diff mbox

drm/i915: Keep track of active forcewake domains in a bitmask

Message ID 1474978310-5423-1-git-send-email-tvrtko.ursulin@linux.intel.com (mailing list archive)
State New, archived
Headers show

Commit Message

Tvrtko Ursulin Sept. 27, 2016, 12:11 p.m. UTC 
From: Tvrtko Ursulin <tvrtko.ursulin@intel.com>

There are current places in the code, and there will be more in the
future, which iterate the forcewake domains to find out which ones
are currently active.

To save them from doing this iteration, we can cheaply keep a mask
of active domains in dev_priv->uncore.fw_domains_active.

This has no cost in terms of object size, even manages to shrink it
overall by 368 bytes on my config.

Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: "Paneri, Praveen" <praveen.paneri@intel.com>
---
 drivers/gpu/drm/i915/i915_drv.h     |  2 ++
 drivers/gpu/drm/i915/intel_uncore.c | 54 ++++++++++++++++---------------------
 2 files changed, 25 insertions(+), 31 deletions(-)

Comments

Chris Wilson Sept. 27, 2016, 12:26 p.m. UTC | #1 
On Tue, Sep 27, 2016 at 01:11:50PM +0100, Tvrtko Ursulin wrote:
> From: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
> 
> There are current places in the code, and there will be more in the
> future, which iterate the forcewake domains to find out which ones
> are currently active.
> 
> To save them from doing this iteration, we can cheaply keep a mask
> of active domains in dev_priv->uncore.fw_domains_active.
> 
> This has no cost in terms of object size, even manages to shrink it
> overall by 368 bytes on my config.
> 
> Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
> Cc: "Paneri, Praveen" <praveen.paneri@intel.com>

Looks solid and should not run afoul of the games played during reset or
suspend.

Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
-Chris
Tvrtko Ursulin Sept. 27, 2016, 1:21 p.m. UTC | #2 
On 27/09/2016 13:26, Chris Wilson wrote:
> On Tue, Sep 27, 2016 at 01:11:50PM +0100, Tvrtko Ursulin wrote:
>> From: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
>>
>> There are current places in the code, and there will be more in the
>> future, which iterate the forcewake domains to find out which ones
>> are currently active.
>>
>> To save them from doing this iteration, we can cheaply keep a mask
>> of active domains in dev_priv->uncore.fw_domains_active.
>>
>> This has no cost in terms of object size, even manages to shrink it
>> overall by 368 bytes on my config.
>>
>> Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
>> Cc: "Paneri, Praveen" <praveen.paneri@intel.com>
> Looks solid and should not run afoul of the games played during reset or
> suspend.
>
> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>

Thanks!

Praveen, you can see if this patch would help for your decoupled MMIO 
patches, and if you like it you can incrorporate it in your series if 
you want?

Regards,

Tvrtko
Praveen Paneri Sept. 28, 2016, 3:33 a.m. UTC | #3 
On 9/27/2016 6:51 PM, Tvrtko Ursulin wrote:
>
> On 27/09/2016 13:26, Chris Wilson wrote:
>> On Tue, Sep 27, 2016 at 01:11:50PM +0100, Tvrtko Ursulin wrote:
>>> From: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
>>>
>>> There are current places in the code, and there will be more in the
>>> future, which iterate the forcewake domains to find out which ones
>>> are currently active.
>>>
>>> To save them from doing this iteration, we can cheaply keep a mask
>>> of active domains in dev_priv->uncore.fw_domains_active.
>>>
>>> This has no cost in terms of object size, even manages to shrink it
>>> overall by 368 bytes on my config.
>>>
>>> Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
>>> Cc: "Paneri, Praveen" <praveen.paneri@intel.com>
>> Looks solid and should not run afoul of the games played during reset or
>> suspend.
>>
>> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
>
> Thanks!
>
> Praveen, you can see if this patch would help for your decoupled MMIO
> patches, and if you like it you can incrorporate it in your series if
> you want?
Thanks for the patch Tvrtko! I can definitely use it :)
>
> Regards,
>
> Tvrtko
>
diff mbox

Patch

diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index 23bc43d23d2c..4cd727376d1d 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -588,7 +588,9 @@  struct intel_uncore {
 	struct intel_uncore_funcs funcs;
 
 	unsigned fifo_count;
+
 	enum forcewake_domains fw_domains;
+	enum forcewake_domains fw_domains_active;
 
 	struct intel_uncore_forcewake_domain {
 		struct drm_i915_private *i915;
diff --git a/drivers/gpu/drm/i915/intel_uncore.c b/drivers/gpu/drm/i915/intel_uncore.c
index a9b6c936aadd..6cd1e78dc08d 100644
--- a/drivers/gpu/drm/i915/intel_uncore.c
+++ b/drivers/gpu/drm/i915/intel_uncore.c
@@ -231,19 +231,21 @@  intel_uncore_fw_release_timer(struct hrtimer *timer)
 {
 	struct intel_uncore_forcewake_domain *domain =
 	       container_of(timer, struct intel_uncore_forcewake_domain, timer);
+	struct drm_i915_private *dev_priv = domain->i915;
 	unsigned long irqflags;
 
-	assert_rpm_device_not_suspended(domain->i915);
+	assert_rpm_device_not_suspended(dev_priv);
 
-	spin_lock_irqsave(&domain->i915->uncore.lock, irqflags);
+	spin_lock_irqsave(&dev_priv->uncore.lock, irqflags);
 	if (WARN_ON(domain->wake_count == 0))
 		domain->wake_count++;
 
-	if (--domain->wake_count == 0)
-		domain->i915->uncore.funcs.force_wake_put(domain->i915,
-							  1 << domain->id);
+	if (--domain->wake_count == 0) {
+		dev_priv->uncore.funcs.force_wake_put(dev_priv, domain->mask);
+		dev_priv->uncore.fw_domains_active &= ~domain->mask;
+	}
 
-	spin_unlock_irqrestore(&domain->i915->uncore.lock, irqflags);
+	spin_unlock_irqrestore(&dev_priv->uncore.lock, irqflags);
 
 	return HRTIMER_NORESTART;
 }
@@ -254,7 +256,7 @@  void intel_uncore_forcewake_reset(struct drm_i915_private *dev_priv,
 	unsigned long irqflags;
 	struct intel_uncore_forcewake_domain *domain;
 	int retry_count = 100;
-	enum forcewake_domains fw = 0, active_domains;
+	enum forcewake_domains fw, active_domains;
 
 	/* Hold uncore.lock across reset to prevent any register access
 	 * with forcewake not set correctly. Wait until all pending
@@ -291,10 +293,7 @@  void intel_uncore_forcewake_reset(struct drm_i915_private *dev_priv,
 
 	WARN_ON(active_domains);
 
-	for_each_fw_domain(domain, dev_priv)
-		if (domain->wake_count)
-			fw |= domain->mask;
-
+	fw = dev_priv->uncore.fw_domains_active;
 	if (fw)
 		dev_priv->uncore.funcs.force_wake_put(dev_priv, fw);
 
@@ -443,9 +442,6 @@  static void __intel_uncore_forcewake_get(struct drm_i915_private *dev_priv,
 {
 	struct intel_uncore_forcewake_domain *domain;
 
-	if (!dev_priv->uncore.funcs.force_wake_get)
-		return;
-
 	fw_domains &= dev_priv->uncore.fw_domains;
 
 	for_each_fw_domain_masked(domain, fw_domains, dev_priv) {
@@ -453,8 +449,10 @@  static void __intel_uncore_forcewake_get(struct drm_i915_private *dev_priv,
 			fw_domains &= ~domain->mask;
 	}
 
-	if (fw_domains)
+	if (fw_domains) {
 		dev_priv->uncore.funcs.force_wake_get(dev_priv, fw_domains);
+		dev_priv->uncore.fw_domains_active |= fw_domains;
+	}
 }
 
 /**
@@ -509,9 +507,6 @@  static void __intel_uncore_forcewake_put(struct drm_i915_private *dev_priv,
 {
 	struct intel_uncore_forcewake_domain *domain;
 
-	if (!dev_priv->uncore.funcs.force_wake_put)
-		return;
-
 	fw_domains &= dev_priv->uncore.fw_domains;
 
 	for_each_fw_domain_masked(domain, fw_domains, dev_priv) {
@@ -567,13 +562,10 @@  void intel_uncore_forcewake_put__locked(struct drm_i915_private *dev_priv,
 
 void assert_forcewakes_inactive(struct drm_i915_private *dev_priv)
 {
-	struct intel_uncore_forcewake_domain *domain;
-
 	if (!dev_priv->uncore.funcs.force_wake_get)
 		return;
 
-	for_each_fw_domain(domain, dev_priv)
-		WARN_ON(domain->wake_count);
+	WARN_ON(dev_priv->uncore.fw_domains_active);
 }
 
 /* We give fast paths for the really cool registers */
@@ -878,18 +870,18 @@  static inline void __force_wake_auto(struct drm_i915_private *dev_priv,
 	if (WARN_ON(!fw_domains))
 		return;
 
-	/* Ideally GCC would be constant-fold and eliminate this loop */
-	for_each_fw_domain_masked(domain, fw_domains, dev_priv) {
-		if (domain->wake_count) {
-			fw_domains &= ~domain->mask;
-			continue;
-		}
+	/* Turn on all requested but inactive supported forcewake domains. */
+	fw_domains &= dev_priv->uncore.fw_domains;
+	fw_domains &= ~dev_priv->uncore.fw_domains_active;
 
-		fw_domain_arm_timer(domain);
-	}
+	if (fw_domains) {
+		/* Ideally GCC would be constant-fold and eliminate this loop */
+		for_each_fw_domain_masked(domain, fw_domains, dev_priv)
+			fw_domain_arm_timer(domain);
 
-	if (fw_domains)
 		dev_priv->uncore.funcs.force_wake_get(dev_priv, fw_domains);
+		dev_priv->uncore.fw_domains_active |= fw_domains;
+	}
 }
 
 #define __gen6_read(x) \