[2/7] net/netvsc: fix VF support with secondary process

Message ID 20190208034407.7865-3-stephen@networkplumber.org
State Accepted
Delegated to: Thomas Monjalon
Headers show
Series
  • vmbus/netvsc: fix multi-process support
Related show

Checks

Context Check Description
ci/Intel-compilation success Compilation OK
ci/checkpatch success coding style OK

Commit Message

Stephen Hemminger Feb. 8, 2019, 3:44 a.m.
From: Stephen Hemminger <sthemmin@microsoft.com>

The VF device management in netvsc was using a pointer to the
rte_eth_devices. But the actual rte_eth_devices array is likely to
be place in the secondary process; which causes a crash.

The solution is to record the port of the VF (instead of a pointer)
and find the device in the per process array as needed.

Fixes: dc7680e8597c ("net/netvsc: support integrated VF")
Signed-off-by: Stephen Hemminger <sthemmin@microsoft.com>
---
 drivers/net/netvsc/hn_ethdev.c |  3 +-
 drivers/net/netvsc/hn_rxtx.c   |  8 ++--
 drivers/net/netvsc/hn_var.h    | 30 ++++++++++++-
 drivers/net/netvsc/hn_vf.c     | 82 +++++++++++++++++-----------------
 4 files changed, 75 insertions(+), 48 deletions(-)

Patch

diff --git a/drivers/net/netvsc/hn_ethdev.c b/drivers/net/netvsc/hn_ethdev.c
index 49b7ca7b2244..407ee484935a 100644
--- a/drivers/net/netvsc/hn_ethdev.c
+++ b/drivers/net/netvsc/hn_ethdev.c
@@ -735,6 +735,7 @@  eth_hn_dev_init(struct rte_eth_dev *eth_dev)
 	hv->port_id = eth_dev->data->port_id;
 	hv->latency = HN_CHAN_LATENCY_NS;
 	hv->max_queues = 1;
+	hv->vf_port = HN_INVALID_PORT;
 
 	err = hn_parse_args(eth_dev);
 	if (err)
@@ -788,7 +789,7 @@  eth_hn_dev_init(struct rte_eth_dev *eth_dev)
 	hv->max_queues = RTE_MIN(rxr_cnt, (unsigned int)max_chan);
 
 	/* If VF was reported but not added, do it now */
-	if (hv->vf_present && !hv->vf_dev) {
+	if (hv->vf_present && !hn_vf_attached(hv)) {
 		PMD_INIT_LOG(DEBUG, "Adding VF device");
 
 		err = hn_vf_add(eth_dev, hv);
diff --git a/drivers/net/netvsc/hn_rxtx.c b/drivers/net/netvsc/hn_rxtx.c
index 6197118b01ee..fecd69887e2b 100644
--- a/drivers/net/netvsc/hn_rxtx.c
+++ b/drivers/net/netvsc/hn_rxtx.c
@@ -1313,8 +1313,8 @@  hn_xmit_pkts(void *ptxq, struct rte_mbuf **tx_pkts, uint16_t nb_pkts)
 		return 0;
 
 	/* Transmit over VF if present and up */
-	vf_dev = hv->vf_dev;
-	rte_compiler_barrier();
+	vf_dev = hn_get_vf_dev(hv);
+
 	if (vf_dev && vf_dev->data->dev_started) {
 		void *sub_q = vf_dev->data->tx_queues[queue_id];
 
@@ -1404,8 +1404,8 @@  hn_recv_pkts(void *prxq, struct rte_mbuf **rx_pkts, uint16_t nb_pkts)
 	if (unlikely(hv->closed))
 		return 0;
 
-	vf_dev = hv->vf_dev;
-	rte_compiler_barrier();
+	/* Transmit over VF if present and up */
+	vf_dev = hn_get_vf_dev(hv);
 
 	if (vf_dev && vf_dev->data->dev_started) {
 		/* Normally, with SR-IOV the ring buffer will be empty */
diff --git a/drivers/net/netvsc/hn_var.h b/drivers/net/netvsc/hn_var.h
index 7f3266c451fb..8383f3246ca4 100644
--- a/drivers/net/netvsc/hn_var.h
+++ b/drivers/net/netvsc/hn_var.h
@@ -91,15 +91,19 @@  struct hn_rx_bufinfo {
 	struct rte_mbuf_ext_shared_info shinfo;
 } __rte_cache_aligned;
 
+#define HN_INVALID_PORT	UINT16_MAX
+
 struct hn_data {
 	struct rte_vmbus_device *vmbus;
 	struct hn_rx_queue *primary;
-	struct rte_eth_dev *vf_dev;		/* Subordinate device */
 	rte_spinlock_t  vf_lock;
 	uint16_t	port_id;
-	uint8_t		closed;
+	uint16_t	vf_port;
+
 	uint8_t		vf_present;
+	uint8_t		closed;
 	uint8_t		vlan_strip;
+
 	uint32_t	link_status;
 	uint32_t	link_speed;
 
@@ -170,6 +174,28 @@  int	hn_dev_rx_queue_setup(struct rte_eth_dev *dev,
 			      struct rte_mempool *mp);
 void	hn_dev_rx_queue_release(void *arg);
 
+/* Check if VF is attached */
+static inline bool
+hn_vf_attached(const struct hn_data *hv)
+{
+	return hv->vf_port != HN_INVALID_PORT;
+}
+
+/* Get VF device for existing netvsc device */
+static inline struct rte_eth_dev *
+hn_get_vf_dev(const struct hn_data *hv)
+{
+	uint16_t vf_port = hv->vf_port;
+
+	/* make sure vf_port is loaded */
+	rte_smp_rmb();
+
+	if (vf_port == HN_INVALID_PORT)
+		return NULL;
+	else
+		return &rte_eth_devices[vf_port];
+}
+
 void	hn_vf_info_get(struct hn_data *hv,
 		       struct rte_eth_dev_info *info);
 int	hn_vf_add(struct rte_eth_dev *dev, struct hn_data *hv);
diff --git a/drivers/net/netvsc/hn_vf.c b/drivers/net/netvsc/hn_vf.c
index 3f714ec99029..de278eb7b403 100644
--- a/drivers/net/netvsc/hn_vf.c
+++ b/drivers/net/netvsc/hn_vf.c
@@ -51,15 +51,20 @@  static int hn_vf_match(const struct rte_eth_dev *dev)
 	return -ENOENT;
 }
 
+
 /*
  * Attach new PCI VF device and return the port_id
  */
-static int hn_vf_attach(struct hn_data *hv, uint16_t port_id,
-			struct rte_eth_dev **vf_dev)
+static int hn_vf_attach(struct hn_data *hv, uint16_t port_id)
 {
 	struct rte_eth_dev_owner owner = { .id = RTE_ETH_DEV_NO_OWNER };
 	int ret;
 
+	if (hn_vf_attached(hv)) {
+		PMD_DRV_LOG(ERR, "VF already attached");
+		return -EEXIST;
+	}
+
 	ret = rte_eth_dev_owner_get(port_id, &owner);
 	if (ret < 0) {
 		PMD_DRV_LOG(ERR, "Can not find owner for port %d", port_id);
@@ -79,8 +84,9 @@  static int hn_vf_attach(struct hn_data *hv, uint16_t port_id,
 	}
 
 	PMD_DRV_LOG(DEBUG, "Attach VF device %u", port_id);
+	hv->vf_port = port_id;
 	rte_smp_wmb();
-	*vf_dev = &rte_eth_devices[port_id];
+
 	return 0;
 }
 
@@ -96,12 +102,7 @@  int hn_vf_add(struct rte_eth_dev *dev, struct hn_data *hv)
 	}
 
 	rte_spinlock_lock(&hv->vf_lock);
-	if (hv->vf_dev) {
-		PMD_DRV_LOG(ERR, "VF already attached");
-		err = -EBUSY;
-	} else {
-		err = hn_vf_attach(hv, port, &hv->vf_dev);
-	}
+	err = hn_vf_attach(hv, port);
 
 	if (err == 0) {
 		dev->data->dev_flags |= RTE_ETH_DEV_INTR_LSC;
@@ -120,22 +121,22 @@  int hn_vf_add(struct rte_eth_dev *dev, struct hn_data *hv)
 /* Remove new VF device */
 static void hn_vf_remove(struct hn_data *hv)
 {
-	struct rte_eth_dev *vf_dev;
 
 	rte_spinlock_lock(&hv->vf_lock);
-	vf_dev = hv->vf_dev;
-	if (!vf_dev) {
+
+	if (!hn_vf_attached(hv)) {
 		PMD_DRV_LOG(ERR, "VF path not active");
-		rte_spinlock_unlock(&hv->vf_lock);
-		return;
-	}
+	} else {
+		/* Stop incoming packets from arriving on VF */
+		hn_nvs_set_datapath(hv, NVS_DATAPATH_SYNTHETIC);
 
-	/* Stop incoming packets from arriving on VF */
-	hn_nvs_set_datapath(hv, NVS_DATAPATH_SYNTHETIC);
-	hv->vf_dev = NULL;
+		/* Stop transmission over VF */
+		hv->vf_port = HN_INVALID_PORT;
+		rte_smp_wmb();
 
-	/* Give back ownership */
-	rte_eth_dev_owner_unset(vf_dev->data->port_id, hv->owner.id);
+		/* Give back ownership */
+		rte_eth_dev_owner_unset(hv->vf_port, hv->owner.id);
+	}
 	rte_spinlock_unlock(&hv->vf_lock);
 }
 
@@ -207,7 +208,7 @@  void hn_vf_info_get(struct hn_data *hv, struct rte_eth_dev_info *info)
 	struct rte_eth_dev *vf_dev;
 
 	rte_spinlock_lock(&hv->vf_lock);
-	vf_dev = hv->vf_dev;
+	vf_dev = hn_get_vf_dev(hv);
 	if (vf_dev)
 		hn_vf_info_merge(vf_dev, info);
 	rte_spinlock_unlock(&hv->vf_lock);
@@ -221,7 +222,7 @@  int hn_vf_link_update(struct rte_eth_dev *dev,
 	int ret = 0;
 
 	rte_spinlock_lock(&hv->vf_lock);
-	vf_dev = hv->vf_dev;
+	vf_dev = hn_get_vf_dev(hv);
 	if (vf_dev && vf_dev->dev_ops->link_update)
 		ret = (*vf_dev->dev_ops->link_update)(vf_dev, wait_to_complete);
 	rte_spinlock_unlock(&hv->vf_lock);
@@ -249,13 +250,14 @@  static int hn_vf_lsc_event(uint16_t port_id __rte_unused,
 }
 
 static int _hn_vf_configure(struct rte_eth_dev *dev,
-			    struct rte_eth_dev *vf_dev,
+			    uint16_t vf_port,
 			    const struct rte_eth_conf *dev_conf)
 {
 	struct rte_eth_conf vf_conf = *dev_conf;
-	uint16_t vf_port = vf_dev->data->port_id;
+	struct rte_eth_dev *vf_dev;
 	int ret;
 
+	vf_dev = &rte_eth_devices[vf_port];
 	if (dev_conf->intr_conf.lsc &&
 	    (vf_dev->data->dev_flags & RTE_ETH_DEV_INTR_LSC)) {
 		PMD_DRV_LOG(DEBUG, "enabling LSC for VF %u",
@@ -294,13 +296,11 @@  int hn_vf_configure(struct rte_eth_dev *dev,
 		    const struct rte_eth_conf *dev_conf)
 {
 	struct hn_data *hv = dev->data->dev_private;
-	struct rte_eth_dev *vf_dev;
 	int ret = 0;
 
 	rte_spinlock_lock(&hv->vf_lock);
-	vf_dev = hv->vf_dev;
-	if (vf_dev)
-		ret = _hn_vf_configure(dev, vf_dev, dev_conf);
+	if (hv->vf_port != HN_INVALID_PORT)
+		ret = _hn_vf_configure(dev, hv->vf_port, dev_conf);
 	rte_spinlock_unlock(&hv->vf_lock);
 	return ret;
 }
@@ -312,7 +312,7 @@  const uint32_t *hn_vf_supported_ptypes(struct rte_eth_dev *dev)
 	const uint32_t *ptypes = NULL;
 
 	rte_spinlock_lock(&hv->vf_lock);
-	vf_dev = hv->vf_dev;
+	vf_dev = hn_get_vf_dev(hv);
 	if (vf_dev && vf_dev->dev_ops->dev_supported_ptypes_get)
 		ptypes = (*vf_dev->dev_ops->dev_supported_ptypes_get)(vf_dev);
 	rte_spinlock_unlock(&hv->vf_lock);
@@ -327,7 +327,7 @@  int hn_vf_start(struct rte_eth_dev *dev)
 	int ret = 0;
 
 	rte_spinlock_lock(&hv->vf_lock);
-	vf_dev = hv->vf_dev;
+	vf_dev = hn_get_vf_dev(hv);
 	if (vf_dev)
 		ret = rte_eth_dev_start(vf_dev->data->port_id);
 	rte_spinlock_unlock(&hv->vf_lock);
@@ -340,7 +340,7 @@  void hn_vf_stop(struct rte_eth_dev *dev)
 	struct rte_eth_dev *vf_dev;
 
 	rte_spinlock_lock(&hv->vf_lock);
-	vf_dev = hv->vf_dev;
+	vf_dev = hn_get_vf_dev(hv);
 	if (vf_dev)
 		rte_eth_dev_stop(vf_dev->data->port_id);
 	rte_spinlock_unlock(&hv->vf_lock);
@@ -352,7 +352,7 @@  void hn_vf_stop(struct rte_eth_dev *dev)
 		struct hn_data *hv = (dev)->data->dev_private;	\
 		struct rte_eth_dev *vf_dev;			\
 		rte_spinlock_lock(&hv->vf_lock);		\
-		vf_dev = hv->vf_dev;				\
+		vf_dev = hn_get_vf_dev(hv);			\
 		if (vf_dev)					\
 			func(vf_dev->data->port_id);		\
 		rte_spinlock_unlock(&hv->vf_lock);		\
@@ -402,7 +402,7 @@  int hn_vf_mc_addr_list(struct rte_eth_dev *dev,
 	int ret = 0;
 
 	rte_spinlock_lock(&hv->vf_lock);
-	vf_dev = hv->vf_dev;
+	vf_dev = hn_get_vf_dev(hv);
 	if (vf_dev)
 		ret = rte_eth_dev_set_mc_addr_list(vf_dev->data->port_id,
 						   mc_addr_set, nb_mc_addr);
@@ -420,7 +420,7 @@  int hn_vf_tx_queue_setup(struct rte_eth_dev *dev,
 	int ret = 0;
 
 	rte_spinlock_lock(&hv->vf_lock);
-	vf_dev = hv->vf_dev;
+	vf_dev = hn_get_vf_dev(hv);
 	if (vf_dev)
 		ret = rte_eth_tx_queue_setup(vf_dev->data->port_id,
 					     queue_idx, nb_desc,
@@ -434,7 +434,7 @@  void hn_vf_tx_queue_release(struct hn_data *hv, uint16_t queue_id)
 	struct rte_eth_dev *vf_dev;
 
 	rte_spinlock_lock(&hv->vf_lock);
-	vf_dev = hv->vf_dev;
+	vf_dev = hn_get_vf_dev(hv);
 	if (vf_dev && vf_dev->dev_ops->tx_queue_release) {
 		void *subq = vf_dev->data->tx_queues[queue_id];
 
@@ -455,7 +455,7 @@  int hn_vf_rx_queue_setup(struct rte_eth_dev *dev,
 	int ret = 0;
 
 	rte_spinlock_lock(&hv->vf_lock);
-	vf_dev = hv->vf_dev;
+	vf_dev = hn_get_vf_dev(hv);
 	if (vf_dev)
 		ret = rte_eth_rx_queue_setup(vf_dev->data->port_id,
 					     queue_idx, nb_desc,
@@ -469,7 +469,7 @@  void hn_vf_rx_queue_release(struct hn_data *hv, uint16_t queue_id)
 	struct rte_eth_dev *vf_dev;
 
 	rte_spinlock_lock(&hv->vf_lock);
-	vf_dev = hv->vf_dev;
+	vf_dev = hn_get_vf_dev(hv);
 	if (vf_dev && vf_dev->dev_ops->rx_queue_release) {
 		void *subq = vf_dev->data->rx_queues[queue_id];
 
@@ -486,7 +486,7 @@  int hn_vf_stats_get(struct rte_eth_dev *dev,
 	int ret = 0;
 
 	rte_spinlock_lock(&hv->vf_lock);
-	vf_dev = hv->vf_dev;
+	vf_dev = hn_get_vf_dev(hv);
 	if (vf_dev)
 		ret = rte_eth_stats_get(vf_dev->data->port_id, stats);
 	rte_spinlock_unlock(&hv->vf_lock);
@@ -503,7 +503,7 @@  int hn_vf_xstats_get_names(struct rte_eth_dev *dev,
 	char tmp[RTE_ETH_XSTATS_NAME_SIZE];
 
 	rte_spinlock_lock(&hv->vf_lock);
-	vf_dev = hv->vf_dev;
+	vf_dev = hn_get_vf_dev(hv);
 	if (vf_dev && vf_dev->dev_ops->xstats_get_names)
 		count = vf_dev->dev_ops->xstats_get_names(vf_dev, names, n);
 	rte_spinlock_unlock(&hv->vf_lock);
@@ -528,7 +528,7 @@  int hn_vf_xstats_get(struct rte_eth_dev *dev,
 	int count = 0;
 
 	rte_spinlock_lock(&hv->vf_lock);
-	vf_dev = hv->vf_dev;
+	vf_dev = hn_get_vf_dev(hv);
 	if (vf_dev && vf_dev->dev_ops->xstats_get)
 		count = vf_dev->dev_ops->xstats_get(vf_dev, xstats, n);
 	rte_spinlock_unlock(&hv->vf_lock);
@@ -542,7 +542,7 @@  void hn_vf_xstats_reset(struct rte_eth_dev *dev)
 	struct rte_eth_dev *vf_dev;
 
 	rte_spinlock_lock(&hv->vf_lock);
-	vf_dev = hv->vf_dev;
+	vf_dev = hn_get_vf_dev(hv);
 	if (vf_dev && vf_dev->dev_ops->xstats_reset)
 		vf_dev->dev_ops->xstats_reset(vf_dev);
 	rte_spinlock_unlock(&hv->vf_lock);