a tbb example

codingTests/cplusplus/build.py

@@ -28,6 +28,7 @@ def shellCommand(command,errorMessage):
              TAU_MAKEFILE=/work/02463/srinathv/tau2/x86_64/lib/Makefile.tau-intelImpiCuda.aac-icpc-papi-mpi-cupti-pdt; \
              export TAU_OPTIONS="-optLinkOnly -optVerbose"'
 OMP_FLAG=' -DOMP -openmp '
+TBB_FLAG=' -DUSE_TBB -tbb '
 
 
 
@@ -55,6 +56,8 @@ def main():
 
   parser.add_argument('-o','--openmp',action="store_true",
                       help='Compile with OpenMP')
+  parser.add_argument('-tbb','--tbb',action="store_true",
+                      help='Compile with Intel Thread building blocks.')
   args = parser.parse_args()
 
   if (args.debug):
@@ -73,6 +76,8 @@ def main():
      IFDEF = IFDEF + '-DDEQ '
   if (args.numelems):
      IFDEF = IFDEF + '-DSET_N='+str(args.numelems)
+  if (args.tbb):
+     IFDEF = IFDEF + TBB_FLAG
   logging.debug('IFDEF is ' + IFDEF)
 
 

codingTests/cplusplus/testVec.cpp

@@ -6,6 +6,10 @@
 #ifdef USE_TAU
 #include <TAU.h>
 #endif
+#ifdef USE_TBB
+#include "tbb/concurrent_vector.h"
+#endif
+
 
 using std::vector;
 

codingTests/intelTbb/example1.cpp

@@ -0,0 +1,49 @@
+#include "tbb/blocked_range.h"
+#include "tbb/parallel_for.h"
+#include "tbb/task_scheduler_init.h"
+#include <iostream>
+#include <vector>
+
+struct mytask {
+  mytask(size_t n)
+    :_n(n)
+  {}
+  void operator()() {
+    for (int i=0;i<1000000;++i) {}  // Deliberately run slow
+    std::cerr << "[" << _n << "]";
+  }
+  size_t _n;
+};
+
+struct executor
+{
+  executor(std::vector<mytask>& t)
+    :_tasks(t)
+  {}
+  executor(executor& e,tbb::split)
+    :_tasks(e._tasks)
+  {}
+
+  void operator()(const tbb::blocked_range<size_t>& r) const {
+    for (size_t i=r.begin();i!=r.end();++i)
+      _tasks[i]();
+  }
+
+  std::vector<mytask>& _tasks;
+};
+
+int main(int,char**) {
+
+  tbb::task_scheduler_init init;  // Automatic number of threads
+//  tbb::task_scheduler_init init(2);  // Explicit number of threads
+
+  std::vector<mytask> tasks;
+  for (int i=0;i<1000;++i)
+     tasks.push_back(mytask(i));
+
+   executor exec(tasks);
+   tbb::parallel_for(tbb::blocked_range<size_t>(0,tasks.size()),exec);
+   std::cerr << std::endl;
+
+   return 0;
+}