From 2bba9d187d32ce18fc0e752f423c637b10ab0335 Mon Sep 17 00:00:00 2001
From: Ismael Gomez <ismagom@gmail.com>
Date: Fri, 7 Jul 2017 16:36:27 +0200
Subject: [PATCH] fixed dotprodconj. Removed unaligned load/store

---
 lib/src/phy/utils/vector.c      |  2 +-
 lib/src/phy/utils/vector_simd.c | 16 ++++++++--------
 2 files changed, 9 insertions(+), 9 deletions(-)

diff --git a/lib/src/phy/utils/vector.c b/lib/src/phy/utils/vector.c
index 61d0271b6..8ceefe3da 100644
--- a/lib/src/phy/utils/vector.c
+++ b/lib/src/phy/utils/vector.c
@@ -608,7 +608,7 @@ cf_t srslte_vec_dot_prod_conj_ccc(cf_t *x, cf_t *y, uint32_t len) {
   uint32_t i;
   cf_t res = 0;
   for (i=0;i<len;i++) {
-    res += x[i]*y[i];
+    res += x[i]*conjf(y[i]);
   }
   return res;
 #else
diff --git a/lib/src/phy/utils/vector_simd.c b/lib/src/phy/utils/vector_simd.c
index c7bf27bf5..abbd2f7ff 100644
--- a/lib/src/phy/utils/vector_simd.c
+++ b/lib/src/phy/utils/vector_simd.c
@@ -206,12 +206,12 @@ void srslte_vec_sub_sss_sse(short *x, short *y, short *z, uint32_t len)
   __m128i xVal, yVal, zVal;
   for(;number < points; number++){
 
-    xVal = _mm_loadu_si128(xPtr);
-    yVal = _mm_loadu_si128(yPtr);
+    xVal = _mm_load_si128(xPtr);
+    yVal = _mm_load_si128(yPtr);
 
     zVal = _mm_sub_epi16(xVal, yVal);
 
-    _mm_storeu_si128(zPtr, zVal); 
+    _mm_store_si128(zPtr, zVal);
 
     xPtr ++;
     yPtr ++;
@@ -273,12 +273,12 @@ void srslte_vec_prod_sss_sse(short *x, short *y, short *z, uint32_t len)
   __m128i xVal, yVal, zVal;
   for(;number < points; number++){
 
-    xVal = _mm_loadu_si128(xPtr);
-    yVal = _mm_loadu_si128(yPtr);
+    xVal = _mm_load_si128(xPtr);
+    yVal = _mm_load_si128(yPtr);
 
     zVal = _mm_mullo_epi16(xVal, yVal);
 
-    _mm_storeu_si128(zPtr, zVal); 
+    _mm_store_si128(zPtr, zVal);
 
     xPtr ++;
     yPtr ++;
@@ -341,11 +341,11 @@ void srslte_vec_sc_div2_sss_sse(short *x, int k, short *z, uint32_t len)
   __m128i xVal, zVal;
   for(;number < points; number++){
 
-    xVal = _mm_loadu_si128(xPtr);
+    xVal = _mm_load_si128(xPtr);
     
     zVal = _mm_srai_epi16(xVal, k);                 
       
-    _mm_storeu_si128(zPtr, zVal); 
+    _mm_store_si128(zPtr, zVal);
 
     xPtr ++;
     zPtr ++;