* 'upstream-linus' of master.kernel.org:/pub/scm/linux/kernel/git/jgarzik/netdev-2.6: (45 commits)
  [PATCH] Restore channel setting after scan.
  [PATCH] hostap: Fix memory leak on PCI probe error path
  [PATCH] hostap: Remove dead code (duplicated idx != 0)
  [PATCH] hostap: Fix unlikely read overrun in CIS parsing
  [PATCH] hostap: Fix double free in prism2_config() error path
  [PATCH] hostap: Fix ap_add_sta() return value verification
  [PATCH] hostap: Fix hw reset after CMDCODE_ACCESS_WRITE timeout
  [PATCH] wireless/airo: cache wireless scans
  [PATCH] wireless/airo: define default MTU
  [PATCH] wireless/airo: clean up printk usage to print device name
  [PATCH] WE-20 for kernel 2.6.16
  [PATCH] softmac: remove function_enter()
  [PATCH] skge: version 1.5
  [PATCH] skge: compute available ring buffers
  [PATCH] skge: dont free skb until multi-part transmit complete
  [PATCH] skge: multicast statistics fix
  [PATCH] skge: rx_reuse called twice
  [PATCH] skge: dont use dev_alloc_skb for rx buffs
  [PATCH] skge: align receive buffers
  [PATCH] sky2: dont need to use dev_kfree_skb_any
  ...

* 'blktrace' of git://brick.kernel.dk/data/git/linux-2.6-block:
  [PATCH] Block queue IO tracing support (blktrace) as of 2006-03-23
  [PATCH] relay: consolidate sendfile() and read() code
  [PATCH] relay: add sendfile() support
  [PATCH] relay: migrate from relayfs to a generic relay API

* master.kernel.org:/pub/scm/linux/kernel/git/mchehab/v4l-dvb:
  V4L/DVB (3568k): zoran: Use i2c_master_send when possible
  V4L/DVB (3568j): adv7175: Drop unused encoder dump command
  V4L/DVB (3568i): adv7175: Drop unused register cache
  V4L/DVB (3568h): cpia: correct email address
  V4L/DVB (3568g): sem2mutex: zoran
  V4L/DVB (3568f): saa7110: Fix array overrun
  V4L/DVB (3568e): bt856: Spare memory
  V4L/DVB (3568d): saa7111.c fix
  V4L/DVB (3568c): zoran: Init cleanups
  V4L/DVB (3568b): saa7111: Prevent array overrun
  V4L/DVB (3568a): saa7114: Fix i2c block write

* master.kernel.org:/pub/scm/linux/kernel/git/gregkh/i2c-2.6: (36 commits)
  [PATCH] hwmon: add required idr locking
  [PATCH] I2C: hwmon: Rename register parameters
  [PATCH] I2C: Drop unneeded i2c-dev.h includes
  [PATCH] I2C: i2c-ixp4xx: Add hwmon class
  [PATCH] I2C: i2c-piix4: Add Broadcom HT-1000 support
  [PATCH] I2C: i2c-amd756-s4882: Improve static mutex initialization
  [PATCH] I2C: i2c-ali1535: Drop redundant mutex
  [PATCH] i2c: Cleanup isp1301_omap
  [PATCH] i2c: Fix i2c-ite name initialization
  [PATCH] i2c: Drop the i2c-frodo bus driver
  [PATCH] i2c: Optimize core_lists mutex usage
  [PATCH] w83781d: Don't reset the chip by default
  [PATCH] w83781d: Document the alarm and beep bits
  [PATCH] w83627ehf: Refactor the sysfs interface
  [PATCH] hwmon: Support the Pentium M VID code
  [PATCH] hwmon: Add support for the Winbond W83687THF
  [PATCH] hwmon: f71805f semaphore to mutex conversions
  [PATCH] hwmon: Semaphore to mutex conversions
  [PATCH] i2c: Semaphore to mutex conversions, part 3
  [PATCH] i2c: Semaphore to mutex conversions, part 2
  ...

* master.kernel.org:/pub/scm/linux/kernel/git/gregkh/pci-2.6: (49 commits)
  [PATCH] acpiphp: fix acpi_path_name
  [PATCH] ibmphp: remove TRUE and FALSE
  [PATCH] PCI Hotplug: add common acpi functions to core
  [PATCH] PCI: kzalloc() conversion in drivers/pci
  [PATCH] acpiphp: Scan slots under the nested P2P bridge
  [PATCH] PCI Hotplug: SN: Fix cleanup on hotplug removal of PPB
  [PATCH] shpchp: cleanup bus speed handling
  [PATCH] PCI: fix pci_request_region[s] arg
  [PATCH] PCI: Provide a boot parameter to disable MSI
  [PATCH] PCI: the scheduled removal of PCI_LEGACY_PROC
  [PATCH] PCI: cpqphp_ctrl.c: board_replaced(): remove dead code
  [PATCH] acpiphp: fix bridge handle
  [PATCH] acpiphp - slot management fix - V4
  [PATCH] acpi: remove dock event handling from ibm_acpi
  [PATCH] acpiphp: add dock event handling
  [PATCH] acpi: export acpi_bus_trim
  [PATCH] acpiphp: add new bus to acpi
  [PATCH] PCI: Move pci_dev_put outside a spinlock
  [PATCH] PCI: PCI/Cardbus cards hidden, needs pci=assign-busses to fix
  [PATCH] PCI: fix problems with MSI-X on ia64
  ...

* master.kernel.org:/pub/scm/linux/kernel/git/davem/net-2.6:
  [TG3]: Update driver version and reldate.
  [TG3]: Add 5755 nvram support
  [TG3]: Add 5755 support
  [IPV6]: ip6_xmit: remove unnecessary NULL ptr check
  [NET_SCHED]: cls_u32: remove unnecessary NULL-ptr check
  [IPV4]: Add fib rule netlink notifications
  [BNX2]: Update version and reldate
  [BNX2]: Separate tx producer and consumer fields
  [BNX2]: Move .h files to bnx2.c
  [BNX2]: Combine small mem allocations
  [BNX2]: Fix link change handling
  [PKTGEN]: Add MPLS extension.

* git://oss.sgi.com:8090/oss/git/xfs-2.6: (71 commits)
  [XFS] Sync up one/two other minor changes missed in previous merges.
  [XFS] Reenable the noikeep (delete inode cluster space) option by default.
  [XFS] Check that a page has dirty buffers before finding it acceptable for
  [XFS] Fixup naming inconsistencies found by Pekka Enberg and one from Jan
  [XFS] Explain the race closed by the addition of vn_iowait() to the start
  [XFS] Fixing the error caused by the conflict between DIO Write's
  [XFS] Fixing KDB's xrwtrc command, also added the current process id into
  [XFS] Fix compiler warning from xfs_file_compat_invis_ioctl prototype. 
  [XFS] remove bogus INT_GET for u8 variables in xfs_dir_leaf.c 
  [XFS] endianess annotations for xfs_da_node_hdr_t 
  [XFS] endianess annotations for xfs_da_node_entry_t 
  [XFS] store xfs_attr_inactive_list_t in native endian 
  [XFS] store xfs_attr_sf_sort in native endian 
  [XFS] endianess annotations for xfs_attr_shortform_t 
  [XFS] endianess annotations for xfs_attr_leaf_name_remote_t 
  [XFS] endianess annotations for xfs_attr_leaf_name_local_t 
  [XFS] endianess annotations for xfs_attr_leaf_entry_t 
  [XFS] endianess annotations for xfs_attr_leaf_hdr_t 
  [XFS] remove bogus INT_GET on u8 variables in xfs_dir2_block.c 
  [XFS] endianess annotations for xfs_da_blkinfo_t 
  ...

I encountered the problem that the insmod of the acpiphp
fails because of the mis-freeing of the memory.

I tested this patch on my tiger4 box.
Signed-off-by: NMUNEDA Takahiro <muneda.takahiro@jp.fujitsu.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

This patch removes the defines TRUE and FALSE and just uses 0 or 1.
Signed-off-by: NKristen Carlson Accardi <kristen.c.accardi@intel.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

shpchprm_acpi.c and pciehprm_acpi.c are nearly identical.  In addition,
there are functions in both these files that are also in acpiphp_glue.c.
This patch will remove duplicate functions from shpchp, pciehp, and
acpiphp and move this functionality to pci_hotplug, as it is not
hardware specific.  Get rid of shpchprm* and pciehprm* files since they
are no longer needed.  shpchprm_nonacpi.c and pciehprm_nonacpi.c are
identical, as well as shpchprm_legacy.c and can be replaced with a
macro.

This patch also changes acpiphp to use the common hpp code.
Signed-off-by: NKristen Carlson Accardi <kristen.c.accardi@intel.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

this patch converts drivers/pci to kzalloc usage.
Compile tested with allyes config.
Signed-off-by: NEric Sesterhenn <snakebyte@gmx.de>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

Current ACPIPHP driver scans only slots under the top level PCI-to-PCI
bridge. So hotplug PCI slots under the nested PCI-to-PCI bridge would
not be detected. For example, if the system has the ACPI namespace
like below, hotplug slots woule not be detected.

Device (PCI0) {						/* Root bridge */
	Name (_HID, "PNP0A03")
	Device (P2PA) {					/* PCI-to-PCI bridge */
		Name (_ADR, ...)
		Device (P2PB) {				/* PCI-to-PCI bridge */
			Name (_ADR, ...)
			Device (S0F0) {			/* hotplug slot */
				Name (_ADR, ...)
				Name (_SUN, ...)
				Method (_EJ0, ...) { ... }
			}
			...
			Device (S0F7) {			/* hotplug slot */
				Name (_ADR, ...)
				Name (_SUN, ...)
				Method (_EJ0, ...) { ... }
			}
			Device (S1F0) {			/* hotplug slot */
				Name (_ADR, ...)
				Name (_SUN, ...)
				Method (_EJ0, ...) { ... }
			}
			...
		}
	}
}

This patch fixes this issue.
Signed-off-by: NKenji Kaneshige <kaneshige.kenji@jp.fujitsu.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

When doing a hotplug removal of a PPB, sn_bus_store_sysdata()
needs to be called for the PPB and all of its children.
Acked-by: NPrarit Bhargava <prarit@sgi.com>
Signed-off-by: NJohn Keller <jpk@sgi.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

The code related to handling bus speed in SHPCHP driver is
unnecessarily complex. This patch cleans up and simplify that.
Signed-off-by: NKenji Kaneshige <kaneshige.kenji@jp.fujitsu.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

    Add missing 'const' to pci_request_region[s] 'res_name' arg,
    since we pass it directly to __request_region(), whose 'name' arg
    is also const.
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

Several drivers are starting to grow options to disable MSI.  However,
it's often a host chipset issue, not something which individual drivers
should handle.  So we add the pci=nomsi kernel parameter to allow the user
to disable MSI modes for systems we haven't added to the quirk list yet.
Signed-off-by: NMatthew Wilcox <matthew@wil.cx>
Signed-off-by: NRandy Dunlap <rdunlap@xenotime.net>
Acked-by: NJeff Garzik <jeff@garzik.org>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

This patch contains the scheduled removal of PCI_LEGACY_PROC.
Signed-off-by: NAdrian Bunk <bunk@stusta.de>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

The Coverity checker correctly noted, that in function board_replaced in
drivers/pci/hotplug/cpqphp_ctrl.c, the variable src always has the
value 8, and therefore much code after the

...
                        if (rc || src) {
...
                                if (rc)
                                        return rc;
                                else
                                        return 1;
                        }
...

can never be called.

This patch removes the unreachable code in this function fixing kernel
Bugzilla #6073.
Signed-off-by: NAdrian Bunk <bunk@stusta.de>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

When hotplug slot is under the host bridge,
DEVICE_ACPI_HANDLE(&bus->self->dev) fails since '&bus->self' was not set.
This patch fixes it.

This patch is based on kristen's latest patches.
I tested this patch on my Tiger4.
Signed-off-by: NMUNEDA Takahiro <muneda.takahiro@jp.fujitsu.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

o This patch removes IDs (for slots management).
o This patch removes the slot register/unregister processes
  from the init/exit phases. Instead, adds these processes
  in the bridge add/cleanup phases.
o Currently, this change doesn't have any meanings. But
  these changes are needed to support p2p bridge(with
  hotplug slot)
Signed-off-by: NMUNEDA Takahiro <muneda.takahiro@jp.fujitsu.com>
Signed-off-by: NKristen Carlson Accardi <kristen.c.accardi@intel.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

Remove dock station support from ibm_acpi by default.  This support has
been put into acpiphp instead.  Allow ibm_acpi to continue to provide
docking station support via config option for laptops/docking stations
that are not supported by acpiphp.
Signed-off-by: NKristen Carlson Accardi <kristen.c.accardi@intel.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

These patches add generic dock event handling to acpiphp.  If there are
pci devices that need to be inserted/removed after the dock event, the
event notification will be handed down to the normal pci hotplug event
handler in acpiphp so that new bridges/devices can be enumerated.

Because some dock stations do not have pci bridges or pci devices that
need to be inserted after a dock, acpiphp will remain loaded to handle
dock events even if no hotpluggable pci slots are discovered.

You probably need to have the pci=assign-busses kernel parameter enabled
to use these patches, and you may not allow ibm_acpi to handle docking
notifications and use this patch.

This patch incorporates feedback provided by many.
Signed-off-by: NKristen Carlson Accardi <kristen.c.accardi@intel.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

Export the acpi_bus_trim function so that the pci hotplug driver can
use it.
Signed-off-by: NKristen Carlson Accardi <kristen.c.accardi@intel.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

If we add a new bridge with subordinate busses, we should call make sure
that acpi is notified so that the PRT (if present) can be read and drivers
who have registered on this bus will be notified when it is started.
Also make sure to use the max reserved bus number for the starting the bus
scan.
Signed-off-by: NKristen Carlson Accardi <kristen.c.accardi@intel.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

This patch (as659) fixes a might_sleep problem in the PCI core, by moving
a call to pci_dev_put() outside the scope of a spinlock.
Signed-off-by: NAlan Stern <stern@rowland.harvard.edu>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

"In some cases, especially on modern laptops with a lot of PCI and cardbus
bridges, we're unable to assign correct secondary/subordinate bus numbers
to all cardbus bridges due to BIOS limitations unless we are using
"pci=assign-busses" boot option." -- Ivan Kokshaysky (from a patch comment)

Without it, Cardbus cards inserted are never seen by PCI because the parent
PCI-PCI Bridge of the Cardbus bridge will not pass and translate Type 1 PCI
configuration cycles correctly and the system will fail to find and
initialise the PCI devices in the system.

Reference: PCI-PCI Bridges: PCI Configuration Cycles and PCI Bus Numbering:
http://www.science.unitn.it/~fiorella/guidelinux/tlk/node72.html

The reason for this is that:
 ``All PCI busses located behind a PCI-PCI bridge must reside between the
secondary bus number and the subordinate bus number (inclusive).''

"pci=assign-busses" makes pcibios_assign_all_busses return 1 and this
turns on PCI renumbering during PCI probing.

Alan suggested to use DMI automatically set assign-busses on problem systems.

The only question for me was where to put it.  I put it directly before
scanning PCI bus into pcibios_scan_root() because it's called from legacy,
acpi and numa and so it can be one place for all systems and configurations
which may need it.

AMD64 Laptops are also affected and fixed by assign-busses, and the code is
also incuded from arch/x86_64/pci/ that place will also work for x86_64
kernels, I only ifdef'-ed the x86-only Laptop in this example.

Affected and known or assumed to be fixed with it are (found by googling):

* ASUS Z71V and L3s
* Samsung X20
* Compaq R3140us and all Compaq R3000 series laptops with TI1620 Controller,
  also Compaq R4000 series (from a kernel.org bugreport)
* HP zv5000z (AMD64 3700+, known that fixup_parent_subordinate_busnr fixes it)
* HP zv5200z
* IBM ThinkPad 240
* An IBM ThinkPad (1.8 GHz Pentium M) debugged by Pavel Machek
  gives the correspondig message which detects the possible problem.
* MSI S260 / Medion SIM 2100 MD 95600

The patch also expands the "try pci=assign-busses" warning so testers will
help us to update the DMI table.

Cc: Ivan Kokshaysky <ink@jurassic.park.msu.ru>
Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>
Cc: Dominik Brodowski <linux@dominikbrodowski.net>
Cc: Russell King <rmk@arm.linux.org.uk>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

Use "unsigned long" when dealing with PCI resources.
The BAR Indicator Register (BIR) can be a 64-bit value
or the resource could be a 64-bit host physical address.

Enables ib_mthca and cciss drivers to use MSI-X on ia64 HW.
Problem showed up now because of new system firmware on one platform.
Symptom will either be memory corruption or MCA.

Second part of this patch deals with "useless" code.
We walk through the steps to find the phys_addr and then
don't use the result. I suspect the intent was to zero
out the respective MSI-X entry but I'm not sure at the moment.
Delete the code inside the #if 0/#endif if it's really
not needed.
Signed-off-by: NGrant Grundler <iod00d@hp.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

On Tue, 21 Feb 2006, Ivan Kokshaysky wrote:
> There are two bogus entries in the BIOS memory map table which are
> conflicting with a prefetchable memory range of the AGP bridge:
>
>  BIOS-e820: 00000000fec00000 - 00000000fec01000 (reserved)
>  BIOS-e820: 00000000fee00000 - 00000000fee01000 (reserved)
>
> 0000:00:02.0 PCI bridge: Silicon Integrated Systems [SiS] Virtual PCI-to-PCI bridge (AGP) (prog-if 00 [Normal decode])
> 	Flags: bus master, fast devsel, latency 0
> 	Bus: primary=00, secondary=01, subordinate=01, sec-latency=0
> 	I/O behind bridge: 0000c000-0000cfff
> 	Memory behind bridge: e7e00000-e7efffff
> 	Prefetchable memory behind bridge: fec00000-ffcfffff
> 					   ^^^^^^^^^^^^^^^^^

Yes. However, it's pretty clear that the e820 entries are there for a
reason. Probably they are a hack by the BIOS maintainers to keep Windows
from stomping/moving that region, exactly because they want to keep the
bridge where it is (or, it's actually for the BIOS itself - the BIOS
tables are a horrid mess, and BIOS engineers are pretty hacky people:
they'll add random entries to make their own broken algorithms do the
"right thing").

> Starting from 2.6.13, kernel tries to resolve that sort of conflicts,
> so that prefetch window of the bridge and the framebuffer memory behind
> it get moved to 0x10000000.

I think we could (and probably should) solve this another way: consider
the ACPI "reserved regions" from the e820 map exactly the same way that we
do other ACPI hints - they should restrict _new_ allocations, but not
impact stuff we figure out on our own.

Basically, right now we assign _unassigned_ resources at "fs_initcall"
time. If we were to add in the e820 "reserved region" stuff before that
(but after we've done PCI discovery), we'd probably do the right thing.

Right now we do the e820 reserved regions very early indeed: we call
"register_memory()" from setup_arch(). We could move at least part of it
(the part that registers the resources) down a bit.

Here's a test-patch. I'm not saying we should absolutely do this, but it
might be interesting to try...

Cc: "Antonino A. Daplas" <adaplas@pol.net>
Cc: Ivan Kokshaysky <ink@jurassic.park.msu.ru>
Cc: <bjk@luxsci.net>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

Remove pcie port driver's msi save/restore code.
Signed-off-by: NShaohua Li <shaohua.li@intel.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

It turns out AMD 8131 quirk only affects MSI for devices behind the 8131 bridge.
Handle this by adding a flags field in pci_bus, inherited from parent to child.
Signed-off-by: NMichael S. Tsirkin <mst@mellanox.co.il>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

On ASUS A8V and A8V Deluxe boards, the onboard AC97 audio controller
and MC97 modem controller are deactivated when a second PCI soundcard
is present.  This patch enables them.
Signed-off-by: NBauke Jan Douma <bjdouma@xs4all.nl>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

I attach a trivial patch for 2.6.15.4 that unhides SMBus controller
on an HP Compaq nx6110 notebook.
Signed-off-by: NTomasz Koprowski <tomek@koprowski.org>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

The extra compatability code is not necessary.  Any code still using
the old shutdown method will trigger the warning in driver_register()
instead.
Signed-off-by: NBrian Gerst <bgerst@didntduck.org>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

This patch adapts SHPCHP driver to the PCI device driver model.
Signed-off-by: NKenji Kaneshige <kaneshige.kenji@jp.fujitsu.com>
Signed-off-by: NKristen Carlson Accardi <kristen.c.accardi@intel.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

Current SHPCHP driver doesn't care about the confliction between
hotplug operation via sysfs and hotplug operation via attention
button. So if those ware conflicted, slot could be an unexpected
state.

This patch changes SHPCHP driver to handle slot state properly. With
this patch, slot events are handled according to the current slot
state as shown at the Table below.

		Table. Slot States and Event Handling
=========================================================================
Slot State		Event and Action
=========================================================================
STATIC			- Go to POWERON state if user initiates
(Slot enabled,		  insertion request via sysfs
 Slot disabled)		- Go to POWEROFF state if user initiates removal
			  request via sysfs
			- Go to BLINKINGON state if user presses
			  attention button when the slot is disabled
			- Go to BLINKINGOFF state if user presses
			  attention button when the slot is enabled

The event handler of SHPCHP driver is unnecessarily very complex. In
addition, current event handler can only a fixed number of events at
the same time, and some of events would be lost if several number of
events happened at the same time.

This patch simplify the event handler by using 'work queue', and it
also fix the above-mentioned issue.
Signed-off-by: NKenji Kaneshige <kaneshige.kenji@jp.fujitsu.com>
Signed-off-by: NKristen Carlson Accardi <kristen.c.accardi@intel.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

The wait_for_ctrl_irq() function in SHPCHP driver is no longer needed.
This patch removes that. This patch has no functional change.
Signed-off-by: NKenji Kaneshige <kaneshige.kenji@jp.fujitsu.com>
Signed-off-by: NKristen Carlson Accardi <kristen.c.accardi@intel.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

This patch removes unused 'pci_bus' member from controller structure.
This patch have no functional change.
Signed-off-by: NKenji Kaneshige <kaneshige.kenji@jp.fujitsu.com>
Signed-off-by: NKristen Carlson Accardi <kristen.c.accardi@intel.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

> On Mon, Feb 13, 2006 at 05:13:21PM -0800, David S. Miller wrote:
> >
> > In drivers/pci/probe.c:pci_scan_bridge(), if this is not the first
> > pass (pass != 0) we don't restore the PCI_BRIDGE_CONTROL_REGISTER and
> > thus leave PCI_BRIDGE_CTL_MASTER_ABORT off:
> >
> > int __devinit pci_scan_bridge(struct pci_bus *bus, struct pci_dev * dev, int max, int pass)
> > {
> >  ...
> > 	/* Disable MasterAbortMode during probing to avoid reporting
> > 	   of bus errors (in some architectures) */
> > 	pci_read_config_word(dev, PCI_BRIDGE_CONTROL, &bctl);
> > 	pci_write_config_word(dev, PCI_BRIDGE_CONTROL,
> > 			      bctl & ~PCI_BRIDGE_CTL_MASTER_ABORT);
> >  ...
> > 	if ((buses & 0xffff00) && !pcibios_assign_all_busses() && !is_cardbus) {
> > 		unsigned int cmax, busnr;
> > 		/*
> > 		 * Bus already configured by firmware, process it in the first
> > 		 * pass and just note the configuration.
> > 		 */
> > 		if (pass)
> > 			return max;
> >  ...
> > 	}
> >
> > 	pci_write_config_word(dev, PCI_BRIDGE_CONTROL, bctl);
> >  ...
> >
> > This doesn't seem intentional.

Agreed, looks like an accident.  The patch [1] originally came from Kip
Walker (Broadcom back then) between 2.6.0-test3 and 2.6.0-test4.  As I
recall it was supposed to fix an issue with with PCI aborts being
signalled by the PCI bridge of the Broadcom BCM1250 family of SOCs when
probing behind pci_scan_bridge.  It is undeseriable to disable
PCI_BRIDGE_CTL_MASTER_ABORT in pci_{read,write)_config_* and the
behaviour wasn't considered a bug in need of a workaround, so this was
put in probe.c.

I don't have an affected system at hand, so can't really test but I
propose something like the below patch.

[1] http://www.linux-mips.org/git?p=linux.git;a=commit;h=599457e0cb702a31a3247ea6a5d9c6c99c4cf195

[PCI] Avoid leaving MASTER_ABORT disabled permanently when returning from pci_scan_bridge.
Signed-off-by: NRalf Baechle <ralf@linux-mips.org>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

I moved it to a separate function which is safer.

This avoids problems with the linker reordering them and the
less useful PCI config space access methods taking priority
over the better ones.

Fixes some problems with broken MMCONFIG

Cc: Dave Hansen <haveblue@us.ibm.com>
Signed-off-by: NAndi Kleen <ak@suse.de>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>