netdev - RE: [RFC] vsock: notify server to shutdown when client has pending signal

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <09562f9b35c3419f9b5844b35b4276ae@huawei.com>
Date:   Mon, 17 May 2021 02:18:51 +0000
From:   "Longpeng (Mike, Cloud Infrastructure Service Product Dept.)" 
        <longpeng2@...wei.com>
To:     Stefano Garzarella <sgarzare@...hat.com>
CC:     "netdev@...r.kernel.org" <netdev@...r.kernel.org>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
        "Gonglei (Arei)" <arei.gonglei@...wei.com>,
        "Subo (Subo, Cloud Infrastructure Service Product Dept.)" 
        <subo7@...wei.com>, "David S . Miller" <davem@...emloft.net>,
        Jakub Kicinski <kuba@...nel.org>,
        Jorgen Hansen <jhansen@...are.com>,
        Norbert Slusarek <nslusarek@....net>,
        Andra Paraschiv <andraprs@...zon.com>,
        Colin Ian King <colin.king@...onical.com>,
        "David Brazdil" <dbrazdil@...gle.com>,
        Alexander Popov <alex.popov@...ux.com>,
        "lixianming (E)" <lixianming5@...wei.com>,
        "Longpeng (Mike, Cloud Infrastructure Service Product Dept.)" 
        <longpeng2@...wei.com>
Subject: RE: [RFC] vsock: notify server to shutdown when client has pending
 signal

Hi Stefano,

> -----Original Message-----
> From: Longpeng (Mike, Cloud Infrastructure Service Product Dept.)
> [mailto:longpeng2@...wei.com]
> Sent: Thursday, May 13, 2021 6:36 PM
> To: Stefano Garzarella <sgarzare@...hat.com>
> Cc: netdev@...r.kernel.org; linux-kernel@...r.kernel.org; Gonglei (Arei)
> <arei.gonglei@...wei.com>; Subo (Subo, Cloud Infrastructure Service Product
> Dept.) <subo7@...wei.com>; David S . Miller <davem@...emloft.net>; Jakub
> Kicinski <kuba@...nel.org>; Jorgen Hansen <jhansen@...are.com>; Norbert
> Slusarek <nslusarek@....net>; Andra Paraschiv <andraprs@...zon.com>;
> Colin Ian King <colin.king@...onical.com>; David Brazdil
> <dbrazdil@...gle.com>; Alexander Popov <alex.popov@...ux.com>;
> lixianming (E) <lixianming5@...wei.com>
> Subject: RE: [RFC] vsock: notify server to shutdown when client has pending
> signal
> 
> Hi Stefano,
> 
> > -----Original Message-----
> > From: Stefano Garzarella [mailto:sgarzare@...hat.com]
> > Sent: Thursday, May 13, 2021 5:42 PM
> > To: Longpeng (Mike, Cloud Infrastructure Service Product Dept.)
> > <longpeng2@...wei.com>
> > Cc: netdev@...r.kernel.org; linux-kernel@...r.kernel.org; Gonglei
> > (Arei) <arei.gonglei@...wei.com>; Subo (Subo, Cloud Infrastructure
> > Service Product
> > Dept.) <subo7@...wei.com>; David S . Miller <davem@...emloft.net>;
> > Jakub Kicinski <kuba@...nel.org>; Jorgen Hansen <jhansen@...are.com>;
> > Norbert Slusarek <nslusarek@....net>; Andra Paraschiv
> > <andraprs@...zon.com>; Colin Ian King <colin.king@...onical.com>;
> > David Brazdil <dbrazdil@...gle.com>; Alexander Popov
> > <alex.popov@...ux.com>; lixianming (E) <lixianming5@...wei.com>
> > Subject: Re: [RFC] vsock: notify server to shutdown when client has
> > pending signal
> >
> > Hi,
> > thanks for this patch, comments below...
> >
> > On Tue, May 11, 2021 at 05:41:27PM +0800, Longpeng(Mike) wrote:
> > >The client's sk_state will be set to TCP_ESTABLISHED if the server
> > >replay the client's connect request.
> > >However, if the client has pending signal, its sk_state will be set
> > >to TCP_CLOSE without notify the server, so the server will hold the
> > >corrupt connection.
> > >
> > >            client                        server
> > >
> > >1. sk_state=TCP_SYN_SENT         |
> > >2. call ->connect()              |
> > >3. wait reply                    |
> > >                                 | 4. sk_state=TCP_ESTABLISHED
> > >                                 | 5. insert to connected list
> > >                                 | 6. reply to the client
> > >7. sk_state=TCP_ESTABLISHED      |
> > >8. insert to connected list      |
> > >9. *signal pending* <--------------------- the user kill client
> > >10. sk_state=TCP_CLOSE           |
> > >client is exiting...             |
> > >11. call ->release()             |
> > >     virtio_transport_close
> > >      if (!(sk->sk_state == TCP_ESTABLISHED ||
> > >	      sk->sk_state == TCP_CLOSING))
> > >		return true; <------------- return at here As a result, the server
> > >cannot notice the connection is corrupt.
> > >So the client should notify the peer in this case.
> > >
> > >Cc: David S. Miller <davem@...emloft.net>
> > >Cc: Jakub Kicinski <kuba@...nel.org>
> > >Cc: Stefano Garzarella <sgarzare@...hat.com>
> > >Cc: Jorgen Hansen <jhansen@...are.com>
> > >Cc: Norbert Slusarek <nslusarek@....net>
> > >Cc: Andra Paraschiv <andraprs@...zon.com>
> > >Cc: Colin Ian King <colin.king@...onical.com>
> > >Cc: David Brazdil <dbrazdil@...gle.com>
> > >Cc: Alexander Popov <alex.popov@...ux.com>
> > >Signed-off-by: lixianming <lixianming5@...wei.com>
> > >Signed-off-by: Longpeng(Mike) <longpeng2@...wei.com>
> > >---
> > > net/vmw_vsock/af_vsock.c | 1 +
> > > 1 file changed, 1 insertion(+)
> > >
> > >diff --git a/net/vmw_vsock/af_vsock.c b/net/vmw_vsock/af_vsock.c
> > >index
> > >92a72f0..d5df908 100644
> > >--- a/net/vmw_vsock/af_vsock.c
> > >+++ b/net/vmw_vsock/af_vsock.c
> > >@@ -1368,6 +1368,7 @@ static int vsock_stream_connect(struct socket
> > >*sock,
> > struct sockaddr *addr,
> > > 		lock_sock(sk);
> > >
> > > 		if (signal_pending(current)) {
> > >+			vsock_send_shutdown(sk, SHUTDOWN_MASK);
> >
> > I see the issue, but I'm not sure is okay to send the shutdown in any
> > case, think about the server didn't setup the connection.
> >
> > Maybe is better to set TCP_CLOSING if the socket state was
> > TCP_ESTABLISHED, so the shutdown will be handled by the
> > transport->release() as usual.
> >
> > What do you think?
> >
> 
> Your method looks more gracefully, we'll try it and get back to you, thanks.
> 

As your suggestion, the following code can solve the problem:

                if (signal_pending(current)) {
                        err = sock_intr_errno(timeout);
-                       sk->sk_state = TCP_CLOSE;
+                       sk->sk_state = TCP_CLOSING;
                        sock->state = SS_UNCONNECTED;
                        vsock_transport_cancel_pkt(vsk);
                        goto out_wait;

This will send shutdown to the server even if the connection is not established, but
I don't see any side effects yet, right ?

The problem is also in the timeout case, we should fix it together ?


> > Anyway, also without the patch, the server should receive a RST if it
> > sends any data to the client, but of course, is better to let it know
> > the socket is closed in advance.
> >
> 
> Yes, agree.
> 
> > Thanks,
> > Stefano