All of lore.kernel.org
 help / color / mirror / Atom feed
From: Chris Wilson <chris@chris-wilson.co.uk>
To: intel-gfx@lists.freedesktop.org
Subject: [PATCH 10/10] drm/i915: Flush idle barriers when waiting
Date: Thu, 10 Oct 2019 08:14:34 +0100	[thread overview]
Message-ID: <20191010071434.31195-10-chris@chris-wilson.co.uk> (raw)
In-Reply-To: <20191010071434.31195-1-chris@chris-wilson.co.uk>

If we do find ourselves with an idle barrier inside our active while
waiting, attempt to flush it by emitting a pulse using the kernel
context.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
---
 .../gpu/drm/i915/gt/intel_engine_heartbeat.c  | 14 +++++++++++++
 .../gpu/drm/i915/gt/intel_engine_heartbeat.h  |  1 +
 drivers/gpu/drm/i915/i915_active.c            | 21 +++++++++++++++++--
 3 files changed, 34 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/i915/gt/intel_engine_heartbeat.c b/drivers/gpu/drm/i915/gt/intel_engine_heartbeat.c
index f68acf9118f3..e27bb7f028bd 100644
--- a/drivers/gpu/drm/i915/gt/intel_engine_heartbeat.c
+++ b/drivers/gpu/drm/i915/gt/intel_engine_heartbeat.c
@@ -169,3 +169,17 @@ int intel_engine_pulse(struct intel_engine_cs *engine)
 	intel_engine_pm_put(engine);
 	return err;
 }
+
+int intel_engine_flush_barriers(struct intel_engine_cs *engine)
+{
+	struct i915_request *rq;
+
+	rq = i915_request_create(engine->kernel_context);
+	if (IS_ERR(rq))
+		return PTR_ERR(rq);
+
+	idle_pulse(engine, rq);
+	i915_request_add(rq);
+
+	return 0;
+}
diff --git a/drivers/gpu/drm/i915/gt/intel_engine_heartbeat.h b/drivers/gpu/drm/i915/gt/intel_engine_heartbeat.h
index 39391004554d..0c1ad0fc091d 100644
--- a/drivers/gpu/drm/i915/gt/intel_engine_heartbeat.h
+++ b/drivers/gpu/drm/i915/gt/intel_engine_heartbeat.h
@@ -15,5 +15,6 @@ void intel_engine_park_heartbeat(struct intel_engine_cs *engine);
 void intel_engine_unpark_heartbeat(struct intel_engine_cs *engine);
 
 int intel_engine_pulse(struct intel_engine_cs *engine);
+int intel_engine_flush_barriers(struct intel_engine_cs *engine);
 
 #endif /* INTEL_ENGINE_HEARTBEAT_H */
diff --git a/drivers/gpu/drm/i915/i915_active.c b/drivers/gpu/drm/i915/i915_active.c
index aa37c07004b9..98d5fe1c7e19 100644
--- a/drivers/gpu/drm/i915/i915_active.c
+++ b/drivers/gpu/drm/i915/i915_active.c
@@ -6,6 +6,7 @@
 
 #include <linux/debugobjects.h>
 
+#include "gt/intel_engine_heartbeat.h"
 #include "gt/intel_engine_pm.h"
 
 #include "i915_drv.h"
@@ -435,6 +436,21 @@ static void enable_signaling(struct i915_active_fence *active)
 	dma_fence_put(fence);
 }
 
+static int flush_barrier(struct active_node *it)
+{
+	struct intel_engine_cs *engine;
+
+	if (!is_barrier(&it->base))
+		return 0;
+
+	engine = __barrier_to_engine(it);
+	smp_rmb(); /* serialise with add_active_barriers */
+	if (!is_barrier(&it->base))
+		return 0;
+
+	return intel_engine_flush_barriers(engine);
+}
+
 int i915_active_wait(struct i915_active *ref)
 {
 	struct active_node *it, *n;
@@ -448,8 +464,9 @@ int i915_active_wait(struct i915_active *ref)
 	/* Flush lazy signals */
 	enable_signaling(&ref->excl);
 	rbtree_postorder_for_each_entry_safe(it, n, &ref->tree, node) {
-		if (is_barrier(&it->base)) /* unconnected idle barrier */
-			continue;
+		err = flush_barrier(it); /* unconnected idle barrier? */
+		if (err)
+			break;
 
 		enable_signaling(&it->base);
 	}
-- 
2.23.0

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

  parent reply	other threads:[~2019-10-10  7:14 UTC|newest]

Thread overview: 43+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-10-10  7:14 [PATCH 01/10] drm/i915: Note the addition of timeslicing to the pretend scheduler Chris Wilson
2019-10-10  7:14 ` [PATCH 02/10] drm/i915/execlists: Leave tell-tales as to why pending[] is bad Chris Wilson
2019-10-11  8:39   ` Tvrtko Ursulin
2019-10-10  7:14 ` [PATCH 03/10] drm/i915: Expose engine properties via sysfs Chris Wilson
2019-10-11  8:44   ` Tvrtko Ursulin
2019-10-11  8:49     ` Chris Wilson
2019-10-11  9:04       ` Tvrtko Ursulin
2019-10-11  9:40   ` [PATCH v2] " Chris Wilson
2019-10-10  7:14 ` [PATCH 04/10] drm/i915/execlists: Force preemption Chris Wilson
2019-10-10  7:14 ` [PATCH 05/10] drm/i915: Mark up "sentinel" requests Chris Wilson
2019-10-11  8:45   ` Tvrtko Ursulin
2019-10-10  7:14 ` [PATCH 06/10] drm/i915/gt: Introduce barrier pulses along engines Chris Wilson
2019-10-11  9:11   ` Tvrtko Ursulin
2019-10-11  9:52     ` Chris Wilson
2019-10-10  7:14 ` [PATCH 07/10] drm/i915/execlists: Cancel banned contexts on schedule-out Chris Wilson
2019-10-11  9:47   ` Tvrtko Ursulin
2019-10-11 10:03     ` Chris Wilson
2019-10-11 10:15     ` Chris Wilson
2019-10-11 10:40       ` Chris Wilson
2019-10-11 11:16   ` [PATCH v2] " Chris Wilson
2019-10-11 13:10     ` Tvrtko Ursulin
2019-10-11 14:10       ` Chris Wilson
2019-10-10  7:14 ` [PATCH 08/10] drm/i915: Cancel non-persistent contexts on close Chris Wilson
2019-10-11 13:55   ` Tvrtko Ursulin
2019-10-11 14:22     ` Chris Wilson
2019-10-11 15:41       ` Chris Wilson
2019-10-10  7:14 ` [PATCH 09/10] drm/i915: Replace hangcheck by heartbeats Chris Wilson
2019-10-11 14:24   ` Tvrtko Ursulin
2019-10-11 15:06     ` Chris Wilson
2019-10-10  7:14 ` Chris Wilson [this message]
2019-10-11 14:56   ` [PATCH 10/10] drm/i915: Flush idle barriers when waiting Tvrtko Ursulin
2019-10-11 15:11     ` Chris Wilson
2019-10-14 13:08       ` Tvrtko Ursulin
2019-10-14 13:38         ` Chris Wilson
2019-10-23 15:33         ` Chris Wilson
2019-10-23 15:33           ` [Intel-gfx] " Chris Wilson
2019-10-10  8:18 ` ✗ Fi.CI.CHECKPATCH: warning for series starting with [01/10] drm/i915: Note the addition of timeslicing to the pretend scheduler Patchwork
2019-10-10  8:42 ` ✓ Fi.CI.BAT: success " Patchwork
2019-10-10 16:19 ` ✗ Fi.CI.IGT: failure " Patchwork
2019-10-11  8:16 ` [PATCH 01/10] " Tvrtko Ursulin
2019-10-11  9:49 ` ✗ Fi.CI.BUILD: failure for series starting with [01/10] drm/i915: Note the addition of timeslicing to the pretend scheduler (rev2) Patchwork
2019-10-11 11:39 ` ✗ Fi.CI.BUILD: failure for series starting with [01/10] drm/i915: Note the addition of timeslicing to the pretend scheduler (rev3) Patchwork
2019-10-14 22:05 [PATCH 01/10] drm/i915/gem: Distinguish each object type Chris Wilson
2019-10-14 22:05 ` [PATCH 10/10] drm/i915: Flush idle barriers when waiting Chris Wilson

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=20191010071434.31195-10-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.