X-Git-Url: http://git.droids-corp.org/?a=blobdiff_plain;f=lib%2Flibrte_vhost%2Fvirtio_net.c;h=edfab3ba618e3db4dce9b0f8265f5b598d1d6fcb;hb=eaf12cccca963a664ff75a8c28454935e237886d;hp=b5d809676a52d965a6b42ab1949808680ddcb73e;hpb=c0583d98a9153549383d236e5cc5464bbbe8dd89;p=dpdk.git diff --git a/lib/librte_vhost/virtio_net.c b/lib/librte_vhost/virtio_net.c index b5d809676a..edfab3ba61 100644 --- a/lib/librte_vhost/virtio_net.c +++ b/lib/librte_vhost/virtio_net.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) 2010-2016 Intel Corporation */ #include @@ -44,11 +15,15 @@ #include #include #include +#include +#include "iotlb.h" #include "vhost.h" #define MAX_PKT_BURST 32 +#define MAX_BATCH_LEN 256 + static bool is_valid_virt_queue_idx(uint32_t idx, int is_tx, uint32_t nr_vring) { @@ -105,6 +80,31 @@ update_shadow_used_ring(struct vhost_virtqueue *vq, vq->shadow_used_ring[i].len = len; } +static inline void +do_data_copy_enqueue(struct virtio_net *dev, struct vhost_virtqueue *vq) +{ + struct batch_copy_elem *elem = vq->batch_copy_elems; + uint16_t count = vq->batch_copy_nb_elems; + int i; + + for (i = 0; i < count; i++) { + rte_memcpy(elem[i].dst, elem[i].src, elem[i].len); + vhost_log_write(dev, elem[i].log_addr, elem[i].len); + PRINT_PACKET(dev, (uintptr_t)elem[i].dst, elem[i].len, 0); + } +} + +static inline void +do_data_copy_dequeue(struct vhost_virtqueue *vq) +{ + struct batch_copy_elem *elem = vq->batch_copy_elems; + uint16_t count = vq->batch_copy_nb_elems; + int i; + + for (i = 0; i < count; i++) + rte_memcpy(elem[i].dst, elem[i].src, elem[i].len); +} + /* avoid write operation when necessary, to lessen cache issues */ #define ASSIGN_UNLESS_EQUAL(var, val) do { \ if ((var) != (val)) \ @@ -114,11 +114,16 @@ update_shadow_used_ring(struct vhost_virtqueue *vq, static void virtio_enqueue_offload(struct rte_mbuf *m_buf, struct virtio_net_hdr *net_hdr) { - if (m_buf->ol_flags & PKT_TX_L4_MASK) { + uint64_t csum_l4 = m_buf->ol_flags & PKT_TX_L4_MASK; + + if (m_buf->ol_flags & PKT_TX_TCP_SEG) + csum_l4 |= PKT_TX_TCP_CKSUM; + + if (csum_l4) { net_hdr->flags = VIRTIO_NET_HDR_F_NEEDS_CSUM; net_hdr->csum_start = m_buf->l2_len + m_buf->l3_len; - switch (m_buf->ol_flags & PKT_TX_L4_MASK) { + switch (csum_l4) { case PKT_TX_TCP_CKSUM: net_hdr->csum_offset = (offsetof(struct tcp_hdr, cksum)); @@ -138,6 +143,15 @@ virtio_enqueue_offload(struct rte_mbuf *m_buf, struct virtio_net_hdr *net_hdr) ASSIGN_UNLESS_EQUAL(net_hdr->flags, 0); } + /* IP cksum verification cannot be bypassed, then calculate here */ + if (m_buf->ol_flags & PKT_TX_IP_CKSUM) { + struct ipv4_hdr *ipv4_hdr; + + ipv4_hdr = rte_pktmbuf_mtod_offset(m_buf, struct ipv4_hdr *, + m_buf->l2_len); + ipv4_hdr->hdr_checksum = rte_ipv4_cksum(ipv4_hdr); + } + if (m_buf->ol_flags & PKT_TX_TCP_SEG) { if (m_buf->ol_flags & PKT_TX_IPV4) net_hdr->gso_type = VIRTIO_NET_HDR_GSO_TCPV4; @@ -146,6 +160,11 @@ virtio_enqueue_offload(struct rte_mbuf *m_buf, struct virtio_net_hdr *net_hdr) net_hdr->gso_size = m_buf->tso_segsz; net_hdr->hdr_len = m_buf->l2_len + m_buf->l3_len + m_buf->l4_len; + } else if (m_buf->ol_flags & PKT_TX_UDP_SEG) { + net_hdr->gso_type = VIRTIO_NET_HDR_GSO_UDP; + net_hdr->gso_size = m_buf->tso_segsz; + net_hdr->hdr_len = m_buf->l2_len + m_buf->l3_len + + m_buf->l4_len; } else { ASSIGN_UNLESS_EQUAL(net_hdr->gso_type, 0); ASSIGN_UNLESS_EQUAL(net_hdr->gso_size, 0); @@ -154,8 +173,9 @@ virtio_enqueue_offload(struct rte_mbuf *m_buf, struct virtio_net_hdr *net_hdr) } static __rte_always_inline int -copy_mbuf_to_desc(struct virtio_net *dev, struct vring_desc *descs, - struct rte_mbuf *m, uint16_t desc_idx, uint32_t size) +copy_mbuf_to_desc(struct virtio_net *dev, struct vhost_virtqueue *vq, + struct vring_desc *descs, struct rte_mbuf *m, + uint16_t desc_idx, uint32_t size) { uint32_t desc_avail, desc_offset; uint32_t mbuf_avail, mbuf_offset; @@ -164,16 +184,22 @@ copy_mbuf_to_desc(struct virtio_net *dev, struct vring_desc *descs, uint64_t desc_addr; /* A counter to avoid desc dead loop chain */ uint16_t nr_desc = 1; + struct batch_copy_elem *batch_copy = vq->batch_copy_elems; + uint16_t copy_nb = vq->batch_copy_nb_elems; + int error = 0; desc = &descs[desc_idx]; - desc_addr = rte_vhost_gpa_to_vva(dev->mem, desc->addr); + desc_addr = vhost_iova_to_vva(dev, vq, desc->addr, + desc->len, VHOST_ACCESS_RW); /* * Checking of 'desc_addr' placed outside of 'unlikely' macro to avoid * performance issue with some versions of gcc (4.8.4 and 5.3.0) which * otherwise stores offset on the stack instead of in a register. */ - if (unlikely(desc->len < dev->vhost_hlen) || !desc_addr) - return -1; + if (unlikely(desc->len < dev->vhost_hlen) || !desc_addr) { + error = -1; + goto out; + } rte_prefetch0((void *)(uintptr_t)desc_addr); @@ -199,27 +225,45 @@ copy_mbuf_to_desc(struct virtio_net *dev, struct vring_desc *descs, if (desc_avail == 0) { if ((desc->flags & VRING_DESC_F_NEXT) == 0) { /* Room in vring buffer is not enough */ - return -1; + error = -1; + goto out; + } + if (unlikely(desc->next >= size || ++nr_desc > size)) { + error = -1; + goto out; } - if (unlikely(desc->next >= size || ++nr_desc > size)) - return -1; desc = &descs[desc->next]; - desc_addr = rte_vhost_gpa_to_vva(dev->mem, desc->addr); - if (unlikely(!desc_addr)) - return -1; + desc_addr = vhost_iova_to_vva(dev, vq, desc->addr, + desc->len, + VHOST_ACCESS_RW); + if (unlikely(!desc_addr)) { + error = -1; + goto out; + } desc_offset = 0; desc_avail = desc->len; } cpy_len = RTE_MIN(desc_avail, mbuf_avail); - rte_memcpy((void *)((uintptr_t)(desc_addr + desc_offset)), - rte_pktmbuf_mtod_offset(m, void *, mbuf_offset), - cpy_len); - vhost_log_write(dev, desc->addr + desc_offset, cpy_len); - PRINT_PACKET(dev, (uintptr_t)(desc_addr + desc_offset), - cpy_len, 0); + if (likely(cpy_len > MAX_BATCH_LEN || copy_nb >= vq->size)) { + rte_memcpy((void *)((uintptr_t)(desc_addr + + desc_offset)), + rte_pktmbuf_mtod_offset(m, void *, mbuf_offset), + cpy_len); + vhost_log_write(dev, desc->addr + desc_offset, cpy_len); + PRINT_PACKET(dev, (uintptr_t)(desc_addr + desc_offset), + cpy_len, 0); + } else { + batch_copy[copy_nb].dst = + (void *)((uintptr_t)(desc_addr + desc_offset)); + batch_copy[copy_nb].src = + rte_pktmbuf_mtod_offset(m, void *, mbuf_offset); + batch_copy[copy_nb].log_addr = desc->addr + desc_offset; + batch_copy[copy_nb].len = cpy_len; + copy_nb++; + } mbuf_avail -= cpy_len; mbuf_offset += cpy_len; @@ -227,13 +271,16 @@ copy_mbuf_to_desc(struct virtio_net *dev, struct vring_desc *descs, desc_offset += cpy_len; } - return 0; +out: + vq->batch_copy_nb_elems = copy_nb; + + return error; } /** * This function adds buffers to the virtio devices RX virtqueue. Buffers can * be received from the physical port or from another virtio device. A packet - * count is returned to indicate the number of packets that are succesfully + * count is returned to indicate the number of packets that are successfully * added to the RX queue. This function works when the mbuf is scattered, but * it doesn't support the mergeable feature. */ @@ -256,8 +303,21 @@ virtio_dev_rx(struct virtio_net *dev, uint16_t queue_id, } vq = dev->virtqueue[queue_id]; + + rte_spinlock_lock(&vq->access_lock); + if (unlikely(vq->enabled == 0)) - return 0; + goto out_access_unlock; + + if (dev->features & (1ULL << VIRTIO_F_IOMMU_PLATFORM)) + vhost_user_iotlb_rd_lock(vq); + + if (unlikely(vq->access_ok == 0)) { + if (unlikely(vring_translate(dev, vq) < 0)) { + count = 0; + goto out; + } + } avail_idx = *((volatile uint16_t *)&vq->avail->idx); start_idx = vq->last_used_idx; @@ -265,11 +325,13 @@ virtio_dev_rx(struct virtio_net *dev, uint16_t queue_id, count = RTE_MIN(count, free_entries); count = RTE_MIN(count, (uint32_t)MAX_PKT_BURST); if (count == 0) - return 0; + goto out; LOG_DEBUG(VHOST_DATA, "(%d) start_idx %d | end_idx %d\n", dev->vid, start_idx, start_idx + count); + vq->batch_copy_nb_elems = 0; + /* Retrieve all of the desc indexes first to avoid caching issues. */ rte_prefetch0(&vq->avail->ring[start_idx & (vq->size - 1)]); for (i = 0; i < count; i++) { @@ -290,8 +352,10 @@ virtio_dev_rx(struct virtio_net *dev, uint16_t queue_id, if (vq->desc[desc_idx].flags & VRING_DESC_F_INDIRECT) { descs = (struct vring_desc *)(uintptr_t) - rte_vhost_gpa_to_vva(dev->mem, - vq->desc[desc_idx].addr); + vhost_iova_to_vva(dev, + vq, vq->desc[desc_idx].addr, + vq->desc[desc_idx].len, + VHOST_ACCESS_RO); if (unlikely(!descs)) { count = i; break; @@ -304,19 +368,18 @@ virtio_dev_rx(struct virtio_net *dev, uint16_t queue_id, sz = vq->size; } - err = copy_mbuf_to_desc(dev, descs, pkts[i], desc_idx, sz); + err = copy_mbuf_to_desc(dev, vq, descs, pkts[i], desc_idx, sz); if (unlikely(err)) { - used_idx = (start_idx + i) & (vq->size - 1); - vq->used->ring[used_idx].len = dev->vhost_hlen; - vhost_log_used_vring(dev, vq, - offsetof(struct vring_used, ring[used_idx]), - sizeof(vq->used->ring[used_idx])); + count = i; + break; } if (i + 1 < count) rte_prefetch0(&vq->desc[desc_indexes[i+1]]); } + do_data_copy_enqueue(dev, vq); + rte_smp_wmb(); *(volatile uint16_t *)&vq->used->idx += count; @@ -325,13 +388,14 @@ virtio_dev_rx(struct virtio_net *dev, uint16_t queue_id, offsetof(struct vring_used, idx), sizeof(vq->used->idx)); - /* flush used->idx update before we read avail->flags. */ - rte_mb(); + vhost_vring_call(dev, vq); +out: + if (dev->features & (1ULL << VIRTIO_F_IOMMU_PLATFORM)) + vhost_user_iotlb_rd_unlock(vq); + +out_access_unlock: + rte_spinlock_unlock(&vq->access_lock); - /* Kick the guest if necessary. */ - if (!(vq->avail->flags & VRING_AVAIL_F_NO_INTERRUPT) - && (vq->callfd >= 0)) - eventfd_write(vq->callfd, (eventfd_t)1); return count; } @@ -350,7 +414,9 @@ fill_vec_buf(struct virtio_net *dev, struct vhost_virtqueue *vq, if (vq->desc[idx].flags & VRING_DESC_F_INDIRECT) { descs = (struct vring_desc *)(uintptr_t) - rte_vhost_gpa_to_vva(dev->mem, vq->desc[idx].addr); + vhost_iova_to_vva(dev, vq, vq->desc[idx].addr, + vq->desc[idx].len, + VHOST_ACCESS_RO); if (unlikely(!descs)) return -1; @@ -425,8 +491,9 @@ reserve_avail_buf_mergeable(struct virtio_net *dev, struct vhost_virtqueue *vq, } static __rte_always_inline int -copy_mbuf_to_desc_mergeable(struct virtio_net *dev, struct rte_mbuf *m, - struct buf_vector *buf_vec, uint16_t num_buffers) +copy_mbuf_to_desc_mergeable(struct virtio_net *dev, struct vhost_virtqueue *vq, + struct rte_mbuf *m, struct buf_vector *buf_vec, + uint16_t num_buffers) { uint32_t vec_idx = 0; uint64_t desc_addr; @@ -435,13 +502,22 @@ copy_mbuf_to_desc_mergeable(struct virtio_net *dev, struct rte_mbuf *m, uint32_t cpy_len; uint64_t hdr_addr, hdr_phys_addr; struct rte_mbuf *hdr_mbuf; + struct batch_copy_elem *batch_copy = vq->batch_copy_elems; + uint16_t copy_nb = vq->batch_copy_nb_elems; + int error = 0; - if (unlikely(m == NULL)) - return -1; + if (unlikely(m == NULL)) { + error = -1; + goto out; + } - desc_addr = rte_vhost_gpa_to_vva(dev->mem, buf_vec[vec_idx].buf_addr); - if (buf_vec[vec_idx].buf_len < dev->vhost_hlen || !desc_addr) - return -1; + desc_addr = vhost_iova_to_vva(dev, vq, buf_vec[vec_idx].buf_addr, + buf_vec[vec_idx].buf_len, + VHOST_ACCESS_RW); + if (buf_vec[vec_idx].buf_len < dev->vhost_hlen || !desc_addr) { + error = -1; + goto out; + } hdr_mbuf = m; hdr_addr = desc_addr; @@ -460,10 +536,15 @@ copy_mbuf_to_desc_mergeable(struct virtio_net *dev, struct rte_mbuf *m, /* done with current desc buf, get the next one */ if (desc_avail == 0) { vec_idx++; - desc_addr = rte_vhost_gpa_to_vva(dev->mem, - buf_vec[vec_idx].buf_addr); - if (unlikely(!desc_addr)) - return -1; + desc_addr = + vhost_iova_to_vva(dev, vq, + buf_vec[vec_idx].buf_addr, + buf_vec[vec_idx].buf_len, + VHOST_ACCESS_RW); + if (unlikely(!desc_addr)) { + error = -1; + goto out; + } /* Prefetch buffer address. */ rte_prefetch0((void *)(uintptr_t)desc_addr); @@ -495,13 +576,27 @@ copy_mbuf_to_desc_mergeable(struct virtio_net *dev, struct rte_mbuf *m, } cpy_len = RTE_MIN(desc_avail, mbuf_avail); - rte_memcpy((void *)((uintptr_t)(desc_addr + desc_offset)), - rte_pktmbuf_mtod_offset(m, void *, mbuf_offset), - cpy_len); - vhost_log_write(dev, buf_vec[vec_idx].buf_addr + desc_offset, - cpy_len); - PRINT_PACKET(dev, (uintptr_t)(desc_addr + desc_offset), - cpy_len, 0); + + if (likely(cpy_len > MAX_BATCH_LEN || copy_nb >= vq->size)) { + rte_memcpy((void *)((uintptr_t)(desc_addr + + desc_offset)), + rte_pktmbuf_mtod_offset(m, void *, mbuf_offset), + cpy_len); + vhost_log_write(dev, + buf_vec[vec_idx].buf_addr + desc_offset, + cpy_len); + PRINT_PACKET(dev, (uintptr_t)(desc_addr + desc_offset), + cpy_len, 0); + } else { + batch_copy[copy_nb].dst = + (void *)((uintptr_t)(desc_addr + desc_offset)); + batch_copy[copy_nb].src = + rte_pktmbuf_mtod_offset(m, void *, mbuf_offset); + batch_copy[copy_nb].log_addr = + buf_vec[vec_idx].buf_addr + desc_offset; + batch_copy[copy_nb].len = cpy_len; + copy_nb++; + } mbuf_avail -= cpy_len; mbuf_offset += cpy_len; @@ -509,7 +604,10 @@ copy_mbuf_to_desc_mergeable(struct virtio_net *dev, struct rte_mbuf *m, desc_offset += cpy_len; } - return 0; +out: + vq->batch_copy_nb_elems = copy_nb; + + return error; } static __rte_always_inline uint32_t @@ -530,12 +628,24 @@ virtio_dev_merge_rx(struct virtio_net *dev, uint16_t queue_id, } vq = dev->virtqueue[queue_id]; + + rte_spinlock_lock(&vq->access_lock); + if (unlikely(vq->enabled == 0)) - return 0; + goto out_access_unlock; + + if (dev->features & (1ULL << VIRTIO_F_IOMMU_PLATFORM)) + vhost_user_iotlb_rd_lock(vq); + + if (unlikely(vq->access_ok == 0)) + if (unlikely(vring_translate(dev, vq) < 0)) + goto out; count = RTE_MIN((uint32_t)MAX_PKT_BURST, count); if (count == 0) - return 0; + goto out; + + vq->batch_copy_nb_elems = 0; rte_prefetch0(&vq->avail->ring[vq->last_avail_idx & (vq->size - 1)]); @@ -558,7 +668,7 @@ virtio_dev_merge_rx(struct virtio_net *dev, uint16_t queue_id, dev->vid, vq->last_avail_idx, vq->last_avail_idx + num_buffers); - if (copy_mbuf_to_desc_mergeable(dev, pkts[pkt_idx], + if (copy_mbuf_to_desc_mergeable(dev, vq, pkts[pkt_idx], buf_vec, num_buffers) < 0) { vq->shadow_used_idx -= num_buffers; break; @@ -567,17 +677,19 @@ virtio_dev_merge_rx(struct virtio_net *dev, uint16_t queue_id, vq->last_avail_idx += num_buffers; } + do_data_copy_enqueue(dev, vq); + if (likely(vq->shadow_used_idx)) { flush_shadow_used_ring(dev, vq); + vhost_vring_call(dev, vq); + } - /* flush used->idx update before we read avail->flags. */ - rte_mb(); +out: + if (dev->features & (1ULL << VIRTIO_F_IOMMU_PLATFORM)) + vhost_user_iotlb_rd_unlock(vq); - /* Kick the guest if necessary. */ - if (!(vq->avail->flags & VRING_AVAIL_F_NO_INTERRUPT) - && (vq->callfd >= 0)) - eventfd_write(vq->callfd, (eventfd_t)1); - } +out_access_unlock: + rte_spinlock_unlock(&vq->access_lock); return pkt_idx; } @@ -601,9 +713,11 @@ static inline bool virtio_net_with_host_offload(struct virtio_net *dev) { if (dev->features & - (VIRTIO_NET_F_CSUM | VIRTIO_NET_F_HOST_ECN | - VIRTIO_NET_F_HOST_TSO4 | VIRTIO_NET_F_HOST_TSO6 | - VIRTIO_NET_F_HOST_UFO)) + ((1ULL << VIRTIO_NET_F_CSUM) | + (1ULL << VIRTIO_NET_F_HOST_ECN) | + (1ULL << VIRTIO_NET_F_HOST_TSO4) | + (1ULL << VIRTIO_NET_F_HOST_TSO6) | + (1ULL << VIRTIO_NET_F_HOST_UFO))) return true; return false; @@ -696,6 +810,11 @@ vhost_dequeue_offload(struct virtio_net_hdr *hdr, struct rte_mbuf *m) m->tso_segsz = hdr->gso_size; m->l4_len = (tcp_hdr->data_off & 0xf0) >> 2; break; + case VIRTIO_NET_HDR_GSO_UDP: + m->ol_flags |= PKT_TX_UDP_SEG; + m->tso_segsz = hdr->gso_size; + m->l4_len = sizeof(struct udp_hdr); + break; default: RTE_LOG(WARNING, VHOST_DATA, "unsupported gso type %u.\n", hdr->gso_type); @@ -704,45 +823,6 @@ vhost_dequeue_offload(struct virtio_net_hdr *hdr, struct rte_mbuf *m) } } -#define RARP_PKT_SIZE 64 - -static int -make_rarp_packet(struct rte_mbuf *rarp_mbuf, const struct ether_addr *mac) -{ - struct ether_hdr *eth_hdr; - struct arp_hdr *rarp; - - if (rarp_mbuf->buf_len < 64) { - RTE_LOG(WARNING, VHOST_DATA, - "failed to make RARP; mbuf size too small %u (< %d)\n", - rarp_mbuf->buf_len, RARP_PKT_SIZE); - return -1; - } - - /* Ethernet header. */ - eth_hdr = rte_pktmbuf_mtod_offset(rarp_mbuf, struct ether_hdr *, 0); - memset(eth_hdr->d_addr.addr_bytes, 0xff, ETHER_ADDR_LEN); - ether_addr_copy(mac, ð_hdr->s_addr); - eth_hdr->ether_type = htons(ETHER_TYPE_RARP); - - /* RARP header. */ - rarp = (struct arp_hdr *)(eth_hdr + 1); - rarp->arp_hrd = htons(ARP_HRD_ETHER); - rarp->arp_pro = htons(ETHER_TYPE_IPv4); - rarp->arp_hln = ETHER_ADDR_LEN; - rarp->arp_pln = 4; - rarp->arp_op = htons(ARP_OP_REVREQUEST); - - ether_addr_copy(mac, &rarp->arp_data.arp_sha); - ether_addr_copy(mac, &rarp->arp_data.arp_tha); - memset(&rarp->arp_data.arp_sip, 0x00, 4); - memset(&rarp->arp_data.arp_tip, 0x00, 4); - - rarp_mbuf->pkt_len = rarp_mbuf->data_len = RARP_PKT_SIZE; - - return 0; -} - static __rte_always_inline void put_zmbuf(struct zcopy_mbuf *zmbuf) { @@ -750,8 +830,9 @@ put_zmbuf(struct zcopy_mbuf *zmbuf) } static __rte_always_inline int -copy_desc_to_mbuf(struct virtio_net *dev, struct vring_desc *descs, - uint16_t max_desc, struct rte_mbuf *m, uint16_t desc_idx, +copy_desc_to_mbuf(struct virtio_net *dev, struct vhost_virtqueue *vq, + struct vring_desc *descs, uint16_t max_desc, + struct rte_mbuf *m, uint16_t desc_idx, struct rte_mempool *mbuf_pool) { struct vring_desc *desc; @@ -763,15 +844,25 @@ copy_desc_to_mbuf(struct virtio_net *dev, struct vring_desc *descs, struct virtio_net_hdr *hdr = NULL; /* A counter to avoid desc dead loop chain */ uint32_t nr_desc = 1; + struct batch_copy_elem *batch_copy = vq->batch_copy_elems; + uint16_t copy_nb = vq->batch_copy_nb_elems; + int error = 0; desc = &descs[desc_idx]; if (unlikely((desc->len < dev->vhost_hlen)) || - (desc->flags & VRING_DESC_F_INDIRECT)) - return -1; + (desc->flags & VRING_DESC_F_INDIRECT)) { + error = -1; + goto out; + } - desc_addr = rte_vhost_gpa_to_vva(dev->mem, desc->addr); - if (unlikely(!desc_addr)) - return -1; + desc_addr = vhost_iova_to_vva(dev, + vq, desc->addr, + desc->len, + VHOST_ACCESS_RO); + if (unlikely(!desc_addr)) { + error = -1; + goto out; + } if (virtio_net_with_host_offload(dev)) { hdr = (struct virtio_net_hdr *)((uintptr_t)desc_addr); @@ -786,12 +877,19 @@ copy_desc_to_mbuf(struct virtio_net *dev, struct vring_desc *descs, if (likely((desc->len == dev->vhost_hlen) && (desc->flags & VRING_DESC_F_NEXT) != 0)) { desc = &descs[desc->next]; - if (unlikely(desc->flags & VRING_DESC_F_INDIRECT)) - return -1; + if (unlikely(desc->flags & VRING_DESC_F_INDIRECT)) { + error = -1; + goto out; + } - desc_addr = rte_vhost_gpa_to_vva(dev->mem, desc->addr); - if (unlikely(!desc_addr)) - return -1; + desc_addr = vhost_iova_to_vva(dev, + vq, desc->addr, + desc->len, + VHOST_ACCESS_RO); + if (unlikely(!desc_addr)) { + error = -1; + goto out; + } desc_offset = 0; desc_avail = desc->len; @@ -821,8 +919,9 @@ copy_desc_to_mbuf(struct virtio_net *dev, struct vring_desc *descs, desc->addr + desc_offset, cpy_len)))) { cur->data_len = cpy_len; cur->data_off = 0; - cur->buf_addr = (void *)(uintptr_t)desc_addr; - cur->buf_physaddr = hpa; + cur->buf_addr = (void *)(uintptr_t)(desc_addr + + desc_offset); + cur->buf_iova = hpa; /* * In zero copy mode, one mbuf can only reference data @@ -830,10 +929,24 @@ copy_desc_to_mbuf(struct virtio_net *dev, struct vring_desc *descs, */ mbuf_avail = cpy_len; } else { - rte_memcpy(rte_pktmbuf_mtod_offset(cur, void *, - mbuf_offset), - (void *)((uintptr_t)(desc_addr + desc_offset)), - cpy_len); + if (likely(cpy_len > MAX_BATCH_LEN || + copy_nb >= vq->size || + (hdr && cur == m))) { + rte_memcpy(rte_pktmbuf_mtod_offset(cur, void *, + mbuf_offset), + (void *)((uintptr_t)(desc_addr + + desc_offset)), + cpy_len); + } else { + batch_copy[copy_nb].dst = + rte_pktmbuf_mtod_offset(cur, void *, + mbuf_offset); + batch_copy[copy_nb].src = + (void *)((uintptr_t)(desc_addr + + desc_offset)); + batch_copy[copy_nb].len = cpy_len; + copy_nb++; + } } mbuf_avail -= cpy_len; @@ -847,15 +960,24 @@ copy_desc_to_mbuf(struct virtio_net *dev, struct vring_desc *descs, break; if (unlikely(desc->next >= max_desc || - ++nr_desc > max_desc)) - return -1; + ++nr_desc > max_desc)) { + error = -1; + goto out; + } desc = &descs[desc->next]; - if (unlikely(desc->flags & VRING_DESC_F_INDIRECT)) - return -1; + if (unlikely(desc->flags & VRING_DESC_F_INDIRECT)) { + error = -1; + goto out; + } - desc_addr = rte_vhost_gpa_to_vva(dev->mem, desc->addr); - if (unlikely(!desc_addr)) - return -1; + desc_addr = vhost_iova_to_vva(dev, + vq, desc->addr, + desc->len, + VHOST_ACCESS_RO); + if (unlikely(!desc_addr)) { + error = -1; + goto out; + } rte_prefetch0((void *)(uintptr_t)desc_addr); @@ -874,7 +996,8 @@ copy_desc_to_mbuf(struct virtio_net *dev, struct vring_desc *descs, if (unlikely(cur == NULL)) { RTE_LOG(ERR, VHOST_DATA, "Failed to " "allocate memory for mbuf.\n"); - return -1; + error = -1; + goto out; } if (unlikely(dev->dequeue_zero_copy)) rte_mbuf_refcnt_update(cur, 1); @@ -896,7 +1019,10 @@ copy_desc_to_mbuf(struct virtio_net *dev, struct vring_desc *descs, if (hdr) vhost_dequeue_offload(hdr, m); - return 0; +out: + vq->batch_copy_nb_elems = copy_nb; + + return error; } static __rte_always_inline void @@ -923,11 +1049,7 @@ update_used_idx(struct virtio_net *dev, struct vhost_virtqueue *vq, vq->used->idx += count; vhost_log_used_vring(dev, vq, offsetof(struct vring_used, idx), sizeof(vq->used->idx)); - - /* Kick guest if required. */ - if (!(vq->avail->flags & VRING_AVAIL_F_NO_INTERRUPT) - && (vq->callfd >= 0)) - eventfd_write(vq->callfd, (eventfd_t)1); + vhost_vring_call(dev, vq); } static __rte_always_inline struct zcopy_mbuf * @@ -973,6 +1095,22 @@ mbuf_is_consumed(struct rte_mbuf *m) return true; } +static __rte_always_inline void +restore_mbuf(struct rte_mbuf *m) +{ + uint32_t mbuf_size, priv_size; + + while (m) { + priv_size = rte_pktmbuf_priv_size(m->pool); + mbuf_size = sizeof(struct rte_mbuf) + priv_size; + /* start of buffer is after mbuf structure and priv data */ + + m->buf_addr = (char *)m + mbuf_size; + m->buf_iova = rte_mempool_virt2iova(m) + mbuf_size; + m = m->next; + } +} + uint16_t rte_vhost_dequeue_burst(int vid, uint16_t queue_id, struct rte_mempool *mbuf_pool, struct rte_mbuf **pkts, uint16_t count) @@ -997,9 +1135,22 @@ rte_vhost_dequeue_burst(int vid, uint16_t queue_id, } vq = dev->virtqueue[queue_id]; - if (unlikely(vq->enabled == 0)) + + if (unlikely(rte_spinlock_trylock(&vq->access_lock) == 0)) return 0; + if (unlikely(vq->enabled == 0)) + goto out_access_unlock; + + vq->batch_copy_nb_elems = 0; + + if (dev->features & (1ULL << VIRTIO_F_IOMMU_PLATFORM)) + vhost_user_iotlb_rd_lock(vq); + + if (unlikely(vq->access_ok == 0)) + if (unlikely(vring_translate(dev, vq) < 0)) + goto out; + if (unlikely(dev->dequeue_zero_copy)) { struct zcopy_mbuf *zmbuf, *next; int nr_updated = 0; @@ -1015,6 +1166,7 @@ rte_vhost_dequeue_burst(int vid, uint16_t queue_id, nr_updated += 1; TAILQ_REMOVE(&vq->zmbuf_list, zmbuf, next); + restore_mbuf(zmbuf->mbuf); rte_pktmbuf_free(zmbuf->mbuf); put_zmbuf(zmbuf); vq->nr_zmbuf -= 1; @@ -1044,19 +1196,13 @@ rte_vhost_dequeue_burst(int vid, uint16_t queue_id, rte_atomic16_cmpset((volatile uint16_t *) &dev->broadcast_rarp.cnt, 1, 0))) { - rarp_mbuf = rte_pktmbuf_alloc(mbuf_pool); + rarp_mbuf = rte_net_make_rarp_packet(mbuf_pool, &dev->mac); if (rarp_mbuf == NULL) { RTE_LOG(ERR, VHOST_DATA, - "Failed to allocate memory for mbuf.\n"); + "Failed to make RARP packet.\n"); return 0; } - - if (make_rarp_packet(rarp_mbuf, &dev->mac)) { - rte_pktmbuf_free(rarp_mbuf); - rarp_mbuf = NULL; - } else { - count -= 1; - } + count -= 1; } free_entries = *((volatile uint16_t *)&vq->avail->idx) - @@ -1099,8 +1245,10 @@ rte_vhost_dequeue_burst(int vid, uint16_t queue_id, if (vq->desc[desc_indexes[i]].flags & VRING_DESC_F_INDIRECT) { desc = (struct vring_desc *)(uintptr_t) - rte_vhost_gpa_to_vva(dev->mem, - vq->desc[desc_indexes[i]].addr); + vhost_iova_to_vva(dev, vq, + vq->desc[desc_indexes[i]].addr, + sizeof(*desc), + VHOST_ACCESS_RO); if (unlikely(!desc)) break; @@ -1120,7 +1268,8 @@ rte_vhost_dequeue_burst(int vid, uint16_t queue_id, break; } - err = copy_desc_to_mbuf(dev, desc, sz, pkts[i], idx, mbuf_pool); + err = copy_desc_to_mbuf(dev, vq, desc, sz, pkts[i], idx, + mbuf_pool); if (unlikely(err)) { rte_pktmbuf_free(pkts[i]); break; @@ -1152,11 +1301,18 @@ rte_vhost_dequeue_burst(int vid, uint16_t queue_id, vq->last_avail_idx += i; if (likely(dev->dequeue_zero_copy == 0)) { + do_data_copy_dequeue(vq); vq->last_used_idx += i; update_used_idx(dev, vq, i); } out: + if (dev->features & (1ULL << VIRTIO_F_IOMMU_PLATFORM)) + vhost_user_iotlb_rd_unlock(vq); + +out_access_unlock: + rte_spinlock_unlock(&vq->access_lock); + if (unlikely(rarp_mbuf != NULL)) { /* * Inject it to the head of "pkts" array, so that switch's mac