( )⚙ D11270 issue6528: also filter delta on the fly when applying a changegroup

This is an archive of the discontinued Mercurial Phabricator instance.

Differential D11270

issue6528: also filter delta on the fly when applying a changegroup
ClosedPublic

Authored by marmoute on Aug 7 2021, 10:07 AM.

Details

Reviewers
pulkit
Group Reviewers
hg-reviewers
Commits
rHGc30ca163b45e: issue6528: also filter delta on the fly when applying a changegroup
Summary

This ensure that corrupted clone does not spread corruption to "fixed" version.

This might come at a performance cost, we will had a config option to control
this behavior in the next changesets.

Diff Detail

Repository
rHG Mercurial
Lint
Automatic diff as part of commit; lint not applicable.
Unit
Automatic diff as part of commit; unit tests not applicable.

Event Timeline

marmoute created this revision.Aug 7 2021, 10:07 AM
baymax updated this revision to Diff 29861.Aug 8 2021, 5:22 AM

✅ refresh by Heptapod after a successful CI run (🐙 💚)
⚠ This patch is intended for stable ⚠

baymax updated this revision to Diff 29863.Aug 8 2021, 6:59 PM

✅ refresh by Heptapod after a successful CI run (🐙 💚)
⚠ This patch is intended for stable ⚠

pulkit accepted this revision.Aug 10 2021, 8:19 AM
This revision is now accepted and ready to land.Aug 10 2021, 8:19 AM

Revision Contents
Changeset List

Diff 29882

mercurial/filelog.py

) )
from .interfaces import ( from .interfaces import (
repository, repository,
util as interfaceutil, util as interfaceutil,
) )
from .utils import storageutil from .utils import storageutil
from .revlogutils import ( from .revlogutils import (
constants as revlog_constants, constants as revlog_constants,
rewrite,
) )
@interfaceutil.implementer(repository.ifilestorage) @interfaceutil.implementer(repository.ifilestorage)
class filelog(object): class filelog(object):
def __init__(self, opener, path): def __init__(self, opener, path):
self._revlog = revlog.revlog( self._revlog = revlog.revlog(
opener, opener,
raise error.Abort( raise error.Abort(
_( _(
b'revlog storage does not support missing ' b'revlog storage does not support missing '
b'parents write mode' b'parents write mode'
) )
) )
with self._revlog._writing(transaction): with self._revlog._writing(transaction):
deltas = rewrite.filter_delta_issue6528(self._revlog, deltas)
return self._revlog.addgroup( return self._revlog.addgroup(
deltas, deltas,
linkmapper, linkmapper,
transaction, transaction,
addrevisioncb=addrevisioncb, addrevisioncb=addrevisioncb,
duplicaterevisioncb=duplicaterevisioncb, duplicaterevisioncb=duplicaterevisioncb,
) )

mercurial/revlogutils/rewrite.py

ENTRY_DELTA_BASE, ENTRY_DELTA_BASE,
ENTRY_LINK_REV, ENTRY_LINK_REV,
ENTRY_NODE_ID, ENTRY_NODE_ID,
ENTRY_PARENT_1, ENTRY_PARENT_1,
ENTRY_PARENT_2, ENTRY_PARENT_2,
ENTRY_SIDEDATA_COMPRESSED_LENGTH, ENTRY_SIDEDATA_COMPRESSED_LENGTH,
ENTRY_SIDEDATA_COMPRESSION_MODE, ENTRY_SIDEDATA_COMPRESSION_MODE,
ENTRY_SIDEDATA_OFFSET, ENTRY_SIDEDATA_OFFSET,
REVIDX_ISCENSORED,
REVLOGV0, REVLOGV0,
REVLOGV1, REVLOGV1,
) )
from ..i18n import _ from ..i18n import _
from .. import ( from .. import (
error, error,
mdiff,
pycompat, pycompat,
revlogutils, revlogutils,
util, util,
) )
from ..utils import ( from ..utils import (
storageutil, storageutil,
) )
from . import ( from . import (
if not to_fix: if not to_fix:
msg = _(b"no affected revisions were found for '%s'\n") msg = _(b"no affected revisions were found for '%s'\n")
ui.write(msg % filename) ui.write(msg % filename)
continue continue
if not dry_run: if not dry_run:
_reorder_filelog_parents(repo, fl, sorted(to_fix)) _reorder_filelog_parents(repo, fl, sorted(to_fix))
def filter_delta_issue6528(revlog, deltas_iter):
"""filter incomind deltas to repaire issue 6528 on the fly"""
metadata_cache = {}
deltacomputer = deltas.deltacomputer(revlog)
for rev, d in enumerate(deltas_iter, len(revlog)):
(
node,
p1_node,
p2_node,
linknode,
deltabase,
delta,
flags,
sidedata,
) = d
if not revlog.index.has_node(deltabase):
raise error.LookupError(
deltabase, revlog.radix, _(b'unknown parent')
)
base_rev = revlog.rev(deltabase)
if not revlog.index.has_node(p1_node):
raise error.LookupError(p1_node, revlog.radix, _(b'unknown parent'))
p1_rev = revlog.rev(p1_node)
if not revlog.index.has_node(p2_node):
raise error.LookupError(p2_node, revlog.radix, _(b'unknown parent'))
p2_rev = revlog.rev(p2_node)
is_censored = lambda: bool(flags & REVIDX_ISCENSORED)
delta_base = lambda: revlog.rev(delta_base)
delta_base = lambda: base_rev
parent_revs = lambda: (p1_rev, p2_rev)
def full_text():
# note: being able to reuse the full text computation in the
# underlying addrevision would be useful however this is a bit too
# intrusive the for the "quick" issue6528 we are writing before the
# 5.8 release
textlen = mdiff.patchedsize(revlog.size(base_rev), delta)
revinfo = revlogutils.revisioninfo(
node,
p1_node,
p2_node,
[None],
textlen,
(base_rev, delta),
flags,
)
# cached by the global "writing" context
assert revlog._writinghandles is not None
if revlog._inline:
fh = revlog._writinghandles[0]
else:
fh = revlog._writinghandles[1]
return deltacomputer.buildtext(revinfo, fh)
is_affected = _is_revision_affected_fast_inner(
is_censored,
delta_base,
lambda: delta,
full_text,
parent_revs,
rev,
metadata_cache,
)
if is_affected:
d = (
node,
p2_node,
p1_node,
linknode,
deltabase,
delta,
flags,
sidedata,
)
yield d
def repair_issue6528( def repair_issue6528(
ui, repo, dry_run=False, to_report=None, from_report=None, paranoid=False ui, repo, dry_run=False, to_report=None, from_report=None, paranoid=False
): ):
from .. import store # avoid cycle from .. import store # avoid cycle
@contextlib.contextmanager @contextlib.contextmanager
def context(): def context():
if dry_run or to_report: # No need for locking if dry_run or to_report: # No need for locking

tests/bundles/issue6528.hg-v1

  • This binary file was added.

tests/bundles/issue6528.hg-v2

  • This binary file was added.

tests/test-issue6528.t

$ hg debugrevlogindex D.txt $ hg debugrevlogindex D.txt
rev linkrev nodeid p1 p2 rev linkrev nodeid p1 p2
0 6 2a8d3833f2fb 000000000000 000000000000 0 6 2a8d3833f2fb 000000000000 000000000000
1 7 2a80419dfc31 000000000000 2a8d3833f2fb 1 7 2a80419dfc31 000000000000 2a8d3833f2fb
2 8 65aecc89bb5d 2a80419dfc31 000000000000 2 8 65aecc89bb5d 2a80419dfc31 000000000000
$ hg debug-repair-issue6528 $ hg debug-repair-issue6528
no affected revisions were found no affected revisions were found
$ hg st $ hg st
$ cd ..
Applying a bad bundle should fix it on the fly
----------------------------------------------
from a v1 bundle
~~~~~~~~~~~~~~~~
$ hg debugbundle --spec "$TESTDIR"/bundles/issue6528.hg-v1
bzip2-v1
$ hg init unbundle-v1
$ cd unbundle-v1
$ hg unbundle "$TESTDIR"/bundles/issue6528.hg-v1
adding changesets
adding manifests
adding file changes
added 8 changesets with 12 changes to 4 files
new changesets f5a5a568022f:3beabb508514 (8 drafts)
(run 'hg update' to get a working copy)
Check that revision were fixed on the fly
$ hg debugrevlogindex b.txt
rev linkrev nodeid p1 p2
0 2 05b806ebe5ea 000000000000 000000000000
1 3 a58b36ad6b65 000000000000 05b806ebe5ea
2 6 216a5fe8b8ed 000000000000 000000000000
3 7 ea4f2f2463cc 000000000000 216a5fe8b8ed
$ hg debugrevlogindex D.txt
rev linkrev nodeid p1 p2
0 6 2a8d3833f2fb 000000000000 000000000000
1 7 2a80419dfc31 000000000000 2a8d3833f2fb
That we don't see the symptoms of the bug
$ hg up -- -1
2 files updated, 0 files merged, 0 files removed, 0 files unresolved
$ hg status
And that the repair command does not find anything to fix
$ hg debug-repair-issue6528
no affected revisions were found
$ cd ..
from a v2 bundle
~~~~~~~~~~~~~~~~
$ hg debugbundle --spec "$TESTDIR"/bundles/issue6528.hg-v2
bzip2-v2
$ hg init unbundle-v2
$ cd unbundle-v2
$ hg unbundle "$TESTDIR"/bundles/issue6528.hg-v2
adding changesets
adding manifests
adding file changes
added 8 changesets with 12 changes to 4 files
new changesets f5a5a568022f:3beabb508514 (8 drafts)
(run 'hg update' to get a working copy)
Check that revision were fixed on the fly
$ hg debugrevlogindex b.txt
rev linkrev nodeid p1 p2
0 2 05b806ebe5ea 000000000000 000000000000
1 3 a58b36ad6b65 000000000000 05b806ebe5ea
2 6 216a5fe8b8ed 000000000000 000000000000
3 7 ea4f2f2463cc 000000000000 216a5fe8b8ed
$ hg debugrevlogindex D.txt
rev linkrev nodeid p1 p2
0 6 2a8d3833f2fb 000000000000 000000000000
1 7 2a80419dfc31 000000000000 2a8d3833f2fb
That we don't see the symptoms of the bug
$ hg up -- -1
2 files updated, 0 files merged, 0 files removed, 0 files unresolved
$ hg status
And that the repair command does not find anything to fix
$ hg debug-repair-issue6528
no affected revisions were found
$ cd ..