From e05273341c573f7b543f45c06e4a232c5b7c5a59 Mon Sep 17 00:00:00 2001 From: "David S. Miller" Date: Tue, 26 Jun 2012 21:43:19 -0700 Subject: infiniband: netlink: Move away from NLMSG_NEW(). And use nlmsg_data() while we're here too. Signed-off-by: David S. Miller --- drivers/infiniband/core/netlink.c | 10 ++++++---- 1 file changed, 6 insertions(+), 4 deletions(-) (limited to 'drivers/infiniband') diff --git a/drivers/infiniband/core/netlink.c b/drivers/infiniband/core/netlink.c index e497dfbee435..1e691dca1820 100644 --- a/drivers/infiniband/core/netlink.c +++ b/drivers/infiniband/core/netlink.c @@ -108,12 +108,14 @@ void *ibnl_put_msg(struct sk_buff *skb, struct nlmsghdr **nlh, int seq, unsigned char *prev_tail; prev_tail = skb_tail_pointer(skb); - *nlh = NLMSG_NEW(skb, 0, seq, RDMA_NL_GET_TYPE(client, op), - len, NLM_F_MULTI); + *nlh = nlmsg_put(skb, 0, seq, RDMA_NL_GET_TYPE(client, op), + len, NLM_F_MULTI); + if (!*nlh) + goto out_nlmsg_trim; (*nlh)->nlmsg_len = skb_tail_pointer(skb) - prev_tail; - return NLMSG_DATA(*nlh); + return nlmsg_data(*nlh); -nlmsg_failure: +out_nlmsg_trim: nlmsg_trim(skb, prev_tail); return NULL; } -- cgit v1.2.3 From a31f2d17b331db970259e875b7223d3aba7e3821 Mon Sep 17 00:00:00 2001 From: Pablo Neira Ayuso Date: Fri, 29 Jun 2012 06:15:21 +0000 Subject: netlink: add netlink_kernel_cfg parameter to netlink_kernel_create This patch adds the following structure: struct netlink_kernel_cfg { unsigned int groups; void (*input)(struct sk_buff *skb); struct mutex *cb_mutex; }; That can be passed to netlink_kernel_create to set optional configurations for netlink kernel sockets. I've populated this structure by looking for NULL and zero parameters at the existing code. The remaining parameters that always need to be set are still left in the original interface. That includes optional parameters for the netlink socket creation. This allows easy extensibility of this interface in the future. This patch also adapts all callers to use this new interface. Signed-off-by: Pablo Neira Ayuso Signed-off-by: David S. Miller --- drivers/infiniband/core/netlink.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) (limited to 'drivers/infiniband') diff --git a/drivers/infiniband/core/netlink.c b/drivers/infiniband/core/netlink.c index 1e691dca1820..3ae2bfd31015 100644 --- a/drivers/infiniband/core/netlink.c +++ b/drivers/infiniband/core/netlink.c @@ -173,8 +173,11 @@ static void ibnl_rcv(struct sk_buff *skb) int __init ibnl_init(void) { - nls = netlink_kernel_create(&init_net, NETLINK_RDMA, 0, ibnl_rcv, - NULL, THIS_MODULE); + struct netlink_kernel_cfg cfg = { + .input = ibnl_rcv, + }; + + nls = netlink_kernel_create(&init_net, NETLINK_RDMA, THIS_MODULE, &cfg); if (!nls) { pr_warn("Failed to create netlink socket\n"); return -ENOMEM; -- cgit v1.2.3 From 178709bbfe9d4fe432c272ed65a34b8582703c23 Mon Sep 17 00:00:00 2001 From: "David S. Miller" Date: Mon, 2 Jul 2012 22:00:03 -0700 Subject: ipoib: Convert over to dev_lookup_neigh_skb(). Signed-off-by: David S. Miller --- drivers/infiniband/ulp/ipoib/ipoib_main.c | 4 +++- drivers/infiniband/ulp/ipoib/ipoib_multicast.c | 22 +++++++++++++--------- 2 files changed, 16 insertions(+), 10 deletions(-) (limited to 'drivers/infiniband') diff --git a/drivers/infiniband/ulp/ipoib/ipoib_main.c b/drivers/infiniband/ulp/ipoib/ipoib_main.c index 3974c290b667..bbee4b2d7a13 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_main.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_main.c @@ -715,7 +715,7 @@ static int ipoib_start_xmit(struct sk_buff *skb, struct net_device *dev) rcu_read_lock(); if (likely(skb_dst(skb))) { - n = dst_get_neighbour_noref(skb_dst(skb)); + n = dst_neigh_lookup_skb(skb_dst(skb), skb); if (!n) { ++dev->stats.tx_dropped; dev_kfree_skb_any(skb); @@ -797,6 +797,8 @@ static int ipoib_start_xmit(struct sk_buff *skb, struct net_device *dev) } } unlock: + if (n) + neigh_release(n); rcu_read_unlock(); return NETDEV_TX_OK; } diff --git a/drivers/infiniband/ulp/ipoib/ipoib_multicast.c b/drivers/infiniband/ulp/ipoib/ipoib_multicast.c index 20ebc6fd1bb9..fbb95ee538b2 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_multicast.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_multicast.c @@ -720,16 +720,20 @@ out: rcu_read_lock(); if (dst) - n = dst_get_neighbour_noref(dst); - if (n && !*to_ipoib_neigh(n)) { - struct ipoib_neigh *neigh = ipoib_neigh_alloc(n, - skb->dev); - - if (neigh) { - kref_get(&mcast->ah->ref); - neigh->ah = mcast->ah; - list_add_tail(&neigh->list, &mcast->neigh_list); + n = dst_neigh_lookup_skb(dst, skb); + if (n) { + if (!*to_ipoib_neigh(n)) { + struct ipoib_neigh *neigh; + + neigh = ipoib_neigh_alloc(n, skb->dev); + if (neigh) { + kref_get(&mcast->ah->ref); + neigh->ah = mcast->ah; + list_add_tail(&neigh->list, + &mcast->neigh_list); + } } + neigh_release(n); } rcu_read_unlock(); spin_unlock_irqrestore(&priv->lock, flags); -- cgit v1.2.3 From 534cb283efef9fdbd9f70f4615054d26aa444dd6 Mon Sep 17 00:00:00 2001 From: "David S. Miller" Date: Mon, 2 Jul 2012 22:35:31 -0700 Subject: cxgb3: Convert t3_l2t_get() over to dst_neigh_lookup(). This means passing in a suitable destination address. Signed-off-by: David S. Miller --- drivers/infiniband/hw/cxgb3/iwch_cm.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) (limited to 'drivers/infiniband') diff --git a/drivers/infiniband/hw/cxgb3/iwch_cm.c b/drivers/infiniband/hw/cxgb3/iwch_cm.c index 740dcc065cf2..77b6b182778a 100644 --- a/drivers/infiniband/hw/cxgb3/iwch_cm.c +++ b/drivers/infiniband/hw/cxgb3/iwch_cm.c @@ -1374,7 +1374,7 @@ static int pass_accept_req(struct t3cdev *tdev, struct sk_buff *skb, void *ctx) goto reject; } dst = &rt->dst; - l2t = t3_l2t_get(tdev, dst, NULL); + l2t = t3_l2t_get(tdev, dst, NULL, &req->peer_ip); if (!l2t) { printk(KERN_ERR MOD "%s - failed to allocate l2t entry!\n", __func__); @@ -1942,7 +1942,8 @@ int iwch_connect(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param) goto fail3; } ep->dst = &rt->dst; - ep->l2t = t3_l2t_get(ep->com.tdev, ep->dst, NULL); + ep->l2t = t3_l2t_get(ep->com.tdev, ep->dst, NULL, + &cm_id->remote_addr.sin_addr.s_addr); if (!ep->l2t) { printk(KERN_ERR MOD "%s - cannot alloc l2e.\n", __func__); err = -ENOMEM; -- cgit v1.2.3 From 700db99d0140e9da2a31e08ebd3e1b121691aa26 Mon Sep 17 00:00:00 2001 From: "David S. Miller" Date: Thu, 5 Jul 2012 21:08:05 -0700 Subject: ipoib: Need to do dst_neigh_lookup_skb() outside of priv->lock. Otherwise local_bh_enable() complains. Reported-by: Or Gerlitz Signed-off-by: David S. Miller --- drivers/infiniband/ulp/ipoib/ipoib_multicast.c | 15 ++++++++------- 1 file changed, 8 insertions(+), 7 deletions(-) (limited to 'drivers/infiniband') diff --git a/drivers/infiniband/ulp/ipoib/ipoib_multicast.c b/drivers/infiniband/ulp/ipoib/ipoib_multicast.c index fbb95ee538b2..7cecb16d3d48 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_multicast.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_multicast.c @@ -658,9 +658,15 @@ static int ipoib_mcast_leave(struct net_device *dev, struct ipoib_mcast *mcast) void ipoib_mcast_send(struct net_device *dev, void *mgid, struct sk_buff *skb) { struct ipoib_dev_priv *priv = netdev_priv(dev); + struct dst_entry *dst = skb_dst(skb); struct ipoib_mcast *mcast; + struct neighbour *n; unsigned long flags; + n = NULL; + if (dst) + n = dst_neigh_lookup_skb(dst, skb); + spin_lock_irqsave(&priv->lock, flags); if (!test_bit(IPOIB_FLAG_OPER_UP, &priv->flags) || @@ -715,12 +721,6 @@ void ipoib_mcast_send(struct net_device *dev, void *mgid, struct sk_buff *skb) out: if (mcast && mcast->ah) { - struct dst_entry *dst = skb_dst(skb); - struct neighbour *n = NULL; - - rcu_read_lock(); - if (dst) - n = dst_neigh_lookup_skb(dst, skb); if (n) { if (!*to_ipoib_neigh(n)) { struct ipoib_neigh *neigh; @@ -735,13 +735,14 @@ out: } neigh_release(n); } - rcu_read_unlock(); spin_unlock_irqrestore(&priv->lock, flags); ipoib_send(dev, skb, mcast->ah, IB_MULTICAST_QPN); return; } unlock: + if (n) + neigh_release(n); spin_unlock_irqrestore(&priv->lock, flags); } -- cgit v1.2.3 From 0ff1fb654bec0cff62ddf81a8a8edec4263604a0 Mon Sep 17 00:00:00 2001 From: Hadar Hen Zion Date: Thu, 5 Jul 2012 04:03:46 +0000 Subject: {NET, IB}/mlx4: Add device managed flow steering firmware API The driver is modified to support three operation modes. If supported by firmware use the device managed flow steering API, that which we call device managed steering mode. Else, if the firmware supports the B0 steering mode use it, and finally, if none of the above, use the A0 steering mode. When the steering mode is device managed, the code is modified such that L2 based rules set by the mlx4_en driver for Ethernet unicast and multicast, and the IB stack multicast attach calls done through the mlx4_ib driver are all routed to use the device managed API. When attaching rule using device managed flow steering API, the firmware returns a 64 bit registration id, which is to be provided during detach. Currently the firmware is always programmed during HCA initialization to use standard L2 hashing. Future work should be done to allow configuring the flow-steering hash function with common, non proprietary means. Signed-off-by: Hadar Hen Zion Signed-off-by: Or Gerlitz Signed-off-by: David S. Miller --- drivers/infiniband/hw/mlx4/main.c | 62 ++++++++++++++++++++++++++++++++---- drivers/infiniband/hw/mlx4/mlx4_ib.h | 1 + drivers/infiniband/hw/mlx4/qp.c | 1 + 3 files changed, 57 insertions(+), 7 deletions(-) (limited to 'drivers/infiniband') diff --git a/drivers/infiniband/hw/mlx4/main.c b/drivers/infiniband/hw/mlx4/main.c index 3530c41fcd1f..8a3a2037b005 100644 --- a/drivers/infiniband/hw/mlx4/main.c +++ b/drivers/infiniband/hw/mlx4/main.c @@ -718,26 +718,53 @@ int mlx4_ib_add_mc(struct mlx4_ib_dev *mdev, struct mlx4_ib_qp *mqp, return ret; } +struct mlx4_ib_steering { + struct list_head list; + u64 reg_id; + union ib_gid gid; +}; + static int mlx4_ib_mcg_attach(struct ib_qp *ibqp, union ib_gid *gid, u16 lid) { int err; struct mlx4_ib_dev *mdev = to_mdev(ibqp->device); struct mlx4_ib_qp *mqp = to_mqp(ibqp); + u64 reg_id; + struct mlx4_ib_steering *ib_steering = NULL; + + if (mdev->dev->caps.steering_mode == + MLX4_STEERING_MODE_DEVICE_MANAGED) { + ib_steering = kmalloc(sizeof(*ib_steering), GFP_KERNEL); + if (!ib_steering) + return -ENOMEM; + } - err = mlx4_multicast_attach(mdev->dev, &mqp->mqp, gid->raw, - !!(mqp->flags & MLX4_IB_QP_BLOCK_MULTICAST_LOOPBACK), - MLX4_PROT_IB_IPV6); + err = mlx4_multicast_attach(mdev->dev, &mqp->mqp, gid->raw, mqp->port, + !!(mqp->flags & + MLX4_IB_QP_BLOCK_MULTICAST_LOOPBACK), + MLX4_PROT_IB_IPV6, ®_id); if (err) - return err; + goto err_malloc; err = add_gid_entry(ibqp, gid); if (err) goto err_add; + if (ib_steering) { + memcpy(ib_steering->gid.raw, gid->raw, 16); + ib_steering->reg_id = reg_id; + mutex_lock(&mqp->mutex); + list_add(&ib_steering->list, &mqp->steering_rules); + mutex_unlock(&mqp->mutex); + } return 0; err_add: - mlx4_multicast_detach(mdev->dev, &mqp->mqp, gid->raw, MLX4_PROT_IB_IPV6); + mlx4_multicast_detach(mdev->dev, &mqp->mqp, gid->raw, + MLX4_PROT_IB_IPV6, reg_id); +err_malloc: + kfree(ib_steering); + return err; } @@ -765,9 +792,30 @@ static int mlx4_ib_mcg_detach(struct ib_qp *ibqp, union ib_gid *gid, u16 lid) u8 mac[6]; struct net_device *ndev; struct mlx4_ib_gid_entry *ge; + u64 reg_id = 0; + + if (mdev->dev->caps.steering_mode == + MLX4_STEERING_MODE_DEVICE_MANAGED) { + struct mlx4_ib_steering *ib_steering; + + mutex_lock(&mqp->mutex); + list_for_each_entry(ib_steering, &mqp->steering_rules, list) { + if (!memcmp(ib_steering->gid.raw, gid->raw, 16)) { + list_del(&ib_steering->list); + break; + } + } + mutex_unlock(&mqp->mutex); + if (&ib_steering->list == &mqp->steering_rules) { + pr_err("Couldn't find reg_id for mgid. Steering rule is left attached\n"); + return -EINVAL; + } + reg_id = ib_steering->reg_id; + kfree(ib_steering); + } - err = mlx4_multicast_detach(mdev->dev, - &mqp->mqp, gid->raw, MLX4_PROT_IB_IPV6); + err = mlx4_multicast_detach(mdev->dev, &mqp->mqp, gid->raw, + MLX4_PROT_IB_IPV6, reg_id); if (err) return err; diff --git a/drivers/infiniband/hw/mlx4/mlx4_ib.h b/drivers/infiniband/hw/mlx4/mlx4_ib.h index ff36655d23d3..42df4f7a6a5b 100644 --- a/drivers/infiniband/hw/mlx4/mlx4_ib.h +++ b/drivers/infiniband/hw/mlx4/mlx4_ib.h @@ -163,6 +163,7 @@ struct mlx4_ib_qp { u8 state; int mlx_type; struct list_head gid_list; + struct list_head steering_rules; }; struct mlx4_ib_srq { diff --git a/drivers/infiniband/hw/mlx4/qp.c b/drivers/infiniband/hw/mlx4/qp.c index 8d4ed24aef93..6af19f6c2b11 100644 --- a/drivers/infiniband/hw/mlx4/qp.c +++ b/drivers/infiniband/hw/mlx4/qp.c @@ -495,6 +495,7 @@ static int create_qp_common(struct mlx4_ib_dev *dev, struct ib_pd *pd, spin_lock_init(&qp->sq.lock); spin_lock_init(&qp->rq.lock); INIT_LIST_HEAD(&qp->gid_list); + INIT_LIST_HEAD(&qp->steering_rules); qp->state = IB_QPS_RESET; if (init_attr->sq_sig_type == IB_SIGNAL_ALL_WR) -- cgit v1.2.3 From 6700c2709c08d74ae2c3c29b84a30da012dbc7f1 Mon Sep 17 00:00:00 2001 From: "David S. Miller" Date: Tue, 17 Jul 2012 03:29:28 -0700 Subject: net: Pass optional SKB and SK arguments to dst_ops->{update_pmtu,redirect}() This will be used so that we can compose a full flow key. Even though we have a route in this context, we need more. In the future the routes will be without destination address, source address, etc. keying. One ipv4 route will cover entire subnets, etc. In this environment we have to have a way to possess persistent storage for redirects and PMTU information. This persistent storage will exist in the FIB tables, and that's why we'll need to be able to rebuild a full lookup flow key here. Using that flow key will do a fib_lookup() and create/update the persistent entry. Signed-off-by: David S. Miller --- drivers/infiniband/ulp/ipoib/ipoib_cm.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'drivers/infiniband') diff --git a/drivers/infiniband/ulp/ipoib/ipoib_cm.c b/drivers/infiniband/ulp/ipoib/ipoib_cm.c index 014504d8e43c..1ca732201f33 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_cm.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_cm.c @@ -1397,7 +1397,7 @@ void ipoib_cm_skb_too_long(struct net_device *dev, struct sk_buff *skb, int e = skb_queue_empty(&priv->cm.skb_queue); if (skb_dst(skb)) - skb_dst(skb)->ops->update_pmtu(skb_dst(skb), mtu); + skb_dst(skb)->ops->update_pmtu(skb_dst(skb), NULL, skb, mtu); skb_queue_tail(&priv->cm.skb_queue, skb); if (e) -- cgit v1.2.3 From d9236c3f10490cd0b3fd4516af12ba62dcbae0b0 Mon Sep 17 00:00:00 2001 From: Amir Vadai Date: Wed, 18 Jul 2012 22:33:51 +0000 Subject: {NET,IB}/mlx4: Add rmap support to mlx4_assign_eq Enable callers of mlx4_assign_eq to supply a pointer to cpu_rmap. If supplied, the assigned IRQ is tracked using rmap infrastructure. Signed-off-by: Amir Vadai Signed-off-by: Or Gerlitz Signed-off-by: David S. Miller --- drivers/infiniband/hw/mlx4/main.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'drivers/infiniband') diff --git a/drivers/infiniband/hw/mlx4/main.c b/drivers/infiniband/hw/mlx4/main.c index 8a3a2037b005..a07b774e7864 100644 --- a/drivers/infiniband/hw/mlx4/main.c +++ b/drivers/infiniband/hw/mlx4/main.c @@ -1159,7 +1159,8 @@ static void mlx4_ib_alloc_eqs(struct mlx4_dev *dev, struct mlx4_ib_dev *ibdev) sprintf(name, "mlx4-ib-%d-%d@%s", i, j, dev->pdev->bus->name); /* Set IRQ for specific name (per ring) */ - if (mlx4_assign_eq(dev, name, &ibdev->eq_table[eq])) { + if (mlx4_assign_eq(dev, name, NULL, + &ibdev->eq_table[eq])) { /* Use legacy (same as mlx4_en driver) */ pr_warn("Can't allocate EQ %d; reverting to legacy\n", eq); ibdev->eq_table[eq] = -- cgit v1.2.3