[Patch net] nfc: use GFP_USER for user-controlled kmalloc

[Patch net] nfc: use GFP_USER for user-controlled kmalloc

[Cong Wang]

Reported-by: Dmitry Vyukov <dvyukov-hpIqsD4AKlfQT0dZR+AlfA@public.gmane.org>
Cc: Lauro Ramos Venancio <lauro.venancio-430g2QfJUUCGglJvpFV4uA@public.gmane.org>
Cc: Aloisio Almeida Jr <aloisio.almeida-430g2QfJUUCGglJvpFV4uA@public.gmane.org>
Cc: Samuel Ortiz <sameo-VuQAYsv1563Yd54FQh9/CA@public.gmane.org>
Signed-off-by: Cong Wang <xiyou.wangcong-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
---
 net/nfc/llcp_commands.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/net/nfc/llcp_commands.c b/net/nfc/llcp_commands.c
index 3621a90..5d94055 100644
--- a/net/nfc/llcp_commands.c
+++ b/net/nfc/llcp_commands.c
@@ -729,7 +729,7 @@ int nfc_llcp_send_ui_frame(struct nfc_llcp_sock *sock, u8 ssap, u8 dsap,
 	if (local == NULL)
 		return -ENODEV;
 
-	msg_data = kzalloc(len, GFP_KERNEL);
+	msg_data = kzalloc(len, GFP_USER | __GFP_NOWARN);
 	if (msg_data == NULL)
 		return -ENOMEM;
 
-- 
1.8.3.1

--
To unsubscribe from this list: send the line "unsubscribe linux-wireless" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[Patch net] nfc: check sock state in llcp_sock_getname()

[Cong Wang]

llcp_sock_getname() checks llcp_sock->dev to make sure
llcp_sock is already connected or bound, however, we could
be in the middle of llcp_sock_bind() where llcp_sock->dev
is bound and llcp_sock->service_name_len is set,
but llcp_sock->service_name is not, in this case we would
lead to copy some bytes from a NULL pointer.

We should just check if sk->sk_state is still closed since
both connect() and bind() will update this state at the end.

Reported-by: Dmitry Vyukov <dvyukov@google.com>
Cc: Lauro Ramos Venancio <lauro.venancio@openbossa.org>
Cc: Aloisio Almeida Jr <aloisio.almeida@openbossa.org>
Cc: Samuel Ortiz <sameo@linux.intel.com>
Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
---
 net/nfc/llcp_sock.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/net/nfc/llcp_sock.c b/net/nfc/llcp_sock.c
index ecf0a01..5a91997 100644
--- a/net/nfc/llcp_sock.c
+++ b/net/nfc/llcp_sock.c
@@ -500,7 +500,7 @@ static int llcp_sock_getname(struct socket *sock, struct sockaddr *uaddr,
 	struct nfc_llcp_sock *llcp_sock = nfc_llcp_sock(sk);
 	DECLARE_SOCKADDR(struct sockaddr_nfc_llcp *, llcp_addr, uaddr);
 
-	if (llcp_sock == NULL || llcp_sock->dev == NULL)
+	if (llcp_sock == NULL || sk->sk_state == LLCP_CLOSED)
 		return -EBADFD;
 
 	pr_debug("%p %d %d %d\n", sk, llcp_sock->target_idx,
-- 
1.8.3.1

Re: [Patch net] nfc: check sock state in llcp_sock_getname()

[Dmitry Vyukov]

On Sat, Jan 2, 2016 at 1:34 AM, Cong Wang <xiyou.wangcong-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org> wrote:
> llcp_sock_getname() checks llcp_sock->dev to make sure
> llcp_sock is already connected or bound, however, we could
> be in the middle of llcp_sock_bind() where llcp_sock->dev
> is bound and llcp_sock->service_name_len is set,
> but llcp_sock->service_name is not, in this case we would
> lead to copy some bytes from a NULL pointer.
>
> We should just check if sk->sk_state is still closed since
> both connect() and bind() will update this state at the end.

Hi Cong,

This is still racy. If you want to play lock-free then you also need
proper memory barriers. Stores to sk_state need to be
smp_store_release, while the load needs to be smp_load_acquire.
Otherwise getname still can see partially initialized socket.


> Reported-by: Dmitry Vyukov <dvyukov-hpIqsD4AKlfQT0dZR+AlfA@public.gmane.org>
> Cc: Lauro Ramos Venancio <lauro.venancio-430g2QfJUUCGglJvpFV4uA@public.gmane.org>
> Cc: Aloisio Almeida Jr <aloisio.almeida-430g2QfJUUCGglJvpFV4uA@public.gmane.org>
> Cc: Samuel Ortiz <sameo-VuQAYsv1563Yd54FQh9/CA@public.gmane.org>
> Signed-off-by: Cong Wang <xiyou.wangcong-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
> ---
>  net/nfc/llcp_sock.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/net/nfc/llcp_sock.c b/net/nfc/llcp_sock.c
> index ecf0a01..5a91997 100644
> --- a/net/nfc/llcp_sock.c
> +++ b/net/nfc/llcp_sock.c
> @@ -500,7 +500,7 @@ static int llcp_sock_getname(struct socket *sock, struct sockaddr *uaddr,
>         struct nfc_llcp_sock *llcp_sock = nfc_llcp_sock(sk);
>         DECLARE_SOCKADDR(struct sockaddr_nfc_llcp *, llcp_addr, uaddr);
>
> -       if (llcp_sock == NULL || llcp_sock->dev == NULL)
> +       if (llcp_sock == NULL || sk->sk_state == LLCP_CLOSED)
>                 return -EBADFD;
>
>         pr_debug("%p %d %d %d\n", sk, llcp_sock->target_idx,
> --
> 1.8.3.1
>
--
To unsubscribe from this list: send the line "unsubscribe linux-wireless" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [Patch net] nfc: check sock state in llcp_sock_getname()

[Cong Wang]

On Sat, Jan 2, 2016 at 12:03 AM, Dmitry Vyukov <dvyukov-hpIqsD4AKlfQT0dZR+AlfA@public.gmane.org> wrote:
> On Sat, Jan 2, 2016 at 1:34 AM, Cong Wang <xiyou.wangcong-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org> wrote:
>> llcp_sock_getname() checks llcp_sock->dev to make sure
>> llcp_sock is already connected or bound, however, we could
>> be in the middle of llcp_sock_bind() where llcp_sock->dev
>> is bound and llcp_sock->service_name_len is set,
>> but llcp_sock->service_name is not, in this case we would
>> lead to copy some bytes from a NULL pointer.
>>
>> We should just check if sk->sk_state is still closed since
>> both connect() and bind() will update this state at the end.
>
> Hi Cong,
>
> This is still racy. If you want to play lock-free then you also need
> proper memory barriers. Stores to sk_state need to be
> smp_store_release, while the load needs to be smp_load_acquire.
> Otherwise getname still can see partially initialized socket.
>

Right... Or just lock sock perhaps. I will update my patch.

Thanks!
--
To unsubscribe from this list: send the line "unsubscribe linux-wireless" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html