netdev - Re: [PATCH net] virtio/vsock: fix sock refcnt bug on owner set failure

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: <ZHbAgkvSHEiQlFs6@bullseye>
Date: Wed, 31 May 2023 03:35:30 +0000
From: Bobby Eshleman <bobbyeshleman@...il.com>
To: Stefano Garzarella <sgarzare@...hat.com>
Cc: Bobby Eshleman <bobby.eshleman@...edance.com>, kvm@...r.kernel.org,
	netdev@...r.kernel.org, linux-kernel@...r.kernel.org,
	virtualization@...ts.linux-foundation.org,
	Eric Dumazet <edumazet@...gle.com>,
	Stefan Hajnoczi <stefanha@...hat.com>,
	Jakub Kicinski <kuba@...nel.org>, Paolo Abeni <pabeni@...hat.com>,
	"David S. Miller" <davem@...emloft.net>
Subject: Re: [PATCH net] virtio/vsock: fix sock refcnt bug on owner set
 failure

On Thu, Jun 01, 2023 at 09:58:47AM +0200, Stefano Garzarella wrote:
> On Wed, May 31, 2023 at 07:47:32PM +0000, Bobby Eshleman wrote:
> > Previous to setting the owner the socket is found via
> > vsock_find_connected_socket(), which returns sk after a call to
> > sock_hold().
> > 
> > If setting the owner fails, then sock_put() needs to be called.
> > 
> > Fixes: f9d2b1e146e0 ("virtio/vsock: fix leaks due to missing skb owner")
> > Signed-off-by: Bobby Eshleman <bobby.eshleman@...edance.com>
> > ---
> > net/vmw_vsock/virtio_transport_common.c | 1 +
> > 1 file changed, 1 insertion(+)
> > 
> > diff --git a/net/vmw_vsock/virtio_transport_common.c b/net/vmw_vsock/virtio_transport_common.c
> > index b769fc258931..f01cd6adc5cb 100644
> > --- a/net/vmw_vsock/virtio_transport_common.c
> > +++ b/net/vmw_vsock/virtio_transport_common.c
> > @@ -1343,6 +1343,7 @@ void virtio_transport_recv_pkt(struct virtio_transport *t,
> > 
> > 	if (!skb_set_owner_sk_safe(skb, sk)) {
> > 		WARN_ONCE(1, "receiving vsock socket has sk_refcnt == 0\n");
> > +		sock_put(sk);
> 
> Did you have any warning, issue here?
> 
> IIUC skb_set_owner_sk_safe() can return false only if the ref counter
> is 0, so calling a sock_put() on it should have no effect except to
> produce a warning.
> 

Oh yeah, you're totally right. I did not recall how
skb_set_owner_sk_safe() worked internally and thought I'd introduced an
uneven hold/put count with that prior patch when reading through the
code again. I haven't seen any live issue, just misread the code.

Sorry about that, feel free to ignore this patch.

Best,
Bobby