
D | 3x8c8-minmax-neon-mlal-padal.c | 127 int16x8_t vprod1x1 = vmull_s8(vb1x0, va1x0); in xnn_qs8_igemm_minmax_ukernel_3x8c8__neon_mlal_padal() local 214 const int16x8_t vprod1x1 = vmull_s8(vb1, va1); in xnn_qs8_igemm_minmax_ukernel_3x8c8__neon_mlal_padal() local
|
D | 3x4c8-minmax-wasmsimd-ld64.c | 111 const v128_t vprod1x1 = wasm_i16x8_mul(vxa1, vxb1); in xnn_qs8_igemm_minmax_ukernel_3x4c8__wasmsimd_ld64() local
|
D | 3x4c8-minmax-wasmsimd-ld128.c | 109 const v128_t vprod1x1 = wasm_i16x8_mul(vxb1, vxa1); in xnn_qs8_igemm_minmax_ukernel_3x4c8__wasmsimd_ld128() local
|
D | 4x8c8-minmax-neon-mlal-padal.c | 148 int16x8_t vprod1x1 = vmull_s8(vb1x0, va1x0); in xnn_qs8_igemm_minmax_ukernel_4x8c8__neon_mlal_padal() local 259 const int16x8_t vprod1x1 = vmull_s8(vb1, va1); in xnn_qs8_igemm_minmax_ukernel_4x8c8__neon_mlal_padal() local
|
D | 2x8c16-minmax-neon-mlal-padal.c | 102 int16x8_t vprod1x1 = vmull_s8(vget_low_s8(vb1), vget_low_s8(va1)); in xnn_qs8_igemm_minmax_ukernel_2x8c16__neon_mlal_padal() local
|
D | 2x8c8-minmax-neon-mull-padal.c | 94 const int16x8_t vprod1x1 = vmull_s8(vb1, va1); in xnn_qs8_igemm_minmax_ukernel_2x8c8__neon_mull_padal() local
|
D | 2x16c8-minmax-neon-mlal-padal.c | 130 int16x8_t vprod1x1 = vmull_s8(vb1x0, va1x0); in xnn_qs8_igemm_minmax_ukernel_2x16c8__neon_mlal_padal() local 249 const int16x8_t vprod1x1 = vmull_s8(vb1, va1); in xnn_qs8_igemm_minmax_ukernel_2x16c8__neon_mlal_padal() local
|
D | 3x8c8-minmax-neon-mull-padal.c | 113 const int16x8_t vprod1x1 = vmull_s8(vb1, va1); in xnn_qs8_igemm_minmax_ukernel_3x8c8__neon_mull_padal() local
|
D | 3x8c16-minmax-neon-mlal-padal.c | 122 int16x8_t vprod1x1 = vmull_s8(vget_low_s8(vb1), vget_low_s8(va1)); in xnn_qs8_igemm_minmax_ukernel_3x8c16__neon_mlal_padal() local
|