1
# Copyright (C) 2006 Canonical Ltd
3
# This program is free software; you can redistribute it and/or modify
4
# it under the terms of the GNU General Public License as published by
5
# the Free Software Foundation; either version 2 of the License, or
6
# (at your option) any later version.
8
# This program is distributed in the hope that it will be useful,
9
# but WITHOUT ANY WARRANTY; without even the implied warranty of
10
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
11
# GNU General Public License for more details.
13
# You should have received a copy of the GNU General Public License
14
# along with this program; if not, write to the Free Software
15
# Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
17
"""On-disk mutex protecting a resource
19
bzr on-disk objects are locked by the existence of a directory with a
20
particular name within the control directory. We use this rather than OS
21
internal locks (such as flock etc) because they can be seen across all
22
transports, including http.
24
Objects can be read if there is only physical read access; therefore
25
readers can never be required to create a lock, though they will
26
check whether a writer is using the lock. Writers can't detect
27
whether anyone else is reading from the resource as they write.
28
This works because of ordering constraints that make sure readers
29
see a consistent view of existing data.
31
Waiting for a lock must be done by polling; this can be aborted after
34
Locks must always be explicitly released, typically from a try/finally
35
block -- they are not released from a finalizer or when Python
38
Locks may fail to be released if the process is abruptly terminated
39
(machine stop, SIGKILL) or if a remote transport becomes permanently
40
disconnected. There is therefore a method to break an existing lock.
41
This should rarely be used, and generally only with user approval.
42
Locks contain some information on when the lock was taken and by who
43
which may guide in deciding whether it can safely be broken. (This is
44
similar to the messages displayed by emacs and vim.) Note that if the
45
lock holder is still alive they will get no notification that the lock
46
has been broken and will continue their work -- so it is important to be
47
sure they are actually dead.
49
A lock is represented on disk by a directory of a particular name,
50
containing an information file. Taking a lock is done by renaming a
51
temporary directory into place. We use temporary directories because
52
for all known transports and filesystems we believe that exactly one
53
attempt to claim the lock will succeed and the others will fail. (Files
54
won't do because some filesystems or transports only have
55
rename-and-overwrite, making it hard to tell who won.)
57
The desired characteristics are:
59
* Locks are not reentrant. (That is, a client that tries to take a
60
lock it already holds may deadlock or fail.)
61
* Stale locks can be guessed at by a heuristic
62
* Lost locks can be broken by any client
63
* Failed lock operations leave little or no mess
64
* Deadlocks are avoided by having a timeout always in use, clients
65
desiring indefinite waits can retry or set a silly big timeout.
67
Storage formats use the locks, and also need to consider concurrency
68
issues underneath the lock. A format may choose not to use a lock
69
at all for some operations.
71
LockDirs always operate over a Transport. The transport may be readonly, in
72
which case the lock can be queried but not acquired.
74
Locks are identified by a path name, relative to a base transport.
76
Calling code will typically want to make sure there is exactly one LockDir
77
object per actual lock on disk. This module does nothing to prevent aliasing
78
and deadlocks will likely occur if the locks are aliased.
80
In the future we may add a "freshen" method which can be called
81
by a lock holder to check that their lock has not been broken, and to
82
update the timestamp within it.
86
>>> from bzrlib.transport.memory import MemoryTransport
87
>>> # typically will be obtained from a BzrDir, Branch, etc
88
>>> t = MemoryTransport()
89
>>> l = LockDir(t, 'sample-lock')
92
>>> # do something here
99
from cStringIO import StringIO
102
from bzrlib.errors import (
114
from bzrlib.trace import mutter, note
115
from bzrlib.transport import Transport
116
from bzrlib.osutils import rand_chars, format_delta
117
from bzrlib.rio import read_stanza, Stanza
119
# XXX: At the moment there is no consideration of thread safety on LockDir
120
# objects. This should perhaps be updated - e.g. if two threads try to take a
121
# lock at the same time they should *both* get it. But then that's unlikely
124
# TODO: Perhaps store some kind of note like the bzr command line in the lock
127
# TODO: Some kind of callback run while polling a lock to show progress
130
# TODO: Make sure to pass the right file and directory mode bits to all
131
# files/dirs created.
133
_DEFAULT_TIMEOUT_SECONDS = 60
134
_DEFAULT_POLL_SECONDS = 0.5
136
class LockDir(object):
137
"""Write-lock guarding access to data."""
139
__INFO_NAME = '/info'
141
def __init__(self, transport, path, file_modebits=0644, dir_modebits=0755):
142
"""Create a new LockDir object.
144
The LockDir is initially unlocked - this just creates the object.
146
:param transport: Transport which will contain the lock
148
:param path: Path to the lock within the base directory of the
151
assert isinstance(transport, Transport), \
152
("not a transport: %r" % transport)
153
self.transport = transport
155
self._lock_held = False
156
self._fake_read_lock = False
157
self._held_dir = path + '/held'
158
self._held_info_path = self._held_dir + self.__INFO_NAME
159
self._file_modebits = file_modebits
160
self._dir_modebits = dir_modebits
161
self.nonce = rand_chars(20)
163
self._report_function = note
166
return '%s(%s%s)' % (self.__class__.__name__,
170
is_held = property(lambda self: self._lock_held)
172
def create(self, mode=None):
173
"""Create the on-disk lock.
175
This is typically only called when the object/directory containing the
176
directory is first created. The lock is not held when it's created.
178
if self.transport.is_readonly():
179
raise UnlockableTransport(self.transport)
180
self.transport.mkdir(self.path, mode=mode)
182
def attempt_lock(self):
183
"""Take the lock; fail if it's already held.
185
If you wish to block until the lock can be obtained, call wait_lock()
188
if self._fake_read_lock:
189
raise LockContention(self)
190
if self.transport.is_readonly():
191
raise UnlockableTransport(self.transport)
193
tmpname = '%s/pending.%s.tmp' % (self.path, rand_chars(20))
195
self.transport.mkdir(tmpname)
197
# This may raise a FileExists exception
198
# which is okay, it will be caught later and determined
199
# to be a LockContention.
200
self.create(mode=self._dir_modebits)
202
# After creating the lock directory, try again
203
self.transport.mkdir(tmpname)
205
info_bytes = self._prepare_info()
206
# We use put_file_non_atomic because we just created a new unique
207
# directory so we don't have to worry about files existing there.
208
# We'll rename the whole directory into place to get atomic
210
self.transport.put_bytes_non_atomic(tmpname + self.__INFO_NAME,
213
self.transport.rename(tmpname, self._held_dir)
214
self._lock_held = True
216
except (PathError, DirectoryNotEmpty, FileExists, ResourceBusy), e:
217
mutter("contention on %r: %s", self, e)
218
raise LockContention(self)
221
"""Release a held lock
223
if self._fake_read_lock:
224
self._fake_read_lock = False
226
if not self._lock_held:
227
raise LockNotHeld(self)
228
# rename before deleting, because we can't atomically remove the whole
230
tmpname = '%s/releasing.%s.tmp' % (self.path, rand_chars(20))
231
# gotta own it to unlock
233
self.transport.rename(self._held_dir, tmpname)
234
self._lock_held = False
235
self.transport.delete(tmpname + self.__INFO_NAME)
236
self.transport.rmdir(tmpname)
238
def break_lock(self):
239
"""Break a lock not held by this instance of LockDir.
241
This is a UI centric function: it uses the bzrlib.ui.ui_factory to
242
prompt for input if a lock is detected and there is any doubt about
243
it possibly being still active.
245
self._check_not_locked()
246
holder_info = self.peek()
247
if holder_info is not None:
248
lock_info = '\n'.join(self._format_lock_info(holder_info))
249
if bzrlib.ui.ui_factory.get_boolean("Break %s" % lock_info):
250
self.force_break(holder_info)
252
def force_break(self, dead_holder_info):
253
"""Release a lock held by another process.
255
WARNING: This should only be used when the other process is dead; if
256
it still thinks it has the lock there will be two concurrent writers.
257
In general the user's approval should be sought for lock breaks.
259
dead_holder_info must be the result of a previous LockDir.peek() call;
260
this is used to check that it's still held by the same process that
261
the user decided was dead. If this is not the current holder,
262
LockBreakMismatch is raised.
264
After the lock is broken it will not be held by any process.
265
It is possible that another process may sneak in and take the
266
lock before the breaking process acquires it.
268
if not isinstance(dead_holder_info, dict):
269
raise ValueError("dead_holder_info: %r" % dead_holder_info)
270
self._check_not_locked()
271
current_info = self.peek()
272
if current_info is None:
273
# must have been recently released
275
if current_info != dead_holder_info:
276
raise LockBreakMismatch(self, current_info, dead_holder_info)
277
tmpname = '%s/broken.%s.tmp' % (self.path, rand_chars(20))
278
self.transport.rename(self._held_dir, tmpname)
279
# check that we actually broke the right lock, not someone else;
280
# there's a small race window between checking it and doing the
282
broken_info_path = tmpname + self.__INFO_NAME
283
broken_info = self._read_info_file(broken_info_path)
284
if broken_info != dead_holder_info:
285
raise LockBreakMismatch(self, broken_info, dead_holder_info)
286
self.transport.delete(broken_info_path)
287
self.transport.rmdir(tmpname)
289
def _check_not_locked(self):
290
"""If the lock is held by this instance, raise an error."""
292
raise AssertionError("can't break own lock: %r" % self)
295
"""Make sure that the lock is still held by this locker.
297
This should only fail if the lock was broken by user intervention,
298
or if the lock has been affected by a bug.
300
If the lock is not thought to be held, raises LockNotHeld. If
301
the lock is thought to be held but has been broken, raises
304
if not self._lock_held:
305
raise LockNotHeld(self)
308
# no lock there anymore!
309
raise LockBroken(self)
310
if info.get('nonce') != self.nonce:
311
# there is a lock, but not ours
312
raise LockBroken(self)
314
def _read_info_file(self, path):
315
"""Read one given info file.
317
peek() reads the info file of the lock holder, if any.
319
return self._parse_info(self.transport.get(path))
322
"""Check if the lock is held by anyone.
324
If it is held, this returns the lock info structure as a rio Stanza,
325
which contains some information about the current lock holder.
326
Otherwise returns None.
329
info = self._read_info_file(self._held_info_path)
330
assert isinstance(info, dict), \
331
"bad parse result %r" % info
333
except NoSuchFile, e:
336
def _prepare_info(self):
337
"""Write information about a pending lock to a temporary file.
340
# XXX: is creating this here inefficient?
341
config = bzrlib.config.GlobalConfig()
342
s = Stanza(hostname=socket.gethostname(),
343
pid=str(os.getpid()),
344
start_time=str(int(time.time())),
346
user=config.user_email(),
350
def _parse_info(self, info_file):
351
return read_stanza(info_file.readlines()).as_dict()
353
def wait_lock(self, timeout=_DEFAULT_TIMEOUT_SECONDS,
354
poll=_DEFAULT_POLL_SECONDS):
355
"""Wait a certain period for a lock.
357
If the lock can be acquired within the bounded time, it
358
is taken and this returns. Otherwise, LockContention
359
is raised. Either way, this function should return within
360
approximately `timeout` seconds. (It may be a bit more if
361
a transport operation takes a long time to complete.)
363
# XXX: the transport interface doesn't let us guard
364
# against operations there taking a long time.
365
deadline = time.time() + timeout
371
except LockContention:
375
self._report_function('Unable to obtain lock on %s\n'
376
'Will continue to try for %s seconds\n',
377
self.transport.abspath(self.path),
380
if time.time() + poll < deadline:
383
raise LockContention(self)
385
def lock_write(self):
386
"""Wait for and acquire the lock."""
390
"""Compatibility-mode shared lock.
392
LockDir doesn't support shared read-only locks, so this
393
just pretends that the lock is taken but really does nothing.
395
# At the moment Branches are commonly locked for read, but
396
# we can't rely on that remotely. Once this is cleaned up,
397
# reenable this warning to prevent it coming back in
399
## warn("LockDir.lock_read falls back to write lock")
400
if self._lock_held or self._fake_read_lock:
401
raise LockContention(self)
402
self._fake_read_lock = True
404
def wait(self, timeout=20, poll=0.5):
405
"""Wait a certain period for a lock to be released."""
406
# XXX: the transport interface doesn't let us guard
407
# against operations there taking a long time.
408
deadline = time.time() + timeout
412
if time.time() + poll < deadline:
415
raise LockContention(self)
417
def _format_lock_info(self, info):
418
"""Turn the contents of peek() into something for the user"""
419
lock_url = self.transport.abspath(self.path)
420
delta = time.time() - int(info['start_time'])
422
'lock %s' % (lock_url,),
423
'held by %(user)s on host %(hostname)s [process #%(pid)s]' % info,
424
'locked %s' % (format_delta(delta),),
added
removed
bzrlib/lockdir.py
