netdev - Re: [PATCH net-next] rtnetlink: return ENODEV when ifname does not exist and group is given

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20220331203637.004709d4@kernel.org>
Date:   Thu, 31 Mar 2022 20:36:37 -0700
From:   Jakub Kicinski <kuba@...nel.org>
To:     Florent Fourcot <florent.fourcot@...irst.fr>,
        netdev@...r.kernel.org
Cc:     Brian Baboch <brian.baboch@...irst.fr>
Subject: Re: [PATCH net-next] rtnetlink: return ENODEV when ifname does not
 exist and group is given

On Thu, 31 Mar 2022 14:35:02 +0200 Florent Fourcot wrote:
> From: Florent Fourcot <florent.fourcot@...irst.fr>
> To: netdev@...r.kernel.org
> Cc: Florent Fourcot <florent.fourcot@...irst.fr>, Brian Baboch <brian.baboch@...irst.fr>

Please run the patches thru ./scripts/get_maintainer.pl.
netdev is high volume (and Gmail hates it), we really need people 
to CC potential reviewers.

> When the interface does not exist, and a group is given, the given
> parameters are being set to all interfaces of the given group. The given
> IFNAME/ALT_IF_NAME are being ignored in that case.
> 
> That can be dangerous since a typo (or a deleted interface) can produce
> weird side effects for caller:
> 
> Case 1:
> 
>  IFLA_IFNAME=valid_interface
>  IFLA_GROUP=1
>  MTU=1234
> 
> Case 1 will update MTU and group of the given interface "valid_interface".
> 
> Case 2:
> 
>  IFLA_IFNAME=doesnotexist
>  IFLA_GROUP=1
>  MTU=1234
> 
> Case 2 will update MTU of all interfaces in group 1. IFLA_IFNAME is
> ignored in this case
> 
> This behaviour is not consistent and dangerous. In order to fix this issue,
> we now return ENODEV when the given IFNAME does not exist.

> diff --git a/net/core/rtnetlink.c b/net/core/rtnetlink.c
> index 159c9c61e6af..3313419bbcba 100644
> --- a/net/core/rtnetlink.c
> +++ b/net/core/rtnetlink.c
> @@ -3417,10 +3417,15 @@ static int __rtnl_newlink(struct sk_buff *skb, struct nlmsghdr *nlh,
>  	}
>  
>  	if (!(nlh->nlmsg_flags & NLM_F_CREATE)) {
> -		if (ifm->ifi_index == 0 && tb[IFLA_GROUP])
> +		if (ifm->ifi_index == 0 && tb[IFLA_GROUP]) {
> +			if (tb[IFLA_IFNAME] || tb[IFLA_ALT_IFNAME]) {
> +				NL_SET_ERR_MSG(extack, "Cannot find interface with given name");
> +				return -ENODEV;
> +			}
>  			return rtnl_group_changelink(skb, net,
>  						nla_get_u32(tb[IFLA_GROUP]),
>  						ifm, extack, tb);
> +		}
>  		return -ENODEV;
>  	}

Would it be slightly cleaner to have a similar check in
validate_linkmsg()? Something like:

	if (!dev && !ifm->ifi_index && tb[IFLA_GROUP] &&
	    (tb[IFLA_IFNAME] || tb[IFLA_ALT_IFNAME]))