2 * Driver giving user-space access to the kernel's xenbus connection
5 * Copyright (c) 2005, Christian Limpach
6 * Copyright (c) 2005, Rusty Russell, IBM Corporation
8 * This program is free software; you can redistribute it and/or
9 * modify it under the terms of the GNU General Public License version 2
10 * as published by the Free Software Foundation; or, when distributed
11 * separately from the Linux kernel or incorporated into other
12 * software packages, subject to the following license:
14 * Permission is hereby granted, free of charge, to any person obtaining a copy
15 * of this source file (the "Software"), to deal in the Software without
16 * restriction, including without limitation the rights to use, copy, modify,
17 * merge, publish, distribute, sublicense, and/or sell copies of the Software,
18 * and to permit persons to whom the Software is furnished to do so, subject to
19 * the following conditions:
21 * The above copyright notice and this permission notice shall be included in
22 * all copies or substantial portions of the Software.
24 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
25 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
26 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
27 * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
28 * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
29 * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
33 * 2008-10-07 Alex Zeffertt Replaced /proc/xen/xenbus with xenfs filesystem
34 * and /proc/xen compatibility mount point.
35 * Turned xenfs into a loadable module.
38 #define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
40 #include <linux/kernel.h>
41 #include <linux/errno.h>
42 #include <linux/uio.h>
43 #include <linux/notifier.h>
44 #include <linux/wait.h>
46 #include <linux/poll.h>
47 #include <linux/mutex.h>
48 #include <linux/sched.h>
49 #include <linux/spinlock.h>
50 #include <linux/mount.h>
51 #include <linux/pagemap.h>
52 #include <linux/uaccess.h>
53 #include <linux/init.h>
54 #include <linux/namei.h>
55 #include <linux/string.h>
56 #include <linux/slab.h>
57 #include <linux/miscdevice.h>
58 #include <linux/workqueue.h>
60 #include <xen/xenbus.h>
62 #include <asm/xen/hypervisor.h>
66 unsigned int xb_dev_generation_id;
69 * An element of a list of outstanding transactions, for which we're
70 * still waiting a reply.
72 struct xenbus_transaction_holder {
73 struct list_head list;
74 struct xenbus_transaction handle;
75 unsigned int generation_id;
79 * A buffer of data on the queue.
82 struct list_head list;
88 struct xenbus_file_priv {
90 * msgbuffer_mutex is held while partial requests are built up
91 * and complete requests are acted on. It therefore protects
92 * the "transactions" and "watches" lists, and the partial
93 * request length and buffer.
95 * reply_mutex protects the reply being built up to return to
96 * usermode. It nests inside msgbuffer_mutex but may be held
97 * alone during a watch callback.
99 struct mutex msgbuffer_mutex;
101 /* In-progress transactions */
102 struct list_head transactions;
104 /* Active watches. */
105 struct list_head watches;
107 /* Partial request. */
110 struct xsd_sockmsg msg;
111 char buffer[XENSTORE_PAYLOAD_MAX];
114 /* Response queue. */
115 struct mutex reply_mutex;
116 struct list_head read_buffers;
117 wait_queue_head_t read_waitq;
121 struct work_struct wq;
124 /* Read out any raw xenbus messages queued up. */
125 static ssize_t xenbus_file_read(struct file *filp,
127 size_t len, loff_t *ppos)
129 struct xenbus_file_priv *u = filp->private_data;
130 struct read_buffer *rb;
134 mutex_lock(&u->reply_mutex);
136 while (list_empty(&u->read_buffers)) {
137 mutex_unlock(&u->reply_mutex);
138 if (filp->f_flags & O_NONBLOCK)
141 ret = wait_event_interruptible(u->read_waitq,
142 !list_empty(&u->read_buffers));
145 mutex_lock(&u->reply_mutex);
148 rb = list_entry(u->read_buffers.next, struct read_buffer, list);
151 unsigned sz = min((unsigned)len - i, rb->len - rb->cons);
153 ret = copy_to_user(ubuf + i, &rb->msg[rb->cons], sz);
156 rb->cons += sz - ret;
164 /* Clear out buffer if it has been consumed */
165 if (rb->cons == rb->len) {
168 if (list_empty(&u->read_buffers))
170 rb = list_entry(u->read_buffers.next,
171 struct read_buffer, list);
178 mutex_unlock(&u->reply_mutex);
183 * Add a buffer to the queue. Caller must hold the appropriate lock
184 * if the queue is not local. (Commonly the caller will build up
185 * multiple queued buffers on a temporary local list, and then add it
186 * to the appropriate list under lock once all the buffers have een
187 * successfully allocated.)
189 static int queue_reply(struct list_head *queue, const void *data, size_t len)
191 struct read_buffer *rb;
195 if (len > XENSTORE_PAYLOAD_MAX)
198 rb = kmalloc(sizeof(*rb) + len, GFP_KERNEL);
205 memcpy(rb->msg, data, len);
207 list_add_tail(&rb->list, queue);
212 * Free all the read_buffer s on a list.
213 * Caller must have sole reference to list.
215 static void queue_cleanup(struct list_head *list)
217 struct read_buffer *rb;
219 while (!list_empty(list)) {
220 rb = list_entry(list->next, struct read_buffer, list);
221 list_del(list->next);
226 struct watch_adapter {
227 struct list_head list;
228 struct xenbus_watch watch;
229 struct xenbus_file_priv *dev_data;
233 static void free_watch_adapter(struct watch_adapter *watch)
235 kfree(watch->watch.node);
240 static struct watch_adapter *alloc_watch_adapter(const char *path,
243 struct watch_adapter *watch;
245 watch = kzalloc(sizeof(*watch), GFP_KERNEL);
249 watch->watch.node = kstrdup(path, GFP_KERNEL);
250 if (watch->watch.node == NULL)
253 watch->token = kstrdup(token, GFP_KERNEL);
254 if (watch->token == NULL)
260 free_watch_adapter(watch);
266 static void watch_fired(struct xenbus_watch *watch,
270 struct watch_adapter *adap;
271 struct xsd_sockmsg hdr;
272 const char *token_caller;
273 int path_len, tok_len, body_len;
275 LIST_HEAD(staging_q);
277 adap = container_of(watch, struct watch_adapter, watch);
279 token_caller = adap->token;
281 path_len = strlen(path) + 1;
282 tok_len = strlen(token_caller) + 1;
283 body_len = path_len + tok_len;
285 hdr.type = XS_WATCH_EVENT;
288 mutex_lock(&adap->dev_data->reply_mutex);
290 ret = queue_reply(&staging_q, &hdr, sizeof(hdr));
292 ret = queue_reply(&staging_q, path, path_len);
294 ret = queue_reply(&staging_q, token_caller, tok_len);
297 /* success: pass reply list onto watcher */
298 list_splice_tail(&staging_q, &adap->dev_data->read_buffers);
299 wake_up(&adap->dev_data->read_waitq);
301 queue_cleanup(&staging_q);
303 mutex_unlock(&adap->dev_data->reply_mutex);
306 static void xenbus_worker(struct work_struct *wq)
308 struct xenbus_file_priv *u;
309 struct xenbus_transaction_holder *trans, *tmp;
310 struct watch_adapter *watch, *tmp_watch;
311 struct read_buffer *rb, *tmp_rb;
313 u = container_of(wq, struct xenbus_file_priv, wq);
316 * No need for locking here because there are no other users,
320 list_for_each_entry_safe(trans, tmp, &u->transactions, list) {
321 xenbus_transaction_end(trans->handle, 1);
322 list_del(&trans->list);
326 list_for_each_entry_safe(watch, tmp_watch, &u->watches, list) {
327 unregister_xenbus_watch(&watch->watch);
328 list_del(&watch->list);
329 free_watch_adapter(watch);
332 list_for_each_entry_safe(rb, tmp_rb, &u->read_buffers, list) {
339 static void xenbus_file_free(struct kref *kref)
341 struct xenbus_file_priv *u;
344 * We might be called in xenbus_thread().
345 * Use workqueue to avoid deadlock.
347 u = container_of(kref, struct xenbus_file_priv, kref);
348 schedule_work(&u->wq);
351 static struct xenbus_transaction_holder *xenbus_get_transaction(
352 struct xenbus_file_priv *u, uint32_t tx_id)
354 struct xenbus_transaction_holder *trans;
356 list_for_each_entry(trans, &u->transactions, list)
357 if (trans->handle.id == tx_id)
363 void xenbus_dev_queue_reply(struct xb_req_data *req)
365 struct xenbus_file_priv *u = req->par;
366 struct xenbus_transaction_holder *trans = NULL;
368 LIST_HEAD(staging_q);
370 xs_request_exit(req);
372 mutex_lock(&u->msgbuffer_mutex);
374 if (req->type == XS_TRANSACTION_START) {
375 trans = xenbus_get_transaction(u, 0);
378 if (req->msg.type == XS_ERROR) {
379 list_del(&trans->list);
382 rc = kstrtou32(req->body, 10, &trans->handle.id);
386 } else if (req->type == XS_TRANSACTION_END) {
387 trans = xenbus_get_transaction(u, req->msg.tx_id);
390 list_del(&trans->list);
394 mutex_unlock(&u->msgbuffer_mutex);
396 mutex_lock(&u->reply_mutex);
397 rc = queue_reply(&staging_q, &req->msg, sizeof(req->msg));
399 rc = queue_reply(&staging_q, req->body, req->msg.len);
401 list_splice_tail(&staging_q, &u->read_buffers);
402 wake_up(&u->read_waitq);
404 queue_cleanup(&staging_q);
406 mutex_unlock(&u->reply_mutex);
411 kref_put(&u->kref, xenbus_file_free);
416 mutex_unlock(&u->msgbuffer_mutex);
419 static int xenbus_command_reply(struct xenbus_file_priv *u,
420 unsigned int msg_type, const char *reply)
423 struct xsd_sockmsg hdr;
429 msg.hdr.type = msg_type;
430 msg.hdr.len = strlen(reply) + 1;
431 if (msg.hdr.len > sizeof(msg.body))
433 memcpy(&msg.body, reply, msg.hdr.len);
435 mutex_lock(&u->reply_mutex);
436 rc = queue_reply(&u->read_buffers, &msg, sizeof(msg.hdr) + msg.hdr.len);
437 wake_up(&u->read_waitq);
438 mutex_unlock(&u->reply_mutex);
441 kref_put(&u->kref, xenbus_file_free);
446 static int xenbus_write_transaction(unsigned msg_type,
447 struct xenbus_file_priv *u)
450 struct xenbus_transaction_holder *trans = NULL;
452 struct xsd_sockmsg hdr;
454 } *msg = (void *)u->u.buffer;
456 if (msg_type == XS_TRANSACTION_START) {
457 trans = kzalloc(sizeof(*trans), GFP_KERNEL);
462 trans->generation_id = xb_dev_generation_id;
463 list_add(&trans->list, &u->transactions);
464 } else if (msg->hdr.tx_id != 0 &&
465 !xenbus_get_transaction(u, msg->hdr.tx_id))
466 return xenbus_command_reply(u, XS_ERROR, "ENOENT");
467 else if (msg_type == XS_TRANSACTION_END &&
468 !(msg->hdr.len == 2 &&
469 (!strcmp(msg->body, "T") || !strcmp(msg->body, "F"))))
470 return xenbus_command_reply(u, XS_ERROR, "EINVAL");
471 else if (msg_type == XS_TRANSACTION_END) {
472 trans = xenbus_get_transaction(u, msg->hdr.tx_id);
473 if (trans && trans->generation_id != xb_dev_generation_id) {
474 list_del(&trans->list);
476 if (!strcmp(msg->body, "T"))
477 return xenbus_command_reply(u, XS_ERROR,
480 return xenbus_command_reply(u,
486 rc = xenbus_dev_request_and_reply(&msg->hdr, u);
488 list_del(&trans->list);
496 static int xenbus_write_watch(unsigned msg_type, struct xenbus_file_priv *u)
498 struct watch_adapter *watch;
502 path = u->u.buffer + sizeof(u->u.msg);
503 token = memchr(path, 0, u->u.msg.len);
505 rc = xenbus_command_reply(u, XS_ERROR, "EINVAL");
509 if (memchr(token, 0, u->u.msg.len - (token - path)) == NULL) {
510 rc = xenbus_command_reply(u, XS_ERROR, "EINVAL");
514 if (msg_type == XS_WATCH) {
515 watch = alloc_watch_adapter(path, token);
521 watch->watch.callback = watch_fired;
524 err = register_xenbus_watch(&watch->watch);
526 free_watch_adapter(watch);
530 list_add(&watch->list, &u->watches);
532 list_for_each_entry(watch, &u->watches, list) {
533 if (!strcmp(watch->token, token) &&
534 !strcmp(watch->watch.node, path)) {
535 unregister_xenbus_watch(&watch->watch);
536 list_del(&watch->list);
537 free_watch_adapter(watch);
543 /* Success. Synthesize a reply to say all is OK. */
544 rc = xenbus_command_reply(u, msg_type, "OK");
550 static ssize_t xenbus_file_write(struct file *filp,
551 const char __user *ubuf,
552 size_t len, loff_t *ppos)
554 struct xenbus_file_priv *u = filp->private_data;
560 * We're expecting usermode to be writing properly formed
561 * xenbus messages. If they write an incomplete message we
562 * buffer it up. Once it is complete, we act on it.
566 * Make sure concurrent writers can't stomp all over each
567 * other's messages and make a mess of our partial message
568 * buffer. We don't make any attemppt to stop multiple
569 * writers from making a mess of each other's incomplete
570 * messages; we're just trying to guarantee our own internal
571 * consistency and make sure that single writes are handled
574 mutex_lock(&u->msgbuffer_mutex);
576 /* Get this out of the way early to avoid confusion */
580 /* Can't write a xenbus message larger we can buffer */
581 if (len > sizeof(u->u.buffer) - u->len) {
582 /* On error, dump existing buffer */
588 ret = copy_from_user(u->u.buffer + u->len, ubuf, len);
595 /* Deal with a partial copy. */
601 /* Return if we haven't got a full message yet */
602 if (u->len < sizeof(u->u.msg))
603 goto out; /* not even the header yet */
605 /* If we're expecting a message that's larger than we can
606 possibly send, dump what we have and return an error. */
607 if ((sizeof(u->u.msg) + u->u.msg.len) > sizeof(u->u.buffer)) {
613 if (u->len < (sizeof(u->u.msg) + u->u.msg.len))
614 goto out; /* incomplete data portion */
617 * OK, now we have a complete message. Do something with it.
622 msg_type = u->u.msg.type;
627 /* (Un)Ask for some path to be watched for changes */
628 ret = xenbus_write_watch(msg_type, u);
632 /* Send out a transaction */
633 ret = xenbus_write_transaction(msg_type, u);
638 kref_put(&u->kref, xenbus_file_free);
641 /* Buffered message consumed */
645 mutex_unlock(&u->msgbuffer_mutex);
649 static int xenbus_file_open(struct inode *inode, struct file *filp)
651 struct xenbus_file_priv *u;
653 if (xen_store_evtchn == 0)
656 stream_open(inode, filp);
658 u = kzalloc(sizeof(*u), GFP_KERNEL);
664 INIT_LIST_HEAD(&u->transactions);
665 INIT_LIST_HEAD(&u->watches);
666 INIT_LIST_HEAD(&u->read_buffers);
667 init_waitqueue_head(&u->read_waitq);
668 INIT_WORK(&u->wq, xenbus_worker);
670 mutex_init(&u->reply_mutex);
671 mutex_init(&u->msgbuffer_mutex);
673 filp->private_data = u;
678 static int xenbus_file_release(struct inode *inode, struct file *filp)
680 struct xenbus_file_priv *u = filp->private_data;
682 kref_put(&u->kref, xenbus_file_free);
687 static __poll_t xenbus_file_poll(struct file *file, poll_table *wait)
689 struct xenbus_file_priv *u = file->private_data;
691 poll_wait(file, &u->read_waitq, wait);
692 if (!list_empty(&u->read_buffers))
693 return EPOLLIN | EPOLLRDNORM;
697 const struct file_operations xen_xenbus_fops = {
698 .read = xenbus_file_read,
699 .write = xenbus_file_write,
700 .open = xenbus_file_open,
701 .release = xenbus_file_release,
702 .poll = xenbus_file_poll,
705 EXPORT_SYMBOL_GPL(xen_xenbus_fops);
707 static struct miscdevice xenbus_dev = {
708 .minor = MISC_DYNAMIC_MINOR,
709 .name = "xen/xenbus",
710 .fops = &xen_xenbus_fops,
713 static int __init xenbus_init(void)
720 err = misc_register(&xenbus_dev);
722 pr_err("Could not register xenbus frontend device\n");
725 device_initcall(xenbus_init);