Re: [PATCH bpf-next v3 2/5] bpf: Add ksock kfuncs

On Tue, Aug 4, 2026 at 9:47 AM Mahe Tardy <[email protected]> wrote:
[...]
> +__bpf_kfunc int bpf_ksock_send(struct bpf_ksock *ks, const void *data,
> +                              u32 data__sz)
> +{
> +       struct msghdr msg = {
> +               .msg_flags = MSG_DONTWAIT,
> +       };
> +       struct kvec iov = {
> +               .iov_base = (void *)data,
> +               .iov_len = data__sz,
> +       };
> +       int ret;
> +
> +       if (!bpf_ksock_has_user_task_context())
> +               return -EOPNOTSUPP;
> +
> +       /* Early check for UDP. Exact limits enforced by kernel_sendmsg(). */
> +       if (data__sz > IP_MAX_MTU)
> +               return -EMSGSIZE;
> +
> +       if (current->in_bpf_ksock_send)
> +               return -EBUSY;

Returning EBUSY is not ideal. Can we avoid recursion by disallowing
calling bpf_ksock_send from certain hooks? You can find examples
check_kfunc_call() or check_special_kfunc(). This probably means
we need to grow special_kfunc_list, which is not ideal.

@Kumar, do you have strong objection that we grow special_kfunc_list
for this use case?

Thanks,
Song


> +
> +       current->in_bpf_ksock_send = true;
> +       ret = kernel_sendmsg(ks->sock, &msg, &iov, 1, data__sz);
> +       current->in_bpf_ksock_send = false;
> +
> +       return ret;
> +}
> +
> +__bpf_kfunc_end_defs();