Re: [PATCH] clocksource: add __ro_after_init to cyclecounter

[Ard Biesheuvel]

On 11 February 2017 at 19:20, Bhumika Goyal <bhumirks@xxxxxxxxx> wrote:
> The object cyclecounter of type cyclecounter is not getting modified
> after getting initialized by arch_counter_register. Apart from
> initialization in arch_counter_register it is also passed as an argument
> to the function timecounter_init but this argument is of type const.
> Therefore, add __ro_after_init to its declaration.
>

I think adding __ro_after_init is fine if this struct is never
modified after init. But the reference in the commit log to the
constness of the timecounter_init() argument  makes no sense: that
only means timecounter_init() will not modify the object, which allows
pointers to const objects to be passed to it as well. The opposite is
not true, though: there is no requirement whatsoever that objects
passed into const pointer arguments should be const themselves.


> Signed-off-by: Bhumika Goyal <bhumirks@xxxxxxxxx>
> ---
>  drivers/clocksource/arm_arch_timer.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/clocksource/arm_arch_timer.c b/drivers/clocksource/arm_arch_timer.c
> index 4c8c3fb..a10506b 100644
> --- a/drivers/clocksource/arm_arch_timer.c
> +++ b/drivers/clocksource/arm_arch_timer.c
> @@ -580,7 +580,7 @@ static u64 arch_counter_read_cc(const struct cyclecounter *cc)
>         .flags  = CLOCK_SOURCE_IS_CONTINUOUS,
>  };
>
> -static struct cyclecounter cyclecounter = {
> +static struct cyclecounter cyclecounter __ro_after_init = {
>         .read   = arch_counter_read_cc,
>         .mask   = CLOCKSOURCE_MASK(56),
>  };
> --
> 1.9.1
>
>
> _______________________________________________
> linux-arm-kernel mailing list
> linux-arm-kernel@xxxxxxxxxxxxxxxxxxx
> http://lists.infradead.org/mailman/listinfo/linux-arm-kernel