Home
last modified time | relevance | path

Searched defs:vsum0xAB (Results 1 – 24 of 24) sorted by relevance

/external/XNNPACK/src/qs8-igemm/gen/
D1x16c8-minmax-neon-mull-padal.c139 const int32x4_t vsum0xAB = vpaddq_s32(vacc0x10, vacc0x11); in xnn_qs8_igemm_minmax_ukernel_1x16c8__neon_mull_padal() local
166 const int32x2_t vsum0xAB = vpadd_s32(vpsum0xA, vpsum0xB); in xnn_qs8_igemm_minmax_ukernel_1x16c8__neon_mull_padal() local
D1x16c16-minmax-neon-mlal-padal.c155 const int32x4_t vsum0xAB = vpaddq_s32(vacc0x10, vacc0x11); in xnn_qs8_igemm_minmax_ukernel_1x16c16__neon_mlal_padal() local
182 const int32x2_t vsum0xAB = vpadd_s32(vpsum0xA, vpsum0xB); in xnn_qs8_igemm_minmax_ukernel_1x16c16__neon_mlal_padal() local
D1x16c8-minmax-neon-mlal-padal.c228 const int32x4_t vsum0xAB = vpaddq_s32(vacc0x10, vacc0x11); in xnn_qs8_igemm_minmax_ukernel_1x16c8__neon_mlal_padal() local
255 const int32x2_t vsum0xAB = vpadd_s32(vpsum0xA, vpsum0xB); in xnn_qs8_igemm_minmax_ukernel_1x16c8__neon_mlal_padal() local
D2x16c8-minmax-neon-mull-padal.c196 const int32x4_t vsum0xAB = vpaddq_s32(vacc0x10, vacc0x11); in xnn_qs8_igemm_minmax_ukernel_2x16c8__neon_mull_padal() local
235 const int32x2_t vsum0xAB = vpadd_s32(vpsum0xA, vpsum0xB); in xnn_qs8_igemm_minmax_ukernel_2x16c8__neon_mull_padal() local
D2x16c16-minmax-neon-mlal-padal.c228 const int32x4_t vsum0xAB = vpaddq_s32(vacc0x10, vacc0x11); in xnn_qs8_igemm_minmax_ukernel_2x16c16__neon_mlal_padal() local
267 const int32x2_t vsum0xAB = vpadd_s32(vpsum0xA, vpsum0xB); in xnn_qs8_igemm_minmax_ukernel_2x16c16__neon_mlal_padal() local
D2x16c8-minmax-neon-mlal-padal.c335 const int32x4_t vsum0xAB = vpaddq_s32(vacc0x10, vacc0x11); in xnn_qs8_igemm_minmax_ukernel_2x16c8__neon_mlal_padal() local
374 const int32x2_t vsum0xAB = vpadd_s32(vpsum0xA, vpsum0xB); in xnn_qs8_igemm_minmax_ukernel_2x16c8__neon_mlal_padal() local
D3x16c8-minmax-neon-mull-padal.c253 const int32x4_t vsum0xAB = vpaddq_s32(vacc0x10, vacc0x11); in xnn_qs8_igemm_minmax_ukernel_3x16c8__neon_mull_padal() local
304 const int32x2_t vsum0xAB = vpadd_s32(vpsum0xA, vpsum0xB); in xnn_qs8_igemm_minmax_ukernel_3x16c8__neon_mull_padal() local
D3x16c16-minmax-neon-mlal-padal.c301 const int32x4_t vsum0xAB = vpaddq_s32(vacc0x10, vacc0x11); in xnn_qs8_igemm_minmax_ukernel_3x16c16__neon_mlal_padal() local
352 const int32x2_t vsum0xAB = vpadd_s32(vpsum0xA, vpsum0xB); in xnn_qs8_igemm_minmax_ukernel_3x16c16__neon_mlal_padal() local
D4x16c8-minmax-neon-mull-padal.c310 const int32x4_t vsum0xAB = vpaddq_s32(vacc0x10, vacc0x11); in xnn_qs8_igemm_minmax_ukernel_4x16c8__neon_mull_padal() local
373 const int32x2_t vsum0xAB = vpadd_s32(vpsum0xA, vpsum0xB); in xnn_qs8_igemm_minmax_ukernel_4x16c8__neon_mull_padal() local
D4x16c16-minmax-neon-mlal-padal.c374 const int32x4_t vsum0xAB = vpaddq_s32(vacc0x10, vacc0x11); in xnn_qs8_igemm_minmax_ukernel_4x16c16__neon_mlal_padal() local
437 const int32x2_t vsum0xAB = vpadd_s32(vpsum0xA, vpsum0xB); in xnn_qs8_igemm_minmax_ukernel_4x16c16__neon_mlal_padal() local
D3x16c8-minmax-neon-mlal-padal.c442 const int32x4_t vsum0xAB = vpaddq_s32(vacc0x10, vacc0x11); in xnn_qs8_igemm_minmax_ukernel_3x16c8__neon_mlal_padal() local
493 const int32x2_t vsum0xAB = vpadd_s32(vpsum0xA, vpsum0xB); in xnn_qs8_igemm_minmax_ukernel_3x16c8__neon_mlal_padal() local
D4x16c8-minmax-neon-mlal-padal.c549 const int32x4_t vsum0xAB = vpaddq_s32(vacc0x10, vacc0x11); in xnn_qs8_igemm_minmax_ukernel_4x16c8__neon_mlal_padal() local
612 const int32x2_t vsum0xAB = vpadd_s32(vpsum0xA, vpsum0xB); in xnn_qs8_igemm_minmax_ukernel_4x16c8__neon_mlal_padal() local
/external/XNNPACK/src/qs8-gemm/gen/
D1x16c16-minmax-neon-mlal-padal.c141 const int32x4_t vsum0xAB = vpaddq_s32(vacc0x10, vacc0x11); in xnn_qs8_gemm_minmax_ukernel_1x16c16__neon_mlal_padal() local
168 const int32x2_t vsum0xAB = vpadd_s32(vpsum0xA, vpsum0xB); in xnn_qs8_gemm_minmax_ukernel_1x16c16__neon_mlal_padal() local
D1x16c8-minmax-neon-mull-padal.c125 const int32x4_t vsum0xAB = vpaddq_s32(vacc0x10, vacc0x11); in xnn_qs8_gemm_minmax_ukernel_1x16c8__neon_mull_padal() local
152 const int32x2_t vsum0xAB = vpadd_s32(vpsum0xA, vpsum0xB); in xnn_qs8_gemm_minmax_ukernel_1x16c8__neon_mull_padal() local
D1x16c8-minmax-neon-mlal-padal.c214 const int32x4_t vsum0xAB = vpaddq_s32(vacc0x10, vacc0x11); in xnn_qs8_gemm_minmax_ukernel_1x16c8__neon_mlal_padal() local
241 const int32x2_t vsum0xAB = vpadd_s32(vpsum0xA, vpsum0xB); in xnn_qs8_gemm_minmax_ukernel_1x16c8__neon_mlal_padal() local
D2x16c16-minmax-neon-mlal-padal.c212 const int32x4_t vsum0xAB = vpaddq_s32(vacc0x10, vacc0x11); in xnn_qs8_gemm_minmax_ukernel_2x16c16__neon_mlal_padal() local
251 const int32x2_t vsum0xAB = vpadd_s32(vpsum0xA, vpsum0xB); in xnn_qs8_gemm_minmax_ukernel_2x16c16__neon_mlal_padal() local
D2x16c8-minmax-neon-mull-padal.c180 const int32x4_t vsum0xAB = vpaddq_s32(vacc0x10, vacc0x11); in xnn_qs8_gemm_minmax_ukernel_2x16c8__neon_mull_padal() local
219 const int32x2_t vsum0xAB = vpadd_s32(vpsum0xA, vpsum0xB); in xnn_qs8_gemm_minmax_ukernel_2x16c8__neon_mull_padal() local
D3x16c8-minmax-neon-mull-padal.c235 const int32x4_t vsum0xAB = vpaddq_s32(vacc0x10, vacc0x11); in xnn_qs8_gemm_minmax_ukernel_3x16c8__neon_mull_padal() local
286 const int32x2_t vsum0xAB = vpadd_s32(vpsum0xA, vpsum0xB); in xnn_qs8_gemm_minmax_ukernel_3x16c8__neon_mull_padal() local
D2x16c8-minmax-neon-mlal-padal.c319 const int32x4_t vsum0xAB = vpaddq_s32(vacc0x10, vacc0x11); in xnn_qs8_gemm_minmax_ukernel_2x16c8__neon_mlal_padal() local
358 const int32x2_t vsum0xAB = vpadd_s32(vpsum0xA, vpsum0xB); in xnn_qs8_gemm_minmax_ukernel_2x16c8__neon_mlal_padal() local
D3x16c16-minmax-neon-mlal-padal.c283 const int32x4_t vsum0xAB = vpaddq_s32(vacc0x10, vacc0x11); in xnn_qs8_gemm_minmax_ukernel_3x16c16__neon_mlal_padal() local
334 const int32x2_t vsum0xAB = vpadd_s32(vpsum0xA, vpsum0xB); in xnn_qs8_gemm_minmax_ukernel_3x16c16__neon_mlal_padal() local
D4x16c16-minmax-neon-mlal-padal.c354 const int32x4_t vsum0xAB = vpaddq_s32(vacc0x10, vacc0x11); in xnn_qs8_gemm_minmax_ukernel_4x16c16__neon_mlal_padal() local
417 const int32x2_t vsum0xAB = vpadd_s32(vpsum0xA, vpsum0xB); in xnn_qs8_gemm_minmax_ukernel_4x16c16__neon_mlal_padal() local
D4x16c8-minmax-neon-mull-padal.c290 const int32x4_t vsum0xAB = vpaddq_s32(vacc0x10, vacc0x11); in xnn_qs8_gemm_minmax_ukernel_4x16c8__neon_mull_padal() local
353 const int32x2_t vsum0xAB = vpadd_s32(vpsum0xA, vpsum0xB); in xnn_qs8_gemm_minmax_ukernel_4x16c8__neon_mull_padal() local
D3x16c8-minmax-neon-mlal-padal.c424 const int32x4_t vsum0xAB = vpaddq_s32(vacc0x10, vacc0x11); in xnn_qs8_gemm_minmax_ukernel_3x16c8__neon_mlal_padal() local
475 const int32x2_t vsum0xAB = vpadd_s32(vpsum0xA, vpsum0xB); in xnn_qs8_gemm_minmax_ukernel_3x16c8__neon_mlal_padal() local
D4x16c8-minmax-neon-mlal-padal.c529 const int32x4_t vsum0xAB = vpaddq_s32(vacc0x10, vacc0x11); in xnn_qs8_gemm_minmax_ukernel_4x16c8__neon_mlal_padal() local
592 const int32x2_t vsum0xAB = vpadd_s32(vpsum0xA, vpsum0xB); in xnn_qs8_gemm_minmax_ukernel_4x16c8__neon_mlal_padal() local