[Valgrind-developers] [PATCH 1/2] x86: optimize XCHG to MOV for same-register forms

[H.J. Lu]

On Tue, Jul 7, 2026 at 1:53 PM Jan Beulich <jbeulich@suse.com> wrote:
>
> On 06.07.2026 23:12, H.J. Lu wrote:
> > We should consider benefits vs drawbacks for this optimization.  Do we have
> > a list of applications which benefit from this optimization?
>
> This question was raised before (in different words), and the answer is still
> the same - how could we possibly know? The same is true though for about any
> other of the optimizations we meanwhile have. IOW to me that's _still_ not a
> viable criteria to work from.

For assembler optimizations, we can give them benefits of doubt.   Since we
now know there is a significant drawback, its benefits should outweigh its
drawbacks to enable this optimization.   We can't assume that this optimization
improves anything without evidence.

-- 
H.J.