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 33/33] drm/i915/execlists: Force preemption
Date: Mon, 20 May 2019 09:01:27 +0100	[thread overview]
Message-ID: <20190520080127.18255-33-chris@chris-wilson.co.uk> (raw)
In-Reply-To: <20190520080127.18255-1-chris@chris-wilson.co.uk>

If the preempted context takes too long to relinquish control, e.g. it
is stuck inside a shader with arbitration disabled, evict that context
with an engine reset. This ensures that preemptions are reasonably
responsive, providing a tighter QoS for the more important context at
the cost of flagging unresponsive contexts more frequently (i.e. instead
of using an ~10s hangcheck, we now evict at ~10ms).  The challenge of
choice being in picky a timeout that can be reasonably serviced by HW for
typical workloads.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
---
 drivers/gpu/drm/i915/Kconfig.profile | 12 +++++++
 drivers/gpu/drm/i915/gt/intel_lrc.c  | 49 ++++++++++++++++++++++++++--
 2 files changed, 58 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/i915/Kconfig.profile b/drivers/gpu/drm/i915/Kconfig.profile
index 0e5db98da8f3..e424823395c4 100644
--- a/drivers/gpu/drm/i915/Kconfig.profile
+++ b/drivers/gpu/drm/i915/Kconfig.profile
@@ -11,3 +11,15 @@ config DRM_I915_SPIN_REQUEST
 	  May be 0 to disable the initial spin. In practice, we estimate
 	  the cost of enabling the interrupt (if currently disabled) to be
 	  a few microseconds.
+
+config DRM_I915_PREEMPT_TIMEOUT
+	int
+	default 10 # milliseconds
+	help
+	  How long to wait (in milliseconds) for a preemption event to occur
+	  when submitting a new context via execlists. If the current context
+	  does not hit an arbitration point and yield to HW before the timer
+	  expires, the HW will be reset to allow the more important context
+	  to execute.
+
+	  May be 0 to disable the timeout.
diff --git a/drivers/gpu/drm/i915/gt/intel_lrc.c b/drivers/gpu/drm/i915/gt/intel_lrc.c
index 109b32c6aead..25cc1706860f 100644
--- a/drivers/gpu/drm/i915/gt/intel_lrc.c
+++ b/drivers/gpu/drm/i915/gt/intel_lrc.c
@@ -1218,6 +1218,11 @@ static void execlists_dequeue(struct intel_engine_cs *engine)
 		*port = execlists_schedule_in(last, port - execlists->pending);
 		memset(port + 1, 0, (last_port - port) * sizeof(*port));
 		execlists_submit_ports(engine);
+
+		if (CONFIG_DRM_I915_PREEMPT_TIMEOUT) {
+			mod_timer(&execlists->timer,
+				  jiffies + msecs_to_jiffies_timeout(CONFIG_DRM_I915_PREEMPT_TIMEOUT));
+		}
 	}
 }
 
@@ -1373,13 +1378,48 @@ static void process_csb(struct intel_engine_cs *engine)
 	invalidate_csb_entries(&buf[0], &buf[num_entries - 1]);
 }
 
-static void __execlists_submission_tasklet(struct intel_engine_cs *const engine)
+static bool __execlists_submission_tasklet(struct intel_engine_cs *const engine)
 {
 	lockdep_assert_held(&engine->active.lock);
 
 	process_csb(engine);
-	if (!engine->execlists.pending[0])
+	if (!engine->execlists.pending[0]) {
 		execlists_dequeue(engine);
+		return true;
+	}
+
+	return false;
+}
+
+static void preempt_reset(struct intel_engine_cs *engine)
+{
+	const unsigned int bit = I915_RESET_ENGINE + engine->id;
+	unsigned long *lock = &engine->i915->gpu_error.flags;
+
+	if (test_and_set_bit(bit, lock))
+		return;
+
+	tasklet_disable_nosync(&engine->execlists.tasklet);
+	spin_unlock(&engine->active.lock);
+
+	i915_reset_engine(engine, "preemption timed out");
+
+	spin_lock(&engine->active.lock);
+	tasklet_enable(&engine->execlists.tasklet);
+
+	clear_bit(bit, lock);
+	wake_up_bit(lock, bit);
+}
+
+static bool preempt_timeout(struct intel_engine_cs *const engine)
+{
+	if (!CONFIG_DRM_I915_PREEMPT_TIMEOUT)
+		return false;
+
+	if (!intel_engine_has_preemption(engine))
+		return false;
+
+	return !timer_pending(&engine->execlists.timer);
 }
 
 /*
@@ -1392,7 +1432,10 @@ static void execlists_submission_tasklet(unsigned long data)
 	unsigned long flags;
 
 	spin_lock_irqsave(&engine->active.lock, flags);
-	__execlists_submission_tasklet(engine);
+
+	if (!__execlists_submission_tasklet(engine) && preempt_timeout(engine))
+		preempt_reset(engine);
+
 	spin_unlock_irqrestore(&engine->active.lock, flags);
 }
 
-- 
2.20.1

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

  parent reply	other threads:[~2019-05-20  9:21 UTC|newest]

Thread overview: 38+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-05-20  8:00 [PATCH 01/33] drm/i915: Restore control over ppgtt for context creation ABI Chris Wilson
2019-05-20  8:00 ` [PATCH 02/33] drm/i915: Allow a context to define its set of engines Chris Wilson
2019-05-20  8:00 ` [PATCH 03/33] drm/i915: Extend I915_CONTEXT_PARAM_SSEU to support local ctx->engine[] Chris Wilson
2019-05-20  8:00 ` [PATCH 04/33] drm/i915: Re-expose SINGLE_TIMELINE flags for context creation Chris Wilson
2019-05-20  8:00 ` [PATCH 05/33] drm/i915: Allow userspace to clone contexts on creation Chris Wilson
2019-05-20  8:01 ` [PATCH 06/33] drm/i915: Load balancing across a virtual engine Chris Wilson
2019-05-20  8:01 ` [PATCH 07/33] drm/i915: Apply an execution_mask to the virtual_engine Chris Wilson
2019-05-20  8:01 ` [PATCH 08/33] drm/i915: Extend execution fence to support a callback Chris Wilson
2019-05-20  8:01 ` [PATCH 09/33] drm/i915/execlists: Virtual engine bonding Chris Wilson
2019-05-20  8:01 ` [PATCH 10/33] drm/i915: Allow specification of parallel execbuf Chris Wilson
2019-05-20  8:01 ` [PATCH 11/33] drm/i915: Split GEM object type definition to its own header Chris Wilson
2019-05-20  8:01 ` [PATCH 12/33] drm/i915: Pull GEM ioctls interface to its own file Chris Wilson
2019-05-20  8:01 ` [PATCH 13/33] drm/i915: Move object->pages API to i915_gem_object.[ch] Chris Wilson
2019-05-20  8:01 ` [PATCH 14/33] drm/i915: Move shmem object setup to its own file Chris Wilson
2019-05-20  8:01 ` [PATCH 15/33] drm/i915: Move phys objects " Chris Wilson
2019-05-20  8:01 ` [PATCH 16/33] drm/i915: Move mmap and friends " Chris Wilson
2019-05-20  8:01 ` [PATCH 17/33] drm/i915: Move GEM domain management " Chris Wilson
2019-05-20  8:01 ` [PATCH 18/33] drm/i915: Move more GEM objects under gem/ Chris Wilson
2019-05-20 12:58   ` Mika Kuoppala
2019-05-20  8:01 ` [PATCH 19/33] drm/i915: Pull scatterlist utils out of i915_gem.h Chris Wilson
2019-05-20  8:01 ` [PATCH 20/33] drm/i915: Move GEM object domain management from struct_mutex to local Chris Wilson
2019-05-20  8:01 ` [PATCH 21/33] drm/i915: Move GEM object waiting to its own file Chris Wilson
2019-05-20  8:01 ` [PATCH 22/33] drm/i915: Move GEM object busy checking " Chris Wilson
2019-05-20  8:01 ` [PATCH 23/33] drm/i915: Move GEM client throttling " Chris Wilson
2019-05-20  8:01 ` [PATCH 24/33] drm/i915: Drop the deferred active reference Chris Wilson
2019-05-20  8:01 ` [PATCH 25/33] drm/i915: Move object close under its own lock Chris Wilson
2019-05-22 14:32   ` Mika Kuoppala
2019-05-22 14:47     ` Chris Wilson
2019-05-22 14:52     ` Chris Wilson
2019-05-20  8:01 ` [PATCH 26/33] drm/i915: Rename intel_context.active to .inflight Chris Wilson
2019-05-20  8:01 ` [PATCH 27/33] drm/i915: Keep contexts pinned until after the next kernel context switch Chris Wilson
2019-05-20  8:01 ` [PATCH 28/33] drm/i915: Stop retiring along engine Chris Wilson
2019-05-20  8:01 ` [PATCH 29/33] drm/i915: Replace engine->timeline with a plain list Chris Wilson
2019-05-20  8:01 ` [PATCH 30/33] drm/i915: Flush the execution-callbacks on retiring Chris Wilson
2019-05-20  8:01 ` [PATCH 31/33] drm/i915/execlists: Preempt-to-busy Chris Wilson
2019-05-20  8:01 ` [PATCH 32/33] drm/i915/execlists: Minimalistic timeslicing Chris Wilson
2019-05-20  8:01 ` Chris Wilson [this message]
2019-05-20 13:05 ` ✗ Fi.CI.BAT: failure for series starting with [01/33] drm/i915: Restore control over ppgtt for context creation ABI 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=20190520080127.18255-33-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.