2012-05-01 11:46:19

by Bryan Wu

[permalink] [raw]
Subject: Re: [PATCH 02/19] led-triggers: create a trigger for CPU activity

On Mon, Apr 30, 2012 at 9:03 PM, Tim Gardner <[email protected]> wrote:
> On 04/30/2012 01:37 AM, Bryan Wu wrote:
>> 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.
>>
>> ([email protected]:
>> It introduces several syscore stubs into this trigger.
>> It also provides ledtrig_cpu trigger event stub in <linux/leds.h>.
>> Although it was inspired by ARM work, it can be used in other arch.)
>>
>> Cc: Richard Purdie <[email protected]>
>> Signed-off-by: Linus Walleij <[email protected]>
>> Signed-off-by: Bryan Wu <[email protected]>
>> Reviewed-by: Jamie Iles <[email protected]>
>> Tested-by: Jochen Friedrich <[email protected]>
>> ---
>> ?drivers/leds/Kconfig ? ? ? | ? 10 +++
>> ?drivers/leds/Makefile ? ? ?| ? ?1 +
>> ?drivers/leds/ledtrig-cpu.c | ?150 ++++++++++++++++++++++++++++++++++++++++++++
>> ?include/linux/leds.h ? ? ? | ? 16 +++++
>> ?4 files changed, 177 insertions(+)
>> ?create mode 100644 drivers/leds/ledtrig-cpu.c
>>
>> diff --git a/drivers/leds/Kconfig b/drivers/leds/Kconfig
>> index 5f12659..dbf8a4c 100644
>> --- a/drivers/leds/Kconfig
>> +++ b/drivers/leds/Kconfig
>> @@ -465,6 +465,16 @@ config LEDS_TRIGGER_BACKLIGHT
>>
>> ? ? ? ? If unsure, say N.
>>
>> +config LEDS_TRIGGER_CPU
>> + ? ? tristate "LED CPU Trigger"
>> + ? ? depends on LEDS_TRIGGERS
>> + ? ? help
>> + ? ? ? This allows LEDs to be controlled by active CPUs. This shows
>> + ? ? ? the active CPUs across an array of LEDs so you can see which
>> + ? ? ? CPUs are active on the system at any given moment.
>> +
>> + ? ? ? If unsure, say N.
>> +
>> ?config LEDS_TRIGGER_GPIO
>> ? ? ? tristate "LED GPIO Trigger"
>> ? ? ? depends on LEDS_TRIGGERS
>> diff --git a/drivers/leds/Makefile b/drivers/leds/Makefile
>> index 9475bbb..ea1efb2 100644
>> --- a/drivers/leds/Makefile
>> +++ b/drivers/leds/Makefile
>> @@ -56,4 +56,5 @@ obj-$(CONFIG_LEDS_TRIGGER_IDE_DISK) += ledtrig-ide-disk.o
>> ?obj-$(CONFIG_LEDS_TRIGGER_HEARTBEAT) += ledtrig-heartbeat.o
>> ?obj-$(CONFIG_LEDS_TRIGGER_BACKLIGHT) += ledtrig-backlight.o
>> ?obj-$(CONFIG_LEDS_TRIGGER_GPIO) ? ? ? ? ? ? ?+= ledtrig-gpio.o
>> +obj-$(CONFIG_LEDS_TRIGGER_CPU) ? ? ? ? ? ? ? += ledtrig-cpu.o
>> ?obj-$(CONFIG_LEDS_TRIGGER_DEFAULT_ON) ? ? ? ?+= ledtrig-default-on.o
>> diff --git a/drivers/leds/ledtrig-cpu.c b/drivers/leds/ledtrig-cpu.c
>> new file mode 100644
>> index 0000000..1f752e1
>> --- /dev/null
>> +++ b/drivers/leds/ledtrig-cpu.c
>> @@ -0,0 +1,150 @@
>> +/*
>> + * ledtrig-cpu.c - LED trigger based on CPU activity
>> + *
>> + * This LED trigger will be registered for each possible CPU and named as
>> + * cpu0, cpu1, cpu2, cpu3, etc.
>> + *
>> + * It can be bound to any LED just like other triggers using either a
>> + * board file or via sysfs interface.
>> + *
>> + * An API named ledtrig_cpu is exported for any user, who want to add CPU
>> + * activity indication in their code
>> + *
>> + * Copyright 2011 Linus Walleij <[email protected]>
>> + * Copyright 2011 - 2012 Bryan Wu <[email protected]>
>> + *
>> + * This program is free software; you can redistribute it and/or modify
>> + * it under the terms of the GNU General Public License version 2 as
>> + * published by the Free Software Foundation.
>> + *
>> + */
>> +
>> +#include <linux/module.h>
>> +#include <linux/kernel.h>
>> +#include <linux/init.h>
>> +#include <linux/slab.h>
>> +#include <linux/percpu.h>
>> +#include <linux/syscore_ops.h>
>> +#include <linux/rwsem.h>
>> +#include "leds.h"
>> +
>> +#define MAX_NAME_LEN 8
>> +
>> +static DEFINE_PER_CPU(struct led_trigger *, cpu_trig);
>> +static DEFINE_PER_CPU(char [MAX_NAME_LEN], trig_name);
>> +static DEFINE_PER_CPU(struct rw_semaphore, trig_lock);
>
> Semaphores are overkill for this. I think a simple 'struct mutex' is
> sufficient. Especially since there is unlikely to be much contention on
> any one of the locks. mutext_init(), mutext_lock(), mutex_unlock(), etc.
>

Indeed, I will change to mutex.

>> +
>> +/**
>> + * ledtrig_cpu - emit a CPU event as a trigger
>> + * @evt: CPU event to be emitted
>> + *
>> + * Emit a CPU event on a CPU core, which will trigger a
>> + * binded LED to turn on or turn off.
>> + */
>> +void ledtrig_cpu(enum cpu_led_event ledevt)
>> +{
>
> I think you still need to acquire trig_lock _before_ reading
> __get_cpu_var(cpu_trig). Otherwise, acquiring the lock in
> ledtrig_cpu_init() is useless.
>

I agree you are right, but call mutex_lock() here will prohibit
system booting, because mutex_lock() should be called after
mutex_init().
mutex_init() is called in module_init function, which is behind some
core functions calling ledtrig_cpu() API.

I plan to introduce a flag variable for checking before calling mutex_lock here.

>> + ? ? struct led_trigger *trig = __get_cpu_var(cpu_trig);
>> +
>> + ? ? if (!trig)
>> + ? ? ? ? ? ? return;
>
> Unnecessary since led_trigger_event() also checks for (!trig).
>
> I've attached the changes I think you should make.
>

Thanks a lot, I will update my patch and post it again.

Best Regards,
--
Bryan Wu <[email protected]>
Kernel Developer ? ?+86.186-168-78255 Mobile
Canonical Ltd. ? ? ?http://www.canonical.com
Ubuntu - Linux for human beings | http://www.ubuntu.com


2012-05-01 12:47:47

by Tim Gardner

[permalink] [raw]
Subject: Re: [PATCH 02/19] led-triggers: create a trigger for CPU activity

On 05/01/2012 05:45 AM, Bryan Wu wrote:
> On Mon, Apr 30, 2012 at 9:03 PM, Tim Gardner <[email protected]> wrote:
>> On 04/30/2012 01:37 AM, Bryan Wu wrote:
>>> 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.
>>>
>>> ([email protected]:
>>> It introduces several syscore stubs into this trigger.
>>> It also provides ledtrig_cpu trigger event stub in <linux/leds.h>.
>>> Although it was inspired by ARM work, it can be used in other arch.)
>>>
>>> Cc: Richard Purdie <[email protected]>
>>> Signed-off-by: Linus Walleij <[email protected]>
>>> Signed-off-by: Bryan Wu <[email protected]>
>>> Reviewed-by: Jamie Iles <[email protected]>
>>> Tested-by: Jochen Friedrich <[email protected]>
>>> ---
>>> drivers/leds/Kconfig | 10 +++
>>> drivers/leds/Makefile | 1 +
>>> drivers/leds/ledtrig-cpu.c | 150 ++++++++++++++++++++++++++++++++++++++++++++
>>> include/linux/leds.h | 16 +++++
>>> 4 files changed, 177 insertions(+)
>>> create mode 100644 drivers/leds/ledtrig-cpu.c
>>>
>>> diff --git a/drivers/leds/Kconfig b/drivers/leds/Kconfig
>>> index 5f12659..dbf8a4c 100644
>>> --- a/drivers/leds/Kconfig
>>> +++ b/drivers/leds/Kconfig
>>> @@ -465,6 +465,16 @@ config LEDS_TRIGGER_BACKLIGHT
>>>
>>> If unsure, say N.
>>>
>>> +config LEDS_TRIGGER_CPU
>>> + tristate "LED CPU Trigger"
>>> + depends on LEDS_TRIGGERS
>>> + help
>>> + This allows LEDs to be controlled by active CPUs. This shows
>>> + the active CPUs across an array of LEDs so you can see which
>>> + CPUs are active on the system at any given moment.
>>> +
>>> + If unsure, say N.
>>> +
>>> config LEDS_TRIGGER_GPIO
>>> tristate "LED GPIO Trigger"
>>> depends on LEDS_TRIGGERS
>>> diff --git a/drivers/leds/Makefile b/drivers/leds/Makefile
>>> index 9475bbb..ea1efb2 100644
>>> --- a/drivers/leds/Makefile
>>> +++ b/drivers/leds/Makefile
>>> @@ -56,4 +56,5 @@ obj-$(CONFIG_LEDS_TRIGGER_IDE_DISK) += ledtrig-ide-disk.o
>>> obj-$(CONFIG_LEDS_TRIGGER_HEARTBEAT) += ledtrig-heartbeat.o
>>> obj-$(CONFIG_LEDS_TRIGGER_BACKLIGHT) += ledtrig-backlight.o
>>> obj-$(CONFIG_LEDS_TRIGGER_GPIO) += ledtrig-gpio.o
>>> +obj-$(CONFIG_LEDS_TRIGGER_CPU) += ledtrig-cpu.o
>>> obj-$(CONFIG_LEDS_TRIGGER_DEFAULT_ON) += ledtrig-default-on.o
>>> diff --git a/drivers/leds/ledtrig-cpu.c b/drivers/leds/ledtrig-cpu.c
>>> new file mode 100644
>>> index 0000000..1f752e1
>>> --- /dev/null
>>> +++ b/drivers/leds/ledtrig-cpu.c
>>> @@ -0,0 +1,150 @@
>>> +/*
>>> + * ledtrig-cpu.c - LED trigger based on CPU activity
>>> + *
>>> + * This LED trigger will be registered for each possible CPU and named as
>>> + * cpu0, cpu1, cpu2, cpu3, etc.
>>> + *
>>> + * It can be bound to any LED just like other triggers using either a
>>> + * board file or via sysfs interface.
>>> + *
>>> + * An API named ledtrig_cpu is exported for any user, who want to add CPU
>>> + * activity indication in their code
>>> + *
>>> + * Copyright 2011 Linus Walleij <[email protected]>
>>> + * Copyright 2011 - 2012 Bryan Wu <[email protected]>
>>> + *
>>> + * This program is free software; you can redistribute it and/or modify
>>> + * it under the terms of the GNU General Public License version 2 as
>>> + * published by the Free Software Foundation.
>>> + *
>>> + */
>>> +
>>> +#include <linux/module.h>
>>> +#include <linux/kernel.h>
>>> +#include <linux/init.h>
>>> +#include <linux/slab.h>
>>> +#include <linux/percpu.h>
>>> +#include <linux/syscore_ops.h>
>>> +#include <linux/rwsem.h>
>>> +#include "leds.h"
>>> +
>>> +#define MAX_NAME_LEN 8
>>> +
>>> +static DEFINE_PER_CPU(struct led_trigger *, cpu_trig);
>>> +static DEFINE_PER_CPU(char [MAX_NAME_LEN], trig_name);
>>> +static DEFINE_PER_CPU(struct rw_semaphore, trig_lock);
>>
>> Semaphores are overkill for this. I think a simple 'struct mutex' is
>> sufficient. Especially since there is unlikely to be much contention on
>> any one of the locks. mutext_init(), mutext_lock(), mutex_unlock(), etc.
>>
>
> Indeed, I will change to mutex.
>
>>> +
>>> +/**
>>> + * ledtrig_cpu - emit a CPU event as a trigger
>>> + * @evt: CPU event to be emitted
>>> + *
>>> + * Emit a CPU event on a CPU core, which will trigger a
>>> + * binded LED to turn on or turn off.
>>> + */
>>> +void ledtrig_cpu(enum cpu_led_event ledevt)
>>> +{
>>
>> I think you still need to acquire trig_lock _before_ reading
>> __get_cpu_var(cpu_trig). Otherwise, acquiring the lock in
>> ledtrig_cpu_init() is useless.
>>
>
> I agree you are right, but call mutex_lock() here will prohibit
> system booting, because mutex_lock() should be called after
> mutex_init().
> mutex_init() is called in module_init function, which is behind some
> core functions calling ledtrig_cpu() API.
>

I think the mutex_lock is OK. If you follow the logic in
kernel/module.c:init_module(), the module init function must complete
_before_ consumers of external symbols (such as ledtrig_cpu) can
proceed. Really, that only makes sense. Otherwise module startup would
be impossibly racy.

Therefore, ledtrig_cpu() cannot be called before ledtrig_cpu_init()
completes.

rtg

--
Tim Gardner [email protected]