| /external/XNNPACK/src/f32-gemm/gen-inc/ |
| D | 2x4-scalar.c | 54 float vacc11 = acc[5]; in xnn_f32_gemminc_ukernel_2x4__scalar() local
75 vacc11 += va1 * vb1; in xnn_f32_gemminc_ukernel_2x4__scalar()
88 vacc11 = math_max_f32(vacc11, vmin); in xnn_f32_gemminc_ukernel_2x4__scalar()
98 vacc11 = math_min_f32(vacc11, vmax); in xnn_f32_gemminc_ukernel_2x4__scalar()
104 c1[1] = vacc11; in xnn_f32_gemminc_ukernel_2x4__scalar()
121 c1[1] = vacc11; in xnn_f32_gemminc_ukernel_2x4__scalar()
|
| D | 2x4-wasm.c | 54 float vacc11 = acc[5]; in xnn_f32_gemminc_ukernel_2x4__wasm() local
75 vacc11 += va1 * vb1; in xnn_f32_gemminc_ukernel_2x4__wasm()
88 vacc11 = __builtin_wasm_max_f32(vacc11, vmin); in xnn_f32_gemminc_ukernel_2x4__wasm()
98 vacc11 = __builtin_wasm_min_f32(vacc11, vmax); in xnn_f32_gemminc_ukernel_2x4__wasm()
104 c1[1] = vacc11; in xnn_f32_gemminc_ukernel_2x4__wasm()
121 c1[1] = vacc11; in xnn_f32_gemminc_ukernel_2x4__wasm()
|
| D | 4x4-wasm.c | 66 float vacc11 = acc[5]; in xnn_f32_gemminc_ukernel_4x4__wasm() local
97 vacc11 += va1 * vb1; in xnn_f32_gemminc_ukernel_4x4__wasm()
118 vacc11 = __builtin_wasm_max_f32(vacc11, vmin); in xnn_f32_gemminc_ukernel_4x4__wasm()
136 vacc11 = __builtin_wasm_min_f32(vacc11, vmax); in xnn_f32_gemminc_ukernel_4x4__wasm()
160 c1[1] = vacc11; in xnn_f32_gemminc_ukernel_4x4__wasm()
187 c1[1] = vacc11; in xnn_f32_gemminc_ukernel_4x4__wasm()
|
| D | 4x4-scalar.c | 66 float vacc11 = acc[5]; in xnn_f32_gemminc_ukernel_4x4__scalar() local
97 vacc11 += va1 * vb1; in xnn_f32_gemminc_ukernel_4x4__scalar()
118 vacc11 = math_max_f32(vacc11, vmin); in xnn_f32_gemminc_ukernel_4x4__scalar()
136 vacc11 = math_min_f32(vacc11, vmax); in xnn_f32_gemminc_ukernel_4x4__scalar()
160 c1[1] = vacc11; in xnn_f32_gemminc_ukernel_4x4__scalar()
187 c1[1] = vacc11; in xnn_f32_gemminc_ukernel_4x4__scalar()
|
| /external/XNNPACK/src/f32-gemm/gen/ |
| D | 2x4-wasm.c | 53 float vacc11 = vacc01; in xnn_f32_gemm_ukernel_2x4__wasm() local
73 vacc11 += va1 * vb1; in xnn_f32_gemm_ukernel_2x4__wasm()
86 vacc11 = __builtin_wasm_max_f32(vacc11, vmin); in xnn_f32_gemm_ukernel_2x4__wasm()
96 vacc11 = __builtin_wasm_min_f32(vacc11, vmax); in xnn_f32_gemm_ukernel_2x4__wasm()
102 c1[1] = vacc11; in xnn_f32_gemm_ukernel_2x4__wasm()
119 c1[1] = vacc11; in xnn_f32_gemm_ukernel_2x4__wasm()
|
| D | 2x4-scalar.c | 53 float vacc11 = vacc01; in xnn_f32_gemm_ukernel_2x4__scalar() local
73 vacc11 += va1 * vb1; in xnn_f32_gemm_ukernel_2x4__scalar()
86 vacc11 = math_max_f32(vacc11, vmin); in xnn_f32_gemm_ukernel_2x4__scalar()
96 vacc11 = math_min_f32(vacc11, vmax); in xnn_f32_gemm_ukernel_2x4__scalar()
102 c1[1] = vacc11; in xnn_f32_gemm_ukernel_2x4__scalar()
119 c1[1] = vacc11; in xnn_f32_gemm_ukernel_2x4__scalar()
|
| D | 4x2-scalar.c | 63 float vacc11 = vacc01; in xnn_f32_gemm_ukernel_4x2__scalar() local
83 vacc11 += va1 * vb1; in xnn_f32_gemm_ukernel_4x2__scalar()
96 vacc11 = math_max_f32(vacc11, vmin); in xnn_f32_gemm_ukernel_4x2__scalar()
106 vacc11 = math_min_f32(vacc11, vmax); in xnn_f32_gemm_ukernel_4x2__scalar()
120 c1[1] = vacc11; in xnn_f32_gemm_ukernel_4x2__scalar()
|
| D | 4x2-wasm.c | 63 float vacc11 = vacc01; in xnn_f32_gemm_ukernel_4x2__wasm() local
83 vacc11 += va1 * vb1; in xnn_f32_gemm_ukernel_4x2__wasm()
96 vacc11 = __builtin_wasm_max_f32(vacc11, vmin); in xnn_f32_gemm_ukernel_4x2__wasm()
106 vacc11 = __builtin_wasm_min_f32(vacc11, vmax); in xnn_f32_gemm_ukernel_4x2__wasm()
120 c1[1] = vacc11; in xnn_f32_gemm_ukernel_4x2__wasm()
|
| D | 4x4-scalar.c | 65 float vacc11 = vacc01; in xnn_f32_gemm_ukernel_4x4__scalar() local
95 vacc11 += va1 * vb1; in xnn_f32_gemm_ukernel_4x4__scalar()
116 vacc11 = math_max_f32(vacc11, vmin); in xnn_f32_gemm_ukernel_4x4__scalar()
134 vacc11 = math_min_f32(vacc11, vmax); in xnn_f32_gemm_ukernel_4x4__scalar()
158 c1[1] = vacc11; in xnn_f32_gemm_ukernel_4x4__scalar()
185 c1[1] = vacc11; in xnn_f32_gemm_ukernel_4x4__scalar()
|
| D | 4x4-wasm.c | 65 float vacc11 = vacc01; in xnn_f32_gemm_ukernel_4x4__wasm() local
95 vacc11 += va1 * vb1; in xnn_f32_gemm_ukernel_4x4__wasm()
116 vacc11 = __builtin_wasm_max_f32(vacc11, vmin); in xnn_f32_gemm_ukernel_4x4__wasm()
134 vacc11 = __builtin_wasm_min_f32(vacc11, vmax); in xnn_f32_gemm_ukernel_4x4__wasm()
158 c1[1] = vacc11; in xnn_f32_gemm_ukernel_4x4__wasm()
185 c1[1] = vacc11; in xnn_f32_gemm_ukernel_4x4__wasm()
|
| /external/XNNPACK/src/f32-igemm/gen/ |
| D | 2x4-wasm.c | 54 float vacc11 = vacc01; in xnn_f32_igemm_ukernel_2x4__wasm() local
89 vacc11 += va1 * vb1; in xnn_f32_igemm_ukernel_2x4__wasm()
104 vacc11 = __builtin_wasm_max_f32(vacc11, vmin); in xnn_f32_igemm_ukernel_2x4__wasm()
114 vacc11 = __builtin_wasm_min_f32(vacc11, vmax); in xnn_f32_igemm_ukernel_2x4__wasm()
120 c1[1] = vacc11; in xnn_f32_igemm_ukernel_2x4__wasm()
135 c1[1] = vacc11; in xnn_f32_igemm_ukernel_2x4__wasm()
|
| D | 2x4-scalar.c | 54 float vacc11 = vacc01; in xnn_f32_igemm_ukernel_2x4__scalar() local
89 vacc11 += va1 * vb1; in xnn_f32_igemm_ukernel_2x4__scalar()
104 vacc11 = math_max_f32(vacc11, vmin); in xnn_f32_igemm_ukernel_2x4__scalar()
114 vacc11 = math_min_f32(vacc11, vmax); in xnn_f32_igemm_ukernel_2x4__scalar()
120 c1[1] = vacc11; in xnn_f32_igemm_ukernel_2x4__scalar()
135 c1[1] = vacc11; in xnn_f32_igemm_ukernel_2x4__scalar()
|
| D | 4x2-scalar.c | 60 float vacc11 = vacc01; in xnn_f32_igemm_ukernel_4x2__scalar() local
105 vacc11 += va1 * vb1; in xnn_f32_igemm_ukernel_4x2__scalar()
120 vacc11 = math_max_f32(vacc11, vmin); in xnn_f32_igemm_ukernel_4x2__scalar()
130 vacc11 = math_min_f32(vacc11, vmax); in xnn_f32_igemm_ukernel_4x2__scalar()
144 c1[1] = vacc11; in xnn_f32_igemm_ukernel_4x2__scalar()
|
| D | 4x2-wasm.c | 60 float vacc11 = vacc01; in xnn_f32_igemm_ukernel_4x2__wasm() local
105 vacc11 += va1 * vb1; in xnn_f32_igemm_ukernel_4x2__wasm()
120 vacc11 = __builtin_wasm_max_f32(vacc11, vmin); in xnn_f32_igemm_ukernel_4x2__wasm()
130 vacc11 = __builtin_wasm_min_f32(vacc11, vmax); in xnn_f32_igemm_ukernel_4x2__wasm()
144 c1[1] = vacc11; in xnn_f32_igemm_ukernel_4x2__wasm()
|
| D | 4x4-scalar.c | 62 float vacc11 = vacc01; in xnn_f32_igemm_ukernel_4x4__scalar() local
117 vacc11 += va1 * vb1; in xnn_f32_igemm_ukernel_4x4__scalar()
140 vacc11 = math_max_f32(vacc11, vmin); in xnn_f32_igemm_ukernel_4x4__scalar()
158 vacc11 = math_min_f32(vacc11, vmax); in xnn_f32_igemm_ukernel_4x4__scalar()
182 c1[1] = vacc11; in xnn_f32_igemm_ukernel_4x4__scalar()
205 c1[1] = vacc11; in xnn_f32_igemm_ukernel_4x4__scalar()
|
| D | 4x4-wasm.c | 62 float vacc11 = vacc01; in xnn_f32_igemm_ukernel_4x4__wasm() local
117 vacc11 += va1 * vb1; in xnn_f32_igemm_ukernel_4x4__wasm()
140 vacc11 = __builtin_wasm_max_f32(vacc11, vmin); in xnn_f32_igemm_ukernel_4x4__wasm()
158 vacc11 = __builtin_wasm_min_f32(vacc11, vmax); in xnn_f32_igemm_ukernel_4x4__wasm()
182 c1[1] = vacc11; in xnn_f32_igemm_ukernel_4x4__wasm()
205 c1[1] = vacc11; in xnn_f32_igemm_ukernel_4x4__wasm()
|
| /external/XNNPACK/src/q8-gemm/ |
| D | 2x4c8-sse2.c | 70 __m128i vacc11 = vacc01; in xnn_q8_gemm_ukernel_2x4c8__sse2() local
99 vacc11 = _mm_add_epi32(vacc11, _mm_madd_epi16(vxa1, vxb1)); in xnn_q8_gemm_ukernel_2x4c8__sse2()
105 __m128i vacc1x0123 = sse_reduce4_i32(vacc10, vacc11, vacc12, vacc13); in xnn_q8_gemm_ukernel_2x4c8__sse2()
|