IB/umad: fix RMPP handling
[pandora-kernel.git] / drivers / infiniband / core / user_mad.c
index a64d6b4..eb7f525 100644 (file)
@@ -31,7 +31,7 @@
  * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
  * SOFTWARE.
  *
- * $Id: user_mad.c 2814 2005-07-06 19:14:09Z halr $
+ * $Id: user_mad.c 4010 2005-11-09 23:11:56Z roland $
  */
 
 #include <linux/module.h>
@@ -64,18 +64,42 @@ enum {
        IB_UMAD_MINOR_BASE = 0
 };
 
+/*
+ * Our lifetime rules for these structs are the following: each time a
+ * device special file is opened, we look up the corresponding struct
+ * ib_umad_port by minor in the umad_port[] table while holding the
+ * port_lock.  If this lookup succeeds, we take a reference on the
+ * ib_umad_port's struct ib_umad_device while still holding the
+ * port_lock; if the lookup fails, we fail the open().  We drop these
+ * references in the corresponding close().
+ *
+ * In addition to references coming from open character devices, there
+ * is one more reference to each ib_umad_device representing the
+ * module's reference taken when allocating the ib_umad_device in
+ * ib_umad_add_one().
+ *
+ * When destroying an ib_umad_device, we clear all of its
+ * ib_umad_ports from umad_port[] while holding port_lock before
+ * dropping the module's reference to the ib_umad_device.  This is
+ * always safe because any open() calls will either succeed and obtain
+ * a reference before we clear the umad_port[] entries, or fail after
+ * we clear the umad_port[] entries.
+ */
+
 struct ib_umad_port {
-       int                    devnum;
-       struct cdev            dev;
-       struct class_device    class_dev;
+       struct cdev           *dev;
+       struct class_device   *class_dev;
 
-       int                    sm_devnum;
-       struct cdev            sm_dev;
-       struct class_device    sm_class_dev;
+       struct cdev           *sm_dev;
+       struct class_device   *sm_class_dev;
        struct semaphore       sm_sem;
 
+       struct rw_semaphore    mutex;
+       struct list_head       file_list;
+
        struct ib_device      *ib_dev;
        struct ib_umad_device *umad_dev;
+       int                    dev_num;
        u8                     port_num;
 };
 
@@ -86,42 +110,59 @@ struct ib_umad_device {
 };
 
 struct ib_umad_file {
-       struct ib_umad_port *port;
-       spinlock_t           recv_lock;
-       struct list_head     recv_list;
-       wait_queue_head_t    recv_wait;
-       struct rw_semaphore  agent_mutex;
-       struct ib_mad_agent *agent[IB_UMAD_MAX_AGENTS];
-       struct ib_mr        *mr[IB_UMAD_MAX_AGENTS];
+       struct ib_umad_port    *port;
+       struct list_head        recv_list;
+       struct list_head        port_list;
+       spinlock_t              recv_lock;
+       wait_queue_head_t       recv_wait;
+       struct ib_mad_agent    *agent[IB_UMAD_MAX_AGENTS];
+       int                     agents_dead;
 };
 
 struct ib_umad_packet {
-       struct ib_ah      *ah;
        struct ib_mad_send_buf *msg;
        struct list_head   list;
        int                length;
-       DECLARE_PCI_UNMAP_ADDR(mapping)
        struct ib_user_mad mad;
 };
 
+static struct class *umad_class;
+
 static const dev_t base_dev = MKDEV(IB_UMAD_MAJOR, IB_UMAD_MINOR_BASE);
-static spinlock_t map_lock;
+
+static DEFINE_SPINLOCK(port_lock);
+static struct ib_umad_port *umad_port[IB_UMAD_MAX_PORTS];
 static DECLARE_BITMAP(dev_map, IB_UMAD_MAX_PORTS * 2);
 
 static void ib_umad_add_one(struct ib_device *device);
 static void ib_umad_remove_one(struct ib_device *device);
 
+static void ib_umad_release_dev(struct kref *ref)
+{
+       struct ib_umad_device *dev =
+               container_of(ref, struct ib_umad_device, ref);
+
+       kfree(dev);
+}
+
+/* caller must hold port->mutex at least for reading */
+static struct ib_mad_agent *__get_agent(struct ib_umad_file *file, int id)
+{
+       return file->agents_dead ? NULL : file->agent[id];
+}
+
 static int queue_packet(struct ib_umad_file *file,
                        struct ib_mad_agent *agent,
                        struct ib_umad_packet *packet)
 {
        int ret = 1;
 
-       down_read(&file->agent_mutex);
+       down_read(&file->port->mutex);
+
        for (packet->mad.hdr.id = 0;
             packet->mad.hdr.id < IB_UMAD_MAX_AGENTS;
             packet->mad.hdr.id++)
-               if (agent == file->agent[packet->mad.hdr.id]) {
+               if (agent == __get_agent(file, packet->mad.hdr.id)) {
                        spin_lock_irq(&file->recv_lock);
                        list_add_tail(&packet->list, &file->recv_list);
                        spin_unlock_irq(&file->recv_lock);
@@ -130,7 +171,7 @@ static int queue_packet(struct ib_umad_file *file,
                        break;
                }
 
-       up_read(&file->agent_mutex);
+       up_read(&file->port->mutex);
 
        return ret;
 }
@@ -139,22 +180,19 @@ static void send_handler(struct ib_mad_agent *agent,
                         struct ib_mad_send_wc *send_wc)
 {
        struct ib_umad_file *file = agent->context;
-       struct ib_umad_packet *timeout, *packet =
-               (void *) (unsigned long) send_wc->wr_id;
+       struct ib_umad_packet *timeout;
+       struct ib_umad_packet *packet = send_wc->send_buf->context[0];
 
-       ib_destroy_ah(packet->msg->send_wr.wr.ud.ah);
+       ib_destroy_ah(packet->msg->ah);
        ib_free_send_mad(packet->msg);
 
        if (send_wc->status == IB_WC_RESP_TIMEOUT_ERR) {
-               timeout = kmalloc(sizeof *timeout + sizeof (struct ib_mad_hdr),
-                                 GFP_KERNEL);
+               timeout = kzalloc(sizeof *timeout + IB_MGMT_MAD_HDR, GFP_KERNEL);
                if (!timeout)
                        goto out;
 
-               memset(timeout, 0, sizeof *timeout + sizeof (struct ib_mad_hdr));
-
-               timeout->length = sizeof (struct ib_mad_hdr);
-               timeout->mad.hdr.id = packet->mad.hdr.id;
+               timeout->length         = IB_MGMT_MAD_HDR;
+               timeout->mad.hdr.id     = packet->mad.hdr.id;
                timeout->mad.hdr.status = ETIMEDOUT;
                memcpy(timeout->mad.data, packet->mad.data,
                       sizeof (struct ib_mad_hdr));
@@ -177,11 +215,10 @@ static void recv_handler(struct ib_mad_agent *agent,
                goto out;
 
        length = mad_recv_wc->mad_len;
-       packet = kmalloc(sizeof *packet + length, GFP_KERNEL);
+       packet = kzalloc(sizeof *packet + length, GFP_KERNEL);
        if (!packet)
                goto out;
 
-       memset(packet, 0, sizeof *packet + length);
        packet->length = length;
 
        ib_coalesce_recv_mad(mad_recv_wc, packet->mad.data);
@@ -247,7 +284,7 @@ static ssize_t ib_umad_read(struct file *filp, char __user *buf,
                else
                        ret = -ENOSPC;
        } else if (copy_to_user(buf, &packet->mad,
-                             packet->length + sizeof (struct ib_user_mad)))
+                               packet->length + sizeof (struct ib_user_mad)))
                ret = -EFAULT;
        else
                ret = packet->length + sizeof (struct ib_user_mad);
@@ -268,26 +305,23 @@ static ssize_t ib_umad_write(struct file *filp, const char __user *buf,
        struct ib_umad_packet *packet;
        struct ib_mad_agent *agent;
        struct ib_ah_attr ah_attr;
-       struct ib_send_wr *bad_wr;
+       struct ib_ah *ah;
        struct ib_rmpp_mad *rmpp_mad;
        u8 method;
        __be64 *tid;
-       int ret, length, hdr_len, data_len, rmpp_hdr_size;
-       int rmpp_active = 0;
+       int ret, length, hdr_len, copy_offset;
+       int rmpp_active, has_rmpp_header;
 
-       if (count < sizeof (struct ib_user_mad))
+       if (count < sizeof (struct ib_user_mad) + IB_MGMT_RMPP_HDR)
                return -EINVAL;
 
        length = count - sizeof (struct ib_user_mad);
-       packet = kmalloc(sizeof *packet + sizeof(struct ib_mad_hdr) +
-                        sizeof(struct ib_rmpp_hdr), GFP_KERNEL);
+       packet = kmalloc(sizeof *packet + IB_MGMT_RMPP_HDR, GFP_KERNEL);
        if (!packet)
                return -ENOMEM;
 
        if (copy_from_user(&packet->mad, buf,
-                           sizeof (struct ib_user_mad) +
-                           sizeof(struct ib_mad_hdr) +
-                           sizeof(struct ib_rmpp_hdr))) {
+                           sizeof (struct ib_user_mad) + IB_MGMT_RMPP_HDR)) {
                ret = -EFAULT;
                goto err;
        }
@@ -298,11 +332,9 @@ static ssize_t ib_umad_write(struct file *filp, const char __user *buf,
                goto err;
        }
 
-       packet->length = length;
-
-       down_read(&file->agent_mutex);
+       down_read(&file->port->mutex);
 
-       agent = file->agent[packet->mad.hdr.id];
+       agent = __get_agent(file, packet->mad.hdr.id);
        if (!agent) {
                ret = -EINVAL;
                goto err_up;
@@ -321,80 +353,63 @@ static ssize_t ib_umad_write(struct file *filp, const char __user *buf,
                ah_attr.grh.traffic_class  = packet->mad.hdr.traffic_class;
        }
 
-       packet->ah = ib_create_ah(agent->qp->pd, &ah_attr);
-       if (IS_ERR(packet->ah)) {
-               ret = PTR_ERR(packet->ah);
+       ah = ib_create_ah(agent->qp->pd, &ah_attr);
+       if (IS_ERR(ah)) {
+               ret = PTR_ERR(ah);
                goto err_up;
        }
 
        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 = offsetof(struct ib_sa_mad, data);
-                       data_len = length - hdr_len;
-               } 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 = offsetof(struct ib_vendor_mad, data);
-                               data_len = length - hdr_len;
-               } 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 {
-               if (length > sizeof(struct ib_mad)) {
-                       ret = -EINVAL;
-                       goto err_ah;
-               }
-               hdr_len = offsetof(struct ib_mad, data);
-               data_len = length - hdr_len;
+               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,
                                         be32_to_cpu(packet->mad.hdr.qpn),
-                                        0, packet->ah, rmpp_active,
-                                        hdr_len, data_len,
+                                        0, rmpp_active,
+                                        hdr_len, length - hdr_len,
                                         GFP_KERNEL);
        if (IS_ERR(packet->msg)) {
                ret = PTR_ERR(packet->msg);
                goto err_ah;
        }
 
-       packet->msg->send_wr.wr.ud.timeout_ms  = packet->mad.hdr.timeout_ms;
-       packet->msg->send_wr.wr.ud.retries = packet->mad.hdr.retries;
-
-       /* Override send WR WRID initialized in ib_create_send_mad */
-       packet->msg->send_wr.wr_id = (unsigned long) packet;
-
-       if (!rmpp_active) {
-               /* Copy message from user into send buffer */
-               if (copy_from_user(packet->msg->mad,
-                                  buf + sizeof(struct ib_user_mad), length)) {
-                       ret = -EFAULT;
-                       goto err_msg;
-               }
-       } else {
-               rmpp_hdr_size = sizeof(struct ib_mad_hdr) +
-                               sizeof(struct ib_rmpp_hdr);
-
-               /* Only copy MAD headers (RMPP header in place) */
-               memcpy(packet->msg->mad, packet->mad.data,
-                      sizeof(struct ib_mad_hdr));
+       packet->msg->ah         = ah;
+       packet->msg->timeout_ms = packet->mad.hdr.timeout_ms;
+       packet->msg->retries    = packet->mad.hdr.retries;
+       packet->msg->context[0] = packet;
 
-               /* Now, copy rest of message from user into send buffer */
-               if (copy_from_user(((struct ib_rmpp_mad *) packet->msg->mad)->data,
-                                  buf + sizeof (struct ib_user_mad) + rmpp_hdr_size,
-                                  length - rmpp_hdr_size)) {
-                       ret = -EFAULT;
-                       goto err_msg;
-               }
+       /* Copy MAD headers (RMPP header in place) */
+       memcpy(packet->msg->mad, packet->mad.data, IB_MGMT_MAD_HDR);
+       /* Now, copy rest of message from user into send buffer */
+       if (copy_from_user(packet->msg->mad + copy_offset,
+                          buf + sizeof (struct ib_user_mad) + copy_offset,
+                          length - copy_offset)) {
+               ret = -EFAULT;
+               goto err_msg;
        }
 
        /*
@@ -403,32 +418,32 @@ static ssize_t ib_umad_write(struct file *filp, const char __user *buf,
         * transaction ID matches the agent being used to send the
         * MAD.
         */
-       method = packet->msg->mad->mad_hdr.method;
+       method = ((struct ib_mad_hdr *) packet->msg->mad)->method;
 
        if (!(method & IB_MGMT_METHOD_RESP)       &&
            method != IB_MGMT_METHOD_TRAP_REPRESS &&
            method != IB_MGMT_METHOD_SEND) {
-               tid = &packet->msg->mad->mad_hdr.tid;
+               tid = &((struct ib_mad_hdr *) packet->msg->mad)->tid;
                *tid = cpu_to_be64(((u64) agent->hi_tid) << 32 |
                                   (be64_to_cpup(tid) & 0xffffffff));
        }
 
-       ret = ib_post_send_mad(agent, &packet->msg->send_wr, &bad_wr);
+       ret = ib_post_send_mad(packet->msg, NULL);
        if (ret)
                goto err_msg;
 
-       up_read(&file->agent_mutex);
+       up_read(&file->port->mutex);
 
-       return sizeof (struct ib_user_mad_hdr) + packet->length;
+       return count;
 
 err_msg:
        ib_free_send_mad(packet->msg);
 
 err_ah:
-       ib_destroy_ah(packet->ah);
+       ib_destroy_ah(ah);
 
 err_up:
-       up_read(&file->agent_mutex);
+       up_read(&file->port->mutex);
 
 err:
        kfree(packet);
@@ -458,7 +473,12 @@ static int ib_umad_reg_agent(struct ib_umad_file *file, unsigned long arg)
        int agent_id;
        int ret;
 
-       down_write(&file->agent_mutex);
+       down_write(&file->port->mutex);
+
+       if (!file->port->ib_dev) {
+               ret = -EPIPE;
+               goto out;
+       }
 
        if (copy_from_user(&ureq, (void __user *) arg, sizeof ureq)) {
                ret = -EFAULT;
@@ -471,7 +491,7 @@ static int ib_umad_reg_agent(struct ib_umad_file *file, unsigned long arg)
        }
 
        for (agent_id = 0; agent_id < IB_UMAD_MAX_AGENTS; ++agent_id)
-               if (!file->agent[agent_id])
+               if (!__get_agent(file, agent_id))
                        goto found;
 
        ret = -ENOMEM;
@@ -495,58 +515,46 @@ found:
                goto out;
        }
 
-       file->agent[agent_id] = agent;
-
-       file->mr[agent_id] = ib_get_dma_mr(agent->qp->pd, IB_ACCESS_LOCAL_WRITE);
-       if (IS_ERR(file->mr[agent_id])) {
-               ret = -ENOMEM;
-               goto err;
-       }
-
        if (put_user(agent_id,
                     (u32 __user *) (arg + offsetof(struct ib_user_mad_reg_req, id)))) {
                ret = -EFAULT;
-               goto err_mr;
+               ib_unregister_mad_agent(agent);
+               goto out;
        }
 
+       file->agent[agent_id] = agent;
        ret = 0;
-       goto out;
-
-err_mr:
-       ib_dereg_mr(file->mr[agent_id]);
-
-err:
-       file->agent[agent_id] = NULL;
-       ib_unregister_mad_agent(agent);
 
 out:
-       up_write(&file->agent_mutex);
+       up_write(&file->port->mutex);
        return ret;
 }
 
 static int ib_umad_unreg_agent(struct ib_umad_file *file, unsigned long arg)
 {
+       struct ib_mad_agent *agent = NULL;
        u32 id;
        int ret = 0;
 
-       down_write(&file->agent_mutex);
+       if (get_user(id, (u32 __user *) arg))
+               return -EFAULT;
 
-       if (get_user(id, (u32 __user *) arg)) {
-               ret = -EFAULT;
-               goto out;
-       }
+       down_write(&file->port->mutex);
 
-       if (id < 0 || id >= IB_UMAD_MAX_AGENTS || !file->agent[id]) {
+       if (id < 0 || id >= IB_UMAD_MAX_AGENTS || !__get_agent(file, id)) {
                ret = -EINVAL;
                goto out;
        }
 
-       ib_dereg_mr(file->mr[id]);
-       ib_unregister_mad_agent(file->agent[id]);
+       agent = file->agent[id];
        file->agent[id] = NULL;
 
 out:
-       up_write(&file->agent_mutex);
+       up_write(&file->port->mutex);
+
+       if (agent)
+               ib_unregister_mad_agent(agent);
+
        return ret;
 }
 
@@ -565,43 +573,76 @@ static long ib_umad_ioctl(struct file *filp, unsigned int cmd,
 
 static int ib_umad_open(struct inode *inode, struct file *filp)
 {
-       struct ib_umad_port *port =
-               container_of(inode->i_cdev, struct ib_umad_port, dev);
+       struct ib_umad_port *port;
        struct ib_umad_file *file;
+       int ret = 0;
 
-       file = kmalloc(sizeof *file, GFP_KERNEL);
-       if (!file)
-               return -ENOMEM;
+       spin_lock(&port_lock);
+       port = umad_port[iminor(inode) - IB_UMAD_MINOR_BASE];
+       if (port)
+               kref_get(&port->umad_dev->ref);
+       spin_unlock(&port_lock);
+
+       if (!port)
+               return -ENXIO;
 
-       memset(file, 0, sizeof *file);
+       down_write(&port->mutex);
+
+       if (!port->ib_dev) {
+               ret = -ENXIO;
+               goto out;
+       }
+
+       file = kzalloc(sizeof *file, GFP_KERNEL);
+       if (!file) {
+               kref_put(&port->umad_dev->ref, ib_umad_release_dev);
+               ret = -ENOMEM;
+               goto out;
+       }
 
        spin_lock_init(&file->recv_lock);
-       init_rwsem(&file->agent_mutex);
        INIT_LIST_HEAD(&file->recv_list);
        init_waitqueue_head(&file->recv_wait);
 
        file->port = port;
        filp->private_data = file;
 
-       return 0;
+       list_add_tail(&file->port_list, &port->file_list);
+
+out:
+       up_write(&port->mutex);
+       return ret;
 }
 
 static int ib_umad_close(struct inode *inode, struct file *filp)
 {
        struct ib_umad_file *file = filp->private_data;
+       struct ib_umad_device *dev = file->port->umad_dev;
        struct ib_umad_packet *packet, *tmp;
+       int already_dead;
        int i;
 
-       for (i = 0; i < IB_UMAD_MAX_AGENTS; ++i)
-               if (file->agent[i]) {
-                       ib_dereg_mr(file->mr[i]);
-                       ib_unregister_mad_agent(file->agent[i]);
-               }
+       down_write(&file->port->mutex);
+
+       already_dead = file->agents_dead;
+       file->agents_dead = 1;
 
        list_for_each_entry_safe(packet, tmp, &file->recv_list, list)
                kfree(packet);
 
+       list_del(&file->port_list);
+
+       downgrade_write(&file->port->mutex);
+
+       if (!already_dead)
+               for (i = 0; i < IB_UMAD_MAX_AGENTS; ++i)
+                       if (file->agent[i])
+                               ib_unregister_mad_agent(file->agent[i]);
+
+       up_read(&file->port->mutex);
+
        kfree(file);
+       kref_put(&dev->ref, ib_umad_release_dev);
 
        return 0;
 }
@@ -619,30 +660,46 @@ static struct file_operations umad_fops = {
 
 static int ib_umad_sm_open(struct inode *inode, struct file *filp)
 {
-       struct ib_umad_port *port =
-               container_of(inode->i_cdev, struct ib_umad_port, sm_dev);
+       struct ib_umad_port *port;
        struct ib_port_modify props = {
                .set_port_cap_mask = IB_PORT_SM
        };
        int ret;
 
+       spin_lock(&port_lock);
+       port = umad_port[iminor(inode) - IB_UMAD_MINOR_BASE - IB_UMAD_MAX_PORTS];
+       if (port)
+               kref_get(&port->umad_dev->ref);
+       spin_unlock(&port_lock);
+
+       if (!port)
+               return -ENXIO;
+
        if (filp->f_flags & O_NONBLOCK) {
-               if (down_trylock(&port->sm_sem))
-                       return -EAGAIN;
+               if (down_trylock(&port->sm_sem)) {
+                       ret = -EAGAIN;
+                       goto fail;
+               }
        } else {
-               if (down_interruptible(&port->sm_sem))
-                       return -ERESTARTSYS;
+               if (down_interruptible(&port->sm_sem)) {
+                       ret = -ERESTARTSYS;
+                       goto fail;
+               }
        }
 
        ret = ib_modify_port(port->ib_dev, port->port_num, 0, &props);
        if (ret) {
                up(&port->sm_sem);
-               return ret;
+               goto fail;
        }
 
        filp->private_data = port;
 
        return 0;
+
+fail:
+       kref_put(&port->umad_dev->ref, ib_umad_release_dev);
+       return ret;
 }
 
 static int ib_umad_sm_close(struct inode *inode, struct file *filp)
@@ -651,11 +708,17 @@ static int ib_umad_sm_close(struct inode *inode, struct file *filp)
        struct ib_port_modify props = {
                .clr_port_cap_mask = IB_PORT_SM
        };
-       int ret;
+       int ret = 0;
+
+       down_write(&port->mutex);
+       if (port->ib_dev)
+               ret = ib_modify_port(port->ib_dev, port->port_num, 0, &props);
+       up_write(&port->mutex);
 
-       ret = ib_modify_port(port->ib_dev, port->port_num, 0, &props);
        up(&port->sm_sem);
 
+       kref_put(&port->umad_dev->ref, ib_umad_release_dev);
+
        return ret;
 }
 
@@ -671,21 +734,13 @@ static struct ib_client umad_client = {
        .remove = ib_umad_remove_one
 };
 
-static ssize_t show_dev(struct class_device *class_dev, char *buf)
-{
-       struct ib_umad_port *port = class_get_devdata(class_dev);
-
-       if (class_dev == &port->class_dev)
-               return print_dev_t(buf, port->dev.dev);
-       else
-               return print_dev_t(buf, port->sm_dev.dev);
-}
-static CLASS_DEVICE_ATTR(dev, S_IRUGO, show_dev, NULL);
-
 static ssize_t show_ibdev(struct class_device *class_dev, char *buf)
 {
        struct ib_umad_port *port = class_get_devdata(class_dev);
 
+       if (!port)
+               return -ENODEV;
+
        return sprintf(buf, "%s\n", port->ib_dev->name);
 }
 static CLASS_DEVICE_ATTR(ibdev, S_IRUGO, show_ibdev, NULL);
@@ -694,38 +749,13 @@ static ssize_t show_port(struct class_device *class_dev, char *buf)
 {
        struct ib_umad_port *port = class_get_devdata(class_dev);
 
+       if (!port)
+               return -ENODEV;
+
        return sprintf(buf, "%d\n", port->port_num);
 }
 static CLASS_DEVICE_ATTR(port, S_IRUGO, show_port, NULL);
 
-static void ib_umad_release_dev(struct kref *ref)
-{
-       struct ib_umad_device *dev =
-               container_of(ref, struct ib_umad_device, ref);
-
-       kfree(dev);
-}
-
-static void ib_umad_release_port(struct class_device *class_dev)
-{
-       struct ib_umad_port *port = class_get_devdata(class_dev);
-
-       if (class_dev == &port->class_dev) {
-               cdev_del(&port->dev);
-               clear_bit(port->devnum, dev_map);
-       } else {
-               cdev_del(&port->sm_dev);
-               clear_bit(port->sm_devnum, dev_map);
-       }
-
-       kref_put(&port->umad_dev->ref, ib_umad_release_dev);
-}
-
-static struct class umad_class = {
-       .name    = "infiniband_mad",
-       .release = ib_umad_release_port
-};
-
 static ssize_t show_abi_version(struct class *class, char *buf)
 {
        return sprintf(buf, "%d\n", IB_USER_MAD_ABI_VERSION);
@@ -735,91 +765,144 @@ static CLASS_ATTR(abi_version, S_IRUGO, show_abi_version, NULL);
 static int ib_umad_init_port(struct ib_device *device, int port_num,
                             struct ib_umad_port *port)
 {
-       spin_lock(&map_lock);
-       port->devnum = find_first_zero_bit(dev_map, IB_UMAD_MAX_PORTS);
-       if (port->devnum >= IB_UMAD_MAX_PORTS) {
-               spin_unlock(&map_lock);
-               return -1;
-       }
-       port->sm_devnum = find_next_zero_bit(dev_map, IB_UMAD_MAX_PORTS * 2, IB_UMAD_MAX_PORTS);
-       if (port->sm_devnum >= IB_UMAD_MAX_PORTS * 2) {
-               spin_unlock(&map_lock);
+       spin_lock(&port_lock);
+       port->dev_num = find_first_zero_bit(dev_map, IB_UMAD_MAX_PORTS);
+       if (port->dev_num >= IB_UMAD_MAX_PORTS) {
+               spin_unlock(&port_lock);
                return -1;
        }
-       set_bit(port->devnum, dev_map);
-       set_bit(port->sm_devnum, dev_map);
-       spin_unlock(&map_lock);
+       set_bit(port->dev_num, dev_map);
+       spin_unlock(&port_lock);
 
        port->ib_dev   = device;
        port->port_num = port_num;
        init_MUTEX(&port->sm_sem);
+       init_rwsem(&port->mutex);
+       INIT_LIST_HEAD(&port->file_list);
 
-       cdev_init(&port->dev, &umad_fops);
-       port->dev.owner = THIS_MODULE;
-       kobject_set_name(&port->dev.kobj, "umad%d", port->devnum);
-       if (cdev_add(&port->dev, base_dev + port->devnum, 1))
+       port->dev = cdev_alloc();
+       if (!port->dev)
                return -1;
-
-       port->class_dev.class = &umad_class;
-       port->class_dev.dev   = device->dma_device;
-
-       snprintf(port->class_dev.class_id, BUS_ID_SIZE, "umad%d", port->devnum);
-
-       if (class_device_register(&port->class_dev))
+       port->dev->owner = THIS_MODULE;
+       port->dev->ops   = &umad_fops;
+       kobject_set_name(&port->dev->kobj, "umad%d", port->dev_num);
+       if (cdev_add(port->dev, base_dev + port->dev_num, 1))
                goto err_cdev;
 
-       class_set_devdata(&port->class_dev, port);
-       kref_get(&port->umad_dev->ref);
+       port->class_dev = class_device_create(umad_class, NULL, port->dev->dev,
+                                             device->dma_device,
+                                             "umad%d", port->dev_num);
+       if (IS_ERR(port->class_dev))
+               goto err_cdev;
 
-       if (class_device_create_file(&port->class_dev, &class_device_attr_dev))
-               goto err_class;
-       if (class_device_create_file(&port->class_dev, &class_device_attr_ibdev))
+       if (class_device_create_file(port->class_dev, &class_device_attr_ibdev))
                goto err_class;
-       if (class_device_create_file(&port->class_dev, &class_device_attr_port))
+       if (class_device_create_file(port->class_dev, &class_device_attr_port))
                goto err_class;
 
-       cdev_init(&port->sm_dev, &umad_sm_fops);
-       port->sm_dev.owner = THIS_MODULE;
-       kobject_set_name(&port->dev.kobj, "issm%d", port->sm_devnum - IB_UMAD_MAX_PORTS);
-       if (cdev_add(&port->sm_dev, base_dev + port->sm_devnum, 1))
-               return -1;
-
-       port->sm_class_dev.class = &umad_class;
-       port->sm_class_dev.dev   = device->dma_device;
-
-       snprintf(port->sm_class_dev.class_id, BUS_ID_SIZE, "issm%d", port->sm_devnum - IB_UMAD_MAX_PORTS);
+       port->sm_dev = cdev_alloc();
+       if (!port->sm_dev)
+               goto err_class;
+       port->sm_dev->owner = THIS_MODULE;
+       port->sm_dev->ops   = &umad_sm_fops;
+       kobject_set_name(&port->sm_dev->kobj, "issm%d", port->dev_num);
+       if (cdev_add(port->sm_dev, base_dev + port->dev_num + IB_UMAD_MAX_PORTS, 1))
+               goto err_sm_cdev;
 
-       if (class_device_register(&port->sm_class_dev))
+       port->sm_class_dev = class_device_create(umad_class, NULL, port->sm_dev->dev,
+                                                device->dma_device,
+                                                "issm%d", port->dev_num);
+       if (IS_ERR(port->sm_class_dev))
                goto err_sm_cdev;
 
-       class_set_devdata(&port->sm_class_dev, port);
-       kref_get(&port->umad_dev->ref);
+       class_set_devdata(port->class_dev,    port);
+       class_set_devdata(port->sm_class_dev, port);
 
-       if (class_device_create_file(&port->sm_class_dev, &class_device_attr_dev))
+       if (class_device_create_file(port->sm_class_dev, &class_device_attr_ibdev))
                goto err_sm_class;
-       if (class_device_create_file(&port->sm_class_dev, &class_device_attr_ibdev))
-               goto err_sm_class;
-       if (class_device_create_file(&port->sm_class_dev, &class_device_attr_port))
+       if (class_device_create_file(port->sm_class_dev, &class_device_attr_port))
                goto err_sm_class;
 
+       spin_lock(&port_lock);
+       umad_port[port->dev_num] = port;
+       spin_unlock(&port_lock);
+
        return 0;
 
 err_sm_class:
-       class_device_unregister(&port->sm_class_dev);
+       class_device_destroy(umad_class, port->sm_dev->dev);
 
 err_sm_cdev:
-       cdev_del(&port->sm_dev);
+       cdev_del(port->sm_dev);
 
 err_class:
-       class_device_unregister(&port->class_dev);
+       class_device_destroy(umad_class, port->dev->dev);
 
 err_cdev:
-       cdev_del(&port->dev);
-       clear_bit(port->devnum, dev_map);
+       cdev_del(port->dev);
+       clear_bit(port->dev_num, dev_map);
 
        return -1;
 }
 
+static void ib_umad_kill_port(struct ib_umad_port *port)
+{
+       struct ib_umad_file *file;
+       int id;
+
+       class_set_devdata(port->class_dev,    NULL);
+       class_set_devdata(port->sm_class_dev, NULL);
+
+       class_device_destroy(umad_class, port->dev->dev);
+       class_device_destroy(umad_class, port->sm_dev->dev);
+
+       cdev_del(port->dev);
+       cdev_del(port->sm_dev);
+
+       spin_lock(&port_lock);
+       umad_port[port->dev_num] = NULL;
+       spin_unlock(&port_lock);
+
+       down_write(&port->mutex);
+
+       port->ib_dev = NULL;
+
+       /*
+        * Now go through the list of files attached to this port and
+        * unregister all of their MAD agents.  We need to hold
+        * port->mutex while doing this to avoid racing with
+        * ib_umad_close(), but we can't hold the mutex for writing
+        * while calling ib_unregister_mad_agent(), since that might
+        * deadlock by calling back into queue_packet().  So we
+        * downgrade our lock to a read lock, and then drop and
+        * reacquire the write lock for the next iteration.
+        *
+        * We do list_del_init() on the file's list_head so that the
+        * list_del in ib_umad_close() is still OK, even after the
+        * file is removed from the list.
+        */
+       while (!list_empty(&port->file_list)) {
+               file = list_entry(port->file_list.next, struct ib_umad_file,
+                                 port_list);
+
+               file->agents_dead = 1;
+               list_del_init(&file->port_list);
+
+               downgrade_write(&port->mutex);
+
+               for (id = 0; id < IB_UMAD_MAX_AGENTS; ++id)
+                       if (file->agent[id])
+                               ib_unregister_mad_agent(file->agent[id]);
+
+               up_read(&port->mutex);
+               down_write(&port->mutex);
+       }
+
+       up_write(&port->mutex);
+
+       clear_bit(port->dev_num, dev_map);
+}
+
 static void ib_umad_add_one(struct ib_device *device)
 {
        struct ib_umad_device *umad_dev;
@@ -832,15 +915,12 @@ static void ib_umad_add_one(struct ib_device *device)
                e = device->phys_port_cnt;
        }
 
-       umad_dev = kmalloc(sizeof *umad_dev +
+       umad_dev = kzalloc(sizeof *umad_dev +
                           (e - s + 1) * sizeof (struct ib_umad_port),
                           GFP_KERNEL);
        if (!umad_dev)
                return;
 
-       memset(umad_dev, 0, sizeof *umad_dev +
-              (e - s + 1) * sizeof (struct ib_umad_port));
-
        kref_init(&umad_dev->ref);
 
        umad_dev->start_port = s;
@@ -858,10 +938,8 @@ static void ib_umad_add_one(struct ib_device *device)
        return;
 
 err:
-       while (--i >= s) {
-               class_device_unregister(&umad_dev->port[i - s].class_dev);
-               class_device_unregister(&umad_dev->port[i - s].sm_class_dev);
-       }
+       while (--i >= s)
+               ib_umad_kill_port(&umad_dev->port[i - s]);
 
        kref_put(&umad_dev->ref, ib_umad_release_dev);
 }
@@ -874,10 +952,8 @@ static void ib_umad_remove_one(struct ib_device *device)
        if (!umad_dev)
                return;
 
-       for (i = 0; i <= umad_dev->end_port - umad_dev->start_port; ++i) {
-               class_device_unregister(&umad_dev->port[i].class_dev);
-               class_device_unregister(&umad_dev->port[i].sm_class_dev);
-       }
+       for (i = 0; i <= umad_dev->end_port - umad_dev->start_port; ++i)
+               ib_umad_kill_port(&umad_dev->port[i]);
 
        kref_put(&umad_dev->ref, ib_umad_release_dev);
 }
@@ -886,8 +962,6 @@ static int __init ib_umad_init(void)
 {
        int ret;
 
-       spin_lock_init(&map_lock);
-
        ret = register_chrdev_region(base_dev, IB_UMAD_MAX_PORTS * 2,
                                     "infiniband_mad");
        if (ret) {
@@ -895,13 +969,14 @@ static int __init ib_umad_init(void)
                goto out;
        }
 
-       ret = class_register(&umad_class);
-       if (ret) {
+       umad_class = class_create(THIS_MODULE, "infiniband_mad");
+       if (IS_ERR(umad_class)) {
+               ret = PTR_ERR(umad_class);
                printk(KERN_ERR "user_mad: couldn't create class infiniband_mad\n");
                goto out_chrdev;
        }
 
-       ret = class_create_file(&umad_class, &class_attr_abi_version);
+       ret = class_create_file(umad_class, &class_attr_abi_version);
        if (ret) {
                printk(KERN_ERR "user_mad: couldn't create abi_version attribute\n");
                goto out_class;
@@ -916,7 +991,7 @@ static int __init ib_umad_init(void)
        return 0;
 
 out_class:
-       class_unregister(&umad_class);
+       class_destroy(umad_class);
 
 out_chrdev:
        unregister_chrdev_region(base_dev, IB_UMAD_MAX_PORTS * 2);
@@ -928,7 +1003,7 @@ out:
 static void __exit ib_umad_cleanup(void)
 {
        ib_unregister_client(&umad_client);
-       class_unregister(&umad_class);
+       class_destroy(umad_class);
        unregister_chrdev_region(base_dev, IB_UMAD_MAX_PORTS * 2);
 }