__m128d test_mm_cvtsi32_sd(__m128d A, int B) {
// DAG-LABEL: test_mm_cvtsi32_sd
// DAG: sitofp i32 %{{.*}} to double
// DAG: insertelement <2 x double> %{{.*}}, double %{{.*}}, i32 0
//
// ASM-LABEL: test_mm_cvtsi32_sd
// ASM: cvtsi2sdl
return _mm_cvtsi32_sd(A, B);
}
/** @brief Rounds floating-point number to the nearest integer not smaller than the original.
The function computes an integer i such that:
\f[i \le \texttt{value} < i+1\f]
@param value floating-point number. If the value is outside of INT_MIN ... INT_MAX range, the
result is not defined.
*/
CV_INLINE int cvCeil( double value )
{
#if (defined _MSC_VER && defined _M_X64 || (defined __GNUC__ && defined __SSE2__&& !defined __APPLE__)) && !defined(__CUDACC__)
__m128d t = _mm_set_sd( value );
int i = _mm_cvtsd_si32(t);
return i + _mm_movemask_pd(_mm_cmplt_sd(_mm_cvtsi32_sd(t,i), t));
#elif defined __GNUC__
int i = (int)value;
return i + (i < value);
#else
int i = cvRound(value);
float diff = (float)(i - value);
return i + (diff < 0);
#endif
}
F64 round(F64 val) {
#ifdef USE_SSE4
__m128d t = _mm_set_sd(val);
t = _mm_round_sd(t, t, _MM_FROUND_TO_NEAREST_INT);
_mm_store_sd(&val, t);
#elif defined(USE_SSE2)
__m128d t = _mm_set_sd(val);
U32 i = (U32)_mm_cvtsd_si32(t);
t = _mm_cvtsi32_sd(t, (int32)i);
_mm_store_sd(&val, t);
#else
val = core_floor(val + 0.5);
#endif
return val;
}
F64 ceil(F64 val) {
#ifdef USE_SSE4
__m128d t = _mm_set_sd(val);
t = _mm_ceil_sd(t, t);
_mm_store_sd(&val, t);
#elif defined(USE_SSE2)
val += 0.5;
__m128d t = _mm_set_sd(val);
U32 i = (U32)_mm_cvtsd_si32(t);
t = _mm_cvtsi32_sd(t, (int32)i);
_mm_store_sd(&val, t);
#else
val = core_ceil(val);
#endif
return val;
}
void convert_dword_to_double(double* result, UINT32 i) {
__m128d mmin, mmout;
mmin = _mm_setzero_pd();
mmout = _mm_cvtsi32_sd(mmin,i);
_mm_store_sd(result, mmout);
}
