From e82153b54d75af31d5d4a84efe441e5719f34cfc Mon Sep 17 00:00:00 2001 From: Krishna Kumar <krkumar2@in.ibm.com> Date: Mon, 16 Oct 2006 10:09:01 +0530 Subject: RDMA/cma: Optimize cma_bind_loopback() to check for empty list Optimize to test for an empty list first. This ends up simplifying the code too. Signed-off-by: Krishna Kumar <krkumar2@in.ibm.com> Acked-by: Sean Hefty <sean.hefty@intel.com> Signed-off-by: Roland Dreier <rolandd@cisco.com> --- drivers/infiniband/core/cma.c | 15 +++++++-------- 1 file changed, 7 insertions(+), 8 deletions(-) (limited to 'drivers/infiniband/core') diff --git a/drivers/infiniband/core/cma.c b/drivers/infiniband/core/cma.c index 845090b0859..4975d8154a7 100644 --- a/drivers/infiniband/core/cma.c +++ b/drivers/infiniband/core/cma.c @@ -1481,19 +1481,18 @@ static int cma_bind_loopback(struct rdma_id_private *id_priv) u8 p; mutex_lock(&lock); + if (list_empty(&dev_list)) { + ret = -ENODEV; + goto out; + } list_for_each_entry(cma_dev, &dev_list, list) for (p = 1; p <= cma_dev->device->phys_port_cnt; ++p) - if (!ib_query_port (cma_dev->device, p, &port_attr) && + if (!ib_query_port(cma_dev->device, p, &port_attr) && port_attr.state == IB_PORT_ACTIVE) goto port_found; - if (!list_empty(&dev_list)) { - p = 1; - cma_dev = list_entry(dev_list.next, struct cma_device, list); - } else { - ret = -ENODEV; - goto out; - } + p = 1; + cma_dev = list_entry(dev_list.next, struct cma_device, list); port_found: ret = ib_get_cached_gid(cma_dev->device, p, 0, &gid); -- cgit v1.2.3-18-g5258 From e4022274cf8df1f78f9e20ba7e199a9edf655422 Mon Sep 17 00:00:00 2001 From: Krishna Kumar <krkumar2@in.ibm.com> Date: Mon, 16 Oct 2006 10:09:08 +0530 Subject: RDMA/cma: Remove redundant check in cma_add_one Remove redundant check of node_guid in cma_add_one(). Signed-off-by: Krishna Kumar <krkumar2@in.ibm.com> Acked-by: Sean Hefty <sean.hefty@intel.com> Signed-off-by: Roland Dreier <rolandd@cisco.com> --- drivers/infiniband/core/cma.c | 5 ----- 1 file changed, 5 deletions(-) (limited to 'drivers/infiniband/core') diff --git a/drivers/infiniband/core/cma.c b/drivers/infiniband/core/cma.c index 4975d8154a7..9aba4ea9493 100644 --- a/drivers/infiniband/core/cma.c +++ b/drivers/infiniband/core/cma.c @@ -2122,8 +2122,6 @@ static void cma_add_one(struct ib_device *device) cma_dev->device = device; cma_dev->node_guid = device->node_guid; - if (!cma_dev->node_guid) - goto err; init_completion(&cma_dev->comp); atomic_set(&cma_dev->refcount, 1); @@ -2135,9 +2133,6 @@ static void cma_add_one(struct ib_device *device) list_for_each_entry(id_priv, &listen_any_list, list) cma_listen_on_dev(id_priv, cma_dev); mutex_unlock(&lock); - return; -err: - kfree(cma_dev); } static int cma_remove_id_dev(struct rdma_id_private *id_priv) -- cgit v1.2.3-18-g5258 From f115db4803effd8207c3169590fb3f13336a4093 Mon Sep 17 00:00:00 2001 From: Krishna Kumar <krkumar2@in.ibm.com> Date: Tue, 17 Oct 2006 10:09:09 +0530 Subject: RDMA/addr: Use time_after_eq() instead of time_after() in queue_req() In queue_req(), use time_after_eq() instead of time_after() for following reasons : - Improves insert time if multiple entries with same time are present. - set_timeout need not be called if entry with same time is added to the list (and that happens to be the entry with the smallest time), saving atomic/locking operations. - Earlier entries with same time are deleted first (fifo). Signed-off-by: Krishna Kumar <krkumar2@in.ibm.com> Acked-by: Sean Hefty <sean.hefty@intel.com> Signed-off-by: Roland Dreier <rolandd@cisco.com> --- drivers/infiniband/core/addr.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'drivers/infiniband/core') diff --git a/drivers/infiniband/core/addr.c b/drivers/infiniband/core/addr.c index e11187ecc93..a68d7c7df4c 100644 --- a/drivers/infiniband/core/addr.c +++ b/drivers/infiniband/core/addr.c @@ -139,7 +139,7 @@ static void queue_req(struct addr_req *req) mutex_lock(&lock); list_for_each_entry_reverse(temp_req, &req_list, list) { - if (time_after(req->timeout, temp_req->timeout)) + if (time_after_eq(req->timeout, temp_req->timeout)) break; } -- cgit v1.2.3-18-g5258 From a1a733f65b091fdad3d0783e648c92b491933ab6 Mon Sep 17 00:00:00 2001 From: Krishna Kumar <krkumar2@in.ibm.com> Date: Tue, 17 Oct 2006 10:09:11 +0530 Subject: RDMA/cma: Rewrite cma_req_handler() to encapsulate common code Rewrite cma_req_handler error handling case to encapsulate common code. Signed-off-by: Krishna Kumar <krkumar2@in.ibm.com> Acked-by: Sean Hefty <sean.hefty@intel.com> Signed-off-by: Roland Dreier <rolandd@cisco.com> --- drivers/infiniband/core/cma.c | 27 +++++++++++++-------------- 1 file changed, 13 insertions(+), 14 deletions(-) (limited to 'drivers/infiniband/core') diff --git a/drivers/infiniband/core/cma.c b/drivers/infiniband/core/cma.c index 9aba4ea9493..b449e26b84d 100644 --- a/drivers/infiniband/core/cma.c +++ b/drivers/infiniband/core/cma.c @@ -935,13 +935,8 @@ static int cma_req_handler(struct ib_cm_id *cm_id, struct ib_cm_event *ib_event) mutex_lock(&lock); ret = cma_acquire_dev(conn_id); mutex_unlock(&lock); - if (ret) { - ret = -ENODEV; - cma_exch(conn_id, CMA_DESTROYING); - cma_release_remove(conn_id); - rdma_destroy_id(&conn_id->id); - goto out; - } + if (ret) + goto release_conn_id; conn_id->cm_id.ib = cm_id; cm_id->context = conn_id; @@ -951,13 +946,17 @@ static int cma_req_handler(struct ib_cm_id *cm_id, struct ib_cm_event *ib_event) ret = cma_notify_user(conn_id, RDMA_CM_EVENT_CONNECT_REQUEST, 0, ib_event->private_data + offset, IB_CM_REQ_PRIVATE_DATA_SIZE - offset); - if (ret) { - /* Destroy the CM ID by returning a non-zero value. */ - conn_id->cm_id.ib = NULL; - cma_exch(conn_id, CMA_DESTROYING); - cma_release_remove(conn_id); - rdma_destroy_id(&conn_id->id); - } + if (!ret) + goto out; + + /* Destroy the CM ID by returning a non-zero value. */ + conn_id->cm_id.ib = NULL; + +release_conn_id: + cma_exch(conn_id, CMA_DESTROYING); + cma_release_remove(conn_id); + rdma_destroy_id(&conn_id->id); + out: cma_release_remove(listen_id); return ret; -- cgit v1.2.3-18-g5258 From bed8bdfddd851657cf9e5fd16bb44abb02ae7f42 Mon Sep 17 00:00:00 2001 From: Eric Sesterhenn <snakebyte@gmx.de> Date: Mon, 23 Oct 2006 22:17:21 +0200 Subject: IB: kmemdup() cleanup Replace open coded kmemdup() to save some screen space, and allow inlining/not inlining to be triggered by gcc. Signed-off-by: Eric Sesterhenn <snakebyte@gmx.de> Signed-off-by: Roland Dreier <rolandd@cisco.com> --- drivers/infiniband/core/cm.c | 3 +-- drivers/infiniband/core/iwcm.c | 3 +-- drivers/infiniband/core/ucm.c | 6 ++---- 3 files changed, 4 insertions(+), 8 deletions(-) (limited to 'drivers/infiniband/core') diff --git a/drivers/infiniband/core/cm.c b/drivers/infiniband/core/cm.c index 25b1018a476..82bc83bacea 100644 --- a/drivers/infiniband/core/cm.c +++ b/drivers/infiniband/core/cm.c @@ -240,11 +240,10 @@ static void * cm_copy_private_data(const void *private_data, if (!private_data || !private_data_len) return NULL; - data = kmalloc(private_data_len, GFP_KERNEL); + data = kmemdup(private_data, private_data_len, GFP_KERNEL); if (!data) return ERR_PTR(-ENOMEM); - memcpy(data, private_data, private_data_len); return data; } diff --git a/drivers/infiniband/core/iwcm.c b/drivers/infiniband/core/iwcm.c index c3fb304a4e8..2bbcfa5c6e2 100644 --- a/drivers/infiniband/core/iwcm.c +++ b/drivers/infiniband/core/iwcm.c @@ -140,10 +140,9 @@ static int copy_private_data(struct iwcm_id_private *cm_id_priv, { void *p; - p = kmalloc(event->private_data_len, GFP_ATOMIC); + p = kmemdup(event->private_data, event->private_data_len, GFP_ATOMIC); if (!p) return -ENOMEM; - memcpy(p, event->private_data, event->private_data_len); event->private_data = p; return 0; } diff --git a/drivers/infiniband/core/ucm.c b/drivers/infiniband/core/ucm.c index ad4f4d5c292..b4894ba223b 100644 --- a/drivers/infiniband/core/ucm.c +++ b/drivers/infiniband/core/ucm.c @@ -328,20 +328,18 @@ static int ib_ucm_event_process(struct ib_cm_event *evt, } if (uvt->data_len) { - uvt->data = kmalloc(uvt->data_len, GFP_KERNEL); + uvt->data = kmemdup(evt->private_data, uvt->data_len, GFP_KERNEL); if (!uvt->data) goto err1; - memcpy(uvt->data, evt->private_data, uvt->data_len); uvt->resp.present |= IB_UCM_PRES_DATA; } if (uvt->info_len) { - uvt->info = kmalloc(uvt->info_len, GFP_KERNEL); + uvt->info = kmemdup(info, uvt->info_len, GFP_KERNEL); if (!uvt->info) goto err2; - memcpy(uvt->info, info, uvt->info_len); uvt->resp.present |= IB_UCM_PRES_INFO; } return 0; -- cgit v1.2.3-18-g5258 From e31353eaeca736981ec13b46089d30147342b28b Mon Sep 17 00:00:00 2001 From: Dotan Barak <dotanb@mellanox.co.il> Date: Tue, 24 Oct 2006 13:35:27 -0700 Subject: RDMA/cm: Remove setting local write as part of QP access flags The qp_access_flags are for remote access permissions only, so IB_ACCESS_LOCAL_WRITE is an invalid value. Remove it from the values set by cm_init_qp_init_attr() and cma_init_ib_qp(). Signed-off-by: Dotan Barak <dotanb@mellanox.co.il> Signed-off-by: Sean Hefty <sean.hefty@intel.com> Signed-off-by: Roland Dreier <rolandd@cisco.com> --- drivers/infiniband/core/cm.c | 3 +-- drivers/infiniband/core/cma.c | 2 +- 2 files changed, 2 insertions(+), 3 deletions(-) (limited to 'drivers/infiniband/core') diff --git a/drivers/infiniband/core/cm.c b/drivers/infiniband/core/cm.c index 82bc83bacea..78d9c0c3314 100644 --- a/drivers/infiniband/core/cm.c +++ b/drivers/infiniband/core/cm.c @@ -3172,8 +3172,7 @@ static int cm_init_qp_init_attr(struct cm_id_private *cm_id_priv, case IB_CM_ESTABLISHED: *qp_attr_mask = IB_QP_STATE | IB_QP_ACCESS_FLAGS | IB_QP_PKEY_INDEX | IB_QP_PORT; - qp_attr->qp_access_flags = IB_ACCESS_LOCAL_WRITE | - IB_ACCESS_REMOTE_WRITE; + qp_attr->qp_access_flags = IB_ACCESS_REMOTE_WRITE; if (cm_id_priv->responder_resources) qp_attr->qp_access_flags |= IB_ACCESS_REMOTE_READ | IB_ACCESS_REMOTE_ATOMIC; diff --git a/drivers/infiniband/core/cma.c b/drivers/infiniband/core/cma.c index b449e26b84d..cf48f269743 100644 --- a/drivers/infiniband/core/cma.c +++ b/drivers/infiniband/core/cma.c @@ -344,7 +344,7 @@ static int cma_init_ib_qp(struct rdma_id_private *id_priv, struct ib_qp *qp) return ret; qp_attr.qp_state = IB_QPS_INIT; - qp_attr.qp_access_flags = IB_ACCESS_LOCAL_WRITE; + qp_attr.qp_access_flags = 0; qp_attr.port_num = id_priv->id.port_num; return ib_modify_qp(qp, &qp_attr, IB_QP_STATE | IB_QP_ACCESS_FLAGS | IB_QP_PKEY_INDEX | IB_QP_PORT); -- cgit v1.2.3-18-g5258 From e54f81889cd5228e7087637c377d76301c7c5663 Mon Sep 17 00:00:00 2001 From: Roland Dreier <rolandd@cisco.com> Date: Wed, 29 Nov 2006 15:33:07 -0800 Subject: IB: Convert kmem_cache_t -> struct kmem_cache Signed-off-by: Roland Dreier <rolandd@cisco.com> --- drivers/infiniband/core/mad.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'drivers/infiniband/core') diff --git a/drivers/infiniband/core/mad.c b/drivers/infiniband/core/mad.c index a72bcea46ff..3f9c16232c4 100644 --- a/drivers/infiniband/core/mad.c +++ b/drivers/infiniband/core/mad.c @@ -46,7 +46,7 @@ MODULE_DESCRIPTION("kernel IB MAD API"); MODULE_AUTHOR("Hal Rosenstock"); MODULE_AUTHOR("Sean Hefty"); -static kmem_cache_t *ib_mad_cache; +static struct kmem_cache *ib_mad_cache; static struct list_head ib_mad_port_list; static u32 ib_mad_client_id = 0; -- cgit v1.2.3-18-g5258 From 33ba0fa9f315ce32fbb86fa671c131f5355b52a1 Mon Sep 17 00:00:00 2001 From: Krishna Kumar <krkumar2@in.ibm.com> Date: Thu, 9 Nov 2006 09:30:34 +0530 Subject: RDMA/iwcm: Fix memory corruption bug in cm_work_handler() Possible memory corruption scenario: after putting the work entry back on the work_free_list, we call process_event() which dereferences work->event, which could have been modified to another value meanwhile. Signed-off-by: Krishna Kumar <krkumar2@in.ibm.com> Acked-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <rolandd@cisco.com> --- drivers/infiniband/core/iwcm.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) (limited to 'drivers/infiniband/core') diff --git a/drivers/infiniband/core/iwcm.c b/drivers/infiniband/core/iwcm.c index 2bbcfa5c6e2..22d498c9a68 100644 --- a/drivers/infiniband/core/iwcm.c +++ b/drivers/infiniband/core/iwcm.c @@ -829,7 +829,8 @@ static int process_event(struct iwcm_id_private *cm_id_priv, */ static void cm_work_handler(void *arg) { - struct iwcm_work *work = arg, lwork; + struct iwcm_work *work = arg; + struct iw_cm_event levent; struct iwcm_id_private *cm_id_priv = work->cm_id; unsigned long flags; int empty; @@ -842,11 +843,11 @@ static void cm_work_handler(void *arg) struct iwcm_work, list); list_del_init(&work->list); empty = list_empty(&cm_id_priv->work_list); - lwork = *work; + levent = work->event; put_work(work); spin_unlock_irqrestore(&cm_id_priv->lock, flags); - ret = process_event(cm_id_priv, &work->event); + ret = process_event(cm_id_priv, &levent); if (ret) { set_bit(IWCM_F_CALLBACK_DESTROY, &cm_id_priv->flags); destroy_cm_id(&cm_id_priv->id); -- cgit v1.2.3-18-g5258 From 83b96586239bf6c719ff640341e1cf83e4a7c046 Mon Sep 17 00:00:00 2001 From: Krishna Kumar <krkumar2@in.ibm.com> Date: Thu, 9 Nov 2006 09:30:41 +0530 Subject: RDMA/iwcm: Fix memory leak If we get IW_CM_EVENT_CONNECT_REQUEST message and encounter an error (not in the LISTEN state, cannot create an id, cannot alloc work_entry, etc), then the memory allocated by cm_event_handler() in the event->private_data gets leaked. Since cm_work_handler has already put the event on the work_free_list, this allocated memory is leaked. High backlog value can allow DoS attacks. Signed-off-by: Krishna Kumar <krkumar2@in.ibm.com> Acked-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <rolandd@cisco.com> --- drivers/infiniband/core/iwcm.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) (limited to 'drivers/infiniband/core') diff --git a/drivers/infiniband/core/iwcm.c b/drivers/infiniband/core/iwcm.c index 22d498c9a68..0cfd7848dd3 100644 --- a/drivers/infiniband/core/iwcm.c +++ b/drivers/infiniband/core/iwcm.c @@ -619,7 +619,7 @@ static void cm_conn_req_handler(struct iwcm_id_private *listen_id_priv, spin_lock_irqsave(&listen_id_priv->lock, flags); if (listen_id_priv->state != IW_CM_STATE_LISTEN) { spin_unlock_irqrestore(&listen_id_priv->lock, flags); - return; + goto out; } spin_unlock_irqrestore(&listen_id_priv->lock, flags); @@ -628,7 +628,7 @@ static void cm_conn_req_handler(struct iwcm_id_private *listen_id_priv, listen_id_priv->id.context); /* If the cm_id could not be created, ignore the request */ if (IS_ERR(cm_id)) - return; + goto out; cm_id->provider_data = iw_event->provider_data; cm_id->local_addr = iw_event->local_addr; @@ -641,7 +641,7 @@ static void cm_conn_req_handler(struct iwcm_id_private *listen_id_priv, if (ret) { iw_cm_reject(cm_id, NULL, 0); iw_destroy_cm_id(cm_id); - return; + goto out; } /* Call the client CM handler */ @@ -653,6 +653,7 @@ static void cm_conn_req_handler(struct iwcm_id_private *listen_id_priv, kfree(cm_id); } +out: if (iw_event->private_data_len) kfree(iw_event->private_data); } -- cgit v1.2.3-18-g5258 From 13fccdb380f88770f05b922c273d907aecda7c12 Mon Sep 17 00:00:00 2001 From: Krishna Kumar <krkumar2@in.ibm.com> Date: Thu, 9 Nov 2006 09:30:43 +0530 Subject: RDMA/iwcm: Remove unnecessary initializations Signed-off-by: Krishna Kumar <krkumar2@in.ibm.com> Acked-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <rolandd@cisco.com> --- drivers/infiniband/core/iwcm.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) (limited to 'drivers/infiniband/core') diff --git a/drivers/infiniband/core/iwcm.c b/drivers/infiniband/core/iwcm.c index 0cfd7848dd3..b02e54136ac 100644 --- a/drivers/infiniband/core/iwcm.c +++ b/drivers/infiniband/core/iwcm.c @@ -407,7 +407,7 @@ int iw_cm_listen(struct iw_cm_id *cm_id, int backlog) { struct iwcm_id_private *cm_id_priv; unsigned long flags; - int ret = 0; + int ret; cm_id_priv = container_of(cm_id, struct iwcm_id_private, id); @@ -534,7 +534,7 @@ EXPORT_SYMBOL(iw_cm_accept); int iw_cm_connect(struct iw_cm_id *cm_id, struct iw_cm_conn_param *iw_param) { struct iwcm_id_private *cm_id_priv; - int ret = 0; + int ret; unsigned long flags; struct ib_qp *qp; @@ -674,7 +674,7 @@ static int cm_conn_est_handler(struct iwcm_id_private *cm_id_priv, struct iw_cm_event *iw_event) { unsigned long flags; - int ret = 0; + int ret; spin_lock_irqsave(&cm_id_priv->lock, flags); @@ -704,7 +704,7 @@ static int cm_conn_rep_handler(struct iwcm_id_private *cm_id_priv, struct iw_cm_event *iw_event) { unsigned long flags; - int ret = 0; + int ret; spin_lock_irqsave(&cm_id_priv->lock, flags); /* -- cgit v1.2.3-18-g5258 From 715a588f420936ecdb813c4fcd40dff7a16b1638 Mon Sep 17 00:00:00 2001 From: Krishna Kumar <krkumar2@in.ibm.com> Date: Thu, 9 Nov 2006 09:30:45 +0530 Subject: RDMA/iwcm: Remove unnecessary function argument Remove unnecessary cm_id_priv argument to copy_private_data(), and change text to reflect the code. Fix couple of typos in comments. Signed-off-by: Krishna Kumar <krkumar2@in.ibm.com> Acked-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <rolandd@cisco.com> --- drivers/infiniband/core/iwcm.c | 13 ++++++------- 1 file changed, 6 insertions(+), 7 deletions(-) (limited to 'drivers/infiniband/core') diff --git a/drivers/infiniband/core/iwcm.c b/drivers/infiniband/core/iwcm.c index b02e54136ac..54faa033d89 100644 --- a/drivers/infiniband/core/iwcm.c +++ b/drivers/infiniband/core/iwcm.c @@ -80,7 +80,7 @@ struct iwcm_work { * 1) in the event upcall, cm_event_handler(), for a listening cm_id. If * the backlog is exceeded, then no more connection request events will * be processed. cm_event_handler() returns -ENOMEM in this case. Its up - * to the provider to reject the connectino request. + * to the provider to reject the connection request. * 2) in the connection request workqueue handler, cm_conn_req_handler(). * If work elements cannot be allocated for the new connect request cm_id, * then IWCM will call the provider reject method. This is ok since @@ -131,12 +131,11 @@ static int alloc_work_entries(struct iwcm_id_private *cm_id_priv, int count) } /* - * Save private data from incoming connection requests in the - * cm_id_priv so the low level driver doesn't have to. Adjust + * Save private data from incoming connection requests to + * iw_cm_event, so the low level driver doesn't have to. Adjust * the event ptr to point to the local copy. */ -static int copy_private_data(struct iwcm_id_private *cm_id_priv, - struct iw_cm_event *event) +static int copy_private_data(struct iw_cm_event *event) { void *p; @@ -242,7 +241,7 @@ static int iwcm_modify_qp_sqd(struct ib_qp *qp) /* * CM_ID <-- CLOSING * - * Block if a passive or active connection is currenlty being processed. Then + * Block if a passive or active connection is currently being processed. Then * process the event as follows: * - If we are ESTABLISHED, move to CLOSING and modify the QP state * based on the abrupt flag @@ -907,7 +906,7 @@ static int cm_event_handler(struct iw_cm_id *cm_id, if ((work->event.event == IW_CM_EVENT_CONNECT_REQUEST || work->event.event == IW_CM_EVENT_CONNECT_REPLY) && work->event.private_data_len) { - ret = copy_private_data(cm_id_priv, &work->event); + ret = copy_private_data(&work->event); if (ret) { put_work(work); goto out; -- cgit v1.2.3-18-g5258 From 9ab1ffa8775d9c677b1301cccce8a7d91e5163d0 Mon Sep 17 00:00:00 2001 From: Krishna Kumar <krkumar2@in.ibm.com> Date: Thu, 9 Nov 2006 09:30:48 +0530 Subject: RDMA/iwcm: Fix comment for iwcm_deref_id() to match code In iwcm_deref_id(), the comment says : "If the last reference is being removed and iw_destroy_cm_id is waiting, wake up the waiting thread". The second part of the comment, "and iw_destroy_cm_id is waiting," is wrong, since this function either wakes the waiter already waiting in iwcm_deref_id, or enables it (so that when wait_for_completion() is performed later, it will immediately return). Signed-off-by: Krishna Kumar <krkumar2@in.ibm.com> Acked-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <rolandd@cisco.com> --- drivers/infiniband/core/iwcm.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) (limited to 'drivers/infiniband/core') diff --git a/drivers/infiniband/core/iwcm.c b/drivers/infiniband/core/iwcm.c index 54faa033d89..cf797d7aea0 100644 --- a/drivers/infiniband/core/iwcm.c +++ b/drivers/infiniband/core/iwcm.c @@ -147,8 +147,9 @@ static int copy_private_data(struct iw_cm_event *event) } /* - * Release a reference on cm_id. If the last reference is being removed - * and iw_destroy_cm_id is waiting, wake up the waiting thread. + * Release a reference on cm_id. If the last reference is being + * released, enable the waiting thread (in iw_destroy_cm_id) to + * get woken up, and return 1 if a thread is already waiting. */ static int iwcm_deref_id(struct iwcm_id_private *cm_id_priv) { -- cgit v1.2.3-18-g5258 From c78bb8442b14ee6704bdb323111ffa874d4bfdaa Mon Sep 17 00:00:00 2001 From: Krishna Kumar <krkumar2@in.ibm.com> Date: Fri, 24 Nov 2006 16:02:34 +0530 Subject: RDMA/addr: Fix some cancellation problems in process_req() Fix following problems in process_req() relating to cancellation: - Function is wrongly doing another addr_remote() when cancelled, which is not required. - Make failure reporting immediate by using time_after_eq(). - On cancellation, -ETIMEDOUT was returned to the callback routine instead of the more appropriate -ECANCELLED (users getting notified may want to print/return this status, eg ucma_event_handler). Signed-off-by: Krishna Kumar <krkumar2@in.ibm.com> Signed-off-by: Roland Dreier <rolandd@cisco.com> --- drivers/infiniband/core/addr.c | 11 +++++------ 1 file changed, 5 insertions(+), 6 deletions(-) (limited to 'drivers/infiniband/core') diff --git a/drivers/infiniband/core/addr.c b/drivers/infiniband/core/addr.c index a68d7c7df4c..15ba51806ff 100644 --- a/drivers/infiniband/core/addr.c +++ b/drivers/infiniband/core/addr.c @@ -225,17 +225,16 @@ static void process_req(void *data) mutex_lock(&lock); list_for_each_entry_safe(req, temp_req, &req_list, list) { - if (req->status) { + if (req->status == -ENODATA) { src_in = (struct sockaddr_in *) &req->src_addr; dst_in = (struct sockaddr_in *) &req->dst_addr; req->status = addr_resolve_remote(src_in, dst_in, req->addr); + if (req->status && time_after_eq(jiffies, req->timeout)) + req->status = -ETIMEDOUT; + else if (req->status == -ENODATA) + continue; } - if (req->status && time_after(jiffies, req->timeout)) - req->status = -ETIMEDOUT; - else if (req->status == -ENODATA) - continue; - list_del(&req->list); list_add_tail(&req->list, &done_list); } -- cgit v1.2.3-18-g5258 From 04699a1f8634a4e89c71b22050b599c72126fa96 Mon Sep 17 00:00:00 2001 From: Roland Dreier <rolandd@cisco.com> Date: Wed, 29 Nov 2006 15:33:09 -0800 Subject: RDMA/addr: list_move() cleanups Replace a couple list_del()/list_add() combos with list_move(). Signed-off-by: Roland Dreier <rolandd@cisco.com> --- drivers/infiniband/core/addr.c | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-) (limited to 'drivers/infiniband/core') diff --git a/drivers/infiniband/core/addr.c b/drivers/infiniband/core/addr.c index 15ba51806ff..7767a11b689 100644 --- a/drivers/infiniband/core/addr.c +++ b/drivers/infiniband/core/addr.c @@ -235,8 +235,7 @@ static void process_req(void *data) else if (req->status == -ENODATA) continue; } - list_del(&req->list); - list_add_tail(&req->list, &done_list); + list_move_tail(&req->list, &done_list); } if (!list_empty(&req_list)) { @@ -346,8 +345,7 @@ void rdma_addr_cancel(struct rdma_dev_addr *addr) if (req->addr == addr) { req->status = -ECANCELED; req->timeout = jiffies; - list_del(&req->list); - list_add(&req->list, &req_list); + list_move(&req->list, &req_list); set_timeout(req->timeout); break; } -- cgit v1.2.3-18-g5258 From e1444b5a163e81138754cab27c4fa1637b5a2239 Mon Sep 17 00:00:00 2001 From: Sean Hefty <sean.hefty@intel.com> Date: Tue, 28 Nov 2006 14:57:13 -0800 Subject: IB/cm: Fix automatic path migration support The ib_cm_establish() function is replaced with a more generic ib_cm_notify(). This routine is used to notify the CM that failover has occurred, so that future CM messages (LAP, DREQ) reach the remote CM. (Currently, we continue to use the original path) This bumps the userspace CM ABI. New alternate path information is captured when a LAP message is sent or received. This allows QP attributes to be initialized for the user when a new path is loaded after failover occurs. Signed-off-by: Sean Hefty <sean.hefty@intel.com> Signed-off-by: Roland Dreier <rolandd@cisco.com> --- drivers/infiniband/core/cm.c | 115 ++++++++++++++++++++++++++++++++---------- drivers/infiniband/core/ucm.c | 12 ++--- 2 files changed, 94 insertions(+), 33 deletions(-) (limited to 'drivers/infiniband/core') diff --git a/drivers/infiniband/core/cm.c b/drivers/infiniband/core/cm.c index 78d9c0c3314..e5dc4530808 100644 --- a/drivers/infiniband/core/cm.c +++ b/drivers/infiniband/core/cm.c @@ -147,12 +147,12 @@ struct cm_id_private { __be32 rq_psn; int timeout_ms; enum ib_mtu path_mtu; + __be16 pkey; u8 private_data_len; u8 max_cm_retries; u8 peer_to_peer; u8 responder_resources; u8 initiator_depth; - u8 local_ack_timeout; u8 retry_count; u8 rnr_retry_count; u8 service_timeout; @@ -690,7 +690,7 @@ static void cm_enter_timewait(struct cm_id_private *cm_id_priv) * timewait before notifying the user that we've exited timewait. */ cm_id_priv->id.state = IB_CM_TIMEWAIT; - wait_time = cm_convert_to_ms(cm_id_priv->local_ack_timeout); + wait_time = cm_convert_to_ms(cm_id_priv->av.packet_life_time + 1); queue_delayed_work(cm.wq, &cm_id_priv->timewait_info->work.work, msecs_to_jiffies(wait_time)); cm_id_priv->timewait_info = NULL; @@ -1009,6 +1009,7 @@ int ib_send_cm_req(struct ib_cm_id *cm_id, cm_id_priv->responder_resources = param->responder_resources; cm_id_priv->retry_count = param->retry_count; cm_id_priv->path_mtu = param->primary_path->mtu; + cm_id_priv->pkey = param->primary_path->pkey; cm_id_priv->qp_type = param->qp_type; ret = cm_alloc_msg(cm_id_priv, &cm_id_priv->msg); @@ -1023,8 +1024,6 @@ int ib_send_cm_req(struct ib_cm_id *cm_id, cm_id_priv->local_qpn = cm_req_get_local_qpn(req_msg); cm_id_priv->rq_psn = cm_req_get_starting_psn(req_msg); - cm_id_priv->local_ack_timeout = - cm_req_get_primary_local_ack_timeout(req_msg); spin_lock_irqsave(&cm_id_priv->lock, flags); ret = ib_post_send_mad(cm_id_priv->msg, NULL); @@ -1409,9 +1408,8 @@ static int cm_req_handler(struct cm_work *work) cm_id_priv->initiator_depth = cm_req_get_resp_res(req_msg); cm_id_priv->responder_resources = cm_req_get_init_depth(req_msg); cm_id_priv->path_mtu = cm_req_get_path_mtu(req_msg); + cm_id_priv->pkey = req_msg->pkey; cm_id_priv->sq_psn = cm_req_get_starting_psn(req_msg); - cm_id_priv->local_ack_timeout = - cm_req_get_primary_local_ack_timeout(req_msg); cm_id_priv->retry_count = cm_req_get_retry_count(req_msg); cm_id_priv->rnr_retry_count = cm_req_get_rnr_retry_count(req_msg); cm_id_priv->qp_type = cm_req_get_qp_type(req_msg); @@ -1715,7 +1713,7 @@ static int cm_establish_handler(struct cm_work *work) unsigned long flags; int ret; - /* See comment in ib_cm_establish about lookup. */ + /* See comment in cm_establish about lookup. */ cm_id_priv = cm_acquire_id(work->local_id, work->remote_id); if (!cm_id_priv) return -EINVAL; @@ -2401,11 +2399,16 @@ int ib_send_cm_lap(struct ib_cm_id *cm_id, cm_id_priv = container_of(cm_id, struct cm_id_private, id); spin_lock_irqsave(&cm_id_priv->lock, flags); if (cm_id->state != IB_CM_ESTABLISHED || - cm_id->lap_state != IB_CM_LAP_IDLE) { + (cm_id->lap_state != IB_CM_LAP_UNINIT && + cm_id->lap_state != IB_CM_LAP_IDLE)) { ret = -EINVAL; goto out; } + ret = cm_init_av_by_path(alternate_path, &cm_id_priv->alt_av); + if (ret) + goto out; + ret = cm_alloc_msg(cm_id_priv, &msg); if (ret) goto out; @@ -2430,7 +2433,8 @@ out: spin_unlock_irqrestore(&cm_id_priv->lock, flags); } EXPORT_SYMBOL(ib_send_cm_lap); -static void cm_format_path_from_lap(struct ib_sa_path_rec *path, +static void cm_format_path_from_lap(struct cm_id_private *cm_id_priv, + struct ib_sa_path_rec *path, struct cm_lap_msg *lap_msg) { memset(path, 0, sizeof *path); @@ -2442,10 +2446,10 @@ static void cm_format_path_from_lap(struct ib_sa_path_rec *path, path->hop_limit = lap_msg->alt_hop_limit; path->traffic_class = cm_lap_get_traffic_class(lap_msg); path->reversible = 1; - /* pkey is same as in REQ */ + path->pkey = cm_id_priv->pkey; path->sl = cm_lap_get_sl(lap_msg); path->mtu_selector = IB_SA_EQ; - /* mtu is same as in REQ */ + path->mtu = cm_id_priv->path_mtu; path->rate_selector = IB_SA_EQ; path->rate = cm_lap_get_packet_rate(lap_msg); path->packet_life_time_selector = IB_SA_EQ; @@ -2471,7 +2475,7 @@ static int cm_lap_handler(struct cm_work *work) param = &work->cm_event.param.lap_rcvd; param->alternate_path = &work->path[0]; - cm_format_path_from_lap(param->alternate_path, lap_msg); + cm_format_path_from_lap(cm_id_priv, param->alternate_path, lap_msg); work->cm_event.private_data = &lap_msg->private_data; spin_lock_irqsave(&cm_id_priv->lock, flags); @@ -2479,6 +2483,7 @@ static int cm_lap_handler(struct cm_work *work) goto unlock; switch (cm_id_priv->id.lap_state) { + case IB_CM_LAP_UNINIT: case IB_CM_LAP_IDLE: break; case IB_CM_MRA_LAP_SENT: @@ -2501,6 +2506,10 @@ static int cm_lap_handler(struct cm_work *work) cm_id_priv->id.lap_state = IB_CM_LAP_RCVD; cm_id_priv->tid = lap_msg->hdr.tid; + cm_init_av_for_response(work->port, work->mad_recv_wc->wc, + work->mad_recv_wc->recv_buf.grh, + &cm_id_priv->av); + cm_init_av_by_path(param->alternate_path, &cm_id_priv->alt_av); ret = atomic_inc_and_test(&cm_id_priv->work_count); if (!ret) list_add_tail(&work->list, &cm_id_priv->work_list); @@ -3039,7 +3048,7 @@ static void cm_work_handler(void *data) cm_free_work(work); } -int ib_cm_establish(struct ib_cm_id *cm_id) +static int cm_establish(struct ib_cm_id *cm_id) { struct cm_id_private *cm_id_priv; struct cm_work *work; @@ -3087,7 +3096,44 @@ int ib_cm_establish(struct ib_cm_id *cm_id) out: return ret; } -EXPORT_SYMBOL(ib_cm_establish); + +static int cm_migrate(struct ib_cm_id *cm_id) +{ + struct cm_id_private *cm_id_priv; + unsigned long flags; + int ret = 0; + + cm_id_priv = container_of(cm_id, struct cm_id_private, id); + spin_lock_irqsave(&cm_id_priv->lock, flags); + if (cm_id->state == IB_CM_ESTABLISHED && + (cm_id->lap_state == IB_CM_LAP_UNINIT || + cm_id->lap_state == IB_CM_LAP_IDLE)) { + cm_id->lap_state = IB_CM_LAP_IDLE; + cm_id_priv->av = cm_id_priv->alt_av; + } else + ret = -EINVAL; + spin_unlock_irqrestore(&cm_id_priv->lock, flags); + + return ret; +} + +int ib_cm_notify(struct ib_cm_id *cm_id, enum ib_event_type event) +{ + int ret; + + switch (event) { + case IB_EVENT_COMM_EST: + ret = cm_establish(cm_id); + break; + case IB_EVENT_PATH_MIG: + ret = cm_migrate(cm_id); + break; + default: + ret = -EINVAL; + } + return ret; +} +EXPORT_SYMBOL(ib_cm_notify); static void cm_recv_handler(struct ib_mad_agent *mad_agent, struct ib_mad_recv_wc *mad_recv_wc) @@ -3220,6 +3266,9 @@ static int cm_init_qp_rtr_attr(struct cm_id_private *cm_id_priv, if (cm_id_priv->alt_av.ah_attr.dlid) { *qp_attr_mask |= IB_QP_ALT_PATH; qp_attr->alt_port_num = cm_id_priv->alt_av.port->port_num; + qp_attr->alt_pkey_index = cm_id_priv->alt_av.pkey_index; + qp_attr->alt_timeout = + cm_id_priv->alt_av.packet_life_time + 1; qp_attr->alt_ah_attr = cm_id_priv->alt_av.ah_attr; } ret = 0; @@ -3246,19 +3295,31 @@ static int cm_init_qp_rts_attr(struct cm_id_private *cm_id_priv, case IB_CM_REP_SENT: case IB_CM_MRA_REP_RCVD: case IB_CM_ESTABLISHED: - *qp_attr_mask = IB_QP_STATE | IB_QP_SQ_PSN; - qp_attr->sq_psn = be32_to_cpu(cm_id_priv->sq_psn); - if (cm_id_priv->qp_type == IB_QPT_RC) { - *qp_attr_mask |= IB_QP_TIMEOUT | IB_QP_RETRY_CNT | - IB_QP_RNR_RETRY | - IB_QP_MAX_QP_RD_ATOMIC; - qp_attr->timeout = cm_id_priv->local_ack_timeout; - qp_attr->retry_cnt = cm_id_priv->retry_count; - qp_attr->rnr_retry = cm_id_priv->rnr_retry_count; - qp_attr->max_rd_atomic = cm_id_priv->initiator_depth; - } - if (cm_id_priv->alt_av.ah_attr.dlid) { - *qp_attr_mask |= IB_QP_PATH_MIG_STATE; + if (cm_id_priv->id.lap_state == IB_CM_LAP_UNINIT) { + *qp_attr_mask = IB_QP_STATE | IB_QP_SQ_PSN; + qp_attr->sq_psn = be32_to_cpu(cm_id_priv->sq_psn); + if (cm_id_priv->qp_type == IB_QPT_RC) { + *qp_attr_mask |= IB_QP_TIMEOUT | IB_QP_RETRY_CNT | + IB_QP_RNR_RETRY | + IB_QP_MAX_QP_RD_ATOMIC; + qp_attr->timeout = + cm_id_priv->av.packet_life_time + 1; + qp_attr->retry_cnt = cm_id_priv->retry_count; + qp_attr->rnr_retry = cm_id_priv->rnr_retry_count; + qp_attr->max_rd_atomic = + cm_id_priv->initiator_depth; + } + if (cm_id_priv->alt_av.ah_attr.dlid) { + *qp_attr_mask |= IB_QP_PATH_MIG_STATE; + qp_attr->path_mig_state = IB_MIG_REARM; + } + } else { + *qp_attr_mask = IB_QP_ALT_PATH | IB_QP_PATH_MIG_STATE; + qp_attr->alt_port_num = cm_id_priv->alt_av.port->port_num; + qp_attr->alt_pkey_index = cm_id_priv->alt_av.pkey_index; + qp_attr->alt_timeout = + cm_id_priv->alt_av.packet_life_time + 1; + qp_attr->alt_ah_attr = cm_id_priv->alt_av.ah_attr; qp_attr->path_mig_state = IB_MIG_REARM; } ret = 0; diff --git a/drivers/infiniband/core/ucm.c b/drivers/infiniband/core/ucm.c index b4894ba223b..1f4f2d2cfa2 100644 --- a/drivers/infiniband/core/ucm.c +++ b/drivers/infiniband/core/ucm.c @@ -683,11 +683,11 @@ out: return result; } -static ssize_t ib_ucm_establish(struct ib_ucm_file *file, - const char __user *inbuf, - int in_len, int out_len) +static ssize_t ib_ucm_notify(struct ib_ucm_file *file, + const char __user *inbuf, + int in_len, int out_len) { - struct ib_ucm_establish cmd; + struct ib_ucm_notify cmd; struct ib_ucm_context *ctx; int result; @@ -698,7 +698,7 @@ static ssize_t ib_ucm_establish(struct ib_ucm_file *file, if (IS_ERR(ctx)) return PTR_ERR(ctx); - result = ib_cm_establish(ctx->cm_id); + result = ib_cm_notify(ctx->cm_id, (enum ib_event_type) cmd.event); ib_ucm_ctx_put(ctx); return result; } @@ -1105,7 +1105,7 @@ static ssize_t (*ucm_cmd_table[])(struct ib_ucm_file *file, [IB_USER_CM_CMD_DESTROY_ID] = ib_ucm_destroy_id, [IB_USER_CM_CMD_ATTR_ID] = ib_ucm_attr_id, [IB_USER_CM_CMD_LISTEN] = ib_ucm_listen, - [IB_USER_CM_CMD_ESTABLISH] = ib_ucm_establish, + [IB_USER_CM_CMD_NOTIFY] = ib_ucm_notify, [IB_USER_CM_CMD_SEND_REQ] = ib_ucm_send_req, [IB_USER_CM_CMD_SEND_REP] = ib_ucm_send_rep, [IB_USER_CM_CMD_SEND_RTU] = ib_ucm_send_rtu, -- cgit v1.2.3-18-g5258 From f469b2626f48829c06e40ac799c1edf62b12048e Mon Sep 17 00:00:00 2001 From: "Michael S. Tsirkin" <mst@mellanox.co.il> Date: Wed, 29 Nov 2006 15:33:10 -0800 Subject: IB/ucm: Fix deadlock in cleanup ib_ucm_cleanup_events() holds file_mutex while calling ib_destroy_cm_id(). This can deadlock since ib_destroy_cm_id() flushes event handlers, and ib_ucm_event_handler() needs file_mutex, too. Therefore, drop the file_mutex during the call to ib_destroy_cm_id(). Signed-off-by: Michael S. Tsirkin <mst@mellanox.co.il> Signed-off-by: Roland Dreier <rolandd@cisco.com> --- drivers/infiniband/core/ucm.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'drivers/infiniband/core') diff --git a/drivers/infiniband/core/ucm.c b/drivers/infiniband/core/ucm.c index 1f4f2d2cfa2..f15220a0ee7 100644 --- a/drivers/infiniband/core/ucm.c +++ b/drivers/infiniband/core/ucm.c @@ -161,12 +161,14 @@ static void ib_ucm_cleanup_events(struct ib_ucm_context *ctx) struct ib_ucm_event, ctx_list); list_del(&uevent->file_list); list_del(&uevent->ctx_list); + mutex_unlock(&ctx->file->file_mutex); /* clear incoming connections. */ if (ib_ucm_new_cm_id(uevent->resp.event)) ib_destroy_cm_id(uevent->cm_id); kfree(uevent); + mutex_lock(&ctx->file->file_mutex); } mutex_unlock(&ctx->file->file_mutex); } -- cgit v1.2.3-18-g5258