Sort all includes in src/gromacs

[alexxy/gromacs.git] / src / gromacs / simd / impl_x86_sse4_1 / impl_x86_sse4_1.h

diff --git a/src/gromacs/simd/impl_x86_sse4_1/impl_x86_sse4_1.h b/src/gromacs/simd/impl_x86_sse4_1/impl_x86_sse4_1.h
index ecb46b2447e9133c3999b4665587562adf95d810..8b8384a835a7e42cedcfeda76abea73d1bb9ba90 100644 (file)
--- a/src/gromacs/simd/impl_x86_sse4_1/impl_x86_sse4_1.h
+++ b/src/gromacs/simd/impl_x86_sse4_1/impl_x86_sse4_1.h
@@ -36,9 +36,11 @@
 #ifndef GMX_SIMD_IMPL_X86_SSE4_1_H
 #define GMX_SIMD_IMPL_X86_SSE4_1_H
 
+#include "config.h"
+
 #include <math.h>
-#include <smmintrin.h>
 
+#include <smmintrin.h>
 
 /* x86 SSE4.1 SIMD instruction wrappers
  *
@@ -96,7 +98,7 @@
 #define gmx_simd4_dotproduct3_f   gmx_simd4_dotproduct3_f_sse4_1
 
 /* SIMD reduction function */
-static gmx_inline float
+static gmx_inline float gmx_simdcall
 gmx_simd_reduce_f_sse4_1(__m128 a)
 {
     float  f;
@@ -108,7 +110,7 @@ gmx_simd_reduce_f_sse4_1(__m128 a)
 }
 
 /* SIMD4 Dotproduct helper function */
-static gmx_inline float
+static gmx_inline float gmx_simdcall
 gmx_simd4_dotproduct3_f_sse4_1(__m128 a, __m128 b)
 {
     float f;
@@ -116,7 +118,7 @@ gmx_simd4_dotproduct3_f_sse4_1(__m128 a, __m128 b)
     return f;
 }
 
-static gmx_inline double
+static gmx_inline double gmx_simdcall
 gmx_simd_reduce_d_sse4_1(__m128d a)
 {
     double  f;