Re: [RESEND PATCH v2 13/33] ARM: davinci: aintc: use readl/writel_relaxed()

[Marc Zyngier <marc.zyngier@arm.com>]

On 11/02/2019 12:25, Bartosz Golaszewski wrote:
> From: Bartosz Golaszewski <bgolaszewski@baylibre.com>
> 
> Raplace all calls to __raw_readl() & __raw_writel() with readl_relaxed()
> and writel_relaxed() respectively. It's safe to do as there's no
> endianness conversion being done in the code.
> 
> Signed-off-by: Bartosz Golaszewski <bgolaszewski@baylibre.com>
> ---
>  arch/arm/mach-davinci/irq.c | 4 ++--
>  1 file changed, 2 insertions(+), 2 deletions(-)
> 
> diff --git a/arch/arm/mach-davinci/irq.c b/arch/arm/mach-davinci/irq.c
> index 148c23728bdb..6a7205a844e1 100644
> --- a/arch/arm/mach-davinci/irq.c
> +++ b/arch/arm/mach-davinci/irq.c
> @@ -36,12 +36,12 @@ static struct irq_domain *davinci_aintc_irq_domain;
>  
>  static inline void davinci_aintc_writel(unsigned long value, int offset)
>  {
> -	__raw_writel(value, davinci_aintc_base + offset);
> +	writel_relaxed(value, davinci_aintc_base + offset);
>  }
>  
>  static inline unsigned long davinci_aintc_readl(int offset)
>  {
> -	return __raw_readl(davinci_aintc_base + offset);
> +	return readl_relaxed(davinci_aintc_base + offset);
>  }
>  
>  static __init void
> 

Right. And that's exactly what I'd like to see in patch #3. There is no
point having a separate patch for this when you can do the right thing
from the beginning.

Thanks,

	M.
-- 
Jazz is not dead. It just smells funny...