diff options
Diffstat (limited to 'drivers/net/virtio')
| -rw-r--r-- | drivers/net/virtio/Makefile | 7 | ||||
| -rw-r--r-- | drivers/net/virtio/virtio_ethdev.c | 477 | ||||
| -rw-r--r-- | drivers/net/virtio/virtio_ethdev.h | 38 | ||||
| -rw-r--r-- | drivers/net/virtio/virtio_pci.c | 5 | ||||
| -rw-r--r-- | drivers/net/virtio/virtio_pci.h | 15 | ||||
| -rw-r--r-- | drivers/net/virtio/virtio_rxtx.c | 564 | ||||
| -rw-r--r-- | drivers/net/virtio/virtio_rxtx.h | 3 | ||||
| -rw-r--r-- | drivers/net/virtio/virtio_rxtx_simple.c | 272 | ||||
| -rw-r--r-- | drivers/net/virtio/virtio_rxtx_simple.h | 136 | ||||
| -rw-r--r-- | drivers/net/virtio/virtio_rxtx_simple_neon.c | 235 | ||||
| -rw-r--r-- | drivers/net/virtio/virtio_rxtx_simple_sse.c | 222 | ||||
| -rw-r--r-- | drivers/net/virtio/virtio_user_ethdev.c | 20 | ||||
| -rw-r--r-- | drivers/net/virtio/virtqueue.h | 9 |
13 files changed, 1267 insertions, 736 deletions
diff --git a/drivers/net/virtio/Makefile b/drivers/net/virtio/Makefile index 3020b688..97972a6c 100644 --- a/drivers/net/virtio/Makefile +++ b/drivers/net/virtio/Makefile @@ -50,9 +50,12 @@ SRCS-$(CONFIG_RTE_LIBRTE_VIRTIO_PMD) += virtqueue.c SRCS-$(CONFIG_RTE_LIBRTE_VIRTIO_PMD) += virtio_pci.c SRCS-$(CONFIG_RTE_LIBRTE_VIRTIO_PMD) += virtio_rxtx.c SRCS-$(CONFIG_RTE_LIBRTE_VIRTIO_PMD) += virtio_ethdev.c - -ifeq ($(findstring RTE_MACHINE_CPUFLAG_SSSE3,$(CFLAGS)),RTE_MACHINE_CPUFLAG_SSSE3) SRCS-$(CONFIG_RTE_LIBRTE_VIRTIO_PMD) += virtio_rxtx_simple.c + +ifeq ($(CONFIG_RTE_ARCH_X86),y) +SRCS-$(CONFIG_RTE_LIBRTE_VIRTIO_PMD) += virtio_rxtx_simple_sse.c +else ifneq ($(filter y,$(CONFIG_RTE_ARCH_ARM) $(CONFIG_RTE_ARCH_ARM64)),) +SRCS-$(CONFIG_RTE_LIBRTE_VIRTIO_PMD) += virtio_rxtx_simple_neon.c endif ifeq ($(CONFIG_RTE_VIRTIO_USER),y) diff --git a/drivers/net/virtio/virtio_ethdev.c b/drivers/net/virtio/virtio_ethdev.c index 86cf8a38..079fd6c8 100644 --- a/drivers/net/virtio/virtio_ethdev.c +++ b/drivers/net/virtio/virtio_ethdev.c @@ -103,7 +103,8 @@ static int virtio_dev_queue_stats_mapping_set( * The set of PCI devices this driver supports */ static const struct rte_pci_id pci_id_virtio_map[] = { - { RTE_PCI_DEVICE(VIRTIO_PCI_VENDORID, VIRTIO_PCI_DEVICEID_MIN) }, + { RTE_PCI_DEVICE(VIRTIO_PCI_VENDORID, VIRTIO_PCI_LEGACY_DEVICEID_NET) }, + { RTE_PCI_DEVICE(VIRTIO_PCI_VENDORID, VIRTIO_PCI_MODERN_DEVICEID_NET) }, { .vendor_id = 0, /* sentinel */ }, }; @@ -279,28 +280,65 @@ virtio_set_multiple_queues(struct rte_eth_dev *dev, uint16_t nb_queues) return 0; } -void -virtio_dev_queue_release(struct virtqueue *vq) +static void +virtio_dev_queue_release(void *queue __rte_unused) { - struct virtio_hw *hw; + /* do nothing */ +} - if (vq) { - hw = vq->hw; - if (vq->configured) - hw->vtpci_ops->del_queue(hw, vq); +static int +virtio_get_queue_type(struct virtio_hw *hw, uint16_t vtpci_queue_idx) +{ + if (vtpci_queue_idx == hw->max_queue_pairs * 2) + return VTNET_CQ; + else if (vtpci_queue_idx % 2 == 0) + return VTNET_RQ; + else + return VTNET_TQ; +} - rte_free(vq->sw_ring); - rte_free(vq); - } +static uint16_t +virtio_get_nr_vq(struct virtio_hw *hw) +{ + uint16_t nr_vq = hw->max_queue_pairs * 2; + + if (vtpci_with_feature(hw, VIRTIO_NET_F_CTRL_VQ)) + nr_vq += 1; + + return nr_vq; } -int virtio_dev_queue_setup(struct rte_eth_dev *dev, - int queue_type, - uint16_t queue_idx, - uint16_t vtpci_queue_idx, - uint16_t nb_desc, - unsigned int socket_id, - void **pvq) +static void +virtio_init_vring(struct virtqueue *vq) +{ + int size = vq->vq_nentries; + struct vring *vr = &vq->vq_ring; + uint8_t *ring_mem = vq->vq_ring_virt_mem; + + PMD_INIT_FUNC_TRACE(); + + /* + * Reinitialise since virtio port might have been stopped and restarted + */ + memset(ring_mem, 0, vq->vq_ring_size); + vring_init(vr, size, ring_mem, VIRTIO_PCI_VRING_ALIGN); + vq->vq_used_cons_idx = 0; + vq->vq_desc_head_idx = 0; + vq->vq_avail_idx = 0; + vq->vq_desc_tail_idx = (uint16_t)(vq->vq_nentries - 1); + vq->vq_free_cnt = vq->vq_nentries; + memset(vq->vq_descx, 0, sizeof(struct vq_desc_extra) * vq->vq_nentries); + + vring_desc_init(vr->desc, size); + + /* + * Disable device(host) interrupting guest + */ + virtqueue_disable_intr(vq); +} + +static int +virtio_init_queue(struct rte_eth_dev *dev, uint16_t vtpci_queue_idx) { char vq_name[VIRTQUEUE_MAX_NAME_SZ]; char vq_hdr_name[VIRTQUEUE_MAX_NAME_SZ]; @@ -311,9 +349,9 @@ int virtio_dev_queue_setup(struct rte_eth_dev *dev, struct virtnet_tx *txvq = NULL; struct virtnet_ctl *cvq = NULL; struct virtqueue *vq; - const char *queue_names[] = {"rvq", "txq", "cvq"}; - size_t sz_vq, sz_q = 0, sz_hdr_mz = 0; + size_t sz_hdr_mz = 0; void *sw_ring = NULL; + int queue_type = virtio_get_queue_type(hw, vtpci_queue_idx); int ret; PMD_INIT_LOG(DEBUG, "setting up queue: %u", vtpci_queue_idx); @@ -323,7 +361,7 @@ int virtio_dev_queue_setup(struct rte_eth_dev *dev, * Always power of 2 and if 0 virtqueue does not exist */ vq_size = hw->vtpci_ops->get_queue_num(hw, vtpci_queue_idx); - PMD_INIT_LOG(DEBUG, "vq_size: %u nb_desc:%u", vq_size, nb_desc); + PMD_INIT_LOG(DEBUG, "vq_size: %u", vq_size); if (vq_size == 0) { PMD_INIT_LOG(ERR, "virtqueue does not exist"); return -EINVAL; @@ -334,40 +372,35 @@ int virtio_dev_queue_setup(struct rte_eth_dev *dev, return -EINVAL; } - snprintf(vq_name, sizeof(vq_name), "port%d_%s%d", - dev->data->port_id, queue_names[queue_type], queue_idx); + snprintf(vq_name, sizeof(vq_name), "port%d_vq%d", + dev->data->port_id, vtpci_queue_idx); - sz_vq = RTE_ALIGN_CEIL(sizeof(*vq) + + size = RTE_ALIGN_CEIL(sizeof(*vq) + vq_size * sizeof(struct vq_desc_extra), RTE_CACHE_LINE_SIZE); - if (queue_type == VTNET_RQ) { - sz_q = sz_vq + sizeof(*rxvq); - } else if (queue_type == VTNET_TQ) { - sz_q = sz_vq + sizeof(*txvq); + if (queue_type == VTNET_TQ) { /* * For each xmit packet, allocate a virtio_net_hdr * and indirect ring elements */ sz_hdr_mz = vq_size * sizeof(struct virtio_tx_region); } else if (queue_type == VTNET_CQ) { - sz_q = sz_vq + sizeof(*cvq); /* Allocate a page for control vq command, data and status */ sz_hdr_mz = PAGE_SIZE; } - vq = rte_zmalloc_socket(vq_name, sz_q, RTE_CACHE_LINE_SIZE, socket_id); + vq = rte_zmalloc_socket(vq_name, size, RTE_CACHE_LINE_SIZE, + SOCKET_ID_ANY); if (vq == NULL) { PMD_INIT_LOG(ERR, "can not allocate vq"); return -ENOMEM; } + hw->vqs[vtpci_queue_idx] = vq; + vq->hw = hw; vq->vq_queue_index = vtpci_queue_idx; vq->vq_nentries = vq_size; - if (nb_desc == 0 || nb_desc > vq_size) - nb_desc = vq_size; - vq->vq_free_cnt = nb_desc; - /* * Reserve a memzone for vring elements */ @@ -376,7 +409,8 @@ int virtio_dev_queue_setup(struct rte_eth_dev *dev, PMD_INIT_LOG(DEBUG, "vring_size: %d, rounded_vring_size: %d", size, vq->vq_ring_size); - mz = rte_memzone_reserve_aligned(vq_name, vq->vq_ring_size, socket_id, + mz = rte_memzone_reserve_aligned(vq_name, vq->vq_ring_size, + SOCKET_ID_ANY, 0, VIRTIO_PCI_VRING_ALIGN); if (mz == NULL) { if (rte_errno == EEXIST) @@ -396,12 +430,13 @@ int virtio_dev_queue_setup(struct rte_eth_dev *dev, PMD_INIT_LOG(DEBUG, "vq->vq_ring_virt_mem: 0x%" PRIx64, (uint64_t)(uintptr_t)mz->addr); + virtio_init_vring(vq); + if (sz_hdr_mz) { - snprintf(vq_hdr_name, sizeof(vq_hdr_name), "port%d_%s%d_hdr", - dev->data->port_id, queue_names[queue_type], - queue_idx); + snprintf(vq_hdr_name, sizeof(vq_hdr_name), "port%d_vq%d_hdr", + dev->data->port_id, vtpci_queue_idx); hdr_mz = rte_memzone_reserve_aligned(vq_hdr_name, sz_hdr_mz, - socket_id, 0, + SOCKET_ID_ANY, 0, RTE_CACHE_LINE_SIZE); if (hdr_mz == NULL) { if (rte_errno == EEXIST) @@ -418,7 +453,7 @@ int virtio_dev_queue_setup(struct rte_eth_dev *dev, sizeof(vq->sw_ring[0]); sw_ring = rte_zmalloc_socket("sw_ring", sz_sw, - RTE_CACHE_LINE_SIZE, socket_id); + RTE_CACHE_LINE_SIZE, SOCKET_ID_ANY); if (!sw_ring) { PMD_INIT_LOG(ERR, "can not allocate RX soft ring"); ret = -ENOMEM; @@ -426,30 +461,26 @@ int virtio_dev_queue_setup(struct rte_eth_dev *dev, } vq->sw_ring = sw_ring; - rxvq = (struct virtnet_rx *)RTE_PTR_ADD(vq, sz_vq); + rxvq = &vq->rxq; rxvq->vq = vq; rxvq->port_id = dev->data->port_id; - rxvq->queue_id = queue_idx; rxvq->mz = mz; - *pvq = rxvq; } else if (queue_type == VTNET_TQ) { - txvq = (struct virtnet_tx *)RTE_PTR_ADD(vq, sz_vq); + txvq = &vq->txq; txvq->vq = vq; txvq->port_id = dev->data->port_id; - txvq->queue_id = queue_idx; txvq->mz = mz; txvq->virtio_net_hdr_mz = hdr_mz; txvq->virtio_net_hdr_mem = hdr_mz->phys_addr; - - *pvq = txvq; } else if (queue_type == VTNET_CQ) { - cvq = (struct virtnet_ctl *)RTE_PTR_ADD(vq, sz_vq); + cvq = &vq->cq; cvq->vq = vq; cvq->mz = mz; cvq->virtio_net_hdr_mz = hdr_mz; cvq->virtio_net_hdr_mem = hdr_mz->phys_addr; memset(cvq->virtio_net_hdr_mz->addr, 0, PAGE_SIZE); - *pvq = cvq; + + hw->cvq = cvq; } /* For virtio_user case (that is when dev->pci_dev is NULL), we use @@ -490,11 +521,9 @@ int virtio_dev_queue_setup(struct rte_eth_dev *dev, if (hw->vtpci_ops->setup_queue(hw, vq) < 0) { PMD_INIT_LOG(ERR, "setup_queue failed"); - virtio_dev_queue_release(vq); return -EINVAL; } - vq->configured = 1; return 0; fail_q_alloc: @@ -506,40 +535,60 @@ fail_q_alloc: return ret; } -static int -virtio_dev_cq_queue_setup(struct rte_eth_dev *dev, uint16_t vtpci_queue_idx, - uint32_t socket_id) +static void +virtio_free_queues(struct virtio_hw *hw) { - struct virtnet_ctl *cvq; - int ret; - struct virtio_hw *hw = dev->data->dev_private; + uint16_t nr_vq = virtio_get_nr_vq(hw); + struct virtqueue *vq; + int queue_type; + uint16_t i; - PMD_INIT_FUNC_TRACE(); - ret = virtio_dev_queue_setup(dev, VTNET_CQ, VTNET_SQ_CQ_QUEUE_IDX, - vtpci_queue_idx, 0, socket_id, (void **)&cvq); - if (ret < 0) { - PMD_INIT_LOG(ERR, "control vq initialization failed"); - return ret; + for (i = 0; i < nr_vq; i++) { + vq = hw->vqs[i]; + if (!vq) + continue; + + queue_type = virtio_get_queue_type(hw, i); + if (queue_type == VTNET_RQ) { + rte_free(vq->sw_ring); + rte_memzone_free(vq->rxq.mz); + } else if (queue_type == VTNET_TQ) { + rte_memzone_free(vq->txq.mz); + rte_memzone_free(vq->txq.virtio_net_hdr_mz); + } else { + rte_memzone_free(vq->cq.mz); + rte_memzone_free(vq->cq.virtio_net_hdr_mz); + } + + rte_free(vq); } - hw->cvq = cvq; - return 0; + rte_free(hw->vqs); } -static void -virtio_free_queues(struct rte_eth_dev *dev) +static int +virtio_alloc_queues(struct rte_eth_dev *dev) { - unsigned int i; - - for (i = 0; i < dev->data->nb_rx_queues; i++) - virtio_dev_rx_queue_release(dev->data->rx_queues[i]); + struct virtio_hw *hw = dev->data->dev_private; + uint16_t nr_vq = virtio_get_nr_vq(hw); + uint16_t i; + int ret; - dev->data->nb_rx_queues = 0; + hw->vqs = rte_zmalloc(NULL, sizeof(struct virtqueue *) * nr_vq, 0); + if (!hw->vqs) { + PMD_INIT_LOG(ERR, "failed to allocate vqs"); + return -ENOMEM; + } - for (i = 0; i < dev->data->nb_tx_queues; i++) - virtio_dev_tx_queue_release(dev->data->tx_queues[i]); + for (i = 0; i < nr_vq; i++) { + ret = virtio_init_queue(dev, i); + if (ret < 0) { + virtio_free_queues(hw); + return ret; + } + } - dev->data->nb_tx_queues = 0; + return 0; } static void @@ -553,9 +602,8 @@ virtio_dev_close(struct rte_eth_dev *dev) if (dev->data->dev_flags & RTE_ETH_DEV_INTR_LSC) vtpci_irq_config(hw, VIRTIO_MSI_NO_VECTOR); vtpci_reset(hw); - hw->started = 0; virtio_dev_free_mbufs(dev); - virtio_free_queues(dev); + virtio_free_queues(hw); } static void @@ -650,6 +698,23 @@ virtio_dev_allmulticast_disable(struct rte_eth_dev *dev) PMD_INIT_LOG(ERR, "Failed to disable allmulticast"); } +#define VLAN_TAG_LEN 4 /* 802.3ac tag (not DMA'd) */ +static int +virtio_mtu_set(struct rte_eth_dev *dev, uint16_t mtu) +{ + struct virtio_hw *hw = dev->data->dev_private; + uint32_t ether_hdr_len = ETHER_HDR_LEN + VLAN_TAG_LEN + + hw->vtnet_hdr_size; + uint32_t frame_size = mtu + ether_hdr_len; + + if (mtu < ETHER_MIN_MTU || frame_size > VIRTIO_MAX_RX_PKTLEN) { + PMD_INIT_LOG(ERR, "MTU should be between %d and %d\n", + ETHER_MIN_MTU, VIRTIO_MAX_RX_PKTLEN - ether_hdr_len); + return -EINVAL; + } + return 0; +} + /* * dev_ops for virtio, bare necessities for basic operation */ @@ -662,7 +727,7 @@ static const struct eth_dev_ops virtio_eth_dev_ops = { .promiscuous_disable = virtio_dev_promiscuous_disable, .allmulticast_enable = virtio_dev_allmulticast_enable, .allmulticast_disable = virtio_dev_allmulticast_disable, - + .mtu_set = virtio_mtu_set, .dev_infos_get = virtio_dev_info_get, .stats_get = virtio_dev_stats_get, .xstats_get = virtio_dev_xstats_get, @@ -671,9 +736,9 @@ static const struct eth_dev_ops virtio_eth_dev_ops = { .xstats_reset = virtio_dev_stats_reset, .link_update = virtio_dev_link_update, .rx_queue_setup = virtio_dev_rx_queue_setup, - .rx_queue_release = virtio_dev_rx_queue_release, + .rx_queue_release = virtio_dev_queue_release, .tx_queue_setup = virtio_dev_tx_queue_setup, - .tx_queue_release = virtio_dev_tx_queue_release, + .tx_queue_release = virtio_dev_queue_release, /* collect stats per queue */ .queue_stats_mapping_set = virtio_dev_queue_stats_mapping_set, .vlan_filter_set = virtio_vlan_filter_set, @@ -1040,14 +1105,13 @@ virtio_vlan_filter_set(struct rte_eth_dev *dev, uint16_t vlan_id, int on) } static int -virtio_negotiate_features(struct virtio_hw *hw) +virtio_negotiate_features(struct virtio_hw *hw, uint64_t req_features) { uint64_t host_features; /* Prepare guest_features: feature that driver wants to support */ - hw->guest_features = VIRTIO_PMD_GUEST_FEATURES; PMD_INIT_LOG(DEBUG, "guest_features before negotiate = %" PRIx64, - hw->guest_features); + req_features); /* Read device(host) feature bits */ host_features = hw->vtpci_ops->get_features(hw); @@ -1058,6 +1122,7 @@ virtio_negotiate_features(struct virtio_hw *hw) * Negotiate features: Subset of device feature bits are written back * guest feature bits. */ + hw->guest_features = req_features; hw->guest_features = vtpci_negotiate_features(hw, host_features); PMD_INIT_LOG(DEBUG, "features after negotiate = %" PRIx64, hw->guest_features); @@ -1076,6 +1141,8 @@ virtio_negotiate_features(struct virtio_hw *hw) } } + hw->req_guest_features = req_features; + return 0; } @@ -1101,7 +1168,7 @@ virtio_interrupt_handler(__rte_unused struct rte_intr_handle *handle, if (isr & VIRTIO_PCI_ISR_CONFIG) { if (virtio_dev_link_update(dev, 0) == 0) _rte_eth_dev_callback_process(dev, - RTE_ETH_EVENT_INTR_LSC); + RTE_ETH_EVENT_INTR_LSC, NULL); } } @@ -1116,47 +1183,16 @@ rx_func_get(struct rte_eth_dev *eth_dev) eth_dev->rx_pkt_burst = &virtio_recv_pkts; } -/* - * This function is based on probe() function in virtio_pci.c - * It returns 0 on success. - */ -int -eth_virtio_dev_init(struct rte_eth_dev *eth_dev) +/* reset device and renegotiate features if needed */ +static int +virtio_init_device(struct rte_eth_dev *eth_dev, uint64_t req_features) { struct virtio_hw *hw = eth_dev->data->dev_private; struct virtio_net_config *config; struct virtio_net_config local_config; - struct rte_pci_device *pci_dev; - uint32_t dev_flags = RTE_ETH_DEV_DETACHABLE; + struct rte_pci_device *pci_dev = eth_dev->pci_dev; int ret; - RTE_BUILD_BUG_ON(RTE_PKTMBUF_HEADROOM < sizeof(struct virtio_net_hdr_mrg_rxbuf)); - - eth_dev->dev_ops = &virtio_eth_dev_ops; - eth_dev->tx_pkt_burst = &virtio_xmit_pkts; - - if (rte_eal_process_type() == RTE_PROC_SECONDARY) { - rx_func_get(eth_dev); - return 0; - } - - /* Allocate memory for storing MAC addresses */ - eth_dev->data->mac_addrs = rte_zmalloc("virtio", VIRTIO_MAX_MAC_ADDRS * ETHER_ADDR_LEN, 0); - if (eth_dev->data->mac_addrs == NULL) { - PMD_INIT_LOG(ERR, - "Failed to allocate %d bytes needed to store MAC addresses", - VIRTIO_MAX_MAC_ADDRS * ETHER_ADDR_LEN); - return -ENOMEM; - } - - pci_dev = eth_dev->pci_dev; - - if (pci_dev) { - ret = vtpci_init(pci_dev, hw, &dev_flags); - if (ret) - return ret; - } - /* Reset the device although not necessary at startup */ vtpci_reset(hw); @@ -1165,15 +1201,16 @@ eth_virtio_dev_init(struct rte_eth_dev *eth_dev) /* Tell the host we've known how to drive the device. */ vtpci_set_status(hw, VIRTIO_CONFIG_STATUS_DRIVER); - if (virtio_negotiate_features(hw) < 0) + if (virtio_negotiate_features(hw, req_features) < 0) return -1; /* If host does not support status then disable LSC */ if (!vtpci_with_feature(hw, VIRTIO_NET_F_STATUS)) - dev_flags &= ~RTE_ETH_DEV_INTR_LSC; + eth_dev->data->dev_flags &= ~RTE_ETH_DEV_INTR_LSC; + else + eth_dev->data->dev_flags |= RTE_ETH_DEV_INTR_LSC; rte_eth_copy_pci_info(eth_dev, pci_dev); - eth_dev->data->dev_flags = dev_flags; rx_func_get(eth_dev); @@ -1221,16 +1258,7 @@ eth_virtio_dev_init(struct rte_eth_dev *eth_dev) config->max_virtqueue_pairs = 1; } - hw->max_rx_queues = - (VIRTIO_MAX_RX_QUEUES < config->max_virtqueue_pairs) ? - VIRTIO_MAX_RX_QUEUES : config->max_virtqueue_pairs; - hw->max_tx_queues = - (VIRTIO_MAX_TX_QUEUES < config->max_virtqueue_pairs) ? - VIRTIO_MAX_TX_QUEUES : config->max_virtqueue_pairs; - - virtio_dev_cq_queue_setup(eth_dev, - config->max_virtqueue_pairs * 2, - SOCKET_ID_ANY); + hw->max_queue_pairs = config->max_virtqueue_pairs; PMD_INIT_LOG(DEBUG, "config->max_virtqueue_pairs=%d", config->max_virtqueue_pairs); @@ -1241,23 +1269,73 @@ eth_virtio_dev_init(struct rte_eth_dev *eth_dev) config->mac[2], config->mac[3], config->mac[4], config->mac[5]); } else { - hw->max_rx_queues = 1; - hw->max_tx_queues = 1; + PMD_INIT_LOG(DEBUG, "config->max_virtqueue_pairs=1"); + hw->max_queue_pairs = 1; } - PMD_INIT_LOG(DEBUG, "hw->max_rx_queues=%d hw->max_tx_queues=%d", - hw->max_rx_queues, hw->max_tx_queues); + ret = virtio_alloc_queues(eth_dev); + if (ret < 0) + return ret; + vtpci_reinit_complete(hw); + if (pci_dev) PMD_INIT_LOG(DEBUG, "port %d vendorID=0x%x deviceID=0x%x", eth_dev->data->port_id, pci_dev->id.vendor_id, pci_dev->id.device_id); + return 0; +} + +/* + * This function is based on probe() function in virtio_pci.c + * It returns 0 on success. + */ +int +eth_virtio_dev_init(struct rte_eth_dev *eth_dev) +{ + struct virtio_hw *hw = eth_dev->data->dev_private; + struct rte_pci_device *pci_dev; + uint32_t dev_flags = RTE_ETH_DEV_DETACHABLE; + int ret; + + RTE_BUILD_BUG_ON(RTE_PKTMBUF_HEADROOM < sizeof(struct virtio_net_hdr_mrg_rxbuf)); + + eth_dev->dev_ops = &virtio_eth_dev_ops; + eth_dev->tx_pkt_burst = &virtio_xmit_pkts; + + if (rte_eal_process_type() == RTE_PROC_SECONDARY) { + rx_func_get(eth_dev); + return 0; + } + + /* Allocate memory for storing MAC addresses */ + eth_dev->data->mac_addrs = rte_zmalloc("virtio", VIRTIO_MAX_MAC_ADDRS * ETHER_ADDR_LEN, 0); + if (eth_dev->data->mac_addrs == NULL) { + PMD_INIT_LOG(ERR, + "Failed to allocate %d bytes needed to store MAC addresses", + VIRTIO_MAX_MAC_ADDRS * ETHER_ADDR_LEN); + return -ENOMEM; + } + + pci_dev = eth_dev->pci_dev; + + if (pci_dev) { + ret = vtpci_init(pci_dev, hw, &dev_flags); + if (ret) + return ret; + } + + eth_dev->data->dev_flags = dev_flags; + + /* reset device and negotiate default features */ + ret = virtio_init_device(eth_dev, VIRTIO_PMD_DEFAULT_GUEST_FEATURES); + if (ret < 0) + return ret; + /* Setup interrupt callback */ if (eth_dev->data->dev_flags & RTE_ETH_DEV_INTR_LSC) rte_intr_callback_register(&pci_dev->intr_handle, - virtio_interrupt_handler, eth_dev); - - virtio_dev_cq_start(eth_dev); + virtio_interrupt_handler, eth_dev); return 0; } @@ -1266,26 +1344,20 @@ static int eth_virtio_dev_uninit(struct rte_eth_dev *eth_dev) { struct rte_pci_device *pci_dev; - struct virtio_hw *hw = eth_dev->data->dev_private; PMD_INIT_FUNC_TRACE(); if (rte_eal_process_type() == RTE_PROC_SECONDARY) return -EPERM; - if (hw->started == 1) { - virtio_dev_stop(eth_dev); - virtio_dev_close(eth_dev); - } + virtio_dev_stop(eth_dev); + virtio_dev_close(eth_dev); pci_dev = eth_dev->pci_dev; eth_dev->dev_ops = NULL; eth_dev->tx_pkt_burst = NULL; eth_dev->rx_pkt_burst = NULL; - if (hw->cvq) - virtio_dev_queue_release(hw->cvq->vq); - rte_free(eth_dev->data->mac_addrs); eth_dev->data->mac_addrs = NULL; @@ -1303,32 +1375,29 @@ eth_virtio_dev_uninit(struct rte_eth_dev *eth_dev) static struct eth_driver rte_virtio_pmd = { .pci_drv = { - .name = "rte_virtio_pmd", + .driver = { + .name = "net_virtio", + }, .id_table = pci_id_virtio_map, .drv_flags = RTE_PCI_DRV_DETACHABLE, + .probe = rte_eth_dev_pci_probe, + .remove = rte_eth_dev_pci_remove, }, .eth_dev_init = eth_virtio_dev_init, .eth_dev_uninit = eth_virtio_dev_uninit, .dev_private_size = sizeof(struct virtio_hw), }; -/* - * Driver initialization routine. - * Invoked once at EAL init time. - * Register itself as the [Poll Mode] Driver of PCI virtio devices. - * Returns 0 on success. - */ -static int -rte_virtio_pmd_init(const char *name __rte_unused, - const char *param __rte_unused) +RTE_INIT(rte_virtio_pmd_init); +static void +rte_virtio_pmd_init(void) { if (rte_eal_iopl_init() != 0) { PMD_INIT_LOG(ERR, "IOPL call failed - cannot use virtio PMD"); - return -1; + return; } - rte_eth_driver_register(&rte_virtio_pmd); - return 0; + rte_eal_pci_register(&rte_virtio_pmd.pci_drv); } /* @@ -1340,14 +1409,44 @@ virtio_dev_configure(struct rte_eth_dev *dev) { const struct rte_eth_rxmode *rxmode = &dev->data->dev_conf.rxmode; struct virtio_hw *hw = dev->data->dev_private; + uint64_t req_features; + int ret; PMD_INIT_LOG(DEBUG, "configure"); + req_features = VIRTIO_PMD_DEFAULT_GUEST_FEATURES; + if (rxmode->hw_ip_checksum) + req_features |= (1ULL << VIRTIO_NET_F_GUEST_CSUM); + if (rxmode->enable_lro) + req_features |= + (1ULL << VIRTIO_NET_F_GUEST_TSO4) | + (1ULL << VIRTIO_NET_F_GUEST_TSO6); + + /* if request features changed, reinit the device */ + if (req_features != hw->req_guest_features) { + ret = virtio_init_device(dev, req_features); + if (ret < 0) + return ret; + } - if (rxmode->hw_ip_checksum) { - PMD_DRV_LOG(ERR, "HW IP checksum not supported"); - return -EINVAL; + if (rxmode->hw_ip_checksum && + !vtpci_with_feature(hw, VIRTIO_NET_F_GUEST_CSUM)) { + PMD_DRV_LOG(NOTICE, + "rx ip checksum not available on this host"); + return -ENOTSUP; + } + + if (rxmode->enable_lro && + (!vtpci_with_feature(hw, VIRTIO_NET_F_GUEST_TSO4) || + !vtpci_with_feature(hw, VIRTIO_NET_F_GUEST_TSO4))) { + PMD_DRV_LOG(NOTICE, + "lro not available on this host"); + return -ENOTSUP; } + /* start control queue */ + if (vtpci_with_feature(hw, VIRTIO_NET_F_CTRL_VQ)) + virtio_dev_cq_start(dev); + hw->vlan_strip = rxmode->hw_vlan_strip; if (rxmode->hw_vlan_filter @@ -1371,9 +1470,9 @@ static int virtio_dev_start(struct rte_eth_dev *dev) { uint16_t nb_queues, i; - struct virtio_hw *hw = dev->data->dev_private; struct virtnet_rx *rxvq; struct virtnet_tx *txvq __rte_unused; + struct virtio_hw *hw = dev->data->dev_private; /* check if lsc interrupt feature is enabled */ if (dev->data->dev_conf.intr_conf.lsc) { @@ -1391,29 +1490,19 @@ virtio_dev_start(struct rte_eth_dev *dev) /* Initialize Link state */ virtio_dev_link_update(dev, 0); - /* On restart after stop do not touch queues */ - if (hw->started) - return 0; - - /* Do final configuration before rx/tx engine starts */ - virtio_dev_rxtx_start(dev); - vtpci_reinit_complete(hw); - - hw->started = 1; - /*Notify the backend *Otherwise the tap backend might already stop its queue due to fullness. *vhost backend will have no chance to be waked up */ - nb_queues = dev->data->nb_rx_queues; - if (nb_queues > 1) { + nb_queues = RTE_MAX(dev->data->nb_rx_queues, dev->data->nb_tx_queues); + if (hw->max_queue_pairs > 1) { if (virtio_set_multiple_queues(dev, nb_queues) != 0) return -EINVAL; } PMD_INIT_LOG(DEBUG, "nb_queues=%d", nb_queues); - for (i = 0; i < nb_queues; i++) { + for (i = 0; i < dev->data->nb_rx_queues; i++) { rxvq = dev->data->rx_queues[i]; virtqueue_notify(rxvq->vq); } @@ -1532,20 +1621,39 @@ virtio_dev_link_update(struct rte_eth_dev *dev, __rte_unused int wait_to_complet static void virtio_dev_info_get(struct rte_eth_dev *dev, struct rte_eth_dev_info *dev_info) { + uint64_t tso_mask; struct virtio_hw *hw = dev->data->dev_private; if (dev->pci_dev) - dev_info->driver_name = dev->driver->pci_drv.name; + dev_info->driver_name = dev->driver->pci_drv.driver.name; else dev_info->driver_name = "virtio_user PMD"; - dev_info->max_rx_queues = (uint16_t)hw->max_rx_queues; - dev_info->max_tx_queues = (uint16_t)hw->max_tx_queues; + dev_info->max_rx_queues = + RTE_MIN(hw->max_queue_pairs, VIRTIO_MAX_RX_QUEUES); + dev_info->max_tx_queues = + RTE_MIN(hw->max_queue_pairs, VIRTIO_MAX_TX_QUEUES); dev_info->min_rx_bufsize = VIRTIO_MIN_RX_BUFSIZE; dev_info->max_rx_pktlen = VIRTIO_MAX_RX_PKTLEN; dev_info->max_mac_addrs = VIRTIO_MAX_MAC_ADDRS; dev_info->default_txconf = (struct rte_eth_txconf) { .txq_flags = ETH_TXQ_FLAGS_NOOFFLOADS }; + dev_info->rx_offload_capa = + DEV_RX_OFFLOAD_TCP_CKSUM | + DEV_RX_OFFLOAD_UDP_CKSUM | + DEV_RX_OFFLOAD_TCP_LRO; + dev_info->tx_offload_capa = 0; + + if (hw->guest_features & (1ULL << VIRTIO_NET_F_CSUM)) { + dev_info->tx_offload_capa |= + DEV_TX_OFFLOAD_UDP_CKSUM | + DEV_TX_OFFLOAD_TCP_CKSUM; + } + + tso_mask = (1ULL << VIRTIO_NET_F_HOST_TSO4) | + (1ULL << VIRTIO_NET_F_HOST_TSO6); + if ((hw->guest_features & tso_mask) == tso_mask) + dev_info->tx_offload_capa |= DEV_TX_OFFLOAD_TCP_TSO; } /* @@ -1559,10 +1667,5 @@ __rte_unused uint8_t is_rx) return 0; } -static struct rte_driver rte_virtio_driver = { - .type = PMD_PDEV, - .init = rte_virtio_pmd_init, -}; - -PMD_REGISTER_DRIVER(rte_virtio_driver, virtio_net); -DRIVER_REGISTER_PCI_TABLE(virtio_net, pci_id_virtio_map); +RTE_PMD_EXPORT_NAME(net_virtio, __COUNTER__); +RTE_PMD_REGISTER_PCI_TABLE(net_virtio, pci_id_virtio_map); diff --git a/drivers/net/virtio/virtio_ethdev.h b/drivers/net/virtio/virtio_ethdev.h index 2ecec6eb..27d9a190 100644 --- a/drivers/net/virtio/virtio_ethdev.h +++ b/drivers/net/virtio/virtio_ethdev.h @@ -47,14 +47,14 @@ #define PAGE_SIZE 4096 #endif -#define VIRTIO_MAX_RX_QUEUES 128 -#define VIRTIO_MAX_TX_QUEUES 128 +#define VIRTIO_MAX_RX_QUEUES 128U +#define VIRTIO_MAX_TX_QUEUES 128U #define VIRTIO_MAX_MAC_ADDRS 64 #define VIRTIO_MIN_RX_BUFSIZE 64 #define VIRTIO_MAX_RX_PKTLEN 9728 /* Features desired/implemented by this driver. */ -#define VIRTIO_PMD_GUEST_FEATURES \ +#define VIRTIO_PMD_DEFAULT_GUEST_FEATURES \ (1u << VIRTIO_NET_F_MAC | \ 1u << VIRTIO_NET_F_STATUS | \ 1u << VIRTIO_NET_F_MQ | \ @@ -62,8 +62,13 @@ 1u << VIRTIO_NET_F_CTRL_VQ | \ 1u << VIRTIO_NET_F_CTRL_RX | \ 1u << VIRTIO_NET_F_CTRL_VLAN | \ + 1u << VIRTIO_NET_F_CSUM | \ + 1u << VIRTIO_NET_F_HOST_TSO4 | \ + 1u << VIRTIO_NET_F_HOST_TSO6 | \ 1u << VIRTIO_NET_F_MRG_RXBUF | \ - 1ULL << VIRTIO_F_VERSION_1) + 1u << VIRTIO_RING_F_INDIRECT_DESC | \ + 1ULL << VIRTIO_F_VERSION_1 | \ + 1ULL << VIRTIO_F_IOMMU_PLATFORM) /* * CQ function prototype @@ -73,31 +78,15 @@ void virtio_dev_cq_start(struct rte_eth_dev *dev); /* * RX/TX function prototypes */ -void virtio_dev_rxtx_start(struct rte_eth_dev *dev); - -int virtio_dev_queue_setup(struct rte_eth_dev *dev, - int queue_type, - uint16_t queue_idx, - uint16_t vtpci_queue_idx, - uint16_t nb_desc, - unsigned int socket_id, - void **pvq); - -void virtio_dev_queue_release(struct virtqueue *vq); - int virtio_dev_rx_queue_setup(struct rte_eth_dev *dev, uint16_t rx_queue_id, uint16_t nb_rx_desc, unsigned int socket_id, const struct rte_eth_rxconf *rx_conf, struct rte_mempool *mb_pool); -void virtio_dev_rx_queue_release(void *rxq); - int virtio_dev_tx_queue_setup(struct rte_eth_dev *dev, uint16_t tx_queue_id, uint16_t nb_tx_desc, unsigned int socket_id, const struct rte_eth_txconf *tx_conf); -void virtio_dev_tx_queue_release(void *txq); - uint16_t virtio_recv_pkts(void *rx_queue, struct rte_mbuf **rx_pkts, uint16_t nb_pkts); @@ -115,13 +104,4 @@ uint16_t virtio_xmit_pkts_simple(void *tx_queue, struct rte_mbuf **tx_pkts, int eth_virtio_dev_init(struct rte_eth_dev *eth_dev); -/* - * The VIRTIO_NET_F_GUEST_TSO[46] features permit the host to send us - * frames larger than 1514 bytes. We do not yet support software LRO - * via tcp_lro_rx(). - */ -#define VTNET_LRO_FEATURES (VIRTIO_NET_F_GUEST_TSO4 | \ - VIRTIO_NET_F_GUEST_TSO6 | VIRTIO_NET_F_GUEST_ECN) - - #endif /* _VIRTIO_ETHDEV_H_ */ diff --git a/drivers/net/virtio/virtio_pci.c b/drivers/net/virtio/virtio_pci.c index f1a7ca7e..9b47165d 100644 --- a/drivers/net/virtio/virtio_pci.c +++ b/drivers/net/virtio/virtio_pci.c @@ -745,8 +745,9 @@ vtpci_init(struct rte_pci_device *dev, struct virtio_hw *hw, PMD_INIT_LOG(INFO, "trying with legacy virtio pci."); if (legacy_virtio_resource_init(dev, hw, dev_flags) < 0) { if (dev->kdrv == RTE_KDRV_UNKNOWN && - (!dev->devargs || - dev->devargs->type != RTE_DEVTYPE_WHITELISTED_PCI)) { + (!dev->device.devargs || + dev->device.devargs->type != + RTE_DEVTYPE_WHITELISTED_PCI)) { PMD_INIT_LOG(INFO, "skip kernel managed virtio device."); return 1; diff --git a/drivers/net/virtio/virtio_pci.h b/drivers/net/virtio/virtio_pci.h index dd7693fe..de271bfe 100644 --- a/drivers/net/virtio/virtio_pci.h +++ b/drivers/net/virtio/virtio_pci.h @@ -44,8 +44,8 @@ struct virtnet_ctl; /* VirtIO PCI vendor/device ID. */ #define VIRTIO_PCI_VENDORID 0x1AF4 -#define VIRTIO_PCI_DEVICEID_MIN 0x1000 -#define VIRTIO_PCI_DEVICEID_MAX 0x103F +#define VIRTIO_PCI_LEGACY_DEVICEID_NET 0x1000 +#define VIRTIO_PCI_MODERN_DEVICEID_NET 0x1041 /* VirtIO ABI version, this must match exactly. */ #define VIRTIO_PCI_ABI_VERSION 0 @@ -138,6 +138,7 @@ struct virtnet_ctl; #define VIRTIO_RING_F_INDIRECT_DESC 28 #define VIRTIO_F_VERSION_1 32 +#define VIRTIO_F_IOMMU_PLATFORM 33 /* * Some VirtIO feature bits (currently bits 28 through 31) are @@ -145,7 +146,7 @@ struct virtnet_ctl; * rest are per-device feature bits. */ #define VIRTIO_TRANSPORT_F_START 28 -#define VIRTIO_TRANSPORT_F_END 32 +#define VIRTIO_TRANSPORT_F_END 34 /* The Guest publishes the used index for which it expects an interrupt * at the end of the avail ring. Host should ignore the avail->flags field. */ @@ -245,14 +246,14 @@ struct virtio_net_config; struct virtio_hw { struct virtnet_ctl *cvq; struct rte_pci_ioport io; + uint64_t req_guest_features; uint64_t guest_features; - uint32_t max_tx_queues; - uint32_t max_rx_queues; + uint32_t max_queue_pairs; uint16_t vtnet_hdr_size; uint8_t vlan_strip; uint8_t use_msix; - uint8_t started; uint8_t modern; + uint8_t use_simple_rxtx; uint8_t mac_addr[ETHER_ADDR_LEN]; uint32_t notify_off_multiplier; uint8_t *isr; @@ -262,6 +263,8 @@ struct virtio_hw { struct virtio_net_config *dev_cfg; const struct virtio_pci_ops *vtpci_ops; void *virtio_user_dev; + + struct virtqueue **vqs; }; /* diff --git a/drivers/net/virtio/virtio_rxtx.c b/drivers/net/virtio/virtio_rxtx.c index 724517e2..22d97a4e 100644 --- a/drivers/net/virtio/virtio_rxtx.c +++ b/drivers/net/virtio/virtio_rxtx.c @@ -50,6 +50,11 @@ #include <rte_string_fns.h> #include <rte_errno.h> #include <rte_byteorder.h> +#include <rte_cpuflags.h> +#include <rte_net.h> +#include <rte_ip.h> +#include <rte_udp.h> +#include <rte_tcp.h> #include "virtio_logs.h" #include "virtio_ethdev.h" @@ -67,10 +72,6 @@ #define VIRTIO_SIMPLE_FLAGS ((uint32_t)ETH_TXQ_FLAGS_NOMULTSEGS | \ ETH_TXQ_FLAGS_NOOFFLOADS) -#ifdef RTE_MACHINE_CPUFLAG_SSSE3 -static int use_simple_rxtx; -#endif - static void vq_ring_free_chain(struct virtqueue *vq, uint16_t desc_idx) { @@ -208,18 +209,70 @@ virtqueue_enqueue_recv_refill(struct virtqueue *vq, struct rte_mbuf *cookie) return 0; } +/* When doing TSO, the IP length is not included in the pseudo header + * checksum of the packet given to the PMD, but for virtio it is + * expected. + */ +static void +virtio_tso_fix_cksum(struct rte_mbuf *m) +{ + /* common case: header is not fragmented */ + if (likely(rte_pktmbuf_data_len(m) >= m->l2_len + m->l3_len + + m->l4_len)) { + struct ipv4_hdr *iph; + struct ipv6_hdr *ip6h; + struct tcp_hdr *th; + uint16_t prev_cksum, new_cksum, ip_len, ip_paylen; + uint32_t tmp; + + iph = rte_pktmbuf_mtod_offset(m, struct ipv4_hdr *, m->l2_len); + th = RTE_PTR_ADD(iph, m->l3_len); + if ((iph->version_ihl >> 4) == 4) { + iph->hdr_checksum = 0; + iph->hdr_checksum = rte_ipv4_cksum(iph); + ip_len = iph->total_length; + ip_paylen = rte_cpu_to_be_16(rte_be_to_cpu_16(ip_len) - + m->l3_len); + } else { + ip6h = (struct ipv6_hdr *)iph; + ip_paylen = ip6h->payload_len; + } + + /* calculate the new phdr checksum not including ip_paylen */ + prev_cksum = th->cksum; + tmp = prev_cksum; + tmp += ip_paylen; + tmp = (tmp & 0xffff) + (tmp >> 16); + new_cksum = tmp; + + /* replace it in the packet */ + th->cksum = new_cksum; + } +} + +static inline int +tx_offload_enabled(struct virtio_hw *hw) +{ + return vtpci_with_feature(hw, VIRTIO_NET_F_CSUM) || + vtpci_with_feature(hw, VIRTIO_NET_F_HOST_TSO4) || + vtpci_with_feature(hw, VIRTIO_NET_F_HOST_TSO6); +} + static inline void virtqueue_enqueue_xmit(struct virtnet_tx *txvq, struct rte_mbuf *cookie, uint16_t needed, int use_indirect, int can_push) { + struct virtio_tx_region *txr = txvq->virtio_net_hdr_mz->addr; struct vq_desc_extra *dxp; struct virtqueue *vq = txvq->vq; struct vring_desc *start_dp; uint16_t seg_num = cookie->nb_segs; uint16_t head_idx, idx; uint16_t head_size = vq->hw->vtnet_hdr_size; - unsigned long offs; + struct virtio_net_hdr *hdr; + int offload; + offload = tx_offload_enabled(vq->hw); head_idx = vq->vq_desc_head_idx; idx = head_idx; dxp = &vq->vq_descx[idx]; @@ -229,10 +282,12 @@ virtqueue_enqueue_xmit(struct virtnet_tx *txvq, struct rte_mbuf *cookie, start_dp = vq->vq_ring.desc; if (can_push) { - /* put on zero'd transmit header (no offloads) */ - void *hdr = rte_pktmbuf_prepend(cookie, head_size); - - memset(hdr, 0, head_size); + /* prepend cannot fail, checked by caller */ + hdr = (struct virtio_net_hdr *) + rte_pktmbuf_prepend(cookie, head_size); + /* if offload disabled, it is not zeroed below, do it now */ + if (offload == 0) + memset(hdr, 0, head_size); } else if (use_indirect) { /* setup tx ring slot to point to indirect * descriptor list stored in reserved region. @@ -240,14 +295,11 @@ virtqueue_enqueue_xmit(struct virtnet_tx *txvq, struct rte_mbuf *cookie, * the first slot in indirect ring is already preset * to point to the header in reserved region */ - struct virtio_tx_region *txr = txvq->virtio_net_hdr_mz->addr; - - offs = idx * sizeof(struct virtio_tx_region) - + offsetof(struct virtio_tx_region, tx_indir); - - start_dp[idx].addr = txvq->virtio_net_hdr_mem + offs; + start_dp[idx].addr = txvq->virtio_net_hdr_mem + + RTE_PTR_DIFF(&txr[idx].tx_indir, txr); start_dp[idx].len = (seg_num + 1) * sizeof(struct vring_desc); start_dp[idx].flags = VRING_DESC_F_INDIRECT; + hdr = (struct virtio_net_hdr *)&txr[idx].tx_hdr; /* loop below will fill in rest of the indirect elements */ start_dp = txr[idx].tx_indir; @@ -256,15 +308,59 @@ virtqueue_enqueue_xmit(struct virtnet_tx *txvq, struct rte_mbuf *cookie, /* setup first tx ring slot to point to header * stored in reserved region. */ - offs = idx * sizeof(struct virtio_tx_region) - + offsetof(struct virtio_tx_region, tx_hdr); - - start_dp[idx].addr = txvq->virtio_net_hdr_mem + offs; + start_dp[idx].addr = txvq->virtio_net_hdr_mem + + RTE_PTR_DIFF(&txr[idx].tx_hdr, txr); start_dp[idx].len = vq->hw->vtnet_hdr_size; start_dp[idx].flags = VRING_DESC_F_NEXT; + hdr = (struct virtio_net_hdr *)&txr[idx].tx_hdr; + idx = start_dp[idx].next; } + /* Checksum Offload / TSO */ + if (offload) { + if (cookie->ol_flags & PKT_TX_TCP_SEG) + cookie->ol_flags |= PKT_TX_TCP_CKSUM; + + switch (cookie->ol_flags & PKT_TX_L4_MASK) { + case PKT_TX_UDP_CKSUM: + hdr->csum_start = cookie->l2_len + cookie->l3_len; + hdr->csum_offset = offsetof(struct udp_hdr, + dgram_cksum); + hdr->flags = VIRTIO_NET_HDR_F_NEEDS_CSUM; + break; + + case PKT_TX_TCP_CKSUM: + hdr->csum_start = cookie->l2_len + cookie->l3_len; + hdr->csum_offset = offsetof(struct tcp_hdr, cksum); + hdr->flags = VIRTIO_NET_HDR_F_NEEDS_CSUM; + break; + + default: + hdr->csum_start = 0; + hdr->csum_offset = 0; + hdr->flags = 0; + break; + } + + /* TCP Segmentation Offload */ + if (cookie->ol_flags & PKT_TX_TCP_SEG) { + virtio_tso_fix_cksum(cookie); + hdr->gso_type = (cookie->ol_flags & PKT_TX_IPV6) ? + VIRTIO_NET_HDR_GSO_TCPV6 : + VIRTIO_NET_HDR_GSO_TCPV4; + hdr->gso_size = cookie->tso_segsz; + hdr->hdr_len = + cookie->l2_len + + cookie->l3_len + + cookie->l4_len; + } else { + hdr->gso_type = 0; + hdr->gso_size = 0; + hdr->hdr_len = 0; + } + } + do { start_dp[idx].addr = VIRTIO_MBUF_DATA_DMA_ADDR(cookie, vq); start_dp[idx].len = cookie->data_len; @@ -282,207 +378,120 @@ virtqueue_enqueue_xmit(struct virtnet_tx *txvq, struct rte_mbuf *cookie, vq_update_avail_ring(vq, head_idx); } -static void -virtio_dev_vring_start(struct virtqueue *vq) -{ - int size = vq->vq_nentries; - struct vring *vr = &vq->vq_ring; - uint8_t *ring_mem = vq->vq_ring_virt_mem; - - PMD_INIT_FUNC_TRACE(); - - /* - * Reinitialise since virtio port might have been stopped and restarted - */ - memset(vq->vq_ring_virt_mem, 0, vq->vq_ring_size); - vring_init(vr, size, ring_mem, VIRTIO_PCI_VRING_ALIGN); - vq->vq_used_cons_idx = 0; - vq->vq_desc_head_idx = 0; - vq->vq_avail_idx = 0; - vq->vq_desc_tail_idx = (uint16_t)(vq->vq_nentries - 1); - vq->vq_free_cnt = vq->vq_nentries; - memset(vq->vq_descx, 0, sizeof(struct vq_desc_extra) * vq->vq_nentries); - - vring_desc_init(vr->desc, size); - - /* - * Disable device(host) interrupting guest - */ - virtqueue_disable_intr(vq); -} - void virtio_dev_cq_start(struct rte_eth_dev *dev) { struct virtio_hw *hw = dev->data->dev_private; if (hw->cvq && hw->cvq->vq) { - virtio_dev_vring_start(hw->cvq->vq); VIRTQUEUE_DUMP((struct virtqueue *)hw->cvq->vq); } } -void -virtio_dev_rxtx_start(struct rte_eth_dev *dev) +int +virtio_dev_rx_queue_setup(struct rte_eth_dev *dev, + uint16_t queue_idx, + uint16_t nb_desc, + unsigned int socket_id __rte_unused, + __rte_unused const struct rte_eth_rxconf *rx_conf, + struct rte_mempool *mp) { - /* - * Start receive and transmit vrings - * - Setup vring structure for all queues - * - Initialize descriptor for the rx vring - * - Allocate blank mbufs for the each rx descriptor - * - */ - uint16_t i; + uint16_t vtpci_queue_idx = 2 * queue_idx + VTNET_SQ_RQ_QUEUE_IDX; + struct virtio_hw *hw = dev->data->dev_private; + struct virtqueue *vq = hw->vqs[vtpci_queue_idx]; + struct virtnet_rx *rxvq; + int error, nbufs; + struct rte_mbuf *m; uint16_t desc_idx; PMD_INIT_FUNC_TRACE(); - /* Start rx vring. */ - for (i = 0; i < dev->data->nb_rx_queues; i++) { - struct virtnet_rx *rxvq = dev->data->rx_queues[i]; - struct virtqueue *vq = rxvq->vq; - int error, nbufs; - struct rte_mbuf *m; - - virtio_dev_vring_start(vq); - if (rxvq->mpool == NULL) { - rte_exit(EXIT_FAILURE, - "Cannot allocate mbufs for rx virtqueue"); - } - - /* Allocate blank mbufs for the each rx descriptor */ - nbufs = 0; - error = ENOSPC; - -#ifdef RTE_MACHINE_CPUFLAG_SSSE3 - if (use_simple_rxtx) { - for (desc_idx = 0; desc_idx < vq->vq_nentries; - desc_idx++) { - vq->vq_ring.avail->ring[desc_idx] = desc_idx; - vq->vq_ring.desc[desc_idx].flags = - VRING_DESC_F_WRITE; - } - } -#endif - memset(&rxvq->fake_mbuf, 0, sizeof(rxvq->fake_mbuf)); - for (desc_idx = 0; desc_idx < RTE_PMD_VIRTIO_RX_MAX_BURST; - desc_idx++) { - vq->sw_ring[vq->vq_nentries + desc_idx] = - &rxvq->fake_mbuf; - } - - while (!virtqueue_full(vq)) { - m = rte_mbuf_raw_alloc(rxvq->mpool); - if (m == NULL) - break; + if (nb_desc == 0 || nb_desc > vq->vq_nentries) + nb_desc = vq->vq_nentries; + vq->vq_free_cnt = RTE_MIN(vq->vq_free_cnt, nb_desc); - /****************************************** - * Enqueue allocated buffers * - *******************************************/ -#ifdef RTE_MACHINE_CPUFLAG_SSSE3 - if (use_simple_rxtx) - error = virtqueue_enqueue_recv_refill_simple(vq, m); - else -#endif - error = virtqueue_enqueue_recv_refill(vq, m); - if (error) { - rte_pktmbuf_free(m); - break; - } - nbufs++; - } + rxvq = &vq->rxq; + rxvq->queue_id = queue_idx; + rxvq->mpool = mp; + if (rxvq->mpool == NULL) { + rte_exit(EXIT_FAILURE, + "Cannot allocate mbufs for rx virtqueue"); + } + dev->data->rx_queues[queue_idx] = rxvq; - vq_update_avail_idx(vq); - PMD_INIT_LOG(DEBUG, "Allocated %d bufs", nbufs); + /* Allocate blank mbufs for the each rx descriptor */ + nbufs = 0; + error = ENOSPC; - VIRTQUEUE_DUMP(vq); + if (hw->use_simple_rxtx) { + for (desc_idx = 0; desc_idx < vq->vq_nentries; + desc_idx++) { + vq->vq_ring.avail->ring[desc_idx] = desc_idx; + vq->vq_ring.desc[desc_idx].flags = + VRING_DESC_F_WRITE; + } } - /* Start tx vring. */ - for (i = 0; i < dev->data->nb_tx_queues; i++) { - struct virtnet_tx *txvq = dev->data->tx_queues[i]; - struct virtqueue *vq = txvq->vq; - - virtio_dev_vring_start(vq); -#ifdef RTE_MACHINE_CPUFLAG_SSSE3 - if (use_simple_rxtx) { - uint16_t mid_idx = vq->vq_nentries >> 1; - - for (desc_idx = 0; desc_idx < mid_idx; desc_idx++) { - vq->vq_ring.avail->ring[desc_idx] = - desc_idx + mid_idx; - vq->vq_ring.desc[desc_idx + mid_idx].next = - desc_idx; - vq->vq_ring.desc[desc_idx + mid_idx].addr = - txvq->virtio_net_hdr_mem + - offsetof(struct virtio_tx_region, tx_hdr); - vq->vq_ring.desc[desc_idx + mid_idx].len = - vq->hw->vtnet_hdr_size; - vq->vq_ring.desc[desc_idx + mid_idx].flags = - VRING_DESC_F_NEXT; - vq->vq_ring.desc[desc_idx].flags = 0; - } - for (desc_idx = mid_idx; desc_idx < vq->vq_nentries; - desc_idx++) - vq->vq_ring.avail->ring[desc_idx] = desc_idx; - } -#endif - VIRTQUEUE_DUMP(vq); + memset(&rxvq->fake_mbuf, 0, sizeof(rxvq->fake_mbuf)); + for (desc_idx = 0; desc_idx < RTE_PMD_VIRTIO_RX_MAX_BURST; + desc_idx++) { + vq->sw_ring[vq->vq_nentries + desc_idx] = + &rxvq->fake_mbuf; } -} -int -virtio_dev_rx_queue_setup(struct rte_eth_dev *dev, - uint16_t queue_idx, - uint16_t nb_desc, - unsigned int socket_id, - __rte_unused const struct rte_eth_rxconf *rx_conf, - struct rte_mempool *mp) -{ - uint16_t vtpci_queue_idx = 2 * queue_idx + VTNET_SQ_RQ_QUEUE_IDX; - struct virtnet_rx *rxvq; - int ret; + while (!virtqueue_full(vq)) { + m = rte_mbuf_raw_alloc(rxvq->mpool); + if (m == NULL) + break; - PMD_INIT_FUNC_TRACE(); - ret = virtio_dev_queue_setup(dev, VTNET_RQ, queue_idx, vtpci_queue_idx, - nb_desc, socket_id, (void **)&rxvq); - if (ret < 0) { - PMD_INIT_LOG(ERR, "rvq initialization failed"); - return ret; + /* Enqueue allocated buffers */ + if (hw->use_simple_rxtx) + error = virtqueue_enqueue_recv_refill_simple(vq, m); + else + error = virtqueue_enqueue_recv_refill(vq, m); + + if (error) { + rte_pktmbuf_free(m); + break; + } + nbufs++; } - /* Create mempool for rx mbuf allocation */ - rxvq->mpool = mp; + vq_update_avail_idx(vq); - dev->data->rx_queues[queue_idx] = rxvq; + PMD_INIT_LOG(DEBUG, "Allocated %d bufs", nbufs); -#ifdef RTE_MACHINE_CPUFLAG_SSSE3 virtio_rxq_vec_setup(rxvq); -#endif + + VIRTQUEUE_DUMP(vq); return 0; } -void -virtio_dev_rx_queue_release(void *rxq) +static void +virtio_update_rxtx_handler(struct rte_eth_dev *dev, + const struct rte_eth_txconf *tx_conf) { - struct virtnet_rx *rxvq = rxq; - struct virtqueue *vq; - const struct rte_memzone *mz; - - if (rxvq == NULL) - return; - - /* - * rxvq is freed when vq is freed, and as mz should be freed after the - * del_queue, so we reserve the mz pointer first. - */ - vq = rxvq->vq; - mz = rxvq->mz; + uint8_t use_simple_rxtx = 0; + struct virtio_hw *hw = dev->data->dev_private; - virtio_dev_queue_release(vq); - rte_memzone_free(mz); +#if defined RTE_ARCH_X86 + if (rte_cpu_get_flag_enabled(RTE_CPUFLAG_SSE3)) + use_simple_rxtx = 1; +#elif defined RTE_ARCH_ARM64 || defined CONFIG_RTE_ARCH_ARM + if (rte_cpu_get_flag_enabled(RTE_CPUFLAG_NEON)) + use_simple_rxtx = 1; +#endif + /* Use simple rx/tx func if single segment and no offloads */ + if (use_simple_rxtx && + (tx_conf->txq_flags & VIRTIO_SIMPLE_FLAGS) == VIRTIO_SIMPLE_FLAGS && + !vtpci_with_feature(hw, VIRTIO_NET_F_MRG_RXBUF)) { + PMD_INIT_LOG(INFO, "Using simple rx/tx path"); + dev->tx_pkt_burst = virtio_xmit_pkts_simple; + dev->rx_pkt_burst = virtio_recv_pkts_vec; + hw->use_simple_rxtx = use_simple_rxtx; + } } /* @@ -496,45 +505,26 @@ int virtio_dev_tx_queue_setup(struct rte_eth_dev *dev, uint16_t queue_idx, uint16_t nb_desc, - unsigned int socket_id, + unsigned int socket_id __rte_unused, const struct rte_eth_txconf *tx_conf) { uint8_t vtpci_queue_idx = 2 * queue_idx + VTNET_SQ_TQ_QUEUE_IDX; - -#ifdef RTE_MACHINE_CPUFLAG_SSSE3 struct virtio_hw *hw = dev->data->dev_private; -#endif + struct virtqueue *vq = hw->vqs[vtpci_queue_idx]; struct virtnet_tx *txvq; - struct virtqueue *vq; uint16_t tx_free_thresh; - int ret; + uint16_t desc_idx; PMD_INIT_FUNC_TRACE(); - if ((tx_conf->txq_flags & ETH_TXQ_FLAGS_NOXSUMS) - != ETH_TXQ_FLAGS_NOXSUMS) { - PMD_INIT_LOG(ERR, "TX checksum offload not supported\n"); - return -EINVAL; - } + virtio_update_rxtx_handler(dev, tx_conf); -#ifdef RTE_MACHINE_CPUFLAG_SSSE3 - /* Use simple rx/tx func if single segment and no offloads */ - if ((tx_conf->txq_flags & VIRTIO_SIMPLE_FLAGS) == VIRTIO_SIMPLE_FLAGS && - !vtpci_with_feature(hw, VIRTIO_NET_F_MRG_RXBUF)) { - PMD_INIT_LOG(INFO, "Using simple rx/tx path"); - dev->tx_pkt_burst = virtio_xmit_pkts_simple; - dev->rx_pkt_burst = virtio_recv_pkts_vec; - use_simple_rxtx = 1; - } -#endif + if (nb_desc == 0 || nb_desc > vq->vq_nentries) + nb_desc = vq->vq_nentries; + vq->vq_free_cnt = RTE_MIN(vq->vq_free_cnt, nb_desc); - ret = virtio_dev_queue_setup(dev, VTNET_TQ, queue_idx, vtpci_queue_idx, - nb_desc, socket_id, (void **)&txvq); - if (ret < 0) { - PMD_INIT_LOG(ERR, "tvq initialization failed"); - return ret; - } - vq = txvq->vq; + txvq = &vq->txq; + txvq->queue_id = queue_idx; tx_free_thresh = tx_conf->tx_free_thresh; if (tx_free_thresh == 0) @@ -552,32 +542,32 @@ virtio_dev_tx_queue_setup(struct rte_eth_dev *dev, vq->vq_free_thresh = tx_free_thresh; - dev->data->tx_queues[queue_idx] = txvq; - return 0; -} - -void -virtio_dev_tx_queue_release(void *txq) -{ - struct virtnet_tx *txvq = txq; - struct virtqueue *vq; - const struct rte_memzone *mz; - const struct rte_memzone *hdr_mz; - - if (txvq == NULL) - return; + if (hw->use_simple_rxtx) { + uint16_t mid_idx = vq->vq_nentries >> 1; + + for (desc_idx = 0; desc_idx < mid_idx; desc_idx++) { + vq->vq_ring.avail->ring[desc_idx] = + desc_idx + mid_idx; + vq->vq_ring.desc[desc_idx + mid_idx].next = + desc_idx; + vq->vq_ring.desc[desc_idx + mid_idx].addr = + txvq->virtio_net_hdr_mem + + offsetof(struct virtio_tx_region, tx_hdr); + vq->vq_ring.desc[desc_idx + mid_idx].len = + vq->hw->vtnet_hdr_size; + vq->vq_ring.desc[desc_idx + mid_idx].flags = + VRING_DESC_F_NEXT; + vq->vq_ring.desc[desc_idx].flags = 0; + } + for (desc_idx = mid_idx; desc_idx < vq->vq_nentries; + desc_idx++) + vq->vq_ring.avail->ring[desc_idx] = desc_idx; + } - /* - * txvq is freed when vq is freed, and as mz should be freed after the - * del_queue, so we reserve the mz pointer first. - */ - vq = txvq->vq; - mz = txvq->mz; - hdr_mz = txvq->virtio_net_hdr_mz; + VIRTQUEUE_DUMP(vq); - virtio_dev_queue_release(vq); - rte_memzone_free(mz); - rte_memzone_free(hdr_mz); + dev->data->tx_queues[queue_idx] = txvq; + return 0; } static void @@ -627,6 +617,86 @@ virtio_update_packet_stats(struct virtnet_stats *stats, struct rte_mbuf *mbuf) } } +/* Optionally fill offload information in structure */ +static int +virtio_rx_offload(struct rte_mbuf *m, struct virtio_net_hdr *hdr) +{ + struct rte_net_hdr_lens hdr_lens; + uint32_t hdrlen, ptype; + int l4_supported = 0; + + /* nothing to do */ + if (hdr->flags == 0 && hdr->gso_type == VIRTIO_NET_HDR_GSO_NONE) + return 0; + + m->ol_flags |= PKT_RX_IP_CKSUM_UNKNOWN; + + ptype = rte_net_get_ptype(m, &hdr_lens, RTE_PTYPE_ALL_MASK); + m->packet_type = ptype; + if ((ptype & RTE_PTYPE_L4_MASK) == RTE_PTYPE_L4_TCP || + (ptype & RTE_PTYPE_L4_MASK) == RTE_PTYPE_L4_UDP || + (ptype & RTE_PTYPE_L4_MASK) == RTE_PTYPE_L4_SCTP) + l4_supported = 1; + + if (hdr->flags & VIRTIO_NET_HDR_F_NEEDS_CSUM) { + hdrlen = hdr_lens.l2_len + hdr_lens.l3_len + hdr_lens.l4_len; + if (hdr->csum_start <= hdrlen && l4_supported) { + m->ol_flags |= PKT_RX_L4_CKSUM_NONE; + } else { + /* Unknown proto or tunnel, do sw cksum. We can assume + * the cksum field is in the first segment since the + * buffers we provided to the host are large enough. + * In case of SCTP, this will be wrong since it's a CRC + * but there's nothing we can do. + */ + uint16_t csum, off; + + rte_raw_cksum_mbuf(m, hdr->csum_start, + rte_pktmbuf_pkt_len(m) - hdr->csum_start, + &csum); + if (likely(csum != 0xffff)) + csum = ~csum; + off = hdr->csum_offset + hdr->csum_start; + if (rte_pktmbuf_data_len(m) >= off + 1) + *rte_pktmbuf_mtod_offset(m, uint16_t *, + off) = csum; + } + } else if (hdr->flags & VIRTIO_NET_HDR_F_DATA_VALID && l4_supported) { + m->ol_flags |= PKT_RX_L4_CKSUM_GOOD; + } + + /* GSO request, save required information in mbuf */ + if (hdr->gso_type != VIRTIO_NET_HDR_GSO_NONE) { + /* Check unsupported modes */ + if ((hdr->gso_type & VIRTIO_NET_HDR_GSO_ECN) || + (hdr->gso_size == 0)) { + return -EINVAL; + } + + /* Update mss lengthes in mbuf */ + m->tso_segsz = hdr->gso_size; + switch (hdr->gso_type & ~VIRTIO_NET_HDR_GSO_ECN) { + case VIRTIO_NET_HDR_GSO_TCPV4: + case VIRTIO_NET_HDR_GSO_TCPV6: + m->ol_flags |= PKT_RX_LRO | \ + PKT_RX_L4_CKSUM_NONE; + break; + default: + return -EINVAL; + } + } + + return 0; +} + +static inline int +rx_offload_enabled(struct virtio_hw *hw) +{ + return vtpci_with_feature(hw, VIRTIO_NET_F_GUEST_CSUM) || + vtpci_with_feature(hw, VIRTIO_NET_F_GUEST_TSO4) || + vtpci_with_feature(hw, VIRTIO_NET_F_GUEST_TSO6); +} + #define VIRTIO_MBUF_BURST_SZ 64 #define DESC_PER_CACHELINE (RTE_CACHE_LINE_SIZE / sizeof(struct vring_desc)) uint16_t @@ -642,6 +712,8 @@ virtio_recv_pkts(void *rx_queue, struct rte_mbuf **rx_pkts, uint16_t nb_pkts) int error; uint32_t i, nb_enqueued; uint32_t hdr_size; + int offload; + struct virtio_net_hdr *hdr; nb_used = VIRTQUEUE_NUSED(vq); @@ -659,6 +731,7 @@ virtio_recv_pkts(void *rx_queue, struct rte_mbuf **rx_pkts, uint16_t nb_pkts) nb_rx = 0; nb_enqueued = 0; hdr_size = hw->vtnet_hdr_size; + offload = rx_offload_enabled(hw); for (i = 0; i < num ; i++) { rxm = rcv_pkts[i]; @@ -683,9 +756,18 @@ virtio_recv_pkts(void *rx_queue, struct rte_mbuf **rx_pkts, uint16_t nb_pkts) rxm->pkt_len = (uint32_t)(len[i] - hdr_size); rxm->data_len = (uint16_t)(len[i] - hdr_size); + hdr = (struct virtio_net_hdr *)((char *)rxm->buf_addr + + RTE_PKTMBUF_HEADROOM - hdr_size); + if (hw->vlan_strip) rte_vlan_strip(rxm); + if (offload && virtio_rx_offload(rxm, hdr) < 0) { + virtio_discard_rxbuf(vq, rxm); + rxvq->stats.errors++; + continue; + } + VIRTIO_DUMP_PACKET(rxm, rxm->data_len); rx_pkts[nb_rx++] = rxm; @@ -745,6 +827,7 @@ virtio_recv_mergeable_pkts(void *rx_queue, uint16_t extra_idx; uint32_t seg_res; uint32_t hdr_size; + int offload; nb_used = VIRTQUEUE_NUSED(vq); @@ -760,6 +843,7 @@ virtio_recv_mergeable_pkts(void *rx_queue, extra_idx = 0; seg_res = 0; hdr_size = hw->vtnet_hdr_size; + offload = rx_offload_enabled(hw); while (i < nb_used) { struct virtio_net_hdr_mrg_rxbuf *header; @@ -805,6 +889,12 @@ virtio_recv_mergeable_pkts(void *rx_queue, rx_pkts[nb_rx] = rxm; prev = rxm; + if (offload && virtio_rx_offload(rxm, &header->hdr) < 0) { + virtio_discard_rxbuf(vq, rxm); + rxvq->stats.errors++; + continue; + } + seg_res = seg_num - 1; while (seg_res != 0) { diff --git a/drivers/net/virtio/virtio_rxtx.h b/drivers/net/virtio/virtio_rxtx.h index 058b56a1..28f82d6a 100644 --- a/drivers/net/virtio/virtio_rxtx.h +++ b/drivers/net/virtio/virtio_rxtx.h @@ -86,10 +86,9 @@ struct virtnet_ctl { const struct rte_memzone *mz; /**< mem zone to populate RX ring. */ }; -#ifdef RTE_MACHINE_CPUFLAG_SSSE3 int virtio_rxq_vec_setup(struct virtnet_rx *rxvq); int virtqueue_enqueue_recv_refill_simple(struct virtqueue *vq, struct rte_mbuf *m); -#endif + #endif /* _VIRTIO_RXTX_H_ */ diff --git a/drivers/net/virtio/virtio_rxtx_simple.c b/drivers/net/virtio/virtio_rxtx_simple.c index 6517aa80..b651e53b 100644 --- a/drivers/net/virtio/virtio_rxtx_simple.c +++ b/drivers/net/virtio/virtio_rxtx_simple.c @@ -37,8 +37,6 @@ #include <string.h> #include <errno.h> -#include <tmmintrin.h> - #include <rte_cycles.h> #include <rte_memory.h> #include <rte_memzone.h> @@ -53,14 +51,7 @@ #include <rte_errno.h> #include <rte_byteorder.h> -#include "virtio_logs.h" -#include "virtio_ethdev.h" -#include "virtqueue.h" -#include "virtio_rxtx.h" - -#define RTE_VIRTIO_VPMD_RX_BURST 32 -#define RTE_VIRTIO_DESC_PER_LOOP 8 -#define RTE_VIRTIO_VPMD_RX_REARM_THRESH RTE_VIRTIO_VPMD_RX_BURST +#include "virtio_rxtx_simple.h" #ifndef __INTEL_COMPILER #pragma GCC diagnostic ignored "-Wcast-qual" @@ -92,257 +83,6 @@ virtqueue_enqueue_recv_refill_simple(struct virtqueue *vq, return 0; } -static inline void -virtio_rxq_rearm_vec(struct virtnet_rx *rxvq) -{ - int i; - uint16_t desc_idx; - struct rte_mbuf **sw_ring; - struct vring_desc *start_dp; - int ret; - struct virtqueue *vq = rxvq->vq; - - desc_idx = vq->vq_avail_idx & (vq->vq_nentries - 1); - sw_ring = &vq->sw_ring[desc_idx]; - start_dp = &vq->vq_ring.desc[desc_idx]; - - ret = rte_mempool_get_bulk(rxvq->mpool, (void **)sw_ring, - RTE_VIRTIO_VPMD_RX_REARM_THRESH); - if (unlikely(ret)) { - rte_eth_devices[rxvq->port_id].data->rx_mbuf_alloc_failed += - RTE_VIRTIO_VPMD_RX_REARM_THRESH; - return; - } - - for (i = 0; i < RTE_VIRTIO_VPMD_RX_REARM_THRESH; i++) { - uintptr_t p; - - p = (uintptr_t)&sw_ring[i]->rearm_data; - *(uint64_t *)p = rxvq->mbuf_initializer; - - start_dp[i].addr = - VIRTIO_MBUF_ADDR(sw_ring[i], vq) + - RTE_PKTMBUF_HEADROOM - vq->hw->vtnet_hdr_size; - start_dp[i].len = sw_ring[i]->buf_len - - RTE_PKTMBUF_HEADROOM + vq->hw->vtnet_hdr_size; - } - - vq->vq_avail_idx += RTE_VIRTIO_VPMD_RX_REARM_THRESH; - vq->vq_free_cnt -= RTE_VIRTIO_VPMD_RX_REARM_THRESH; - vq_update_avail_idx(vq); -} - -/* virtio vPMD receive routine, only accept(nb_pkts >= RTE_VIRTIO_DESC_PER_LOOP) - * - * This routine is for non-mergeable RX, one desc for each guest buffer. - * This routine is based on the RX ring layout optimization. Each entry in the - * avail ring points to the desc with the same index in the desc ring and this - * will never be changed in the driver. - * - * - nb_pkts < RTE_VIRTIO_DESC_PER_LOOP, just return no packet - */ -uint16_t -virtio_recv_pkts_vec(void *rx_queue, struct rte_mbuf **rx_pkts, - uint16_t nb_pkts) -{ - struct virtnet_rx *rxvq = rx_queue; - struct virtqueue *vq = rxvq->vq; - uint16_t nb_used; - uint16_t desc_idx; - struct vring_used_elem *rused; - struct rte_mbuf **sw_ring; - struct rte_mbuf **sw_ring_end; - uint16_t nb_pkts_received; - __m128i shuf_msk1, shuf_msk2, len_adjust; - - shuf_msk1 = _mm_set_epi8( - 0xFF, 0xFF, 0xFF, 0xFF, - 0xFF, 0xFF, /* vlan tci */ - 5, 4, /* dat len */ - 0xFF, 0xFF, 5, 4, /* pkt len */ - 0xFF, 0xFF, 0xFF, 0xFF /* packet type */ - - ); - - shuf_msk2 = _mm_set_epi8( - 0xFF, 0xFF, 0xFF, 0xFF, - 0xFF, 0xFF, /* vlan tci */ - 13, 12, /* dat len */ - 0xFF, 0xFF, 13, 12, /* pkt len */ - 0xFF, 0xFF, 0xFF, 0xFF /* packet type */ - ); - - /* Subtract the header length. - * In which case do we need the header length in used->len ? - */ - len_adjust = _mm_set_epi16( - 0, 0, - 0, - (uint16_t)-vq->hw->vtnet_hdr_size, - 0, (uint16_t)-vq->hw->vtnet_hdr_size, - 0, 0); - - if (unlikely(nb_pkts < RTE_VIRTIO_DESC_PER_LOOP)) - return 0; - - nb_used = VIRTQUEUE_NUSED(vq); - - rte_compiler_barrier(); - - if (unlikely(nb_used == 0)) - return 0; - - nb_pkts = RTE_ALIGN_FLOOR(nb_pkts, RTE_VIRTIO_DESC_PER_LOOP); - nb_used = RTE_MIN(nb_used, nb_pkts); - - desc_idx = (uint16_t)(vq->vq_used_cons_idx & (vq->vq_nentries - 1)); - rused = &vq->vq_ring.used->ring[desc_idx]; - sw_ring = &vq->sw_ring[desc_idx]; - sw_ring_end = &vq->sw_ring[vq->vq_nentries]; - - _mm_prefetch((const void *)rused, _MM_HINT_T0); - - if (vq->vq_free_cnt >= RTE_VIRTIO_VPMD_RX_REARM_THRESH) { - virtio_rxq_rearm_vec(rxvq); - if (unlikely(virtqueue_kick_prepare(vq))) - virtqueue_notify(vq); - } - - for (nb_pkts_received = 0; - nb_pkts_received < nb_used;) { - __m128i desc[RTE_VIRTIO_DESC_PER_LOOP / 2]; - __m128i mbp[RTE_VIRTIO_DESC_PER_LOOP / 2]; - __m128i pkt_mb[RTE_VIRTIO_DESC_PER_LOOP]; - - mbp[0] = _mm_loadu_si128((__m128i *)(sw_ring + 0)); - desc[0] = _mm_loadu_si128((__m128i *)(rused + 0)); - _mm_storeu_si128((__m128i *)&rx_pkts[0], mbp[0]); - - mbp[1] = _mm_loadu_si128((__m128i *)(sw_ring + 2)); - desc[1] = _mm_loadu_si128((__m128i *)(rused + 2)); - _mm_storeu_si128((__m128i *)&rx_pkts[2], mbp[1]); - - mbp[2] = _mm_loadu_si128((__m128i *)(sw_ring + 4)); - desc[2] = _mm_loadu_si128((__m128i *)(rused + 4)); - _mm_storeu_si128((__m128i *)&rx_pkts[4], mbp[2]); - - mbp[3] = _mm_loadu_si128((__m128i *)(sw_ring + 6)); - desc[3] = _mm_loadu_si128((__m128i *)(rused + 6)); - _mm_storeu_si128((__m128i *)&rx_pkts[6], mbp[3]); - - pkt_mb[1] = _mm_shuffle_epi8(desc[0], shuf_msk2); - pkt_mb[0] = _mm_shuffle_epi8(desc[0], shuf_msk1); - pkt_mb[1] = _mm_add_epi16(pkt_mb[1], len_adjust); - pkt_mb[0] = _mm_add_epi16(pkt_mb[0], len_adjust); - _mm_storeu_si128((void *)&rx_pkts[1]->rx_descriptor_fields1, - pkt_mb[1]); - _mm_storeu_si128((void *)&rx_pkts[0]->rx_descriptor_fields1, - pkt_mb[0]); - - pkt_mb[3] = _mm_shuffle_epi8(desc[1], shuf_msk2); - pkt_mb[2] = _mm_shuffle_epi8(desc[1], shuf_msk1); - pkt_mb[3] = _mm_add_epi16(pkt_mb[3], len_adjust); - pkt_mb[2] = _mm_add_epi16(pkt_mb[2], len_adjust); - _mm_storeu_si128((void *)&rx_pkts[3]->rx_descriptor_fields1, - pkt_mb[3]); - _mm_storeu_si128((void *)&rx_pkts[2]->rx_descriptor_fields1, - pkt_mb[2]); - - pkt_mb[5] = _mm_shuffle_epi8(desc[2], shuf_msk2); - pkt_mb[4] = _mm_shuffle_epi8(desc[2], shuf_msk1); - pkt_mb[5] = _mm_add_epi16(pkt_mb[5], len_adjust); - pkt_mb[4] = _mm_add_epi16(pkt_mb[4], len_adjust); - _mm_storeu_si128((void *)&rx_pkts[5]->rx_descriptor_fields1, - pkt_mb[5]); - _mm_storeu_si128((void *)&rx_pkts[4]->rx_descriptor_fields1, - pkt_mb[4]); - - pkt_mb[7] = _mm_shuffle_epi8(desc[3], shuf_msk2); - pkt_mb[6] = _mm_shuffle_epi8(desc[3], shuf_msk1); - pkt_mb[7] = _mm_add_epi16(pkt_mb[7], len_adjust); - pkt_mb[6] = _mm_add_epi16(pkt_mb[6], len_adjust); - _mm_storeu_si128((void *)&rx_pkts[7]->rx_descriptor_fields1, - pkt_mb[7]); - _mm_storeu_si128((void *)&rx_pkts[6]->rx_descriptor_fields1, - pkt_mb[6]); - - if (unlikely(nb_used <= RTE_VIRTIO_DESC_PER_LOOP)) { - if (sw_ring + nb_used <= sw_ring_end) - nb_pkts_received += nb_used; - else - nb_pkts_received += sw_ring_end - sw_ring; - break; - } else { - if (unlikely(sw_ring + RTE_VIRTIO_DESC_PER_LOOP >= - sw_ring_end)) { - nb_pkts_received += sw_ring_end - sw_ring; - break; - } else { - nb_pkts_received += RTE_VIRTIO_DESC_PER_LOOP; - - rx_pkts += RTE_VIRTIO_DESC_PER_LOOP; - sw_ring += RTE_VIRTIO_DESC_PER_LOOP; - rused += RTE_VIRTIO_DESC_PER_LOOP; - nb_used -= RTE_VIRTIO_DESC_PER_LOOP; - } - } - } - - vq->vq_used_cons_idx += nb_pkts_received; - vq->vq_free_cnt += nb_pkts_received; - rxvq->stats.packets += nb_pkts_received; - return nb_pkts_received; -} - -#define VIRTIO_TX_FREE_THRESH 32 -#define VIRTIO_TX_MAX_FREE_BUF_SZ 32 -#define VIRTIO_TX_FREE_NR 32 -/* TODO: vq->tx_free_cnt could mean num of free slots so we could avoid shift */ -static inline void -virtio_xmit_cleanup(struct virtqueue *vq) -{ - uint16_t i, desc_idx; - uint32_t nb_free = 0; - struct rte_mbuf *m, *free[VIRTIO_TX_MAX_FREE_BUF_SZ]; - - desc_idx = (uint16_t)(vq->vq_used_cons_idx & - ((vq->vq_nentries >> 1) - 1)); - m = (struct rte_mbuf *)vq->vq_descx[desc_idx++].cookie; - m = __rte_pktmbuf_prefree_seg(m); - if (likely(m != NULL)) { - free[0] = m; - nb_free = 1; - for (i = 1; i < VIRTIO_TX_FREE_NR; i++) { - m = (struct rte_mbuf *)vq->vq_descx[desc_idx++].cookie; - m = __rte_pktmbuf_prefree_seg(m); - if (likely(m != NULL)) { - if (likely(m->pool == free[0]->pool)) - free[nb_free++] = m; - else { - rte_mempool_put_bulk(free[0]->pool, - (void **)free, - RTE_MIN(RTE_DIM(free), - nb_free)); - free[0] = m; - nb_free = 1; - } - } - } - rte_mempool_put_bulk(free[0]->pool, (void **)free, - RTE_MIN(RTE_DIM(free), nb_free)); - } else { - for (i = 1; i < VIRTIO_TX_FREE_NR; i++) { - m = (struct rte_mbuf *)vq->vq_descx[desc_idx++].cookie; - m = __rte_pktmbuf_prefree_seg(m); - if (m != NULL) - rte_mempool_put(m->pool, m); - } - } - - vq->vq_used_cons_idx += VIRTIO_TX_FREE_NR; - vq->vq_free_cnt += (VIRTIO_TX_FREE_NR << 1); -} - uint16_t virtio_xmit_pkts_simple(void *tx_queue, struct rte_mbuf **tx_pkts, uint16_t nb_pkts) @@ -423,3 +163,13 @@ virtio_rxq_vec_setup(struct virtnet_rx *rxq) return 0; } + +/* Stub for linkage when arch specific implementation is not available */ +uint16_t __attribute__((weak)) +virtio_recv_pkts_vec(void *rx_queue __rte_unused, + struct rte_mbuf **rx_pkts __rte_unused, + uint16_t nb_pkts __rte_unused) +{ + rte_panic("Wrong weak function linked by linker\n"); + return 0; +} diff --git a/drivers/net/virtio/virtio_rxtx_simple.h b/drivers/net/virtio/virtio_rxtx_simple.h new file mode 100644 index 00000000..b08f8594 --- /dev/null +++ b/drivers/net/virtio/virtio_rxtx_simple.h @@ -0,0 +1,136 @@ +/*- + * 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. + */ + +#ifndef _VIRTIO_RXTX_SIMPLE_H_ +#define _VIRTIO_RXTX_SIMPLE_H_ + +#include <stdint.h> + +#include "virtio_logs.h" +#include "virtio_ethdev.h" +#include "virtqueue.h" +#include "virtio_rxtx.h" + +#define RTE_VIRTIO_VPMD_RX_BURST 32 +#define RTE_VIRTIO_VPMD_RX_REARM_THRESH RTE_VIRTIO_VPMD_RX_BURST + +static inline void +virtio_rxq_rearm_vec(struct virtnet_rx *rxvq) +{ + int i; + uint16_t desc_idx; + struct rte_mbuf **sw_ring; + struct vring_desc *start_dp; + int ret; + struct virtqueue *vq = rxvq->vq; + + desc_idx = vq->vq_avail_idx & (vq->vq_nentries - 1); + sw_ring = &vq->sw_ring[desc_idx]; + start_dp = &vq->vq_ring.desc[desc_idx]; + + ret = rte_mempool_get_bulk(rxvq->mpool, (void **)sw_ring, + RTE_VIRTIO_VPMD_RX_REARM_THRESH); + if (unlikely(ret)) { + rte_eth_devices[rxvq->port_id].data->rx_mbuf_alloc_failed += + RTE_VIRTIO_VPMD_RX_REARM_THRESH; + return; + } + + for (i = 0; i < RTE_VIRTIO_VPMD_RX_REARM_THRESH; i++) { + uintptr_t p; + + p = (uintptr_t)&sw_ring[i]->rearm_data; + *(uint64_t *)p = rxvq->mbuf_initializer; + + start_dp[i].addr = + VIRTIO_MBUF_ADDR(sw_ring[i], vq) + + RTE_PKTMBUF_HEADROOM - vq->hw->vtnet_hdr_size; + start_dp[i].len = sw_ring[i]->buf_len - + RTE_PKTMBUF_HEADROOM + vq->hw->vtnet_hdr_size; + } + + vq->vq_avail_idx += RTE_VIRTIO_VPMD_RX_REARM_THRESH; + vq->vq_free_cnt -= RTE_VIRTIO_VPMD_RX_REARM_THRESH; + vq_update_avail_idx(vq); +} + +#define VIRTIO_TX_FREE_THRESH 32 +#define VIRTIO_TX_MAX_FREE_BUF_SZ 32 +#define VIRTIO_TX_FREE_NR 32 +/* TODO: vq->tx_free_cnt could mean num of free slots so we could avoid shift */ +static inline void +virtio_xmit_cleanup(struct virtqueue *vq) +{ + uint16_t i, desc_idx; + uint32_t nb_free = 0; + struct rte_mbuf *m, *free[VIRTIO_TX_MAX_FREE_BUF_SZ]; + + desc_idx = (uint16_t)(vq->vq_used_cons_idx & + ((vq->vq_nentries >> 1) - 1)); + m = (struct rte_mbuf *)vq->vq_descx[desc_idx++].cookie; + m = __rte_pktmbuf_prefree_seg(m); + if (likely(m != NULL)) { + free[0] = m; + nb_free = 1; + for (i = 1; i < VIRTIO_TX_FREE_NR; i++) { + m = (struct rte_mbuf *)vq->vq_descx[desc_idx++].cookie; + m = __rte_pktmbuf_prefree_seg(m); + if (likely(m != NULL)) { + if (likely(m->pool == free[0]->pool)) + free[nb_free++] = m; + else { + rte_mempool_put_bulk(free[0]->pool, + (void **)free, + RTE_MIN(RTE_DIM(free), + nb_free)); + free[0] = m; + nb_free = 1; + } + } + } + rte_mempool_put_bulk(free[0]->pool, (void **)free, + RTE_MIN(RTE_DIM(free), nb_free)); + } else { + for (i = 1; i < VIRTIO_TX_FREE_NR; i++) { + m = (struct rte_mbuf *)vq->vq_descx[desc_idx++].cookie; + m = __rte_pktmbuf_prefree_seg(m); + if (m != NULL) + rte_mempool_put(m->pool, m); + } + } + + vq->vq_used_cons_idx += VIRTIO_TX_FREE_NR; + vq->vq_free_cnt += (VIRTIO_TX_FREE_NR << 1); +} + +#endif /* _VIRTIO_RXTX_SIMPLE_H_ */ diff --git a/drivers/net/virtio/virtio_rxtx_simple_neon.c b/drivers/net/virtio/virtio_rxtx_simple_neon.c new file mode 100644 index 00000000..793eefbe --- /dev/null +++ b/drivers/net/virtio/virtio_rxtx_simple_neon.c @@ -0,0 +1,235 @@ +/* + * BSD LICENSE + * + * Copyright (C) Cavium networks Ltd. 2016 + * + * 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 Cavium networks 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. +*/ + +#include <stdint.h> +#include <stdio.h> +#include <stdlib.h> +#include <string.h> +#include <errno.h> + +#include <rte_byteorder.h> +#include <rte_branch_prediction.h> +#include <rte_cycles.h> +#include <rte_ether.h> +#include <rte_ethdev.h> +#include <rte_errno.h> +#include <rte_memory.h> +#include <rte_memzone.h> +#include <rte_mempool.h> +#include <rte_malloc.h> +#include <rte_mbuf.h> +#include <rte_prefetch.h> +#include <rte_string_fns.h> +#include <rte_vect.h> + +#include "virtio_rxtx_simple.h" + +#define RTE_VIRTIO_VPMD_RX_BURST 32 +#define RTE_VIRTIO_DESC_PER_LOOP 8 +#define RTE_VIRTIO_VPMD_RX_REARM_THRESH RTE_VIRTIO_VPMD_RX_BURST + +/* virtio vPMD receive routine, only accept(nb_pkts >= RTE_VIRTIO_DESC_PER_LOOP) + * + * This routine is for non-mergeable RX, one desc for each guest buffer. + * This routine is based on the RX ring layout optimization. Each entry in the + * avail ring points to the desc with the same index in the desc ring and this + * will never be changed in the driver. + * + * - nb_pkts < RTE_VIRTIO_DESC_PER_LOOP, just return no packet + */ +uint16_t +virtio_recv_pkts_vec(void *rx_queue, struct rte_mbuf **rx_pkts, + uint16_t nb_pkts) +{ + struct virtnet_rx *rxvq = rx_queue; + struct virtqueue *vq = rxvq->vq; + uint16_t nb_used; + uint16_t desc_idx; + struct vring_used_elem *rused; + struct rte_mbuf **sw_ring; + struct rte_mbuf **sw_ring_end; + uint16_t nb_pkts_received; + + uint8x16_t shuf_msk1 = { + 0xFF, 0xFF, 0xFF, 0xFF, /* packet type */ + 4, 5, 0xFF, 0xFF, /* pkt len */ + 4, 5, /* dat len */ + 0xFF, 0xFF, /* vlan tci */ + 0xFF, 0xFF, 0xFF, 0xFF + }; + + uint8x16_t shuf_msk2 = { + 0xFF, 0xFF, 0xFF, 0xFF, /* packet type */ + 12, 13, 0xFF, 0xFF, /* pkt len */ + 12, 13, /* dat len */ + 0xFF, 0xFF, /* vlan tci */ + 0xFF, 0xFF, 0xFF, 0xFF + }; + + /* Subtract the header length. + * In which case do we need the header length in used->len ? + */ + uint16x8_t len_adjust = { + 0, 0, + (uint16_t)vq->hw->vtnet_hdr_size, 0, + (uint16_t)vq->hw->vtnet_hdr_size, + 0, + 0, 0 + }; + + if (unlikely(nb_pkts < RTE_VIRTIO_DESC_PER_LOOP)) + return 0; + + nb_used = VIRTQUEUE_NUSED(vq); + + rte_rmb(); + + if (unlikely(nb_used == 0)) + return 0; + + nb_pkts = RTE_ALIGN_FLOOR(nb_pkts, RTE_VIRTIO_DESC_PER_LOOP); + nb_used = RTE_MIN(nb_used, nb_pkts); + + desc_idx = (uint16_t)(vq->vq_used_cons_idx & (vq->vq_nentries - 1)); + rused = &vq->vq_ring.used->ring[desc_idx]; + sw_ring = &vq->sw_ring[desc_idx]; + sw_ring_end = &vq->sw_ring[vq->vq_nentries]; + + rte_prefetch_non_temporal(rused); + + if (vq->vq_free_cnt >= RTE_VIRTIO_VPMD_RX_REARM_THRESH) { + virtio_rxq_rearm_vec(rxvq); + if (unlikely(virtqueue_kick_prepare(vq))) + virtqueue_notify(vq); + } + + for (nb_pkts_received = 0; + nb_pkts_received < nb_used;) { + uint64x2_t desc[RTE_VIRTIO_DESC_PER_LOOP / 2]; + uint64x2_t mbp[RTE_VIRTIO_DESC_PER_LOOP / 2]; + uint64x2_t pkt_mb[RTE_VIRTIO_DESC_PER_LOOP]; + + mbp[0] = vld1q_u64((uint64_t *)(sw_ring + 0)); + desc[0] = vld1q_u64((uint64_t *)(rused + 0)); + vst1q_u64((uint64_t *)&rx_pkts[0], mbp[0]); + + mbp[1] = vld1q_u64((uint64_t *)(sw_ring + 2)); + desc[1] = vld1q_u64((uint64_t *)(rused + 2)); + vst1q_u64((uint64_t *)&rx_pkts[2], mbp[1]); + + mbp[2] = vld1q_u64((uint64_t *)(sw_ring + 4)); + desc[2] = vld1q_u64((uint64_t *)(rused + 4)); + vst1q_u64((uint64_t *)&rx_pkts[4], mbp[2]); + + mbp[3] = vld1q_u64((uint64_t *)(sw_ring + 6)); + desc[3] = vld1q_u64((uint64_t *)(rused + 6)); + vst1q_u64((uint64_t *)&rx_pkts[6], mbp[3]); + + pkt_mb[1] = vreinterpretq_u64_u8(vqtbl1q_u8( + vreinterpretq_u8_u64(desc[0]), shuf_msk2)); + pkt_mb[0] = vreinterpretq_u64_u8(vqtbl1q_u8( + vreinterpretq_u8_u64(desc[0]), shuf_msk1)); + pkt_mb[1] = vreinterpretq_u64_u16(vsubq_u16( + vreinterpretq_u16_u64(pkt_mb[1]), len_adjust)); + pkt_mb[0] = vreinterpretq_u64_u16(vsubq_u16( + vreinterpretq_u16_u64(pkt_mb[0]), len_adjust)); + vst1q_u64((void *)&rx_pkts[1]->rx_descriptor_fields1, + pkt_mb[1]); + vst1q_u64((void *)&rx_pkts[0]->rx_descriptor_fields1, + pkt_mb[0]); + + pkt_mb[3] = vreinterpretq_u64_u8(vqtbl1q_u8( + vreinterpretq_u8_u64(desc[1]), shuf_msk2)); + pkt_mb[2] = vreinterpretq_u64_u8(vqtbl1q_u8( + vreinterpretq_u8_u64(desc[1]), shuf_msk1)); + pkt_mb[3] = vreinterpretq_u64_u16(vsubq_u16( + vreinterpretq_u16_u64(pkt_mb[3]), len_adjust)); + pkt_mb[2] = vreinterpretq_u64_u16(vsubq_u16( + vreinterpretq_u16_u64(pkt_mb[2]), len_adjust)); + vst1q_u64((void *)&rx_pkts[3]->rx_descriptor_fields1, + pkt_mb[3]); + vst1q_u64((void *)&rx_pkts[2]->rx_descriptor_fields1, + pkt_mb[2]); + + pkt_mb[5] = vreinterpretq_u64_u8(vqtbl1q_u8( + vreinterpretq_u8_u64(desc[2]), shuf_msk2)); + pkt_mb[4] = vreinterpretq_u64_u8(vqtbl1q_u8( + vreinterpretq_u8_u64(desc[2]), shuf_msk1)); + pkt_mb[5] = vreinterpretq_u64_u16(vsubq_u16( + vreinterpretq_u16_u64(pkt_mb[5]), len_adjust)); + pkt_mb[4] = vreinterpretq_u64_u16(vsubq_u16( + vreinterpretq_u16_u64(pkt_mb[4]), len_adjust)); + vst1q_u64((void *)&rx_pkts[5]->rx_descriptor_fields1, + pkt_mb[5]); + vst1q_u64((void *)&rx_pkts[4]->rx_descriptor_fields1, + pkt_mb[4]); + + pkt_mb[7] = vreinterpretq_u64_u8(vqtbl1q_u8( + vreinterpretq_u8_u64(desc[3]), shuf_msk2)); + pkt_mb[6] = vreinterpretq_u64_u8(vqtbl1q_u8( + vreinterpretq_u8_u64(desc[3]), shuf_msk1)); + pkt_mb[7] = vreinterpretq_u64_u16(vsubq_u16( + vreinterpretq_u16_u64(pkt_mb[7]), len_adjust)); + pkt_mb[6] = vreinterpretq_u64_u16(vsubq_u16( + vreinterpretq_u16_u64(pkt_mb[6]), len_adjust)); + vst1q_u64((void *)&rx_pkts[7]->rx_descriptor_fields1, + pkt_mb[7]); + vst1q_u64((void *)&rx_pkts[6]->rx_descriptor_fields1, + pkt_mb[6]); + + if (unlikely(nb_used <= RTE_VIRTIO_DESC_PER_LOOP)) { + if (sw_ring + nb_used <= sw_ring_end) + nb_pkts_received += nb_used; + else + nb_pkts_received += sw_ring_end - sw_ring; + break; + } else { + if (unlikely(sw_ring + RTE_VIRTIO_DESC_PER_LOOP >= + sw_ring_end)) { + nb_pkts_received += sw_ring_end - sw_ring; + break; + } else { + nb_pkts_received += RTE_VIRTIO_DESC_PER_LOOP; + + rx_pkts += RTE_VIRTIO_DESC_PER_LOOP; + sw_ring += RTE_VIRTIO_DESC_PER_LOOP; + rused += RTE_VIRTIO_DESC_PER_LOOP; + nb_used -= RTE_VIRTIO_DESC_PER_LOOP; + } + } + } + + vq->vq_used_cons_idx += nb_pkts_received; + vq->vq_free_cnt += nb_pkts_received; + rxvq->stats.packets += nb_pkts_received; + return nb_pkts_received; +} diff --git a/drivers/net/virtio/virtio_rxtx_simple_sse.c b/drivers/net/virtio/virtio_rxtx_simple_sse.c new file mode 100644 index 00000000..87bb5c63 --- /dev/null +++ b/drivers/net/virtio/virtio_rxtx_simple_sse.c @@ -0,0 +1,222 @@ +/*- + * BSD LICENSE + * + * Copyright(c) 2010-2015 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. + */ + +#include <stdint.h> +#include <stdio.h> +#include <stdlib.h> +#include <string.h> +#include <errno.h> + +#include <tmmintrin.h> + +#include <rte_byteorder.h> +#include <rte_branch_prediction.h> +#include <rte_cycles.h> +#include <rte_ether.h> +#include <rte_ethdev.h> +#include <rte_errno.h> +#include <rte_memory.h> +#include <rte_memzone.h> +#include <rte_mempool.h> +#include <rte_malloc.h> +#include <rte_mbuf.h> +#include <rte_prefetch.h> +#include <rte_string_fns.h> + +#include "virtio_rxtx_simple.h" + +#define RTE_VIRTIO_VPMD_RX_BURST 32 +#define RTE_VIRTIO_DESC_PER_LOOP 8 +#define RTE_VIRTIO_VPMD_RX_REARM_THRESH RTE_VIRTIO_VPMD_RX_BURST + +/* virtio vPMD receive routine, only accept(nb_pkts >= RTE_VIRTIO_DESC_PER_LOOP) + * + * This routine is for non-mergeable RX, one desc for each guest buffer. + * This routine is based on the RX ring layout optimization. Each entry in the + * avail ring points to the desc with the same index in the desc ring and this + * will never be changed in the driver. + * + * - nb_pkts < RTE_VIRTIO_DESC_PER_LOOP, just return no packet + */ +uint16_t +virtio_recv_pkts_vec(void *rx_queue, struct rte_mbuf **rx_pkts, + uint16_t nb_pkts) +{ + struct virtnet_rx *rxvq = rx_queue; + struct virtqueue *vq = rxvq->vq; + uint16_t nb_used; + uint16_t desc_idx; + struct vring_used_elem *rused; + struct rte_mbuf **sw_ring; + struct rte_mbuf **sw_ring_end; + uint16_t nb_pkts_received; + __m128i shuf_msk1, shuf_msk2, len_adjust; + + shuf_msk1 = _mm_set_epi8( + 0xFF, 0xFF, 0xFF, 0xFF, + 0xFF, 0xFF, /* vlan tci */ + 5, 4, /* dat len */ + 0xFF, 0xFF, 5, 4, /* pkt len */ + 0xFF, 0xFF, 0xFF, 0xFF /* packet type */ + + ); + + shuf_msk2 = _mm_set_epi8( + 0xFF, 0xFF, 0xFF, 0xFF, + 0xFF, 0xFF, /* vlan tci */ + 13, 12, /* dat len */ + 0xFF, 0xFF, 13, 12, /* pkt len */ + 0xFF, 0xFF, 0xFF, 0xFF /* packet type */ + ); + + /* Subtract the header length. + * In which case do we need the header length in used->len ? + */ + len_adjust = _mm_set_epi16( + 0, 0, + 0, + (uint16_t)-vq->hw->vtnet_hdr_size, + 0, (uint16_t)-vq->hw->vtnet_hdr_size, + 0, 0); + + if (unlikely(nb_pkts < RTE_VIRTIO_DESC_PER_LOOP)) + return 0; + + nb_used = VIRTQUEUE_NUSED(vq); + + rte_compiler_barrier(); + + if (unlikely(nb_used == 0)) + return 0; + + nb_pkts = RTE_ALIGN_FLOOR(nb_pkts, RTE_VIRTIO_DESC_PER_LOOP); + nb_used = RTE_MIN(nb_used, nb_pkts); + + desc_idx = (uint16_t)(vq->vq_used_cons_idx & (vq->vq_nentries - 1)); + rused = &vq->vq_ring.used->ring[desc_idx]; + sw_ring = &vq->sw_ring[desc_idx]; + sw_ring_end = &vq->sw_ring[vq->vq_nentries]; + + rte_prefetch0(rused); + + if (vq->vq_free_cnt >= RTE_VIRTIO_VPMD_RX_REARM_THRESH) { + virtio_rxq_rearm_vec(rxvq); + if (unlikely(virtqueue_kick_prepare(vq))) + virtqueue_notify(vq); + } + + for (nb_pkts_received = 0; + nb_pkts_received < nb_used;) { + __m128i desc[RTE_VIRTIO_DESC_PER_LOOP / 2]; + __m128i mbp[RTE_VIRTIO_DESC_PER_LOOP / 2]; + __m128i pkt_mb[RTE_VIRTIO_DESC_PER_LOOP]; + + mbp[0] = _mm_loadu_si128((__m128i *)(sw_ring + 0)); + desc[0] = _mm_loadu_si128((__m128i *)(rused + 0)); + _mm_storeu_si128((__m128i *)&rx_pkts[0], mbp[0]); + + mbp[1] = _mm_loadu_si128((__m128i *)(sw_ring + 2)); + desc[1] = _mm_loadu_si128((__m128i *)(rused + 2)); + _mm_storeu_si128((__m128i *)&rx_pkts[2], mbp[1]); + + mbp[2] = _mm_loadu_si128((__m128i *)(sw_ring + 4)); + desc[2] = _mm_loadu_si128((__m128i *)(rused + 4)); + _mm_storeu_si128((__m128i *)&rx_pkts[4], mbp[2]); + + mbp[3] = _mm_loadu_si128((__m128i *)(sw_ring + 6)); + desc[3] = _mm_loadu_si128((__m128i *)(rused + 6)); + _mm_storeu_si128((__m128i *)&rx_pkts[6], mbp[3]); + + pkt_mb[1] = _mm_shuffle_epi8(desc[0], shuf_msk2); + pkt_mb[0] = _mm_shuffle_epi8(desc[0], shuf_msk1); + pkt_mb[1] = _mm_add_epi16(pkt_mb[1], len_adjust); + pkt_mb[0] = _mm_add_epi16(pkt_mb[0], len_adjust); + _mm_storeu_si128((void *)&rx_pkts[1]->rx_descriptor_fields1, + pkt_mb[1]); + _mm_storeu_si128((void *)&rx_pkts[0]->rx_descriptor_fields1, + pkt_mb[0]); + + pkt_mb[3] = _mm_shuffle_epi8(desc[1], shuf_msk2); + pkt_mb[2] = _mm_shuffle_epi8(desc[1], shuf_msk1); + pkt_mb[3] = _mm_add_epi16(pkt_mb[3], len_adjust); + pkt_mb[2] = _mm_add_epi16(pkt_mb[2], len_adjust); + _mm_storeu_si128((void *)&rx_pkts[3]->rx_descriptor_fields1, + pkt_mb[3]); + _mm_storeu_si128((void *)&rx_pkts[2]->rx_descriptor_fields1, + pkt_mb[2]); + + pkt_mb[5] = _mm_shuffle_epi8(desc[2], shuf_msk2); + pkt_mb[4] = _mm_shuffle_epi8(desc[2], shuf_msk1); + pkt_mb[5] = _mm_add_epi16(pkt_mb[5], len_adjust); + pkt_mb[4] = _mm_add_epi16(pkt_mb[4], len_adjust); + _mm_storeu_si128((void *)&rx_pkts[5]->rx_descriptor_fields1, + pkt_mb[5]); + _mm_storeu_si128((void *)&rx_pkts[4]->rx_descriptor_fields1, + pkt_mb[4]); + + pkt_mb[7] = _mm_shuffle_epi8(desc[3], shuf_msk2); + pkt_mb[6] = _mm_shuffle_epi8(desc[3], shuf_msk1); + pkt_mb[7] = _mm_add_epi16(pkt_mb[7], len_adjust); + pkt_mb[6] = _mm_add_epi16(pkt_mb[6], len_adjust); + _mm_storeu_si128((void *)&rx_pkts[7]->rx_descriptor_fields1, + pkt_mb[7]); + _mm_storeu_si128((void *)&rx_pkts[6]->rx_descriptor_fields1, + pkt_mb[6]); + + if (unlikely(nb_used <= RTE_VIRTIO_DESC_PER_LOOP)) { + if (sw_ring + nb_used <= sw_ring_end) + nb_pkts_received += nb_used; + else + nb_pkts_received += sw_ring_end - sw_ring; + break; + } else { + if (unlikely(sw_ring + RTE_VIRTIO_DESC_PER_LOOP >= + sw_ring_end)) { + nb_pkts_received += sw_ring_end - sw_ring; + break; + } else { + nb_pkts_received += RTE_VIRTIO_DESC_PER_LOOP; + + rx_pkts += RTE_VIRTIO_DESC_PER_LOOP; + sw_ring += RTE_VIRTIO_DESC_PER_LOOP; + rused += RTE_VIRTIO_DESC_PER_LOOP; + nb_used -= RTE_VIRTIO_DESC_PER_LOOP; + } + } + } + + vq->vq_used_cons_idx += nb_pkts_received; + vq->vq_free_cnt += nb_pkts_received; + rxvq->stats.packets += nb_pkts_received; + return nb_pkts_received; +} diff --git a/drivers/net/virtio/virtio_user_ethdev.c b/drivers/net/virtio/virtio_user_ethdev.c index bba74028..406beeac 100644 --- a/drivers/net/virtio/virtio_user_ethdev.c +++ b/drivers/net/virtio/virtio_user_ethdev.c @@ -37,6 +37,7 @@ #include <rte_malloc.h> #include <rte_kvargs.h> +#include <rte_vdev.h> #include "virtio_ethdev.h" #include "virtio_logs.h" @@ -277,7 +278,7 @@ virtio_user_eth_dev_alloc(const char *name) struct virtio_hw *hw; struct virtio_user_dev *dev; - eth_dev = rte_eth_dev_allocate(name, RTE_ETH_DEV_VIRTUAL); + eth_dev = rte_eth_dev_allocate(name); if (!eth_dev) { PMD_INIT_LOG(ERR, "cannot alloc rte_eth_dev"); return NULL; @@ -303,6 +304,7 @@ virtio_user_eth_dev_alloc(const char *name) hw->vtpci_ops = &virtio_user_ops; hw->use_msix = 0; hw->modern = 0; + hw->use_simple_rxtx = 0; hw->virtio_user_dev = dev; data->dev_private = hw; data->numa_node = SOCKET_ID_ANY; @@ -329,7 +331,7 @@ virtio_user_eth_dev_free(struct rte_eth_dev *eth_dev) * Returns 0 on success. */ static int -virtio_user_pmd_devinit(const char *name, const char *params) +virtio_user_pmd_probe(const char *name, const char *params) { struct rte_kvargs *kvlist = NULL; struct rte_eth_dev *eth_dev; @@ -443,7 +445,7 @@ end: /** Called by rte_eth_dev_detach() */ static int -virtio_user_pmd_devuninit(const char *name) +virtio_user_pmd_remove(const char *name) { struct rte_eth_dev *eth_dev; struct virtio_hw *hw; @@ -471,14 +473,14 @@ virtio_user_pmd_devuninit(const char *name) return 0; } -static struct rte_driver virtio_user_driver = { - .type = PMD_VDEV, - .init = virtio_user_pmd_devinit, - .uninit = virtio_user_pmd_devuninit, +static struct rte_vdev_driver virtio_user_driver = { + .probe = virtio_user_pmd_probe, + .remove = virtio_user_pmd_remove, }; -PMD_REGISTER_DRIVER(virtio_user_driver, virtio_user); -DRIVER_REGISTER_PARAM_STRING(virtio_user, +RTE_PMD_REGISTER_VDEV(net_virtio_user, virtio_user_driver); +RTE_PMD_REGISTER_ALIAS(net_virtio_user, virtio_user); +RTE_PMD_REGISTER_PARAM_STRING(net_virtio_user, "path=<path> " "mac=<mac addr> " "cq=<int> " diff --git a/drivers/net/virtio/virtqueue.h b/drivers/net/virtio/virtqueue.h index 6737b81d..f0bb0899 100644 --- a/drivers/net/virtio/virtqueue.h +++ b/drivers/net/virtio/virtqueue.h @@ -44,6 +44,7 @@ #include "virtio_pci.h" #include "virtio_ring.h" #include "virtio_logs.h" +#include "virtio_rxtx.h" struct rte_mbuf; @@ -191,6 +192,12 @@ struct virtqueue { void *vq_ring_virt_mem; /**< linear address of vring*/ unsigned int vq_ring_size; + union { + struct virtnet_rx rxq; + struct virtnet_tx txq; + struct virtnet_ctl cq; + }; + phys_addr_t vq_ring_mem; /**< physical address of vring, * or virtual address for virtio_user. */ @@ -204,7 +211,6 @@ struct virtqueue { uint16_t vq_queue_index; /**< PCI queue index */ uint16_t offset; /**< relative offset to obtain addr in mbuf */ uint16_t *notify_addr; - int configured; struct rte_mbuf **sw_ring; /**< RX software ring. */ struct vq_desc_extra vq_descx[0]; }; @@ -223,6 +229,7 @@ struct virtqueue { */ struct virtio_net_hdr { #define VIRTIO_NET_HDR_F_NEEDS_CSUM 1 /**< Use csum_start,csum_offset*/ +#define VIRTIO_NET_HDR_F_DATA_VALID 2 /**< Checksum is valid */ uint8_t flags; #define VIRTIO_NET_HDR_GSO_NONE 0 /**< Not a GSO frame */ #define VIRTIO_NET_HDR_GSO_TCPV4 1 /**< GSO frame, IPv4 TCP (TSO) */ |