/external/XNNPACK/src/qs8-gemm/gen/ |
D | 3x8c4s2-minmax-rndnu-neon-mlal.c | 165 int16x8_t vprod2x67c1 = vmull_s8(vb67c1x0, va2x0); in xnn_qs8_gemm_minmax_rndnu_ukernel_3x8c4s2__neon_mlal() local 237 int16x8_t vprod2x67c1 = vmull_s8(vb67c1x0, va2x0); in xnn_qs8_gemm_minmax_rndnu_ukernel_3x8c4s2__neon_mlal() local
|
D | 4x8c4s2-minmax-rndnu-neon-mlal.c | 200 int16x8_t vprod2x67c1 = vmull_s8(vb67c1x0, va2x0); in xnn_qs8_gemm_minmax_rndnu_ukernel_4x8c4s2__neon_mlal() local 291 int16x8_t vprod2x67c1 = vmull_s8(vb67c1x0, va2x0); in xnn_qs8_gemm_minmax_rndnu_ukernel_4x8c4s2__neon_mlal() local
|
D | 3x8c4-minmax-rndnu-neon-mlal-ld2r.c | 174 int16x8_t vprod2x67c1 = vmull_s8(vb67c1x0, va2c1x0); in xnn_qs8_gemm_minmax_rndnu_ukernel_3x8c4__neon_mlal_ld2r() local 252 const int16x8_t vprod2x67c1 = vmull_s8(vb67c1, va2c1); in xnn_qs8_gemm_minmax_rndnu_ukernel_3x8c4__neon_mlal_ld2r() local
|
D | 3x8c4-minmax-rndnu-neon-mlal-dup.c | 174 int16x8_t vprod2x67c1 = vmull_s8(vb67c1x0, va2c1x0); in xnn_qs8_gemm_minmax_rndnu_ukernel_3x8c4__neon_mlal_dup() local 252 const int16x8_t vprod2x67c1 = vmull_s8(vb67c1, va2c1); in xnn_qs8_gemm_minmax_rndnu_ukernel_3x8c4__neon_mlal_dup() local
|
D | 3x8c4-minmax-rndnu-neon-mlal-ld1r.c | 180 int16x8_t vprod2x67c1 = vmull_s8(vb67c1x0, va2c1x0); in xnn_qs8_gemm_minmax_rndnu_ukernel_3x8c4__neon_mlal_ld1r() local 261 const int16x8_t vprod2x67c1 = vmull_s8(vb67c1, va2c1); in xnn_qs8_gemm_minmax_rndnu_ukernel_3x8c4__neon_mlal_ld1r() local
|
D | 4x8c4-minmax-rndnu-neon-mlal-ld2r.c | 211 int16x8_t vprod2x67c1 = vmull_s8(vb67c1x0, va2c1x0); in xnn_qs8_gemm_minmax_rndnu_ukernel_4x8c4__neon_mlal_ld2r() local 309 const int16x8_t vprod2x67c1 = vmull_s8(vb67c1, va2c1); in xnn_qs8_gemm_minmax_rndnu_ukernel_4x8c4__neon_mlal_ld2r() local
|
D | 4x8c4-minmax-rndnu-neon-mlal-dup.c | 211 int16x8_t vprod2x67c1 = vmull_s8(vb67c1x0, va2c1x0); in xnn_qs8_gemm_minmax_rndnu_ukernel_4x8c4__neon_mlal_dup() local 309 const int16x8_t vprod2x67c1 = vmull_s8(vb67c1, va2c1); in xnn_qs8_gemm_minmax_rndnu_ukernel_4x8c4__neon_mlal_dup() local
|
D | 3x8c4s2-minmax-rndnu-neon-mull.c | 131 int16x8_t vprod2x67c1 = vmull_s8(vb67c1x0, va2x0); in xnn_qs8_gemm_minmax_rndnu_ukernel_3x8c4s2__neon_mull() local
|
D | 4x8c4-minmax-rndnu-neon-mlal-ld1r.c | 219 int16x8_t vprod2x67c1 = vmull_s8(vb67c1x0, va2c1x0); in xnn_qs8_gemm_minmax_rndnu_ukernel_4x8c4__neon_mlal_ld1r() local 321 const int16x8_t vprod2x67c1 = vmull_s8(vb67c1, va2c1); in xnn_qs8_gemm_minmax_rndnu_ukernel_4x8c4__neon_mlal_ld1r() local
|
D | 3x16c4s2-minmax-rndnu-neon-mlal.c | 225 int16x8_t vprod2x67c1 = vmull_s8(vb67c1x0, va2x0); in xnn_qs8_gemm_minmax_rndnu_ukernel_3x16c4s2__neon_mlal() local 369 int16x8_t vprod2x67c1 = vmull_s8(vb67c1x0, va2x0); in xnn_qs8_gemm_minmax_rndnu_ukernel_3x16c4s2__neon_mlal() local
|
D | 3x8c4-minmax-rndnu-neon-mull-dup.c | 138 const int16x8_t vprod2x67c1 = vmull_s8(vb67c1, va2c1); in xnn_qs8_gemm_minmax_rndnu_ukernel_3x8c4__neon_mull_dup() local
|
D | 3x8c4-minmax-rndnu-neon-mull-ld2r.c | 138 const int16x8_t vprod2x67c1 = vmull_s8(vb67c1, va2c1); in xnn_qs8_gemm_minmax_rndnu_ukernel_3x8c4__neon_mull_ld2r() local
|
D | 4x8c4s2-minmax-rndnu-neon-mull.c | 157 int16x8_t vprod2x67c1 = vmull_s8(vb67c1x0, va2x0); in xnn_qs8_gemm_minmax_rndnu_ukernel_4x8c4s2__neon_mull() local
|
D | 3x8c4-minmax-rndnu-neon-mull-ld1r.c | 141 const int16x8_t vprod2x67c1 = vmull_s8(vb67c1, va2c1); in xnn_qs8_gemm_minmax_rndnu_ukernel_3x8c4__neon_mull_ld1r() local
|
/external/XNNPACK/src/qs8-igemm/gen/ |
D | 3x8c4s2-minmax-rndnu-neon-mlal.c | 180 int16x8_t vprod2x67c1 = vmull_s8(vb67c1x0, va2x0); in xnn_qs8_igemm_minmax_rndnu_ukernel_3x8c4s2__neon_mlal() local 252 int16x8_t vprod2x67c1 = vmull_s8(vb67c1x0, va2x0); in xnn_qs8_igemm_minmax_rndnu_ukernel_3x8c4s2__neon_mlal() local
|
D | 4x8c4s2-minmax-rndnu-neon-mlal.c | 217 int16x8_t vprod2x67c1 = vmull_s8(vb67c1x0, va2x0); in xnn_qs8_igemm_minmax_rndnu_ukernel_4x8c4s2__neon_mlal() local 308 int16x8_t vprod2x67c1 = vmull_s8(vb67c1x0, va2x0); in xnn_qs8_igemm_minmax_rndnu_ukernel_4x8c4s2__neon_mlal() local
|
D | 3x8c4-minmax-rndnu-neon-mlal-dup.c | 189 int16x8_t vprod2x67c1 = vmull_s8(vb67c1x0, va2c1x0); in xnn_qs8_igemm_minmax_rndnu_ukernel_3x8c4__neon_mlal_dup() local 267 const int16x8_t vprod2x67c1 = vmull_s8(vb67c1, va2c1); in xnn_qs8_igemm_minmax_rndnu_ukernel_3x8c4__neon_mlal_dup() local
|
D | 3x8c4-minmax-rndnu-neon-mlal-ld2r.c | 189 int16x8_t vprod2x67c1 = vmull_s8(vb67c1x0, va2c1x0); in xnn_qs8_igemm_minmax_rndnu_ukernel_3x8c4__neon_mlal_ld2r() local 267 const int16x8_t vprod2x67c1 = vmull_s8(vb67c1, va2c1); in xnn_qs8_igemm_minmax_rndnu_ukernel_3x8c4__neon_mlal_ld2r() local
|
D | 3x8c4-minmax-rndnu-neon-mlal-ld1r.c | 195 int16x8_t vprod2x67c1 = vmull_s8(vb67c1x0, va2c1x0); in xnn_qs8_igemm_minmax_rndnu_ukernel_3x8c4__neon_mlal_ld1r() local 276 const int16x8_t vprod2x67c1 = vmull_s8(vb67c1, va2c1); in xnn_qs8_igemm_minmax_rndnu_ukernel_3x8c4__neon_mlal_ld1r() local
|
D | 4x8c4-minmax-rndnu-neon-mlal-dup.c | 228 int16x8_t vprod2x67c1 = vmull_s8(vb67c1x0, va2c1x0); in xnn_qs8_igemm_minmax_rndnu_ukernel_4x8c4__neon_mlal_dup() local 326 const int16x8_t vprod2x67c1 = vmull_s8(vb67c1, va2c1); in xnn_qs8_igemm_minmax_rndnu_ukernel_4x8c4__neon_mlal_dup() local
|
D | 4x8c4-minmax-rndnu-neon-mlal-ld2r.c | 228 int16x8_t vprod2x67c1 = vmull_s8(vb67c1x0, va2c1x0); in xnn_qs8_igemm_minmax_rndnu_ukernel_4x8c4__neon_mlal_ld2r() local 326 const int16x8_t vprod2x67c1 = vmull_s8(vb67c1, va2c1); in xnn_qs8_igemm_minmax_rndnu_ukernel_4x8c4__neon_mlal_ld2r() local
|
D | 3x8c4s2-minmax-rndnu-neon-mull.c | 146 int16x8_t vprod2x67c1 = vmull_s8(vb67c1x0, va2x0); in xnn_qs8_igemm_minmax_rndnu_ukernel_3x8c4s2__neon_mull() local
|
D | 4x8c4-minmax-rndnu-neon-mlal-ld1r.c | 236 int16x8_t vprod2x67c1 = vmull_s8(vb67c1x0, va2c1x0); in xnn_qs8_igemm_minmax_rndnu_ukernel_4x8c4__neon_mlal_ld1r() local 338 const int16x8_t vprod2x67c1 = vmull_s8(vb67c1, va2c1); in xnn_qs8_igemm_minmax_rndnu_ukernel_4x8c4__neon_mlal_ld1r() local
|
D | 3x16c4s2-minmax-rndnu-neon-mlal.c | 240 int16x8_t vprod2x67c1 = vmull_s8(vb67c1x0, va2x0); in xnn_qs8_igemm_minmax_rndnu_ukernel_3x16c4s2__neon_mlal() local 384 int16x8_t vprod2x67c1 = vmull_s8(vb67c1x0, va2x0); in xnn_qs8_igemm_minmax_rndnu_ukernel_3x16c4s2__neon_mlal() local
|
D | 4x8c4s2-minmax-rndnu-neon-mull.c | 174 int16x8_t vprod2x67c1 = vmull_s8(vb67c1x0, va2x0); in xnn_qs8_igemm_minmax_rndnu_ukernel_4x8c4s2__neon_mull() local
|