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 07/11] drm/i915/gt: Expose preempt reset timeout via sysfs
Date: Thu, 24 Oct 2019 12:40:24 +0100	[thread overview]
Message-ID: <20191024114028.6170-8-chris@chris-wilson.co.uk> (raw)
In-Reply-To: <20191024114028.6170-1-chris@chris-wilson.co.uk>

After initialising a preemption request, we give the current resident a
small amount of time to vacate the GPU. The preemption request is for a
higher priority context and should be immediate to maintain high
quality of service (and avoid priority inversion). However, the
preemption granularity of the GPU can be quite coarse and so we need a
compromise.

The preempt timeout can be adjusted per-engine using,

	/sys/class/drm/card?/engine/*/preempt_timeout_ms

and can be disabled by setting it to 0.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
---
 drivers/gpu/drm/i915/Kconfig.profile         |  3 ++
 drivers/gpu/drm/i915/gt/intel_engine_sysfs.c | 48 ++++++++++++++++++++
 2 files changed, 51 insertions(+)

diff --git a/drivers/gpu/drm/i915/Kconfig.profile b/drivers/gpu/drm/i915/Kconfig.profile
index 76145d25ce65..066ea9ba2756 100644
--- a/drivers/gpu/drm/i915/Kconfig.profile
+++ b/drivers/gpu/drm/i915/Kconfig.profile
@@ -33,6 +33,9 @@ config DRM_I915_PREEMPT_TIMEOUT
 	  expires, the HW will be reset to allow the more important context
 	  to execute.
 
+	  This is adjustable via
+	  /sys/class/drm/card?/engine/*/preempt_timeout_ms
+
 	  May be 0 to disable the timeout.
 
 config DRM_I915_SPIN_REQUEST
diff --git a/drivers/gpu/drm/i915/gt/intel_engine_sysfs.c b/drivers/gpu/drm/i915/gt/intel_engine_sysfs.c
index 86377a4ffe70..25012297aa9e 100644
--- a/drivers/gpu/drm/i915/gt/intel_engine_sysfs.c
+++ b/drivers/gpu/drm/i915/gt/intel_engine_sysfs.c
@@ -223,6 +223,50 @@ stop_show(struct kobject *kobj, struct kobj_attribute *attr, char *buf)
 static struct kobj_attribute stop_timeout_attr =
 __ATTR(stop_timeout_ms, 0644, stop_show, stop_store);
 
+static ssize_t
+preempt_timeout_store(struct kobject *kobj, struct kobj_attribute *attr,
+		      const char *buf, size_t count)
+{
+	struct intel_engine_cs *engine = kobj_to_engine(kobj);
+	unsigned long long timeout;
+	int err;
+
+	/*
+	 * After initialising a preemption request, we give the current
+	 * resident a small amount of time to vacate the GPU. The preemption
+	 * request is for a higher priority context and should be immediate to
+	 * maintain high quality of service (and avoid priority inversion).
+	 * However, the preemption granularity of the GPU can be quite coarse
+	 * and so we need a compromise.
+	 */
+
+	err = kstrtoull(buf, 0, &timeout);
+	if (err)
+		return err;
+
+	if (timeout > jiffies_to_msecs(MAX_SCHEDULE_TIMEOUT))
+		return -EINVAL;
+
+	WRITE_ONCE(engine->props.preempt_timeout_ms, timeout);
+
+	if (READ_ONCE(engine->execlists.pending[0]))
+		set_timer_ms(&engine->execlists.preempt, timeout);
+
+	return count;
+}
+
+static ssize_t
+preempt_timeout_show(struct kobject *kobj, struct kobj_attribute *attr,
+		     char *buf)
+{
+	struct intel_engine_cs *engine = kobj_to_engine(kobj);
+
+	return sprintf(buf, "%lu\n", engine->props.preempt_timeout_ms);
+}
+
+static struct kobj_attribute preempt_timeout_attr =
+__ATTR(preempt_timeout_ms, 0644, preempt_timeout_show, preempt_timeout_store);
+
 static void kobj_engine_release(struct kobject *kobj)
 {
 	kfree(kobj);
@@ -292,6 +336,10 @@ void intel_engines_add_sysfs(struct drm_i915_private *i915)
 		    sysfs_create_file(kobj, &timeslice_duration_attr.attr))
 			goto err_engine;
 
+		if (intel_engine_has_preempt_reset(engine) &&
+		    sysfs_create_file(kobj, &preempt_timeout_attr.attr))
+			goto err_engine;
+
 		if (0) {
 err_object:
 			kobject_put(kobj);
-- 
2.24.0.rc0

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

WARNING: multiple messages have this Message-ID (diff)
From: Chris Wilson <chris@chris-wilson.co.uk>
To: intel-gfx@lists.freedesktop.org
Subject: [Intel-gfx] [PATCH 07/11] drm/i915/gt: Expose preempt reset timeout via sysfs
Date: Thu, 24 Oct 2019 12:40:24 +0100	[thread overview]
Message-ID: <20191024114028.6170-8-chris@chris-wilson.co.uk> (raw)
Message-ID: <20191024114024.4VyV3RlCKkxz716uEBWX4LwzlnUgmHXbA3KqiyzoTbU@z> (raw)
In-Reply-To: <20191024114028.6170-1-chris@chris-wilson.co.uk>

After initialising a preemption request, we give the current resident a
small amount of time to vacate the GPU. The preemption request is for a
higher priority context and should be immediate to maintain high
quality of service (and avoid priority inversion). However, the
preemption granularity of the GPU can be quite coarse and so we need a
compromise.

The preempt timeout can be adjusted per-engine using,

	/sys/class/drm/card?/engine/*/preempt_timeout_ms

and can be disabled by setting it to 0.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
---
 drivers/gpu/drm/i915/Kconfig.profile         |  3 ++
 drivers/gpu/drm/i915/gt/intel_engine_sysfs.c | 48 ++++++++++++++++++++
 2 files changed, 51 insertions(+)

diff --git a/drivers/gpu/drm/i915/Kconfig.profile b/drivers/gpu/drm/i915/Kconfig.profile
index 76145d25ce65..066ea9ba2756 100644
--- a/drivers/gpu/drm/i915/Kconfig.profile
+++ b/drivers/gpu/drm/i915/Kconfig.profile
@@ -33,6 +33,9 @@ config DRM_I915_PREEMPT_TIMEOUT
 	  expires, the HW will be reset to allow the more important context
 	  to execute.
 
+	  This is adjustable via
+	  /sys/class/drm/card?/engine/*/preempt_timeout_ms
+
 	  May be 0 to disable the timeout.
 
 config DRM_I915_SPIN_REQUEST
diff --git a/drivers/gpu/drm/i915/gt/intel_engine_sysfs.c b/drivers/gpu/drm/i915/gt/intel_engine_sysfs.c
index 86377a4ffe70..25012297aa9e 100644
--- a/drivers/gpu/drm/i915/gt/intel_engine_sysfs.c
+++ b/drivers/gpu/drm/i915/gt/intel_engine_sysfs.c
@@ -223,6 +223,50 @@ stop_show(struct kobject *kobj, struct kobj_attribute *attr, char *buf)
 static struct kobj_attribute stop_timeout_attr =
 __ATTR(stop_timeout_ms, 0644, stop_show, stop_store);
 
+static ssize_t
+preempt_timeout_store(struct kobject *kobj, struct kobj_attribute *attr,
+		      const char *buf, size_t count)
+{
+	struct intel_engine_cs *engine = kobj_to_engine(kobj);
+	unsigned long long timeout;
+	int err;
+
+	/*
+	 * After initialising a preemption request, we give the current
+	 * resident a small amount of time to vacate the GPU. The preemption
+	 * request is for a higher priority context and should be immediate to
+	 * maintain high quality of service (and avoid priority inversion).
+	 * However, the preemption granularity of the GPU can be quite coarse
+	 * and so we need a compromise.
+	 */
+
+	err = kstrtoull(buf, 0, &timeout);
+	if (err)
+		return err;
+
+	if (timeout > jiffies_to_msecs(MAX_SCHEDULE_TIMEOUT))
+		return -EINVAL;
+
+	WRITE_ONCE(engine->props.preempt_timeout_ms, timeout);
+
+	if (READ_ONCE(engine->execlists.pending[0]))
+		set_timer_ms(&engine->execlists.preempt, timeout);
+
+	return count;
+}
+
+static ssize_t
+preempt_timeout_show(struct kobject *kobj, struct kobj_attribute *attr,
+		     char *buf)
+{
+	struct intel_engine_cs *engine = kobj_to_engine(kobj);
+
+	return sprintf(buf, "%lu\n", engine->props.preempt_timeout_ms);
+}
+
+static struct kobj_attribute preempt_timeout_attr =
+__ATTR(preempt_timeout_ms, 0644, preempt_timeout_show, preempt_timeout_store);
+
 static void kobj_engine_release(struct kobject *kobj)
 {
 	kfree(kobj);
@@ -292,6 +336,10 @@ void intel_engines_add_sysfs(struct drm_i915_private *i915)
 		    sysfs_create_file(kobj, &timeslice_duration_attr.attr))
 			goto err_engine;
 
+		if (intel_engine_has_preempt_reset(engine) &&
+		    sysfs_create_file(kobj, &preempt_timeout_attr.attr))
+			goto err_engine;
+
 		if (0) {
 err_object:
 			kobject_put(kobj);
-- 
2.24.0.rc0

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

  parent reply	other threads:[~2019-10-24 11:40 UTC|newest]

Thread overview: 40+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-10-24 11:40 CI testing of persistence and sysfs Chris Wilson
2019-10-24 11:40 ` [Intel-gfx] " Chris Wilson
2019-10-24 11:40 ` [PATCH 01/11] drm/i915/gem: Make context persistence optional Chris Wilson
2019-10-24 11:40   ` [Intel-gfx] " Chris Wilson
2019-10-25  9:10   ` Joonas Lahtinen
2019-10-25  9:10     ` [Intel-gfx] " Joonas Lahtinen
2019-10-25 18:22   ` Jason Ekstrand
2019-10-25 18:22     ` [Intel-gfx] " Jason Ekstrand
2019-10-25 21:29     ` Chris Wilson
2019-10-25 21:29       ` [Intel-gfx] " Chris Wilson
2019-10-29 16:19       ` Jason Ekstrand
2019-10-29 16:19         ` [Intel-gfx] " Jason Ekstrand
2019-10-29 18:02         ` Chris Wilson
2019-10-29 18:02           ` [Intel-gfx] " Chris Wilson
2019-10-24 11:40 ` [PATCH 02/11] drm/i915: Put future HW and their uAPIs under STAGING & BROKEN Chris Wilson
2019-10-24 11:40   ` [Intel-gfx] " Chris Wilson
2019-10-24 23:25   ` David Airlie
2019-10-24 23:25     ` [Intel-gfx] " David Airlie
2019-10-24 23:35   ` Rodrigo Vivi
2019-10-24 23:35     ` [Intel-gfx] " Rodrigo Vivi
2019-10-25  7:09   ` Jani Nikula
2019-10-25  7:09     ` [Intel-gfx] " Jani Nikula
2019-10-24 11:40 ` [PATCH 03/11] drm/i915/gt: Expose engine properties via sysfs Chris Wilson
2019-10-24 11:40   ` [Intel-gfx] " Chris Wilson
2019-10-24 11:40 ` [PATCH 04/11] drm/i915/gt: Expose engine->mmio_base " Chris Wilson
2019-10-24 11:40   ` [Intel-gfx] " Chris Wilson
2019-10-24 11:40 ` [PATCH 05/11] drm/i915/gt: Expose timeslice duration to sysfs Chris Wilson
2019-10-24 11:40   ` [Intel-gfx] " Chris Wilson
2019-10-24 11:40 ` [PATCH 06/11] drm/i915/gt: Expose reset stop timeout via sysfs Chris Wilson
2019-10-24 11:40   ` [Intel-gfx] " Chris Wilson
2019-10-24 11:40 ` Chris Wilson [this message]
2019-10-24 11:40   ` [Intel-gfx] [PATCH 07/11] drm/i915/gt: Expose preempt reset " Chris Wilson
2019-10-24 11:40 ` [PATCH 08/11] drm/i915/gt: Expose heartbeat interval " Chris Wilson
2019-10-24 11:40   ` [Intel-gfx] " Chris Wilson
2019-10-24 11:40 ` [PATCH 09/11] drm/i915: Flush idle barriers when waiting Chris Wilson
2019-10-24 11:40   ` [Intel-gfx] " Chris Wilson
2019-10-24 11:40 ` [PATCH 10/11] drm/i915: Allow userspace to specify ringsize on construction Chris Wilson
2019-10-24 11:40   ` [Intel-gfx] " Chris Wilson
2019-10-24 11:40 ` [PATCH 11/11] drm/i915/gem: Honour O_NONBLOCK before throttling execbuf submissions Chris Wilson
2019-10-24 11:40   ` [Intel-gfx] " 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=20191024114028.6170-8-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.