Re: [PATCH 1/1] prefetch result for ARM64 atomic ops

[Joe Perches <joe@perches.com>]

On Mon, 2014-04-14 at 14:11 -0400, Pranith Kumar wrote:
> On Mon, Apr 14, 2014 at 2:04 PM, Joe Perches <joe@perches.com> wrote:
> > Are there some benchmark comparison results
> > you neglected to attach?
> 
> I was trying to get these ops to be similar to the 32 bit atomic ops which
> all have prefetch. I did not see any reason why we shouldn't do the same
> here. But no, no hard numbers yet.

Then you should probably send this as an RFC patch.

> >> diff --git a/arch/arm64/include/asm/atomic.h b/arch/arm64/include/asm/atomic.h
> >> +    prefetchw(&v->counter);
> >
> > Does this compile?

> Argh, my mistake! I overlooked this. Please find fixed patch below:
[]
> diff --git a/arch/arm64/include/asm/atomic.h b/arch/arm64/include/asm/atomic.h
[]
> @@ -235,6 +244,7 @@ static inline long atomic64_cmpxchg(atomic64_t
> *ptr, long old, long new)
>      unsigned long res;
> 
>      smp_mb();
> +    prefetchw(&v->counter);

It seems you sent the original patch again, still not compiled.

An arm64 cross-compiler is available here:
http://linuxdriverproject.org/mediawiki/index.php/Cross-compiling_Linux_kernel_on_x86_64#Installing_arm64_cross-compiler

It'd be nice if an arm64 toolchain was on the kernel.org list too
https://www.kernel.org/pub/tools/crosstool/