All of lore.kernel.org
 help / color / mirror / Atom feed
From: Chris Wilson <chris@chris-wilson.co.uk>
To: intel-gfx@lists.freedesktop.org
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Subject: [Intel-gfx] [PATCH 26/28] drm/i915/gt: Shrink the critical section for irq signaling
Date: Tue, 17 Nov 2020 11:31:01 +0000	[thread overview]
Message-ID: <20201117113103.21480-26-chris@chris-wilson.co.uk> (raw)
In-Reply-To: <20201117113103.21480-1-chris@chris-wilson.co.uk>

Let's only wait for the list iterator when decoupling the virtual
breadcrumb, as the signaling of all the requests may take a long time,
during which we do not want to keep the tasklet spinning.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
---
 drivers/gpu/drm/i915/gt/intel_breadcrumbs.c       | 2 ++
 drivers/gpu/drm/i915/gt/intel_breadcrumbs_types.h | 1 +
 drivers/gpu/drm/i915/gt/intel_lrc.c               | 3 ++-
 3 files changed, 5 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/i915/gt/intel_breadcrumbs.c b/drivers/gpu/drm/i915/gt/intel_breadcrumbs.c
index 5b84f51931d9..3e69d6c3b197 100644
--- a/drivers/gpu/drm/i915/gt/intel_breadcrumbs.c
+++ b/drivers/gpu/drm/i915/gt/intel_breadcrumbs.c
@@ -251,6 +251,7 @@ static void signal_irq_work(struct irq_work *work)
 		intel_breadcrumbs_disarm_irq(b);
 
 	rcu_read_lock();
+	atomic_inc(&b->signaler_active);
 	list_for_each_entry_rcu(ce, &b->signalers, signal_link) {
 		struct i915_request *rq;
 
@@ -284,6 +285,7 @@ static void signal_irq_work(struct irq_work *work)
 			}
 		}
 	}
+	atomic_dec(&b->signaler_active);
 	rcu_read_unlock();
 
 	llist_for_each_safe(signal, sn, signal) {
diff --git a/drivers/gpu/drm/i915/gt/intel_breadcrumbs_types.h b/drivers/gpu/drm/i915/gt/intel_breadcrumbs_types.h
index a74bb3062bd8..f672053d694d 100644
--- a/drivers/gpu/drm/i915/gt/intel_breadcrumbs_types.h
+++ b/drivers/gpu/drm/i915/gt/intel_breadcrumbs_types.h
@@ -35,6 +35,7 @@ struct intel_breadcrumbs {
 	spinlock_t signalers_lock; /* protects the list of signalers */
 	struct list_head signalers;
 	struct llist_head signaled_requests;
+	atomic_t signaler_active;
 
 	spinlock_t irq_lock; /* protects the interrupt from hardirq context */
 	struct irq_work irq_work; /* for use from inside irq_lock */
diff --git a/drivers/gpu/drm/i915/gt/intel_lrc.c b/drivers/gpu/drm/i915/gt/intel_lrc.c
index a94daf4c70a3..87ede3513e4c 100644
--- a/drivers/gpu/drm/i915/gt/intel_lrc.c
+++ b/drivers/gpu/drm/i915/gt/intel_lrc.c
@@ -1400,7 +1400,8 @@ static void kick_siblings(struct i915_request *rq, struct intel_context *ce)
 		 * ce->signal_link.
 		 */
 		i915_request_cancel_breadcrumb(rq);
-		irq_work_sync(&engine->breadcrumbs->irq_work);
+		while (atomic_read(&engine->breadcrumbs->signaler_active))
+			cpu_relax();
 	}
 
 	if (READ_ONCE(ve->request))
-- 
2.20.1

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

  parent reply	other threads:[~2020-11-17 11:31 UTC|newest]

Thread overview: 52+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-11-17 11:30 [Intel-gfx] [PATCH 01/28] drm/i915/selftests: Improve granularity for mocs reset checks Chris Wilson
2020-11-17 11:30 ` [Intel-gfx] [PATCH 02/28] drm/i915/selftests: Small tweak to put the termination conditions together Chris Wilson
2020-11-17 11:30 ` [Intel-gfx] [PATCH 03/28] drm/i915/gem: Drop free_work for GEM contexts Chris Wilson
2020-11-17 11:30 ` [Intel-gfx] [PATCH 04/28] drm/i915/gt: Ignore dt==0 for reporting underflows Chris Wilson
2020-11-17 11:42   ` Tvrtko Ursulin
2020-11-17 11:30 ` [Intel-gfx] [PATCH 05/28] drm/i915/gt: Track the overall busy time Chris Wilson
2020-11-17 12:44   ` Tvrtko Ursulin
2020-11-17 13:05     ` Chris Wilson
2020-11-17 11:30 ` [Intel-gfx] [PATCH 06/28] drm/i915/gt: Include semaphore status in print_request() Chris Wilson
2020-11-17 11:30 ` [Intel-gfx] [PATCH 07/28] drm/i915: Lift i915_request_show() Chris Wilson
2020-11-17 12:51   ` Tvrtko Ursulin
2020-11-17 11:30 ` [Intel-gfx] [PATCH 08/28] drm/i915/gt: Show all active timelines for debugging Chris Wilson
2020-11-17 12:59   ` Tvrtko Ursulin
2020-11-17 13:25     ` Chris Wilson
2020-11-18 15:51       ` Tvrtko Ursulin
2020-11-19 10:47         ` Chris Wilson
2020-11-17 11:30 ` [Intel-gfx] [PATCH 09/28] drm/i915: Lift waiter/signaler iterators Chris Wilson
2020-11-17 13:00   ` Tvrtko Ursulin
2020-11-17 11:30 ` [Intel-gfx] [PATCH 10/28] drm/i915: Show timeline dependencies for debug Chris Wilson
2020-11-17 13:06   ` Tvrtko Ursulin
2020-11-17 13:30     ` Chris Wilson
2020-11-17 11:30 ` [Intel-gfx] [PATCH 11/28] drm/i915/gt: Defer enabling the breadcrumb interrupt to after submission Chris Wilson
2020-11-17 11:30 ` [Intel-gfx] [PATCH 12/28] drm/i915/gt: Track signaled breadcrumbs outside of the breadcrumb spinlock Chris Wilson
2020-11-17 11:30 ` [Intel-gfx] [PATCH 13/28] drm/i915/gt: Don't cancel the interrupt shadow too early Chris Wilson
2020-11-17 11:30 ` [Intel-gfx] [PATCH 14/28] drm/i915/gt: Free stale request on destroying the virtual engine Chris Wilson
2020-11-18 11:05   ` Tvrtko Ursulin
2020-11-18 11:24     ` Chris Wilson
2020-11-18 11:38       ` Tvrtko Ursulin
2020-11-18 12:10         ` Chris Wilson
2020-11-19 14:06           ` Tvrtko Ursulin
2020-11-19 14:22             ` Chris Wilson
2020-11-19 16:17               ` Tvrtko Ursulin
2020-11-17 11:30 ` [Intel-gfx] [PATCH 15/28] drm/i915/gt: Protect context lifetime with RCU Chris Wilson
2020-11-18 11:36   ` Tvrtko Ursulin
2020-11-17 11:30 ` [Intel-gfx] [PATCH 16/28] drm/i915/gt: Split the breadcrumb spinlock between global and contexts Chris Wilson
2020-11-18 11:35   ` Tvrtko Ursulin
2020-11-17 11:30 ` [Intel-gfx] [PATCH 17/28] drm/i915/gt: Move the breadcrumb to the signaler if completed upon cancel Chris Wilson
2020-11-17 11:30 ` [Intel-gfx] [PATCH 18/28] drm/i915/gt: Decouple completed requests on unwind Chris Wilson
2020-11-17 11:30 ` [Intel-gfx] [PATCH 19/28] drm/i915/gt: Check for a completed last request once Chris Wilson
2020-11-17 11:30 ` [Intel-gfx] [PATCH 20/28] drm/i915/gt: Replace direct submit with direct call to tasklet Chris Wilson
2020-11-17 11:30 ` [Intel-gfx] [PATCH 21/28] drm/i915/gt: ce->inflight updates are now serialised Chris Wilson
2020-11-17 11:30 ` [Intel-gfx] [PATCH 22/28] drm/i915/gt: Use virtual_engine during execlists_dequeue Chris Wilson
2020-11-17 11:30 ` [Intel-gfx] [PATCH 23/28] drm/i915/gt: Decouple inflight virtual engines Chris Wilson
2020-11-17 11:30 ` [Intel-gfx] [PATCH 24/28] drm/i915/gt: Defer schedule_out until after the next dequeue Chris Wilson
2020-11-17 11:31 ` [Intel-gfx] [PATCH 25/28] drm/i915/gt: Remove virtual breadcrumb before transfer Chris Wilson
2020-11-17 11:31 ` Chris Wilson [this message]
2020-11-17 11:31 ` [Intel-gfx] [PATCH 27/28] drm/i915/gt: Resubmit the virtual engine on schedule-out Chris Wilson
2020-11-17 11:31 ` [Intel-gfx] [PATCH 28/28] drm/i915/gt: Simplify virtual engine handling for execlists_hold() Chris Wilson
2020-11-17 18:54 ` [Intel-gfx] ✗ Fi.CI.CHECKPATCH: warning for series starting with [01/28] drm/i915/selftests: Improve granularity for mocs reset checks Patchwork
2020-11-17 18:56 ` [Intel-gfx] ✗ Fi.CI.SPARSE: " Patchwork
2020-11-17 19:24 ` [Intel-gfx] ✓ Fi.CI.BAT: success " Patchwork
2020-11-17 22:56 ` [Intel-gfx] ✗ Fi.CI.IGT: failure " Patchwork

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20201117113103.21480-26-chris@chris-wilson.co.uk \
    --to=chris@chris-wilson.co.uk \
    --cc=intel-gfx@lists.freedesktop.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.