diff options
Diffstat (limited to '')
-rw-r--r-- | net/xdp/xdp_umem.c | 8 | ||||
-rw-r--r-- | net/xdp/xsk.c | 143 | ||||
-rw-r--r-- | net/xdp/xsk_buff_pool.c | 46 | ||||
-rw-r--r-- | net/xdp/xsk_queue.h | 47 | ||||
-rw-r--r-- | net/xdp/xskmap.c | 7 |
5 files changed, 152 insertions, 99 deletions
diff --git a/net/xdp/xdp_umem.c b/net/xdp/xdp_umem.c index f01ef6bda390..4681e8e8ad94 100644 --- a/net/xdp/xdp_umem.c +++ b/net/xdp/xdp_umem.c @@ -19,8 +19,6 @@ #include "xdp_umem.h" #include "xsk_queue.h" -#define XDP_UMEM_MIN_CHUNK_SIZE 2048 - static DEFINE_IDA(umem_ida); static void xdp_umem_unpin_pages(struct xdp_umem *umem) @@ -57,7 +55,7 @@ static int xdp_umem_addr_map(struct xdp_umem *umem, struct page **pages, static void xdp_umem_release(struct xdp_umem *umem) { umem->zc = false; - ida_simple_remove(&umem_ida, umem->id); + ida_free(&umem_ida, umem->id); xdp_umem_addr_unmap(umem); xdp_umem_unpin_pages(umem); @@ -242,7 +240,7 @@ struct xdp_umem *xdp_umem_create(struct xdp_umem_reg *mr) if (!umem) return ERR_PTR(-ENOMEM); - err = ida_simple_get(&umem_ida, 0, 0, GFP_KERNEL); + err = ida_alloc(&umem_ida, GFP_KERNEL); if (err < 0) { kfree(umem); return ERR_PTR(err); @@ -251,7 +249,7 @@ struct xdp_umem *xdp_umem_create(struct xdp_umem_reg *mr) err = xdp_umem_reg(umem, mr); if (err) { - ida_simple_remove(&umem_ida, umem->id); + ida_free(&umem_ida, umem->id); kfree(umem); return ERR_PTR(err); } diff --git a/net/xdp/xsk.c b/net/xdp/xsk.c index f16074eb53c7..9f0561b67c12 100644 --- a/net/xdp/xsk.c +++ b/net/xdp/xsk.c @@ -184,7 +184,7 @@ static int __xsk_rcv(struct xdp_sock *xs, struct xdp_buff *xdp) xsk_xdp = xsk_buff_alloc(xs->pool); if (!xsk_xdp) { xs->rx_dropped++; - return -ENOSPC; + return -ENOMEM; } xsk_copy_xdp(xsk_xdp, xdp, len); @@ -217,7 +217,7 @@ static bool xsk_is_bound(struct xdp_sock *xs) static int xsk_rcv_check(struct xdp_sock *xs, struct xdp_buff *xdp) { if (!xsk_is_bound(xs)) - return -EINVAL; + return -ENXIO; if (xs->dev != xdp->rxq->dev || xs->queue_id != xdp->rxq->queue_index) return -EINVAL; @@ -343,9 +343,9 @@ out: } EXPORT_SYMBOL(xsk_tx_peek_desc); -static u32 xsk_tx_peek_release_fallback(struct xsk_buff_pool *pool, struct xdp_desc *descs, - u32 max_entries) +static u32 xsk_tx_peek_release_fallback(struct xsk_buff_pool *pool, u32 max_entries) { + struct xdp_desc *descs = pool->tx_descs; u32 nb_pkts = 0; while (nb_pkts < max_entries && xsk_tx_peek_desc(pool, &descs[nb_pkts])) @@ -355,17 +355,15 @@ static u32 xsk_tx_peek_release_fallback(struct xsk_buff_pool *pool, struct xdp_d return nb_pkts; } -u32 xsk_tx_peek_release_desc_batch(struct xsk_buff_pool *pool, struct xdp_desc *descs, - u32 max_entries) +u32 xsk_tx_peek_release_desc_batch(struct xsk_buff_pool *pool, u32 nb_pkts) { struct xdp_sock *xs; - u32 nb_pkts; rcu_read_lock(); if (!list_is_singular(&pool->xsk_tx_list)) { /* Fallback to the non-batched version */ rcu_read_unlock(); - return xsk_tx_peek_release_fallback(pool, descs, max_entries); + return xsk_tx_peek_release_fallback(pool, nb_pkts); } xs = list_first_or_null_rcu(&pool->xsk_tx_list, struct xdp_sock, tx_list); @@ -374,11 +372,7 @@ u32 xsk_tx_peek_release_desc_batch(struct xsk_buff_pool *pool, struct xdp_desc * goto out; } - nb_pkts = xskq_cons_peek_desc_batch(xs->tx, descs, pool, max_entries); - if (!nb_pkts) { - xs->tx->queue_empty_descs++; - goto out; - } + nb_pkts = xskq_cons_nb_entries(xs->tx, nb_pkts); /* This is the backpressure mechanism for the Tx path. Try to * reserve space in the completion queue for all packets, but @@ -386,12 +380,18 @@ u32 xsk_tx_peek_release_desc_batch(struct xsk_buff_pool *pool, struct xdp_desc * * packets. This avoids having to implement any buffering in * the Tx path. */ - nb_pkts = xskq_prod_reserve_addr_batch(pool->cq, descs, nb_pkts); + nb_pkts = xskq_prod_nb_free(pool->cq, nb_pkts); if (!nb_pkts) goto out; - xskq_cons_release_n(xs->tx, nb_pkts); + nb_pkts = xskq_cons_read_desc_batch(xs->tx, pool, nb_pkts); + if (!nb_pkts) { + xs->tx->queue_empty_descs++; + goto out; + } + __xskq_cons_release(xs->tx); + xskq_prod_write_addr_batch(pool->cq, pool->tx_descs, nb_pkts); xs->sk.sk_write_space(&xs->sk); out: @@ -403,18 +403,8 @@ EXPORT_SYMBOL(xsk_tx_peek_release_desc_batch); static int xsk_wakeup(struct xdp_sock *xs, u8 flags) { struct net_device *dev = xs->dev; - int err; - rcu_read_lock(); - err = dev->netdev_ops->ndo_xsk_wakeup(dev, xs->queue_id, flags); - rcu_read_unlock(); - - return err; -} - -static int xsk_zc_xmit(struct xdp_sock *xs) -{ - return xsk_wakeup(xs, XDP_WAKEUP_TX); + return dev->netdev_ops->ndo_xsk_wakeup(dev, xs->queue_id, flags); } static void xsk_destruct_skb(struct sk_buff *skb) @@ -533,6 +523,12 @@ static int xsk_generic_xmit(struct sock *sk) mutex_lock(&xs->mutex); + /* Since we dropped the RCU read lock, the socket state might have changed. */ + if (unlikely(!xsk_is_bound(xs))) { + err = -ENXIO; + goto out; + } + if (xs->queue_id >= xs->dev->real_num_tx_queues) goto out; @@ -542,12 +538,6 @@ static int xsk_generic_xmit(struct sock *sk) goto out; } - skb = xsk_build_skb(xs, &desc); - if (IS_ERR(skb)) { - err = PTR_ERR(skb); - goto out; - } - /* This is the backpressure mechanism for the Tx path. * Reserve space in the completion queue and only proceed * if there is space in it. This avoids having to implement @@ -556,11 +546,19 @@ static int xsk_generic_xmit(struct sock *sk) spin_lock_irqsave(&xs->pool->cq_lock, flags); if (xskq_prod_reserve(xs->pool->cq)) { spin_unlock_irqrestore(&xs->pool->cq_lock, flags); - kfree_skb(skb); goto out; } spin_unlock_irqrestore(&xs->pool->cq_lock, flags); + skb = xsk_build_skb(xs, &desc); + if (IS_ERR(skb)) { + err = PTR_ERR(skb); + spin_lock_irqsave(&xs->pool->cq_lock, flags); + xskq_prod_cancel(xs->pool->cq); + spin_unlock_irqrestore(&xs->pool->cq_lock, flags); + goto out; + } + err = __dev_direct_xmit(skb, xs->queue_id); if (err == NETDEV_TX_BUSY) { /* Tell user-space to retry the send */ @@ -596,16 +594,26 @@ out: return err; } -static int __xsk_sendmsg(struct sock *sk) +static int xsk_xmit(struct sock *sk) { struct xdp_sock *xs = xdp_sk(sk); + int ret; if (unlikely(!(xs->dev->flags & IFF_UP))) return -ENETDOWN; if (unlikely(!xs->tx)) return -ENOBUFS; - return xs->zc ? xsk_zc_xmit(xs) : xsk_generic_xmit(sk); + if (xs->zc) + return xsk_wakeup(xs, XDP_WAKEUP_TX); + + /* Drop the RCU lock since the SKB path might sleep. */ + rcu_read_unlock(); + ret = xsk_generic_xmit(sk); + /* Reaquire RCU lock before going into common code. */ + rcu_read_lock(); + + return ret; } static bool xsk_no_wakeup(struct sock *sk) @@ -619,7 +627,7 @@ static bool xsk_no_wakeup(struct sock *sk) #endif } -static int xsk_sendmsg(struct socket *sock, struct msghdr *m, size_t total_len) +static int __xsk_sendmsg(struct socket *sock, struct msghdr *m, size_t total_len) { bool need_wait = !(m->msg_flags & MSG_DONTWAIT); struct sock *sk = sock->sk; @@ -631,19 +639,33 @@ static int xsk_sendmsg(struct socket *sock, struct msghdr *m, size_t total_len) if (unlikely(need_wait)) return -EOPNOTSUPP; - if (sk_can_busy_loop(sk)) + if (sk_can_busy_loop(sk)) { + if (xs->zc) + __sk_mark_napi_id_once(sk, xsk_pool_get_napi_id(xs->pool)); sk_busy_loop(sk, 1); /* only support non-blocking sockets */ + } - if (xsk_no_wakeup(sk)) + if (xs->zc && xsk_no_wakeup(sk)) return 0; pool = xs->pool; if (pool->cached_need_wakeup & XDP_WAKEUP_TX) - return __xsk_sendmsg(sk); + return xsk_xmit(sk); return 0; } -static int xsk_recvmsg(struct socket *sock, struct msghdr *m, size_t len, int flags) +static int xsk_sendmsg(struct socket *sock, struct msghdr *m, size_t total_len) +{ + int ret; + + rcu_read_lock(); + ret = __xsk_sendmsg(sock, m, total_len); + rcu_read_unlock(); + + return ret; +} + +static int __xsk_recvmsg(struct socket *sock, struct msghdr *m, size_t len, int flags) { bool need_wait = !(flags & MSG_DONTWAIT); struct sock *sk = sock->sk; @@ -669,6 +691,17 @@ static int xsk_recvmsg(struct socket *sock, struct msghdr *m, size_t len, int fl return 0; } +static int xsk_recvmsg(struct socket *sock, struct msghdr *m, size_t len, int flags) +{ + int ret; + + rcu_read_lock(); + ret = __xsk_recvmsg(sock, m, len, flags); + rcu_read_unlock(); + + return ret; +} + static __poll_t xsk_poll(struct file *file, struct socket *sock, struct poll_table_struct *wait) { @@ -679,8 +712,11 @@ static __poll_t xsk_poll(struct file *file, struct socket *sock, sock_poll_wait(file, sock, wait); - if (unlikely(!xsk_is_bound(xs))) + rcu_read_lock(); + if (unlikely(!xsk_is_bound(xs))) { + rcu_read_unlock(); return mask; + } pool = xs->pool; @@ -689,7 +725,7 @@ static __poll_t xsk_poll(struct file *file, struct socket *sock, xsk_wakeup(xs, pool->cached_need_wakeup); else /* Poll needs to drive Tx also in copy mode */ - __xsk_sendmsg(sk); + xsk_xmit(sk); } if (xs->rx && !xskq_prod_is_empty(xs->rx)) @@ -697,6 +733,7 @@ static __poll_t xsk_poll(struct file *file, struct socket *sock, if (xs->tx && xsk_tx_writeable(xs)) mask |= EPOLLOUT | EPOLLWRNORM; + rcu_read_unlock(); return mask; } @@ -728,7 +765,6 @@ static void xsk_unbind_dev(struct xdp_sock *xs) /* Wait for driver to stop using the xdp socket. */ xp_del_xsk(xs->pool, xs); - xs->dev = NULL; synchronize_net(); dev_put(dev); } @@ -794,9 +830,7 @@ static int xsk_release(struct socket *sock) sk_del_node_init_rcu(sk); mutex_unlock(&net->xdp.lock); - local_bh_disable(); sock_prot_inuse_add(net, sk->sk_prot, -1); - local_bh_enable(); xsk_delete_from_maps(xs); mutex_lock(&xs->mutex); @@ -920,8 +954,8 @@ static int xsk_bind(struct socket *sock, struct sockaddr *addr, int addr_len) goto out_unlock; } - err = xp_assign_dev_shared(xs->pool, umem_xs->umem, - dev, qid); + err = xp_assign_dev_shared(xs->pool, umem_xs, dev, + qid); if (err) { xp_destroy(xs->pool); xs->pool = NULL; @@ -939,6 +973,19 @@ static int xsk_bind(struct socket *sock, struct sockaddr *addr, int addr_len) xp_get_pool(umem_xs->pool); xs->pool = umem_xs->pool; + + /* If underlying shared umem was created without Tx + * ring, allocate Tx descs array that Tx batching API + * utilizes + */ + if (xs->tx && !xs->pool->tx_descs) { + err = xp_alloc_tx_descs(xs->pool, xs); + if (err) { + xp_put_pool(xs->pool); + sockfd_put(sock); + goto out_unlock; + } + } } xdp_get_umem(umem_xs->umem); @@ -1396,9 +1443,7 @@ static int xsk_create(struct net *net, struct socket *sock, int protocol, sk_add_node_rcu(sk, &net->xdp.list); mutex_unlock(&net->xdp.lock); - local_bh_disable(); sock_prot_inuse_add(net, &xsk_proto, 1); - local_bh_enable(); return 0; } diff --git a/net/xdp/xsk_buff_pool.c b/net/xdp/xsk_buff_pool.c index bc4ad48ea4f0..ed6c71826d31 100644 --- a/net/xdp/xsk_buff_pool.c +++ b/net/xdp/xsk_buff_pool.c @@ -37,10 +37,21 @@ void xp_destroy(struct xsk_buff_pool *pool) if (!pool) return; + kvfree(pool->tx_descs); kvfree(pool->heads); kvfree(pool); } +int xp_alloc_tx_descs(struct xsk_buff_pool *pool, struct xdp_sock *xs) +{ + pool->tx_descs = kvcalloc(xs->tx->nentries, sizeof(*pool->tx_descs), + GFP_KERNEL); + if (!pool->tx_descs) + return -ENOMEM; + + return 0; +} + struct xsk_buff_pool *xp_create_and_assign_umem(struct xdp_sock *xs, struct xdp_umem *umem) { @@ -58,6 +69,10 @@ struct xsk_buff_pool *xp_create_and_assign_umem(struct xdp_sock *xs, if (!pool->heads) goto out; + if (xs->tx) + if (xp_alloc_tx_descs(pool, xs)) + goto out; + pool->chunk_mask = ~((u64)umem->chunk_size - 1); pool->addrs_cnt = umem->size; pool->heads_cnt = umem->chunks; @@ -83,6 +98,7 @@ struct xsk_buff_pool *xp_create_and_assign_umem(struct xdp_sock *xs, xskb = &pool->heads[i]; xskb->pool = pool; xskb->xdp.frame_sz = umem->chunk_size - umem->headroom; + INIT_LIST_HEAD(&xskb->free_list_node); if (pool->unaligned) pool->free_heads[i] = xskb; else @@ -196,17 +212,18 @@ err_unreg_pool: return err; } -int xp_assign_dev_shared(struct xsk_buff_pool *pool, struct xdp_umem *umem, +int xp_assign_dev_shared(struct xsk_buff_pool *pool, struct xdp_sock *umem_xs, struct net_device *dev, u16 queue_id) { u16 flags; + struct xdp_umem *umem = umem_xs->umem; /* One fill and completion ring required for each queue id. */ if (!pool->fq || !pool->cq) return -EINVAL; flags = umem->zc ? XDP_ZEROCOPY : XDP_COPY; - if (pool->uses_need_wakeup) + if (umem_xs->pool->uses_need_wakeup) flags |= XDP_USE_NEED_WAKEUP; return xp_assign_dev(pool, dev, queue_id, flags); @@ -316,6 +333,7 @@ static void __xp_dma_unmap(struct xsk_dma_map *dma_map, unsigned long attrs) for (i = 0; i < dma_map->dma_pages_cnt; i++) { dma = &dma_map->dma_pages[i]; if (*dma) { + *dma &= ~XSK_NEXT_PG_CONTIG_MASK; dma_unmap_page_attrs(dma_map->dev, *dma, PAGE_SIZE, DMA_BIDIRECTIONAL, attrs); *dma = 0; @@ -362,6 +380,16 @@ static void xp_check_dma_contiguity(struct xsk_dma_map *dma_map) static int xp_init_dma_info(struct xsk_buff_pool *pool, struct xsk_dma_map *dma_map) { + if (!pool->unaligned) { + u32 i; + + for (i = 0; i < pool->heads_cnt; i++) { + struct xdp_buff_xsk *xskb = &pool->heads[i]; + + xp_init_xskb_dma(xskb, pool, dma_map->dma_pages, xskb->orig_addr); + } + } + pool->dma_pages = kvcalloc(dma_map->dma_pages_cnt, sizeof(*pool->dma_pages), GFP_KERNEL); if (!pool->dma_pages) return -ENOMEM; @@ -411,12 +439,6 @@ int xp_dma_map(struct xsk_buff_pool *pool, struct device *dev, if (pool->unaligned) xp_check_dma_contiguity(dma_map); - else - for (i = 0; i < pool->heads_cnt; i++) { - struct xdp_buff_xsk *xskb = &pool->heads[i]; - - xp_init_xskb_dma(xskb, pool, dma_map->dma_pages, xskb->orig_addr); - } err = xp_init_dma_info(pool, dma_map); if (err) { @@ -583,9 +605,13 @@ u32 xp_alloc_batch(struct xsk_buff_pool *pool, struct xdp_buff **xdp, u32 max) u32 nb_entries1 = 0, nb_entries2; if (unlikely(pool->dma_need_sync)) { + struct xdp_buff *buff; + /* Slow path */ - *xdp = xp_alloc(pool); - return !!*xdp; + buff = xp_alloc(pool); + if (buff) + *xdp = buff; + return !!buff; } if (unlikely(pool->free_list_cnt)) { diff --git a/net/xdp/xsk_queue.h b/net/xdp/xsk_queue.h index e9aa2c236356..c6fb6b763658 100644 --- a/net/xdp/xsk_queue.h +++ b/net/xdp/xsk_queue.h @@ -205,11 +205,16 @@ static inline bool xskq_cons_read_desc(struct xsk_queue *q, return false; } -static inline u32 xskq_cons_read_desc_batch(struct xsk_queue *q, - struct xdp_desc *descs, - struct xsk_buff_pool *pool, u32 max) +static inline void xskq_cons_release_n(struct xsk_queue *q, u32 cnt) +{ + q->cached_cons += cnt; +} + +static inline u32 xskq_cons_read_desc_batch(struct xsk_queue *q, struct xsk_buff_pool *pool, + u32 max) { u32 cached_cons = q->cached_cons, nb_entries = 0; + struct xdp_desc *descs = pool->tx_descs; while (cached_cons != q->cached_prod && nb_entries < max) { struct xdp_rxtx_ring *ring = (struct xdp_rxtx_ring *)q->ring; @@ -226,6 +231,8 @@ static inline u32 xskq_cons_read_desc_batch(struct xsk_queue *q, cached_cons++; } + /* Release valid plus any invalid entries */ + xskq_cons_release_n(q, cached_cons - q->cached_cons); return nb_entries; } @@ -263,7 +270,7 @@ static inline u32 xskq_cons_nb_entries(struct xsk_queue *q, u32 max) static inline bool xskq_cons_has_entries(struct xsk_queue *q, u32 cnt) { - return xskq_cons_nb_entries(q, cnt) >= cnt ? true : false; + return xskq_cons_nb_entries(q, cnt) >= cnt; } static inline bool xskq_cons_peek_addr_unchecked(struct xsk_queue *q, u64 *addr) @@ -282,14 +289,6 @@ static inline bool xskq_cons_peek_desc(struct xsk_queue *q, return xskq_cons_read_desc(q, desc, pool); } -static inline u32 xskq_cons_peek_desc_batch(struct xsk_queue *q, struct xdp_desc *descs, - struct xsk_buff_pool *pool, u32 max) -{ - u32 entries = xskq_cons_nb_entries(q, max); - - return xskq_cons_read_desc_batch(q, descs, pool, entries); -} - /* To improve performance in the xskq_cons_release functions, only update local state here. * Reflect this to global state when we get new entries from the ring in * xskq_cons_get_entries() and whenever Rx or Tx processing are completed in the NAPI loop. @@ -299,18 +298,6 @@ static inline void xskq_cons_release(struct xsk_queue *q) q->cached_cons++; } -static inline void xskq_cons_release_n(struct xsk_queue *q, u32 cnt) -{ - q->cached_cons += cnt; -} - -static inline bool xskq_cons_is_full(struct xsk_queue *q) -{ - /* No barriers needed since data is not accessed */ - return READ_ONCE(q->ring->producer) - READ_ONCE(q->ring->consumer) == - q->nentries; -} - static inline u32 xskq_cons_present_entries(struct xsk_queue *q) { /* No barriers needed since data is not accessed */ @@ -365,21 +352,17 @@ static inline int xskq_prod_reserve_addr(struct xsk_queue *q, u64 addr) return 0; } -static inline u32 xskq_prod_reserve_addr_batch(struct xsk_queue *q, struct xdp_desc *descs, - u32 max) +static inline void xskq_prod_write_addr_batch(struct xsk_queue *q, struct xdp_desc *descs, + u32 nb_entries) { struct xdp_umem_ring *ring = (struct xdp_umem_ring *)q->ring; - u32 nb_entries, i, cached_prod; - - nb_entries = xskq_prod_nb_free(q, max); + u32 i, cached_prod; /* A, matches D */ cached_prod = q->cached_prod; for (i = 0; i < nb_entries; i++) ring->desc[cached_prod++ & q->ring_mask] = descs[i].addr; q->cached_prod = cached_prod; - - return nb_entries; } static inline int xskq_prod_reserve_desc(struct xsk_queue *q, @@ -389,7 +372,7 @@ static inline int xskq_prod_reserve_desc(struct xsk_queue *q, u32 idx; if (xskq_prod_is_full(q)) - return -ENOSPC; + return -ENOBUFS; /* A, matches D */ idx = q->cached_prod++ & q->ring_mask; diff --git a/net/xdp/xskmap.c b/net/xdp/xskmap.c index 2e48d0e094d9..acc8e52a4f5f 100644 --- a/net/xdp/xskmap.c +++ b/net/xdp/xskmap.c @@ -4,10 +4,12 @@ */ #include <linux/bpf.h> +#include <linux/filter.h> #include <linux/capability.h> #include <net/xdp_sock.h> #include <linux/slab.h> #include <linux/sched.h> +#include <linux/btf_ids.h> #include "xsk.h" @@ -253,7 +255,7 @@ static bool xsk_map_meta_equal(const struct bpf_map *meta0, bpf_map_meta_equal(meta0, meta1); } -static int xsk_map_btf_id; +BTF_ID_LIST_SINGLE(xsk_map_btf_ids, struct, xsk_map) const struct bpf_map_ops xsk_map_ops = { .map_meta_equal = xsk_map_meta_equal, .map_alloc = xsk_map_alloc, @@ -265,7 +267,6 @@ const struct bpf_map_ops xsk_map_ops = { .map_update_elem = xsk_map_update_elem, .map_delete_elem = xsk_map_delete_elem, .map_check_btf = map_check_no_btf, - .map_btf_name = "xsk_map", - .map_btf_id = &xsk_map_btf_id, + .map_btf_id = &xsk_map_btf_ids[0], .map_redirect = xsk_map_redirect, }; |