Mercurial > hg
view tests/fakedirstatewritetime.py @ 46607:e9901d01d135
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
author | Kyle Lippincott <spectral@google.com> |
---|---|
date | Wed, 03 Feb 2021 16:33:10 -0800 |
parents | 89a2afe31e82 |
children | 8b7e47802deb |
line wrap: on
line source
# extension to emulate invoking 'dirstate.write()' at the time # specified by '[fakedirstatewritetime] fakenow', only when # 'dirstate.write()' is invoked via functions below: # # - 'workingctx._poststatusfixup()' (= 'repo.status()') # - 'committablectx.markcommitted()' from __future__ import absolute_import from mercurial import ( context, dirstate, extensions, policy, registrar, ) from mercurial.utils import dateutil try: from mercurial import rustext rustext.__name__ # force actual import (see hgdemandimport) except ImportError: rustext = None configtable = {} configitem = registrar.configitem(configtable) configitem( b'fakedirstatewritetime', b'fakenow', default=None, ) parsers = policy.importmod('parsers') rustmod = policy.importrust('parsers') def pack_dirstate(fakenow, orig, dmap, copymap, pl, now): # execute what original parsers.pack_dirstate should do actually # for consistency actualnow = int(now) for f, e in dmap.items(): if e[0] == 'n' and e[3] == actualnow: e = parsers.dirstatetuple(e[0], e[1], e[2], -1) dmap[f] = e return orig(dmap, copymap, pl, fakenow) def fakewrite(ui, func): # fake "now" of 'pack_dirstate' only if it is invoked while 'func' fakenow = ui.config(b'fakedirstatewritetime', b'fakenow') if not fakenow: # Execute original one, if fakenow isn't configured. This is # useful to prevent subrepos from executing replaced one, # because replacing 'parsers.pack_dirstate' is also effective # in subrepos. return func() # parsing 'fakenow' in YYYYmmddHHMM format makes comparison between # 'fakenow' value and 'touch -t YYYYmmddHHMM' argument easy fakenow = dateutil.parsedate(fakenow, [b'%Y%m%d%H%M'])[0] if rustmod is not None: # The Rust implementation does not use public parse/pack dirstate # to prevent conversion round-trips orig_dirstatemap_write = dirstate.dirstatemap.write wrapper = lambda self, st, now: orig_dirstatemap_write( self, st, fakenow ) dirstate.dirstatemap.write = wrapper orig_dirstate_getfsnow = dirstate._getfsnow wrapper = lambda *args: pack_dirstate(fakenow, orig_pack_dirstate, *args) orig_module = parsers orig_pack_dirstate = parsers.pack_dirstate orig_module.pack_dirstate = wrapper dirstate._getfsnow = lambda *args: fakenow try: return func() finally: orig_module.pack_dirstate = orig_pack_dirstate dirstate._getfsnow = orig_dirstate_getfsnow if rustmod is not None: dirstate.dirstatemap.write = orig_dirstatemap_write def _poststatusfixup(orig, workingctx, status, fixup): ui = workingctx.repo().ui return fakewrite(ui, lambda: orig(workingctx, status, fixup)) def markcommitted(orig, committablectx, node): ui = committablectx.repo().ui return fakewrite(ui, lambda: orig(committablectx, node)) def extsetup(ui): extensions.wrapfunction( context.workingctx, '_poststatusfixup', _poststatusfixup ) extensions.wrapfunction(context.workingctx, 'markcommitted', markcommitted)