summaryrefslogtreecommitdiff
path: root/Drivers/CMSIS/DSP/Source/SupportFunctions/arm_float_to_q7.c
blob: 3b7d6c6c19aa720021b1e3962a1a2977c51f2bff (plain) (blame)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
/* ----------------------------------------------------------------------
 * Project:      CMSIS DSP Library
 * Title:        arm_float_to_q7.c
 * Description:  Converts the elements of the floating-point vector to Q7 vector
 *
 * $Date:        18. March 2019
 * $Revision:    V1.6.0
 *
 * Target Processor: Cortex-M cores
 * -------------------------------------------------------------------- */
/*
 * Copyright (C) 2010-2019 ARM Limited or its affiliates. All rights reserved.
 *
 * SPDX-License-Identifier: Apache-2.0
 *
 * Licensed under the Apache License, Version 2.0 (the License); you may
 * not use this file except in compliance with the License.
 * You may obtain a copy of the License at
 *
 * www.apache.org/licenses/LICENSE-2.0
 *
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an AS IS BASIS, WITHOUT
 * WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 * See the License for the specific language governing permissions and
 * limitations under the License.
 */

#include "arm_math.h"

/**
  @ingroup groupSupport
 */

/**
  @addtogroup float_to_x
  @{
 */

/**
 * @brief Converts the elements of the floating-point vector to Q7 vector.
 * @param[in]       *pSrc points to the floating-point input vector
 * @param[out]      *pDst points to the Q7 output vector
 * @param[in]       blockSize length of the input vector
 * @return none.
 *
 *\par Description:
 * \par
 * The equation used for the conversion process is:
 * <pre>
 * 	pDst[n] = (q7_t)(pSrc[n] * 128);   0 <= n < blockSize.
 * </pre>
 * \par Scaling and Overflow Behavior:
 * \par
 * The function uses saturating arithmetic.
 * Results outside of the allowable Q7 range [0x80 0x7F] will be saturated.
 * \note
 * In order to apply rounding, the library should be rebuilt with the ROUNDING macro
 * defined in the preprocessor section of project options.
 */

#if defined(ARM_MATH_NEON)
void arm_float_to_q7(
  const float32_t * pSrc,
  q7_t * pDst,
  uint32_t blockSize)
{
  const float32_t *pIn = pSrc;                         /* Src pointer */
  uint32_t blkCnt;                               /* loop counter */

  float32_t in;
  float32x4_t inV;
  #ifdef ARM_MATH_ROUNDING
  float32x4_t zeroV = vdupq_n_f32(0.0f);
  float32x4_t pHalf = vdupq_n_f32(0.5f / 128.0f);
  float32x4_t mHalf = vdupq_n_f32(-0.5f / 128.0f);
  float32x4_t r;
  uint32x4_t cmp;
  #endif

  int32x4_t cvt;
  int16x4_t cvt1,cvt2;
  int8x8_t outV;

  blkCnt = blockSize >> 3U;

  /* Compute 8 outputs at a time.
   ** a second loop below computes the remaining 1 to 7 samples. */
  while (blkCnt > 0U)
  {

#ifdef ARM_MATH_ROUNDING
    /* C = A * 128 */
    /* Convert from float to q7 and then store the results in the destination buffer */
    inV = vld1q_f32(pIn);
    cmp = vcgtq_f32(inV,zeroV);
    r = vbslq_f32(cmp,pHalf,mHalf);
    inV = vaddq_f32(inV, r);
    cvt1 = vqmovn_s32(vcvtq_n_s32_f32(inV,7));
    pIn += 4;

    inV = vld1q_f32(pIn);
    cmp = vcgtq_f32(inV,zeroV);
    r = vbslq_f32(cmp,pHalf,mHalf);
    inV = vaddq_f32(inV, r);
    cvt2 = vqmovn_s32(vcvtq_n_s32_f32(inV,7));
    pIn += 4;
    
    outV = vqmovn_s16(vcombine_s16(cvt1,cvt2));
    vst1_s8(pDst, outV);
    pDst += 8;

#else

    /* C = A * 128 */
    /* Convert from float to q7 and then store the results in the destination buffer */
    inV = vld1q_f32(pIn);
    cvt1 = vqmovn_s32(vcvtq_n_s32_f32(inV,7));
    pIn += 4;

    inV = vld1q_f32(pIn);
    cvt2 = vqmovn_s32(vcvtq_n_s32_f32(inV,7));
    pIn += 4;

    outV = vqmovn_s16(vcombine_s16(cvt1,cvt2));

    vst1_s8(pDst, outV);
    pDst += 8;
#endif /*      #ifdef ARM_MATH_ROUNDING        */

    /* Decrement the loop counter */
    blkCnt--;
  }

  /* If the blockSize is not a multiple of 4, compute any remaining output samples here.
   ** No loop unrolling is used. */
  blkCnt = blockSize & 7;

  while (blkCnt > 0U)
  {

#ifdef ARM_MATH_ROUNDING
    /* C = A * 128 */
    /* Convert from float to q7 and then store the results in the destination buffer */
    in = *pIn++;
    in = (in * 128);
    in += in > 0.0f ? 0.5f : -0.5f;
    *pDst++ = (q7_t) (__SSAT((q15_t) (in), 8));

#else

    /* C = A * 128 */
    /* Convert from float to q7 and then store the results in the destination buffer */
    *pDst++ = __SSAT((q31_t) (*pIn++ * 128.0f), 8);

#endif /*      #ifdef ARM_MATH_ROUNDING        */

    /* Decrement the loop counter */
    blkCnt--;
  }

}
#else
void arm_float_to_q7(
  const float32_t * pSrc,
        q7_t * pDst,
        uint32_t blockSize)
{
        uint32_t blkCnt;                               /* Loop counter */
  const float32_t *pIn = pSrc;                         /* Source pointer */

#ifdef ARM_MATH_ROUNDING
        float32_t in;
#endif /* #ifdef ARM_MATH_ROUNDING */

#if defined (ARM_MATH_LOOPUNROLL)

  /* Loop unrolling: Compute 4 outputs at a time */
  blkCnt = blockSize >> 2U;

  while (blkCnt > 0U)
  {
    /* C = A * 128 */

    /* Convert from float to q7 and store result in destination buffer */
#ifdef ARM_MATH_ROUNDING

    in = (*pIn++ * 128);
    in += in > 0.0f ? 0.5f : -0.5f;
    *pDst++ = (q7_t) (__SSAT((q15_t) (in), 8));

    in = (*pIn++ * 128);
    in += in > 0.0f ? 0.5f : -0.5f;
    *pDst++ = (q7_t) (__SSAT((q15_t) (in), 8));

    in = (*pIn++ * 128);
    in += in > 0.0f ? 0.5f : -0.5f;
    *pDst++ = (q7_t) (__SSAT((q15_t) (in), 8));

    in = (*pIn++ * 128);
    in += in > 0.0f ? 0.5f : -0.5f;
    *pDst++ = (q7_t) (__SSAT((q15_t) (in), 8));

#else

    *pDst++ = __SSAT((q31_t) (*pIn++ * 128.0f), 8);
    *pDst++ = __SSAT((q31_t) (*pIn++ * 128.0f), 8);
    *pDst++ = __SSAT((q31_t) (*pIn++ * 128.0f), 8);
    *pDst++ = __SSAT((q31_t) (*pIn++ * 128.0f), 8);

#endif /* #ifdef ARM_MATH_ROUNDING */

    /* Decrement loop counter */
    blkCnt--;
  }

  /* Loop unrolling: Compute remaining outputs */
  blkCnt = blockSize % 0x4U;

#else

  /* Initialize blkCnt with number of samples */
  blkCnt = blockSize;

#endif /* #if defined (ARM_MATH_LOOPUNROLL) */

  while (blkCnt > 0U)
  {
    /* C = A * 128 */

    /* Convert from float to q7 and store result in destination buffer */
#ifdef ARM_MATH_ROUNDING

    in = (*pIn++ * 128);
    in += in > 0.0f ? 0.5f : -0.5f;
    *pDst++ = (q7_t) (__SSAT((q15_t) (in), 8));

#else

    *pDst++ = (q7_t) __SSAT((q31_t) (*pIn++ * 128.0f), 8);

#endif /* #ifdef ARM_MATH_ROUNDING */

    /* Decrement loop counter */
    blkCnt--;
  }

}
#endif /* #if defined(ARM_MATH_NEON) */

/**
  @} end of float_to_x group
 */