Lines Matching full:semaphores

642 	 * Are we using semaphores when the gpu is already saturated?  in __i915_request_submit()
644 	 * Using semaphores incurs a cost in having the GPU poll a  in __i915_request_submit()
652 	 * system is overloaded and using semaphores at this time only  in __i915_request_submit()
654 	 * further use of semaphores until we are idle again, whence we  in __i915_request_submit()
657 	if (request->sched.semaphores &&  in __i915_request_submit()
659 		engine->saturated |= request->sched.semaphores;  in __i915_request_submit()
736 	if (request->sched.semaphores && __i915_request_has_started(request))  in __i915_request_unsubmit()
737 		request->sched.semaphores = 0;  in __i915_request_unsubmit()
1139 	return rq->sched.semaphores | READ_ONCE(rq->engine->saturated);  in already_busywaiting()
1219 	 * semaphores as they bypass the fence signaling metadata, and we  in emit_semaphore_wait()
1239 	to->sched.semaphores |= mask;  in emit_semaphore_wait()
1295 	 * Ensure both start together [after all semaphores in signal]  in __i915_request_await_execution()
1337 	 * The downside of using semaphores is that we lose metadata passing  in mark_external()
1751 	 * precludes optimising to use semaphores serialisation of a single  in __i915_request_add_to_timeline()
2206 	 *    - a ready request may still need to wait on semaphores  in i915_request_show()