| Message ID | 20170717084235.2982-2-chris@chris-wilson.co.uk (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
On Mon, Jul 17, 2017 at 09:42:33AM +0100, Chris Wilson wrote: > With preemption, we will want to "unsubmit" a request, taking it back > from the hw and returning it to the priority sorted execution list. In > order to know where to insert it into that list, we need to remember > its adjust priority (which may change even as it was being executed). > > Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> > Cc: Michal Winiarski <michal.winiarski@intel.com> We should also change GuC dequeue/irq_handler. With that: Reviewed-by: Michał Winiarski <michal.winiarski@intel.com> -Michał > --- > drivers/gpu/drm/i915/intel_lrc.c | 3 +-- > 1 file changed, 1 insertion(+), 2 deletions(-) > > diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c > index 1b2f0e3d383a..8ab0c4b76c98 100644 > --- a/drivers/gpu/drm/i915/intel_lrc.c > +++ b/drivers/gpu/drm/i915/intel_lrc.c > @@ -552,8 +552,6 @@ static void execlists_dequeue(struct intel_engine_cs *engine) > } > > INIT_LIST_HEAD(&rq->priotree.link); > - rq->priotree.priority = INT_MAX; > - > __i915_gem_request_submit(rq); > trace_i915_gem_request_in(rq, port_index(port, engine)); > last = rq; > @@ -687,6 +685,7 @@ static void intel_lrc_irq_handler(unsigned long data) > execlists_context_status_change(rq, INTEL_CONTEXT_SCHEDULE_OUT); > > trace_i915_gem_request_out(rq); > + rq->priotree.priority = INT_MAX; > i915_gem_request_put(rq); > > port[0] = port[1]; > -- > 2.13.2 >
Quoting Michał Winiarski (2017-07-19 15:25:51) > On Mon, Jul 17, 2017 at 09:42:33AM +0100, Chris Wilson wrote: > > With preemption, we will want to "unsubmit" a request, taking it back > > from the hw and returning it to the priority sorted execution list. In > > order to know where to insert it into that list, we need to remember > > its adjust priority (which may change even as it was being executed). > > > > Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> > > Cc: Michal Winiarski <michal.winiarski@intel.com> > > We should also change GuC dequeue/irq_handler. I wasn't sure if that was necessary as the current shortcut of sealing the priority once submitted applies to guc dequeue as it currently doesn't unsubmit. (Wasn't much point in penalising the guc until it was ready.) -Chris
diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c index 1b2f0e3d383a..8ab0c4b76c98 100644 --- a/drivers/gpu/drm/i915/intel_lrc.c +++ b/drivers/gpu/drm/i915/intel_lrc.c @@ -552,8 +552,6 @@ static void execlists_dequeue(struct intel_engine_cs *engine) } INIT_LIST_HEAD(&rq->priotree.link); - rq->priotree.priority = INT_MAX; - __i915_gem_request_submit(rq); trace_i915_gem_request_in(rq, port_index(port, engine)); last = rq; @@ -687,6 +685,7 @@ static void intel_lrc_irq_handler(unsigned long data) execlists_context_status_change(rq, INTEL_CONTEXT_SCHEDULE_OUT); trace_i915_gem_request_out(rq); + rq->priotree.priority = INT_MAX; i915_gem_request_put(rq); port[0] = port[1];
With preemption, we will want to "unsubmit" a request, taking it back from the hw and returning it to the priority sorted execution list. In order to know where to insert it into that list, we need to remember its adjust priority (which may change even as it was being executed). Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Michal Winiarski <michal.winiarski@intel.com> --- drivers/gpu/drm/i915/intel_lrc.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-)