/*
- * Copyright (c) 2009, 2010, 2011, 2012, 2013 Nicira, Inc.
+ * Copyright (c) 2009, 2010, 2011, 2012, 2013, 2014 Nicira, Inc.
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
#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>
#include <linux/pkt_sched.h>
#include <linux/rtnetlink.h>
#include <linux/sockios.h>
-#include <linux/version.h>
#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>
#include <string.h>
#include <unistd.h>
-#include "connectivity.h"
#include "coverage.h"
#include "dpif-linux.h"
+#include "dpif-netdev.h"
#include "dynamic-string.h"
#include "fatal-signal.h"
#include "hash.h"
#include "packets.h"
#include "poll-loop.h"
#include "rtnetlink-link.h"
-#include "seq.h"
#include "shash.h"
#include "socket-util.h"
#include "sset.h"
* With all this churn it's easiest to unconditionally define a replacement
* structure that has everything we want.
*/
+#ifndef PACKET_AUXDATA
+#define PACKET_AUXDATA 8
+#endif
#ifndef TP_STATUS_VLAN_VALID
#define TP_STATUS_VLAN_VALID (1 << 4)
#endif
int tap_fd;
};
-struct netdev_rx_linux {
- struct netdev_rx up;
+struct netdev_rxq_linux {
+ struct netdev_rxq up;
bool is_tap;
int fd;
};
static bool netdev_linux_miimon_enabled(void);
static void netdev_linux_miimon_run(void);
static void netdev_linux_miimon_wait(void);
+static int netdev_linux_get_mtu__(struct netdev_linux *netdev, int *mtup);
static bool
is_netdev_linux_class(const struct netdev_class *netdev_class)
return CONTAINER_OF(netdev, struct netdev_linux, up);
}
-static struct netdev_rx_linux *
-netdev_rx_linux_cast(const struct netdev_rx *rx)
+static struct netdev_rxq_linux *
+netdev_rxq_linux_cast(const struct netdev_rxq *rx)
{
ovs_assert(is_netdev_linux_class(netdev_get_class(rx->netdev)));
- return CONTAINER_OF(rx, struct netdev_rx_linux, up);
+ return CONTAINER_OF(rx, struct netdev_rxq_linux, up);
}
\f
static void netdev_linux_update(struct netdev_linux *netdev,
unsigned int ifi_flags, unsigned int mask)
OVS_REQUIRES(dev->mutex)
{
- seq_change(connectivity_seq_get());
+ netdev_change_seq_changed(&dev->up);
if ((dev->ifi_flags ^ ifi_flags) & IFF_RUNNING) {
dev->carrier_resets++;
free(netdev);
}
-static struct netdev_rx *
-netdev_linux_rx_alloc(void)
+static struct netdev_rxq *
+netdev_linux_rxq_alloc(void)
{
- struct netdev_rx_linux *rx = xzalloc(sizeof *rx);
+ struct netdev_rxq_linux *rx = xzalloc(sizeof *rx);
return &rx->up;
}
static int
-netdev_linux_rx_construct(struct netdev_rx *rx_)
+netdev_linux_rxq_construct(struct netdev_rxq *rxq_)
{
- struct netdev_rx_linux *rx = netdev_rx_linux_cast(rx_);
+ struct netdev_rxq_linux *rx = netdev_rxq_linux_cast(rxq_);
struct netdev *netdev_ = rx->up.netdev;
struct netdev_linux *netdev = netdev_linux_cast(netdev_);
int error;
}
static void
-netdev_linux_rx_destruct(struct netdev_rx *rx_)
+netdev_linux_rxq_destruct(struct netdev_rxq *rxq_)
{
- struct netdev_rx_linux *rx = netdev_rx_linux_cast(rx_);
+ struct netdev_rxq_linux *rx = netdev_rxq_linux_cast(rxq_);
if (!rx->is_tap) {
close(rx->fd);
}
static void
-netdev_linux_rx_dealloc(struct netdev_rx *rx_)
+netdev_linux_rxq_dealloc(struct netdev_rxq *rxq_)
{
- struct netdev_rx_linux *rx = netdev_rx_linux_cast(rx_);
+ struct netdev_rxq_linux *rx = netdev_rxq_linux_cast(rxq_);
free(rx);
}
}
static int
-netdev_linux_rx_recv_sock(int fd, struct ofpbuf *buffer)
+netdev_linux_rxq_recv_sock(int fd, struct ofpbuf *buffer)
{
size_t size;
ssize_t retval;
ofpbuf_reserve(buffer, VLAN_HEADER_LEN);
size = ofpbuf_tailroom(buffer);
- iov.iov_base = buffer->data;
+ iov.iov_base = ofpbuf_data(buffer);
iov.iov_len = size;
msgh.msg_name = NULL;
msgh.msg_namelen = 0;
return EMSGSIZE;
}
- buffer->size += retval;
+ ofpbuf_set_size(buffer, ofpbuf_size(buffer) + retval);
for (cmsg = CMSG_FIRSTHDR(&msgh); cmsg; cmsg = CMSG_NXTHDR(&msgh, cmsg)) {
const struct tpacket_auxdata *aux;
}
static int
-netdev_linux_rx_recv_tap(int fd, struct ofpbuf *buffer)
+netdev_linux_rxq_recv_tap(int fd, struct ofpbuf *buffer)
{
ssize_t retval;
size_t size = ofpbuf_tailroom(buffer);
do {
- retval = read(fd, buffer->data, size);
+ retval = read(fd, ofpbuf_data(buffer), size);
} while (retval < 0 && errno == EINTR);
if (retval < 0) {
return EMSGSIZE;
}
- buffer->size += retval;
+ ofpbuf_set_size(buffer, ofpbuf_size(buffer) + retval);
return 0;
}
static int
-netdev_linux_rx_recv(struct netdev_rx *rx_, struct ofpbuf *buffer)
+netdev_linux_rxq_recv(struct netdev_rxq *rxq_, struct ofpbuf **packet, int *c)
{
- struct netdev_rx_linux *rx = netdev_rx_linux_cast(rx_);
- int retval;
+ struct netdev_rxq_linux *rx = netdev_rxq_linux_cast(rxq_);
+ struct netdev *netdev = rx->up.netdev;
+ struct ofpbuf *buffer;
+ ssize_t retval;
+ int mtu;
+
+ if (netdev_linux_get_mtu__(netdev_linux_cast(netdev), &mtu)) {
+ mtu = ETH_PAYLOAD_MAX;
+ }
+
+ buffer = ofpbuf_new_with_headroom(VLAN_ETH_HEADER_LEN + mtu, DP_NETDEV_HEADROOM);
retval = (rx->is_tap
- ? netdev_linux_rx_recv_tap(rx->fd, buffer)
- : netdev_linux_rx_recv_sock(rx->fd, buffer));
- if (retval && retval != EAGAIN && retval != EMSGSIZE) {
- VLOG_WARN_RL(&rl, "error receiving Ethernet packet on %s: %s",
- ovs_strerror(errno), netdev_rx_get_name(rx_));
+ ? netdev_linux_rxq_recv_tap(rx->fd, buffer)
+ : netdev_linux_rxq_recv_sock(rx->fd, buffer));
+
+ if (retval) {
+ if (retval != EAGAIN && retval != EMSGSIZE) {
+ VLOG_WARN_RL(&rl, "error receiving Ethernet packet on %s: %s",
+ ovs_strerror(errno), netdev_rxq_get_name(rxq_));
+ }
+ ofpbuf_delete(buffer);
+ } else {
+ dp_packet_pad(buffer);
+ packet[0] = buffer;
+ *c = 1;
}
return retval;
}
static void
-netdev_linux_rx_wait(struct netdev_rx *rx_)
+netdev_linux_rxq_wait(struct netdev_rxq *rxq_)
{
- struct netdev_rx_linux *rx = netdev_rx_linux_cast(rx_);
+ struct netdev_rxq_linux *rx = netdev_rxq_linux_cast(rxq_);
poll_fd_wait(rx->fd, POLLIN);
}
static int
-netdev_linux_rx_drain(struct netdev_rx *rx_)
+netdev_linux_rxq_drain(struct netdev_rxq *rxq_)
{
- struct netdev_rx_linux *rx = netdev_rx_linux_cast(rx_);
+ struct netdev_rxq_linux *rx = netdev_rxq_linux_cast(rxq_);
if (rx->is_tap) {
struct ifreq ifr;
- int error = af_inet_ifreq_ioctl(netdev_rx_get_name(rx_), &ifr,
+ int error = af_inet_ifreq_ioctl(netdev_rxq_get_name(rxq_), &ifr,
SIOCGIFTXQLEN, "SIOCGIFTXQLEN");
if (error) {
return error;
* The kernel maintains a packet transmission queue, so the caller is not
* expected to do additional queuing of packets. */
static int
-netdev_linux_send(struct netdev *netdev_, const void *data, size_t size)
+netdev_linux_send(struct netdev *netdev_, struct ofpbuf *pkt, bool may_steal)
{
+ const void *data = ofpbuf_data(pkt);
+ size_t size = ofpbuf_size(pkt);
+
for (;;) {
ssize_t retval;
retval = write(netdev->tap_fd, data, size);
}
+ if (may_steal) {
+ ofpbuf_delete(pkt);
+ }
+
if (retval < 0) {
/* The Linux AF_PACKET implementation never blocks waiting for room
* for packets, instead returning ENOBUFS. Translate this into
error = 0;
}
} else if (netdev->vport_stats_error) {
- /* stats not available from OVS then use ioctl stats. */
+ /* stats not available from OVS then use netdev stats. */
*stats = dev_stats;
} else {
+ /* Use kernel netdev's packet and byte counts since vport's counters
+ * do not reflect packet counts on the wire when GSO, TSO or GRO are
+ * enabled. */
+ stats->rx_packets = dev_stats.rx_packets;
+ stats->rx_bytes = dev_stats.rx_bytes;
+ stats->tx_packets = dev_stats.tx_packets;
+ stats->tx_bytes = dev_stats.tx_bytes;
+
stats->rx_errors += dev_stats.rx_errors;
stats->tx_errors += dev_stats.tx_errors;
stats->rx_dropped += dev_stats.rx_dropped;
stats->tx_heartbeat_errors = 0;
stats->tx_window_errors = 0;
} else {
+ /* Use kernel netdev's packet and byte counts since vport counters
+ * do not reflect packet counts on the wire when GSO, TSO or GRO
+ * are enabled. */
+ stats->rx_packets = dev_stats.tx_packets;
+ stats->rx_bytes = dev_stats.tx_bytes;
+ stats->tx_packets = dev_stats.rx_packets;
+ stats->tx_bytes = dev_stats.rx_bytes;
+
stats->rx_dropped += dev_stats.tx_dropped;
stats->tx_dropped += dev_stats.rx_dropped;
return error;
}
+struct queue_dump_state {
+ struct nl_dump dump;
+ struct ofpbuf buf;
+};
+
static bool
-start_queue_dump(const struct netdev *netdev, struct nl_dump *dump)
+start_queue_dump(const struct netdev *netdev, struct queue_dump_state *state)
{
struct ofpbuf request;
struct tcmsg *tcmsg;
return false;
}
tcmsg->tcm_parent = 0;
- nl_dump_start(dump, NETLINK_ROUTE, &request);
+ nl_dump_start(&state->dump, NETLINK_ROUTE, &request);
ofpbuf_uninit(&request);
+
+ ofpbuf_init(&state->buf, NL_DUMP_BUFSIZE);
return true;
}
+static int
+finish_queue_dump(struct queue_dump_state *state)
+{
+ ofpbuf_uninit(&state->buf);
+ return nl_dump_done(&state->dump);
+}
+
struct netdev_linux_queue_state {
unsigned int *queues;
size_t cur_queue;
ovs_mutex_lock(&netdev->mutex);
error = tc_query_qdisc(netdev_);
if (!error) {
- struct nl_dump dump;
+ struct queue_dump_state state;
if (!netdev->tc->ops->class_dump_stats) {
error = EOPNOTSUPP;
- } else if (!start_queue_dump(netdev_, &dump)) {
+ } else if (!start_queue_dump(netdev_, &state)) {
error = ENODEV;
} else {
struct ofpbuf msg;
int retval;
- while (nl_dump_next(&dump, &msg)) {
+ while (nl_dump_next(&state.dump, &msg, &state.buf)) {
retval = netdev->tc->ops->class_dump_stats(netdev_, &msg,
cb, aux);
if (retval) {
}
}
- retval = nl_dump_done(&dump);
+ retval = finish_queue_dump(&state);
if (retval) {
error = retval;
}
\
netdev_linux_update_flags, \
\
- netdev_linux_rx_alloc, \
- netdev_linux_rx_construct, \
- netdev_linux_rx_destruct, \
- netdev_linux_rx_dealloc, \
- netdev_linux_rx_recv, \
- netdev_linux_rx_wait, \
- netdev_linux_rx_drain, \
+ netdev_linux_rxq_alloc, \
+ netdev_linux_rxq_construct, \
+ netdev_linux_rxq_destruct, \
+ netdev_linux_rxq_dealloc, \
+ netdev_linux_rxq_recv, \
+ netdev_linux_rxq_wait, \
+ netdev_linux_rxq_drain, \
}
const struct netdev_class netdev_linux_class =
htb_tc_load(struct netdev *netdev, struct ofpbuf *nlmsg OVS_UNUSED)
{
struct ofpbuf msg;
- struct nl_dump dump;
+ struct queue_dump_state state;
struct htb_class hc;
/* Get qdisc options. */
htb_install__(netdev, hc.max_rate);
/* Get queues. */
- if (!start_queue_dump(netdev, &dump)) {
+ if (!start_queue_dump(netdev, &state)) {
return ENODEV;
}
- while (nl_dump_next(&dump, &msg)) {
+ while (nl_dump_next(&state.dump, &msg, &state.buf)) {
unsigned int queue_id;
if (!htb_parse_tcmsg__(&msg, &queue_id, &hc, NULL)) {
htb_update_queue__(netdev, queue_id, &hc);
}
}
- nl_dump_done(&dump);
+ finish_queue_dump(&state);
return 0;
}
hfsc_tc_load(struct netdev *netdev, struct ofpbuf *nlmsg OVS_UNUSED)
{
struct ofpbuf msg;
- struct nl_dump dump;
+ struct queue_dump_state state;
struct hfsc_class hc;
hc.max_rate = 0;
hfsc_query_class__(netdev, tc_make_handle(1, 0xfffe), 0, &hc, NULL);
hfsc_install__(netdev, hc.max_rate);
- if (!start_queue_dump(netdev, &dump)) {
+ if (!start_queue_dump(netdev, &state)) {
return ENODEV;
}
- while (nl_dump_next(&dump, &msg)) {
+ while (nl_dump_next(&state.dump, &msg, &state.buf)) {
unsigned int queue_id;
if (!hfsc_parse_tcmsg__(&msg, &queue_id, &hc, NULL)) {
}
}
- nl_dump_done(&dump);
+ finish_queue_dump(&state);
return 0;
}
git://git.onelab.eu / sliver-openvswitch.git / blobdiff