All of lore.kernel.org
 help / color / mirror / Atom feed
* [RFC] PCI/PM: Keep runtime PM enabled for unbound PCI devices
@ 2012-11-19  3:14 Huang Ying
  2012-11-19 16:31   ` Alan Stern
  0 siblings, 1 reply; 5+ messages in thread
From: Huang Ying @ 2012-11-19  3:14 UTC (permalink / raw)
  To: Bjorn Helgaas
  Cc: linux-kernel, linux-pci, linux-pm, Rafael J. Wysocki, Huang Ying,
	Rafael J. Wysocki, Alan Stern

For unbound PCI devices, what we need is:

 - Always in D0 state, because some devices does not work again after
   being put into D3 by the PCI bus.

 - In SUSPENDED state if allowed, so that the parent devices can still
   be put into low power state.

To satisfy these requirement, the runtime PM for the unbound PCI
devices are disabled and set to SUSPENDED state.  One issue of this
solution is that the PCI devices will be put into SUSPENDED state even
if the SUSPENDED state is forbidden via the sysfs interface
(.../power/control) of the device.  This is not an issue for most
devices, because most PCI devices are not used at all if unbounded.
But there are exceptions.  For example, unbound VGA card can be used
for display, but suspend its parents make it stop working.

To fix the issue, we keep the runtime PM enabled when the PCI devices
are unbound.  But the runtime PM callbacks will do nothing if the PCI
devices are unbound.  This way, we can put the PCI devices into
SUSPENDED state without put the PCI devices into D3 state.

Known issue: after some changing, pci_dev->driver is used to indicate
whether the PCI devices are bound and whether the runtime PM callbacks
should do nothing.  Maybe it is better to use a dedicated flag such as
.skip_rtpm_callbacks.  That may improve code readability.

Signed-off-by: Huang Ying <ying.huang@intel.com>
CC: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
CC: Alan Stern <stern@rowland.harvard.edu>
CC: stable@vger.kernel.org          # v3.6+
---
 drivers/pci/pci-driver.c |   57 +++++++++++++++++++++++------------------------
 drivers/pci/pci.c        |    2 +
 2 files changed, 31 insertions(+), 28 deletions(-)

--- a/drivers/pci/pci.c
+++ b/drivers/pci/pci.c
@@ -1900,6 +1900,8 @@ void pci_pm_init(struct pci_dev *dev)
 	u16 pmc;
 
 	pm_runtime_forbid(&dev->dev);
+	pm_runtime_set_active(&dev->dev);
+	pm_runtime_enable(&dev->dev);
 	device_enable_async_suspend(&dev->dev);
 	dev->wakeup_prepared = false;
 
--- a/drivers/pci/pci-driver.c
+++ b/drivers/pci/pci-driver.c
@@ -256,31 +256,26 @@ struct drv_dev_and_id {
 static long local_pci_probe(void *_ddi)
 {
 	struct drv_dev_and_id *ddi = _ddi;
-	struct device *dev = &ddi->dev->dev;
-	struct device *parent = dev->parent;
+	struct pci_dev *pci_dev = ddi->dev;
+	struct pci_driver *pci_drv = ddi->drv;
+	struct device *dev = &pci_dev->dev;
 	int rc;
 
-	/* The parent bridge must be in active state when probing */
-	if (parent)
-		pm_runtime_get_sync(parent);
-	/* Unbound PCI devices are always set to disabled and suspended.
-	 * During probe, the device is set to enabled and active and the
-	 * usage count is incremented.  If the driver supports runtime PM,
-	 * it should call pm_runtime_put_noidle() in its probe routine and
-	 * pm_runtime_get_noresume() in its remove routine.
-	 */
-	pm_runtime_get_noresume(dev);
-	pm_runtime_set_active(dev);
-	pm_runtime_enable(dev);
-
-	rc = ddi->drv->probe(ddi->dev, ddi->id);
+	/*
+	 * Unbound PCI devices are always set to suspended and put in
+	 * D0.  During probe, the device is set to active and the
+	 * usage count is incremented.  If the driver supports runtime
+	 * PM, it should call pm_runtime_put_noidle() in its probe
+	 * routine and pm_runtime_get_noresume() in its remove
+	 * routine.
+	 */
+	pm_runtime_get_sync(dev);
+	pci_dev->driver = pci_drv;
+	rc = pci_drv->probe(pci_dev, ddi->id);
 	if (rc) {
-		pm_runtime_disable(dev);
-		pm_runtime_set_suspended(dev);
-		pm_runtime_put_noidle(dev);
+		pci_dev->driver = NULL;
+		pm_runtime_put_sync(dev);
 	}
-	if (parent)
-		pm_runtime_put(parent);
 	return rc;
 }
 
@@ -330,10 +325,8 @@ __pci_device_probe(struct pci_driver *dr
 		id = pci_match_device(drv, pci_dev);
 		if (id)
 			error = pci_call_probe(drv, pci_dev, id);
-		if (error >= 0) {
-			pci_dev->driver = drv;
+		if (error >= 0)
 			error = 0;
-		}
 	}
 	return error;
 }
@@ -369,9 +362,7 @@ static int pci_device_remove(struct devi
 	}
 
 	/* Undo the runtime PM settings in local_pci_probe() */
-	pm_runtime_disable(dev);
-	pm_runtime_set_suspended(dev);
-	pm_runtime_put_noidle(dev);
+	pm_runtime_put_sync(dev);
 
 	/*
 	 * If the device is still on, set the power state as "unknown",
@@ -994,6 +985,9 @@ static int pci_pm_runtime_suspend(struct
 	pci_power_t prev = pci_dev->current_state;
 	int error;
 
+	if (!pci_dev->driver)
+		return 0;
+
 	if (!pm || !pm->runtime_suspend)
 		return -ENOSYS;
 
@@ -1029,6 +1023,9 @@ static int pci_pm_runtime_resume(struct
 	struct pci_dev *pci_dev = to_pci_dev(dev);
 	const struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL;
 
+	if (!pci_dev->driver)
+		return 0;
+
 	if (!pm || !pm->runtime_resume)
 		return -ENOSYS;
 
@@ -1046,8 +1043,12 @@ static int pci_pm_runtime_resume(struct
 
 static int pci_pm_runtime_idle(struct device *dev)
 {
+	struct pci_dev *pci_dev = to_pci_dev(dev);
 	const struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL;
 
+	if (!pci_dev->driver)
+		goto out;
+
 	if (!pm)
 		return -ENOSYS;
 
@@ -1057,8 +1058,8 @@ static int pci_pm_runtime_idle(struct de
 			return ret;
 	}
 
+out:
 	pm_runtime_suspend(dev);
-
 	return 0;
 }
 

^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: [RFC] PCI/PM: Keep runtime PM enabled for unbound PCI devices
  2012-11-19  3:14 [RFC] PCI/PM: Keep runtime PM enabled for unbound PCI devices Huang Ying
@ 2012-11-19 16:31   ` Alan Stern
  0 siblings, 0 replies; 5+ messages in thread
From: Alan Stern @ 2012-11-19 16:31 UTC (permalink / raw)
  To: Huang Ying
  Cc: Bjorn Helgaas, linux-kernel, linux-pci, linux-pm,
	Rafael J. Wysocki, Rafael J. Wysocki

On Mon, 19 Nov 2012, Huang Ying wrote:

> For unbound PCI devices, what we need is:
> 
>  - Always in D0 state, because some devices does not work again after
>    being put into D3 by the PCI bus.
> 
>  - In SUSPENDED state if allowed, so that the parent devices can still
>    be put into low power state.
> 
> To satisfy these requirement, the runtime PM for the unbound PCI
> devices are disabled and set to SUSPENDED state.  One issue of this
> solution is that the PCI devices will be put into SUSPENDED state even
> if the SUSPENDED state is forbidden via the sysfs interface
> (.../power/control) of the device.  This is not an issue for most
> devices, because most PCI devices are not used at all if unbounded.
> But there are exceptions.  For example, unbound VGA card can be used
> for display, but suspend its parents make it stop working.
> 
> To fix the issue, we keep the runtime PM enabled when the PCI devices
> are unbound.  But the runtime PM callbacks will do nothing if the PCI
> devices are unbound.  This way, we can put the PCI devices into
> SUSPENDED state without put the PCI devices into D3 state.
> 
> Known issue: after some changing, pci_dev->driver is used to indicate
> whether the PCI devices are bound and whether the runtime PM callbacks
> should do nothing.  Maybe it is better to use a dedicated flag such as
> .skip_rtpm_callbacks.  That may improve code readability.

I think it's okay like this, especially if you add a comment in 
pci_runtime_suspend, pci_runtime_resume, and pci_runtime_idle 
explaining that when pci_dev->driver isn't set, the device should 
always remain in D0 regardless of the runtime status.

Alan Stern


^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: [RFC] PCI/PM: Keep runtime PM enabled for unbound PCI devices
@ 2012-11-19 16:31   ` Alan Stern
  0 siblings, 0 replies; 5+ messages in thread
From: Alan Stern @ 2012-11-19 16:31 UTC (permalink / raw)
  To: Huang Ying
  Cc: Bjorn Helgaas, linux-kernel, linux-pci, linux-pm,
	Rafael J. Wysocki, Rafael J. Wysocki

On Mon, 19 Nov 2012, Huang Ying wrote:

> For unbound PCI devices, what we need is:
> 
>  - Always in D0 state, because some devices does not work again after
>    being put into D3 by the PCI bus.
> 
>  - In SUSPENDED state if allowed, so that the parent devices can still
>    be put into low power state.
> 
> To satisfy these requirement, the runtime PM for the unbound PCI
> devices are disabled and set to SUSPENDED state.  One issue of this
> solution is that the PCI devices will be put into SUSPENDED state even
> if the SUSPENDED state is forbidden via the sysfs interface
> (.../power/control) of the device.  This is not an issue for most
> devices, because most PCI devices are not used at all if unbounded.
> But there are exceptions.  For example, unbound VGA card can be used
> for display, but suspend its parents make it stop working.
> 
> To fix the issue, we keep the runtime PM enabled when the PCI devices
> are unbound.  But the runtime PM callbacks will do nothing if the PCI
> devices are unbound.  This way, we can put the PCI devices into
> SUSPENDED state without put the PCI devices into D3 state.
> 
> Known issue: after some changing, pci_dev->driver is used to indicate
> whether the PCI devices are bound and whether the runtime PM callbacks
> should do nothing.  Maybe it is better to use a dedicated flag such as
> .skip_rtpm_callbacks.  That may improve code readability.

I think it's okay like this, especially if you add a comment in 
pci_runtime_suspend, pci_runtime_resume, and pci_runtime_idle 
explaining that when pci_dev->driver isn't set, the device should 
always remain in D0 regardless of the runtime status.

Alan Stern

^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: [RFC] PCI/PM: Keep runtime PM enabled for unbound PCI devices
  2012-11-19 16:31   ` Alan Stern
  (?)
@ 2012-11-19 22:00   ` Rafael J. Wysocki
  2012-11-20  0:52     ` Huang Ying
  -1 siblings, 1 reply; 5+ messages in thread
From: Rafael J. Wysocki @ 2012-11-19 22:00 UTC (permalink / raw)
  To: Alan Stern, Huang Ying
  Cc: Bjorn Helgaas, linux-kernel, linux-pci, linux-pm, Rafael J. Wysocki

On Monday, November 19, 2012 11:31:01 AM Alan Stern wrote:
> On Mon, 19 Nov 2012, Huang Ying wrote:
> 
> > For unbound PCI devices, what we need is:
> > 
> >  - Always in D0 state, because some devices does not work again after
> >    being put into D3 by the PCI bus.
> > 
> >  - In SUSPENDED state if allowed, so that the parent devices can still
> >    be put into low power state.
> > 
> > To satisfy these requirement, the runtime PM for the unbound PCI
> > devices are disabled and set to SUSPENDED state.  One issue of this
> > solution is that the PCI devices will be put into SUSPENDED state even
> > if the SUSPENDED state is forbidden via the sysfs interface
> > (.../power/control) of the device.  This is not an issue for most
> > devices, because most PCI devices are not used at all if unbounded.
> > But there are exceptions.  For example, unbound VGA card can be used
> > for display, but suspend its parents make it stop working.
> > 
> > To fix the issue, we keep the runtime PM enabled when the PCI devices
> > are unbound.  But the runtime PM callbacks will do nothing if the PCI
> > devices are unbound.  This way, we can put the PCI devices into
> > SUSPENDED state without put the PCI devices into D3 state.
> > 
> > Known issue: after some changing, pci_dev->driver is used to indicate
> > whether the PCI devices are bound and whether the runtime PM callbacks
> > should do nothing.  Maybe it is better to use a dedicated flag such as
> > .skip_rtpm_callbacks.  That may improve code readability.
> 
> I think it's okay like this, especially if you add a comment in 
> pci_runtime_suspend, pci_runtime_resume, and pci_runtime_idle 
> explaining that when pci_dev->driver isn't set, the device should 
> always remain in D0 regardless of the runtime status.

Yes, I agree with Alan.  Please add comments as Alan's suggesting and it
should be fine.

Thanks,
Rafael



-- 
I speak only for myself.
Rafael J. Wysocki, Intel Open Source Technology Center.

^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: [RFC] PCI/PM: Keep runtime PM enabled for unbound PCI devices
  2012-11-19 22:00   ` Rafael J. Wysocki
@ 2012-11-20  0:52     ` Huang Ying
  0 siblings, 0 replies; 5+ messages in thread
From: Huang Ying @ 2012-11-20  0:52 UTC (permalink / raw)
  To: Rafael J. Wysocki
  Cc: Alan Stern, Bjorn Helgaas, linux-kernel, linux-pci, linux-pm,
	Rafael J. Wysocki

On Mon, 2012-11-19 at 23:00 +0100, Rafael J. Wysocki wrote:
> On Monday, November 19, 2012 11:31:01 AM Alan Stern wrote:
> > On Mon, 19 Nov 2012, Huang Ying wrote:
> > 
> > > For unbound PCI devices, what we need is:
> > > 
> > >  - Always in D0 state, because some devices does not work again after
> > >    being put into D3 by the PCI bus.
> > > 
> > >  - In SUSPENDED state if allowed, so that the parent devices can still
> > >    be put into low power state.
> > > 
> > > To satisfy these requirement, the runtime PM for the unbound PCI
> > > devices are disabled and set to SUSPENDED state.  One issue of this
> > > solution is that the PCI devices will be put into SUSPENDED state even
> > > if the SUSPENDED state is forbidden via the sysfs interface
> > > (.../power/control) of the device.  This is not an issue for most
> > > devices, because most PCI devices are not used at all if unbounded.
> > > But there are exceptions.  For example, unbound VGA card can be used
> > > for display, but suspend its parents make it stop working.
> > > 
> > > To fix the issue, we keep the runtime PM enabled when the PCI devices
> > > are unbound.  But the runtime PM callbacks will do nothing if the PCI
> > > devices are unbound.  This way, we can put the PCI devices into
> > > SUSPENDED state without put the PCI devices into D3 state.
> > > 
> > > Known issue: after some changing, pci_dev->driver is used to indicate
> > > whether the PCI devices are bound and whether the runtime PM callbacks
> > > should do nothing.  Maybe it is better to use a dedicated flag such as
> > > .skip_rtpm_callbacks.  That may improve code readability.
> > 
> > I think it's okay like this, especially if you add a comment in 
> > pci_runtime_suspend, pci_runtime_resume, and pci_runtime_idle 
> > explaining that when pci_dev->driver isn't set, the device should 
> > always remain in D0 regardless of the runtime status.
> 
> Yes, I agree with Alan.  Please add comments as Alan's suggesting and it
> should be fine.

Sure.  Will do that.

Best Regards,
Huang Ying



^ permalink raw reply	[flat|nested] 5+ messages in thread

end of thread, other threads:[~2012-11-20  0:52 UTC | newest]

Thread overview: 5+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2012-11-19  3:14 [RFC] PCI/PM: Keep runtime PM enabled for unbound PCI devices Huang Ying
2012-11-19 16:31 ` Alan Stern
2012-11-19 16:31   ` Alan Stern
2012-11-19 22:00   ` Rafael J. Wysocki
2012-11-20  0:52     ` Huang Ying

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.