The PNP system board driver (drivers/pnp/system.c) contains all the
same functionality, so we don't need the ACPI version.
Previously, a motherboard device would be claimed by *both* the ACPI and
PNP drivers, resulting in stuff like this in /proc/ioports:
1200-121f : motherboard <-- from drivers/acpi/motherboard.c
1200-121f : pnp 00:0d <-- from drivers/pnp/system.c
Make sure to enable CONFIG_PNP (and CONFIG_PNPACPI) to include the
PNP system board driver.
Signed-off-by: Bjorn Helgaas <bjorn.helgaas@hp.com>
Signed-off-by: Len Brown <len.brown@intel.com>
Resources described by the FADT aren't really a good fit for the
ACPI motherboard driver.
The motherboard driver cares about PNP0C01 and PNP0C02 devices and
their resources.
The FADT describes some resources used by the ACPI core. Often, they
are also described by by the _CRS of a motherboard device, but I think
it's better to reserve them specifically in the ACPI osl.c because
(a) the motherboard driver is optional and ACPI uses the resources even
if the driver is absent, and (b) I want to remove the ACPI motherboard
driver because it's mostly redundant with the PNP system.c driver.
Signed-off-by: Bjorn Helgaas <bjorn.helgaas@hp.com>
Signed-off-by: Len Brown <len.brown@intel.com>
Seems to be some left-over debug code.
Cc: Len Brown <lenb@kernel.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Move acpi_get_pci_rootbridge_handle() from glue.c to pci_root.c and get the
root bridge ACPI handles by searching the &acpi_pci_roots list instead of
walking through the ACPI name space. This significantly reduces boot time
on large I/O systems.
Signed-off-by: Justin Chen <justin.chen@hp.com>
Signed-off-by: Bjorn Helgaas <bjorn.helgaas@hp.com>
Signed-off-by: Len Brown <len.brown@intel.com>
This reverts commit 2df910b4c3.
ACPI_BAY has not been merged into mainline yet, so the changes to ibm-acpi
related Kconfig entries that depend on ACPI_BAY were permanently disabling
ibm-acpi bay support. This is a serious regression for ThinkPad users.
Signed-off-by: Henrique de Moraes Holschuh <hmh@hmh.eng.br>
Signed-off-by: Len Brown <len.brown@intel.com>
write_lcd() in toshiba_acpi returns 0 on success since the big ACPI patch
merged in 2.6.20-rc2. It should return count.
Signed-off-by: Matthijs van Otterdijk <thotter@gmail.com>
Cc: Len Brown <lenb@kernel.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Some of the ACPI devices use the internal fake hids
which are exposed to userspace as devces' bus_id after sysfs conversion.
To make it more friendly, we convert them to more understandable strings.
For those devices w/o PNPids, we use "device:instance_no" as the bus_id
instead of "PNPIDNON:instance_no".
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>
Provide ACPI _PRT support for SN Altix systems.
The SN Altix platform does not conform to the
IOSAPIC IRQ routing model, so a new acpi_irq_model
(ACPI_IRQ_MODEL_PLATFORM) has been defined. The SN
platform specific code sets acpi_irq_model to
this new value, and keys off of it in acpi_register_gsi()
to avoid the iosapic code path.
Signed-off-by: John Keller <jpk@sgi.com>
Signed-off-by: Len Brown <len.brown@intel.com>
The recent EC cleanup left a printk enabled on handler evaluation
resulting in a bunch of messages on normal operation, like so:
ACPI: EC: evaluating _Q60
Signed-off-by: Len Brown <len.brown@intel.com>
* 'release' of git://git.kernel.org/pub/scm/linux/kernel/git/lenb/linux-acpi-2.6: (68 commits)
ACPI: replace kmalloc+memset with kzalloc
ACPI: Add support for acpi_load_table/acpi_unload_table_id
fbdev: update after backlight argument change
ACPI: video: Add dev argument for backlight_device_register
ACPI: Implement acpi_video_get_next_level()
ACPI: Kconfig - depend on PM rather than selecting it
ACPI: fix NULL check in drivers/acpi/osl.c
ACPI: make drivers/acpi/ec.c:ec_ecdt static
ACPI: prevent processor module from loading on failures
ACPI: fix single linked list manipulation
ACPI: ibm_acpi: allow clean removal
ACPI: fix git automerge failure
ACPI: ibm_acpi: respond to workqueue update
ACPI: dock: add uevent to indicate change in device status
ACPI: ec: Lindent once again
ACPI: ec: Change #define to enums there possible.
ACPI: ec: Style changes.
ACPI: ec: Acquire Global Lock under EC mutex.
ACPI: ec: Drop udelay() from poll mode. Loop by reading status field instead.
ACPI: ec: Rename gpe_bit to gpe
...
Fernando Lopez-Lezcano reported frequent scheduling latencies and audio
xruns starting at the 2.6.18-rt kernel, and those problems persisted all
until current -rt kernels. The latencies were serious and unjustified by
system load, often in the milliseconds range.
After a patient and heroic multi-month effort of Fernando, where he
tested dozens of kernels, tried various configs, boot options,
test-patches of mine and provided latency traces of those incidents, the
following 'smoking gun' trace was captured by him:
_------=> CPU#
/ _-----=> irqs-off
| / _----=> need-resched
|| / _---=> hardirq/softirq
||| / _--=> preempt-depth
|||| /
||||| delay
cmd pid ||||| time | caller
\ / ||||| \ | /
IRQ_19-1479 1D..1 0us : __trace_start_sched_wakeup (try_to_wake_up)
IRQ_19-1479 1D..1 0us : __trace_start_sched_wakeup <<...>-5856> (37 0)
IRQ_19-1479 1D..1 0us : __trace_start_sched_wakeup (c01262ba 0 0)
IRQ_19-1479 1D..1 0us : resched_task (try_to_wake_up)
IRQ_19-1479 1D..1 0us : __spin_unlock_irqrestore (try_to_wake_up)
...
<idle>-0 1...1 11us!: default_idle (cpu_idle)
...
<idle>-0 0Dn.1 602us : smp_apic_timer_interrupt (c0103baf 1 0)
...
<...>-5856 0D..2 618us : __switch_to (__schedule)
<...>-5856 0D..2 618us : __schedule <<idle>-0> (20 162)
<...>-5856 0D..2 619us : __spin_unlock_irq (__schedule)
<...>-5856 0...1 619us : trace_stop_sched_switched (__schedule)
<...>-5856 0D..1 619us : trace_stop_sched_switched <<...>-5856> (37 0)
what is visible in this trace is that CPU#1 ran try_to_wake_up() for
PID:5856, it placed PID:5856 on CPU#0's runqueue and ran resched_task()
for CPU#0. But it decided to not send an IPI that no CPU - due to
TS_POLLING. But CPU#0 never woke up after its NEED_RESCHED bit was set,
and only rescheduled to PID:5856 upon the next lapic timer IRQ. The
result was a 600+ usecs latency and a missed wakeup!
the bug turned out to be an idle-wakeup bug introduced into the mainline
kernel this summer via an optimization in the x86_64 tree:
commit 495ab9c045
Author: Andi Kleen <ak@suse.de>
Date: Mon Jun 26 13:59:11 2006 +0200
[PATCH] i386/x86-64/ia64: Move polling flag into thread_info_status
During some profiling I noticed that default_idle causes a lot of
memory traffic. I think that is caused by the atomic operations
to clear/set the polling flag in thread_info. There is actually
no reason to make this atomic - only the idle thread does it
to itself, other CPUs only read it. So I moved it into ti->status.
the problem is this type of change:
if (!hlt_counter && boot_cpu_data.hlt_works_ok) {
- clear_thread_flag(TIF_POLLING_NRFLAG);
+ current_thread_info()->status &= ~TS_POLLING;
smp_mb__after_clear_bit();
while (!need_resched()) {
local_irq_disable();
this changes clear_thread_flag() to an explicit clearing of TS_POLLING.
clear_thread_flag() is defined as:
clear_bit(flag, &ti->flags);
and clear_bit() is a LOCK-ed atomic instruction on all x86 platforms:
static inline void clear_bit(int nr, volatile unsigned long * addr)
{
__asm__ __volatile__( LOCK_PREFIX
"btrl %1,%0"
hence smp_mb__after_clear_bit() is defined as a simple compile barrier:
#define smp_mb__after_clear_bit() barrier()
but the explicit TS_POLLING clearing introduced by the patch:
+ current_thread_info()->status &= ~TS_POLLING;
is not an atomic op! So the clearing of the TS_POLLING bit is freely
reorderable with the reading of the NEED_RESCHED bit - and both now
reside in different memory addresses.
CPU idle wakeup very much depends on ordered memory ops, the clearing of
the TS_POLLING flag must always be done before we test need_resched()
and hit the idle instruction(s). [Symmetrically, the wakeup code needs
to set NEED_RESCHED before it tests the TS_POLLING flag, so memory
ordering is paramount.]
Fernando's dual-core Athlon64 system has a sufficiently advanced memory
ordering model so that it triggered this scenario very often.
( And it also turned out that the reason why these latencies never
triggered on my testsystems is that i routinely use idle=poll, which
was the only idle variant not affected by this bug. )
The fix is to change the smp_mb__after_clear_bit() to an smp_mb(), to
act as an absolute barrier between the TS_POLLING write and the
NEED_RESCHED read. This affects almost all idling methods (default,
ACPI, APM), on all 3 x86 architectures: i386, x86_64, ia64.
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Tested-by: Fernando Lopez-Lezcano <nando@ccrma.Stanford.EDU>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
http://bugzilla.kernel.org/show_bug.cgi?id=7695
Originally we converted bind/unbind to use a new pci bridge driver.
The driver will add/remove _PRT, so we can eventually remove
.bind/.unbind methods.
But we found that some of the _ADR-Based devices don't have _PRT,
i.e. they are not managed by the new ACPI PCI bridge driver.
So that .bind method is not called for some _ADR-Based devices,
which leads to a failure.
Now we make ACPI PCI Root Bridge Driver scan and binds all _ADR-Based devices
once the driver is loaded, in the .add method of ACPI PCI Root Bridge driver.
Extra code path for calling .bind/.unbind when _ADR-Based devices
are hot added/removed is also added.
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>
Previously we used the device name in the DSDT, but would
crash upon encountering a duplicate. Also, exposing the
DSDT device name to the user in a patch isn't a good idea,
because it is arbitrary.
After some discussion, we finally decided to use
"PNPID:instance_no" as the bus_id of ACPI devices.
Two attributes for each device are added at the same time,
the full pathname in ACPI namespace and hardware_id if it has.
NOTE: acpi_bus_id_list is used to keep the information of PNPID
and instance number of the given PNPID. Loop the
acpi_bus_id_list to find the instance_no of the same PNPID
when register a device. If failed, i.e. we don't have a
node with this PNPID, allocate one and link it to this list.
NOTE: Now I don't take the memory free work in charge.
If necessary, I can add a reference count in
struct acpi_device_bus_id, and check the reference and
when unregister a device, i.e. memory is freed when
the reference count of a given PNPID is 0.
Signed-off-by: Li Shaohua <shaohua.li@intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>
Add a prototype for driver_init() in include/linux/device.h.
Also remove a static function of the same name in drivers/acpi/ibm_acpi.c to
ibm_acpi_driver_init() to fix the namespace collision.
Signed-off-by: Adrian Bunk <bunk@stusta.de>
Acked-by: Henrique de Moraes Holschuh <hmh@hmh.eng.br>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Make acpi_load_table() available for use by removing it from the #ifdef
ACPI_FUTURE_USAGE.
Also add a new routine used to unload an ACPI table of a given type and "id" -
acpi_unload_table_id(). The implementation of this new routine was almost a
direct copy of existing routine acpi_unload_table() - only difference being
that it only removes a specific table id instead of ALL tables of a given
type. The SN hotplug driver (sgi_hotplug.c) now uses both of these interfaces
to dynamically load and unload SSDT ACPI tables.
Also, a few other ACPI routines now used by the SN hotplug driver are exported
(since the driver can be a loadable module):
acpi_ns_map_handle_to_node
acpi_ns_convert_entry_to_handle
acpi_ns_get_next_node
Signed-off-by: Aaron Young <ayoung@sgi.com>
Cc: Greg KH <greg@kroah.com>
Cc: "Luck, Tony" <tony.luck@intel.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Len Brown <len.brown@intel.com>
This patch set adds generic abstract layer support for acpi video driver to
have generic user interface to control backlight and output switch control by
leveraging the existing backlight sysfs class driver, and by adding a new
video output sysfs class driver.
This patch:
Add dev argument for backlight_device_register to link the class device to
real device object. The platform specific driver should find a way to get the
real device object for their video device.
[akpm@osdl.org: build fix]
[akpm@osdl.org: fix msi-laptop.c]
Signed-off-by: Luming Yu <Luming.yu@intel.com>
Cc: "Antonino A. Daplas" <adaplas@pol.net>
Cc: Greg KH <greg@kroah.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Len Brown <len.brown@intel.com>
acpi_video_get_next_level was supposed to implement an algorithm to select
a new brightness level based on the old brightness level of an ACPI video
device, but it simply says "/* Fix me */" and returns the current
brightness.
This patch implements acpi_video_get_next_level properly. It had to change
a few constants at the top of the file because they were (apparently)
wrong, but it appears to work on my Dell Inspiron e1405 (with BIOS A05
only--BIOS A04 doesn't seem to send ACPI video hotkey events).
[akpm@osdl.org: cleanups]
Signed-off-by: Thomas Tuttle <linux-kernel@ttuttle.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Len Brown <len.brown@intel.com>
Make ACPI depend on PM rather than selecting it.
Otherwise it's a nightmare working out why CONFIG_PM keeps getting set.
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Len Brown <len.brown@intel.com>
Spotted by the Coverity checker.
Signed-off-by: Adrian Bunk <bunk@stusta.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Len Brown <len.brown@intel.com>
Make loading processor.ko fail when an error happens.
Signed-off-by: Akinobu Mita <akinobu.mita@gmail.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Len Brown <len.brown@intel.com>
Fix single linked list manipulation for sub_driver. If the remving entry
is not on the head of the sub_driver list, it goes into infinate loop.
Though that infinite loop doesn't happen. Because the only user of
acpi_pci_register_dirver() is acpiphp.
Signed-off-by: Akinobu Mita <akinobu.mita@gmail.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Len Brown <len.brown@intel.com>
Allow clean removal by setting notify_installed in the right place.
Signed-off-by: Alexey Starikovskiy <alexey.y.starikovskiy@intel.com>
Signed-off-by: Henrique de Moraes Holschuh <hmh@hmh.eng.br>
Signed-off-by: Len Brown <len.brown@intel.com>
We do this mainly because:
1. hid is used to match ACPI devices and drivers.
.match method which is incompatible to driver model
can be deleted from acpi_driver.ops then.
2. As the .uevent method mark ACPI drivers by PNPID,
fake hid is set to non-PNPID devices so that udev script
can load the right ACPI driver by looking for
"HWID = " or "COMPTID = ".
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>
acpi_device had a .bind/.unbind methods, but Linux driver model does not.
Cut ACPI PCI code over to use the Linux driver model methods.
Convert bind/unbind to use a new pci bridge driver.
The driver will add/remove _PRT, so we can eventually
remove .bind/.unbind methods.
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>
Consolidate motherboard1 and motherboard2 drivers into one
so that driver core doesn't complain that two drivers have
the same name.
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>
Add removal_type in structure acpi_device for hot removal.
ACPI_BUS_REMOVAL_EJECT is used for ACPI device hot removal.
Only one parameter is allowed in .remove method due to driver model.
So removal_type is added to indicate different removal type.
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>
Add acpi_bus_ops in acpi_device to support acpi hot plug.
NOTE: Two methods .add and .start in acpi_driver.ops are
called separately to probe ACPI devices, while only
.probe method is called in driver model.
As executing .add and .start separately is critical
for ACPI device hot plug, we use acpi_bus_ops to
distinguish different code path.
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>
Setup new sysfs framework
1. Remove /sys/firmware/acpi
2. Add ACPI device in device tree.
File "eject" for every device that has _EJ0 method is moved from
/sys/firmware to /sys/devices.
Operation on this file is exactly the same as before.
i.e. echo 1 to "eject" will cause hot removal of this device.
Corresponding changes should be made in userspace for hot removal.
Signed-off-by: Li Shaohua <shaohua.li@intel.com>
Signed-off-by: Zhang Rui<rui.zhang@intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>
Adjust link order to add ACPI devices to global list before PCI devices.
In addition, acpi_bus type must be initialized before any driver loads.
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>
ACPI device/driver registration Interfaces are modified
to follow Linux driver model.
Signed-off-by: Li Shaohua <shaohua.li@intel.com>
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>
Add ACPI bus_type for Linux driver model.
1. .shutdown method is added into acpi_driver.ops
needed by bus_type operations.
2. remove useless parameter 'int state' in .resume method.
3. change parameter 'int state'
to 'pm_message_t state' in .suspend method.
Note: The new .uevent method mark ACPI drivers by PNPID instead of by name.
Udev script needs to look for "HWID=" or "COMPTID=" to load
ACPI drivers as a result.
Signed-off-by: Li Shaohua <shaohua.li@intel.com>
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>
Add device_driver into acpi_driver for driver model.
Add helper functions 'to_acpi_device' and 'to_acpi_driver'
to get structure acpi_device/acpi_driver by device/device_driver.
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>
We want the name 'to_acpi_device'.
And the current macro 'to_acpi_device' will be removed
after device model is setup.
So just simply rename them.
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>
Adjust the code and make code doing similar things together.
No logic changes.
Signed-off-by : Zhang Rui <rui.zhang@intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>
Send a uevent to indicate a device change whenever we dock or
undock, so that userspace may now check the dock status via sysfs.
Signed-off-by: Kristen Carlson Accardi <kristen.c.accardi@intel.com>
Signed-off-by: Holger Macht <hmacht@suse.de>
Signed-off-by: Len Brown <len.brown@intel.com>
Fix bug which will cause acpiphp to not be able to load when dock.ko
cannot load.
Signed-off-by: Prarit Bhargava <prarit@redhat.com>
Signed-off-by: Kristen Carlson Accardi <kristen.c.accardi@intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>
Add 2 sysfs files for user interface.
1) docked - 1/0 (read only) - indicates whether the software believes the
laptop is docked in a docking station.
2) undock - (write only) - writing to this file causes the software to
initiate an undock request to the firmware.
Signed-off-by: Brandon Philips <brandon@ifup.org>
Signed-off-by: Kristen Carlson Accardi <kristen.c.accardi@intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>
Make the dock station driver a platform device driver so that
we can create sysfs entries under /sys/device/platform.
Signed-off-by: Kristen Carlson Accardi <kristen.c.accardi@intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>
This patch cleans up the recently added backlight device support by Holger
Macht <hmacht@suse.de> to fit well with the rest of the code, using the
ibms struct as the other "subdrivers" in ibm-acpi.
Signed-off-by: Henrique de Moraes Holschuh <hmh@hmh.eng.br>
This patch makes it possible to disable ibm-acpi non-generic bay support,
as generic bay support already works well for a number of ThinkPads.
Signed-off-by: Henrique de Moraes Holschuh <hmh@hmh.eng.br>
This patch adds support for the ultrabay on the T60, X60 and other new
ThinkPads that have a SATA ultrabay.
I intend to keep bay and dock support in ibm-acpi working and updated until
it finally gets deprecated and removed in favour of the generic dock and
bay support. But we aren't there yet.
Signed-off-by: Henrique de Moraes Holschuh <hmh@hmh.eng.br>
This patch implements a fan control safety watchdog, by request of the
authors of userspace fan control scripts.
When the watchdog timer expires, the equivalent action of a "fan enable"
command is executed. The watchdog timer is reset at every reception of a
fan control command that could change the state of the fan itself.
This command is meant to be used by userspace fan control daemons, to make
sure the fan is never left set to an unsafe level because of userspace
problems.
Users of the X31/X40/X41 "speed" command are on their own, the current
implementation of "speed" is just too incomplete to be used safely,
anyway. Better to never use it, and just use the "level" command instead.
The watchdog is programmed using echo "watchdog <number>" > fan, where
number is the number of seconds to wait before doing an "enable", and zero
disables the watchdog.
Signed-off-by: Henrique de Moraes Holschuh <hmh@hmh.eng.br>
A few ThinkPads fail to initialize EC register 0x2f both in the EC
firmware and ACPI DSDT. If the BIOS and the ACPI DSDT also do not
initialize it, then the initial status of that register does not
correspond to reality.
On all reported buggy machines, EC 0x2f will read 0x07 (fan level 7) upon
cold boot, when the EC is actually in mode 0x80 (auto mode). Since
returning a text string ("unknown") would break a number of userspace
programs, instead we correct the reading for the most probably correct
answer, and return it is in auto mode.
The workaround flags the status and level as unknown on module load/kernel
boot, until we are certain at least one fan control command was issued,
either by us, or by something else.
We don't work around the bug by doing a "fan enable" at module
load/startup (which would initialize the EC register) because it is not
known if these ThinkPad ACPI DSDT might have set the fan to level 7
instead of "auto" (we don't know if they can do this or not) due to a
thermal condition, and we don't want to override that, should they be
capable of it.
We should be setting the workaround flag to "status known" upon resume, as
both reports and a exaustive search on the DSDT tables at acpi.sf.net show
that the DSDTs always enable the fan on resume, thus working around the
bug. But since we don't have suspend/resume handlers in ibm-acpi yet and
the "EC register 0x2f was modified" logic is likely to catch the change
anyway, we don't.
Signed-off-by: Henrique de Moraes Holschuh <hmh@hmh.eng.br>
This patch changes the ThinkPad Embedded Controller DMI matching
code to store the firmware version of the EC for later usage, e.g.
for quirks.
It also prints the firmware version when starting up.
Signed-off-by: Henrique de Moraes Holschuh <hmh@hmh.eng.br>
This patch extend fan control functions, implementing enable/disable for
all write access modes, implementing level control for all level-capable
write access modes.
The patch also updates the documentation, explaining levels auto and
disengaged.
ABI changes:
1. Support level 0 as an equivalent to disable
2. Add support for level auto and level disengaged when doing
EC 0x2f fan control
3. Support enable/disable for all level-based write access modes
4. Add support for level command on FANS thinkpads, as per
thinkwiki reports
Signed-off-by: Henrique de Moraes Holschuh <hmh@hmh.eng.br>
This patch fix fan enable to attempt to do the right thing and not slow
down the fan if it is forced to the maximum speed. It also extends fan
enable to work on older thinkpads.
ABI changes:
1. Support enable/disable for all level-based write access modes
Signed-off-by: Henrique de Moraes Holschuh <hmh@hmh.eng.br>
This patch fixes fan_read to return correct values for all fan access
modes. It also implements some fan access mode status output that was
missing, and normalizes the proc fan abi to return consistent data across
all fan read/write modes.
Userspace ABI changes and extensions:
1. Return status: enable/disable for *all* modes
(this actually improves compatibility with userspace utils!)
2. Return level: auto and level: disengaged for EC 2f access mode
3. Return level: <number> for EC 0x2f access mode
4. Return level 0 as well as "disabled" in level-aware modes
Signed-off-by: Henrique de Moraes Holschuh <hmh@hmh.eng.br>
This patch cleans up fan_write so that it is much easier to read and
extend. It separates the proc api handling from the operations themselves.
Signed-off-by: Henrique de Moraes Holschuh <hmh@hmh.eng.br>
This patch breaks fan_read mechanics into a generic function to get fan
status and speed, and leaves only the procfs interface code in fan_read.
Signed-off-by: Henrique de Moraes Holschuh <hmh@hmh.eng.br>
This patch cleans up fan_read so that it is much easier to read and
extend.
The patch fixes the userspace ABI to return "status: not supported" (like
all other ibm-acpi functions) when neither fan status or fan control are
possible.
It also fixes the userspace ABI to return EIO if ACPI access to the EC
fails, instead of returning "status: unreadable" or "speed: unreadable".
Signed-off-by: Henrique de Moraes Holschuh <hmh@hmh.eng.br>
This patch lays some groundwork for a fan_read and fan_write cleanup in the
next patches. To do so, it provides a new fan_init initializer, and also some
constants (through enums).
Signed-off-by: Henrique de Moraes Holschuh <hmh@hmh.eng.br>
This patch extends ibm-acpi to support reading thermal sensors directly
through ACPI EC register access. It uses a DMI match to detect ThinkPads
with a new-style embedded controller, that are known to have forward-
compatible register maps and use 0x00 to fill in non-used registers and
export thermal sensors at EC offsets 0x78-7F and 0xC0-C7.
Direct ACPI EC register access is implemented for 8-sensor and 16-sensor
new-style ThinkPad controller firmwares as an experimental feature. The
code does some limited sanity checks on the temperatures read through EC
access, and will default to the old ACPI TMP0-7 mode if anything is amiss.
Userspace ABI is not changed for 8 sensors, but /proc/acpi/ibm/thermal is
extended for 16 sensors if the firmware supports 16 sensors.
A documentation update is also provided.
The information about the ThinkPad register map was determined by studying
ibm-acpi "ecdump" output from various ThinkPad models, submitted by
subscribers of the linux-thinkpad mailinglist. Futher information was
gathered from the DSDT tables, as they describe the EC register map in
recent ThinkPads.
DSDT source shows that TMP0-7 access and direct register access are
actually the same thing on these firmwares, but unfortunately IBM never
did update their DSDT EC register map to export TMP8-TMP15 for the second
range of sensors.
Signed-off-by: Henrique de Moraes Holschuh <hmh@hmh.eng.br>
This patch consolidades all decisions regarding the strategy to be used to
read thinkpad thermal sensors into a single enum, and refactors the
thermal sensor reading code to use a much more readable (and easier to
extend) switch() construct, in a separate function.
Signed-off-by: Henrique de Moraes Holschuh <hmh@hmh.eng.br>
ibm-acpi uses sub-device names like ibm/hotkey, which get in the way of
a sysfs conversion. Fix it to use ibm_hotkey instead. Thanks to Zhang
Rui for noticing this.
Signed-off-by: Henrique de Moraes Holschuh <hmh@hmh.eng.br>
Conflicts:
drivers/ata/libata-scsi.c
include/linux/libata.h
Futher merge of Linus's head and compilation fixups.
Signed-Off-By: David Howells <dhowells@redhat.com>
Conflicts:
drivers/infiniband/core/iwcm.c
drivers/net/chelsio/cxgb2.c
drivers/net/wireless/bcm43xx/bcm43xx_main.c
drivers/net/wireless/prism54/islpci_eth.c
drivers/usb/core/hub.h
drivers/usb/input/hid-core.c
net/core/netpoll.c
Fix up merge failures with Linus's head and fix new compilation failures.
Signed-Off-By: David Howells <dhowells@redhat.com>
Change ACPI to use dev_archdata instead of firmware_data
This patch changes ACPI to use the new dev_archdata on i386, x86_64
and ia64 (is there any other arch using ACPI ?) to store it's
acpi_handle.
It also removes the firmware_data field from struct device as this
was the only user.
Only build-tested on x86
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Len Brown <lenb@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
processor_perflib.c::acpi_processor_ppc_notifier() check if the value
returned by the processor's _PPC method is 0 and return failed if so.
This is wrong since 0 indicate that the bios think the processor can go
to the highest frequency. This patch for example fix the HP NX 6125 to
allow its highest frequency to be available.
Signed-off-by: Bruno Ducrot <ducrot@poupinou.org>
Cc: "Pallipadi, Venkatesh" <venkatesh.pallipadi@intel.com>
Signed-off-by: Dave Jones <davej@redhat.com>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Pass the work_struct pointer to the work function rather than context data.
The work function can use container_of() to work out the data.
For the cases where the container of the work_struct may go away the moment the
pending bit is cleared, it is made possible to defer the release of the
structure by deferring the clearing of the pending bit.
To make this work, an extra flag is introduced into the management side of the
work_struct. This governs auto-release of the structure upon execution.
Ordinarily, the work queue executor would release the work_struct for further
scheduling or deallocation by clearing the pending bit prior to jumping to the
work function. This means that, unless the driver makes some guarantee itself
that the work_struct won't go away, the work function may not access anything
else in the work_struct or its container lest they be deallocated.. This is a
problem if the auxiliary data is taken away (as done by the last patch).
However, if the pending bit is *not* cleared before jumping to the work
function, then the work function *may* access the work_struct and its container
with no problems. But then the work function must itself release the
work_struct by calling work_release().
In most cases, automatic release is fine, so this is the default. Special
initiators exist for the non-auto-release case (ending in _NAR).
Signed-Off-By: David Howells <dhowells@redhat.com>
This reverts commit 37605a6900.
Again.
This same bug has now been introduced twice: it was done earlier by
commit b8d35192c5, only to be reverted
last time in commit 72945b2b90.
We must NOT try to queue up notify handlers to another thread than the
normal ACPI execution thread, because the notifications on some systems
seem to just keep on accumulating until we run out of memory and/or
threads.
Keeping events within the one deferred execution thread automatically
throttles the events properly.
At least the Compaq N620c will lock up completely on the first thermal
event without this patch reverted.
Cc: David Brownell <david-b@pacbell.net>
Cc: Len Brown <len.brown@intel.com>
Cc: Alexey Starikovskiy <alexey.y.starikovskiy@linux.intel.com>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
In addition to signalling button/lid events through /proc/acpi/event,
create separate input devices and report KEY_POWER, KEY_SLEEP and
SW_LID through input layer. Also remove unnecessary casts and variable
initializations, clean up formatting.
Sleep button may autorepeat but userspace will have to filter duplicate
sleep requests anyway (and discard unprocessed events right after
wakeup).
Unlike /proc/acpi/event interface input device corresponding to LID
switch reports true lid state instead of just a counter. SW_LID is
active when lid is closed.
The driver now depends on CONFIG_INPUT.
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>
Signed-off-by: Len Brown <len.brown@intel.com>
Fixing wrong description for acpi_gpe_sleep_prepare().
acpi_gpe_sleep_prepare() had only used on power off and was changed
to also used on entering some sleep state. However its description
isn't changed yet.
Signed-off-by: Satoru Takeuchi <takeuchi_satoru@jp.fujitsu.com>
Signed-off-by: Len Brown <len.brown@intel.com>
Fix this warning :
drivers/acpi/events/evmisc.c: In function `acpi_ev_global_lock_handler':
drivers/acpi/events/evmisc.c:334: warning: unused variable `status'
Signed-off-by: Jesper Juhl <jesper.juhl@gmail.com>
Signed-off-by: Len Brown <len.brown@intel.com>
http://bugzilla.kernel.org/show_bug.cgi?id=7303
Use a mutex instead of a spinlock for locking the
hotplug list because we need to call into the ACPI
subsystem which might sleep.
Signed-off-by: Kristen Carlson Accardi <kristen.c.accardi@intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>
acpi_get_pci_rootbridge_handle() walks the ACPI name space
searching for seg, bus and the PCI_ROOT_HID_STRING --
returning the handle as soon as if find the match.
But the current codes always parses through the whole namespace because
the user_function find_pci_rootbridge() returns status=AE_OK when it finds the match.
Make the find_pci_rootbridge() return AE_CTRL_TERMINATE when it finds the match.
This reduces the ACPI namespace walk for acpi_get_pci_rootbridge_handle().
Signed-off-by: Justin Chen <justin.chen@hp.com>
Signed-off-by: Len Brown <len.brown@intel.com>
Fix sparse warning:
drivers/acpi/ec.c:372:12: warning: function 'ec_transaction' with external linkage has definition
Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Len Brown <len.brown@intel.com>
Add support for the generic backlight interface below /sys/class/backlight.
Keep the procfs brightness handling for backward compatibility.
To achive this, add two generic functions get_lcd and set_lcd
to be used both by the procfs related and the sysfs related methods.
[apw@shadowen.org: backlight users need to select BACKLIGHT_CLASS_DEVICE]
Signed-off-by: Holger Macht <hmacht@suse.de>
Signed-off-by: Andy Whitcroft <apw@shadowen.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Len Brown <len.brown@intel.com>
Add support for the generic backlight interface below /sys/class/backlight.
Keep the procfs brightness handling for backward compatibility.
[apw@shadowen.org: backlight users need to select BACKLIGHT_CLASS_DEVICE]
Signed-off-by: Holger Macht <hmacht@suse.de>
Signed-off-by: Andy Whitcroft <apw@shadowen.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Len Brown <len.brown@intel.com>
Add support for the generic backlight interface below /sys/class/backlight.
The patch keeps the procfs brightness handling for backward compatibility.
Add two generic functions brightness_get and brightness_set
to be used both by the procfs related and the sysfs related methods.
[apw@shadowen.org: backlight users need to select BACKLIGHT_CLASS_DEVICE]
Signed-off-by: Holger Macht <hmacht@suse.de>
Signed-off-by: Andy Whitcroft <apw@shadowen.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Len Brown <len.brown@intel.com>
32bit vs 64 bit issues. sizeof(sizeof) and sizeof(pointer) is variable,
but we're trying to print it as unsigned int or u32.
Casts to unsigned long are used because type acpi_thread_id can be any one of
typedef u64 acpi_native_uint;
typedef u32 acpi_native_uint;
typedef u16 acpi_native_uint;
#define acpi_thread_id struct task_struct *
Signed-off-by: Martin J. Bligh <mbligh@google.com>
Acked-by: Jeff Garzik <jeff@garzik.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Len Brown <len.brown@intel.com>
I wrote a patch to avoid redundant memory hot-add call at boot time. This
was cause of strange fail message of memory hotplug like "ACPI: add_memory
failed". Memory is recognized by early boot code with EFI/E820.
But, if DSDT describes memory devices for them, then hot-add code is called
for already recognized memory, and it shows fail messages with -EEXIST.
So, sys admin will misunderstand this message as something wrong by it.
This patch avoids them by preventing redundant hot-add call until
completion of driver initialization.
[akpm@osdl.org: cleanups]
Signed-off-by: Yasunori Goto <y-goto@jp.fujitsu.com>
Cc: "Brown, Len" <len.brown@intel.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
I suppose this message seems quite useless except debugging. It just shows
"Hotplug Mem Device". System admin can't know anything by this message.
So, I would like to change it to KERN_DEBUG.
Signed-off-by: Yasunori Goto <y-goto@jp.fujitsu.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This patch breaks C-state discovery on my IBM IntelliStation Z30 because
the return value of acpi_processor_get_power_info_fadt is not assigned to
"result" in the case that acpi_processor_get_power_info_cst returns
-ENODEV. Thus, if ACPI provides C-state data via the FADT and not _CST (as
is the case on this machine), we incorrectly exit the function with -ENODEV
after reading the FADT. The attached patch sets the value of result so
that we don't exit early.
Signed-off-by: Darrick J. Wong <djwong@us.ibm.com>
Acked-by: "Pallipadi, Venkatesh" <venkatesh.pallipadi@intel.com>
Acked-by: "Brown, Len" <len.brown@intel.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
drivers/acpi/processor_idle.c:1112: warning: 'smp_callback' defined but not used
Cc: Len Brown <lenb@kernel.org>
Cc: Arjan van de Ven <arjan@infradead.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Add ->resume method to the ACPI battery handler to check
if the battery state has changed during sleep.
If yes, update the ACPI internal data structures
for benefit of /proc/acpi/battery/.
Signed-off-by: Jiri Kosina <jikos@jikos.cz>
Cc: Stefan Seyfried <seife@suse.de>
Acked-by: Pavel Machek <pavel@ucw.cz>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Len Brown <len.brown@intel.com>
I could not get correct PCI Express bus number from the structure of
acpi_object_extra. I always get zero as bus number regardless of bus
location. I found that there is incorrect comparison with _HID (PNP0A08) in
acpi/events/evrgnini.c and PCI Express _BBN method always fail.
Therefore, we always get zero as PCI Express bus number.
http://bugzilla.kernel.org/show_bug.cgi?id=7145
Signed-off-by: Bob Moore <robert.moore@intel.com>
Signed-off-by: Alexey Starikovskiy <alexey.y.starikovskiy@intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>
This reporting is useless (we errno anyway).
Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Len Brown <len.brown@intel.com>
ICC complains about a "Pointless comparsion of unsigned interger with zero"
@ line 760 & 808 of asus_acpi.c
parse_arg() mentioned below returns -E but it's copied into unsigned variable...
Signed-off-by: Darren Jenkins <darrenrjenkins@gmail.com>
Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Len Brown <len.brown@intel.com>
Fix printk format warnings in drivers/acpi:
drivers/acpi/tables/tbget.c:326: warning: format '%X' expects type 'unsigned int', but argument 5 has type 'long unsigned int'
drivers/acpi/tables/tbrsdt.c:189: warning: format '%X' expects type 'unsigned int', but argument 5 has type 'long unsigned int'
Signed-off-by: Randy Dunlap <rdunlap@xenotime.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Len Brown <len.brown@intel.com>
The ACPI processor init functions should be marked as __cpuinit as they use
structures marked with __cpuinitdata.
Signed-off-by: Pierre Ossman <drzeus@drzeus.cx>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Len Brown <len.brown@intel.com>
acpi_pci_link_set() allocates both with interrupts on
and with interrupts off (resume-time), so check interrupts
and decide on GFP_ATOMIC or GFP_KERNEL at run-time.
Signed-off-by: Jiri Kosina <jikos@jikos.cz>
Signed-off-by: Len Brown <len.brown@intel.com>
device was set to null and used before set in a debug printk
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>
Signed-off-by: Len Brown <len.brown@intel.com>
On acquiring the ACPI global lock, if there were sleepers on the lock,
we used to use acpi_os_execute() to defer a thread which would signal
sleepers. Now just signal the semaphore directly.
http://bugzilla.kernel.org/show_bug.cgi?id=5534#c159
Signed-off-by: Len Brown <len.brown@intel.com>
Simplify acpi_hw_low_level_xxx() functions to inb() and outb().
Signed-off-by: Alexey Y. Starikovskiy <alexey.y.starikovskiy@intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>
Remove unnecessary delay (50 ms) while reading data from EC in interrupt mode.
Signed-off-by: Alexey Y. Starikovskiy <alexey.y.starikovskiy@intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>
Unify the following functions:
acpi_ec_poll_read()
acpi_ec_poll_write()
acpi_ec_poll_query()
acpi_ec_intr_read()
acpi_ec_intr_write()
acpi_ec_intr_query()
into:
acpi_ec_poll_transaction()
acpi_ec_intr_transaction()
These new functions take as arguments an ACPI EC command, a few bytes
to write to the EC data register and a buffer for a few bytes to read
from the EC data register. The old _read(), _write(), _query() are
just special cases of these functions.
Then unified the code in acpi_ec_poll_transaction() and
acpi_ec_intr_transaction() a little more. Both functions are now just
wrappers around the new acpi_ec_transaction_unlocked() function. The
latter contains the EC access logic, the two original
function now just do their special way of locking and call the the
new function for the actual work.
This saves a lot of very similar code. The primary reason for doing
this, however, is that my driver for MSI 270 laptops needs to issue
some non-standard EC commands in a safe way. Due to this I added a new
exported function similar to ec_write()/ec_write() which is called
ec_transaction() and is essentially just a wrapper around
acpi_ec_{poll,intr}_transaction().
Signed-off-by: Lennart Poettering <mzxreary@0pointer.de>
Acked-by: Luming Yu <luming.yu@intel.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Len Brown <len.brown@intel.com>
Intel processors starting with the Core Duo support
support processor native C-state using the MWAIT instruction.
Refer: Intel Architecture Software Developer's Manual
http://www.intel.com/design/Pentium4/manuals/253668.htm
Platform firmware exports the support for Native C-state to OS using
ACPI _PDC and _CST methods.
Refer: Intel Processor Vendor-Specific ACPI: Interface Specification
http://www.intel.com/technology/iapc/acpi/downloads/302223.htm
With Processor Native C-state, we use 'MWAIT' instruction on the processor
to enter different C-states (C1, C2, C3). We won't use the special IO
ports to enter C-state and no SMM mode etc required to enter C-state.
Overall this will mean better C-state support.
One major advantage of using MWAIT for all C-states is, with this and
"treat interrupt as break event" feature of MWAIT, we can now get accurate
timing for the time spent in C1, C2, .. states.
Signed-off-by: Venkatesh Pallipadi <venkatesh.pallipadi@intel.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Len Brown <len.brown@intel.com>
The brightness and volume features from ibm-acpi are stable.
The experimental flag is no longer needed.
Signed-off-by: Stefan Schmidt <stefan@datenfreihafen.org>
Acked-by: Borislav Deianov <borislav@users.sourceforge.net>
Signed-off-by: Len Brown <len.brown@intel.com>
Add support for W3000 (W3V) and indirectly fixes an issue with kmilo under KDE
(it was triggering excessive LCD read error messages by querying asus_acpi
module) allowing people (I am probably the only one who tested this) with
W3000 to run kmilo.
Cc: Karol Kozimor <sziwan@hell.org.pl>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Len Brown <len.brown@intel.com>
This is needed by at least the Mac Mini's, which (incorrectly) come back
from suspend with SCI_EN clear.
Thanks to Frédéric Riss for hunting this down.
Acked-by: Frédéric Riss <frederic.riss@gmail.com>
Cc: Pavel Machek <pavel@ucw.cz>
Cc: Len Brown <len.brown@intel.com>
Cc: Arjan van de Ven <arjan@infradead.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Maintain a per-CPU global "struct pt_regs *" variable which can be used instead
of passing regs around manually through all ~1800 interrupt handlers in the
Linux kernel.
The regs pointer is used in few places, but it potentially costs both stack
space and code to pass it around. On the FRV arch, removing the regs parameter
from all the genirq function results in a 20% speed up of the IRQ exit path
(ie: from leaving timer_interrupt() to leaving do_IRQ()).
Where appropriate, an arch may override the generic storage facility and do
something different with the variable. On FRV, for instance, the address is
maintained in GR28 at all times inside the kernel as part of general exception
handling.
Having looked over the code, it appears that the parameter may be handed down
through up to twenty or so layers of functions. Consider a USB character
device attached to a USB hub, attached to a USB controller that posts its
interrupts through a cascaded auxiliary interrupt controller. A character
device driver may want to pass regs to the sysrq handler through the input
layer which adds another few layers of parameter passing.
I've build this code with allyesconfig for x86_64 and i386. I've runtested the
main part of the code on FRV and i386, though I can't test most of the drivers.
I've also done partial conversion for powerpc and MIPS - these at least compile
with minimal configurations.
This will affect all archs. Mostly the changes should be relatively easy.
Take do_IRQ(), store the regs pointer at the beginning, saving the old one:
struct pt_regs *old_regs = set_irq_regs(regs);
And put the old one back at the end:
set_irq_regs(old_regs);
Don't pass regs through to generic_handle_irq() or __do_IRQ().
In timer_interrupt(), this sort of change will be necessary:
- update_process_times(user_mode(regs));
- profile_tick(CPU_PROFILING, regs);
+ update_process_times(user_mode(get_irq_regs()));
+ profile_tick(CPU_PROFILING);
I'd like to move update_process_times()'s use of get_irq_regs() into itself,
except that i386, alone of the archs, uses something other than user_mode().
Some notes on the interrupt handling in the drivers:
(*) input_dev() is now gone entirely. The regs pointer is no longer stored in
the input_dev struct.
(*) finish_unlinks() in drivers/usb/host/ohci-q.c needs checking. It does
something different depending on whether it's been supplied with a regs
pointer or not.
(*) Various IRQ handler function pointers have been moved to type
irq_handler_t.
Signed-Off-By: David Howells <dhowells@redhat.com>
(cherry picked from 1b16e7ac850969f38b375e511e3fa2f474a33867 commit)
Add infrastructure to track "maximum allowable latency" for power saving
policies.
The reason for adding this infrastructure is that power management in the
idle loop needs to make a tradeoff between latency and power savings
(deeper power save modes have a longer latency to running code again). The
code that today makes this tradeoff just does a rather simple algorithm;
however this is not good enough: There are devices and use cases where a
lower latency is required than that the higher power saving states provide.
An example would be audio playback, but another example is the ipw2100
wireless driver that right now has a very direct and ugly acpi hook to
disable some higher power states randomly when it gets certain types of
error.
The proposed solution is to have an interface where drivers can
* announce the maximum latency (in microseconds) that they can deal with
* modify this latency
* give up their constraint
and a function where the code that decides on power saving strategy can
query the current global desired maximum.
This patch has a user of each side: on the consumer side, ACPI is patched
to use this, on the producer side the ipw2100 driver is patched.
A generic maximum latency is also registered of 2 timer ticks (more and you
lose accurate time tracking after all).
While the existing users of the patch are x86 specific, the infrastructure
is not. I'd like to ask the arch maintainers of other architectures if the
infrastructure is generic enough for their use (assuming the architecture
has such a tradeoff as concept at all), and the sound/multimedia driver
owners to look at the driver facing API to see if this is something they
can use.
[akpm@osdl.org: cleanups]
Signed-off-by: Arjan van de Ven <arjan@linux.intel.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Acked-by: Jesse Barnes <jesse.barnes@intel.com>
Cc: "Brown, Len" <len.brown@intel.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
In cases where the acpi memory-add event does not containe the pxm (node)
infomation allow the driver to look up node info based on the address. The
acpi_get_node call returns -1 if it can't decode the pxm info, this causes
add_memory to panic. acpi_get_node would have to decode the resource from the
handle (a lenghty proposition). This seems to be the cleanist point to
interject the hook.
[kamezawa.hiroyu@jp.fujitsu.com: build fixes]
[y-goto@jp.fujitsu.com: build fixes]
Signed-off-by: Keith Mannthey <kmannth@us.ibm.com>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Andi Kleen <ak@muc.de>
Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Signed-off-by: Yasunori Goto <y-goto@jp.fujitsu.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
