Complete the feature of clGetEventProfilingInfo API

The profiling feature is now all supported. We use
drm_intel_reg_read to get the current time of GPU
when the event is queued and submitted, and use
PIPI_CONTROL cmd to get the executing time of the
GPU for kernel start and end.
One trivial problem is that:
The GPU timer counter is 36 bits with resolution of
80ns, so 2^36*80 = 5500s, about half an hour.
Some test may last about 2~5 min and if it starts at
about half an hour, this may cause a wrap back problem
and cause the case fail.

Signed-off-by: Junyan He <junyan.he@linux.intel.com>
Reviewed-by: Zhigang Gong <zhigang.gong@linux.intel.com>

1 files changed, 2 insertions, 1 deletions

diff --git a/src/cl_event.h b/src/cl_event.h
index 722486a6..3c611109 100644
--- a/src/cl_event.h
+++ b/src/cl_event.h
@@ -68,6 +68,7 @@ struct _cl_event {
   enqueue_callback*  enqueue_cb;  /* This event's enqueue */
   enqueue_callback*  waits_head;  /* The head of enqueues list wait on this event */
   cl_bool            emplict;     /* Identify this event whether created by api emplict*/
+  cl_ulong           timestamp[4];/* The time stamps for profiling. */
 };
 
 /* Create a new event object */
@@ -91,6 +92,6 @@ void cl_event_update_status(cl_event);
 /* Create the marker event */
 cl_int cl_event_marker(cl_command_queue, cl_event*);
 /* Do the event profiling */
-cl_int cl_event_profiling(cl_event event, cl_profiling_info param_name, cl_ulong *ret_val);
+cl_int cl_event_get_timestamp(cl_event event, cl_profiling_info param_name);
 #endif /* __CL_EVENT_H__ */