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: tvrtko.ursulin@intel.com, Chris Wilson <chris@chris-wilson.co.uk>,
	stable@vger.kernel.org
Subject: [PATCH 03/24] drm/i915/gt: Cancel the preemption timeout on responding to it
Date: Fri,  4 Dec 2020 14:02:54 +0000	[thread overview]
Message-ID: <20201204140315.24341-3-chris@chris-wilson.co.uk> (raw)
In-Reply-To: <20201204140315.24341-1-chris@chris-wilson.co.uk>

We currently presume that the engine reset is successful, cancelling the
expired preemption timer in the process. However, engine resets can
fail, leaving the timeout still pending and we will then respond to the
timeout again next time the tasklet fires. What we want is for the
failed engine reset to be promoted to a full device reset, which is
kicked by the heartbeat once the engine stops processing events.

Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/1168
Fixes: 3a7a92aba8fb ("drm/i915/execlists: Force preemption")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: <stable@vger.kernel.org> # v5.5+
---
 drivers/gpu/drm/i915/gt/intel_lrc.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/i915/gt/intel_lrc.c b/drivers/gpu/drm/i915/gt/intel_lrc.c
index 1d209a8a95e8..7f25894e41d5 100644
--- a/drivers/gpu/drm/i915/gt/intel_lrc.c
+++ b/drivers/gpu/drm/i915/gt/intel_lrc.c
@@ -3209,8 +3209,10 @@ static void execlists_submission_tasklet(unsigned long data)
 		spin_unlock_irqrestore(&engine->active.lock, flags);
 
 		/* Recheck after serialising with direct-submission */
-		if (unlikely(timeout && preempt_timeout(engine)))
+		if (unlikely(timeout && preempt_timeout(engine))) {
+			cancel_timer(&engine->execlists.preempt);
 			execlists_reset(engine, "preemption time out");
+		}
 	}
 }
 
-- 
2.20.1


WARNING: multiple messages have this Message-ID (diff)
From: Chris Wilson <chris@chris-wilson.co.uk>
To: intel-gfx@lists.freedesktop.org
Cc: Chris Wilson <chris@chris-wilson.co.uk>, stable@vger.kernel.org
Subject: [Intel-gfx] [PATCH 03/24] drm/i915/gt: Cancel the preemption timeout on responding to it
Date: Fri,  4 Dec 2020 14:02:54 +0000	[thread overview]
Message-ID: <20201204140315.24341-3-chris@chris-wilson.co.uk> (raw)
In-Reply-To: <20201204140315.24341-1-chris@chris-wilson.co.uk>

We currently presume that the engine reset is successful, cancelling the
expired preemption timer in the process. However, engine resets can
fail, leaving the timeout still pending and we will then respond to the
timeout again next time the tasklet fires. What we want is for the
failed engine reset to be promoted to a full device reset, which is
kicked by the heartbeat once the engine stops processing events.

Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/1168
Fixes: 3a7a92aba8fb ("drm/i915/execlists: Force preemption")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: <stable@vger.kernel.org> # v5.5+
---
 drivers/gpu/drm/i915/gt/intel_lrc.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/i915/gt/intel_lrc.c b/drivers/gpu/drm/i915/gt/intel_lrc.c
index 1d209a8a95e8..7f25894e41d5 100644
--- a/drivers/gpu/drm/i915/gt/intel_lrc.c
+++ b/drivers/gpu/drm/i915/gt/intel_lrc.c
@@ -3209,8 +3209,10 @@ static void execlists_submission_tasklet(unsigned long data)
 		spin_unlock_irqrestore(&engine->active.lock, flags);
 
 		/* Recheck after serialising with direct-submission */
-		if (unlikely(timeout && preempt_timeout(engine)))
+		if (unlikely(timeout && preempt_timeout(engine))) {
+			cancel_timer(&engine->execlists.preempt);
 			execlists_reset(engine, "preemption time out");
+		}
 	}
 }
 
-- 
2.20.1

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

  parent reply	other threads:[~2020-12-04 14:04 UTC|newest]

Thread overview: 33+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-12-04 14:02 [Intel-gfx] [PATCH 01/24] drm/i915: Disable outputs during unregister Chris Wilson
2020-12-04 14:02 ` [PATCH 02/24] drm/i915/gt: Ignore repeated attempts to suspend request flow across reset Chris Wilson
2020-12-04 14:02   ` [Intel-gfx] " Chris Wilson
2020-12-04 15:03   ` Mika Kuoppala
2020-12-04 15:03     ` Mika Kuoppala
2020-12-04 14:02 ` Chris Wilson [this message]
2020-12-04 14:02   ` [Intel-gfx] [PATCH 03/24] drm/i915/gt: Cancel the preemption timeout on responding to it Chris Wilson
2020-12-04 15:04   ` Mika Kuoppala
2020-12-04 15:04     ` Mika Kuoppala
2020-12-04 14:02 ` [Intel-gfx] [PATCH 04/24] drm/i915/gt: Include reset failures in the trace Chris Wilson
2020-12-04 14:02 ` [Intel-gfx] [PATCH 05/24] drm/i915/gt: Clear the execlists timers upon reset Chris Wilson
2020-12-04 14:02 ` [Intel-gfx] [PATCH 06/24] drm/i915/gt: Replace direct submit with direct call to tasklet Chris Wilson
2020-12-04 14:02 ` [Intel-gfx] [PATCH 07/24] drm/i915/gt: Use virtual_engine during execlists_dequeue Chris Wilson
2020-12-04 14:02 ` [Intel-gfx] [PATCH 08/24] drm/i915/gt: Decouple inflight virtual engines Chris Wilson
2020-12-04 14:03 ` [Intel-gfx] [PATCH 09/24] drm/i915/gt: Defer schedule_out until after the next dequeue Chris Wilson
2020-12-04 14:03 ` [Intel-gfx] [PATCH 10/24] drm/i915/gt: Remove virtual breadcrumb before transfer Chris Wilson
2020-12-04 14:03 ` [Intel-gfx] [PATCH 11/24] drm/i915/gt: Shrink the critical section for irq signaling Chris Wilson
2020-12-04 14:03 ` [Intel-gfx] [PATCH 12/24] drm/i915/gt: Resubmit the virtual engine on schedule-out Chris Wilson
2020-12-04 14:03 ` [Intel-gfx] [PATCH 13/24] drm/i915/gt: Simplify virtual engine handling for execlists_hold() Chris Wilson
2020-12-04 14:03 ` [Intel-gfx] [PATCH 14/24] drm/i915/gt: ce->inflight updates are now serialised Chris Wilson
2020-12-04 14:03 ` [Intel-gfx] [PATCH 15/24] drm/i915/gem: Drop free_work for GEM contexts Chris Wilson
2020-12-04 14:03 ` [Intel-gfx] [PATCH 16/24] drm/i915/gt: Track the overall awake/busy time Chris Wilson
2020-12-04 14:03 ` [Intel-gfx] [PATCH 17/24] drm/i915: Encode fence specific waitqueue behaviour into the wait.flags Chris Wilson
2020-12-04 14:03 ` [Intel-gfx] [PATCH 18/24] drm/i915/gt: Track all timelines created using the HWSP Chris Wilson
2020-12-04 14:03 ` [Intel-gfx] [PATCH 19/24] drm/i915/gt: Wrap intel_timeline.has_initial_breadcrumb Chris Wilson
2020-12-04 14:03 ` [Intel-gfx] [PATCH 20/24] drm/i915/gt: Track timeline GGTT offset separately from subpage offset Chris Wilson
2020-12-04 14:03 ` [Intel-gfx] [PATCH 21/24] drm/i915/gt: Add timeline "mode" Chris Wilson
2020-12-04 14:03 ` [Intel-gfx] [PATCH 22/24] drm/i915/gt: Use indices for writing into relative timelines Chris Wilson
2020-12-04 14:03 ` [Intel-gfx] [PATCH 23/24] drm/i915/selftests: Exercise relative timeline modes Chris Wilson
2020-12-04 14:03 ` [Intel-gfx] [PATCH 24/24] drm/i915/gt: Use ppHWSP for unshared non-semaphore related timelines Chris Wilson
2020-12-04 15:52 ` [Intel-gfx] ✗ Fi.CI.CHECKPATCH: warning for series starting with [01/24] drm/i915: Disable outputs during unregister Patchwork
2020-12-04 15:53 ` [Intel-gfx] ✗ Fi.CI.SPARSE: " Patchwork
2020-12-04 16:23 ` [Intel-gfx] ✗ Fi.CI.BAT: 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=20201204140315.24341-3-chris@chris-wilson.co.uk \
    --to=chris@chris-wilson.co.uk \
    --cc=intel-gfx@lists.freedesktop.org \
    --cc=stable@vger.kernel.org \
    --cc=tvrtko.ursulin@intel.com \
    /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.