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 #include <linux/kernel.h>
39 #include <linux/errno.h>
40 #include <linux/uio.h>
41 #include <linux/notifier.h>
42 #include <linux/wait.h>
44 #include <linux/poll.h>
45 #include <linux/mutex.h>
46 #include <linux/sched.h>
47 #include <linux/spinlock.h>
48 #include <linux/mount.h>
49 #include <linux/pagemap.h>
50 #include <linux/uaccess.h>
51 #include <linux/init.h>
52 #include <linux/namei.h>
53 #include <linux/string.h>
56 #include "../xenbus/xenbus_comms.h"
58 #include <xen/xenbus.h>
59 #include <asm/xen/hypervisor.h>
62 * An element of a list of outstanding transactions, for which we're
63 * still waiting a reply.
65 struct xenbus_transaction_holder {
66 struct list_head list;
67 struct xenbus_transaction handle;
71 * A buffer of data on the queue.
74 struct list_head list;
80 struct xenbus_file_priv {
82 * msgbuffer_mutex is held while partial requests are built up
83 * and complete requests are acted on. It therefore protects
84 * the "transactions" and "watches" lists, and the partial
85 * request length and buffer.
87 * reply_mutex protects the reply being built up to return to
88 * usermode. It nests inside msgbuffer_mutex but may be held
89 * alone during a watch callback.
91 struct mutex msgbuffer_mutex;
93 /* In-progress transactions */
94 struct list_head transactions;
97 struct list_head watches;
99 /* Partial request. */
102 struct xsd_sockmsg msg;
103 char buffer[PAGE_SIZE];
106 /* Response queue. */
107 struct mutex reply_mutex;
108 struct list_head read_buffers;
109 wait_queue_head_t read_waitq;
113 /* Read out any raw xenbus messages queued up. */
114 static ssize_t xenbus_file_read(struct file *filp,
116 size_t len, loff_t *ppos)
118 struct xenbus_file_priv *u = filp->private_data;
119 struct read_buffer *rb;
123 mutex_lock(&u->reply_mutex);
124 while (list_empty(&u->read_buffers)) {
125 mutex_unlock(&u->reply_mutex);
126 ret = wait_event_interruptible(u->read_waitq,
127 !list_empty(&u->read_buffers));
130 mutex_lock(&u->reply_mutex);
133 rb = list_entry(u->read_buffers.next, struct read_buffer, list);
136 unsigned sz = min((unsigned)len - i, rb->len - rb->cons);
138 ret = copy_to_user(ubuf + i, &rb->msg[rb->cons], sz);
141 rb->cons += sz - ret;
149 /* Clear out buffer if it has been consumed */
150 if (rb->cons == rb->len) {
153 if (list_empty(&u->read_buffers))
155 rb = list_entry(u->read_buffers.next,
156 struct read_buffer, list);
161 mutex_unlock(&u->reply_mutex);
166 * Add a buffer to the queue. Caller must hold the appropriate lock
167 * if the queue is not local. (Commonly the caller will build up
168 * multiple queued buffers on a temporary local list, and then add it
169 * to the appropriate list under lock once all the buffers have een
170 * successfully allocated.)
172 static int queue_reply(struct list_head *queue, const void *data, size_t len)
174 struct read_buffer *rb;
179 rb = kmalloc(sizeof(*rb) + len, GFP_KERNEL);
186 memcpy(rb->msg, data, len);
188 list_add_tail(&rb->list, queue);
193 * Free all the read_buffer s on a list.
194 * Caller must have sole reference to list.
196 static void queue_cleanup(struct list_head *list)
198 struct read_buffer *rb;
200 while (!list_empty(list)) {
201 rb = list_entry(list->next, struct read_buffer, list);
202 list_del(list->next);
207 struct watch_adapter {
208 struct list_head list;
209 struct xenbus_watch watch;
210 struct xenbus_file_priv *dev_data;
214 static void free_watch_adapter(struct watch_adapter *watch)
216 kfree(watch->watch.node);
221 static struct watch_adapter *alloc_watch_adapter(const char *path,
224 struct watch_adapter *watch;
226 watch = kzalloc(sizeof(*watch), GFP_KERNEL);
230 watch->watch.node = kstrdup(path, GFP_KERNEL);
231 if (watch->watch.node == NULL)
234 watch->token = kstrdup(token, GFP_KERNEL);
235 if (watch->token == NULL)
241 free_watch_adapter(watch);
247 static void watch_fired(struct xenbus_watch *watch,
251 struct watch_adapter *adap;
252 struct xsd_sockmsg hdr;
253 const char *path, *token;
254 int path_len, tok_len, body_len, data_len = 0;
256 LIST_HEAD(staging_q);
258 adap = container_of(watch, struct watch_adapter, watch);
260 path = vec[XS_WATCH_PATH];
263 path_len = strlen(path) + 1;
264 tok_len = strlen(token) + 1;
266 data_len = vec[len] - vec[2] + 1;
267 body_len = path_len + tok_len + data_len;
269 hdr.type = XS_WATCH_EVENT;
272 mutex_lock(&adap->dev_data->reply_mutex);
274 ret = queue_reply(&staging_q, &hdr, sizeof(hdr));
276 ret = queue_reply(&staging_q, path, path_len);
278 ret = queue_reply(&staging_q, token, tok_len);
280 ret = queue_reply(&staging_q, vec[2], data_len);
283 /* success: pass reply list onto watcher */
284 list_splice_tail(&staging_q, &adap->dev_data->read_buffers);
285 wake_up(&adap->dev_data->read_waitq);
287 queue_cleanup(&staging_q);
289 mutex_unlock(&adap->dev_data->reply_mutex);
292 static int xenbus_write_transaction(unsigned msg_type,
293 struct xenbus_file_priv *u)
297 struct xenbus_transaction_holder *trans = NULL;
298 LIST_HEAD(staging_q);
300 if (msg_type == XS_TRANSACTION_START) {
301 trans = kmalloc(sizeof(*trans), GFP_KERNEL);
308 reply = xenbus_dev_request_and_reply(&u->u.msg);
315 if (msg_type == XS_TRANSACTION_START) {
316 trans->handle.id = simple_strtoul(reply, NULL, 0);
318 list_add(&trans->list, &u->transactions);
319 } else if (msg_type == XS_TRANSACTION_END) {
320 list_for_each_entry(trans, &u->transactions, list)
321 if (trans->handle.id == u->u.msg.tx_id)
323 BUG_ON(&trans->list == &u->transactions);
324 list_del(&trans->list);
329 mutex_lock(&u->reply_mutex);
330 rc = queue_reply(&staging_q, &u->u.msg, sizeof(u->u.msg));
332 rc = queue_reply(&staging_q, reply, u->u.msg.len);
334 list_splice_tail(&staging_q, &u->read_buffers);
335 wake_up(&u->read_waitq);
337 queue_cleanup(&staging_q);
339 mutex_unlock(&u->reply_mutex);
347 static int xenbus_write_watch(unsigned msg_type, struct xenbus_file_priv *u)
349 struct watch_adapter *watch, *tmp_watch;
352 LIST_HEAD(staging_q);
354 path = u->u.buffer + sizeof(u->u.msg);
355 token = memchr(path, 0, u->u.msg.len);
362 if (msg_type == XS_WATCH) {
363 watch = alloc_watch_adapter(path, token);
369 watch->watch.callback = watch_fired;
372 err = register_xenbus_watch(&watch->watch);
374 free_watch_adapter(watch);
378 list_add(&watch->list, &u->watches);
380 list_for_each_entry_safe(watch, tmp_watch, &u->watches, list) {
381 if (!strcmp(watch->token, token) &&
382 !strcmp(watch->watch.node, path)) {
383 unregister_xenbus_watch(&watch->watch);
384 list_del(&watch->list);
385 free_watch_adapter(watch);
391 /* Success. Synthesize a reply to say all is OK. */
394 struct xsd_sockmsg hdr;
399 .len = sizeof(reply.body)
404 mutex_lock(&u->reply_mutex);
405 rc = queue_reply(&u->read_buffers, &reply, sizeof(reply));
406 mutex_unlock(&u->reply_mutex);
413 static ssize_t xenbus_file_write(struct file *filp,
414 const char __user *ubuf,
415 size_t len, loff_t *ppos)
417 struct xenbus_file_priv *u = filp->private_data;
421 LIST_HEAD(staging_q);
424 * We're expecting usermode to be writing properly formed
425 * xenbus messages. If they write an incomplete message we
426 * buffer it up. Once it is complete, we act on it.
430 * Make sure concurrent writers can't stomp all over each
431 * other's messages and make a mess of our partial message
432 * buffer. We don't make any attemppt to stop multiple
433 * writers from making a mess of each other's incomplete
434 * messages; we're just trying to guarantee our own internal
435 * consistency and make sure that single writes are handled
438 mutex_lock(&u->msgbuffer_mutex);
440 /* Get this out of the way early to avoid confusion */
444 /* Can't write a xenbus message larger we can buffer */
445 if ((len + u->len) > sizeof(u->u.buffer)) {
446 /* On error, dump existing buffer */
452 ret = copy_from_user(u->u.buffer + u->len, ubuf, len);
459 /* Deal with a partial copy. */
465 /* Return if we haven't got a full message yet */
466 if (u->len < sizeof(u->u.msg))
467 goto out; /* not even the header yet */
469 /* If we're expecting a message that's larger than we can
470 possibly send, dump what we have and return an error. */
471 if ((sizeof(u->u.msg) + u->u.msg.len) > sizeof(u->u.buffer)) {
477 if (u->len < (sizeof(u->u.msg) + u->u.msg.len))
478 goto out; /* incomplete data portion */
481 * OK, now we have a complete message. Do something with it.
484 msg_type = u->u.msg.type;
487 case XS_TRANSACTION_START:
488 case XS_TRANSACTION_END:
493 case XS_GET_DOMAIN_PATH:
498 /* Send out a transaction */
499 ret = xenbus_write_transaction(msg_type, u);
504 /* (Un)Ask for some path to be watched for changes */
505 ret = xenbus_write_watch(msg_type, u);
515 /* Buffered message consumed */
519 mutex_unlock(&u->msgbuffer_mutex);
523 static int xenbus_file_open(struct inode *inode, struct file *filp)
525 struct xenbus_file_priv *u;
527 if (xen_store_evtchn == 0)
530 nonseekable_open(inode, filp);
532 u = kzalloc(sizeof(*u), GFP_KERNEL);
536 INIT_LIST_HEAD(&u->transactions);
537 INIT_LIST_HEAD(&u->watches);
538 INIT_LIST_HEAD(&u->read_buffers);
539 init_waitqueue_head(&u->read_waitq);
541 mutex_init(&u->reply_mutex);
542 mutex_init(&u->msgbuffer_mutex);
544 filp->private_data = u;
549 static int xenbus_file_release(struct inode *inode, struct file *filp)
551 struct xenbus_file_priv *u = filp->private_data;
552 struct xenbus_transaction_holder *trans, *tmp;
553 struct watch_adapter *watch, *tmp_watch;
556 * No need for locking here because there are no other users,
560 list_for_each_entry_safe(trans, tmp, &u->transactions, list) {
561 xenbus_transaction_end(trans->handle, 1);
562 list_del(&trans->list);
566 list_for_each_entry_safe(watch, tmp_watch, &u->watches, list) {
567 unregister_xenbus_watch(&watch->watch);
568 list_del(&watch->list);
569 free_watch_adapter(watch);
577 static unsigned int xenbus_file_poll(struct file *file, poll_table *wait)
579 struct xenbus_file_priv *u = file->private_data;
581 poll_wait(file, &u->read_waitq, wait);
582 if (!list_empty(&u->read_buffers))
583 return POLLIN | POLLRDNORM;
587 const struct file_operations xenbus_file_ops = {
588 .read = xenbus_file_read,
589 .write = xenbus_file_write,
590 .open = xenbus_file_open,
591 .release = xenbus_file_release,
592 .poll = xenbus_file_poll,