diff options
author | Linus Torvalds <torvalds@linux-foundation.org> | 2009-09-11 08:58:32 -0700 |
---|---|---|
committer | Linus Torvalds <torvalds@linux-foundation.org> | 2009-09-11 08:58:32 -0700 |
commit | 2490138cb785d299d898b579fa2874a59a3d321a (patch) | |
tree | fc9acc19ce1e089e383f901e4eaa406a739f4837 /drivers | |
parent | f6f79190866d5b2d06a2114d673f91f54e7c7ce4 (diff) | |
parent | 73f526da0260db5376951373c267596993dc13a8 (diff) |
Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband: (48 commits)
RDMA/iwcm: Reject the connection when the cm_id is destroyed
RDMA/cxgb3: Clean up properly on FW mismatch failures
RDMA/cxgb3: Don't ignore insert_handle() failures
MAINTAINERS: InfiniBand/RDMA mailing list transition to vger
IB/mad: Allow tuning of QP0 and QP1 sizes
IB/mad: Fix possible lock-lock-timer deadlock
RDMA/nes: Map MTU to IB_MTU_* and correctly report link state
RDMA/nes: Rework the disconn routine for terminate and flushing
RDMA/nes: Use the flush code to fill in cqe error
RDMA/nes: Make poll_cq return correct number of wqes during flush
RDMA/nes: Use flush mechanism to set status for wqe in error
RDMA/nes: Implement Terminate Packet
RDMA/nes: Add CQ error handling
RDMA/nes: Clean out CQ completions when QP is destroyed
RDMA/nes: Change memory allocation for cqp request to GFP_ATOMIC
RDMA/nes: Allocate work item for disconnect event handling
RDMA/nes: Update refcnt during disconnect
IB/mthca: Don't allow userspace open while recovering from catastrophic error
IB/mthca: Distinguish multiple devices in /proc/interrupts
IB/mthca: Annotate CQ locking
...
Diffstat (limited to 'drivers')
62 files changed, 1272 insertions, 589 deletions
diff --git a/drivers/infiniband/core/iwcm.c b/drivers/infiniband/core/iwcm.c index 8f9509e1ebf..55d093a36ae 100644 --- a/drivers/infiniband/core/iwcm.c +++ b/drivers/infiniband/core/iwcm.c @@ -362,6 +362,7 @@ static void destroy_cm_id(struct iw_cm_id *cm_id) * In either case, must tell the provider to reject. */ cm_id_priv->state = IW_CM_STATE_DESTROYING; + cm_id->device->iwcm->reject(cm_id, NULL, 0); break; case IW_CM_STATE_CONN_SENT: case IW_CM_STATE_DESTROYING: diff --git a/drivers/infiniband/core/mad.c b/drivers/infiniband/core/mad.c index de922a04ca2..7522008fda8 100644 --- a/drivers/infiniband/core/mad.c +++ b/drivers/infiniband/core/mad.c @@ -2,6 +2,7 @@ * Copyright (c) 2004-2007 Voltaire, Inc. All rights reserved. * Copyright (c) 2005 Intel Corporation. All rights reserved. * Copyright (c) 2005 Mellanox Technologies Ltd. All rights reserved. + * Copyright (c) 2009 HNR Consulting. All rights reserved. * * This software is available to you under a choice of one of two * licenses. You may choose to be licensed under the terms of the GNU @@ -45,14 +46,21 @@ MODULE_DESCRIPTION("kernel IB MAD API"); MODULE_AUTHOR("Hal Rosenstock"); MODULE_AUTHOR("Sean Hefty"); +int mad_sendq_size = IB_MAD_QP_SEND_SIZE; +int mad_recvq_size = IB_MAD_QP_RECV_SIZE; + +module_param_named(send_queue_size, mad_sendq_size, int, 0444); +MODULE_PARM_DESC(send_queue_size, "Size of send queue in number of work requests"); +module_param_named(recv_queue_size, mad_recvq_size, int, 0444); +MODULE_PARM_DESC(recv_queue_size, "Size of receive queue in number of work requests"); + static struct kmem_cache *ib_mad_cache; static struct list_head ib_mad_port_list; static u32 ib_mad_client_id = 0; /* Port list lock */ -static spinlock_t ib_mad_port_list_lock; - +static DEFINE_SPINLOCK(ib_mad_port_list_lock); /* Forward declarations */ static int method_in_use(struct ib_mad_mgmt_method_table **method, @@ -1974,7 +1982,7 @@ static void adjust_timeout(struct ib_mad_agent_private *mad_agent_priv) unsigned long delay; if (list_empty(&mad_agent_priv->wait_list)) { - cancel_delayed_work(&mad_agent_priv->timed_work); + __cancel_delayed_work(&mad_agent_priv->timed_work); } else { mad_send_wr = list_entry(mad_agent_priv->wait_list.next, struct ib_mad_send_wr_private, @@ -1983,7 +1991,7 @@ static void adjust_timeout(struct ib_mad_agent_private *mad_agent_priv) if (time_after(mad_agent_priv->timeout, mad_send_wr->timeout)) { mad_agent_priv->timeout = mad_send_wr->timeout; - cancel_delayed_work(&mad_agent_priv->timed_work); + __cancel_delayed_work(&mad_agent_priv->timed_work); delay = mad_send_wr->timeout - jiffies; if ((long)delay <= 0) delay = 1; @@ -2023,7 +2031,7 @@ static void wait_for_response(struct ib_mad_send_wr_private *mad_send_wr) /* Reschedule a work item if we have a shorter timeout */ if (mad_agent_priv->wait_list.next == &mad_send_wr->agent_list) { - cancel_delayed_work(&mad_agent_priv->timed_work); + __cancel_delayed_work(&mad_agent_priv->timed_work); queue_delayed_work(mad_agent_priv->qp_info->port_priv->wq, &mad_agent_priv->timed_work, delay); } @@ -2736,8 +2744,8 @@ static int create_mad_qp(struct ib_mad_qp_info *qp_info, qp_init_attr.send_cq = qp_info->port_priv->cq; qp_init_attr.recv_cq = qp_info->port_priv->cq; qp_init_attr.sq_sig_type = IB_SIGNAL_ALL_WR; - qp_init_attr.cap.max_send_wr = IB_MAD_QP_SEND_SIZE; - qp_init_attr.cap.max_recv_wr = IB_MAD_QP_RECV_SIZE; + qp_init_attr.cap.max_send_wr = mad_sendq_size; + qp_init_attr.cap.max_recv_wr = mad_recvq_size; qp_init_attr.cap.max_send_sge = IB_MAD_SEND_REQ_MAX_SG; qp_init_attr.cap.max_recv_sge = IB_MAD_RECV_REQ_MAX_SG; qp_init_attr.qp_type = qp_type; @@ -2752,8 +2760,8 @@ static int create_mad_qp(struct ib_mad_qp_info *qp_info, goto error; } /* Use minimum queue sizes unless the CQ is resized */ - qp_info->send_queue.max_active = IB_MAD_QP_SEND_SIZE; - qp_info->recv_queue.max_active = IB_MAD_QP_RECV_SIZE; + qp_info->send_queue.max_active = mad_sendq_size; + qp_info->recv_queue.max_active = mad_recvq_size; return 0; error: @@ -2792,7 +2800,7 @@ static int ib_mad_port_open(struct ib_device *device, init_mad_qp(port_priv, &port_priv->qp_info[0]); init_mad_qp(port_priv, &port_priv->qp_info[1]); - cq_size = (IB_MAD_QP_SEND_SIZE + IB_MAD_QP_RECV_SIZE) * 2; + cq_size = (mad_sendq_size + mad_recvq_size) * 2; port_priv->cq = ib_create_cq(port_priv->device, ib_mad_thread_completion_handler, NULL, port_priv, cq_size, 0); @@ -2984,7 +2992,11 @@ static int __init ib_mad_init_module(void) { int ret; - spin_lock_init(&ib_mad_port_list_lock); + mad_recvq_size = min(mad_recvq_size, IB_MAD_QP_MAX_SIZE); + mad_recvq_size = max(mad_recvq_size, IB_MAD_QP_MIN_SIZE); + + mad_sendq_size = min(mad_sendq_size, IB_MAD_QP_MAX_SIZE); + mad_sendq_size = max(mad_sendq_size, IB_MAD_QP_MIN_SIZE); ib_mad_cache = kmem_cache_create("ib_mad", sizeof(struct ib_mad_private), @@ -3021,4 +3033,3 @@ static void __exit ib_mad_cleanup_module(void) module_init(ib_mad_init_module); module_exit(ib_mad_cleanup_module); - diff --git a/drivers/infiniband/core/mad_priv.h b/drivers/infiniband/core/mad_priv.h index 05ce331733b..9430ab4969c 100644 --- a/drivers/infiniband/core/mad_priv.h +++ b/drivers/infiniband/core/mad_priv.h @@ -2,6 +2,7 @@ * Copyright (c) 2004, 2005, Voltaire, Inc. All rights reserved. * Copyright (c) 2005 Intel Corporation. All rights reserved. * Copyright (c) 2005 Sun Microsystems, Inc. All rights reserved. + * Copyright (c) 2009 HNR Consulting. All rights reserved. * * This software is available to you under a choice of one of two * licenses. You may choose to be licensed under the terms of the GNU @@ -49,6 +50,8 @@ /* QP and CQ parameters */ #define IB_MAD_QP_SEND_SIZE 128 #define IB_MAD_QP_RECV_SIZE 512 +#define IB_MAD_QP_MIN_SIZE 64 +#define IB_MAD_QP_MAX_SIZE 8192 #define IB_MAD_SEND_REQ_MAX_SG 2 #define IB_MAD_RECV_REQ_MAX_SG 1 diff --git a/drivers/infiniband/core/multicast.c b/drivers/infiniband/core/multicast.c index 107f170c57c..8d82ba17135 100644 --- a/drivers/infiniband/core/multicast.c +++ b/drivers/infiniband/core/multicast.c @@ -106,6 +106,8 @@ struct mcast_group { struct ib_sa_query *query; int query_id; u16 pkey_index; + u8 leave_state; + int retries; }; struct mcast_member { @@ -350,6 +352,7 @@ static int send_leave(struct mcast_group *group, u8 leave_state) rec = group->rec; rec.join_state = leave_state; + group->leave_state = leave_state; ret = ib_sa_mcmember_rec_query(&sa_client, port->dev->device, port->port_num, IB_SA_METHOD_DELETE, &rec, @@ -542,7 +545,11 @@ static void leave_handler(int status, struct ib_sa_mcmember_rec *rec, { struct mcast_group *group = context; - mcast_work_handler(&group->work); + if (status && group->retries > 0 && + !send_leave(group, group->leave_state)) + group->retries--; + else + mcast_work_handler(&group->work); } static struct mcast_group *acquire_group(struct mcast_port *port, @@ -565,6 +572,7 @@ static struct mcast_group *acquire_group(struct mcast_port *port, if (!group) return NULL; + group->retries = 3; group->port = port; group->rec.mgid = *mgid; group->pkey_index = MCAST_INVALID_PKEY_INDEX; diff --git a/drivers/infiniband/core/sa_query.c b/drivers/infiniband/core/sa_query.c index 1865049e80f..82543716d59 100644 --- a/drivers/infiniband/core/sa_query.c +++ b/drivers/infiniband/core/sa_query.c @@ -109,10 +109,10 @@ static struct ib_client sa_client = { .remove = ib_sa_remove_one }; -static spinlock_t idr_lock; +static DEFINE_SPINLOCK(idr_lock); static DEFINE_IDR(query_idr); -static spinlock_t tid_lock; +static DEFINE_SPINLOCK(tid_lock); static u32 tid; #define PATH_REC_FIELD(field) \ @@ -1077,9 +1077,6 @@ static int __init ib_sa_init(void) { int ret; - spin_lock_init(&idr_lock); - spin_lock_init(&tid_lock); - get_random_bytes(&tid, sizeof tid); ret = ib_register_client(&sa_client); diff --git a/drivers/infiniband/core/smi.c b/drivers/infiniband/core/smi.c index 87236753bce..5855e4405d9 100644 --- a/drivers/infiniband/core/smi.c +++ b/drivers/infiniband/core/smi.c @@ -52,6 +52,10 @@ enum smi_action smi_handle_dr_smp_send(struct ib_smp *smp, hop_cnt = smp->hop_cnt; /* See section 14.2.2.2, Vol 1 IB spec */ + /* C14-6 -- valid hop_cnt values are from 0 to 63 */ + if (hop_cnt >= IB_SMP_MAX_PATH_HOPS) + return IB_SMI_DISCARD; + if (!ib_get_smp_direction(smp)) { /* C14-9:1 */ if (hop_cnt && hop_ptr == 0) { @@ -133,6 +137,10 @@ enum smi_action smi_handle_dr_smp_recv(struct ib_smp *smp, u8 node_type, hop_cnt = smp->hop_cnt; /* See section 14.2.2.2, Vol 1 IB spec */ + /* C14-6 -- valid hop_cnt values are from 0 to 63 */ + if (hop_cnt >= IB_SMP_MAX_PATH_HOPS) + return IB_SMI_DISCARD; + if (!ib_get_smp_direction(smp)) { /* C14-9:1 -- sender should have incremented hop_ptr */ if (hop_cnt && hop_ptr == 0) diff --git a/drivers/infiniband/core/uverbs_main.c b/drivers/infiniband/core/uverbs_main.c index eb36a81dd09..d3fff9e008a 100644 --- a/drivers/infiniband/core/uverbs_main.c +++ b/drivers/infiniband/core/uverbs_main.c @@ -73,7 +73,7 @@ DEFINE_IDR(ib_uverbs_cq_idr); DEFINE_IDR(ib_uverbs_qp_idr); DEFINE_IDR(ib_uverbs_srq_idr); -static spinlock_t map_lock; +static DEFINE_SPINLOCK(map_lock); static struct ib_uverbs_device *dev_table[IB_UVERBS_MAX_DEVICES]; static DECLARE_BITMAP(dev_map, IB_UVERBS_MAX_DEVICES); @@ -584,14 +584,16 @@ static ssize_t ib_uverbs_write(struct file *filp, const char __user *buf, if (hdr.command < 0 || hdr.command >= ARRAY_SIZE(uverbs_cmd_table) || - !uverbs_cmd_table[hdr.command] || - !(file->device->ib_dev->uverbs_cmd_mask & (1ull << hdr.command))) + !uverbs_cmd_table[hdr.command]) return -EINVAL; if (!file->ucontext && hdr.command != IB_USER_VERBS_CMD_GET_CONTEXT) return -EINVAL; + if (!(file->device->ib_dev->uverbs_cmd_mask & (1ull << hdr.command))) + return -ENOSYS; + return uverbs_cmd_table[hdr.command](file, buf + sizeof hdr, hdr.in_words * 4, hdr.out_words * 4); } @@ -836,8 +838,6 @@ static int __init ib_uverbs_init(void) { int ret; - spin_lock_init(&map_lock); - ret = register_chrdev_region(IB_UVERBS_BASE_DEV, IB_UVERBS_MAX_DEVICES, "infiniband_verbs"); if (ret) { diff --git a/drivers/infiniband/hw/amso1100/c2.c b/drivers/infiniband/hw/amso1100/c2.c index 0cfbb6d2f76..8250740c94b 100644 --- a/drivers/infiniband/hw/amso1100/c2.c +++ b/drivers/infiniband/hw/amso1100/c2.c @@ -86,11 +86,7 @@ MODULE_DEVICE_TABLE(pci, c2_pci_table); static void c2_print_macaddr(struct net_device *netdev) { - pr_debug("%s: MAC %02X:%02X:%02X:%02X:%02X:%02X, " - "IRQ %u\n", netdev->name, - netdev->dev_addr[0], netdev->dev_addr[1], netdev->dev_addr[2], - netdev->dev_addr[3], netdev->dev_addr[4], netdev->dev_addr[5], - netdev->irq); + pr_debug("%s: MAC %pM, IRQ %u\n", netdev->name, netdev->dev_addr, netdev->irq); } static void c2_set_rxbufsize(struct c2_port *c2_port) diff --git a/drivers/infiniband/hw/amso1100/c2_provider.c b/drivers/infiniband/hw/amso1100/c2_provider.c index f1948fad85d..ad723bd8bf4 100644 --- a/drivers/infiniband/hw/amso1100/c2_provider.c +++ b/drivers/infiniband/hw/amso1100/c2_provider.c @@ -780,11 +780,11 @@ int c2_register_device(struct c2_dev *dev) /* Register pseudo network device */ dev->pseudo_netdev = c2_pseudo_netdev_init(dev); if (!dev->pseudo_netdev) - goto out3; + goto out; ret = register_netdev(dev->pseudo_netdev); if (ret) - goto out2; + goto out_free_netdev; pr_debug("%s:%u\n", __func__, __LINE__); strlcpy(dev->ibdev.name, "amso%d", IB_DEVICE_NAME_MAX); @@ -851,6 +851,10 @@ int c2_register_device(struct c2_dev *dev) dev->ibdev.post_recv = c2_post_receive; dev->ibdev.iwcm = kmalloc(sizeof(*dev->ibdev.iwcm), GFP_KERNEL); + if (dev->ibdev.iwcm == NULL) { + ret = -ENOMEM; + goto out_unregister_netdev; + } dev->ibdev.iwcm->add_ref = c2_add_ref; dev->ibdev.iwcm->rem_ref = c2_rem_ref; dev->ibdev.iwcm->get_qp = c2_get_qp; @@ -862,23 +866,25 @@ int c2_register_device(struct c2_dev *dev) ret = ib_register_device(&dev->ibdev); if (ret) - goto out1; + goto out_free_iwcm; for (i = 0; i < ARRAY_SIZE(c2_dev_attributes); ++i) { ret = device_create_file(&dev->ibdev.dev, c2_dev_attributes[i]); if (ret) - goto out0; + goto out_unregister_ibdev; } - goto out3; + goto out; -out0: +out_unregister_ibdev: ib_unregister_device(&dev->ibdev); -out1: +out_free_iwcm: + kfree(dev->ibdev.iwcm); +out_unregister_netdev: unregister_netdev(dev->pseudo_netdev); -out2: +out_free_netdev: free_netdev(dev->pseudo_netdev); -out3: +out: pr_debug("%s:%u ret=%d\n", __func__, __LINE__, ret); return ret; } diff --git a/drivers/infiniband/hw/cxgb3/cxio_hal.c b/drivers/infiniband/hw/cxgb3/cxio_hal.c index 62f9cf2f94e..72ed3396b72 100644 --- a/drivers/infiniband/hw/cxgb3/cxio_hal.c +++ b/drivers/infiniband/hw/cxgb3/cxio_hal.c @@ -852,7 +852,9 @@ int cxio_rdma_init(struct cxio_rdev *rdev_p, struct t3_rdma_init_attr *attr) wqe->qpcaps = attr->qpcaps; wqe->ulpdu_size = cpu_to_be16(attr->tcp_emss); wqe->rqe_count = cpu_to_be16(attr->rqe_count); - wqe->flags_rtr_type = cpu_to_be16(attr->flags|V_RTR_TYPE(attr->rtr_type)); + wqe->flags_rtr_type = cpu_to_be16(attr->flags | + V_RTR_TYPE(attr->rtr_type) | + V_CHAN(attr->chan)); wqe->ord = cpu_to_be32(attr->ord); wqe->ird = cpu_to_be32(attr->ird); wqe->qp_dma_addr = cpu_to_be64(attr->qp_dma_addr); @@ -1032,6 +1034,7 @@ err3: err2: cxio_hal_destroy_ctrl_qp(rdev_p); err1: + rdev_p->t3cdev_p->ulp = NULL; list_del(&rdev_p->entry); return err; } diff --git a/drivers/infiniband/hw/cxgb3/cxio_wr.h b/drivers/infiniband/hw/cxgb3/cxio_wr.h index 32e3b1461d8..a197a5b7ac7 100644 --- a/drivers/infiniband/hw/cxgb3/cxio_wr.h +++ b/drivers/infiniband/hw/cxgb3/cxio_wr.h @@ -327,6 +327,11 @@ enum rdma_init_rtr_types { #define V_RTR_TYPE(x) ((x) << S_RTR_TYPE) #define G_RTR_TYPE(x) ((((x) >> S_RTR_TYPE)) & M_RTR_TYPE) +#define S_CHAN 4 +#define M_CHAN 0x3 +#define V_CHAN(x) ((x) << S_CHAN) +#define G_CHAN(x) ((((x) >> S_CHAN)) & M_CHAN) + struct t3_rdma_init_attr { u32 tid; u32 qpid; @@ -346,6 +351,7 @@ struct t3_rdma_init_attr { u16 flags; u16 rqe_count; u32 irs; + u32 chan; }; struct t3_rdma_init_wr { diff --git a/drivers/infiniband/hw/cxgb3/iwch.c b/drivers/infiniband/hw/cxgb3/iwch.c index 26fc0a4eaa7..b0ea0105ddf 100644 --- a/drivers/infiniband/hw/cxgb3/iwch.c +++ b/drivers/infiniband/hw/cxgb3/iwch.c @@ -51,7 +51,7 @@ cxgb3_cpl_handler_func t3c_handlers[NUM_CPL_CMDS]; static void open_rnic_dev(struct t3cdev *); static void close_rnic_dev(struct t3cdev *); -static void iwch_err_handler(struct t3cdev *, u32, u32); +static void iwch_event_handler(struct t3cdev *, u32, u32); struct cxgb3_client t3c_client = { .name = "iw_cxgb3", @@ -59,7 +59,7 @@ struct cxgb3_client t3c_client = { .remove = close_rnic_dev, .handlers = t3c_handlers, .redirect = iwch_ep_redirect, - .err_handler = iwch_err_handler + .event_handler = iwch_event_handler }; static LIST_HEAD(dev_list); @@ -105,11 +105,9 @@ static void rnic_init(struct iwch_dev *rnicp) static void open_rnic_dev(struct t3cdev *tdev) { struct iwch_dev *rnicp; - static int vers_printed; PDBG("%s t3cdev %p\n", __func__, tdev); - if (!vers_printed++) - printk(KERN_INFO MOD "Chelsio T3 RDMA Driver - version %s\n", + printk_once(KERN_INFO MOD "Chelsio T3 RDMA Driver - version %s\n", DRV_VERSION); rnicp = (struct iwch_dev *)ib_alloc_device(sizeof(*rnicp)); if (!rnicp) { @@ -162,21 +160,36 @@ static void close_rnic_dev(struct t3cdev *tdev) mutex_unlock(&dev_mutex); } -static void iwch_err_handler(struct t3cdev *tdev, u32 status, u32 error) +static void iwch_event_handler(struct t3cdev *tdev, u32 evt, u32 port_id) { struct cxio_rdev *rdev = tdev->ulp; - struct iwch_dev *rnicp = rdev_to_iwch_dev(rdev); + struct iwch_dev *rnicp; struct ib_event event; + u32 portnum = port_id + 1; - if (status == OFFLOAD_STATUS_DOWN) { + if (!rdev) + return; + rnicp = rdev_to_iwch_dev(rdev); + switch (evt) { + case OFFLOAD_STATUS_DOWN: { rdev->flags = CXIO_ERROR_FATAL; - - event.device = &rnicp->ibdev; event.event = IB_EVENT_DEVICE_FATAL; - event.element.port_num = 0; - ib_dispatch_event(&event); + break; + } + case OFFLOAD_PORT_DOWN: { + event.event = IB_EVENT_PORT_ERR; + break; + } + case OFFLOAD_PORT_UP: { + event.event = IB_EVENT_PORT_ACTIVE; + break; + } } + event.device = &rnicp->ibdev; + event.element.port_num = portnum; + ib_dispatch_event(&event); + return; } diff --git a/drivers/infiniband/hw/cxgb3/iwch_cm.c b/drivers/infiniband/hw/cxgb3/iwch_cm.c index 52d7bb0c2a1..66b41351910 100644 --- a/drivers/infinib |