I just checked the test case. This may not the best fix. The issue should be
the first time to access all the events which also inclues the uninitialized
event 3,4,5, And the following patch should be better.
- for (cl_uint i = 0; i != sizeof(ev) / sizeof(cl_event); ++i) {
+ for (cl_uint i = 0; i < 3; ++i) {
clGetEventInfo(ev[i], CL_EVENT_COMMAND_EXECUTION_STATUS, sizeof(status),
&status, NULL);
OCL_ASSERT(status >= CL_SUBMITTED);
}
Any thought?
On Mon, Sep 01, 2014 at 10:04:46AM +0800, [email protected] wrote:
> From: Junyan He <[email protected]>
>
> Signed-off-by: Junyan He <[email protected]>
> ---
> utests/runtime_barrier_list.cpp | 3 +++
> 1 file changed, 3 insertions(+)
>
> diff --git a/utests/runtime_barrier_list.cpp b/utests/runtime_barrier_list.cpp
> index 6987d5e..e176771 100644
> --- a/utests/runtime_barrier_list.cpp
> +++ b/utests/runtime_barrier_list.cpp
> @@ -1,3 +1,4 @@
> +#include <string.h>
> #include "utest_helper.hpp"
>
> #define BUFFERSIZE 32*1024
> @@ -10,6 +11,8 @@ void runtime_barrier_list(void)
> cl_int status = 0;
> cl_int value = 34;
>
> + memset(ev, 0, sizeof(cl_event)*5);
> +
> // Setup kernel and buffers
> OCL_CREATE_KERNEL("compiler_event");
> OCL_CREATE_BUFFER(buf[0], 0, BUFFERSIZE*sizeof(int), NULL);
> --
> 1.7.9.5
>
>
>
> _______________________________________________
> Beignet mailing list
> [email protected]
> http://lists.freedesktop.org/mailman/listinfo/beignet
_______________________________________________
Beignet mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/beignet
