linux/drivers/xen/xenfs/xenbus.c

598 lines
14 KiB
C
Raw Normal View History

/*
* Driver giving user-space access to the kernel's xenbus connection
* to xenstore.
*
* Copyright (c) 2005, Christian Limpach
* Copyright (c) 2005, Rusty Russell, IBM Corporation
*
* This program is free software; you can redistribute it and/or
* modify it under the terms of the GNU General Public License version 2
* as published by the Free Software Foundation; or, when distributed
* separately from the Linux kernel or incorporated into other
* software packages, subject to the following license:
*
* Permission is hereby granted, free of charge, to any person obtaining a copy
* of this source file (the "Software"), to deal in the Software without
* restriction, including without limitation the rights to use, copy, modify,
* merge, publish, distribute, sublicense, and/or sell copies of the Software,
* and to permit persons to whom the Software is furnished to do so, subject to
* the following conditions:
*
* The above copyright notice and this permission notice shall be included in
* all copies or substantial portions of the Software.
*
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
* AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
* LIABILITY, WHETHER IN AN 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.
*
* Changes:
* 2008-10-07 Alex Zeffertt Replaced /proc/xen/xenbus with xenfs filesystem
* and /proc/xen compatibility mount point.
* Turned xenfs into a loadable module.
*/
#include <linux/kernel.h>
#include <linux/errno.h>
#include <linux/uio.h>
#include <linux/notifier.h>
#include <linux/wait.h>
#include <linux/fs.h>
#include <linux/poll.h>
#include <linux/mutex.h>
#include <linux/sched.h>
#include <linux/spinlock.h>
#include <linux/mount.h>
#include <linux/pagemap.h>
#include <linux/uaccess.h>
#include <linux/init.h>
#include <linux/namei.h>
#include <linux/string.h>
include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h percpu.h is included by sched.h and module.h and thus ends up being included when building most .c files. percpu.h includes slab.h which in turn includes gfp.h making everything defined by the two files universally available and complicating inclusion dependencies. percpu.h -> slab.h dependency is about to be removed. Prepare for this change by updating users of gfp and slab facilities include those headers directly instead of assuming availability. As this conversion needs to touch large number of source files, the following script is used as the basis of conversion. http://userweb.kernel.org/~tj/misc/slabh-sweep.py The script does the followings. * Scan files for gfp and slab usages and update includes such that only the necessary includes are there. ie. if only gfp is used, gfp.h, if slab is used, slab.h. * When the script inserts a new include, it looks at the include blocks and try to put the new include such that its order conforms to its surrounding. It's put in the include block which contains core kernel includes, in the same order that the rest are ordered - alphabetical, Christmas tree, rev-Xmas-tree or at the end if there doesn't seem to be any matching order. * If the script can't find a place to put a new include (mostly because the file doesn't have fitting include block), it prints out an error message indicating which .h file needs to be added to the file. The conversion was done in the following steps. 1. The initial automatic conversion of all .c files updated slightly over 4000 files, deleting around 700 includes and adding ~480 gfp.h and ~3000 slab.h inclusions. The script emitted errors for ~400 files. 2. Each error was manually checked. Some didn't need the inclusion, some needed manual addition while adding it to implementation .h or embedding .c file was more appropriate for others. This step added inclusions to around 150 files. 3. The script was run again and the output was compared to the edits from #2 to make sure no file was left behind. 4. Several build tests were done and a couple of problems were fixed. e.g. lib/decompress_*.c used malloc/free() wrappers around slab APIs requiring slab.h to be added manually. 5. The script was run on all .h files but without automatically editing them as sprinkling gfp.h and slab.h inclusions around .h files could easily lead to inclusion dependency hell. Most gfp.h inclusion directives were ignored as stuff from gfp.h was usually wildly available and often used in preprocessor macros. Each slab.h inclusion directive was examined and added manually as necessary. 6. percpu.h was updated not to include slab.h. 7. Build test were done on the following configurations and failures were fixed. CONFIG_GCOV_KERNEL was turned off for all tests (as my distributed build env didn't work with gcov compiles) and a few more options had to be turned off depending on archs to make things build (like ipr on powerpc/64 which failed due to missing writeq). * x86 and x86_64 UP and SMP allmodconfig and a custom test config. * powerpc and powerpc64 SMP allmodconfig * sparc and sparc64 SMP allmodconfig * ia64 SMP allmodconfig * s390 SMP allmodconfig * alpha SMP allmodconfig * um on x86_64 SMP allmodconfig 8. percpu.h modifications were reverted so that it could be applied as a separate patch and serve as bisection point. Given the fact that I had only a couple of failures from tests on step 6, I'm fairly confident about the coverage of this conversion patch. If there is a breakage, it's likely to be something in one of the arch headers which should be easily discoverable easily on most builds of the specific arch. Signed-off-by: Tejun Heo <tj@kernel.org> Guess-its-ok-by: Christoph Lameter <cl@linux-foundation.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
2010-03-24 16:04:11 +08:00
#include <linux/slab.h>
#include "xenfs.h"
#include "../xenbus/xenbus_comms.h"
#include <xen/xenbus.h>
#include <asm/xen/hypervisor.h>
/*
* An element of a list of outstanding transactions, for which we're
* still waiting a reply.
*/
struct xenbus_transaction_holder {
struct list_head list;
struct xenbus_transaction handle;
};
/*
* A buffer of data on the queue.
*/
struct read_buffer {
struct list_head list;
unsigned int cons;
unsigned int len;
char msg[];
};
struct xenbus_file_priv {
/*
* msgbuffer_mutex is held while partial requests are built up
* and complete requests are acted on. It therefore protects
* the "transactions" and "watches" lists, and the partial
* request length and buffer.
*
* reply_mutex protects the reply being built up to return to
* usermode. It nests inside msgbuffer_mutex but may be held
* alone during a watch callback.
*/
struct mutex msgbuffer_mutex;
/* In-progress transactions */
struct list_head transactions;
/* Active watches. */
struct list_head watches;
/* Partial request. */
unsigned int len;
union {
struct xsd_sockmsg msg;
char buffer[PAGE_SIZE];
} u;
/* Response queue. */
struct mutex reply_mutex;
struct list_head read_buffers;
wait_queue_head_t read_waitq;
};
/* Read out any raw xenbus messages queued up. */
static ssize_t xenbus_file_read(struct file *filp,
char __user *ubuf,
size_t len, loff_t *ppos)
{
struct xenbus_file_priv *u = filp->private_data;
struct read_buffer *rb;
unsigned i;
int ret;
mutex_lock(&u->reply_mutex);
while (list_empty(&u->read_buffers)) {
mutex_unlock(&u->reply_mutex);
if (filp->f_flags & O_NONBLOCK)
return -EAGAIN;
ret = wait_event_interruptible(u->read_waitq,
!list_empty(&u->read_buffers));
if (ret)
return ret;
mutex_lock(&u->reply_mutex);
}
rb = list_entry(u->read_buffers.next, struct read_buffer, list);
i = 0;
while (i < len) {
unsigned sz = min((unsigned)len - i, rb->len - rb->cons);
ret = copy_to_user(ubuf + i, &rb->msg[rb->cons], sz);
i += sz - ret;
rb->cons += sz - ret;
if (ret != sz) {
if (i == 0)
i = -EFAULT;
goto out;
}
/* Clear out buffer if it has been consumed */
if (rb->cons == rb->len) {
list_del(&rb->list);
kfree(rb);
if (list_empty(&u->read_buffers))
break;
rb = list_entry(u->read_buffers.next,
struct read_buffer, list);
}
}
out:
mutex_unlock(&u->reply_mutex);
return i;
}
/*
* Add a buffer to the queue. Caller must hold the appropriate lock
* if the queue is not local. (Commonly the caller will build up
* multiple queued buffers on a temporary local list, and then add it
* to the appropriate list under lock once all the buffers have een
* successfully allocated.)
*/
static int queue_reply(struct list_head *queue, const void *data, size_t len)
{
struct read_buffer *rb;
if (len == 0)
return 0;
rb = kmalloc(sizeof(*rb) + len, GFP_KERNEL);
if (rb == NULL)
return -ENOMEM;
rb->cons = 0;
rb->len = len;
memcpy(rb->msg, data, len);
list_add_tail(&rb->list, queue);
return 0;
}
/*
* Free all the read_buffer s on a list.
* Caller must have sole reference to list.
*/
static void queue_cleanup(struct list_head *list)
{
struct read_buffer *rb;
while (!list_empty(list)) {
rb = list_entry(list->next, struct read_buffer, list);
list_del(list->next);
kfree(rb);
}
}
struct watch_adapter {
struct list_head list;
struct xenbus_watch watch;
struct xenbus_file_priv *dev_data;
char *token;
};
static void free_watch_adapter(struct watch_adapter *watch)
{
kfree(watch->watch.node);
kfree(watch->token);
kfree(watch);
}
static struct watch_adapter *alloc_watch_adapter(const char *path,
const char *token)
{
struct watch_adapter *watch;
watch = kzalloc(sizeof(*watch), GFP_KERNEL);
if (watch == NULL)
goto out_fail;
watch->watch.node = kstrdup(path, GFP_KERNEL);
if (watch->watch.node == NULL)
goto out_free;
watch->token = kstrdup(token, GFP_KERNEL);
if (watch->token == NULL)
goto out_free;
return watch;
out_free:
free_watch_adapter(watch);
out_fail:
return NULL;
}
static void watch_fired(struct xenbus_watch *watch,
const char **vec,
unsigned int len)
{
struct watch_adapter *adap;
struct xsd_sockmsg hdr;
const char *path, *token;
int path_len, tok_len, body_len, data_len = 0;
int ret;
LIST_HEAD(staging_q);
adap = container_of(watch, struct watch_adapter, watch);
path = vec[XS_WATCH_PATH];
token = adap->token;
path_len = strlen(path) + 1;
tok_len = strlen(token) + 1;
if (len > 2)
data_len = vec[len] - vec[2] + 1;
body_len = path_len + tok_len + data_len;
hdr.type = XS_WATCH_EVENT;
hdr.len = body_len;
mutex_lock(&adap->dev_data->reply_mutex);
ret = queue_reply(&staging_q, &hdr, sizeof(hdr));
if (!ret)
ret = queue_reply(&staging_q, path, path_len);
if (!ret)
ret = queue_reply(&staging_q, token, tok_len);
if (!ret && len > 2)
ret = queue_reply(&staging_q, vec[2], data_len);
if (!ret) {
/* success: pass reply list onto watcher */
list_splice_tail(&staging_q, &adap->dev_data->read_buffers);
wake_up(&adap->dev_data->read_waitq);
} else
queue_cleanup(&staging_q);
mutex_unlock(&adap->dev_data->reply_mutex);
}
static int xenbus_write_transaction(unsigned msg_type,
struct xenbus_file_priv *u)
{
int rc;
void *reply;
struct xenbus_transaction_holder *trans = NULL;
LIST_HEAD(staging_q);
if (msg_type == XS_TRANSACTION_START) {
trans = kmalloc(sizeof(*trans), GFP_KERNEL);
if (!trans) {
rc = -ENOMEM;
goto out;
}
}
reply = xenbus_dev_request_and_reply(&u->u.msg);
if (IS_ERR(reply)) {
kfree(trans);
rc = PTR_ERR(reply);
goto out;
}
if (msg_type == XS_TRANSACTION_START) {
trans->handle.id = simple_strtoul(reply, NULL, 0);
list_add(&trans->list, &u->transactions);
} else if (msg_type == XS_TRANSACTION_END) {
list_for_each_entry(trans, &u->transactions, list)
if (trans->handle.id == u->u.msg.tx_id)
break;
BUG_ON(&trans->list == &u->transactions);
list_del(&trans->list);
kfree(trans);
}
mutex_lock(&u->reply_mutex);
rc = queue_reply(&staging_q, &u->u.msg, sizeof(u->u.msg));
if (!rc)
rc = queue_reply(&staging_q, reply, u->u.msg.len);
if (!rc) {
list_splice_tail(&staging_q, &u->read_buffers);
wake_up(&u->read_waitq);
} else {
queue_cleanup(&staging_q);
}
mutex_unlock(&u->reply_mutex);
kfree(reply);
out:
return rc;
}
static int xenbus_write_watch(unsigned msg_type, struct xenbus_file_priv *u)
{
struct watch_adapter *watch, *tmp_watch;
char *path, *token;
int err, rc;
LIST_HEAD(staging_q);
path = u->u.buffer + sizeof(u->u.msg);
token = memchr(path, 0, u->u.msg.len);
if (token == NULL) {
rc = -EILSEQ;
goto out;
}
token++;
if (msg_type == XS_WATCH) {
watch = alloc_watch_adapter(path, token);
if (watch == NULL) {
rc = -ENOMEM;
goto out;
}
watch->watch.callback = watch_fired;
watch->dev_data = u;
err = register_xenbus_watch(&watch->watch);
if (err) {
free_watch_adapter(watch);
rc = err;
goto out;
}
list_add(&watch->list, &u->watches);
} else {
list_for_each_entry_safe(watch, tmp_watch, &u->watches, list) {
if (!strcmp(watch->token, token) &&
!strcmp(watch->watch.node, path)) {
unregister_xenbus_watch(&watch->watch);
list_del(&watch->list);
free_watch_adapter(watch);
break;
}
}
}
/* Success. Synthesize a reply to say all is OK. */
{
struct {
struct xsd_sockmsg hdr;
char body[3];
} __packed reply = {
{
.type = msg_type,
.len = sizeof(reply.body)
},
"OK"
};
mutex_lock(&u->reply_mutex);
rc = queue_reply(&u->read_buffers, &reply, sizeof(reply));
mutex_unlock(&u->reply_mutex);
}
out:
return rc;
}
static ssize_t xenbus_file_write(struct file *filp,
const char __user *ubuf,
size_t len, loff_t *ppos)
{
struct xenbus_file_priv *u = filp->private_data;
uint32_t msg_type;
int rc = len;
int ret;
LIST_HEAD(staging_q);
/*
* We're expecting usermode to be writing properly formed
* xenbus messages. If they write an incomplete message we
* buffer it up. Once it is complete, we act on it.
*/
/*
* Make sure concurrent writers can't stomp all over each
* other's messages and make a mess of our partial message
* buffer. We don't make any attemppt to stop multiple
* writers from making a mess of each other's incomplete
* messages; we're just trying to guarantee our own internal
* consistency and make sure that single writes are handled
* atomically.
*/
mutex_lock(&u->msgbuffer_mutex);
/* Get this out of the way early to avoid confusion */
if (len == 0)
goto out;
/* Can't write a xenbus message larger we can buffer */
if ((len + u->len) > sizeof(u->u.buffer)) {
/* On error, dump existing buffer */
u->len = 0;
rc = -EINVAL;
goto out;
}
ret = copy_from_user(u->u.buffer + u->len, ubuf, len);
if (ret == len) {
rc = -EFAULT;
goto out;
}
/* Deal with a partial copy. */
len -= ret;
rc = len;
u->len += len;
/* Return if we haven't got a full message yet */
if (u->len < sizeof(u->u.msg))
goto out; /* not even the header yet */
/* If we're expecting a message that's larger than we can
possibly send, dump what we have and return an error. */
if ((sizeof(u->u.msg) + u->u.msg.len) > sizeof(u->u.buffer)) {
rc = -E2BIG;
u->len = 0;
goto out;
}
if (u->len < (sizeof(u->u.msg) + u->u.msg.len))
goto out; /* incomplete data portion */
/*
* OK, now we have a complete message. Do something with it.
*/
msg_type = u->u.msg.type;
switch (msg_type) {
case XS_TRANSACTION_START:
case XS_TRANSACTION_END:
case XS_DIRECTORY:
case XS_READ:
case XS_GET_PERMS:
case XS_RELEASE:
case XS_GET_DOMAIN_PATH:
case XS_WRITE:
case XS_MKDIR:
case XS_RM:
case XS_SET_PERMS:
/* Send out a transaction */
ret = xenbus_write_transaction(msg_type, u);
break;
case XS_WATCH:
case XS_UNWATCH:
/* (Un)Ask for some path to be watched for changes */
ret = xenbus_write_watch(msg_type, u);
break;
default:
ret = -EINVAL;
break;
}
if (ret != 0)
rc = ret;
/* Buffered message consumed */
u->len = 0;
out:
mutex_unlock(&u->msgbuffer_mutex);
return rc;
}
static int xenbus_file_open(struct inode *inode, struct file *filp)
{
struct xenbus_file_priv *u;
if (xen_store_evtchn == 0)
return -ENOENT;
nonseekable_open(inode, filp);
u = kzalloc(sizeof(*u), GFP_KERNEL);
if (u == NULL)
return -ENOMEM;
INIT_LIST_HEAD(&u->transactions);
INIT_LIST_HEAD(&u->watches);
INIT_LIST_HEAD(&u->read_buffers);
init_waitqueue_head(&u->read_waitq);
mutex_init(&u->reply_mutex);
mutex_init(&u->msgbuffer_mutex);
filp->private_data = u;
return 0;
}
static int xenbus_file_release(struct inode *inode, struct file *filp)
{
struct xenbus_file_priv *u = filp->private_data;
struct xenbus_transaction_holder *trans, *tmp;
struct watch_adapter *watch, *tmp_watch;
/*
* No need for locking here because there are no other users,
* by definition.
*/
list_for_each_entry_safe(trans, tmp, &u->transactions, list) {
xenbus_transaction_end(trans->handle, 1);
list_del(&trans->list);
kfree(trans);
}
list_for_each_entry_safe(watch, tmp_watch, &u->watches, list) {
unregister_xenbus_watch(&watch->watch);
list_del(&watch->list);
free_watch_adapter(watch);
}
kfree(u);
return 0;
}
static unsigned int xenbus_file_poll(struct file *file, poll_table *wait)
{
struct xenbus_file_priv *u = file->private_data;
poll_wait(file, &u->read_waitq, wait);
if (!list_empty(&u->read_buffers))
return POLLIN | POLLRDNORM;
return 0;
}
const struct file_operations xenbus_file_ops = {
.read = xenbus_file_read,
.write = xenbus_file_write,
.open = xenbus_file_open,
.release = xenbus_file_release,
.poll = xenbus_file_poll,
};