/*
- * Copyright (c) 2014 Nicira, Inc.
+ * Copyright (c) 2014, 2015, 2016 Nicira, Inc.
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
/*
* need to reserve tons of extra space in the mbufs so we can align the
* DMA addresses to 4KB.
+ * The minimum mbuf size is limited to avoid scatter behaviour and drop in
+ * performance for standard Ethernet MTU.
*/
-
#define MTU_TO_MAX_LEN(mtu) ((mtu) + ETHER_HDR_LEN + ETHER_CRC_LEN)
-#define MBUF_SIZE(mtu) (MTU_TO_MAX_LEN(mtu) + (512) + \
- sizeof(struct rte_mbuf) + RTE_PKTMBUF_HEADROOM)
+#define MBUF_SIZE_MTU(mtu) (MTU_TO_MAX_LEN(mtu) \
+ + sizeof(struct dp_packet) \
+ + RTE_PKTMBUF_HEADROOM)
+#define MBUF_SIZE_DRIVER (2048 \
+ + sizeof (struct rte_mbuf) \
+ + RTE_PKTMBUF_HEADROOM)
+#define MBUF_SIZE(mtu) MAX(MBUF_SIZE_MTU(mtu), MBUF_SIZE_DRIVER)
/* Max and min number of packets in the mempool. OVS tries to allocate a
* mempool with MAX_NB_MBUF: if this fails (because the system doesn't have
int port_id;
};
-static bool thread_is_pmd(void);
+static bool dpdk_thread_is_pmd(void);
static int netdev_dpdk_construct(struct netdev *);
return;
}
+ if (rte_vhost_driver_unregister(dev->vhost_id)) {
+ VLOG_ERR("Unable to remove vhost-user socket %s", dev->vhost_id);
+ }
+
ovs_mutex_lock(&dpdk_mutex);
list_remove(&dev->list_node);
dpdk_mp_put(dev->dpdk_mp);
return (dev != NULL && (dev->flags & VIRTIO_DEV_RUNNING));
}
+static inline void
+netdev_dpdk_vhost_update_rx_counters(struct netdev_stats *stats,
+ struct dp_packet **packets, int count)
+{
+ int i;
+ struct dp_packet *packet;
+
+ stats->rx_packets += count;
+ for (i = 0; i < count; i++) {
+ packet = packets[i];
+
+ if (OVS_UNLIKELY(dp_packet_size(packet) < ETH_HEADER_LEN)) {
+ /* This only protects the following multicast counting from
+ * too short packets, but it does not stop the packet from
+ * further processing. */
+ stats->rx_errors++;
+ stats->rx_length_errors++;
+ continue;
+ }
+
+ struct eth_header *eh = (struct eth_header *) dp_packet_data(packet);
+ if (OVS_UNLIKELY(eth_addr_is_multicast(eh->eth_dst))) {
+ stats->multicast++;
+ }
+
+ stats->rx_bytes += dp_packet_size(packet);
+ }
+}
+
/*
* The receive path for the vhost port is the TX path out from guest.
*/
}
rte_spinlock_lock(&vhost_dev->stats_lock);
- vhost_dev->stats.rx_packets += (uint64_t)nb_rx;
+ netdev_dpdk_vhost_update_rx_counters(&vhost_dev->stats, packets, nb_rx);
rte_spinlock_unlock(&vhost_dev->stats_lock);
*c = (int) nb_rx;
return 0;
}
+static inline void
+netdev_dpdk_vhost_update_tx_counters(struct netdev_stats *stats,
+ struct dp_packet **packets,
+ int attempted,
+ int dropped)
+{
+ int i;
+ int sent = attempted - dropped;
+
+ stats->tx_packets += sent;
+ stats->tx_dropped += dropped;
+
+ for (i = 0; i < sent; i++) {
+ stats->tx_bytes += dp_packet_size(packets[i]);
+ }
+}
+
static void
__netdev_dpdk_vhost_send(struct netdev *netdev, struct dp_packet **pkts,
int cnt, bool may_steal)
rte_spinlock_unlock(&vhost_dev->vhost_tx_lock);
rte_spinlock_lock(&vhost_dev->stats_lock);
- vhost_dev->stats.tx_packets += (total_pkts - cnt);
- vhost_dev->stats.tx_dropped += cnt;
+ netdev_dpdk_vhost_update_tx_counters(&vhost_dev->stats, pkts, total_pkts,
+ cnt);
rte_spinlock_unlock(&vhost_dev->stats_lock);
out:
/* If we are on a non pmd thread we have to use the mempool mutex, because
* every non pmd thread shares the same mempool cache */
- if (!thread_is_pmd()) {
+ if (!dpdk_thread_is_pmd()) {
ovs_mutex_lock(&nonpmd_mempool_mutex);
}
dpdk_queue_flush(dev, qid);
}
- if (!thread_is_pmd()) {
+ if (!dpdk_thread_is_pmd()) {
ovs_mutex_unlock(&nonpmd_mempool_mutex);
}
}
ovs_mutex_lock(&dev->mutex);
memset(stats, 0, sizeof(*stats));
/* Unsupported Stats */
- stats->rx_errors = UINT64_MAX;
- stats->tx_errors = UINT64_MAX;
- stats->multicast = UINT64_MAX;
stats->collisions = UINT64_MAX;
stats->rx_crc_errors = UINT64_MAX;
stats->rx_fifo_errors = UINT64_MAX;
stats->rx_frame_errors = UINT64_MAX;
- stats->rx_length_errors = UINT64_MAX;
stats->rx_missed_errors = UINT64_MAX;
stats->rx_over_errors = UINT64_MAX;
stats->tx_aborted_errors = UINT64_MAX;
stats->tx_fifo_errors = UINT64_MAX;
stats->tx_heartbeat_errors = UINT64_MAX;
stats->tx_window_errors = UINT64_MAX;
- stats->rx_bytes += UINT64_MAX;
stats->rx_dropped += UINT64_MAX;
- stats->tx_bytes += UINT64_MAX;
rte_spinlock_lock(&dev->stats_lock);
/* Supported Stats */
stats->rx_packets += dev->stats.rx_packets;
stats->tx_packets += dev->stats.tx_packets;
stats->tx_dropped += dev->stats.tx_dropped;
+ stats->multicast = dev->stats.multicast;
+ stats->rx_bytes = dev->stats.rx_bytes;
+ stats->tx_bytes = dev->stats.tx_bytes;
+ stats->rx_errors = dev->stats.rx_errors;
+ stats->rx_length_errors = dev->stats.rx_length_errors;
rte_spinlock_unlock(&dev->stats_lock);
+
ovs_mutex_unlock(&dev->mutex);
return 0;
stats->tx_packets = rte_stats.opackets;
stats->rx_bytes = rte_stats.ibytes;
stats->tx_bytes = rte_stats.obytes;
- stats->rx_errors = rte_stats.ierrors;
+ /* DPDK counts imissed as errors, but count them here as dropped instead */
+ stats->rx_errors = rte_stats.ierrors - rte_stats.imissed;
stats->tx_errors = rte_stats.oerrors;
stats->multicast = rte_stats.imcasts;
rte_spinlock_lock(&dev->stats_lock);
stats->tx_dropped = dev->stats.tx_dropped;
rte_spinlock_unlock(&dev->stats_lock);
+
+ /* These are the available DPDK counters for packets not received due to
+ * local resource constraints in DPDK and NIC respectively. */
+ stats->rx_dropped = rte_stats.rx_nombuf + rte_stats.imissed;
+ stats->collisions = UINT64_MAX;
+
+ stats->rx_length_errors = rte_stats.ibadlen;
+ stats->rx_over_errors = UINT64_MAX;
+ stats->rx_crc_errors = rte_stats.ibadcrc;
+ stats->rx_frame_errors = UINT64_MAX;
+ stats->rx_fifo_errors = UINT64_MAX;
+ stats->rx_missed_errors = rte_stats.imissed;
+
+ stats->tx_aborted_errors = UINT64_MAX;
+ stats->tx_carrier_errors = UINT64_MAX;
+ stats->tx_fifo_errors = UINT64_MAX;
+ stats->tx_heartbeat_errors = UINT64_MAX;
+ stats->tx_window_errors = UINT64_MAX;
+
ovs_mutex_unlock(&dev->mutex);
return 0;
smap_add_format(args, "max_vfs", "%u", dev_info.max_vfs);
smap_add_format(args, "max_vmdq_pools", "%u", dev_info.max_vmdq_pools);
- smap_add_format(args, "pci-vendor_id", "0x%u", dev_info.pci_dev->id.vendor_id);
- smap_add_format(args, "pci-device_id", "0x%x", dev_info.pci_dev->id.device_id);
+ if (dev_info.pci_dev) {
+ smap_add_format(args, "pci-vendor_id", "0x%u",
+ dev_info.pci_dev->id.vendor_id);
+ smap_add_format(args, "pci-device_id", "0x%x",
+ dev_info.pci_dev->id.device_id);
+ }
return 0;
}
ovs_mutex_unlock(&dpdk_mutex);
if (!exists) {
- VLOG_INFO("vHost Device '%s' (%ld) can't be added - name not found",
- dev->ifname, dev->device_fh);
+ VLOG_INFO("vHost Device '%s' %"PRIu64" can't be added - name not "
+ "found", dev->ifname, dev->device_fh);
return -1;
}
- VLOG_INFO("vHost Device '%s' (%ld) has been added",
- dev->ifname, dev->device_fh);
+ VLOG_INFO("vHost Device '%s' %"PRIu64" has been added", dev->ifname,
+ dev->device_fh);
return 0;
}
}
ovs_mutex_unlock(&dpdk_mutex);
- VLOG_INFO("vHost Device '%s' (%ld) has been removed",
- dev->ifname, dev->device_fh);
+ VLOG_INFO("vHost Device '%s' %"PRIu64" has been removed", dev->ifname,
+ dev->device_fh);
}
struct virtio_net *
return -err;
}
- /* Create single consumer/producer rings, netdev does explicit locking. */
+ /* Create single producer tx ring, netdev does explicit locking. */
ivshmem->cring_tx = rte_ring_create(ring_name, DPDK_RING_SIZE, SOCKET0,
- RING_F_SP_ENQ | RING_F_SC_DEQ);
+ RING_F_SP_ENQ);
if (ivshmem->cring_tx == NULL) {
rte_free(ivshmem);
return ENOMEM;
return -err;
}
- /* Create single consumer/producer rings, netdev does explicit locking. */
+ /* Create single consumer rx ring, netdev does explicit locking. */
ivshmem->cring_rx = rte_ring_create(ring_name, DPDK_RING_SIZE, SOCKET0,
- RING_F_SP_ENQ | RING_F_SC_DEQ);
+ RING_F_SC_DEQ);
if (ivshmem->cring_rx == NULL) {
rte_free(ivshmem);
return ENOMEM;
* the consumer of the ring and return into the datapath without recalculating
* the RSS hash. */
for (i = 0; i < cnt; i++) {
- dp_packet_set_rss_hash(pkts[i], 0);
+ dp_packet_rss_invalidate(pkts[i]);
}
netdev_dpdk_send__(netdev, qid, pkts, cnt, may_steal);
* flag if it is provided on the vswitchd command line, otherwise resort to
* a default value.
*
- * For vhost-user: Process "-cuse_dev_name" to set the custom location of
+ * For vhost-user: Process "-vhost_sock_dir" to set the custom location of
* the vhost-user socket(s).
- * For vhost-cuse: Process "-vhost_sock_dir" to set the custom name of the
+ * For vhost-cuse: Process "-cuse_dev_name" to set the custom name of the
* vhost-cuse character device.
*/
if (!strcmp(argv[1], flag) && (strlen(argv[2]) <= size)) {
changed = 1;
- *new_val = strdup(argv[2]);
+ *new_val = xstrdup(argv[2]);
VLOG_INFO("User-provided %s in use: %s", flag, *new_val);
} else {
VLOG_INFO("No %s provided - defaulting to %s", flag, default_val);
argc--;
argv++;
+ /* Reject --user option */
+ int i;
+ for (i = 0; i < argc; i++) {
+ if (!strcmp(argv[i], "--user")) {
+ VLOG_ERR("Can not mix --dpdk and --user options, aborting.");
+ }
+ }
+
#ifdef VHOST_CUSE
- if (process_vhost_flags("-cuse_dev_name", strdup("vhost-net"),
+ if (process_vhost_flags("-cuse_dev_name", xstrdup("vhost-net"),
PATH_MAX, argv, &cuse_dev_name)) {
#else
- if (process_vhost_flags("-vhost_sock_dir", strdup(ovs_rundir()),
+ if (process_vhost_flags("-vhost_sock_dir", xstrdup(ovs_rundir()),
NAME_MAX, argv, &vhost_sock_dir)) {
struct stat s;
int err;
}
static bool
-thread_is_pmd(void)
+dpdk_thread_is_pmd(void)
{
return rte_lcore_id() != NON_PMD_CORE_ID;
}