extern void lud_omp(
_IEEE32 * a,
_INT32 size)
{
register _INT32 _w2c_trip_count;
register _INT32 _w2c___ompv_ok_to_fork;
register _UINT64 _w2c_reg3;
register _INT32 _w2c_trip_count0;
register _INT32 _w2c___ompv_ok_to_fork0;
_INT32 __localized_common_j;
_IEEE32 __localized_common_sum;
_INT32 __localized_common_k;
_INT32 _w2c___localized_common_j0;
_IEEE32 _w2c___localized_common_sum0;
_INT32 _w2c___localized_common_k0;
_INT32 __ompv_gtid_s1;
/*Begin_of_nested_PU(s)*/
a_g = a;
size_g = size;
i = 0;
while(size_g > i)
{
_514 :;
_w2c_trip_count = size_g - i;
_w2c___ompv_ok_to_fork = _w2c_trip_count > 1;
if(_w2c___ompv_ok_to_fork)
{
_w2c___ompv_ok_to_fork = __ompc_can_fork();
}
if(_w2c___ompv_ok_to_fork)
{
__ompc_fork(0, &__ompdo_lud_omp_1, _w2c_reg3);
}
else
{
__ompv_gtid_s1 = __ompc_get_local_thread_num();
__ompc_serialized_parallel();
for(__localized_common_j = i; __localized_common_j < size_g; __localized_common_j = __localized_common_j + 1)
{
__localized_common_sum = *(a_g + (_INT64)(__localized_common_j + (size_g * i)));
__localized_common_k = 0;
while(__localized_common_k < i)
{
_1026 :;
__localized_common_sum = __localized_common_sum - (*(a_g + (_INT64)(__localized_common_j + (__localized_common_k * size_g))) ** (a_g + (_INT64)(__localized_common_k + (size_g * i))));
__localized_common_k = __localized_common_k + 1;
_770 :;
}
goto _1282;
_1282 :;
* (a_g + (_INT64)(__localized_common_j + (size_g * i))) = __localized_common_sum;
}
__ompc_end_serialized_parallel();
}
_w2c_trip_count0 = (size_g - i) + -1;
_w2c___ompv_ok_to_fork0 = _w2c_trip_count0 > 1;
if(_w2c___ompv_ok_to_fork0)
{
_w2c___ompv_ok_to_fork0 = __ompc_can_fork();
}
if(_w2c___ompv_ok_to_fork0)
{
__ompc_fork(0, &__ompdo_lud_omp_2, _w2c_reg3);
}
else
{
__ompv_gtid_s1 = __ompc_get_local_thread_num();
__ompc_serialized_parallel();
for(_w2c___localized_common_j0 = i + 1; _w2c___localized_common_j0 < size_g; _w2c___localized_common_j0 = _w2c___localized_common_j0 + 1)
{
_w2c___localized_common_sum0 = *(a_g + (_INT64)(i + (_w2c___localized_common_j0 * size_g)));
_w2c___localized_common_k0 = 0;
while(_w2c___localized_common_k0 < i)
{
_1794 :;
_w2c___localized_common_sum0 = _w2c___localized_common_sum0 - (*(a_g + (_INT64)(_w2c___localized_common_k0 + (_w2c___localized_common_j0 * size_g))) ** (a_g + (_INT64)(i + (_w2c___localized_common_k0 * size_g))));
_w2c___localized_common_k0 = _w2c___localized_common_k0 + 1;
_1538 :;
}
goto _2050;
_2050 :;
* (a_g + (_INT64)(i + (_w2c___localized_common_j0 * size_g))) = _w2c___localized_common_sum0 / *(a_g + (_INT64)(i * (size_g + 1)));
}
__ompc_end_serialized_parallel();
}
i = i + 1;
_258 :;
}
goto _2306;
_2306 :;
return;
} /* lud_omp */
OMPT_API ompt_thread_id_t ompt_get_thread_id(void) {
return __ompc_get_local_thread_num() + 1;
// return __omp_myid + 1;
}
extern void fft_float(
_UINT32 __NumSamples,
_INT32 __InverseTransform,
_IEEE32 * __RealIn,
_IEEE32 * __ImagIn,
_IEEE32 * __RealOut,
_IEEE32 * __ImagOut)
{
register _INT32 _w2c___comma;
register _UINT32 _w2c___comma0;
register _INT32 _w2c___ompv_ok_to_fork;
register _UINT64 _w2c_reg3;
register _UINT32 _w2c___comma1;
register _IEEE32 _w2c___cselect;
register _IEEE64 _w2c___comma2;
register _IEEE64 _w2c___comma3;
register _IEEE64 _w2c___comma4;
register _INT32 _w2c_trip_count;
register _INT32 _w2c___ompv_ok_to_fork0;
_IEEE64 delta_angle;
_IEEE64 denom;
_INT32 __localized_common_i;
_INT32 __localized_common_j;
_INT32 _w2c___localized_common_i0;
_IEEE64 __localized_common_ar[3LL];
_IEEE64 __localized_common_ai[3LL];
_INT32 _w2c___localized_common_j0;
_INT32 __localized_common_n;
_INT32 __ompv_gtid_s1;
/*Begin_of_nested_PU(s)*/
NumSamples = __NumSamples;
InverseTransform = __InverseTransform;
* RealIn = *__RealIn;
* ImagIn = *__ImagIn;
* RealOut = *__RealOut;
* ImagOut = *__ImagOut;
_w2c___comma = IsPowerOfTwo(NumSamples);
if(_w2c___comma == 0)
{
printf("Error in fft(): NumSamples=%u is not power of two\n", NumSamples);
exit(1);
}
if(InverseTransform != 0)
{
angle_numerator = -angle_numerator;
}
CheckPointer(RealIn, "RealIn");
CheckPointer(RealOut, "RealOut");
CheckPointer(ImagOut, "ImagOut");
_w2c___comma0 = NumberOfBitsNeeded(NumSamples);
NumBits = _w2c___comma0;
_w2c___ompv_ok_to_fork = 1;
if(_w2c___ompv_ok_to_fork)
{
_w2c___ompv_ok_to_fork = __ompc_can_fork();
}
if(_w2c___ompv_ok_to_fork)
{
__ompc_fork(0, &__omprg_fft_float_1, _w2c_reg3);
}
else
{
__ompv_gtid_s1 = __ompc_get_local_thread_num();
__ompc_serialized_parallel();
for(__localized_common_i = 0; __localized_common_i < (_INT32) NumSamples; __localized_common_i = __localized_common_i + 1)
{
_w2c___comma1 = ReverseBits((_UINT32) __localized_common_i, NumBits);
__localized_common_j = (_INT32)(_w2c___comma1);
* (RealOut + (_UINT64)((_UINT64) __localized_common_j)) = *(RealIn + (_UINT64)((_UINT64) __localized_common_i));
if((_UINT64)(ImagIn) != 0ULL)
{
_w2c___cselect = *(ImagIn + (_UINT64)((_UINT64) __localized_common_i));
}
else
{
_w2c___cselect = 0.0F;
}
* (ImagOut + (_UINT64)((_UINT64) __localized_common_j)) = _w2c___cselect;
}
__ompc_end_serialized_parallel();
}
_w2c___comma2 = log2((_IEEE64)(NumSamples));
NumIter = _U4F8TRUNC(_w2c___comma2);
m = 1;
while(NumIter >= (_UINT32) m)
{
_514 :;
_w2c___comma3 = pow(2.0, (_IEEE64)(m));
BlockSize = _U4F8TRUNC(_w2c___comma3);
_w2c___comma4 = pow(2.0, (_IEEE64)(m + -1));
BlockEnd = _U4F8TRUNC(_w2c___comma4);
delta_angle = angle_numerator / (_IEEE64)(BlockSize);
sm2 = sin(delta_angle * -2.0);
sm1 = sin(-delta_angle);
cm2 = cos(delta_angle * -2.0);
cm1 = cos(delta_angle);
w = cm1 * 2.0;
_w2c_trip_count = (((_INT32) NumSamples + (_INT32) BlockSize) + -1) / (_INT32) BlockSize;
_w2c___ompv_ok_to_fork0 = _w2c_trip_count > 1;
if(_w2c___ompv_ok_to_fork0)
{
_w2c___ompv_ok_to_fork0 = __ompc_can_fork();
}
if(_w2c___ompv_ok_to_fork0)
{
__ompc_fork(0, &__ompdo_fft_float_11, _w2c_reg3);
}
else
{
__ompv_gtid_s1 = __ompc_get_local_thread_num();
__ompc_serialized_parallel();
for(_w2c___localized_common_i0 = 0; _w2c___localized_common_i0 < (_INT32) NumSamples; _w2c___localized_common_i0 = _w2c___localized_common_i0 + (_INT32) BlockSize)
{
(__localized_common_ar)[2] = cm2;
(__localized_common_ar)[1] = cm1;
(__localized_common_ai)[2] = sm2;
(__localized_common_ai)[1] = sm1;
_w2c___localized_common_j0 = _w2c___localized_common_i0;
__localized_common_n = 0;
while((_UINT32) __localized_common_n < BlockEnd)
{
_1026 :;
(__localized_common_ar)[0] = ((__localized_common_ar)[1] * w) - (__localized_common_ar)[2];
(__localized_common_ar)[2] = (__localized_common_ar)[1];
(__localized_common_ar)[1] = (__localized_common_ar)[0];
(__localized_common_ai)[0] = ((__localized_common_ai)[1] * w) - (__localized_common_ai)[2];
(__localized_common_ai)[2] = (__localized_common_ai)[1];
(__localized_common_ai)[1] = (__localized_common_ai)[0];
k = (_INT32)((_UINT32) _w2c___localized_common_j0 + BlockEnd);
tr = ((_IEEE64)(*(RealOut + (_UINT64)((_UINT64) k))) * (__localized_common_ar)[0]) - ((_IEEE64)(*(ImagOut + (_UINT64)((_UINT64) k))) * (__localized_common_ai)[0]);
ti = ((_IEEE64)(*(RealOut + (_UINT64)((_UINT64) k))) * (__localized_common_ai)[0]) + ((_IEEE64)(*(ImagOut + (_UINT64)((_UINT64) k))) * (__localized_common_ar)[0]);
* (RealOut + (_UINT64)((_UINT64) k)) = (_IEEE32)((_IEEE64)(*(RealOut + (_UINT64)((_UINT64) _w2c___localized_common_j0))) - tr);
* (ImagOut + (_UINT64)((_UINT64) k)) = (_IEEE32)((_IEEE64)(*(ImagOut + (_UINT64)((_UINT64) _w2c___localized_common_j0))) - ti);
* (RealOut + (_UINT64)((_UINT64) _w2c___localized_common_j0)) = (_IEEE32)((_IEEE64)(*(RealOut + (_UINT64)((_UINT64) _w2c___localized_common_j0))) + tr);
* (ImagOut + (_UINT64)((_UINT64) _w2c___localized_common_j0)) = (_IEEE32)((_IEEE64)(*(ImagOut + (_UINT64)((_UINT64) _w2c___localized_common_j0))) + ti);
_w2c___localized_common_j0 = _w2c___localized_common_j0 + 1;
__localized_common_n = __localized_common_n + 1;
_770 :;
}
goto _1282;
_1282 :;
}
__ompc_end_serialized_parallel();
}
m = m + 1;
_258 :;
}
goto _1538;
_1538 :;
if(InverseTransform != 0)
{
denom = (_IEEE64)(NumSamples);
i = 0;
while(NumSamples > (_UINT32) i)
{
_2050 :;
* (RealOut + (_UINT64)((_UINT64) i)) = (_IEEE32)((_IEEE64)(*(RealOut + (_UINT64)((_UINT64) i))) / denom);
* (ImagOut + (_UINT64)((_UINT64) i)) = (_IEEE32)((_IEEE64)(*(ImagOut + (_UINT64)((_UINT64) i))) / denom);
i = i + 1;
_1794 :;
}
_2306 :;
}
return;
} /* fft_float */
