Commit b03e7495 ("PCI: Set PCI-E Max Payload Size on fabric")
introduced a potential NULL pointer dereference in calls to
pcie_bus_configure_settings due to attempts to access pci_bus self
variables when the self pointer is NULL.

To correct this, verify that the self pointer in pci_bus is non-NULL
before dereferencing it.
Reported-by: NStanislaw Gruszka <sgruszka@redhat.com>
Signed-off-by: NShyam Iyer <shyam_iyer@dell.com>
Signed-off-by: NJon Mason <mason@myri.com>
Acked-by: NJesse Barnes <jbarnes@virtuousgeek.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

On a given PCI-E fabric, each device, bridge, and root port can have a
different PCI-E maximum payload size.  There is a sizable performance
boost for having the largest possible maximum payload size on each PCI-E
device.  However, if improperly configured, fatal bus errors can occur.
Thus, it is important to ensure that PCI-E payloads sends by a device
are never larger than the MPS setting of all devices on the way to the
destination.

This can be achieved two ways:

- A conservative approach is to use the smallest common denominator of
  the entire tree below a root complex for every device on that fabric.

This means for example that having a 128 bytes MPS USB controller on one
leg of a switch will dramatically reduce performances of a video card or
10GE adapter on another leg of that same switch.

It also means that any hierarchy supporting hotplug slots (including
expresscard or thunderbolt I suppose, dbl check that) will have to be
entirely clamped to 128 bytes since we cannot predict what will be
plugged into those slots, and we cannot change the MPS on a "live"
system.

- A more optimal way is possible, if it falls within a couple of
  constraints:
* The top-level host bridge will never generate packets larger than the
  smallest TLP (or if it can be controlled independently from its MPS at
  least)
* The device will never generate packets larger than MPS (which can be
  configured via MRRS)
* No support of direct PCI-E <-> PCI-E transfers between devices without
  some additional code to specifically deal with that case

Then we can use an approach that basically ignores downstream requests
and focuses exclusively on upstream requests. In that case, all we need
to care about is that a device MPS is no larger than its parent MPS,
which allows us to keep all switches/bridges to the max MPS supported by
their parent and eventually the PHB.

In this case, your USB controller would no longer "starve" your 10GE
Ethernet and your hotplug slots won't affect your global MPS.
Additionally, the hotplugged devices themselves can be configured to a
larger MPS up to the value configured in the hotplug bridge.

To choose between the two available options, two PCI kernel boot args
have been added to the PCI calls.  "pcie_bus_safe" will provide the
former behavior, while "pcie_bus_perf" will perform the latter behavior.
By default, the latter behavior is used.

NOTE: due to the location of the enablement, each arch will need to add
calls to this function.  This patch only enables x86.

This patch includes a number of changes recommended by Benjamin
Herrenschmidt.

Tested-by: Jordan_Hargrave@dell.com
Signed-off-by: NJon Mason <mason@myri.com>
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>

Commit 6e8af08d enables pci=bfsort on
future Dell systems. But the identification string 'Dell System' matches
on already existing whitelist, which do not have SMBIOS type 0xB1,
causing pci=bfsort not being set on existing whitelist.

This patch fixes the regression by moving the type 0xB1 check beyond the
existing whitelist so that existing whitelist is walked before.
Signed-off-by: NShyam Iyer <shyam_iyer@dell.com>
Signed-off-by: NNarendra K <narendra_k@dell.com>
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>

Aside of the usual motivation for constification,  this function has a
history of being abused a hook for interrupt and other fixups so I turned
this function const ages ago in the MIPS code but it should be done
treewide.

Due to function pointer passing in varous places a few other functions
had to be constified as well.
Signed-off-by: NRalf Baechle <ralf@linux-mips.org>
To: Anton Vorontsov <avorontsov@mvista.com>
To: Chris Metcalf <cmetcalf@tilera.com>
To: Colin Cross <ccross@android.com>
Acked-by: N"David S. Miller" <davem@davemloft.net>
To: Eric Miao <eric.y.miao@gmail.com>
To: Erik Gilling <konkers@android.com>
Acked-by: NGuan Xuetao <gxt@mprc.pku.edu.cn>
To: "H. Peter Anvin" <hpa@zytor.com>
To: Imre Kaloz <kaloz@openwrt.org>
To: Ingo Molnar <mingo@redhat.com>
To: Ivan Kokshaysky <ink@jurassic.park.msu.ru>
To: Jesse Barnes <jbarnes@virtuousgeek.org>
To: Krzysztof Halasa <khc@pm.waw.pl>
To: Lennert Buytenhek <kernel@wantstofly.org>
To: Matt Turner <mattst88@gmail.com>
To: Nicolas Pitre <nico@fluxnic.net>
To: Olof Johansson <olof@lixom.net>
Acked-by: NPaul Mundt <lethal@linux-sh.org>
To: Richard Henderson <rth@twiddle.net>
To: Russell King <linux@arm.linux.org.uk>
To: Thomas Gleixner <tglx@linutronix.de>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: linux-alpha@vger.kernel.org
Cc: linux-arm-kernel@lists.infradead.org
Cc: linux-kernel@vger.kernel.org
Cc: linux-mips@linux-mips.org
Cc: linux-pci@vger.kernel.org
Cc: linux-sh@vger.kernel.org
Cc: linux-tegra@vger.kernel.org
Cc: sparclinux@vger.kernel.org
Cc: x86@kernel.org
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>

Without this change, the majority of the raw PCI config space access
functions silently ignore a non-zero segment argument, which is
certainly wrong.

Apart from pci_direct_conf1, all other non-MMCFG access methods get
used only for non-extended accesses (i.e. assigned to raw_pci_ops
only). Consequently, with the way raw_pci_{read,write}() work, it would
be a coding error to call these functions with a non-zero segment (with
the current call flow this cannot happen afaict).

The access method 1 accessor, as it can be used for extended accesses
(on AMD systems) instead gets checks added for the passed in segment to
be zero. This would be the case when on such a system having multiple
PCI segments (don't know whether any exist in practice) MMCFG for some
reason is not usable, and method 1 gets selected for doing extended
accesses. Rather than accessing the wrong device's config space, the
function will now error out.

v2: Convert BUG_ON() to WARN_ON(), and extend description as per Ingo's
request.
Signed-off-by: NJan Beulich <jbeulich@novell.com>
Reviewed-by: NIngo Molnar <mingo@elte.hu>
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>

Host bridge windows are top-level resources, so if we find a host bridge
window conflict, it's probably with a hard-coded legacy reservation.
Moving host bridge windows is theoretically possible, but we don't support
it; we just ignore windows with conflicts, and it's not worth making this
a user-visible error.
Reported-and-tested-by: NJools Wills <jools@oxfordinspire.co.uk>
References: https://bugzilla.kernel.org/show_bug.cgi?id=38522Reported-by: NDas <dasfox@gmail.com>
References: https://bugzilla.kernel.org/show_bug.cgi?id=16497Signed-off-by: NBjorn Helgaas <bhelgaas@google.com>
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>

In the past we would only use the function's value if the
returned value was not equal to 'acpi_sci_override_gsi'. Meaning
that the INT_SRV_OVR values for global and source irq were different.
But it is OK to use the function's value even when the global
and source irq are the same.
Signed-off-by: NKonrad Rzeszutek Wilk <konrad.wilk@oracle.com>

In the past (2.6.38) the 'xen_allocate_pirq_gsi' would allocate
an entry in a Linux IRQ -> {XEN_IRQ, type, event, ..} array. All
of that has been removed in 2.6.39 and the Xen IRQ subsystem uses
an linked list that is populated when the call to
'xen_allocate_irq_gsi' (universally done from any of the xen_bind_*
calls) is done. The 'xen_allocate_pirq_gsi' is a NOP and there is
no need for it anymore so lets remove it.
Signed-off-by: NKonrad Rzeszutek Wilk <konrad.wilk@oracle.com>

As the code paths that are guarded by CONFIG_XEN_DOM0 already depend
on CONFIG_ACPI so the extra #ifdef is not required. The earlier
patch that added them in had done its job.
Signed-off-by: NKonrad Rzeszutek Wilk <konrad.wilk@oracle.com>

.. which means we can preset of NR_IRQS_LEGACY interrupts using
the 'acpi_get_override_irq' API before this loop.
This means that we can get the IRQ's polarity (and trigger) from either
the ACPI (or MP); or use the default values. This fixes a bug if we did
not have an IOAPIC we would not been able to preset the IRQ's polarity
if the MP table existed.
Signed-off-by: NKonrad Rzeszutek Wilk <konrad.wilk@oracle.com>

Since they are only called once and the rest of the pci_xen_*
functions follow the same pattern of setup.
Signed-off-by: NKonrad Rzeszutek Wilk <konrad.wilk@oracle.com>

.. to cut down on the code duplicity.
Signed-off-by: NKonrad Rzeszutek Wilk <konrad.wilk@oracle.com>

Not before .. also that code segment starts looking like the HVM one.
Signed-off-by: NKonrad Rzeszutek Wilk <konrad.wilk@oracle.com>

In the past we would guard those code segments to be dependent
on CONFIG_XEN_DOM0 (which depends on CONFIG_ACPI) so this patch is
not stricly necessary. But the next patch will merge common
HVM and initial domain code and we want to make sure the CONFIG_ACPI
dependency is preserved - as HVM code does not depend on CONFIG_XEN_DOM0.
Signed-off-by: NKonrad Rzeszutek Wilk <konrad.wilk@oracle.com>

Update the out-dated comment at the beginning of the file.
Also provide the copyrights of folks who have been contributing
to this code lately.
Signed-off-by: NKonrad Rzeszutek Wilk <konrad.wilk@oracle.com>

The file is hard to read. Move the code around so that
the contents of it follows a uniform format:
 - setup GSIs - PV, HVM, and initial domain case
 - then MSI/MSI-x setup - PV, HVM and then initial domain case.
 - then MSI/MSI-x teardown - same order.
 - lastly, the __init functions in PV, HVM, and initial domain order.
Signed-off-by: NKonrad Rzeszutek Wilk <konrad.wilk@oracle.com>

Previously we would check for acpi_sci_override_gsi == gsi every time
a PCI device was enabled. That works during early bootup, but later
on it could lead to triggering unnecessarily the acpi_gsi_to_irq(..) lookup.
The reason is that acpi_sci_override_gsi was declared in __initdata and
after early bootup could contain bogus values.

This patch moves the check for acpi_sci_override_gsi to the
site where the ACPI SCI is preset.

CC: stable@kernel.org
Reported-by: NRaghavendra D Prabhu <rprabhu@wnohang.net>
Tested-by: NRaghavendra D Prabhu <rprabhu@wnohang.net>
[http://lists.xensource.com/archives/html/xen-devel/2011-07/msg00154.html]
Suggested-by: NIan Campbell <ijc@hellion.org.uk>
Signed-off-by: NKonrad Rzeszutek Wilk <konrad.wilk@oracle.com>

In the past we would use the GSI value to preset the ACPI SCI
IRQ which worked great as GSI == IRQ:

ACPI: INT_SRC_OVR (bus 0 bus_irq 9 global_irq 9 low level)

While that is most often seen, there are some oddities:

ACPI: INT_SRC_OVR (bus 0 bus_irq 9 global_irq 20 low level)

which means that GSI 20 (or pin 20) is to be overriden for IRQ 9.
Our code that presets the interrupt for ACPI SCI however would
use the GSI 20 instead of IRQ 9 ending up with:

xen: sci override: global_irq=20 trigger=0 polarity=1
xen: registering gsi 20 triggering 0 polarity 1
xen: --> pirq=20 -> irq=20
xen: acpi sci 20
.. snip..
calling  acpi_init+0x0/0xbc @ 1
ACPI: SCI (IRQ9) allocation failed
ACPI Exception: AE_NOT_ACQUIRED, Unable to install System Control Interrupt handler (20110413/evevent-119)
ACPI: Unable to start the ACPI Interpreter

as the ACPI interpreter made a call to 'acpi_gsi_to_irq' which got nine.
It used that value to request an IRQ (request_irq) and since that was not
present it failed.

The fix is to recognize that for interrupts that are overriden (in our
case we only care about the ACPI SCI) we should use the IRQ number
to present the IRQ instead of the using GSI. End result is that we get:

xen: sci override: global_irq=20 trigger=0 polarity=1
xen: registering gsi 20 triggering 0 polarity 1
xen: --> pirq=20 -> irq=9 (gsi=9)
xen: acpi sci 9

which fixes the ACPI interpreter failing on startup.

CC: stable@kernel.org
Reported-by: NLiwei <xieliwei@gmail.com>
Tested-by: NLiwei <xieliwei@gmail.com>
[http://lists.xensource.com/archives/html/xen-devel/2011-06/msg01727.html]
Signed-off-by: NKonrad Rzeszutek Wilk <konrad.wilk@oracle.com>

Enable 64-bit ACPI MFCG support for SGI UV2 platform. The check
is similar to the check on UV1. UV2 has a different oem_id
string.
Signed-off-by: NJack Steiner <steiner@sgi.com>
Link: http://lkml.kernel.org/r/20110602195943.GA27079@sgi.comSigned-off-by: NIngo Molnar <mingo@elte.hu>

The flags field of struct resource from linux/ioport.h is "unsigned
long". Change the "type" parameter of coalesce_windows() function to
match that field. This fixes the following warning messages when
compiling with "make C=1 W=1 bzImage modules":

arch/x86/pci/acpi.c: In function ‘coalesce_windows’:
arch/x86/pci/acpi.c:198: warning: conversion to ‘long unsigned int’ from ‘int’ may change the sign of the result
arch/x86/pci/acpi.c:203: warning: conversion to ‘long unsigned int’ from ‘int’ may change the sign of the result
Signed-off-by: NMárton Németh <nm127@freemail.hu>
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>

On various newer Intel systems the PCI bus(ses) the non-core devices
live on aren't getting announced by ACPI except through the bus range
covered by mmconfig. At least the i7core-edac driver depends on these
devices getting detected.

Mauro, could you check whether with this change the Xeon 55xx hack in
that driver can go away altogether, and with it the bogus exporting of
pcibios_scan_specific_bus()?
Signed-off-by: NJan Beulich <jbeulich@novell.com>
Cc: Mauro Carvalho Chehab <mchehab@redhat.com>
Cc: Aristeu Sergio <arozansk@redhat.com>
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>

If we have CONFIG_XEN and the other parameters to build an
Linux kernel that is non-privileged, the xen_[find|register|unregister]_
device_domain_owner functions should not be compiled. They should
use the nops defined in arch/x86/include/asm/xen/pci.h instead.

This fixes:

arch/x86/pci/xen.c:496: error: redefinition of ‘xen_find_device_domain_owner’
arch/x86/include/asm/xen/pci.h:25: note: previous definition of ‘xen_find_device_domain_owner’ was here
arch/x86/pci/xen.c:510: error: redefinition of ‘xen_register_device_domain_owner’
arch/x86/include/asm/xen/pci.h:29: note: previous definition of ‘xen_register_device_domain_owner’ was here
arch/x86/pci/xen.c:532: error: redefinition of ‘xen_unregister_device_domain_owner’
arch/x86/include/asm/xen/pci.h:34: note: previous definition of ‘xen_unregister_device_domain_owner’ was here
Signed-off-by: NKonrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Reported-by: NRandy Dunlap <randy.dunlap@oracle.com>

This patch adds the LPC Controller DeviceIDs for the Intel Panther Point PCH.
Acked-by: NJean Delvare <khali@linux-fr.org>
Signed-off-by: NSeth Heasley <seth.heasley@intel.com>
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>

Request_region should be used with release_region, not release_resource.

The local variables region and region2 are dropped and the calls to
release_resource are replaced with calls to release_region, using the first
two arguments of the corresponding calls to request_region.

The semantic match that finds this problem is as follows:
(http://coccinelle.lip6.fr/)

// <smpl>
@@
expression x,E;
@@
(
*x = request_region(...)
|
*x = request_mem_region(...)
)
... when != release_region(x)
    when != x = E
* release_resource(x);
// </smpl>
Signed-off-by: NJulia Lawall <julia@diku.dk>
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>

We check if there is a domain owner for the PCI device. In case of failure
(meaning no domain has registered for this device) we make DOMID_SELF the owner.
Signed-off-by: NKonrad Rzeszutek Wilk <konrad.wilk@oracle.com>
[v2: deal with rebasing on v2.6.37-1]
[v3: deal with rebasing on stable/irq.cleanup]
[v4: deal with rebasing on stable/irq.ween_of_nr_irqs]
[v5: deal with rebasing on v2.6.39-rc3]
Signed-off-by: NJeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
Acked-by: NXiantao Zhang <xiantao.zhang@intel.com>

When the Xen PCI backend is told to enable or disable MSI/MSI-X functions,
the initial domain performs these operations. The initial domain needs
to know which domain (guest) is going to use the PCI device so when it
makes the appropiate hypercall to retrieve the MSI/MSI-X vector it will
also assign the PCI device to the appropiate domain (guest).

This boils down to us needing a mechanism to find, set and unset the domain
id that will be using the device.

[v2: EXPORT_SYMBOL -> EXPORT_SYMBOL_GPL.]
Signed-off-by: NKonrad Rzeszutek Wilk <konrad.wilk@oracle.com>

They were generated by 'codespell' and then manually reviewed.
Signed-off-by: NLucas De Marchi <lucas.demarchi@profusion.mobi>
Cc: trivial@kernel.org
LKML-Reference: <1300389856-1099-3-git-send-email-lucas.demarchi@profusion.mobi>
Signed-off-by: NIngo Molnar <mingo@elte.hu>

Setting the pci ops on subsys initcall unconditionally will break
multi platform kernels on anything except ce4100.

Use x86_init.pci.init ops to call this only on real ce4100 platforms.
Signed-off-by: NSebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: sodaville@linutronix.de
LKML-Reference: <20110314093340.GA21026@www.tglx.de>
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>

genirq is switching to a consistent name space for the irq related
functions. Convert x86. Conversion was done with coccinelle.
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>

Following the example set by xen_allocate_pirq_msi and
xen_bind_pirq_msi_to_irq:

xen_allocate_pirq becomes xen_allocate_pirq_gsi and now only allocates
a pirq number and does not bind it.

xen_map_pirq_gsi becomes xen_bind_pirq_gsi_to_irq and binds an
existing pirq.
Signed-off-by: NIan Campbell <ian.campbell@citrix.com>
Signed-off-by: NKonrad Rzeszutek Wilk <konrad.wilk@oracle.com>

The function name does not distinguish it from xen_allocate_pirq_msi
(which operates on domU and pvhvm domains rather than dom0).

Hoist domain 0 specific functionality up into the only caller leaving
functionality common to all guest types in xen_bind_pirq_msi_to_irq.
Signed-off-by: NIan Campbell <ian.campbell@citrix.com>
Signed-off-by: NKonrad Rzeszutek Wilk <konrad.wilk@oracle.com>

Signed-off-by: NIan Campbell <ian.campbell@citrix.com>
Signed-off-by: NKonrad Rzeszutek Wilk <konrad.wilk@oracle.com>

Makes the tail end of this function look even more like
xen_bind_pirq_msi_to_irq.
Signed-off-by: NIan Campbell <ian.campbell@citrix.com>
Signed-off-by: NKonrad Rzeszutek Wilk <konrad.wilk@oracle.com>

Split the binding aspect of xen_allocate_pirq_msi out into a new
xen_bind_pirq_to_irq function.

In xen_hvm_setup_msi_irq when allocating a pirq write the MSI message
to signal the PIRQ as soon as the pirq is obtained. There is no way to
free the pirq back so if the subsequent binding to an IRQ fails we
want to ensure that we will reuse the PIRQ next time rather than leak
it.
Signed-off-by: NIan Campbell <ian.campbell@citrix.com>
Signed-off-by: NKonrad Rzeszutek Wilk <konrad.wilk@oracle.com>

apic_register_gsi_xen_hvm is a tiny wrapper around
xen_hvm_register_pirq.
Signed-off-by: NIan Campbell <ian.campbell@citrix.com>
Signed-off-by: NKonrad Rzeszutek Wilk <konrad.wilk@oracle.com>

consistent with other similar functions.
Signed-off-by: NIan Campbell <ian.campbell@citrix.com>
Signed-off-by: NKonrad Rzeszutek Wilk <konrad.wilk@oracle.com>

All callers pass this flag so it is pointless.
Signed-off-by: NIan Campbell <ian.campbell@citrix.com>
Signed-off-by: NKonrad Rzeszutek Wilk <konrad.wilk@oracle.com>

Fixes:
 CC      arch/x86/pci/xen.o
arch/x86/pci/xen.c:183: warning: 'xen_initdom_setup_msi_irqs' defined but not used
Signed-off-by: NIan Campbell <ian.campbell@citrix.com>
Signed-off-by: NKonrad Rzeszutek Wilk <konrad.wilk@oracle.com>

This happens to not be an issue currently because we take pains to try
to ensure that the GSI-IRQ mapping is 1-1 in a PV guest and that
regular event channels do not clash. However a subsequent patch is
going to break this 1-1 mapping.
Signed-off-by: NIan Campbell <ian.campbell@citrix.com>
Signed-off-by: NKonrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Cc: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
Cc: Jeremy Fitzhardinge <jeremy@goop.org>

This code path is only run when an MSI/MSI-X PCI device is passed
in to PV DomU.

In 2.6.37 time-frame we over-wrote the default cleanup handler for
MSI/MSI-X irq->desc to be "xen_teardown_msi_irqs". That function
calls the the xen-pcifront driver which can tell the backend to
cleanup/take back the MSI/MSI-X device.

However, we forgot to continue the process of free-ing the MSI/MSI-X
device resources (irq->desc) in the PV domU side. Which is what
the default cleanup handler: default_teardown_msi_irqs did.

Hence we would leak IRQ descriptors.

Without this patch, doing "rmmod igbvf;modprobe igbvf" multiple
times ends with abandoned IRQ descriptors:

 28:          5  xen-pirq-pcifront-msi-x
 29:          8  xen-pirq-pcifront-msi-x
...
130:         10  xen-pirq-pcifront-msi-x

with the end result of running out of IRQ descriptors.
Reviewed-by: NIan Campbell <Ian.Campbell@citrix.com>
Signed-off-by: NKonrad Rzeszutek Wilk <konrad.wilk@oracle.com>