/external/XNNPACK/src/f32-gemm/gen/ |
D | 2x4-minmax-scalar.c | 54 float vacc10 = vacc00; in xnn_f32_gemm_minmax_ukernel_2x4__scalar() local 74 vacc10 += va1 * vb0; in xnn_f32_gemm_minmax_ukernel_2x4__scalar() 86 vacc10 = math_max_f32(vacc10, vmin); in xnn_f32_gemm_minmax_ukernel_2x4__scalar() 95 vacc10 = math_min_f32(vacc10, vmax); in xnn_f32_gemm_minmax_ukernel_2x4__scalar() 101 c1[0] = vacc10; in xnn_f32_gemm_minmax_ukernel_2x4__scalar() 118 c1[0] = vacc10; in xnn_f32_gemm_minmax_ukernel_2x4__scalar() 120 vacc10 = vacc12; in xnn_f32_gemm_minmax_ukernel_2x4__scalar() 128 c1[0] = vacc10; in xnn_f32_gemm_minmax_ukernel_2x4__scalar()
|
D | 2x4-minmax-wasm.c | 54 float vacc10 = vacc00; in xnn_f32_gemm_minmax_ukernel_2x4__wasm() local 74 vacc10 += va1 * vb0; in xnn_f32_gemm_minmax_ukernel_2x4__wasm() 86 vacc10 = __builtin_wasm_max_f32(vacc10, vmin); in xnn_f32_gemm_minmax_ukernel_2x4__wasm() 95 vacc10 = __builtin_wasm_min_f32(vacc10, vmax); in xnn_f32_gemm_minmax_ukernel_2x4__wasm() 101 c1[0] = vacc10; in xnn_f32_gemm_minmax_ukernel_2x4__wasm() 118 c1[0] = vacc10; in xnn_f32_gemm_minmax_ukernel_2x4__wasm() 120 vacc10 = vacc12; in xnn_f32_gemm_minmax_ukernel_2x4__wasm() 128 c1[0] = vacc10; in xnn_f32_gemm_minmax_ukernel_2x4__wasm()
|
D | 2x4-relu-scalar.c | 52 float vacc10 = vacc00; in xnn_f32_gemm_relu_ukernel_2x4__scalar() local 72 vacc10 += va1 * vb0; in xnn_f32_gemm_relu_ukernel_2x4__scalar() 84 vacc10 = math_max_f32(vacc10, 0.0f); in xnn_f32_gemm_relu_ukernel_2x4__scalar() 90 c1[0] = vacc10; in xnn_f32_gemm_relu_ukernel_2x4__scalar() 107 c1[0] = vacc10; in xnn_f32_gemm_relu_ukernel_2x4__scalar() 109 vacc10 = vacc12; in xnn_f32_gemm_relu_ukernel_2x4__scalar() 117 c1[0] = vacc10; in xnn_f32_gemm_relu_ukernel_2x4__scalar()
|
D | 2x4-relu-wasm.c | 52 float vacc10 = vacc00; in xnn_f32_gemm_relu_ukernel_2x4__wasm() local 72 vacc10 += va1 * vb0; in xnn_f32_gemm_relu_ukernel_2x4__wasm() 84 vacc10 = __builtin_wasm_max_f32(vacc10, 0.0f); in xnn_f32_gemm_relu_ukernel_2x4__wasm() 90 c1[0] = vacc10; in xnn_f32_gemm_relu_ukernel_2x4__wasm() 107 c1[0] = vacc10; in xnn_f32_gemm_relu_ukernel_2x4__wasm() 109 vacc10 = vacc12; in xnn_f32_gemm_relu_ukernel_2x4__wasm() 117 c1[0] = vacc10; in xnn_f32_gemm_relu_ukernel_2x4__wasm()
|
D | 4x2-minmax-scalar.c | 64 float vacc10 = vacc00; in xnn_f32_gemm_minmax_ukernel_4x2__scalar() local 84 vacc10 += va1 * vb0; in xnn_f32_gemm_minmax_ukernel_4x2__scalar() 96 vacc10 = math_max_f32(vacc10, vmin); in xnn_f32_gemm_minmax_ukernel_4x2__scalar() 105 vacc10 = math_min_f32(vacc10, vmax); in xnn_f32_gemm_minmax_ukernel_4x2__scalar() 119 c1[0] = vacc10; in xnn_f32_gemm_minmax_ukernel_4x2__scalar() 136 c1[0] = vacc10; in xnn_f32_gemm_minmax_ukernel_4x2__scalar()
|
D | 4x2-minmax-wasm.c | 64 float vacc10 = vacc00; in xnn_f32_gemm_minmax_ukernel_4x2__wasm() local 84 vacc10 += va1 * vb0; in xnn_f32_gemm_minmax_ukernel_4x2__wasm() 96 vacc10 = __builtin_wasm_max_f32(vacc10, vmin); in xnn_f32_gemm_minmax_ukernel_4x2__wasm() 105 vacc10 = __builtin_wasm_min_f32(vacc10, vmax); in xnn_f32_gemm_minmax_ukernel_4x2__wasm() 119 c1[0] = vacc10; in xnn_f32_gemm_minmax_ukernel_4x2__wasm() 136 c1[0] = vacc10; in xnn_f32_gemm_minmax_ukernel_4x2__wasm()
|
D | 4x4-minmax-scalar.c | 66 float vacc10 = vacc00; in xnn_f32_gemm_minmax_ukernel_4x4__scalar() local 96 vacc10 += va1 * vb0; in xnn_f32_gemm_minmax_ukernel_4x4__scalar() 116 vacc10 = math_max_f32(vacc10, vmin); in xnn_f32_gemm_minmax_ukernel_4x4__scalar() 133 vacc10 = math_min_f32(vacc10, vmax); in xnn_f32_gemm_minmax_ukernel_4x4__scalar() 157 c1[0] = vacc10; in xnn_f32_gemm_minmax_ukernel_4x4__scalar() 184 c1[0] = vacc10; in xnn_f32_gemm_minmax_ukernel_4x4__scalar() 186 vacc10 = vacc12; in xnn_f32_gemm_minmax_ukernel_4x4__scalar() 196 c1[0] = vacc10; in xnn_f32_gemm_minmax_ukernel_4x4__scalar()
|
D | 4x4-minmax-wasm.c | 66 float vacc10 = vacc00; in xnn_f32_gemm_minmax_ukernel_4x4__wasm() local 96 vacc10 += va1 * vb0; in xnn_f32_gemm_minmax_ukernel_4x4__wasm() 116 vacc10 = __builtin_wasm_max_f32(vacc10, vmin); in xnn_f32_gemm_minmax_ukernel_4x4__wasm() 133 vacc10 = __builtin_wasm_min_f32(vacc10, vmax); in xnn_f32_gemm_minmax_ukernel_4x4__wasm() 157 c1[0] = vacc10; in xnn_f32_gemm_minmax_ukernel_4x4__wasm() 184 c1[0] = vacc10; in xnn_f32_gemm_minmax_ukernel_4x4__wasm() 186 vacc10 = vacc12; in xnn_f32_gemm_minmax_ukernel_4x4__wasm() 196 c1[0] = vacc10; in xnn_f32_gemm_minmax_ukernel_4x4__wasm()
|
D | 2x4-wasm.c | 52 float vacc10 = vacc00; in xnn_f32_gemm_ukernel_2x4__wasm() local 72 vacc10 += va1 * vb0; in xnn_f32_gemm_ukernel_2x4__wasm() 82 c1[0] = vacc10; in xnn_f32_gemm_ukernel_2x4__wasm() 99 c1[0] = vacc10; in xnn_f32_gemm_ukernel_2x4__wasm() 101 vacc10 = vacc12; in xnn_f32_gemm_ukernel_2x4__wasm() 109 c1[0] = vacc10; in xnn_f32_gemm_ukernel_2x4__wasm()
|
D | 2x4-scalar.c | 52 float vacc10 = vacc00; in xnn_f32_gemm_ukernel_2x4__scalar() local 72 vacc10 += va1 * vb0; in xnn_f32_gemm_ukernel_2x4__scalar() 82 c1[0] = vacc10; in xnn_f32_gemm_ukernel_2x4__scalar() 99 c1[0] = vacc10; in xnn_f32_gemm_ukernel_2x4__scalar() 101 vacc10 = vacc12; in xnn_f32_gemm_ukernel_2x4__scalar() 109 c1[0] = vacc10; in xnn_f32_gemm_ukernel_2x4__scalar()
|
D | 4x4-relu-wasm.c | 64 float vacc10 = vacc00; in xnn_f32_gemm_relu_ukernel_4x4__wasm() local 94 vacc10 += va1 * vb0; in xnn_f32_gemm_relu_ukernel_4x4__wasm() 114 vacc10 = __builtin_wasm_max_f32(vacc10, 0.0f); in xnn_f32_gemm_relu_ukernel_4x4__wasm() 138 c1[0] = vacc10; in xnn_f32_gemm_relu_ukernel_4x4__wasm() 165 c1[0] = vacc10; in xnn_f32_gemm_relu_ukernel_4x4__wasm() 167 vacc10 = vacc12; in xnn_f32_gemm_relu_ukernel_4x4__wasm() 177 c1[0] = vacc10; in xnn_f32_gemm_relu_ukernel_4x4__wasm()
|
D | 4x4-relu-scalar.c | 64 float vacc10 = vacc00; in xnn_f32_gemm_relu_ukernel_4x4__scalar() local 94 vacc10 += va1 * vb0; in xnn_f32_gemm_relu_ukernel_4x4__scalar() 114 vacc10 = math_max_f32(vacc10, 0.0f); in xnn_f32_gemm_relu_ukernel_4x4__scalar() 138 c1[0] = vacc10; in xnn_f32_gemm_relu_ukernel_4x4__scalar() 165 c1[0] = vacc10; in xnn_f32_gemm_relu_ukernel_4x4__scalar() 167 vacc10 = vacc12; in xnn_f32_gemm_relu_ukernel_4x4__scalar() 177 c1[0] = vacc10; in xnn_f32_gemm_relu_ukernel_4x4__scalar()
|
/external/XNNPACK/src/f32-gemm/gen-inc/ |
D | 2x4inc-minmax-scalar.c | 55 float vacc10 = acc[4]; in xnn_f32_gemminc_minmax_ukernel_2x4__scalar() local 76 vacc10 += va1 * vb0; in xnn_f32_gemminc_minmax_ukernel_2x4__scalar() 88 vacc10 = math_max_f32(vacc10, vmin); in xnn_f32_gemminc_minmax_ukernel_2x4__scalar() 97 vacc10 = math_min_f32(vacc10, vmax); in xnn_f32_gemminc_minmax_ukernel_2x4__scalar() 103 c1[0] = vacc10; in xnn_f32_gemminc_minmax_ukernel_2x4__scalar() 120 c1[0] = vacc10; in xnn_f32_gemminc_minmax_ukernel_2x4__scalar() 122 vacc10 = vacc12; in xnn_f32_gemminc_minmax_ukernel_2x4__scalar() 130 c1[0] = vacc10; in xnn_f32_gemminc_minmax_ukernel_2x4__scalar()
|
D | 2x4inc-minmax-wasm.c | 55 float vacc10 = acc[4]; in xnn_f32_gemminc_minmax_ukernel_2x4__wasm() local 76 vacc10 += va1 * vb0; in xnn_f32_gemminc_minmax_ukernel_2x4__wasm() 88 vacc10 = __builtin_wasm_max_f32(vacc10, vmin); in xnn_f32_gemminc_minmax_ukernel_2x4__wasm() 97 vacc10 = __builtin_wasm_min_f32(vacc10, vmax); in xnn_f32_gemminc_minmax_ukernel_2x4__wasm() 103 c1[0] = vacc10; in xnn_f32_gemminc_minmax_ukernel_2x4__wasm() 120 c1[0] = vacc10; in xnn_f32_gemminc_minmax_ukernel_2x4__wasm() 122 vacc10 = vacc12; in xnn_f32_gemminc_minmax_ukernel_2x4__wasm() 130 c1[0] = vacc10; in xnn_f32_gemminc_minmax_ukernel_2x4__wasm()
|
D | 4x4inc-minmax-scalar.c | 67 float vacc10 = acc[4]; in xnn_f32_gemminc_minmax_ukernel_4x4__scalar() local 98 vacc10 += va1 * vb0; in xnn_f32_gemminc_minmax_ukernel_4x4__scalar() 118 vacc10 = math_max_f32(vacc10, vmin); in xnn_f32_gemminc_minmax_ukernel_4x4__scalar() 135 vacc10 = math_min_f32(vacc10, vmax); in xnn_f32_gemminc_minmax_ukernel_4x4__scalar() 159 c1[0] = vacc10; in xnn_f32_gemminc_minmax_ukernel_4x4__scalar() 186 c1[0] = vacc10; in xnn_f32_gemminc_minmax_ukernel_4x4__scalar() 188 vacc10 = vacc12; in xnn_f32_gemminc_minmax_ukernel_4x4__scalar() 198 c1[0] = vacc10; in xnn_f32_gemminc_minmax_ukernel_4x4__scalar()
|
D | 4x4inc-minmax-wasm.c | 67 float vacc10 = acc[4]; in xnn_f32_gemminc_minmax_ukernel_4x4__wasm() local 98 vacc10 += va1 * vb0; in xnn_f32_gemminc_minmax_ukernel_4x4__wasm() 118 vacc10 = __builtin_wasm_max_f32(vacc10, vmin); in xnn_f32_gemminc_minmax_ukernel_4x4__wasm() 135 vacc10 = __builtin_wasm_min_f32(vacc10, vmax); in xnn_f32_gemminc_minmax_ukernel_4x4__wasm() 159 c1[0] = vacc10; in xnn_f32_gemminc_minmax_ukernel_4x4__wasm() 186 c1[0] = vacc10; in xnn_f32_gemminc_minmax_ukernel_4x4__wasm() 188 vacc10 = vacc12; in xnn_f32_gemminc_minmax_ukernel_4x4__wasm() 198 c1[0] = vacc10; in xnn_f32_gemminc_minmax_ukernel_4x4__wasm()
|
/external/XNNPACK/src/f32-igemm/gen/ |
D | 2x4-minmax-wasm.c | 55 float vacc10 = vacc00; in xnn_f32_igemm_minmax_ukernel_2x4__wasm() local 90 vacc10 += va1 * vb0; in xnn_f32_igemm_minmax_ukernel_2x4__wasm() 104 vacc10 = __builtin_wasm_max_f32(vacc10, vmin); in xnn_f32_igemm_minmax_ukernel_2x4__wasm() 113 vacc10 = __builtin_wasm_min_f32(vacc10, vmax); in xnn_f32_igemm_minmax_ukernel_2x4__wasm() 119 c1[0] = vacc10; in xnn_f32_igemm_minmax_ukernel_2x4__wasm() 134 c1[0] = vacc10; in xnn_f32_igemm_minmax_ukernel_2x4__wasm() 136 vacc10 = vacc12; in xnn_f32_igemm_minmax_ukernel_2x4__wasm() 144 c1[0] = vacc10; in xnn_f32_igemm_minmax_ukernel_2x4__wasm()
|
D | 2x4-minmax-scalar.c | 55 float vacc10 = vacc00; in xnn_f32_igemm_minmax_ukernel_2x4__scalar() local 90 vacc10 += va1 * vb0; in xnn_f32_igemm_minmax_ukernel_2x4__scalar() 104 vacc10 = math_max_f32(vacc10, vmin); in xnn_f32_igemm_minmax_ukernel_2x4__scalar() 113 vacc10 = math_min_f32(vacc10, vmax); in xnn_f32_igemm_minmax_ukernel_2x4__scalar() 119 c1[0] = vacc10; in xnn_f32_igemm_minmax_ukernel_2x4__scalar() 134 c1[0] = vacc10; in xnn_f32_igemm_minmax_ukernel_2x4__scalar() 136 vacc10 = vacc12; in xnn_f32_igemm_minmax_ukernel_2x4__scalar() 144 c1[0] = vacc10; in xnn_f32_igemm_minmax_ukernel_2x4__scalar()
|
D | 2x4-relu-scalar.c | 53 float vacc10 = vacc00; in xnn_f32_igemm_relu_ukernel_2x4__scalar() local 88 vacc10 += va1 * vb0; in xnn_f32_igemm_relu_ukernel_2x4__scalar() 102 vacc10 = math_max_f32(vacc10, 0.0f); in xnn_f32_igemm_relu_ukernel_2x4__scalar() 108 c1[0] = vacc10; in xnn_f32_igemm_relu_ukernel_2x4__scalar() 123 c1[0] = vacc10; in xnn_f32_igemm_relu_ukernel_2x4__scalar() 125 vacc10 = vacc12; in xnn_f32_igemm_relu_ukernel_2x4__scalar() 133 c1[0] = vacc10; in xnn_f32_igemm_relu_ukernel_2x4__scalar()
|
D | 2x4-relu-wasm.c | 53 float vacc10 = vacc00; in xnn_f32_igemm_relu_ukernel_2x4__wasm() local 88 vacc10 += va1 * vb0; in xnn_f32_igemm_relu_ukernel_2x4__wasm() 102 vacc10 = __builtin_wasm_max_f32(vacc10, 0.0f); in xnn_f32_igemm_relu_ukernel_2x4__wasm() 108 c1[0] = vacc10; in xnn_f32_igemm_relu_ukernel_2x4__wasm() 123 c1[0] = vacc10; in xnn_f32_igemm_relu_ukernel_2x4__wasm() 125 vacc10 = vacc12; in xnn_f32_igemm_relu_ukernel_2x4__wasm() 133 c1[0] = vacc10; in xnn_f32_igemm_relu_ukernel_2x4__wasm()
|
D | 4x2-minmax-wasm.c | 61 float vacc10 = vacc00; in xnn_f32_igemm_minmax_ukernel_4x2__wasm() local 106 vacc10 += va1 * vb0; in xnn_f32_igemm_minmax_ukernel_4x2__wasm() 120 vacc10 = __builtin_wasm_max_f32(vacc10, vmin); in xnn_f32_igemm_minmax_ukernel_4x2__wasm() 129 vacc10 = __builtin_wasm_min_f32(vacc10, vmax); in xnn_f32_igemm_minmax_ukernel_4x2__wasm() 143 c1[0] = vacc10; in xnn_f32_igemm_minmax_ukernel_4x2__wasm() 156 c1[0] = vacc10; in xnn_f32_igemm_minmax_ukernel_4x2__wasm()
|
D | 4x2-minmax-scalar.c | 61 float vacc10 = vacc00; in xnn_f32_igemm_minmax_ukernel_4x2__scalar() local 106 vacc10 += va1 * vb0; in xnn_f32_igemm_minmax_ukernel_4x2__scalar() 120 vacc10 = math_max_f32(vacc10, vmin); in xnn_f32_igemm_minmax_ukernel_4x2__scalar() 129 vacc10 = math_min_f32(vacc10, vmax); in xnn_f32_igemm_minmax_ukernel_4x2__scalar() 143 c1[0] = vacc10; in xnn_f32_igemm_minmax_ukernel_4x2__scalar() 156 c1[0] = vacc10; in xnn_f32_igemm_minmax_ukernel_4x2__scalar()
|
D | 4x4-minmax-scalar.c | 63 float vacc10 = vacc00; in xnn_f32_igemm_minmax_ukernel_4x4__scalar() local 118 vacc10 += va1 * vb0; in xnn_f32_igemm_minmax_ukernel_4x4__scalar() 140 vacc10 = math_max_f32(vacc10, vmin); in xnn_f32_igemm_minmax_ukernel_4x4__scalar() 157 vacc10 = math_min_f32(vacc10, vmax); in xnn_f32_igemm_minmax_ukernel_4x4__scalar() 181 c1[0] = vacc10; in xnn_f32_igemm_minmax_ukernel_4x4__scalar() 204 c1[0] = vacc10; in xnn_f32_igemm_minmax_ukernel_4x4__scalar() 206 vacc10 = vacc12; in xnn_f32_igemm_minmax_ukernel_4x4__scalar() 216 c1[0] = vacc10; in xnn_f32_igemm_minmax_ukernel_4x4__scalar()
|
D | 4x4-minmax-wasm.c | 63 float vacc10 = vacc00; in xnn_f32_igemm_minmax_ukernel_4x4__wasm() local 118 vacc10 += va1 * vb0; in xnn_f32_igemm_minmax_ukernel_4x4__wasm() 140 vacc10 = __builtin_wasm_max_f32(vacc10, vmin); in xnn_f32_igemm_minmax_ukernel_4x4__wasm() 157 vacc10 = __builtin_wasm_min_f32(vacc10, vmax); in xnn_f32_igemm_minmax_ukernel_4x4__wasm() 181 c1[0] = vacc10; in xnn_f32_igemm_minmax_ukernel_4x4__wasm() 204 c1[0] = vacc10; in xnn_f32_igemm_minmax_ukernel_4x4__wasm() 206 vacc10 = vacc12; in xnn_f32_igemm_minmax_ukernel_4x4__wasm() 216 c1[0] = vacc10; in xnn_f32_igemm_minmax_ukernel_4x4__wasm()
|
D | 2x4-wasm.c | 53 float vacc10 = vacc00; in xnn_f32_igemm_ukernel_2x4__wasm() local 88 vacc10 += va1 * vb0; in xnn_f32_igemm_ukernel_2x4__wasm() 100 c1[0] = vacc10; in xnn_f32_igemm_ukernel_2x4__wasm() 115 c1[0] = vacc10; in xnn_f32_igemm_ukernel_2x4__wasm() 117 vacc10 = vacc12; in xnn_f32_igemm_ukernel_2x4__wasm() 125 c1[0] = vacc10; in xnn_f32_igemm_ukernel_2x4__wasm()
|