Merge tag 'for-3.8' of git://git.kernel.org/pub/scm/linux/kernel/git/helgaas/pci

Pull PCI update from Bjorn Helgaas:
 "Host bridge hotplug:
   - Untangle _PRT from struct pci_bus (Bjorn Helgaas)
   - Request _OSC control before scanning root bus (Taku Izumi)
   - Assign resources when adding host bridge (Yinghai Lu)
   - Remove root bus when removing host bridge (Yinghai Lu)
   - Remove _PRT during hot remove (Yinghai Lu)

  SRIOV
    - Add sysfs knobs to control numVFs (Don Dutile)

  Power management
   - Notify devices when power resource turned on (Huang Ying)

  Bug fixes
   - Work around broken _SEG on HP xw9300 (Bjorn Helgaas)
   - Keep runtime PM enabled for unbound PCI devices (Huang Ying)
   - Fix Optimus dual-GPU runtime D3 suspend issue (Dave Airlie)
   - Fix xen frontend shutdown issue (David Vrabel)
   - Work around PLX PCI 9050 BAR alignment erratum (Ian Abbott)

  Miscellaneous
   - Add GPL license for drivers/pci/ioapic (Andrew Cooks)
   - Add standard PCI-X, PCIe ASPM register #defines (Bjorn Helgaas)
   - NumaChip remote PCI support (Daniel Blueman)
   - Fix PCIe Link Capabilities Supported Link Speed definition (Jingoo
     Han)
   - Convert dev_printk() to dev_info(), etc (Joe Perches)
   - Add support for non PCI BAR ROM data (Matthew Garrett)
   - Add x86 support for host bridge translation offset (Mike Yoknis)
   - Report success only when every driver supports AER (Vijay
     Pandarathil)"

Fix up trivial conflicts.

* tag 'for-3.8' of git://git.kernel.org/pub/scm/linux/kernel/git/helgaas/pci: (48 commits)
  PCI: Use phys_addr_t for physical ROM address
  x86/PCI: Add NumaChip remote PCI support
  ath9k: Use standard #defines for PCIe Capability ASPM fields
  iwlwifi: Use standard #defines for PCIe Capability ASPM fields
  iwlwifi: collapse wrapper for pcie_capability_read_word()
  iwlegacy: Use standard #defines for PCIe Capability ASPM fields
  iwlegacy: collapse wrapper for pcie_capability_read_word()
  cxgb3: Use standard #defines for PCIe Capability ASPM fields
  PCI: Add standard PCIe Capability Link ASPM field names
  PCI/portdrv: Use PCI Express Capability accessors
  PCI: Use standard PCIe Capability Link register field names
  x86: Use PCI setup data
  PCI: Add support for non-BAR ROMs
  PCI: Add pcibios_add_device
  EFI: Stash ROMs if they're not in the PCI BAR
  PCI: Add and use standard PCI-X Capability register names
  PCI/PM: Keep runtime PM enabled for unbound PCI devices
  xen-pcifront: Handle backend CLOSED without CLOSING
  PCI: SRIOV control and status via sysfs (documentation)
  PCI/AER: Report success only when every device has AER-aware driver
  ...

1 files changed, 36 insertions, 12 deletions

diff --git a/drivers/pci/pci.c b/drivers/pci/pci.c
index bdf66b500f2..5cb5820fae4 100644
--- a/drivers/pci/pci.c
+++ b/drivers/pci/pci.c
@@ -1333,6 +1333,19 @@ void pcim_pin_device(struct pci_dev *pdev)
 		dr->pinned = 1;
 }
 
+/*
+ * pcibios_add_device - provide arch specific hooks when adding device dev
+ * @dev: the PCI device being added
+ *
+ * Permits the platform to provide architecture specific functionality when
+ * devices are added. This is the default implementation. Architecture
+ * implementations can override this.
+ */
+int __weak pcibios_add_device (struct pci_dev *dev)
+{
+	return 0;
+}
+
 /**
  * pcibios_disable_device - disable arch specific PCI resources for device dev
  * @dev: the PCI device to disable
@@ -1578,15 +1591,25 @@ void pci_pme_active(struct pci_dev *dev, bool enable)
 
 	pci_write_config_word(dev, dev->pm_cap + PCI_PM_CTRL, pmcsr);
 
-	/* PCI (as opposed to PCIe) PME requires that the device have
-	   its PME# line hooked up correctly. Not all hardware vendors
-	   do this, so the PME never gets delivered and the device
-	   remains asleep. The easiest way around this is to
-	   periodically walk the list of suspended devices and check
-	   whether any have their PME flag set. The assumption is that
-	   we'll wake up often enough anyway that this won't be a huge
-	   hit, and the power savings from the devices will still be a
-	   win. */
+	/*
+	 * PCI (as opposed to PCIe) PME requires that the device have
+	 * its PME# line hooked up correctly. Not all hardware vendors
+	 * do this, so the PME never gets delivered and the device
+	 * remains asleep. The easiest way around this is to
+	 * periodically walk the list of suspended devices and check
+	 * whether any have their PME flag set. The assumption is that
+	 * we'll wake up often enough anyway that this won't be a huge
+	 * hit, and the power savings from the devices will still be a
+	 * win.
+	 *
+	 * Although PCIe uses in-band PME message instead of PME# line
+	 * to report PME, PME does not work for some PCIe devices in
+	 * reality.  For example, there are devices that set their PME
+	 * status bits, but don't really bother to send a PME message;
+	 * there are PCI Express Root Ports that don't bother to
+	 * trigger interrupts when they receive PME messages from the
+	 * devices below.  So PME poll is used for PCIe devices too.
+	 */
 
 	if (dev->pme_poll) {
 		struct pci_pme_device *pme_dev;
@@ -1900,6 +1923,8 @@ 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;
 
@@ -3865,14 +3890,13 @@ static void pci_no_domains(void)
 }
 
 /**
- * pci_ext_cfg_enabled - can we access extended PCI config space?
- * @dev: The PCI device of the root bridge.
+ * pci_ext_cfg_avail - can we access extended PCI config space?
  *
  * Returns 1 if we can access PCI extended config space (offsets
  * greater than 0xff). This is the default implementation. Architecture
  * implementations can override this.
  */
-int __weak pci_ext_cfg_avail(struct pci_dev *dev)
+int __weak pci_ext_cfg_avail(void)
 {
 	return 1;
 }