X-Git-Url: http://git.droids-corp.org/?a=blobdiff_plain;f=examples%2Fvhost%2Fmain.c;h=e7fee5aa1bc25b7082d273544aa27e7f16815acf;hb=ae835aba40349ee9631ef6b52e68a2893febe7e0;hp=91000e8c30fa328d38ce1482e06ca30fce5ddbd1;hpb=00b8b70666f82cb5aae50f182cfc90ac3fdad739;p=dpdk.git diff --git a/examples/vhost/main.c b/examples/vhost/main.c index 91000e8c30..e7fee5aa1b 100644 --- a/examples/vhost/main.c +++ b/examples/vhost/main.c @@ -1,34 +1,5 @@ -/*- - * BSD LICENSE - * - * Copyright(c) 2010-2016 Intel Corporation. All rights reserved. - * All rights reserved. - * - * Redistribution and use in source and binary forms, with or without - * modification, are permitted provided that the following conditions - * are met: - * - * * Redistributions of source code must retain the above copyright - * notice, this list of conditions and the following disclaimer. - * * Redistributions in binary form must reproduce the above copyright - * notice, this list of conditions and the following disclaimer in - * the documentation and/or other materials provided with the - * distribution. - * * Neither the name of Intel Corporation nor the names of its - * contributors may be used to endorse or promote products derived - * from this software without specific prior written permission. - * - * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS - * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT - * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR - * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT - * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, - * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT - * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, - * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY - * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT - * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE - * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. +/* SPDX-License-Identifier: BSD-3-Clause + * Copyright(c) 2010-2017 Intel Corporation */ #include @@ -43,15 +14,18 @@ #include #include -#include #include #include #include #include #include -#include +#include +#include #include #include +#include +#include +#include #include "main.h" @@ -59,19 +33,21 @@ #define MAX_QUEUES 128 #endif +#define NUM_MBUFS_DEFAULT 0x24000 + /* the maximum number of external ports supported */ #define MAX_SUP_PORTS 1 #define MBUF_CACHE_SIZE 128 #define MBUF_DATA_SIZE RTE_MBUF_DEFAULT_BUF_SIZE -#define MAX_PKT_BURST 32 /* Max burst size for RX/TX */ #define BURST_TX_DRAIN_US 100 /* TX drain every ~100us */ #define BURST_RX_WAIT_US 15 /* Defines how long we wait between retries on RX */ #define BURST_RX_RETRIES 4 /* Number of retries on RX. */ #define JUMBO_FRAME_MAX_SIZE 0x2600 +#define MAX_MTU (JUMBO_FRAME_MAX_SIZE - (RTE_ETHER_HDR_LEN + RTE_ETHER_CRC_LEN)) /* State of virtio device. */ #define DEVICE_MAC_LEARNING 0 @@ -83,15 +59,19 @@ #define RTE_TEST_TX_DESC_DEFAULT 512 #define INVALID_PORT_ID 0xFF +#define INVALID_DMA_ID -1 + +#define DMA_RING_SIZE 4096 -/* Max number of devices. Limited by vmdq. */ -#define MAX_DEVICES 64 +#define ASYNC_ENQUEUE_VHOST 1 +#define ASYNC_DEQUEUE_VHOST 2 -/* Size of buffers used for snprintfs. */ -#define MAX_PRINT_BUFF 6072 +/* number of mbufs in all pools - if specified on command-line. */ +static int total_num_mbufs = NUM_MBUFS_DEFAULT; -/* Maximum long option length for option parsing. */ -#define MAX_LONG_OPT_SZ 64 +struct dma_for_vhost dma_bind[RTE_MAX_VHOST_DEVICE]; +int16_t dmas_id[RTE_DMADEV_DEFAULT_MAX]; +static int dma_count; /* mask of enabled ports */ static uint32_t enabled_port_mask = 0; @@ -127,7 +107,8 @@ static uint32_t enable_tx_csum; static uint32_t enable_tso; static int client_mode; -static int dequeue_zero_copy; + +static int builtin_net_driver; /* Specify timeout (in useconds) between retries on RX. */ static uint32_t burst_rx_delay_time = BURST_RX_WAIT_US; @@ -138,26 +119,28 @@ static uint32_t burst_rx_retry_num = BURST_RX_RETRIES; static char *socket_files; static int nb_sockets; -/* empty vmdq configuration structure. Filled in programatically */ +static struct vhost_queue_ops vdev_queue_ops[RTE_MAX_VHOST_DEVICE]; + +/* empty VMDq configuration structure. Filled in programmatically */ static struct rte_eth_conf vmdq_conf_default = { .rxmode = { - .mq_mode = ETH_MQ_RX_VMDQ_ONLY, + .mq_mode = RTE_ETH_MQ_RX_VMDQ_ONLY, .split_hdr_size = 0, - .header_split = 0, /**< Header Split disabled */ - .hw_ip_checksum = 0, /**< IP checksum offload disabled */ - .hw_vlan_filter = 0, /**< VLAN filtering disabled */ /* - * It is necessary for 1G NIC such as I350, + * VLAN strip is necessary for 1G NIC such as I350, * this fixes bug of ipv4 forwarding in guest can't - * forward pakets from one virtio dev to another virtio dev. + * forward packets from one virtio dev to another virtio dev. */ - .hw_vlan_strip = 1, /**< VLAN strip enabled. */ - .jumbo_frame = 0, /**< Jumbo Frame Support disabled */ - .hw_strip_crc = 0, /**< CRC stripped by hardware */ + .offloads = RTE_ETH_RX_OFFLOAD_VLAN_STRIP, }, .txmode = { - .mq_mode = ETH_MQ_TX_NONE, + .mq_mode = RTE_ETH_MQ_TX_NONE, + .offloads = (RTE_ETH_TX_OFFLOAD_IPV4_CKSUM | + RTE_ETH_TX_OFFLOAD_TCP_CKSUM | + RTE_ETH_TX_OFFLOAD_VLAN_INSERT | + RTE_ETH_TX_OFFLOAD_MULTI_SEGS | + RTE_ETH_TX_OFFLOAD_TCP_TSO), }, .rx_adv_conf = { /* @@ -165,7 +148,7 @@ static struct rte_eth_conf vmdq_conf_default = { * appropriate values */ .vmdq_rx_conf = { - .nb_queue_pools = ETH_8_POOLS, + .nb_queue_pools = RTE_ETH_8_POOLS, .enable_default_pool = 0, .default_pool = 0, .nb_pool_maps = 0, @@ -174,8 +157,9 @@ static struct rte_eth_conf vmdq_conf_default = { }, }; + static unsigned lcore_ids[RTE_MAX_LCORE]; -static uint8_t ports[RTE_MAX_ETHPORTS]; +static uint16_t ports[RTE_MAX_ETHPORTS]; static unsigned num_ports = 0; /**< The number of ports specified in command line */ static uint16_t num_pf_queues, num_vmdq_queues; static uint16_t vmdq_pool_base, vmdq_queue_base; @@ -193,7 +177,7 @@ const uint16_t vlan_tags[] = { }; /* ethernet addresses of ports */ -static struct ether_addr vmdq_ports_eth_addr[RTE_MAX_ETHPORTS]; +static struct rte_ether_addr vmdq_ports_eth_addr[RTE_MAX_ETHPORTS]; static struct vhost_dev_tailq_list vhost_dev_list = TAILQ_HEAD_INITIALIZER(vhost_dev_list); @@ -207,12 +191,191 @@ struct mbuf_table { struct rte_mbuf *m_table[MAX_PKT_BURST]; }; +struct vhost_bufftable { + uint32_t len; + uint64_t pre_tsc; + struct rte_mbuf *m_table[MAX_PKT_BURST]; +}; + /* TX queue for each data core. */ struct mbuf_table lcore_tx_queue[RTE_MAX_LCORE]; +/* + * Vhost TX buffer for each data core. + * Every data core maintains a TX buffer for every vhost device, + * which is used for batch pkts enqueue for higher performance. + */ +struct vhost_bufftable *vhost_txbuff[RTE_MAX_LCORE * RTE_MAX_VHOST_DEVICE]; + #define MBUF_TABLE_DRAIN_TSC ((rte_get_tsc_hz() + US_PER_S - 1) \ / US_PER_S * BURST_TX_DRAIN_US) -#define VLAN_HLEN 4 + +static int vid2socketid[RTE_MAX_VHOST_DEVICE]; + +static inline uint32_t +get_async_flag_by_socketid(int socketid) +{ + return dma_bind[socketid].async_flag; +} + +static inline void +init_vid2socketid_array(int vid, int socketid) +{ + vid2socketid[vid] = socketid; +} + +static inline bool +is_dma_configured(int16_t dev_id) +{ + int i; + + for (i = 0; i < dma_count; i++) + if (dmas_id[i] == dev_id) + return true; + return false; +} + +static inline int +open_dma(const char *value) +{ + struct dma_for_vhost *dma_info = dma_bind; + char *input = strndup(value, strlen(value) + 1); + char *addrs = input; + char *ptrs[2]; + char *start, *end, *substr; + int64_t socketid, vring_id; + + struct rte_dma_info info; + struct rte_dma_conf dev_config = { .nb_vchans = 1 }; + struct rte_dma_vchan_conf qconf = { + .direction = RTE_DMA_DIR_MEM_TO_MEM, + .nb_desc = DMA_RING_SIZE + }; + + int dev_id; + int ret = 0; + uint16_t i = 0; + char *dma_arg[RTE_MAX_VHOST_DEVICE]; + int args_nr; + + while (isblank(*addrs)) + addrs++; + if (*addrs == '\0') { + ret = -1; + goto out; + } + + /* process DMA devices within bracket. */ + addrs++; + substr = strtok(addrs, ";]"); + if (!substr) { + ret = -1; + goto out; + } + + args_nr = rte_strsplit(substr, strlen(substr), dma_arg, RTE_MAX_VHOST_DEVICE, ','); + if (args_nr <= 0) { + ret = -1; + goto out; + } + + while (i < args_nr) { + char *arg_temp = dma_arg[i]; + char *txd, *rxd; + uint8_t sub_nr; + int async_flag; + + sub_nr = rte_strsplit(arg_temp, strlen(arg_temp), ptrs, 2, '@'); + if (sub_nr != 2) { + ret = -1; + goto out; + } + + txd = strstr(ptrs[0], "txd"); + rxd = strstr(ptrs[0], "rxd"); + if (txd) { + start = txd; + vring_id = VIRTIO_RXQ; + async_flag = ASYNC_ENQUEUE_VHOST; + } else if (rxd) { + start = rxd; + vring_id = VIRTIO_TXQ; + async_flag = ASYNC_DEQUEUE_VHOST; + } else { + ret = -1; + goto out; + } + + start += 3; + socketid = strtol(start, &end, 0); + if (end == start) { + ret = -1; + goto out; + } + + dev_id = rte_dma_get_dev_id_by_name(ptrs[1]); + if (dev_id < 0) { + RTE_LOG(ERR, VHOST_CONFIG, "Fail to find DMA %s.\n", ptrs[1]); + ret = -1; + goto out; + } + + /* DMA device is already configured, so skip */ + if (is_dma_configured(dev_id)) + goto done; + + if (rte_dma_info_get(dev_id, &info) != 0) { + RTE_LOG(ERR, VHOST_CONFIG, "Error with rte_dma_info_get()\n"); + ret = -1; + goto out; + } + + if (info.max_vchans < 1) { + RTE_LOG(ERR, VHOST_CONFIG, "No channels available on device %d\n", dev_id); + ret = -1; + goto out; + } + + if (rte_dma_configure(dev_id, &dev_config) != 0) { + RTE_LOG(ERR, VHOST_CONFIG, "Fail to configure DMA %d.\n", dev_id); + ret = -1; + goto out; + } + + /* Check the max desc supported by DMA device */ + rte_dma_info_get(dev_id, &info); + if (info.nb_vchans != 1) { + RTE_LOG(ERR, VHOST_CONFIG, "No configured queues reported by DMA %d.\n", + dev_id); + ret = -1; + goto out; + } + + qconf.nb_desc = RTE_MIN(DMA_RING_SIZE, info.max_desc); + + if (rte_dma_vchan_setup(dev_id, 0, &qconf) != 0) { + RTE_LOG(ERR, VHOST_CONFIG, "Fail to set up DMA %d.\n", dev_id); + ret = -1; + goto out; + } + + if (rte_dma_start(dev_id) != 0) { + RTE_LOG(ERR, VHOST_CONFIG, "Fail to start DMA %u.\n", dev_id); + ret = -1; + goto out; + } + + dmas_id[dma_count++] = dev_id; + +done: + (dma_info + socketid)->dmas[vring_id].dev_id = dev_id; + (dma_info + socketid)->async_flag |= async_flag; + i++; + } +out: + free(input); + return ret; +} /* * Builds up the correct configuration for VMDQ VLAN pool map @@ -243,27 +406,12 @@ get_eth_conf(struct rte_eth_conf *eth_conf, uint32_t num_devices) return 0; } -/* - * Validate the device number according to the max pool number gotten form - * dev_info. If the device number is invalid, give the error message and - * return -1. Each device must have its own pool. - */ -static inline int -validate_num_devices(uint32_t max_nb_devices) -{ - if (num_devices > max_nb_devices) { - RTE_LOG(ERR, VHOST_PORT, "invalid number of devices\n"); - return -1; - } - return 0; -} - /* * Initialises a given port using global settings and with the rx buffers * coming from the mbuf_pool passed as parameter */ static inline int -port_init(uint8_t port) +port_init(uint16_t port) { struct rte_eth_dev_info dev_info; struct rte_eth_conf port_conf; @@ -275,42 +423,37 @@ port_init(uint8_t port) uint16_t q; /* The max pool number from dev_info will be used to validate the pool number specified in cmd line */ - rte_eth_dev_info_get (port, &dev_info); + retval = rte_eth_dev_info_get(port, &dev_info); + if (retval != 0) { + RTE_LOG(ERR, VHOST_PORT, + "Error during getting device (port %u) info: %s\n", + port, strerror(-retval)); - if (dev_info.max_rx_queues > MAX_QUEUES) { - rte_exit(EXIT_FAILURE, - "please define MAX_QUEUES no less than %u in %s\n", - dev_info.max_rx_queues, __FILE__); + return retval; + } + if (dev_info.max_vmdq_pools == 0) { + RTE_LOG(ERR, VHOST_PORT, "Failed to get VMDq info.\n"); + return -1; } rxconf = &dev_info.default_rxconf; txconf = &dev_info.default_txconf; rxconf->rx_drop_en = 1; - /* Enable vlan offload */ - txconf->txq_flags &= ~ETH_TXQ_FLAGS_NOVLANOFFL; - /*configure the number of supported virtio devices based on VMDQ limits */ num_devices = dev_info.max_vmdq_pools; rx_ring_size = RTE_TEST_RX_DESC_DEFAULT; tx_ring_size = RTE_TEST_TX_DESC_DEFAULT; - /* - * When dequeue zero copy is enabled, guest Tx used vring will be - * updated only when corresponding mbuf is freed. Thus, the nb_tx_desc - * (tx_ring_size here) must be small enough so that the driver will - * hit the free threshold easily and free mbufs timely. Otherwise, - * guest Tx vring would be starved. - */ - if (dequeue_zero_copy) - tx_ring_size = 64; - tx_rings = (uint16_t)rte_lcore_count(); - retval = validate_num_devices(MAX_DEVICES); - if (retval < 0) - return retval; + if (mergeable) { + if (dev_info.max_mtu != UINT16_MAX && dev_info.max_rx_pktlen > dev_info.max_mtu) + vmdq_conf_default.rxmode.mtu = dev_info.max_mtu; + else + vmdq_conf_default.rxmode.mtu = MAX_MTU; + } /* Get port configuration. */ retval = get_eth_conf(&port_conf, num_devices); @@ -326,19 +469,13 @@ port_init(uint8_t port) printf("pf queue num: %u, configured vmdq pool num: %u, each vmdq pool has %u queues\n", num_pf_queues, num_devices, queues_per_pool); - if (port >= rte_eth_dev_count()) return -1; - - if (enable_tx_csum == 0) - rte_vhost_feature_disable(1ULL << VIRTIO_NET_F_CSUM); - - if (enable_tso == 0) { - rte_vhost_feature_disable(1ULL << VIRTIO_NET_F_HOST_TSO4); - rte_vhost_feature_disable(1ULL << VIRTIO_NET_F_HOST_TSO6); - rte_vhost_feature_disable(1ULL << VIRTIO_NET_F_GUEST_TSO4); - rte_vhost_feature_disable(1ULL << VIRTIO_NET_F_GUEST_TSO6); - } + if (!rte_eth_dev_is_valid_port(port)) + return -1; rx_rings = (uint16_t)dev_info.max_rx_queues; + if (dev_info.tx_offload_capa & RTE_ETH_TX_OFFLOAD_MBUF_FAST_FREE) + port_conf.txmode.offloads |= + RTE_ETH_TX_OFFLOAD_MBUF_FAST_FREE; /* Configure ethernet device. */ retval = rte_eth_dev_configure(port, rx_rings, tx_rings, &port_conf); if (retval != 0) { @@ -347,7 +484,21 @@ port_init(uint8_t port) return retval; } + retval = rte_eth_dev_adjust_nb_rx_tx_desc(port, &rx_ring_size, + &tx_ring_size); + if (retval != 0) { + RTE_LOG(ERR, VHOST_PORT, "Failed to adjust number of descriptors " + "for port %u: %s.\n", port, strerror(-retval)); + return retval; + } + if (rx_ring_size > RTE_TEST_RX_DESC_DEFAULT) { + RTE_LOG(ERR, VHOST_PORT, "Mbuf pool has an insufficient size " + "for Rx queues on port %u.\n", port); + return -1; + } + /* Setup the queues. */ + rxconf->offloads = port_conf.rxmode.offloads; for (q = 0; q < rx_rings; q ++) { retval = rte_eth_rx_queue_setup(port, q, rx_ring_size, rte_eth_dev_socket_id(port), @@ -360,6 +511,7 @@ port_init(uint8_t port) return retval; } } + txconf->offloads = port_conf.txmode.offloads; for (q = 0; q < tx_rings; q ++) { retval = rte_eth_tx_queue_setup(port, q, tx_ring_size, rte_eth_dev_socket_id(port), @@ -380,20 +532,28 @@ port_init(uint8_t port) return retval; } - if (promiscuous) - rte_eth_promiscuous_enable(port); + if (promiscuous) { + retval = rte_eth_promiscuous_enable(port); + if (retval != 0) { + RTE_LOG(ERR, VHOST_PORT, + "Failed to enable promiscuous mode on port %u: %s\n", + port, rte_strerror(-retval)); + return retval; + } + } + + retval = rte_eth_macaddr_get(port, &vmdq_ports_eth_addr[port]); + if (retval < 0) { + RTE_LOG(ERR, VHOST_PORT, + "Failed to get MAC address on port %u: %s\n", + port, rte_strerror(-retval)); + return retval; + } - rte_eth_macaddr_get(port, &vmdq_ports_eth_addr[port]); RTE_LOG(INFO, VHOST_PORT, "Max virtio devices supported: %u\n", num_devices); RTE_LOG(INFO, VHOST_PORT, "Port %u MAC: %02"PRIx8" %02"PRIx8" %02"PRIx8 - " %02"PRIx8" %02"PRIx8" %02"PRIx8"\n", - (unsigned)port, - vmdq_ports_eth_addr[port].addr_bytes[0], - vmdq_ports_eth_addr[port].addr_bytes[1], - vmdq_ports_eth_addr[port].addr_bytes[2], - vmdq_ports_eth_addr[port].addr_bytes[3], - vmdq_ports_eth_addr[port].addr_bytes[4], - vmdq_ports_eth_addr[port].addr_bytes[5]); + " %02"PRIx8" %02"PRIx8" %02"PRIx8"\n", + port, RTE_ETHER_ADDR_BYTES(&vmdq_ports_eth_addr[port])); return 0; } @@ -404,12 +564,20 @@ port_init(uint8_t port) static int us_vhost_parse_socket_path(const char *q_arg) { + char *old; + /* parse number string */ - if (strnlen(q_arg, PATH_MAX) > PATH_MAX) + if (strnlen(q_arg, PATH_MAX) == PATH_MAX) return -1; + old = socket_files; socket_files = realloc(socket_files, PATH_MAX * (nb_sockets + 1)); - snprintf(socket_files + nb_sockets * PATH_MAX, PATH_MAX, "%s", q_arg); + if (socket_files == NULL) { + free(old); + return -1; + } + + strlcpy(socket_files + nb_sockets * PATH_MAX, q_arg, PATH_MAX); nb_sockets++; return 0; @@ -429,10 +597,7 @@ parse_portmask(const char *portmask) /* parse hexadecimal string */ pm = strtoul(portmask, &end, 16); if ((portmask[0] == '\0') || (end == NULL) || (*end != '\0') || (errno != 0)) - return -1; - - if (pm == 0) - return -1; + return 0; return pm; @@ -474,7 +639,7 @@ us_vhost_usage(const char *prgname) " --nb-devices ND\n" " -p PORTMASK: Set mask for ports to be used by application\n" " --vm2vm [0|1|2]: disable/software(default)/hardware vm2vm comms\n" - " --rx-retry [0|1]: disable/enable(default) retries on rx. Enable retry if destintation queue is full\n" + " --rx-retry [0|1]: disable/enable(default) retries on Rx. Enable retry if destination queue is full\n" " --rx-retry-delay [0-N]: timeout(in usecond) between retries on RX. This makes effect only if retries on rx enabled\n" " --rx-retry-num [0-N]: the number of retries on rx. This makes effect only if retries on rx enabled\n" " --mergeable [0|1]: disable(default)/enable RX mergeable buffers\n" @@ -483,10 +648,40 @@ us_vhost_usage(const char *prgname) " --tx-csum [0|1] disable/enable TX checksum offload.\n" " --tso [0|1] disable/enable TCP segment offload.\n" " --client register a vhost-user socket as client mode.\n" - " --dequeue-zero-copy enables dequeue zero copy\n", + " --dmas register dma channel for specific vhost device.\n" + " --total-num-mbufs [0-N] set the number of mbufs to be allocated in mbuf pools, the default value is 147456.\n", prgname); } +enum { +#define OPT_VM2VM "vm2vm" + OPT_VM2VM_NUM = 256, +#define OPT_RX_RETRY "rx-retry" + OPT_RX_RETRY_NUM, +#define OPT_RX_RETRY_DELAY "rx-retry-delay" + OPT_RX_RETRY_DELAY_NUM, +#define OPT_RX_RETRY_NUMB "rx-retry-num" + OPT_RX_RETRY_NUMB_NUM, +#define OPT_MERGEABLE "mergeable" + OPT_MERGEABLE_NUM, +#define OPT_STATS "stats" + OPT_STATS_NUM, +#define OPT_SOCKET_FILE "socket-file" + OPT_SOCKET_FILE_NUM, +#define OPT_TX_CSUM "tx-csum" + OPT_TX_CSUM_NUM, +#define OPT_TSO "tso" + OPT_TSO_NUM, +#define OPT_CLIENT "client" + OPT_CLIENT_NUM, +#define OPT_BUILTIN_NET_DRIVER "builtin-net-driver" + OPT_BUILTIN_NET_DRIVER_NUM, +#define OPT_DMAS "dmas" + OPT_DMAS_NUM, +#define OPT_NUM_MBUFS "total-num-mbufs" + OPT_NUM_MBUFS_NUM, +}; + /* * Parse the arguments given in the command line of the application. */ @@ -498,17 +693,32 @@ us_vhost_parse_args(int argc, char **argv) unsigned i; const char *prgname = argv[0]; static struct option long_option[] = { - {"vm2vm", required_argument, NULL, 0}, - {"rx-retry", required_argument, NULL, 0}, - {"rx-retry-delay", required_argument, NULL, 0}, - {"rx-retry-num", required_argument, NULL, 0}, - {"mergeable", required_argument, NULL, 0}, - {"stats", required_argument, NULL, 0}, - {"socket-file", required_argument, NULL, 0}, - {"tx-csum", required_argument, NULL, 0}, - {"tso", required_argument, NULL, 0}, - {"client", no_argument, &client_mode, 1}, - {"dequeue-zero-copy", no_argument, &dequeue_zero_copy, 1}, + {OPT_VM2VM, required_argument, + NULL, OPT_VM2VM_NUM}, + {OPT_RX_RETRY, required_argument, + NULL, OPT_RX_RETRY_NUM}, + {OPT_RX_RETRY_DELAY, required_argument, + NULL, OPT_RX_RETRY_DELAY_NUM}, + {OPT_RX_RETRY_NUMB, required_argument, + NULL, OPT_RX_RETRY_NUMB_NUM}, + {OPT_MERGEABLE, required_argument, + NULL, OPT_MERGEABLE_NUM}, + {OPT_STATS, required_argument, + NULL, OPT_STATS_NUM}, + {OPT_SOCKET_FILE, required_argument, + NULL, OPT_SOCKET_FILE_NUM}, + {OPT_TX_CSUM, required_argument, + NULL, OPT_TX_CSUM_NUM}, + {OPT_TSO, required_argument, + NULL, OPT_TSO_NUM}, + {OPT_CLIENT, no_argument, + NULL, OPT_CLIENT_NUM}, + {OPT_BUILTIN_NET_DRIVER, no_argument, + NULL, OPT_BUILTIN_NET_DRIVER_NUM}, + {OPT_DMAS, required_argument, + NULL, OPT_DMAS_NUM}, + {OPT_NUM_MBUFS, required_argument, + NULL, OPT_NUM_MBUFS_NUM}, {NULL, 0, 0, 0}, }; @@ -529,131 +739,135 @@ us_vhost_parse_args(int argc, char **argv) case 'P': promiscuous = 1; vmdq_conf_default.rx_adv_conf.vmdq_rx_conf.rx_mode = - ETH_VMDQ_ACCEPT_BROADCAST | - ETH_VMDQ_ACCEPT_MULTICAST; - rte_vhost_feature_enable(1ULL << VIRTIO_NET_F_CTRL_RX); + RTE_ETH_VMDQ_ACCEPT_BROADCAST | + RTE_ETH_VMDQ_ACCEPT_MULTICAST; + break; + case OPT_VM2VM_NUM: + ret = parse_num_opt(optarg, (VM2VM_LAST - 1)); + if (ret == -1) { + RTE_LOG(INFO, VHOST_CONFIG, + "Invalid argument for " + "vm2vm [0|1|2]\n"); + us_vhost_usage(prgname); + return -1; + } + vm2vm_mode = (vm2vm_type)ret; + break; + + case OPT_RX_RETRY_NUM: + ret = parse_num_opt(optarg, 1); + if (ret == -1) { + RTE_LOG(INFO, VHOST_CONFIG, "Invalid argument for rx-retry [0|1]\n"); + us_vhost_usage(prgname); + return -1; + } + enable_retry = ret; break; - case 0: - /* Enable/disable vm2vm comms. */ - if (!strncmp(long_option[option_index].name, "vm2vm", - MAX_LONG_OPT_SZ)) { - ret = parse_num_opt(optarg, (VM2VM_LAST - 1)); - if (ret == -1) { - RTE_LOG(INFO, VHOST_CONFIG, - "Invalid argument for " - "vm2vm [0|1|2]\n"); - us_vhost_usage(prgname); - return -1; - } else { - vm2vm_mode = (vm2vm_type)ret; - } + case OPT_TX_CSUM_NUM: + ret = parse_num_opt(optarg, 1); + if (ret == -1) { + RTE_LOG(INFO, VHOST_CONFIG, "Invalid argument for tx-csum [0|1]\n"); + us_vhost_usage(prgname); + return -1; } + enable_tx_csum = ret; + break; - /* Enable/disable retries on RX. */ - if (!strncmp(long_option[option_index].name, "rx-retry", MAX_LONG_OPT_SZ)) { - ret = parse_num_opt(optarg, 1); - if (ret == -1) { - RTE_LOG(INFO, VHOST_CONFIG, "Invalid argument for rx-retry [0|1]\n"); - us_vhost_usage(prgname); - return -1; - } else { - enable_retry = ret; - } + case OPT_TSO_NUM: + ret = parse_num_opt(optarg, 1); + if (ret == -1) { + RTE_LOG(INFO, VHOST_CONFIG, "Invalid argument for tso [0|1]\n"); + us_vhost_usage(prgname); + return -1; } + enable_tso = ret; + break; - /* Enable/disable TX checksum offload. */ - if (!strncmp(long_option[option_index].name, "tx-csum", MAX_LONG_OPT_SZ)) { - ret = parse_num_opt(optarg, 1); - if (ret == -1) { - RTE_LOG(INFO, VHOST_CONFIG, "Invalid argument for tx-csum [0|1]\n"); - us_vhost_usage(prgname); - return -1; - } else - enable_tx_csum = ret; + case OPT_RX_RETRY_DELAY_NUM: + ret = parse_num_opt(optarg, INT32_MAX); + if (ret == -1) { + RTE_LOG(INFO, VHOST_CONFIG, "Invalid argument for rx-retry-delay [0-N]\n"); + us_vhost_usage(prgname); + return -1; } + burst_rx_delay_time = ret; + break; - /* Enable/disable TSO offload. */ - if (!strncmp(long_option[option_index].name, "tso", MAX_LONG_OPT_SZ)) { - ret = parse_num_opt(optarg, 1); - if (ret == -1) { - RTE_LOG(INFO, VHOST_CONFIG, "Invalid argument for tso [0|1]\n"); - us_vhost_usage(prgname); - return -1; - } else - enable_tso = ret; + case OPT_RX_RETRY_NUMB_NUM: + ret = parse_num_opt(optarg, INT32_MAX); + if (ret == -1) { + RTE_LOG(INFO, VHOST_CONFIG, "Invalid argument for rx-retry-num [0-N]\n"); + us_vhost_usage(prgname); + return -1; } + burst_rx_retry_num = ret; + break; - /* Specify the retries delay time (in useconds) on RX. */ - if (!strncmp(long_option[option_index].name, "rx-retry-delay", MAX_LONG_OPT_SZ)) { - ret = parse_num_opt(optarg, INT32_MAX); - if (ret == -1) { - RTE_LOG(INFO, VHOST_CONFIG, "Invalid argument for rx-retry-delay [0-N]\n"); - us_vhost_usage(prgname); - return -1; - } else { - burst_rx_delay_time = ret; - } + case OPT_MERGEABLE_NUM: + ret = parse_num_opt(optarg, 1); + if (ret == -1) { + RTE_LOG(INFO, VHOST_CONFIG, "Invalid argument for mergeable [0|1]\n"); + us_vhost_usage(prgname); + return -1; } + mergeable = !!ret; + break; - /* Specify the retries number on RX. */ - if (!strncmp(long_option[option_index].name, "rx-retry-num", MAX_LONG_OPT_SZ)) { - ret = parse_num_opt(optarg, INT32_MAX); - if (ret == -1) { - RTE_LOG(INFO, VHOST_CONFIG, "Invalid argument for rx-retry-num [0-N]\n"); - us_vhost_usage(prgname); - return -1; - } else { - burst_rx_retry_num = ret; - } + case OPT_STATS_NUM: + ret = parse_num_opt(optarg, INT32_MAX); + if (ret == -1) { + RTE_LOG(INFO, VHOST_CONFIG, + "Invalid argument for stats [0..N]\n"); + us_vhost_usage(prgname); + return -1; } + enable_stats = ret; + break; - /* Enable/disable RX mergeable buffers. */ - if (!strncmp(long_option[option_index].name, "mergeable", MAX_LONG_OPT_SZ)) { - ret = parse_num_opt(optarg, 1); - if (ret == -1) { - RTE_LOG(INFO, VHOST_CONFIG, "Invalid argument for mergeable [0|1]\n"); - us_vhost_usage(prgname); - return -1; - } else { - mergeable = !!ret; - if (ret) { - vmdq_conf_default.rxmode.jumbo_frame = 1; - vmdq_conf_default.rxmode.max_rx_pkt_len - = JUMBO_FRAME_MAX_SIZE; - } - } + /* Set socket file path. */ + case OPT_SOCKET_FILE_NUM: + if (us_vhost_parse_socket_path(optarg) == -1) { + RTE_LOG(INFO, VHOST_CONFIG, + "Invalid argument for socket name (Max %d characters)\n", + PATH_MAX); + us_vhost_usage(prgname); + return -1; } + break; - /* Enable/disable stats. */ - if (!strncmp(long_option[option_index].name, "stats", MAX_LONG_OPT_SZ)) { - ret = parse_num_opt(optarg, INT32_MAX); - if (ret == -1) { - RTE_LOG(INFO, VHOST_CONFIG, - "Invalid argument for stats [0..N]\n"); - us_vhost_usage(prgname); - return -1; - } else { - enable_stats = ret; - } + case OPT_DMAS_NUM: + if (open_dma(optarg) == -1) { + RTE_LOG(INFO, VHOST_CONFIG, + "Wrong DMA args\n"); + us_vhost_usage(prgname); + return -1; } + break; - /* Set socket file path. */ - if (!strncmp(long_option[option_index].name, - "socket-file", MAX_LONG_OPT_SZ)) { - if (us_vhost_parse_socket_path(optarg) == -1) { - RTE_LOG(INFO, VHOST_CONFIG, - "Invalid argument for socket name (Max %d characters)\n", - PATH_MAX); - us_vhost_usage(prgname); - return -1; - } + case OPT_NUM_MBUFS_NUM: + ret = parse_num_opt(optarg, INT32_MAX); + if (ret == -1) { + RTE_LOG(INFO, VHOST_CONFIG, + "Invalid argument for total-num-mbufs [0..N]\n"); + us_vhost_usage(prgname); + return -1; } + if (total_num_mbufs < ret) + total_num_mbufs = ret; break; - /* Invalid option - print options. */ + case OPT_CLIENT_NUM: + client_mode = 1; + break; + + case OPT_BUILTIN_NET_DRIVER_NUM: + builtin_net_driver = 1; + break; + + /* Invalid option - print options. */ default: us_vhost_usage(prgname); return -1; @@ -662,7 +876,7 @@ us_vhost_parse_args(int argc, char **argv) for (i = 0; i < RTE_MAX_ETHPORTS; i++) { if (enabled_port_mask & (1 << i)) - ports[num_ports++] = (uint8_t)i; + ports[num_ports++] = i; } if ((num_ports == 0) || (num_ports > MAX_SUP_PORTS)) { @@ -690,9 +904,10 @@ static unsigned check_ports_num(unsigned nb_ports) } for (portid = 0; portid < num_ports; portid ++) { - if (ports[portid] >= nb_ports) { - RTE_LOG(INFO, VHOST_PORT, "\nSpecified port ID(%u) exceeds max system port ID(%u)\n", - ports[portid], (nb_ports - 1)); + if (!rte_eth_dev_is_valid_port(ports[portid])) { + RTE_LOG(INFO, VHOST_PORT, + "\nSpecified port ID(%u) is not valid\n", + ports[portid]); ports[portid] = INVALID_PORT_ID; valid_num_ports--; } @@ -700,14 +915,14 @@ static unsigned check_ports_num(unsigned nb_ports) return valid_num_ports; } -static inline struct vhost_dev *__attribute__((always_inline)) -find_vhost_dev(struct ether_addr *mac) +static __rte_always_inline struct vhost_dev * +find_vhost_dev(struct rte_ether_addr *mac) { struct vhost_dev *vdev; TAILQ_FOREACH(vdev, &vhost_dev_list, global_vdev_entry) { if (vdev->ready == DEVICE_RX && - is_same_ether_addr(mac, &vdev->mac_address)) + rte_is_same_ether_addr(mac, &vdev->mac_address)) return vdev; } @@ -721,32 +936,30 @@ find_vhost_dev(struct ether_addr *mac) static int link_vmdq(struct vhost_dev *vdev, struct rte_mbuf *m) { - struct ether_hdr *pkt_hdr; + struct rte_ether_hdr *pkt_hdr; int i, ret; /* Learn MAC address of guest device from packet */ - pkt_hdr = rte_pktmbuf_mtod(m, struct ether_hdr *); + pkt_hdr = rte_pktmbuf_mtod(m, struct rte_ether_hdr *); - if (find_vhost_dev(&pkt_hdr->s_addr)) { + if (find_vhost_dev(&pkt_hdr->src_addr)) { RTE_LOG(ERR, VHOST_DATA, "(%d) device is using a registered MAC!\n", vdev->vid); return -1; } - for (i = 0; i < ETHER_ADDR_LEN; i++) - vdev->mac_address.addr_bytes[i] = pkt_hdr->s_addr.addr_bytes[i]; + for (i = 0; i < RTE_ETHER_ADDR_LEN; i++) + vdev->mac_address.addr_bytes[i] = + pkt_hdr->src_addr.addr_bytes[i]; /* vlan_tag currently uses the device_id. */ vdev->vlan_tag = vlan_tags[vdev->vid]; /* Print out VMDQ registration info. */ RTE_LOG(INFO, VHOST_DATA, - "(%d) mac %02x:%02x:%02x:%02x:%02x:%02x and vlan %d registered\n", - vdev->vid, - vdev->mac_address.addr_bytes[0], vdev->mac_address.addr_bytes[1], - vdev->mac_address.addr_bytes[2], vdev->mac_address.addr_bytes[3], - vdev->mac_address.addr_bytes[4], vdev->mac_address.addr_bytes[5], + "(%d) mac " RTE_ETHER_ADDR_PRT_FMT " and vlan %d registered\n", + vdev->vid, RTE_ETHER_ADDR_BYTES(&vdev->mac_address), vdev->vlan_tag); /* Register the MAC address. */ @@ -800,54 +1013,143 @@ unlink_vmdq(struct vhost_dev *vdev) } } -static inline void __attribute__((always_inline)) -virtio_xmit(struct vhost_dev *dst_vdev, struct vhost_dev *src_vdev, +static inline void +free_pkts(struct rte_mbuf **pkts, uint16_t n) +{ + while (n--) + rte_pktmbuf_free(pkts[n]); +} + +static __rte_always_inline void +complete_async_pkts(struct vhost_dev *vdev) +{ + struct rte_mbuf *p_cpl[MAX_PKT_BURST]; + uint16_t complete_count; + int16_t dma_id = dma_bind[vid2socketid[vdev->vid]].dmas[VIRTIO_RXQ].dev_id; + + complete_count = rte_vhost_poll_enqueue_completed(vdev->vid, + VIRTIO_RXQ, p_cpl, MAX_PKT_BURST, dma_id, 0); + if (complete_count) + free_pkts(p_cpl, complete_count); + +} + +static __rte_always_inline void +sync_virtio_xmit(struct vhost_dev *dst_vdev, struct vhost_dev *src_vdev, struct rte_mbuf *m) { uint16_t ret; - ret = rte_vhost_enqueue_burst(dst_vdev->vid, VIRTIO_RXQ, &m, 1); + if (builtin_net_driver) { + ret = vs_enqueue_pkts(dst_vdev, VIRTIO_RXQ, &m, 1); + } else { + ret = rte_vhost_enqueue_burst(dst_vdev->vid, VIRTIO_RXQ, &m, 1); + } + if (enable_stats) { - rte_atomic64_inc(&dst_vdev->stats.rx_total_atomic); - rte_atomic64_add(&dst_vdev->stats.rx_atomic, ret); + __atomic_add_fetch(&dst_vdev->stats.rx_total_atomic, 1, + __ATOMIC_SEQ_CST); + __atomic_add_fetch(&dst_vdev->stats.rx_atomic, ret, + __ATOMIC_SEQ_CST); src_vdev->stats.tx_total++; src_vdev->stats.tx += ret; } } +static __rte_always_inline void +drain_vhost(struct vhost_dev *vdev) +{ + uint16_t ret; + uint32_t buff_idx = rte_lcore_id() * RTE_MAX_VHOST_DEVICE + vdev->vid; + uint16_t nr_xmit = vhost_txbuff[buff_idx]->len; + struct rte_mbuf **m = vhost_txbuff[buff_idx]->m_table; + + ret = vdev_queue_ops[vdev->vid].enqueue_pkt_burst(vdev, VIRTIO_RXQ, m, nr_xmit); + + if (enable_stats) { + __atomic_add_fetch(&vdev->stats.rx_total_atomic, nr_xmit, + __ATOMIC_SEQ_CST); + __atomic_add_fetch(&vdev->stats.rx_atomic, ret, + __ATOMIC_SEQ_CST); + } + + if (!dma_bind[vid2socketid[vdev->vid]].dmas[VIRTIO_RXQ].async_enabled) + free_pkts(m, nr_xmit); +} + +static __rte_always_inline void +drain_vhost_table(void) +{ + uint16_t lcore_id = rte_lcore_id(); + struct vhost_bufftable *vhost_txq; + struct vhost_dev *vdev; + uint64_t cur_tsc; + + TAILQ_FOREACH(vdev, &vhost_dev_list, global_vdev_entry) { + if (unlikely(vdev->remove == 1)) + continue; + + vhost_txq = vhost_txbuff[lcore_id * RTE_MAX_VHOST_DEVICE + vdev->vid]; + + cur_tsc = rte_rdtsc(); + if (unlikely(cur_tsc - vhost_txq->pre_tsc + > MBUF_TABLE_DRAIN_TSC)) { + RTE_LOG_DP(DEBUG, VHOST_DATA, + "Vhost TX queue drained after timeout with burst size %u\n", + vhost_txq->len); + drain_vhost(vdev); + vhost_txq->len = 0; + vhost_txq->pre_tsc = cur_tsc; + } + } +} + /* * Check if the packet destination MAC address is for a local device. If so then put * the packet on that devices RX queue. If not then return. */ -static inline int __attribute__((always_inline)) +static __rte_always_inline int virtio_tx_local(struct vhost_dev *vdev, struct rte_mbuf *m) { - struct ether_hdr *pkt_hdr; + struct rte_ether_hdr *pkt_hdr; struct vhost_dev *dst_vdev; + struct vhost_bufftable *vhost_txq; + uint16_t lcore_id = rte_lcore_id(); + pkt_hdr = rte_pktmbuf_mtod(m, struct rte_ether_hdr *); - pkt_hdr = rte_pktmbuf_mtod(m, struct ether_hdr *); - - dst_vdev = find_vhost_dev(&pkt_hdr->d_addr); + dst_vdev = find_vhost_dev(&pkt_hdr->dst_addr); if (!dst_vdev) return -1; if (vdev->vid == dst_vdev->vid) { - RTE_LOG(DEBUG, VHOST_DATA, + RTE_LOG_DP(DEBUG, VHOST_DATA, "(%d) TX: src and dst MAC is same. Dropping packet.\n", vdev->vid); return 0; } - RTE_LOG(DEBUG, VHOST_DATA, + RTE_LOG_DP(DEBUG, VHOST_DATA, "(%d) TX: MAC address is local\n", dst_vdev->vid); if (unlikely(dst_vdev->remove)) { - RTE_LOG(DEBUG, VHOST_DATA, + RTE_LOG_DP(DEBUG, VHOST_DATA, "(%d) device is marked for removal\n", dst_vdev->vid); return 0; } - virtio_xmit(dst_vdev, vdev, m); + vhost_txq = vhost_txbuff[lcore_id * RTE_MAX_VHOST_DEVICE + dst_vdev->vid]; + vhost_txq->m_table[vhost_txq->len++] = m; + + if (enable_stats) { + vdev->stats.tx_total++; + vdev->stats.tx++; + } + + if (unlikely(vhost_txq->len == MAX_PKT_BURST)) { + drain_vhost(dst_vdev); + vhost_txq->len = 0; + vhost_txq->pre_tsc = rte_rdtsc(); + } return 0; } @@ -855,19 +1157,20 @@ virtio_tx_local(struct vhost_dev *vdev, struct rte_mbuf *m) * Check if the destination MAC of a packet is one local VM, * and get its vlan tag, and offset if it is. */ -static inline int __attribute__((always_inline)) +static __rte_always_inline int find_local_dest(struct vhost_dev *vdev, struct rte_mbuf *m, uint32_t *offset, uint16_t *vlan_tag) { struct vhost_dev *dst_vdev; - struct ether_hdr *pkt_hdr = rte_pktmbuf_mtod(m, struct ether_hdr *); + struct rte_ether_hdr *pkt_hdr = + rte_pktmbuf_mtod(m, struct rte_ether_hdr *); - dst_vdev = find_vhost_dev(&pkt_hdr->d_addr); + dst_vdev = find_vhost_dev(&pkt_hdr->dst_addr); if (!dst_vdev) return 0; if (vdev->vid == dst_vdev->vid) { - RTE_LOG(DEBUG, VHOST_DATA, + RTE_LOG_DP(DEBUG, VHOST_DATA, "(%d) TX: src and dst MAC is same. Dropping packet.\n", vdev->vid); return -1; @@ -878,52 +1181,47 @@ find_local_dest(struct vhost_dev *vdev, struct rte_mbuf *m, * by minus length of vlan tag, so need restore * the packet length by plus it. */ - *offset = VLAN_HLEN; + *offset = RTE_VLAN_HLEN; *vlan_tag = vlan_tags[vdev->vid]; - RTE_LOG(DEBUG, VHOST_DATA, + RTE_LOG_DP(DEBUG, VHOST_DATA, "(%d) TX: pkt to local VM device id: (%d), vlan tag: %u.\n", vdev->vid, dst_vdev->vid, *vlan_tag); return 0; } -static uint16_t -get_psd_sum(void *l3_hdr, uint64_t ol_flags) -{ - if (ol_flags & PKT_TX_IPV4) - return rte_ipv4_phdr_cksum(l3_hdr, ol_flags); - else /* assume ethertype == ETHER_TYPE_IPv6 */ - return rte_ipv6_phdr_cksum(l3_hdr, ol_flags); -} - static void virtio_tx_offload(struct rte_mbuf *m) { + struct rte_net_hdr_lens hdr_lens; + struct rte_ipv4_hdr *ipv4_hdr; + struct rte_tcp_hdr *tcp_hdr; + uint32_t ptype; void *l3_hdr; - struct ipv4_hdr *ipv4_hdr = NULL; - struct tcp_hdr *tcp_hdr = NULL; - struct ether_hdr *eth_hdr = rte_pktmbuf_mtod(m, struct ether_hdr *); - l3_hdr = (char *)eth_hdr + m->l2_len; + ptype = rte_net_get_ptype(m, &hdr_lens, RTE_PTYPE_ALL_MASK); + m->l2_len = hdr_lens.l2_len; + m->l3_len = hdr_lens.l3_len; + m->l4_len = hdr_lens.l4_len; + + l3_hdr = rte_pktmbuf_mtod_offset(m, void *, m->l2_len); + tcp_hdr = rte_pktmbuf_mtod_offset(m, struct rte_tcp_hdr *, + m->l2_len + m->l3_len); - if (m->ol_flags & PKT_TX_IPV4) { + m->ol_flags |= RTE_MBUF_F_TX_TCP_SEG; + if ((ptype & RTE_PTYPE_L3_MASK) == RTE_PTYPE_L3_IPV4) { + m->ol_flags |= RTE_MBUF_F_TX_IPV4; + m->ol_flags |= RTE_MBUF_F_TX_IP_CKSUM; ipv4_hdr = l3_hdr; ipv4_hdr->hdr_checksum = 0; - m->ol_flags |= PKT_TX_IP_CKSUM; + tcp_hdr->cksum = rte_ipv4_phdr_cksum(l3_hdr, m->ol_flags); + } else { /* assume ethertype == RTE_ETHER_TYPE_IPV6 */ + m->ol_flags |= RTE_MBUF_F_TX_IPV6; + tcp_hdr->cksum = rte_ipv6_phdr_cksum(l3_hdr, m->ol_flags); } - - tcp_hdr = (struct tcp_hdr *)((char *)l3_hdr + m->l3_len); - tcp_hdr->cksum = get_psd_sum(l3_hdr, m->ol_flags); } -static inline void -free_pkts(struct rte_mbuf **pkts, uint16_t n) -{ - while (n--) - rte_pktmbuf_free(pkts[n]); -} - -static inline void __attribute__((always_inline)) +static __rte_always_inline void do_drain_mbuf_table(struct mbuf_table *tx_q) { uint16_t count; @@ -940,30 +1238,29 @@ do_drain_mbuf_table(struct mbuf_table *tx_q) * This function routes the TX packet to the correct interface. This * may be a local device or the physical port. */ -static inline void __attribute__((always_inline)) +static __rte_always_inline void virtio_tx_route(struct vhost_dev *vdev, struct rte_mbuf *m, uint16_t vlan_tag) { struct mbuf_table *tx_q; unsigned offset = 0; const uint16_t lcore_id = rte_lcore_id(); - struct ether_hdr *nh; + struct rte_ether_hdr *nh; - nh = rte_pktmbuf_mtod(m, struct ether_hdr *); - if (unlikely(is_broadcast_ether_addr(&nh->d_addr))) { + nh = rte_pktmbuf_mtod(m, struct rte_ether_hdr *); + if (unlikely(rte_is_broadcast_ether_addr(&nh->dst_addr))) { struct vhost_dev *vdev2; TAILQ_FOREACH(vdev2, &vhost_dev_list, global_vdev_entry) { - virtio_xmit(vdev2, vdev, m); + if (vdev2 != vdev) + sync_virtio_xmit(vdev2, vdev, m); } goto queue2nic; } /*check if destination is local VM*/ - if ((vm2vm_mode == VM2VM_SOFTWARE) && (virtio_tx_local(vdev, m) == 0)) { - rte_pktmbuf_free(m); + if ((vm2vm_mode == VM2VM_SOFTWARE) && (virtio_tx_local(vdev, m) == 0)) return; - } if (unlikely(vm2vm_mode == VM2VM_HARDWARE)) { if (unlikely(find_local_dest(vdev, m, &offset, @@ -973,7 +1270,7 @@ virtio_tx_route(struct vhost_dev *vdev, struct rte_mbuf *m, uint16_t vlan_tag) } } - RTE_LOG(DEBUG, VHOST_DATA, + RTE_LOG_DP(DEBUG, VHOST_DATA, "(%d) TX: MAC address is external\n", vdev->vid); queue2nic: @@ -981,16 +1278,16 @@ queue2nic: /*Add packet to the port tx queue*/ tx_q = &lcore_tx_queue[lcore_id]; - nh = rte_pktmbuf_mtod(m, struct ether_hdr *); - if (unlikely(nh->ether_type == rte_cpu_to_be_16(ETHER_TYPE_VLAN))) { + nh = rte_pktmbuf_mtod(m, struct rte_ether_hdr *); + if (unlikely(nh->ether_type == rte_cpu_to_be_16(RTE_ETHER_TYPE_VLAN))) { /* Guest has inserted the vlan tag. */ - struct vlan_hdr *vh = (struct vlan_hdr *) (nh + 1); + struct rte_vlan_hdr *vh = (struct rte_vlan_hdr *) (nh + 1); uint16_t vlan_tag_be = rte_cpu_to_be_16(vlan_tag); if ((vm2vm_mode == VM2VM_HARDWARE) && (vh->vlan_tci != vlan_tag_be)) vh->vlan_tci = vlan_tag_be; } else { - m->ol_flags |= PKT_TX_VLAN_PKT; + m->ol_flags |= RTE_MBUF_F_TX_VLAN; /* * Find the right seg to adjust the data len when offset is @@ -1014,7 +1311,7 @@ queue2nic: m->vlan_tci = vlan_tag; } - if (m->ol_flags & PKT_TX_TCP_SEG) + if (m->ol_flags & RTE_MBUF_F_RX_LRO) virtio_tx_offload(m); tx_q->m_table[tx_q->len++] = m; @@ -1028,7 +1325,7 @@ queue2nic: } -static inline void __attribute__((always_inline)) +static __rte_always_inline void drain_mbuf_table(struct mbuf_table *tx_q) { static uint64_t prev_tsc; @@ -1041,14 +1338,40 @@ drain_mbuf_table(struct mbuf_table *tx_q) if (unlikely(cur_tsc - prev_tsc > MBUF_TABLE_DRAIN_TSC)) { prev_tsc = cur_tsc; - RTE_LOG(DEBUG, VHOST_DATA, + RTE_LOG_DP(DEBUG, VHOST_DATA, "TX queue drained after timeout with burst size %u\n", tx_q->len); do_drain_mbuf_table(tx_q); } } -static inline void __attribute__((always_inline)) +uint16_t +async_enqueue_pkts(struct vhost_dev *dev, uint16_t queue_id, + struct rte_mbuf **pkts, uint32_t rx_count) +{ + uint16_t enqueue_count; + uint16_t enqueue_fail = 0; + uint16_t dma_id = dma_bind[vid2socketid[dev->vid]].dmas[VIRTIO_RXQ].dev_id; + + complete_async_pkts(dev); + enqueue_count = rte_vhost_submit_enqueue_burst(dev->vid, queue_id, + pkts, rx_count, dma_id, 0); + + enqueue_fail = rx_count - enqueue_count; + if (enqueue_fail) + free_pkts(&pkts[enqueue_count], enqueue_fail); + + return enqueue_count; +} + +uint16_t +sync_enqueue_pkts(struct vhost_dev *dev, uint16_t queue_id, + struct rte_mbuf **pkts, uint32_t rx_count) +{ + return rte_vhost_enqueue_burst(dev->vid, queue_id, pkts, rx_count); +} + +static __rte_always_inline void drain_eth_rx(struct vhost_dev *vdev) { uint16_t rx_count, enqueue_count; @@ -1056,6 +1379,7 @@ drain_eth_rx(struct vhost_dev *vdev) rx_count = rte_eth_rx_burst(ports[0], vdev->vmdq_rx_q, pkts, MAX_PKT_BURST); + if (!rx_count) return; @@ -1077,25 +1401,50 @@ drain_eth_rx(struct vhost_dev *vdev) } } - enqueue_count = rte_vhost_enqueue_burst(vdev->vid, VIRTIO_RXQ, - pkts, rx_count); + enqueue_count = vdev_queue_ops[vdev->vid].enqueue_pkt_burst(vdev, + VIRTIO_RXQ, pkts, rx_count); + if (enable_stats) { - rte_atomic64_add(&vdev->stats.rx_total_atomic, rx_count); - rte_atomic64_add(&vdev->stats.rx_atomic, enqueue_count); + __atomic_add_fetch(&vdev->stats.rx_total_atomic, rx_count, + __ATOMIC_SEQ_CST); + __atomic_add_fetch(&vdev->stats.rx_atomic, enqueue_count, + __ATOMIC_SEQ_CST); } - free_pkts(pkts, rx_count); + if (!dma_bind[vid2socketid[vdev->vid]].dmas[VIRTIO_RXQ].async_enabled) + free_pkts(pkts, rx_count); } -static inline void __attribute__((always_inline)) +uint16_t async_dequeue_pkts(struct vhost_dev *dev, uint16_t queue_id, + struct rte_mempool *mbuf_pool, + struct rte_mbuf **pkts, uint16_t count) +{ + int nr_inflight; + uint16_t dequeue_count; + int16_t dma_id = dma_bind[vid2socketid[dev->vid]].dmas[VIRTIO_TXQ].dev_id; + + dequeue_count = rte_vhost_async_try_dequeue_burst(dev->vid, queue_id, + mbuf_pool, pkts, count, &nr_inflight, dma_id, 0); + + return dequeue_count; +} + +uint16_t sync_dequeue_pkts(struct vhost_dev *dev, uint16_t queue_id, + struct rte_mempool *mbuf_pool, + struct rte_mbuf **pkts, uint16_t count) +{ + return rte_vhost_dequeue_burst(dev->vid, queue_id, mbuf_pool, pkts, count); +} + +static __rte_always_inline void drain_virtio_tx(struct vhost_dev *vdev) { struct rte_mbuf *pkts[MAX_PKT_BURST]; uint16_t count; uint16_t i; - count = rte_vhost_dequeue_burst(vdev->vid, VIRTIO_TXQ, mbuf_pool, - pkts, MAX_PKT_BURST); + count = vdev_queue_ops[vdev->vid].dequeue_pkt_burst(vdev, + VIRTIO_TXQ, mbuf_pool, pkts, MAX_PKT_BURST); /* setup VMDq for the first packet */ if (unlikely(vdev->ready == DEVICE_MAC_LEARNING) && count) { @@ -1132,7 +1481,7 @@ switch_worker(void *arg __rte_unused) struct vhost_dev *vdev; struct mbuf_table *tx_q; - RTE_LOG(INFO, VHOST_DATA, "Procesing on Core %u started\n", lcore_id); + RTE_LOG(INFO, VHOST_DATA, "Processing on Core %u started\n", lcore_id); tx_q = &lcore_tx_queue[lcore_id]; for (i = 0; i < rte_lcore_count(); i++) { @@ -1144,7 +1493,7 @@ switch_worker(void *arg __rte_unused) while(1) { drain_mbuf_table(tx_q); - + drain_vhost_table(); /* * Inform the configuration core that we have exited the * linked list and that no devices are in use if requested. @@ -1174,9 +1523,29 @@ switch_worker(void *arg __rte_unused) return 0; } +static void +vhost_clear_queue_thread_unsafe(struct vhost_dev *vdev, uint16_t queue_id) +{ + uint16_t n_pkt = 0; + int pkts_inflight; + + int16_t dma_id = dma_bind[vid2socketid[vdev->vid]].dmas[queue_id].dev_id; + pkts_inflight = rte_vhost_async_get_inflight_thread_unsafe(vdev->vid, queue_id); + + struct rte_mbuf *m_cpl[pkts_inflight]; + + while (pkts_inflight) { + n_pkt = rte_vhost_clear_queue_thread_unsafe(vdev->vid, queue_id, m_cpl, + pkts_inflight, dma_id, 0); + free_pkts(m_cpl, n_pkt); + pkts_inflight = rte_vhost_async_get_inflight_thread_unsafe(vdev->vid, + queue_id); + } +} + /* * Remove a device from the specific data core linked list and from the - * main linked list. Synchonization occurs through the use of the + * main linked list. Synchronization occurs through the use of the * lcore dev_removal_flag. Device is made volatile here to avoid re-ordering * of dev->remove=1 which can cause an infinite loop in the rte_pause loop. */ @@ -1185,6 +1554,7 @@ destroy_device(int vid) { struct vhost_dev *vdev = NULL; int lcore; + uint16_t i; TAILQ_FOREACH(vdev, &vhost_dev_list, global_vdev_entry) { if (vdev->vid == vid) @@ -1198,13 +1568,19 @@ destroy_device(int vid) rte_pause(); } + for (i = 0; i < RTE_MAX_LCORE; i++) + rte_free(vhost_txbuff[i * RTE_MAX_VHOST_DEVICE + vid]); + + if (builtin_net_driver) + vs_vhost_net_remove(vdev); + TAILQ_REMOVE(&lcore_info[vdev->coreid].vdev_list, vdev, lcore_vdev_entry); TAILQ_REMOVE(&vhost_dev_list, vdev, global_vdev_entry); /* Set the dev_removal_flag on each lcore. */ - RTE_LCORE_FOREACH_SLAVE(lcore) + RTE_LCORE_FOREACH_WORKER(lcore) lcore_info[lcore].dev_removal_flag = REQUEST_DEV_REMOVAL; /* @@ -1212,7 +1588,7 @@ destroy_device(int vid) * we can be sure that they can no longer access the device removed * from the linked lists and that the devices are no longer in use. */ - RTE_LCORE_FOREACH_SLAVE(lcore) { + RTE_LCORE_FOREACH_WORKER(lcore) { while (lcore_info[lcore].dev_removal_flag != ACK_DEV_REMOVAL) rte_pause(); } @@ -1223,19 +1599,92 @@ destroy_device(int vid) "(%d) device has been removed from data core\n", vdev->vid); + if (dma_bind[vid].dmas[VIRTIO_RXQ].async_enabled) { + vhost_clear_queue_thread_unsafe(vdev, VIRTIO_RXQ); + rte_vhost_async_channel_unregister(vid, VIRTIO_RXQ); + dma_bind[vid].dmas[VIRTIO_RXQ].async_enabled = false; + } + + if (dma_bind[vid].dmas[VIRTIO_TXQ].async_enabled) { + vhost_clear_queue_thread_unsafe(vdev, VIRTIO_TXQ); + rte_vhost_async_channel_unregister(vid, VIRTIO_TXQ); + dma_bind[vid].dmas[VIRTIO_TXQ].async_enabled = false; + } + rte_free(vdev); } +static inline int +get_socketid_by_vid(int vid) +{ + int i; + char ifname[PATH_MAX]; + rte_vhost_get_ifname(vid, ifname, sizeof(ifname)); + + for (i = 0; i < nb_sockets; i++) { + char *file = socket_files + i * PATH_MAX; + if (strcmp(file, ifname) == 0) + return i; + } + + return -1; +} + +static int +init_vhost_queue_ops(int vid) +{ + if (builtin_net_driver) { + vdev_queue_ops[vid].enqueue_pkt_burst = builtin_enqueue_pkts; + vdev_queue_ops[vid].dequeue_pkt_burst = builtin_dequeue_pkts; + } else { + if (dma_bind[vid2socketid[vid]].dmas[VIRTIO_RXQ].async_enabled) + vdev_queue_ops[vid].enqueue_pkt_burst = async_enqueue_pkts; + else + vdev_queue_ops[vid].enqueue_pkt_burst = sync_enqueue_pkts; + + if (dma_bind[vid2socketid[vid]].dmas[VIRTIO_TXQ].async_enabled) + vdev_queue_ops[vid].dequeue_pkt_burst = async_dequeue_pkts; + else + vdev_queue_ops[vid].dequeue_pkt_burst = sync_dequeue_pkts; + } + + return 0; +} + +static inline int +vhost_async_channel_register(int vid) +{ + int rx_ret = 0, tx_ret = 0; + + if (dma_bind[vid2socketid[vid]].dmas[VIRTIO_RXQ].dev_id != INVALID_DMA_ID) { + rx_ret = rte_vhost_async_channel_register(vid, VIRTIO_RXQ); + if (rx_ret == 0) + dma_bind[vid2socketid[vid]].dmas[VIRTIO_RXQ].async_enabled = true; + } + + if (dma_bind[vid2socketid[vid]].dmas[VIRTIO_TXQ].dev_id != INVALID_DMA_ID) { + tx_ret = rte_vhost_async_channel_register(vid, VIRTIO_TXQ); + if (tx_ret == 0) + dma_bind[vid2socketid[vid]].dmas[VIRTIO_TXQ].async_enabled = true; + } + + return rx_ret | tx_ret; +} + + + /* * A new device is added to a data core. First the device is added to the main linked list - * and the allocated to a specific data core. + * and then allocated to a specific data core. */ static int new_device(int vid) { int lcore, core_add = 0; + uint16_t i; uint32_t device_num_min = num_devices; struct vhost_dev *vdev; + int ret; vdev = rte_zmalloc("vhost device", sizeof(*vdev), RTE_CACHE_LINE_SIZE); if (vdev == NULL) { @@ -1246,6 +1695,33 @@ new_device(int vid) } vdev->vid = vid; + for (i = 0; i < RTE_MAX_LCORE; i++) { + vhost_txbuff[i * RTE_MAX_VHOST_DEVICE + vid] + = rte_zmalloc("vhost bufftable", + sizeof(struct vhost_bufftable), + RTE_CACHE_LINE_SIZE); + + if (vhost_txbuff[i * RTE_MAX_VHOST_DEVICE + vid] == NULL) { + RTE_LOG(INFO, VHOST_DATA, + "(%d) couldn't allocate memory for vhost TX\n", vid); + return -1; + } + } + + int socketid = get_socketid_by_vid(vid); + if (socketid == -1) + return -1; + + init_vid2socketid_array(vid, socketid); + + ret = vhost_async_channel_register(vid); + + if (init_vhost_queue_ops(vid) != 0) + return -1; + + if (builtin_net_driver) + vs_vhost_net_setup(vdev); + TAILQ_INSERT_TAIL(&vhost_dev_list, vdev, global_vdev_entry); vdev->vmdq_rx_q = vid * queues_per_pool + vmdq_queue_base; @@ -1254,7 +1730,7 @@ new_device(int vid) vdev->remove = 0; /* Find a suitable lcore to add the device. */ - RTE_LCORE_FOREACH_SLAVE(lcore) { + RTE_LCORE_FOREACH_WORKER(lcore) { if (lcore_info[lcore].device_num < device_num_min) { device_num_min = lcore_info[lcore].device_num; core_add = lcore; @@ -1274,6 +1750,29 @@ new_device(int vid) "(%d) device has been added to data core %d\n", vid, vdev->coreid); + return ret; +} + +static int +vring_state_changed(int vid, uint16_t queue_id, int enable) +{ + struct vhost_dev *vdev = NULL; + + TAILQ_FOREACH(vdev, &vhost_dev_list, global_vdev_entry) { + if (vdev->vid == vid) + break; + } + if (!vdev) + return -1; + + if (queue_id != VIRTIO_RXQ) + return 0; + + if (dma_bind[vid2socketid[vid]].dmas[queue_id].async_enabled) { + if (!enable) + vhost_clear_queue_thread_unsafe(vdev, queue_id); + } + return 0; } @@ -1281,18 +1780,19 @@ new_device(int vid) * These callback allow devices to be added to the data core when configuration * has been fully complete. */ -static const struct virtio_net_device_ops virtio_net_device_ops = +static const struct rte_vhost_device_ops virtio_net_device_ops = { .new_device = new_device, .destroy_device = destroy_device, + .vring_state_changed = vring_state_changed, }; /* * This is a thread will wake up after a period to print stats if the user has * enabled them. */ -static void -print_stats(void) +static void * +print_stats(__rte_unused void *arg) { struct vhost_dev *vdev; uint64_t tx_dropped, rx_dropped; @@ -1312,8 +1812,10 @@ print_stats(void) tx = vdev->stats.tx; tx_dropped = tx_total - tx; - rx_total = rte_atomic64_read(&vdev->stats.rx_total_atomic); - rx = rte_atomic64_read(&vdev->stats.rx_atomic); + rx_total = __atomic_load_n(&vdev->stats.rx_total_atomic, + __ATOMIC_SEQ_CST); + rx = __atomic_load_n(&vdev->stats.rx_atomic, + __ATOMIC_SEQ_CST); rx_dropped = rx_total - rx; printf("Statistics for device %d\n" @@ -1330,7 +1832,11 @@ print_stats(void) } printf("===================================================\n"); + + fflush(stdout); } + + return NULL; } static void @@ -1357,60 +1863,26 @@ sigint_handler(__rte_unused int signum) exit(0); } -/* - * While creating an mbuf pool, one key thing is to figure out how - * many mbuf entries is enough for our use. FYI, here are some - * guidelines: - * - * - Each rx queue would reserve @nr_rx_desc mbufs at queue setup stage - * - * - For each switch core (A CPU core does the packet switch), we need - * also make some reservation for receiving the packets from virtio - * Tx queue. How many is enough depends on the usage. It's normally - * a simple calculation like following: - * - * MAX_PKT_BURST * max packet size / mbuf size - * - * So, we definitely need allocate more mbufs when TSO is enabled. - * - * - Similarly, for each switching core, we should serve @nr_rx_desc - * mbufs for receiving the packets from physical NIC device. - * - * - We also need make sure, for each switch core, we have allocated - * enough mbufs to fill up the mbuf cache. - */ static void -create_mbuf_pool(uint16_t nr_port, uint32_t nr_switch_core, uint32_t mbuf_size, - uint32_t nr_queues, uint32_t nr_rx_desc, uint32_t nr_mbuf_cache) +reset_dma(void) { - uint32_t nr_mbufs; - uint32_t nr_mbufs_per_core; - uint32_t mtu = 1500; - - if (mergeable) - mtu = 9000; - if (enable_tso) - mtu = 64 * 1024; - - nr_mbufs_per_core = (mtu + mbuf_size) * MAX_PKT_BURST / - (mbuf_size - RTE_PKTMBUF_HEADROOM) * MAX_PKT_BURST; - nr_mbufs_per_core += nr_rx_desc; - nr_mbufs_per_core = RTE_MAX(nr_mbufs_per_core, nr_mbuf_cache); - - nr_mbufs = nr_queues * nr_rx_desc; - nr_mbufs += nr_mbufs_per_core * nr_switch_core; - nr_mbufs *= nr_port; - - mbuf_pool = rte_pktmbuf_pool_create("MBUF_POOL", nr_mbufs, - nr_mbuf_cache, 0, mbuf_size, - rte_socket_id()); - if (mbuf_pool == NULL) - rte_exit(EXIT_FAILURE, "Cannot create mbuf pool\n"); + int i; + + for (i = 0; i < RTE_MAX_VHOST_DEVICE; i++) { + int j; + + for (j = 0; j < RTE_MAX_QUEUES_PER_PORT * 2; j++) { + dma_bind[i].dmas[j].dev_id = INVALID_DMA_ID; + dma_bind[i].dmas[j].async_enabled = false; + } + } + + for (i = 0; i < RTE_DMADEV_DEFAULT_MAX; i++) + dmas_id[i] = INVALID_DMA_ID; } /* - * Main function, does initialisation and calls the per-lcore functions. The CUSE - * device is also registered here to handle the IOCTLs. + * Main function, does initialisation and calls the per-lcore functions. */ int main(int argc, char *argv[]) @@ -1418,10 +1890,9 @@ main(int argc, char *argv[]) unsigned lcore_id, core_id = 0; unsigned nb_ports, valid_num_ports; int ret, i; - uint8_t portid; + uint16_t portid; static pthread_t tid; - char thread_name[RTE_MAX_THREAD_NAME_LEN]; - uint64_t flags = 0; + uint64_t flags = RTE_VHOST_USER_NET_COMPLIANT_OL_FLAGS; signal(SIGINT, sigint_handler); @@ -1432,22 +1903,26 @@ main(int argc, char *argv[]) argc -= ret; argv += ret; + /* initialize dma structures */ + reset_dma(); + /* parse app arguments */ ret = us_vhost_parse_args(argc, argv); if (ret < 0) rte_exit(EXIT_FAILURE, "Invalid argument\n"); - for (lcore_id = 0; lcore_id < RTE_MAX_LCORE; lcore_id ++) + for (lcore_id = 0; lcore_id < RTE_MAX_LCORE; lcore_id++) { TAILQ_INIT(&lcore_info[lcore_id].vdev_list); if (rte_lcore_is_enabled(lcore_id)) - lcore_ids[core_id ++] = lcore_id; + lcore_ids[core_id++] = lcore_id; + } if (rte_lcore_count() > RTE_MAX_LCORE) rte_exit(EXIT_FAILURE,"Not enough cores\n"); /* Get the number of physical ports. */ - nb_ports = rte_eth_dev_count(); + nb_ports = rte_eth_dev_count_avail(); /* * Update the global var NUM_PORTS and global array PORTS @@ -1467,8 +1942,11 @@ main(int argc, char *argv[]) * many queues here. We probably should only do allocation for * those queues we are going to use. */ - create_mbuf_pool(valid_num_ports, rte_lcore_count() - 1, MBUF_DATA_SIZE, - MAX_QUEUES, RTE_TEST_RX_DESC_DEFAULT, MBUF_CACHE_SIZE); + mbuf_pool = rte_pktmbuf_pool_create("MBUF_POOL", total_num_mbufs, + MBUF_CACHE_SIZE, 0, MBUF_DATA_SIZE, + rte_socket_id()); + if (mbuf_pool == NULL) + rte_exit(EXIT_FAILURE, "Cannot create mbuf pool\n"); if (vm2vm_mode == VM2VM_HARDWARE) { /* Enable VT loop back to let L2 switch to do it. */ @@ -1478,7 +1956,7 @@ main(int argc, char *argv[]) } /* initialize all ports */ - for (portid = 0; portid < nb_ports; portid++) { + RTE_ETH_FOREACH_DEV(portid) { /* skip ports that are not enabled */ if ((enabled_port_mask & (1 << portid)) == 0) { RTE_LOG(INFO, VHOST_PORT, @@ -1492,47 +1970,88 @@ main(int argc, char *argv[]) /* Enable stats if the user option is set. */ if (enable_stats) { - ret = pthread_create(&tid, NULL, (void *)print_stats, NULL); - if (ret != 0) + ret = rte_ctrl_thread_create(&tid, "print-stats", NULL, + print_stats, NULL); + if (ret < 0) rte_exit(EXIT_FAILURE, "Cannot create print-stats thread\n"); - - /* Set thread_name for aid in debugging. */ - snprintf(thread_name, RTE_MAX_THREAD_NAME_LEN, "print-stats"); - ret = rte_thread_setname(tid, thread_name); - if (ret != 0) - RTE_LOG(DEBUG, VHOST_CONFIG, - "Cannot set print-stats name\n"); } /* Launch all data cores. */ - RTE_LCORE_FOREACH_SLAVE(lcore_id) + RTE_LCORE_FOREACH_WORKER(lcore_id) rte_eal_remote_launch(switch_worker, NULL, lcore_id); - if (mergeable == 0) - rte_vhost_feature_disable(1ULL << VIRTIO_NET_F_MRG_RXBUF); - if (client_mode) flags |= RTE_VHOST_USER_CLIENT; - if (dequeue_zero_copy) - flags |= RTE_VHOST_USER_DEQUEUE_ZERO_COPY; + for (i = 0; i < dma_count; i++) { + if (rte_vhost_async_dma_configure(dmas_id[i], 0) < 0) { + RTE_LOG(ERR, VHOST_PORT, "Failed to configure DMA in vhost.\n"); + rte_exit(EXIT_FAILURE, "Cannot use given DMA device\n"); + } + } /* Register vhost user driver to handle vhost messages. */ for (i = 0; i < nb_sockets; i++) { - ret = rte_vhost_driver_register - (socket_files + i * PATH_MAX, flags); + char *file = socket_files + i * PATH_MAX; + + if (dma_count && get_async_flag_by_socketid(i) != 0) + flags = flags | RTE_VHOST_USER_ASYNC_COPY; + + ret = rte_vhost_driver_register(file, flags); if (ret != 0) { unregister_drivers(i); rte_exit(EXIT_FAILURE, "vhost driver register failure.\n"); } + + if (builtin_net_driver) + rte_vhost_driver_set_features(file, VIRTIO_NET_FEATURES); + + if (mergeable == 0) { + rte_vhost_driver_disable_features(file, + 1ULL << VIRTIO_NET_F_MRG_RXBUF); + } + + if (enable_tx_csum == 0) { + rte_vhost_driver_disable_features(file, + 1ULL << VIRTIO_NET_F_CSUM); + } + + if (enable_tso == 0) { + rte_vhost_driver_disable_features(file, + 1ULL << VIRTIO_NET_F_HOST_TSO4); + rte_vhost_driver_disable_features(file, + 1ULL << VIRTIO_NET_F_HOST_TSO6); + rte_vhost_driver_disable_features(file, + 1ULL << VIRTIO_NET_F_GUEST_TSO4); + rte_vhost_driver_disable_features(file, + 1ULL << VIRTIO_NET_F_GUEST_TSO6); + } + + if (promiscuous) { + rte_vhost_driver_enable_features(file, + 1ULL << VIRTIO_NET_F_CTRL_RX); + } + + ret = rte_vhost_driver_callback_register(file, + &virtio_net_device_ops); + if (ret != 0) { + rte_exit(EXIT_FAILURE, + "failed to register vhost driver callbacks.\n"); + } + + if (rte_vhost_driver_start(file) < 0) { + rte_exit(EXIT_FAILURE, + "failed to start vhost driver.\n"); + } } - rte_vhost_driver_callback_register(&virtio_net_device_ops); + RTE_LCORE_FOREACH_WORKER(lcore_id) + rte_eal_wait_lcore(lcore_id); - /* Start CUSE session. */ - rte_vhost_driver_session_start(); - return 0; + /* clean up the EAL */ + rte_eal_cleanup(); + return 0; }