Skip to contentSkip to conversation

net: sock: add the case if sk is NULL

PATCH In review Yajun Deng 4 participants 6 messages Aug 6, 2021 – Aug 9, 2021

Subsystems (3)
lore
Share & tools
lore

Yajun Deng

author
#0 lore raw

PATCH 1/1 net: sock: add the case if sk is NULL

Add the case if sk is NULL in sock_{put, hold},
The caller is free to use it.

---
 include/net/sock.h | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)
include/net/sock.h +3 -2
index 6e761451c927..8821ec0d4147 100644
@@ -699,7 +699,8 @@ static inline bool __sk_del_node_init(struct sock *sk)
 
 static __always_inline void sock_hold(struct sock *sk)
 {
-	refcount_inc(&sk->sk_refcnt);
+	if (sk)
+		refcount_inc(&sk->sk_refcnt);
 }
 
 /* Ungrab socket in the context, which assumes that socket refcnt
@@ -1811,7 +1812,7 @@ void sock_init_data(struct socket *sock, struct sock *sk);
 /* Ungrab socket and destroy it, if it was the last reference. */
 static inline void sock_put(struct sock *sk)
 {
-	if (refcount_dec_and_test(&sk->sk_refcnt))
+	if (sk && refcount_dec_and_test(&sk->sk_refcnt))
 		sk_free(sk);
 }
 /* Generic version of sock_put(), dealing with all sockets
Signature · 1 line
2.32.0
maintainer
#1 lore raw
Quoted 4 lines · On Fri, 6 Aug 2021 14:38:15 +0800 Yajun Deng wrote:

Add the case if sk is NULL in sock_{put, hold},
The caller is free to use it.

The obvious complaint about this patch (and your previous netdev patch) is that you're spraying branches everywhere in the code. Sure, it may be okay for free(), given how expensive of an operation that is but is having refcounting functions accept NULL really the best practice?

Can you give us examples in the kernel where that's the case?

#2 lore raw
Quoted 9 lines · Jakub Kicinski
Quoted 4 lines · On Fri, 6 Aug 2021 14:38:15 +0800 Yajun Deng wrote:

Add the case if sk is NULL in sock_{put, hold},
The caller is free to use it.

The obvious complaint about this patch (and your previous netdev patch) is that you're spraying branches everywhere in the code. Sure, it may

Sorry for that, I'll be more normative in later submission.

Quoted 4 lines

be okay for free(), given how expensive of an operation that is but is having refcounting functions accept NULL really the best practice?

Can you give us examples in the kernel where that's the case?

0   include/net/neighbour.h         neigh_clone()
1   include/linux/cgroup.h          get_cgroup_ns() and put_cgroup_ns()  (This is very similar to my submission)
2   include/linux/ipc_namespace.h   get_ipc_ns()
3   include/linux/posix_acl.h       posix_acl_dup()
4   include/linux/pid.h             get_pid()
5   include/linux/user_namespace.h  get_user_ns()
#3 lore raw
Quoted 16 lines · On 8/9/21 8:12 AM, [email protected] wrote:
Quoted 9 lines · Jakub Kicinski
Quoted 4 lines · On Fri, 6 Aug 2021 14:38:15 +0800 Yajun Deng wrote:

Add the case if sk is NULL in sock_{put, hold},
The caller is free to use it.

The obvious complaint about this patch (and your previous netdev patch) is that you're spraying branches everywhere in the code. Sure, it may

Sorry for that, I'll be more normative in later submission.

Quoted 3 lines

be okay for free(), given how expensive of an operation that is but is having refcounting functions accept NULL really the best practice?

Quoted 9 lines
Quoted 1 line

Can you give us examples in the kernel where that's the case?

0   include/net/neighbour.h         neigh_clone()
1   include/linux/cgroup.h          get_cgroup_ns() and put_cgroup_ns()  (This is very similar to my submission)
2   include/linux/ipc_namespace.h   get_ipc_ns()
3   include/linux/posix_acl.h       posix_acl_dup()
4   include/linux/pid.h             get_pid()
5   include/linux/user_namespace.h  get_user_ns()

These helpers might be called with NULL pointers by design.

sock_put() and sock_hold() are never called with NULL.

Same for put_page() and hundreds of other functions.

By factorizing a conditional in the function, hoping to remove one in few callers, we add more conditional branches (and increase code size)

#4 lore raw
Quoted 29 lines · Eric Dumazet
Quoted 16 lines · On 8/9/21 8:12 AM, [email protected] wrote:
Quoted 9 lines · Jakub Kicinski
Quoted 4 lines · On Fri, 6 Aug 2021 14:38:15 +0800 Yajun Deng wrote:

Add the case if sk is NULL in sock_{put, hold},
The caller is free to use it.

The obvious complaint about this patch (and your previous netdev patch) is that you're spraying branches everywhere in the code. Sure, it may

Sorry for that, I'll be more normative in later submission.

Quoted 3 lines

be okay for free(), given how expensive of an operation that is but is having refcounting functions accept NULL really the best practice?

Quoted 9 lines
Quoted 1 line

Can you give us examples in the kernel where that's the case?

0   include/net/neighbour.h         neigh_clone()
1   include/linux/cgroup.h          get_cgroup_ns() and put_cgroup_ns()  (This is very similar to my submission)
2   include/linux/ipc_namespace.h   get_ipc_ns()
3   include/linux/posix_acl.h       posix_acl_dup()
4   include/linux/pid.h             get_pid()
5   include/linux/user_namespace.h  get_user_ns()
Quoted 8 lines

These helpers might be called with NULL pointers by design.

sock_put() and sock_hold() are never called with NULL.

Same for put_page() and hundreds of other functions.

By factorizing a conditional in the function, hoping to remove one in few callers, we add more conditional branches (and increase code size)

You can add into your list that such "if NULL" checks add false impression that NULL can be there and it is valid.

Thanks

Quoted 1 line
#5 lore raw
In reply to Yajun Deng
Quoted 3 lines · Yajun Deng

Add the case if sk is NULL in sock_{put, hold},
The caller is free to use it.

Can we please stop adding code like that all over the places ?

This is wrong, fix the callers that are lazy, or fix the real bug if this is a syzbot report.