X-Git-Url: http://git.droids-corp.org/?a=blobdiff_plain;f=drivers%2Fnet%2Faf_xdp%2Frte_eth_af_xdp.c;h=3957227bf01cb592f8721fd95177539d35f418b6;hb=0984d196be2a92eb6e2e0b926fdb4a06a1d7d823;hp=01f462b4656e431b216f84c7829b1ed8a41898e5;hpb=74b46340e2d4f6b4b00476ed3790580004f21341;p=dpdk.git

diff --git a/drivers/net/af_xdp/rte_eth_af_xdp.c b/drivers/net/af_xdp/rte_eth_af_xdp.c
index 01f462b465..3957227bf0 100644
--- a/drivers/net/af_xdp/rte_eth_af_xdp.c
+++ b/drivers/net/af_xdp/rte_eth_af_xdp.c
@@ -19,8 +19,8 @@
 #include <bpf/xsk.h>
 
 #include <rte_ethdev.h>
-#include <rte_ethdev_driver.h>
-#include <rte_ethdev_vdev.h>
+#include <ethdev_driver.h>
+#include <ethdev_vdev.h>
 #include <rte_kvargs.h>
 #include <rte_bus_vdev.h>
 #include <rte_string_fns.h>
@@ -126,6 +126,8 @@ struct pmd_internals {
 	int max_queue_cnt;
 	int combined_queue_cnt;
 	bool shared_umem;
+	char prog_path[PATH_MAX];
+	bool custom_prog_configured;
 
 	struct rte_ether_addr eth_addr;
 
@@ -137,12 +139,14 @@ struct pmd_internals {
 #define ETH_AF_XDP_START_QUEUE_ARG		"start_queue"
 #define ETH_AF_XDP_QUEUE_COUNT_ARG		"queue_count"
 #define ETH_AF_XDP_SHARED_UMEM_ARG		"shared_umem"
+#define ETH_AF_XDP_PROG_ARG			"xdp_prog"
 
 static const char * const valid_arguments[] = {
 	ETH_AF_XDP_IFACE_ARG,
 	ETH_AF_XDP_START_QUEUE_ARG,
 	ETH_AF_XDP_QUEUE_COUNT_ARG,
 	ETH_AF_XDP_SHARED_UMEM_ARG,
+	ETH_AF_XDP_PROG_ARG,
 	NULL
 };
 
@@ -251,28 +255,32 @@ af_xdp_rx_zc(void *queue, struct rte_mbuf **bufs, uint16_t nb_pkts)
 	struct xsk_umem_info *umem = rxq->umem;
 	uint32_t idx_rx = 0;
 	unsigned long rx_bytes = 0;
-	int rcvd, i;
+	int i;
 	struct rte_mbuf *fq_bufs[ETH_AF_XDP_RX_BATCH_SIZE];
 
-	/* allocate bufs for fill queue replenishment after rx */
-	if (rte_pktmbuf_alloc_bulk(umem->mb_pool, fq_bufs, nb_pkts)) {
-		AF_XDP_LOG(DEBUG,
-			"Failed to get enough buffers for fq.\n");
-		return 0;
-	}
+	nb_pkts = xsk_ring_cons__peek(rx, nb_pkts, &idx_rx);
 
-	rcvd = xsk_ring_cons__peek(rx, nb_pkts, &idx_rx);
-
-	if (rcvd == 0) {
+	if (nb_pkts == 0) {
 #if defined(XDP_USE_NEED_WAKEUP)
 		if (xsk_ring_prod__needs_wakeup(fq))
 			(void)poll(rxq->fds, 1, 1000);
 #endif
 
-		goto out;
+		return 0;
+	}
+
+	/* allocate bufs for fill queue replenishment after rx */
+	if (rte_pktmbuf_alloc_bulk(umem->mb_pool, fq_bufs, nb_pkts)) {
+		AF_XDP_LOG(DEBUG,
+			"Failed to get enough buffers for fq.\n");
+		/* rollback cached_cons which is added by
+		 * xsk_ring_cons__peek
+		 */
+		rx->cached_cons -= nb_pkts;
+		return 0;
 	}
 
-	for (i = 0; i < rcvd; i++) {
+	for (i = 0; i < nb_pkts; i++) {
 		const struct xdp_desc *desc;
 		uint64_t addr;
 		uint32_t len;
@@ -297,20 +305,14 @@ af_xdp_rx_zc(void *queue, struct rte_mbuf **bufs, uint16_t nb_pkts)
 		rx_bytes += len;
 	}
 
-	xsk_ring_cons__release(rx, rcvd);
-
-	(void)reserve_fill_queue(umem, rcvd, fq_bufs, fq);
+	xsk_ring_cons__release(rx, nb_pkts);
+	(void)reserve_fill_queue(umem, nb_pkts, fq_bufs, fq);
 
 	/* statistics */
-	rxq->stats.rx_pkts += rcvd;
+	rxq->stats.rx_pkts += nb_pkts;
 	rxq->stats.rx_bytes += rx_bytes;
 
-out:
-	if (rcvd != nb_pkts)
-		rte_mempool_put_bulk(umem->mb_pool, (void **)&fq_bufs[rcvd],
-				     nb_pkts - rcvd);
-
-	return rcvd;
+	return nb_pkts;
 }
 #else
 static uint16_t
@@ -322,7 +324,7 @@ af_xdp_rx_cp(void *queue, struct rte_mbuf **bufs, uint16_t nb_pkts)
 	struct xsk_ring_prod *fq = &rxq->fq;
 	uint32_t idx_rx = 0;
 	unsigned long rx_bytes = 0;
-	int rcvd, i;
+	int i;
 	uint32_t free_thresh = fq->size >> 1;
 	struct rte_mbuf *mbufs[ETH_AF_XDP_RX_BATCH_SIZE];
 
@@ -330,20 +332,24 @@ af_xdp_rx_cp(void *queue, struct rte_mbuf **bufs, uint16_t nb_pkts)
 		(void)reserve_fill_queue(umem, ETH_AF_XDP_RX_BATCH_SIZE,
 					 NULL, fq);
 
-	if (unlikely(rte_pktmbuf_alloc_bulk(rxq->mb_pool, mbufs, nb_pkts) != 0))
-		return 0;
-
-	rcvd = xsk_ring_cons__peek(rx, nb_pkts, &idx_rx);
-	if (rcvd == 0) {
+	nb_pkts = xsk_ring_cons__peek(rx, nb_pkts, &idx_rx);
+	if (nb_pkts == 0) {
 #if defined(XDP_USE_NEED_WAKEUP)
 		if (xsk_ring_prod__needs_wakeup(fq))
 			(void)poll(rxq->fds, 1, 1000);
 #endif
+		return 0;
+	}
 
-		goto out;
+	if (unlikely(rte_pktmbuf_alloc_bulk(rxq->mb_pool, mbufs, nb_pkts))) {
+		/* rollback cached_cons which is added by
+		 * xsk_ring_cons__peek
+		 */
+		rx->cached_cons -= nb_pkts;
+		return 0;
 	}
 
-	for (i = 0; i < rcvd; i++) {
+	for (i = 0; i < nb_pkts; i++) {
 		const struct xdp_desc *desc;
 		uint64_t addr;
 		uint32_t len;
@@ -362,18 +368,13 @@ af_xdp_rx_cp(void *queue, struct rte_mbuf **bufs, uint16_t nb_pkts)
 		bufs[i] = mbufs[i];
 	}
 
-	xsk_ring_cons__release(rx, rcvd);
+	xsk_ring_cons__release(rx, nb_pkts);
 
 	/* statistics */
-	rxq->stats.rx_pkts += rcvd;
+	rxq->stats.rx_pkts += nb_pkts;
 	rxq->stats.rx_bytes += rx_bytes;
 
-out:
-	if (rcvd != nb_pkts)
-		rte_mempool_put_bulk(rxq->mb_pool, (void **)&mbufs[rcvd],
-				     nb_pkts - rcvd);
-
-	return rcvd;
+	return nb_pkts;
 }
 #endif
 
@@ -595,10 +596,11 @@ eth_dev_start(struct rte_eth_dev *dev)
 }
 
 /* This function gets called when the current port gets stopped. */
-static void
+static int
 eth_dev_stop(struct rte_eth_dev *dev)
 {
 	dev->data->dev_link.link_status = ETH_LINK_DOWN;
+	return 0;
 }
 
 /* Find ethdev in list */
@@ -630,16 +632,35 @@ find_internal_resource(struct pmd_internals *port_int)
 	return list;
 }
 
+/* Check if the netdev,qid context already exists */
+static inline bool
+ctx_exists(struct pkt_rx_queue *rxq, const char *ifname,
+		struct pkt_rx_queue *list_rxq, const char *list_ifname)
+{
+	bool exists = false;
+
+	if (rxq->xsk_queue_idx == list_rxq->xsk_queue_idx &&
+			!strncmp(ifname, list_ifname, IFNAMSIZ)) {
+		AF_XDP_LOG(ERR, "ctx %s,%i already exists, cannot share umem\n",
+					ifname, rxq->xsk_queue_idx);
+		exists = true;
+	}
+
+	return exists;
+}
+
 /* Get a pointer to an existing UMEM which overlays the rxq's mb_pool */
-static inline struct xsk_umem_info *
-get_shared_umem(struct pkt_rx_queue *rxq) {
+static inline int
+get_shared_umem(struct pkt_rx_queue *rxq, const char *ifname,
+			struct xsk_umem_info **umem)
+{
 	struct internal_list *list;
 	struct pmd_internals *internals;
-	int i = 0;
+	int i = 0, ret = 0;
 	struct rte_mempool *mb_pool = rxq->mb_pool;
 
 	if (mb_pool == NULL)
-		return NULL;
+		return ret;
 
 	pthread_mutex_lock(&internal_list_lock);
 
@@ -651,20 +672,25 @@ get_shared_umem(struct pkt_rx_queue *rxq) {
 			if (rxq == list_rxq)
 				continue;
 			if (mb_pool == internals->rx_queues[i].mb_pool) {
+				if (ctx_exists(rxq, ifname, list_rxq,
+						internals->if_name)) {
+					ret = -1;
+					goto out;
+				}
 				if (__atomic_load_n(
 					&internals->rx_queues[i].umem->refcnt,
 							__ATOMIC_ACQUIRE)) {
-					pthread_mutex_unlock(
-							&internal_list_lock);
-					return internals->rx_queues[i].umem;
+					*umem = internals->rx_queues[i].umem;
+					goto out;
 				}
 			}
 		}
 	}
 
+out:
 	pthread_mutex_unlock(&internal_list_lock);
 
-	return NULL;
+	return ret;
 }
 
 static int
@@ -811,16 +837,18 @@ xdp_umem_destroy(struct xsk_umem_info *umem)
 #endif
 
 	rte_free(umem);
-	umem = NULL;
 }
 
-static void
+static int
 eth_dev_close(struct rte_eth_dev *dev)
 {
 	struct pmd_internals *internals = dev->data->dev_private;
 	struct pkt_rx_queue *rxq;
 	int i;
 
+	if (rte_eal_process_type() != RTE_PROC_PRIMARY)
+		return 0;
+
 	AF_XDP_LOG(INFO, "Closing AF_XDP ethdev on numa socket %u\n",
 		rte_socket_id());
 
@@ -861,6 +889,8 @@ eth_dev_close(struct rte_eth_dev *dev)
 			rte_free(list);
 		}
 	}
+
+	return 0;
 }
 
 static void
@@ -876,13 +906,13 @@ eth_link_update(struct rte_eth_dev *dev __rte_unused,
 }
 
 #if defined(XDP_UMEM_UNALIGNED_CHUNK_FLAG)
-static inline uint64_t get_base_addr(struct rte_mempool *mp, uint64_t *align)
+static inline uintptr_t get_base_addr(struct rte_mempool *mp, uint64_t *align)
 {
 	struct rte_mempool_memhdr *memhdr;
-	uint64_t memhdr_addr, aligned_addr;
+	uintptr_t memhdr_addr, aligned_addr;
 
 	memhdr = STAILQ_FIRST(&mp->mem_list);
-	memhdr_addr = (uint64_t)memhdr->addr;
+	memhdr_addr = (uintptr_t)memhdr->addr;
 	aligned_addr = memhdr_addr & ~(getpagesize() - 1);
 	*align = memhdr_addr - aligned_addr;
 
@@ -904,7 +934,9 @@ xsk_umem_info *xdp_umem_configure(struct pmd_internals *internals,
 	uint64_t umem_size, align = 0;
 
 	if (internals->shared_umem) {
-		umem = get_shared_umem(rxq);
+		if (get_shared_umem(rxq, internals->if_name, &umem) < 0)
+			return NULL;
+
 		if (umem != NULL &&
 			__atomic_load_n(&umem->refcnt, __ATOMIC_ACQUIRE) <
 					umem->max_xsks) {
@@ -932,7 +964,8 @@ xsk_umem_info *xdp_umem_configure(struct pmd_internals *internals,
 
 		umem->mb_pool = mb_pool;
 		base_addr = (void *)get_base_addr(mb_pool, &align);
-		umem_size = mb_pool->populated_size * usr_config.frame_size +
+		umem_size = (uint64_t)mb_pool->populated_size *
+				(uint64_t)usr_config.frame_size +
 				align;
 
 		ret = xsk_umem__create(&umem->umem, base_addr, umem_size,
@@ -1021,6 +1054,45 @@ err:
 	return NULL;
 }
 
+static int
+load_custom_xdp_prog(const char *prog_path, int if_index)
+{
+	int ret, prog_fd = -1;
+	struct bpf_object *obj;
+	struct bpf_map *map;
+
+	ret = bpf_prog_load(prog_path, BPF_PROG_TYPE_XDP, &obj, &prog_fd);
+	if (ret) {
+		AF_XDP_LOG(ERR, "Failed to load program %s\n", prog_path);
+		return ret;
+	}
+
+	/*
+	 * The loaded program must provision for a map of xsks, such that some
+	 * traffic can be redirected to userspace. When the xsk is created,
+	 * libbpf inserts it into the map.
+	 */
+	map = bpf_object__find_map_by_name(obj, "xsks_map");
+	if (!map) {
+		AF_XDP_LOG(ERR, "Failed to find xsks_map in %s\n", prog_path);
+		return -1;
+	}
+
+	/* Link the program with the given network device */
+	ret = bpf_set_link_xdp_fd(if_index, prog_fd,
+					XDP_FLAGS_UPDATE_IF_NOEXIST);
+	if (ret) {
+		AF_XDP_LOG(ERR, "Failed to set prog fd %d on interface\n",
+				prog_fd);
+		return -1;
+	}
+
+	AF_XDP_LOG(INFO, "Successfully loaded XDP program %s with fd %d\n",
+				prog_path, prog_fd);
+
+	return 0;
+}
+
 static int
 xsk_configure(struct pmd_internals *internals, struct pkt_rx_queue *rxq,
 	      int ring_size)
@@ -1046,6 +1118,18 @@ xsk_configure(struct pmd_internals *internals, struct pkt_rx_queue *rxq,
 	cfg.bind_flags |= XDP_USE_NEED_WAKEUP;
 #endif
 
+	if (strnlen(internals->prog_path, PATH_MAX) &&
+				!internals->custom_prog_configured) {
+		ret = load_custom_xdp_prog(internals->prog_path,
+					   internals->if_index);
+		if (ret) {
+			AF_XDP_LOG(ERR, "Failed to load custom XDP program %s\n",
+					internals->prog_path);
+			goto err;
+		}
+		internals->custom_prog_configured = 1;
+	}
+
 	if (internals->shared_umem)
 		ret = create_shared_socket(&rxq->xsk, internals->if_name,
 				rxq->xsk_queue_idx, rxq->umem->umem, &rxq->rx,
@@ -1061,7 +1145,8 @@ xsk_configure(struct pmd_internals *internals, struct pkt_rx_queue *rxq,
 	}
 
 #if defined(XDP_UMEM_UNALIGNED_CHUNK_FLAG)
-	if (rte_pktmbuf_alloc_bulk(rxq->umem->mb_pool, fq_bufs, reserve_size)) {
+	ret = rte_pktmbuf_alloc_bulk(rxq->umem->mb_pool, fq_bufs, reserve_size);
+	if (ret) {
 		AF_XDP_LOG(DEBUG, "Failed to get enough buffers for fq.\n");
 		goto err;
 	}
@@ -1264,6 +1349,30 @@ parse_name_arg(const char *key __rte_unused,
 	return 0;
 }
 
+/** parse xdp prog argument */
+static int
+parse_prog_arg(const char *key __rte_unused,
+	       const char *value, void *extra_args)
+{
+	char *path = extra_args;
+
+	if (strnlen(value, PATH_MAX) == PATH_MAX) {
+		AF_XDP_LOG(ERR, "Invalid path %s, should be less than %u bytes.\n",
+			   value, PATH_MAX);
+		return -EINVAL;
+	}
+
+	if (access(value, F_OK) != 0) {
+		AF_XDP_LOG(ERR, "Error accessing %s: %s\n",
+			   value, strerror(errno));
+		return -EINVAL;
+	}
+
+	strlcpy(path, value, PATH_MAX);
+
+	return 0;
+}
+
 static int
 xdp_get_channels_info(const char *if_name, int *max_queues,
 				int *combined_queues)
@@ -1278,7 +1387,7 @@ xdp_get_channels_info(const char *if_name, int *max_queues,
 
 	channels.cmd = ETHTOOL_GCHANNELS;
 	ifr.ifr_data = (void *)&channels;
-	strncpy(ifr.ifr_name, if_name, IFNAMSIZ);
+	strlcpy(ifr.ifr_name, if_name, IFNAMSIZ);
 	ret = ioctl(fd, SIOCETHTOOL, &ifr);
 	if (ret) {
 		if (errno == EOPNOTSUPP) {
@@ -1307,7 +1416,7 @@ xdp_get_channels_info(const char *if_name, int *max_queues,
 
 static int
 parse_parameters(struct rte_kvargs *kvlist, char *if_name, int *start_queue,
-			int *queue_cnt, int *shared_umem)
+			int *queue_cnt, int *shared_umem, char *prog_path)
 {
 	int ret;
 
@@ -1333,6 +1442,11 @@ parse_parameters(struct rte_kvargs *kvlist, char *if_name, int *start_queue,
 	if (ret < 0)
 		goto free_kvlist;
 
+	ret = rte_kvargs_process(kvlist, ETH_AF_XDP_PROG_ARG,
+				 &parse_prog_arg, prog_path);
+	if (ret < 0)
+		goto free_kvlist;
+
 free_kvlist:
 	rte_kvargs_free(kvlist);
 	return ret;
@@ -1370,7 +1484,8 @@ error:
 
 static struct rte_eth_dev *
 init_internals(struct rte_vdev_device *dev, const char *if_name,
-			int start_queue_idx, int queue_cnt, int shared_umem)
+		int start_queue_idx, int queue_cnt, int shared_umem,
+		const char *prog_path)
 {
 	const char *name = rte_vdev_device_name(dev);
 	const unsigned int numa_node = dev->device.numa_node;
@@ -1386,6 +1501,8 @@ init_internals(struct rte_vdev_device *dev, const char *if_name,
 	internals->start_queue_idx = start_queue_idx;
 	internals->queue_cnt = queue_cnt;
 	strlcpy(internals->if_name, if_name, IFNAMSIZ);
+	strlcpy(internals->prog_path, prog_path, PATH_MAX);
+	internals->custom_prog_configured = 0;
 
 #ifndef ETH_AF_XDP_SHARED_UMEM
 	if (shared_umem) {
@@ -1443,11 +1560,10 @@ init_internals(struct rte_vdev_device *dev, const char *if_name,
 	eth_dev->data->dev_private = internals;
 	eth_dev->data->dev_link = pmd_link;
 	eth_dev->data->mac_addrs = &internals->eth_addr;
+	eth_dev->data->dev_flags |= RTE_ETH_DEV_AUTOFILL_QUEUE_XSTATS;
 	eth_dev->dev_ops = &ops;
 	eth_dev->rx_pkt_burst = eth_af_xdp_rx;
 	eth_dev->tx_pkt_burst = eth_af_xdp_tx;
-	/* Let rte_eth_dev_close() release the port resources. */
-	eth_dev->data->dev_flags |= RTE_ETH_DEV_CLOSE_REMOVE;
 
 #if defined(XDP_UMEM_UNALIGNED_CHUNK_FLAG)
 	AF_XDP_LOG(INFO, "Zero copy between umem and mbuf enabled.\n");
@@ -1472,6 +1588,7 @@ rte_pmd_af_xdp_probe(struct rte_vdev_device *dev)
 	int xsk_start_queue_idx = ETH_AF_XDP_DFLT_START_QUEUE_IDX;
 	int xsk_queue_cnt = ETH_AF_XDP_DFLT_QUEUE_COUNT;
 	int shared_umem = 0;
+	char prog_path[PATH_MAX] = {'\0'};
 	struct rte_eth_dev *eth_dev = NULL;
 	const char *name;
 
@@ -1501,7 +1618,7 @@ rte_pmd_af_xdp_probe(struct rte_vdev_device *dev)
 		dev->device.numa_node = rte_socket_id();
 
 	if (parse_parameters(kvlist, if_name, &xsk_start_queue_idx,
-			     &xsk_queue_cnt, &shared_umem) < 0) {
+			     &xsk_queue_cnt, &shared_umem, prog_path) < 0) {
 		AF_XDP_LOG(ERR, "Invalid kvargs value\n");
 		return -EINVAL;
 	}
@@ -1512,7 +1629,7 @@ rte_pmd_af_xdp_probe(struct rte_vdev_device *dev)
 	}
 
 	eth_dev = init_internals(dev, if_name, xsk_start_queue_idx,
-					xsk_queue_cnt, shared_umem);
+					xsk_queue_cnt, shared_umem, prog_path);
 	if (eth_dev == NULL) {
 		AF_XDP_LOG(ERR, "Failed to init internals\n");
 		return -1;
@@ -1556,4 +1673,5 @@ RTE_PMD_REGISTER_PARAM_STRING(net_af_xdp,
 			      "iface=<string> "
 			      "start_queue=<int> "
 			      "queue_count=<int> "
-			      "shared_umem=<int> ");
+			      "shared_umem=<int> "
+			      "xdp_prog=<string> ");