On 11/04/2018 12:34, Chris Wilson wrote:
Quoting Chris Wilson (2018-04-11 11:39:29)
We can refine our current execlists->queue_priority if we inspect
ELSP rather than the head of the unsubmitted queue. Currently, we use
the unsubmitted queue and say that if a subsequent request is more
important than the current queue, we will rerun the submission tasklet
to evaluate the need for preemption. However, we only want to preempt if
we need to jump ahead of a currently executing request in ELSP. The
second reason for running the submission tasklet is amalgamate requests
into the active context on ELSP to avoid a stall when ELSP drains.
(Though repeatedly amalgamating requests into the active context and
triggering many lite-restore is off question gain, the goal really is to
put a context into ELSP to cover the interrupt.) So if instead of
looking at the head of the queue, we look at the context in ELSP we
can answer both of the questions more accurately -- we don't need to
rerun the submission tasklet unless our new request is important enough
to feed into, at least, ELSP.
v2: Add some comments from the discussion with Tvrtko.
v3: More commentary to cross-reference queue_request()
References: f6322eddaff7 ("drm/i915/preemption: Allow preemption between submission
Signed-off-by: Chris Wilson <ch...@chris-wilson.co.uk>
Cc: Michał Winiarski <michal.winiar...@intel.com>
Cc: Michel Thierry <michel.thie...@intel.com>
Cc: Mika Kuoppala <mika.kuopp...@linux.intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursu...@intel.com>
From the discussion thread,
Reviewed-by: Tvrtko Ursulin <tvrtko.ursu...@intel.com>
I assume Tvrtko is happy with the comments added based on that
Yes, all fine!
Intel-gfx mailing list