DL Wiener: optimized with inlining

master
Xavier Arteaga 5 years ago committed by Xavier Arteaga
parent 9d0a3268e8
commit 928439b94c

@ -33,7 +33,6 @@
#define M_4_3 1.33333333333333333333f /* 4 / 3 */
#define M_5_3 1.66666666666666666666f /* 5 / 3 */
#define SRSLTE_WIENER_HALFREF_IDX (q->nof_ref / 2 - 1)
#define SRSLTE_WIENER_LOCAL
// Constants
const float hlsv_sum_norm[SRSLTE_WIENER_DL_MIN_RE] = {0.0625f,
@ -86,15 +85,15 @@ const float hlsv_sum_norm[SRSLTE_WIENER_DL_MIN_RE] = {0.0625f,
2.99999985900001};
// Local state function prototypes
SRSLTE_WIENER_LOCAL srslte_wiener_dl_state_t* srslte_wiener_dl_state_malloc(srslte_wiener_dl_t* q);
SRSLTE_WIENER_LOCAL void srslte_wiener_dl_state_free(srslte_wiener_dl_state_t* q);
SRSLTE_WIENER_LOCAL void srslte_wiener_dl_state_reset(srslte_wiener_dl_t* q, srslte_wiener_dl_state_t* state);
static srslte_wiener_dl_state_t* srslte_wiener_dl_state_malloc(srslte_wiener_dl_t* q);
static void srslte_wiener_dl_state_free(srslte_wiener_dl_state_t* q);
static void srslte_wiener_dl_state_reset(srslte_wiener_dl_t* q, srslte_wiener_dl_state_t* state);
// Local run function prototypes
SRSLTE_WIENER_LOCAL void
static void
srslte_wiener_dl_run_symbol_1_8(srslte_wiener_dl_t* q, srslte_wiener_dl_state_t* state, cf_t* pilots, float snr_lin);
SRSLTE_WIENER_LOCAL void srslte_wiener_dl_run_symbol_2_9(srslte_wiener_dl_t* q, srslte_wiener_dl_state_t* state);
SRSLTE_WIENER_LOCAL void srslte_wiener_dl_run_symbol_5_12(srslte_wiener_dl_t* q,
static void srslte_wiener_dl_run_symbol_2_9(srslte_wiener_dl_t* q, srslte_wiener_dl_state_t* state);
static void srslte_wiener_dl_run_symbol_5_12(srslte_wiener_dl_t* q,
srslte_wiener_dl_state_t* state,
cf_t* pilots,
uint32_t tx,
@ -103,7 +102,7 @@ SRSLTE_WIENER_LOCAL void srslte_wiener_dl_run_symbol_5_12(srslte_wiener_dl_t*
float snr_lin);
// Local state related functions
SRSLTE_WIENER_LOCAL srslte_wiener_dl_state_t* srslte_wiener_dl_state_malloc(srslte_wiener_dl_t* q)
static srslte_wiener_dl_state_t* srslte_wiener_dl_state_malloc(srslte_wiener_dl_t* q)
{
// Allocate Channel state
srslte_wiener_dl_state_t* state = calloc(sizeof(srslte_wiener_dl_state_t), 1);
@ -183,7 +182,7 @@ SRSLTE_WIENER_LOCAL srslte_wiener_dl_state_t* srslte_wiener_dl_state_malloc(srsl
return state;
}
SRSLTE_WIENER_LOCAL void srslte_wiener_dl_state_reset(srslte_wiener_dl_t* q, srslte_wiener_dl_state_t* state)
static void srslte_wiener_dl_state_reset(srslte_wiener_dl_t* q, srslte_wiener_dl_state_t* state)
{
if (q && state) {
// Initialise memory
@ -214,7 +213,7 @@ SRSLTE_WIENER_LOCAL void srslte_wiener_dl_state_reset(srslte_wiener_dl_t* q, srs
}
}
SRSLTE_WIENER_LOCAL void srslte_wiener_dl_state_free(srslte_wiener_dl_state_t* q)
static void srslte_wiener_dl_state_free(srslte_wiener_dl_state_t* q)
{
if (q) {
@ -372,7 +371,7 @@ void srslte_wiener_dl_reset(srslte_wiener_dl_t* q)
}
}
SRSLTE_WIENER_LOCAL void circshift_dim1(cf_t** matrix, uint32_t ndim1, int32_t k)
static void circshift_dim1(cf_t** matrix, uint32_t ndim1, int32_t k)
{
// Check valid inputs
if (matrix != NULL && ndim1 != 0 && k != 0) {
@ -397,7 +396,7 @@ SRSLTE_WIENER_LOCAL void circshift_dim1(cf_t** matrix, uint32_t ndim1, int32_t k
}
}
SRSLTE_WIENER_LOCAL void circshift_dim2(cf_t** matrix, uint32_t ndim1, uint32_t ndim2, int32_t k)
static void circshift_dim2(cf_t** matrix, uint32_t ndim1, uint32_t ndim2, int32_t k)
{
// Wrap k
k = (k + ndim2) % ndim2;
@ -419,7 +418,7 @@ SRSLTE_WIENER_LOCAL void circshift_dim2(cf_t** matrix, uint32_t ndim1, uint32_t
}
}
SRSLTE_WIENER_LOCAL void matrix_acc_dim1_cc(cf_t** matrix, cf_t* res, uint32_t ndim1, uint32_t ndim2)
static void matrix_acc_dim1_cc(cf_t** matrix, cf_t* res, uint32_t ndim1, uint32_t ndim2)
{
int dim2 = 0;
@ -452,7 +451,7 @@ SRSLTE_WIENER_LOCAL void matrix_acc_dim1_cc(cf_t** matrix, cf_t* res, uint32_t n
}
}*/
SRSLTE_WIENER_LOCAL uint32_t vec_find_first_smaller_than_cf(cf_t* x, float y, uint32_t n, uint32_t pos)
static inline uint32_t vec_find_first_smaller_than_cf(cf_t* x, float y, uint32_t n, uint32_t pos)
{
uint32_t ret = n;
@ -465,7 +464,7 @@ SRSLTE_WIENER_LOCAL uint32_t vec_find_first_smaller_than_cf(cf_t* x, float y, ui
return ret;
}
SRSLTE_WIENER_LOCAL cf_t _srslte_vec_dot_prod_ccc_simd(const cf_t* x, const cf_t* y, const int len)
static inline cf_t _srslte_vec_dot_prod_ccc_simd(const cf_t* x, const cf_t* y, const int len)
{
int i = 0;
cf_t result = 0;
@ -533,7 +532,17 @@ static void estimate_wiener(srslte_wiener_dl_t* q,
}
}
SRSLTE_WIENER_LOCAL void
inline static cf_t _cmul(cf_t a, cf_t b)
{
cf_t ret = 0;
__real__ ret = __real__ a * __real__ b - __imag__ a * __imag__ b;
__imag__ ret = __real__ a * __imag__ b + __imag__ a * __real__ b;
return ret;
}
static void
srslte_wiener_dl_run_symbol_1_8(srslte_wiener_dl_t* q, srslte_wiener_dl_state_t* state, cf_t* pilots, float snr_lin)
{
@ -561,7 +570,7 @@ srslte_wiener_dl_run_symbol_1_8(srslte_wiener_dl_t* q, srslte_wiener_dl_state_t*
state->skip = SRSLTE_MAX(1, floorf(halfcx / 4.0f * SRSLTE_MIN(1, snr_lin / 16.0f)));
}
SRSLTE_WIENER_LOCAL void srslte_wiener_dl_run_symbol_2_9(srslte_wiener_dl_t* q, srslte_wiener_dl_state_t* state)
static void srslte_wiener_dl_run_symbol_2_9(srslte_wiener_dl_t* q, srslte_wiener_dl_state_t* state)
{
// here we only shift and feed TD interpolation fifo
@ -579,7 +588,7 @@ SRSLTE_WIENER_LOCAL void srslte_wiener_dl_run_symbol_2_9(srslte_wiener_dl_t* q,
state->invtpilotoff = M_1_3;
}
SRSLTE_WIENER_LOCAL void srslte_wiener_dl_run_symbol_5_12(srslte_wiener_dl_t* q,
static void srslte_wiener_dl_run_symbol_5_12(srslte_wiener_dl_t* q,
srslte_wiener_dl_state_t* state,
cf_t* pilots,
uint32_t tx,
@ -728,13 +737,11 @@ SRSLTE_WIENER_LOCAL void srslte_wiener_dl_run_symbol_5_12(srslte_wiener_dl_t*
// Compute Wiener matrices
for (uint32_t dim1 = 0; dim1 < SRSLTE_WIENER_DL_MIN_RE; dim1++) {
for (uint32_t dim2 = 0; dim2 < SRSLTE_WIENER_DL_MIN_REF; dim2++) {
q->wm2[dim1][dim2] = 0;
for (int i = 0; i < SRSLTE_WIENER_DL_MIN_REF; i++) {
q->wm2[dim1][dim2] += q->hH2[dim1][i] * q->invRH.m[i][dim2];
}
q->wm1[dim1][dim2] = 0;
q->wm2[dim1][dim2] = 0;
for (int i = 0; i < SRSLTE_WIENER_DL_MIN_REF; i++) {
q->wm1[dim1][dim2] += q->hH1[dim1][i] * q->invRH.m[i][dim2];
q->wm1[dim1][dim2] += _cmul(q->hH1[dim1][i], q->invRH.m[i][dim2]);
q->wm2[dim1][dim2] += _cmul(q->hH2[dim1][i], q->invRH.m[i][dim2]);
}
}
}

Loading…
Cancel
Save