Re: [PATCH net-next] __udp_enqueue_schedule_skb() drops packets when there is no buffer space available, but currently does not update UDP SNMP counters.

[Willem de Bruijn <willemdebruijn.kernel@gmail.com>]

Kshitiz Bartariya wrote:
> Update UDP_MIB_MEMERRORS and UDP_MIB_INERRORS when packets are dropped
> due to memory pressure, for both UDP and UDPLite sockets.
> 
> This removes a long-standing TODO and makes UDP statistics consistent
> with actual drop behavior.
> 
> Signed-off-by: Kshitiz Bartariya <kshitiz.bartariya@zohomail.in>

This is addressing the same open TODO as the patch under review

https://lore.kernel.org/netdev/20260105114732.140719-1-mahdifrmx@gmail.com/

It does so in the single basic block, as suggested here

https://lore.kernel.org/netdev/willemdebruijn.kernel.21c4d3b7b8f9d@gmail.com/

But these updates are expensive, so better to batch them as in the
other patch.

> ---
>  net/ipv4/udp.c | 7 ++++++-
>  1 file changed, 6 insertions(+), 1 deletion(-)
> 
> diff --git a/net/ipv4/udp.c b/net/ipv4/udp.c
> index 9c87067c74bc..66c06f468240 100644
> --- a/net/ipv4/udp.c
> +++ b/net/ipv4/udp.c
> @@ -1794,11 +1794,16 @@ int __udp_enqueue_schedule_skb(struct sock *sk, struct sk_buff *skb)
>  	}
>  
>  	if (unlikely(to_drop)) {
> +		const bool is_udplite = IS_UDPLITE(sk);
> +
>  		for (nb = 0; to_drop != NULL; nb++) {
>  			skb = to_drop;
>  			to_drop = skb->next;
>  			skb_mark_not_on_list(skb);
> -			/* TODO: update SNMP values. */
> +
> +			UDP_INC_STATS(sock_net(sk), UDP_MIB_MEMERRORS, is_udplite);
> +			UDP_INC_STATS(sock_net(sk), UDP_MIB_INERRORS, is_udplite);
> +
>  			sk_skb_reason_drop(sk, skb, SKB_DROP_REASON_PROTO_MEM);
>  		}
>  		numa_drop_add(&udp_sk(sk)->drop_counters, nb);
> -- 
> 2.50.1 (Apple Git-155)
>