Re: [PATCH v2 10/10] clocksource: import ARC timer driver

[Vineet Gupta]

On 11/03/2016 04:01 PM, Daniel Lezcano wrote:
> On Thu, Nov 03, 2016 at 03:50:21PM -0700, Vineet Gupta wrote:
>> On 11/03/2016 03:38 PM, Daniel Lezcano wrote:
>>> On Thu, Nov 03, 2016 at 02:31:41PM -0700, Vineet Gupta wrote:
>>>> This adds support for
>>>>
>>>>  - CONFIG_ARC_TIMERS : legacy 32-bit TIMER0 and TIMER1 which count UP
>>>>    from @CNT to @LIMIT, before optionally triggering an interrupt.
>>>>    These are programmed using ARC auxiliary register interface.
>>>>    These are present in all ARC cores (ARC700 and ARC HS38)
>>>>    TIMER0 serves as clockevent for all ARC linux builds.
>>>>    TIMER1 is used for clocksource in arc700 builds.
>>>>
>>>>  - CONFIG_ARC_TIMERS_64BIT: 64-bit counters, RTC and GFRC found in
>>>>    ARC HS38 cores. These are independnet IP blocks with different
>>>>    programming model respectively.
>>>>
>>>> Signed-off-by: Vineet Gupta <vgupta@xxxxxxxxxxxx>
>>>> ---
>>>
>>> [ ... ]
>>>
>>>>  #include <linux/of_irq.h>
>>>> -#include <asm/irq.h>
>>>>  
>>>>  #include <soc/arc/timers.h>
>>>>  #include <soc/arc/mcip.h>
>>>> @@ -263,7 +248,7 @@ static irqreturn_t timer_irq_handler(int irq, void *dev_id)
>>>>  	 * irq_set_chip_and_handler() asked for handle_percpu_devid_irq()
>>>>  	 */
>>>>  	struct clock_event_device *evt = this_cpu_ptr(&arc_clockevent_device);
>>>> -	int irq_reenable = clockevent_state_periodic(evt);
>>>> +	int irq_reenable __maybe_unused = clockevent_state_periodic(evt);
>>>
>>> Why is needed __maybe_unused ? I see in the previous driver 'irq_reenable' is
>>> used or is there a change in the previous patches I missed ?
>>
>> This is needed when not building for CONFIG_ARC (saw this when building for ARM)
>> write_aux_reg() becomes a no-op which causes a warning:
>>
>> 	write_aux_reg(ARC_REG_TIMER0_CTRL, irq_reenable | TIMER_CTRL_NH);
> 
> Instead of adding the __maybe_unused, changing in patch 7/10:
> 
> #define read_aux_reg(r) 0
> #define write_aux_reg(r, v)
> 
> by
> 
> static inline int read_aux_reg(void *)
> {
> 	return 0;
> }
> 
> static inline void write_aux_reg(void *, u32)
> {
> 	;
> }
> 
> Should fix the warning.

Good point, slight mod preferred as @reg argument is not really a MMIO register so
not a pointer but a number instead so I'd prefer u32 for that as well.