Re: [PATCH 1/4] xfrm: Add support for per cpu xfrm state handling.

[Steffen Klassert <steffen.klassert@secunet.com>]

On Mon, Nov 11, 2024 at 09:42:02PM +0100, Kees Bakker wrote:
> Hi Steffen,
> 
> Sorry for the direct email. Did you perhaps forgot a "goto out_cancel" here?

Yes, looks like that. Do you want to send a patch?

> 
> diff --git a/net/xfrm/xfrm_user.c b/net/xfrm/xfrm_user.c
> [...]
> @@ -2576,6 +2603,8 @@ static int build_aevent(struct sk_buff *skb, struct
> xfrm_state *x, const struct
>      err = xfrm_if_id_put(skb, x->if_id);
>      if (err)
>          goto out_cancel;
> +    if (x->pcpu_num != UINT_MAX)
> +        err = nla_put_u32(skb, XFRMA_SA_PCPU, x->pcpu_num);
> 
>      if (x->dir) {
>          err = nla_put_u8(skb, XFRMA_SA_DIR, x->dir);
> 
> -- 
> Kees Bakker