From 00ec9a7db62b83b4fec59e32f113901c4a72b985 Mon Sep 17 00:00:00 2001
From: svkurowski <sascha.kurowski@rwth-aachen.de>
Date: Thu, 27 Nov 2014 10:15:20 +0100
Subject: [PATCH] Integrate CUDA into buildsystem and add example function

Former-commit-id: 392acb148a0033cf3cb24ed9f558a7c9b625935b
---
 CMakeLists.txt        | 45 +++++++++++++++++++++++++++++++++
 cuda/kernels/graph.cu | 58 +++++++++++++++++++++++++++++++++++++++++++
 cuda/kernels/graph.h  |  6 +++++
 storm-config.h.in     |  3 +++
 4 files changed, 112 insertions(+)
 create mode 100644 cuda/kernels/graph.cu
 create mode 100644 cuda/kernels/graph.h

diff --git a/CMakeLists.txt b/CMakeLists.txt
index 4c2d132b8..73c025b88 100644
--- a/CMakeLists.txt
+++ b/CMakeLists.txt
@@ -73,6 +73,12 @@ else()
     set(ENABLE_GUROBI ON)
 endif()
 
+if ("${CUDA_ROOT}" STREQUAL "")
+    set(ENABLE_CUDA OFF)
+else()
+    set(ENABLE_CUDA ON)
+endif()
+
 if ("${Z3_ROOT}" STREQUAL "")
     set(ENABLE_Z3 OFF)
 else()
@@ -195,6 +201,13 @@ else()
 	set(STORM_CPP_GUROBI_DEF "undef")
 endif()
 
+# CUDA Defines
+if (ENABLE_CUDA)
+	set(STORM_CPP_CUDA_DEF "define")
+else()
+	set(STORM_CPP_CUDA_DEF "undef")
+endif()
+
 # glpk defines
 set(STORM_CPP_GLPK_DEF "define")
 
@@ -227,6 +240,28 @@ configure_file (
 # Add the binary dir include directory for storm-config.h
 include_directories("${PROJECT_BINARY_DIR}/include")
 
+#############################################################
+##
+##	CUDA Library generation
+##
+#############################################################
+
+if(ENABLE_CUDA)
+	find_package(CUDA REQUIRED)
+
+	set(STORM_CUDA_LIB_NAME "storm-cuda")
+
+	file(GLOB_RECURSE STORM_CUDA_KERNEL_FILES ${PROJECT_SOURCE_DIR}/cuda/kernels/*.cu)
+
+	set(CUDA_PROPAGATE_HOST_FLAGS OFF)
+	set(CUDA_NVCC_FLAGS "-arch=sm_13")
+
+	cuda_add_library(${STORM_CUDA_LIB_NAME}
+	    ${STORM_CUDA_KERNEL_FILES}
+	    SHARED
+	)
+endif()
+
 #############################################################
 ##
 ##	Source file aggregation and clustering
@@ -387,6 +422,16 @@ if (ENABLE_GUROBI)
     target_link_libraries(storm-performance-tests "gurobi56")
 endif(ENABLE_GUROBI)
 
+#############################################################
+##
+##	CUDA (optional)
+##
+#############################################################
+if (ENABLE_CUDA)
+    message (STATUS "StoRM - Linking with CUDA")
+	target_link_libraries(storm ${STORM_CUDA_LIB_NAME})
+endif(ENABLE_CUDA)
+
 #############################################################
 ##
 ##	glpk
diff --git a/cuda/kernels/graph.cu b/cuda/kernels/graph.cu
new file mode 100644
index 000000000..c3d708b14
--- /dev/null
+++ b/cuda/kernels/graph.cu
@@ -0,0 +1,58 @@
+#include "cuda/kernels/graph.h"
+
+#include <stdio.h>
+ 
+const int N = 16; 
+const int blocksize = 16;
+
+#define gpuErrchk(ans) { gpuAssert((ans), __FILE__, __LINE__); }
+inline void gpuAssert(cudaError_t code, const char *file, int line, bool abort=true)
+{
+   if (code != cudaSuccess) 
+   {
+      fprintf(stderr,"GPUassert: %s %s %d\n", cudaGetErrorString(code), file, line);
+      if (abort) exit(code);
+   }
+}
+
+__global__ 
+void hello(char *a, int *b) 
+{
+    a[threadIdx.x] += b[threadIdx.x];
+}
+
+void helloWorldCuda()
+{
+    printf("CUDA TEST START\n");
+    printf("Should print \"Hello World\"\n");
+
+    char a[N] = "Hello \0\0\0\0\0\0";
+    int b[N] = {15, 10, 6, 0, -11, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0};
+    char c[N] = "YELLO \0\0\0\0\0\0";
+ 
+    char *ad;
+    int *bd;
+    const int csize = N*sizeof(char);
+    const int isize = N*sizeof(int);
+ 
+    printf("%s", a);
+ 
+    cudaMalloc( (void**)&ad, csize ); 
+    cudaMalloc( (void**)&bd, isize ); 
+    cudaMemcpy( ad, a, csize, cudaMemcpyHostToDevice ); 
+    cudaMemcpy( bd, b, isize, cudaMemcpyHostToDevice ); 
+    
+    dim3 dimBlock( blocksize, 1 );
+    dim3 dimGrid( 1, 1 );
+    hello<<<dimGrid, dimBlock>>>(ad, bd);
+
+    gpuErrchk( cudaPeekAtLastError() );
+    gpuErrchk( cudaDeviceSynchronize() );
+
+    cudaMemcpy( c, ad, csize, cudaMemcpyDeviceToHost ); 
+    cudaFree( ad );
+    cudaFree( bd );
+    
+    printf("%s\n", c);
+    printf("CUDA TEST END\n");
+}
diff --git a/cuda/kernels/graph.h b/cuda/kernels/graph.h
new file mode 100644
index 000000000..8f9cf8e8b
--- /dev/null
+++ b/cuda/kernels/graph.h
@@ -0,0 +1,6 @@
+#ifndef CUDA_KERNELS_GRAPH_H
+#define	CUDA_KERNELS_GRAPH_H
+
+void helloWorldCuda();
+
+#endif /* CUDA_KERNELS_GRAPH_H */
\ No newline at end of file
diff --git a/storm-config.h.in b/storm-config.h.in
index cb90e5af6..40db93554 100644
--- a/storm-config.h.in
+++ b/storm-config.h.in
@@ -17,6 +17,9 @@
 // Whether Gurobi is available and to be used (define/undef)
 #@STORM_CPP_GUROBI_DEF@ STORM_HAVE_GUROBI
 
+// Whether CUDA is available (define/undef)
+#@STORM_CPP_CUDA_DEF@ STORM_HAVE_CUDA
+
 // Whether GLPK is available and to be used (define/undef)
 #@STORM_CPP_GLPK_DEF@ STORM_HAVE_GLPK