moebiusband/MD-Bench
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Update script

Browse Source 
Signed-off-by: Rafael Ravedutti <rafaelravedutti@gmail.com>
This commit is contained in:
Rafael Ravedutti
2022-12-21 18:04:18 +01:00
parent 8d5e10f635
commit e5c233e072
3 changed files with 30 additions and 18 deletions
Download Patch File Download Diff File Expand all files Collapse all files

11
common/includes/simd/avx512_float.h
View File

@@ -7,10 +7,13 @@
#include <stdlib.h>
#include <string.h>
#include <immintrin.h>
#include <zmmintrin.h>
#ifndef NO_ZMM_INTRIN
# include <zmmintrin.h>
#endif
#define MD_SIMD_FLOAT __m512
#define MD_SIMD_MASK __mmask16
#define MD_SIMD_FLOAT __m512
#define MD_SIMD_MASK __mmask16
#define MD_SIMD_INT __m256i
static inline MD_SIMD_FLOAT simd_broadcast(float scalar) { return _mm512_set1_ps(scalar); }
static inline MD_SIMD_FLOAT simd_zero() { return _mm512_set1_ps(0.0f); }
@@ -69,7 +72,7 @@ static inline MD_FLOAT simd_h_dual_incr_reduced_sum(float* m, MD_SIMD_FLOAT v0,
return _mm_cvtss_f32(t3);
}
inline void simd_h_decr(MD_FLOAT *m, MD_SIMD_FLOAT a) {
static inline void simd_h_decr(MD_FLOAT *m, MD_SIMD_FLOAT a) {
__m256 t;
a = _mm512_add_ps(a, _mm512_shuffle_f32x4(a, a, 0xee));
t = _mm256_load_ps(m);