| /hal_nxp-3.5.0/mcux/mcux-sdk/CMSIS/DSP/Source/QuaternionMathFunctions/ |
| D | arm_quaternion_inverse_f32.c | 64 f32x4_t vec1,vec2; in arm_quaternion_inverse_f32() local
70 vec1 = vld1q(pInputQuaternions); in arm_quaternion_inverse_f32()
71 vec2 = vmulq(vec1,vec1); in arm_quaternion_inverse_f32()
75 vec1 = vmulq_n_f32(vec1, 1.0f / squaredSum); in arm_quaternion_inverse_f32()
76 vec1 = vsetq_lane_f32(-vgetq_lane(vec1, 0),vec1,0); in arm_quaternion_inverse_f32()
77 vec1 = vnegq_f32(vec1); in arm_quaternion_inverse_f32()
79 vst1q(pInverseQuaternions, vec1); in arm_quaternion_inverse_f32()
|
| D | arm_quaternion_conjugate_f32.c | 62 f32x4_t vec1; in arm_quaternion_conjugate_f32() local
66 vec1 = vld1q(pInputQuaternions); in arm_quaternion_conjugate_f32()
69 vec1 = vsetq_lane_f32(-vgetq_lane(vec1, 0),vec1,0); in arm_quaternion_conjugate_f32()
70 vec1 = vnegq_f32(vec1); in arm_quaternion_conjugate_f32()
72 vst1q(pConjugateQuaternions, vec1); in arm_quaternion_conjugate_f32()
|
| D | arm_quaternion_normalize_f32.c | 63 f32x4_t vec1,vec2; in arm_quaternion_normalize_f32() local
68 vec1 = vld1q(pInputQuaternions); in arm_quaternion_normalize_f32()
69 vec2 = vmulq(vec1,vec1); in arm_quaternion_normalize_f32()
72 vec1 = vmulq_n_f32(vec1, 1.0f / norm); in arm_quaternion_normalize_f32()
73 vst1q(pNormalizedQuaternions, vec1); in arm_quaternion_normalize_f32()
|
| D | arm_quaternion2rotation_f32.c | 85 f32x4_t vec0,vec1, vec2 ,vec3; in arm_quaternion2rotation_f32() local
95 vec1 = vmulq(vec0,vec0); in arm_quaternion2rotation_f32()
114 vec0 = vsetq_lane(vgetq_lane(vec1,0) + vgetq_lane(vec1,1),vec0,0); in arm_quaternion2rotation_f32()
115 vec0 = vsetq_lane(vgetq_lane(vec0,0) - vgetq_lane(vec1,2),vec0,0); in arm_quaternion2rotation_f32()
116 vec0 = vsetq_lane(vgetq_lane(vec0,0) - vgetq_lane(vec1,3),vec0,0); in arm_quaternion2rotation_f32()
124 tmp1 = vgetq_lane(vec1,0) - vgetq_lane(vec1,1); in arm_quaternion2rotation_f32()
125 tmp2 = vgetq_lane(vec1,2) - vgetq_lane(vec1,3); in arm_quaternion2rotation_f32()
|
| D | arm_quaternion_norm_f32.c | 64 f32x4_t vec1; in arm_quaternion_norm_f32() local
69 vec1 = vld1q(pInputQuaternions); in arm_quaternion_norm_f32()
70 vec1 = vmulq(vec1,vec1); in arm_quaternion_norm_f32()
71 squaredSum = vecAddAcrossF32Mve(vec1); in arm_quaternion_norm_f32()
|
| /hal_nxp-3.5.0/mcux/mcux-sdk/CMSIS/DSP/Source/MatrixFunctions/ |
| D | arm_mat_inverse_f32.c | 456 f32x4_t vec1, vec2; in arm_mat_inverse_f32() local
461 vec1 = vldrwq_f32(pInT1); in arm_mat_inverse_f32()
463 vec1 = vfmsq_f32(vec1, tmpV, vec2); in arm_mat_inverse_f32()
464 vstrwq_f32(pInT1, vec1); in arm_mat_inverse_f32()
479 f32x4_t vec1, vec2; in arm_mat_inverse_f32() local
482 vec1 = vldrwq_f32(pInT1); in arm_mat_inverse_f32()
484 vec1 = vfmsq_f32(vec1, tmpV, vec2); in arm_mat_inverse_f32()
485 vstrwq_p_f32(pInT1, vec1, p0); in arm_mat_inverse_f32()
492 f32x4_t vec1, vec2; in arm_mat_inverse_f32() local
498 vec1 = vldrwq_f32(pOutT1); in arm_mat_inverse_f32()
[all …]
|
| D | arm_mat_inverse_f16.c | 438 f16x8_t vec1, vec2; in arm_mat_inverse_f16() local
443 vec1 = vldrhq_f16(pInT1); in arm_mat_inverse_f16()
445 vec1 = vfmsq_f16(vec1, tmpV, vec2); in arm_mat_inverse_f16()
446 vstrhq_f16(pInT1, vec1); in arm_mat_inverse_f16()
461 f16x8_t vec1, vec2; in arm_mat_inverse_f16() local
464 vec1 = vldrhq_f16(pInT1); in arm_mat_inverse_f16()
466 vec1 = vfmsq_f16(vec1, tmpV, vec2); in arm_mat_inverse_f16()
467 vstrhq_p_f16(pInT1, vec1, p0); in arm_mat_inverse_f16()
474 f16x8_t vec1, vec2; in arm_mat_inverse_f16() local
480 vec1 = vldrhq_f16(pOutT1); in arm_mat_inverse_f16()
[all …]
|
| /hal_nxp-3.5.0/mcux/mcux-sdk/CMSIS/DSP/Source/BasicMathFunctions/ |
| D | arm_abs_f16.c | 74 f16x8_t vec1; in arm_abs_f16() local
86 vec1 = vld1q(pSrc); in arm_abs_f16()
87 res = vabsq(vec1); in arm_abs_f16()
106 vec1 = vld1q(pSrc); in arm_abs_f16()
107 vstrhq_p(pDst, vabsq(vec1), p0); in arm_abs_f16()
122 f16x8_t vec1; in arm_abs_f16() local
133 vec1 = vld1q_f16(pSrc); in arm_abs_f16()
134 res = vabsq_f16(vec1); in arm_abs_f16()
|
| D | arm_abs_f32.c | 74 f32x4_t vec1; in arm_abs_f32() local
86 vec1 = vld1q(pSrc); in arm_abs_f32()
87 res = vabsq(vec1); in arm_abs_f32()
106 vec1 = vld1q(pSrc); in arm_abs_f32()
107 vstrwq_p(pDst, vabsq(vec1), p0); in arm_abs_f32()
121 f32x4_t vec1; in arm_abs_f32() local
132 vec1 = vld1q_f32(pSrc); in arm_abs_f32()
133 res = vabsq_f32(vec1); in arm_abs_f32()
|
| D | arm_negate_f32.c | 72 f32x4_t vec1; in arm_negate_f32() local
83 vec1 = vld1q(pSrc); in arm_negate_f32()
84 res = vnegq(vec1); in arm_negate_f32()
101 vec1 = vld1q((float32_t const *) pSrc); in arm_negate_f32()
102 vstrwq_p(pDst, vnegq(vec1), p0); in arm_negate_f32()
116 f32x4_t vec1; in arm_negate_f32() local
127 vec1 = vld1q_f32(pSrc); in arm_negate_f32()
128 res = vnegq_f32(vec1); in arm_negate_f32()
|
| D | arm_offset_f32.c | 75 f32x4_t vec1; in arm_offset_f32() local
85 vec1 = vld1q(pSrc); in arm_offset_f32()
86 res = vaddq(vec1,offset); in arm_offset_f32()
103 vec1 = vld1q((float32_t const *) pSrc); in arm_offset_f32()
104 vstrwq_p(pDst, vaddq(vec1, offset), p0); in arm_offset_f32()
120 f32x4_t vec1; in arm_offset_f32() local
131 vec1 = vld1q_f32(pSrc); in arm_offset_f32()
132 res = vaddq_f32(vec1,vdupq_n_f32(offset)); in arm_offset_f32()
|
| D | arm_add_f32.c | 73 f32x4_t vec1; in arm_add_f32() local
85 vec1 = vld1q(pSrcA); in arm_add_f32()
87 res = vaddq(vec1, vec2); in arm_add_f32()
106 vec1 = vld1q(pSrcA); in arm_add_f32()
108 vstrwq_p(pDst, vaddq(vec1,vec2), p0); in arm_add_f32()
123 f32x4_t vec1; in arm_add_f32() local
135 vec1 = vld1q_f32(pSrcA); in arm_add_f32()
137 res = vaddq_f32(vec1, vec2); in arm_add_f32()
|
| D | arm_mult_f32.c | 73 f32x4_t vec1; in arm_mult_f32() local
84 vec1 = vld1q(pSrcA); in arm_mult_f32()
86 res = vmulq(vec1, vec2); in arm_mult_f32()
104 vec1 = vld1q(pSrcA); in arm_mult_f32()
106 vstrwq_p(pDst, vmulq(vec1,vec2), p0); in arm_mult_f32()
121 f32x4_t vec1; in arm_mult_f32() local
133 vec1 = vld1q_f32(pSrcA); in arm_mult_f32()
135 res = vmulq_f32(vec1, vec2); in arm_mult_f32()
|
| D | arm_sub_f32.c | 73 f32x4_t vec1; in arm_sub_f32() local
85 vec1 = vld1q(pSrcA); in arm_sub_f32()
87 res = vsubq(vec1, vec2); in arm_sub_f32()
106 vec1 = vld1q(pSrcA); in arm_sub_f32()
108 vstrwq_p(pDst, vsubq(vec1,vec2), p0); in arm_sub_f32()
123 f32x4_t vec1; in arm_sub_f32() local
135 vec1 = vld1q_f32(pSrcA); in arm_sub_f32()
137 res = vsubq_f32(vec1, vec2); in arm_sub_f32()
|
| D | arm_scale_f32.c | 88 f32x4_t vec1; in arm_scale_f32() local
99 vec1 = vld1q(pSrc); in arm_scale_f32()
100 res = vmulq(vec1,scale); in arm_scale_f32()
117 vec1 = vld1q((float32_t const *) pSrc); in arm_scale_f32()
118 vstrwq_p(pDst, vmulq(vec1, scale), p0); in arm_scale_f32()
133 f32x4_t vec1; in arm_scale_f32() local
144 vec1 = vld1q_f32(pSrc); in arm_scale_f32()
145 res = vmulq_f32(vec1, vdupq_n_f32(scale)); in arm_scale_f32()
|
| D | arm_negate_f16.c | 72 f16x8_t vec1; in arm_negate_f16() local
83 vec1 = vld1q(pSrc); in arm_negate_f16()
84 res = vnegq(vec1); in arm_negate_f16()
101 vec1 = vld1q((float16_t const *) pSrc); in arm_negate_f16()
102 vstrhq_p(pDst, vnegq(vec1), p0); in arm_negate_f16()
|
| D | arm_scale_f16.c | 88 f16x8_t vec1; in arm_scale_f16() local
99 vec1 = vld1q(pSrc); in arm_scale_f16()
100 res = vmulq(vec1,scale); in arm_scale_f16()
117 vec1 = vld1q((float16_t const *) pSrc); in arm_scale_f16()
118 vstrhq_p(pDst, vmulq(vec1, scale), p0); in arm_scale_f16()
|
| D | arm_offset_f16.c | 75 f16x8_t vec1; in arm_offset_f16() local
85 vec1 = vld1q(pSrc); in arm_offset_f16()
86 res = vaddq(vec1,offset); in arm_offset_f16()
103 vec1 = vld1q((float16_t const *) pSrc); in arm_offset_f16()
104 vstrhq_p(pDst, vaddq(vec1, offset), p0); in arm_offset_f16()
|
| D | arm_add_f16.c | 73 f16x8_t vec1; in arm_add_f16() local
85 vec1 = vld1q(pSrcA); in arm_add_f16()
87 res = vaddq(vec1, vec2); in arm_add_f16()
106 vec1 = vld1q(pSrcA); in arm_add_f16()
108 vstrhq_p(pDst, vaddq(vec1,vec2), p0); in arm_add_f16()
|
| D | arm_mult_f16.c | 74 f16x8_t vec1; in arm_mult_f16() local
85 vec1 = vld1q(pSrcA); in arm_mult_f16()
87 res = vmulq(vec1, vec2); in arm_mult_f16()
105 vec1 = vld1q(pSrcA); in arm_mult_f16()
107 vstrhq_p(pDst, vmulq(vec1,vec2), p0); in arm_mult_f16()
|
| D | arm_sub_f16.c | 72 f16x8_t vec1; in arm_sub_f16() local
84 vec1 = vld1q(pSrcA); in arm_sub_f16()
86 res = vsubq(vec1, vec2); in arm_sub_f16()
105 vec1 = vld1q(pSrcA); in arm_sub_f16()
107 vstrhq_p(pDst, vsubq(vec1,vec2), p0); in arm_sub_f16()
|
| D | arm_dot_prod_f32.c | 132 f32x4_t vec1; in arm_dot_prod_f32() local
140 vec1 = vld1q_f32(pSrcA); in arm_dot_prod_f32()
148 accum = vmlaq_f32(accum, vec1, vec2); in arm_dot_prod_f32()
154 vec1 = vld1q_f32(pSrcA); in arm_dot_prod_f32()
|
| /hal_nxp-3.5.0/mcux/mcux-sdk/CMSIS/DSP/Source/SVMFunctions/ |
| D | arm_svm_linear_predict_f32.c | 298 float32x4_t vec1; in arm_svm_linear_predict_f32() local
335 vec1 = vld1q_f32(pIn); in arm_svm_linear_predict_f32()
347 accuma = vmlaq_f32(accuma, vec1,vec2a); in arm_svm_linear_predict_f32()
348 accumb = vmlaq_f32(accumb, vec1,vec2b); in arm_svm_linear_predict_f32()
349 accumc = vmlaq_f32(accumc, vec1,vec2c); in arm_svm_linear_predict_f32()
350 accumd = vmlaq_f32(accumd, vec1,vec2d); in arm_svm_linear_predict_f32()
380 vec1 = vld1q_f32(pDualCoefs); in arm_svm_linear_predict_f32()
383 accum = vmulq_f32(vec1,dotV); in arm_svm_linear_predict_f32()
407 vec1 = vld1q_f32(pIn); in arm_svm_linear_predict_f32()
412 accum = vmlaq_f32(accum, vec1,vec2); in arm_svm_linear_predict_f32()
|
| D | arm_svm_sigmoid_predict_f32.c | 316 float32x4_t vec1; in arm_svm_sigmoid_predict_f32() local
353 vec1 = vld1q_f32(pIn); in arm_svm_sigmoid_predict_f32()
365 accuma = vmlaq_f32(accuma, vec1,vec2a); in arm_svm_sigmoid_predict_f32()
366 accumb = vmlaq_f32(accumb, vec1,vec2b); in arm_svm_sigmoid_predict_f32()
367 accumc = vmlaq_f32(accumc, vec1,vec2c); in arm_svm_sigmoid_predict_f32()
368 accumd = vmlaq_f32(accumd, vec1,vec2d); in arm_svm_sigmoid_predict_f32()
398 vec1 = vld1q_f32(pDualCoefs); in arm_svm_sigmoid_predict_f32()
407 accum = vmulq_f32(vec1,dotV); in arm_svm_sigmoid_predict_f32()
432 vec1 = vld1q_f32(pIn); in arm_svm_sigmoid_predict_f32()
437 accum = vmlaq_f32(accum, vec1,vec2); in arm_svm_sigmoid_predict_f32()
|
| D | arm_svm_polynomial_predict_f32.c | 318 float32x4_t vec1; in arm_svm_polynomial_predict_f32() local
355 vec1 = vld1q_f32(pIn); in arm_svm_polynomial_predict_f32()
367 accuma = vmlaq_f32(accuma, vec1,vec2a); in arm_svm_polynomial_predict_f32()
368 accumb = vmlaq_f32(accumb, vec1,vec2b); in arm_svm_polynomial_predict_f32()
369 accumc = vmlaq_f32(accumc, vec1,vec2c); in arm_svm_polynomial_predict_f32()
370 accumd = vmlaq_f32(accumd, vec1,vec2d); in arm_svm_polynomial_predict_f32()
400 vec1 = vld1q_f32(pDualCoefs); in arm_svm_polynomial_predict_f32()
409 accum = vmulq_f32(vec1,dotV); in arm_svm_polynomial_predict_f32()
434 vec1 = vld1q_f32(pIn); in arm_svm_polynomial_predict_f32()
439 accum = vmlaq_f32(accum, vec1,vec2); in arm_svm_polynomial_predict_f32()
|
