public inbox for intel-gfx@lists.freedesktop.org
 help / color / mirror / Atom feed
From: Paulo Zanoni <przanoni@gmail.com>
To: intel-gfx@lists.freedesktop.org
Cc: Paulo Zanoni <paulo.r.zanoni@intel.com>
Subject: [PATCH 01/23] drm/i915: Accurately track when we mark the hardware as idle/busy
Date: Thu, 27 Feb 2014 19:26:28 -0300	[thread overview]
Message-ID: <1393540010-1582-2-git-send-email-przanoni@gmail.com> (raw)
In-Reply-To: <1393540010-1582-1-git-send-email-przanoni@gmail.com>

From: Chris Wilson <chris@chris-wilson.co.uk>

We currently call intel_mark_idle() too often, as we do so as a
side-effect of processing the request queue. However, we the calls to
intel_mark_idle() are expected to be paired with a call to
intel_mark_busy() (or else we try to idle the hardware by accessing
registers that are already disabled). Make the idle/busy tracking
explicit to prevent the multiple calls.

v2: We can drop some of the complexity in __i915_add_request() as
queue_delayed_work() already behaves as we want (not requeuing the item
if it is already in the queue) and mark_busy/mark_idle imply that the
idle task is inactive.

v3: We do still need to cancel the pending idle task so that it is sent
again after the current busy load completes (not in the middle of it).

Reported-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Tested-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Paulo Zanoni <paulo.r.zanoni@intel.com>
---
 drivers/gpu/drm/i915/i915_drv.h      |  8 ++++++++
 drivers/gpu/drm/i915/i915_gem.c      | 14 +++++---------
 drivers/gpu/drm/i915/intel_display.c |  9 +++++++++
 3 files changed, 22 insertions(+), 9 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index 8c64831..a5caa7e 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -1124,6 +1124,14 @@ struct i915_gem_mm {
 	 */
 	bool interruptible;
 
+	/**
+	 * Is the GPU currently considered idle, or busy executing userspace
+	 * requests?  Whilst idle, we attempt to power down the hardware and
+	 * display clocks. In order to reduce the effect on performance, there
+	 * is a slight delay before we do so.
+	 */
+	bool busy;
+
 	/** Bit 6 swizzling required for X tiling */
 	uint32_t bit_6_swizzle_x;
 	/** Bit 6 swizzling required for Y tiling */
diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index 3618bb0..6978e69 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -2148,7 +2148,6 @@ int __i915_add_request(struct intel_ring_buffer *ring,
 	drm_i915_private_t *dev_priv = ring->dev->dev_private;
 	struct drm_i915_gem_request *request;
 	u32 request_ring_position, request_start;
-	int was_empty;
 	int ret;
 
 	request_start = intel_ring_get_tail(ring);
@@ -2199,7 +2198,6 @@ int __i915_add_request(struct intel_ring_buffer *ring,
 		i915_gem_context_reference(request->ctx);
 
 	request->emitted_jiffies = jiffies;
-	was_empty = list_empty(&ring->request_list);
 	list_add_tail(&request->list, &ring->request_list);
 	request->file_priv = NULL;
 
@@ -2220,13 +2218,11 @@ int __i915_add_request(struct intel_ring_buffer *ring,
 	if (!dev_priv->ums.mm_suspended) {
 		i915_queue_hangcheck(ring->dev);
 
-		if (was_empty) {
-			cancel_delayed_work_sync(&dev_priv->mm.idle_work);
-			queue_delayed_work(dev_priv->wq,
-					   &dev_priv->mm.retire_work,
-					   round_jiffies_up_relative(HZ));
-			intel_mark_busy(dev_priv->dev);
-		}
+		cancel_delayed_work_sync(&dev_priv->mm.idle_work);
+		queue_delayed_work(dev_priv->wq,
+				   &dev_priv->mm.retire_work,
+				   round_jiffies_up_relative(HZ));
+		intel_mark_busy(dev_priv->dev);
 	}
 
 	if (out_seqno)
diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
index f19e6ea..dd416f2 100644
--- a/drivers/gpu/drm/i915/intel_display.c
+++ b/drivers/gpu/drm/i915/intel_display.c
@@ -8192,8 +8192,12 @@ void intel_mark_busy(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
+	if (dev_priv->mm.busy)
+		return;
+
 	hsw_package_c8_gpu_busy(dev_priv);
 	i915_update_gfx_val(dev_priv);
+	dev_priv->mm.busy = true;
 }
 
 void intel_mark_idle(struct drm_device *dev)
@@ -8201,6 +8205,11 @@ void intel_mark_idle(struct drm_device *dev)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct drm_crtc *crtc;
 
+	if (!dev_priv->mm.busy)
+		return;
+
+	dev_priv->mm.busy = false;
+
 	hsw_package_c8_gpu_idle(dev_priv);
 
 	if (!i915.powersave)
-- 
1.8.5.3

  reply	other threads:[~2014-02-27 22:27 UTC|newest]

Thread overview: 55+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-02-27 22:26 [PATCH 00/23] Merge PC8 with runtime PM, v2 Paulo Zanoni
2014-02-27 22:26 ` Paulo Zanoni [this message]
2014-02-27 22:26 ` [PATCH 02/23] drm/i915: put runtime PM only at the end of intel_mark_idle Paulo Zanoni
2014-02-27 22:26 ` [PATCH 03/23] drm/i915: put runtime PM only when we actually release force_wake Paulo Zanoni
2014-02-27 22:26 ` [PATCH 04/23] drm/i915: kill dev_priv->pc8.gpu_idle Paulo Zanoni
2014-02-28 15:23   ` Imre Deak
2014-02-27 22:26 ` [PATCH 05/23] drm/i915: rename modeset_update_power_wells Paulo Zanoni
2014-02-27 22:26 ` [PATCH 06/23] drm/i915: get/put runtime PM without holding rps.hw_lock Paulo Zanoni
2014-02-27 22:26 ` [PATCH 07/23] drm/i915: add forcewake functions that don't touch runtime PM Paulo Zanoni
2014-02-28  8:44   ` Chris Wilson
2014-02-28 19:38     ` Paulo Zanoni
2014-02-28 19:46       ` Chris Wilson
2014-03-05 16:56   ` Daniel Vetter
2014-02-27 22:26 ` [PATCH 08/23] drm/i915: extract __hsw_do_{en, dis}able_package_c8 Paulo Zanoni
2014-02-27 22:26 ` [PATCH 09/23] drm/i915: make PC8 be part of runtime PM suspend/resume Paulo Zanoni
2014-02-28  9:42   ` Chris Wilson
2014-03-06 22:48     ` Daniel Vetter
2014-02-27 22:26 ` [PATCH 10/23] drm/i915: get/put runtime PM when we get/put a power domain Paulo Zanoni
2014-02-28 15:45   ` Imre Deak
2014-02-28 19:54     ` Paulo Zanoni
2014-02-27 22:26 ` [PATCH 11/23] drm/i915: remove dev_priv->pc8.requirements_met Paulo Zanoni
2014-02-27 22:26 ` [PATCH 12/23] drm/i915: get runtime PM references when the GPU is idle/busy Paulo Zanoni
2014-02-28 17:08   ` Jesse Barnes
2014-02-27 22:26 ` [PATCH 13/23] drm/i915: kill pc8.disable_count Paulo Zanoni
2014-02-28 17:10   ` Jesse Barnes
2014-02-27 22:26 ` [PATCH 14/23] drm/i915: remove an indirection level on PC8 functions Paulo Zanoni
2014-02-28 17:11   ` Jesse Barnes
2014-02-27 22:26 ` [PATCH 15/23] drm/i915: don't get/put PC8 reference on freeze/thaw Paulo Zanoni
2014-02-28 17:11   ` Jesse Barnes
2014-02-27 22:26 ` [PATCH 16/23] drm/i915: get/put runtime PM references for GMBUS and DP AUX Paulo Zanoni
2014-02-28 17:13   ` Jesse Barnes
2014-02-28 17:38     ` Imre Deak
2014-02-28 17:55       ` Jesse Barnes
2014-02-28 18:20         ` Imre Deak
2014-02-28 19:07           ` Paulo Zanoni
2014-03-05 17:04         ` Daniel Vetter
2014-02-27 22:26 ` [PATCH 17/23] drm/i915: don't get/put PC8 when getting/putting power wells Paulo Zanoni
2014-02-28 17:13   ` Jesse Barnes
2014-02-27 22:26 ` [PATCH 18/23] drm/i915: remove dev_priv->pc8.enabled Paulo Zanoni
2014-02-28 17:14   ` Jesse Barnes
2014-02-27 22:26 ` [PATCH 19/23] drm/i915: move pc8.irqs_disabled to pm.irqs_disabled Paulo Zanoni
2014-02-28 17:15   ` Jesse Barnes
2014-02-28 18:17     ` Paulo Zanoni
2014-02-27 22:26 ` [PATCH 20/23] drm/i915: kill struct i915_package_c8 Paulo Zanoni
2014-02-28 17:16   ` Jesse Barnes
2014-02-27 22:26 ` [PATCH 21/23] drm/i915: rename __hsw_do_{en, dis}able_pc8 Paulo Zanoni
2014-02-28 17:17   ` Jesse Barnes
2014-02-27 22:26 ` [PATCH 22/23] drm/i915: update the PC8 and runtime PM documentation Paulo Zanoni
2014-02-28 17:19   ` Jesse Barnes
2014-02-27 22:26 ` [PATCH 23/23] drm/i915: init pm.suspended earlier Paulo Zanoni
2014-02-28 17:20   ` Jesse Barnes
2014-03-05 17:09 ` [PATCH 00/23] Merge PC8 with runtime PM, v2 Daniel Vetter
2014-03-06 22:23   ` Paulo Zanoni
2014-03-06 22:45     ` Daniel Vetter
2014-03-07  8:51       ` Daniel Vetter

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=1393540010-1582-2-git-send-email-przanoni@gmail.com \
    --to=przanoni@gmail.com \
    --cc=intel-gfx@lists.freedesktop.org \
    --cc=paulo.r.zanoni@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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox