Donate to e Foundation | Murena handsets with /e/OS | Own a part of Murena! Learn more

Commit e1fee72c authored by Oscar Mateo's avatar Oscar Mateo Committed by Daniel Vetter
Browse files

drm/i915/bdw: Avoid non-lite-restore preemptions



In the current Execlists feeding mechanism, full preemption is not
supported yet: only lite-restores are allowed (this is: the GPU
simply samples a new tail pointer for the context currently in
execution).

But we have identified an scenario in which a full preemption occurs:
1) We submit two contexts for execution (A & B).
2) The GPU finishes with the first one (A), switches to the second one
(B) and informs us.
3) We submit B again (hoping to cause a lite restore) together with C,
but in the time we spend writing to the ELSP, the GPU finishes B.
4) The GPU start executing B again (since we told it so).
5) We receive a B finished interrupt and, mistakenly, we submit C (again)
and D, causing a full preemption of B.

The race is avoided by keeping track of how many times a context has been
submitted to the hardware and by better discriminating the received context
switch interrupts: in the example, when we have submitted B twice, we won´t
submit C and D as soon as we receive the notification that B is completed
because we were expecting to get a LITE_RESTORE and we didn´t, so we know a
second completion will be received shortly.

Without this explicit checking, somehow, the batch buffer execution order
gets messed with. This can be verified with the IGT test I sent together with
the series. I don´t know the exact mechanism by which the pre-emption messes
with the execution order but, since other people is working on the Scheduler
+ Preemption on Execlists, I didn´t try to fix it. In these series, only Lite
Restores are supported (other kind of preemptions WARN).

v2: elsp_submitted belongs in the new intel_ctx_submit_request. Several
rebase changes.

v3: Clarify how the race is avoided, as requested by Daniel.

Signed-off-by: default avatarOscar Mateo <oscar.mateo@intel.com>
Reviewed-by: default avatarDamien Lespiau <damien.lespiau@intel.com>
[danvet: Align function parameters ...]
Signed-off-by: default avatarDaniel Vetter <daniel.vetter@ffwll.ch>
parent e981e7b1
Loading
Loading
Loading
Loading
+25 −4
Original line number Original line Diff line number Diff line
@@ -268,6 +268,7 @@ static void execlists_context_unqueue(struct intel_engine_cs *ring)
		} else if (req0->ctx == cursor->ctx) {
		} else if (req0->ctx == cursor->ctx) {
			/* Same ctx: ignore first request, as second request
			/* Same ctx: ignore first request, as second request
			 * will update tail past first request's workload */
			 * will update tail past first request's workload */
			cursor->elsp_submitted = req0->elsp_submitted;
			list_del(&req0->execlist_link);
			list_del(&req0->execlist_link);
			queue_work(dev_priv->wq, &req0->work);
			queue_work(dev_priv->wq, &req0->work);
			req0 = cursor;
			req0 = cursor;
@@ -277,9 +278,15 @@ static void execlists_context_unqueue(struct intel_engine_cs *ring)
		}
		}
	}
	}


	WARN_ON(req1 && req1->elsp_submitted);

	WARN_ON(execlists_submit_context(ring, req0->ctx, req0->tail,
	WARN_ON(execlists_submit_context(ring, req0->ctx, req0->tail,
					 req1 ? req1->ctx : NULL,
					 req1 ? req1->ctx : NULL,
					 req1 ? req1->tail : 0));
					 req1 ? req1->tail : 0));

	req0->elsp_submitted++;
	if (req1)
		req1->elsp_submitted++;
}
}


static bool execlists_check_remove_request(struct intel_engine_cs *ring,
static bool execlists_check_remove_request(struct intel_engine_cs *ring,
@@ -298,11 +305,16 @@ static bool execlists_check_remove_request(struct intel_engine_cs *ring,
		struct drm_i915_gem_object *ctx_obj =
		struct drm_i915_gem_object *ctx_obj =
				head_req->ctx->engine[ring->id].state;
				head_req->ctx->engine[ring->id].state;
		if (intel_execlists_ctx_id(ctx_obj) == request_id) {
		if (intel_execlists_ctx_id(ctx_obj) == request_id) {
			WARN(head_req->elsp_submitted == 0,
			     "Never submitted head request\n");

			if (--head_req->elsp_submitted <= 0) {
				list_del(&head_req->execlist_link);
				list_del(&head_req->execlist_link);
				queue_work(dev_priv->wq, &head_req->work);
				queue_work(dev_priv->wq, &head_req->work);
				return true;
				return true;
			}
			}
		}
		}
	}


	return false;
	return false;
}
}
@@ -333,7 +345,16 @@ void intel_execlists_handle_ctx_events(struct intel_engine_cs *ring)
		status_id = I915_READ(RING_CONTEXT_STATUS_BUF(ring) +
		status_id = I915_READ(RING_CONTEXT_STATUS_BUF(ring) +
				(read_pointer % 6) * 8 + 4);
				(read_pointer % 6) * 8 + 4);


		if (status & GEN8_CTX_STATUS_COMPLETE) {
		if (status & GEN8_CTX_STATUS_PREEMPTED) {
			if (status & GEN8_CTX_STATUS_LITE_RESTORE) {
				if (execlists_check_remove_request(ring, status_id))
					WARN(1, "Lite Restored request removed from queue\n");
			} else
				WARN(1, "Preemption without Lite Restore\n");
		}

		 if ((status & GEN8_CTX_STATUS_ACTIVE_IDLE) ||
		     (status & GEN8_CTX_STATUS_ELEMENT_SWITCH)) {
			if (execlists_check_remove_request(ring, status_id))
			if (execlists_check_remove_request(ring, status_id))
				submit_contexts++;
				submit_contexts++;
		}
		}
+2 −0
Original line number Original line Diff line number Diff line
@@ -66,6 +66,8 @@ struct intel_ctx_submit_request {


	struct list_head execlist_link;
	struct list_head execlist_link;
	struct work_struct work;
	struct work_struct work;

	int elsp_submitted;
};
};


void intel_execlists_handle_ctx_events(struct intel_engine_cs *ring);
void intel_execlists_handle_ctx_events(struct intel_engine_cs *ring);