Move cra_writel(), cra_readl(), and altera_pcie_link_is_up() so a future
patch can use them in altera_pcie_retrain(). No functional change
intended.
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Set the affinity_mask in the PCI device before allocating vectors so that
the affinity can be propagated through the MSI descriptor structures to the
core IRQ code. To facilitate this, new __pci_enable_msi_range() and
__pci_enable_msix_range() helpers are factored out of their not prefixed
variants which assigning the new IRQ affinity mask in the PCI device so
that the low-level interrupt code can perform the interrupt affinity
assignment and do node-local allocations.
A new PCI_IRQ_NOAFFINITY flag is added to pci_alloc_irq_vectors() so that
this function can also be used by drivers that don't wish to use the
automatic affinity assignment.
[bhelgaas: omit "else" after "return" consistently]
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Alexander Gordeev <agordeev@redhat.com>
When assign new PCI platform PM operations check for all mandatory fields to
prevent NULL pointer dereference.
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Add a function to allocate and free a range of interrupt vectors, using
MSI-X, MSI or legacy vectors (in that order) based on the capabilities of
the underlying device and PCIe complex.
Additionally a new helper is provided to get the Linux IRQ number for given
device-relative vector so that the drivers don't need to allocate their own
arrays to keep track of the vectors for the multi vector MSI-X case.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Alexander Gordeev <agordeev@redhat.com>
The "entries" argument isn't needed if the list of entries does not contain
any holes. Make it optional so that we can avoid the need to allocate a
msix_entry structure for this (common) case.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Alexander Gordeev <agordeev@redhat.com>
Instead of relying on the msix_entry structure for the vector number, read
it from the msi_desc.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Alexander Gordeev <agordeev@redhat.com>
Add a pci_msix_desc_addr() helper to factor out the calculation of the base
address for a given MSI-X vector.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Alexander Gordeev <agordeev@redhat.com>
The underlying slot hotplug registration code assumed multiple slots, but
the actual implementation is broken for multiple slots.
This went unnoticed for years do to the fact that PowerVM seems to only
ever provide a single hotplug slot per PHB.
Under qemu/kvm the hotplug slot model aligns more with x86 where
multiple slots are presented under a single PHB. As seen in the
following each additional slot after the first fails to register due to
each slot always being compared against the first child node of the PHB
in the device tree.
rpaphp: RPA HOT Plug PCI Controller Driver version: 0.1
rpaphp: Slot [Slot 0] registered
rpaphp: pci_hp_register failed with error -16
rpaphp: pci_hp_register failed with error -16
rpaphp: pci_hp_register failed with error -16
rpaphp: pci_hp_register failed with error -16
The registration logic is fixed so that each slot is compared
against the existing child devices of the PHB in the device tree to
determine present slots vs empty slots.
rpaphp: RPA HOT Plug PCI Controller Driver version: 0.1
rpaphp: Slot [C0] registered
rpaphp: Slot [C1] registered
rpaphp: Slot [C2] registered
rpaphp: Slot [C3] registered
rpaphp: Slot [C4] registered
Signed-off-by: Tyrel Datwyler <tyreld@linux.vnet.ibm.com>
Reviewed-by: Nathan Fontenot <nfont@linux.vnet.ibm.com>
[mpe: Massage changelog]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
At least on arm, <asm/dma.h> does not get included when building
drivers/pci/pci.o. This causes the following build warning which can be
fixed by including <asm/dma.h>:
drivers/pci/pci.c:37:5: warning: symbol 'isa_dma_bridge_buggy' was not declared. Should it be static?
Signed-off-by: Ben Dooks <ben.dooks@codethink.co.uk>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
The cxl driver will use infrastructure from pnv_php to handle device tree
updates when switching bi-modal CAPI cards into CAPI mode.
To enable this, export pnv_php_find_slot() and
pnv_php_set_slot_power_state(), and add corresponding declarations, as well
as the definition of struct pnv_php_slot, to asm/pnv-pci.h.
Cc: Gavin Shan <gwshan@linux.vnet.ibm.com>
Cc: linux-pci@vger.kernel.org
Cc: Bjorn Helgaas <bhelgaas@google.com>
Signed-off-by: Andrew Donnellan <andrew.donnellan@au1.ibm.com>
Signed-off-by: Ian Munsie <imunsie@au1.ibm.com>
Acked-by: Gavin Shan <gwshan@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
The switch is the only statement in the resource_list_for_each_entry()
loop, so remove unnecessary "continue" statements in the switch. Simplify
checking for the required non-prefetchable memory aperture.
No functional change intended.
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Use devm_request_pci_bus_resources() to request host bridge window
resources instead of doing it by hand in the driver.
No functional change intended.
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Use devm_request_pci_bus_resources() to request host bridge window
resources instead of doing it by hand in the driver.
No functional change intended.
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
41534e5378 ("PCI: tegra: Implement a proper resource hierarchy") did two
things:
1) It added a top-level resource that encloses all resources declared in
the DT description, including registers and bridge apertures, and
2) It requested the bridge apertures, which means the PCI core can track
the resources used by PCI devices below the bridge.
The latter is necessary, but the former is questionable because there's no
guarantee that the bridge registers and the apertures are contiguous. In
this example:
# cat /proc/iomem
00000000-3fffffff : /pcie-controller@00003000
00000000-00000fff : /pcie-controller@00003000/pci@1,0
00003000-000037ff : pads
00003800-000039ff : afi
10000000-1fffffff : cs
the resource tree claims that [mem 0x00003a00-0x0fffffff] is consumed by
/pcie-controller@00003000, but it's not mentioned in the DT, and it might
actually be used by other devices.
Remove the top-level resource so we don't claim more than the device
actually consumes.
This reintroduces the problem that we can't match the resources, e.g.,
"pads", "afi", "cs", etc., to the DT device. I think this should be solved
by having the DT core request all resources of all devices in the DT (it
does not do that today). If a driver claims the device, it can request the
resources it uses. For example:
# cat /proc/iomem
00000000-00000fff : /pcie-controller@00003000
00000000-00000fff : /pcie-controller@00003000/pci@1,0
00003000-000037ff : /pcie-controller@00003000
00003000-000037ff : pads
00003800-000039ff : /pcie-controller@00003000
00003800-000039ff : afi
10000000-1fffffff : /pcie-controller@00003000
10000000-1fffffff : cs
...
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
The switch is the only statement in the resource_list_for_each_entry()
loop, so remove unnecessary cases and "continue" statements in the switch.
Inline rcar_pcie_release_of_pci_ranges(), which is only called once.
No functional change intended.
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Use devm_request_pci_bus_resources() to request host bridge window
resources instead of doing it by hand in the driver.
No functional change intended.
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Request host bridge window resources so they appear in ioport_resource and
iomem_resource and are reflected in /proc/ioports and /proc/iomem.
Tested-by: Geert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Previously we added a dummy I/O port region even though the R-Car
controller doesn't support PCI port I/O. This resulted in bogus root bus
resources like this:
pci_bus 0000:00: root bus resource [io 0xee080000-0xee0810ff]
pci_bus 0000:00: root bus resource [mem 0xee080000-0xee0810ff]
Drop the unused dummy I/O port region and set struct hw_pci.io_optional so
the ARM PCI code doesn't add a default one for us.
Tested-by: Geert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
We claim PCI BAR and bridge window resources in pci_bus_assign_resources(),
but when PCI_PROBE_ONLY is set, we treat those resources as immutable and
don't call pci_bus_assign_resources(), so the resources aren't put in the
resource tree.
When the resources aren't in the tree, they don't show up in /proc/iomem,
we can't detect conflicts, and we need special cases elsewhere for
PCI_PROBE_ONLY or resources without a parent pointer.
Claim all PCI BAR and window resources in the PCI_PROBE_ONLY case.
If a PCI_PROBE_ONLY platform assigns conflicting resources, Linux can't fix
the conflicts. Previously we didn't notice the conflicts, but now we will,
which may expose new failures.
[bhelgaas: changelog, summarize comment]
Signed-off-by: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by: Will Deacon <will.deacon@arm.com>
CC: Arnd Bergmann <arnd@arndb.de>
CC: David Daney <david.daney@cavium.com>
All PCI resources (bridge windows and BARs) should be inserted in the
iomem_resource and ioport_resource trees so we know what space is occupied
and what is available for other devices. There's nothing arch-specific
about this, but it is currently done by arch-specific code.
Add a generic pci_bus_claim_resources() interface so we can migrate away
from the arch-specific code.
[bhelgaas: changelog]
Signed-off-by: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
CC: Arnd Bergmann <arnd@arndb.de>
CC: Yinghai Lu <yinghai@kernel.org>
Some uio-based PCI drivers, e.g., uio_cif do not work if the assigned PCI
memory resources are not page aligned.
By using the kernel option "pci=resource_alignment" it is possible to force
single PCI boards to use page alignment for their memory resources.
However, this is fairly cumbersome if several of these boards are in use
as the specification of the cards has to be done via PCI bus/slot/function
number which might change, e.g., by adding another board.
Extend the kernel option "pci=resource_alignment" to allow specification of
relevant devices via PCI device/vendor (and subdevice/subvendor) IDs. The
specification of the devices via device/vendor is indicated by a leading
string "pci:" as argument to "pci=resource_alignment". The format of the
specification is pci:<vendor>:<device>[:<subvendor>:<subdevice>]
Signed-off-by: Mathias Koehrer <mathias.koehrer@etas.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Use the device resource management (devm) interfaces so we don't need to
explicitly release resources on failure paths or when the driver is
removed.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by: Keith Busch <keith.busch@intel.com>
Use devm_request_pci_bus_resources() to request host bridge window
resources instead of doing it by hand in the driver.
No functional change intended.
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
The switch is the only statement in the resource_list_for_each_entry()
loop, so remove unnecessary "continue" statements in the switch. Remove
unnecessary "goto" statements and label. Simplify checking for the
required non-prefetchable memory aperture.
No functional change intended.
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Use devm_request_pci_bus_resources() to request host bridge window
resources instead of doing it by hand in the driver.
No functional change intended.
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
This adds standalone driver to support PCI hotplug for PowerPC PowerNV
platform that runs on top of skiboot firmware. The firmware identifies
hotpluggable slots and marked their device tree node with proper
"ibm,slot-pluggable" and "ibm,reset-by-firmware". The driver scans
device tree nodes to create/register PCI hotplug slot accordingly.
The PCI slots are organized in fashion of tree, which means one
PCI slot might have parent PCI slot and parent PCI slot possibly
contains multiple child PCI slots. At the plugging time, the parent
PCI slot is populated before its children. The child PCI slots are
removed before their parent PCI slot can be removed from the system.
If the skiboot firmware doesn't support slot status retrieval, the PCI
slot device node shouldn't have property "ibm,reset-by-firmware". In
that case, none of valid PCI slots will be detected from device tree.
The skiboot firmware doesn't export the capability to access attention
LEDs yet and it's something for TBD.
Signed-off-by: Gavin Shan <gwshan@linux.vnet.ibm.com>
Acked-by: Bjorn Helgaas <bhelgaas@google.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Currently, PowerPC PowerNV platform utilizes ppc_md.pcibios_fixup(),
which is called for once after PCI probing and resource assignment
are completed, to allocate platform required resources for PCI devices:
PE#, IO and MMIO mapping, DMA address translation (TCE) table etc.
Obviously, it's not hotplug friendly.
This adds weak function pcibios_setup_bridge(), which is called by
pci_setup_bridge(). PowerPC PowerNV platform will reuse the function
to assign above platform required resources to newly plugged PCI devices
during PCI hotplug in subsequent patches.
Signed-off-by: Gavin Shan <gwshan@linux.vnet.ibm.com>
Acked-by: Bjorn Helgaas <bhelgaas@google.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
The switch is the only statement in the resource_list_for_each_entry()
loop, so remove unnecessary "continue" statements in the switch. Simplify
checking for the required non-prefetchable memory aperture. Inline
altera_pcie_release_of_pci_ranges(), which is only called once.
No functional change intended.
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Use devm_request_pci_bus_resources() to request host bridge window
resources instead of doing it by hand in the driver.
No functional change intended.
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Request host bridge window resources so they appear in ioport_resource and
iomem_resource and are reflected in /proc/ioports and /proc/iomem.
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
of_pci_get_host_bridge_resources() allocates a list of resources for host
bridge windows. If we fail after allocating that list, free it before we
return error.
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Request host bridge window resources so they appear in ioport_resource and
iomem_resource and are reflected in /proc/ioports and /proc/iomem.
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
of_pci_get_host_bridge_resources() allocates a list of resources for host
bridge windows. If we fail after allocating that list, free it before we
return error.
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Request host bridge window resources so they appear in ioport_resource and
iomem_resource and are reflected in /proc/ioports and /proc/iomem.
For example, the following entries did not previously appear in /proc/iomem:
e180000000-e1ffffffff : /soc/pcie@1f2b0000
e180000000-e182ffffff : PCI Bus 0000:01
e180000000-e181ffffff : 0000:01:00.0
e182000000-e1820fffff : 0000:01:00.0
e182100000-e1821fffff : 0000:01:00.0
f000000000-ffffffffff : /soc/pcie@1f2b0000
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
of_pci_get_host_bridge_resources() allocates a list of resources for host
bridge windows. If we fail after allocating that list, free it before we
return error.
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Request host bridge window resources so they appear in ioport_resource and
iomem_resource and are reflected in /proc/ioports and /proc/iomem.
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
The switch is the only statement in the resource_list_for_each_entry()
loop, so remove unnecessary "continue" statements in the switch.
No functional change intended.
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Request host bridge window resources so they appear in ioport_resource and
iomem_resource and are reflected in /proc/ioports and /proc/iomem.
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
of_pci_get_host_bridge_resources() allocates a list of resources for host
bridge windows. If we fail after allocating that list, free it before we
return error.
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
If a hotplug port is suspended to D3cold, its slot status register cannot
be read. If that hotplug port happens to share its IRQ with other devices,
whenever an interrupt occurs for one of these devices, pciehp logs a
"no response from device" message and tries to read the PCI_EXP_SLTSTA
register, even though we know that will fail.
Ignore interrupts while we're in D3cold.
[bhelgaas: changelog]
Signed-off-by: Lukas Wunner <lukas@wunner.de>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
The save/restore buffers for VC state is first composed of a 2-byte control
register, then a bunch of 4-byte words.
This causes unaligned accesses which trap on platform such as sparc.
This is easy to fix by simply moving the buffer pointer forward by 4 bytes
instead of 2 after dealing with the control register. The length
adjustment needs to be changed likewise as well.
Fixes: 5f8fc43217 ("PCI: Include pci/pcie/Kconfig directly from pci/Kconfig")
Reported-by: Meelis Roos <mroos@linux.ee>
Reported-by: Anatoly Pugachev <matorola@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
CC: stable@vger.kernel.org # v4.6+
PCI exposes files like /proc/bus/pci/00/00.0 in procfs. These files
support operations like this:
ioctl(fd, PCIIOC_MMAP_IS_IO); # request I/O port space
ioctl(fd, PCIIOC_WRITE_COMBINE, 1); # request write-combining
mmap(fd, ...)
Write combining is useful on PCI memory space, but I don't think it makes
sense on PCI I/O port space.
We *could* change proc_bus_pci_ioctl() to make it impossible to set
mmap_state == pci_mmap_io and write_combine at the same time, but that
would break the following sequence, which is currently legal:
mmap(fd, ...) # default is I/O, non-combining
ioctl(fd, PCIIOC_WRITE_COMBINE, 1); # request write-combining
ioctl(fd, PCIIOC_MMAP_IS_MEM); # request memory space
mmap(fd, ...) # get write-combining mapping
Ignore the write-combining flag when mapping I/O port space.
This patch should have no functional effect, based on this analysis of all
implementations of pci_mmap_page_range():
- ia64 mips parisc sh unicore32 x86 do not support mapping of I/O port
space at all.
- arm cris microblaze mn10300 sparc xtensa support mapping of I/O port
space, but ignore the write_combine argument to pci_mmap_page_range().
- powerpc supports mapping of I/O port space and uses write_combine, and
it disables write combining for I/O port space in
__pci_mmap_set_pgprot().
This patch makes it possible to remove __pci_mmap_set_pgprot() from
powerpc, which simplifies that path.
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
When we have an interrupt from the host we have a bit set in event page
indicating there are messages for the particular channel. We need to read
them all as we won't get signaled for what was on the queue before we
cleared the bit in vmbus_on_event(). This applies to all Hyper-V drivers
and the pass-through driver should do the same.
I did not meet any bugs; the issue was found by code inspection. We don't
have many events going through hv_pci_onchannelcallback(), which explains
why nobody reported the issue before.
While on it, fix handling non-zero vmbus_recvpacket_raw() return values by
dropping out. If the return value is not zero, it is wrong to inspect
buffer or bytes_recvd as these may contain invalid data.
Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by: Jake Oshins <jakeo@microsoft.com>
We don't free buffer on several code paths in hv_pci_onchannelcallback(),
put kfree() to the end of the function to fix the issue. Direct { kfree();
return; } can now be replaced with a simple 'break';
Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by: Jake Oshins <jakeo@microsoft.com>
The PCI_MSI symbol is used inconsistently throughout the tree, with some
drivers using 'select' and others using 'depends on', or using conditional
selects. This keeps causing problems; the latest one is a result of
ARCH_ALPINE using a 'select' statement to enable its platform-specific MSI
driver without enabling MSI:
warning: (ARCH_ALPINE) selects ALPINE_MSI which has unmet direct dependencies (PCI && PCI_MSI)
drivers/irqchip/irq-alpine-msi.c:104:15: error: variable 'alpine_msix_domain_info' has initializer but incomplete type
static struct msi_domain_info alpine_msix_domain_info = {
^~~~~~~~~~~~~~~
drivers/irqchip/irq-alpine-msi.c:105:2: error: unknown field 'flags' specified in initializer
.flags = MSI_FLAG_USE_DEF_DOM_OPS | MSI_FLAG_USE_DEF_CHIP_OPS |
^
drivers/irqchip/irq-alpine-msi.c:105:11: error: 'MSI_FLAG_USE_DEF_DOM_OPS' undeclared here (not in a function)
.flags = MSI_FLAG_USE_DEF_DOM_OPS | MSI_FLAG_USE_DEF_CHIP_OPS |
^~~~~~~~~~~~~~~~~~~~~~~~
There is little reason to enable PCI support for a platform that uses MSI
but then leave MSI disabled at compile time.
Select PCI_MSI from irqchips that implement MSI, and make PCI host bridges
that use MSI on ARM depend on PCI_MSI_IRQ_DOMAIN.
For all three architectures that support PCI_MSI_IRQ_DOMAIN (ARM, ARM64,
X86), enable it by default whenever MSI is enabled.
[bhelgaas: changelog, omit crypto config change]
Suggested-by: Marc Zyngier <marc.zyngier@arm.com>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by: Marc Zyngier <marc.zyngier@arm.com>
