Re: [PATCH bpf] bpf: Avoid soft lockup in __htab_map_lookup_and_delete_batch()

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Low] Multi-line comments do not follow BPF subsystem guidelines.
--

--- Patch [1]: [PATCH bpf] bpf: Avoid soft lockup in __htab_map_lookup_and_delete_batch() ---
commit a1114e030dd1f8d190db49effffca17e5da09d3f
Author: Jose Fernandez (Anthropic) <jose.fernandez@linux.dev>

bpf: Avoid soft lockup in __htab_map_lookup_and_delete_batch()

This commit limits the number of consecutive empty buckets visited in
__htab_map_lookup_and_delete_batch() before yielding. It drops locks and
calls cond_resched() to prevent soft lockup watchdogs from firing on
large sparse maps.

> diff --git a/kernel/bpf/hashtab.c b/kernel/bpf/hashtab.c
> --- a/kernel/bpf/hashtab.c
> +++ b/kernel/bpf/hashtab.c
> @@ -1769,6 +1769,11 @@ static int htab_lru_percpu_map_lookup_and_delete_elem(struct bpf_map *map,
>  						 flags);
>  }
>  
> +/* Max consecutive empty buckets to walk in one RCU +
> + * instrumentation-disabled section before rescheduling.
> + */
> +#define HTAB_BATCH_EMPTY_RESCHED 64

[Severity: Low]
This isn't a bug, but does this multi-line comment follow the preferred BPF
subsystem comment style? The guidelines specify that the opening /* should
be on its own line, rather than having text on the same line.

[ ... ]

> @@ -1969,11 +1975,19 @@ __htab_map_lookup_and_delete_batch(struct bpf_map *map,
>  
>  next_batch:
> -	/* If we are not copying data, we can go to next bucket and avoid
> -	 * unlocking the rcu.
> +	/* If we are not copying data, we can go to next bucket and avoid
> +	 * unlocking the rcu. Bound the walk though: after
> +	 * HTAB_BATCH_EMPTY_RESCHED consecutive empty buckets, fully exit
> +	 * the critical section (no locks are held here) and reschedule.
>  	 */

[Severity: Low]
Similarly, could this multi-line comment be updated to place the opening /*
on its own line to match the modern BPF comment style?

>  	if (!bucket_cnt && (batch + 1 < htab->n_buckets)) {
>  		batch++;
> -		goto again_nocopy;
> +		if (++empty_cnt < HTAB_BATCH_EMPTY_RESCHED)
> +			goto again_nocopy;

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260709-b4-htab-batch-resched-v1-1-ad7a6b3b4513@linux.dev?part=1