[v1] sctp: Hold sock lock while iterating over address list

[PATCH net] sctp: Hold sock lock while iterating over address list

Posted by Stefan Wiehler 3 months, 1 week ago

Fixes: 8f840e47f190 ("sctp: add the sctp_diag.c file")
Signed-off-by: Stefan Wiehler <stefan.wiehler@nokia.com>
---
 net/sctp/diag.c | 10 ++++++----
 1 file changed, 6 insertions(+), 4 deletions(-)

diff --git a/net/sctp/diag.c b/net/sctp/diag.c
index 996c2018f0e6..7f7e2773e047 100644
--- a/net/sctp/diag.c
+++ b/net/sctp/diag.c
@@ -223,14 +223,15 @@ struct sctp_comm_param {
 	bool net_admin;
 };
 
-static size_t inet_assoc_attr_size(struct sctp_association *asoc)
+static size_t inet_assoc_attr_size(struct sock *sk,
+				   struct sctp_association *asoc)
 {
 	int addrlen = sizeof(struct sockaddr_storage);
 	int addrcnt = 0;
 	struct sctp_sockaddr_entry *laddr;
 
 	list_for_each_entry_rcu(laddr, &asoc->base.bind_addr.address_list,
-				list)
+				list, lockdep_sock_is_held(sk))
 		addrcnt++;
 
 	return	  nla_total_size(sizeof(struct sctp_info))
@@ -256,11 +257,12 @@ static int sctp_sock_dump_one(struct sctp_endpoint *ep, struct sctp_transport *t
 	if (err)
 		return err;
 
-	rep = nlmsg_new(inet_assoc_attr_size(assoc), GFP_KERNEL);
+	lock_sock(sk);
+
+	rep = nlmsg_new(inet_assoc_attr_size(sk, assoc), GFP_KERNEL);
 	if (!rep)
 		return -ENOMEM;
 
-	lock_sock(sk);
 	if (ep != assoc->ep) {
 		err = -EAGAIN;
 		goto out;
-- 
2.51.0

Re: [PATCH net] sctp: Hold sock lock while iterating over address list

Posted by Eric Dumazet 3 months, 1 week ago

We need a changelog.

On Mon, Oct 27, 2025 at 1:50 AM Stefan Wiehler <stefan.wiehler@nokia.com> wrote:
>
> Fixes: 8f840e47f190 ("sctp: add the sctp_diag.c file")
> Signed-off-by: Stefan Wiehler <stefan.wiehler@nokia.com>
> ---
>  net/sctp/diag.c | 10 ++++++----
>  1 file changed, 6 insertions(+), 4 deletions(-)
>
> diff --git a/net/sctp/diag.c b/net/sctp/diag.c
> index 996c2018f0e6..7f7e2773e047 100644
> --- a/net/sctp/diag.c
> +++ b/net/sctp/diag.c
> @@ -223,14 +223,15 @@ struct sctp_comm_param {
>         bool net_admin;
>  };
>
> -static size_t inet_assoc_attr_size(struct sctp_association *asoc)
> +static size_t inet_assoc_attr_size(struct sock *sk,
> +                                  struct sctp_association *asoc)
>  {
>         int addrlen = sizeof(struct sockaddr_storage);
>         int addrcnt = 0;
>         struct sctp_sockaddr_entry *laddr;
>
>         list_for_each_entry_rcu(laddr, &asoc->base.bind_addr.address_list,
> -                               list)
> +                               list, lockdep_sock_is_held(sk))
>                 addrcnt++;
>
>         return    nla_total_size(sizeof(struct sctp_info))
> @@ -256,11 +257,12 @@ static int sctp_sock_dump_one(struct sctp_endpoint *ep, struct sctp_transport *t
>         if (err)
>                 return err;
>
> -       rep = nlmsg_new(inet_assoc_attr_size(assoc), GFP_KERNEL);
> +       lock_sock(sk);
> +
> +       rep = nlmsg_new(inet_assoc_attr_size(sk, assoc), GFP_KERNEL);
>         if (!rep)
>                 return -ENOMEM;

If -ENOMEM is returned, the lock needs to be released ?

Please do not rush patches like this.