Commit 762c330d authored by Jason Wang's avatar Jason Wang Committed by David S. Miller

tuntap: add missing xdp flush

When using devmap to redirect packets between interfaces,
xdp_do_flush() is usually a must to flush any batched
packets. Unfortunately this is missed in current tuntap
implementation.

Unlike most hardware driver which did XDP inside NAPI loop and call
xdp_do_flush() at then end of each round of poll. TAP did it in the
context of process e.g tun_get_user(). So fix this by count the
pending redirected packets and flush when it exceeds NAPI_POLL_WEIGHT
or MSG_MORE was cleared by sendmsg() caller.

With this fix, xdp_redirect_map works again between two TAPs.

Fixes: 761876c8 ("tap: XDP support")
Signed-off-by: default avatarJason Wang <jasowang@redhat.com>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
parent cb9f7a9a
...@@ -181,6 +181,7 @@ struct tun_file { ...@@ -181,6 +181,7 @@ struct tun_file {
struct tun_struct *detached; struct tun_struct *detached;
struct ptr_ring tx_ring; struct ptr_ring tx_ring;
struct xdp_rxq_info xdp_rxq; struct xdp_rxq_info xdp_rxq;
int xdp_pending_pkts;
}; };
struct tun_flow_entry { struct tun_flow_entry {
...@@ -1665,6 +1666,7 @@ static struct sk_buff *tun_build_skb(struct tun_struct *tun, ...@@ -1665,6 +1666,7 @@ static struct sk_buff *tun_build_skb(struct tun_struct *tun,
case XDP_REDIRECT: case XDP_REDIRECT:
get_page(alloc_frag->page); get_page(alloc_frag->page);
alloc_frag->offset += buflen; alloc_frag->offset += buflen;
++tfile->xdp_pending_pkts;
err = xdp_do_redirect(tun->dev, &xdp, xdp_prog); err = xdp_do_redirect(tun->dev, &xdp, xdp_prog);
if (err) if (err)
goto err_redirect; goto err_redirect;
...@@ -1986,6 +1988,11 @@ static ssize_t tun_chr_write_iter(struct kiocb *iocb, struct iov_iter *from) ...@@ -1986,6 +1988,11 @@ static ssize_t tun_chr_write_iter(struct kiocb *iocb, struct iov_iter *from)
result = tun_get_user(tun, tfile, NULL, from, result = tun_get_user(tun, tfile, NULL, from,
file->f_flags & O_NONBLOCK, false); file->f_flags & O_NONBLOCK, false);
if (tfile->xdp_pending_pkts) {
tfile->xdp_pending_pkts = 0;
xdp_do_flush_map();
}
tun_put(tun); tun_put(tun);
return result; return result;
} }
...@@ -2322,6 +2329,13 @@ static int tun_sendmsg(struct socket *sock, struct msghdr *m, size_t total_len) ...@@ -2322,6 +2329,13 @@ static int tun_sendmsg(struct socket *sock, struct msghdr *m, size_t total_len)
ret = tun_get_user(tun, tfile, m->msg_control, &m->msg_iter, ret = tun_get_user(tun, tfile, m->msg_control, &m->msg_iter,
m->msg_flags & MSG_DONTWAIT, m->msg_flags & MSG_DONTWAIT,
m->msg_flags & MSG_MORE); m->msg_flags & MSG_MORE);
if (tfile->xdp_pending_pkts >= NAPI_POLL_WEIGHT ||
!(m->msg_flags & MSG_MORE)) {
tfile->xdp_pending_pkts = 0;
xdp_do_flush_map();
}
tun_put(tun); tun_put(tun);
return ret; return ret;
} }
...@@ -3153,6 +3167,7 @@ static int tun_chr_open(struct inode *inode, struct file * file) ...@@ -3153,6 +3167,7 @@ static int tun_chr_open(struct inode *inode, struct file * file)
sock_set_flag(&tfile->sk, SOCK_ZEROCOPY); sock_set_flag(&tfile->sk, SOCK_ZEROCOPY);
memset(&tfile->tx_ring, 0, sizeof(tfile->tx_ring)); memset(&tfile->tx_ring, 0, sizeof(tfile->tx_ring));
tfile->xdp_pending_pkts = 0;
return 0; return 0;
} }
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment