Callback methods should not refer to a variable like "eeepc" (formally
"ehotk"). Instead, they should extract the data they need either from
a "driver data" parameter, or the "driver data" field of the object
which they operate on. The "eeepc" variable can then be removed.
In practice, drivers under "drivers/platform" can get away without using
driver data, because it doesn't make sense to have more than one
instance of them. However this makes it harder to review them for
correctness. This is especially true for core ACPI developers who have
not previously been exposed to this anti-pattern :-).
This will serve as an example of best practice for new driver writers
(whether they find it themselves, or have it pointed out during review
:-).
The hwmon sub-device is a special case. It uses ec_{read,write} which
are defined to communicate with the (first) EC, so it does not require
any driver data. It should still only be instantiated in the context of
an ASUS010 device because we don't have a safe way to probe for it.
Signed-off-by: Alan Jenkins <alan-jenkins@tuffmail.co.uk>
CC: Bjorn Helgaas <bjorn.helgaas@hp.com>
Signed-off-by: Len Brown <len.brown@intel.com>
eeepc-laptop now does a lot more than just hotkeys. Replace the "hotk"
names used throughout the driver with some slightly more appropriate
names. The actual strings used in kernel messages and sysfs are left
unchanged.
e.g.
EEEPC_HOTK_FILE -> EEEPC_LAPTOP_FILE
EEEPC_HOTK_HID -> EEEPC_ACPI_HID
eeepc_hotk_notify -> eeepc_acpi_notify
struct eeepc_hotk -> struct eeepc_laptop
ehotk -> eeepc
I'm about to refactor the entire driver to remove the global "ehotk"
variable, and I don't wish to add "struct eeepc_hotk *ehotk" to
functions which have nothing to do with hotkeys.
Also
- fix the name of "eepc_get_entry_by_keycode()"
- remove the unused definition of NOTIFY_WLAN_ON.
Signed-off-by: Alan Jenkins <alan-jenkins@tuffmail.co.uk>
Signed-off-by: Len Brown <len.brown@intel.com>
Move e.g. backlight_init() and backlight_exit() together along with the
other backlight functions, instead of grouping init() and exit()
functions. Move e.g. backlight_ops to follow the functions it refers
to, and remove the forward declarations. The code itself should remain
unchanged.
The eeepc-laptop driver implements a number of interfaces like the
backlight class driver. This change makes it easier to examine the
implementation of one interface at at a time, without having to search
through the file to find init() and exit() functions etc.
Signed-off-by: Alan Jenkins <alan-jenkins@tuffmail.co.uk>
Signed-off-by: Len Brown <len.brown@intel.com>
This moves the sysfs_create_group() call just after the declaration of
the platform device attributes. It should make it easier to examine
the implementation of the platform device attributes in isolation
from the rest of the code. (The next commit will apply this pattern
to all of the sub-devices as well).
Signed-off-by: Alan Jenkins <alan-jenkins@tuffmail.co.uk
Signed-off-by: Len Brown <len.brown@intel.com>
Strictly speaking we should register the platform driver exactly once,
whether there are zero, one, or multiple matching acpi devices.
Signed-off-by: Alan Jenkins <alan-jenkins@tuffmail.co.uk>
Signed-off-by: Len Brown <len.brown@intel.com>
Separate out input_notify(), in a similar way to how notify_brn()
is already separated. This will allow all the functions which refer to
the input device to be grouped together.
This includes a small behaviour change - we now synthesize brightness
up/down key events even if the brightness is already at the
maximum/minimum value. This is consistent with the new uevent
interface.
Signed-off-by: Alan Jenkins <alan-jenkins@tuffmail.co.uk>
Signed-off-by: Len Brown <len.brown@intel.com>
The hwmon device uses ec_write() to write values to the EC. So for
consistency it should use ec_read() to read values. The extra layers
of indirection used did not add any value.
This may mean we no longer take the ACPI global lock for such reads
(if the EC operation region requires the lock and the EC does not).
But there is no point locking each one-byte read individually, when
write operations do not use the lock at all.
Signed-off-by: Alan Jenkins <alan-jenkins@tuffmail.co.uk>
Signed-off-by: Len Brown <len.brown@intel.com>
We don't need to store init_flags after using them. And we don't use
the result of INIT, so we don't need to allocate a buffer for it.
Signed-off-by: Alan Jenkins <alan-jenkins@tuffmail.co.uk>
Signed-off-by: Len Brown <len.brown@intel.com>
We already tell the backlight class our maximum brightness value; it
will validate the user requested values for us.
Signed-off-by: Alan Jenkins <alan-jenkins@tuffmail.co.uk>
Signed-off-by: Len Brown <len.brown@intel.com>
eeepc_hotk_notify() cannot be called with ehotk == NULL or bd == NULL.
We check both variables for allocation failure and would bail out before
the notifier is registered.
Signed-off-by: Alan Jenkins <alan-jenkins@tuffmail.co.uk>
Signed-off-by: Len Brown <len.brown@intel.com>
If the control method does not exist, return -ENODEV for consistency
with get_acpi()
Signed-off-by: Alan Jenkins <alan-jenkins@tuffmail.co.uk>
Signed-off-by: Len Brown <len.brown@intel.com>
If we bail out because we can't create the led class device, we need to
ensure the led workqueue is cleaned up.
Signed-off-by: Alan Jenkins <alan-jenkins@tuffmail.co.uk>
Signed-off-by: Len Brown <len.brown@intel.com>
Create the workqueue thread used by tpd_led_set() *before* we register
the led device. (And vice versa for unregistration).
Signed-off-by: Alan Jenkins <alan-jenkins@tuffmail.co.uk>
Signed-off-by: Len Brown <len.brown@intel.com>
Documentation/hwmon/sysfs-interface tells us that automatic fan speed
control should be represented by a value of 2 or above for pwm1_enable.
Fix eeepc_get_fan_ctrl() to return 2 for automatic fan control.
Setting "1" for manual control is already consistent with the
documentation, so this remains unchanged.
Let's preserve the ABI for this specific driver, so that writing "0"
will still invoke automatic control.
(The documentation says setting "0" should leave the fan at full speed
all the time. This mode is not directly supported by our hardware. Full
speed is rather noisy on my 701 and the automatic control has never used
it. If you really want this e.g. to prolong the life of an EeePC used
as a server, you can always use manual mode. hwmon has always been
fairly machine-specific, and you're in a tiny minority (or elite :-).
I'm sure you're smart enough to notice that the fan doesn't turn on to
full speed when you try this mode, either by ear or checking
fan_input1.
We could even claim to be honouring the spirit of the documentation.
"0" really means "safe mode". EeePCs default to automatic mode, ie that
is what Asus will actually test. Since we do not provide any way to
tamper with the temperature threshold, automatic mode _is_ the safe
option).
Signed-off-by: Alan Jenkins <alan-jenkins@tuffmail.co.uk>
Signed-off-by: Len Brown <len.brown@intel.com>
The owner field provides the link between drivers and modules in sysfs,
but no ACPI driver was setting it.
After setting the owner field, we can see which module provides which
driver and vice versa by looking at /sys/bus/acpi/driver/*/module and
/sys/module/*/drivers/acpi:*.
Signed-off-by: Alan Jenkins <alan-jenkins@tuffmail.co.uk>
Signed-off-by: Len Brown <len.brown@intel.com>
acpi_bus_register_driver() already checks acpi_disabled, so acpi bus
drivers don't need to.
Signed-off-by: Alan Jenkins <alan-jenkins@tuffmail.co.uk>
Signed-off-by: Len Brown <len.brown@intel.com>
The acpi device callbacks add, start, remove, suspend and resume can
never be called with a NULL acpi_device. Each callsite in acpi/scan.c
has to dereference the device in order to get the ops structure, e.g.
struct acpi_device *acpi_dev = to_acpi_device(dev);
struct acpi_driver *acpi_drv = acpi_dev->driver;
if (acpi_drv && acpi_drv->ops.suspend)
return acpi_drv->ops.suspend(acpi_dev, state);
Remove all checks for acpi_dev == NULL within these callbacks.
Also remove the checks for acpi_driver_data(acpi_dev) == NULL. None of
these checks could fail unless the driver does something strange
(which none of them do), the acpi core did something terribly wrong,
or we have a memory corruption issue. If this does happen then it's
best to dereference the pointer and crash noisily.
Signed-off-by: Alan Jenkins <alan-jenkins@tuffmail.co.uk>
Signed-off-by: Len Brown <len.brown@intel.com>
The owner field provides the link between drivers and modules in sysfs,
but no ACPI driver was setting it.
After setting the owner field, we can see which module provides which
driver and vice versa by looking at /sys/bus/acpi/driver/*/module and
/sys/module/*/drivers/acpi:*.
Signed-off-by: Alan Jenkins <alan-jenkins@tuffmail.co.uk>
Signed-off-by: Len Brown <len.brown@intel.com>
The owner field provides the link between drivers and modules in sysfs,
but no ACPI driver was setting it.
After setting the owner field, we can see which module provides which
driver and vice versa by looking at /sys/bus/acpi/driver/*/module and
/sys/module/*/drivers/acpi:*.
Signed-off-by: Alan Jenkins <alan-jenkins@tuffmail.co.uk>
Signed-off-by: Len Brown <len.brown@intel.com>
acpi_bus_register_driver() already checks acpi_disabled, so acpi bus
drivers don't need to.
Signed-off-by: Alan Jenkins <alan-jenkins@tuffmail.co.uk>
Signed-off-by: Len Brown <len.brown@intel.com>
acpi_bus_register_driver() already checks acpi_disabled, so acpi bus
drivers don't need to.
Signed-off-by: Alan Jenkins <alan-jenkins@tuffmail.co.uk>
Signed-off-by: Len Brown <len.brown@intel.com>
The acpi device callbacks add, start, remove, suspend and resume can
never be called with a NULL acpi_device. Each callsite in acpi/scan.c
has to dereference the device in order to get the ops structure, e.g.
struct acpi_device *acpi_dev = to_acpi_device(dev);
struct acpi_driver *acpi_drv = acpi_dev->driver;
if (acpi_drv && acpi_drv->ops.suspend)
return acpi_drv->ops.suspend(acpi_dev, state);
Remove all checks for acpi_dev == NULL within these callbacks.
Also remove the checks for acpi_driver_data(acpi_dev) == NULL. None of
these checks could fail unless the driver does something strange
(which none of them do), the acpi core did something terribly wrong,
or we have a memory corruption issue. If this does happen then it's
best to dereference the pointer and crash noisily.
Signed-off-by: Alan Jenkins <alan-jenkins@tuffmail.co.uk>
Signed-off-by: Len Brown <len.brown@intel.com>
The acpi device callbacks add, start, remove, suspend and resume can
never be called with a NULL acpi_device. Each callsite in acpi/scan.c
has to dereference the device in order to get the ops structure, e.g.
struct acpi_device *acpi_dev = to_acpi_device(dev);
struct acpi_driver *acpi_drv = acpi_dev->driver;
if (acpi_drv && acpi_drv->ops.suspend)
return acpi_drv->ops.suspend(acpi_dev, state);
Remove all checks for acpi_dev == NULL within these callbacks.
Also remove the checks for acpi_driver_data(acpi_dev) == NULL. None of
these checks could fail unless the driver does something strange
(which none of them do), the acpi core did something terribly wrong,
or we have a memory corruption issue. If this does happen then it's
best to dereference the pointer and crash noisily.
Signed-off-by: Alan Jenkins <alan-jenkins@tuffmail.co.uk>
Signed-off-by: Len Brown <len.brown@intel.com>
Currently, reading from the disp attribute fails with "No such device",
which is misleading. According to CMSG table on acpi4asus project site,
no models have a getter method corresponding to SDSP. Change the file
permission to disallow reads.
If some joker changes the permission to permit reads, then return -EIO
to be consistent with sysfs' behaviour when no show() method is
provided.
Signed-off-by: Alan Jenkins <alan-jenkins@tuffmail.co.uk>
Signed-off-by: Corentin Chary <corentincj@iksaif.net>
Signed-off-by: Len Brown <len.brown@intel.com>
Use input_set_capability() instead of set_bit.
Signed-off-by: Henrique de Moraes Holschuh <hmh@hmh.eng.br>
Cc: Dmitry Torokhov <dtor@mail.ru>
Signed-off-by: Len Brown <len.brown@intel.com>
Log temperatures on any of the EC thermal alarms. It could be
useful to help tracking down what is happening...
Signed-off-by: Henrique de Moraes Holschuh <hmh@hmh.eng.br>
Acked-by: Pavel Machek <pavel@ucw.cz>
Signed-off-by: Len Brown <len.brown@intel.com>
Export the normal (non-command) module paramenters as mode 0444, so
that they will show up in sysfs.
These parameters must not be changed at runtime as a rule, with very
few exceptions.
Reported-by: Ferenc Wagner <wferi@niif.hu>
Signed-off-by: Henrique de Moraes Holschuh <hmh@hmh.eng.br>
Signed-off-by: Len Brown <len.brown@intel.com>
Properly init the parent field of the input device. Thanks to Alan
Jenkins, who noted this problem in a different driver.
Reported-by: Alan Jenkins <alan-jenkins@tuffmail.co.uk>
Signed-off-by: Henrique de Moraes Holschuh <hmh@hmh.eng.br>
Signed-off-by: Len Brown <len.brown@intel.com>
Fix this bogus warning during module shutdown, when
backlight event reporting is enabled:
"thinkpad_acpi: required events 0x00018000 not enabled!"
Signed-off-by: Henrique de Moraes Holschuh <hmh@hmh.eng.br>
Signed-off-by: Len Brown <len.brown@intel.com>
Take advantage of the new events capabilities of the backlight class to
notify userspace of backlight changes.
This depends on "backlight: Allow drivers to update the core, and
generate events on changes", by Matthew Garrett.
Signed-off-by: Henrique de Moraes Holschuh <hmh@hmh.eng.br>
Cc: Matthew Garrett <mjg@redhat.com>
Cc: Richard Purdie <rpurdie@linux.intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>
Since the rfkill rework in 2.6.31, the driver is always resuming with
the radios disabled.
Change thinkpad-acpi to ask the firmware to resume with the radios in
the last state. This fixes the Bluetooth and WWAN rfkill switches.
Note that it means we respect the firmware's oddities. Should the
user toggle the hardware rfkill switch on and off, it might cause the
radios to resume enabled.
UWB is an unknown quantity since it has nowhere the same level of
firmware support (no control over state storage in NVRAM, for
example), and might need further fixing. Testers welcome.
This change fixes a regression from 2.6.30.
Reported-by: Jerone Young <jerone.young@canonical.com>
Reported-by: Ian Molton <ian.molton@collabora.co.uk>
Signed-off-by: Henrique de Moraes Holschuh <hmh@hmh.eng.br>
Tested-by: Ian Molton <ian.molton@collabora.co.uk>
Cc: stable@kernel.org
Signed-off-by: Len Brown <len.brown@intel.com>
According to a report, the R50e wants EC-based brightness control,
even if it uses an Intel GPU. The current driver default was reported
to not work at all.
This bug can be worked around by the "brightness_mode=3" module
parameter.
Change the default of the R50e and R51 2xxx models (which use the same
EC firmware, 1V) to TPACPI_BRGHT_Q_EC, but keep TPACPI_BRGHT_Q_ASK set
for now, as I'd like to get more reports.
This fixes a regression caused by commit
59fe4fe34d7afdf63208124f313be9056feaa2f4,
"thinkpad-acpi: fix incorrect use of TPACPI_BRGHT_MODE_ECNVRAM"
Kernel 2.6.31 also needs this fix.
Reported-by: Ferenc Wagner <wferi@niif.hu>
Tested-by: Ferenc Wagner <wferi@niif.hu>
Signed-off-by: Henrique de Moraes Holschuh <hmh@hmh.eng.br>
Cc: stable@kernel.org
Signed-off-by: Len Brown <len.brown@intel.com>
Return temperature in milidegree instead of degree, as sysfs-api requires
the temperature in milidegree.
Signed-off-by: Peter Feuerer <peter@piie.net>
Tested-by: Borislav Petkov <petkovbb@gmail.com>
Cc: Andreas Mohr <andi@lisas.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Len Brown <len.brown@intel.com>
There is a problem in the quirk tables used by tpacpi_is_fw_known() and
tpacpi_check_outdated_fw(), which causes outdated BIOSes that are lacking
the EC firmware ID DMI field to never match.
This breaks module loading on, e.g. a T23 with outdated BIOS, and the
module will refuse to load unless the "force_load=1" parameter is given.
Fix the quirk tables so that they can also match the outdated BIOSes,
which in turn will both fix the module loading, and also warn the user
that he is using outdated firmware and should upgrade.
This fixes a serious regression, introduced by commit
e675abafcc0df38125e6e94a9ba91c92fe774f52, "thinkpad-acpi: be more strict
when detecting a ThinkPad".
http://bugzilla.kernel.org/show_bug.cgi?id=14597
Reported-by: Paul Kimoto <kimoto@lightlink.com>
Signed-off-by: Henrique de Moraes Holschuh <hmh@hmh.eng.br>
Tested-by: Paul Kimoto <kimoto@lightlink.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Len Brown <len.brown@intel.com>
The returned error should be negative
Signed-off-by: Roel Kluin <roel.kluin@gmail.com>
Acked-by: Henrique de Moraes Holschuh <hmh@hmh.eng.br>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Cc: <stable@kernel.org>
Signed-off-by: Len Brown <len.brown@intel.com>
The existing interface only has a pre-order callback. This change
adds an additional parameter for a post-order callback which will
be more useful for bus scans. ACPICA BZ 779.
Also update the external calls to acpi_walk_namespace.
http://www.acpica.org/bugzilla/show_bug.cgi?id=779
Signed-off-by: Lin Ming <ming.m.lin@intel.com>
Signed-off-by: Bob Moore <robert.moore@intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>
WMI provides interface-specific GUIDs that are exported from modules as
modalises, but the core currently generates no events to trigger module
loading. This patch adds support for registering devices for each WMI GUID
and generating the appropriate uevent.
Based heavily on a patch by Carlos Corbacho (<carlos@strangeworlds.co.uk>).
Signed-off-by: Matthew Garrett <mjg@redhat.com>
Tested-by: Carlos Corbacho <carlos@strangeworlds.co.uk>
Acked-by: Carlos Corbacho <carlos@strangeworlds.co.uk>
Signed-off-by: Len Brown <len.brown@intel.com>
Switching the camera takes 500ms, checking if it's on is almost free...
The BIOS remembers the setting through reboots, so there's good chance the
camera is already enabled.
Signed-off-by: Luca Niccoli <lultimouomo@gmail.com>
Cc: Corentin Chary <corentincj@iksaif.net>
Cc: Alan Jenkins <alan-jenkins@tuffmail.co.uk>
Cc: Matthew Garrett <mjg@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Corentin Chary <corentincj@iksaif.net>
Signed-off-by: Len Brown <len.brown@intel.com>
rt2860sta is fine with the patch as is, but iwl3945 isn't
(eeepc_rfkill_set() needs to call eeepc_rfkill_hotplug(true) – which means
that we're back to causing the rt2860sta panic
This reverts commit b56ab33d68638e6aafdbfc694025e8354a628f49.
Signed-off-by: Corentin Chary <corentincj@iksaif.net>
Signed-off-by: Darren Salt <linux@youmustbejoking.demon.co.uk>
Signed-off-by: Len Brown <len.brown@intel.com>
This works around what I think is actually a bug in rt2860sta which is
triggered when the hardware "disappears" from beneath the driver, i.e. when
wireless is toggled off via ACPI. It does so by ensuring that the rfkill
soft-block flag is set before the hardware is disabled.
I do not know whether this patch is required if rt2800pci is in use instead
of rt2860sta; at the time of submission of this patch, I've not been able to
test this.
(Ref. http://bugzilla.kernel.org/show_bug.cgi?id=13390)
Signed-off-by: Darren Salt <linux@youmustbejoking.demon.co.uk>
Signed-off-by: Len Brown <len.brown@intel.com>
Currently the annotation for function eeepc_enable_camera() is
__init, and refers to a
function eeepc_hotk_add() which is non-init. Use __devinit for both
functions which is
more appropriate and fixes a section mismatch warning.
We were warned by the following warning:
LD drivers/platform/x86/built-in.o
WARNING: drivers/platform/x86/built-in.o(.text+0x12e1): Section
mismatch in reference from the function eeepc_hotk_add() to the
function .init.text:eeepc_enable_camera()
The function eeepc_hotk_add() references
the function __init eeepc_enable_camera().
This is often because eeepc_hotk_add lacks a __init
annotation or the annotation of eeepc_enable_camera is wrong.
Signed-off-by: Rakib Mullick <rakib.mullick@gmail.com>
Signed-off-by: Corentin Chary <corentincj@iksaif.net>
Signed-off-by: Len Brown <len.brown@intel.com>
A follow-up 2.6.32-rc1's
1e384cb0f9a940f2a431d1708f963987e61d71e3
"fujitsu-laptop: support led-class as module"
It's a trivial fix for one of the CONFIG_LEDS_CLASS ifdefs
which was somehow missed in the original patch.
Signed-off-by: Jonathan Woithe <jwoithe@physics.adelaide.edu.au>
Signed-off-by: Len Brown <len.brown@intel.com>
