1
0
Fork 0
mirror of https://github.com/betaflight/betaflight.git synced 2025-07-23 16:25:31 +03:00

Remove superfluous double ;

This commit is contained in:
Steve Evans 2022-07-24 14:49:35 +01:00
parent 1f821b6d56
commit cf3615938b
15 changed files with 310 additions and 310 deletions

View file

@ -289,7 +289,7 @@ void arm_conv_fast_q15(
x0 = *__SIMD32(px);
/* read x[1], x[2] samples */
x1 = _SIMD32_OFFSET(px+1);
px+= 2U;
px+= 2U;
/* Apply loop unrolling and compute 4 MACs simultaneously. */
@ -335,7 +335,7 @@ void arm_conv_fast_q15(
/* Read x[5], x[6] */
x1 = _SIMD32_OFFSET(px+3);
px += 4U;
px += 4U;
/* acc2 += x[4] * y[srcBLen - 3] + x[5] * y[srcBLen - 4] */
acc2 = __SMLADX(x0, c0, acc2);
@ -369,7 +369,7 @@ void arm_conv_fast_q15(
/* Read x[7] */
x3 = *__SIMD32(px);
px++;
px++;
/* Perform the multiply-accumulates */
acc0 = __SMLAD(x0, c0, acc0);
@ -388,7 +388,7 @@ void arm_conv_fast_q15(
/* Read x[9] */
x2 = _SIMD32_OFFSET(px+1);
px += 2U;
px += 2U;
/* Perform the multiply-accumulates */
acc0 = __SMLADX(x0, c0, acc0);
@ -415,7 +415,7 @@ void arm_conv_fast_q15(
acc3 = __SMLADX(x2, c0, acc3);
/* Read y[srcBLen - 7] */
c0 = *(py-1);
c0 = *(py-1);
#ifdef ARM_MATH_BIG_ENDIAN
c0 = c0 << 16U;
@ -426,7 +426,7 @@ void arm_conv_fast_q15(
/* Read x[10] */
x3 = _SIMD32_OFFSET(px+2);
px += 3U;
px += 3U;
/* Perform the multiply-accumulates */
acc0 = __SMLADX(x1, c0, acc0);
@ -802,7 +802,7 @@ void arm_conv_fast_q15(
/* First part of the processing with loop unrolling. Compute 4 MACs at a time.
** a second loop below computes MACs for the remaining 1 to 3 samples. */
py++;
py++;
while (k > 0U)
{
@ -887,22 +887,22 @@ void arm_conv_fast_q15(
acc3 = 0;
/* read x[0], x[1] samples */
a = *px++;
b = *px++;
a = *px++;
b = *px++;
#ifndef ARM_MATH_BIG_ENDIAN
x0 = __PKHBT(a, b, 16);
a = *px;
x1 = __PKHBT(b, a, 16);
x0 = __PKHBT(a, b, 16);
a = *px;
x1 = __PKHBT(b, a, 16);
#else
x0 = __PKHBT(b, a, 16);
a = *px;
x1 = __PKHBT(a, b, 16);
x0 = __PKHBT(b, a, 16);
a = *px;
x1 = __PKHBT(a, b, 16);
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
/* Apply loop unrolling and compute 4 MACs simultaneously. */
k = srcBLen >> 2U;
@ -913,19 +913,19 @@ void arm_conv_fast_q15(
{
/* Read the last two inputB samples using SIMD:
* y[srcBLen - 1] and y[srcBLen - 2] */
a = *py;
b = *(py+1);
py -= 2;
a = *py;
b = *(py+1);
py -= 2;
#ifndef ARM_MATH_BIG_ENDIAN
c0 = __PKHBT(a, b, 16);
c0 = __PKHBT(a, b, 16);
#else
c0 = __PKHBT(b, a, 16);;
c0 = __PKHBT(b, a, 16);
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
/* acc0 += x[0] * y[srcBLen - 1] + x[1] * y[srcBLen - 2] */
acc0 = __SMLADX(x0, c0, acc0);
@ -933,22 +933,22 @@ void arm_conv_fast_q15(
/* acc1 += x[1] * y[srcBLen - 1] + x[2] * y[srcBLen - 2] */
acc1 = __SMLADX(x1, c0, acc1);
a = *px;
b = *(px + 1);
a = *px;
b = *(px + 1);
#ifndef ARM_MATH_BIG_ENDIAN
x2 = __PKHBT(a, b, 16);
a = *(px + 2);
x3 = __PKHBT(b, a, 16);
x2 = __PKHBT(a, b, 16);
a = *(px + 2);
x3 = __PKHBT(b, a, 16);
#else
x2 = __PKHBT(b, a, 16);
a = *(px + 2);
x3 = __PKHBT(a, b, 16);
x2 = __PKHBT(b, a, 16);
a = *(px + 2);
x3 = __PKHBT(a, b, 16);
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
/* acc2 += x[2] * y[srcBLen - 1] + x[3] * y[srcBLen - 2] */
acc2 = __SMLADX(x2, c0, acc2);
@ -957,19 +957,19 @@ void arm_conv_fast_q15(
acc3 = __SMLADX(x3, c0, acc3);
/* Read y[srcBLen - 3] and y[srcBLen - 4] */
a = *py;
b = *(py+1);
py -= 2;
a = *py;
b = *(py+1);
py -= 2;
#ifndef ARM_MATH_BIG_ENDIAN
c0 = __PKHBT(a, b, 16);
c0 = __PKHBT(a, b, 16);
#else
c0 = __PKHBT(b, a, 16);;
c0 = __PKHBT(b, a, 16);
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
/* acc0 += x[2] * y[srcBLen - 3] + x[3] * y[srcBLen - 4] */
acc0 = __SMLADX(x2, c0, acc0);
@ -978,24 +978,24 @@ void arm_conv_fast_q15(
acc1 = __SMLADX(x3, c0, acc1);
/* Read x[4], x[5], x[6] */
a = *(px + 2);
b = *(px + 3);
a = *(px + 2);
b = *(px + 3);
#ifndef ARM_MATH_BIG_ENDIAN
x0 = __PKHBT(a, b, 16);
a = *(px + 4);
x1 = __PKHBT(b, a, 16);
x0 = __PKHBT(a, b, 16);
a = *(px + 4);
x1 = __PKHBT(b, a, 16);
#else
x0 = __PKHBT(b, a, 16);
a = *(px + 4);
x1 = __PKHBT(a, b, 16);
x0 = __PKHBT(b, a, 16);
a = *(px + 4);
x1 = __PKHBT(a, b, 16);
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
px += 4U;
px += 4U;
/* acc2 += x[4] * y[srcBLen - 3] + x[5] * y[srcBLen - 4] */
acc2 = __SMLADX(x0, c0, acc2);
@ -1028,19 +1028,19 @@ void arm_conv_fast_q15(
#endif /* #ifdef ARM_MATH_BIG_ENDIAN */
/* Read x[7] */
a = *px;
b = *(px+1);
px++;
a = *px;
b = *(px+1);
px++;
#ifndef ARM_MATH_BIG_ENDIAN
x3 = __PKHBT(a, b, 16);
x3 = __PKHBT(a, b, 16);
#else
x3 = __PKHBT(b, a, 16);;
x3 = __PKHBT(b, a, 16);
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
/* Perform the multiply-accumulates */
@ -1053,37 +1053,37 @@ void arm_conv_fast_q15(
if (k == 2U)
{
/* Read y[srcBLen - 5], y[srcBLen - 6] */
a = *py;
b = *(py+1);
a = *py;
b = *(py+1);
#ifndef ARM_MATH_BIG_ENDIAN
c0 = __PKHBT(a, b, 16);
c0 = __PKHBT(a, b, 16);
#else
c0 = __PKHBT(b, a, 16);;
c0 = __PKHBT(b, a, 16);
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
/* Read x[7], x[8], x[9] */
a = *px;
b = *(px + 1);
a = *px;
b = *(px + 1);
#ifndef ARM_MATH_BIG_ENDIAN
x3 = __PKHBT(a, b, 16);
a = *(px + 2);
x2 = __PKHBT(b, a, 16);
x3 = __PKHBT(a, b, 16);
a = *(px + 2);
x2 = __PKHBT(b, a, 16);
#else
x3 = __PKHBT(b, a, 16);
a = *(px + 2);
x2 = __PKHBT(a, b, 16);
x3 = __PKHBT(b, a, 16);
a = *(px + 2);
x2 = __PKHBT(a, b, 16);
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
px += 2U;
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
px += 2U;
/* Perform the multiply-accumulates */
acc0 = __SMLADX(x0, c0, acc0);
@ -1095,36 +1095,36 @@ void arm_conv_fast_q15(
if (k == 3U)
{
/* Read y[srcBLen - 5], y[srcBLen - 6] */
a = *py;
b = *(py+1);
a = *py;
b = *(py+1);
#ifndef ARM_MATH_BIG_ENDIAN
c0 = __PKHBT(a, b, 16);
c0 = __PKHBT(a, b, 16);
#else
c0 = __PKHBT(b, a, 16);;
c0 = __PKHBT(b, a, 16);
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
/* Read x[7], x[8], x[9] */
a = *px;
b = *(px + 1);
a = *px;
b = *(px + 1);
#ifndef ARM_MATH_BIG_ENDIAN
x3 = __PKHBT(a, b, 16);
a = *(px + 2);
x2 = __PKHBT(b, a, 16);
x3 = __PKHBT(a, b, 16);
a = *(px + 2);
x2 = __PKHBT(b, a, 16);
#else
x3 = __PKHBT(b, a, 16);
a = *(px + 2);
x2 = __PKHBT(a, b, 16);
x3 = __PKHBT(b, a, 16);
a = *(px + 2);
x2 = __PKHBT(a, b, 16);
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
/* Perform the multiply-accumulates */
acc0 = __SMLADX(x0, c0, acc0);
@ -1133,7 +1133,7 @@ void arm_conv_fast_q15(
acc3 = __SMLADX(x2, c0, acc3);
/* Read y[srcBLen - 7] */
c0 = *(py-1);
c0 = *(py-1);
#ifdef ARM_MATH_BIG_ENDIAN
c0 = c0 << 16U;
@ -1143,20 +1143,20 @@ void arm_conv_fast_q15(
#endif /* #ifdef ARM_MATH_BIG_ENDIAN */
/* Read x[10] */
a = *(px+2);
b = *(px+3);
a = *(px+2);
b = *(px+3);
#ifndef ARM_MATH_BIG_ENDIAN
x3 = __PKHBT(a, b, 16);
x3 = __PKHBT(a, b, 16);
#else
x3 = __PKHBT(b, a, 16);;
x3 = __PKHBT(b, a, 16);
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
px += 3U;
px += 3U;
/* Perform the multiply-accumulates */
acc0 = __SMLADX(x1, c0, acc0);
@ -1166,10 +1166,10 @@ void arm_conv_fast_q15(
}
/* Store the results in the accumulators in the destination buffer. */
*pOut++ = (q15_t)(acc0 >> 15);
*pOut++ = (q15_t)(acc1 >> 15);
*pOut++ = (q15_t)(acc2 >> 15);
*pOut++ = (q15_t)(acc3 >> 15);
*pOut++ = (q15_t)(acc0 >> 15);
*pOut++ = (q15_t)(acc1 >> 15);
*pOut++ = (q15_t)(acc2 >> 15);
*pOut++ = (q15_t)(acc3 >> 15);
/* Increment the pointer pIn1 index, count by 4 */
count += 4U;
@ -1318,7 +1318,7 @@ void arm_conv_fast_q15(
/* First part of the processing with loop unrolling. Compute 4 MACs at a time.
** a second loop below computes MACs for the remaining 1 to 3 samples. */
py++;
py++;
while (k > 0U)
{
@ -1390,7 +1390,7 @@ void arm_conv_fast_q15(
blockSize3--;
}
#endif /* #ifndef UNALIGNED_SUPPORT_DISABLE */
#endif /* #ifndef UNALIGNED_SUPPORT_DISABLE */
}
/**

View file

@ -312,7 +312,7 @@ arm_status arm_conv_partial_fast_q15(
x0 = *__SIMD32(px);
/* read x[1], x[2] samples */
x1 = _SIMD32_OFFSET(px+1);
px+= 2U;
px+= 2U;
/* Apply loop unrolling and compute 4 MACs simultaneously. */
@ -358,7 +358,7 @@ arm_status arm_conv_partial_fast_q15(
/* Read x[5], x[6] */
x1 = _SIMD32_OFFSET(px+3);
px += 4U;
px += 4U;
/* acc2 += x[4] * y[srcBLen - 3] + x[5] * y[srcBLen - 4] */
acc2 = __SMLADX(x0, c0, acc2);
@ -391,7 +391,7 @@ arm_status arm_conv_partial_fast_q15(
/* Read x[7] */
x3 = *__SIMD32(px);
px++;
px++;
/* Perform the multiply-accumulates */
acc0 = __SMLAD(x0, c0, acc0);
@ -410,7 +410,7 @@ arm_status arm_conv_partial_fast_q15(
/* Read x[9] */
x2 = _SIMD32_OFFSET(px+1);
px += 2U;
px += 2U;
/* Perform the multiply-accumulates */
acc0 = __SMLADX(x0, c0, acc0);
@ -436,7 +436,7 @@ arm_status arm_conv_partial_fast_q15(
acc2 = __SMLADX(x3, c0, acc2);
acc3 = __SMLADX(x2, c0, acc3);
c0 = *(py-1);
c0 = *(py-1);
#ifdef ARM_MATH_BIG_ENDIAN
c0 = c0 << 16U;
@ -447,7 +447,7 @@ arm_status arm_conv_partial_fast_q15(
/* Read x[10] */
x3 = _SIMD32_OFFSET(px+2);
px += 3U;
px += 3U;
/* Perform the multiply-accumulates */
acc0 = __SMLADX(x1, c0, acc0);
@ -875,7 +875,7 @@ arm_status arm_conv_partial_fast_q15(
/* First part of the processing with loop unrolling. Compute 4 MACs at a time.
** a second loop below computes MACs for the remaining 1 to 3 samples. */
py++;
py++;
while (k > 0U)
{
@ -967,22 +967,22 @@ arm_status arm_conv_partial_fast_q15(
acc3 = 0;
/* read x[0], x[1] samples */
a = *px++;
b = *px++;
a = *px++;
b = *px++;
#ifndef ARM_MATH_BIG_ENDIAN
x0 = __PKHBT(a, b, 16);
a = *px;
x1 = __PKHBT(b, a, 16);
x0 = __PKHBT(a, b, 16);
a = *px;
x1 = __PKHBT(b, a, 16);
#else
x0 = __PKHBT(b, a, 16);
a = *px;
x1 = __PKHBT(a, b, 16);
x0 = __PKHBT(b, a, 16);
a = *px;
x1 = __PKHBT(a, b, 16);
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
/* Apply loop unrolling and compute 4 MACs simultaneously. */
k = srcBLen >> 2U;
@ -993,19 +993,19 @@ arm_status arm_conv_partial_fast_q15(
{
/* Read the last two inputB samples using SIMD:
* y[srcBLen - 1] and y[srcBLen - 2] */
a = *py;
b = *(py+1);
py -= 2;
a = *py;
b = *(py+1);
py -= 2;
#ifndef ARM_MATH_BIG_ENDIAN
c0 = __PKHBT(a, b, 16);
c0 = __PKHBT(a, b, 16);
#else
c0 = __PKHBT(b, a, 16);;
c0 = __PKHBT(b, a, 16);
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
/* acc0 += x[0] * y[srcBLen - 1] + x[1] * y[srcBLen - 2] */
acc0 = __SMLADX(x0, c0, acc0);
@ -1013,22 +1013,22 @@ arm_status arm_conv_partial_fast_q15(
/* acc1 += x[1] * y[srcBLen - 1] + x[2] * y[srcBLen - 2] */
acc1 = __SMLADX(x1, c0, acc1);
a = *px;
b = *(px + 1);
a = *px;
b = *(px + 1);
#ifndef ARM_MATH_BIG_ENDIAN
x2 = __PKHBT(a, b, 16);
a = *(px + 2);
x3 = __PKHBT(b, a, 16);
x2 = __PKHBT(a, b, 16);
a = *(px + 2);
x3 = __PKHBT(b, a, 16);
#else
x2 = __PKHBT(b, a, 16);
a = *(px + 2);
x3 = __PKHBT(a, b, 16);
x2 = __PKHBT(b, a, 16);
a = *(px + 2);
x3 = __PKHBT(a, b, 16);
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
/* acc2 += x[2] * y[srcBLen - 1] + x[3] * y[srcBLen - 2] */
acc2 = __SMLADX(x2, c0, acc2);
@ -1037,19 +1037,19 @@ arm_status arm_conv_partial_fast_q15(
acc3 = __SMLADX(x3, c0, acc3);
/* Read y[srcBLen - 3] and y[srcBLen - 4] */
a = *py;
b = *(py+1);
py -= 2;
a = *py;
b = *(py+1);
py -= 2;
#ifndef ARM_MATH_BIG_ENDIAN
c0 = __PKHBT(a, b, 16);
c0 = __PKHBT(a, b, 16);
#else
c0 = __PKHBT(b, a, 16);;
c0 = __PKHBT(b, a, 16);
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
/* acc0 += x[2] * y[srcBLen - 3] + x[3] * y[srcBLen - 4] */
acc0 = __SMLADX(x2, c0, acc0);
@ -1058,24 +1058,24 @@ arm_status arm_conv_partial_fast_q15(
acc1 = __SMLADX(x3, c0, acc1);
/* Read x[4], x[5], x[6] */
a = *(px + 2);
b = *(px + 3);
a = *(px + 2);
b = *(px + 3);
#ifndef ARM_MATH_BIG_ENDIAN
x0 = __PKHBT(a, b, 16);
a = *(px + 4);
x1 = __PKHBT(b, a, 16);
x0 = __PKHBT(a, b, 16);
a = *(px + 4);
x1 = __PKHBT(b, a, 16);
#else
x0 = __PKHBT(b, a, 16);
a = *(px + 4);
x1 = __PKHBT(a, b, 16);
x0 = __PKHBT(b, a, 16);
a = *(px + 4);
x1 = __PKHBT(a, b, 16);
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
px += 4U;
px += 4U;
/* acc2 += x[4] * y[srcBLen - 3] + x[5] * y[srcBLen - 4] */
acc2 = __SMLADX(x0, c0, acc2);
@ -1108,19 +1108,19 @@ arm_status arm_conv_partial_fast_q15(
#endif /* #ifdef ARM_MATH_BIG_ENDIAN */
/* Read x[7] */
a = *px;
b = *(px+1);
px++;
a = *px;
b = *(px+1);
px++;
#ifndef ARM_MATH_BIG_ENDIAN
x3 = __PKHBT(a, b, 16);
x3 = __PKHBT(a, b, 16);
#else
x3 = __PKHBT(b, a, 16);;
x3 = __PKHBT(b, a, 16);
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
/* Perform the multiply-accumulates */
@ -1133,37 +1133,37 @@ arm_status arm_conv_partial_fast_q15(
if (k == 2U)
{
/* Read y[srcBLen - 5], y[srcBLen - 6] */
a = *py;
b = *(py+1);
a = *py;
b = *(py+1);
#ifndef ARM_MATH_BIG_ENDIAN
c0 = __PKHBT(a, b, 16);
c0 = __PKHBT(a, b, 16);
#else
c0 = __PKHBT(b, a, 16);;
c0 = __PKHBT(b, a, 16);
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
/* Read x[7], x[8], x[9] */
a = *px;
b = *(px + 1);
a = *px;
b = *(px + 1);
#ifndef ARM_MATH_BIG_ENDIAN
x3 = __PKHBT(a, b, 16);
a = *(px + 2);
x2 = __PKHBT(b, a, 16);
x3 = __PKHBT(a, b, 16);
a = *(px + 2);
x2 = __PKHBT(b, a, 16);
#else
x3 = __PKHBT(b, a, 16);
a = *(px + 2);
x2 = __PKHBT(a, b, 16);
x3 = __PKHBT(b, a, 16);
a = *(px + 2);
x2 = __PKHBT(a, b, 16);
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
px += 2U;
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
px += 2U;
/* Perform the multiply-accumulates */
acc0 = __SMLADX(x0, c0, acc0);
@ -1175,36 +1175,36 @@ arm_status arm_conv_partial_fast_q15(
if (k == 3U)
{
/* Read y[srcBLen - 5], y[srcBLen - 6] */
a = *py;
b = *(py+1);
a = *py;
b = *(py+1);
#ifndef ARM_MATH_BIG_ENDIAN
c0 = __PKHBT(a, b, 16);
c0 = __PKHBT(a, b, 16);
#else
c0 = __PKHBT(b, a, 16);;
c0 = __PKHBT(b, a, 16);
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
/* Read x[7], x[8], x[9] */
a = *px;
b = *(px + 1);
a = *px;
b = *(px + 1);
#ifndef ARM_MATH_BIG_ENDIAN
x3 = __PKHBT(a, b, 16);
a = *(px + 2);
x2 = __PKHBT(b, a, 16);
x3 = __PKHBT(a, b, 16);
a = *(px + 2);
x2 = __PKHBT(b, a, 16);
#else
x3 = __PKHBT(b, a, 16);
a = *(px + 2);
x2 = __PKHBT(a, b, 16);
x3 = __PKHBT(b, a, 16);
a = *(px + 2);
x2 = __PKHBT(a, b, 16);
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
/* Perform the multiply-accumulates */
acc0 = __SMLADX(x0, c0, acc0);
@ -1213,7 +1213,7 @@ arm_status arm_conv_partial_fast_q15(
acc3 = __SMLADX(x2, c0, acc3);
/* Read y[srcBLen - 7] */
c0 = *(py-1);
c0 = *(py-1);
#ifdef ARM_MATH_BIG_ENDIAN
c0 = c0 << 16U;
@ -1223,20 +1223,20 @@ arm_status arm_conv_partial_fast_q15(
#endif /* #ifdef ARM_MATH_BIG_ENDIAN */
/* Read x[10] */
a = *(px+2);
b = *(px+3);
a = *(px+2);
b = *(px+3);
#ifndef ARM_MATH_BIG_ENDIAN
x3 = __PKHBT(a, b, 16);
x3 = __PKHBT(a, b, 16);
#else
x3 = __PKHBT(b, a, 16);;
x3 = __PKHBT(b, a, 16);
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
px += 3U;
px += 3U;
/* Perform the multiply-accumulates */
acc0 = __SMLADX(x1, c0, acc0);
@ -1246,10 +1246,10 @@ arm_status arm_conv_partial_fast_q15(
}
/* Store the results in the accumulators in the destination buffer. */
*pOut++ = (q15_t)(acc0 >> 15);
*pOut++ = (q15_t)(acc1 >> 15);
*pOut++ = (q15_t)(acc2 >> 15);
*pOut++ = (q15_t)(acc3 >> 15);
*pOut++ = (q15_t)(acc0 >> 15);
*pOut++ = (q15_t)(acc1 >> 15);
*pOut++ = (q15_t)(acc2 >> 15);
*pOut++ = (q15_t)(acc3 >> 15);
/* Increment the pointer pIn1 index, count by 4 */
count += 4U;
@ -1399,7 +1399,7 @@ arm_status arm_conv_partial_fast_q15(
/* First part of the processing with loop unrolling. Compute 4 MACs at a time.
** a second loop below computes MACs for the remaining 1 to 3 samples. */
py++;
py++;
while (k > 0U)
{

View file

@ -272,7 +272,7 @@ void arm_correlate_fast_q15(
x0 = *__SIMD32(px);
/* read x[1], x[2] samples */
x1 = _SIMD32_OFFSET(px + 1);
px += 2U;
px += 2U;
/* Apply loop unrolling and compute 4 MACs simultaneously. */
k = srcBLen >> 2U;
@ -317,7 +317,7 @@ void arm_correlate_fast_q15(
/* Read x[5], x[6] */
x1 = _SIMD32_OFFSET(px + 3);
px += 4U;
px += 4U;
/* acc2 += x[4] * y[2] + x[5] * y[3] */
acc2 = __SMLAD(x0, c0, acc2);
@ -350,7 +350,7 @@ void arm_correlate_fast_q15(
/* Read x[7] */
x3 = *__SIMD32(px);
px++;
px++;
/* Perform the multiply-accumulates */
acc0 = __SMLAD(x0, c0, acc0);
@ -369,7 +369,7 @@ void arm_correlate_fast_q15(
/* Read x[9] */
x2 = _SIMD32_OFFSET(px + 1);
px += 2U;
px += 2U;
/* Perform the multiply-accumulates */
acc0 = __SMLAD(x0, c0, acc0);
@ -407,7 +407,7 @@ void arm_correlate_fast_q15(
/* Read x[10] */
x3 = _SIMD32_OFFSET(px + 2);
px += 3U;
px += 3U;
/* Perform the multiply-accumulates */
acc0 = __SMLADX(x1, c0, acc0);
@ -816,24 +816,24 @@ void arm_correlate_fast_q15(
acc3 = 0;
/* read x[0], x[1], x[2] samples */
a = *px;
b = *(px + 1);
a = *px;
b = *(px + 1);
#ifndef ARM_MATH_BIG_ENDIAN
x0 = __PKHBT(a, b, 16);
a = *(px + 2);
x1 = __PKHBT(b, a, 16);
x0 = __PKHBT(a, b, 16);
a = *(px + 2);
x1 = __PKHBT(b, a, 16);
#else
x0 = __PKHBT(b, a, 16);
a = *(px + 2);
x1 = __PKHBT(a, b, 16);
x0 = __PKHBT(b, a, 16);
a = *(px + 2);
x1 = __PKHBT(a, b, 16);
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
px += 2U;
px += 2U;
/* Apply loop unrolling and compute 4 MACs simultaneously. */
k = srcBLen >> 2U;
@ -844,18 +844,18 @@ void arm_correlate_fast_q15(
{
/* Read the first two inputB samples using SIMD:
* y[0] and y[1] */
a = *py;
b = *(py + 1);
a = *py;
b = *(py + 1);
#ifndef ARM_MATH_BIG_ENDIAN
c0 = __PKHBT(a, b, 16);
c0 = __PKHBT(a, b, 16);
#else
c0 = __PKHBT(b, a, 16);
c0 = __PKHBT(b, a, 16);
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
/* acc0 += x[0] * y[0] + x[1] * y[1] */
acc0 = __SMLAD(x0, c0, acc0);
@ -864,22 +864,22 @@ void arm_correlate_fast_q15(
acc1 = __SMLAD(x1, c0, acc1);
/* Read x[2], x[3], x[4] */
a = *px;
b = *(px + 1);
a = *px;
b = *(px + 1);
#ifndef ARM_MATH_BIG_ENDIAN
x2 = __PKHBT(a, b, 16);
a = *(px + 2);
x3 = __PKHBT(b, a, 16);
x2 = __PKHBT(a, b, 16);
a = *(px + 2);
x3 = __PKHBT(b, a, 16);
#else
x2 = __PKHBT(b, a, 16);
a = *(px + 2);
x3 = __PKHBT(a, b, 16);
x2 = __PKHBT(b, a, 16);
a = *(px + 2);
x3 = __PKHBT(a, b, 16);
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
/* acc2 += x[2] * y[0] + x[3] * y[1] */
acc2 = __SMLAD(x2, c0, acc2);
@ -888,20 +888,20 @@ void arm_correlate_fast_q15(
acc3 = __SMLAD(x3, c0, acc3);
/* Read y[2] and y[3] */
a = *(py + 2);
b = *(py + 3);
a = *(py + 2);
b = *(py + 3);
py += 4U;
py += 4U;
#ifndef ARM_MATH_BIG_ENDIAN
c0 = __PKHBT(a, b, 16);
c0 = __PKHBT(a, b, 16);
#else
c0 = __PKHBT(b, a, 16);
c0 = __PKHBT(b, a, 16);
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
/* acc0 += x[2] * y[2] + x[3] * y[3] */
acc0 = __SMLAD(x2, c0, acc0);
@ -910,24 +910,24 @@ void arm_correlate_fast_q15(
acc1 = __SMLAD(x3, c0, acc1);
/* Read x[4], x[5], x[6] */
a = *(px + 2);
b = *(px + 3);
a = *(px + 2);
b = *(px + 3);
#ifndef ARM_MATH_BIG_ENDIAN
x0 = __PKHBT(a, b, 16);
a = *(px + 4);
x1 = __PKHBT(b, a, 16);
x0 = __PKHBT(a, b, 16);
a = *(px + 4);
x1 = __PKHBT(b, a, 16);
#else
x0 = __PKHBT(b, a, 16);
a = *(px + 4);
x1 = __PKHBT(a, b, 16);
x0 = __PKHBT(b, a, 16);
a = *(px + 4);
x1 = __PKHBT(a, b, 16);
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
px += 4U;
px += 4U;
/* acc2 += x[4] * y[2] + x[5] * y[3] */
acc2 = __SMLAD(x0, c0, acc2);
@ -959,22 +959,22 @@ void arm_correlate_fast_q15(
#endif /* #ifdef ARM_MATH_BIG_ENDIAN */
/* Read x[7] */
a = *px;
b = *(px + 1);
a = *px;
b = *(px + 1);
px++;;
px++;
#ifndef ARM_MATH_BIG_ENDIAN
x3 = __PKHBT(a, b, 16);
x3 = __PKHBT(a, b, 16);
#else
x3 = __PKHBT(b, a, 16);
x3 = __PKHBT(b, a, 16);
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
px++;
px++;
/* Perform the multiply-accumulates */
acc0 = __SMLAD(x0, c0, acc0);
@ -986,38 +986,38 @@ void arm_correlate_fast_q15(
if (k == 2U)
{
/* Read y[4], y[5] */
a = *py;
b = *(py + 1);
a = *py;
b = *(py + 1);
#ifndef ARM_MATH_BIG_ENDIAN
c0 = __PKHBT(a, b, 16);
c0 = __PKHBT(a, b, 16);
#else
c0 = __PKHBT(b, a, 16);
c0 = __PKHBT(b, a, 16);
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
/* Read x[7], x[8], x[9] */
a = *px;
b = *(px + 1);
a = *px;
b = *(px + 1);
#ifndef ARM_MATH_BIG_ENDIAN
x3 = __PKHBT(a, b, 16);
a = *(px + 2);
x2 = __PKHBT(b, a, 16);
x3 = __PKHBT(a, b, 16);
a = *(px + 2);
x2 = __PKHBT(b, a, 16);
#else
x3 = __PKHBT(b, a, 16);
a = *(px + 2);
x2 = __PKHBT(a, b, 16);
x3 = __PKHBT(b, a, 16);
a = *(px + 2);
x2 = __PKHBT(a, b, 16);
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
px += 2U;
px += 2U;
/* Perform the multiply-accumulates */
acc0 = __SMLAD(x0, c0, acc0);
@ -1029,38 +1029,38 @@ void arm_correlate_fast_q15(
if (k == 3U)
{
/* Read y[4], y[5] */
a = *py;
b = *(py + 1);
a = *py;
b = *(py + 1);
#ifndef ARM_MATH_BIG_ENDIAN
c0 = __PKHBT(a, b, 16);
c0 = __PKHBT(a, b, 16);
#else
c0 = __PKHBT(b, a, 16);
c0 = __PKHBT(b, a, 16);
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
py += 2U;
py += 2U;
/* Read x[7], x[8], x[9] */
a = *px;
b = *(px + 1);
a = *px;
b = *(px + 1);
#ifndef ARM_MATH_BIG_ENDIAN
x3 = __PKHBT(a, b, 16);
a = *(px + 2);
x2 = __PKHBT(b, a, 16);
x3 = __PKHBT(a, b, 16);
a = *(px + 2);
x2 = __PKHBT(b, a, 16);
#else
x3 = __PKHBT(b, a, 16);
a = *(px + 2);
x2 = __PKHBT(a, b, 16);
x3 = __PKHBT(b, a, 16);
a = *(px + 2);
x2 = __PKHBT(a, b, 16);
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
/* Perform the multiply-accumulates */
acc0 = __SMLAD(x0, c0, acc0);
@ -1079,19 +1079,19 @@ void arm_correlate_fast_q15(
#endif /* #ifdef ARM_MATH_BIG_ENDIAN */
/* Read x[10] */
b = *(px + 3);
b = *(px + 3);
#ifndef ARM_MATH_BIG_ENDIAN
x3 = __PKHBT(a, b, 16);
x3 = __PKHBT(a, b, 16);
#else
x3 = __PKHBT(b, a, 16);
x3 = __PKHBT(b, a, 16);
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
#endif /* #ifndef ARM_MATH_BIG_ENDIAN */
px += 3U;
px += 3U;
/* Perform the multiply-accumulates */
acc0 = __SMLADX(x1, c0, acc0);