Searched refs:va2x3o (Results 1 – 6 of 6) sorted by relevance
| /external/XNNPACK/src/bf16-gemm/gen/ |
| D | 3x4c8-minmax-neonfma-shland.c | 200 const float32x4_t va2x3o = vreinterpretq_f32_u16(vandq_u16(va2x3, vmask)); in xnn_bf16_gemm_minmax_ukernel_3x4c8__neonfma_shland() local
213 vacc2x3 = vfmaq_f32(vacc2x3, va2x3o, vb3o); in xnn_bf16_gemm_minmax_ukernel_3x4c8__neonfma_shland()
|
| D | 3x4c8-minmax-neonfma-zip.c | 200 const float32x4_t va2x3o = vreinterpretq_f32_u16(vzip2q_u16(vzero, va2x3)); in xnn_bf16_gemm_minmax_ukernel_3x4c8__neonfma_zip() local
213 vacc2x3 = vfmaq_f32(vacc2x3, va2x3o, vb3o); in xnn_bf16_gemm_minmax_ukernel_3x4c8__neonfma_zip()
|
| D | 4x4c8-minmax-neonfma-zip.c | 237 const float32x4_t va2x3o = vreinterpretq_f32_u16(vzip2q_u16(vzero, va2x3)); in xnn_bf16_gemm_minmax_ukernel_4x4c8__neonfma_zip() local
254 vacc2x3 = vfmaq_f32(vacc2x3, va2x3o, vb3o); in xnn_bf16_gemm_minmax_ukernel_4x4c8__neonfma_zip()
|
| D | 4x4c8-minmax-neonfma-shland.c | 237 const float32x4_t va2x3o = vreinterpretq_f32_u16(vandq_u16(va2x3, vmask)); in xnn_bf16_gemm_minmax_ukernel_4x4c8__neonfma_shland() local
254 vacc2x3 = vfmaq_f32(vacc2x3, va2x3o, vb3o); in xnn_bf16_gemm_minmax_ukernel_4x4c8__neonfma_shland()
|
| D | 5x4c8-minmax-neonfma-zip.c | 274 const float32x4_t va2x3o = vreinterpretq_f32_u16(vzip2q_u16(vzero, va2x3)); in xnn_bf16_gemm_minmax_ukernel_5x4c8__neonfma_zip() local
295 vacc2x3 = vfmaq_f32(vacc2x3, va2x3o, vb3o); in xnn_bf16_gemm_minmax_ukernel_5x4c8__neonfma_zip()
|
| D | 5x4c8-minmax-neonfma-shland.c | 274 const float32x4_t va2x3o = vreinterpretq_f32_u16(vandq_u16(va2x3, vmask)); in xnn_bf16_gemm_minmax_ukernel_5x4c8__neonfma_shland() local
295 vacc2x3 = vfmaq_f32(vacc2x3, va2x3o, vb3o); in xnn_bf16_gemm_minmax_ukernel_5x4c8__neonfma_shland()
|