mirrors/gnss-sdr
1
0
Fork 0
mirror of https://github.com/gnss-sdr/gnss-sdr synced 2025-07-04 11:02:57 +00:00
Code Issues Releases Wiki Activity

Adding AVX protokernel

Browse Source
This commit is contained in:
Carles Fernandez 2016-04-23 16:31:41 +02:00
parent c6359c3313
commit b2cc18bfa5
1 changed files with 53 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

58
src/algorithms/libs/volk_gnsssdr_module/volk_gnsssdr/kernels/volk_gnsssdr/volk_gnsssdr_16ic_convert_32fc.h
View File

@ -96,7 +96,6 @@ static inline void volk_gnsssdr_16ic_convert_32fc_a_sse2(lv_32fc_t* outputVector
*_out++ = lv_cmake((float)lv_creal(*_in), (float)lv_cimag(*_in)); *_out++ = lv_cmake((float)lv_creal(*_in), (float)lv_cimag(*_in));
_in++; _in++;
} }
} }
#endif /* LV_HAVE_SSE2 */ #endif /* LV_HAVE_SSE2 */
@ -123,13 +122,62 @@ static inline void volk_gnsssdr_16ic_convert_32fc_u_sse2(lv_32fc_t* outputVector
*_out++ = lv_cmake((float)lv_creal(*_in), (float)lv_cimag(*_in)); *_out++ = lv_cmake((float)lv_creal(*_in), (float)lv_cimag(*_in));
_in++; _in++;
} }
} }
#endif /* LV_HAVE_SSE2 */ #endif /* LV_HAVE_SSE2 */
// SSE4.1
// a = _mm_load_si128((__m128i*)_in); //load (2 byte imag, 2 byte real) x 4 into 128 bits reg #ifdef LV_HAVE_AVX
//use _mm_cvtepi16_epi32 !!!! #include <immintrin.h>
static inline void volk_gnsssdr_16ic_convert_32fc_u_axv(lv_32fc_t* outputVector, const lv_16sc_t* inputVector, unsigned int num_points)
{
const unsigned int sse_iters = num_points / 4;
const lv_16sc_t* _in = inputVector;
lv_32fc_t* _out = outputVector;
__m256 a;
for(unsigned int number = 0; number < sse_iters; number++)
{
a = _mm256_set_ps((float)(lv_cimag(_in[3])), (float)(lv_creal(_in[3])), (float)(lv_cimag(_in[2])), (float)(lv_creal(_in[2])), (float)(lv_cimag(_in[1])), (float)(lv_creal(_in[1])), (float)(lv_cimag(_in[0])), (float)(lv_creal(_in[0]))); // //load (2 byte imag, 2 byte real) x 2 into 128 bits reg
_mm256_storeu_ps((float*)_out, a);
_in += 4;
_out += 4;
}
_mm256_zeroupper();
for (unsigned int i = 0; i < (num_points % 4); ++i)
{
*_out++ = lv_cmake((float)lv_creal(*_in), (float)lv_cimag(*_in));
_in++;
}
}
#endif /* LV_HAVE_AVX */
#ifdef LV_HAVE_AVX
#include <immintrin.h>
static inline void volk_gnsssdr_16ic_convert_32fc_a_axv(lv_32fc_t* outputVector, const lv_16sc_t* inputVector, unsigned int num_points)
{
const unsigned int sse_iters = num_points / 4;
const lv_16sc_t* _in = inputVector;
lv_32fc_t* _out = outputVector;
__m256 a;
for(unsigned int number = 0; number < sse_iters; number++)
{
a = _mm256_set_ps((float)(lv_cimag(_in[3])), (float)(lv_creal(_in[3])), (float)(lv_cimag(_in[2])), (float)(lv_creal(_in[2])), (float)(lv_cimag(_in[1])), (float)(lv_creal(_in[1])), (float)(lv_cimag(_in[0])), (float)(lv_creal(_in[0]))); // //load (2 byte imag, 2 byte real) x 2 into 128 bits reg
_mm256_store_ps((float*)_out, a);
_in += 4;
_out += 4;
}
_mm256_zeroupper();
for (unsigned int i = 0; i < (num_points % 4); ++i)
{
*_out++ = lv_cmake((float)lv_creal(*_in), (float)lv_cimag(*_in));
_in++;
}
}
#endif /* LV_HAVE_AVX */
#ifdef LV_HAVE_NEON #ifdef LV_HAVE_NEON
#include <arm_neon.h> #include <arm_neon.h>