Re: [PATCH] net: Only allow init netns to set default tcp cong to a restricted algo

[Jonathon Reinhart <jonathon.reinhart@gmail.com>]

On Sat, May 1, 2021 at 4:29 AM Jonathon Reinhart
<jonathon.reinhart@gmail.com> wrote:
>
> tcp_set_default_congestion_control() is netns-safe in that it writes
> to &net->ipv4.tcp_congestion_control, but it also sets
> ca->flags |= TCP_CONG_NON_RESTRICTED which is not namespaced.
> This has the unintended side-effect of changing the global
> net.ipv4.tcp_allowed_congestion_control sysctl, despite the fact that it
> is read-only: 97684f0970f6 ("net: Make tcp_allowed_congestion_control
> readonly in non-init netns")
>
> Resolve this netns "leak" by only allowing the init netns to set the
> default algorithm to one that is restricted. This restriction could be
> removed if tcp_allowed_congestion_control were namespace-ified in the
> future.
>
> This bug was uncovered with
> https://github.com/JonathonReinhart/linux-netns-sysctl-verify
>
> Fixes: 6670e1524477 ("tcp: Namespace-ify sysctl_tcp_default_congestion_control")
> Signed-off-by: Jonathon Reinhart <jonathon.reinhart@gmail.com>
> ---
>  net/ipv4/tcp_cong.c | 4 ++++
>  1 file changed, 4 insertions(+)
>
> diff --git a/net/ipv4/tcp_cong.c b/net/ipv4/tcp_cong.c
> index 563d016e7478..db5831e6c136 100644
> --- a/net/ipv4/tcp_cong.c
> +++ b/net/ipv4/tcp_cong.c
> @@ -230,6 +230,10 @@ int tcp_set_default_congestion_control(struct net *net, const char *name)
>                 ret = -ENOENT;
>         } else if (!bpf_try_module_get(ca, ca->owner)) {
>                 ret = -EBUSY;
> +       } else if (!net_eq(net, &init_net) &&
> +                       !(ca->flags & TCP_CONG_NON_RESTRICTED)) {
> +               /* Only init netns can set default to a restricted algorithm */
> +               ret = -EPERM;
>         } else {
>                 prev = xchg(&net->ipv4.tcp_congestion_control, ca);
>                 if (prev)
> --
> 2.20.1
>

This should be targeting "net" (and stable), not "net-next". Sorry about that.