1 // Auto-generated file. Do not edit! 2 // Template: src/f32-gemm/neon-shuffle.c.in 3 // Generator: tools/xngen 4 // 5 // Copyright 2019 Google LLC 6 // 7 // This source code is licensed under the BSD-style license found in the 8 // LICENSE file in the root directory of this source tree. 9 10 11 #include <assert.h> 12 13 #include <arm_neon.h> 14 15 #include <xnnpack/gemm.h> 16 17 xnn_f32_gemm_minmax_ukernel_4x8s4__neon(size_t mr,size_t nc,size_t kc,const float * restrict a,size_t a_stride,const float * restrict w,float * restrict c,size_t cm_stride,size_t cn_stride,const union xnn_f32_minmax_params params[restrict XNN_MIN_ELEMENTS (1)])18 void xnn_f32_gemm_minmax_ukernel_4x8s4__neon( 19 size_t mr, 20 size_t nc, 21 size_t kc, 22 const float* restrict a, 23 size_t a_stride, 24 const float* restrict w, 25 float* restrict c, 26 size_t cm_stride, 27 size_t cn_stride, 28 const union xnn_f32_minmax_params params[restrict XNN_MIN_ELEMENTS(1)]) 29 { 30 assert(mr != 0); 31 assert(mr <= 4); 32 assert(nc != 0); 33 assert(kc != 0); 34 assert(kc % sizeof(float) == 0); 35 assert(a != NULL); 36 assert(w != NULL); 37 assert(c != NULL); 38 39 const float* a0 = a; 40 float* c0 = c; 41 const float* a1 = (const float*) ((uintptr_t) a0 + a_stride); 42 float* c1 = (float*) ((uintptr_t) c0 + cm_stride); 43 if XNN_UNPREDICTABLE(mr < 2) { 44 a1 = a0; 45 c1 = c0; 46 } 47 const float* a2 = (const float*) ((uintptr_t) a1 + a_stride); 48 float* c2 = (float*) ((uintptr_t) c1 + cm_stride); 49 if XNN_UNPREDICTABLE(mr <= 2) { 50 a2 = a1; 51 c2 = c1; 52 } 53 const float* a3 = (const float*) ((uintptr_t) a2 + a_stride); 54 float* c3 = (float*) ((uintptr_t) c2 + cm_stride); 55 if XNN_UNPREDICTABLE(mr != 4) { 56 a3 = a2; 57 c3 = c2; 58 } 59 60 do { 61 float32x4_t vacc0x0123 = vld1q_f32(w); w += 4; 62 float32x4_t vacc0x4567 = vld1q_f32(w); w += 4; 63 float32x4_t vacc1x0123 = vacc0x0123; 64 float32x4_t vacc1x4567 = vacc0x4567; 65 float32x4_t vacc2x0123 = vacc0x0123; 66 float32x4_t vacc2x4567 = vacc0x4567; 67 float32x4_t vacc3x0123 = vacc0x0123; 68 float32x4_t vacc3x4567 = vacc0x4567; 69 70 size_t k = kc; 71 while (k >= 4 * sizeof(float)) { 72 float32x4_t va0 = vld1q_f32(a0); a0 += 4; 73 float32x4_t va1 = vld1q_f32(a1); a1 += 4; 74 float32x4_t va2 = vld1q_f32(a2); a2 += 4; 75 float32x4_t va3 = vld1q_f32(a3); a3 += 4; 76 77 78 const float32x4_t vb0123c0 = vld1q_f32(w + 0); 79 const float32x4_t vb4567c0 = vld1q_f32(w + 4); 80 81 vacc0x0123 = vmlaq_f32(vacc0x0123, va0, vb0123c0); 82 vacc1x0123 = vmlaq_f32(vacc1x0123, va1, vb0123c0); 83 vacc2x0123 = vmlaq_f32(vacc2x0123, va2, vb0123c0); 84 vacc3x0123 = vmlaq_f32(vacc3x0123, va3, vb0123c0); 85 vacc0x4567 = vmlaq_f32(vacc0x4567, va0, vb4567c0); 86 vacc1x4567 = vmlaq_f32(vacc1x4567, va1, vb4567c0); 87 vacc2x4567 = vmlaq_f32(vacc2x4567, va2, vb4567c0); 88 vacc3x4567 = vmlaq_f32(vacc3x4567, va3, vb4567c0); 89 90 va0 = vextq_f32(va0, va0, 1); 91 va1 = vextq_f32(va1, va1, 1); 92 va2 = vextq_f32(va2, va2, 1); 93 va3 = vextq_f32(va3, va3, 1); 94 95 const float32x4_t vb0123c1 = vld1q_f32(w + 8); 96 const float32x4_t vb4567c1 = vld1q_f32(w + 12); 97 98 vacc0x0123 = vmlaq_f32(vacc0x0123, va0, vb0123c1); 99 vacc1x0123 = vmlaq_f32(vacc1x0123, va1, vb0123c1); 100 vacc2x0123 = vmlaq_f32(vacc2x0123, va2, vb0123c1); 101 vacc3x0123 = vmlaq_f32(vacc3x0123, va3, vb0123c1); 102 vacc0x4567 = vmlaq_f32(vacc0x4567, va0, vb4567c1); 103 vacc1x4567 = vmlaq_f32(vacc1x4567, va1, vb4567c1); 104 vacc2x4567 = vmlaq_f32(vacc2x4567, va2, vb4567c1); 105 vacc3x4567 = vmlaq_f32(vacc3x4567, va3, vb4567c1); 106 107 va0 = vextq_f32(va0, va0, 1); 108 va1 = vextq_f32(va1, va1, 1); 109 va2 = vextq_f32(va2, va2, 1); 110 va3 = vextq_f32(va3, va3, 1); 111 112 const float32x4_t vb0123c2 = vld1q_f32(w + 16); 113 const float32x4_t vb4567c2 = vld1q_f32(w + 20); 114 115 vacc0x0123 = vmlaq_f32(vacc0x0123, va0, vb0123c2); 116 vacc1x0123 = vmlaq_f32(vacc1x0123, va1, vb0123c2); 117 vacc2x0123 = vmlaq_f32(vacc2x0123, va2, vb0123c2); 118 vacc3x0123 = vmlaq_f32(vacc3x0123, va3, vb0123c2); 119 vacc0x4567 = vmlaq_f32(vacc0x4567, va0, vb4567c2); 120 vacc1x4567 = vmlaq_f32(vacc1x4567, va1, vb4567c2); 121 vacc2x4567 = vmlaq_f32(vacc2x4567, va2, vb4567c2); 122 vacc3x4567 = vmlaq_f32(vacc3x4567, va3, vb4567c2); 123 124 va0 = vextq_f32(va0, va0, 1); 125 va1 = vextq_f32(va1, va1, 1); 126 va2 = vextq_f32(va2, va2, 1); 127 va3 = vextq_f32(va3, va3, 1); 128 129 const float32x4_t vb0123c3 = vld1q_f32(w + 24); 130 const float32x4_t vb4567c3 = vld1q_f32(w + 28); 131 132 vacc0x0123 = vmlaq_f32(vacc0x0123, va0, vb0123c3); 133 vacc1x0123 = vmlaq_f32(vacc1x0123, va1, vb0123c3); 134 vacc2x0123 = vmlaq_f32(vacc2x0123, va2, vb0123c3); 135 vacc3x0123 = vmlaq_f32(vacc3x0123, va3, vb0123c3); 136 vacc0x4567 = vmlaq_f32(vacc0x4567, va0, vb4567c3); 137 vacc1x4567 = vmlaq_f32(vacc1x4567, va1, vb4567c3); 138 vacc2x4567 = vmlaq_f32(vacc2x4567, va2, vb4567c3); 139 vacc3x4567 = vmlaq_f32(vacc3x4567, va3, vb4567c3); 140 141 142 w += 32; 143 k -= 4 * sizeof(float); 144 } 145 if XNN_UNLIKELY(k != 0) { 146 do { 147 const float32x4_t va0 = vld1q_dup_f32(a0); a0 += 1; 148 const float32x4_t va1 = vld1q_dup_f32(a1); a1 += 1; 149 const float32x4_t va2 = vld1q_dup_f32(a2); a2 += 1; 150 const float32x4_t va3 = vld1q_dup_f32(a3); a3 += 1; 151 152 const float32x4_t vb0123 = vld1q_f32(w); w += 4; 153 const float32x4_t vb4567 = vld1q_f32(w); w += 4; 154 155 vacc0x0123 = vmlaq_f32(vacc0x0123, va0, vb0123); 156 vacc1x0123 = vmlaq_f32(vacc1x0123, va1, vb0123); 157 vacc2x0123 = vmlaq_f32(vacc2x0123, va2, vb0123); 158 vacc3x0123 = vmlaq_f32(vacc3x0123, va3, vb0123); 159 vacc0x4567 = vmlaq_f32(vacc0x4567, va0, vb4567); 160 vacc1x4567 = vmlaq_f32(vacc1x4567, va1, vb4567); 161 vacc2x4567 = vmlaq_f32(vacc2x4567, va2, vb4567); 162 vacc3x4567 = vmlaq_f32(vacc3x4567, va3, vb4567); 163 164 k -= sizeof(float); 165 } while (k != 0); 166 } 167 const float32x4_t vmax = vld1q_dup_f32(¶ms->scalar.max); 168 vacc0x0123 = vminq_f32(vacc0x0123, vmax); 169 vacc1x0123 = vminq_f32(vacc1x0123, vmax); 170 vacc2x0123 = vminq_f32(vacc2x0123, vmax); 171 vacc3x0123 = vminq_f32(vacc3x0123, vmax); 172 vacc0x4567 = vminq_f32(vacc0x4567, vmax); 173 vacc1x4567 = vminq_f32(vacc1x4567, vmax); 174 vacc2x4567 = vminq_f32(vacc2x4567, vmax); 175 vacc3x4567 = vminq_f32(vacc3x4567, vmax); 176 177 const float32x4_t vmin = vld1q_dup_f32(¶ms->scalar.min); 178 vacc0x0123 = vmaxq_f32(vacc0x0123, vmin); 179 vacc1x0123 = vmaxq_f32(vacc1x0123, vmin); 180 vacc2x0123 = vmaxq_f32(vacc2x0123, vmin); 181 vacc3x0123 = vmaxq_f32(vacc3x0123, vmin); 182 vacc0x4567 = vmaxq_f32(vacc0x4567, vmin); 183 vacc1x4567 = vmaxq_f32(vacc1x4567, vmin); 184 vacc2x4567 = vmaxq_f32(vacc2x4567, vmin); 185 vacc3x4567 = vmaxq_f32(vacc3x4567, vmin); 186 187 if XNN_LIKELY(nc >= 8) { 188 vst1q_f32(c3, vacc3x0123); 189 vst1q_f32(c3 + 4, vacc3x4567); 190 c3 = (float*) ((uintptr_t) c3 + cn_stride); 191 vst1q_f32(c2, vacc2x0123); 192 vst1q_f32(c2 + 4, vacc2x4567); 193 c2 = (float*) ((uintptr_t) c2 + cn_stride); 194 vst1q_f32(c1, vacc1x0123); 195 vst1q_f32(c1 + 4, vacc1x4567); 196 c1 = (float*) ((uintptr_t) c1 + cn_stride); 197 vst1q_f32(c0, vacc0x0123); 198 vst1q_f32(c0 + 4, vacc0x4567); 199 c0 = (float*) ((uintptr_t) c0 + cn_stride); 200 201 a3 = (const float*) ((uintptr_t) a3 - kc); 202 a2 = (const float*) ((uintptr_t) a2 - kc); 203 a1 = (const float*) ((uintptr_t) a1 - kc); 204 a0 = (const float*) ((uintptr_t) a0 - kc); 205 206 nc -= 8; 207 208 } else { 209 if (nc & 4) { 210 vst1q_f32(c3, vacc3x0123); c3 += 4; 211 vst1q_f32(c2, vacc2x0123); c2 += 4; 212 vst1q_f32(c1, vacc1x0123); c1 += 4; 213 vst1q_f32(c0, vacc0x0123); c0 += 4; 214 215 vacc3x0123 = vacc3x4567; 216 vacc2x0123 = vacc2x4567; 217 vacc1x0123 = vacc1x4567; 218 vacc0x0123 = vacc0x4567; 219 } 220 float32x2_t vacc3x01 = vget_low_f32(vacc3x0123); 221 float32x2_t vacc2x01 = vget_low_f32(vacc2x0123); 222 float32x2_t vacc1x01 = vget_low_f32(vacc1x0123); 223 float32x2_t vacc0x01 = vget_low_f32(vacc0x0123); 224 if (nc & 2) { 225 vst1_f32(c3, vacc3x01); c3 += 2; 226 vst1_f32(c2, vacc2x01); c2 += 2; 227 vst1_f32(c1, vacc1x01); c1 += 2; 228 vst1_f32(c0, vacc0x01); c0 += 2; 229 230 vacc3x01 = vget_high_f32(vacc3x0123); 231 vacc2x01 = vget_high_f32(vacc2x0123); 232 vacc1x01 = vget_high_f32(vacc1x0123); 233 vacc0x01 = vget_high_f32(vacc0x0123); 234 } 235 if (nc & 1) { 236 vst1_lane_f32(c3, vacc3x01, 0); 237 vst1_lane_f32(c2, vacc2x01, 0); 238 vst1_lane_f32(c1, vacc1x01, 0); 239 vst1_lane_f32(c0, vacc0x01, 0); 240 } 241 242 nc = 0; 243 } 244 } while (nc != 0); 245 } 246