[NET]: Add netlink connector.

Kernel connector - new userspace <-> kernel space easy to use
communication module which implements easy to use bidirectional
message bus using netlink as it's backend.  Connector was created to
eliminate complex skb handling both in send and receive message bus
direction.

Connector driver adds possibility to connect various agents using as
one of it's backends netlink based network.  One must register
callback and identifier. When driver receives special netlink message
with appropriate identifier, appropriate callback will be called.

From the userspace point of view it's quite straightforward:

	socket();
	bind();
	send();
	recv();

But if kernelspace want to use full power of such connections, driver
writer must create special sockets, must know about struct sk_buff
handling...  Connector allows any kernelspace agents to use netlink
based networking for inter-process communication in a significantly
easier way:

int cn_add_callback(struct cb_id *id, char *name, void (*callback) (void *));
void cn_netlink_send(struct cn_msg *msg, u32 __groups, int gfp_mask);

struct cb_id
{
	__u32			idx;
	__u32			val;
};

idx and val are unique identifiers which must be registered in
connector.h for in-kernel usage.  void (*callback) (void *) - is a
callback function which will be called when message with above idx.val
will be received by connector core.

Using connector completely hides low-level transport layer from it's
users.

Connector uses new netlink ability to have many groups in one socket.

[ Incorporating many cleanups and fixes by myself and
  Andrew Morton -DaveM ]

Signed-off-by: Evgeniy Polyakov <johnpol@2ka.mipt.ru>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

4 files changed, 675 insertions, 0 deletions

diff --git a/drivers/connector/Kconfig b/drivers/connector/Kconfig
new file mode 100644
index 00000000000..0bc2059c1e0
--- /dev/null
+++ b/drivers/connector/Kconfig
@@ -0,0 +1,13 @@
+menu "Connector - unified userspace <-> kernelspace linker"
+
+config CONNECTOR
+	tristate "Connector - unified userspace <-> kernelspace linker"
+	depends on NET
+	---help---
+	  This is unified userspace <-> kernelspace connector working on top
+	  of the netlink socket protocol.
+
+	  Connector support can also be built as a module.  If so, the module
+	  will be called cn.ko.
+
+endmenu
diff --git a/drivers/connector/Makefile b/drivers/connector/Makefile
new file mode 100644
index 00000000000..12ca79e8234
--- /dev/null
+++ b/drivers/connector/Makefile
@@ -0,0 +1,3 @@
+obj-$(CONFIG_CONNECTOR)		+= cn.o
+
+cn-y				+= cn_queue.o connector.o
diff --git a/drivers/connector/cn_queue.c b/drivers/connector/cn_queue.c
new file mode 100644
index 00000000000..966632182e2
--- /dev/null
+++ b/drivers/connector/cn_queue.c
@@ -0,0 +1,173 @@
+/*
+ * 	cn_queue.c
+ * 
+ * 2004-2005 Copyright (c) Evgeniy Polyakov <johnpol@2ka.mipt.ru>
+ * All rights reserved.
+ * 
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ *
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/list.h>
+#include <linux/workqueue.h>
+#include <linux/spinlock.h>
+#include <linux/slab.h>
+#include <linux/skbuff.h>
+#include <linux/suspend.h>
+#include <linux/connector.h>
+#include <linux/delay.h>
+
+static void cn_queue_wrapper(void *data)
+{
+	struct cn_callback_entry *cbq = data;
+
+	cbq->cb->callback(cbq->cb->priv);
+	cbq->destruct_data(cbq->ddata);
+	cbq->ddata = NULL;
+}
+
+static struct cn_callback_entry *cn_queue_alloc_callback_entry(struct cn_callback *cb)
+{
+	struct cn_callback_entry *cbq;
+
+	cbq = kzalloc(sizeof(*cbq), GFP_KERNEL);
+	if (!cbq) {
+		printk(KERN_ERR "Failed to create new callback queue.\n");
+		return NULL;
+	}
+
+	cbq->cb = cb;
+	INIT_WORK(&cbq->work, &cn_queue_wrapper, cbq);
+	return cbq;
+}
+
+static void cn_queue_free_callback(struct cn_callback_entry *cbq)
+{
+	cancel_delayed_work(&cbq->work);
+	flush_workqueue(cbq->pdev->cn_queue);
+
+	kfree(cbq);
+}
+
+int cn_cb_equal(struct cb_id *i1, struct cb_id *i2)
+{
+	return ((i1->idx == i2->idx) && (i1->val == i2->val));
+}
+
+int cn_queue_add_callback(struct cn_queue_dev *dev, struct cn_callback *cb)
+{
+	struct cn_callback_entry *cbq, *__cbq;
+	int found = 0;
+
+	cbq = cn_queue_alloc_callback_entry(cb);
+	if (!cbq)
+		return -ENOMEM;
+
+	atomic_inc(&dev->refcnt);
+	cbq->pdev = dev;
+
+	spin_lock_bh(&dev->queue_lock);
+	list_for_each_entry(__cbq, &dev->queue_list, callback_entry) {
+		if (cn_cb_equal(&__cbq->cb->id, &cb->id)) {
+			found = 1;
+			break;
+		}
+	}
+	if (!found)
+		list_add_tail(&cbq->callback_entry, &dev->queue_list);
+	spin_unlock_bh(&dev->queue_lock);
+
+	if (found) {
+		atomic_dec(&dev->refcnt);
+		cn_queue_free_callback(cbq);
+		return -EINVAL;
+	}
+
+	cbq->nls = dev->nls;
+	cbq->seq = 0;
+	cbq->group = cbq->cb->id.idx;
+
+	return 0;
+}
+
+void cn_queue_del_callback(struct cn_queue_dev *dev, struct cb_id *id)
+{
+	struct cn_callback_entry *cbq, *n;
+	int found = 0;
+
+	spin_lock_bh(&dev->queue_lock);
+	list_for_each_entry_safe(cbq, n, &dev->queue_list, callback_entry) {
+		if (cn_cb_equal(&cbq->cb->id, id)) {
+			list_del(&cbq->callback_entry);
+			found = 1;
+			break;
+		}
+	}
+	spin_unlock_bh(&dev->queue_lock);
+
+	if (found) {
+		cn_queue_free_callback(cbq);
+		atomic_dec_and_test(&dev->refcnt);
+	}
+}
+
+struct cn_queue_dev *cn_queue_alloc_dev(char *name, struct sock *nls)
+{
+	struct cn_queue_dev *dev;
+
+	dev = kzalloc(sizeof(*dev), GFP_KERNEL);
+	if (!dev)
+		return NULL;
+
+	snprintf(dev->name, sizeof(dev->name), "%s", name);
+	atomic_set(&dev->refcnt, 0);
+	INIT_LIST_HEAD(&dev->queue_list);
+	spin_lock_init(&dev->queue_lock);
+
+	dev->nls = nls;
+	dev->netlink_groups = 0;
+
+	dev->cn_queue = create_workqueue(dev->name);
+	if (!dev->cn_queue) {
+		kfree(dev);
+		return NULL;
+	}
+
+	return dev;
+}
+
+void cn_queue_free_dev(struct cn_queue_dev *dev)
+{
+	struct cn_callback_entry *cbq, *n;
+
+	flush_workqueue(dev->cn_queue);
+	destroy_workqueue(dev->cn_queue);
+
+	spin_lock_bh(&dev->queue_lock);
+	list_for_each_entry_safe(cbq, n, &dev->queue_list, callback_entry)
+		list_del(&cbq->callback_entry);
+	spin_unlock_bh(&dev->queue_lock);
+
+	while (atomic_read(&dev->refcnt)) {
+		printk(KERN_INFO "Waiting for %s to become free: refcnt=%d.\n",
+		       dev->name, atomic_read(&dev->refcnt));
+		msleep(1000);
+	}
+
+	kfree(dev);
+	dev = NULL;
+}
diff --git a/drivers/connector/connector.c b/drivers/connector/connector.c
new file mode 100644
index 00000000000..aaf6d468a8b
--- /dev/null
+++ b/drivers/connector/connector.c
@@ -0,0 +1,486 @@
+/*
+ * 	connector.c
+ * 
+ * 2004-2005 Copyright (c) Evgeniy Polyakov <johnpol@2ka.mipt.ru>
+ * All rights reserved.
+ * 
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/list.h>
+#include <linux/skbuff.h>
+#include <linux/netlink.h>
+#include <linux/moduleparam.h>
+#include <linux/connector.h>
+
+#include <net/sock.h>
+
+MODULE_LICENSE("GPL");
+MODULE_AUTHOR("Evgeniy Polyakov <johnpol@2ka.mipt.ru>");
+MODULE_DESCRIPTION("Generic userspace <-> kernelspace connector.");
+
+static u32 cn_idx = CN_IDX_CONNECTOR;
+static u32 cn_val = CN_VAL_CONNECTOR;
+
+module_param(cn_idx, uint, 0);
+module_param(cn_val, uint, 0);
+MODULE_PARM_DESC(cn_idx, "Connector's main device idx.");
+MODULE_PARM_DESC(cn_val, "Connector's main device val.");
+
+static DECLARE_MUTEX(notify_lock);
+static LIST_HEAD(notify_list);
+
+static struct cn_dev cdev;
+
+int cn_already_initialized = 0;
+
+/*
+ * msg->seq and msg->ack are used to determine message genealogy.
+ * When someone sends message it puts there locally unique sequence
+ * and random acknowledge numbers.  Sequence number may be copied into
+ * nlmsghdr->nlmsg_seq too.
+ *
+ * Sequence number is incremented with each message to be sent.
+ *
+ * If we expect reply to our message then the sequence number in
+ * received message MUST be the same as in original message, and
+ * acknowledge number MUST be the same + 1.
+ *
+ * If we receive a message and its sequence number is not equal to the
+ * one we are expecting then it is a new message.
+ *
+ * If we receive a message and its sequence number is the same as one
+ * we are expecting but it's acknowledgement number is not equal to
+ * the acknowledgement number in the original message + 1, then it is
+ * a new message.
+ *
+ */
+int cn_netlink_send(struct cn_msg *msg, u32 __group, int gfp_mask)
+{
+	struct cn_callback_entry *__cbq;
+	unsigned int size;
+	struct sk_buff *skb;
+	struct nlmsghdr *nlh;
+	struct cn_msg *data;
+	struct cn_dev *dev = &cdev;
+	u32 group = 0;
+	int found = 0;
+
+	if (!__group) {
+		spin_lock_bh(&dev->cbdev->queue_lock);
+		list_for_each_entry(__cbq, &dev->cbdev->queue_list,
+				    callback_entry) {
+			if (cn_cb_equal(&__cbq->cb->id, &msg->id)) {
+				found = 1;
+				group = __cbq->group;
+			}
+		}
+		spin_unlock_bh(&dev->cbdev->queue_lock);
+
+		if (!found)
+			return -ENODEV;
+	} else {
+		group = __group;
+	}
+
+	size = NLMSG_SPACE(sizeof(*msg) + msg->len);
+
+	skb = alloc_skb(size, gfp_mask);
+	if (!skb)
+		return -ENOMEM;
+
+	nlh = NLMSG_PUT(skb, 0, msg->seq, NLMSG_DONE, size - sizeof(*nlh));
+
+	data = NLMSG_DATA(nlh);
+
+	memcpy(data, msg, sizeof(*data) + msg->len);
+
+	NETLINK_CB(skb).dst_group = group;
+
+	netlink_broadcast(dev->nls, skb, 0, group, gfp_mask);
+
+	return 0;
+
+nlmsg_failure:
+	kfree_skb(skb);
+	return -EINVAL;
+}
+
+/*
+ * Callback helper - queues work and setup destructor for given data.
+ */
+static int cn_call_callback(struct cn_msg *msg, void (*destruct_data)(void *), void *data)
+{
+	struct cn_callback_entry *__cbq;
+	struct cn_dev *dev = &cdev;
+	int found = 0;
+
+	spin_lock_bh(&dev->cbdev->queue_lock);
+	list_for_each_entry(__cbq, &dev->cbdev->queue_list, callback_entry) {
+		if (cn_cb_equal(&__cbq->cb->id, &msg->id)) {
+			/*
+			 * Let's scream if there is some magic and the
+			 * data will arrive asynchronously here.
+			 * [i.e. netlink messages will be queued].
+			 * After the first warning I will fix it
+			 * quickly, but now I think it is
+			 * impossible. --zbr (2004_04_27).
+			 */
+			if (likely(!test_bit(0, &__cbq->work.pending) &&
+					__cbq->ddata == NULL)) {
+				__cbq->cb->priv = msg;
+
+				__cbq->ddata = data;
+				__cbq->destruct_data = destruct_data;
+
+				if (queue_work(dev->cbdev->cn_queue,
+						&__cbq->work))
+					found = 1;
+			} else {
+				printk("%s: cbq->data=%p, "
+				       "work->pending=%08lx.\n",
+				       __func__, __cbq->ddata,
+				       __cbq->work.pending);
+				WARN_ON(1);
+			}
+			break;
+		}
+	}
+	spin_unlock_bh(&dev->cbdev->queue_lock);
+
+	return found ? 0 : -ENODEV;
+}
+
+/*
+ * Skb receive helper - checks skb and msg size and calls callback
+ * helper.
+ */
+static int __cn_rx_skb(struct sk_buff *skb, struct nlmsghdr *nlh)
+{
+	u32 pid, uid, seq, group;
+	struct cn_msg *msg;
+
+	pid = NETLINK_CREDS(skb)->pid;
+	uid = NETLINK_CREDS(skb)->uid;
+	seq = nlh->nlmsg_seq;
+	group = NETLINK_CB((skb)).dst_group;
+	msg = NLMSG_DATA(nlh);
+
+	return cn_call_callback(msg, (void (*)(void *))kfree_skb, skb);
+}
+
+/*
+ * Main netlink receiving function.
+ *
+ * It checks skb and netlink header sizes and calls the skb receive
+ * helper with a shared skb.
+ */
+static void cn_rx_skb(struct sk_buff *__skb)
+{
+	struct nlmsghdr *nlh;
+	u32 len;
+	int err;
+	struct sk_buff *skb;
+
+	skb = skb_get(__skb);
+
+	if (skb->len >= NLMSG_SPACE(0)) {
+		nlh = (struct nlmsghdr *)skb->data;
+
+		if (nlh->nlmsg_len < sizeof(struct cn_msg) ||
+		    skb->len < nlh->nlmsg_len ||
+		    nlh->nlmsg_len > CONNECTOR_MAX_MSG_SIZE) {
+			kfree_skb(skb);
+			goto out;
+		}
+
+		len = NLMSG_ALIGN(nlh->nlmsg_len);
+		if (len > skb->len)
+			len = skb->len;
+
+		err = __cn_rx_skb(skb, nlh);
+		if (err < 0)
+			kfree_skb(skb);
+	}
+
+out:
+	kfree_skb(__skb);
+}
+
+/*
+ * Netlink socket input callback - dequeues the skbs and calls the
+ * main netlink receiving function.
+ */
+static void cn_input(struct sock *sk, int len)
+{
+	struct sk_buff *skb;
+
+	while ((skb = skb_dequeue(&sk->sk_receive_queue)) != NULL)
+		cn_rx_skb(skb);
+}
+
+/*
+ * Notification routing.
+ *
+ * Gets id and checks if there are notification request for it's idx
+ * and val.  If there are such requests notify the listeners with the
+ * given notify event.
+ *
+ */
+static void cn_notify(struct cb_id *id, u32 notify_event)
+{
+	struct cn_ctl_entry *ent;
+
+	down(&notify_lock);
+	list_for_each_entry(ent, &notify_list, notify_entry) {
+		int i;
+		struct cn_notify_req *req;
+		struct cn_ctl_msg *ctl = ent->msg;
+		int idx_found, val_found;
+
+		idx_found = val_found = 0;
+
+		req = (struct cn_notify_req *)ctl->data;
+		for (i = 0; i < ctl->idx_notify_num; ++i, ++req) {
+			if (id->idx >= req->first && 
+					id->idx < req->first + req->range) {
+				idx_found = 1;
+				break;
+			}
+		}
+
+		for (i = 0; i < ctl->val_notify_num; ++i, ++req) {
+			if (id->val >= req->first && 
+					id->val < req->first + req->range) {
+				val_found = 1;
+				break;
+			}
+		}
+
+		if (idx_found && val_found) {
+			struct cn_msg m = { .ack = notify_event, };
+
+			memcpy(&m.id, id, sizeof(m.id));
+			cn_netlink_send(&m, ctl->group, GFP_KERNEL);
+		}
+	}
+	up(&notify_lock);
+}
+
+/*
+ * Callback add routing - adds callback with given ID and name.
+ * If there is registered callback with the same ID it will not be added.
+ *
+ * May sleep.
+ */
+int cn_add_callback(struct cb_id *id, char *name, void (*callback)(void *))
+{
+	int err;
+	struct cn_dev *dev = &cdev;
+	struct cn_callback *cb;
+
+	cb = kzalloc(sizeof(*cb), GFP_KERNEL);
+	if (!cb)
+		return -ENOMEM;
+
+	scnprintf(cb->name, sizeof(cb->name), "%s", name);
+
+	memcpy(&cb->id, id, sizeof(cb->id));
+	cb->callback = callback;
+
+	err = cn_queue_add_callback(dev->cbdev, cb);
+	if (err) {
+		kfree(cb);
+		return err;
+	}
+
+	cn_notify(id, 0);
+
+	return 0;
+}
+
+/*
+ * Callback remove routing - removes callback
+ * with given ID.
+ * If there is no registered callback with given
+ * ID nothing happens.
+ *
+ * May sleep while waiting for reference counter to become zero.
+ */
+void cn_del_callback(struct cb_id *id)
+{
+	struct cn_dev *dev = &cdev;
+
+	cn_queue_del_callback(dev->cbdev, id);
+	cn_notify(id, 1);
+}
+
+/*
+ * Checks two connector's control messages to be the same.
+ * Returns 1 if they are the same or if the first one is corrupted.
+ */
+static int cn_ctl_msg_equals(struct cn_ctl_msg *m1, struct cn_ctl_msg *m2)
+{
+	int i;
+	struct cn_notify_req *req1, *req2;
+
+	if (m1->idx_notify_num != m2->idx_notify_num)
+		return 0;
+
+	if (m1->val_notify_num != m2->val_notify_num)
+		return 0;
+
+	if (m1->len != m2->len)
+		return 0;
+
+	if ((m1->idx_notify_num + m1->val_notify_num) * sizeof(*req1) !=
+	    m1->len)
+		return 1;
+
+	req1 = (struct cn_notify_req *)m1->data;
+	req2 = (struct cn_notify_req *)m2->data;
+
+	for (i = 0; i < m1->idx_notify_num; ++i) {
+		if (req1->first != req2->first || req1->range != req2->range)
+			return 0;
+		req1++;
+		req2++;
+	}
+
+	for (i = 0; i < m1->val_notify_num; ++i) {
+		if (req1->first != req2->first || req1->range != req2->range)
+			return 0;
+		req1++;
+		req2++;
+	}
+
+	return 1;
+}
+
+/*
+ * Main connector device's callback.
+ *
+ * Used for notification of a request's processing.
+ */
+static void cn_callback(void *data)
+{
+	struct cn_msg *msg = data;
+	struct cn_ctl_msg *ctl;
+	struct cn_ctl_entry *ent;
+	u32 size;
+
+	if (msg->len < sizeof(*ctl))
+		return;
+
+	ctl = (struct cn_ctl_msg *)msg->data;
+
+	size = (sizeof(*ctl) + ((ctl->idx_notify_num +
+				 ctl->val_notify_num) *
+				sizeof(struct cn_notify_req)));
+
+	if (msg->len != size)
+		return;
+
+	if (ctl->len + sizeof(*ctl) != msg->len)
+		return;
+
+	/*
+	 * Remove notification.
+	 */
+	if (ctl->group == 0) {
+		struct cn_ctl_entry *n;
+
+		down(&notify_lock);
+		list_for_each_entry_safe(ent, n, &notify_list, notify_entry) {
+			if (cn_ctl_msg_equals(ent->msg, ctl)) {
+				list_del(&ent->notify_entry);
+				kfree(ent);
+			}
+		}
+		up(&notify_lock);
+
+		return;
+	}
+
+	size += sizeof(*ent);
+
+	ent = kzalloc(size, GFP_KERNEL);
+	if (!ent)
+		return;
+
+	ent->msg = (struct cn_ctl_msg *)(ent + 1);
+
+	memcpy(ent->msg, ctl, size - sizeof(*ent));
+
+	down(&notify_lock);
+	list_add(&ent->notify_entry, &notify_list);
+	up(&notify_lock);
+}
+
+static int __init cn_init(void)
+{
+	struct cn_dev *dev = &cdev;
+	int err;
+
+	dev->input = cn_input;
+	dev->id.idx = cn_idx;
+	dev->id.val = cn_val;
+
+	dev->nls = netlink_kernel_create(NETLINK_CONNECTOR,
+					 CN_NETLINK_USERS + 0xf,
+					 dev->input, THIS_MODULE);
+	if (!dev->nls)
+		return -EIO;
+
+	dev->cbdev = cn_queue_alloc_dev("cqueue", dev->nls);
+	if (!dev->cbdev) {
+		if (dev->nls->sk_socket)
+			sock_release(dev->nls->sk_socket);
+		return -EINVAL;
+	}
+
+	err = cn_add_callback(&dev->id, "connector", &cn_callback);
+	if (err) {
+		cn_queue_free_dev(dev->cbdev);
+		if (dev->nls->sk_socket)
+			sock_release(dev->nls->sk_socket);
+		return -EINVAL;
+	}
+
+	cn_already_initialized = 1;
+
+	return 0;
+}
+
+static void __exit cn_fini(void)
+{
+	struct cn_dev *dev = &cdev;
+
+	cn_already_initialized = 0;
+
+	cn_del_callback(&dev->id);
+	cn_queue_free_dev(dev->cbdev);
+	if (dev->nls->sk_socket)
+		sock_release(dev->nls->sk_socket);
+}
+
+module_init(cn_init);
+module_exit(cn_fini);
+
+EXPORT_SYMBOL_GPL(cn_add_callback);
+EXPORT_SYMBOL_GPL(cn_del_callback);
+EXPORT_SYMBOL_GPL(cn_netlink_send);