Pull x86 core updates from Ingo Molnar:
"There were so many changes in the x86/asm, x86/apic and x86/mm topics
in this cycle that the topical separation of -tip broke down somewhat -
so the result is a more traditional architecture pull request,
collected into the 'x86/core' topic.
The topics were still maintained separately as far as possible, so
bisectability and conceptual separation should still be pretty good -
but there were a handful of merge points to avoid excessive
dependencies (and conflicts) that would have been poorly tested in the
end.
The next cycle will hopefully be much more quiet (or at least will
have fewer dependencies).
The main changes in this cycle were:
* x86/apic changes, with related IRQ core changes: (Jiang Liu, Thomas
Gleixner)
- This is the second and most intrusive part of changes to the x86
interrupt handling - full conversion to hierarchical interrupt
domains:
[IOAPIC domain] -----
|
[MSI domain] --------[Remapping domain] ----- [ Vector domain ]
| (optional) |
[HPET MSI domain] ----- |
|
[DMAR domain] -----------------------------
|
[Legacy domain] -----------------------------
This now reflects the actual hardware and allowed us to distangle
the domain specific code from the underlying parent domain, which
can be optional in the case of interrupt remapping. It's a clear
separation of functionality and removes quite some duct tape
constructs which plugged the remap code between ioapic/msi/hpet
and the vector management.
- Intel IOMMU IRQ remapping enhancements, to allow direct interrupt
injection into guests (Feng Wu)
* x86/asm changes:
- Tons of cleanups and small speedups, micro-optimizations. This
is in preparation to move a good chunk of the low level entry
code from assembly to C code (Denys Vlasenko, Andy Lutomirski,
Brian Gerst)
- Moved all system entry related code to a new home under
arch/x86/entry/ (Ingo Molnar)
- Removal of the fragile and ugly CFI dwarf debuginfo annotations.
Conversion to C will reintroduce many of them - but meanwhile
they are only getting in the way, and the upstream kernel does
not rely on them (Ingo Molnar)
- NOP handling refinements. (Borislav Petkov)
* x86/mm changes:
- Big PAT and MTRR rework: making the code more robust and
preparing to phase out exposing direct MTRR interfaces to drivers -
in favor of using PAT driven interfaces (Toshi Kani, Luis R
Rodriguez, Borislav Petkov)
- New ioremap_wt()/set_memory_wt() interfaces to support
Write-Through cached memory mappings. This is especially
important for good performance on NVDIMM hardware (Toshi Kani)
* x86/ras changes:
- Add support for deferred errors on AMD (Aravind Gopalakrishnan)
This is an important RAS feature which adds hardware support for
poisoned data. That means roughly that the hardware marks data
which it has detected as corrupted but wasn't able to correct, as
poisoned data and raises an APIC interrupt to signal that in the
form of a deferred error. It is the OS's responsibility then to
take proper recovery action and thus prolonge system lifetime as
far as possible.
- Add support for Intel "Local MCE"s: upcoming CPUs will support
CPU-local MCE interrupts, as opposed to the traditional system-
wide broadcasted MCE interrupts (Ashok Raj)
- Misc cleanups (Borislav Petkov)
* x86/platform changes:
- Intel Atom SoC updates
... and lots of other cleanups, fixlets and other changes - see the
shortlog and the Git log for details"
* 'x86-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (222 commits)
x86/hpet: Use proper hpet device number for MSI allocation
x86/hpet: Check for irq==0 when allocating hpet MSI interrupts
x86/mm/pat, drivers/infiniband/ipath: Use arch_phys_wc_add() and require PAT disabled
x86/mm/pat, drivers/media/ivtv: Use arch_phys_wc_add() and require PAT disabled
x86/platform/intel/baytrail: Add comments about why we disabled HPET on Baytrail
genirq: Prevent crash in irq_move_irq()
genirq: Enhance irq_data_to_desc() to support hierarchy irqdomain
iommu, x86: Properly handle posted interrupts for IOMMU hotplug
iommu, x86: Provide irq_remapping_cap() interface
iommu, x86: Setup Posted-Interrupts capability for Intel iommu
iommu, x86: Add cap_pi_support() to detect VT-d PI capability
iommu, x86: Avoid migrating VT-d posted interrupts
iommu, x86: Save the mode (posted or remapped) of an IRTE
iommu, x86: Implement irq_set_vcpu_affinity for intel_ir_chip
iommu: dmar: Provide helper to copy shared irte fields
iommu: dmar: Extend struct irte for VT-d Posted-Interrupts
iommu: Add new member capability to struct irq_remap_ops
x86/asm/entry/64: Disentangle error_entry/exit gsbase/ebx/usermode code
x86/asm/entry/32: Shorten __audit_syscall_entry() args preparation
x86/asm/entry/32: Explain reloading of registers after __audit_syscall_entry()
...
* pci/host-xgene:
PCI: xgene: Allow config access to Root Port even when link is down
PCI: xgene: Disable Configuration Request Retry Status for v1 silicon
* pci/hotplug:
PCI: pciehp: Inline the "handle event" functions into the ISR
PCI: pciehp: Rename queue_interrupt_event() to pciehp_queue_interrupt_event()
PCI: pciehp: Make queue_interrupt_event() void
PCI: pciehp: Clean up debug logging
The pciehp_handle_*() functions (pciehp_handle_attention_button(), etc.)
only contain a line or two of useful code, so it's clumsy to put
them in separate functions. All they so is add an event to a work queue,
and it's clearer to see that directly in the ISR.
Inline them directly into pcie_isr(). No functional change.
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Rajat Jain <rajatja@google.com>
Acked-by: Yinghai Lu <yinghai@kernel.org>
Rename queue_interrupt_event() to pciehp_queue_interrupt_event() so we can
make it extern and call it from pcie_isr().
No functional change.
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Rajat Jain <rajatja@google.com>
Acked-by: Yinghai Lu <yinghai@kernel.org>
Nobody looks at the return value from queue_interrupt_event(), so errors
were silently ignored. Convert it to a "void" function and note the error
in the dmesg log.
No functional change except the new message.
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Rajat Jain <rajatja@google.com>
Acked-by: Yinghai Lu <yinghai@kernel.org>
Previously, when a Root Port's link was down, we didn't allow config access
to the Root Port, which meant that if the Root Port led to an empty slot,
"lspci" didn't even show the Root Port.
Allow config access to Root Port even when link is down.
[bhelgaas: changelog, fold in unused var fix]
Suggested-by: Bjorn Helgaas <bhelgaas@google.com>
Signed-off-by: Duc Dang <dhdang@apm.com>
Signed-off-by: Tanmay Inamdar <tinamdar@apm.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
When a CPU reads the Vendor and Device ID of a non-existent device, the
controller should fabricate return data of 0xFFFFFFFF. Configuration
Request Retry Status (CRS) is not applicable in this case because the
device doesn't exist at all.
The X-Gene v1 PCIe controller has a bug in the CRS logic such that when CRS
is enabled, it fabricates return data of 0xFFFF0001 for this case, which
means "the device exists but is not ready." That causes the PCI core to
retry the read until it times out after 60 seconds.
Disable CRS capability advertisement by clearing the CRS Software
Visibility bit in the Root Capabilities Register.
[bhelgaas: changelog and comment]
Tested-by: Ian Campbell <ian.campbell@citrix.com>
Tested-by: Marcin Juszkiewicz <mjuszkiewicz@redhat.com>
Signed-off-by: Duc Dang <dhdang@apm.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by: Tanmay Inamdar <tinamdar@apm.com>
The pciehp debug logging is overly verbose and often redundant. Almost all
of the information printed by dbg_ctrl() is also printed by the normal PCI
core enumeration code and by pcie_init().
Remove the redundant debug info.
When claiming a pciehp bridge, we print the slot characteristics, e.g.,
Slot #6 AttnBtn- AttnInd- PwrInd- PwrCtrl- MRL- Interlock- NoCompl+ LLActRep+
Add the Hot-Plug Capable and Hot-Plug Surprise bits to this information,
and print it all in the same order as lspci does.
No functional change except the message text changes.
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Rajat Jain <rajatja@google.com>
Acked-by: Yinghai Lu <yinghai@kernel.org>
Use "u32", not "uint32_t", for consistency. Use "tmp", not "temp", for
consistency within the driver.
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by: Richard Zhu <Richard.Zhu@freescale.com>
No one uses pci_scan_bus_parented() any more, remove it.
Signed-off-by: Yijing Wang <wangyijing@huawei.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Currently, the timeout is never detected as count has a value of -1 if a
timeout happens, but the code is checking for 0. Also, this patch removes
the unneeded final wait if a timeout occurs.
[bhelgaas: reworked starting from http://lkml.kernel.org/r/1433543864-7252-1-git-send-email-troy.kisky@boundarydevices.com]
Signed-off-by: Troy Kisky <troy.kisky@boundarydevices.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Update the Link Control Enable Clock Power Management bit the same
way we update the ASPM Control bits, with a single call of
pcie_capability_clear_and_set_word().
No functional change; this just makes both paths use the same style.
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
All the DesignWare-based host drivers loop waiting for the link to come up,
but they do it several ways that are needlessly different.
Wait for the link to come up in a consistent style across all the
DesignWare drivers. No functional change.
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by: Pratyush Anand <pratyush.anand@gmail.com>
All other DesignWare-based drivers have a *_establish_link() function.
This functionality is trivial for Layerscape, but factor out a
ls_pcie_establish_link() for consistency with the other drivers. No
functional change.
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by: Pratyush Anand <pratyush.anand@gmail.com>
All the other DesignWare-based drivers use dw_pcie_link_up(), so use it in
this driver, too, for consistency. No functional change.
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by: Pratyush Anand <pratyush.anand@gmail.com>
We already use dw_pcie_link_up() once in dra7xx_pcie_establish_link(), but
we duplicate its code later. Use dw_pcie_link_up() for consistency. No
functional change.
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by: Kishon Vijay Abraham I <kishon@ti.com>
Acked-by: Pratyush Anand <pratyush.anand@gmail.com>
The commit referenced below deferred waiting for command completion until
the start of the next command, allowing hardware to do the latching
asynchronously. Unfortunately, being ready to accept a new command is the
only indication we have that the previous command is completed. In cases
where we need that state change to be enabled, we must still wait for
completion. For instance, pciehp_reset_slot() attempts to disable anything
that might generate a surprise hotplug on slots that support presence
detection. If we don't wait for those settings to latch before the
secondary bus reset, we negate any value in attempting to prevent the
spurious hotplug.
Create a base function with optional wait and helper functions so that
pcie_write_cmd() turns back into the "safe" interface which waits before
and after issuing a command and add pcie_write_cmd_nowait(), which
eliminates the trailing wait for asynchronous completion. The following
functions are returned to their previous behavior:
pciehp_power_on_slot
pciehp_power_off_slot
pcie_disable_notification
pciehp_reset_slot
The rationale is that pciehp_power_on_slot() enables the link and therefore
relies on completion of power-on. pciehp_power_off_slot() and
pcie_disable_notification() need a wait because data structures may be
freed after these calls and continued signaling from the device would be
unexpected. And, of course, pciehp_reset_slot() needs to wait for the
scenario outlined above.
Fixes: 3461a06866 ("PCI: pciehp: Wait for hotplug command completion lazily")
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
CC: stable@vger.kernel.org # v3.17+
APM X-Gene v1 SoC supports its own implementation of MSI, which is not
compliant to GIC V2M specification for MSI Termination.
There is a single MSI block in X-Gene v1 SOC which serves all 5 PCIe ports.
This MSI block supports 2048 MSI termination ports coalesced into 16
physical HW IRQ lines and shared across all 5 PCIe ports.
As there are only 16 HW IRQs to serve 2048 MSI vectors, to support
set_affinity correctly for each MSI vectors, the 16 HW IRQs are statically
allocated to 8 X-Gene v1 cores (2 HW IRQs for each cores). To steer MSI
interrupt to target CPU, MSI vector is moved around these HW IRQs lines.
With this approach, the total MSI vectors this driver supports is reduced
to 256.
[bhelgaas: squash doc, driver, maintainer update]
Signed-off-by: Duc Dang <dhdang@apm.com>
Signed-off-by: Tanmay Inamdar <tinamdar@apm.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Marc Zyngier <marc.zyngier@arm.com>
Rename imx6_pcie_start_link() to imx6_pcie_establish_link() to follow the
convention of other DesignWare-based host drivers. No functional change.
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by: Pratyush Anand <pratyush.anand@gmail.com>
In d74b9027a4 ("PCI: Consider additional PF's IOV BAR alignment in sizing
and assigning"), we store additional alignment in realloc_head and take
this into consideration for assignment.
In __assign_resources_sorted(), we changed dev_res->res->start, then used
resource_start() (which depends on res->start), so the recomputed res->end
was completely bogus. Even if we'd had the correct size, the end would
have been off by one.
Preserve the resource size when we adjust its alignment.
[bhelgaas: changelog]
Fixes: d74b9027a4 ("PCI: Consider additional PF's IOV BAR alignment in sizing and assigning")
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by: Wei Yang <weiyang@linux.vnet.ibm.com>
CC: Benjamin Herrenschmidt <benh@kernel.crashing.org>
David Ahern reported that d63e2e1f3d ("sparc/PCI: Clip bridge windows
to fit in upstream windows") fails to boot on sparc/T5-8:
pci 0000:06:00.0: reg 0x184: can't handle BAR above 4GB (bus address 0x110204000)
The problem is that sparc64 assumed that dma_addr_t only needed to hold DMA
addresses, i.e., bus addresses returned via the DMA API (dma_map_single(),
etc.), while the PCI core assumed dma_addr_t could hold *any* bus address,
including raw BAR values. On sparc64, all DMA addresses fit in 32 bits, so
dma_addr_t is a 32-bit type. However, BAR values can be 64 bits wide, so
they don't fit in a dma_addr_t. d63e2e1f3d added new checking that
tripped over this mismatch.
Add pci_bus_addr_t, which is wide enough to hold any PCI bus address,
including both raw BAR values and DMA addresses. This will be 64 bits
on 64-bit platforms and on platforms with a 64-bit dma_addr_t. Then
dma_addr_t only needs to be wide enough to hold addresses from the DMA API.
[bhelgaas: changelog, bugzilla, Kconfig to ensure pci_bus_addr_t is at
least as wide as dma_addr_t, documentation]
Fixes: d63e2e1f3d ("sparc/PCI: Clip bridge windows to fit in upstream windows")
Fixes: 23b13bc76f ("PCI: Fail safely if we can't handle BARs larger than 4GB")
Link: http://lkml.kernel.org/r/CAE9FiQU1gJY1LYrxs+ma5LCTEEe4xmtjRG0aXJ9K_Tsu+m9Wuw@mail.gmail.com
Link: http://lkml.kernel.org/r/1427857069-6789-1-git-send-email-yinghai@kernel.org
Link: https://bugzilla.kernel.org/show_bug.cgi?id=96231
Reported-by: David Ahern <david.ahern@oracle.com>
Tested-by: David Ahern <david.ahern@oracle.com>
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by: David S. Miller <davem@davemloft.net>
CC: stable@vger.kernel.org # v3.19+
pci_ari_enabled() is useful outside of drivers/pci, particularly for
deriving INTx routing via ACPI _PRT, so move it to the global header.
Also convert to bool return.
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Don Dutile <ddutile@redhat.com>
Acked-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Previously we assumed that PCIe Root Ports and Downstream Ports had Links
on their secondary side. That is true in most systems, but it is possible
to connect a switch with either an Upstream or a Downstream Port leading
downstream.
Instead of relying on the component type to identify devices that have
links leading downstream, use the "dev->has_secondary_link" field.
[bhelgaas: changelog]
Signed-off-by: Yijing Wang <wangyijing@huawei.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
The resource list is only used in the setup process and was never freed.
pci_add_resource() allocates a memory area to store the list item.
Fix the memory leak.
Tested-by: Ray Jui <rjui@broadcom.com>
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Ray Jui <rjui@broadcom.com>
The struct iproc_pcie.resources member was pointing to a stack variable and
is invalid after the registration function returned.
Remove this pointer and add a parameter to the function.
Tested-by: Ray Jui <rjui@broadcom.com>
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Ray Jui <rjui@broadcom.com>
In d74b9027a4 ("PCI: Consider additional PF's IOV BAR alignment in sizing
and assigning"), it stores additional alignment in realloc_head and takes
this into consideration for assignment.
After getting the additional alignment, it reorders the head list so
resources with bigger alignment are ahead of resources with smaller
alignment. It does this by iterating over the head list and inserting
ahead of any resource with smaller alignment. This should be done for the
first occurrence, but the code currently iterates over the whole list.
Fix this by terminating the loop when we find the first smaller resource in
the head list.
[bhelgaas: changelog]
Fixes: d74b9027a4 ("PCI: Consider additional PF's IOV BAR alignment in sizing and assigning")
Signed-off-by: Wei Yang <weiyang@linux.vnet.ibm.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
After b97ea289cf ("PCI: Assign resources before drivers claim devices
(pci_scan_root_bus())"), pci_scan_root_bus() no longer adds the devices, so
it is equivalent to:
pci_create_root_bus()
pci_scan_child_bus()
Use pci_scan_root_bus() to simplify the code.
[bhelgaas: changelog]
Signed-off-by: Yijing Wang <wangyijing@huawei.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Lucas Stach <l.stach@pengutronix.de>
Acked-by: Jingoo Han <jingoohan1@gmail.com>
CC: Mohit Kumar <mohit.kumar@st.com>
After b97ea289cf ("PCI: Assign resources before drivers claim devices
(pci_scan_root_bus())"), pci_scan_root_bus() no longer adds the devices, so
it is equivalent to tegra_pcie_scan_bus().
Remove tegra_pcie_scan_bus() (the hw.scan method), so we use the generic
pci_scan_root_bus() path.
[bhelgaas: changelog]
Tested-by: Thierry Reding <treding@nvidia.com>
Signed-off-by: Yijing Wang <wangyijing@huawei.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by: Thierry Reding <treding@nvidia.com>
After b97ea289cf ("PCI: Assign resources before drivers claim devices
(pci_scan_root_bus())"), pci_scan_root_bus() no longer adds the devices, so
it is equivalent to mvebu_pcie_scan_bus().
Remove mvebu_pcie_scan_bus() (the hw.scan method), so we use the generic
pci_scan_root_bus() path. We also need to use pci_common_init_dev()
instead of pci_common_init() so we can supply the host bridge device
pointer.
[bhelgaas: changelog]
Tested-by: Gregory CLEMENT <gregory.clement@free-electrons.com>
Signed-off-by: Yijing Wang <wangyijing@huawei.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Gregory CLEMENT <gregory.clement@free-electrons.com>
CC: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
CC: Jason Cooper <jason@lakedaemon.net>
We allocate pcie_link_state for the component at the upstream end of a
Link. Previously we did this by allocating pcie_link_state for Root Ports
and Downstream Ports. This works fine for the typical topology:
00:1c.0 Root Port [bridge to bus 02]
02:00.0 Upstream Port [bridge to bus 03]
03:00.0 Downstream Port [bridge to bus 04]
04:00.0 Endpoint or Switch Port
However, it is possible to have a Root Port connected to a Downstream Port
instead of an Upstream Port, as in Robert White's ATCA system:
00:1c.0 Root Port [bridge to bus 02]
02:00.0 Downstream Port [bridge to bus 03]
03:01.0 Downstream Port [bridge to bus 04]
04:00.0 Endpoint or Switch Port
In this topology, we wrongly allocated pcie_link_state for the 02:00.0
Downstream Port, which is actually the *downstream* end of a link. This
led to the following NULL pointer dereference when we tried to connect this
link into the tree of links starting at the 00:1c.0 Root Port:
BUG: unable to handle kernel NULL pointer dereference at 0000000000000088
IP: [<ffffffff81550324>] pcie_aspm_init_link_state+0x744/0x850
Hardware name: Kontron B3001/B3001, BIOS 4.6.3 08/07/2012
Call Trace:
[<ffffffff8153b865>] pci_scan_slot+0xd5/0x120
[<ffffffff8153ca1d>] pci_scan_child_bus+0x2d/0xd0
...
Instead of relying on the component type to identify the upstream end of a
link, use the "dev->has_secondary_link" field.
This means it's now possible for an Upstream Port to have a link on its
secondary side, so alloc_pcie_link_state() needs to connect links
originating from both Upstream and Downstream Ports into the tree.
[bhelgaas: changelog, add comment]
Link: https://bugzilla.kernel.org/show_bug.cgi?id=94361
Link: http://lkml.kernel.org/r/54EB81B2.4050904@pobox.com
Reported-by: Robert White <rwhite@pobox.com>
Signed-off-by: Yijing Wang <wangyijing@huawei.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Refine the mechanism introduced by commit f244d8b623 ("ACPIPHP / radeon /
nouveau: Fix VGA switcheroo problem related to hotplug") to propagate the
ignore_hotplug setting of the device to its parent bridge in case hotplug
notifications related to the graphics adapter switching are given for the
bridge rather than for the device itself (they need to be ignored in both
cases).
Link: https://bugzilla.kernel.org/show_bug.cgi?id=61891
Link: https://bugs.freedesktop.org/show_bug.cgi?id=88927
Fixes: b440bde74f ("PCI: Add pci_ignore_hotplug() to ignore hotplug events for a device")
Reported-and-tested-by: tiagdtd-lava <tiagdtd-lava@yahoo.de>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
CC: stable@vger.kernel.org # v3.17+
If the ignore_hotplug flag is set for a PCI device without an ACPI
companion and a bus check notification is received for an ancestor bridge
that is not the device's parent, ACPIPHP will ignore that flag.
Namely, in that case acpiphp_check_bridge() is called for the target bridge
and if all of the devices immediately below the bridge are still present,
trim_stale_devices() will be called for each of them. That function
recursively walks the hierarchy downwards and removes device objects
corresponding to devices that don't appear to be present any more.
Unfortunately, it only checks ignore_hotplug for devices having ACPI
companions, so it will remove the others (if they don't respond) regardless
of the ignore_hotplug value.
Fix the problem by making trim_stale_devices() take ignore_hotplug into
consideration regardless of whether or not an ACPI companion is present for
the device it has been called for.
[bhelgaas: This may fix bug 61891, depending on whether the bridge above a
device is removed along with the device]
Link: https://bugzilla.kernel.org/show_bug.cgi?id=61891
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
The err_out_none label in pciehp_probe() only leads to a return statement,
so use return statements instead of jumps to it and drop it.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
A PCIe Port is an interface to a Link. A Root Port is a PCI-PCI bridge in
a Root Complex and has a Link on its secondary (downstream) side. For
other Ports, the Link may be on either the upstream (closer to the Root
Complex) or downstream side of the Port.
The usual topology has a Root Port connected to an Upstream Port. We
previously assumed this was the only possible topology, and that a
Downstream Port's Link was always on its downstream side, like this:
+---------------------+
+------+ | Downstream |
| Root | | Upstream Port +--Link--
| Port +--Link--+ Port |
+------+ | Downstream |
| Port +--Link--
+---------------------+
But systems do exist (see URL below) where the Root Port is connected to a
Downstream Port. In this case, a Downstream Port's Link may be on either
the upstream or downstream side:
+---------------------+
+------+ | Upstream |
| Root | | Downstream Port +--Link--
| Port +--Link--+ Port |
+------+ | Downstream |
| Port +--Link--
+---------------------+
We can't use the Port type to determine which side the Link is on, so add a
bit in struct pci_dev to keep track.
A Root Port's Link is always on the Port's secondary side. A component
(Endpoint or Port) on the other end of the Link obviously has the Link on
its upstream side. If that component is a Port, it is part of a Switch or
a Bridge. A Bridge has a PCI or PCI-X bus on its secondary side, not a
Link. The internal bus of a Switch connects the Port to another Port whose
Link is on the downstream side.
[bhelgaas: changelog, comment, cache "type", use if/else]
Link: http://lkml.kernel.org/r/54EB81B2.4050904@pobox.com
Link: https://bugzilla.kernel.org/show_bug.cgi?id=94361
Suggested-by: Bjorn Helgaas <bhelgaas@google.com>
Signed-off-by: Yijing Wang <wangyijing@huawei.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Marvell 9120 SATA controller has the same issue as a number of others, so
use the same quirk for this one. The other quirks were added by
cc346a4714 ("PCI: Add function 1 DMA alias quirk for Marvell devices").
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by: Alex Williamson <alex.williamson@redhat.com>
Jarod Wilson reports that ExpressCard hotplug doesn't work on HP ZBook G2.
The problem turns out to be the ACPI-based "slot detection" code called
from pciehp_probe() which uses questionable heuristics based on what ACPI
objects are present for the PCIe port device to figure out whether to
register a hotplug slot for that port.
That code is used if there is at least one PCIe port having an ACPI device
configuration object related to hotplug (such as _EJ0 or _RMV), and the
Thunderbolt port on the ZBook has _RMV. Of course, Thunderbolt and PCIe
native hotplug need not be mutually exclusive (as they aren't on the
ZBook), so that rule is simply incorrect.
Moreover, the ACPI-based "slot detection" check does not add any value if
pciehp_probe() is called at all and the service type of the device object
it has been called for is PCIE_PORT_SERVICE_HP, because PCIe hotplug
services are only registered if the _OSC handshake in acpi_pci_root_add()
allows the kernel to control the PCIe native hotplug feature. No more
checks need to be carried out to decide whether or not to register a native
PCIe hotlug slot in that case.
For the above reasons, make pciehp_probe() check if it has been called for
the right service type and drop the pointless ACPI-based "slot detection"
check from it. Also remove the entire code whose only user is that check
(the entire pciehp_acpi.c file goes away as a result) and drop function
headers related to it from the internal pciehp header file.
Link: http://lkml.kernel.org/r/1431632038-39917-1-git-send-email-jarod@redhat.com
Link: https://bugzilla.kernel.org/show_bug.cgi?id=98581
Reported-by: Jarod Wilson <jarod@redhat.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Jarod Wilson <jarod@redhat.com>
Tested-by: Jarod Wilson <jarod@redhat.com>
Most transactions' type are cfg0 and MEM, so the current iATU usage is not
balanced: iATU0 is hot while iATU1 is rarely used.
Refactor the iATU usage so we use iATU0 for cfg and IO and iATU1 for MEM.
This allocation idea comes from Minghuan Lian
<Minghuan.Lian@freescale.com>:
[bhelgaas: use link with Message-ID]
Link: http://lkml.kernel.org/r/1429091315-31891-3-git-send-email-Minghuan.Lian@freescale.com
Signed-off-by: Jisheng Zhang <jszhang@marvell.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by: Pratyush Anand <pratyush.anand@gmail.com>
Currently, the outbound iATU programming functions are similar: the only
difference is index, type, addr and size. Consolidate these functions into
one. This saves about 1700 bytes in text:
text data bss dec hex filename
9276 204 4 9484 250c pcie-designware.o-before
7532 204 4 7740 1e3c pcie-designware.o
Signed-off-by: Jisheng Zhang <jszhang@marvell.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by: Pratyush Anand <pratyush.anand@gmail.com>
We decide in alloc_pcie_link_state() whether to allocate a pcie_link_state
for a device. After that, it's sufficient to check pdev->link_state. We
don't need to check the PCIe port type again.
Remove the redundant PCIe port type checking.
[bhelgaas: changelog]
Signed-off-by: Yijing Wang <wangyijing@huawei.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
After 387d37577f ("PCI: Don't clear ASPM bits when the FADT declares it's
unsupported"), the "force" parameter to __pci_disable_link_state() is
always "false".
Remove the "force" parameter and assume it's always false.
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
This driver adds support for the PCIe 2.0 controller found on the BCMA bus.
This controller can be found on (mostly) all Broadcom BCM470X / BCM5301X
ARM SoCs.
The driver found in the Broadcom SDK does some more stuff, like setting up
some DMA memory areas, chaining MPS and MRRS to 512 and also some PHY
changes like "improving" the PCIe jitter and doing some special
initialization for the 3rd PCIe port.
This was tested on a bcm4708 board with 2 PCIe ports and wireless cards
connected to them.
PCI_DOMAINS is needed by this driver, because normally there is more than
one PCIe controller and without PCI_DOMAINS only the first controller gets
registered. This controller gets 6 IRQs; the last one is trigged by all
IRQ events.
[bhelgaas: fix "GPLv2" MODULE_LICENSE typo]
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by: Rafał Miłecki <zajec5@gmail.com>
Acked-by: Ray Jui <rjui@broadcom.com.com>
The iProc core PCIe driver defaults to using of_irq_parse_and_map_pci() for
IRQ mapping. Add iproc_pcie.map_irq so bus interfaces that don't use
device tree can override this by supplying their own IRQ mapping function.
[bhelgaas: changelog]
Posting: http://lkml.kernel.org/r/1431465781-10753-1-git-send-email-hauke@hauke-m.de
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Ray Jui <rjui@broadcom.com.com>
Intel confirms that 9-series chipset root ports provide ACS-equivalent
isolation when configured via the existing Intel PCH ACS quirk setup.
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by: Don Dugger <donald.d.dugger@intel.com>
Linus Torvalds