From mboxrd@z Thu Jan  1 00:00:00 1970
From: "Rafael J. Wysocki" <rjw@sisk.pl>
Subject: Re: [PATCH 4/5]PCIe native PME support
Date: Thu, 20 Aug 2009 22:07:20 +0200
Message-ID: <200908202207.20923.rjw@sisk.pl>
References: <1250666659.23178.119.camel@sli10-desk.sh.intel.com> <20090819115811.GD12216@srcf.ucam.org> <20090820031056.GB26357@sli10-desk.sh.intel.com>
Mime-Version: 1.0
Content-Type: Text/Plain;
  charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
Return-path: <linux-acpi-owner@vger.kernel.org>
In-Reply-To: <20090820031056.GB26357@sli10-desk.sh.intel.com>
Sender: linux-acpi-owner@vger.kernel.org
To: Shaohua Li <shaohua.li@intel.com>
Cc: Matthew Garrett <mjg59@srcf.ucam.org>, linux acpi <linux-acpi@vger.kernel.org>, linux-pm <linux-pm@lists.osdl.org>, Alan Stern <stern@rowland.harvard.edu>
List-Id: linux-pm@vger.kernel.org

On Thursday 20 August 2009, Shaohua Li wrote:
> On Wed, Aug 19, 2009 at 07:58:11PM +0800, Matthew Garrett wrote:
> > On Wed, Aug 19, 2009 at 03:24:19PM +0800, Shaohua Li wrote:
> > 
> > > +static inline void npme_enable_pme(struct pci_dev *pdev, bool enable)
> > > +{
> > > +	int pos;
> > > +	u16 rtctl;
> > > +
> > > +	pos = pci_find_capability(pdev, PCI_CAP_ID_EXP);
> > > +
> > > +	pci_read_config_word(pdev, pos + PCI_EXP_RTCTL, &rtctl);
> > > +	if (!enable)
> > > +		rtctl &= ~PCI_EXP_RTCTL_PMEIE;
> > > +	else
> > > +		rtctl |= PCI_EXP_RTCTL_PMEIE;
> > > +	pci_write_config_word(pdev, pos + PCI_EXP_RTCTL, rtctl);
> > > +}
> > 
> > This seems to duplicate the existing pci_pme_active() function?
> No, these registers are completely different. pci_pme_active is to handle
> the PM capability regiser. This routine is to handle the PCI express capability.
>  
> > > +static inline void npme_clear_pme(struct pci_dev *pdev)
> > > +{
> > > +	int pos;
> > > +	u32 rtsta;
> > > +
> > > +	pos = pci_find_capability(pdev, PCI_CAP_ID_EXP);
> > > +
> > > +	pci_read_config_dword(pdev, pos + PCI_EXP_RTSTA, &rtsta);
> > > +	rtsta |= PCI_EXP_RTSTA_PME;
> > > +	pci_write_config_dword(pdev, pos + PCI_EXP_RTSTA, rtsta);
> > > +}
> > 
> > Ditto.
> Ditto.
> 
> > > +static bool npme_pme_target(struct pci_dev *target)
> > > +{
> > > +	bool ret = false;
> > > +	if (target->dev.bus->pm && target->dev.bus->pm->wakeup_event)
> > > +		ret = target->dev.bus->pm->wakeup_event(&target->dev);
> > > +	return ret;
> > > +}
> > 
> > Is there any situation in which we wouldn't want to just perform a 
> > runtime resume of the device here?
> The devices pcie-to-pci bridge can send wakeup event, but the event might use
> root port or the bridge pci id, so can't directly send the event to the device.
> 
> Surely we can move the logic in my patch 3 here if we don't want to duplicate
> code, and then we can directly perform a runtime resume for target devices.

Do that, please.

Rafael