key might contain private information, so better use
kfree_sensitive to free it.
In tcp_md5_twsk_free_rcu() use kfree_sensitive().

Signed-off-by: Minjie Du <duminjie@vivo.com>
---
 net/ipv4/tcp_minisocks.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/net/ipv4/tcp_minisocks.c b/net/ipv4/tcp_minisocks.c
index a9807eeb311c..a7be78096783 100644
--- a/net/ipv4/tcp_minisocks.c
+++ b/net/ipv4/tcp_minisocks.c
@@ -368,7 +368,7 @@ static void tcp_md5_twsk_free_rcu(struct rcu_head *head)
 	struct tcp_md5sig_key *key;
 
 	key = container_of(head, struct tcp_md5sig_key, rcu);
-	kfree(key);
+	kfree_sensitive(key);
 	static_branch_slow_dec_deferred(&tcp_md5_needed);
 	tcp_md5_release_sigpool();
 }
-- 
2.39.0

On Tue, Nov 7, 2023 at 3:35 AM Minjie Du <duminjie@vivo.com> wrote:
>
> key might contain private information, so better use
> kfree_sensitive to free it.
> In tcp_md5_twsk_free_rcu() use kfree_sensitive().
>
> Signed-off-by: Minjie Du <duminjie@vivo.com>
> ---
>  net/ipv4/tcp_minisocks.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/net/ipv4/tcp_minisocks.c b/net/ipv4/tcp_minisocks.c
> index a9807eeb311c..a7be78096783 100644
> --- a/net/ipv4/tcp_minisocks.c
> +++ b/net/ipv4/tcp_minisocks.c
> @@ -368,7 +368,7 @@ static void tcp_md5_twsk_free_rcu(struct rcu_head *head)
>         struct tcp_md5sig_key *key;
>
>         key = container_of(head, struct tcp_md5sig_key, rcu);
> -       kfree(key);
> +       kfree_sensitive(key);
>         static_branch_slow_dec_deferred(&tcp_md5_needed);
>         tcp_md5_release_sigpool();
>  }
> --
> 2.39.0
>



1) net-next is currently closed.

2) such patch could send a wrong signal (false sense of security with MD5)

3) You forgot tcp_time_wait_init(), tcp_md5_do_del(), tcp_md5_key_copy(),
    tcp_md5_do_add(), tcp_clear_md5_list().

More work is needed I am afraid, please wait until next week when
6-7-rc1 is tagged and net-next opens again.