[PATCH 1/2] net: ethernet address comparison optimizations

[Bruce Richardson]

On Fri, Jan 30, 2026 at 10:46:16AM +0000, Morten Brørup wrote:
> For CPU architectures without strict alignment requirements, operations on
> 6-byte Ethernet addresses using three 2-byte operations were replaced by a
> 4-byte and a 2-byte operation, i.e. two operations instead of three.
> 
> Comparison functions are pure, so added __rte_pure.
> 
> Removed superfluous parentheses. (No functional change.)
> 
> Signed-off-by: Morten Brørup <mb at smartsharesystems.com>
> ---
>  lib/net/rte_ether.h | 19 ++++++++++++++++++-
>  1 file changed, 18 insertions(+), 1 deletion(-)
> 
> diff --git a/lib/net/rte_ether.h b/lib/net/rte_ether.h
> index c9a0b536c3..5552d3c1f6 100644
> --- a/lib/net/rte_ether.h
> +++ b/lib/net/rte_ether.h
> @@ -99,13 +99,19 @@ static_assert(alignof(struct rte_ether_addr) == 2,
>   *  True  (1) if the given two ethernet address are the same;
>   *  False (0) otherwise.
>   */
> +__rte_pure
>  static inline int rte_is_same_ether_addr(const struct rte_ether_addr *ea1,
>  				     const struct rte_ether_addr *ea2)
>  {
> +#if !defined(RTE_ARCH_STRICT_ALIGN)
> +	return ((((const unaligned_uint32_t *)ea1)[0] ^ ((const unaligned_uint32_t *)ea2)[0]) |
> +			(((const uint16_t *)ea1)[2] ^ ((const uint16_t *)ea2)[2])) == 0;
> +#else
>  	const uint16_t *w1 = (const uint16_t *)ea1;
>  	const uint16_t *w2 = (const uint16_t *)ea2;
>  
>  	return ((w1[0] ^ w2[0]) | (w1[1] ^ w2[1]) | (w1[2] ^ w2[2])) == 0;
> +#endif
>  }

Is this actually faster? For architectures that support strict alignment,
this looks like something that the compilers should be doing using proper
cost-benefit evaluation based on target architecture, rather than us doing
it in our code.

/Bruce