| /cmsis-dsp-latest/dsppp/Include/dsppp/Neon/ |
| D | float.hpp | 52 inline float32x4_t vload1(const float32_t *p) in vload1() argument
54 return(vld1q(p)); in vload1()
59 inline float32x4_t vload1(const float32_t *p) in vload1() argument
62 res[0] = *p; in vload1()
63 p += S; in vload1()
65 res[1] = *p; in vload1()
66 p += S; in vload1()
68 res[2] = *p; in vload1()
69 p += S; in vload1()
71 res[3] = *p; in vload1()
[all …]
|
| /cmsis-dsp-latest/dsppp/Include/dsppp/DSP/ |
| D | memory.hpp | 30 const int16_t *p=reinterpret_cast<const int16_t* >(pQ15); in read_q15x2() local
33 memcpy (&val, p, 4); in read_q15x2()
35 val = (p[1] << 16) | (p[0] & 0x0FFFF) ; in read_q15x2()
48 int16_t *p=reinterpret_cast<int16_t* >(pQ15); in write_q15x2() local
51 memcpy (p, &val, 4); in write_q15x2()
53 p[0] = (int16_t)(val & 0x0FFFF); in write_q15x2()
54 p[1] = (int16_t)(val >> 16); in write_q15x2()
63 const int8_t *p=reinterpret_cast<const int8_t*>(pQ7); in read_q7x4() local
66 memcpy (&val, p, 4); in read_q7x4()
68 val =((p[3] & 0x0FF) << 24) | ((p[2] & 0x0FF) << 16) | ((p[1] & 0x0FF) << 8) | (p[0] & 0x0FF); in read_q7x4()
[all …]
|
| D | q7.hpp | 166 inline Q7DSPVector vload1(const Q7 *p) in vload1() argument
168 return(Q7DSPVector(read_q7x4(p))); in vload1()
174 inline Q7DSPVector vload1(const Q7 *p) in vload1() argument
176 Q7 a = p[0]; in vload1()
177 Q7 b = p[S]; in vload1()
178 Q7 c = p[2*S]; in vload1()
179 Q7 d = p[3*S]; in vload1()
186 inline Q7DSPVector vload1(const Q7 *p,index_t stride) in vload1() argument
188 Q7 a = p[0]; in vload1()
189 Q7 b = *(p+stride); in vload1()
[all …]
|
| D | q15.hpp | 226 inline Q15DSPVector vload1(const Q15 *p) in vload1() argument
228 return(Q15DSPVector(read_q15x2(p))); in vload1()
234 inline Q15DSPVector vload1(const Q15 *p) in vload1() argument
236 Q15 a = p[0]; in vload1()
237 Q15 b = p[S]; in vload1()
244 inline Q15DSPVector vload1(const Q15 *p,index_t stride) in vload1() argument
246 Q15 a = p[0]; in vload1()
247 Q15 b = *(p+stride); in vload1()
254 inline void vstore1(Q15 *p,const Q15DSPVector val) in vstore1() argument
256 write_q15x2 (p, val.v); in vstore1()
[all …]
|
| /cmsis-dsp-latest/Source/FilteringFunctions/ |
| D | arm_levinson_durbin_f32.c | 74 for(int p=1; p < nbCoefs; p++) in arm_levinson_durbin_f32() local
90 pRevPhi = &phi[p-3]; in arm_levinson_durbin_f32()
95 blkCnt = p >> 2; in arm_levinson_durbin_f32()
118 blkCnt = p & 3; in arm_levinson_durbin_f32()
121 suma += a[i] * phi[p - i]; in arm_levinson_durbin_f32()
128 k = (phi[p+1] - suma)/(phi[0] - sumb); in arm_levinson_durbin_f32()
138 vecTmp = vdupq_n_u32(p); in arm_levinson_durbin_f32()
143 nb = p >> 2; in arm_levinson_durbin_f32()
179 switch(p & 3) in arm_levinson_durbin_f32()
184 x = a[j] - k * a[p-1-j]; in arm_levinson_durbin_f32()
[all …]
|
| D | arm_levinson_durbin_q31.c | 136 for(int p=1; p < nbCoefs; p++) in arm_levinson_durbin_q31() local
151 pRevPhi = &phi[p-3]; in arm_levinson_durbin_q31()
156 blkCnt = p >> 2; in arm_levinson_durbin_q31()
176 blkCnt = p & 3; in arm_levinson_durbin_q31()
179 suma += ((q63_t)a[i] * phi[p - i]); in arm_levinson_durbin_q31()
192 k = divide(phi[p+1]-(q31_t)suma,phi[0] - (q31_t)sumb); in arm_levinson_durbin_q31()
202 vecTmp = vdupq_n_u32(p); in arm_levinson_durbin_q31()
208 nb = p >> 2; in arm_levinson_durbin_q31()
247 switch(p & 3) in arm_levinson_durbin_q31()
254 x = a[j] - mul32x32(k,a[p-1-j]); in arm_levinson_durbin_q31()
[all …]
|
| D | arm_levinson_durbin_f16.c | 71 for(int p=1; p < nbCoefs; p++) in arm_levinson_durbin_f16() local
87 pRevPhi = &phi[p-7]; in arm_levinson_durbin_f16()
92 blkCnt = p >> 3; in arm_levinson_durbin_f16()
115 blkCnt = p & 7; in arm_levinson_durbin_f16()
118 suma += (_Float16)a[i] * (_Float16)phi[p - i]; in arm_levinson_durbin_f16()
125 k = ((_Float16)phi[p+1] - suma)/((_Float16)phi[0] - sumb); in arm_levinson_durbin_f16()
135 vecTmp = vdupq_n_u16(p); in arm_levinson_durbin_f16()
140 nb = p >> 3; in arm_levinson_durbin_f16()
176 blkCnt = p & 7; in arm_levinson_durbin_f16()
185 x=(_Float16)a[j] - (_Float16)k * (_Float16)a[p-1-j]; in arm_levinson_durbin_f16()
[all …]
|
| /cmsis-dsp-latest/dsppp/Include/dsppp/Helium/ |
| D | half.hpp | 399 inline float16x8_t vload1(const float16_t *p) in vload1() argument
401 return(vld1q(p)); in vload1()
406 inline float16x8_t vload1(const float16_t *p) in vload1() argument
409 return(vldrhq_gather_shifted_offset_f16(p,offset)); in vload1()
414 inline float16x8_t vload1(const float16_t *p) in vload1() argument
419 res[i] = *p; in vload1()
420 p += S; in vload1()
427 inline float16x8_t vload1(const float16_t *p,const index_t stride) in vload1() argument
433 return(vldrhq_gather_shifted_offset_f16(p,offset)); in vload1()
440 res[i] = *p; in vload1()
[all …]
|
| D | q15.hpp | 236 inline int16x8_t vload1(const Q15 *p) in vload1() argument
238 return(vld1q(reinterpret_cast<const int16_t*>(p))); in vload1()
269 inline int16x8_t vload1(const Q15 *p) in vload1() argument
274 return(vldrhq_gather_shifted_offset_s16(reinterpret_cast<const int16_t*>(p),offset)); in vload1()
279 inline int16x8_t vload1(const Q15 *p) in vload1() argument
284 res[i] = p->v; in vload1()
285 p += S; in vload1()
292 inline int16x8_t vload1(const Q15 *p,index_t stride) in vload1() argument
298 return(vldrhq_gather_shifted_offset_s16(reinterpret_cast<const int16_t*>(p),offset)); in vload1()
305 res[i] = p->v; in vload1()
[all …]
|
| D | float.hpp | 499 inline float32x4_t vload1(const float32_t *p) in vload1() argument
501 return(vld1q(p)); in vload1()
507 inline float32x4_t vload1(const float32_t *p) in vload1() argument
510 return(vldrwq_gather_shifted_offset_f32(p,offset)); in vload1()
524 inline float32x4_t vload1(const float32_t *p,const index_t stride) in vload1() argument
528 return(vldrwq_gather_shifted_offset_f32(p,offset)); in vload1()
546 inline float32x4_t vload1_z(const float32_t *p,const std::size_t nb,const mve_pred16_t p0) in vload1_z() argument
549 return(vld1q_z(p,p0)); in vload1_z()
555 inline float32x4_t vload1_z(const float32_t *p,const std::size_t nb,const mve_pred16_t p0) in vload1_z() argument
562 return(vldrwq_gather_shifted_offset_z_f32(p,offset,p0)); in vload1_z()
[all …]
|
| D | q7.hpp | 199 inline int8x16_t vload1(const Q7 *p) in vload1() argument
201 return(vld1q(reinterpret_cast<const int8_t*>(p))); in vload1()
211 inline int8x16_t vload1(const Q7 *p) in vload1() argument
217 return(vldrbq_gather_offset_s8(reinterpret_cast<const int8_t*>(p),offset)); in vload1()
222 inline int8x16_t vload1(const Q7 *p) in vload1() argument
227 res[i] = p->v; in vload1()
228 p += S; in vload1()
235 inline int8x16_t vload1(const Q7 *p,index_t stride) in vload1() argument
241 return(vldrbq_gather_offset_s8(reinterpret_cast<const int8_t*>(p),offset)); in vload1()
248 res[i] = p->v; in vload1()
[all …]
|
| D | q31.hpp | 199 inline int32x4_t vload1(const Q31 *p) in vload1() argument
201 return(vld1q(reinterpret_cast<const int32_t*>(p))); in vload1()
206 inline int32x4_t vload1(const Q31 *p) in vload1() argument
209 return(vldrwq_gather_shifted_offset_s32(reinterpret_cast<const int32_t*>(p),offset)); in vload1()
214 inline int32x4_t vload1(const Q31 *p,index_t stride) in vload1() argument
218 return(vldrwq_gather_shifted_offset_s32(reinterpret_cast<const int32_t*>(p),offset)); in vload1()
224 inline int32x4_t vload1_z(const Q31 *p,std::size_t nb,mve_pred16_t p0) in vload1_z() argument
227 return(vld1q_z(reinterpret_cast<const int32_t*>(p),p0)); in vload1_z()
233 inline int32x4_t vload1_z(const Q31 *p,std::size_t nb,mve_pred16_t p0) in vload1_z() argument
238 return(vldrwq_gather_shifted_offset_z_s32(reinterpret_cast<const int32_t*>(p),offset,p0)); in vload1_z()
[all …]
|
| /cmsis-dsp-latest/Source/StatisticsFunctions/ |
| D | arm_entropy_f32.c | 57 float32_t accum=0.0f,p; in arm_entropy_f32() local
86 p = *pSrcA++; in arm_entropy_f32()
87 accum += p * logf(p); in arm_entropy_f32()
105 float32_t accum, p; in arm_entropy_f32() local
136 p = *pIn++; in arm_entropy_f32()
137 accum += p * logf(p); in arm_entropy_f32()
151 float32_t accum, p; in arm_entropy_f32() local
160 p = *pIn++; in arm_entropy_f32()
161 accum += p * logf(p); in arm_entropy_f32()
|
| D | arm_entropy_f16.c | 70 _Float16 accum=0.0f16,p; in arm_entropy_f16() local
99 p = *pSrcA++; in arm_entropy_f16()
100 accum += p * (_Float16)logf((float32_t)p); in arm_entropy_f16()
115 _Float16 accum, p; in arm_entropy_f16() local
124 p = *pIn++; in arm_entropy_f16()
125 accum += p * (_Float16)logf((float32_t)p); in arm_entropy_f16()
|
| /cmsis-dsp-latest/Source/TransformFunctions/ |
| D | arm_rfft_fast_f64.c | 33 const float64_t * p, in stage_rfft_f64() argument
39 const float64_t *pA = p; /* increasing pointer */ in stage_rfft_f64()
40 const float64_t *pB = p; /* decreasing pointer */ in stage_rfft_f64()
70 pB = p + 2*k; in stage_rfft_f64()
120 const float64_t * p, in merge_rfft_f64() argument
126 const float64_t *pA = p; /* increasing pointer */ in merge_rfft_f64()
127 const float64_t *pB = p; /* decreasing pointer */ in merge_rfft_f64()
141 pB = p + 2*k ; in merge_rfft_f64()
202 float64_t * p, in arm_rfft_fast_f64() argument
213 merge_rfft_f64(S, p, pOut); in arm_rfft_fast_f64()
[all …]
|
| D | arm_rfft_fast_f32.c | 34 const float32_t * p, in stage_rfft_f32() argument
40 const float32_t *pA = p; /* increasing pointer */ in stage_rfft_f32()
41 const float32_t *pB = p; /* decreasing pointer */ in stage_rfft_f32()
81 pB = p + 2*k; in stage_rfft_f32()
193 const float32_t * p, in merge_rfft_f32() argument
199 const float32_t *pA = p; /* increasing pointer */ in merge_rfft_f32()
200 const float32_t *pB = p; /* decreasing pointer */ in merge_rfft_f32()
224 pB = p + 2*k ; in merge_rfft_f32()
312 const float32_t * p, in stage_rfft_f32() argument
318 const float32_t *pA = p; /* increasing pointer */ in stage_rfft_f32()
[all …]
|
| D | arm_rfft_fast_f16.c | 39 const float16_t * p, in stage_rfft_f16() argument
45 const float16_t *pA = p; /* increasing pointer */ in stage_rfft_f16()
46 const float16_t *pB = p; /* decreasing pointer */ in stage_rfft_f16()
86 pB = p + 2*k - 14; in stage_rfft_f16()
199 const float16_t * p, in merge_rfft_f16() argument
205 const float16_t *pA = p; /* increasing pointer */ in merge_rfft_f16()
206 const float16_t *pB = p; /* decreasing pointer */ in merge_rfft_f16()
229 pB = p + 2*k - 14; in merge_rfft_f16()
318 const float16_t * p, in stage_rfft_f16() argument
324 const float16_t *pA = p; /* increasing pointer */ in stage_rfft_f16()
[all …]
|
| /cmsis-dsp-latest/Source/ComplexMathFunctions/ |
| D | arm_cmplx_mult_cmplx_q31.c | 126 mve_pred16_t p = vctp32q(blkCnt); in arm_cmplx_mult_cmplx_q31() local
131 vecSrcA = vldrwq_z_s32(pSrcA, p); in arm_cmplx_mult_cmplx_q31()
132 vecSrcB = vldrwq_z_s32(pSrcB, p); in arm_cmplx_mult_cmplx_q31()
134 vecDst = vqdmlsdhq_m(vuninitializedq_s32(), vecSrcA, vecSrcB, p); in arm_cmplx_mult_cmplx_q31()
135 vecDst = vqdmladhxq_m(vecDst, vecSrcA, vecSrcB, p); in arm_cmplx_mult_cmplx_q31()
137 vecDst = vshrq_m(vuninitializedq_s32(), vecDst, 2, p); in arm_cmplx_mult_cmplx_q31()
138 vstrwq_p_s32(pDst, vecDst, p); in arm_cmplx_mult_cmplx_q31()
147 mve_pred16_t p = vctp32q(blkCnt); in arm_cmplx_mult_cmplx_q31() local
149 vecSrcA = vldrwq_z_s32(pSrcA, p); in arm_cmplx_mult_cmplx_q31()
150 vecSrcB = vldrwq_z_s32(pSrcB, p); in arm_cmplx_mult_cmplx_q31()
[all …]
|
| D | arm_cmplx_mult_cmplx_q15.c | 129 mve_pred16_t p = vctp16q(blkCnt); in arm_cmplx_mult_cmplx_q15() local
134 vecSrcA = vldrhq_z_s16(pSrcA, p); in arm_cmplx_mult_cmplx_q15()
135 vecSrcB = vldrhq_z_s16(pSrcB, p); in arm_cmplx_mult_cmplx_q15()
137 vecDst = vqdmlsdhq_m(vuninitializedq_s16(), vecSrcA, vecSrcB, p); in arm_cmplx_mult_cmplx_q15()
138 vecDst = vqdmladhxq_m(vecDst, vecSrcA, vecSrcB, p); in arm_cmplx_mult_cmplx_q15()
140 vecDst = vshrq_m(vuninitializedq_s16(), vecDst, 2, p); in arm_cmplx_mult_cmplx_q15()
141 vstrhq_p_s16(pDst, vecDst, p); in arm_cmplx_mult_cmplx_q15()
152 mve_pred16_t p = vctp16q(blkCnt); in arm_cmplx_mult_cmplx_q15() local
154 vecSrcA = vldrhq_z_s16(pSrcA, p); in arm_cmplx_mult_cmplx_q15()
155 vecSrcB = vldrhq_z_s16(pSrcB, p); in arm_cmplx_mult_cmplx_q15()
[all …]
|
| D | arm_cmplx_mult_cmplx_f16.c | 119 mve_pred16_t p = vctp16q(blkCnt); in arm_cmplx_mult_cmplx_f16() local
123 vecSrcA = vldrhq_z_f16(pSrcA, p); in arm_cmplx_mult_cmplx_f16()
124 vecSrcB = vldrhq_z_f16(pSrcB, p); in arm_cmplx_mult_cmplx_f16()
125 vec_acc = vcmulq_m(vuninitializedq_f16(),vecSrcA, vecSrcB, p); in arm_cmplx_mult_cmplx_f16()
126 vec_acc = vcmlaq_rot90_m(vec_acc, vecSrcA, vecSrcB, p); in arm_cmplx_mult_cmplx_f16()
128 vstrhq_p_f16(pDst, vec_acc, p); in arm_cmplx_mult_cmplx_f16()
138 mve_pred16_t p = vctp16q(blkCnt); in arm_cmplx_mult_cmplx_f16() local
140 vecSrcA = vldrhq_z_f16(pSrcA, p); in arm_cmplx_mult_cmplx_f16()
141 vecSrcB = vldrhq_z_f16(pSrcB, p); in arm_cmplx_mult_cmplx_f16()
143 vec_acc = vcmulq_m(vuninitializedq_f16(),vecSrcA, vecSrcB, p); in arm_cmplx_mult_cmplx_f16()
[all …]
|
| D | arm_cmplx_mult_cmplx_f32.c | 137 mve_pred16_t p = vctp32q(blkCnt); in arm_cmplx_mult_cmplx_f32() local
141 vecSrcA = vldrwq_z_f32(pSrcA, p); in arm_cmplx_mult_cmplx_f32()
142 vecSrcB = vldrwq_z_f32(pSrcB, p); in arm_cmplx_mult_cmplx_f32()
143 vec_acc = vcmulq_m(vuninitializedq_f32(),vecSrcA, vecSrcB, p); in arm_cmplx_mult_cmplx_f32()
144 vec_acc = vcmlaq_rot90_m(vec_acc, vecSrcA, vecSrcB, p); in arm_cmplx_mult_cmplx_f32()
146 vstrwq_p_f32(pDst, vec_acc, p); in arm_cmplx_mult_cmplx_f32()
157 mve_pred16_t p = vctp32q(blkCnt); in arm_cmplx_mult_cmplx_f32() local
159 vecSrcA = vldrwq_z_f32(pSrcA, p); in arm_cmplx_mult_cmplx_f32()
160 vecSrcB = vldrwq_z_f32(pSrcB, p); in arm_cmplx_mult_cmplx_f32()
162 vec_acc = vcmulq_m(vuninitializedq_f32(),vecSrcA, vecSrcB, p); in arm_cmplx_mult_cmplx_f32()
[all …]
|
| /cmsis-dsp-latest/Testing/Source/Benchmarks/ |
| D | UnaryF64.cpp | 91 float64_t *p; in setUp() local
102 p = input1.ptr(); in setUp()
107 memcpy(aPtr,p + offset,sizeof(float64_t)*this->nbr*this->nbr); in setUp()
125 float64_t *p; in setUp() local
140 p = input1.ptr(); in setUp()
142 memcpy(aPtr,&p[offset],sizeof(float64_t)*this->nbr*this->nbr); in setUp()
144 p = input2.ptr(); in setUp()
146 memcpy(bPtr,&p[offset],sizeof(float64_t)*this->nbr*this->nbr); in setUp()
165 float64_t *p; in setUp() local
180 p = input1.ptr(); in setUp()
[all …]
|
| D | UnaryF32.cpp | 149 float32_t *p; in setUp() local
160 p = input1.ptr(); in setUp()
165 memcpy(aPtr,p + offset,sizeof(float32_t)*this->nbr*this->nbr); in setUp()
183 float32_t *p; in setUp() local
198 p = input1.ptr(); in setUp()
200 memcpy(aPtr,&p[offset],sizeof(float32_t)*this->nbr*this->nbr); in setUp()
202 p = input2.ptr(); in setUp()
204 memcpy(bPtr,&p[offset],sizeof(float32_t)*this->nbr*this->nbr); in setUp()
223 float32_t *p; in setUp() local
238 p = input1.ptr(); in setUp()
[all …]
|
| D | UnaryF16.cpp | 140 float16_t *p; in setUp() local
151 p = input1.ptr(); in setUp()
156 memcpy(aPtr,p + offset,sizeof(float16_t)*this->nbr*this->nbr); in setUp()
174 float16_t *p; in setUp() local
189 p = input1.ptr(); in setUp()
191 memcpy(aPtr,&p[offset],sizeof(float16_t)*this->nbr*this->nbr); in setUp()
193 p = input2.ptr(); in setUp()
195 memcpy(bPtr,&p[offset],sizeof(float16_t)*this->nbr*this->nbr); in setUp()
214 float16_t *p; in setUp() local
229 p = input1.ptr(); in setUp()
[all …]
|
| /cmsis-dsp-latest/Testing/FrameworkSource/ |
| D | FPGA.cpp | 126 char *p = str; in readStr() local
129 *p++ = *this->currentDesc++; in readStr()
131 *p++ = 0; in readStr()
354 Testing::param_t *p,*current; in ReadParameterList() local
363 p=(Testing::param_t*)malloc(sizeof(Testing::param_t)*(nbInputSamples)); in ReadParameterList()
364 current=p; in ReadParameterList()
374 gen.data=p; in ReadParameterList()
505 void FPGA::ImportPattern_f64(Testing::PatternID_t id,char* p,Testing::nbSamples_t nb) in ImportPattern_f64() argument
513 float64_t *dst = (float64_t*)p; in ImportPattern_f64()
524 void FPGA::ImportPattern_f32(Testing::PatternID_t id,char* p,Testing::nbSamples_t nb) in ImportPattern_f32() argument
[all …]
|
