linux.git/drivers/input/touchscreen/tsc2007.c, branch v2.6.33 Linux kernel source tree Input: tsc2007 - check if I2C communication works during probe 2009-08-05T05:35:49+00:00 Richard Röjfors richard.rojfors.ext@mocean-labs.com 2009-08-05T05:34:10+00:00 cf5f439b48f82c230dcd81d0061e00664cbb6d39 Check the result when sending the power down command to the controller. Signed-off-by: Richard Röjfors <richard.rojfors.ext@mocean-labs.com> Signed-off-by: Dmitry Torokhov <dtor@mail.ru> 
Check the result when sending the power down command to the controller.

Signed-off-by: Richard Röjfors <richard.rojfors.ext@mocean-labs.com>
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>
Input: tsc2007 - make get_pendown_state platform callback optional 2009-08-05T05:35:34+00:00 Dmitry Torokhov dmitry.torokhov@gmail.com 2009-08-05T05:07:26+00:00 d570e9ef84e559b09e729f27f5381b6868f6cc5f In cases when get_pendown_state callback is not available have the driver to fallback on pressure calculation to determine if the pen is up. Signed-off-by: Dmitry Torokhov <dtor@mail.ru> 
In cases when get_pendown_state callback is not available have
the driver to fallback on pressure calculation to determine if
the pen is up.

Signed-off-by: Dmitry Torokhov <dtor@mail.ru>
Input: tsc2007 - do not read coordinates during probe 2009-07-26T18:21:03+00:00 Richard Röjfors richard.rojfors.ext@mocean-labs.com 2009-07-25T06:14:17+00:00 703490ff7eaff03e412683da3d8367b5190a71ca Don't read coordinates during probe of the driver, just power down the controller and wait for interrupts. Signed-off-by: Richard Röjfors<richard.rojfors.ext@mocean-labs.com> Signed-off-by: Dmitry Torokhov <dtor@mail.ru> 
Don't read coordinates during probe of the driver, just power down
the controller and wait for interrupts.

Signed-off-by: Richard Röjfors<richard.rojfors.ext@mocean-labs.com>
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>
Input: tsc2007 - make init/exit platform hw callbacks optional 2009-07-26T18:21:00+00:00 Richard Röjfors richard.rojfors.ext@mocean-labs.com 2009-07-25T06:14:17+00:00 cad065fd1e95003444e8528f801f52164cb78561 Make init_platform_hw and exit_platform_hw callbacks optional since they are not needed on all platforms. Signed-off-by: Richard Röjfors <richard.rojfors.ext@mocean-labs.com> Signed-off-by: Dmitry Torokhov <dtor@mail.ru> 
Make init_platform_hw and exit_platform_hw callbacks optional since
they are not needed on all platforms.

Signed-off-by: Richard Röjfors <richard.rojfors.ext@mocean-labs.com>
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>
Input: tsc2007 - properly shut off interrupts/delayed work 2009-07-26T18:20:57+00:00 Dmitry Torokhov dmitry.torokhov@gmail.com 2009-07-25T06:14:16+00:00 141586bc57f6083f36c18d86e1cfa5916a1e7c05 Properly shut off interrupts/delayed work by free-ing IRQ first and then ensuring that enable/disable is balanced. Also add __devinit/__devexit markings, restore poll delay/period scheduling logic, make sure we call exit_platform_hw() method when probe fails. Tested-by: Richard Röjfors <richard.rojfors.ext@mocean-labs.com> Signed-off-by: Dmitry Torokhov <dtor@mail.ru> 
Properly shut off interrupts/delayed work by free-ing IRQ first
and then ensuring that enable/disable is balanced. Also add
__devinit/__devexit markings, restore poll delay/period scheduling
logic, make sure we call exit_platform_hw() method when probe
fails.

Tested-by: Richard Röjfors <richard.rojfors.ext@mocean-labs.com>
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>
Input: tsc2007 - remove HR timer 2009-07-26T18:20:55+00:00 Richard Röjfors richard.rojfors.ext@mocean-labs.com 2009-07-25T05:01:39+00:00 75fba3b05d6ed82b975c1f8df8f8e08d5d81dee5 Since it's not allowed to do synchronous I2C in the HR timer callback context we have to switch to using the global workqueue. The work is scheduled every 1ms when polling rather than 5 us. Signed-off-by: Richard Röjfors <richard.rojfors.ext@mocean-labs.com> Signed-off-by: Dmitry Torokhov <dtor@mail.ru> 
Since it's not allowed to do synchronous I2C in the HR timer callback
context we have to switch to using the global workqueue. The work is
scheduled every 1ms when polling rather than 5 us.

Signed-off-by: Richard Röjfors <richard.rojfors.ext@mocean-labs.com>
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>
Merge commit 'v2.6.30' into next 2009-06-11T08:58:01+00:00 Dmitry Torokhov dmitry.torokhov@gmail.com 2009-06-11T08:58:01+00:00 4894e4aca82aca927d0404ce61f021f790de4b1e 






Input: tsc2007 - make sure platform provides get_pendown_state()
2009-05-16T02:23:54+00:00

Kwangwoo Lee
kwangwoo.lee@gmail.com

2009-05-16T02:14:35+00:00

78f7f36711396991431a1d7ceab6103d2c54694c

The platform codes must provide get_pendown_state() for the driver
to work properly.

Signed-off-by: Kwangwoo Lee <kwangwoo.lee@gmail.com>
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>





The platform codes must provide get_pendown_state() for the driver
to work properly.

Signed-off-by: Kwangwoo Lee <kwangwoo.lee@gmail.com>
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>







Input: tsc2007 - fix locking in hrtimer handler
2009-05-11T22:35:48+00:00

Thierry Reding
thierry.reding@avionic-design.de

2009-05-11T15:36:39+00:00

705a76d2d22a2824f45f07d023a380293554e989

Now that hrtimers are always running in hard irq context we can't
unconditionally enable interrupts at the end of the timer function.

Signed-off-by: Thierry Reding <thierry.reding@avionic-design.de>
Signed-off-by: Kwangwoo Lee <kwangwoo.lee@gmail.com>
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>





Now that hrtimers are always running in hard irq context we can't
unconditionally enable interrupts at the end of the timer function.

Signed-off-by: Thierry Reding <thierry.reding@avionic-design.de>
Signed-off-by: Kwangwoo Lee <kwangwoo.lee@gmail.com>
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>







Input: tsc2007 - use disable_irq_nosync() in irq handler
2009-04-18T03:42:00+00:00

Ben Nizette
bn@niasdigital.com

2009-04-18T03:35:57+00:00

29fa98bd9b50b5e9c17d99e961640ffafb765d71

disable_irq() waits for all running handlers to complete before
returning.  As such, if it's used to disable an interrupt from
that interrupt's handler it will deadlock.  This replaces the
dangerous instances with the _nosync() variant which doesn't have
this problem.

Signed-off-by: Ben Nizette <bn@niasdigital.com>
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>





disable_irq() waits for all running handlers to complete before
returning.  As such, if it's used to disable an interrupt from
that interrupt's handler it will deadlock.  This replaces the
dangerous instances with the _nosync() variant which doesn't have
this problem.

Signed-off-by: Ben Nizette <bn@niasdigital.com>
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>