[dpdk-dev] [PATCH v3 2/5] Link Bonding PMD Library (librte_eal/librte_ether link bonding support changes)

Declan Doherty declan.doherty at intel.com
Fri Jun 13 16:41:59 CEST 2014


Updating functionality in EAL to support adding link bonding
devices via –vdev option. Link bonding devices will be
initialized after all physical devices have been probed and
initialized.

Signed-off-by: Declan Doherty <declan.doherty at intel.com>
---
 lib/librte_eal/common/eal_common_dev.c      | 66 +++++++++++++++++++++++++++--
 lib/librte_eal/common/eal_common_pci.c      |  6 +++
 lib/librte_eal/common/include/eal_private.h |  7 +++
 lib/librte_eal/common/include/rte_dev.h     |  1 +
 lib/librte_ether/rte_ethdev.c               | 34 +++++++++++++--
 lib/librte_ether/rte_ethdev.h               |  7 ++-
 lib/librte_pmd_pcap/rte_eth_pcap.c          | 22 +++++-----
 lib/librte_pmd_ring/rte_eth_ring.c          | 32 +++++++-------
 lib/librte_pmd_ring/rte_eth_ring.h          |  3 +-
 lib/librte_pmd_xenvirt/rte_eth_xenvirt.c    |  2 +-
 10 files changed, 144 insertions(+), 36 deletions(-)

diff --git a/lib/librte_eal/common/eal_common_dev.c b/lib/librte_eal/common/eal_common_dev.c
index eae5656..b50c908 100644
--- a/lib/librte_eal/common/eal_common_dev.c
+++ b/lib/librte_eal/common/eal_common_dev.c
@@ -75,14 +75,28 @@ rte_eal_dev_init(void)
 
 	/* call the init function for each virtual device */
 	TAILQ_FOREACH(devargs, &devargs_list, next) {
+		uint8_t bdev = 0;
 
 		if (devargs->type != RTE_DEVTYPE_VIRTUAL)
 			continue;
 
 		TAILQ_FOREACH(driver, &dev_driver_list, next) {
-			if (driver->type != PMD_VDEV)
+			/* RTE_DEVTYPE_VIRTUAL can only be a virtual or bonded device*/
+			if (driver->type != PMD_VDEV && driver->type != PMD_BDEV)
 				continue;
 
+			/*
+			 * Bonded devices are not initialize here, we do it later in
+			 * rte_eal_bonded_dev_init() after all physical devices have been
+			 * probed and initialized
+			 */
+			if (driver->type == PMD_BDEV &&
+					!strncmp(driver->name, devargs->virtual.drv_name,
+							strlen(driver->name))) {
+				bdev = 1;
+				break;
+			}
+
 			/* search a driver prefix in virtual device name */
 			if (!strncmp(driver->name, devargs->virtual.drv_name,
 					strlen(driver->name))) {
@@ -92,9 +106,9 @@ rte_eal_dev_init(void)
 			}
 		}
 
-		if (driver == NULL) {
-			rte_panic("no driver found for %s\n",
-				  devargs->virtual.drv_name);
+		if (driver == NULL && !bdev) {
+			rte_panic("no driver found for %s and is not a bonded vdev %d\n",
+				  devargs->virtual.drv_name, bdev);
 		}
 	}
 
@@ -107,3 +121,47 @@ rte_eal_dev_init(void)
 	}
 	return 0;
 }
+
+#ifdef RTE_LIBRTE_PMD_BOND
+int
+rte_eal_bonded_dev_init(void)
+{
+	struct rte_devargs *devargs;
+	struct rte_driver *driver;
+
+	TAILQ_FOREACH(devargs, &devargs_list, next) {
+		int vdev = 0;
+
+		if (devargs->type != RTE_DEVTYPE_VIRTUAL)
+			continue;
+
+		TAILQ_FOREACH(driver, &dev_driver_list, next) {
+			if (driver->type != PMD_VDEV && driver->type != PMD_BDEV)
+				continue;
+
+			/* Virtual devices have already been initialized so we skip them
+			 * here*/
+			if (driver->type == PMD_VDEV &&
+					!strncmp(driver->name, devargs->virtual.drv_name,
+							strlen(driver->name))) {
+				vdev = 1;
+				break;
+			}
+
+			/* search a driver prefix in bonded device name */
+			if (!strncmp(driver->name, devargs->virtual.drv_name,
+					strlen(driver->name))) {
+				driver->init(devargs->virtual.drv_name, devargs->args);
+				break;
+			}
+		}
+
+		if (driver == NULL && !vdev) {
+			rte_panic("no driver found for %s\n",
+					devargs->virtual.drv_name);
+		}
+	}
+	return 0;
+}
+#endif
+
diff --git a/lib/librte_eal/common/eal_common_pci.c b/lib/librte_eal/common/eal_common_pci.c
index 4d877ea..9b584f5 100644
--- a/lib/librte_eal/common/eal_common_pci.c
+++ b/lib/librte_eal/common/eal_common_pci.c
@@ -166,7 +166,13 @@ rte_eal_pci_probe(void)
 				 dev->addr.devid, dev->addr.function);
 	}
 
+#ifdef RTE_LIBRTE_PMD_BOND
+	/* After all physical PCI devices have been probed and initialized then we
+	 * initialize the bonded devices */
+	return rte_eal_bonded_dev_init();
+#else
 	return 0;
+#endif
 }
 
 /* dump one device */
diff --git a/lib/librte_eal/common/include/eal_private.h b/lib/librte_eal/common/include/eal_private.h
index 232fcec..f6081bb 100644
--- a/lib/librte_eal/common/include/eal_private.h
+++ b/lib/librte_eal/common/include/eal_private.h
@@ -203,4 +203,11 @@ int rte_eal_alarm_init(void);
  */
 int rte_eal_dev_init(void);
 
+#ifdef RTE_LIBRTE_PMD_BOND
+/**
+ * Initialize the bonded devices
+ */
+int rte_eal_bonded_dev_init(void);
+#endif
+
 #endif /* _EAL_PRIVATE_H_ */
diff --git a/lib/librte_eal/common/include/rte_dev.h b/lib/librte_eal/common/include/rte_dev.h
index f7e3a10..f0a780a 100644
--- a/lib/librte_eal/common/include/rte_dev.h
+++ b/lib/librte_eal/common/include/rte_dev.h
@@ -62,6 +62,7 @@ typedef int (rte_dev_init_t)(const char *name, const char *args);
 enum pmd_type {
 	PMD_VDEV = 0,
 	PMD_PDEV = 1,
+	PMD_BDEV = 2,	/**< Poll Mode Driver Bonded Device*/
 };
 
 /**
diff --git a/lib/librte_ether/rte_ethdev.c b/lib/librte_ether/rte_ethdev.c
index 8011b8b..4c2f1d3 100644
--- a/lib/librte_ether/rte_ethdev.c
+++ b/lib/librte_ether/rte_ethdev.c
@@ -64,6 +64,7 @@
 #include <rte_mbuf.h>
 #include <rte_errno.h>
 #include <rte_spinlock.h>
+#include <rte_string_fns.h>
 
 #include "rte_ether.h"
 #include "rte_ethdev.h"
@@ -152,8 +153,21 @@ rte_eth_dev_data_alloc(void)
 				RTE_MAX_ETHPORTS * sizeof(*rte_eth_dev_data));
 }
 
+static int
+rte_eth_dev_name_unique(const char* name)
+{
+	unsigned i;
+
+	for (i = 0; i < nb_ports; i++) {
+		if (strcmp(rte_eth_devices[i].data->name, name) == 0)
+			return -1;
+	}
+
+	return 0;
+}
+
 struct rte_eth_dev *
-rte_eth_dev_allocate(void)
+rte_eth_dev_allocate(const char* name)
 {
 	struct rte_eth_dev *eth_dev;
 
@@ -165,23 +179,37 @@ rte_eth_dev_allocate(void)
 	if (rte_eth_dev_data == NULL)
 		rte_eth_dev_data_alloc();
 
+	if (rte_eth_dev_name_unique(name)) {
+		PMD_DEBUG_TRACE("Ethernet Device with name %s already allocated!\n");
+		return NULL;
+	}
+
 	eth_dev = &rte_eth_devices[nb_ports];
 	eth_dev->data = &rte_eth_dev_data[nb_ports];
+	rte_snprintf(eth_dev->data->name , sizeof(eth_dev->data->name ),
+			"%s", name);
 	eth_dev->data->port_id = nb_ports++;
 	return eth_dev;
 }
 
 static int
 rte_eth_dev_init(struct rte_pci_driver *pci_drv,
-		 struct rte_pci_device *pci_dev)
+		struct rte_pci_device *pci_dev)
 {
 	struct eth_driver    *eth_drv;
 	struct rte_eth_dev *eth_dev;
+	char ethdev_name[RTE_ETH_NAME_MAX_LEN];
+
 	int diag;
 
 	eth_drv = (struct eth_driver *)pci_drv;
 
-	eth_dev = rte_eth_dev_allocate();
+	/* Create unique ethdev name by concatenating drive name and number of
+	 * ports */
+	rte_snprintf(ethdev_name, RTE_ETH_NAME_MAX_LEN, "%d:%d.%d",
+			pci_dev->addr.bus, pci_dev->addr.devid, pci_dev->addr.function);
+
+	eth_dev = rte_eth_dev_allocate(ethdev_name);
 	if (eth_dev == NULL)
 		return -ENOMEM;
 
diff --git a/lib/librte_ether/rte_ethdev.h b/lib/librte_ether/rte_ethdev.h
index 67eda50..27ed0ab 100644
--- a/lib/librte_ether/rte_ethdev.h
+++ b/lib/librte_ether/rte_ethdev.h
@@ -1233,6 +1233,8 @@ struct rte_eth_dev_sriov {
 };
 #define RTE_ETH_DEV_SRIOV(dev)         ((dev)->data->sriov)
 
+#define RTE_ETH_NAME_MAX_LEN (32)
+
 /**
  * @internal
  * The data part, with no function pointers, associated with each ethernet device.
@@ -1241,6 +1243,8 @@ struct rte_eth_dev_sriov {
  * processes in a multi-process configuration.
  */
 struct rte_eth_dev_data {
+	char name[RTE_ETH_NAME_MAX_LEN]; /**< Unique identifier name */
+
 	void **rx_queues; /**< Array of pointers to RX queues. */
 	void **tx_queues; /**< Array of pointers to TX queues. */
 	uint16_t nb_rx_queues; /**< Number of RX queues. */
@@ -1293,10 +1297,11 @@ extern uint8_t rte_eth_dev_count(void);
  * Allocates a new ethdev slot for an ethernet device and returns the pointer
  * to that slot for the driver to use.
  *
+ * @param	name	Unique identifier name for each Ethernet device
  * @return
  *   - Slot in the rte_dev_devices array for a new device;
  */
-struct rte_eth_dev *rte_eth_dev_allocate(void);
+struct rte_eth_dev *rte_eth_dev_allocate(const char *name);
 
 struct eth_driver;
 /**
diff --git a/lib/librte_pmd_pcap/rte_eth_pcap.c b/lib/librte_pmd_pcap/rte_eth_pcap.c
index b3dbbda..cd6c0e1 100644
--- a/lib/librte_pmd_pcap/rte_eth_pcap.c
+++ b/lib/librte_pmd_pcap/rte_eth_pcap.c
@@ -534,7 +534,7 @@ open_tx_iface(const char *key __rte_unused, const char *value, void *extra_args)
 
 
 static int
-rte_pmd_init_internals(const unsigned nb_rx_queues,
+rte_pmd_init_internals(const char* name, const unsigned nb_rx_queues,
 		const unsigned nb_tx_queues,
 		const unsigned numa_node,
 		struct pmd_internals **internals,
@@ -558,20 +558,20 @@ rte_pmd_init_internals(const unsigned nb_rx_queues,
 	/* now do all data allocation - for eth_dev structure, dummy pci driver
 	 * and internal (private) data
 	 */
-	data = rte_zmalloc_socket(NULL, sizeof(*data), 0, numa_node);
+	data = rte_zmalloc_socket(name, sizeof(*data), 0, numa_node);
 	if (data == NULL)
 		goto error;
 
-	pci_dev = rte_zmalloc_socket(NULL, sizeof(*pci_dev), 0, numa_node);
+	pci_dev = rte_zmalloc_socket(name, sizeof(*pci_dev), 0, numa_node);
 	if (pci_dev == NULL)
 		goto error;
 
-	*internals = rte_zmalloc_socket(NULL, sizeof(**internals), 0, numa_node);
+	*internals = rte_zmalloc_socket(name, sizeof(**internals), 0, numa_node);
 	if (*internals == NULL)
 		goto error;
 
 	/* reserve an ethdev entry */
-	*eth_dev = rte_eth_dev_allocate();
+	*eth_dev = rte_eth_dev_allocate(name);
 	if (*eth_dev == NULL)
 		goto error;
 
@@ -617,7 +617,7 @@ rte_pmd_init_internals(const unsigned nb_rx_queues,
 }
 
 static int
-rte_eth_from_pcaps_n_dumpers(pcap_t * const rx_queues[],
+rte_eth_from_pcaps_n_dumpers(const char *name, pcap_t * const rx_queues[],
 		const unsigned nb_rx_queues,
 		pcap_dumper_t * const tx_queues[],
 		const unsigned nb_tx_queues,
@@ -634,7 +634,7 @@ rte_eth_from_pcaps_n_dumpers(pcap_t * const rx_queues[],
 	if (tx_queues == NULL && nb_tx_queues > 0)
 		return -1;
 
-	if (rte_pmd_init_internals(nb_rx_queues, nb_tx_queues, numa_node,
+	if (rte_pmd_init_internals(name, nb_rx_queues, nb_tx_queues, numa_node,
 			&internals, &eth_dev, kvlist) < 0)
 		return -1;
 
@@ -652,7 +652,7 @@ rte_eth_from_pcaps_n_dumpers(pcap_t * const rx_queues[],
 }
 
 static int
-rte_eth_from_pcaps(pcap_t * const rx_queues[],
+rte_eth_from_pcaps(const char *name, pcap_t * const rx_queues[],
 		const unsigned nb_rx_queues,
 		pcap_t * const tx_queues[],
 		const unsigned nb_tx_queues,
@@ -669,7 +669,7 @@ rte_eth_from_pcaps(pcap_t * const rx_queues[],
 	if (tx_queues == NULL && nb_tx_queues > 0)
 		return -1;
 
-	if (rte_pmd_init_internals(nb_rx_queues, nb_tx_queues, numa_node,
+	if (rte_pmd_init_internals(name, nb_rx_queues, nb_tx_queues, numa_node,
 			&internals, &eth_dev, kvlist) < 0)
 		return -1;
 
@@ -760,10 +760,10 @@ rte_pmd_pcap_devinit(const char *name, const char *params)
 		return -1;
 
 	if (using_dumpers)
-		return rte_eth_from_pcaps_n_dumpers(pcaps.pcaps, pcaps.num_of_rx,
+		return rte_eth_from_pcaps_n_dumpers(name, pcaps.pcaps, pcaps.num_of_rx,
 				dumpers.dumpers, dumpers.num_of_tx, numa_node, kvlist);
 
-	return rte_eth_from_pcaps(pcaps.pcaps, pcaps.num_of_rx, dumpers.pcaps,
+	return rte_eth_from_pcaps(name, pcaps.pcaps, pcaps.num_of_rx, dumpers.pcaps,
 			dumpers.num_of_tx, numa_node, kvlist);
 
 }
diff --git a/lib/librte_pmd_ring/rte_eth_ring.c b/lib/librte_pmd_ring/rte_eth_ring.c
index 10d4e24..e6779c4 100644
--- a/lib/librte_pmd_ring/rte_eth_ring.c
+++ b/lib/librte_pmd_ring/rte_eth_ring.c
@@ -219,7 +219,7 @@ static struct eth_dev_ops ops = {
 };
 
 int
-rte_eth_from_rings(struct rte_ring *const rx_queues[],
+rte_eth_from_rings(const char* name, struct rte_ring *const rx_queues[],
 		const unsigned nb_rx_queues,
 		struct rte_ring *const tx_queues[],
 		const unsigned nb_tx_queues,
@@ -243,20 +243,20 @@ rte_eth_from_rings(struct rte_ring *const rx_queues[],
 	/* now do all data allocation - for eth_dev structure, dummy pci driver
 	 * and internal (private) data
 	 */
-	data = rte_zmalloc_socket(NULL, sizeof(*data), 0, numa_node);
+	data = rte_zmalloc_socket(name, sizeof(*data), 0, numa_node);
 	if (data == NULL)
 		goto error;
 
-	pci_dev = rte_zmalloc_socket(NULL, sizeof(*pci_dev), 0, numa_node);
+	pci_dev = rte_zmalloc_socket(name, sizeof(*pci_dev), 0, numa_node);
 	if (pci_dev == NULL)
 		goto error;
 
-	internals = rte_zmalloc_socket(NULL, sizeof(*internals), 0, numa_node);
+	internals = rte_zmalloc_socket(name, sizeof(*internals), 0, numa_node);
 	if (internals == NULL)
 		goto error;
 
 	/* reserve an ethdev entry */
-	eth_dev = rte_eth_dev_allocate();
+	eth_dev = rte_eth_dev_allocate(name);
 	if (eth_dev == NULL)
 		goto error;
 
@@ -335,7 +335,7 @@ eth_dev_ring_create(const char *name, const unsigned numa_node,
 			return -1;
 	}
 
-	if (rte_eth_from_rings(rxtx, num_rings, rxtx, num_rings, numa_node))
+	if (rte_eth_from_rings(name, rxtx, num_rings, rxtx, num_rings, numa_node))
 		return -1;
 
 	return 0;
@@ -352,29 +352,31 @@ eth_dev_ring_pair_create(const char *name, const unsigned numa_node,
 	struct rte_ring *rx[RTE_PMD_RING_MAX_RX_RINGS];
 	struct rte_ring *tx[RTE_PMD_RING_MAX_TX_RINGS];
 	unsigned i;
-	char rng_name[RTE_RING_NAMESIZE];
+	char rx_rng_name[RTE_RING_NAMESIZE];
+	char tx_rng_name[RTE_RING_NAMESIZE];
 	unsigned num_rings = RTE_MIN(RTE_PMD_RING_MAX_RX_RINGS,
 			RTE_PMD_RING_MAX_TX_RINGS);
 
 	for (i = 0; i < num_rings; i++) {
-		rte_snprintf(rng_name, sizeof(rng_name), "ETH_RX%u_%s", i, name);
+		rte_snprintf(rx_rng_name, sizeof(rx_rng_name), "ETH_RX%u_%s", i, name);
 		rx[i] = (action == DEV_CREATE) ?
-				rte_ring_create(rng_name, 1024, numa_node,
+				rte_ring_create(rx_rng_name, 1024, numa_node,
 						RING_F_SP_ENQ|RING_F_SC_DEQ) :
-				rte_ring_lookup(rng_name);
+				rte_ring_lookup(rx_rng_name);
 		if (rx[i] == NULL)
 			return -1;
-		rte_snprintf(rng_name, sizeof(rng_name), "ETH_TX%u_%s", i, name);
+		rte_snprintf(tx_rng_name, sizeof(tx_rng_name), "ETH_TX%u_%s", i, name);
 		tx[i] = (action == DEV_CREATE) ?
-				rte_ring_create(rng_name, 1024, numa_node,
+				rte_ring_create(tx_rng_name, 1024, numa_node,
 						RING_F_SP_ENQ|RING_F_SC_DEQ):
-				rte_ring_lookup(rng_name);
+				rte_ring_lookup(tx_rng_name);
 		if (tx[i] == NULL)
 			return -1;
 	}
 
-	if (rte_eth_from_rings(rx, num_rings, tx, num_rings, numa_node) ||
-			rte_eth_from_rings(tx, num_rings, rx, num_rings, numa_node) )
+	if (rte_eth_from_rings(rx_rng_name, rx, num_rings, tx, num_rings,
+			numa_node) || rte_eth_from_rings(tx_rng_name, tx, num_rings, rx,
+					num_rings, numa_node) )
 		return -1;
 
 	return 0;
diff --git a/lib/librte_pmd_ring/rte_eth_ring.h b/lib/librte_pmd_ring/rte_eth_ring.h
index ef29344..e6ae19e 100644
--- a/lib/librte_pmd_ring/rte_eth_ring.h
+++ b/lib/librte_pmd_ring/rte_eth_ring.h
@@ -40,7 +40,8 @@ extern "C" {
 
 #include <rte_ring.h>
 
-int rte_eth_from_rings(struct rte_ring * const rx_queues[],
+int rte_eth_from_rings(const char *name,
+		struct rte_ring * const rx_queues[],
 		const unsigned nb_rx_queues,
 		struct rte_ring *const tx_queues[],
 		const unsigned nb_tx_queues,
diff --git a/lib/librte_pmd_xenvirt/rte_eth_xenvirt.c b/lib/librte_pmd_xenvirt/rte_eth_xenvirt.c
index 7c4d3fe..a0b4bdd 100644
--- a/lib/librte_pmd_xenvirt/rte_eth_xenvirt.c
+++ b/lib/librte_pmd_xenvirt/rte_eth_xenvirt.c
@@ -647,7 +647,7 @@ eth_dev_xenvirt_create(const char *name, const char *params,
 		goto err;
 
 	/* reserve an ethdev entry */
-	eth_dev = rte_eth_dev_allocate();
+	eth_dev = rte_eth_dev_allocate(name);
 	if (eth_dev == NULL)
 		goto err;
 
-- 
1.8.5.3



More information about the dev mailing list