Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Date:   Wed, 2 Nov 2022 14:39:53 -0700
From:   Jakub Kicinski <kuba@kernel.org>
To:     Tao Chen <chentao.kernel@linux.alibaba.com>
Cc:     "David S. Miller" <davem@davemloft.net>,
        Eric Dumazet <edumazet@google.com>,
        Paolo Abeni <pabeni@redhat.com>,
        Johannes Berg <johannes@sipsolutions.net>,
        Oliver Hartkopp <socketcan@hartkopp.net>,
        Petr Machata <petrm@nvidia.com>,
        Kees Cook <keescook@chromium.org>,
        Harshit Mogalapalli <harshit.m.mogalapalli@oracle.com>,
        netdev@vger.kernel.org, linux-kernel@vger.kernel.org
Subject: Re: [PATCH net-next] netlink: Fix potential skb memleak in
 netlink_ack
Message-ID: <20221102143953.001f1247@kernel.org>
In-Reply-To: <7a382b9503d10d235238ca55938bc933d92a1de7.1667389213.git.chentao.kernel@linux.alibaba.com>
References: <7a382b9503d10d235238ca55938bc933d92a1de7.1667389213.git.chentao.kernel@linux.alibaba.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Precedence: bulk

On Wed,  2 Nov 2022 20:08:20 +0800 Tao Chen wrote:
> We should clean the skb resource if nlmsg_put/append failed
> , so fix it.

The comma should be at the end of the previous line.
But really the entire ", so fix it." is redundant.

> Fiexs: commit 738136a0e375 ("netlink: split up copies in the
> ack construction")

Please look around to see how to correctly format a Fixes tag
(including not line wrapping it).

How did you find this bug? An automated tool? Syzbot?

One more note below on the code itself.

> Signed-off-by: Tao Chen <chentao.kernel@linux.alibaba.com>
> ---
>  net/netlink/af_netlink.c | 4 +++-
>  1 file changed, 3 insertions(+), 1 deletion(-)
> 
> diff --git a/net/netlink/af_netlink.c b/net/netlink/af_netlink.c
> index c6b8207e..9d73dae 100644
> --- a/net/netlink/af_netlink.c
> +++ b/net/netlink/af_netlink.c
> @@ -2500,7 +2500,7 @@ void netlink_ack(struct sk_buff *in_skb, struct nlmsghdr *nlh, int err,
>  
>  	skb = nlmsg_new(payload + tlvlen, GFP_KERNEL);
>  	if (!skb)
> -		goto err_bad_put;
> +		goto err_skb;
>  
>  	rep = nlmsg_put(skb, NETLINK_CB(in_skb).portid, nlh->nlmsg_seq,
>  			NLMSG_ERROR, sizeof(*errmsg), flags);
> @@ -2528,6 +2528,8 @@ void netlink_ack(struct sk_buff *in_skb, struct nlmsghdr *nlh, int err,
>  	return;
>  
>  err_bad_put:
> +	kfree_skb(skb);

Please use nlmsg_free() since we allocated with nlmsg_new().

> +err_skb:
>  	NETLINK_CB(in_skb).sk->sk_err = ENOBUFS;
>  	sk_error_report(NETLINK_CB(in_skb).sk);
>  }