dirstate: Remove the flat Rust DirstateMap implementation
Before this changeset we had two Rust implementations of `DirstateMap`.
This removes the "flat" DirstateMap so that the "tree" DirstateMap is always
used when Rust enabled. This simplifies the code a lot, and will enable
(in the next changeset) further removal of a trait abstraction.
This is a performance regression when:
* Rust is enabled, and
* The repository uses the legacy dirstate-v1 file format, and
* For `hg status`, unknown files are not listed (such as with `-mard`)
The regression is about 100 milliseconds for `hg status -mard` on a
semi-large repository (mozilla-central), from ~320ms to ~420ms.
We deem this to be small enough to be worth it.
The new dirstate-v2 is still experimental at this point, but we aim to
stabilize it (though not yet enable it by default for new repositories)
in Mercurial 6.0. Eventually, upgrating repositories to dirsate-v2 will
eliminate this regression (and enable other performance improvements).
# Background
The flat DirstateMap was introduced with the first Rust implementation of the
status algorithm. It works similarly to the previous Python + C one, with a
single `HashMap` that associates file paths to a `DirstateEntry` (where Python
has a dict).
We later added the tree DirstateMap where the root of the tree contains nodes
for files and directories that are directly at the root of the repository,
and nodes for directories can contain child nodes representing the files and
directly that *they* contain directly. The shape of this tree mirrors that of
the working directory in the filesystem. This enables the status algorithm to
traverse this tree in tandem with traversing the filesystem tree, which in
turns enables a more efficient algorithm.
Furthermore, the new dirstate-v2 file format is also based on a tree of the
same shape. The tree DirstateMap can access a dirstate-v2 file without parsing
it: binary data in a single large (possibly memory-mapped) bytes buffer is
traversed on demand. This allows `DirstateMap` creation to take `O(1)` time.
(Mutation works by creating new in-memory nodes with copy-on-write semantics,
and serialization is append-mostly.)
The tradeoff is that for "legacy" repositories that use the dirstate-v1 file
format, parsing that file into a tree DirstateMap takes more time. Profiling
shows that this time is dominated by `HashMap`. For a dirstate containing `F`
files with an average `D` directory depth, the flat DirstateMap does parsing
in `O(F)` number of HashMap operations but the tree DirstateMap in `O(F × D)`
operations, since each node has its own HashMap containing its child nodes.
This slower costs ~140ms on an old snapshot of mozilla-central, and ~80ms
on an old snapshot of the Netbeans repository.
The status algorithm is faster, but with `-mard` (when not listing unknown
files) it is typically not faster *enough* to compensate the slower parsing.
Both Rust implementations are always faster than the Python + C implementation
# Benchmark results
All benchmarks are run on changeset
98c0408324e6, with repositories that use
the dirstate-v1 file format, on a server with 4 CPU cores and 4 CPU threads
(no HyperThreading).
`hg status` benchmarks show wall clock times of the entire command as the
average and standard deviation of serveral runs, collected by
https://github.com/sharkdp/hyperfine and reformated.
Parsing benchmarks are wall clock time of the Rust function that converts a
bytes buffer of the dirstate file into the `DirstateMap` data structure as
used by the status algorithm. A single run each, collected by running
`hg status` this environment variable:
RUST_LOG=hg::dirstate::dirstate_map=trace,hg::dirstate_tree::dirstate_map=trace
Benchmark 1: Rust flat DirstateMap → Rust tree DirstateMap
hg status
mozilla-clean 562.3 ms ± 2.0 ms → 462.5 ms ± 0.6 ms 1.22 ± 0.00 times faster
mozilla-dirty 859.6 ms ± 2.2 ms → 719.5 ms ± 3.2 ms 1.19 ± 0.01 times faster
mozilla-ignored 558.2 ms ± 3.0 ms → 457.9 ms ± 2.9 ms 1.22 ± 0.01 times faster
mozilla-unknowns 859.4 ms ± 5.7 ms → 716.0 ms ± 4.7 ms 1.20 ± 0.01 times faster
netbeans-clean 336.5 ms ± 0.9 ms → 339.5 ms ± 0.4 ms 0.99 ± 0.00 times faster
netbeans-dirty 491.4 ms ± 1.6 ms → 475.1 ms ± 1.2 ms 1.03 ± 0.00 times faster
netbeans-ignored 343.7 ms ± 1.0 ms → 347.8 ms ± 0.4 ms 0.99 ± 0.00 times faster
netbeans-unknowns 484.3 ms ± 1.0 ms → 466.0 ms ± 1.2 ms 1.04 ± 0.00 times faster
hg status -mard
mozilla-clean 317.3 ms ± 0.6 ms → 422.5 ms ± 1.2 ms 0.75 ± 0.00 times faster
mozilla-dirty 315.4 ms ± 0.6 ms → 417.7 ms ± 1.1 ms 0.76 ± 0.00 times faster
mozilla-ignored 314.6 ms ± 0.6 ms → 417.4 ms ± 1.0 ms 0.75 ± 0.00 times faster
mozilla-unknowns 312.9 ms ± 0.9 ms → 417.3 ms ± 1.6 ms 0.75 ± 0.00 times faster
netbeans-clean 212.0 ms ± 0.6 ms → 283.6 ms ± 0.8 ms 0.75 ± 0.00 times faster
netbeans-dirty 211.4 ms ± 1.0 ms → 283.4 ms ± 1.6 ms 0.75 ± 0.01 times faster
netbeans-ignored 211.4 ms ± 0.9 ms → 283.9 ms ± 0.8 ms 0.74 ± 0.01 times faster
netbeans-unknowns 211.1 ms ± 0.6 ms → 283.4 ms ± 1.0 ms 0.74 ± 0.00 times faster
Parsing
mozilla-clean 38.4ms → 177.6ms
mozilla-dirty 38.8ms → 177.0ms
mozilla-ignored 38.8ms → 178.0ms
mozilla-unknowns 38.7ms → 176.9ms
netbeans-clean 16.5ms → 97.3ms
netbeans-dirty 16.5ms → 98.4ms
netbeans-ignored 16.9ms → 97.4ms
netbeans-unknowns 16.9ms → 96.3ms
Benchmark 2: Python + C dirstatemap → Rust tree DirstateMap
hg status
mozilla-clean 1261.0 ms ± 3.6 ms → 461.1 ms ± 0.5 ms 2.73 ± 0.00 times faster
mozilla-dirty 2293.4 ms ± 9.1 ms → 719.6 ms ± 3.6 ms 3.19 ± 0.01 times faster
mozilla-ignored 1240.4 ms ± 2.3 ms → 457.7 ms ± 1.9 ms 2.71 ± 0.00 times faster
mozilla-unknowns 2283.3 ms ± 9.0 ms → 719.7 ms ± 3.8 ms 3.17 ± 0.01 times faster
netbeans-clean 879.7 ms ± 3.5 ms → 339.9 ms ± 0.5 ms 2.59 ± 0.00 times faster
netbeans-dirty 1257.3 ms ± 4.7 ms → 474.6 ms ± 1.6 ms 2.65 ± 0.01 times faster
netbeans-ignored 943.9 ms ± 1.9 ms → 347.3 ms ± 1.1 ms 2.72 ± 0.00 times faster
netbeans-unknowns 1188.1 ms ± 5.0 ms → 465.2 ms ± 2.3 ms 2.55 ± 0.01 times faster
hg status -mard
mozilla-clean 903.2 ms ± 3.6 ms → 423.4 ms ± 2.2 ms 2.13 ± 0.01 times faster
mozilla-dirty 884.6 ms ± 4.5 ms → 417.3 ms ± 1.4 ms 2.12 ± 0.01 times faster
mozilla-ignored 881.9 ms ± 1.3 ms → 417.3 ms ± 0.8 ms 2.11 ± 0.00 times faster
mozilla-unknowns 878.5 ms ± 1.9 ms → 416.4 ms ± 0.9 ms 2.11 ± 0.00 times faster
netbeans-clean 434.9 ms ± 1.8 ms → 284.0 ms ± 0.8 ms 1.53 ± 0.01 times faster
netbeans-dirty 434.1 ms ± 0.8 ms → 283.1 ms ± 0.8 ms 1.53 ± 0.00 times faster
netbeans-ignored 431.7 ms ± 1.1 ms → 283.6 ms ± 1.8 ms 1.52 ± 0.01 times faster
netbeans-unknowns 433.0 ms ± 1.3 ms → 283.5 ms ± 0.7 ms 1.53 ± 0.00 times faster
Differential Revision: https://phab.mercurial-scm.org/D11516
# upgrade.py - functions for in place upgrade of Mercurial repository
#
# Copyright (c) 2016-present, Gregory Szorc
#
# This software may be used and distributed according to the terms of the
# GNU General Public License version 2 or any later version.
from __future__ import absolute_import
import stat
from ..i18n import _
from ..pycompat import getattr
from .. import (
changelog,
error,
filelog,
manifest,
metadata,
pycompat,
requirements,
scmutil,
store,
util,
vfs as vfsmod,
)
from ..revlogutils import (
constants as revlogconst,
flagutil,
nodemap,
sidedata as sidedatamod,
)
from . import actions as upgrade_actions
def get_sidedata_helpers(srcrepo, dstrepo):
use_w = srcrepo.ui.configbool(b'experimental', b'worker.repository-upgrade')
sequential = pycompat.iswindows or not use_w
if not sequential:
srcrepo.register_sidedata_computer(
revlogconst.KIND_CHANGELOG,
sidedatamod.SD_FILES,
(sidedatamod.SD_FILES,),
metadata._get_worker_sidedata_adder(srcrepo, dstrepo),
flagutil.REVIDX_HASCOPIESINFO,
replace=True,
)
return sidedatamod.get_sidedata_helpers(srcrepo, dstrepo._wanted_sidedata)
def _revlogfrompath(repo, rl_type, path):
"""Obtain a revlog from a repo path.
An instance of the appropriate class is returned.
"""
if rl_type & store.FILEFLAGS_CHANGELOG:
return changelog.changelog(repo.svfs)
elif rl_type & store.FILEFLAGS_MANIFESTLOG:
mandir = b''
if b'/' in path:
mandir = path.rsplit(b'/', 1)[0]
return manifest.manifestrevlog(
repo.nodeconstants, repo.svfs, tree=mandir
)
else:
# drop the extension and the `data/` prefix
path_part = path.rsplit(b'.', 1)[0].split(b'/', 1)
if len(path_part) < 2:
msg = _(b'cannot recognize revlog from filename: %s')
msg %= path
raise error.Abort(msg)
path = path_part[1]
return filelog.filelog(repo.svfs, path)
def _copyrevlog(tr, destrepo, oldrl, rl_type, unencodedname):
"""copy all relevant files for `oldrl` into `destrepo` store
Files are copied "as is" without any transformation. The copy is performed
without extra checks. Callers are responsible for making sure the copied
content is compatible with format of the destination repository.
"""
oldrl = getattr(oldrl, '_revlog', oldrl)
newrl = _revlogfrompath(destrepo, rl_type, unencodedname)
newrl = getattr(newrl, '_revlog', newrl)
oldvfs = oldrl.opener
newvfs = newrl.opener
oldindex = oldvfs.join(oldrl._indexfile)
newindex = newvfs.join(newrl._indexfile)
olddata = oldvfs.join(oldrl._datafile)
newdata = newvfs.join(newrl._datafile)
with newvfs(newrl._indexfile, b'w'):
pass # create all the directories
util.copyfile(oldindex, newindex)
copydata = oldrl.opener.exists(oldrl._datafile)
if copydata:
util.copyfile(olddata, newdata)
if rl_type & store.FILEFLAGS_FILELOG:
destrepo.svfs.fncache.add(unencodedname)
if copydata:
destrepo.svfs.fncache.add(unencodedname[:-2] + b'.d')
UPGRADE_CHANGELOG = b"changelog"
UPGRADE_MANIFEST = b"manifest"
UPGRADE_FILELOGS = b"all-filelogs"
UPGRADE_ALL_REVLOGS = frozenset(
[UPGRADE_CHANGELOG, UPGRADE_MANIFEST, UPGRADE_FILELOGS]
)
def matchrevlog(revlogfilter, rl_type):
"""check if a revlog is selected for cloning.
In other words, are there any updates which need to be done on revlog
or it can be blindly copied.
The store entry is checked against the passed filter"""
if rl_type & store.FILEFLAGS_CHANGELOG:
return UPGRADE_CHANGELOG in revlogfilter
elif rl_type & store.FILEFLAGS_MANIFESTLOG:
return UPGRADE_MANIFEST in revlogfilter
assert rl_type & store.FILEFLAGS_FILELOG
return UPGRADE_FILELOGS in revlogfilter
def _perform_clone(
ui,
dstrepo,
tr,
old_revlog,
rl_type,
unencoded,
upgrade_op,
sidedata_helpers,
oncopiedrevision,
):
"""returns the new revlog object created"""
newrl = None
if matchrevlog(upgrade_op.revlogs_to_process, rl_type):
ui.note(
_(b'cloning %d revisions from %s\n') % (len(old_revlog), unencoded)
)
newrl = _revlogfrompath(dstrepo, rl_type, unencoded)
old_revlog.clone(
tr,
newrl,
addrevisioncb=oncopiedrevision,
deltareuse=upgrade_op.delta_reuse_mode,
forcedeltabothparents=upgrade_op.force_re_delta_both_parents,
sidedata_helpers=sidedata_helpers,
)
else:
msg = _(b'blindly copying %s containing %i revisions\n')
ui.note(msg % (unencoded, len(old_revlog)))
_copyrevlog(tr, dstrepo, old_revlog, rl_type, unencoded)
newrl = _revlogfrompath(dstrepo, rl_type, unencoded)
return newrl
def _clonerevlogs(
ui,
srcrepo,
dstrepo,
tr,
upgrade_op,
):
"""Copy revlogs between 2 repos."""
revcount = 0
srcsize = 0
srcrawsize = 0
dstsize = 0
fcount = 0
frevcount = 0
fsrcsize = 0
frawsize = 0
fdstsize = 0
mcount = 0
mrevcount = 0
msrcsize = 0
mrawsize = 0
mdstsize = 0
crevcount = 0
csrcsize = 0
crawsize = 0
cdstsize = 0
alldatafiles = list(srcrepo.store.walk())
# mapping of data files which needs to be cloned
# key is unencoded filename
# value is revlog_object_from_srcrepo
manifests = {}
changelogs = {}
filelogs = {}
# Perform a pass to collect metadata. This validates we can open all
# source files and allows a unified progress bar to be displayed.
for rl_type, unencoded, size in alldatafiles:
if not rl_type & store.FILEFLAGS_REVLOG_MAIN:
continue
# the store.walk function will wrongly pickup transaction backup and
# get confused. As a quick fix for 5.9 release, we ignore those.
# (this is not a module constants because it seems better to keep the
# hack together)
skip_undo = (
b'undo.backup.00changelog.i',
b'undo.backup.00manifest.i',
)
if unencoded in skip_undo:
continue
rl = _revlogfrompath(srcrepo, rl_type, unencoded)
info = rl.storageinfo(
exclusivefiles=True,
revisionscount=True,
trackedsize=True,
storedsize=True,
)
revcount += info[b'revisionscount'] or 0
datasize = info[b'storedsize'] or 0
rawsize = info[b'trackedsize'] or 0
srcsize += datasize
srcrawsize += rawsize
# This is for the separate progress bars.
if rl_type & store.FILEFLAGS_CHANGELOG:
changelogs[unencoded] = (rl_type, rl)
crevcount += len(rl)
csrcsize += datasize
crawsize += rawsize
elif rl_type & store.FILEFLAGS_MANIFESTLOG:
manifests[unencoded] = (rl_type, rl)
mcount += 1
mrevcount += len(rl)
msrcsize += datasize
mrawsize += rawsize
elif rl_type & store.FILEFLAGS_FILELOG:
filelogs[unencoded] = (rl_type, rl)
fcount += 1
frevcount += len(rl)
fsrcsize += datasize
frawsize += rawsize
else:
error.ProgrammingError(b'unknown revlog type')
if not revcount:
return
ui.status(
_(
b'migrating %d total revisions (%d in filelogs, %d in manifests, '
b'%d in changelog)\n'
)
% (revcount, frevcount, mrevcount, crevcount)
)
ui.status(
_(b'migrating %s in store; %s tracked data\n')
% ((util.bytecount(srcsize), util.bytecount(srcrawsize)))
)
# Used to keep track of progress.
progress = None
def oncopiedrevision(rl, rev, node):
progress.increment()
sidedata_helpers = get_sidedata_helpers(srcrepo, dstrepo)
# Migrating filelogs
ui.status(
_(
b'migrating %d filelogs containing %d revisions '
b'(%s in store; %s tracked data)\n'
)
% (
fcount,
frevcount,
util.bytecount(fsrcsize),
util.bytecount(frawsize),
)
)
progress = srcrepo.ui.makeprogress(_(b'file revisions'), total=frevcount)
for unencoded, (rl_type, oldrl) in sorted(filelogs.items()):
newrl = _perform_clone(
ui,
dstrepo,
tr,
oldrl,
rl_type,
unencoded,
upgrade_op,
sidedata_helpers,
oncopiedrevision,
)
info = newrl.storageinfo(storedsize=True)
fdstsize += info[b'storedsize'] or 0
ui.status(
_(
b'finished migrating %d filelog revisions across %d '
b'filelogs; change in size: %s\n'
)
% (frevcount, fcount, util.bytecount(fdstsize - fsrcsize))
)
# Migrating manifests
ui.status(
_(
b'migrating %d manifests containing %d revisions '
b'(%s in store; %s tracked data)\n'
)
% (
mcount,
mrevcount,
util.bytecount(msrcsize),
util.bytecount(mrawsize),
)
)
if progress:
progress.complete()
progress = srcrepo.ui.makeprogress(
_(b'manifest revisions'), total=mrevcount
)
for unencoded, (rl_type, oldrl) in sorted(manifests.items()):
newrl = _perform_clone(
ui,
dstrepo,
tr,
oldrl,
rl_type,
unencoded,
upgrade_op,
sidedata_helpers,
oncopiedrevision,
)
info = newrl.storageinfo(storedsize=True)
mdstsize += info[b'storedsize'] or 0
ui.status(
_(
b'finished migrating %d manifest revisions across %d '
b'manifests; change in size: %s\n'
)
% (mrevcount, mcount, util.bytecount(mdstsize - msrcsize))
)
# Migrating changelog
ui.status(
_(
b'migrating changelog containing %d revisions '
b'(%s in store; %s tracked data)\n'
)
% (
crevcount,
util.bytecount(csrcsize),
util.bytecount(crawsize),
)
)
if progress:
progress.complete()
progress = srcrepo.ui.makeprogress(
_(b'changelog revisions'), total=crevcount
)
for unencoded, (rl_type, oldrl) in sorted(changelogs.items()):
newrl = _perform_clone(
ui,
dstrepo,
tr,
oldrl,
rl_type,
unencoded,
upgrade_op,
sidedata_helpers,
oncopiedrevision,
)
info = newrl.storageinfo(storedsize=True)
cdstsize += info[b'storedsize'] or 0
progress.complete()
ui.status(
_(
b'finished migrating %d changelog revisions; change in size: '
b'%s\n'
)
% (crevcount, util.bytecount(cdstsize - csrcsize))
)
dstsize = fdstsize + mdstsize + cdstsize
ui.status(
_(
b'finished migrating %d total revisions; total change in store '
b'size: %s\n'
)
% (revcount, util.bytecount(dstsize - srcsize))
)
def _files_to_copy_post_revlog_clone(srcrepo):
"""yields files which should be copied to destination after revlogs
are cloned"""
for path, kind, st in sorted(srcrepo.store.vfs.readdir(b'', stat=True)):
# don't copy revlogs as they are already cloned
if store.revlog_type(path) is not None:
continue
# Skip transaction related files.
if path.startswith(b'undo'):
continue
# Only copy regular files.
if kind != stat.S_IFREG:
continue
# Skip other skipped files.
if path in (b'lock', b'fncache'):
continue
# TODO: should we skip cache too?
yield path
def _replacestores(currentrepo, upgradedrepo, backupvfs, upgrade_op):
"""Replace the stores after current repository is upgraded
Creates a backup of current repository store at backup path
Replaces upgraded store files in current repo from upgraded one
Arguments:
currentrepo: repo object of current repository
upgradedrepo: repo object of the upgraded data
backupvfs: vfs object for the backup path
upgrade_op: upgrade operation object
to be used to decide what all is upgraded
"""
# TODO: don't blindly rename everything in store
# There can be upgrades where store is not touched at all
if upgrade_op.backup_store:
util.rename(currentrepo.spath, backupvfs.join(b'store'))
else:
currentrepo.vfs.rmtree(b'store', forcibly=True)
util.rename(upgradedrepo.spath, currentrepo.spath)
def finishdatamigration(ui, srcrepo, dstrepo, requirements):
"""Hook point for extensions to perform additional actions during upgrade.
This function is called after revlogs and store files have been copied but
before the new store is swapped into the original location.
"""
def upgrade(ui, srcrepo, dstrepo, upgrade_op):
"""Do the low-level work of upgrading a repository.
The upgrade is effectively performed as a copy between a source
repository and a temporary destination repository.
The source repository is unmodified for as long as possible so the
upgrade can abort at any time without causing loss of service for
readers and without corrupting the source repository.
"""
assert srcrepo.currentwlock()
assert dstrepo.currentwlock()
backuppath = None
backupvfs = None
ui.status(
_(
b'(it is safe to interrupt this process any time before '
b'data migration completes)\n'
)
)
if upgrade_actions.dirstatev2 in upgrade_op.upgrade_actions:
ui.status(_(b'upgrading to dirstate-v2 from v1\n'))
upgrade_dirstate(ui, srcrepo, upgrade_op, b'v1', b'v2')
upgrade_op.upgrade_actions.remove(upgrade_actions.dirstatev2)
if upgrade_actions.dirstatev2 in upgrade_op.removed_actions:
ui.status(_(b'downgrading from dirstate-v2 to v1\n'))
upgrade_dirstate(ui, srcrepo, upgrade_op, b'v2', b'v1')
upgrade_op.removed_actions.remove(upgrade_actions.dirstatev2)
if not (upgrade_op.upgrade_actions or upgrade_op.removed_actions):
return
if upgrade_op.requirements_only:
ui.status(_(b'upgrading repository requirements\n'))
scmutil.writereporequirements(srcrepo, upgrade_op.new_requirements)
# if there is only one action and that is persistent nodemap upgrade
# directly write the nodemap file and update requirements instead of going
# through the whole cloning process
elif (
len(upgrade_op.upgrade_actions) == 1
and b'persistent-nodemap' in upgrade_op.upgrade_actions_names
and not upgrade_op.removed_actions
):
ui.status(
_(b'upgrading repository to use persistent nodemap feature\n')
)
with srcrepo.transaction(b'upgrade') as tr:
unfi = srcrepo.unfiltered()
cl = unfi.changelog
nodemap.persist_nodemap(tr, cl, force=True)
# we want to directly operate on the underlying revlog to force
# create a nodemap file. This is fine since this is upgrade code
# and it heavily relies on repository being revlog based
# hence accessing private attributes can be justified
nodemap.persist_nodemap(
tr, unfi.manifestlog._rootstore._revlog, force=True
)
scmutil.writereporequirements(srcrepo, upgrade_op.new_requirements)
elif (
len(upgrade_op.removed_actions) == 1
and [
x
for x in upgrade_op.removed_actions
if x.name == b'persistent-nodemap'
]
and not upgrade_op.upgrade_actions
):
ui.status(
_(b'downgrading repository to not use persistent nodemap feature\n')
)
with srcrepo.transaction(b'upgrade') as tr:
unfi = srcrepo.unfiltered()
cl = unfi.changelog
nodemap.delete_nodemap(tr, srcrepo, cl)
# check comment 20 lines above for accessing private attributes
nodemap.delete_nodemap(
tr, srcrepo, unfi.manifestlog._rootstore._revlog
)
scmutil.writereporequirements(srcrepo, upgrade_op.new_requirements)
else:
with dstrepo.transaction(b'upgrade') as tr:
_clonerevlogs(
ui,
srcrepo,
dstrepo,
tr,
upgrade_op,
)
# Now copy other files in the store directory.
for p in _files_to_copy_post_revlog_clone(srcrepo):
srcrepo.ui.status(_(b'copying %s\n') % p)
src = srcrepo.store.rawvfs.join(p)
dst = dstrepo.store.rawvfs.join(p)
util.copyfile(src, dst, copystat=True)
finishdatamigration(ui, srcrepo, dstrepo, requirements)
ui.status(_(b'data fully upgraded in a temporary repository\n'))
if upgrade_op.backup_store:
backuppath = pycompat.mkdtemp(
prefix=b'upgradebackup.', dir=srcrepo.path
)
backupvfs = vfsmod.vfs(backuppath)
# Make a backup of requires file first, as it is the first to be modified.
util.copyfile(
srcrepo.vfs.join(b'requires'), backupvfs.join(b'requires')
)
# We install an arbitrary requirement that clients must not support
# as a mechanism to lock out new clients during the data swap. This is
# better than allowing a client to continue while the repository is in
# an inconsistent state.
ui.status(
_(
b'marking source repository as being upgraded; clients will be '
b'unable to read from repository\n'
)
)
scmutil.writereporequirements(
srcrepo, srcrepo.requirements | {b'upgradeinprogress'}
)
ui.status(_(b'starting in-place swap of repository data\n'))
if upgrade_op.backup_store:
ui.status(
_(b'replaced files will be backed up at %s\n') % backuppath
)
# Now swap in the new store directory. Doing it as a rename should make
# the operation nearly instantaneous and atomic (at least in well-behaved
# environments).
ui.status(_(b'replacing store...\n'))
tstart = util.timer()
_replacestores(srcrepo, dstrepo, backupvfs, upgrade_op)
elapsed = util.timer() - tstart
ui.status(
_(
b'store replacement complete; repository was inconsistent for '
b'%0.1fs\n'
)
% elapsed
)
# We first write the requirements file. Any new requirements will lock
# out legacy clients.
ui.status(
_(
b'finalizing requirements file and making repository readable '
b'again\n'
)
)
scmutil.writereporequirements(srcrepo, upgrade_op.new_requirements)
if upgrade_op.backup_store:
# The lock file from the old store won't be removed because nothing has a
# reference to its new location. So clean it up manually. Alternatively, we
# could update srcrepo.svfs and other variables to point to the new
# location. This is simpler.
assert backupvfs is not None # help pytype
backupvfs.unlink(b'store/lock')
return backuppath
def upgrade_dirstate(ui, srcrepo, upgrade_op, old, new):
if upgrade_op.backup_store:
backuppath = pycompat.mkdtemp(
prefix=b'upgradebackup.', dir=srcrepo.path
)
ui.status(_(b'replaced files will be backed up at %s\n') % backuppath)
backupvfs = vfsmod.vfs(backuppath)
util.copyfile(
srcrepo.vfs.join(b'requires'), backupvfs.join(b'requires')
)
util.copyfile(
srcrepo.vfs.join(b'dirstate'), backupvfs.join(b'dirstate')
)
assert srcrepo.dirstate._use_dirstate_v2 == (old == b'v2')
srcrepo.dirstate._map.preload()
srcrepo.dirstate._use_dirstate_v2 = new == b'v2'
srcrepo.dirstate._map._use_dirstate_v2 = srcrepo.dirstate._use_dirstate_v2
srcrepo.dirstate._dirty = True
srcrepo.vfs.unlink(b'dirstate')
srcrepo.dirstate.write(None)
scmutil.writereporequirements(srcrepo, upgrade_op.new_requirements)