Using ACPI to find free address space allows us to find a gap for the
unallocated PCI resources or MMIO resources for hotplug devices within
the BIOS allowed PCI regions.
It works by evaluating the _CRS object under PCI0 looking for producer
resources. Then searches the e820 memory space for a gap within these
producer resources.
Signed-off-by: Alok N Kataria <akataria@vmware.com> Cc: Andi Kleen <ak@linux.intel.com> Cc: Len Brown <lenb@kernel.org> Cc: Ingo Molnar <mingo@elte.hu> Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
Stephen Rothwell [Thu, 10 Jul 2008 00:16:44 +0000 (02:16 +0200)]
PCI: include linux/pm_wakeup.h for device_set_wakeup_capable
drivers/pci/pci.c needs pm_wakeup.h since it uses device_set_wakup_capable().
The latter also needs to be stubbed out for !CONFIG_PM.
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl> Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
The recently introduced pci_prepare_to_sleep() needs the following fix,
because there are systems which are not power manageable by ACPI (ie.
ACPI doesn't provide methods to put the device into low power states and
back), but require ACPI hooks to be executed for wake-up to work.
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl> Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
Matthew Wilcox [Fri, 11 Jul 2008 21:21:17 +0000 (15:21 -0600)]
x86/PCI: Fix PCI config space for domains > 0
John Keller reports that PCI config space access is broken on machines
with more than one domain. conf1 accesses only work for domain 0, so make sure
we check the domain number in the raw routines before trying conf1.
Reported-by: John Keller <jpk@sgi.com> Signed-off-by: Matthew Wilcox <willy@linux.intel.com> Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
If the offset of PCI device's PM capability in its configuration space,
the mask of states that the device supports PME# from and the D1 and D2
support bits are cached in the corresponding struct pci_dev, the PCI
device PM code can be simplified quite a bit.
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl> Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
PCI PM: Introduce pci_prepare_to_sleep and pci_back_from_sleep
Introduce functions pci_prepare_to_sleep() and pci_back_from_sleep(),
to be used by the PCI drivers that want to place their devices into
the lowest power state appropiate for them (PCI_D3hot, if the device
is not supposed to wake up the system, or the deepest state from
which the wake-up is possible, otherwise) while the system is being
prepared to go into a sleeping state and to put them back into D0
during the subsequent transition to the working state.
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl> Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
* Introduce function acpi_pm_device_sleep_wake() for enabling and
disabling the system wake-up capability of devices that are power
manageable by ACPI.
* Introduce function acpi_bus_can_wakeup() allowing other (dependent)
subsystems to check if ACPI is able to enable the system wake-up
capability of given device.
* Introduce callback .sleep_wake() in struct pci_platform_pm_ops and
for the ACPI PCI 'driver' make it use acpi_pm_device_sleep_wake().
* Introduce callback .can_wakeup() in struct pci_platform_pm_ops and
for the ACPI 'driver' make it use acpi_bus_can_wakeup().
* Move the PME# handlig code out of pci_enable_wake() and split it
into two functions, pci_pme_capable() and pci_pme_active(),
allowing the caller to check if given device is capable of
generating PME# from given power state and to enable/disable the
device's PME# functionality, respectively.
* Modify pci_enable_wake() to use the new ACPI callbacks and the new
PME#-related functions.
* Drop the generic .platform_enable_wakeup() callback that is not
used any more.
* Introduce device_set_wakeup_capable() that will set the
power.can_wakeup flag of given device.
* Rework PCI device PM initialization so that, if given device is
capable of generating wake-up events, either natively through the
PME# mechanism, or with the help of the platform, its
power.can_wakeup flag is set and its power.should_wakeup flag is
unset as appropriate.
* Make ACPI set the power.can_wakeup flag for devices found to be
wake-up capable by it.
* Make the ACPI wake-up code enable/disable GPEs for devices that
have the wakeup.flags.prepared flag set (which means that their
wake-up power has been enabled).
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl> Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
Introduce additional flag 'prepared' in struct acpi_device_wakeup_flags
and use it to prevent devices from being enable/disabled do wake up the
system multiple times in a row (this does not happen currently, but will
be possible after some of the following patches).
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl> Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
The currect ACPI code attempts to execute _PSW at three different
places and in one of them only it tries to execute _DSW before _PSW,
which is inconsistent with the other two cases.
Move the execution of _DSW and _PSW into a separate function called
acpi_device_sleep_wake() and call it wherever appropriate instead of
executing _DSW and/or _PSW directly.
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl> Acked-by: Pavel Machek <pavel@suse.cz> Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
PCI: rework pci_set_power_state function to call platform first
Rework pci_set_power_state() so that the platform callback is
invoked before the native mechanism, if necessary. Also, make
the function check if the device is power manageable by the
platform before invoking the platform callback.
This may matter if the device dependent on additional power
resources controlled by the platform is being put into D0, in which
case those power resources must be turned on before we attempt to
handle the device itself.
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl> Acked-by: Pavel Machek <pavel@suse.cz> Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
PCI: Introduce platform_pci_power_manageable function
Introduce function pointer platform_pci_power_manageable to be used
by the platform-related code to point to a function allowing us to
check if given device is power manageable by the platform.
Introduce acpi_pci_power_manageable() playing that role for ACPI.
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl> Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
ACPI: Introduce acpi_bus_power_manageable function
Introduce function acpi_bus_power_manageable() allowing other
(dependent) subsystems to check if ACPI is able to power manage given
device. This may be useful, for example, for PCI device power
management.
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl> Acked-by: Pavel Machek <pavel@suse.cz> Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
Also fixes up the sparc code that was assuming this is not a constant.
Acked-by: David S. Miller <davem@davemloft.net> Signed-off-by: Kay Sievers <kay.sievers@vrfy.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de> Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
Randy Dunlap [Mon, 30 Jun 2008 18:35:53 +0000 (11:35 -0700)]
PCI: add stub for pci_set_consistent_dma_mask()
When CONFIG_PCI=n, there is no stub for pci_set_consistent_dma_mask(),
so add one like other similar stubs. Otherwise there can be build errors,
as here:
linux-next-20080630/drivers/ssb/main.c:1175: error: implicit declaration of
function 'pci_set_consistent_dma_mask'
Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com> Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
Kill pcibios_update_resource(), replacing it with pci_update_resource().
pci_update_resource() uses pcibios_resource_to_bus() to convert a
resource to a device BAR - the transformation should be exactly the
same as the transformation used for the PCI bridges.
pci_update_resource "knows" about 64-bit BARs, but doesn't attempt to
set the high 32-bits to anything non-zero - currently no architecture
attempts to do something different. If anyone cares, please fix; I'm
going to reflect current behaviour for the time being.
Ivan pointed out the following architectures need to examine their
pcibios_update_resource() implementation - they should make sure that
this new implementation does the right thing. #warning's have been
added where appropriate.
ia64
mips
mips64
This cset also includes a fix for the problem reported by AKPM where
64-bit arch compilers complain about the resource mask being placed
in a u32.
<-- snip -->
This patch removes the unused pcibios_update_resource() functions the
kernel gained since, from FRV, m68k, mips & sh architectures.
Signed-off-by: Adrian Bunk <bunk@kernel.org> Acked-by: David Howells <dhowells@redhat.com> Acked-by: Greg Ungerer <gerg@uclinux.org> Acked-by: Paul Mundt <lethal@linux-sh.org> Acked-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
David Howells [Fri, 27 Jun 2008 12:23:20 +0000 (13:23 +0100)]
PCI: fix pci_setup_device()'s sprinting into a const buffer
Make pci_setup_device() write the bus ID directly into the allotted storage,
rather than using pci_name() as the address as that now returns a const
pointer.
Signed-off-by: David Howells <dhowells@redhat.com> Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
Kenji Kaneshige [Thu, 26 Jun 2008 11:06:24 +0000 (20:06 +0900)]
pciehp: use get_service_data
Current pciehp driver saves its private data pointer into pci_dev
structure using pci_set_drvdata()/pci_get_drvdata(). But because
pciehp is not a pci device driver but a PCI Express service driver, it
should save its private data pointer into pcie_device structure using
set_service_data()/get_service_data().
Currently, pciehp driver enables command completed interrupt as follows.
(1) Don't enable at initialization.
(2) Enable command completed interrupt whenever pciehp issues a
command, if the command doesn't attempt to disable the interrupt.
(3) Disable command completed interrupt at driver unloading.
Once we enable command completed interrupt, we don't need to re-enable
it for every command. So we can simplify above steps as follows:
(1) Enable command completed interrupt at initialization.
(2) No special sequence for command completed interrupt.
(3) Disable command completed interrupt at driver unloading.
Kenji Kaneshige [Fri, 20 Jun 2008 03:07:08 +0000 (12:07 +0900)]
pciehp: fix interrupt initialization
Current pciehp driver's intialization sequence is as follows:
(1) initialize controller data structure
(2) install interrupt handler
(3) enable software notification
(4) initialize controller specific slot data structure
(5) initialize generic slot data structure and register it to pci hotplug core
The interrupt handler of pciehp assumes that controller specific slot
data structure is already initialized. However, it is installed at (2)
before initializing controller specific slot data structure at
(4). Because of this, pciehp driver cannot handle the following cases
properly.
- If devices that shares IRQ with pciehp raise interrupts between (2) and (4).
- If hotplug events (e.g. MRL open) happen between (3) and (4).
We already have a workaround for this problem ("pciehp: fix NULL
dereference in interrupt handler: dbd79aed1aea2bece0bf43cc2ff3b2f9baf48a08).
But we still need fundamental fix.
This patch fix the problem by changing the initilization sequence as follows:
(1) initialize controller data structure
(2) initialize controller specific slot data structure
(3) install interrupt handler
(4) enable software notification
(5) initialize generic slot data structure and register it to pci hotplug core
Bjorn Helgaas [Fri, 13 Jun 2008 16:52:11 +0000 (10:52 -0600)]
PCI: use dev_printk when possible
Convert printks to use dev_printk().
I converted pr_debug() to dev_dbg(). Both use KERN_DEBUG and are enabled
only when DEBUG is defined.
I converted printk(KERN_DEBUG) to dev_printk(KERN_DEBUG), not to dev_dbg(),
because dev_dbg() is only enabled when DEBUG is defined.
I converted DBG(KERN_INFO) (only in setup-bus.c) to dev_info(). The DBG()
name makes it sound like debug, but it's been enabled forever, so dev_info()
preserves the previous behavior.
I tried to make the resource assignment formats more consistent, e.g.,
"BAR %d: got res [%#llx-%#llx] bus [%#llx-%#llx] flags %#lx\n"
instead of sometimes using "start-end" and sometimes using "size@start".
I'm not attached to one or the other; I'd just like them consistent.
Jesse Barnes [Mon, 16 Jun 2008 22:29:45 +0000 (15:29 -0700)]
x86/PCI: fixup early quirk probing
On x86, we do early PCI probing to apply some quirks for chipset bugs.
However, in a recent cleanup (7bcbc78dea92fdf0947fa48e248da3c993a5690f) a
thinko was introduced that causes us to probe all subfunctions of even single
function devices (a function was factored out of an inner loop and a "break"
became a "return"). Fix that up by making check_dev_quirk() return a value so
we can keep the factored code intact.
ACPI PM: Add possibility to change suspend sequence
There are some systems out there that don't work correctly with
our current suspend/hibernation code ordering. Provide a workaround
for these systems allowing them to pass 'acpi_sleep=old_ordering' in
the kernel command line so that it will use the pre-ACPI 2.0 ("old")
suspend code ordering.
Unfortunately, this requires us to add a platform hook to the
resuming of devices for recovering the platform in case one of the
device drivers' .suspend() routines returns error code. Namely,
ACPI 1.0 specifies that _PTS should be called before suspending
devices, but _WAK still should be called before resuming them in
order to undo the changes made by _PTS. However, if there is an
error during suspending devices, they are automatically resumed
without returning control to the PM core, so the _WAK has to be
called from within device_resume() in that cases.
The patch also reorders and refactors the ACPI suspend/hibernation
code to avoid duplication as far as reasonably possible.
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl> Acked-by: Pavel Machek <pavel@suse.cz> Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
David Brownell [Wed, 4 Jun 2008 23:15:40 +0000 (01:15 +0200)]
ACPI PM: acpi_pm_device_sleep_state() cleanup
Get rid of a superfluous acpi_pm_device_sleep_state() parameter. The
only legitimate value of that parameter must be derived from the first
parameter, which is what all the callers already do. (However, this
does not address the fact that ACPI still doesn't set up those flags.)
Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Acked-by: Pavel Machek <pavel@ucw.cz> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl> Signed-off-by: Len Brown <len.brown@intel.com>
Alex Chiang [Tue, 10 Jun 2008 21:30:42 +0000 (15:30 -0600)]
PCI: ACPI PCI slot detection driver
Detect all physical PCI slots as described by ACPI, and create entries in
/sys/bus/pci/slots/.
Not all physical slots are hotpluggable, and the acpiphp module does not
detect them. Now we know the physical PCI geography of our system, without
caring about hotplug.
[kaneshige.kenji@jp.fujitsu.com: export-kobject_rename-for-pci_hotplug_core] Signed-off-by: Kenji Kaneshige <kaneshige.kenji@jp.fujitsu.com> Acked-by: Greg KH <greg@kroah.com>
[akpm@linux-foundation.org: build fix]
[akpm@linux-foundation.org: fix build with CONFIG_DMI=n] Signed-off-by: Alex Chiang <achiang@hp.com> Cc: Greg KH <greg@kroah.com> Cc: Kristen Carlson Accardi <kristen.c.accardi@intel.com> Cc: Len Brown <lenb@kernel.org> Acked-by: Len Brown <len.brown@intel.com> Acked-by: Kenji Kaneshige <kaneshige.kenji@jp.fujitsu.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
Alex Chiang [Tue, 10 Jun 2008 21:28:50 +0000 (15:28 -0600)]
PCI: introduce pci_slot
Currently, /sys/bus/pci/slots/ only exposes hotplug attributes when a
hotplug driver is loaded, but PCI slots have attributes such as address,
speed, width, etc. that are not related to hotplug at all.
Introduce pci_slot as the primary data structure and kobject model.
Hotplug attributes described in hotplug_slot become a secondary
structure associated with the pci_slot.
This patch only creates the infrastructure that allows the separation of
PCI slot attributes and hotplug attributes. In this patch, the PCI
hotplug core remains the only user of this infrastructure, and thus,
/sys/bus/pci/slots/ will still only become populated when a hotplug
driver is loaded.
A later patch in this series will add a second user of this new
infrastructure and demonstrate splitting the task of exposing pci_slot
attributes from hotplug_slot attributes.
- Make pci_slot the primary sysfs entity. hotplug_slot becomes a
subsidiary structure.
o pci_create_slot() creates and registers a slot with the PCI core
o pci_slot_add_hotplug() gives it hotplug capability
- Change the prototype of pci_hp_register() to take the bus and
slot number (on parent bus) as parameters.
- Remove all the ->get_address methods since this functionality is
now handled by pci_slot directly.
[achiang@hp.com: rpaphp-correctly-pci_hp_register-for-empty-pci-slots] Tested-by: Badari Pulavarty <pbadari@us.ibm.com> Acked-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
[akpm@linux-foundation.org: build fix]
[akpm@linux-foundation.org: make headers_check happy]
[akpm@linux-foundation.org: nuther build fix]
[akpm@linux-foundation.org: fix typo in #include] Signed-off-by: Alex Chiang <achiang@hp.com> Signed-off-by: Matthew Wilcox <matthew@wil.cx> Cc: Greg KH <greg@kroah.com> Cc: Kristen Carlson Accardi <kristen.c.accardi@intel.com> Cc: Len Brown <lenb@kernel.org> Acked-by: Kenji Kaneshige <kaneshige.kenji@jp.fujitsu.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
Adrian Bunk [Tue, 10 Jun 2008 18:56:38 +0000 (20:56 +0200)]
fix BLK_DEV_HD_ONLY on ARM dependencies
This patch limits BLK_DEV_HD_ONLY to the ARM platforms offering
IRQ_HARDDISK, fixing the following compile error on others:
<-- snip -->
...
CC drivers/ide/legacy/hd.o
...
/home/bunk/linux/kernel-2.6/git/linux-2.6/drivers/ide/legacy/hd.c: In function 'hd_times_out':
/home/bunk/linux/kernel-2.6/git/linux-2.6/drivers/ide/legacy/hd.c:542: error: 'IRQ_HARDDISK' undeclared (first use in this function)
/home/bunk/linux/kernel-2.6/git/linux-2.6/drivers/ide/legacy/hd.c:542: error: (Each undeclared identifier is reported only once
/home/bunk/linux/kernel-2.6/git/linux-2.6/drivers/ide/legacy/hd.c:542: error: for each function it appears in.)
/home/bunk/linux/kernel-2.6/git/linux-2.6/drivers/ide/legacy/hd.c: In function 'do_hd_request':
/home/bunk/linux/kernel-2.6/git/linux-2.6/drivers/ide/legacy/hd.c:661: error: 'IRQ_HARDDISK' undeclared (first use in this function)
/home/bunk/linux/kernel-2.6/git/linux-2.6/drivers/ide/legacy/hd.c: In function 'hd_init':
/home/bunk/linux/kernel-2.6/git/linux-2.6/drivers/ide/legacy/hd.c:765: error: 'IRQ_HARDDISK' undeclared (first use in this function)
make[3]: *** [drivers/ide/legacy/hd.o] Error 1
<-- snip -->
Reported-by: Adrian Bunk <bunk@kernel.org> Cc: Russell King <rmk@arm.linux.org.uk> Signed-off-by: Adrian Bunk <bunk@kernel.org> Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
Convert the driver to use struct ide_port_info - as a nice side-effect
this fixes racy setup of ->io_32bit/unmask settings (after ide_device_add()
call device can be already in use).
Alex Chiang [Wed, 4 Jun 2008 17:39:07 +0000 (11:39 -0600)]
PCIe: fix 'symbol not declared' sparse warnings
While refreshing my physical PCI slot series against upstream, I
noticed a few simple sparse/compile warnings that were easy to
fix.
Fix the following sparse warnings in PCIe:
drivers/pci/pcie/aer/aerdrv.c:86:6: warning: symbol 'pci_no_aer'
was not declared. Should it be static?
drivers/pci/pcie/portdrv_bus.c:21:17: warning: symbol
'pcie_port_bus_type' was not declared. Should it be static?
Signed-off-by: Alex Chiang <achiang@hp.com> Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
Jean Delvare [Wed, 4 Jun 2008 11:53:31 +0000 (13:53 +0200)]
PCI: Unhide the SMBus on the Compaq Evo D510
One more machine with a hidden Intel SMBus. Unhiding it reveals a SMSC
EMC6D100 hardware monitoring chip. I have checked that this machine
has no ACPI magic touching the SMBus nor the hardware monitoring chip,
so this should be safe.
Signed-off-by: Jean Delvare <khali@linux-fr.org> Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
Kenji Kaneshige [Wed, 28 May 2008 06:01:03 +0000 (15:01 +0900)]
shpchp: check firmware before taking control
Fix the following problems of shpchp driver about getting hotplug
control from firmware.
- The shpchp driver must not control the hotplug controller if it
fails to get control from the firmware. But current shpchp
controls the hotplug controller regardless the result, because it
doesn't check the return value of get_hp_hw_control_from_firmware().
- Current shpchp driver doesn't support _OSC.
The pciehp driver already have the code for evaluating _OSC and OSHP
and shpchp and pciehp can share it. So this patch move that code from
pciehp to acpi_pcihp.c.
Kenji Kaneshige [Wed, 28 May 2008 05:57:30 +0000 (14:57 +0900)]
pciehp: evaluate _OSC/OSHP before controller init
Current pciehp evaluates _OSC/OSHP method after some controller
initialization is done. So if evaluating _OSC/OSHP is failed, we need
to cleanup already initialized data structures or hardware. This
clearly is not robust way. With this patch, _OSC/OSHP evaluation is
done first.
Introduce new top level suspend and hibernation callbacks
Introduce 'struct pm_ops' and 'struct pm_ext_ops' ('ext' meaning
'extended') representing suspend and hibernation operations for bus
types, device classes, device types and device drivers.
Modify the PM core to use 'struct pm_ops' and 'struct pm_ext_ops'
objects, if defined, instead of the ->suspend(), ->resume(),
->suspend_late(), and ->resume_early() callbacks (the old callbacks
will be considered as legacy and gradually phased out).
The main purpose of doing this is to separate suspend (aka S2RAM and
standby) callbacks from hibernation callbacks in such a way that the
new callbacks won't take arguments and the semantics of each of them
will be clearly specified. This has been requested for multiple
times by many people, including Linus himself, and the reason is that
within the current scheme if ->resume() is called, for example, it's
difficult to say why it's been called (ie. is it a resume from RAM or
from hibernation or a suspend/hibernation failure etc.?).
The second purpose is to make the suspend/hibernation callbacks more
flexible so that device drivers can handle more than they can within
the current scheme. For example, some drivers may need to prevent
new children of the device from being registered before their
->suspend() callbacks are executed or they may want to carry out some
operations requiring the availability of some other devices, not
directly bound via the parent-child relationship, in order to prepare
for the execution of ->suspend(), etc.
Ultimately, we'd like to stop using the freezing of tasks for suspend
and therefore the drivers' suspend/hibernation code will have to take
care of the handling of the user space during suspend/hibernation.
That, in turn, would be difficult within the current scheme, without
the new ->prepare() and ->complete() callbacks.
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl> Acked-by: Pavel Machek <pavel@ucw.cz> Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
Gary Hade [Mon, 12 May 2008 20:57:46 +0000 (13:57 -0700)]
PCI: boot parameter to avoid expansion ROM memory allocation
Contention for scarce PCI memory resources has been growing
due to an increasing number of PCI slots in large multi-node
systems. The kernel currently attempts by default to
allocate memory for all PCI expansion ROMs so there has
also been an increasing number of PCI memory allocation
failures seen on these systems. This occurs because the
BIOS either (1) provides insufficient PCI memory resource
for all the expansion ROMs or (2) provides adequate PCI
memory resource for expansion ROMs but provides the
space in kernel unexpected BIOS assigned P2P non-prefetch
windows.
The resulting PCI memory allocation failures may be benign
when related to memory requests for expansion ROMs themselves
but in some cases they can occur when attempting to allocate
space for more critical BARs. This can happen when a successful
expansion ROM allocation request consumes memory resource
that was intended for a non-ROM BAR. We have seen this
happen during PCI hotplug of an adapter that contains a
P2P bridge where successful memory allocation for an
expansion ROM BAR on device behind the bridge consumed
memory that was intended for a non-ROM BAR on the P2P bridge.
In all cases the allocation failure messages can be very
confusing for users.
This patch provides a new 'pci=norom' kernel boot parameter
that can be used to disable the default PCI expansion ROM memory
resource allocation. This provides a way to avoid the above
described issues on systems that do not contain PCI devices
for which drivers or user-level applications depend on the
default PCI expansion ROM memory resource allocation behavior.
Signed-off-by: Gary Hade <garyhade@us.ibm.com> Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
Kenji Kaneshige [Thu, 15 May 2008 06:21:16 +0000 (15:21 +0900)]
pci-acpi: add flag to indicate query had been done
Current pci-acpi implementation checks osc_data->support_stat to see
if control bits had been already queried. It is not good from the
viewpoint of easy understanding. So this patch adds new 'is_queried'
flag to indicate query had been done.
Kenji Kaneshige [Thu, 15 May 2008 06:20:11 +0000 (15:20 +0900)]
pci-acpi: use local buffer for _OSC
Current pci-acpi implementation uses array in osc_data directly to
evaluate _OSC. It needs to save the old data and restore it if _OSC
evaluation fails. To make it more robust, we should use local array to
evaluate _OSC.
Zhao Yakui [Tue, 13 May 2008 03:15:05 +0000 (11:15 +0800)]
PCI: Disable PME during PCI scan
If a device supports #PME and can generate PME events from D0, we may see
superfluous events before a driver is loaded (drivers should only enable PME as
needed), preventing suspend from working if the corresponding GPE was enabled.
Likewise, if the ACPI device has the _PRW object, the _PSW/_DSW object will be
called in order to disable the wakeup functionality. But when it is allowed to
wake up the sleeping state, OSPM will enable it again.
So we should disable PME in the course of scanning PCI devices and enable it
again only when PME events are actually required to be generated from the
requested PCI state (for example, D3_hot or D3_cold). It is also safe to
disable PME again when the PME is disabled for the PCI devices.
Some quirks should be called with interrupt disabled, we can't directly
call them in .resume_early. Also the patch introduces
pci_fixup_resume_early and pci_fixup_suspend, which matches current
device core callbacks (.suspend/.resume_early).
TBD: Somebody knows why we need quirk resume should double check if a
quirk should be called in resume or resume_early. I changed some per my
understanding, but can't make sure I fixed all.
Signed-off-by: Shaohua Li <shaohua.li@intel.com> Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
Julia Lawall [Tue, 10 Jun 2008 08:03:25 +0000 (10:03 +0200)]
[S390] tape_3590.c: introduce missing kfree
The semantic match that finds the problem is as follows:
(http://www.emn.fr/x-info/coccinelle/)
@r exists@
expression E,E1;
statement S;
position p1,p2,p3;
@@
E =@p1 \(kmalloc\|kcalloc\|kzalloc\)(...)
... when != E = E1
if (E == NULL || ...) S
... when != E = E1
if@p2 (...) {
... when != kfree(E)
}
... when != E = E1
kfree@p3(E);
@forall@
position r.p2;
expression r.E;
int E1 != 0;
@@
* if@p2 (...) {
... when != kfree(E)
when strict
return E1; }
Signed-off-by: Julia Lawall <julia@diku.dk> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
__ctl_load/__ctl_store are called with either an array of unsigned long or
a single unsigned long value. Add an address operator to the "m"/"=m"
contraints to make them work for unsigned long arguments as well.
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
The first argument to __ctl_store() should be the array to store
stuff in, not just the first element of that array. With the
current code in __cpu_up(), mainline GCC dies with an internal
compiler error. I didn't diagnose that further, but just fixed
the kernel bug.
The issue is caused by a list_empty() check in __sclp_vt220_cleanup, which
usually fails on non-initialized list heads that contain {NULL,NULL} instead.
Signed-off-by: Carsten Otte <cotte@de.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cornelia Huck [Tue, 10 Jun 2008 08:03:21 +0000 (10:03 +0200)]
[S390] cio: Fix inverted isc priorities.
Priorities for I/O interruption subclasses range from 0 (highest)
to 7 (lowest). Unfortunately, the console has been using isc 7
instead of an isc with a higher priority than regular I/O
subchannels (which use 3). Fix this by making the console use
isc 1.
Signed-off-by: Cornelia Huck <cornelia.huck@de.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Heiko Carstens [Tue, 10 Jun 2008 08:03:20 +0000 (10:03 +0200)]
[S390] vmemmap: fix off-by-one bug.
If a memory range is supposed to be added to the 1:1 mapping and it
ends just below the maximum supported physical address it won't
succeed. This is because a test doesn't consider that the end address
is 1 smaller than start + size.
Fix the comparison.
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cornelia Huck [Tue, 10 Jun 2008 08:03:19 +0000 (10:03 +0200)]
[S390] cio: Fix sparse warnings in blacklist.c.
sparse complains about signedness:
drivers/s390/cio/blacklist.c:132:28: warning: incorrect type in
argument 2 (different signedness)
drivers/s390/cio/blacklist.c:132:28: expected unsigned int *val
drivers/s390/cio/blacklist.c:132:28: got int *cssid
drivers/s390/cio/blacklist.c:136:28: warning: incorrect type in
argument 2 (different signedness)
drivers/s390/cio/blacklist.c:136:28: expected unsigned int *val
drivers/s390/cio/blacklist.c:136:28: got int *ssid
drivers/s390/cio/blacklist.c:140:28: warning: incorrect type in
argument 2 (different signedness)
drivers/s390/cio/blacklist.c:140:28: expected unsigned int *val
drivers/s390/cio/blacklist.c:140:28: got int *devno
cssid, ssid and devno are of course unsigned, so let's make the
variables unsigned as well.
Signed-off-by: Cornelia Huck <cornelia.huck@de.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Heiko Carstens [Tue, 10 Jun 2008 08:03:18 +0000 (10:03 +0200)]
[S390] sparsemem: use SPARSEMEM_STATIC if !64BIT.
In case of !64BIT kernel we end up with a zero sized mem_section array.
This happens because NR_MEM_SECTIONS is smaller than SECTIONS_PER_ROOT
but we have:
Linus Torvalds [Tue, 10 Jun 2008 02:28:54 +0000 (19:28 -0700)]
Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband:
IB/core: Remove IB_DEVICE_SEND_W_INV capability flag
IB/umem: Avoid sign problems when demoting npages to integer
Linus Torvalds [Tue, 10 Jun 2008 02:27:52 +0000 (19:27 -0700)]
Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/cooloney/blackfin-2.6
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/cooloney/blackfin-2.6:
Blackfin serial driver: fix up tty core set_ldisc API change breakage bug
Blackfin arch: protect only the SPI bus controller with CONFIG_SPI_BFIN
Blackfin arch: fixup warnings with the new cplb saved values
Blackfin Serial Driver: Clean up BF54x macro in blackfin UART driver.
Commit 54d29ad33e3483bcc7ca433a21cf294854e5154a (Power Supply: fix race
in device_create) introduced a race in power_supply_uevent. Previously it
checked that power_supply is available by checking for dev->driver_data.
But now dev->driver_data is set before power_supply->dev is initialised.
Signed-off-by: Dmitry Baryshkov <dbaryshkov@gmail.com> Signed-off-by: Anton Vorontsov <cbouatmailru@gmail.com>
projects / linux-2.6 / log