diff mbox series

[net,v3] net/tls: missing received data after fast remote close

Message ID 1605801588-12236-1-git-send-email-vfedorenko@novek.ru
State Superseded
Headers show
Series [net,v3] net/tls: missing received data after fast remote close | expand

Commit Message

Vadim Fedorenko Nov. 19, 2020, 3:59 p.m. UTC
In case when tcp socket received FIN after some data and the
parser haven't started before reading data caller will receive
an empty buffer. This behavior differs from plain TCP socket and
leads to special treating in user-space.
The flow that triggers the race is simple. Server sends small
amount of data right after the connection is configured to use TLS
and closes the connection. In this case receiver sees TLS Handshake
data, configures TLS socket right after Change Cipher Spec record.
While the configuration is in process, TCP socket receives small
Application Data record, Encrypted Alert record and FIN packet. So
the TCP socket changes sk_shutdown to RCV_SHUTDOWN and sk_flag with
SK_DONE bit set. The received data is not parsed upon arrival and is
never sent to user-space.

Patch unpauses parser directly if we have unparsed data in tcp
receive queue.

Signed-off-by: Vadim Fedorenko <vfedorenko@novek.ru>
---
 net/tls/tls_sw.c | 6 ++++++
 1 file changed, 6 insertions(+)

Comments

Jakub Kicinski Nov. 20, 2020, 6:26 p.m. UTC | #1
On Thu, 19 Nov 2020 18:59:48 +0300 Vadim Fedorenko wrote:
> In case when tcp socket received FIN after some data and the
> parser haven't started before reading data caller will receive
> an empty buffer. This behavior differs from plain TCP socket and
> leads to special treating in user-space.
> The flow that triggers the race is simple. Server sends small
> amount of data right after the connection is configured to use TLS
> and closes the connection. In this case receiver sees TLS Handshake
> data, configures TLS socket right after Change Cipher Spec record.
> While the configuration is in process, TCP socket receives small
> Application Data record, Encrypted Alert record and FIN packet. So
> the TCP socket changes sk_shutdown to RCV_SHUTDOWN and sk_flag with
> SK_DONE bit set. The received data is not parsed upon arrival and is
> never sent to user-space.
> 
> Patch unpauses parser directly if we have unparsed data in tcp
> receive queue.
> 
> Signed-off-by: Vadim Fedorenko <vfedorenko@novek.ru>

Applied, thanks!
Vadim Fedorenko Nov. 23, 2020, 1:40 p.m. UTC | #2
On 20.11.2020 18:26, Jakub Kicinski wrote:
> On Thu, 19 Nov 2020 18:59:48 +0300 Vadim Fedorenko wrote:
>> In case when tcp socket received FIN after some data and the
>> parser haven't started before reading data caller will receive
>> an empty buffer. This behavior differs from plain TCP socket and
>> leads to special treating in user-space.
>> The flow that triggers the race is simple. Server sends small
>> amount of data right after the connection is configured to use TLS
>> and closes the connection. In this case receiver sees TLS Handshake
>> data, configures TLS socket right after Change Cipher Spec record.
>> While the configuration is in process, TCP socket receives small
>> Application Data record, Encrypted Alert record and FIN packet. So
>> the TCP socket changes sk_shutdown to RCV_SHUTDOWN and sk_flag with
>> SK_DONE bit set. The received data is not parsed upon arrival and is
>> never sent to user-space.
>>
>> Patch unpauses parser directly if we have unparsed data in tcp
>> receive queue.
>>
>> Signed-off-by: Vadim Fedorenko <vfedorenko@novek.ru>
> Applied, thanks!
Looks like I missed fixes tag to queue this patch to -stable.

Fixes: c46234ebb4d1 ("tls: RX path for ktls")
Jakub Kicinski Nov. 23, 2020, 4:34 p.m. UTC | #3
On Mon, 23 Nov 2020 13:40:46 +0000 Vadim Fedorenko wrote:
> On 20.11.2020 18:26, Jakub Kicinski wrote:
> > On Thu, 19 Nov 2020 18:59:48 +0300 Vadim Fedorenko wrote:  
> >> In case when tcp socket received FIN after some data and the
> >> parser haven't started before reading data caller will receive
> >> an empty buffer. This behavior differs from plain TCP socket and
> >> leads to special treating in user-space.
> >> The flow that triggers the race is simple. Server sends small
> >> amount of data right after the connection is configured to use TLS
> >> and closes the connection. In this case receiver sees TLS Handshake
> >> data, configures TLS socket right after Change Cipher Spec record.
> >> While the configuration is in process, TCP socket receives small
> >> Application Data record, Encrypted Alert record and FIN packet. So
> >> the TCP socket changes sk_shutdown to RCV_SHUTDOWN and sk_flag with
> >> SK_DONE bit set. The received data is not parsed upon arrival and is
> >> never sent to user-space.
> >>
> >> Patch unpauses parser directly if we have unparsed data in tcp
> >> receive queue.
> >>
> >> Signed-off-by: Vadim Fedorenko <vfedorenko@novek.ru>  
> > Applied, thanks!  
> Looks like I missed fixes tag to queue this patch to -stable.
> 
> Fixes: c46234ebb4d1 ("tls: RX path for ktls")

I put this on:

Fixes: fcf4793e278e ("tls: check RCV_SHUTDOWN in tls_wait_data")

It's queued for stable, but it needs to hit Linus' tree first, so it'll
take another week or so to show up in stable releases.
diff mbox series

Patch

diff --git a/net/tls/tls_sw.c b/net/tls/tls_sw.c
index 2fe9e2c..845c628 100644
--- a/net/tls/tls_sw.c
+++ b/net/tls/tls_sw.c
@@ -1295,6 +1295,12 @@  static struct sk_buff *tls_wait_data(struct sock *sk, struct sk_psock *psock,
 			return NULL;
 		}
 
+		if (!skb_queue_empty(&sk->sk_receive_queue)) {
+			__strp_unpause(&ctx->strp);
+			if (ctx->recv_pkt)
+				return ctx->recv_pkt;
+		}
+
 		if (sk->sk_shutdown & RCV_SHUTDOWN)
 			return NULL;