Re: [PATCH net-next v11] net: reduce ARFS flow updates by checking LLC affinity

[Jakub Kicinski]

On Mon, 12 Oct 2026 14:01:08 +0800 Chuang Wang wrote:
> +		/*

nit: empty line not needed in networking, start with the "title":

	/* ARFS LLC Affinity Feature:

> +		 * ARFS LLC Affinity Feature:
> +		 * Reduce ARFS flow updates by checking LLC affinity.
> +		 *
> +		 * Frequent flow table updates can trigger constant hardware steering
> +		 * reconfigurations (e.g., ndo_rx_flow_steer), leading to significant
> +		 * contention on driver internal locks (like mlx5's arfs_lock).
> +		 *
> +		 * Skip the expensive hardware steering update if the flow's previous
> +		 * target CPU and the new one share the last level cache: cache locality
> +		 * for the application is preserved either way.
> +		 */

Honestly, I don't think this comment is needed at all. The check 
is pretty understandable to anyone with any system config experience.

> +		if (static_branch_unlikely(&arfs_llc_affinity) &&
> +		    old_cpu < nr_cpu_ids && cpu_online(old_cpu) &&
> +			cpus_share_cache(old_cpu, next_cpu))
> +			goto out;

The last line of the condition is mis-aligned 

> +	{
> +		.procname	= "arfs_llc_affinity",
> +		.data		= &arfs_llc_affinity.key,
> +		.maxlen		= sizeof(arfs_llc_affinity.key),
> +		.mode		= 0644,
> +		.proc_handler	= proc_do_static_key
> +	},

You're adding a uAPI / profcs knob. It needs to be documented
appropriately under Documentation/

Please note that net next is closed:
https://netdev.bots.linux.dev/net-next.html
-- 
pw-bot: cr