/hal_nxp-3.6.0/mcux/mcux-sdk/CMSIS/DSP/Source/MatrixFunctions/ |
D | arm_mat_cmplx_mult_q31.c | 79 q31x4_t vecB, vecA; in arm_mat_cmplx_mult_q31_2x2_mve() local 91 vecA = vldrwq_s32(pInA0); in arm_mat_cmplx_mult_q31_2x2_mve() 92 acc0 = vmlsldavq_s32(vecA, vecB); in arm_mat_cmplx_mult_q31_2x2_mve() 93 acc1 = vmlaldavxq_s32(vecA, vecB); in arm_mat_cmplx_mult_q31_2x2_mve() 95 vecA = vldrwq_s32(pInA1); in arm_mat_cmplx_mult_q31_2x2_mve() 96 acc2 = vmlsldavq_s32(vecA, vecB); in arm_mat_cmplx_mult_q31_2x2_mve() 97 acc3 = vmlaldavxq_s32(vecA, vecB); in arm_mat_cmplx_mult_q31_2x2_mve() 109 vecA = vldrwq_s32(pInA0); in arm_mat_cmplx_mult_q31_2x2_mve() 110 acc0 = vmlsldavq_s32(vecA, vecB); in arm_mat_cmplx_mult_q31_2x2_mve() 111 acc1 = vmlaldavxq_s32(vecA, vecB); in arm_mat_cmplx_mult_q31_2x2_mve() [all …]
|
D | arm_mat_cmplx_mult_f32.c | 83 f32x4_t vecB, vecA; in arm_mat_cmplx_mult_f32_2x2_mve() local 95 vecA = vldrwq_f32(pInA0); in arm_mat_cmplx_mult_f32_2x2_mve() 96 acc0 = vcmulq(vecA, vecB); in arm_mat_cmplx_mult_f32_2x2_mve() 97 acc0 = vcmlaq_rot90(acc0, vecA, vecB); in arm_mat_cmplx_mult_f32_2x2_mve() 99 vecA = vldrwq_f32(pInA1); in arm_mat_cmplx_mult_f32_2x2_mve() 100 acc1 = vcmulq(vecA, vecB); in arm_mat_cmplx_mult_f32_2x2_mve() 101 acc1 = vcmlaq_rot90(acc1, vecA, vecB); in arm_mat_cmplx_mult_f32_2x2_mve() 116 vecA = vldrwq_f32(pInA0); in arm_mat_cmplx_mult_f32_2x2_mve() 117 acc0 = vcmulq(vecA, vecB); in arm_mat_cmplx_mult_f32_2x2_mve() 118 acc0 = vcmlaq_rot90(acc0, vecA, vecB); in arm_mat_cmplx_mult_f32_2x2_mve() [all …]
|
D | arm_mat_mult_q31.c | 139 q31x4_t vecB, vecA; in arm_mat_mult_q31_3x3_mve() local 150 vecA = vldrwq_s32(pInA0); in arm_mat_mult_q31_3x3_mve() 151 acc0 = vrmlaldavhq(vecA, vecB); in arm_mat_mult_q31_3x3_mve() 152 vecA = vldrwq_s32(pInA1); in arm_mat_mult_q31_3x3_mve() 153 acc1 = vrmlaldavhq(vecA, vecB); in arm_mat_mult_q31_3x3_mve() 154 vecA = vldrwq_s32(pInA2); in arm_mat_mult_q31_3x3_mve() 155 acc2 = vrmlaldavhq(vecA, vecB); in arm_mat_mult_q31_3x3_mve() 171 vecA = vldrwq_s32(pInA0); in arm_mat_mult_q31_3x3_mve() 172 acc0 = vrmlaldavhq(vecA, vecB); in arm_mat_mult_q31_3x3_mve() 173 vecA = vldrwq_s32(pInA1); in arm_mat_mult_q31_3x3_mve() [all …]
|
D | arm_mat_cmplx_mult_f16.c | 249 f16x8_t vecB, vecA; in arm_mat_cmplx_mult_f16_4x4_mve() local 262 vecA = vldrhq_f16(pInA0); in arm_mat_cmplx_mult_f16_4x4_mve() 263 acc0 = vcmulq(vecA, vecB); in arm_mat_cmplx_mult_f16_4x4_mve() 264 acc0 = vcmlaq_rot90(acc0, vecA, vecB); in arm_mat_cmplx_mult_f16_4x4_mve() 266 vecA = vldrhq_f16(pInA1); in arm_mat_cmplx_mult_f16_4x4_mve() 267 acc1 = vcmulq(vecA, vecB); in arm_mat_cmplx_mult_f16_4x4_mve() 268 acc1 = vcmlaq_rot90(acc1, vecA, vecB); in arm_mat_cmplx_mult_f16_4x4_mve() 270 vecA = vldrhq_f16(pInA2); in arm_mat_cmplx_mult_f16_4x4_mve() 271 acc2 = vcmulq(vecA, vecB); in arm_mat_cmplx_mult_f16_4x4_mve() 272 acc2 = vcmlaq_rot90(acc2, vecA, vecB); in arm_mat_cmplx_mult_f16_4x4_mve() [all …]
|
D | arm_mat_vec_mult_f16.c | 109 f16x8_t vecA; in arm_mat_vec_mult_f16() local 113 vecA = vld1q(pSrcA0Vec); in arm_mat_vec_mult_f16() 115 acc0 = vfmaq(acc0, vecIn, vecA); in arm_mat_vec_mult_f16() 116 vecA = vld1q(pSrcA1Vec); in arm_mat_vec_mult_f16() 118 acc1 = vfmaq(acc1, vecIn, vecA); in arm_mat_vec_mult_f16() 119 vecA = vld1q(pSrcA2Vec); in arm_mat_vec_mult_f16() 121 acc2 = vfmaq(acc2, vecIn, vecA); in arm_mat_vec_mult_f16() 122 vecA = vld1q(pSrcA3Vec); in arm_mat_vec_mult_f16() 124 acc3 = vfmaq(acc3, vecIn, vecA); in arm_mat_vec_mult_f16() 136 f16x8_t vecA; in arm_mat_vec_mult_f16() local [all …]
|
D | arm_mat_vec_mult_f32.c | 113 f32x4_t vecA; in arm_mat_vec_mult_f32() local 117 vecA = vld1q(pSrcA0Vec); in arm_mat_vec_mult_f32() 119 acc0 = vfmaq(acc0, vecIn, vecA); in arm_mat_vec_mult_f32() 120 vecA = vld1q(pSrcA1Vec); in arm_mat_vec_mult_f32() 122 acc1 = vfmaq(acc1, vecIn, vecA); in arm_mat_vec_mult_f32() 123 vecA = vld1q(pSrcA2Vec); in arm_mat_vec_mult_f32() 125 acc2 = vfmaq(acc2, vecIn, vecA); in arm_mat_vec_mult_f32() 126 vecA = vld1q(pSrcA3Vec); in arm_mat_vec_mult_f32() 128 acc3 = vfmaq(acc3, vecIn, vecA); in arm_mat_vec_mult_f32() 140 f32x4_t vecA; in arm_mat_vec_mult_f32() local [all …]
|
D | arm_mat_ldlt_f32.c | 190 f32x4_t vecA; in arm_mat_ldlt_f32() local 211 vecA = vldrwq_z_f32(&pA[(w + 0)*n+x],p0); in arm_mat_ldlt_f32() 212 vecA = vfmsq_m(vecA, vecW0, vecX, p0); in arm_mat_ldlt_f32() 213 vstrwq_p(&pA[(w + 0)*n+x], vecA, p0); in arm_mat_ldlt_f32() 215 vecA = vldrwq_z_f32(&pA[(w + 1)*n+x],p0); in arm_mat_ldlt_f32() 216 vecA = vfmsq_m(vecA, vecW1, vecX, p0); in arm_mat_ldlt_f32() 217 vstrwq_p(&pA[(w + 1)*n+x], vecA, p0); in arm_mat_ldlt_f32() 219 vecA = vldrwq_z_f32(&pA[(w + 2)*n+x],p0); in arm_mat_ldlt_f32() 220 vecA = vfmsq_m(vecA, vecW2, vecX, p0); in arm_mat_ldlt_f32() 221 vstrwq_p(&pA[(w + 2)*n+x], vecA, p0); in arm_mat_ldlt_f32() [all …]
|
D | arm_mat_solve_upper_triangular_f32.c | 91 f32x4_t vecA; in arm_mat_solve_upper_triangular_f32() local 98 vecA = vld1q_f32(&pA[i * n + j]); in arm_mat_solve_upper_triangular_f32() 103 vecA = vfmsq(vecA,vdupq_n_f32(pUT[n*i + k]),vecX); in arm_mat_solve_upper_triangular_f32() 112 vecA = vmulq(vecA,vdupq_n_f32(invUT)); in arm_mat_solve_upper_triangular_f32() 115 vst1q(&pX[i*n+j],vecA); in arm_mat_solve_upper_triangular_f32() 188 f32x4_t vecA; in arm_mat_solve_upper_triangular_f32() local 195 vecA = vld1q_f32(&pA[i * n + j]); in arm_mat_solve_upper_triangular_f32() 200 vecA = vfmsq_f32(vecA,vdupq_n_f32(pUT[n*i + k]),vecX); in arm_mat_solve_upper_triangular_f32() 209 vecA = vmulq_f32(vecA,vdupq_n_f32(invUT)); in arm_mat_solve_upper_triangular_f32() 212 vst1q_f32(&pX[i*n+j],vecA); in arm_mat_solve_upper_triangular_f32()
|
D | arm_mat_solve_lower_triangular_f32.c | 98 f32x4_t vecA; in arm_mat_solve_lower_triangular_f32() local 106 vecA = vld1q_f32(&pA[i * n + j]); in arm_mat_solve_lower_triangular_f32() 111 vecA = vfmsq(vecA,vdupq_n_f32(pLT[n*i + k]),vecX); in arm_mat_solve_lower_triangular_f32() 120 vecA = vmulq(vecA,vdupq_n_f32(invLT)); in arm_mat_solve_lower_triangular_f32() 121 vst1q(&pX[i*n+j],vecA); in arm_mat_solve_lower_triangular_f32() 199 f32x4_t vecA; in arm_mat_solve_lower_triangular_f32() local 207 vecA = vld1q_f32(&pA[i * n + j]); in arm_mat_solve_lower_triangular_f32() 212 vecA = vfmsq_f32(vecA,vdupq_n_f32(pLT[n*i + k]),vecX); in arm_mat_solve_lower_triangular_f32() 221 vecA = vmulq_f32(vecA,vdupq_n_f32(invLT)); in arm_mat_solve_lower_triangular_f32() 222 vst1q_f32(&pX[i*n+j],vecA); in arm_mat_solve_lower_triangular_f32()
|
D | arm_mat_cmplx_mult_q15.c | 136 q15x8_t vecA, vecB, vecB2; in arm_mat_cmplx_mult_q15() local 170 vecA = vld1q(pSrcAVec); in arm_mat_cmplx_mult_q15() 174 acc0 = vmlsldavaq_s16(acc0, vecA, vecB); in arm_mat_cmplx_mult_q15() 175 acc1 = vmlaldavaxq_s16(acc1, vecA, vecB); in arm_mat_cmplx_mult_q15() 182 acc2 = vmlsldavaq_s16(acc2, vecA, vecB2); in arm_mat_cmplx_mult_q15() 183 acc3 = vmlaldavaxq_s16(acc3, vecA, vecB2); in arm_mat_cmplx_mult_q15() 197 vecA = vldrhq_z_s16(pSrcAVec, p0); in arm_mat_cmplx_mult_q15() 199 acc0 = vmlsldavaq_s16(acc0, vecA, vecB); in arm_mat_cmplx_mult_q15() 200 acc1 = vmlaldavaxq_s16(acc1, vecA, vecB); in arm_mat_cmplx_mult_q15() 208 acc2 = vmlsldavaq_s16(acc2, vecA, vecB2); in arm_mat_cmplx_mult_q15() [all …]
|
D | arm_mat_inverse_f16.c | 217 f16x8_t vecA, vecB; in arm_mat_inverse_f16() local 228 vecA = vldrhq_f16(pTmpA); in arm_mat_inverse_f16() 230 vstrhq_f16(pTmpB, vecA); in arm_mat_inverse_f16() 249 vecA = vldrhq_f16(pTmpA); in arm_mat_inverse_f16() 251 vstrhq_p_f16(pTmpB, vecA, p0); in arm_mat_inverse_f16() 263 vecA = vldrhq_f16(pTmpA); in arm_mat_inverse_f16() 265 vstrhq_f16(pTmpB, vecA); in arm_mat_inverse_f16() 282 vecA = vldrhq_f16(pTmpA); in arm_mat_inverse_f16() 284 vstrhq_p_f16(pTmpB, vecA, p0); in arm_mat_inverse_f16() 334 f16x8_t vecA; in arm_mat_inverse_f16() local [all …]
|
D | arm_mat_mult_q15.c | 429 q15x8_t vecB, vecA; in arm_mat_mult_q15() local 434 vecA = vld1q(pSrcA0Vec); pSrcA0Vec += 8; in arm_mat_mult_q15() 435 acc0 = vmlaldavaq(acc0, vecA, vecB); in arm_mat_mult_q15() 436 vecA = vld1q(pSrcA1Vec); pSrcA1Vec += 8; in arm_mat_mult_q15() 437 acc1 = vmlaldavaq(acc1, vecA, vecB); in arm_mat_mult_q15() 438 vecA = vld1q(pSrcA2Vec); pSrcA2Vec += 8; in arm_mat_mult_q15() 439 acc2 = vmlaldavaq(acc2, vecA, vecB); in arm_mat_mult_q15() 440 vecA = vld1q(pSrcA3Vec); pSrcA3Vec += 8; in arm_mat_mult_q15() 441 acc3 = vmlaldavaq(acc3, vecA, vecB); in arm_mat_mult_q15() 452 q15x8_t vecB, vecA; in arm_mat_mult_q15() local [all …]
|
D | arm_mat_mult_q7.c | 361 q7x16_t vecA, vecA2, vecB, vecB2; in arm_mat_mult_q7() local 382 vecA = vld1q(pSrcAVec); in arm_mat_mult_q7() 390 acc0 = vmladavaq_s8(acc0, vecA, vecB); in arm_mat_mult_q7() 396 acc2 = vmladavaq_s8(acc2, vecA, vecB2); in arm_mat_mult_q7() 397 vecA = vld1q(pSrcAVec); in arm_mat_mult_q7() 412 acc0 = vmladavaq_p_s8(acc0, vecA, vecB, p0); in arm_mat_mult_q7() 416 acc2 = vmladavaq_p_s8(acc2, vecA, vecB2, p0); in arm_mat_mult_q7() 417 vecA = vld1q(pSrcAVec); in arm_mat_mult_q7() 458 q7x16_t vecA, vecB; in arm_mat_mult_q7() local 474 vecA = vld1q(pSrcAVec); in arm_mat_mult_q7() [all …]
|
D | arm_mat_inverse_f32.c | 235 f32x4_t vecA, vecB; in arm_mat_inverse_f32() local 246 vecA = vldrwq_f32(pTmpA); in arm_mat_inverse_f32() 248 vstrwq_f32(pTmpB, vecA); in arm_mat_inverse_f32() 267 vecA = vldrwq_f32(pTmpA); in arm_mat_inverse_f32() 269 vstrwq_p_f32(pTmpB, vecA, p0); in arm_mat_inverse_f32() 281 vecA = vldrwq_f32(pTmpA); in arm_mat_inverse_f32() 283 vstrwq_f32(pTmpB, vecA); in arm_mat_inverse_f32() 300 vecA = vldrwq_f32(pTmpA); in arm_mat_inverse_f32() 302 vstrwq_p_f32(pTmpB, vecA, p0); in arm_mat_inverse_f32() 352 f32x4_t vecA; in arm_mat_inverse_f32() local [all …]
|
/hal_nxp-3.6.0/mcux/mcux-sdk/CMSIS/DSP/Source/SVMFunctions/ |
D | arm_svm_linear_predict_f16.c | 109 f16x8_t vecA; in arm_svm_linear_predict_f16() local 113 vecA = vld1q(pSrcA0Vec); in arm_svm_linear_predict_f16() 115 acc0 = vfmaq(acc0, vecIn, vecA); in arm_svm_linear_predict_f16() 116 vecA = vld1q(pSrcA1Vec); in arm_svm_linear_predict_f16() 118 acc1 = vfmaq(acc1, vecIn, vecA); in arm_svm_linear_predict_f16() 119 vecA = vld1q(pSrcA2Vec); in arm_svm_linear_predict_f16() 121 acc2 = vfmaq(acc2, vecIn, vecA); in arm_svm_linear_predict_f16() 122 vecA = vld1q(pSrcA3Vec); in arm_svm_linear_predict_f16() 124 acc3 = vfmaq(acc3, vecIn, vecA); in arm_svm_linear_predict_f16() 135 f16x8_t vecA; in arm_svm_linear_predict_f16() local [all …]
|
D | arm_svm_sigmoid_predict_f16.c | 112 f16x8_t vecA; in arm_svm_sigmoid_predict_f16() local 116 vecA = vld1q(pSrcA0Vec); in arm_svm_sigmoid_predict_f16() 118 acc0 = vfmaq(acc0, vecIn, vecA); in arm_svm_sigmoid_predict_f16() 119 vecA = vld1q(pSrcA1Vec); in arm_svm_sigmoid_predict_f16() 121 acc1 = vfmaq(acc1, vecIn, vecA); in arm_svm_sigmoid_predict_f16() 122 vecA = vld1q(pSrcA2Vec); in arm_svm_sigmoid_predict_f16() 124 acc2 = vfmaq(acc2, vecIn, vecA); in arm_svm_sigmoid_predict_f16() 125 vecA = vld1q(pSrcA3Vec); in arm_svm_sigmoid_predict_f16() 127 acc3 = vfmaq(acc3, vecIn, vecA); in arm_svm_sigmoid_predict_f16() 138 f16x8_t vecA; in arm_svm_sigmoid_predict_f16() local [all …]
|
D | arm_svm_polynomial_predict_f16.c | 112 f16x8_t vecA; in arm_svm_polynomial_predict_f16() local 116 vecA = vld1q(pSrcA0Vec); in arm_svm_polynomial_predict_f16() 118 acc0 = vfmaq(acc0, vecIn, vecA); in arm_svm_polynomial_predict_f16() 119 vecA = vld1q(pSrcA1Vec); in arm_svm_polynomial_predict_f16() 121 acc1 = vfmaq(acc1, vecIn, vecA); in arm_svm_polynomial_predict_f16() 122 vecA = vld1q(pSrcA2Vec); in arm_svm_polynomial_predict_f16() 124 acc2 = vfmaq(acc2, vecIn, vecA); in arm_svm_polynomial_predict_f16() 125 vecA = vld1q(pSrcA3Vec); in arm_svm_polynomial_predict_f16() 127 acc3 = vfmaq(acc3, vecIn, vecA); in arm_svm_polynomial_predict_f16() 138 f16x8_t vecA; in arm_svm_polynomial_predict_f16() local [all …]
|
D | arm_svm_rbf_predict_f16.c | 112 f16x8_t vecA; in arm_svm_rbf_predict_f16() local 117 vecA = vld1q(pSrcA0Vec); in arm_svm_rbf_predict_f16() 119 vecDif = vsubq(vecIn, vecA); in arm_svm_rbf_predict_f16() 121 vecA = vld1q(pSrcA1Vec); in arm_svm_rbf_predict_f16() 123 vecDif = vsubq(vecIn, vecA); in arm_svm_rbf_predict_f16() 125 vecA = vld1q(pSrcA2Vec); in arm_svm_rbf_predict_f16() 127 vecDif = vsubq(vecIn, vecA); in arm_svm_rbf_predict_f16() 129 vecA = vld1q(pSrcA3Vec); in arm_svm_rbf_predict_f16() 131 vecDif = vsubq(vecIn, vecA); in arm_svm_rbf_predict_f16() 143 f16x8_t vecA; in arm_svm_rbf_predict_f16() local [all …]
|
D | arm_svm_linear_predict_f32.c | 106 f32x4_t vecA; in arm_svm_linear_predict_f32() local 110 vecA = vld1q(pSrcA0Vec); in arm_svm_linear_predict_f32() 112 acc0 = vfmaq(acc0, vecIn, vecA); in arm_svm_linear_predict_f32() 113 vecA = vld1q(pSrcA1Vec); in arm_svm_linear_predict_f32() 115 acc1 = vfmaq(acc1, vecIn, vecA); in arm_svm_linear_predict_f32() 116 vecA = vld1q(pSrcA2Vec); in arm_svm_linear_predict_f32() 118 acc2 = vfmaq(acc2, vecIn, vecA); in arm_svm_linear_predict_f32() 119 vecA = vld1q(pSrcA3Vec); in arm_svm_linear_predict_f32() 121 acc3 = vfmaq(acc3, vecIn, vecA); in arm_svm_linear_predict_f32() 132 f32x4_t vecA; in arm_svm_linear_predict_f32() local [all …]
|
D | arm_svm_sigmoid_predict_f32.c | 109 f32x4_t vecA; in arm_svm_sigmoid_predict_f32() local 113 vecA = vld1q(pSrcA0Vec); in arm_svm_sigmoid_predict_f32() 115 acc0 = vfmaq(acc0, vecIn, vecA); in arm_svm_sigmoid_predict_f32() 116 vecA = vld1q(pSrcA1Vec); in arm_svm_sigmoid_predict_f32() 118 acc1 = vfmaq(acc1, vecIn, vecA); in arm_svm_sigmoid_predict_f32() 119 vecA = vld1q(pSrcA2Vec); in arm_svm_sigmoid_predict_f32() 121 acc2 = vfmaq(acc2, vecIn, vecA); in arm_svm_sigmoid_predict_f32() 122 vecA = vld1q(pSrcA3Vec); in arm_svm_sigmoid_predict_f32() 124 acc3 = vfmaq(acc3, vecIn, vecA); in arm_svm_sigmoid_predict_f32() 135 f32x4_t vecA; in arm_svm_sigmoid_predict_f32() local [all …]
|
D | arm_svm_polynomial_predict_f32.c | 112 f32x4_t vecA; in arm_svm_polynomial_predict_f32() local 116 vecA = vld1q(pSrcA0Vec); in arm_svm_polynomial_predict_f32() 118 acc0 = vfmaq(acc0, vecIn, vecA); in arm_svm_polynomial_predict_f32() 119 vecA = vld1q(pSrcA1Vec); in arm_svm_polynomial_predict_f32() 121 acc1 = vfmaq(acc1, vecIn, vecA); in arm_svm_polynomial_predict_f32() 122 vecA = vld1q(pSrcA2Vec); in arm_svm_polynomial_predict_f32() 124 acc2 = vfmaq(acc2, vecIn, vecA); in arm_svm_polynomial_predict_f32() 125 vecA = vld1q(pSrcA3Vec); in arm_svm_polynomial_predict_f32() 127 acc3 = vfmaq(acc3, vecIn, vecA); in arm_svm_polynomial_predict_f32() 138 f32x4_t vecA; in arm_svm_polynomial_predict_f32() local [all …]
|
D | arm_svm_rbf_predict_f32.c | 109 f32x4_t vecA; in arm_svm_rbf_predict_f32() local 114 vecA = vld1q(pSrcA0Vec); in arm_svm_rbf_predict_f32() 116 vecDif = vsubq(vecIn, vecA); in arm_svm_rbf_predict_f32() 118 vecA = vld1q(pSrcA1Vec); in arm_svm_rbf_predict_f32() 120 vecDif = vsubq(vecIn, vecA); in arm_svm_rbf_predict_f32() 122 vecA = vld1q(pSrcA2Vec); in arm_svm_rbf_predict_f32() 124 vecDif = vsubq(vecIn, vecA); in arm_svm_rbf_predict_f32() 126 vecA = vld1q(pSrcA3Vec); in arm_svm_rbf_predict_f32() 128 vecDif = vsubq(vecIn, vecA); in arm_svm_rbf_predict_f32() 140 f32x4_t vecA; in arm_svm_rbf_predict_f32() local [all …]
|
/hal_nxp-3.6.0/mcux/mcux-sdk/CMSIS/DSP/Source/TransformFunctions/ |
D | arm_cfft_f16.c | 95 f16x8_t vecA, vecB, vecC, vecD; in _arm_radix4_butterfly_f16_mve() local 138 vecA = vldrhq_f16(inA); in _arm_radix4_butterfly_f16_mve() 145 vecSum0 = vecA + vecC; /* vecSum0 = vaddq(vecA, vecC) */ in _arm_radix4_butterfly_f16_mve() 146 vecDiff0 = vecA - vecC; /* vecSum0 = vsubq(vecA, vecC) */ in _arm_radix4_butterfly_f16_mve() 196 vecA = vldrhq_f16(inA); in _arm_radix4_butterfly_f16_mve() 215 vecA = (f16x8_t)vldrwq_gather_base_wb_f32(&vecScGathAddr, 64); in _arm_radix4_butterfly_f16_mve() 221 vecSum0 = vecA + vecC; /* vecSum0 = vaddq(vecA, vecC) */ in _arm_radix4_butterfly_f16_mve() 222 vecDiff0 = vecA - vecC; /* vecSum0 = vsubq(vecA, vecC) */ in _arm_radix4_butterfly_f16_mve() 231 vecA = (f16x8_t)vldrwq_gather_base_wb_f32(&vecScGathAddr, 64); in _arm_radix4_butterfly_f16_mve() 302 f16x8_t vecA, vecB, vecC, vecD; in _arm_radix4_butterfly_inverse_f16_mve() local [all …]
|
D | arm_cfft_q31.c | 45 q31x4_t vecA, vecB, vecC, vecD; in _arm_radix4_butterfly_q31_mve() local 91 vecA = vldrwq_s32(inA); in _arm_radix4_butterfly_q31_mve() 98 vecSum0 = vhaddq(vecA, vecC); in _arm_radix4_butterfly_q31_mve() 99 vecDiff0 = vhsubq(vecA, vecC); in _arm_radix4_butterfly_q31_mve() 147 vecA = vldrwq_s32(inA); in _arm_radix4_butterfly_q31_mve() 176 vecA = vldrwq_gather_base_wb_s32(&vecScGathAddr, 64); in _arm_radix4_butterfly_q31_mve() 182 vecSum0 = vhaddq(vecA, vecC); in _arm_radix4_butterfly_q31_mve() 183 vecDiff0 = vhsubq(vecA, vecC); in _arm_radix4_butterfly_q31_mve() 193 vecA = vldrwq_gather_base_wb_s32(&vecScGathAddr, 64); in _arm_radix4_butterfly_q31_mve() 295 q31x4_t vecA, vecB, vecC, vecD; in _arm_radix4_butterfly_inverse_q31_mve() local [all …]
|
D | arm_cfft_q15.c | 43 q15x8_t vecA, vecB, vecC, vecD; in _arm_radix4_butterfly_q15_mve() local 88 vecA = vldrhq_s16(inA); in _arm_radix4_butterfly_q15_mve() 95 vecSum0 = vhaddq(vecA, vecC); in _arm_radix4_butterfly_q15_mve() 96 vecDiff0 = vhsubq(vecA, vecC); in _arm_radix4_butterfly_q15_mve() 145 vecA = vldrhq_s16(inA); in _arm_radix4_butterfly_q15_mve() 166 vecA = (q15x8_t) vldrwq_gather_base_wb_s32(&vecScGathAddr, 64); in _arm_radix4_butterfly_q15_mve() 172 vecSum0 = vhaddq(vecA, vecC); in _arm_radix4_butterfly_q15_mve() 173 vecDiff0 = vhsubq(vecA, vecC); in _arm_radix4_butterfly_q15_mve() 183 vecA = (q15x8_t) vldrwq_gather_base_wb_s32(&vecScGathAddr, 64); in _arm_radix4_butterfly_q15_mve() 278 q15x8_t vecA, vecB, vecC, vecD; in _arm_radix4_butterfly_inverse_q15_mve() local [all …]
|