##// END OF EJS Templates
revlog: add a mechanism to verify expected file position before appending...
revlog: add a mechanism to verify expected file position before appending If someone uses `hg debuglocks`, or some non-hg process writes to the .hg directory without respecting the locks, or if the repo's on a networked filesystem, it's possible for the revlog code to write out corrupted data. The form of this corruption can vary depending on what data was written and how that happened. We are in the "networked filesystem" case (though I've had users also do this to themselves with the "`hg debuglocks`" scenario), and most often see this with the changelog. What ends up happening is we produce two items (let's call them rev1 and rev2) in the .i file that have the same linkrev, baserev, and offset into the .d file, while the data in the .d file is appended properly. rev2's compressed_size is accurate for rev2, but when we go to decompress the data in the .d file, we use the offset that's recorded in the index file, which is the same as rev1, and attempt to decompress rev2.compressed_size bytes of rev1's data. This usually does not succeed. :) When using inline data, this also fails, though I haven't investigated why too closely. This shows up as a "patch decode" error. I believe what's happening there is that we're basically ignoring the offset field, getting the data properly, but since baserev != rev, it thinks this is a delta based on rev (instead of a full text) and can't actually apply it as such. For now, I'm going to make this an optional component and default it to entirely off. I may increase the default severity of this in the future, once I've enabled it for my users and we gain more experience with it. Luckily, most of my users have a versioned filesystem and can roll back to before the corruption has been written, it's just a hassle to do so and not everyone knows how (so it's a support burden). Users on other filesystems will not have that luxury, and this can cause them to have a corrupted repository that they are unlikely to know how to resolve, and they'll see this as a data-loss event. Refusing to create the corruption is a much better user experience. This mechanism is not perfect. There may be false-negatives (racy writes that are not detected). There should not be any false-positives (non-racy writes that are detected as such). This is not a mechanism that makes putting a repo on a networked filesystem "safe" or "supported", just *less* likely to cause corruption. Differential Revision: https://phab.mercurial-scm.org/D9952

File last commit:

r46554:89a2afe3 default
r47349:e9901d01 default
Show More
fetch.py
198 lines | 6.4 KiB | text/x-python | PythonLexer
# fetch.py - pull and merge remote changes
#
# Copyright 2006 Vadim Gelfer <vadim.gelfer@gmail.com>
#
# This software may be used and distributed according to the terms of the
# GNU General Public License version 2 or any later version.
'''pull, update and merge in one command (DEPRECATED)'''
from __future__ import absolute_import
from mercurial.i18n import _
from mercurial.node import short
from mercurial import (
cmdutil,
error,
exchange,
hg,
lock,
pycompat,
registrar,
util,
)
from mercurial.utils import dateutil
release = lock.release
cmdtable = {}
command = registrar.command(cmdtable)
# Note for extension authors: ONLY specify testedwith = 'ships-with-hg-core' for
# extensions which SHIP WITH MERCURIAL. Non-mainline extensions should
# be specifying the version(s) of Mercurial they are tested with, or
# leave the attribute unspecified.
testedwith = b'ships-with-hg-core'
@command(
b'fetch',
[
(
b'r',
b'rev',
[],
_(b'a specific revision you would like to pull'),
_(b'REV'),
),
(b'', b'edit', None, _(b'invoke editor on commit messages')),
(b'', b'force-editor', None, _(b'edit commit message (DEPRECATED)')),
(b'', b'switch-parent', None, _(b'switch parents when merging')),
]
+ cmdutil.commitopts
+ cmdutil.commitopts2
+ cmdutil.remoteopts,
_(b'hg fetch [SOURCE]'),
helpcategory=command.CATEGORY_REMOTE_REPO_MANAGEMENT,
)
def fetch(ui, repo, source=b'default', **opts):
"""pull changes from a remote repository, merge new changes if needed.
This finds all changes from the repository at the specified path
or URL and adds them to the local repository.
If the pulled changes add a new branch head, the head is
automatically merged, and the result of the merge is committed.
Otherwise, the working directory is updated to include the new
changes.
When a merge is needed, the working directory is first updated to
the newly pulled changes. Local changes are then merged into the
pulled changes. To switch the merge order, use --switch-parent.
See :hg:`help dates` for a list of formats valid for -d/--date.
Returns 0 on success.
"""
opts = pycompat.byteskwargs(opts)
date = opts.get(b'date')
if date:
opts[b'date'] = dateutil.parsedate(date)
parent = repo.dirstate.p1()
branch = repo.dirstate.branch()
try:
branchnode = repo.branchtip(branch)
except error.RepoLookupError:
branchnode = None
if parent != branchnode:
raise error.Abort(
_(b'working directory not at branch tip'),
hint=_(b"use 'hg update' to check out branch tip"),
)
wlock = lock = None
try:
wlock = repo.wlock()
lock = repo.lock()
cmdutil.bailifchanged(repo)
bheads = repo.branchheads(branch)
bheads = [head for head in bheads if len(repo[head].children()) == 0]
if len(bheads) > 1:
raise error.Abort(
_(
b'multiple heads in this branch '
b'(use "hg heads ." and "hg merge" to merge)'
)
)
other = hg.peer(repo, opts, ui.expandpath(source))
ui.status(
_(b'pulling from %s\n') % util.hidepassword(ui.expandpath(source))
)
revs = None
if opts[b'rev']:
try:
revs = [other.lookup(rev) for rev in opts[b'rev']]
except error.CapabilityError:
err = _(
b"other repository doesn't support revision lookup, "
b"so a rev cannot be specified."
)
raise error.Abort(err)
# Are there any changes at all?
modheads = exchange.pull(repo, other, heads=revs).cgresult
if modheads == 0:
return 0
# Is this a simple fast-forward along the current branch?
newheads = repo.branchheads(branch)
newchildren = repo.changelog.nodesbetween([parent], newheads)[2]
if len(newheads) == 1 and len(newchildren):
if newchildren[0] != parent:
return hg.update(repo, newchildren[0])
else:
return 0
# Are there more than one additional branch heads?
newchildren = [n for n in newchildren if n != parent]
newparent = parent
if newchildren:
newparent = newchildren[0]
hg.clean(repo, newparent)
newheads = [n for n in newheads if n != newparent]
if len(newheads) > 1:
ui.status(
_(
b'not merging with %d other new branch heads '
b'(use "hg heads ." and "hg merge" to merge them)\n'
)
% (len(newheads) - 1)
)
return 1
if not newheads:
return 0
# Otherwise, let's merge.
err = False
if newheads:
# By default, we consider the repository we're pulling
# *from* as authoritative, so we merge our changes into
# theirs.
if opts[b'switch_parent']:
firstparent, secondparent = newparent, newheads[0]
else:
firstparent, secondparent = newheads[0], newparent
ui.status(
_(b'updating to %d:%s\n')
% (repo.changelog.rev(firstparent), short(firstparent))
)
hg.clean(repo, firstparent)
p2ctx = repo[secondparent]
ui.status(
_(b'merging with %d:%s\n') % (p2ctx.rev(), short(secondparent))
)
err = hg.merge(p2ctx, remind=False)
if not err:
# we don't translate commit messages
message = cmdutil.logmessage(ui, opts) or (
b'Automated merge with %s' % util.removeauth(other.url())
)
editopt = opts.get(b'edit') or opts.get(b'force_editor')
editor = cmdutil.getcommiteditor(edit=editopt, editform=b'fetch')
n = repo.commit(
message, opts[b'user'], opts[b'date'], editor=editor
)
ui.status(
_(b'new changeset %d:%s merges remote changes with local\n')
% (repo.changelog.rev(n), short(n))
)
return err
finally:
release(lock, wlock)