If VLAN acceleration is used when the kernel receives a packet
then the outer-most VLAN tag will not be present in the packet
when it is received by netdev-linux. Rather, it will be present
in auxdata.
This patch uses recvmsg() instead of recv() to read auxdata for
each packet and if the vlan_tid is set then it is added to the packet.
Adding the vlan_tid to the packet involves copying most of the packet
and may be rather expensive. There is ample scope to avoid this by
passing the vlan_tid back to the caller separately to the packet itself
or providing access headroom in the packet. This would most likely
involve updating the netdev-class API.
Signed-off-by: Simon Horman <ho...@verge.net.au>
---
include/sparse/sys/socket.h | 1 +
lib/netdev-linux.c | 86 +++++++++++++++++++++++++++++++++++++++++----
2 files changed, 81 insertions(+), 6 deletions(-)
diff --git a/include/sparse/sys/socket.h b/include/sparse/sys/socket.h
index 75ee43c..3212bf4 100644
--- a/include/sparse/sys/socket.h
+++ b/include/sparse/sys/socket.h
@@ -87,6 +87,7 @@ enum {
};
enum {
+ SOL_PACKET,
SOL_SOCKET
};
diff --git a/lib/netdev-linux.c b/lib/netdev-linux.c
index 3e0da48..5034907 100644
--- a/lib/netdev-linux.c
+++ b/lib/netdev-linux.c
@@ -20,11 +20,11 @@
#include <errno.h>
#include <fcntl.h>
-#include <arpa/inet.h>
#include <inttypes.h>
#include <linux/filter.h>
#include <linux/gen_stats.h>
#include <linux/if_ether.h>
+#include <linux/if_packet.h>
#include <linux/if_tun.h>
#include <linux/types.h>
#include <linux/ethtool.h>
@@ -37,10 +37,8 @@
#include <sys/types.h>
#include <sys/ioctl.h>
#include <sys/socket.h>
-#include <netpacket/packet.h>
#include <net/if.h>
#include <net/if_arp.h>
-#include <net/if_packet.h>
#include <net/route.h>
#include <netinet/in.h>
#include <poll.h>
@@ -767,7 +765,7 @@ netdev_linux_rx_construct(struct netdev_rx *rx_)
rx->fd = netdev->tap_fd;
} else {
struct sockaddr_ll sll;
- int ifindex;
+ int ifindex, val;
/* Result of tcpdump -dd inbound */
static const struct sock_filter filt[] = {
{ 0x28, 0, 0, 0xfffff004 }, /* ldh [0] */
@@ -787,6 +785,16 @@ netdev_linux_rx_construct(struct netdev_rx *rx_)
goto error;
}
+ val = 1;
+ error = setsockopt(rx->fd, SOL_PACKET, PACKET_AUXDATA,
+ &val, sizeof val);
+ if (error) {
+ error = errno;
+ VLOG_ERR("%s: failed to mark socket for auxdata (%s)",
+ netdev_get_name(netdev_), ovs_strerror(error));
+ goto error;
+ }
+
/* Set non-blocking mode. */
error = set_nonblocking(rx->fd);
if (error) {
@@ -803,7 +811,7 @@ netdev_linux_rx_construct(struct netdev_rx *rx_)
memset(&sll, 0, sizeof sll);
sll.sll_family = AF_PACKET;
sll.sll_ifindex = ifindex;
- sll.sll_protocol = (OVS_FORCE unsigned short int) htons(ETH_P_ALL);
+ sll.sll_protocol = htons(ETH_P_ALL);
if (bind(rx->fd, (struct sockaddr *) &sll, sizeof sll) < 0) {
error = errno;
VLOG_ERR("%s: failed to bind raw socket (%s)",
@@ -852,6 +860,72 @@ netdev_linux_rx_dealloc(struct netdev_rx *rx_)
}
static int
+netdev_linux_rx_recv_sock(int fd, void *data, size_t size)
+{
+ ssize_t retval;
+ struct cmsghdr *cmsg;
+ union {
+ struct cmsghdr cmsg;
+ char buf[CMSG_SPACE(sizeof(struct tpacket_auxdata))];
+ } cmsg_buf;
+ struct msghdr msg;
+ struct iovec iov;
+
+ iov.iov_base = data;
+ iov.iov_len = size;
+ msg.msg_iov = &iov;
+ msg.msg_iovlen = 1;
+ msg.msg_control = &cmsg_buf;
+ msg.msg_controllen = sizeof cmsg_buf;
+ msg.msg_name = NULL;
+
+ do {
+ retval = recvmsg(fd, &msg, MSG_TRUNC);
+ } while (retval < 0 && errno == EINTR);
+
+ if (retval > size || retval < 0) {
+ return retval;
+ }
+
+ for (cmsg = CMSG_FIRSTHDR(&msg); cmsg; cmsg = CMSG_NXTHDR(&msg, cmsg)) {
+ struct tpacket_auxdata *aux;
+
+ if (cmsg->cmsg_len < CMSG_LEN(sizeof(struct tpacket_auxdata))
+ || cmsg->cmsg_level != SOL_PACKET
+ || cmsg->cmsg_type != PACKET_AUXDATA) {
+ continue;
+ }
+
+ aux = (struct tpacket_auxdata *)(void *)CMSG_DATA(cmsg);
+ if (aux->tp_vlan_tci) {
+ struct ofpbuf b;
+
+ VLOG_WARN("vlan tci = %u", aux->tp_vlan_tci);
+
+ if (retval + VLAN_HEADER_LEN > size) {
+ errno = EMSGSIZE;
+ return -errno;
+ }
+ if (size < ETH_ADDR_LEN) {
+ errno = EINVAL;
+ return -errno;
+ }
+
+ /* XXX: This may be quite expensive! */
+ memmove((char *)data + VLAN_HEADER_LEN, data, retval);
+
+ ofpbuf_use_stack(&b, data, size);
+ b.data = (char *)b.data + VLAN_HEADER_LEN;
+ b.size = retval;
+ eth_push_vlan(&b, htons(aux->tp_vlan_tci));
+ retval = b.size;
+ }
+ }
+
+ return retval;
+}
+
+static int
netdev_linux_rx_recv(struct netdev_rx *rx_, void *data, size_t size)
{
struct netdev_rx_linux *rx = netdev_rx_linux_cast(rx_);
@@ -860,7 +934,7 @@ netdev_linux_rx_recv(struct netdev_rx *rx_, void *data,
size_t size)
do {
retval = (rx->is_tap
? read(rx->fd, data, size)
- : recv(rx->fd, data, size, MSG_TRUNC));
+ : netdev_linux_rx_recv_sock(rx->fd, data, size));
} while (retval < 0 && errno == EINTR);
if (retval >= 0) {
--
1.8.4
_______________________________________________
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
