Corrections to the documentation.

pull/100/head
Christophe Favergeon 3 years ago
parent a5048f6ab9
commit e13bd5a664

@ -56,6 +56,9 @@
and then saturated to 1.15 format.
@par
Refer to \ref arm_mat_mult_fast_q15() for a faster but less precise version of this function.
@par pState
pState will contain the transpose of pSrcB
*/
#if defined(ARM_MATH_MVEI) && !defined(ARM_MATH_AUTOVECTORIZE)

@ -68,29 +68,29 @@ void arm_split_rifft_q15(
| RFFT Size | Input Format | Output Format | Number of bits to upscale |
| ---------: | ------------: | -------------: | ------------------------: |
| 32 | 1.15 | 5.11 | 5 |
| 64 | 1.15 | 6.10 | 6 |
| 128 | 1.15 | 7.9 | 7 |
| 256 | 1.15 | 8.8 | 8 |
| 512 | 1.15 | 9.7 | 9 |
| 1024 | 1.15 | 10.6 | 10 |
| 2048 | 1.15 | 11.5 | 11 |
| 4096 | 1.15 | 12.4 | 12 |
| 8192 | 1.15 | 13.3 | 13 |
| 32 | 1.15 | 6.10 | 5 |
| 64 | 1.15 | 7.9 | 6 |
| 128 | 1.15 | 8.8 | 7 |
| 256 | 1.15 | 9.7 | 8 |
| 512 | 1.15 | 10.6 | 9 |
| 1024 | 1.15 | 11.5 | 10 |
| 2048 | 1.15 | 12.4 | 11 |
| 4096 | 1.15 | 13.3 | 12 |
| 8192 | 1.15 | 14.2 | 13 |
@par Input and Output formats for RIFFT Q15
| RIFFT Size | Input Format | Output Format | Number of bits to upscale |
| ----------: | ------------: | -------------: | ------------------------: |
| 32 | 1.15 | 5.11 | 0 |
| 64 | 1.15 | 6.10 | 0 |
| 128 | 1.15 | 7.9 | 0 |
| 256 | 1.15 | 8.8 | 0 |
| 512 | 1.15 | 9.7 | 0 |
| 1024 | 1.15 | 10.6 | 0 |
| 2048 | 1.15 | 11.5 | 0 |
| 4096 | 1.15 | 12.4 | 0 |
| 8192 | 1.15 | 13.3 | 0 |
| 32 | 1.15 | 6.10 | 0 |
| 64 | 1.15 | 7.9 | 0 |
| 128 | 1.15 | 8.8 | 0 |
| 256 | 1.15 | 9.7 | 0 |
| 512 | 1.15 | 10.6 | 0 |
| 1024 | 1.15 | 11.5 | 0 |
| 2048 | 1.15 | 12.4 | 0 |
| 4096 | 1.15 | 13.3 | 0 |
| 8192 | 1.15 | 14.2 | 0 |
@par
If the input buffer is of length N (fftLenReal), the output buffer must have length 2N

@ -68,29 +68,29 @@ void arm_split_rifft_q31(
| RFFT Size | Input Format | Output Format | Number of bits to upscale |
| ---------: | ------------: | -------------: | ------------------------: |
| 32 | 1.31 | 5.27 | 5 |
| 64 | 1.31 | 6.26 | 6 |
| 128 | 1.31 | 7.25 | 7 |
| 256 | 1.31 | 8.24 | 8 |
| 512 | 1.31 | 9.23 | 9 |
| 1024 | 1.31 | 10.22 | 10 |
| 2048 | 1.31 | 11.21 | 11 |
| 4096 | 1.31 | 12.20 | 12 |
| 8192 | 1.31 | 13.19 | 13 |
| 32 | 1.31 | 6.26 | 5 |
| 64 | 1.31 | 7.25 | 6 |
| 128 | 1.31 | 8.24 | 7 |
| 256 | 1.31 | 9.23 | 8 |
| 512 | 1.31 | 10.22 | 9 |
| 1024 | 1.31 | 11.21 | 10 |
| 2048 | 1.31 | 12.20 | 11 |
| 4096 | 1.31 | 13.19 | 12 |
| 8192 | 1.31 | 14.18 | 13 |
@par Input and Output formats for RIFFT Q31
| RIFFT Size | Input Format | Output Format | Number of bits to upscale |
| ----------: | ------------: | -------------: | ------------------------: |
| 32 | 1.31 | 5.27 | 0 |
| 64 | 1.31 | 6.26 | 0 |
| 128 | 1.31 | 7.25 | 0 |
| 256 | 1.31 | 8.24 | 0 |
| 512 | 1.31 | 9.23 | 0 |
| 1024 | 1.31 | 10.22 | 0 |
| 2048 | 1.31 | 11.21 | 0 |
| 4096 | 1.31 | 12.20 | 0 |
| 8192 | 1.31 | 13.19 | 0 |
| 32 | 1.31 | 6.26 | 0 |
| 64 | 1.31 | 7.25 | 0 |
| 128 | 1.31 | 8.24 | 0 |
| 256 | 1.31 | 9.23 | 0 |
| 512 | 1.31 | 10.22 | 0 |
| 1024 | 1.31 | 11.21 | 0 |
| 2048 | 1.31 | 12.20 | 0 |
| 4096 | 1.31 | 13.19 | 0 |
| 8192 | 1.31 | 14.18 | 0 |
@par
If the input buffer is of length N (fftLenReal), the output buffer must have length 2N

Loading…
Cancel
Save