Donate to e Foundation | Murena handsets with /e/OS | Own a part of Murena! Learn more

Commit ef1b5dad authored by Bjorn Helgaas's avatar Bjorn Helgaas
Browse files

Merge branch 'pci/virtualization' into next 

* pci/virtualization:
  ixgbe: Use pcie_flr() instead of duplicating it
  IB/hfi1: Use pcie_flr() instead of duplicating it
  PCI: Call pcie_flr() from reset_chelsio_generic_dev()
  PCI: Call pcie_flr() from reset_intel_82599_sfp_virtfn()
  PCI: Export pcie_flr()
  PCI: Add sysfs sriov_drivers_autoprobe to control VF driver binding
  PCI: Avoid FLR for Intel 82579 NICs

Conflicts:
	include/linux/pci.h
parents 889e4dd9 63af8f7a

Documentation/ABI/testing/sysfs-bus-pci

+22 −0
Original line number Diff line number Diff line
@@ -301,3 +301,25 @@ Contact: Emil Velikov <emil.l.velikov@gmail.com> 
Description:

		This file contains the revision field of the the PCI device.

		The value comes from device config space. The file is read only.



What:		/sys/bus/pci/devices/.../sriov_drivers_autoprobe

Date:		April 2017

Contact:	Bodong Wang<bodong@mellanox.com>

Description:

		This file is associated with the PF of a device that

		supports SR-IOV.  It determines whether newly-enabled VFs

		are immediately bound to a driver.  It initially contains

		1, which means the kernel automatically binds VFs to a

		compatible driver immediately after they are enabled.  If

		an application writes 0 to the file before enabling VFs,

		the kernel will not bind VFs to a driver.



		A typical use case is to write 0 to this file, then enable

		VFs, then assign the newly-created VFs to virtual machines.

		Note that changing this file does not affect already-

		enabled VFs.  In this scenario, the user must first disable

		the VFs, write 0 to sriov_drivers_autoprobe, then re-enable

		the VFs.



		This is similar to /sys/bus/pci/drivers_autoprobe, but

		affects only the VFs associated with a specific PF.

Documentation/PCI/pci-iov-howto.txt

+12 −0
Original line number Diff line number Diff line
@@ -68,6 +68,18 @@ To disable SR-IOV capability: 
	echo  0 > \

        /sys/bus/pci/devices/<DOMAIN:BUS:DEVICE.FUNCTION>/sriov_numvfs



To enable auto probing VFs by a compatible driver on the host, run

command below before enabling SR-IOV capabilities. This is the

default behavior.

	echo 1 > \

        /sys/bus/pci/devices/<DOMAIN:BUS:DEVICE.FUNCTION>/sriov_drivers_autoprobe



To disable auto probing VFs by a compatible driver on the host, run

command below before enabling SR-IOV capabilities. Updating this

entry will not affect VFs which are already probed.

	echo  0 > \

        /sys/bus/pci/devices/<DOMAIN:BUS:DEVICE.FUNCTION>/sriov_drivers_autoprobe



3.2 Usage example



Following piece of code illustrates the usage of the SR-IOV API.

drivers/infiniband/hw/hfi1/chip.c

+2 −2
Original line number Diff line number Diff line
@@ -13610,14 +13610,14 @@ static void init_chip(struct hfi1_devdata *dd) 
		dd_dev_info(dd, "Resetting CSRs with FLR\n");



		/* do the FLR, the DC reset will remain */

		hfi1_pcie_flr(dd);

		pcie_flr(dd->pcidev);



		/* restore command and BARs */

		restore_pci_variables(dd);



		if (is_ax(dd)) {

			dd_dev_info(dd, "Resetting CSRs with FLR\n");

			hfi1_pcie_flr(dd);

			pcie_flr(dd->pcidev);

			restore_pci_variables(dd);

		}

	} else {

drivers/infiniband/hw/hfi1/hfi.h

+0 −1
Original line number Diff line number Diff line
@@ -1764,7 +1764,6 @@ int hfi1_pcie_init(struct pci_dev *, const struct pci_device_id *); 
void hfi1_pcie_cleanup(struct pci_dev *);

int hfi1_pcie_ddinit(struct hfi1_devdata *, struct pci_dev *);

void hfi1_pcie_ddcleanup(struct hfi1_devdata *);

void hfi1_pcie_flr(struct hfi1_devdata *);

int pcie_speeds(struct hfi1_devdata *);

void request_msix(struct hfi1_devdata *, u32 *, struct hfi1_msix_entry *);

void hfi1_enable_intx(struct pci_dev *);

drivers/infiniband/hw/hfi1/pcie.c

+0 −30
Original line number Diff line number Diff line
@@ -240,36 +240,6 @@ void hfi1_pcie_ddcleanup(struct hfi1_devdata *dd) 
		iounmap(dd->piobase);

}



/*

 * Do a Function Level Reset (FLR) on the device.

 * Based on static function drivers/pci/pci.c:pcie_flr().

 */

void hfi1_pcie_flr(struct hfi1_devdata *dd)

{

	int i;

	u16 status;



	/* no need to check for the capability - we know the device has it */



	/* wait for Transaction Pending bit to clear, at most a few ms */

	for (i = 0; i < 4; i++) {

		if (i)

			msleep((1 << (i - 1)) * 100);



		pcie_capability_read_word(dd->pcidev, PCI_EXP_DEVSTA, &status);

		if (!(status & PCI_EXP_DEVSTA_TRPND))

			goto clear;

	}



	dd_dev_err(dd, "Transaction Pending bit is not clearing, proceeding with reset anyway\n");



clear:

	pcie_capability_set_word(dd->pcidev, PCI_EXP_DEVCTL,

				 PCI_EXP_DEVCTL_BCR_FLR);

	/* PCIe spec requires the function to be back within 100ms */

	msleep(100);

}



static void msix_setup(struct hfi1_devdata *dd, int pos, u32 *msixcnt,

		       struct hfi1_msix_entry *hfi1_msix_entry)

{