RMUL2025/lib/cmsis_5/CMSIS/DSP/Testing/Source/Tests/BinaryTestsQ15.cpp

195 lines
6.1 KiB
C++
Executable File

#include "BinaryTestsQ15.h"
#include <stdio.h>
#include "Error.h"
#define SNR_THRESHOLD 70
#define SNR_LOW_THRESHOLD 30
/*
Reference patterns are generated with
a double precision computation.
*/
#define ABS_HIGH_ERROR_Q15 ((q15_t)2000)
#define ABS_ERROR_Q15 ((q15_t)1000)
#define ABS_ERROR_Q63 ((q63_t)(1<<16))
#define MULT_SNR_THRESHOLD 60
#define ONEHALF 0x4000
/* Upper bound of maximum matrix dimension used by Python */
#define MAXMATRIXDIM 40
static void checkInnerTail(q15_t *b)
{
ASSERT_TRUE(b[0] == 0);
ASSERT_TRUE(b[1] == 0);
ASSERT_TRUE(b[2] == 0);
ASSERT_TRUE(b[3] == 0);
ASSERT_TRUE(b[4] == 0);
ASSERT_TRUE(b[5] == 0);
ASSERT_TRUE(b[6] == 0);
ASSERT_TRUE(b[7] == 0);
}
#define LOADDATA2() \
const q15_t *inp1=input1.ptr(); \
const q15_t *inp2=input2.ptr(); \
\
q15_t *ap=a.ptr(); \
q15_t *bp=b.ptr(); \
\
q15_t *outp=output.ptr(); \
q15_t *tmpPtr=tmp.ptr(); \
int16_t *dimsp = dims.ptr(); \
int nbMatrixes = dims.nbSamples() / 3;\
int rows,internal,columns; \
int i;
#define PREPAREDATA2C() \
in1.numRows=rows; \
in1.numCols=internal; \
memcpy((void*)ap,(const void*)inp1,2*sizeof(q15_t)*rows*internal);\
in1.pData = ap; \
\
in2.numRows=internal; \
in2.numCols=columns; \
memcpy((void*)bp,(const void*)inp2,2*sizeof(q15_t)*internal*columns);\
in2.pData = bp; \
\
out.numRows=rows; \
out.numCols=columns; \
out.pData = outp;
#define PREPAREDATA2R() \
in1.numRows=rows; \
in1.numCols=internal; \
memcpy((void*)ap,(const void*)inp1,sizeof(q15_t)*rows*internal);\
in1.pData = ap; \
\
in2.numRows=internal; \
in2.numCols=columns; \
memcpy((void*)bp,(const void*)inp2,sizeof(q15_t)*internal*columns);\
in2.pData = bp; \
\
out.numRows=rows; \
out.numCols=columns; \
out.pData = outp;
void BinaryTestsQ15::test_mat_mult_q15()
{
LOADDATA2();
arm_status status;
for(i=0;i < nbMatrixes ; i ++)
{
rows = *dimsp++;
internal = *dimsp++;
columns = *dimsp++;
PREPAREDATA2R();
memset(tmpPtr,0,sizeof(q15_t)*internal*columns + 16);
status=arm_mat_mult_q15(&this->in1,&this->in2,&this->out,tmpPtr);
ASSERT_TRUE(status==ARM_MATH_SUCCESS);
outp += (rows * columns);
checkInnerTail(outp);
checkInnerTail(tmpPtr + internal * columns);
}
ASSERT_SNR(output,ref,(q15_t)SNR_LOW_THRESHOLD);
ASSERT_NEAR_EQ(output,ref,ABS_HIGH_ERROR_Q15);
}
void BinaryTestsQ15::test_mat_cmplx_mult_q15()
{
LOADDATA2();
arm_status status;
for(i=0;i < nbMatrixes ; i ++)
{
rows = *dimsp++;
internal = *dimsp++;
columns = *dimsp++;
PREPAREDATA2C();
status=arm_mat_cmplx_mult_q15(&this->in1,&this->in2,&this->out,tmpPtr);
ASSERT_TRUE(status==ARM_MATH_SUCCESS);
outp += (2*rows * columns);
checkInnerTail(outp);
}
ASSERT_SNR(output,ref,(q15_t)MULT_SNR_THRESHOLD);
ASSERT_NEAR_EQ(output,ref,ABS_ERROR_Q15);
}
void BinaryTestsQ15::setUp(Testing::testID_t id,std::vector<Testing::param_t>& params,Client::PatternMgr *mgr)
{
(void)params;
switch(id)
{
case TEST_MAT_MULT_Q15_1:
input1.reload(BinaryTestsQ15::INPUTS1_Q15_ID,mgr);
input2.reload(BinaryTestsQ15::INPUTS2_Q15_ID,mgr);
dims.reload(BinaryTestsQ15::DIMSBINARY1_S16_ID,mgr);
ref.reload(BinaryTestsQ15::REFMUL1_Q15_ID,mgr);
output.create(ref.nbSamples(),BinaryTestsQ15::OUT_Q15_ID,mgr);
a.create(MAXMATRIXDIM*MAXMATRIXDIM,BinaryTestsQ15::TMPA_Q15_ID,mgr);
b.create(MAXMATRIXDIM*MAXMATRIXDIM,BinaryTestsQ15::TMPB_Q15_ID,mgr);
tmp.create(MAXMATRIXDIM*MAXMATRIXDIM,BinaryTestsQ15::TMP_Q15_ID,mgr);
break;
case TEST_MAT_CMPLX_MULT_Q15_2:
input1.reload(BinaryTestsQ15::INPUTSC1_Q15_ID,mgr);
input2.reload(BinaryTestsQ15::INPUTSC2_Q15_ID,mgr);
dims.reload(BinaryTestsQ15::DIMSBINARY1_S16_ID,mgr);
ref.reload(BinaryTestsQ15::REFCMPLXMUL1_Q15_ID,mgr);
output.create(ref.nbSamples(),BinaryTestsQ15::OUT_Q15_ID,mgr);
a.create(2*MAXMATRIXDIM*MAXMATRIXDIM,BinaryTestsQ15::TMPA_Q15_ID,mgr);
b.create(2*MAXMATRIXDIM*MAXMATRIXDIM,BinaryTestsQ15::TMPB_Q15_ID,mgr);
tmp.create(2*MAXMATRIXDIM*MAXMATRIXDIM,BinaryTestsQ15::TMP_Q15_ID,mgr);
break;
}
}
void BinaryTestsQ15::tearDown(Testing::testID_t id,Client::PatternMgr *mgr)
{
(void)id;
output.dump(mgr);
}