/external/XNNPACK/src/f32-gemm/gen/ |
D | 4x2-minmax-neon-lane-ld64.c | 94 vacc0x01 = vmla_f32(vacc0x01, va0, vb01); in xnn_f32_gemm_minmax_ukernel_4x2__neon_lane_ld64() 95 vacc1x01 = vmla_f32(vacc1x01, va1, vb01); in xnn_f32_gemm_minmax_ukernel_4x2__neon_lane_ld64() 96 vacc2x01 = vmla_f32(vacc2x01, va2, vb01); in xnn_f32_gemm_minmax_ukernel_4x2__neon_lane_ld64() 97 vacc3x01 = vmla_f32(vacc3x01, va3, vb01); in xnn_f32_gemm_minmax_ukernel_4x2__neon_lane_ld64()
|
/external/XNNPACK/src/f32-ibilinear-chw/gen/ |
D | neon-p4.c | 117 const float32x2_t vl = vmla_f32(vtl, vld, valphav); in xnn_f32_ibilinear_chw_ukernel__neon_p4() 118 const float32x2_t vr = vmla_f32(vtr, vrd, valphav); in xnn_f32_ibilinear_chw_ukernel__neon_p4() 121 const float32x2_t vo = vmla_f32(vl, vd, valphah); in xnn_f32_ibilinear_chw_ukernel__neon_p4() 155 const float32x2_t vlr = vmla_f32(vtltr, vldrd, valphav); in xnn_f32_ibilinear_chw_ukernel__neon_p4()
|
D | neon-p8.c | 205 const float32x2_t vl = vmla_f32(vtl, vld, valphav); in xnn_f32_ibilinear_chw_ukernel__neon_p8() 206 const float32x2_t vr = vmla_f32(vtr, vrd, valphav); in xnn_f32_ibilinear_chw_ukernel__neon_p8() 209 const float32x2_t vo = vmla_f32(vl, vd, valphah); in xnn_f32_ibilinear_chw_ukernel__neon_p8() 243 const float32x2_t vlr = vmla_f32(vtltr, vldrd, valphav); in xnn_f32_ibilinear_chw_ukernel__neon_p8()
|
/external/XNNPACK/src/f32-igemm/gen/ |
D | 4x2-minmax-neon-lane-ld64.c | 116 vacc0x01 = vmla_f32(vacc0x01, va0, vb01); in xnn_f32_igemm_minmax_ukernel_4x2__neon_lane_ld64() 117 vacc1x01 = vmla_f32(vacc1x01, va1, vb01); in xnn_f32_igemm_minmax_ukernel_4x2__neon_lane_ld64() 118 vacc2x01 = vmla_f32(vacc2x01, va2, vb01); in xnn_f32_igemm_minmax_ukernel_4x2__neon_lane_ld64() 119 vacc3x01 = vmla_f32(vacc3x01, va3, vb01); in xnn_f32_igemm_minmax_ukernel_4x2__neon_lane_ld64()
|
/external/XNNPACK/src/f32-spmm/gen/ |
D | 4x1-minmax-neon.c | 79 vacc01 = vmla_f32(vacc01, vi01, vw); in xnn_f32_spmm_minmax_ukernel_4x1__neon() 105 vacc0 = vmla_f32(vacc0, vi0, vw); in xnn_f32_spmm_minmax_ukernel_4x1__neon()
|
D | 4x1-minmax-neon-pipelined.c | 85 vacc01 = vmla_f32(vacc01, vi01, vb); in xnn_f32_spmm_minmax_ukernel_4x1__neon_pipelined() 113 vacc0 = vmla_f32(vacc0, vi0, vb); in xnn_f32_spmm_minmax_ukernel_4x1__neon_pipelined()
|
D | 4x1-minmax-neon-x2.c | 99 vacc01 = vmla_f32(vacc01, vi01, vw); in xnn_f32_spmm_minmax_ukernel_4x1__neon_x2() 125 vacc0 = vmla_f32(vacc0, vi0, vw); in xnn_f32_spmm_minmax_ukernel_4x1__neon_x2()
|
D | 8x1-minmax-neon.c | 111 vacc01 = vmla_f32(vacc01, vi01, vw); in xnn_f32_spmm_minmax_ukernel_8x1__neon() 137 vacc0 = vmla_f32(vacc0, vi0, vw); in xnn_f32_spmm_minmax_ukernel_8x1__neon()
|
D | 8x1-minmax-neon-pipelined.c | 120 vacc01 = vmla_f32(vacc01, vi01, vb); in xnn_f32_spmm_minmax_ukernel_8x1__neon_pipelined() 148 vacc0 = vmla_f32(vacc0, vi0, vb); in xnn_f32_spmm_minmax_ukernel_8x1__neon_pipelined()
|
D | 8x1-minmax-neon-x2.c | 138 vacc01 = vmla_f32(vacc01, vi01, vw); in xnn_f32_spmm_minmax_ukernel_8x1__neon_x2() 164 vacc0 = vmla_f32(vacc0, vi0, vw); in xnn_f32_spmm_minmax_ukernel_8x1__neon_x2()
|
D | 12x1-minmax-neon.c | 149 vacc01 = vmla_f32(vacc01, vi01, vw); in xnn_f32_spmm_minmax_ukernel_12x1__neon() 175 vacc0 = vmla_f32(vacc0, vi0, vw); in xnn_f32_spmm_minmax_ukernel_12x1__neon()
|
D | 16x1-minmax-neon.c | 155 vacc01 = vmla_f32(vacc01, vi01, vw); in xnn_f32_spmm_minmax_ukernel_16x1__neon() 181 vacc0 = vmla_f32(vacc0, vi0, vw); in xnn_f32_spmm_minmax_ukernel_16x1__neon()
|
D | 16x1-minmax-neon-pipelined.c | 168 vacc01 = vmla_f32(vacc01, vi01, vb); in xnn_f32_spmm_minmax_ukernel_16x1__neon_pipelined() 196 vacc0 = vmla_f32(vacc0, vi0, vb); in xnn_f32_spmm_minmax_ukernel_16x1__neon_pipelined()
|
D | 16x1-minmax-neon-x2.c | 196 vacc01 = vmla_f32(vacc01, vi01, vw); in xnn_f32_spmm_minmax_ukernel_16x1__neon_x2() 222 vacc0 = vmla_f32(vacc0, vi0, vw); in xnn_f32_spmm_minmax_ukernel_16x1__neon_x2()
|
D | 32x1-minmax-neon.c | 224 vacc01 = vmla_f32(vacc01, vi01, vw); in xnn_f32_spmm_minmax_ukernel_32x1__neon() 250 vacc0 = vmla_f32(vacc0, vi0, vw); in xnn_f32_spmm_minmax_ukernel_32x1__neon()
|
D | 32x1-minmax-neon-pipelined.c | 247 vacc01 = vmla_f32(vacc01, vi01, vb); in xnn_f32_spmm_minmax_ukernel_32x1__neon_pipelined() 276 vacc0 = vmla_f32(vacc0, vi0, vb); in xnn_f32_spmm_minmax_ukernel_32x1__neon_pipelined()
|
D | 32x1-minmax-neon-x2.c | 295 vacc01 = vmla_f32(vacc01, vi01, vw); in xnn_f32_spmm_minmax_ukernel_32x1__neon_x2() 321 vacc0 = vmla_f32(vacc0, vi0, vw); in xnn_f32_spmm_minmax_ukernel_32x1__neon_x2()
|
/external/XNNPACK/src/f32-igemm/ |
D | MRx2-neon-ld64.c.in | 104 vacc${M}x01 = vmla_f32(vacc${M}x01, va${M}, vb01);
|
/external/XNNPACK/src/f32-gemm/ |
D | MRx2-neon-ld64.c.in | 102 vacc${M}x01 = vmla_f32(vacc${M}x01, va${M}, vb01);
|
/external/XNNPACK/src/f32-spmm/ |
D | neon-pipelined.c.in | 8 $VMULADD_F32 = "vfma_f32" if FMA else "vmla_f32"
|
D | neon.c.in | 8 $VMULADD_F32 = "vfma_f32" if FMA else "vmla_f32"
|
/external/llvm-project/clang/test/CodeGen/ |
D | aarch64-neon-intrinsics-constrained.c | 110 return vmla_f32(v1, v2, v3); in test_vmla_f32()
|
/external/XNNPACK/src/f32-ibilinear-chw/ |
D | neon.c.in | 10 $VMULADD_F32 = "vfma_f32" if FMA else "vmla_f32"
|
/external/tensorflow/tensorflow/lite/kernels/internal/optimized/ |
D | depthwiseconv_float.h | 168 acc = vmla_f32(acc, input, filters); 728 acc = vmla_f32(acc, input, filter);
|
/external/tensorflow/tensorflow/core/kernels/neon/ |
D | depthwiseconv_float.h | 172 acc = vmla_f32(acc, input, filters);
|