From: Jason Wang jasowang@redhat.com
[ Upstream commit 2f3ab6221e4c87960347d65c7cab9bd917d1f637 ]
When link is down, writes to the device might fail with -EIO. Userspace needs an indication when the status is resolved. As a fix, tun_net_open() attempts to wake up writers - but that is only effective if SOCKWQ_ASYNC_NOSPACE has been set in the past. This is not the case of vhost_net which only poll for EPOLLOUT after it meets errors during sendmsg().
This patch fixes this by making sure SOCKWQ_ASYNC_NOSPACE is set when socket is not writable or device is down to guarantee EPOLLOUT will be raised in either tun_chr_poll() or tun_sock_write_space() after device is up.
Cc: Hannes Frederic Sowa hannes@stressinduktion.org Cc: Eric Dumazet edumazet@google.com Fixes: 1bd4978a88ac2 ("tun: honor IFF_UP in tun_get_user()") Signed-off-by: Jason Wang jasowang@redhat.com Signed-off-by: David S. Miller davem@davemloft.net Signed-off-by: Tommi Rantala tommi.t.rantala@nokia.com --- drivers/net/tun.c | 19 +++++++++++++++---- 1 file changed, 15 insertions(+), 4 deletions(-)
diff --git a/drivers/net/tun.c b/drivers/net/tun.c index 3086211829a7..ba34f61d70de 100644 --- a/drivers/net/tun.c +++ b/drivers/net/tun.c @@ -1134,6 +1134,13 @@ static void tun_net_init(struct net_device *dev) dev->max_mtu = MAX_MTU - dev->hard_header_len; }
+static bool tun_sock_writeable(struct tun_struct *tun, struct tun_file *tfile) +{ + struct sock *sk = tfile->socket.sk; + + return (tun->dev->flags & IFF_UP) && sock_writeable(sk); +} + /* Character device part */
/* Poll */ @@ -1156,10 +1163,14 @@ static unsigned int tun_chr_poll(struct file *file, poll_table *wait) if (!skb_array_empty(&tfile->tx_array)) mask |= POLLIN | POLLRDNORM;
- if (tun->dev->flags & IFF_UP && - (sock_writeable(sk) || - (!test_and_set_bit(SOCKWQ_ASYNC_NOSPACE, &sk->sk_socket->flags) && - sock_writeable(sk)))) + /* Make sure SOCKWQ_ASYNC_NOSPACE is set if not writable to + * guarantee EPOLLOUT to be raised by either here or + * tun_sock_write_space(). Then process could get notification + * after it writes to a down device and meets -EIO. + */ + if (tun_sock_writeable(tun, tfile) || + (!test_and_set_bit(SOCKWQ_ASYNC_NOSPACE, &sk->sk_socket->flags) && + tun_sock_writeable(tun, tfile))) mask |= POLLOUT | POLLWRNORM;
if (tun->dev->reg_state != NETREG_REGISTERED)
On Fri, Feb 28, 2020 at 10:42:16AM +0200, Tommi Rantala wrote:
From: Jason Wang jasowang@redhat.com
[ Upstream commit 2f3ab6221e4c87960347d65c7cab9bd917d1f637 ]
When link is down, writes to the device might fail with -EIO. Userspace needs an indication when the status is resolved. As a fix, tun_net_open() attempts to wake up writers - but that is only effective if SOCKWQ_ASYNC_NOSPACE has been set in the past. This is not the case of vhost_net which only poll for EPOLLOUT after it meets errors during sendmsg().
This patch fixes this by making sure SOCKWQ_ASYNC_NOSPACE is set when socket is not writable or device is down to guarantee EPOLLOUT will be raised in either tun_chr_poll() or tun_sock_write_space() after device is up.
Cc: Hannes Frederic Sowa hannes@stressinduktion.org Cc: Eric Dumazet edumazet@google.com Fixes: 1bd4978a88ac2 ("tun: honor IFF_UP in tun_get_user()") Signed-off-by: Jason Wang jasowang@redhat.com Signed-off-by: David S. Miller davem@davemloft.net Signed-off-by: Tommi Rantala tommi.t.rantala@nokia.com
drivers/net/tun.c | 19 +++++++++++++++---- 1 file changed, 15 insertions(+), 4 deletions(-)
Thanks for the backport, now queued up to 4.9.y and 4.14.y.
greg k-h
linux-stable-mirror@lists.linaro.org