1 files changed, 310 insertions, 195 deletions

diff --git a/drivers/infiniband/core/uverbs_main.c b/drivers/infiniband/core/uverbs_main.c
index 7c2ac390558..08219fb3338 100644
--- a/drivers/infiniband/core/uverbs_main.c
+++ b/drivers/infiniband/core/uverbs_main.c
@@ -32,8 +32,6 @@
  * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
  * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
  * SOFTWARE.
- *
- * $Id: uverbs_main.c 2733 2005-06-28 19:14:34Z roland $
  */
 
 #include <linux/module.h>
@@ -42,9 +40,11 @@
 #include <linux/err.h>
 #include <linux/fs.h>
 #include <linux/poll.h>
+#include <linux/sched.h>
 #include <linux/file.h>
-#include <linux/mount.h>
 #include <linux/cdev.h>
+#include <linux/anon_inodes.h>
+#include <linux/slab.h>
 
 #include <asm/uaccess.h>
 
@@ -54,8 +54,6 @@ MODULE_AUTHOR("Roland Dreier");
 MODULE_DESCRIPTION("InfiniBand userspace verbs access");
 MODULE_LICENSE("Dual BSD/GPL");
 
-#define INFINIBANDEVENTFS_MAGIC	0x49426576	/* "IBev" */
-
 enum {
 	IB_UVERBS_MAJOR       = 231,
 	IB_UVERBS_BASE_MINOR  = 192,
@@ -74,45 +72,57 @@ DEFINE_IDR(ib_uverbs_ah_idr);
 DEFINE_IDR(ib_uverbs_cq_idr);
 DEFINE_IDR(ib_uverbs_qp_idr);
 DEFINE_IDR(ib_uverbs_srq_idr);
+DEFINE_IDR(ib_uverbs_xrcd_idr);
+DEFINE_IDR(ib_uverbs_rule_idr);
 
-static spinlock_t map_lock;
-static struct ib_uverbs_device *dev_table[IB_UVERBS_MAX_DEVICES];
+static DEFINE_SPINLOCK(map_lock);
 static DECLARE_BITMAP(dev_map, IB_UVERBS_MAX_DEVICES);
 
 static ssize_t (*uverbs_cmd_table[])(struct ib_uverbs_file *file,
 				     const char __user *buf, int in_len,
 				     int out_len) = {
-	[IB_USER_VERBS_CMD_GET_CONTEXT]   	= ib_uverbs_get_context,
-	[IB_USER_VERBS_CMD_QUERY_DEVICE]  	= ib_uverbs_query_device,
-	[IB_USER_VERBS_CMD_QUERY_PORT]    	= ib_uverbs_query_port,
-	[IB_USER_VERBS_CMD_ALLOC_PD]      	= ib_uverbs_alloc_pd,
-	[IB_USER_VERBS_CMD_DEALLOC_PD]    	= ib_uverbs_dealloc_pd,
-	[IB_USER_VERBS_CMD_REG_MR]        	= ib_uverbs_reg_mr,
-	[IB_USER_VERBS_CMD_DEREG_MR]      	= ib_uverbs_dereg_mr,
+	[IB_USER_VERBS_CMD_GET_CONTEXT]		= ib_uverbs_get_context,
+	[IB_USER_VERBS_CMD_QUERY_DEVICE]	= ib_uverbs_query_device,
+	[IB_USER_VERBS_CMD_QUERY_PORT]		= ib_uverbs_query_port,
+	[IB_USER_VERBS_CMD_ALLOC_PD]		= ib_uverbs_alloc_pd,
+	[IB_USER_VERBS_CMD_DEALLOC_PD]		= ib_uverbs_dealloc_pd,
+	[IB_USER_VERBS_CMD_REG_MR]		= ib_uverbs_reg_mr,
+	[IB_USER_VERBS_CMD_DEREG_MR]		= ib_uverbs_dereg_mr,
+	[IB_USER_VERBS_CMD_ALLOC_MW]		= ib_uverbs_alloc_mw,
+	[IB_USER_VERBS_CMD_DEALLOC_MW]		= ib_uverbs_dealloc_mw,
 	[IB_USER_VERBS_CMD_CREATE_COMP_CHANNEL] = ib_uverbs_create_comp_channel,
-	[IB_USER_VERBS_CMD_CREATE_CQ]     	= ib_uverbs_create_cq,
-	[IB_USER_VERBS_CMD_RESIZE_CQ]     	= ib_uverbs_resize_cq,
-	[IB_USER_VERBS_CMD_POLL_CQ]     	= ib_uverbs_poll_cq,
-	[IB_USER_VERBS_CMD_REQ_NOTIFY_CQ]     	= ib_uverbs_req_notify_cq,
-	[IB_USER_VERBS_CMD_DESTROY_CQ]    	= ib_uverbs_destroy_cq,
-	[IB_USER_VERBS_CMD_CREATE_QP]     	= ib_uverbs_create_qp,
-	[IB_USER_VERBS_CMD_QUERY_QP]     	= ib_uverbs_query_qp,
-	[IB_USER_VERBS_CMD_MODIFY_QP]     	= ib_uverbs_modify_qp,
-	[IB_USER_VERBS_CMD_DESTROY_QP]    	= ib_uverbs_destroy_qp,
-	[IB_USER_VERBS_CMD_POST_SEND]    	= ib_uverbs_post_send,
-	[IB_USER_VERBS_CMD_POST_RECV]    	= ib_uverbs_post_recv,
-	[IB_USER_VERBS_CMD_POST_SRQ_RECV]    	= ib_uverbs_post_srq_recv,
-	[IB_USER_VERBS_CMD_CREATE_AH]    	= ib_uverbs_create_ah,
-	[IB_USER_VERBS_CMD_DESTROY_AH]    	= ib_uverbs_destroy_ah,
-	[IB_USER_VERBS_CMD_ATTACH_MCAST]  	= ib_uverbs_attach_mcast,
-	[IB_USER_VERBS_CMD_DETACH_MCAST]  	= ib_uverbs_detach_mcast,
-	[IB_USER_VERBS_CMD_CREATE_SRQ]    	= ib_uverbs_create_srq,
-	[IB_USER_VERBS_CMD_MODIFY_SRQ]    	= ib_uverbs_modify_srq,
-	[IB_USER_VERBS_CMD_QUERY_SRQ]     	= ib_uverbs_query_srq,
-	[IB_USER_VERBS_CMD_DESTROY_SRQ]   	= ib_uverbs_destroy_srq,
+	[IB_USER_VERBS_CMD_CREATE_CQ]		= ib_uverbs_create_cq,
+	[IB_USER_VERBS_CMD_RESIZE_CQ]		= ib_uverbs_resize_cq,
+	[IB_USER_VERBS_CMD_POLL_CQ]		= ib_uverbs_poll_cq,
+	[IB_USER_VERBS_CMD_REQ_NOTIFY_CQ]	= ib_uverbs_req_notify_cq,
+	[IB_USER_VERBS_CMD_DESTROY_CQ]		= ib_uverbs_destroy_cq,
+	[IB_USER_VERBS_CMD_CREATE_QP]		= ib_uverbs_create_qp,
+	[IB_USER_VERBS_CMD_QUERY_QP]		= ib_uverbs_query_qp,
+	[IB_USER_VERBS_CMD_MODIFY_QP]		= ib_uverbs_modify_qp,
+	[IB_USER_VERBS_CMD_DESTROY_QP]		= ib_uverbs_destroy_qp,
+	[IB_USER_VERBS_CMD_POST_SEND]		= ib_uverbs_post_send,
+	[IB_USER_VERBS_CMD_POST_RECV]		= ib_uverbs_post_recv,
+	[IB_USER_VERBS_CMD_POST_SRQ_RECV]	= ib_uverbs_post_srq_recv,
+	[IB_USER_VERBS_CMD_CREATE_AH]		= ib_uverbs_create_ah,
+	[IB_USER_VERBS_CMD_DESTROY_AH]		= ib_uverbs_destroy_ah,
+	[IB_USER_VERBS_CMD_ATTACH_MCAST]	= ib_uverbs_attach_mcast,
+	[IB_USER_VERBS_CMD_DETACH_MCAST]	= ib_uverbs_detach_mcast,
+	[IB_USER_VERBS_CMD_CREATE_SRQ]		= ib_uverbs_create_srq,
+	[IB_USER_VERBS_CMD_MODIFY_SRQ]		= ib_uverbs_modify_srq,
+	[IB_USER_VERBS_CMD_QUERY_SRQ]		= ib_uverbs_query_srq,
+	[IB_USER_VERBS_CMD_DESTROY_SRQ]		= ib_uverbs_destroy_srq,
+	[IB_USER_VERBS_CMD_OPEN_XRCD]		= ib_uverbs_open_xrcd,
+	[IB_USER_VERBS_CMD_CLOSE_XRCD]		= ib_uverbs_close_xrcd,
+	[IB_USER_VERBS_CMD_CREATE_XSRQ]		= ib_uverbs_create_xsrq,
+	[IB_USER_VERBS_CMD_OPEN_QP]		= ib_uverbs_open_qp,
 };
 
-static struct vfsmount *uverbs_event_mnt;
+static int (*uverbs_ex_cmd_table[])(struct ib_uverbs_file *file,
+				    struct ib_udata *ucore,
+				    struct ib_udata *uhw) = {
+	[IB_USER_VERBS_EX_CMD_CREATE_FLOW]	= ib_uverbs_ex_create_flow,
+	[IB_USER_VERBS_EX_CMD_DESTROY_FLOW]	= ib_uverbs_ex_destroy_flow
+};
 
 static void ib_uverbs_add_one(struct ib_device *device);
 static void ib_uverbs_remove_one(struct ib_device *device);
@@ -201,14 +211,35 @@ static int ib_uverbs_cleanup_ucontext(struct ib_uverbs_file *file,
 		kfree(uobj);
 	}
 
+	/* Remove MWs before QPs, in order to support type 2A MWs. */
+	list_for_each_entry_safe(uobj, tmp, &context->mw_list, list) {
+		struct ib_mw *mw = uobj->object;
+
+		idr_remove_uobj(&ib_uverbs_mw_idr, uobj);
+		ib_dealloc_mw(mw);
+		kfree(uobj);
+	}
+
+	list_for_each_entry_safe(uobj, tmp, &context->rule_list, list) {
+		struct ib_flow *flow_id = uobj->object;
+
+		idr_remove_uobj(&ib_uverbs_rule_idr, uobj);
+		ib_destroy_flow(flow_id);
+		kfree(uobj);
+	}
+
 	list_for_each_entry_safe(uobj, tmp, &context->qp_list, list) {
 		struct ib_qp *qp = uobj->object;
 		struct ib_uqp_object *uqp =
 			container_of(uobj, struct ib_uqp_object, uevent.uobject);
 
 		idr_remove_uobj(&ib_uverbs_qp_idr, uobj);
-		ib_uverbs_detach_umcast(qp, uqp);
-		ib_destroy_qp(qp);
+		if (qp != qp->real_qp) {
+			ib_close_qp(qp);
+		} else {
+			ib_uverbs_detach_umcast(qp, uqp);
+			ib_destroy_qp(qp);
+		}
 		ib_uverbs_release_uevent(file, &uqp->uevent);
 		kfree(uqp);
 	}
@@ -236,8 +267,6 @@ static int ib_uverbs_cleanup_ucontext(struct ib_uverbs_file *file,
 		kfree(uevent);
 	}
 
-	/* XXX Free MWs */
-
 	list_for_each_entry_safe(uobj, tmp, &context->mr_list, list) {
 		struct ib_mr *mr = uobj->object;
 
@@ -246,6 +275,18 @@ static int ib_uverbs_cleanup_ucontext(struct ib_uverbs_file *file,
 		kfree(uobj);
 	}
 
+	mutex_lock(&file->device->xrcd_tree_mutex);
+	list_for_each_entry_safe(uobj, tmp, &context->xrcd_list, list) {
+		struct ib_xrcd *xrcd = uobj->object;
+		struct ib_uxrcd_object *uxrcd =
+			container_of(uobj, struct ib_uxrcd_object, uobject);
+
+		idr_remove_uobj(&ib_uverbs_xrcd_idr, uobj);
+		ib_uverbs_dealloc_xrcd(file->device, xrcd);
+		kfree(uxrcd);
+	}
+	mutex_unlock(&file->device->xrcd_tree_mutex);
+
 	list_for_each_entry_safe(uobj, tmp, &context->pd_list, list) {
 		struct ib_pd *pd = uobj->object;
 
@@ -352,7 +393,7 @@ static int ib_uverbs_event_close(struct inode *inode, struct file *filp)
 	struct ib_uverbs_event *entry, *tmp;
 
 	spin_lock_irq(&file->lock);
-	file->file = NULL;
+	file->is_closed = 1;
 	list_for_each_entry_safe(entry, tmp, &file->event_list, list) {
 		if (entry->counter)
 			list_del(&entry->obj_list);
@@ -360,8 +401,6 @@ static int ib_uverbs_event_close(struct inode *inode, struct file *filp)
 	}
 	spin_unlock_irq(&file->lock);
 
-	ib_uverbs_event_fasync(-1, filp, 0);
-
 	if (file->is_async) {
 		ib_unregister_event_handler(&file->uverbs_file->event_handler);
 		kref_put(&file->uverbs_file->ref, ib_uverbs_release_file);
@@ -373,10 +412,11 @@ static int ib_uverbs_event_close(struct inode *inode, struct file *filp)
 
 static const struct file_operations uverbs_event_fops = {
 	.owner	 = THIS_MODULE,
-	.read 	 = ib_uverbs_event_read,
+	.read	 = ib_uverbs_event_read,
 	.poll    = ib_uverbs_event_poll,
 	.release = ib_uverbs_event_close,
-	.fasync  = ib_uverbs_event_fasync
+	.fasync  = ib_uverbs_event_fasync,
+	.llseek	 = no_llseek,
 };
 
 void ib_uverbs_comp_handler(struct ib_cq *cq, void *cq_context)
@@ -390,7 +430,7 @@ void ib_uverbs_comp_handler(struct ib_cq *cq, void *cq_context)
 		return;
 
 	spin_lock_irqsave(&file->lock, flags);
-	if (!file->file) {
+	if (file->is_closed) {
 		spin_unlock_irqrestore(&file->lock, flags);
 		return;
 	}
@@ -423,7 +463,7 @@ static void ib_uverbs_async_handler(struct ib_uverbs_file *file,
 	unsigned long flags;
 
 	spin_lock_irqsave(&file->async_file->lock, flags);
-	if (!file->async_file->file) {
+	if (file->async_file->is_closed) {
 		spin_unlock_irqrestore(&file->async_file->lock, flags);
 		return;
 	}
@@ -492,11 +532,10 @@ void ib_uverbs_event_handler(struct ib_event_handler *handler,
 }
 
 struct file *ib_uverbs_alloc_event_file(struct ib_uverbs_file *uverbs_file,
-					int is_async, int *fd)
+					int is_async)
 {
 	struct ib_uverbs_event_file *ev_file;
 	struct file *filp;
-	int ret;
 
 	ev_file = kmalloc(sizeof *ev_file, GFP_KERNEL);
 	if (!ev_file)
@@ -509,42 +548,14 @@ struct file *ib_uverbs_alloc_event_file(struct ib_uverbs_file *uverbs_file,
 	ev_file->uverbs_file = uverbs_file;
 	ev_file->async_queue = NULL;
 	ev_file->is_async    = is_async;
+	ev_file->is_closed   = 0;
 
-	*fd = get_unused_fd();
-	if (*fd < 0) {
-		ret = *fd;
-		goto err;
-	}
-
-	filp = get_empty_filp();
-	if (!filp) {
-		ret = -ENFILE;
-		goto err_fd;
-	}
-
-	ev_file->file      = filp;
-
-	/*
-	 * fops_get() can't fail here, because we're coming from a
-	 * system call on a uverbs file, which will already have a
-	 * module reference.
-	 */
-	filp->f_op 	   = fops_get(&uverbs_event_fops);
-	filp->f_path.mnt 	   = mntget(uverbs_event_mnt);
-	filp->f_path.dentry 	   = dget(uverbs_event_mnt->mnt_root);
-	filp->f_mapping    = filp->f_path.dentry->d_inode->i_mapping;
-	filp->f_flags      = O_RDONLY;
-	filp->f_mode       = FMODE_READ;
-	filp->private_data = ev_file;
+	filp = anon_inode_getfile("[infinibandevent]", &uverbs_event_fops,
+				  ev_file, O_RDONLY);
+	if (IS_ERR(filp))
+		kfree(ev_file);
 
 	return filp;
-
-err_fd:
-	put_unused_fd(*fd);
-
-err:
-	kfree(ev_file);
-	return ERR_PTR(ret);
 }
 
 /*
@@ -555,16 +566,15 @@ err:
 struct ib_uverbs_event_file *ib_uverbs_lookup_comp_file(int fd)
 {
 	struct ib_uverbs_event_file *ev_file = NULL;
-	struct file *filp;
+	struct fd f = fdget(fd);
 
-	filp = fget(fd);
-	if (!filp)
+	if (!f.file)
 		return NULL;
 
-	if (filp->f_op != &uverbs_event_fops)
+	if (f.file->f_op != &uverbs_event_fops)
 		goto out;
 
-	ev_file = filp->private_data;
+	ev_file = f.file->private_data;
 	if (ev_file->is_async) {
 		ev_file = NULL;
 		goto out;
@@ -573,7 +583,7 @@ struct ib_uverbs_event_file *ib_uverbs_lookup_comp_file(int fd)
 	kref_get(&ev_file->ref);
 
 out:
-	fput(filp);
+	fdput(f);
 	return ev_file;
 }
 
@@ -582,6 +592,7 @@ static ssize_t ib_uverbs_write(struct file *filp, const char __user *buf,
 {
 	struct ib_uverbs_file *file = filp->private_data;
 	struct ib_uverbs_cmd_hdr hdr;
+	__u32 flags;
 
 	if (count < sizeof hdr)
 		return -EINVAL;
@@ -589,21 +600,110 @@ static ssize_t ib_uverbs_write(struct file *filp, const char __user *buf,
 	if (copy_from_user(&hdr, buf, sizeof hdr))
 		return -EFAULT;
 
-	if (hdr.in_words * 4 != count)
-		return -EINVAL;
+	flags = (hdr.command &
+		 IB_USER_VERBS_CMD_FLAGS_MASK) >> IB_USER_VERBS_CMD_FLAGS_SHIFT;
 
-	if (hdr.command < 0				||
-	    hdr.command >= ARRAY_SIZE(uverbs_cmd_table) ||
-	    !uverbs_cmd_table[hdr.command]		||
-	    !(file->device->ib_dev->uverbs_cmd_mask & (1ull << hdr.command)))
-		return -EINVAL;
+	if (!flags) {
+		__u32 command;
 
-	if (!file->ucontext &&
-	    hdr.command != IB_USER_VERBS_CMD_GET_CONTEXT)
-		return -EINVAL;
+		if (hdr.command & ~(__u32)(IB_USER_VERBS_CMD_FLAGS_MASK |
+					   IB_USER_VERBS_CMD_COMMAND_MASK))
+			return -EINVAL;
+
+		command = hdr.command & IB_USER_VERBS_CMD_COMMAND_MASK;
+
+		if (command >= ARRAY_SIZE(uverbs_cmd_table) ||
+		    !uverbs_cmd_table[command])
+			return -EINVAL;
+
+		if (!file->ucontext &&
+		    command != IB_USER_VERBS_CMD_GET_CONTEXT)
+			return -EINVAL;
+
+		if (!(file->device->ib_dev->uverbs_cmd_mask & (1ull << command)))
+			return -ENOSYS;
+
+		if (hdr.in_words * 4 != count)
+			return -EINVAL;
+
+		return uverbs_cmd_table[command](file,
+						 buf + sizeof(hdr),
+						 hdr.in_words * 4,
+						 hdr.out_words * 4);
+
+	} else if (flags == IB_USER_VERBS_CMD_FLAG_EXTENDED) {
+		__u32 command;
+
+		struct ib_uverbs_ex_cmd_hdr ex_hdr;
+		struct ib_udata ucore;
+		struct ib_udata uhw;
+		int err;
+		size_t written_count = count;
+
+		if (hdr.command & ~(__u32)(IB_USER_VERBS_CMD_FLAGS_MASK |
+					   IB_USER_VERBS_CMD_COMMAND_MASK))
+			return -EINVAL;
+
+		command = hdr.command & IB_USER_VERBS_CMD_COMMAND_MASK;
+
+		if (command >= ARRAY_SIZE(uverbs_ex_cmd_table) ||
+		    !uverbs_ex_cmd_table[command])
+			return -ENOSYS;
+
+		if (!file->ucontext)
+			return -EINVAL;
+
+		if (!(file->device->ib_dev->uverbs_ex_cmd_mask & (1ull << command)))
+			return -ENOSYS;
+
+		if (count < (sizeof(hdr) + sizeof(ex_hdr)))
+			return -EINVAL;
+
+		if (copy_from_user(&ex_hdr, buf + sizeof(hdr), sizeof(ex_hdr)))
+			return -EFAULT;
+
+		count -= sizeof(hdr) + sizeof(ex_hdr);
+		buf += sizeof(hdr) + sizeof(ex_hdr);
+
+		if ((hdr.in_words + ex_hdr.provider_in_words) * 8 != count)
+			return -EINVAL;
+
+		if (ex_hdr.cmd_hdr_reserved)
+			return -EINVAL;
+
+		if (ex_hdr.response) {
+			if (!hdr.out_words && !ex_hdr.provider_out_words)
+				return -EINVAL;
+
+			if (!access_ok(VERIFY_WRITE,
+				       (void __user *) (unsigned long) ex_hdr.response,
+				       (hdr.out_words + ex_hdr.provider_out_words) * 8))
+				return -EFAULT;
+		} else {
+			if (hdr.out_words || ex_hdr.provider_out_words)
+				return -EINVAL;
+		}
+
+		INIT_UDATA_BUF_OR_NULL(&ucore, buf, (unsigned long) ex_hdr.response,
+				       hdr.in_words * 8, hdr.out_words * 8);
 
-	return uverbs_cmd_table[hdr.command](file, buf + sizeof hdr,
-					     hdr.in_words * 4, hdr.out_words * 4);
+		INIT_UDATA_BUF_OR_NULL(&uhw,
+				       buf + ucore.inlen,
+				       (unsigned long) ex_hdr.response + ucore.outlen,
+				       ex_hdr.provider_in_words * 8,
+				       ex_hdr.provider_out_words * 8);
+
+		err = uverbs_ex_cmd_table[command](file,
+						   &ucore,
+						   &uhw);
+
+		if (err)
+			return err;
+
+		return written_count;
+	}
+
+	return -ENOSYS;
 }
 
 static int ib_uverbs_mmap(struct file *filp, struct vm_area_struct *vma)
@@ -616,19 +716,26 @@ static int ib_uverbs_mmap(struct file *filp, struct vm_area_struct *vma)
 		return file->device->ib_dev->mmap(file->ucontext, vma);
 }
 
+/*
+ * ib_uverbs_open() does not need the BKL:
+ *
+ *  - the ib_uverbs_device structures are properly reference counted and
+ *    everything else is purely local to the file being created, so
+ *    races against other open calls are not a problem;
+ *  - there is no ioctl method to race against;
+ *  - the open method will either immediately run -ENXIO, or all
+ *    required initialization will be done.
+ */
 static int ib_uverbs_open(struct inode *inode, struct file *filp)
 {
 	struct ib_uverbs_device *dev;
 	struct ib_uverbs_file *file;
 	int ret;
 
-	spin_lock(&map_lock);
-	dev = dev_table[iminor(inode) - IB_UVERBS_BASE_MINOR];
+	dev = container_of(inode->i_cdev, struct ib_uverbs_device, cdev);
 	if (dev)
 		kref_get(&dev->ref);
-	spin_unlock(&map_lock);
-
-	if (!dev)
+	else
 		return -ENXIO;
 
 	if (!try_module_get(dev->ib_dev->owner)) {
@@ -650,14 +757,13 @@ static int ib_uverbs_open(struct inode *inode, struct file *filp)
 
 	filp->private_data = file;
 
-	return 0;
+	return nonseekable_open(inode, filp);
 
 err_module:
 	module_put(dev->ib_dev->owner);
 
 err:
 	kref_put(&dev->ref, ib_uverbs_release_dev);
-
 	return ret;
 }
 
@@ -676,18 +782,20 @@ static int ib_uverbs_close(struct inode *inode, struct file *filp)
 }
 
 static const struct file_operations uverbs_fops = {
-	.owner 	 = THIS_MODULE,
-	.write 	 = ib_uverbs_write,
-	.open 	 = ib_uverbs_open,
-	.release = ib_uverbs_close
+	.owner	 = THIS_MODULE,
+	.write	 = ib_uverbs_write,
+	.open	 = ib_uverbs_open,
+	.release = ib_uverbs_close,
+	.llseek	 = no_llseek,
 };
 
 static const struct file_operations uverbs_mmap_fops = {
-	.owner 	 = THIS_MODULE,
-	.write 	 = ib_uverbs_write,
+	.owner	 = THIS_MODULE,
+	.write	 = ib_uverbs_write,
 	.mmap    = ib_uverbs_mmap,
-	.open 	 = ib_uverbs_open,
-	.release = ib_uverbs_close
+	.open	 = ib_uverbs_open,
+	.release = ib_uverbs_close,
+	.llseek	 = no_llseek,
 };
 
 static struct ib_client uverbs_client = {
@@ -696,36 +804,65 @@ static struct ib_client uverbs_client = {
 	.remove = ib_uverbs_remove_one
 };
 
-static ssize_t show_ibdev(struct class_device *class_dev, char *buf)
+static ssize_t show_ibdev(struct device *device, struct device_attribute *attr,
+			  char *buf)
 {
-	struct ib_uverbs_device *dev = class_get_devdata(class_dev);
+	struct ib_uverbs_device *dev = dev_get_drvdata(device);
 
 	if (!dev)
 		return -ENODEV;
 
 	return sprintf(buf, "%s\n", dev->ib_dev->name);
 }
-static CLASS_DEVICE_ATTR(ibdev, S_IRUGO, show_ibdev, NULL);
+static DEVICE_ATTR(ibdev, S_IRUGO, show_ibdev, NULL);
 
-static ssize_t show_dev_abi_version(struct class_device *class_dev, char *buf)
+static ssize_t show_dev_abi_version(struct device *device,
+				    struct device_attribute *attr, char *buf)
 {
-	struct ib_uverbs_device *dev = class_get_devdata(class_dev);
+	struct ib_uverbs_device *dev = dev_get_drvdata(device);
 
 	if (!dev)
 		return -ENODEV;
 
 	return sprintf(buf, "%d\n", dev->ib_dev->uverbs_abi_ver);
 }
-static CLASS_DEVICE_ATTR(abi_version, S_IRUGO, show_dev_abi_version, NULL);
+static DEVICE_ATTR(abi_version, S_IRUGO, show_dev_abi_version, NULL);
+
+static CLASS_ATTR_STRING(abi_version, S_IRUGO,
+			 __stringify(IB_USER_VERBS_ABI_VERSION));
 
-static ssize_t show_abi_version(struct class *class, char *buf)
+static dev_t overflow_maj;
+static DECLARE_BITMAP(overflow_map, IB_UVERBS_MAX_DEVICES);
+
+/*
+ * If we have more than IB_UVERBS_MAX_DEVICES, dynamically overflow by
+ * requesting a new major number and doubling the number of max devices we
+ * support. It's stupid, but simple.
+ */
+static int find_overflow_devnum(void)
 {
-	return sprintf(buf, "%d\n", IB_USER_VERBS_ABI_VERSION);
+	int ret;
+
+	if (!overflow_maj) {
+		ret = alloc_chrdev_region(&overflow_maj, 0, IB_UVERBS_MAX_DEVICES,
+					  "infiniband_verbs");
+		if (ret) {
+			printk(KERN_ERR "user_verbs: couldn't register dynamic device number\n");
+			return ret;
+		}
+	}
+
+	ret = find_first_zero_bit(overflow_map, IB_UVERBS_MAX_DEVICES);
+	if (ret >= IB_UVERBS_MAX_DEVICES)
+		return -1;
+
+	return ret;
 }
-static CLASS_ATTR(abi_version, S_IRUGO, show_abi_version, NULL);
 
 static void ib_uverbs_add_one(struct ib_device *device)
 {
+	int devnum;
+	dev_t base;
 	struct ib_uverbs_device *uverbs_dev;
 
 	if (!device->alloc_ucontext)
@@ -737,56 +874,62 @@ static void ib_uverbs_add_one(struct ib_device *device)
 
 	kref_init(&uverbs_dev->ref);
 	init_completion(&uverbs_dev->comp);
+	uverbs_dev->xrcd_tree = RB_ROOT;
+	mutex_init(&uverbs_dev->xrcd_tree_mutex);
 
 	spin_lock(&map_lock);
-	uverbs_dev->devnum = find_first_zero_bit(dev_map, IB_UVERBS_MAX_DEVICES);
-	if (uverbs_dev->devnum >= IB_UVERBS_MAX_DEVICES) {
+	devnum = find_first_zero_bit(dev_map, IB_UVERBS_MAX_DEVICES);
+	if (devnum >= IB_UVERBS_MAX_DEVICES) {
 		spin_unlock(&map_lock);
-		goto err;
+		devnum = find_overflow_devnum();
+		if (devnum < 0)
+			goto err;
+
+		spin_lock(&map_lock);
+		uverbs_dev->devnum = devnum + IB_UVERBS_MAX_DEVICES;
+		base = devnum + overflow_maj;
+		set_bit(devnum, overflow_map);
+	} else {
+		uverbs_dev->devnum = devnum;
+		base = devnum + IB_UVERBS_BASE_DEV;
+		set_bit(devnum, dev_map);
 	}
-	set_bit(uverbs_dev->devnum, dev_map);
 	spin_unlock(&map_lock);
 
 	uverbs_dev->ib_dev           = device;
 	uverbs_dev->num_comp_vectors = device->num_comp_vectors;
 
-	uverbs_dev->dev = cdev_alloc();
-	if (!uverbs_dev->dev)
-		goto err;
-	uverbs_dev->dev->owner = THIS_MODULE;
-	uverbs_dev->dev->ops = device->mmap ? &uverbs_mmap_fops : &uverbs_fops;
-	kobject_set_name(&uverbs_dev->dev->kobj, "uverbs%d", uverbs_dev->devnum);
-	if (cdev_add(uverbs_dev->dev, IB_UVERBS_BASE_DEV + uverbs_dev->devnum, 1))
+	cdev_init(&uverbs_dev->cdev, NULL);
+	uverbs_dev->cdev.owner = THIS_MODULE;
+	uverbs_dev->cdev.ops = device->mmap ? &uverbs_mmap_fops : &uverbs_fops;
+	kobject_set_name(&uverbs_dev->cdev.kobj, "uverbs%d", uverbs_dev->devnum);
+	if (cdev_add(&uverbs_dev->cdev, base, 1))
 		goto err_cdev;
 
-	uverbs_dev->class_dev = class_device_create(uverbs_class, NULL,
-						    uverbs_dev->dev->dev,
-						    device->dma_device,
-						    "uverbs%d", uverbs_dev->devnum);
-	if (IS_ERR(uverbs_dev->class_dev))
+	uverbs_dev->dev = device_create(uverbs_class, device->dma_device,
+					uverbs_dev->cdev.dev, uverbs_dev,
+					"uverbs%d", uverbs_dev->devnum);
+	if (IS_ERR(uverbs_dev->dev))
 		goto err_cdev;
 
-	class_set_devdata(uverbs_dev->class_dev, uverbs_dev);
-
-	if (class_device_create_file(uverbs_dev->class_dev, &class_device_attr_ibdev))
+	if (device_create_file(uverbs_dev->dev, &dev_attr_ibdev))
 		goto err_class;
-	if (class_device_create_file(uverbs_dev->class_dev, &class_device_attr_abi_version))
+	if (device_create_file(uverbs_dev->dev, &dev_attr_abi_version))
 		goto err_class;
 
-	spin_lock(&map_lock);
-	dev_table[uverbs_dev->devnum] = uverbs_dev;
-	spin_unlock(&map_lock);
-
 	ib_set_client_data(device, &uverbs_client, uverbs_dev);
 
 	return;
 
 err_class:
-	class_device_destroy(uverbs_class, uverbs_dev->dev->dev);
+	device_destroy(uverbs_class, uverbs_dev->cdev.dev);
 
 err_cdev:
-	cdev_del(uverbs_dev->dev);
-	clear_bit(uverbs_dev->devnum, dev_map);
+	cdev_del(&uverbs_dev->cdev);
+	if (uverbs_dev->devnum < IB_UVERBS_MAX_DEVICES)
+		clear_bit(devnum, dev_map);
+	else
+		clear_bit(devnum, overflow_map);
 
 err:
 	kref_put(&uverbs_dev->ref, ib_uverbs_release_dev);
@@ -802,42 +945,31 @@ static void ib_uverbs_remove_one(struct ib_device *device)
 	if (!uverbs_dev)
 		return;
 
-	class_set_devdata(uverbs_dev->class_dev, NULL);
-	class_device_destroy(uverbs_class, uverbs_dev->dev->dev);
-	cdev_del(uverbs_dev->dev);
-
-	spin_lock(&map_lock);
-	dev_table[uverbs_dev->devnum] = NULL;
-	spin_unlock(&map_lock);
+	dev_set_drvdata(uverbs_dev->dev, NULL);
+	device_destroy(uverbs_class, uverbs_dev->cdev.dev);
+	cdev_del(&uverbs_dev->cdev);
 
-	clear_bit(uverbs_dev->devnum, dev_map);
+	if (uverbs_dev->devnum < IB_UVERBS_MAX_DEVICES)
+		clear_bit(uverbs_dev->devnum, dev_map);
+	else
+		clear_bit(uverbs_dev->devnum - IB_UVERBS_MAX_DEVICES, overflow_map);
 
 	kref_put(&uverbs_dev->ref, ib_uverbs_release_dev);
 	wait_for_completion(&uverbs_dev->comp);
 	kfree(uverbs_dev);
 }
 
-static int uverbs_event_get_sb(struct file_system_type *fs_type, int flags,
-			       const char *dev_name, void *data,
-			       struct vfsmount *mnt)
+static char *uverbs_devnode(struct device *dev, umode_t *mode)
 {
-	return get_sb_pseudo(fs_type, "infinibandevent:", NULL,
-			     INFINIBANDEVENTFS_MAGIC, mnt);
+	if (mode)
+		*mode = 0666;
+	return kasprintf(GFP_KERNEL, "infiniband/%s", dev_name(dev));
 }
 
-static struct file_system_type uverbs_event_fs = {
-	/* No owner field so module can be unloaded */
-	.name    = "infinibandeventfs",
-	.get_sb  = uverbs_event_get_sb,
-	.kill_sb = kill_litter_super
-};
-
 static int __init ib_uverbs_init(void)
 {
 	int ret;
 
-	spin_lock_init(&map_lock);
-
 	ret = register_chrdev_region(IB_UVERBS_BASE_DEV, IB_UVERBS_MAX_DEVICES,
 				     "infiniband_verbs");
 	if (ret) {
@@ -852,39 +984,22 @@ static int __init ib_uverbs_init(void)
 		goto out_chrdev;
 	}
 
-	ret = class_create_file(uverbs_class, &class_attr_abi_version);
-	if (ret) {
-		printk(KERN_ERR "user_verbs: couldn't create abi_version attribute\n");
-		goto out_class;
-	}
+	uverbs_class->devnode = uverbs_devnode;
 
-	ret = register_filesystem(&uverbs_event_fs);
+	ret = class_create_file(uverbs_class, &class_attr_abi_version.attr);
 	if (ret) {
-		printk(KERN_ERR "user_verbs: couldn't register infinibandeventfs\n");
+		printk(KERN_ERR "user_verbs: couldn't create abi_version attribute\n");
 		goto out_class;
 	}
 
-	uverbs_event_mnt = kern_mount(&uverbs_event_fs);
-	if (IS_ERR(uverbs_event_mnt)) {
-		ret = PTR_ERR(uverbs_event_mnt);
-		printk(KERN_ERR "user_verbs: couldn't mount infinibandeventfs\n");
-		goto out_fs;
-	}
-
 	ret = ib_register_client(&uverbs_client);
 	if (ret) {
 		printk(KERN_ERR "user_verbs: couldn't register client\n");
-		goto out_mnt;
+		goto out_class;
 	}
 
 	return 0;
 
-out_mnt:
-	mntput(uverbs_event_mnt);
-
-out_fs:
-	unregister_filesystem(&uverbs_event_fs);
-
 out_class:
 	class_destroy(uverbs_class);
 
@@ -898,10 +1013,10 @@ out:
 static void __exit ib_uverbs_cleanup(void)
 {
 	ib_unregister_client(&uverbs_client);
-	mntput(uverbs_event_mnt);
-	unregister_filesystem(&uverbs_event_fs);
 	class_destroy(uverbs_class);
 	unregister_chrdev_region(IB_UVERBS_BASE_DEV, IB_UVERBS_MAX_DEVICES);
+	if (overflow_maj)
+		unregister_chrdev_region(overflow_maj, IB_UVERBS_MAX_DEVICES);
 	idr_destroy(&ib_uverbs_pd_idr);
 	idr_destroy(&ib_uverbs_mr_idr);
 	idr_destroy(&ib_uverbs_mw_idr);