Hi,
When are out-of-order command queues going to be allowed?
Am I right in thinking that at the moment, you need multiple command queues to run native kernels in parallel?
eg. clEnqueueNativeKernel(queues[0], ..., &events[0]); clEnqueueNativeKernel(queues[1], ..., &events[1]); clEnqueueNativeKernel(queues[2], ..., &events[2]); clEnqueueNativeKernel(queues[3], ..., &events[3]); clWaitForEvents(4, events);