Currently vsock_connectible_has_data() may miss a wakeup operation
between vsock_connectible_has_data() == 0 and the prepare_to_wait().

Fix the race by adding the process to the wait qeuue before checking
vsock_connectible_has_data().

Fixes: b3f7fd54881b ("af_vsock: separate wait data loop")
Signed-off-by: Dexuan Cui <decui@microsoft.com>
---
 net/vmw_vsock/af_vsock.c | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)

diff --git a/net/vmw_vsock/af_vsock.c b/net/vmw_vsock/af_vsock.c
index d258fd43092e..03a6b5bc6ba7 100644
--- a/net/vmw_vsock/af_vsock.c
+++ b/net/vmw_vsock/af_vsock.c
@@ -1905,8 +1905,11 @@ static int vsock_connectible_wait_data(struct sock *sk,
 	err = 0;
 	transport = vsk->transport;
 
-	while ((data = vsock_connectible_has_data(vsk)) == 0) {
+	while (1) {
 		prepare_to_wait(sk_sleep(sk), wait, TASK_INTERRUPTIBLE);
+		data = vsock_connectible_has_data(vsk);
+		if (data != 0)
+			break;
 
 		if (sk->sk_err != 0 ||
 		    (sk->sk_shutdown & RCV_SHUTDOWN) ||
@@ -1937,6 +1940,8 @@ static int vsock_connectible_wait_data(struct sock *sk,
 			err = -EAGAIN;
 			break;
 		}
+
+		finish_wait(sk_sleep(sk), wait);
 	}
 
 	finish_wait(sk_sleep(sk), wait);
-- 
2.25.1

On Fri, Oct 28, 2022 at 01:56:46PM -0700, Dexuan Cui wrote:
>Currently vsock_connectible_has_data() may miss a wakeup operation
>between vsock_connectible_has_data() == 0 and the prepare_to_wait().
>
>Fix the race by adding the process to the wait qeuue before checking

s/qeuue/queue

>vsock_connectible_has_data().
>
>Fixes: b3f7fd54881b ("af_vsock: separate wait data loop")
>Signed-off-by: Dexuan Cui <decui@microsoft.com>
>---
> net/vmw_vsock/af_vsock.c | 7 ++++++-
> 1 file changed, 6 insertions(+), 1 deletion(-)
>
>diff --git a/net/vmw_vsock/af_vsock.c b/net/vmw_vsock/af_vsock.c
>index d258fd43092e..03a6b5bc6ba7 100644
>--- a/net/vmw_vsock/af_vsock.c
>+++ b/net/vmw_vsock/af_vsock.c
>@@ -1905,8 +1905,11 @@ static int vsock_connectible_wait_data(struct sock *sk,
> 	err = 0;
> 	transport = vsk->transport;
>
>-	while ((data = vsock_connectible_has_data(vsk)) == 0) {
>+	while (1) {
> 		prepare_to_wait(sk_sleep(sk), wait, TASK_INTERRUPTIBLE);
>+		data = vsock_connectible_has_data(vsk);
>+		if (data != 0)
>+			break;
>
> 		if (sk->sk_err != 0 ||
> 		    (sk->sk_shutdown & RCV_SHUTDOWN) ||
>@@ -1937,6 +1940,8 @@ static int vsock_connectible_wait_data(struct sock *sk,
> 			err = -EAGAIN;
> 			break;
> 		}
>+
>+		finish_wait(sk_sleep(sk), wait);

Since we are going to call again prepare_to_wait() on top of the loop, 
is finish_wait() call here really needed?

What about following what we do in vsock_accept and vsock_connect?

     prepare_to_wait()

     while (condition) {
         ...
         prepare_to_wait();
     }

     finish_wait()

I find it a little more readable, but your solution is fine too.

Thanks,
Stefano

> From: Stefano Garzarella <sgarzare@redhat.com>
> Sent: Monday, October 31, 2022 1:43 AM
>  ...
> s/qeuue/queue
Will fix this.
 
> >@@ -1905,8 +1905,11 @@ static int vsock_connectible_wait_data(struct
> sock *sk,
> > 	err = 0;
> > 	transport = vsk->transport;
> >
> >-	while ((data = vsock_connectible_has_data(vsk)) == 0) {
> >+	while (1) {
> > 		prepare_to_wait(sk_sleep(sk), wait, TASK_INTERRUPTIBLE);
> >+		data = vsock_connectible_has_data(vsk);
> >+		if (data != 0)
> >+			break;
> >
> > 		if (sk->sk_err != 0 ||
> > 		    (sk->sk_shutdown & RCV_SHUTDOWN) ||
> >@@ -1937,6 +1940,8 @@ static int vsock_connectible_wait_data(struct sock
> *sk,
> > 			err = -EAGAIN;
> > 			break;
> > 		}
> >+
> >+		finish_wait(sk_sleep(sk), wait);
> 
> Since we are going to call again prepare_to_wait() on top of the loop,
> is finish_wait() call here really needed?

It's not needed. Will remove this and send v2.

> What about following what we do in vsock_accept and vsock_connect?
> 
>      prepare_to_wait()
> 
>      while (condition) {
>          ...
>          prepare_to_wait();
>      }
> 
>      finish_wait()
> 
> I find it a little more readable, but your solution is fine too.
> 
> Thanks,
> Stefano

I'd like to stay with my version, as it only needs one line of
prepare_to_wait(), and IMO it's more readable if we only exit from
inside the while loop.