linux.git/drivers/pci, branch v4.4-rc7

Merge tag 'pci-v4.4-fixes-2' of git://git.kernel.org/pub/scm/linux/kernel/git/helgaas/pci

2015-12-09T17:26:06+00:00

Pull PCI fixes from Bjorn Helgaas:
 "These are more fixes I'd like to have in v4.4.  Several for the Altera
  driver added for v4.4, and one for an MSI domain problem that affects
  several arm64 platforms:

  MSI:
   - Only use the generic MSI layer when domain is hierarchical (Marc
     Zyngier)

  Altera host bridge driver:
   - Fix loop in tlp_read_packet() (Dan Carpenter)
   - Fix Requester ID for config accesses (Ley Foon Tan)
   - Check TLP completion status (Ley Foon Tan)
   - Fix error when INTx is 4 (Ley Foon Tan)"

* tag 'pci-v4.4-fixes-2' of git://git.kernel.org/pub/scm/linux/kernel/git/helgaas/pci:
  PCI: altera: Fix error when INTx is 4
  PCI: altera: Check TLP completion status
  PCI: altera: Fix Requester ID for config accesses
  PCI: altera: Fix loop in tlp_read_packet()
  PCI/MSI: Only use the generic MSI layer when domain is hierarchical

PCI: altera: Fix error when INTx is 4

2015-12-04T22:21:21+00:00

PCI interrupt lines start at 1, not at 0.  So, creates additional one
interrupt when register for irq domain.

Error when PCIe devices have 4 INTx:

  WARNING: CPU: 1 PID: 1 at kernel/irq/irqdomain.c:280
    irq_domain_associate+0x17c/0x1cc()
  error: hwirq 0x4 is too large for dummy

Tested on Ethernet adapter card with multi-functions.

Signed-off-by: Ley Foon Tan 
Signed-off-by: Bjorn Helgaas

PCI: altera: Check TLP completion status

2015-12-04T22:21:16+00:00

Check TLP packet successful completion status.  This fix the issue when
accessing multi-function devices in enumeration process, TLP will return
error when accessing non-exist function number.  Returns PCI error code
instead of generic errno.

Tested on Ethernet adapter card with multi-functions.

[bhelgaas: simplify completion status checking code]
Signed-off-by: Ley Foon Tan 
Signed-off-by: Bjorn Helgaas

PCI: altera: Fix Requester ID for config accesses

2015-12-04T22:21:12+00:00

The Requester ID should use the Root Port devfn and it should be always 0.
Previously we constructed the Requester ID using the *Completer* devfn,
i.e., the devfn of the Function we expect to respond to the config access.
This causes issues when accessing configuration space for devices other
than the Root Port.

Build the Requester ID using the Root Port devfn.

Tested on Ethernet adapter card with multi-functions.

Signed-off-by: Ley Foon Tan 
Signed-off-by: Bjorn Helgaas

PCI: altera: Fix loop in tlp_read_packet()

2015-12-04T22:21:08+00:00

TLP_LOOP is 500 and the "loop" variable was a u8 so "loop < TLP_LOOP" is
always true.  We only need this condition to work if there is a problem so
it would have been easy to miss this in testing.

Make it a normal for loop with "int i" instead of over thinking things and
making it complicated.

Fixes: 6bb4dd154ae8 ("PCI: altera: Add Altera PCIe host controller driver")
Signed-off-by: Dan Carpenter 
Signed-off-by: Bjorn Helgaas 
Acked-by: Ley Foon Tan

PCI/MSI: Only use the generic MSI layer when domain is hierarchical

2015-12-04T16:28:14+00:00

Since d8a1cb757550 ("PCI/MSI: Let pci_msi_get_domain use struct
device::msi_domain"), we use the MSI domain associated with the PCI device.

But finding an MSI domain doesn't mean that the domain is implemented using
the generic MSI domain API, and a number of MSI controllers are still using
arch_setup_msi_irq() and arch_teardown_msi_irqs().

Check that the domain we just obtained is hierarchical.  If it is, we can
use the new generic MSI stuff.  Otherwise we have to fall back to the old
arch_setup_msi_irq() and arch_teardown_msi_irqs() interfaces.

This avoids an oops in msi_domain_alloc_irqs() on systems with R-Car,
Tegra, Armada 370, and probably other DesignWare-based host controllers.

Fixes: d8a1cb757550 ("PCI/MSI: Let pci_msi_get_domain use struct device::msi_domain")
Reported-by: Phil Edworthy 
Tested-by: Phil Edworthy 
Signed-off-by: Marc Zyngier 
Signed-off-by: Bjorn Helgaas 
Acked-by: Thomas Gleixner 
CC: stable@vger.kernel.org	# v4.3+

PCI / PM: Tune down retryable runtime suspend error messages

2015-12-02T14:24:21+00:00

The runtime PM core doesn't treat EBUSY and EAGAIN retvals from the driver
suspend hooks as errors, but they still show up as errors in dmesg. Tune
them down. See rpm_suspend() for details of handling these return values.

Note that we use dev_dbg() for the retryable retvals, so after this
change you'll need either CONFIG_DYNAMIC_DEBUG or CONFIG_PCI_DEBUG
for them to show up in the log.

One problem caused by this was noticed by Daniel: the i915 driver
returns EAGAIN to signal a temporary failure to suspend and as a request
towards the RPM core for scheduling a suspend again. This is a normal
event, but the resulting error message flags a breakage during the
driver's automated testing which parses dmesg and picks up the error.

Reported-by: Daniel Vetter 
Link: https://bugs.freedesktop.org/show_bug.cgi?id=92992
Signed-off-by: Imre Deak 
Acked-by: Bjorn Helgaas 
Signed-off-by: Rafael J. Wysocki

ARM/PCI: Move align_resource function pointer to pci_host_bridge structure

2015-11-25T19:23:38+00:00

Commit b3a72384fe29 ("ARM/PCI: Replace pci_sys_data->align_resource with
global function pointer") introduced an ARM-specific align_resource()
function pointer.  This is not portable to other arches and doesn't work
for platforms with two different PCIe host bridge controllers.

Move the function pointer to the pci_host_bridge structure so each host
bridge driver can specify its own align_resource() function.

Signed-off-by: Gabriele Paoloni 
Signed-off-by: Bjorn Helgaas 
Reviewed-by: Arnd Bergmann

PCI: hisi: Fix deferred probing

2015-11-24T21:38:07+00:00

The hisi_pcie_probe() function is incorrectly marked as __init, as Kconfig
tells us:

  WARNING: drivers/pci/host/built-in.o(.data+0x7780): Section mismatch in reference from the variable hisi_pcie_driver to the function .init.text:hisi_pcie_probe()

If the probe for this device gets deferred past the point where __init
functions are removed, or the device is unbound and then reattached to the
driver, we branch into uninitialized memory, which is bad.

Remove the __init annotation from hisi_pcie_probe() and
hisi_add_pcie_port().

Fixes: 500a1d9a43e0 ("PCI: hisi: Add HiSilicon SoC Hip05 PCIe driver")
Signed-off-by: Arnd Bergmann 
Signed-off-by: Bjorn Helgaas 
Reviewed-by: Hanjun Guo 
Acked-by: Zhou Wang

PCI: designware: Remove incorrect io_base assignment

2015-11-24T20:06:41+00:00

"pp->io" is an I/O resource, e.g., "[io 0x0000-0xffff]"; "pp->io_base" is
the CPU physical address of a region where the host bridge converts CPU
memory accesses into PCI I/O transactions.

Corrupting pp->io_base by assigning pp->io->start to it breaks access to
the PCI I/O space, as reported by Kishon.

Remove the invalid assignment.

[bhelgaas: changelog]
Fixes: 0021d22b73d6 ("PCI: designware: Use of_pci_get_host_bridge_resources() to parse DT")
Reported-and-tested-by: Kishon Vijay Abraham I 
Signed-off-by: Stanimir Varbanov 
Signed-off-by: Bjorn Helgaas 
Reviewed-by: Arnd Bergmann