new 2D convolution benchmark

diff --git a/Robust/src/Benchmarks/Prefetch/2DConv/dsm/Convolution.java b/Robust/src/Benchmarks/Prefetch/2DConv/dsm/Convolution.java
new file mode 100644 (file)
index 0000000..edcf3e6
--- /dev/null
+++ b/Robust/src/Benchmarks/Prefetch/2DConv/dsm/Convolution.java
@@ -0,0 +1,162 @@
+public class Convolution extends Thread {
+  Image img;
+  int x0,x1,y0,y1;
+
+  public Convolution(Image img, int x0, int x1, int y0, int y1) {
+    this.img = img;
+    this.x0 = x0;
+    this.x1 = x1;
+    this.y0 = y0;
+    this.y1 = y1;
+  }
+
+  public void run() {
+    int kernelHeight = 5;
+    int kernelWidth = 5;
+    double[][] kernel = new double[kernelHeight][kernelWidth];
+
+    atomic {
+      initKernel(kernel);
+      double tempinput[][] = img.inputImage;
+      double tempout[][] = img.outputImage;
+
+      double tinput0[] = tempinput[x0];
+      double tinput1[] = tempinput[x0+1];
+      double tinput2[] = tempinput[x0+2];
+      double tinput3[] = tempinput[x0+3];
+      double tinput4[] = tempinput[x0+4];
+
+      for(int i=x0;i<x1;++i){
+        double tout[] = tempout[x0];
+        for(int j=y0;j<y1;++j){
+          tout[y0] = 0;
+          for(int b=0;b<kernelHeight;++b){
+            tout[y0] = tout[y0] + (tinput0[j+b] * kernel[0][b] + tinput1[j+b] * kernel[1][b] + tinput2[j+b]*kernel[2][b] +
+                tinput3[j+b]*kernel[3][b] + tinput4[j+b]*kernel[4][b]);
+          }
+        }
+        if(i != 4095) {
+          tinput0 = tinput1; tinput1=tinput2; tinput2=tinput3; tinput3=tinput4; tinput4=tempinput[i+5];
+        }
+      }
+    }
+  }
+
+  public static void main(String[] args) {
+    int SIZE = 256;
+    int NUM_THREADS = 1;
+    if(args.length>0) {
+      NUM_THREADS = Integer.parseInt(args[0]);
+      if(args.length>1)
+        SIZE = Integer.parseInt(args[1]);
+    }
+
+    int[] mid = new int[8];
+    mid[0] = (128<<24)|(195<<16)|(175<<8)|84; //dw-10
+    mid[1] = (128<<24)|(195<<16)|(175<<8)|85; //dw-11
+    mid[2] = (128<<24)|(195<<16)|(175<<8)|86; //dw-12
+    mid[3] = (128<<24)|(195<<16)|(175<<8)|87; //dw-13
+    mid[4] = (128<<24)|(195<<16)|(175<<8)|88; //dw-14
+    mid[5] = (128<<24)|(195<<16)|(175<<8)|89; //dw-15
+    mid[6] = (128<<24)|(195<<16)|(175<<8)|90; //dw-16
+    mid[7] = (128<<24)|(195<<16)|(175<<8)|91; //dw-17
+
+    Image img;
+    Convolution[] conv;
+    Convolution tmp;
+
+    int kernelHeight = 5;
+    int kernelWidth = 5;
+    atomic {
+      img = global new Image(SIZE,SIZE,kernelHeight,kernelWidth);
+      img.setValues();
+      conv = global new Convolution[NUM_THREADS];
+      int increment=SIZE/NUM_THREADS;
+      int base = 0;
+      for(int i = 0; i<NUM_THREADS; i++) {
+        if((i+1)==NUM_THREADS)
+          conv[i] = global new Convolution(img, base, SIZE, 0, SIZE);
+        else 
+          conv[i] = global new Convolution(img, base, base+increment, 0, SIZE);
+        base+=increment;
+      }
+    }
+
+    for(int i = 0; i <NUM_THREADS; i++) {
+      atomic {
+        tmp = conv[i];
+      }
+      tmp.start(mid[i]);
+    }
+
+    for(int i = 0; i < NUM_THREADS; i++) {
+      atomic {
+        tmp = conv[i];
+      }
+      tmp.join();
+    }
+    System.printString("Done!");
+  }
+
+  //define 5X5 Gaussian kernel
+  public void initKernel(double[][] kernel) {
+    kernel[0][0] = 1/256.0;
+    kernel[0][1] = 4/256.0;
+    kernel[0][2] = 6/256.0;
+    kernel[0][3] = 4/256.0;
+    kernel[0][4] = 1/256.0;
+    kernel[1][0] = 4/256.0;
+    kernel[1][1] = 16/256.0;
+    kernel[1][2] = 24/256.0;
+    kernel[1][3] = 16/256.0;
+    kernel[1][4] = 4/256.0;
+    kernel[2][0] = 6/256.0;
+    kernel[2][1] = 24/256.0;
+    kernel[2][2] = 36/256.0;
+    kernel[2][3] = 24/256.0;
+    kernel[2][4] = 6/256.0;
+    kernel[3][0] = 4/256.0;
+    kernel[3][1] = 16/256.0;
+    kernel[3][2] = 24/256.0;
+    kernel[3][3] = 16/256.0;
+    kernel[3][4] = 4/256.0;
+    kernel[4][0] = 1/256.0;
+    kernel[4][1] = 4/256.0;
+    kernel[4][2] = 6/256.0;
+    kernel[4][3] = 4/256.0;
+    kernel[4][4] = 1/256.0;
+  }
+}
+
+public class Image {
+  int width, height;
+  int kernelWidth, kernelHeight;
+  double[][] inputImage;
+  double[][] outputImage;
+
+  public Image(int width, int height, int kernelWidth, int kernelHeight) {
+    this.width = width;
+    this.height = height;
+    this.kernelWidth = kernelWidth;
+    this.kernelHeight = kernelHeight;
+    inputImage = global new double[height+kernelHeight-1][width+kernelWidth-1];
+    outputImage = global new double[height][width];
+  }
+
+  /* Create a valid image */
+  public void setValues() {
+    for (int i = 0; i < (height+kernelHeight - 1); i++) {
+      double ainput[] = inputImage[i];
+      for(int j = 0; j < (width+kernelWidth - 1); j++) {
+        ainput[j] = 256-j;
+      }
+    }
+
+    for (int i = 0; i < height; i++){
+      double aout[] = outputImage[i];
+      for(int j = 0; j < width; j++) {
+        aout[j] = 0;
+      }
+    }
+  }
+}

diff --git a/Robust/src/Benchmarks/Prefetch/2DConv/dsm/Makefile b/Robust/src/Benchmarks/Prefetch/2DConv/dsm/Makefile
new file mode 100644 (file)
index 0000000..4147536
--- /dev/null
+++ b/Robust/src/Benchmarks/Prefetch/2DConv/dsm/Makefile
@@ -0,0 +1,27 @@
+MAINCLASS=Convolution
+SRC=${MAINCLASS}.java
+FLAGS =-dsm -dsmcaching -prefetch -optimize -excprefetch Image.initKernel -excprefetch Image.setValues -excprefetch Convolution.main -trueprob 0.95 -mainclass ${MAINCLASS}
+FLAGS1=-dsm -optimize -mainclass ${MAINCLASS}
+
+default:
+       ../../../../buildscript ${FLAGS1} ${SRC} -o ${MAINCLASS}NPNC
+       ../../../../buildscript ${FLAGS} ${SRC} -o ${MAINCLASS}N
+       cp ${MAINCLASS}NPNC.bin ${MAINCLASS}1NPNC.bin
+       cp ${MAINCLASS}NPNC.bin ${MAINCLASS}2NPNC.bin
+       cp ${MAINCLASS}N.bin ${MAINCLASS}2.bin
+       cp ${MAINCLASS}NPNC.bin ${MAINCLASS}3NPNC.bin
+       cp ${MAINCLASS}N.bin ${MAINCLASS}3.bin
+       cp ${MAINCLASS}NPNC.bin ${MAINCLASS}4NPNC.bin
+       cp ${MAINCLASS}N.bin ${MAINCLASS}4.bin
+       cp ${MAINCLASS}NPNC.bin ${MAINCLASS}5NPNC.bin
+       cp ${MAINCLASS}N.bin ${MAINCLASS}5.bin
+       cp ${MAINCLASS}NPNC.bin ${MAINCLASS}6NPNC.bin
+       cp ${MAINCLASS}N.bin ${MAINCLASS}6.bin
+       cp ${MAINCLASS}NPNC.bin ${MAINCLASS}7NPNC.bin
+       cp ${MAINCLASS}N.bin ${MAINCLASS}7.bin
+       cp ${MAINCLASS}NPNC.bin ${MAINCLASS}8NPNC.bin
+       cp ${MAINCLASS}N.bin ${MAINCLASS}8.bin
+
+clean:
+       rm -rf tmpbuilddirectory
+       rm *.bin

diff --git a/Robust/src/Benchmarks/Prefetch/2DConv/javasingle/Convolution.java b/Robust/src/Benchmarks/Prefetch/2DConv/javasingle/Convolution.java
new file mode 100644 (file)
index 0000000..95ae9d7
--- /dev/null
+++ b/Robust/src/Benchmarks/Prefetch/2DConv/javasingle/Convolution.java
@@ -0,0 +1,140 @@
+public class Convolution {
+  Image img;
+  int x0,x1,y0,y1;
+
+  public Convolution(Image img, int x0, int x1, int y0, int y1) {
+    this.img = img;
+    this.x0 = x0;
+    this.x1 = x1;
+    this.y0 = y0;
+    this.y1 = y1;
+  }
+
+  public void run() {
+    int kernelHeight = 5;
+    int kernelWidth = 5;
+    double[][] kernel = new double[kernelHeight][kernelWidth];
+
+    initKernel(kernel);
+    double tempinput[][] = img.inputImage;
+    double tempout[][] = img.outputImage;
+
+    double tinput0[] = tempinput[x0];
+    double tinput1[] = tempinput[x0+1];
+    double tinput2[] = tempinput[x0+2];
+    double tinput3[] = tempinput[x0+3];
+    double tinput4[] = tempinput[x0+4];
+
+    for(int i=x0;i<x1;++i){
+      double tout[] = tempout[x0];
+      for(int j=y0;j<y1;++j){
+        tout[y0] = 0;
+        for(int b=0;b<kernelHeight;++b){
+          tout[y0] = tout[y0] + (tinput0[j+b] * kernel[0][b] + tinput1[j+b] * kernel[1][b] + tinput2[j+b]*kernel[2][b] +
+              tinput3[j+b]*kernel[3][b] + tinput4[j+b]*kernel[4][b]);
+        }
+      }
+      if(i != 4095) {
+        tinput0 = tinput1; tinput1=tinput2; tinput2=tinput3; tinput3=tinput4; tinput4=tempinput[i+5];
+      }
+    }
+  }
+
+  public static void main(String[] args) {
+    int SIZE = 256;
+    int NUM_THREADS = 1;
+    if(args.length>0) {
+      NUM_THREADS = Integer.parseInt(args[0]);
+      if(args.length>1)
+        SIZE = Integer.parseInt(args[1]);
+    }
+
+    Image img;
+    Convolution[] conv;
+    Convolution tmp;
+
+    int kernelHeight = 5;
+    int kernelWidth = 5;
+    img = new Image(SIZE,SIZE,kernelHeight,kernelWidth);
+    img.setValues();
+    conv = new Convolution[NUM_THREADS];
+    int increment=SIZE/NUM_THREADS;
+    int base = 0;
+    for(int i = 0; i<NUM_THREADS; i++) {
+      if((i+1)==NUM_THREADS)
+        conv[i] = new Convolution(img, base, SIZE, 0, SIZE);
+      else 
+        conv[i] = new Convolution(img, base, base+increment, 0, SIZE);
+      base+=increment;
+    }
+
+    for(int i = 0; i <NUM_THREADS; i++) {
+      tmp = conv[i];
+      tmp.run();
+    }
+
+    System.printString("Done!");
+  }
+
+  //define 5X5 Gaussian kernel
+  public void initKernel(double[][] kernel) {
+    kernel[0][0] = 1/256.0;
+    kernel[0][1] = 4/256.0;
+    kernel[0][2] = 6/256.0;
+    kernel[0][3] = 4/256.0;
+    kernel[0][4] = 1/256.0;
+    kernel[1][0] = 4/256.0;
+    kernel[1][1] = 16/256.0;
+    kernel[1][2] = 24/256.0;
+    kernel[1][3] = 16/256.0;
+    kernel[1][4] = 4/256.0;
+    kernel[2][0] = 6/256.0;
+    kernel[2][1] = 24/256.0;
+    kernel[2][2] = 36/256.0;
+    kernel[2][3] = 24/256.0;
+    kernel[2][4] = 6/256.0;
+    kernel[3][0] = 4/256.0;
+    kernel[3][1] = 16/256.0;
+    kernel[3][2] = 24/256.0;
+    kernel[3][3] = 16/256.0;
+    kernel[3][4] = 4/256.0;
+    kernel[4][0] = 1/256.0;
+    kernel[4][1] = 4/256.0;
+    kernel[4][2] = 6/256.0;
+    kernel[4][3] = 4/256.0;
+    kernel[4][4] = 1/256.0;
+  }
+}
+
+public class Image {
+  int width, height;
+  int kernelWidth, kernelHeight;
+  double[][] inputImage;
+  double[][] outputImage;
+
+  public Image(int width, int height, int kernelWidth, int kernelHeight) {
+    this.width = width;
+    this.height = height;
+    this.kernelWidth = kernelWidth;
+    this.kernelHeight = kernelHeight;
+    inputImage = new double[height+kernelHeight-1][width+kernelWidth-1];
+    outputImage = new double[height][width];
+  }
+
+  /* Create a valid image */
+  public void setValues() {
+    for (int i = 0; i < (height+kernelHeight - 1); i++) {
+      double ainput[] = inputImage[i];
+      for(int j = 0; j < (width+kernelWidth - 1); j++) {
+        ainput[j] = 256-j;
+      }
+    }
+
+    for (int i = 0; i < height; i++){
+      double aout[] = outputImage[i];
+      for(int j = 0; j < width; j++) {
+        aout[j] = 0;
+      }
+    }
+  }
+}

diff --git a/Robust/src/Benchmarks/Prefetch/2DConv/javasingle/makefile b/Robust/src/Benchmarks/Prefetch/2DConv/javasingle/makefile
new file mode 100644 (file)
index 0000000..df60825
--- /dev/null
+++ b/Robust/src/Benchmarks/Prefetch/2DConv/javasingle/makefile
@@ -0,0 +1,8 @@
+MAINCLASS=Convolution
+SRC=${MAINCLASS}.java 
+default:
+       ../../../../buildscript -optimize -mainclass ${MAINCLASS} ${SRC} -o ${MAINCLASS}
+
+clean:
+       rm -rf tmpbuilddirectory
+       rm *.bin