Skip to content

O
OpenXG-RAN
Commit 02f1ec69 authored by Luis Ariza's avatar Luis Ariza
Browse files
Options

SSE Ziggurat version improved

parent fc657e21
Hide whitespace changes
Inline Side-by-side
Showing with 391 additions and 123 deletions
openair1/SCHED/phy_procedures_lte_ue.c
View file @ 02f1ec69
......@@ -2896,7 +2896,7 @@ void ue_pbch_procedures(uint8_t eNB_id,PHY_VARS_UE *ue,UE_rxtx_proc_t *proc, uin
int pbch_tx_ant=0;
uint8_t pbch_phase;
uint16_t frame_tx;
static uint8_t first_run = 1;
static uint8_t first_run = 0;
uint8_t pbch_trials = 0;
DevAssert(ue);
......@@ -2985,8 +2985,8 @@ void ue_pbch_procedures(uint8_t eNB_id,PHY_VARS_UE *ue,UE_rxtx_proc_t *proc, uin
//emos_dump_UE.mimo_mode = ue->pbch_vars[eNB_id]->decoded_output[1];
#endif
if (first_run) {
first_run = 0;
if (first_run<=ue->Mod_id) {
first_run++;
proc->frame_rx = (proc->frame_rx & 0xFFFFFC00) | (frame_tx & 0x000003FF);
proc->frame_tx = proc->frame_rx;
......@@ -5300,7 +5300,6 @@ else
if ((ue->dlsch_SI[eNB_id]) && (ue->dlsch_SI[eNB_id]->active == 1)) {
VCD_SIGNAL_DUMPER_DUMP_FUNCTION_BY_NAME(VCD_SIGNAL_DUMPER_FUNCTIONS_PDSCH_PROC_SI, VCD_FUNCTION_IN);
ue_pdsch_procedures(ue,
proc,
eNB_id,
SI_PDSCH,
......@@ -5410,6 +5409,7 @@ else
if ( (subframe_rx == 0) && (ue->decode_MIB == 1))
{
//printf("[ue_pbch_procedures] subframe_rx %d ,ue->decode_MIB %d, UE %d\n",subframe_rx,ue->decode_MIB,ue->Mod_id);
ue_pbch_procedures(eNB_id,ue,proc,abstraction_flag);
}
......
openair1/SIMULATION/RF/rf.c
View file @ 02f1ec69
......@@ -303,9 +303,9 @@ clock_t start=clock();*/
rx128_re = _mm_loadu_pd(&r_re[a][2*i]);//r_re[a][i],r_re[a][i+1]
rx128_im = _mm_loadu_pd(&r_im[a][2*i]);//r_im[a][i],r_im[a][i+1]
rx128_gain_lin = _mm_set1_pd(rx_gain_lin);
//gauss_0_128_sqrt_NOW = _mm_set_pd(ziggurat(0.0,1.0),ziggurat(0.0,1.0));
//gauss_1_128_sqrt_NOW = _mm_set_pd(ziggurat(0.0,1.0),ziggurat(0.0,1.0));
boxmuller_SSE_float(&gauss_0_128_sqrt_NOW, &gauss_1_128_sqrt_NOW);
gauss_0_128_sqrt_NOW = _mm_set_pd(ziggurat(0.0,1.0),ziggurat(0.0,1.0));
gauss_1_128_sqrt_NOW = _mm_set_pd(ziggurat(0.0,1.0),ziggurat(0.0,1.0));
//boxmuller_SSE_float(&gauss_0_128_sqrt_NOW, &gauss_1_128_sqrt_NOW);
gauss_0_128_sqrt_NOW = _mm_mul_pd(gauss_0_128_sqrt_NOW,_mm_set1_pd(sqrt_NOW));
gauss_1_128_sqrt_NOW = _mm_mul_pd(gauss_1_128_sqrt_NOW,_mm_set1_pd(sqrt_NOW));
// Amplify by receiver gain and apply 3rd order non-linearity
......@@ -435,7 +435,6 @@ clock_t start=clock();*/
//boxmuller_SSE_float(&gauss_0_128_sqrt_NOW, &gauss_1_128_sqrt_NOW);
gauss_0_128_sqrt_NOW = ziggurat_SSE_float();
gauss_1_128_sqrt_NOW = ziggurat_SSE_float();
//stop_meas(&desc->ziggurat);
gauss_0_128_sqrt_NOW = _mm_mul_ps(gauss_0_128_sqrt_NOW,_mm_set1_ps(sqrt_NOW));
gauss_1_128_sqrt_NOW = _mm_mul_ps(gauss_1_128_sqrt_NOW,_mm_set1_ps(sqrt_NOW));
......
openair1/SIMULATION/TOOLS/rangen_double.c
View file @ 02f1ec69
......@@ -128,156 +128,119 @@ double nfix(void)
static uint32_t jsr4[4] __attribute__((aligned(16))) = {123456789,112548569,985584512,452236879};//This initialization depends on the seed for nor_table function in oaisim_functions.c file.
static uint32_t iz4[4] __attribute__((aligned(16)));
static float out[4] __attribute__((aligned(16)));
static uint32_t ssh3_sse4[4] __attribute__((aligned(16)));
static int32_t ifabs4[4] __attribute__((aligned(16)));
static uint32_t iz1[4] __attribute__((aligned(16)));
static uint32_t iz2[4] __attribute__((aligned(16)));
//static float out[4] __attribute__((aligned(16)));
//static int32_t ifabs4[4] __attribute__((aligned(16)));
static int32_t hz4[4] __attribute__((aligned(16)));
static int32_t abshz4[4] __attribute__((aligned(16)));
static float x4_option0[4] __attribute__((aligned(16)));
static float x4[4] __attribute__((aligned(16)));
static int32_t hz1[4] __attribute__((aligned(16)));
static int32_t hz2[4] __attribute__((aligned(16)));
static int32_t abshz[4] __attribute__((aligned(16)));
static int32_t abshz1[4] __attribute__((aligned(16)));
static int32_t abshz2[4] __attribute__((aligned(16)));
static __m128i jsr_128 __attribute__((aligned(16)));
static __m128i jz_128 __attribute__((aligned(16)));
static __m128i hz_128 __attribute__((aligned(16)));
static __m128i hz1_128 __attribute__((aligned(16)));
static __m128i hz2_128 __attribute__((aligned(16)));
static __m128i abs_hz_128 __attribute__((aligned(16)));
static __m128i abs_hz1_128 __attribute__((aligned(16)));
static __m128i abs_hz2_128 __attribute__((aligned(16)));
static __m128i iz_128 __attribute__((aligned(16)));
static __m128 x128 __attribute__((aligned(16)));
static __m128i ifabs __attribute__((aligned(16)));
static __m128i iz1_128 __attribute__((aligned(16)));
static __m128i iz2_128 __attribute__((aligned(16)));
static __m128i cmplt_option0_128 __attribute__((aligned(16)));
static int count99=0;
static int count0=0;
static int nfix_first_run=0;
static __m128 x __attribute__((aligned(16)));
#define SHR3_SSE (jsr_128=_mm_loadu_si128((__m128i *)jsr4),jz_128=jsr_128, jsr_128=_mm_xor_si128(_mm_slli_epi32(jsr_128,13),jsr_128),jsr_128=_mm_xor_si128(_mm_srli_epi32(jsr_128,17),jsr_128),jsr_128=_mm_xor_si128(_mm_slli_epi32(jsr_128,5),jsr_128),_mm_storeu_si128((__m128i *)jsr4,jsr_128),_mm_add_epi32(jz_128,jsr_128))
#define UNI_SSE (_mm_add_ps(_mm_mul_ps(_mm_set1_ps(0.2328306e-9),_mm_cvtepi32_ps(SHR3_SSE)),_mm_set1_ps(0.5)))
#define NOR_SSE (hz_128=SHR3_SSE,_mm_storeu_si128((__m128i *)hz4,hz_128),iz_128=_mm_and_si128(hz_128,_mm_set1_epi32(127)),_mm_storeu_si128((__m128i *)iz4,iz_128),abs_hz_128=_mm_and_si128(hz_128, _mm_set1_epi32(~0x80000000)),_mm_storeu_si128((__m128i *)abshz4,abs_hz_128))
#define NOR_SSE (hz_128=SHR3_SSE,_mm_storeu_si128((__m128i *)hz4,hz_128),iz_128=_mm_and_si128(hz_128,_mm_set1_epi32(127)),_mm_storeu_si128((__m128i *)iz4,iz_128),abs_hz_128=_mm_and_si128(hz_128, _mm_set1_epi32(~0x80000000)),cmplt_option0_128 = _mm_cmplt_epi32(abs_hz_128,_mm_setr_epi32(kn[iz4[0]],kn[iz4[1]],kn[iz4[2]],kn[iz4[3]])),count99=(count99>99)?0:count99+4,nfix_first_run=(count99>99)?0:1,(_mm_testc_si128(cmplt_option0_128,_mm_setr_epi32(0xFFFFFFFF,0xFFFFFFFF,0xFFFFFFFF,0xFFFFFFFF)))?x=_mm_mul_ps(_mm_cvtepi32_ps(hz_128),_mm_setr_ps(wn[iz4[0]],wn[iz4[1]],wn[iz4[2]],wn[iz4[3]])):nfix_SSE())
//,ifabs=_mm_cmplt_epi32(_mm_max_epi32(_mm_sub_epi32(_mm_setzero_si128(),hz_128),hz_128),_mm_setr_epi32(kn[iz4[0]],kn[iz4[1]],kn[iz4[2]],kn[iz4[3]])),_mm_storeu_si128((__m128i *)ifabs4,ifabs),abs_hz_128=_mm_and_si128(hz_128, _mm_set1_epi32(~0x80000000)),_mm_storeu_si128((__m128i *)abshz4,abs_hz_128),printf("abs_hz_128 %d,%d,%d,%d\n",abshz4[0],abshz4[1],abshz4[2],abshz4[3]),printf("kn %d,%d,%d,%d\n",kn[iz4[0]],kn[iz4[1]],kn[iz4[2]],kn[iz4[3]]),printf("ifabs %x,%x,%x,%x\n",ifabs4[0],ifabs4[1],ifabs4[2],ifabs4[3]),x128=_mm_and_ps(_mm_cvtepi32_ps(_mm_cmplt_epi32(_mm_max_epi32(_mm_sub_epi32(_mm_setzero_si128(),hz_128),hz_128),_mm_setr_epi32(kn[iz4[0]],kn[iz4[1]],kn[iz4[2]],kn[iz4[3]]))),_mm_mul_ps(_mm_cvtepi32_ps(hz_128),_mm_setr_ps(wn[iz4[0]],wn[iz4[1]],wn[iz4[2]],wn[iz4[3]]))),printf("x128 %e,%e,%e,%e\n",x128[0],x128[1],x128[2],x128[3]),printf("iz %d,%d,%d,%d\n",iz4[0],iz4[1],iz4[2],iz4[3]),printf("wn*hz %e,%e,%e,%e\n",hz4[0]*wn[iz4[0]],hz4[1]*wn[iz4[1]],hz4[2]*wn[iz4[2]],hz4[3]*wn[iz4[3]]))
//#define NOR1_SSE (hz1_128=SHR3_SSE,_mm_storeu_si128((__m128i *)hz1,hz1_128),iz1_128=_mm_and_si128(hz1_128,_mm_set1_epi32(127)),_mm_storeu_si128((__m128i *)iz1,iz1_128),abs_hz1_128=_mm_and_si128(hz1_128, _mm_set1_epi32(~0x80000000)),_mm_storeu_si128((__m128i *)abshz1,abs_hz1_128))
//,_mm_storeu_si128(ssh3_sse4,hz_128),printf("ssh3_sse4 %lu,%lu,%lu,%lu\n",ssh3_sse4[0],ssh3_sse4[1],ssh3_sse4[2],ssh3_sse4[3])
//#define NOR (hz=SHR3, printf("hz %d\n",hz),sign=(hz&128)>>7,printf("sign %s\n",(sign)?"-":"+"),iz=hz&127,printf("iz %d\n",iz), (abs(hz)<kn[iz])? (sign)?(-1)*hz*wn[iz]:hz*wn[iz] : (sign)?(-1)*nfix():nfix())
//#define NOR2_SSE (hz2_128=SHR3_SSE,_mm_storeu_si128((__m128i *)hz2,hz2_128),iz2_128=_mm_and_si128(hz2_128,_mm_set1_epi32(127)),_mm_storeu_si128((__m128i *)iz2,iz2_128),abs_hz2_128=_mm_and_si128(hz2_128, _mm_set1_epi32(~0x80000000)),_mm_storeu_si128((__m128i *)abshz2,abs_hz2_128))
//#define NOR_SSE (hz_128=SHR3_SSE,_mm_storeu_si128((__m128i *)hz4,hz_128),iz_128=_mm_and_si128(hz_128,_mm_set1_epi32(127)),_mm_storeu_si128((__m128i *)iz4,iz_128),abs_hz_128=_mm_and_si128(hz_128, _mm_set1_epi32(~0x80000000)),_mm_storeu_si128((__m128i *)abshz4,abs_hz_128),cmplt_option0_128 = _mm_cmplt_epi32(abs_hz_128,_mm_setr_epi32(kn[iz4[0]],kn[iz4[1]],kn[iz4[2]],kn[iz4[3]])),_mm_storeu_si128((__m128i *)cmplt_option0,cmplt_option0_128),count0=0,(cmplt_option0[0]==0xFFFFFFFF)?count99+=count0++:count0,(cmplt_option0[1]==0xFFFFFFFF)?count99+=count0++:count0,(cmplt_option0[2]==0xFFFFFFFF)?count99+=count0++:count0,(cmplt_option0[3]==0xFFFFFFFF)?count99+=count0++:count0,(cmplt_option0[0]==0xFFFFFFFF && cmplt_option0[1]==0xFFFFFFFF && cmplt_option0[2]==0xFFFFFFFF && cmplt_option0[3]==0xFFFFFFFF && count99<95 && count0==4)?_mm_mul_ps(_mm_cvtepi32_ps(hz_128),_mm_setr_ps(wn[iz4[0]],wn[iz4[1]],wn[iz4[2]],wn[iz4[3]])):nfix_SSE())
static __m128 x __attribute__((aligned(16)));
static __m128 y __attribute__((aligned(16)));
static __m128i cmplt_option0_128 __attribute__((aligned(16)));
static __m128i cmplt_option1_128 __attribute__((aligned(16)));
static __m128i cmplt_option2_128 __attribute__((aligned(16)));
static int32_t cmplt_option0[4] __attribute__((aligned(16)));
static int32_t cmplt_option1[4] __attribute__((aligned(16)));
static int32_t cmplt_option2[4] __attribute__((aligned(16)));
static float output0[4] __attribute__((aligned(16)));
static float output1[4] __attribute__((aligned(16)));
static float output2[4] __attribute__((aligned(16)));
static float output[12] __attribute__((aligned(16)));
static int option=-1;
__m128 option012(void)
{
int i;
for (i=0;i<4;i++)
{
if (abshz4[i]<kn[iz4[i]])
{
//,ifabs=_mm_cmplt_epi32(_mm_max_epi32(_mm_sub_epi32(_mm_setzero_si128(),hz_128),hz_128),_mm_setr_epi32(kn[iz4[0]],kn[iz4[1]],kn[iz4[2]],kn[iz4[3]])),_mm_storeu_si128((__m128i *)ifabs4,ifabs),abs_hz_128=_mm_and_si128(hz_128, _mm_set1_epi32(~0x80000000)),_mm_storeu_si128((__m128i *)abshz4,abs_hz_128),printf("abs_hz_128 %d,%d,%d,%d\n",abshz4[0],abshz4[1],abshz4[2],abshz4[3]),printf("kn %d,%d,%d,%d\n",kn[iz4[0]],kn[iz4[1]],kn[iz4[2]],kn[iz4[3]]),printf("ifabs %x,%x,%x,%x\n",ifabs4[0],ifabs4[1],ifabs4[2],ifabs4[3]),x128=_mm_and_ps(_mm_cvtepi32_ps(_mm_cmplt_epi32(_mm_max_epi32(_mm_sub_epi32(_mm_setzero_si128(),hz_128),hz_128),_mm_setr_epi32(kn[iz4[0]],kn[iz4[1]],kn[iz4[2]],kn[iz4[3]]))),_mm_mul_ps(_mm_cvtepi32_ps(hz_128),_mm_setr_ps(wn[iz4[0]],wn[iz4[1]],wn[iz4[2]],wn[iz4[3]]))),printf("x128 %e,%e,%e,%e\n",x128[0],x128[1],x128[2],x128[3]),printf("iz %d,%d,%d,%d\n",iz4[0],iz4[1],iz4[2],iz4[3]),printf("wn*hz %e,%e,%e,%e\n",hz4[0]*wn[iz4[0]],hz4[1]*wn[iz4[1]],hz4[2]*wn[iz4[2]],hz4[3]*wn[iz4[3]]))
}
}
}
//,_mm_storeu_si128(ssh3_sse4,hz_128),printf("ssh3_sse4 %lu,%lu,%lu,%lu\n",ssh3_sse4[0],ssh3_sse4[1],ssh3_sse4[2],ssh3_sse4[3])
//#define NOR (hz=SHR3, printf("hz %d\n",hz),sign=(hz&128)>>7,printf("sign %s\n",(sign)?"-":"+"),iz=hz&127,printf("iz %d\n",iz), (abs(hz)<kn[iz])? (sign)?(-1)*hz*wn[iz]:hz*wn[iz] : (sign)?(-1)*nfix():nfix())
__m128 nfix_SSE(void)
{
static int count0=0;
static int count1=0;
static int count2=0;
static int count=0;
static int rand0=0;
static int rand1=0;
static int rand2=0;
static int rand3=0;
__m128 y __attribute__((aligned(16)));
__m128i cmplt_option1_128 __attribute__((aligned(16)));
__m128i cmplt_option2_128 __attribute__((aligned(16)));
int32_t cmplt_option0[4] __attribute__((aligned(16)));
int32_t cmplt_option1[4] __attribute__((aligned(16)));
int32_t cmplt_option2[4] __attribute__((aligned(16)));
float output[12] __attribute__((aligned(16)));
float x4_option0[4] __attribute__((aligned(16)));
float x4[4] __attribute__((aligned(16)));
int i;
static float r = 3.442620;
for (;;)
{
if (count0+count1+count2>3)
{
return _mm_setr_ps(output[rand0],output[rand1],output[rand2],output[rand3]);
}
uint32_t iz4_i[4] __attribute__((aligned(16))) ;
_mm_storeu_si128((__m128i *)iz4_i,iz_128);
NOR_SSE;
//(abs(hz)<kn[iz])? hz*wn[iz]
cmplt_option0_128 = _mm_cmplt_epi32(abs_hz_128,_mm_setr_epi32(kn[iz4[0]],kn[iz4[1]],kn[iz4[2]],kn[iz4[3]]));
_mm_storeu_si128((__m128i *)cmplt_option0,cmplt_option0_128);
//x=hz * wn[iz];
x=_mm_mul_ps(_mm_cvtepi32_ps(hz_128),_mm_setr_ps(wn[iz4[0]],wn[iz4[1]],wn[iz4[2]],wn[iz4[3]]));
_mm_storeu_si128((__m128i *)cmplt_option0,cmplt_option0_128);
_mm_storeu_ps(x4_option0,x);
count0=0;
for (i=0;i<4;i++)
{
if (cmplt_option0[i]==0xFFFFFFFF)
{
//printf("count0 %d\n",count0);
output[count0]=hz4[i]*wn[iz4[i]];
output[count0]=hz4[i]*wn[iz4_i[i]];
count0++;
}
}
// if (fn[iz]+UNI*(fn[iz-1]-fn[iz])<exp(-0.5*x*x))
cmplt_option2_128 = _mm_cvtps_epi32(_mm_cmplt_ps(_mm_add_ps(_mm_setr_ps(fn[iz4[0]],fn[iz4[1]],fn[iz4[2]],fn[iz4[3]]),_mm_mul_ps(UNI_SSE,_mm_sub_ps(_mm_setr_ps(fn[iz4[0]-1],fn[iz4[1]-1],fn[iz4[2]-1],fn[iz4[3]-1]),_mm_setr_ps(fn[iz4[0]],fn[iz4[1]],fn[iz4[2]],fn[iz4[3]])))),exp_ps(_mm_mul_ps(_mm_mul_ps(x,x),_mm_set1_ps(-0.5f)))));
_mm_storeu_si128((__m128i *)cmplt_option2,cmplt_option2_128);
for (i=0;i<4;i++)
if ((iz4_i[0]==0||iz4_i[1]==0||iz4_i[2]==0||iz4_i[3]==0)&&nfix_first_run==0&&count0>0)
{
if (cmplt_option2[i]==0x80000000)
{
//printf("count1 %d\n",count1);
output[count0+count1]=x4_option0[i];
count1++;
}
nfix_first_run=1;
do
{
//x = - 0.2904764 * log (UNI);
x = _mm_mul_ps(_mm_set1_ps(-0.2904764f), log_ps(UNI_SSE));
_mm_storeu_ps(x4,x);
//y = - log (UNI);
y = _mm_mul_ps(_mm_set1_ps(-1.0f), log_ps(UNI_SSE));
//(y+y < x*x)?
cmplt_option1_128 = _mm_cvtps_epi32(_mm_cmplt_ps(_mm_add_ps(y,y),_mm_mul_ps(x,x)));
_mm_storeu_si128((__m128i *)cmplt_option1,cmplt_option1_128);
for (i=0;i<4;i++)
{
if (cmplt_option1[i]==0x80000000)
{
output[3]=(hz4[i]>0)? x4[i]+r:-x4[i]-r;
break;
}
}
}
while (cmplt_option1[0]!=0x80000000 && cmplt_option1[1]!=0x80000000 && cmplt_option1[2]!=0x80000000 && cmplt_option1[3]!=0x80000000);
return _mm_setr_ps(output[0],output[1],output[2],output[3]);
}
/*if (iz==0)
{
do
{
x = - 0.2904764 * log (UNI);
y = - log (UNI);
}
while (y+y < x*x);
return (hz>0)? r+x : -r-x;
}*/
//if (iz==0)
if (iz4[0]==0 ||iz4[1]==0 ||iz4[2]==0 ||iz4[3]==0)
else if (iz4[0]>0&&iz4[1]>0&&iz4[2]>0&&iz4[3]>0&&nfix_first_run==0&&count0>0)
{
do
nfix_first_run=1;
cmplt_option2_128 = _mm_cvtps_epi32(_mm_cmplt_ps(_mm_add_ps(_mm_setr_ps(fn[iz4_i[0]],fn[iz4_i[1]],fn[iz4_i[2]],fn[iz4_i[3]]),_mm_mul_ps(UNI_SSE,_mm_sub_ps(_mm_setr_ps(fn[iz4_i[0]-1],fn[iz4_i[1]-1],fn[iz4_i[2]-1],fn[iz4_i[3]-1]),_mm_setr_ps(fn[iz4_i[0]],fn[iz4_i[1]],fn[iz4_i[2]],fn[iz4_i[3]])))),exp_ps(_mm_mul_ps(_mm_mul_ps(x,x),_mm_set1_ps(-0.5f)))));
_mm_storeu_si128((__m128i *)cmplt_option2,cmplt_option2_128);
for (i=0;i<4;i++)
{
//x = - 0.2904764 * log (UNI);
x = _mm_mul_ps(_mm_set1_ps(-0.2904764f), log_ps(UNI_SSE));
_mm_storeu_ps(x4,x);
//y = - log (UNI);
y = _mm_mul_ps(_mm_set1_ps(-1.0f), log_ps(UNI_SSE));
//(y+y < x*x)?
cmplt_option1_128 = _mm_cvtps_epi32(_mm_cmplt_ps(_mm_add_ps(y,y),_mm_mul_ps(x,x)));
_mm_storeu_si128((__m128i *)cmplt_option1,cmplt_option1_128);
for (i=0;i<4;i++)
{
if (cmplt_option1[i]==0x80000000)
{
printf("count2 %d\n",count2);
output[count0+count1+count2]=(hz4[i]>0)? x4[i]+r:-x4[i]-r;
count2++;
}
}
if (cmplt_option2[i]==0x80000000)
{
output[3]=x4_option0[i];
break;
}
}
while (cmplt_option1[0]!=0x80000000 && cmplt_option1[1]!=0x80000000 && cmplt_option1[2]!=0x80000000 && cmplt_option1[3]!=0x80000000);
}
count=count0+count1+count2;
rand0=iz4[0]%count;
rand1=iz4[1]%count;
rand2=iz4[2]%count;
rand3=iz4[3]%count;
}
return _mm_setr_ps(output[0],output[1],output[2],output[3]);
}
}
/*!\Procedure to create tables for normal distribution kn,wn and fn. */
void table_nor(unsigned long seed)
{
......@@ -316,7 +279,7 @@ double ziggurat(double mean, double variance)
}
__m128 ziggurat_SSE_float(void)
{
return nfix_SSE();
return NOR_SSE;
}
void boxmuller_SSE_float(__m128 *data1, __m128 *data2) {
......@@ -516,3 +479,308 @@ void randominit(unsigned seed_init)
}
}*/
/*__m128 nfix1_SSE(void)
{
__m128 x1 __attribute__((aligned(16)));
__m128 y1 __attribute__((aligned(16)));
__m128i cmplt_option0_128 __attribute__((aligned(16)));
__m128i cmplt_option1_128 __attribute__((aligned(16)));
__m128i cmplt_option2_128 __attribute__((aligned(16)));
int32_t cmplt_option0[4] __attribute__((aligned(16)));
int32_t cmplt_option1[4] __attribute__((aligned(16)));
int32_t cmplt_option2[4] __attribute__((aligned(16)));
float output1[12] __attribute__((aligned(16)));
float x1_option0[4] __attribute__((aligned(16)));
float x4[4] __attribute__((aligned(16)));
int count0=0;
int count1=0;
int count2=0;
int i;
static float r = 3.442620;
static int nfix_first_run=0;
for (;;)
{
NOR1_SSE;
//(abs(hz)<kn[iz])? hz*wn[iz]
cmplt_option0_128 = _mm_cmplt_epi32(abs_hz1_128,_mm_setr_epi32(kn[iz1[0]],kn[iz1[1]],kn[iz1[2]],kn[iz1[3]]));
_mm_storeu_si128((__m128i *)cmplt_option0,cmplt_option0_128);
//x=hz * wn[iz];
for (i=0;i<4;i++)
{
if (cmplt_option0[i]==0xFFFFFFFF)
{
//printf("count0 %d\n",count0);
output1[count0]=hz1[i]*wn[iz1[i]];
count0++;
}
}
if (count0>3)
{
count99+=4;
if (count99>99)
{
count99=0;
nfix_first_run=0;
}
return _mm_setr_ps(output1[0],output1[1],output1[2],output1[3]);
}
//x=hz * wn[iz];
x1=_mm_mul_ps(_mm_cvtepi32_ps(hz1_128),_mm_setr_ps(wn[iz1[0]],wn[iz1[1]],wn[iz1[2]],wn[iz1[3]]));
_mm_storeu_ps(x1_option0,x1);
//printf("count0 is %d, count1 is %d, count2 is %d,count99 is %d\n",count0,count1,count2,count99);
if ((iz1[0]==0||iz1[1]==0||iz1[2]==0||iz1[3]==0)&&nfix_first_run==0&&count0>0)
{
//printf("\niz == 0 [%d,%d,%d,%d]\n\n",iz4[0],iz4[1],iz4[2],iz4[3]);
nfix_first_run=1;
do
{
//x = - 0.2904764 * log (UNI);
x1 = _mm_mul_ps(_mm_set1_ps(-0.2904764f), log_ps(UNI_SSE));
_mm_storeu_ps(x4,x1);
//y = - log (UNI);
y1 = _mm_mul_ps(_mm_set1_ps(-1.0f), log_ps(UNI_SSE));
//(y+y < x*x)?
cmplt_option1_128 = _mm_cvtps_epi32(_mm_cmplt_ps(_mm_add_ps(y1,y1),_mm_mul_ps(x1,x1)));
_mm_storeu_si128((__m128i *)cmplt_option1,cmplt_option1_128);
for (i=0;i<4;i++)
{
if (cmplt_option1[i]==0x80000000)
{
//printf("count22 %d\n",count2);
output1[3]=(hz1[i]>0)? x4[i]+r:-x4[i]-r;
count2++;
break;
}
}
}
while (cmplt_option1[0]!=0x80000000 || cmplt_option1[1]!=0x80000000 || cmplt_option1[2]!=0x80000000 || cmplt_option1[3]!=0x80000000);
if (count0+count2>3)
{
count99+=4;
if (count99>99)
{
count99=0;
nfix_first_run=0;
}
return _mm_setr_ps(output1[0],output1[1],output1[2],output1[3]);
}
}
if (iz1[0]>0&&iz1[1]>0&&iz1[2]>0&&iz1[3]>0&&nfix_first_run==0&&count0>0)
{
//printf("\niz > 0 [%d,%d,%d,%d]\n\n",iz4[0],iz4[1],iz4[2],iz4[3]);
nfix_first_run=1;
printf("\niz1 > 0 [%d,%d,%d,%d].\nfn [%e,%e,%e,%e].\n\n",iz1[0],iz1[1],iz1[2],iz1[3],fn[iz1[0]],fn[iz1[1]],fn[iz1[2]],fn[iz1[3]]);
printf("fn1 - 1 [%e,%e,%e,%e]\n",fn[iz1[0]-1],fn[iz1[1]-1],fn[iz1[2]-1],fn[iz1[3]-1]);
//if (iz==0)
printf("\niz [%d,%d,%d,%d]\n",iz4[0],iz4[1],iz4[2],iz4[3]);
printf("iz==0 [%d,%d,%d,%d]\n",iz4[0]==0,iz4[1]==0,iz4[2]==0,iz4[3]==0);
printf("iz>0 [%d,%d,%d,%d]\n\n",iz4[0]>0,iz4[1]>0,iz4[2]>0,iz4[3]>0);//
// if (fn[iz]+UNI*(fn[iz-1]-fn[iz])<exp(-0.5*x*x))
//printf("iz [%d,%d,%d,%d] is ok? %d\n",iz4[0],iz4[1],iz4[2],iz4[3],iz4[0]==0&&iz4[1]==0&&iz4[2]==0&&iz4[3]==0);
//printf("iz>0 inside [%d,%d,%d,%d]\n",iz4[0]>0,iz4[1]>0,iz4[2]>0,iz4[3]>0);
//printf("iz-1 [%d,%d,%d,%d]\n",iz4[0]-1,iz4[1]-1,iz4[2]-1,iz4[3]-1);
//printf("x [%e,%e,%e,%e]\n",x[0],x[1],x[2],x[3]);
//printf("exp [%e,%e,%e,%e]\n",exp(-0.5*x[0]*x[0]),exp(-0.5*x[1]*x[1]),exp(-0.5*x[2]*x[2]),exp(-0.5*x[3]*x[3]));//
cmplt_option2_128 = _mm_cvtps_epi32(_mm_cmplt_ps(_mm_add_ps(_mm_setr_ps(fn[iz1[0]],fn[iz1[1]],fn[iz1[2]],fn[iz1[3]]),_mm_mul_ps(UNI_SSE,_mm_sub_ps(_mm_setr_ps(fn[iz1[0]-1],fn[iz1[1]-1],fn[iz1[2]-1],fn[iz1[3]-1]),_mm_setr_ps(fn[iz1[0]],fn[iz1[1]],fn[iz1[2]],fn[iz1[3]])))),exp_ps(_mm_mul_ps(_mm_mul_ps(x1,x1),_mm_set1_ps(-0.5f)))));
//cmplt_option2_128 = _mm_cvtps_epi32(_mm_cmplt_ps(_mm_setr_ps(fn[iz4[0]],fn[iz4[1]],fn[iz4[2]],fn[iz4[3]]),exp_ps(_mm_mul_ps(_mm_mul_ps(x,x),_mm_set1_ps(-0.5f)))));
_mm_storeu_si128((__m128i *)cmplt_option2,cmplt_option2_128);
for (i=0;i<4;i++)
{
if (cmplt_option2[i]==0x80000000)
{
//printf("count1 %d\n",count1);
output1[3]=x1_option0[i];
count1++;
break;
}
}
if (count0+count1>3)
{
count99+=4;
if (count99>109)
{
count99=0;
nfix_first_run=0;
}
return _mm_setr_ps(output1[0],output1[1],output1[2],output1[3]);
}
}
NOR1_SSE;
//(abs(hz)<kn[iz])? hz*wn[iz]
cmplt_option0_128 = _mm_cmplt_epi32(abs_hz1_128,_mm_setr_epi32(kn[iz1[0]],kn[iz1[1]],kn[iz1[2]],kn[iz1[3]]));
_mm_storeu_si128((__m128i *)cmplt_option0,cmplt_option0_128);
for (i=count0;i<3;i++)
{
if (cmplt_option0[i-count0]==0xFFFFFFFF)
{
//printf("count0 %d\n",count0);
output1[count0]=hz1[i-count0]*wn[iz1[i-count0]];
count0++;
}
}
count99+=4;
if (count99>109)
{
count99=0;
nfix_first_run=0;
}
return _mm_setr_ps(output1[0],output1[1],output1[2],output1[3]);
}
}
__m128 nfix2_SSE(void)
{
__m128 x2 __attribute__((aligned(16)));
__m128 y2 __attribute__((aligned(16)));
__m128i cmplt_option0_128 __attribute__((aligned(16)));
__m128i cmplt_option1_128 __attribute__((aligned(16)));
__m128i cmplt_option2_128 __attribute__((aligned(16)));
int32_t cmplt_option0[4] __attribute__((aligned(16)));
int32_t cmplt_option1[4] __attribute__((aligned(16)));
int32_t cmplt_option2[4] __attribute__((aligned(16)));
float output2[12] __attribute__((aligned(16)));
float x2_option0[4] __attribute__((aligned(16)));
float x4[4] __attribute__((aligned(16)));
static int count0=0;
static int count1=0;
static int count2=0;
static int count99=0;
int i;
static float r = 3.442620;
static int nfix_first_run=0;
for (;;)
{
NOR2_SSE;
//(abs(hz)<kn[iz])? hz*wn[iz]
cmplt_option0_128 = _mm_cmplt_epi32(abs_hz2_128,_mm_setr_epi32(kn[iz2[0]],kn[iz2[1]],kn[iz2[2]],kn[iz2[3]]));
_mm_storeu_si128((__m128i *)cmplt_option0,cmplt_option0_128);
//x=hz * wn[iz];
for (i=0;i<4;i++)
{
if (cmplt_option0[i]==0xFFFFFFFF)
{
//printf("count0 %d\n",count0);
output2[count0]=hz2[i]*wn[iz2[i]];
count0++;
}
}
if (count0>3)
{
count99+=4;
if (count99>99)
{
count99=0;
nfix_first_run=0;
}
return _mm_setr_ps(output2[0],output2[1],output2[2],output2[3]);
}
//x=hz * wn[iz];
x2=_mm_mul_ps(_mm_cvtepi32_ps(hz2_128),_mm_setr_ps(wn[iz2[0]],wn[iz2[1]],wn[iz2[2]],wn[iz2[3]]));
_mm_storeu_ps(x2_option0,x2);
//printf("count0 is %d, count1 is %d, count2 is %d,count99 is %d\n",count0,count1,count2,count99);
if ((iz2[0]==0||iz2[1]==0||iz2[2]==0||iz2[3]==0)&&nfix_first_run==0&&count0>0)
{
//printf("\niz == 0 [%d,%d,%d,%d]\n\n",iz4[0],iz4[1],iz4[2],iz4[3]);
nfix_first_run=1;
do
{
//x = - 0.2904764 * log (UNI);
x2 = _mm_mul_ps(_mm_set1_ps(-0.2904764f), log_ps(UNI_SSE));
_mm_storeu_ps(x4,x2);
//y = - log (UNI);
y2 = _mm_mul_ps(_mm_set1_ps(-1.0f), log_ps(UNI_SSE));
//(y+y < x*x)?
cmplt_option1_128 = _mm_cvtps_epi32(_mm_cmplt_ps(_mm_add_ps(y2,y2),_mm_mul_ps(x2,x2)));
_mm_storeu_si128((__m128i *)cmplt_option1,cmplt_option1_128);
for (i=0;i<4;i++)
{
if (cmplt_option1[i]==0x80000000)
{
//printf("count22 %d\n",count2);
output2[3]=(hz2[i]>0)? x4[i]+r:-x4[i]-r;
count2++;
break;
}
}
}
while (cmplt_option1[0]!=0x80000000 || cmplt_option1[1]!=0x80000000 || cmplt_option1[2]!=0x80000000 || cmplt_option1[3]!=0x80000000);
if (count0+count2>3)
{
count99+=4;
if (count99>109)
{
count99=0;
nfix_first_run=0;
}
return _mm_setr_ps(output2[0],output2[1],output2[2],output2[3]);
}
}
if (iz2[0]>0&&iz2[1]>0&&iz2[2]>0&&iz2[3]>0&&nfix_first_run==0&&count0>0)
{
//printf("\niz > 0 [%d,%d,%d,%d]\n\n",iz4[0],iz4[1],iz4[2],iz4[3]);
nfix_first_run=1;
printf("\niz2 > 0 [%d,%d,%d,%d].\nfn [%e,%e,%e,%e].\n\n",iz2[0],iz2[1],iz2[2],iz2[3],fn[iz2[0]],fn[iz2[1]],fn[iz2[2]],fn[iz2[3]]);
printf("fn2 - 1 [%e,%e,%e,%e]\n",fn[iz2[0]-1],fn[iz2[1]-1],fn[iz2[2]-1],fn[iz2[3]-1]);
//if (iz==0)
printf("\niz [%d,%d,%d,%d]\n",iz4[0],iz4[1],iz4[2],iz4[3]);
printf("iz==0 [%d,%d,%d,%d]\n",iz4[0]==0,iz4[1]==0,iz4[2]==0,iz4[3]==0);
printf("iz>0 [%d,%d,%d,%d]\n\n",iz4[0]>0,iz4[1]>0,iz4[2]>0,iz4[3]>0);//
// if (fn[iz]+UNI*(fn[iz-1]-fn[iz])<exp(-0.5*x*x))
//printf("iz [%d,%d,%d,%d] is ok? %d\n",iz4[0],iz4[1],iz4[2],iz4[3],iz4[0]==0&&iz4[1]==0&&iz4[2]==0&&iz4[3]==0);
printf("iz>0 inside [%d,%d,%d,%d]\n",iz4[0]>0,iz4[1]>0,iz4[2]>0,iz4[3]>0);
printf("iz-1 [%d,%d,%d,%d]\n",iz4[0]-1,iz4[1]-1,iz4[2]-1,iz4[3]-1);
printf("x [%e,%e,%e,%e]\n",x[0],x[1],x[2],x[3]);
printf("exp [%e,%e,%e,%e]\n",exp(-0.5*x[0]*x[0]),exp(-0.5*x[1]*x[1]),exp(-0.5*x[2]*x[2]),exp(-0.5*x[3]*x[3]));//
cmplt_option2_128 = _mm_cvtps_epi32(_mm_cmplt_ps(_mm_add_ps(_mm_setr_ps(fn[iz2[0]],fn[iz2[1]],fn[iz2[2]],fn[iz2[3]]),_mm_mul_ps(UNI_SSE,_mm_sub_ps(_mm_setr_ps(fn[iz2[0]-1],fn[iz2[1]-1],fn[iz2[2]-1],fn[iz2[3]-1]),_mm_setr_ps(fn[iz2[0]],fn[iz2[1]],fn[iz2[2]],fn[iz2[3]])))),exp_ps(_mm_mul_ps(_mm_mul_ps(x2,x2),_mm_set1_ps(-0.5f)))));
//cmplt_option2_128 = _mm_cvtps_epi32(_mm_cmplt_ps(_mm_setr_ps(fn[iz4[0]],fn[iz4[1]],fn[iz4[2]],fn[iz4[3]]),exp_ps(_mm_mul_ps(_mm_mul_ps(x,x),_mm_set1_ps(-0.5f)))));
_mm_storeu_si128((__m128i *)cmplt_option2,cmplt_option2_128);
for (i=0;i<4;i++)
{
if (cmplt_option2[i]==0x80000000)
{
//printf("count1 %d\n",count1);
output2[3]=x2_option0[i];
count1++;
break;
}
}
if (count0+count1>3)
{
count99+=4;
if (count99>109)
{
count99=0;
nfix_first_run=0;
}
return _mm_setr_ps(output2[0],output2[1],output2[2],output2[3]);
}
}
NOR2_SSE;
//(abs(hz)<kn[iz])? hz*wn[iz]
cmplt_option0_128 = _mm_cmplt_epi32(abs_hz2_128,_mm_setr_epi32(kn[iz2[0]],kn[iz2[1]],kn[iz2[2]],kn[iz2[3]]));
_mm_storeu_si128((__m128i *)cmplt_option0,cmplt_option0_128);
for (i=count0;i<3;i++)
{
if (cmplt_option0[i-count0]==0xFFFFFFFF)
{
//printf("count0 %d\n",count0);
output2[count0]=hz2[i-count0]*wn[iz2[i-count0]];
count0++;
}
}
count99+=4;
if (count99>109)
{
count99=0;
nfix_first_run=0;
}
return _mm_setr_ps(output2[0],output2[1],output2[2],output2[3]);
}
}*/
openair2/RRC/LITE/rrc_UE.c
View file @ 02f1ec69
......@@ -2717,6 +2717,7 @@ int decode_BCCH_DLSCH_Message(
sizeof(SystemInformationBlockType1_t) );
LOG_D( RRC, "[UE %"PRIu8"] Decoding First SIB1\n", ctxt_pP->module_id );
decode_SIB1( ctxt_pP, eNB_index, rsrq, rsrp );
//printf("decode_BCCH_DLSCH_Message \n");
}
}
......@@ -2731,7 +2732,7 @@ int decode_BCCH_DLSCH_Message(
&bcch_message->message.choice.c1.choice.systemInformation,
sizeof(SystemInformation_t) );
LOG_D( RRC, "[UE %"PRIu8"] Decoding SI for frameP %"PRIu32"\n",
LOG_I( RRC, "[UE %"PRIu8"] Decoding SI for frameP %"PRIu32"\n",
ctxt_pP->module_id,
ctxt_pP->frame );
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment