Re: [PATCH v5 5/6] arm64: move ARM64_HAS_CACHE_DIC/_IDC from asm to C

[Will Deacon]

On Thu, Jan 02, 2020 at 04:13:56PM -0500, Pavel Tatashin wrote:
> The assmbly functions __asm_flush_cache_user_range and
> __asm_invalidate_icache_range have alternatives:
> 
> alternative_if ARM64_HAS_CACHE_DIC
> ...
> 
> alternative_if ARM64_HAS_CACHE_IDC
> ...
> 
> But, the implementation of those alternatives is trivial and therefore
> can be done in the C inline wrappers.
> 
> Signed-off-by: Pavel Tatashin <pasha.tatashin@xxxxxxxxxx>
> ---
>  arch/arm64/include/asm/cacheflush.h | 19 +++++++++++++++++++
>  arch/arm64/mm/cache.S               | 27 +++++----------------------
>  arch/arm64/mm/flush.c               |  1 +
>  3 files changed, 25 insertions(+), 22 deletions(-)
> 
> diff --git a/arch/arm64/include/asm/cacheflush.h b/arch/arm64/include/asm/cacheflush.h
> index 047af338ba15..fc5217a18398 100644
> --- a/arch/arm64/include/asm/cacheflush.h
> +++ b/arch/arm64/include/asm/cacheflush.h
> @@ -77,8 +77,22 @@ static inline long __flush_cache_user_range(unsigned long start,
>  {
>  	int ret;
>  
> +	if (cpus_have_const_cap(ARM64_HAS_CACHE_IDC)) {
> +		dsb(ishst);
> +		if (cpus_have_const_cap(ARM64_HAS_CACHE_DIC)) {
> +			isb();
> +			return 0;
> +		}
> +	}
> +
>  	uaccess_ttbr0_enable();
>  	ret = __asm_flush_cache_user_range(start, end);

I don't understand this. Doesn't it mean a CPU with IDC but not DIC will
end up with doing the D-cache maintenance?

Will