mirrors/gnss-sdr
1
0
Fork 0
mirror of https://github.com/gnss-sdr/gnss-sdr synced 2025-02-24 15:00:13 +00:00
Code Issues Releases Wiki Activity

kk

Browse Source
This commit is contained in:
Carles Fernandez 2016-02-11 22:58:56 +01:00
parent e66ac2f476
commit 787c28480b
1 changed files with 7 additions and 7 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

14
src/algorithms/libs/volk_gnsssdr_module/volk_gnsssdr/kernels/volk_gnsssdr/volk_gnsssdr_16ic_x2_rotator_dot_prod_16ic_xn.h
View File

@ -441,7 +441,7 @@ static inline void volk_gnsssdr_16ic_x2_rotator_dot_prod_16ic_xn_neon(lv_16sc_t*
int16x4x2_t* accumulator; int16x4x2_t* accumulator;
accumulator = (int16x4x2_t*)calloc(num_a_vectors, sizeof(int16x4x2_t)); accumulator = (int16x4x2_t*)calloc(num_a_vectors, sizeof(int16x4x2_t));
int16x4x2_t tmp_real, tmp_imag; int16x4x2_t tmp_real16, tmp_imag16;
for(int n_vec = 0; n_vec < num_a_vectors; n_vec++) for(int n_vec = 0; n_vec < num_a_vectors; n_vec++)
{ {
@ -511,18 +511,18 @@ static inline void volk_gnsssdr_16ic_x2_rotator_dot_prod_16ic_xn_neon(lv_16sc_t*
// multiply the real*real and imag*imag to get real result // multiply the real*real and imag*imag to get real result
// a0r*b0r|a1r*b1r|a2r*b2r|a3r*b3r // a0r*b0r|a1r*b1r|a2r*b2r|a3r*b3r
tmp_real.val[0] = vmul_s16(a_val.val[0], b_val.val[0]); tmp_real16.val[0] = vmul_s16(a_val.val[0], b_val.val[0]);
// a0i*b0i|a1i*b1i|a2i*b2i|a3i*b3i // a0i*b0i|a1i*b1i|a2i*b2i|a3i*b3i
tmp_real.val[1] = vmul_s16(a_val.val[1], b_val.val[1]); tmp_real16.val[1] = vmul_s16(a_val.val[1], b_val.val[1]);
// Multiply cross terms to get the imaginary result // Multiply cross terms to get the imaginary result
// a0r*b0i|a1r*b1i|a2r*b2i|a3r*b3i // a0r*b0i|a1r*b1i|a2r*b2i|a3r*b3i
tmp_imag.val[0] = vmul_s16(a_val.val[0], b_val.val[1]); tmp_imag16.val[0] = vmul_s16(a_val.val[0], b_val.val[1]);
// a0i*b0r|a1i*b1r|a2i*b2r|a3i*b3r // a0i*b0r|a1i*b1r|a2i*b2r|a3i*b3r
tmp_imag.val[1] = vmul_s16(a_val.val[1], b_val.val[0]); tmp_imag16.val[1] = vmul_s16(a_val.val[1], b_val.val[0]);
c_val.val[0] = vsub_s16(tmp_real.val[0], tmp_real.val[1]); c_val.val[0] = vsub_s16(tmp_real16.val[0], tmp_real16.val[1]);
c_val.val[1] = vadd_s16(tmp_imag.val[0], tmp_imag.val[1]); c_val.val[1] = vadd_s16(tmp_imag16.val[0], tmp_imag16.val[1]);
accumulator[n_vec].val[0] = vadd_s16(accumulator[n_vec].val[0], c_val.val[0]); accumulator[n_vec].val[0] = vadd_s16(accumulator[n_vec].val[0], c_val.val[0]);
accumulator[n_vec].val[1] = vadd_s16(accumulator[n_vec].val[1], c_val.val[1]); accumulator[n_vec].val[1] = vadd_s16(accumulator[n_vec].val[1], c_val.val[1]);