ledtrig-cpu: kill useless mutex to fix sleep in atomic context

2012-11-11T20:09:43Z

Seeing the following every time the CPU enters or leaves idle on a Beagleboard: BUG: sleeping function called from invalid context at kernel/mutex.c:269 in_atomic(): 1, irqs_disabled(): 0, pid: 0, name: swapper/0 no locks held by swapper/0/0. [] (unwind_backtrace+0x0/0xf8) from [] (mutex_lock_nested+0x24/0x380) [] (mutex_lock_nested+0x24/0x380) from [] (ledtrig_cpu+0x38/0x88) [] (ledtrig_cpu+0x38/0x88) from [] (cpu_idle+0xf4/0x120) [] (cpu_idle+0xf4/0x120) from [] (start_kernel+0x2bc/0x30c) Miles Lane has reported seeing similar splats during system suspend. The mutex in struct led_trigger_cpu appears to have no function: it resides in a per-cpu data structure which never changes after the trigger is registered. So just remove it. Reported-by: Miles Lane Signed-off-by: Nathan Lynch Signed-off-by: Bryan Wu

led-triggers: create a trigger for CPU activity

2012-07-31T13:23:31Z

Attempting to consolidate the ARM LED code, this removes the custom RealView LED trigger code to turn LEDs on and off in response to CPU activity and replace it with a standard trigger. (bryan.wu@canonical.com: It introduces several syscore stubs into this trigger. It also provides ledtrig_cpu trigger event stub in . Although it was inspired by ARM work, it can be used in other arch.) Cc: Richard Purdie Signed-off-by: Linus Walleij Signed-off-by: Bryan Wu Reviewed-by: Jamie Iles Tested-by: Jochen Friedrich

kernel.git/drivers/leds/ledtrig-cpu.c, branch linux-3.8.y

ledtrig-cpu: kill useless mutex to fix sleep in atomic context

led-triggers: create a trigger for CPU activity