From b7a4fa754aead0a75bcf133fc00509226dd99460 Mon Sep 17 00:00:00 2001
From: GLeonhardt <guileonhardt2@gmail.com>
Date: Wed, 21 Mar 2018 09:24:03 -0300
Subject: [PATCH 1/2] Work done

---
 Atividade2.c | 39 +++++++++++++++++++++++++++++++++++++++
 1 file changed, 39 insertions(+)
 create mode 100644 Atividade2.c

diff --git a/Atividade2.c b/Atividade2.c
new file mode 100644
index 0000000..d011715
--- /dev/null
+++ b/Atividade2.c
@@ -0,0 +1,39 @@
+#include <stdio.h>
+
+#include <immintrin.h> // AVX
+
+void add_intel_intrinsics(float *imgIn, float *imgOut, __m256 bright)
+{
+  __m256 vimgIn = _mm256_load_ps (imgIn);
+  __m256 vimgOut = _mm256_add_ps (vimgIn, bright);
+  _mm256_store_ps(imgOut, vimgOut); 
+}
+
+int main (int argc, char *argv[])
+{
+  float *imgIn = (float*)_mm_malloc (sizeof(float) * 128, 32); //aloca um vetor de 32 bytes (256 bits) alinhado em endereços múltiplos de 16 bytes.
+  float *imgOut = (float*)_mm_malloc (sizeof(float) * 128, 32);
+  __m256 bright = _mm256_set1_ps(20);
+  
+  int i = 0;
+  
+   for (i = 0; i < 128; ++i) {
+     imgIn[i] = i;
+   }
+  
+  printf("\nExecutando\n");
+
+    for(i = 0; i<=128-8; i+=8)
+    {
+      add_intel_intrinsics(&imgIn[i], &imgOut[i], bright);   
+    }
+   
+  
+  for (i = 0; i < 128; ++i) {
+    printf("%f\n", imgOut[i]);
+  }
+  
+  _mm_free(imgIn);
+  _mm_free(imgOut);
+}
+  
\ No newline at end of file

From 55866aa8827c2438f6bf3f8faf2c5e922bdeff8f Mon Sep 17 00:00:00 2001
From: GLeonhardt <guileonhardt2@gmail.com>
Date: Wed, 28 Mar 2018 09:44:23 -0300
Subject: [PATCH 2/2] produto escalar

---
 Atividade2.c      |  2 ++
 escalarParalelo.c | 48 +++++++++++++++++++++++++++++++++++++++++++++++
 produtoEscalar.c  | 44 +++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 94 insertions(+)
 create mode 100644 escalarParalelo.c
 create mode 100644 produtoEscalar.c

diff --git a/Atividade2.c b/Atividade2.c
index d011715..c3a5d27 100644
--- a/Atividade2.c
+++ b/Atividade2.c
@@ -2,6 +2,8 @@
 
 #include <immintrin.h> // AVX
 
+// gcc Atividade2.c -o atividade2 -mavx -O3
+
 void add_intel_intrinsics(float *imgIn, float *imgOut, __m256 bright)
 {
   __m256 vimgIn = _mm256_load_ps (imgIn);
diff --git a/escalarParalelo.c b/escalarParalelo.c
new file mode 100644
index 0000000..3958af1
--- /dev/null
+++ b/escalarParalelo.c
@@ -0,0 +1,48 @@
+#include <stdio.h>
+#include <stdlib.h>
+#include <time.h>
+#include <immintrin.h>
+// gcc escalarParalelo.c -o escalarParalelo -mavx -O3
+#define size  200000
+
+void produtoEscalar();
+
+int main(){
+    clock_t tic = clock();
+
+    produtoEscalar();
+
+    clock_t toc = clock();
+    
+    printf("Elapsed: %f seconds\n", (double)(toc - tic) / CLOCKS_PER_SEC);
+    
+    return 0;
+}
+
+void produtoEscalar()
+{
+    float *In = (float*)_mm_malloc (sizeof(float) * size, 32); //aloca um vetor de 32 bytes (256 bits) alinhado em endereços múltiplos de 16 bytes.
+    float *Out = (float*)_mm_malloc (sizeof(float) * size, 32);
+    float *mult = (float*)_mm_malloc (sizeof(float) * 8, 32);
+    mult[0] = 1;
+    mult[1] = 2;
+    mult[2] = 3;
+
+    int i=0, j=0, k=0;
+
+    for(i=0; i<size;i++)
+    {
+        Out[i] = 0;
+    }
+
+    __m256 vetIn = _mm256_load_ps(In);
+    __m256 vetOut = _mm256_load_ps(Out);
+    __m256 Vmult = _mm256_load_ps(mult);
+
+    for(i=0; i<size-8; i+=8){
+        for(j=0; j<3;j++){
+            vetOut += _mm256_mul_ps(vetIn, Vmult);
+            // _mm256_store_ps(imgOut, vimgOut); 
+        }
+    }
+}
\ No newline at end of file
diff --git a/produtoEscalar.c b/produtoEscalar.c
new file mode 100644
index 0000000..0f9e74e
--- /dev/null
+++ b/produtoEscalar.c
@@ -0,0 +1,44 @@
+// gcc produtoEscalar.c -o escalar
+
+#include <stdio.h>
+#include <stdlib.h>
+#include <time.h>
+
+#define size  200000
+
+void produtoEscalar();
+
+int main(){
+    clock_t tic = clock();
+
+    produtoEscalar();
+
+    clock_t toc = clock();
+    
+    printf("Elapsed: %f seconds\n", (double)(toc - tic) / CLOCKS_PER_SEC);
+    
+    return 0;
+}
+
+void produtoEscalar()
+{
+    int i, esc[3] = {2, 2, 2}, x=0;
+    long vetA[size]={0}, vetB[size]; 
+    srand(time(NULL));
+    for (i=0; i<size; i++){
+        vetA[i] = rand() % 10;
+        // printf("\n %d fak \n", vetA[i]); 
+    }
+
+    // printf("\n\n %d \n\n", x);
+
+    for (i=0; i<size-2; i++)
+    {
+        x=0;
+        x = vetA[i]*esc[0];
+        x += vetA[i+1]*esc[1];
+        x += vetA[i+2]*esc[2];
+        vetB[i+1] = x;
+        // printf("\n %d", vetB[i]);
+    }
+}