20 Nov, 2011
1 commit
-
Signed-off-by: Thomas Meyer
Signed-off-by: Jiri Kosina
07 Sep, 2011
3 commits
-
Profile changes were only reported when issued mouse internal. Now all
changes are reported.Signed-off-by: Stefan Achatz
Signed-off-by: Jiri Kosina -
Introduced function kone_profile_activated() to reduce code duplication.
This by the way fixes a wrong initialization value.
Also fixes early mutex unlocks.Signed-off-by: Stefan Achatz
Signed-off-by: Jiri Kosina -
Actual dpi resolution was set on wrong occassion.
Signed-off-by: Stefan Achatz
Signed-off-by: Jiri Kosina
13 Jun, 2011
2 commits
-
On rare occassions raw events can be triggered before drvdata gets set up
which leads to NULL pointer dereferences. This was only observed with pyra on
2.6.39, but is fixed for all devices now to play it save.
kovaplus returned wrong actual values when profile change was initiated from host.
Added range checks for setting actual profile on all devices.Signed-off-by: Stefan Achatz
Signed-off-by: Jiri Kosina -
Removed analog feature report enums and modified code in roccat_common
to reflect this. Non standard conform Kone got its own copy of the old
code. That helps extracting more generalizations for newer devices.Signed-off-by: Stefan Achatz
Signed-off-by: Jiri Kosina
03 Feb, 2011
4 commits
-
It was desired that the header roccat.h should be named hid-roccat.h
Signed-off-by: Stefan Achatz
Signed-off-by: Jiri Kosina -
Class was destroyed before starting the unregistering driver chain.
Disconnecting a device from roccat chardev in this process then
raised a NULL pointer dereference.
Fixed this by destroying class after unregistering driver.Signed-off-by: Stefan Achatz
Signed-off-by: Jiri Kosina -
Roccat chardev was reworked to support only a defined report size per
device and this can be retreived by an ioctl now to enable future changes
in report definitions.
Header was moved/renamed from drivers/hid to include/linux for accessibility.Signed-off-by: Stefan Achatz
Signed-off-by: Jiri Kosina -
Module hid-roccat-common contains functions used by roccat device driver
modules to reduce code duplication.
At the moment it contains just two wrapper methods for usb_control_msg
that ensure that the buffer used for transfer is dma capable which wasn't
the case before.
The kconfig option is not visible to the user but will be selected by the
device specific drivers.Signed-off-by: Stefan Achatz
Signed-off-by: Jiri Kosina
08 Jan, 2011
2 commits
-
The profile number is now passed via bin_attribute->private instead
of function parameter to reduce number of functions.Signed-off-by: Stefan Achatz
Signed-off-by: Jiri Kosina -
Adding sysfs attributes to an already created device raises no userland
notification. Now the device drivers associate the devices attributes
with a class and use this for roccat event char device creation.Signed-off-by: Stefan Achatz
Signed-off-by: Jiri Kosina
10 Dec, 2010
1 commit
-
Neaten current uses of dev_ by adding and using
hid specific hid_ macros.Convert existing uses of dev_ uses to hid_.
Convert hid-pidff printk uses to hid_.Remove err_hid and use hid_err instead.
Add missing newlines to logging messages where necessary.
Coalesce format strings.Add and use pr_fmt(fmt) KBUILD_MODNAME ": " fmt
Other miscellaneous changes:
Add const struct hid_device * argument to hid-core functions
extract() and implement() so hid_ can be used by them.
Fix bad indentation in hid-core hid_input_field function
that calls extract() function above.Signed-off-by: Joe Perches
Signed-off-by: Jiri Kosina
21 Jun, 2010
2 commits
-
The newest version of the accompanying userland tools cuts backward
compatibility and uses libudev to find its devices superseding the
quirky kone_abi_version sysfs attribute. Therefore it should be removed.Signed-off-by: Stefan Achatz
Signed-off-by: Jiri Kosina -
Fixing wrong calculated offsets in bin_attribute read functions.
Signed-off-by: Stefan Achatz
Signed-off-by: Jiri Kosina
31 May, 2010
2 commits
-
Renamed the sysfs attribute kone_driver_version to kone_abi_version and
simplified returned data to integer.Signed-off-by: Stefan Achatz
Signed-off-by: Jiri Kosina -
Removed comment that is obsolete since roccat char device is
implementedSigned-off-by: Stefan Achatz
Signed-off-by: Jiri Kosina
25 May, 2010
1 commit
-
Module roccat is a char device used to report special events of roccat hardware
to userland. These events include requests for on-screen-display of profile or
dpi settings or requests for execution of macro sequences that are not stored
in device. The information in these events depends on hid device implementation
and contains data that is not available in a single hid event or else hidraw
could have been used.It is inspired by hidraw, but uses only one circular buffer for all readers.
The device is as generic as possible so that the functionality is usable by all
(kone and upcomming) roccat device drivers.Signed-off-by: Stefan Achatz
Signed-off-by: Jiri Kosina
24 May, 2010
1 commit
-
After merging the driver-core tree, today's linux-next build (x86_64
allmodconfig) produced these warnings:drivers/hid/hid-roccat-kone.c:694: warning: initialization from incompatible pointer type
drivers/hid/hid-roccat-kone.c:696: warning: initialization from incompatible pointer type
drivers/hid/hid-roccat-kone.c:701: warning: initialization from incompatible pointer type
drivers/hid/hid-roccat-kone.c:703: warning: initialization from incompatible pointer type
drivers/hid/hid-roccat-kone.c:708: warning: initialization from incompatible pointer type
drivers/hid/hid-roccat-kone.c:710: warning: initialization from incompatible pointer type
drivers/hid/hid-roccat-kone.c:715: warning: initialization from incompatible pointer type
drivers/hid/hid-roccat-kone.c:717: warning: initialization from incompatible pointer type
drivers/hid/hid-roccat-kone.c:722: warning: initialization from incompatible pointer type
drivers/hid/hid-roccat-kone.c:724: warning: initialization from incompatible pointer type
drivers/hid/hid-roccat-kone.c:729: warning: initialization from incompatible pointer type
drivers/hid/hid-roccat-kone.c:731: warning: initialization from incompatible pointer typeIntroduced by commit 867040163f10f2b52b45bc573f330d6eb28f5914 ("sysfs:
add struct file* to bin_attr callbacks") from the driver-core tree
interacting with commit 14bf62cde79423a02a590e02664ed29a36facec1 ("HID:
add driver for Roccat Kone gaming mouse") from the hid tree.Signed-off-by: Stephen Rothwell
Signed-off-by: Jiri Kosina
19 May, 2010
3 commits
-
Removed useless preprocessor macros and renamed remaining one to be
more qualified.Signed-off-by: Stefan Achatz
Signed-off-by: Jiri Kosina -
As special events are reported along with hid event information all
events are now processed further by standard handler.
Also cleaned up this code.Signed-off-by: Stefan Achatz
Signed-off-by: Jiri Kosina -
Added new data and changed workaround for abnormal button behaviour
according to new gained knowledge about Roccat Kone device.Signed-off-by: Stefan Achatz
Signed-off-by: Jiri Kosina
18 May, 2010
1 commit
-
Added the __init and __exit hints for module functions.
Signed-off-by: Stefan Achatz
Signed-off-by: Jiri Kosina
30 Mar, 2010
1 commit
-
Implicit slab.h inclusion via percpu.h is about to go away. Make sure
gfp.h or slab.h is included as necessary.Signed-off-by: Tejun Heo
Cc: Stephen Rothwell
Cc: Jiri Kosina
Signed-off-by: Jiri Kosina
22 Mar, 2010
1 commit
-
This Patch adds support for Kone gaming mouse from Roccat.
It provides access to profiles, settings, firmware, weight,
actual settings etc. through sysfs attributes.
Event handling of this mouse differs from standard hid behaviour
in that tilt button press is reported in each move event which
results in strange behaviour if not handled by the driver.This is a heavily reworked version of the previously introduced driver.
The changes include most of the previously raised concerns,
memory leak and other fixes, code cleanups, adoption of additional
achieved knowlege about the hardware and is (IMHO) a much better version
than before even when I exchanged reduced USB-IO with a bigger memory
consumption.I refused to implement one mentioned point:
Removing the 'just-because-we-can' attributes. Motivation:
Reading the clipped in weight: I'm no gamer and can't determine the
usefulness of this feature but if the manufacturer implements such a
feature it might make sense to someone and I would unwillingly limit the
functionality besides its such a small feature.
Reading the actual profile and dpi settings: Here I can testify that one
can get lost of the actual settings when switching back and forth.
The manufacturers windows driver has the ability for on-screen-display
of the values and there is a mouse in the market that has an lcd on the
underside of it to show these values. So I think this feature makes sense
not only for me and shouldn't be removed.Signed-off-by: Stefan Achatz
Signed-off-by: Jiri Kosina