Enable the clFlush.

We don't need to do anything now, as current it alwasy flushs
all the commands each time. We may need to revisit here after we
optmize the clEnqueueNDRangeKernel's behaviour and don't flush
pipeline every time.

Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com>
Tested-by: Lu, Guanqun <guanqun.lu@intel.com>

diff --git a/src/cl_api.c b/src/cl_api.c
index 61a00d1..88e1096 100644 (file)
--- a/src/cl_api.c
+++ b/src/cl_api.c
@@ -754,7 +754,10 @@ clGetEventProfilingInfo(cl_event             event,
 cl_int
 clFlush(cl_command_queue command_queue)
 {
-  NOT_IMPLEMENTED;
+  /* have nothing to do now, as currently
+   * clEnqueueNDRangeKernel will flush at
+   * the end of each calling. we may need
+   * to optimize it latter.*/
   return 0;
 }