Re: [RFC 1/3] rv: add per-edge dwell-time statistics primitive

[Gabriele Monaco]

On Thu, 2026-08-27 at 09:23 +0200, Tobias Schaffner wrote:
> +     while (dwell_ns > (u64)max) {
> +             s64 prev = local64_cmpxchg(&s->max_ns, max, dwell_ns);
> +
> +             if (prev == max)
> +                     break;
> +             max = prev;
> +     }

try_cmpxchg() functions produce arguably cleaner code, would you mind doing:

  while (dwell_ns > (u64)max) {
      if (local64_try_cmpxchg(&s->max_ns, &max, dwell_ns))
          break;
  }

Also, although issues are practically impossible, I'd avoid an unbound loop in
hot paths, you could go up to MAX_DA_RETRY_RACING_EVENTS (already existing in
da_monitor) and put some warning in case you reach the maximum amount of
retries.

Thanks,
Gabriele