summaryrefslogtreecommitdiff
path: root/drivers/infiniband
diff options
context:
space:
mode:
Diffstat (limited to 'drivers/infiniband')
-rw-r--r--drivers/infiniband/core/cm.c49
-rw-r--r--drivers/infiniband/core/device.c23
-rw-r--r--drivers/infiniband/core/sysfs.c32
-rw-r--r--drivers/infiniband/core/ucm.c23
-rw-r--r--drivers/infiniband/core/user_mad.c45
-rw-r--r--drivers/infiniband/core/uverbs.h16
-rw-r--r--drivers/infiniband/core/uverbs_cmd.c263
-rw-r--r--drivers/infiniband/core/uverbs_main.c29
-rw-r--r--drivers/infiniband/core/verbs.c4
-rw-r--r--drivers/infiniband/hw/mthca/mthca_av.c10
-rw-r--r--drivers/infiniband/hw/mthca/mthca_cmd.c19
-rw-r--r--drivers/infiniband/hw/mthca/mthca_cq.c23
-rw-r--r--drivers/infiniband/hw/mthca/mthca_dev.h2
-rw-r--r--drivers/infiniband/hw/mthca/mthca_eq.c32
-rw-r--r--drivers/infiniband/hw/mthca/mthca_main.c4
-rw-r--r--drivers/infiniband/hw/mthca/mthca_mcg.c54
-rw-r--r--drivers/infiniband/hw/mthca/mthca_memfree.c4
-rw-r--r--drivers/infiniband/hw/mthca/mthca_provider.c132
-rw-r--r--drivers/infiniband/hw/mthca/mthca_qp.c301
-rw-r--r--drivers/infiniband/hw/mthca/mthca_srq.c2
-rw-r--r--drivers/infiniband/hw/mthca/mthca_wqe.h3
-rw-r--r--drivers/infiniband/ulp/ipoib/ipoib.h6
-rw-r--r--drivers/infiniband/ulp/ipoib/ipoib_ib.c31
-rw-r--r--drivers/infiniband/ulp/ipoib/ipoib_main.c25
-rw-r--r--drivers/infiniband/ulp/ipoib/ipoib_multicast.c117
-rw-r--r--drivers/infiniband/ulp/ipoib/ipoib_verbs.c8
-rw-r--r--drivers/infiniband/ulp/ipoib/ipoib_vlan.c10
-rw-r--r--drivers/infiniband/ulp/srp/ib_srp.c24
28 files changed, 717 insertions, 574 deletions
diff --git a/drivers/infiniband/core/cm.c b/drivers/infiniband/core/cm.c
index 02110e00d145..2514de3480d8 100644
--- a/drivers/infiniband/core/cm.c
+++ b/drivers/infiniband/core/cm.c
@@ -308,10 +308,11 @@ static int cm_alloc_id(struct cm_id_private *cm_id_priv)
{
unsigned long flags;
int ret;
+ static int next_id;
do {
spin_lock_irqsave(&cm.lock, flags);
- ret = idr_get_new_above(&cm.local_id_table, cm_id_priv, 1,
+ ret = idr_get_new_above(&cm.local_id_table, cm_id_priv, next_id++,
(__force int *) &cm_id_priv->id.local_id);
spin_unlock_irqrestore(&cm.lock, flags);
} while( (ret == -EAGAIN) && idr_pre_get(&cm.local_id_table, GFP_KERNEL) );
@@ -684,6 +685,13 @@ retest:
cm_reject_sidr_req(cm_id_priv, IB_SIDR_REJECT);
break;
case IB_CM_REQ_SENT:
+ ib_cancel_mad(cm_id_priv->av.port->mad_agent, cm_id_priv->msg);
+ spin_unlock_irqrestore(&cm_id_priv->lock, flags);
+ ib_send_cm_rej(cm_id, IB_CM_REJ_TIMEOUT,
+ &cm_id_priv->av.port->cm_dev->ca_guid,
+ sizeof cm_id_priv->av.port->cm_dev->ca_guid,
+ NULL, 0);
+ break;
case IB_CM_MRA_REQ_RCVD:
case IB_CM_REP_SENT:
case IB_CM_MRA_REP_RCVD:
@@ -694,10 +702,8 @@ retest:
case IB_CM_REP_RCVD:
case IB_CM_MRA_REP_SENT:
spin_unlock_irqrestore(&cm_id_priv->lock, flags);
- ib_send_cm_rej(cm_id, IB_CM_REJ_TIMEOUT,
- &cm_id_priv->av.port->cm_dev->ca_guid,
- sizeof cm_id_priv->av.port->cm_dev->ca_guid,
- NULL, 0);
+ ib_send_cm_rej(cm_id, IB_CM_REJ_CONSUMER_DEFINED,
+ NULL, 0, NULL, 0);
break;
case IB_CM_ESTABLISHED:
spin_unlock_irqrestore(&cm_id_priv->lock, flags);
@@ -850,7 +856,7 @@ static void cm_format_req(struct cm_req_msg *req_msg,
param->private_data_len);
}
-static inline int cm_validate_req_param(struct ib_cm_req_param *param)
+static int cm_validate_req_param(struct ib_cm_req_param *param)
{
/* peer-to-peer not supported */
if (param->peer_to_peer)
@@ -999,7 +1005,7 @@ static inline int cm_is_active_peer(__be64 local_ca_guid, __be64 remote_ca_guid,
(be32_to_cpu(local_qpn) > be32_to_cpu(remote_qpn))));
}
-static inline void cm_format_paths_from_req(struct cm_req_msg *req_msg,
+static void cm_format_paths_from_req(struct cm_req_msg *req_msg,
struct ib_sa_path_rec *primary_path,
struct ib_sa_path_rec *alt_path)
{
@@ -3157,22 +3163,6 @@ int ib_cm_init_qp_attr(struct ib_cm_id *cm_id,
}
EXPORT_SYMBOL(ib_cm_init_qp_attr);
-static __be64 cm_get_ca_guid(struct ib_device *device)
-{
- struct ib_device_attr *device_attr;
- __be64 guid;
- int ret;
-
- device_attr = kmalloc(sizeof *device_attr, GFP_KERNEL);
- if (!device_attr)
- return 0;
-
- ret = ib_query_device(device, device_attr);
- guid = ret ? 0 : device_attr->node_guid;
- kfree(device_attr);
- return guid;
-}
-
static void cm_add_one(struct ib_device *device)
{
struct cm_device *cm_dev;
@@ -3194,9 +3184,7 @@ static void cm_add_one(struct ib_device *device)
return;
cm_dev->device = device;
- cm_dev->ca_guid = cm_get_ca_guid(device);
- if (!cm_dev->ca_guid)
- goto error1;
+ cm_dev->ca_guid = device->node_guid;
set_bit(IB_MGMT_METHOD_SEND, reg_req.method_mask);
for (i = 1; i <= device->phys_port_cnt; i++) {
@@ -3211,11 +3199,11 @@ static void cm_add_one(struct ib_device *device)
cm_recv_handler,
port);
if (IS_ERR(port->mad_agent))
- goto error2;
+ goto error1;
ret = ib_modify_port(device, i, 0, &port_modify);
if (ret)
- goto error3;
+ goto error2;
}
ib_set_client_data(device, &cm_client, cm_dev);
@@ -3224,9 +3212,9 @@ static void cm_add_one(struct ib_device *device)
write_unlock_irqrestore(&cm.device_lock, flags);
return;
-error3:
- ib_unregister_mad_agent(port->mad_agent);
error2:
+ ib_unregister_mad_agent(port->mad_agent);
+error1:
port_modify.set_port_cap_mask = 0;
port_modify.clr_port_cap_mask = IB_PORT_CM_SUP;
while (--i) {
@@ -3234,7 +3222,6 @@ error2:
ib_modify_port(device, port->port_num, 0, &port_modify);
ib_unregister_mad_agent(port->mad_agent);
}
-error1:
kfree(cm_dev);
}
diff --git a/drivers/infiniband/core/device.c b/drivers/infiniband/core/device.c
index e169e798354b..b2f3cb91d9bc 100644
--- a/drivers/infiniband/core/device.c
+++ b/drivers/infiniband/core/device.c
@@ -38,8 +38,7 @@
#include <linux/errno.h>
#include <linux/slab.h>
#include <linux/init.h>
-
-#include <asm/semaphore.h>
+#include <linux/mutex.h>
#include "core_priv.h"
@@ -57,13 +56,13 @@ static LIST_HEAD(device_list);
static LIST_HEAD(client_list);
/*
- * device_sem protects access to both device_list and client_list.
+ * device_mutex protects access to both device_list and client_list.
* There's no real point to using multiple locks or something fancier
* like an rwsem: we always access both lists, and we're always
* modifying one list or the other list. In any case this is not a
* hot path so there's no point in trying to optimize.
*/
-static DECLARE_MUTEX(device_sem);
+static DEFINE_MUTEX(device_mutex);
static int ib_device_check_mandatory(struct ib_device *device)
{
@@ -221,7 +220,7 @@ int ib_register_device(struct ib_device *device)
{
int ret;
- down(&device_sem);
+ mutex_lock(&device_mutex);
if (strchr(device->name, '%')) {
ret = alloc_name(device->name);
@@ -259,7 +258,7 @@ int ib_register_device(struct ib_device *device)
}
out:
- up(&device_sem);
+ mutex_unlock(&device_mutex);
return ret;
}
EXPORT_SYMBOL(ib_register_device);
@@ -276,7 +275,7 @@ void ib_unregister_device(struct ib_device *device)
struct ib_client_data *context, *tmp;
unsigned long flags;
- down(&device_sem);
+ mutex_lock(&device_mutex);
list_for_each_entry_reverse(client, &client_list, list)
if (client->remove)
@@ -284,7 +283,7 @@ void ib_unregister_device(struct ib_device *device)
list_del(&device->core_list);
- up(&device_sem);
+ mutex_unlock(&device_mutex);
spin_lock_irqsave(&device->client_data_lock, flags);
list_for_each_entry_safe(context, tmp, &device->client_data_list, list)
@@ -312,14 +311,14 @@ int ib_register_client(struct ib_client *client)
{
struct ib_device *device;
- down(&device_sem);
+ mutex_lock(&device_mutex);
list_add_tail(&client->list, &client_list);
list_for_each_entry(device, &device_list, core_list)
if (client->add && !add_client_context(device, client))
client->add(device);
- up(&device_sem);
+ mutex_unlock(&device_mutex);
return 0;
}
@@ -339,7 +338,7 @@ void ib_unregister_client(struct ib_client *client)
struct ib_device *device;
unsigned long flags;
- down(&device_sem);
+ mutex_lock(&device_mutex);
list_for_each_entry(device, &device_list, core_list) {
if (client->remove)
@@ -355,7 +354,7 @@ void ib_unregister_client(struct ib_client *client)
}
list_del(&client->list);
- up(&device_sem);
+ mutex_unlock(&device_mutex);
}
EXPORT_SYMBOL(ib_unregister_client);
diff --git a/drivers/infiniband/core/sysfs.c b/drivers/infiniband/core/sysfs.c
index 08648b1a387e..5982d687a000 100644
--- a/drivers/infiniband/core/sysfs.c
+++ b/drivers/infiniband/core/sysfs.c
@@ -434,24 +434,18 @@ static void ib_device_release(struct class_device *cdev)
kfree(dev);
}
-static int ib_device_hotplug(struct class_device *cdev, char **envp,
- int num_envp, char *buf, int size)
+static int ib_device_uevent(struct class_device *cdev, char **envp,
+ int num_envp, char *buf, int size)
{
struct ib_device *dev = container_of(cdev, struct ib_device, class_dev);
int i = 0, len = 0;
- if (add_hotplug_env_var(envp, num_envp, &i, buf, size, &len,
- "NAME=%s", dev->name))
+ if (add_uevent_var(envp, num_envp, &i, buf, size, &len,
+ "NAME=%s", dev->name))
return -ENOMEM;
/*
- * It might be nice to pass the node GUID to hotplug, but
- * right now the only way to get it is to query the device
- * provider, and this can crash during device removal because
- * we are will be running after driver removal has started.
- * We could add a node_guid field to struct ib_device, or we
- * could just let the hotplug script read the node GUID from
- * sysfs when devices are added.
+ * It would be nice to pass the node GUID with the event...
*/
envp[i] = NULL;
@@ -623,21 +617,15 @@ static ssize_t show_sys_image_guid(struct class_device *cdev, char *buf)
static ssize_t show_node_guid(struct class_device *cdev, char *buf)
{
struct ib_device *dev = container_of(cdev, struct ib_device, class_dev);
- struct ib_device_attr attr;
- ssize_t ret;
if (!ibdev_is_alive(dev))
return -ENODEV;
- ret = ib_query_device(dev, &attr);
- if (ret)
- return ret;
-
return sprintf(buf, "%04x:%04x:%04x:%04x\n",
- be16_to_cpu(((__be16 *) &attr.node_guid)[0]),
- be16_to_cpu(((__be16 *) &attr.node_guid)[1]),
- be16_to_cpu(((__be16 *) &attr.node_guid)[2]),
- be16_to_cpu(((__be16 *) &attr.node_guid)[3]));
+ be16_to_cpu(((__be16 *) &dev->node_guid)[0]),
+ be16_to_cpu(((__be16 *) &dev->node_guid)[1]),
+ be16_to_cpu(((__be16 *) &dev->node_guid)[2]),
+ be16_to_cpu(((__be16 *) &dev->node_guid)[3]));
}
static CLASS_DEVICE_ATTR(node_type, S_IRUGO, show_node_type, NULL);
@@ -653,7 +641,7 @@ static struct class_device_attribute *ib_class_attributes[] = {
static struct class ib_class = {
.name = "infiniband",
.release = ib_device_release,
- .hotplug = ib_device_hotplug,
+ .uevent = ib_device_uevent,
};
int ib_device_register_sysfs(struct ib_device *device)
diff --git a/drivers/infiniband/core/ucm.c b/drivers/infiniband/core/ucm.c
index 6e15787d1de1..e95c4293a496 100644
--- a/drivers/infiniband/core/ucm.c
+++ b/drivers/infiniband/core/ucm.c
@@ -42,6 +42,7 @@
#include <linux/mount.h>
#include <linux/cdev.h>
#include <linux/idr.h>
+#include <linux/mutex.h>
#include <asm/uaccess.h>
@@ -113,7 +114,7 @@ static struct ib_client ucm_client = {
.remove = ib_ucm_remove_one
};
-static DECLARE_MUTEX(ctx_id_mutex);
+static DEFINE_MUTEX(ctx_id_mutex);
static DEFINE_IDR(ctx_id_table);
static DECLARE_BITMAP(dev_map, IB_UCM_MAX_DEVICES);
@@ -121,7 +122,7 @@ static struct ib_ucm_context *ib_ucm_ctx_get(struct ib_ucm_file *file, int id)
{
struct ib_ucm_context *ctx;
- down(&ctx_id_mutex);
+ mutex_lock(&ctx_id_mutex);
ctx = idr_find(&ctx_id_table, id);
if (!ctx)
ctx = ERR_PTR(-ENOENT);
@@ -129,7 +130,7 @@ static struct ib_ucm_context *ib_ucm_ctx_get(struct ib_ucm_file *file, int id)
ctx = ERR_PTR(-EINVAL);
else
atomic_inc(&ctx->ref);
- up(&ctx_id_mutex);
+ mutex_unlock(&ctx_id_mutex);
return ctx;
}
@@ -186,9 +187,9 @@ static struct ib_ucm_context *ib_ucm_ctx_alloc(struct ib_ucm_file *file)
if (!result)
goto error;
- down(&ctx_id_mutex);
+ mutex_lock(&ctx_id_mutex);
result = idr_get_new(&ctx_id_table, ctx, &ctx->id);
- up(&ctx_id_mutex);
+ mutex_unlock(&ctx_id_mutex);
} while (result == -EAGAIN);
if (result)
@@ -550,9 +551,9 @@ static ssize_t ib_ucm_create_id(struct ib_ucm_file *file,
err2:
ib_destroy_cm_id(ctx->cm_id);
err1:
- down(&ctx_id_mutex);
+ mutex_lock(&ctx_id_mutex);
idr_remove(&ctx_id_table, ctx->id);
- up(&ctx_id_mutex);
+ mutex_unlock(&ctx_id_mutex);
kfree(ctx);
return result;
}
@@ -572,7 +573,7 @@ static ssize_t ib_ucm_destroy_id(struct ib_ucm_file *file,
if (copy_from_user(&cmd, inbuf, sizeof(cmd)))
return -EFAULT;
- down(&ctx_id_mutex);
+ mutex_lock(&ctx_id_mutex);
ctx = idr_find(&ctx_id_table, cmd.id);
if (!ctx)
ctx = ERR_PTR(-ENOENT);
@@ -580,7 +581,7 @@ static ssize_t ib_ucm_destroy_id(struct ib_ucm_file *file,
ctx = ERR_PTR(-EINVAL);
else
idr_remove(&ctx_id_table, ctx->id);
- up(&ctx_id_mutex);
+ mutex_unlock(&ctx_id_mutex);
if (IS_ERR(ctx))
return PTR_ERR(ctx);
@@ -1280,9 +1281,9 @@ static int ib_ucm_close(struct inode *inode, struct file *filp)
struct ib_ucm_context, file_list);
up(&file->mutex);
- down(&ctx_id_mutex);
+ mutex_lock(&ctx_id_mutex);
idr_remove(&ctx_id_table, ctx->id);
- up(&ctx_id_mutex);
+ mutex_unlock(&ctx_id_mutex);
ib_destroy_cm_id(ctx->cm_id);
ib_ucm_cleanup_events(ctx);
diff --git a/drivers/infiniband/core/user_mad.c b/drivers/infiniband/core/user_mad.c
index e73f81c22381..c908de8db5a9 100644
--- a/drivers/infiniband/core/user_mad.c
+++ b/drivers/infiniband/core/user_mad.c
@@ -197,8 +197,8 @@ static void send_handler(struct ib_mad_agent *agent,
memcpy(timeout->mad.data, packet->mad.data,
sizeof (struct ib_mad_hdr));
- if (!queue_packet(file, agent, timeout))
- return;
+ if (queue_packet(file, agent, timeout))
+ kfree(timeout);
}
out:
kfree(packet);
@@ -310,7 +310,7 @@ static ssize_t ib_umad_write(struct file *filp, const char __user *buf,
u8 method;
__be64 *tid;
int ret, length, hdr_len, copy_offset;
- int rmpp_active = 0;
+ int rmpp_active, has_rmpp_header;
if (count < sizeof (struct ib_user_mad) + IB_MGMT_RMPP_HDR)
return -EINVAL;
@@ -360,28 +360,31 @@ static ssize_t ib_umad_write(struct file *filp, const char __user *buf,
}
rmpp_mad = (struct ib_rmpp_mad *) packet->mad.data;
- if (ib_get_rmpp_flags(&rmpp_mad->rmpp_hdr) & IB_MGMT_RMPP_FLAG_ACTIVE) {
- /* RMPP active */
- if (!agent->rmpp_version) {
- ret = -EINVAL;
- goto err_ah;
- }
-
- /* Validate that the management class can support RMPP */
- if (rmpp_mad->mad_hdr.mgmt_class == IB_MGMT_CLASS_SUBN_ADM) {
- hdr_len = IB_MGMT_SA_HDR;
- } else if ((rmpp_mad->mad_hdr.mgmt_class >= IB_MGMT_CLASS_VENDOR_RANGE2_START) &&
- (rmpp_mad->mad_hdr.mgmt_class <= IB_MGMT_CLASS_VENDOR_RANGE2_END)) {
- hdr_len = IB_MGMT_VENDOR_HDR;
- } else {
- ret = -EINVAL;
- goto err_ah;
- }
- rmpp_active = 1;
+ if (rmpp_mad->mad_hdr.mgmt_class == IB_MGMT_CLASS_SUBN_ADM) {
+ hdr_len = IB_MGMT_SA_HDR;
copy_offset = IB_MGMT_RMPP_HDR;
+ has_rmpp_header = 1;
+ } else if (rmpp_mad->mad_hdr.mgmt_class >= IB_MGMT_CLASS_VENDOR_RANGE2_START &&
+ rmpp_mad->mad_hdr.mgmt_class <= IB_MGMT_CLASS_VENDOR_RANGE2_END) {
+ hdr_len = IB_MGMT_VENDOR_HDR;
+ copy_offset = IB_MGMT_RMPP_HDR;
+ has_rmpp_header = 1;
} else {
hdr_len = IB_MGMT_MAD_HDR;
copy_offset = IB_MGMT_MAD_HDR;
+ has_rmpp_header = 0;
+ }
+
+ if (has_rmpp_header)
+ rmpp_active = ib_get_rmpp_flags(&rmpp_mad->rmpp_hdr) &
+ IB_MGMT_RMPP_FLAG_ACTIVE;
+ else
+ rmpp_active = 0;
+
+ /* Validate that the management class can support RMPP */
+ if (rmpp_active && !agent->rmpp_version) {
+ ret = -EINVAL;
+ goto err_ah;
}
packet->msg = ib_create_send_mad(agent,
diff --git a/drivers/infiniband/core/uverbs.h b/drivers/infiniband/core/uverbs.h
index ecb830127865..f7eecbc6af6c 100644
--- a/drivers/infiniband/core/uverbs.h
+++ b/drivers/infiniband/core/uverbs.h
@@ -41,6 +41,7 @@
#include <linux/kref.h>
#include <linux/idr.h>
+#include <linux/mutex.h>
#include <rdma/ib_verbs.h>
#include <rdma/ib_user_verbs.h>
@@ -88,7 +89,7 @@ struct ib_uverbs_event_file {
struct ib_uverbs_file {
struct kref ref;
- struct semaphore mutex;
+ struct mutex mutex;
struct ib_uverbs_device *device;
struct ib_ucontext *ucontext;
struct ib_event_handler event_handler;
@@ -105,12 +106,23 @@ struct ib_uverbs_event {
u32 *counter;
};
+struct ib_uverbs_mcast_entry {
+ struct list_head list;
+ union ib_gid gid;
+ u16 lid;
+};
+
struct ib_uevent_object {
struct ib_uobject uobject;
struct list_head event_list;
u32 events_reported;
};
+struct ib_uqp_object {
+ struct ib_uevent_object uevent;
+ struct list_head mcast_list;
+};
+
struct ib_ucq_object {
struct ib_uobject uobject;
struct ib_uverbs_file *uverbs_file;
@@ -120,7 +132,7 @@ struct ib_ucq_object {
u32 async_events_reported;
};
-extern struct semaphore ib_uverbs_idr_mutex;
+extern struct mutex ib_uverbs_idr_mutex;
extern struct idr ib_uverbs_pd_idr;
extern struct idr ib_uverbs_mr_idr;
extern struct idr ib_uverbs_mw_idr;
diff --git a/drivers/infiniband/core/uverbs_cmd.c b/drivers/infiniband/core/uverbs_cmd.c
index ed45da892b1c..407b6284d7d5 100644
--- a/drivers/infiniband/core/uverbs_cmd.c
+++ b/drivers/infiniband/core/uverbs_cmd.c
@@ -67,7 +67,7 @@ ssize_t ib_uverbs_get_context(struct ib_uverbs_file *file,
if (copy_from_user(&cmd, buf, sizeof cmd))
return -EFAULT;
- down(&file->mutex);
+ mutex_lock(&file->mutex);
if (file->ucontext) {
ret = -EINVAL;
@@ -119,7 +119,7 @@ ssize_t ib_uverbs_get_context(struct ib_uverbs_file *file,
fd_install(resp.async_fd, filp);
- up(&file->mutex);
+ mutex_unlock(&file->mutex);
return in_len;
@@ -131,7 +131,7 @@ err_free:
ibdev->dealloc_ucontext(ucontext);
err:
- up(&file->mutex);
+ mutex_unlock(&file->mutex);
return ret;
}
@@ -157,7 +157,7 @@ ssize_t ib_uverbs_query_device(struct ib_uverbs_file *file,
memset(&resp, 0, sizeof resp);
resp.fw_ver = attr.fw_ver;
- resp.node_guid = attr.node_guid;
+ resp.node_guid = file->device->ib_dev->node_guid;
resp.sys_image_guid = attr.sys_image_guid;
resp.max_mr_size = attr.max_mr_size;
resp.page_size_cap = attr.page_size_cap;
@@ -290,7 +290,7 @@ ssize_t ib_uverbs_alloc_pd(struct ib_uverbs_file *file,
pd->uobject = uobj;
atomic_set(&pd->usecnt, 0);
- down(&ib_uverbs_idr_mutex);
+ mutex_lock(&ib_uverbs_idr_mutex);
retry:
if (!idr_pre_get(&ib_uverbs_pd_idr, GFP_KERNEL)) {
@@ -314,11 +314,11 @@ retry:
goto err_idr;
}
- down(&file->mutex);
+ mutex_lock(&file->mutex);
list_add_tail(&uobj->list, &file->ucontext->pd_list);
- up(&file->mutex);
+ mutex_unlock(&file->mutex);
- up(&ib_uverbs_idr_mutex);
+ mutex_unlock(&ib_uverbs_idr_mutex);
return in_len;
@@ -326,7 +326,7 @@ err_idr:
idr_remove(&ib_uverbs_pd_idr, uobj->id);
err_up:
- up(&ib_uverbs_idr_mutex);
+ mutex_unlock(&ib_uverbs_idr_mutex);
ib_dealloc_pd(pd);
err:
@@ -346,7 +346,7 @@ ssize_t ib_uverbs_dealloc_pd(struct ib_uverbs_file *file,
if (copy_from_user(&cmd, buf, sizeof cmd))
return -EFAULT;
- down(&ib_uverbs_idr_mutex);
+ mutex_lock(&ib_uverbs_idr_mutex);
pd = idr_find(&ib_uverbs_pd_idr, cmd.pd_handle);
if (!pd || pd->uobject->context != file->ucontext)
@@ -360,14 +360,14 @@ ssize_t ib_uverbs_dealloc_pd(struct ib_uverbs_file *file,
idr_remove(&ib_uverbs_pd_idr, cmd.pd_handle);
- down(&file->mutex);
+ mutex_lock(&file->mutex);
list_del(&uobj->list);
- up(&file->mutex);
+ mutex_unlock(&file->mutex);
kfree(uobj);
out:
- up(&ib_uverbs_idr_mutex);
+ mutex_unlock(&ib_uverbs_idr_mutex);
return ret ? ret : in_len;
}
@@ -426,7 +426,7 @@ ssize_t ib_uverbs_reg_mr(struct ib_uverbs_file *file,
obj->umem.virt_base = cmd.hca_va;
- down(&ib_uverbs_idr_mutex);
+ mutex_lock(&ib_uverbs_idr_mutex);
pd = idr_find(&ib_uverbs_pd_idr, cmd.pd_handle);
if (!pd || pd->uobject->context != file->ucontext) {
@@ -476,11 +476,11 @@ retry:
goto err_idr;
}
- down(&file->mutex);
+ mutex_lock(&file->mutex);
list_add_tail(&obj->uobject.list, &file->ucontext->mr_list);
- up(&file->mutex);
+ mutex_unlock(&file->mutex);
- up(&ib_uverbs_idr_mutex);
+ mutex_unlock(&ib_uverbs_idr_mutex);
return in_len;
@@ -489,9 +489,10 @@ err_idr:
err_unreg:
ib_dereg_mr(mr);
+ atomic_dec(&pd->usecnt);
err_up:
- up(&ib_uverbs_idr_mutex);
+ mutex_unlock(&ib_uverbs_idr_mutex);
ib_umem_release(file->device->ib_dev, &obj->umem);
@@ -512,7 +513,7 @@ ssize_t ib_uverbs_dereg_mr(struct ib_uverbs_file *file,
if (copy_from_user(&cmd, buf, sizeof cmd))
return -EFAULT;
- down(&ib_uverbs_idr_mutex);
+ mutex_lock(&ib_uverbs_idr_mutex);
mr = idr_find(&ib_uverbs_mr_idr, cmd.mr_handle);
if (!mr || mr->uobject->context != file->ucontext)
@@ -526,15 +527,15 @@ ssize_t ib_uverbs_dereg_mr(struct ib_uverbs_file *file,
idr_remove(&ib_uverbs_mr_idr, cmd.mr_handle);
- down(&file->mutex);
+ mutex_lock(&file->mutex);
list_del(&memobj->uobject.list);
- up(&file->mutex);
+ mutex_unlock(&file->mutex);
ib_umem_release(file->device->ib_dev, &memobj->umem);
kfree(memobj);
out:
- up(&ib_uverbs_idr_mutex);
+ mutex_unlock(&ib_uverbs_idr_mutex);
return ret ? ret : in_len;
}
@@ -593,13 +594,18 @@ ssize_t ib_uverbs_create_cq(struct ib_uverbs_file *file,
if (cmd.comp_vector >= file->device->num_comp_vectors)
return -EINVAL;
- if (cmd.comp_channel >= 0)
- ev_file = ib_uverbs_lookup_comp_file(cmd.comp_channel);
-
uobj = kmalloc(sizeof *uobj, GFP_KERNEL);
if (!uobj)
return -ENOMEM;
+ if (cmd.comp_channel >= 0) {
+ ev_file = ib_uverbs_lookup_comp_file(cmd.comp_channel);
+ if (!ev_file) {
+ ret = -EINVAL;
+ goto err;
+ }
+ }
+
uobj->uobject.user_handle = cmd.user_handle;
uobj->uobject.context = file->ucontext;
uobj->uverbs_file = file;
@@ -622,7 +628,7 @@ ssize_t ib_uverbs_create_cq(struct ib_uverbs_file *file,
cq->cq_context = ev_file;
atomic_set(&cq->usecnt, 0);
- down(&ib_uverbs_idr_mutex);
+ mutex_lock(&ib_uverbs_idr_mutex);
retry:
if (!idr_pre_get(&ib_uverbs_cq_idr, GFP_KERNEL)) {
@@ -647,11 +653,11 @@ retry:
goto err_idr;
}
- down(&file->mutex);
+ mutex_lock(&file->mutex);
list_add_tail(&uobj->uobject.list, &file->ucontext->cq_list);
- up(&file->mutex);
+ mutex_unlock(&file->mutex);
- up(&ib_uverbs_idr_mutex);
+ mutex_unlock(&ib_uverbs_idr_mutex);
return in_len;
@@ -659,10 +665,12 @@ err_idr:
idr_remove(&ib_uverbs_cq_idr, uobj->uobject.id);
err_up:
- up(&ib_uverbs_idr_mutex);
+ mutex_unlock(&ib_uverbs_idr_mutex);
ib_destroy_cq(cq);
err:
+ if (ev_file)
+ ib_uverbs_release_ucq(file, ev_file, uobj);
kfree(uobj);
return ret;
}
@@ -693,7 +701,7 @@ ssize_t ib_uverbs_poll_cq(struct ib_uverbs_file *file,
goto out_wc;
}
- down(&ib_uverbs_idr_mutex);
+ mutex_lock(&ib_uverbs_idr_mutex);
cq = idr_find(&ib_uverbs_cq_idr, cmd.cq_handle);
if (!cq || cq->uobject->context != file->ucontext) {
ret = -EINVAL;
@@ -723,7 +731,7 @@ ssize_t ib_uverbs_poll_cq(struct ib_uverbs_file *file,
ret = -EFAULT;
out:
- up(&ib_uverbs_idr_mutex);
+ mutex_unlock(&ib_uverbs_idr_mutex);
kfree(resp);
out_wc:
@@ -742,14 +750,14 @@ ssize_t ib_uverbs_req_notify_cq(struct ib_uverbs_file *file,
if (copy_from_user(&cmd, buf, sizeof cmd))
return -EFAULT;
- down(&ib_uverbs_idr_mutex);
+ mutex_lock(&ib_uverbs_idr_mutex);
cq = idr_find(&ib_uverbs_cq_idr, cmd.cq_handle);
if (cq && cq->uobject->context == file->ucontext) {
ib_req_notify_cq(cq, cmd.solicited_only ?
IB_CQ_SOLICITED : IB_CQ_NEXT_COMP);
ret = in_len;
}
- up(&ib_uverbs_idr_mutex);
+ mutex_unlock(&ib_uverbs_idr_mutex);
return ret;
}
@@ -771,7 +779,7 @@ ssize_t ib_uverbs_destroy_cq(struct ib_uverbs_file *file,
memset(&resp, 0, sizeof resp);
- down(&ib_uverbs_idr_mutex);
+ mutex_lock(&ib_uverbs_idr_mutex);
cq = idr_find(&ib_uverbs_cq_idr, cmd.cq_handle);
if (!cq || cq->uobject->context != file->ucontext)
@@ -787,9 +795,9 @@ ssize_t ib_uverbs_destroy_cq(struct ib_uverbs_file *file,
idr_remove(&ib_uverbs_cq_idr, cmd.cq_handle);
- down(&file->mutex);
+ mutex_lock(&file->mutex);
list_del(&uobj->uobject.list);
- up(&file->mutex);
+ mutex_unlock(&file->mutex);
ib_uverbs_release_ucq(file, ev_file, uobj);
@@ -803,7 +811,7 @@ ssize_t ib_uverbs_destroy_cq(struct ib_uverbs_file *file,
ret = -EFAULT;
out:
- up(&ib_uverbs_idr_mutex);
+ mutex_unlock(&ib_uverbs_idr_mutex);
return ret ? ret : in_len;
}
@@ -815,7 +823,7 @@ ssize_t ib_uverbs_create_qp(struct ib_uverbs_file *file,
struct ib_uverbs_create_qp cmd;
struct ib_uverbs_create_qp_resp resp;
struct ib_udata udata;
- struct ib_uevent_object *uobj;
+ struct ib_uqp_object *uobj;
struct ib_pd *pd;
struct ib_cq *scq, *rcq;
struct ib_srq *srq;
@@ -837,7 +845,7 @@ ssize_t ib_uverbs_create_qp(struct ib_uverbs_file *file,
if (!uobj)
return -ENOMEM;
- down(&ib_uverbs_idr_mutex);
+ mutex_lock(&ib_uverbs_idr_mutex);
pd = idr_find(&ib_uverbs_pd_idr, cmd.pd_handle);
scq = idr_find(&ib_uverbs_cq_idr, cmd.send_cq_handle);
@@ -866,10 +874,11 @@ ssize_t ib_uverbs_create_qp(struct ib_uverbs_file *file,
attr.cap.max_recv_sge = cmd.max_recv_sge;
attr.cap.max_inline_data = cmd.max_inline_data;
- uobj->uobject.user_handle = cmd.user_handle;
- uobj->uobject.context = file->ucontext;
- uobj->events_reported = 0;
- INIT_LIST_HEAD(&uobj->event_list);
+ uobj->uevent.uobject.user_handle = cmd.user_handle;
+ uobj->uevent.uobject.context = file->ucontext;
+ uobj->uevent.events_reported = 0;
+ INIT_LIST_HEAD(&uobj->uevent.event_list);
+ INIT_LIST_HEAD(&uobj->mcast_list);
qp = pd->device->create_qp(pd, &attr, &udata);
if (IS_ERR(qp)) {
@@ -882,7 +891,7 @@ ssize_t ib_uverbs_create_qp(struct ib_uverbs_file *file,
qp->send_cq = attr.send_cq;
qp->recv_cq = attr.recv_cq;
qp->srq = attr.srq;
- qp->uobject = &uobj->uobject;
+ qp->uobject = &uobj->uevent.uobject;
qp->event_handler = attr.event_handler;
qp->qp_context = attr.qp_context;
qp->qp_type = attr.qp_type;
@@ -901,14 +910,14 @@ retry:
goto err_destroy;
}
- ret = idr_get_new(&ib_uverbs_qp_idr, qp, &uobj->uobject.id);
+ ret = idr_get_new(&ib_uverbs_qp_idr, qp, &uobj->uevent.uobject.id);
if (ret == -EAGAIN)
goto retry;
if (ret)
goto err_destroy;
- resp.qp_handle = uobj->uobject.id;
+ resp.qp_handle = uobj->uevent.uobject.id;
resp.max_recv_sge = attr.cap.max_recv_sge;
resp.max_send_sge = attr.cap.max_send_sge;
resp.max_recv_wr = attr.cap.max_recv_wr;
@@ -921,22 +930,27 @@ retry:
goto err_idr;
}
- down(&file->mutex);
- list_add_tail(&uobj->uobject.list, &file->ucontext->qp_list);
- up(&file->mutex);
+ mutex_lock(&file->mutex);
+ list_add_tail(&uobj->uevent.uobject.list, &file->ucontext->qp_list);
+ mutex_unlock(&file->mutex);
- up(&ib_uverbs_idr_mutex);
+ mutex_unlock(&ib_uverbs_idr_mutex);
return in_len;
err_idr:
- idr_remove(&ib_uverbs_qp_idr, uobj->uobject.id);
+ idr_remove(&ib_uverbs_qp_idr, uobj->uevent.uobject.id);
err_destroy:
ib_destroy_qp(qp);
+ atomic_dec(&pd->usecnt);
+ atomic_dec(&attr.send_cq->usecnt);
+ atomic_dec(&attr.recv_cq->usecnt);
+ if (attr.srq)
+ atomic_dec(&attr.srq->usecnt);
err_up:
- up(&ib_uverbs_idr_mutex);
+ mutex_unlock(&ib_uverbs_idr_mutex);
kfree(uobj);
return ret;
@@ -958,7 +972,7 @@ ssize_t ib_uverbs_modify_qp(struct ib_uverbs_file *file,
if (!attr)
return -ENOMEM;
- down(&ib_uverbs_idr_mutex);
+ mutex_lock(&ib_uverbs_idr_mutex);
qp = idr_find(&ib_uverbs_qp_idr, cmd.qp_handle);
if (!qp || qp->uobject->context != file->ucontext) {
@@ -1019,7 +1033,7 @@ ssize_t ib_uverbs_modify_qp(struct ib_uverbs_file *file,
ret = in_len;
out:
- up(&ib_uverbs_idr_mutex);
+ mutex_unlock(&ib_uverbs_idr_mutex);
kfree(attr);
return ret;
@@ -1032,7 +1046,7 @@ ssize_t ib_uverbs_destroy_qp(struct ib_uverbs_file *file,
struct ib_uverbs_destroy_qp cmd;
struct ib_uverbs_destroy_qp_resp resp;
struct ib_qp *qp;
- struct ib_uevent_object *uobj;
+ struct ib_uqp_object *uobj;
int ret = -EINVAL;
if (copy_from_user(&cmd, buf, sizeof cmd))
@@ -1040,13 +1054,18 @@ ssize_t ib_uverbs_destroy_qp(struct ib_uverbs_file *file,
memset(&resp, 0, sizeof resp);
- down(&ib_uverbs_idr_mutex);
+ mutex_lock(&ib_uverbs_idr_mutex);
qp = idr_find(&ib_uverbs_qp_idr, cmd.qp_handle);
if (!qp || qp->uobject->context != file->ucontext)
goto out;
- uobj = container_of(qp->uobject, struct ib_uevent_object, uobject);
+ uobj = container_of(qp->uobject, struct ib_uqp_object, uevent.uobject);
+
+ if (!list_empty(&uobj->mcast_list)) {
+ ret = -EBUSY;
+ goto out;
+ }
ret = ib_destroy_qp(qp);
if (ret)
@@ -1054,13 +1073,13 @@ ssize_t ib_uverbs_destroy_qp(struct ib_uverbs_file *file,
idr_remove(&ib_uverbs_qp_idr, cmd.qp_handle);
- down(&file->mutex);
- list_del(&uobj->uobject.list);
- up(&file->mutex);
+ mutex_lock(&file->mutex);
+ list_del(&uobj->uevent.uobject.list);
+ mutex_unlock(&file->mutex);
- ib_uverbs_release_uevent(file, uobj);
+ ib_uverbs_release_uevent(file, &uobj->uevent);
- resp.events_reported = uobj->events_reported;
+ resp.events_reported = uobj->uevent.events_reported;
kfree(uobj);
@@ -1069,7 +1088,7 @@ ssize_t ib_uverbs_destroy_qp(struct ib_uverbs_file *file,
ret = -EFAULT;
out:
- up(&ib_uverbs_idr_mutex);
+ mutex_unlock(&ib_uverbs_idr_mutex);
return ret ? ret : in_len;
}
@@ -1100,7 +1119,7 @@ ssize_t ib_uverbs_post_send(struct ib_uverbs_file *file,
if (!user_wr)
return -ENOMEM;
- down(&ib_uverbs_idr_mutex);
+ mutex_lock(&ib_uverbs_idr_mutex);
qp = idr_find(&ib_uverbs_qp_idr, cmd.qp_handle);
if (!qp || qp->uobject->context != file->ucontext)
@@ -1205,7 +1224,7 @@ ssize_t ib_uverbs_post_send(struct ib_uverbs_file *file,
ret = -EFAULT;
out:
- up(&ib_uverbs_idr_mutex);
+ mutex_unlock(&ib_uverbs_idr_mutex);
while (wr) {
next = wr->next;
@@ -1322,7 +1341,7 @@ ssize_t ib_uverbs_post_recv(struct ib_uverbs_file *file,
if (IS_ERR(wr))
return PTR_ERR(wr);
- down(&ib_uverbs_idr_mutex);
+ mutex_lock(&ib_uverbs_idr_mutex);
qp = idr_find(&ib_uverbs_qp_idr, cmd.qp_handle);
if (!qp || qp->uobject->context != file->ucontext)
@@ -1343,7 +1362,7 @@ ssize_t ib_uverbs_post_recv(struct ib_uverbs_file *file,
ret = -EFAULT;
out:
- up(&ib_uverbs_idr_mutex);
+ mutex_unlock(&ib_uverbs_idr_mutex);
while (wr) {
next = wr->next;
@@ -1373,7 +1392,7 @@ ssize_t ib_uverbs_post_srq_recv(struct ib_uverbs_file *file,
if (IS_ERR(wr))
return PTR_ERR(wr);
- down(&ib_uverbs_idr_mutex);
+ mutex_lock(&ib_uverbs_idr_mutex);
srq = idr_find(&ib_uverbs_srq_idr, cmd.srq_handle);
if (!srq || srq->uobject->context != file->ucontext)
@@ -1394,7 +1413,7 @@ ssize_t ib_uverbs_post_srq_recv(struct ib_uverbs_file *file,
ret = -EFAULT;
out:
- up(&ib_uverbs_idr_mutex);
+ mutex_unlock(&ib_uverbs_idr_mutex);
while (wr) {
next = wr->next;
@@ -1427,7 +1446,7 @@ ssize_t ib_uverbs_create_ah(struct ib_uverbs_file *file,
if (!uobj)
return -ENOMEM;
- down(&ib_uverbs_idr_mutex);
+ mutex_lock(&ib_uverbs_idr_mutex);
pd = idr_find(&ib_uverbs_pd_idr, cmd.pd_handle);
if (!pd || pd->uobject->context != file->ucontext) {
@@ -1442,6 +1461,7 @@ ssize_t ib_uverbs_create_ah(struct ib_uverbs_file *file,
attr.sl = cmd.attr.sl;
attr.src_path_bits = cmd.attr.src_path_bits;
attr.static_rate = cmd.attr.static_rate;
+ attr.ah_flags = cmd.attr.is_global ? IB_AH_GRH : 0;
attr.port_num = cmd.attr.port_num;
attr.grh.flow_label = cmd.attr.grh.flow_label;
attr.grh.sgid_index = cmd.attr.grh.sgid_index;
@@ -1478,11 +1498,11 @@ retry:
goto err_idr;
}
- down(&file->mutex);
+ mutex_lock(&file->mutex);
list_add_tail(&uobj->list, &file->ucontext->ah_list);
- up(&file->mutex);
+ mutex_unlock(&file->mutex);
- up(&ib_uverbs_idr_mutex);
+ mutex_unlock(&ib_uverbs_idr_mutex);
return in_len;
@@ -1493,7 +1513,7 @@ err_destroy:
ib_destroy_ah(ah);
err_up:
- up(&ib_uverbs_idr_mutex);
+ mutex_unlock(&ib_uverbs_idr_mutex);
kfree(uobj);
return ret;
@@ -1510,7 +1530,7 @@ ssize_t ib_uverbs_destroy_ah(struct ib_uverbs_file *file,
if (copy_from_user(&cmd, buf, sizeof cmd))
return -EFAULT;
- down(&ib_uverbs_idr_mutex);
+ mutex_lock(&ib_uverbs_idr_mutex);
ah = idr_find(&ib_uverbs_ah_idr, cmd.ah_handle);
if (!ah || ah->uobject->context != file->ucontext)
@@ -1524,14 +1544,14 @@ ssize_t ib_uverbs_destroy_ah(struct ib_uverbs_file *file,
idr_remove(&ib_uverbs_ah_idr, cmd.ah_handle);
- down(&file->mutex);
+ mutex_lock(&file->mutex);
list_del(&uobj->list);
- up(&file->mutex);
+ mutex_unlock(&file->mutex);
kfree(uobj);
out:
- up(&ib_uverbs_idr_mutex);
+ mutex_unlock(&ib_uverbs_idr_mutex);
return ret ? ret : in_len;
}
@@ -1542,18 +1562,47 @@ ssize_t ib_uverbs_attach_mcast(struct ib_uverbs_file *file,
{
struct ib_uverbs_attach_mcast cmd;
struct ib_qp *qp;
+ struct ib_uqp_object *uobj;
+ struct ib_uverbs_mcast_entry *mcast;
int ret = -EINVAL;
if (copy_from_user(&cmd, buf, sizeof cmd))
return -EFAULT;
- down(&ib_uverbs_idr_mutex);
+ mutex_lock(&ib_uverbs_idr_mutex);
qp = idr_find(&ib_uverbs_qp_idr, cmd.qp_handle);
- if (qp && qp->uobject->context == file->ucontext)
- ret = ib_attach_mcast(qp, (union ib_gid *) cmd.gid, cmd.mlid);
+ if (!qp || qp->uobject->context != file->ucontext)
+ goto out;
- up(&ib_uverbs_idr_mutex);
+ uobj = container_of(qp->uobject, struct ib_uqp_object, uevent.uobject);
+
+ list_for_each_entry(mcast, &uobj->mcast_list, list)
+ if (cmd.mlid == mcast->lid &&
+ !memcmp(cmd.gid, mcast->gid.raw, sizeof mcast->gid.raw)) {
+ ret = 0;
+ goto out;
+ }
+
+ mcast = kmalloc(sizeof *mcast, GFP_KERNEL);
+ if (!mcast) {
+ ret = -ENOMEM;
+ goto out;
+ }
+
+ mcast->lid = cmd.mlid;
+ memcpy(mcast->gid.raw, cmd.gid, sizeof mcast->gid.raw);
+
+ ret = ib_attach_mcast(qp, &mcast->gid, cmd.mlid);
+ if (!ret) {
+ uobj = container_of(qp->uobject, struct ib_uqp_object,
+ uevent.uobject);
+ list_add_tail(&mcast->list, &uobj->mcast_list);
+ } else
+ kfree(mcast);
+
+out:
+ mutex_unlock(&ib_uverbs_idr_mutex);
return ret ? ret : in_len;
}
@@ -1563,19 +1612,36 @@ ssize_t ib_uverbs_detach_mcast(struct ib_uverbs_file *file,
int out_len)
{
struct ib_uverbs_detach_mcast cmd;
+ struct ib_uqp_object *uobj;
struct ib_qp *qp;
+ struct ib_uverbs_mcast_entry *mcast;
int ret = -EINVAL;
if (copy_from_user(&cmd, buf, sizeof cmd))
return -EFAULT;
- down(&ib_uverbs_idr_mutex);
+ mutex_lock(&ib_uverbs_idr_mutex);
qp = idr_find(&ib_uverbs_qp_idr, cmd.qp_handle);
- if (qp && qp->uobject->context == file->ucontext)
- ret = ib_detach_mcast(qp, (union ib_gid *) cmd.gid, cmd.mlid);
+ if (!qp || qp->uobject->context != file->ucontext)
+ goto out;
+
+ ret = ib_detach_mcast(qp, (union ib_gid *) cmd.gid, cmd.mlid);
+ if (ret)
+ goto out;
+
+ uobj = container_of(qp->uobject, struct ib_uqp_object, uevent.uobject);
- up(&ib_uverbs_idr_mutex);
+ list_for_each_entry(mcast, &uobj->mcast_list, list)
+ if (cmd.mlid == mcast->lid &&
+ !memcmp(cmd.gid, mcast->gid.raw, sizeof mcast->gid.raw)) {
+ list_del(&mcast->list);
+ kfree(mcast);
+ break;
+ }
+
+out:
+ mutex_unlock(&ib_uverbs_idr_mutex);
return ret ? ret : in_len;
}
@@ -1607,7 +1673,7 @@ ssize_t ib_uverbs_create_srq(struct ib_uverbs_file *file,
if (!uobj)
return -ENOMEM;
- down(&ib_uverbs_idr_mutex);
+ mutex_lock(&ib_uverbs_idr_mutex);
pd = idr_find(&ib_uverbs_pd_idr, cmd.pd_handle);
@@ -1664,11 +1730,11 @@ retry:
goto err_idr;
}
- down(&file->mutex);
+ mutex_lock(&file->mutex);
list_add_tail(&uobj->uobject.list, &file->ucontext->srq_list);
- up(&file->mutex);
+ mutex_unlock(&file->mutex);
- up(&ib_uverbs_idr_mutex);
+ mutex_unlock(&ib_uverbs_idr_mutex);
return in_len;
@@ -1677,9 +1743,10 @@ err_idr:
err_destroy:
ib_destroy_srq(srq);
+ atomic_dec(&pd->usecnt);
err_up:
- up(&ib_uverbs_idr_mutex);
+ mutex_unlock(&ib_uverbs_idr_mutex);
kfree(uobj);
return ret;
@@ -1697,7 +1764,7 @@ ssize_t ib_uverbs_modify_srq(struct ib_uverbs_file *file,
if (copy_from_user(&cmd, buf, sizeof cmd))
return -EFAULT;
- down(&ib_uverbs_idr_mutex);
+ mutex_lock(&ib_uverbs_idr_mutex);
srq = idr_find(&ib_uverbs_srq_idr, cmd.srq_handle);
if (!srq || srq->uobject->context != file->ucontext) {
@@ -1711,7 +1778,7 @@ ssize_t ib_uverbs_modify_srq(struct ib_uverbs_file *file,
ret = ib_modify_srq(srq, &attr, cmd.attr_mask);
out:
- up(&ib_uverbs_idr_mutex);
+ mutex_unlock(&ib_uverbs_idr_mutex);
return ret ? ret : in_len;
}
@@ -1729,7 +1796,7 @@ ssize_t ib_uverbs_destroy_srq(struct ib_uverbs_file *file,
if (copy_from_user(&cmd, buf, sizeof cmd))
return -EFAULT;
- down(&ib_uverbs_idr_mutex);
+ mutex_lock(&ib_uverbs_idr_mutex);
memset(&resp, 0, sizeof resp);
@@ -1745,9 +1812,9 @@ ssize_t ib_uverbs_destroy_srq(struct ib_uverbs_file *file,
idr_remove(&ib_uverbs_srq_idr, cmd.srq_handle);
- down(&file->mutex);
+ mutex_lock(&file->mutex);
list_del(&uobj->uobject.list);
- up(&file->mutex);
+ mutex_unlock(&file->mutex);
ib_uverbs_release_uevent(file, uobj);
@@ -1760,7 +1827,7 @@ ssize_t ib_uverbs_destroy_srq(struct ib_uverbs_file *file,
ret = -EFAULT;
out:
- up(&ib_uverbs_idr_mutex);
+ mutex_unlock(&ib_uverbs_idr_mutex);
return ret ? ret : in_len;
}
diff --git a/drivers/infiniband/core/uverbs_main.c b/drivers/infiniband/core/uverbs_main.c
index de6581d7cb8d..96ea79b63df7 100644
--- a/drivers/infiniband/core/uverbs_main.c
+++ b/drivers/infiniband/core/uverbs_main.c
@@ -66,7 +66,7 @@ enum {
static struct class *uverbs_class;
-DECLARE_MUTEX(ib_uverbs_idr_mutex);
+DEFINE_MUTEX(ib_uverbs_idr_mutex);
DEFINE_IDR(ib_uverbs_pd_idr);
DEFINE_IDR(ib_uverbs_mr_idr);
DEFINE_IDR(ib_uverbs_mw_idr);
@@ -160,6 +160,18 @@ void ib_uverbs_release_uevent(struct ib_uverbs_file *file,
spin_unlock_irq(&file->async_file->lock);
}
+static void ib_uverbs_detach_umcast(struct ib_qp *qp,
+ struct ib_uqp_object *uobj)
+{
+ struct ib_uverbs_mcast_entry *mcast, *tmp;
+
+ list_for_each_entry_safe(mcast, tmp, &uobj->mcast_list, list) {
+ ib_detach_mcast(qp, &mcast->gid, mcast->lid);
+ list_del(&mcast->list);
+ kfree(mcast);
+ }
+}
+
static int ib_uverbs_cleanup_ucontext(struct ib_uverbs_file *file,
struct ib_ucontext *context)
{
@@ -168,7 +180,7 @@ static int ib_uverbs_cleanup_ucontext(struct ib_uverbs_file *file,
if (!context)
return 0;
- down(&ib_uverbs_idr_mutex);
+ mutex_lock(&ib_uverbs_idr_mutex);
list_for_each_entry_safe(uobj, tmp, &context->ah_list, list) {
struct ib_ah *ah = idr_find(&ib_uverbs_ah_idr, uobj->id);
@@ -180,13 +192,14 @@ static int ib_uverbs_cleanup_ucontext(struct ib_uverbs_file *file,
list_for_each_entry_safe(uobj, tmp, &context->qp_list, list) {
struct ib_qp *qp = idr_find(&ib_uverbs_qp_idr, uobj->id);
- struct ib_uevent_object *uevent =
- container_of(uobj, struct ib_uevent_object, uobject);
+ struct ib_uqp_object *uqp =
+ container_of(uobj, struct ib_uqp_object, uevent.uobject);
idr_remove(&ib_uverbs_qp_idr, uobj->id);
+ ib_uverbs_detach_umcast(qp, uqp);
ib_destroy_qp(qp);
list_del(&uobj->list);
- ib_uverbs_release_uevent(file, uevent);
- kfree(uevent);
+ ib_uverbs_release_uevent(file, &uqp->uevent);
+ kfree(uqp);
}
list_for_each_entry_safe(uobj, tmp, &context->cq_list, list) {
@@ -237,7 +250,7 @@ static int ib_uverbs_cleanup_ucontext(struct ib_uverbs_file *file,
kfree(uobj);
}
- up(&ib_uverbs_idr_mutex);
+ mutex_unlock(&ib_uverbs_idr_mutex);
return context->device->dealloc_ucontext(context);
}
@@ -640,7 +653,7 @@ static int ib_uverbs_open(struct inode *inode, struct file *filp)
file->ucontext = NULL;
file->async_file = NULL;
kref_init(&file->ref);
- init_MUTEX(&file->mutex);
+ mutex_init(&file->mutex);
filp->private_data = file;
diff --git a/drivers/infiniband/core/verbs.c b/drivers/infiniband/core/verbs.c
index 4c15e112736c..c857361be449 100644
--- a/drivers/infiniband/core/verbs.c
+++ b/drivers/infiniband/core/verbs.c
@@ -107,9 +107,9 @@ struct ib_ah *ib_create_ah_from_wc(struct ib_pd *pd, struct ib_wc *wc,
if (wc->wc_flags & IB_WC_GRH) {
ah_attr.ah_flags = IB_AH_GRH;
- ah_attr.grh.dgid = grh->dgid;
+ ah_attr.grh.dgid = grh->sgid;
- ret = ib_find_cached_gid(pd->device, &grh->sgid, &port_num,
+ ret = ib_find_cached_gid(pd->device, &grh->dgid, &port_num,
&gid_index);
if (ret)
return ERR_PTR(ret);
diff --git a/drivers/infiniband/hw/mthca/mthca_av.c b/drivers/infiniband/hw/mthca/mthca_av.c
index 22fdc446f25c..a14eed08a0fc 100644
--- a/drivers/infiniband/hw/mthca/mthca_av.c
+++ b/drivers/infiniband/hw/mthca/mthca_av.c
@@ -163,6 +163,11 @@ int mthca_destroy_ah(struct mthca_dev *dev, struct mthca_ah *ah)
return 0;
}
+int mthca_ah_grh_present(struct mthca_ah *ah)
+{
+ return !!(ah->av->g_slid & 0x80);
+}
+
int mthca_read_ah(struct mthca_dev *dev, struct mthca_ah *ah,
struct ib_ud_header *header)
{
@@ -172,8 +177,7 @@ int mthca_read_ah(struct mthca_dev *dev, struct mthca_ah *ah,
header->lrh.service_level = be32_to_cpu(ah->av->sl_tclass_flowlabel) >> 28;
header->lrh.destination_lid = ah->av->dlid;
header->lrh.source_lid = cpu_to_be16(ah->av->g_slid & 0x7f);
- if (ah->av->g_slid & 0x80) {
- header->grh_present = 1;
+ if (mthca_ah_grh_present(ah)) {
header->grh.traffic_class =
(be32_to_cpu(ah->av->sl_tclass_flowlabel) >> 20) & 0xff;
header->grh.flow_label =
@@ -184,8 +188,6 @@ int mthca_read_ah(struct mthca_dev *dev, struct mthca_ah *ah,
&header->grh.source_gid);
memcpy(header->grh.destination_gid.raw,
ah->av->dgid, 16);
- } else {
- header->grh_present = 0;
}
return 0;
diff --git a/drivers/infiniband/hw/mthca/mthca_cmd.c b/drivers/infiniband/hw/mthca/mthca_cmd.c
index 9ed34587fc5c..be1791be627b 100644
--- a/drivers/infiniband/hw/mthca/mthca_cmd.c
+++ b/drivers/infiniband/hw/mthca/mthca_cmd.c
@@ -606,7 +606,7 @@ static int mthca_map_cmd(struct mthca_dev *dev, u16 op, struct mthca_icm *icm,
err = -EINVAL;
goto out;
}
- for (i = 0; i < mthca_icm_size(&iter) / (1 << lg); ++i) {
+ for (i = 0; i < mthca_icm_size(&iter) >> lg; ++i) {
if (virt != -1) {
pages[nent * 2] = cpu_to_be64(virt);
virt += 1 << lg;
@@ -727,8 +727,8 @@ int mthca_QUERY_FW(struct mthca_dev *dev, u8 *status)
* system pages needed.
*/
dev->fw.arbel.fw_pages =
- (dev->fw.arbel.fw_pages + (1 << (PAGE_SHIFT - 12)) - 1) >>
- (PAGE_SHIFT - 12);
+ ALIGN(dev->fw.arbel.fw_pages, PAGE_SIZE >> 12) >>
+ (PAGE_SHIFT - 12);
mthca_dbg(dev, "Clear int @ %llx, EQ arm @ %llx, EQ set CI @ %llx\n",
(unsigned long long) dev->fw.arbel.clr_int_base,
@@ -937,10 +937,6 @@ int mthca_QUERY_DEV_LIM(struct mthca_dev *dev,
if (err)
goto out;
- MTHCA_GET(field, outbox, QUERY_DEV_LIM_MAX_SRQ_SZ_OFFSET);
- dev_lim->max_srq_sz = (1 << field) - 1;
- MTHCA_GET(field, outbox, QUERY_DEV_LIM_MAX_QP_SZ_OFFSET);
- dev_lim->max_qp_sz = (1 << field) - 1;
MTHCA_GET(field, outbox, QUERY_DEV_LIM_RSVD_QP_OFFSET);
dev_lim->reserved_qps = 1 << (field & 0xf);
MTHCA_GET(field, outbox, QUERY_DEV_LIM_MAX_QP_OFFSET);
@@ -1056,6 +1052,10 @@ int mthca_QUERY_DEV_LIM(struct mthca_dev *dev,
mthca_dbg(dev, "Flags: %08x\n", dev_lim->flags);
if (mthca_is_memfree(dev)) {
+ MTHCA_GET(field, outbox, QUERY_DEV_LIM_MAX_SRQ_SZ_OFFSET);
+ dev_lim->max_srq_sz = 1 << field;
+ MTHCA_GET(field, outbox, QUERY_DEV_LIM_MAX_QP_SZ_OFFSET);
+ dev_lim->max_qp_sz = 1 << field;
MTHCA_GET(field, outbox, QUERY_DEV_LIM_RSZ_SRQ_OFFSET);
dev_lim->hca.arbel.resize_srq = field & 1;
MTHCA_GET(field, outbox, QUERY_DEV_LIM_MAX_SG_RQ_OFFSET);
@@ -1087,6 +1087,10 @@ int mthca_QUERY_DEV_LIM(struct mthca_dev *dev,
mthca_dbg(dev, "Max ICM size %lld MB\n",
(unsigned long long) dev_lim->hca.arbel.max_icm_sz >> 20);
} else {
+ MTHCA_GET(field, outbox, QUERY_DEV_LIM_MAX_SRQ_SZ_OFFSET);
+ dev_lim->max_srq_sz = (1 << field) - 1;
+ MTHCA_GET(field, outbox, QUERY_DEV_LIM_MAX_QP_SZ_OFFSET);
+ dev_lim->max_qp_sz = (1 << field) - 1;
MTHCA_GET(field, outbox, QUERY_DEV_LIM_MAX_AV_OFFSET);
dev_lim->hca.tavor.max_avs = 1 << (field & 0x3f);
dev_lim->mpt_entry_sz = MTHCA_MPT_ENTRY_SIZE;
@@ -1441,6 +1445,7 @@ int mthca_SET_ICM_SIZE(struct mthca_dev *dev, u64 icm_size, u64 *aux_pages,
* pages needed.
*/
*aux_pages = (*aux_pages + (1 << (PAGE_SHIFT - 12)) - 1) >> (PAGE_SHIFT - 12);
+ *aux_pages = ALIGN(*aux_pages, PAGE_SIZE >> 12) >> (PAGE_SHIFT - 12);
return 0;
}
diff --git a/drivers/infiniband/hw/mthca/mthca_cq.c b/drivers/infiniband/hw/mthca/mthca_cq.c
index 4a8adcef2079..96f1a86bf049 100644
--- a/drivers/infiniband/hw/mthca/mthca_cq.c
+++ b/drivers/infiniband/hw/mthca/mthca_cq.c
@@ -128,12 +128,12 @@ struct mthca_err_cqe {
__be32 my_qpn;
u32 reserved1[3];
u8 syndrome;
- u8 reserved2;
+ u8 vendor_err;
__be16 db_cnt;
- u32 reserved3;
+ u32 reserved2;
__be32 wqe;
u8 opcode;
- u8 reserved4[2];
+ u8 reserved3[2];
u8 owner;
};
@@ -253,6 +253,15 @@ void mthca_cq_event(struct mthca_dev *dev, u32 cqn,
wake_up(&cq->wait);
}
+static inline int is_recv_cqe(struct mthca_cqe *cqe)
+{
+ if ((cqe->opcode & MTHCA_ERROR_CQE_OPCODE_MASK) ==
+ MTHCA_ERROR_CQE_OPCODE_MASK)
+ return !(cqe->opcode & 0x01);
+ else
+ return !(cqe->is_send & 0x80);
+}
+
void mthca_cq_clean(struct mthca_dev *dev, u32 cqn, u32 qpn,
struct mthca_srq *srq)
{
@@ -296,7 +305,7 @@ void mthca_cq_clean(struct mthca_dev *dev, u32 cqn, u32 qpn,
while ((int) --prod_index - (int) cq->cons_index >= 0) {
cqe = get_cqe(cq, prod_index & cq->ibcq.cqe);
if (cqe->my_qpn == cpu_to_be32(qpn)) {
- if (srq)
+ if (srq && is_recv_cqe(cqe))
mthca_free_srq_wqe(srq, be32_to_cpu(cqe->wqe));
++nfreed;
} else if (nfreed)
@@ -333,8 +342,8 @@ static int handle_error_cqe(struct mthca_dev *dev, struct mthca_cq *cq,
}
/*
- * For completions in error, only work request ID, status (and
- * freed resource count for RD) have to be set.
+ * For completions in error, only work request ID, status, vendor error
+ * (and freed resource count for RD) have to be set.
*/
switch (cqe->syndrome) {
case SYNDROME_LOCAL_LENGTH_ERR:
@@ -396,6 +405,8 @@ static int handle_error_cqe(struct mthca_dev *dev, struct mthca_cq *cq,
break;
}
+ entry->vendor_err = cqe->vendor_err;
+
/*
* Mem-free HCAs always generate one CQE per WQE, even in the
* error case, so we don't have to check the doorbell count, etc.
diff --git a/drivers/infiniband/hw/mthca/mthca_dev.h b/drivers/infiniband/hw/mthca/mthca_dev.h
index 497ff794ef6a..a104ab041ea3 100644
--- a/drivers/infiniband/hw/mthca/mthca_dev.h
+++ b/drivers/infiniband/hw/mthca/mthca_dev.h
@@ -43,6 +43,7 @@
#include <linux/kernel.h>
#include <linux/pci.h>
#include <linux/dma-mapping.h>
+#include <linux/timer.h>
#include <asm/semaphore.h>
#include "mthca_provider.h"
@@ -519,6 +520,7 @@ int mthca_create_ah(struct mthca_dev *dev,
int mthca_destroy_ah(struct mthca_dev *dev, struct mthca_ah *ah);
int mthca_read_ah(struct mthca_dev *dev, struct mthca_ah *ah,
struct ib_ud_header *header);
+int mthca_ah_grh_present(struct mthca_ah *ah);
int mthca_multicast_attach(struct ib_qp *ibqp, union ib_gid *gid, u16 lid);
int mthca_multicast_detach(struct ib_qp *ibqp, union ib_gid *gid, u16 lid);
diff --git a/drivers/infiniband/hw/mthca/mthca_eq.c b/drivers/infiniband/hw/mthca/mthca_eq.c
index 34d68e5a72d8..2eabb27804cd 100644
--- a/drivers/infiniband/hw/mthca/mthca_eq.c
+++ b/drivers/infiniband/hw/mthca/mthca_eq.c
@@ -45,6 +45,7 @@
enum {
MTHCA_NUM_ASYNC_EQE = 0x80,
MTHCA_NUM_CMD_EQE = 0x80,
+ MTHCA_NUM_SPARE_EQE = 0x80,
MTHCA_EQ_ENTRY_SIZE = 0x20
};
@@ -277,11 +278,10 @@ static int mthca_eq_int(struct mthca_dev *dev, struct mthca_eq *eq)
{
struct mthca_eqe *eqe;
int disarm_cqn;
- int eqes_found = 0;
+ int eqes_found = 0;
+ int set_ci = 0;
while ((eqe = next_eqe_sw(eq))) {
- int set_ci = 0;
-
/*
* Make sure we read EQ entry contents after we've
* checked the ownership bit.
@@ -345,12 +345,6 @@ static int mthca_eq_int(struct mthca_dev *dev, struct mthca_eq *eq)
be16_to_cpu(eqe->event.cmd.token),
eqe->event.cmd.status,
be64_to_cpu(eqe->event.cmd.out_param));
- /*
- * cmd_event() may add more commands.
- * The card will think the queue has overflowed if
- * we don't tell it we've been processing events.
- */
- set_ci = 1;
break;
case MTHCA_EVENT_TYPE_PORT_CHANGE:
@@ -385,8 +379,16 @@ static int mthca_eq_int(struct mthca_dev *dev, struct mthca_eq *eq)
set_eqe_hw(eqe);
++eq->cons_index;
eqes_found = 1;
+ ++set_ci;
- if (unlikely(set_ci)) {
+ /*
+ * The HCA will think the queue has overflowed if we
+ * don't tell it we've been processing events. We
+ * create our EQs with MTHCA_NUM_SPARE_EQE extra
+ * entries, so we must update our consumer index at
+ * least that often.
+ */
+ if (unlikely(set_ci >= MTHCA_NUM_SPARE_EQE)) {
/*
* Conditional on hca_type is OK here because
* this is a rare case, not the fast path.
@@ -484,8 +486,7 @@ static int __devinit mthca_create_eq(struct mthca_dev *dev,
u8 intr,
struct mthca_eq *eq)
{
- int npages = (nent * MTHCA_EQ_ENTRY_SIZE + PAGE_SIZE - 1) /
- PAGE_SIZE;
+ int npages;
u64 *dma_list = NULL;
dma_addr_t t;
struct mthca_mailbox *mailbox;
@@ -496,6 +497,7 @@ static int __devinit mthca_create_eq(struct mthca_dev *dev,
eq->dev = dev;
eq->nent = roundup_pow_of_two(max(nent, 2));
+ npages = ALIGN(eq->nent * MTHCA_EQ_ENTRY_SIZE, PAGE_SIZE) / PAGE_SIZE;
eq->page_list = kmalloc(npages * sizeof *eq->page_list,
GFP_KERNEL);
@@ -862,19 +864,19 @@ int __devinit mthca_init_eq_table(struct mthca_dev *dev)
intr = (dev->mthca_flags & MTHCA_FLAG_MSI) ?
128 : dev->eq_table.inta_pin;
- err = mthca_create_eq(dev, dev->limits.num_cqs,
+ err = mthca_create_eq(dev, dev->limits.num_cqs + MTHCA_NUM_SPARE_EQE,
(dev->mthca_flags & MTHCA_FLAG_MSI_X) ? 128 : intr,
&dev->eq_table.eq[MTHCA_EQ_COMP]);
if (err)
goto err_out_unmap;
- err = mthca_create_eq(dev, MTHCA_NUM_ASYNC_EQE,
+ err = mthca_create_eq(dev, MTHCA_NUM_ASYNC_EQE + MTHCA_NUM_SPARE_EQE,
(dev->mthca_flags & MTHCA_FLAG_MSI_X) ? 129 : intr,
&dev->eq_table.eq[MTHCA_EQ_ASYNC]);
if (err)
goto err_out_comp;
- err = mthca_create_eq(dev, MTHCA_NUM_CMD_EQE,
+ err = mthca_create_eq(dev, MTHCA_NUM_CMD_EQE + MTHCA_NUM_SPARE_EQE,
(dev->mthca_flags & MTHCA_FLAG_MSI_X) ? 130 : intr,
&dev->eq_table.eq[MTHCA_EQ_CMD]);
if (err)
diff --git a/drivers/infiniband/hw/mthca/mthca_main.c b/drivers/infiniband/hw/mthca/mthca_main.c
index 6f94b25f3acd..8b00d9a0f6f4 100644
--- a/drivers/infiniband/hw/mthca/mthca_main.c
+++ b/drivers/infiniband/hw/mthca/mthca_main.c
@@ -261,6 +261,10 @@ static int __devinit mthca_init_tavor(struct mthca_dev *mdev)
}
err = mthca_dev_lim(mdev, &dev_lim);
+ if (err) {
+ mthca_err(mdev, "QUERY_DEV_LIM command failed, aborting.\n");
+ goto err_disable;
+ }
profile = default_profile;
profile.num_uar = dev_lim.uar_size / PAGE_SIZE;
diff --git a/drivers/infiniband/hw/mthca/mthca_mcg.c b/drivers/infiniband/hw/mthca/mthca_mcg.c
index 2fc449da418d..77bc6c746f43 100644
--- a/drivers/infiniband/hw/mthca/mthca_mcg.c
+++ b/drivers/infiniband/hw/mthca/mthca_mcg.c
@@ -111,7 +111,8 @@ static int find_mgm(struct mthca_dev *dev,
goto out;
if (status) {
mthca_err(dev, "READ_MGM returned status %02x\n", status);
- return -EINVAL;
+ err = -EINVAL;
+ goto out;
}
if (!memcmp(mgm->gid, zero_gid, 16)) {
@@ -126,7 +127,7 @@ static int find_mgm(struct mthca_dev *dev,
goto out;
*prev = *index;
- *index = be32_to_cpu(mgm->next_gid_index) >> 5;
+ *index = be32_to_cpu(mgm->next_gid_index) >> 6;
} while (*index);
*index = -1;
@@ -153,8 +154,10 @@ int mthca_multicast_attach(struct ib_qp *ibqp, union ib_gid *gid, u16 lid)
return PTR_ERR(mailbox);
mgm = mailbox->buf;
- if (down_interruptible(&dev->mcg_table.sem))
- return -EINTR;
+ if (down_interruptible(&dev->mcg_table.sem)) {
+ err = -EINTR;
+ goto err_sem;
+ }
err = find_mgm(dev, gid->raw, mailbox, &hash, &prev, &index);
if (err)
@@ -181,9 +184,8 @@ int mthca_multicast_attach(struct ib_qp *ibqp, union ib_gid *gid, u16 lid)
err = -EINVAL;
goto out;
}
-
+ memset(mgm, 0, sizeof *mgm);
memcpy(mgm->gid, gid->raw, 16);
- mgm->next_gid_index = 0;
}
for (i = 0; i < MTHCA_QP_PER_MGM; ++i)
@@ -209,6 +211,7 @@ int mthca_multicast_attach(struct ib_qp *ibqp, union ib_gid *gid, u16 lid)
if (status) {
mthca_err(dev, "WRITE_MGM returned status %02x\n", status);
err = -EINVAL;
+ goto out;
}
if (!link)
@@ -223,7 +226,7 @@ int mthca_multicast_attach(struct ib_qp *ibqp, union ib_gid *gid, u16 lid)
goto out;
}
- mgm->next_gid_index = cpu_to_be32(index << 5);
+ mgm->next_gid_index = cpu_to_be32(index << 6);
err = mthca_WRITE_MGM(dev, prev, mailbox, &status);
if (err)
@@ -234,7 +237,12 @@ int mthca_multicast_attach(struct ib_qp *ibqp, union ib_gid *gid, u16 lid)
}
out:
+ if (err && link && index != -1) {
+ BUG_ON(index < dev->limits.num_mgms);
+ mthca_free(&dev->mcg_table.alloc, index);
+ }
up(&dev->mcg_table.sem);
+ err_sem:
mthca_free_mailbox(dev, mailbox);
return err;
}
@@ -255,8 +263,10 @@ int mthca_multicast_detach(struct ib_qp *ibqp, union ib_gid *gid, u16 lid)
return PTR_ERR(mailbox);
mgm = mailbox->buf;
- if (down_interruptible(&dev->mcg_table.sem))
- return -EINTR;
+ if (down_interruptible(&dev->mcg_table.sem)) {
+ err = -EINTR;
+ goto err_sem;
+ }
err = find_mgm(dev, gid->raw, mailbox, &hash, &prev, &index);
if (err)
@@ -305,13 +315,11 @@ int mthca_multicast_detach(struct ib_qp *ibqp, union ib_gid *gid, u16 lid)
if (i != 1)
goto out;
- goto out;
-
if (prev == -1) {
/* Remove entry from MGM */
- if (be32_to_cpu(mgm->next_gid_index) >> 5) {
- err = mthca_READ_MGM(dev,
- be32_to_cpu(mgm->next_gid_index) >> 5,
+ int amgm_index_to_free = be32_to_cpu(mgm->next_gid_index) >> 6;
+ if (amgm_index_to_free) {
+ err = mthca_READ_MGM(dev, amgm_index_to_free,
mailbox, &status);
if (err)
goto out;
@@ -332,9 +340,13 @@ int mthca_multicast_detach(struct ib_qp *ibqp, union ib_gid *gid, u16 lid)
err = -EINVAL;
goto out;
}
+ if (amgm_index_to_free) {
+ BUG_ON(amgm_index_to_free < dev->limits.num_mgms);
+ mthca_free(&dev->mcg_table.alloc, amgm_index_to_free);
+ }
} else {
/* Remove entry from AMGM */
- index = be32_to_cpu(mgm->next_gid_index) >> 5;
+ int curr_next_index = be32_to_cpu(mgm->next_gid_index) >> 6;
err = mthca_READ_MGM(dev, prev, mailbox, &status);
if (err)
goto out;
@@ -344,7 +356,7 @@ int mthca_multicast_detach(struct ib_qp *ibqp, union ib_gid *gid, u16 lid)
goto out;
}
- mgm->next_gid_index = cpu_to_be32(index << 5);
+ mgm->next_gid_index = cpu_to_be32(curr_next_index << 6);
err = mthca_WRITE_MGM(dev, prev, mailbox, &status);
if (err)
@@ -354,10 +366,13 @@ int mthca_multicast_detach(struct ib_qp *ibqp, union ib_gid *gid, u16 lid)
err = -EINVAL;
goto out;
}
+ BUG_ON(index < dev->limits.num_mgms);
+ mthca_free(&dev->mcg_table.alloc, index);
}
out:
up(&dev->mcg_table.sem);
+ err_sem:
mthca_free_mailbox(dev, mailbox);
return err;
}
@@ -365,11 +380,12 @@ int mthca_multicast_detach(struct ib_qp *ibqp, union ib_gid *gid, u16 lid)
int __devinit mthca_init_mcg_table(struct mthca_dev *dev)
{
int err;
+ int table_size = dev->limits.num_mgms + dev->limits.num_amgms;
err = mthca_alloc_init(&dev->mcg_table.alloc,
- dev->limits.num_amgms,
- dev->limits.num_amgms - 1,
- 0);
+ table_size,
+ table_size - 1,
+ dev->limits.num_mgms);
if (err)
return err;
diff --git a/drivers/infiniband/hw/mthca/mthca_memfree.c b/drivers/infiniband/hw/mthca/mthca_memfree.c
index d72fe95cba08..9fb985a016e9 100644
--- a/drivers/infiniband/hw/mthca/mthca_memfree.c
+++ b/drivers/infiniband/hw/mthca/mthca_memfree.c
@@ -233,7 +233,7 @@ void *mthca_table_find(struct mthca_icm_table *table, int obj)
for (i = 0; i < chunk->npages; ++i) {
if (chunk->mem[i].length >= offset) {
page = chunk->mem[i].page;
- break;
+ goto out;
}
offset -= chunk->mem[i].length;
}
@@ -485,6 +485,8 @@ void mthca_cleanup_user_db_tab(struct mthca_dev *dev, struct mthca_uar *uar,
put_page(db_tab->page[i].mem.page);
}
}
+
+ kfree(db_tab);
}
int mthca_alloc_db(struct mthca_dev *dev, enum mthca_db_type type,
diff --git a/drivers/infiniband/hw/mthca/mthca_provider.c b/drivers/infiniband/hw/mthca/mthca_provider.c
index 4cc7e2846df1..484a7e6b7f8c 100644
--- a/drivers/infiniband/hw/mthca/mthca_provider.c
+++ b/drivers/infiniband/hw/mthca/mthca_provider.c
@@ -33,7 +33,7 @@
* CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
* SOFTWARE.
*
- * $Id: mthca_provider.c 1397 2004-12-28 05:09:00Z roland $
+ * $Id: mthca_provider.c 4859 2006-01-09 21:55:10Z roland $
*/
#include <rdma/ib_smi.h>
@@ -45,6 +45,14 @@
#include "mthca_user.h"
#include "mthca_memfree.h"
+static void init_query_mad(struct ib_smp *mad)
+{
+ mad->base_version = 1;
+ mad->mgmt_class = IB_MGMT_CLASS_SUBN_LID_ROUTED;
+ mad->class_version = 1;
+ mad->method = IB_MGMT_METHOD_GET;
+}
+
static int mthca_query_device(struct ib_device *ibdev,
struct ib_device_attr *props)
{
@@ -55,7 +63,7 @@ static int mthca_query_device(struct ib_device *ibdev,
u8 status;
- in_mad = kmalloc(sizeof *in_mad, GFP_KERNEL);
+ in_mad = kzalloc(sizeof *in_mad, GFP_KERNEL);
out_mad = kmalloc(sizeof *out_mad, GFP_KERNEL);
if (!in_mad || !out_mad)
goto out;
@@ -64,12 +72,8 @@ static int mthca_query_device(struct ib_device *ibdev,
props->fw_ver = mdev->fw_ver;
- memset(in_mad, 0, sizeof *in_mad);
- in_mad->base_version = 1;
- in_mad->mgmt_class = IB_MGMT_CLASS_SUBN_LID_ROUTED;
- in_mad->class_version = 1;
- in_mad->method = IB_MGMT_METHOD_GET;
- in_mad->attr_id = IB_SMP_ATTR_NODE_INFO;
+ init_query_mad(in_mad);
+ in_mad->attr_id = IB_SMP_ATTR_NODE_INFO;
err = mthca_MAD_IFC(mdev, 1, 1,
1, NULL, NULL, in_mad, out_mad,
@@ -87,7 +91,6 @@ static int mthca_query_device(struct ib_device *ibdev,
props->vendor_part_id = be16_to_cpup((__be16 *) (out_mad->data + 30));
props->hw_ver = be32_to_cpup((__be32 *) (out_mad->data + 32));
memcpy(&props->sys_image_guid, out_mad->data + 4, 8);
- memcpy(&props->node_guid, out_mad->data + 12, 8);
props->max_mr_size = ~0ull;
props->page_size_cap = mdev->limits.page_size_cap;
@@ -128,20 +131,16 @@ static int mthca_query_port(struct ib_device *ibdev,
int err = -ENOMEM;
u8 status;
- in_mad = kmalloc(sizeof *in_mad, GFP_KERNEL);
+ in_mad = kzalloc(sizeof *in_mad, GFP_KERNEL);
out_mad = kmalloc(sizeof *out_mad, GFP_KERNEL);
if (!in_mad || !out_mad)
goto out;
memset(props, 0, sizeof *props);
- memset(in_mad, 0, sizeof *in_mad);
- in_mad->base_version = 1;
- in_mad->mgmt_class = IB_MGMT_CLASS_SUBN_LID_ROUTED;
- in_mad->class_version = 1;
- in_mad->method = IB_MGMT_METHOD_GET;
- in_mad->attr_id = IB_SMP_ATTR_PORT_INFO;
- in_mad->attr_mod = cpu_to_be32(port);
+ init_query_mad(in_mad);
+ in_mad->attr_id = IB_SMP_ATTR_PORT_INFO;
+ in_mad->attr_mod = cpu_to_be32(port);
err = mthca_MAD_IFC(to_mdev(ibdev), 1, 1,
port, NULL, NULL, in_mad, out_mad,
@@ -220,18 +219,14 @@ static int mthca_query_pkey(struct ib_device *ibdev,
int err = -ENOMEM;
u8 status;
- in_mad = kmalloc(sizeof *in_mad, GFP_KERNEL);
+ in_mad = kzalloc(sizeof *in_mad, GFP_KERNEL);
out_mad = kmalloc(sizeof *out_mad, GFP_KERNEL);
if (!in_mad || !out_mad)
goto out;
- memset(in_mad, 0, sizeof *in_mad);
- in_mad->base_version = 1;
- in_mad->mgmt_class = IB_MGMT_CLASS_SUBN_LID_ROUTED;
- in_mad->class_version = 1;
- in_mad->method = IB_MGMT_METHOD_GET;
- in_mad->attr_id = IB_SMP_ATTR_PKEY_TABLE;
- in_mad->attr_mod = cpu_to_be32(index / 32);
+ init_query_mad(in_mad);
+ in_mad->attr_id = IB_SMP_ATTR_PKEY_TABLE;
+ in_mad->attr_mod = cpu_to_be32(index / 32);
err = mthca_MAD_IFC(to_mdev(ibdev), 1, 1,
port, NULL, NULL, in_mad, out_mad,
@@ -259,18 +254,14 @@ static int mthca_query_gid(struct ib_device *ibdev, u8 port,
int err = -ENOMEM;
u8 status;
- in_mad = kmalloc(sizeof *in_mad, GFP_KERNEL);
+ in_mad = kzalloc(sizeof *in_mad, GFP_KERNEL);
out_mad = kmalloc(sizeof *out_mad, GFP_KERNEL);
if (!in_mad || !out_mad)
goto out;
- memset(in_mad, 0, sizeof *in_mad);
- in_mad->base_version = 1;
- in_mad->mgmt_class = IB_MGMT_CLASS_SUBN_LID_ROUTED;
- in_mad->class_version = 1;
- in_mad->method = IB_MGMT_METHOD_GET;
- in_mad->attr_id = IB_SMP_ATTR_PORT_INFO;
- in_mad->attr_mod = cpu_to_be32(port);
+ init_query_mad(in_mad);
+ in_mad->attr_id = IB_SMP_ATTR_PORT_INFO;
+ in_mad->attr_mod = cpu_to_be32(port);
err = mthca_MAD_IFC(to_mdev(ibdev), 1, 1,
port, NULL, NULL, in_mad, out_mad,
@@ -284,13 +275,9 @@ static int mthca_query_gid(struct ib_device *ibdev, u8 port,
memcpy(gid->raw, out_mad->data + 8, 8);
- memset(in_mad, 0, sizeof *in_mad);
- in_mad->base_version = 1;
- in_mad->mgmt_class = IB_MGMT_CLASS_SUBN_LID_ROUTED;
- in_mad->class_version = 1;
- in_mad->method = IB_MGMT_METHOD_GET;
- in_mad->attr_id = IB_SMP_ATTR_GUID_INFO;
- in_mad->attr_mod = cpu_to_be32(index / 8);
+ init_query_mad(in_mad);
+ in_mad->attr_id = IB_SMP_ATTR_GUID_INFO;
+ in_mad->attr_mod = cpu_to_be32(index / 8);
err = mthca_MAD_IFC(to_mdev(ibdev), 1, 1,
port, NULL, NULL, in_mad, out_mad,
@@ -458,8 +445,10 @@ static struct ib_srq *mthca_create_srq(struct ib_pd *pd,
if (pd->uobject) {
context = to_mucontext(pd->uobject->context);
- if (ib_copy_from_udata(&ucmd, udata, sizeof ucmd))
- return ERR_PTR(-EFAULT);
+ if (ib_copy_from_udata(&ucmd, udata, sizeof ucmd)) {
+ err = -EFAULT;
+ goto err_free;
+ }
err = mthca_map_user_db(to_mdev(pd->device), &context->uar,
context->db_tab, ucmd.db_index,
@@ -535,8 +524,10 @@ static struct ib_qp *mthca_create_qp(struct ib_pd *pd,
if (pd->uobject) {
context = to_mucontext(pd->uobject->context);
- if (ib_copy_from_udata(&ucmd, udata, sizeof ucmd))
+ if (ib_copy_from_udata(&ucmd, udata, sizeof ucmd)) {
+ kfree(qp);
return ERR_PTR(-EFAULT);
+ }
err = mthca_map_user_db(to_mdev(pd->device), &context->uar,
context->db_tab,
@@ -783,24 +774,20 @@ static struct ib_mr *mthca_reg_phys_mr(struct ib_pd *pd,
if ((*iova_start & ~PAGE_MASK) != (buffer_list[0].addr & ~PAGE_MASK))
return ERR_PTR(-EINVAL);
- if (num_phys_buf > 1 &&
- ((buffer_list[0].addr + buffer_list[0].size) & ~PAGE_MASK))
- return ERR_PTR(-EINVAL);
-
mask = 0;
total_size = 0;
for (i = 0; i < num_phys_buf; ++i) {
- if (i != 0 && buffer_list[i].addr & ~PAGE_MASK)
- return ERR_PTR(-EINVAL);
- if (i != 0 && i != num_phys_buf - 1 &&
- (buffer_list[i].size & ~PAGE_MASK))
- return ERR_PTR(-EINVAL);
+ if (i != 0)
+ mask |= buffer_list[i].addr;
+ if (i != num_phys_buf - 1)
+ mask |= buffer_list[i].addr + buffer_list[i].size;
total_size += buffer_list[i].size;
- if (i > 0)
- mask |= buffer_list[i].addr;
}
+ if (mask & ~PAGE_MASK)
+ return ERR_PTR(-EINVAL);
+
/* Find largest page shift we can use to cover buffers */
for (shift = PAGE_SHIFT; shift < 31; ++shift)
if (num_phys_buf > 1) {
@@ -1070,11 +1057,48 @@ static struct class_device_attribute *mthca_class_attributes[] = {
&class_device_attr_board_id
};
+static int mthca_init_node_data(struct mthca_dev *dev)
+{
+ struct ib_smp *in_mad = NULL;
+ struct ib_smp *out_mad = NULL;
+ int err = -ENOMEM;
+ u8 status;
+
+ in_mad = kzalloc(sizeof *in_mad, GFP_KERNEL);
+ out_mad = kmalloc(sizeof *out_mad, GFP_KERNEL);
+ if (!in_mad || !out_mad)
+ goto out;
+
+ init_query_mad(in_mad);
+ in_mad->attr_id = IB_SMP_ATTR_NODE_INFO;
+
+ err = mthca_MAD_IFC(dev, 1, 1,
+ 1, NULL, NULL, in_mad, out_mad,
+ &status);
+ if (err)
+ goto out;
+ if (status) {
+ err = -EINVAL;
+ goto out;
+ }
+
+ memcpy(&dev->ib_dev.node_guid, out_mad->data + 12, 8);
+
+out:
+ kfree(in_mad);
+ kfree(out_mad);
+ return err;
+}
+
int mthca_register_device(struct mthca_dev *dev)
{
int ret;
int i;
+ ret = mthca_init_node_data(dev);
+ if (ret)
+ return ret;
+
strlcpy(dev->ib_dev.name, "mthca%d", IB_DEVICE_NAME_MAX);
dev->ib_dev.owner = THIS_MODULE;
diff --git a/drivers/infiniband/hw/mthca/mthca_qp.c b/drivers/infiniband/hw/mthca/mthca_qp.c
index dd4e13303e96..fba608ed7df2 100644
--- a/drivers/infiniband/hw/mthca/mthca_qp.c
+++ b/drivers/infiniband/hw/mthca/mthca_qp.c
@@ -383,12 +383,10 @@ static const struct {
[UC] = (IB_QP_CUR_STATE |
IB_QP_ALT_PATH |
IB_QP_ACCESS_FLAGS |
- IB_QP_PKEY_INDEX |
IB_QP_PATH_MIG_STATE),
[RC] = (IB_QP_CUR_STATE |
IB_QP_ALT_PATH |
IB_QP_ACCESS_FLAGS |
- IB_QP_PKEY_INDEX |
IB_QP_MIN_RNR_TIMER |
IB_QP_PATH_MIG_STATE),
[MLX] = (IB_QP_CUR_STATE |
@@ -476,9 +474,8 @@ static const struct {
.opt_param = {
[UD] = (IB_QP_CUR_STATE |
IB_QP_QKEY),
- [UC] = IB_QP_CUR_STATE,
- [RC] = (IB_QP_CUR_STATE |
- IB_QP_MIN_RNR_TIMER),
+ [UC] = (IB_QP_CUR_STATE |
+ IB_QP_ACCESS_FLAGS),
[MLX] = (IB_QP_CUR_STATE |
IB_QP_QKEY),
}
@@ -522,6 +519,55 @@ static void init_port(struct mthca_dev *dev, int port)
mthca_warn(dev, "INIT_IB returned status %02x.\n", status);
}
+static __be32 get_hw_access_flags(struct mthca_qp *qp, struct ib_qp_attr *attr,
+ int attr_mask)
+{
+ u8 dest_rd_atomic;
+ u32 access_flags;
+ u32 hw_access_flags = 0;
+
+ if (attr_mask & IB_QP_MAX_DEST_RD_ATOMIC)
+ dest_rd_atomic = attr->max_dest_rd_atomic;
+ else
+ dest_rd_atomic = qp->resp_depth;
+
+ if (attr_mask & IB_QP_ACCESS_FLAGS)
+ access_flags = attr->qp_access_flags;
+ else
+ access_flags = qp->atomic_rd_en;
+
+ if (!dest_rd_atomic)
+ access_flags &= IB_ACCESS_REMOTE_WRITE;
+
+ if (access_flags & IB_ACCESS_REMOTE_READ)
+ hw_access_flags |= MTHCA_QP_BIT_RRE;
+ if (access_flags & IB_ACCESS_REMOTE_ATOMIC)
+ hw_access_flags |= MTHCA_QP_BIT_RAE;
+ if (access_flags & IB_ACCESS_REMOTE_WRITE)
+ hw_access_flags |= MTHCA_QP_BIT_RWE;
+
+ return cpu_to_be32(hw_access_flags);
+}
+
+static void mthca_path_set(struct ib_ah_attr *ah, struct mthca_qp_path *path)
+{
+ path->g_mylmc = ah->src_path_bits & 0x7f;
+ path->rlid = cpu_to_be16(ah->dlid);
+ path->static_rate = !!ah->static_rate;
+
+ if (ah->ah_flags & IB_AH_GRH) {
+ path->g_mylmc |= 1 << 7;
+ path->mgid_index = ah->grh.sgid_index;
+ path->hop_limit = ah->grh.hop_limit;
+ path->sl_tclass_flowlabel =
+ cpu_to_be32((ah->sl << 28) |
+ (ah->grh.traffic_class << 20) |
+ (ah->grh.flow_label));
+ memcpy(path->rgid, ah->grh.dgid.raw, 16);
+ } else
+ path->sl_tclass_flowlabel = cpu_to_be32(ah->sl << 28);
+}
+
int mthca_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr, int attr_mask)
{
struct mthca_dev *dev = to_mdev(ibqp->device);
@@ -591,6 +637,26 @@ int mthca_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr, int attr_mask)
return -EINVAL;
}
+ if ((attr_mask & IB_QP_PORT) &&
+ (attr->port_num == 0 || attr->port_num > dev->limits.num_ports)) {
+ mthca_dbg(dev, "Port number (%u) is invalid\n", attr->port_num);
+ return -EINVAL;
+ }
+
+ if (attr_mask & IB_QP_MAX_QP_RD_ATOMIC &&
+ attr->max_rd_atomic > dev->limits.max_qp_init_rdma) {
+ mthca_dbg(dev, "Max rdma_atomic as initiator %u too large (max is %d)\n",
+ attr->max_rd_atomic, dev->limits.max_qp_init_rdma);
+ return -EINVAL;
+ }
+
+ if (attr_mask & IB_QP_MAX_DEST_RD_ATOMIC &&
+ attr->max_dest_rd_atomic > 1 << dev->qp_table.rdb_shift) {
+ mthca_dbg(dev, "Max rdma_atomic as responder %u too large (max %d)\n",
+ attr->max_dest_rd_atomic, 1 << dev->qp_table.rdb_shift);
+ return -EINVAL;
+ }
+
mailbox = mthca_alloc_mailbox(dev, GFP_KERNEL);
if (IS_ERR(mailbox))
return PTR_ERR(mailbox);
@@ -665,28 +731,14 @@ int mthca_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr, int attr_mask)
}
if (attr_mask & IB_QP_RNR_RETRY) {
- qp_context->pri_path.rnr_retry = attr->rnr_retry << 5;
- qp_param->opt_param_mask |= cpu_to_be32(MTHCA_QP_OPTPAR_RNR_RETRY);
+ qp_context->alt_path.rnr_retry = qp_context->pri_path.rnr_retry =
+ attr->rnr_retry << 5;
+ qp_param->opt_param_mask |= cpu_to_be32(MTHCA_QP_OPTPAR_RNR_RETRY |
+ MTHCA_QP_OPTPAR_ALT_RNR_RETRY);
}
if (attr_mask & IB_QP_AV) {
- qp_context->pri_path.g_mylmc = attr->ah_attr.src_path_bits & 0x7f;
- qp_context->pri_path.rlid = cpu_to_be16(attr->ah_attr.dlid);
- qp_context->pri_path.static_rate = !!attr->ah_attr.static_rate;
- if (attr->ah_attr.ah_flags & IB_AH_GRH) {
- qp_context->pri_path.g_mylmc |= 1 << 7;
- qp_context->pri_path.mgid_index = attr->ah_attr.grh.sgid_index;
- qp_context->pri_path.hop_limit = attr->ah_attr.grh.hop_limit;
- qp_context->pri_path.sl_tclass_flowlabel =
- cpu_to_be32((attr->ah_attr.sl << 28) |
- (attr->ah_attr.grh.traffic_class << 20) |
- (attr->ah_attr.grh.flow_label));
- memcpy(qp_context->pri_path.rgid,
- attr->ah_attr.grh.dgid.raw, 16);
- } else {
- qp_context->pri_path.sl_tclass_flowlabel =
- cpu_to_be32(attr->ah_attr.sl << 28);
- }
+ mthca_path_set(&attr->ah_attr, &qp_context->pri_path);
qp_param->opt_param_mask |= cpu_to_be32(MTHCA_QP_OPTPAR_PRIMARY_ADDR_PATH);
}
@@ -695,7 +747,19 @@ int mthca_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr, int attr_mask)
qp_param->opt_param_mask |= cpu_to_be32(MTHCA_QP_OPTPAR_ACK_TIMEOUT);
}
- /* XXX alt_path */
+ if (attr_mask & IB_QP_ALT_PATH) {
+ if (attr->alt_port_num == 0 || attr->alt_port_num > dev->limits.num_ports) {
+ mthca_dbg(dev, "Alternate port number (%u) is invalid\n",
+ attr->alt_port_num);
+ return -EINVAL;
+ }
+
+ mthca_path_set(&attr->alt_ah_attr, &qp_context->alt_path);
+ qp_context->alt_path.port_pkey |= cpu_to_be32(attr->alt_pkey_index |
+ attr->alt_port_num << 24);
+ qp_context->alt_path.ackto = attr->alt_timeout << 3;
+ qp_param->opt_param_mask |= cpu_to_be32(MTHCA_QP_OPTPAR_ALT_ADDR_PATH);
+ }
/* leave rdd as 0 */
qp_context->pd = cpu_to_be32(to_mpd(ibqp->pd)->pd_num);
@@ -703,9 +767,7 @@ int mthca_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr, int attr_mask)
qp_context->wqe_lkey = cpu_to_be32(qp->mr.ibmr.lkey);
qp_context->params1 = cpu_to_be32((MTHCA_ACK_REQ_FREQ << 28) |
(MTHCA_FLIGHT_LIMIT << 24) |
- MTHCA_QP_BIT_SRE |
- MTHCA_QP_BIT_SWE |
- MTHCA_QP_BIT_SAE);
+ MTHCA_QP_BIT_SWE);
if (qp->sq_policy == IB_SIGNAL_ALL_WR)
qp_context->params1 |= cpu_to_be32(MTHCA_QP_BIT_SSC);
if (attr_mask & IB_QP_RETRY_CNT) {
@@ -714,9 +776,13 @@ int mthca_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr, int attr_mask)
}
if (attr_mask & IB_QP_MAX_QP_RD_ATOMIC) {
- qp_context->params1 |= cpu_to_be32(min(attr->max_rd_atomic ?
- ffs(attr->max_rd_atomic) - 1 : 0,
- 7) << 21);
+ if (attr->max_rd_atomic) {
+ qp_context->params1 |=
+ cpu_to_be32(MTHCA_QP_BIT_SRE |
+ MTHCA_QP_BIT_SAE);
+ qp_context->params1 |=
+ cpu_to_be32(fls(attr->max_rd_atomic - 1) << 21);
+ }
qp_param->opt_param_mask |= cpu_to_be32(MTHCA_QP_OPTPAR_SRA_MAX);
}
@@ -729,71 +795,19 @@ int mthca_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr, int attr_mask)
qp_context->snd_db_index = cpu_to_be32(qp->sq.db_index);
}
- if (attr_mask & IB_QP_ACCESS_FLAGS) {
- qp_context->params2 |=
- cpu_to_be32(attr->qp_access_flags & IB_ACCESS_REMOTE_WRITE ?
- MTHCA_QP_BIT_RWE : 0);
-
- /*
- * Only enable RDMA reads and atomics if we have
- * responder resources set to a non-zero value.
- */
- if (qp->resp_depth) {
- qp_context->params2 |=
- cpu_to_be32(attr->qp_access_flags & IB_ACCESS_REMOTE_READ ?
- MTHCA_QP_BIT_RRE : 0);
- qp_context->params2 |=
- cpu_to_be32(attr->qp_access_flags & IB_ACCESS_REMOTE_ATOMIC ?
- MTHCA_QP_BIT_RAE : 0);
- }
-
- qp_param->opt_param_mask |= cpu_to_be32(MTHCA_QP_OPTPAR_RWE |
- MTHCA_QP_OPTPAR_RRE |
- MTHCA_QP_OPTPAR_RAE);
-
- qp->atomic_rd_en = attr->qp_access_flags;
- }
-
if (attr_mask & IB_QP_MAX_DEST_RD_ATOMIC) {
- u8 rra_max;
-
- if (qp->resp_depth && !attr->max_dest_rd_atomic) {
- /*
- * Lowering our responder resources to zero.
- * Turn off reads RDMA and atomics as responder.
- * (RRE/RAE in params2 already zero)
- */
- qp_param->opt_param_mask |= cpu_to_be32(MTHCA_QP_OPTPAR_RRE |
- MTHCA_QP_OPTPAR_RAE);
- }
-
- if (!qp->resp_depth && attr->max_dest_rd_atomic) {
- /*
- * Increasing our responder resources from
- * zero. Turn on RDMA reads and atomics as
- * appropriate.
- */
- qp_context->params2 |=
- cpu_to_be32(qp->atomic_rd_en & IB_ACCESS_REMOTE_READ ?
- MTHCA_QP_BIT_RRE : 0);
+ if (attr->max_dest_rd_atomic)
qp_context->params2 |=
- cpu_to_be32(qp->atomic_rd_en & IB_ACCESS_REMOTE_ATOMIC ?
- MTHCA_QP_BIT_RAE : 0);
+ cpu_to_be32(fls(attr->max_dest_rd_atomic - 1) << 21);
- qp_param->opt_param_mask |= cpu_to_be32(MTHCA_QP_OPTPAR_RRE |
- MTHCA_QP_OPTPAR_RAE);
- }
-
- for (rra_max = 0;
- 1 << rra_max < attr->max_dest_rd_atomic &&
- rra_max < dev->qp_table.rdb_shift;
- ++rra_max)
- ; /* nothing */
-
- qp_context->params2 |= cpu_to_be32(rra_max << 21);
qp_param->opt_param_mask |= cpu_to_be32(MTHCA_QP_OPTPAR_RRA_MAX);
+ }
- qp->resp_depth = attr->max_dest_rd_atomic;
+ if (attr_mask & (IB_QP_ACCESS_FLAGS | IB_QP_MAX_DEST_RD_ATOMIC)) {
+ qp_context->params2 |= get_hw_access_flags(qp, attr, attr_mask);
+ qp_param->opt_param_mask |= cpu_to_be32(MTHCA_QP_OPTPAR_RWE |
+ MTHCA_QP_OPTPAR_RRE |
+ MTHCA_QP_OPTPAR_RAE);
}
qp_context->params2 |= cpu_to_be32(MTHCA_QP_BIT_RSC);
@@ -835,8 +849,13 @@ int mthca_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr, int attr_mask)
err = -EINVAL;
}
- if (!err)
+ if (!err) {
qp->state = new_state;
+ if (attr_mask & IB_QP_ACCESS_FLAGS)
+ qp->atomic_rd_en = attr->qp_access_flags;
+ if (attr_mask & IB_QP_MAX_DEST_RD_ATOMIC)
+ qp->resp_depth = attr->max_dest_rd_atomic;
+ }
mthca_free_mailbox(dev, mailbox);
@@ -871,7 +890,10 @@ int mthca_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr, int attr_mask)
qp->ibqp.srq ? to_msrq(qp->ibqp.srq) : NULL);
mthca_wq_init(&qp->sq);
+ qp->sq.last = get_send_wqe(qp, qp->sq.max - 1);
+
mthca_wq_init(&qp->rq);
+ qp->rq.last = get_recv_wqe(qp, qp->rq.max - 1);
if (mthca_is_memfree(dev)) {
*qp->sq.db = 0;
@@ -882,18 +904,13 @@ int mthca_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr, int attr_mask)
return err;
}
-static void mthca_adjust_qp_caps(struct mthca_dev *dev,
- struct mthca_pd *pd,
- struct mthca_qp *qp)
+static int mthca_max_data_size(struct mthca_dev *dev, struct mthca_qp *qp, int desc_sz)
{
- int max_data_size;
-
/*
* Calculate the maximum size of WQE s/g segments, excluding
* the next segment and other non-data segments.
*/
- max_data_size = min(dev->limits.max_desc_sz, 1 << qp->sq.wqe_shift) -
- sizeof (struct mthca_next_seg);
+ int max_data_size = desc_sz - sizeof (struct mthca_next_seg);
switch (qp->transport) {
case MLX:
@@ -912,11 +929,24 @@ static void mthca_adjust_qp_caps(struct mthca_dev *dev,
break;
}
+ return max_data_size;
+}
+
+static inline int mthca_max_inline_data(struct mthca_pd *pd, int max_data_size)
+{
/* We don't support inline data for kernel QPs (yet). */
- if (!pd->ibpd.uobject)
- qp->max_inline_data = 0;
- else
- qp->max_inline_data = max_data_size - MTHCA_INLINE_HEADER_SIZE;
+ return pd->ibpd.uobject ? max_data_size - MTHCA_INLINE_HEADER_SIZE : 0;
+}
+
+static void mthca_adjust_qp_caps(struct mthca_dev *dev,
+ struct mthca_pd *pd,
+ struct mthca_qp *qp)
+{
+ int max_data_size = mthca_max_data_size(dev, qp,
+ min(dev->limits.max_desc_sz,
+ 1 << qp->sq.wqe_shift));
+
+ qp->max_inline_data = mthca_max_inline_data(pd, max_data_size);
qp->sq.max_gs = min_t(int, dev->limits.max_sg,
max_data_size / sizeof (struct mthca_data_seg));
@@ -1183,13 +1213,23 @@ static int mthca_alloc_qp_common(struct mthca_dev *dev,
}
static int mthca_set_qp_size(struct mthca_dev *dev, struct ib_qp_cap *cap,
- struct mthca_qp *qp)
+ struct mthca_pd *pd, struct mthca_qp *qp)
{
+ int max_data_size = mthca_max_data_size(dev, qp, dev->limits.max_desc_sz);
+
/* Sanity check QP size before proceeding */
- if (cap->max_send_wr > dev->limits.max_wqes ||
- cap->max_recv_wr > dev->limits.max_wqes ||
- cap->max_send_sge > dev->limits.max_sg ||
- cap->max_recv_sge > dev->limits.max_sg)
+ if (cap->max_send_wr > dev->limits.max_wqes ||
+ cap->max_recv_wr > dev->limits.max_wqes ||
+ cap->max_send_sge > dev->limits.max_sg ||
+ cap->max_recv_sge > dev->limits.max_sg ||
+ cap->max_inline_data > mthca_max_inline_data(pd, max_data_size))
+ return -EINVAL;
+
+ /*
+ * For MLX transport we need 2 extra S/G entries:
+ * one for the header and one for the checksum at the end
+ */
+ if (qp->transport == MLX && cap->max_recv_sge + 2 > dev->limits.max_sg)
return -EINVAL;
if (mthca_is_memfree(dev)) {
@@ -1208,14 +1248,6 @@ static int mthca_set_qp_size(struct mthca_dev *dev, struct ib_qp_cap *cap,
MTHCA_INLINE_CHUNK_SIZE) /
sizeof (struct mthca_data_seg));
- /*
- * For MLX transport we need 2 extra S/G entries:
- * one for the header and one for the checksum at the end
- */
- if ((qp->transport == MLX && qp->sq.max_gs + 2 > dev->limits.max_sg) ||
- qp->sq.max_gs > dev->limits.max_sg || qp->rq.max_gs > dev->limits.max_sg)
- return -EINVAL;
-
return 0;
}
@@ -1230,7 +1262,7 @@ int mthca_alloc_qp(struct mthca_dev *dev,
{
int err;
- err = mthca_set_qp_size(dev, cap, qp);
+ err = mthca_set_qp_size(dev, cap, pd, qp);
if (err)
return err;
@@ -1273,7 +1305,7 @@ int mthca_alloc_sqp(struct mthca_dev *dev,
u32 mqpn = qpn * 2 + dev->qp_table.sqp_start + port - 1;
int err;
- err = mthca_set_qp_size(dev, cap, &sqp->qp);
+ err = mthca_set_qp_size(dev, cap, pd, &sqp->qp);
if (err)
return err;
@@ -1402,7 +1434,7 @@ static int build_mlx_header(struct mthca_dev *dev, struct mthca_sqp *sqp,
u16 pkey;
ib_ud_header_init(256, /* assume a MAD */
- sqp->ud_header.grh_present,
+ mthca_ah_grh_present(to_mah(wr->wr.ud.ah)),
&sqp->ud_header);
err = mthca_read_ah(dev, to_mah(wr->wr.ud.ah), &sqp->ud_header);
@@ -1819,6 +1851,7 @@ int mthca_arbel_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr,
{
struct mthca_dev *dev = to_mdev(ibqp->device);
struct mthca_qp *qp = to_mqp(ibqp);
+ __be32 doorbell[2];
void *wqe;
void *prev_wqe;
unsigned long flags;
@@ -1838,6 +1871,34 @@ int mthca_arbel_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr,
ind = qp->sq.head & (qp->sq.max - 1);
for (nreq = 0; wr; ++nreq, wr = wr->next) {
+ if (unlikely(nreq == MTHCA_ARBEL_MAX_WQES_PER_SEND_DB)) {
+ nreq = 0;
+
+ doorbell[0] = cpu_to_be32((MTHCA_ARBEL_MAX_WQES_PER_SEND_DB << 24) |
+ ((qp->sq.head & 0xffff) << 8) |
+ f0 | op0);
+ doorbell[1] = cpu_to_be32((qp->qpn << 8) | size0);
+
+ qp->sq.head += MTHCA_ARBEL_MAX_WQES_PER_SEND_DB;
+ size0 = 0;
+
+ /*
+ * Make sure that descriptors are written before
+ * doorbell record.
+ */
+ wmb();
+ *qp->sq.db = cpu_to_be32(qp->sq.head & 0xffff);
+
+ /*
+ * Make sure doorbell record is written before we
+ * write MMIO send doorbell.
+ */
+ wmb();
+ mthca_write64(doorbell,
+ dev->kar + MTHCA_SEND_DOORBELL,
+ MTHCA_GET_DOORBELL_LOCK(&dev->doorbell_lock));
+ }
+
if (mthca_wq_overflow(&qp->sq, nreq, qp->ibqp.send_cq)) {
mthca_err(dev, "SQ %06x full (%u head, %u tail,"
" %d max, %d nreq)\n", qp->qpn,
@@ -2014,8 +2075,6 @@ int mthca_arbel_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr,
out:
if (likely(nreq)) {
- __be32 doorbell[2];
-
doorbell[0] = cpu_to_be32((nreq << 24) |
((qp->sq.head & 0xffff) << 8) |
f0 | op0);
diff --git a/drivers/infiniband/hw/mthca/mthca_srq.c b/drivers/infiniband/hw/mthca/mthca_srq.c
index f7d234295efe..e7e153d9c4c6 100644
--- a/drivers/infiniband/hw/mthca/mthca_srq.c
+++ b/drivers/infiniband/hw/mthca/mthca_srq.c
@@ -201,7 +201,7 @@ int mthca_alloc_srq(struct mthca_dev *dev, struct mthca_pd *pd,
if (mthca_is_memfree(dev))
srq->max = roundup_pow_of_two(srq->max + 1);
- ds = min(64UL,
+ ds = max(64UL,
roundup_pow_of_two(sizeof (struct mthca_next_seg) +
srq->max_gs * sizeof (struct mthca_data_seg)));
srq->wqe_shift = long_log2(ds);
diff --git a/drivers/infiniband/hw/mthca/mthca_wqe.h b/drivers/infiniband/hw/mthca/mthca_wqe.h
index 73f1c0b9021e..e7d2c1e86199 100644
--- a/drivers/infiniband/hw/mthca/mthca_wqe.h
+++ b/drivers/infiniband/hw/mthca/mthca_wqe.h
@@ -50,7 +50,8 @@ enum {
enum {
MTHCA_INVAL_LKEY = 0x100,
- MTHCA_TAVOR_MAX_WQES_PER_RECV_DB = 256
+ MTHCA_TAVOR_MAX_WQES_PER_RECV_DB = 256,
+ MTHCA_ARBEL_MAX_WQES_PER_SEND_DB = 255
};
struct mthca_next_seg {
diff --git a/drivers/infiniband/ulp/ipoib/ipoib.h b/drivers/infiniband/ulp/ipoib/ipoib.h
index 9923a15a9996..e0a5412b7e68 100644
--- a/drivers/infiniband/ulp/ipoib/ipoib.h
+++ b/drivers/infiniband/ulp/ipoib/ipoib.h
@@ -45,11 +45,11 @@
#include <linux/config.h>
#include <linux/kref.h>
#include <linux/if_infiniband.h>
+#include <linux/mutex.h>
#include <net/neighbour.h>
#include <asm/atomic.h>
-#include <asm/semaphore.h>
#include <rdma/ib_verbs.h>
#include <rdma/ib_pack.h>
@@ -123,8 +123,8 @@ struct ipoib_dev_priv {
unsigned long flags;
- struct semaphore mcast_mutex;
- struct semaphore vlan_mutex;
+ struct mutex mcast_mutex;
+ struct mutex vlan_mutex;
struct rb_root path_tree;
struct list_head path_list;
diff --git a/drivers/infiniband/ulp/ipoib/ipoib_ib.c b/drivers/infiniband/ulp/ipoib/ipoib_ib.c
index 54ef2fea530f..86bcdd72a107 100644
--- a/drivers/infiniband/ulp/ipoib/ipoib_ib.c
+++ b/drivers/infiniband/ulp/ipoib/ipoib_ib.c
@@ -52,7 +52,7 @@ MODULE_PARM_DESC(data_debug_level,
#define IPOIB_OP_RECV (1ul << 31)
-static DECLARE_MUTEX(pkey_sem);
+static DEFINE_MUTEX(pkey_mutex);
struct ipoib_ah *ipoib_create_ah(struct net_device *dev,
struct ib_pd *pd, struct ib_ah_attr *attr)
@@ -445,25 +445,16 @@ int ipoib_ib_dev_down(struct net_device *dev)
/* Shutdown the P_Key thread if still active */
if (!test_bit(IPOIB_PKEY_ASSIGNED, &priv->flags)) {
- down(&pkey_sem);
+ mutex_lock(&pkey_mutex);
set_bit(IPOIB_PKEY_STOP, &priv->flags);
cancel_delayed_work(&priv->pkey_task);
- up(&pkey_sem);
+ mutex_unlock(&pkey_mutex);
flush_workqueue(ipoib_workqueue);
}
ipoib_mcast_stop_thread(dev, 1);
-
- /*
- * Flush the multicast groups first so we stop any multicast joins. The
- * completion thread may have already died and we may deadlock waiting
- * for the completion thread to finish some multicast joins.
- */
ipoib_mcast_dev_flush(dev);
- /* Delete broadcast and local addresses since they will be recreated */
- ipoib_mcast_dev_down(dev);
-
ipoib_flush_paths(dev);
return 0;
@@ -608,9 +599,13 @@ void ipoib_ib_dev_flush(void *_dev)
if (test_bit(IPOIB_FLAG_ADMIN_UP, &priv->flags))
ipoib_ib_dev_up(dev);
+ mutex_lock(&priv->vlan_mutex);
+
/* Flush any child interfaces too */
list_for_each_entry(cpriv, &priv->child_intfs, list)
ipoib_ib_dev_flush(&cpriv->dev);
+
+ mutex_unlock(&priv->vlan_mutex);
}
void ipoib_ib_dev_cleanup(struct net_device *dev)
@@ -620,9 +615,7 @@ void ipoib_ib_dev_cleanup(struct net_device *dev)
ipoib_dbg(priv, "cleaning up ib_dev\n");
ipoib_mcast_stop_thread(dev, 1);
-
- /* Delete the broadcast address and the local address */
- ipoib_mcast_dev_down(dev);
+ ipoib_mcast_dev_flush(dev);
ipoib_transport_dev_cleanup(dev);
}
@@ -658,12 +651,12 @@ void ipoib_pkey_poll(void *dev_ptr)
if (test_bit(IPOIB_PKEY_ASSIGNED, &priv->flags))
ipoib_open(dev);
else {
- down(&pkey_sem);
+ mutex_lock(&pkey_mutex);
if (!test_bit(IPOIB_PKEY_STOP, &priv->flags))
queue_delayed_work(ipoib_workqueue,
&priv->pkey_task,
HZ);
- up(&pkey_sem);
+ mutex_unlock(&pkey_mutex);
}
}
@@ -677,12 +670,12 @@ int ipoib_pkey_dev_delay_open(struct net_device *dev)
/* P_Key value not assigned yet - start polling */
if (!test_bit(IPOIB_PKEY_ASSIGNED, &priv->flags)) {
- down(&pkey_sem);
+ mutex_lock(&pkey_mutex);
clear_bit(IPOIB_PKEY_STOP, &priv->flags);
queue_delayed_work(ipoib_workqueue,
&priv->pkey_task,
HZ);
- up(&pkey_sem);
+ mutex_unlock(&pkey_mutex);
return 1;
}
diff --git a/drivers/infiniband/ulp/ipoib/ipoib_main.c b/drivers/infiniband/ulp/ipoib/ipoib_main.c
index 2fa30751f362..fd3f5c862a5d 100644
--- a/drivers/infiniband/ulp/ipoib/ipoib_main.c
+++ b/drivers/infiniband/ulp/ipoib/ipoib_main.c
@@ -47,6 +47,8 @@
#include <linux/ip.h>
#include <linux/in.h>
+#include <net/dst.h>
+
MODULE_AUTHOR("Roland Dreier");
MODULE_DESCRIPTION("IP-over-InfiniBand net driver");
MODULE_LICENSE("Dual BSD/GPL");
@@ -94,14 +96,16 @@ int ipoib_open(struct net_device *dev)
if (ipoib_ib_dev_open(dev))
return -EINVAL;
- if (ipoib_ib_dev_up(dev))
+ if (ipoib_ib_dev_up(dev)) {
+ ipoib_ib_dev_stop(dev);
return -EINVAL;
+ }
if (!test_bit(IPOIB_FLAG_SUBINTERFACE, &priv->flags)) {
struct ipoib_dev_priv *cpriv;
/* Bring up any child interfaces too */
- down(&priv->vlan_mutex);
+ mutex_lock(&priv->vlan_mutex);
list_for_each_entry(cpriv, &priv->child_intfs, list) {
int flags;
@@ -111,7 +115,7 @@ int ipoib_open(struct net_device *dev)
dev_change_flags(cpriv->dev, flags | IFF_UP);
}
- up(&priv->vlan_mutex);
+ mutex_unlock(&priv->vlan_mutex);
}
netif_start_queue(dev);
@@ -136,7 +140,7 @@ static int ipoib_stop(struct net_device *dev)
struct ipoib_dev_priv *cpriv;
/* Bring down any child interfaces too */
- down(&priv->vlan_mutex);
+ mutex_lock(&priv->vlan_mutex);
list_for_each_entry(cpriv, &priv->child_intfs, list) {
int flags;
@@ -146,7 +150,7 @@ static int ipoib_stop(struct net_device *dev)
dev_change_flags(cpriv->dev, flags & ~IFF_UP);
}
- up(&priv->vlan_mutex);
+ mutex_unlock(&priv->vlan_mutex);
}
return 0;
@@ -398,9 +402,9 @@ static void path_rec_completion(int status,
while ((skb = __skb_dequeue(&neigh->queue)))
__skb_queue_tail(&skqueue, skb);
}
- } else
- path->query = NULL;
+ }
+ path->query = NULL;
complete(&path->done);
spin_unlock_irqrestore(&priv->lock, flags);
@@ -428,7 +432,6 @@ static struct ipoib_path *path_rec_create(struct net_device *dev,
skb_queue_head_init(&path->queue);
INIT_LIST_HEAD(&path->neigh_list);
- init_completion(&path->done);
memcpy(path->pathrec.dgid.raw, gid->raw, sizeof (union ib_gid));
path->pathrec.sgid = priv->local_gid;
@@ -446,6 +449,8 @@ static int path_rec_start(struct net_device *dev,
ipoib_dbg(priv, "Start path record lookup for " IPOIB_GID_FMT "\n",
IPOIB_GID_ARG(path->pathrec.dgid));
+ init_completion(&path->done);
+
path->query_id =
ib_sa_path_rec_get(priv->ca, priv->port,
&path->pathrec,
@@ -887,8 +892,8 @@ static void ipoib_setup(struct net_device *dev)
spin_lock_init(&priv->lock);
spin_lock_init(&priv->tx_lock);
- init_MUTEX(&priv->mcast_mutex);
- init_MUTEX(&priv->vlan_mutex);
+ mutex_init(&priv->mcast_mutex);
+ mutex_init(&priv->vlan_mutex);
INIT_LIST_HEAD(&priv->path_list);
INIT_LIST_HEAD(&priv->child_intfs);
diff --git a/drivers/infiniband/ulp/ipoib/ipoib_multicast.c b/drivers/infiniband/ulp/ipoib/ipoib_multicast.c
index c33ed87f9dff..98039da0caf0 100644
--- a/drivers/infiniband/ulp/ipoib/ipoib_multicast.c
+++ b/drivers/infiniband/ulp/ipoib/ipoib_multicast.c
@@ -43,6 +43,8 @@
#include <linux/delay.h>
#include <linux/completion.h>
+#include <net/dst.h>
+
#include "ipoib.h"
#ifdef CONFIG_INFINIBAND_IPOIB_DEBUG
@@ -53,7 +55,7 @@ MODULE_PARM_DESC(mcast_debug_level,
"Enable multicast debug tracing if > 0");
#endif
-static DECLARE_MUTEX(mcast_mutex);
+static DEFINE_MUTEX(mcast_mutex);
/* Used for all multicast joins (broadcast, IPv4 mcast and IPv6 mcast) */
struct ipoib_mcast {
@@ -95,8 +97,6 @@ static void ipoib_mcast_free(struct ipoib_mcast *mcast)
struct ipoib_dev_priv *priv = netdev_priv(dev);
struct ipoib_neigh *neigh, *tmp;
unsigned long flags;
- LIST_HEAD(ah_list);
- struct ipoib_ah *ah, *tah;
ipoib_dbg_mcast(netdev_priv(dev),
"deleting multicast group " IPOIB_GID_FMT "\n",
@@ -105,8 +105,14 @@ static void ipoib_mcast_free(struct ipoib_mcast *mcast)
spin_lock_irqsave(&priv->lock, flags);
list_for_each_entry_safe(neigh, tmp, &mcast->neigh_list, list) {
+ /*
+ * It's safe to call ipoib_put_ah() inside priv->lock
+ * here, because we know that mcast->ah will always
+ * hold one more reference, so ipoib_put_ah() will
+ * never do more than decrement the ref count.
+ */
if (neigh->ah)
- list_add_tail(&neigh->ah->list, &ah_list);
+ ipoib_put_ah(neigh->ah);
*to_ipoib_neigh(neigh->neighbour) = NULL;
neigh->neighbour->ops->destructor = NULL;
kfree(neigh);
@@ -114,9 +120,6 @@ static void ipoib_mcast_free(struct ipoib_mcast *mcast)
spin_unlock_irqrestore(&priv->lock, flags);
- list_for_each_entry_safe(ah, tah, &ah_list, list)
- ipoib_put_ah(ah);
-
if (mcast->ah)
ipoib_put_ah(mcast->ah);
@@ -135,20 +138,14 @@ static struct ipoib_mcast *ipoib_mcast_alloc(struct net_device *dev,
if (!mcast)
return NULL;
- init_completion(&mcast->done);
-
mcast->dev = dev;
mcast->created = jiffies;
mcast->backoff = 1;
- mcast->logcount = 0;
INIT_LIST_HEAD(&mcast->list);
INIT_LIST_HEAD(&mcast->neigh_list);
skb_queue_head_init(&mcast->pkt_queue);
- mcast->ah = NULL;
- mcast->query = NULL;
-
return mcast;
}
@@ -350,6 +347,8 @@ static int ipoib_mcast_sendonly_join(struct ipoib_mcast *mcast)
rec.port_gid = priv->local_gid;
rec.pkey = cpu_to_be16(priv->pkey);
+ init_completion(&mcast->done);
+
ret = ib_sa_mcmember_rec_set(priv->ca, priv->port, &rec,
IB_SA_MCMEMBER_REC_MGID |
IB_SA_MCMEMBER_REC_PORT_GID |
@@ -386,10 +385,10 @@ static void ipoib_mcast_join_complete(int status,
if (!status && !ipoib_mcast_join_finish(mcast, mcmember)) {
mcast->backoff = 1;
- down(&mcast_mutex);
+ mutex_lock(&mcast_mutex);
if (test_bit(IPOIB_MCAST_RUN, &priv->flags))
queue_work(ipoib_workqueue, &priv->mcast_task);
- up(&mcast_mutex);
+ mutex_unlock(&mcast_mutex);
complete(&mcast->done);
return;
}
@@ -419,7 +418,7 @@ static void ipoib_mcast_join_complete(int status,
mcast->query = NULL;
- down(&mcast_mutex);
+ mutex_lock(&mcast_mutex);
if (test_bit(IPOIB_MCAST_RUN, &priv->flags)) {
if (status == -ETIMEDOUT)
queue_work(ipoib_workqueue, &priv->mcast_task);
@@ -428,7 +427,7 @@ static void ipoib_mcast_join_complete(int status,
mcast->backoff * HZ);
} else
complete(&mcast->done);
- up(&mcast_mutex);
+ mutex_unlock(&mcast_mutex);
return;
}
@@ -469,6 +468,8 @@ static void ipoib_mcast_join(struct net_device *dev, struct ipoib_mcast *mcast,
rec.traffic_class = priv->broadcast->mcmember.traffic_class;
}
+ init_completion(&mcast->done);
+
ret = ib_sa_mcmember_rec_set(priv->ca, priv->port, &rec, comp_mask,
mcast->backoff * 1000, GFP_ATOMIC,
ipoib_mcast_join_complete,
@@ -481,12 +482,12 @@ static void ipoib_mcast_join(struct net_device *dev, struct ipoib_mcast *mcast,
if (mcast->backoff > IPOIB_MAX_BACKOFF_SECONDS)
mcast->backoff = IPOIB_MAX_BACKOFF_SECONDS;
- down(&mcast_mutex);
+ mutex_lock(&mcast_mutex);
if (test_bit(IPOIB_MCAST_RUN, &priv->flags))
queue_delayed_work(ipoib_workqueue,
&priv->mcast_task,
mcast->backoff * HZ);
- up(&mcast_mutex);
+ mutex_unlock(&mcast_mutex);
} else
mcast->query_id = ret;
}
@@ -519,11 +520,11 @@ void ipoib_mcast_join_task(void *dev_ptr)
priv->broadcast = ipoib_mcast_alloc(dev, 1);
if (!priv->broadcast) {
ipoib_warn(priv, "failed to allocate broadcast group\n");
- down(&mcast_mutex);
+ mutex_lock(&mcast_mutex);
if (test_bit(IPOIB_MCAST_RUN, &priv->flags))
queue_delayed_work(ipoib_workqueue,
&priv->mcast_task, HZ);
- up(&mcast_mutex);
+ mutex_unlock(&mcast_mutex);
return;
}
@@ -579,10 +580,10 @@ int ipoib_mcast_start_thread(struct net_device *dev)
ipoib_dbg_mcast(priv, "starting multicast thread\n");
- down(&mcast_mutex);
+ mutex_lock(&mcast_mutex);
if (!test_and_set_bit(IPOIB_MCAST_RUN, &priv->flags))
queue_work(ipoib_workqueue, &priv->mcast_task);
- up(&mcast_mutex);
+ mutex_unlock(&mcast_mutex);
return 0;
}
@@ -594,10 +595,10 @@ int ipoib_mcast_stop_thread(struct net_device *dev, int flush)
ipoib_dbg_mcast(priv, "stopping multicast thread\n");
- down(&mcast_mutex);
+ mutex_lock(&mcast_mutex);
clear_bit(IPOIB_MCAST_RUN, &priv->flags);
cancel_delayed_work(&priv->mcast_task);
- up(&mcast_mutex);
+ mutex_unlock(&mcast_mutex);
if (flush)
flush_workqueue(ipoib_workqueue);
@@ -741,48 +742,23 @@ void ipoib_mcast_dev_flush(struct net_device *dev)
{
struct ipoib_dev_priv *priv = netdev_priv(dev);
LIST_HEAD(remove_list);
- struct ipoib_mcast *mcast, *tmcast, *nmcast;
+ struct ipoib_mcast *mcast, *tmcast;
unsigned long flags;
ipoib_dbg_mcast(priv, "flushing multicast list\n");
spin_lock_irqsave(&priv->lock, flags);
- list_for_each_entry_safe(mcast, tmcast, &priv->multicast_list, list) {
- nmcast = ipoib_mcast_alloc(dev, 0);
- if (nmcast) {
- nmcast->flags =
- mcast->flags & (1 << IPOIB_MCAST_FLAG_SENDONLY);
-
- nmcast->mcmember.mgid = mcast->mcmember.mgid;
-
- /* Add the new group in before the to-be-destroyed group */
- list_add_tail(&nmcast->list, &mcast->list);
- list_del_init(&mcast->list);
-
- rb_replace_node(&mcast->rb_node, &nmcast->rb_node,
- &priv->multicast_tree);
- list_add_tail(&mcast->list, &remove_list);
- } else {
- ipoib_warn(priv, "could not reallocate multicast group "
- IPOIB_GID_FMT "\n",
- IPOIB_GID_ARG(mcast->mcmember.mgid));
- }
+ list_for_each_entry_safe(mcast, tmcast, &priv->multicast_list, list) {
+ list_del(&mcast->list);
+ rb_erase(&mcast->rb_node, &priv->multicast_tree);
+ list_add_tail(&mcast->list, &remove_list);
}
if (priv->broadcast) {
- nmcast = ipoib_mcast_alloc(dev, 0);
- if (nmcast) {
- nmcast->mcmember.mgid = priv->broadcast->mcmember.mgid;
-
- rb_replace_node(&priv->broadcast->rb_node,
- &nmcast->rb_node,
- &priv->multicast_tree);
-
- list_add_tail(&priv->broadcast->list, &remove_list);
- }
-
- priv->broadcast = nmcast;
+ rb_erase(&priv->broadcast->rb_node, &priv->multicast_tree);
+ list_add_tail(&priv->broadcast->list, &remove_list);
+ priv->broadcast = NULL;
}
spin_unlock_irqrestore(&priv->lock, flags);
@@ -793,24 +769,6 @@ void ipoib_mcast_dev_flush(struct net_device *dev)
}
}
-void ipoib_mcast_dev_down(struct net_device *dev)
-{
- struct ipoib_dev_priv *priv = netdev_priv(dev);
- unsigned long flags;
-
- /* Delete broadcast since it will be recreated */
- if (priv->broadcast) {
- ipoib_dbg_mcast(priv, "deleting broadcast group\n");
-
- spin_lock_irqsave(&priv->lock, flags);
- rb_erase(&priv->broadcast->rb_node, &priv->multicast_tree);
- spin_unlock_irqrestore(&priv->lock, flags);
- ipoib_mcast_leave(dev, priv->broadcast);
- ipoib_mcast_free(priv->broadcast);
- priv->broadcast = NULL;
- }
-}
-
void ipoib_mcast_restart_task(void *dev_ptr)
{
struct net_device *dev = dev_ptr;
@@ -824,7 +782,8 @@ void ipoib_mcast_restart_task(void *dev_ptr)
ipoib_mcast_stop_thread(dev, 0);
- spin_lock_irqsave(&priv->lock, flags);
+ spin_lock_irqsave(&dev->xmit_lock, flags);
+ spin_lock(&priv->lock);
/*
* Unfortunately, the networking core only gives us a list of all of
@@ -896,7 +855,9 @@ void ipoib_mcast_restart_task(void *dev_ptr)
list_add_tail(&mcast->list, &remove_list);
}
}
- spin_unlock_irqrestore(&priv->lock, flags);
+
+ spin_unlock(&priv->lock);
+ spin_unlock_irqrestore(&dev->xmit_lock, flags);
/* We have to cancel outside of the spinlock */
list_for_each_entry_safe(mcast, tmcast, &remove_list, list) {
diff --git a/drivers/infiniband/ulp/ipoib/ipoib_verbs.c b/drivers/infiniband/ulp/ipoib/ipoib_verbs.c
index e829e10400e3..faaf10e5fc7b 100644
--- a/drivers/infiniband/ulp/ipoib/ipoib_verbs.c
+++ b/drivers/infiniband/ulp/ipoib/ipoib_verbs.c
@@ -65,9 +65,9 @@ int ipoib_mcast_attach(struct net_device *dev, u16 mlid, union ib_gid *mgid)
}
/* attach QP to multicast group */
- down(&priv->mcast_mutex);
+ mutex_lock(&priv->mcast_mutex);
ret = ib_attach_mcast(priv->qp, mgid, mlid);
- up(&priv->mcast_mutex);
+ mutex_unlock(&priv->mcast_mutex);
if (ret)
ipoib_warn(priv, "failed to attach to multicast group, ret = %d\n", ret);
@@ -81,9 +81,9 @@ int ipoib_mcast_detach(struct net_device *dev, u16 mlid, union ib_gid *mgid)
struct ipoib_dev_priv *priv = netdev_priv(dev);
int ret;
- down(&priv->mcast_mutex);
+ mutex_lock(&priv->mcast_mutex);
ret = ib_detach_mcast(priv->qp, mgid, mlid);
- up(&priv->mcast_mutex);
+ mutex_unlock(&priv->mcast_mutex);
if (ret)
ipoib_warn(priv, "ib_detach_mcast failed (result = %d)\n", ret);
diff --git a/drivers/infiniband/ulp/ipoib/ipoib_vlan.c b/drivers/infiniband/ulp/ipoib/ipoib_vlan.c
index d280b341a37f..4ca175553f9f 100644
--- a/drivers/infiniband/ulp/ipoib/ipoib_vlan.c
+++ b/drivers/infiniband/ulp/ipoib/ipoib_vlan.c
@@ -63,7 +63,7 @@ int ipoib_vlan_add(struct net_device *pdev, unsigned short pkey)
ppriv = netdev_priv(pdev);
- down(&ppriv->vlan_mutex);
+ mutex_lock(&ppriv->vlan_mutex);
/*
* First ensure this isn't a duplicate. We check the parent device and
@@ -124,7 +124,7 @@ int ipoib_vlan_add(struct net_device *pdev, unsigned short pkey)
list_add_tail(&priv->list, &ppriv->child_intfs);
- up(&ppriv->vlan_mutex);
+ mutex_unlock(&ppriv->vlan_mutex);
return 0;
@@ -139,7 +139,7 @@ device_init_failed:
free_netdev(priv->dev);
err:
- up(&ppriv->vlan_mutex);
+ mutex_unlock(&ppriv->vlan_mutex);
return result;
}
@@ -153,7 +153,7 @@ int ipoib_vlan_delete(struct net_device *pdev, unsigned short pkey)
ppriv = netdev_priv(pdev);
- down(&ppriv->vlan_mutex);
+ mutex_lock(&ppriv->vlan_mutex);
list_for_each_entry_safe(priv, tpriv, &ppriv->child_intfs, list) {
if (priv->pkey == pkey) {
unregister_netdev(priv->dev);
@@ -167,7 +167,7 @@ int ipoib_vlan_delete(struct net_device *pdev, unsigned short pkey)
break;
}
}
- up(&ppriv->vlan_mutex);
+ mutex_unlock(&ppriv->vlan_mutex);
return ret;
}
diff --git a/drivers/infiniband/ulp/srp/ib_srp.c b/drivers/infiniband/ulp/srp/ib_srp.c
index ee9fe226ae99..31207e664148 100644
--- a/drivers/infiniband/ulp/srp/ib_srp.c
+++ b/drivers/infiniband/ulp/srp/ib_srp.c
@@ -39,6 +39,7 @@
#include <linux/string.h>
#include <linux/parser.h>
#include <linux/random.h>
+#include <linux/jiffies.h>
#include <asm/atomic.h>
@@ -1515,8 +1516,7 @@ static ssize_t show_port(struct class_device *class_dev, char *buf)
static CLASS_DEVICE_ATTR(port, S_IRUGO, show_port, NULL);
-static struct srp_host *srp_add_port(struct ib_device *device,
- __be64 node_guid, u8 port)
+static struct srp_host *srp_add_port(struct ib_device *device, u8 port)
{
struct srp_host *host;
@@ -1531,7 +1531,7 @@ static struct srp_host *srp_add_port(struct ib_device *device,
host->port = port;
host->initiator_port_id[7] = port;
- memcpy(host->initiator_port_id + 8, &node_guid, 8);
+ memcpy(host->initiator_port_id + 8, &device->node_guid, 8);
host->pd = ib_alloc_pd(device);
if (IS_ERR(host->pd))
@@ -1579,22 +1579,11 @@ static void srp_add_one(struct ib_device *device)
{
struct list_head *dev_list;
struct srp_host *host;
- struct ib_device_attr *dev_attr;
int s, e, p;
- dev_attr = kmalloc(sizeof *dev_attr, GFP_KERNEL);
- if (!dev_attr)
- return;
-
- if (ib_query_device(device, dev_attr)) {
- printk(KERN_WARNING PFX "Couldn't query node GUID for %s.\n",
- device->name);
- goto out;
- }
-
dev_list = kmalloc(sizeof *dev_list, GFP_KERNEL);
if (!dev_list)
- goto out;
+ return;
INIT_LIST_HEAD(dev_list);
@@ -1607,15 +1596,12 @@ static void srp_add_one(struct ib_device *device)
}
for (p = s; p <= e; ++p) {
- host = srp_add_port(device, dev_attr->node_guid, p);
+ host = srp_add_port(device, p);
if (host)
list_add_tail(&host->list, dev_list);
}
ib_set_client_data(device, &srp_client, dev_list);
-
-out:
- kfree(dev_attr);
}
static void srp_remove_one(struct ib_device *device)