dirstate: avoid a race with multiple commits in the same process

(issue2264, issue2516)

The race happens when two commits in a row change the same file
without changing its size, *if* those two commits happen in the same
second in the same process while holding the same repo lock.  For
example:

  commit 1:
    M a
    M b
  commit 2:           # same process, same second, same repo lock
    M b               # modify b without changing its size
    M c

This first manifested in transplant, which is the most common way to
do multiple commits in the same process. But it can manifest in any
script or extension that does multiple commits under the same repo
lock. (Thus, the test script tests both transplant and a custom script.)

The problem was that dirstate.status() failed to notice the change to
b when localrepo is about to do the second commit, meaning that change
gets left in the working directory. In the context of transplant, that
means either a crash ("RuntimeError: nothing committed after
transplant") or a silently inaccurate transplant, depending on whether
any other files were modified by the second transplanted changeset.

The fix is to make status() work a little harder when we have
previously marked files as clean (state 'normal') in the same process.
Specifically, dirstate.normal() adds files to self._lastnormal, and
other state-changing methods remove them. Then dirstate.status() puts
any files in self._lastnormal into state 'lookup', which will make
localrepository.status() read file contents to see if it has really
changed.  So we pay a small performance penalty for the second (and
subsequent) commits in the same process, without affecting the common
case.  Anything that does lots of status updates and checks in the
same process could suffer a performance hit.

Incidentally, there is a simpler fix: call dirstate.normallookup() on
every file updated by commit() at the end of the commit.  The trouble
with that solution is that it imposes a performance penalty on the
common case: it means the next status-dependent hg command after every
"hg commit" will be a little bit slower.  The patch here is more
complex, but only affects performance for the uncommon case.

tests/test-commit-multiple.t

@@ -0,0 +1,119
+# reproduce issue2264, issue2516
+
+create test repo
+  $ cat <<EOF >> $HGRCPATH
+  > [extensions]
+  > transplant =
+  > graphlog =
+  > EOF
+  $ hg init repo
+  $ cd repo
+  $ template="{rev}  {desc|firstline}  [{branch}]\n"
+
+# we need to start out with two changesets on the default branch
+# in order to avoid the cute little optimization where transplant
+# pulls rather than transplants
+add initial changesets
+  $ echo feature1 > file1
+  $ hg ci -Am"feature 1"
+  adding file1
+  $ echo feature2 >> file2
+  $ hg ci -Am"feature 2"
+  adding file2
+
+# The changes to 'bugfix' are enough to show the bug: in fact, with only
+# those changes, it's a very noisy crash ("RuntimeError: nothing
+# committed after transplant").  But if we modify a second file in the
+# transplanted changesets, the bug is much more subtle: transplant
+# silently drops the second change to 'bugfix' on the floor, and we only
+# see it when we run 'hg status' after transplanting.  Subtle data loss
+# bugs are worse than crashes, so reproduce the subtle case here.
+commit bug fixes on bug fix branch
+  $ hg branch fixes
+  marked working directory as branch fixes
+  $ echo fix1 > bugfix
+  $ echo fix1 >> file1
+  $ hg ci -Am"fix 1"
+  adding bugfix
+  $ echo fix2 > bugfix
+  $ echo fix2 >> file1
+  $ hg ci -Am"fix 2"
+  $ hg glog --template="$template"
+  @  3  fix 2  [fixes]
+  |
+  o  2  fix 1  [fixes]
+  |
+  o  1  feature 2  [default]
+  |
+  o  0  feature 1  [default]
+  
+transplant bug fixes onto release branch
+  $ hg update 0
+  1 files updated, 0 files merged, 2 files removed, 0 files unresolved
+  $ hg branch release
+  marked working directory as branch release
+  $ hg transplant 2 3
+  applying [0-9a-f]{12} (re)
+  [0-9a-f]{12} transplanted to [0-9a-f]{12} (re)
+  applying [0-9a-f]{12} (re)
+  [0-9a-f]{12} transplanted to [0-9a-f]{12} (re)
+  $ hg glog --template="$template"
+  @  5  fix 2  [release]
+  |
+  o  4  fix 1  [release]
+  |
+  | o  3  fix 2  [fixes]
+  | |
+  | o  2  fix 1  [fixes]
+  | |
+  | o  1  feature 2  [default]
+  |/
+  o  0  feature 1  [default]
+  
+  $ hg status
+  $ hg status --rev 0:4
+  M file1
+  A bugfix
+  $ hg status --rev 4:5
+  M bugfix
+  M file1
+
+now test that we fixed the bug for all scripts/extensions
+  $ cat > $TESTTMP/committwice.py <<__EOF__
+  > from mercurial import ui, hg, match, node
+  > 
+  > def replacebyte(fn, b):
+  >     f = open("file1", "rb+")
+  >     f.seek(0, 0)
+  >     f.write(b)
+  >     f.close()
+  > 
+  > repo = hg.repository(ui.ui(), '.')
+  > assert len(repo) == 6, \
+  >        "initial: len(repo) == %d, expected 6" % len(repo)
+  > try:
+  >     wlock = repo.wlock()
+  >     lock = repo.lock()
+  >     m = match.exact(repo.root, '', ['file1'])
+  >     replacebyte("file1", "x")
+  >     n = repo.commit(text="x", user="test", date=(0, 0), match=m)
+  >     print "commit 1: len(repo) == %d" % len(repo)
+  >     replacebyte("file1", "y")
+  >     n = repo.commit(text="y", user="test", date=(0, 0), match=m)
+  >     print "commit 2: len(repo) == %d" % len(repo)
+  > finally:
+  >     lock.release()
+  >     wlock.release()
+  > __EOF__
+  $ $PYTHON $TESTTMP/committwice.py
+  commit 1: len(repo) == 7
+  commit 2: len(repo) == 8
+
+Do a size-preserving modification outside of that process
+  $ echo abcd > bugfix
+  $ hg status
+  M bugfix
+  $ hg log --template "{rev}  {desc}  {files}\n" -r5:
+  5  fix 2  bugfix file1
+  6  x  file1
+  7  y  file1

mercurial/dirstate.py

@@ -1,685 +1,711
 # dirstate.py - working directory tracking for mercurial
 #
 # Copyright 2005-2007 Matt Mackall <[e-mail]>
 #
 # This software may be used and distributed according to the terms of the
 # GNU General Public License version 2 or any later version.
 
 from node import nullid
 from i18n import _
 import util, ignore, osutil, parsers, encoding
 import struct, os, stat, errno
 import cStringIO
 
 _format = ">cllll"
 propertycache = util.propertycache
 
 def _finddirs(path):
     pos = path.rfind('/')
     while pos != -1:
         yield path[:pos]
         pos = path.rfind('/', 0, pos)
 
 def _incdirs(dirs, path):
     for base in _finddirs(path):
         if base in dirs:
             dirs[base] += 1
             return
         dirs[base] = 1
 
 def _decdirs(dirs, path):
     for base in _finddirs(path):
         if dirs[base] > 1:
             dirs[base] -= 1
             return
         del dirs[base]
 
 class dirstate(object):
 
     def __init__(self, opener, ui, root, validate):
         '''Create a new dirstate object.
 
         opener is an open()-like callable that can be used to open the
         dirstate file; root is the root of the directory tracked by
         the dirstate.
         '''
         self._opener = opener
         self._validate = validate
         self._root = root
         self._rootdir = os.path.join(root, '')
         self._dirty = False
         self._dirtypl = False
+        self._lastnormal = set()        # files believed to be normal
         self._ui = ui
 
     @propertycache
     def _map(self):
         '''Return the dirstate contents as a map from filename to
         (state, mode, size, time).'''
         self._read()
         return self._map
 
     @propertycache
     def _copymap(self):
         self._read()
         return self._copymap
 
     @propertycache
     def _foldmap(self):
         f = {}
         for name in self._map:
             f[os.path.normcase(name)] = name
         return f
 
     @propertycache
     def _branch(self):
         try:
             return self._opener("branch").read().strip() or "default"
         except IOError:
             return "default"
 
     @propertycache
     def _pl(self):
         try:
             fp = self._opener("dirstate")
             st = fp.read(40)
             fp.close()
             l = len(st)
             if l == 40:
                 return st[:20], st[20:40]
             elif l > 0 and l < 40:
                 raise util.Abort(_('working directory state appears damaged!'))
         except IOError, err:
             if err.errno != errno.ENOENT:
                 raise
         return [nullid, nullid]
 
     @propertycache
     def _dirs(self):
         dirs = {}
         for f, s in self._map.iteritems():
             if s[0] != 'r':
                 _incdirs(dirs, f)
         return dirs
 
     @propertycache
     def _ignore(self):
         files = [self._join('.hgignore')]
         for name, path in self._ui.configitems("ui"):
             if name == 'ignore' or name.startswith('ignore.'):
                 files.append(util.expandpath(path))
         return ignore.ignore(self._root, files, self._ui.warn)
 
     @propertycache
     def _slash(self):
         return self._ui.configbool('ui', 'slash') and os.sep != '/'
 
     @propertycache
     def _checklink(self):
         return util.checklink(self._root)
 
     @propertycache
     def _checkexec(self):
         return util.checkexec(self._root)
 
     @propertycache
     def _checkcase(self):
         return not util.checkcase(self._join('.hg'))
 
     def _join(self, f):
         # much faster than os.path.join()
         # it's safe because f is always a relative path
         return self._rootdir + f
 
     def flagfunc(self, fallback):
         if self._checklink:
             if self._checkexec:
                 def f(x):
                     p = self._join(x)
                     if os.path.islink(p):
                         return 'l'
                     if util.is_exec(p):
                         return 'x'
                     return ''
                 return f
             def f(x):
                 if os.path.islink(self._join(x)):
                     return 'l'
                 if 'x' in fallback(x):
                     return 'x'
                 return ''
             return f
         if self._checkexec:
             def f(x):
                 if 'l' in fallback(x):
                     return 'l'
                 if util.is_exec(self._join(x)):
                     return 'x'
                 return ''
             return f
         return fallback
 
     def getcwd(self):
         cwd = os.getcwd()
         if cwd == self._root:
             return ''
         # self._root ends with a path separator if self._root is '/' or 'C:\'
         rootsep = self._root
         if not util.endswithsep(rootsep):
             rootsep += os.sep
         if cwd.startswith(rootsep):
             return cwd[len(rootsep):]
         else:
             # we're outside the repo. return an absolute path.
             return cwd
 
     def pathto(self, f, cwd=None):
         if cwd is None:
             cwd = self.getcwd()
         path = util.pathto(self._root, cwd, f)
         if self._slash:
             return util.normpath(path)
         return path
 
     def __getitem__(self, key):
         '''Return the current state of key (a filename) in the dirstate.
 
         States are:
           n  normal
           m  needs merging
           r  marked for removal
           a  marked for addition
           ?  not tracked
         '''
         return self._map.get(key, ("?",))[0]
 
     def __contains__(self, key):
         return key in self._map
 
     def __iter__(self):
         for x in sorted(self._map):
             yield x
 
     def parents(self):
         return [self._validate(p) for p in self._pl]
 
     def branch(self):
         return encoding.tolocal(self._branch)
 
     def setparents(self, p1, p2=nullid):
         self._dirty = self._dirtypl = True
         self._pl = p1, p2
 
     def setbranch(self, branch):
         if branch in ['tip', '.', 'null']:
             raise util.Abort(_('the name \'%s\' is reserved') % branch)
         self._branch = encoding.fromlocal(branch)
         self._opener("branch", "w").write(self._branch + '\n')
 
     def _read(self):
         self._map = {}
         self._copymap = {}
         try:
             st = self._opener("dirstate").read()
         except IOError, err:
             if err.errno != errno.ENOENT:
                 raise
             return
         if not st:
             return
 
         p = parsers.parse_dirstate(self._map, self._copymap, st)
         if not self._dirtypl:
             self._pl = p
 
     def invalidate(self):
         for a in ("_map", "_copymap", "_foldmap", "_branch", "_pl", "_dirs",
                 "_ignore"):
             if a in self.__dict__:
                 delattr(self, a)
         self._dirty = False
 
     def copy(self, source, dest):
         """Mark dest as a copy of source. Unmark dest if source is None."""
         if source == dest:
             return
         self._dirty = True
         if source is not None:
             self._copymap[dest] = source
         elif dest in self._copymap:
             del self._copymap[dest]
 
     def copied(self, file):
         return self._copymap.get(file, None)
 
     def copies(self):
         return self._copymap
 
     def _droppath(self, f):
         if self[f] not in "?r" and "_dirs" in self.__dict__:
             _decdirs(self._dirs, f)
 
     def _addpath(self, f, check=False):
         oldstate = self[f]
         if check or oldstate == "r":
             if '\r' in f or '\n' in f:
                 raise util.Abort(
                     _("'\\n' and '\\r' disallowed in filenames: %r") % f)
             if f in self._dirs:
                 raise util.Abort(_('directory %r already in dirstate') % f)
             # shadows
             for d in _finddirs(f):
                 if d in self._dirs:
                     break
                 if d in self._map and self[d] != 'r':
                     raise util.Abort(
                         _('file %r in dirstate clashes with %r') % (d, f))
         if oldstate in "?r" and "_dirs" in self.__dict__:
             _incdirs(self._dirs, f)
 
     def normal(self, f):
         '''Mark a file normal and clean.'''
         self._dirty = True
         self._addpath(f)
         s = os.lstat(self._join(f))
         self._map[f] = ('n', s.st_mode, s.st_size, int(s.st_mtime))
         if f in self._copymap:
             del self._copymap[f]
 
+        # Right now, this file is clean: but if some code in this
+        # process modifies it without changing its size before the clock
+        # ticks over to the next second, then it won't be clean anymore.
+        # So make sure that status() will look harder at it.
+        self._lastnormal.add(f)
+
     def normallookup(self, f):
         '''Mark a file normal, but possibly dirty.'''
         if self._pl[1] != nullid and f in self._map:
             # if there is a merge going on and the file was either
             # in state 'm' (-1) or coming from other parent (-2) before
             # being removed, restore that state.
             entry = self._map[f]
             if entry[0] == 'r' and entry[2] in (-1, -2):
                 source = self._copymap.get(f)
                 if entry[2] == -1:
                     self.merge(f)
                 elif entry[2] == -2:
                     self.otherparent(f)
                 if source:
                     self.copy(source, f)
                 return
             if entry[0] == 'm' or entry[0] == 'n' and entry[2] == -2:
                 return
         self._dirty = True
         self._addpath(f)
         self._map[f] = ('n', 0, -1, -1)
         if f in self._copymap:
             del self._copymap[f]
+        self._lastnormal.discard(f)
 
     def otherparent(self, f):
         '''Mark as coming from the other parent, always dirty.'''
         if self._pl[1] == nullid:
             raise util.Abort(_("setting %r to other parent "
                                "only allowed in merges") % f)
         self._dirty = True
         self._addpath(f)
         self._map[f] = ('n', 0, -2, -1)
         if f in self._copymap:
             del self._copymap[f]
+        self._lastnormal.discard(f)
 
     def add(self, f):
         '''Mark a file added.'''
         self._dirty = True
         self._addpath(f, True)
         self._map[f] = ('a', 0, -1, -1)
         if f in self._copymap:
             del self._copymap[f]
+        self._lastnormal.discard(f)
 
     def remove(self, f):
         '''Mark a file removed.'''
         self._dirty = True
         self._droppath(f)
         size = 0
         if self._pl[1] != nullid and f in self._map:
             # backup the previous state
             entry = self._map[f]
             if entry[0] == 'm': # merge
                 size = -1
             elif entry[0] == 'n' and entry[2] == -2: # other parent
                 size = -2
         self._map[f] = ('r', 0, size, 0)
         if size == 0 and f in self._copymap:
             del self._copymap[f]
+        self._lastnormal.discard(f)
 
     def merge(self, f):
         '''Mark a file merged.'''
         self._dirty = True
         s = os.lstat(self._join(f))
         self._addpath(f)
         self._map[f] = ('m', s.st_mode, s.st_size, int(s.st_mtime))
         if f in self._copymap:
             del self._copymap[f]
+        self._lastnormal.discard(f)
 
     def forget(self, f):
         '''Forget a file.'''
         self._dirty = True
         try:
             self._droppath(f)
             del self._map[f]
         except KeyError:
             self._ui.warn(_("not in dirstate: %s\n") % f)
+        self._lastnormal.discard(f)
 
     def _normalize(self, path, knownpath):
         norm_path = os.path.normcase(path)
         fold_path = self._foldmap.get(norm_path, None)
         if fold_path is None:
             if knownpath or not os.path.lexists(os.path.join(self._root, path)):
                 fold_path = path
             else:
                 fold_path = self._foldmap.setdefault(norm_path,
                                 util.fspath(path, self._root))
         return fold_path
 
     def clear(self):
         self._map = {}
         if "_dirs" in self.__dict__:
             delattr(self, "_dirs")
         self._copymap = {}
         self._pl = [nullid, nullid]
         self._dirty = True
 
     def rebuild(self, parent, files):
         self.clear()
         for f in files:
             if 'x' in files.flags(f):
                 self._map[f] = ('n', 0777, -1, 0)
             else:
                 self._map[f] = ('n', 0666, -1, 0)
         self._pl = (parent, nullid)
         self._dirty = True
 
     def write(self):
         if not self._dirty:
             return
         st = self._opener("dirstate", "w", atomictemp=True)
 
         # use the modification time of the newly created temporary file as the
         # filesystem's notion of 'now'
         now = int(util.fstat(st).st_mtime)
 
         cs = cStringIO.StringIO()
         copymap = self._copymap
         pack = struct.pack
         write = cs.write
         write("".join(self._pl))
         for f, e in self._map.iteritems():
             if e[0] == 'n' and e[3] == now:
                 # The file was last modified "simultaneously" with the current
                 # write to dirstate (i.e. within the same second for file-
                 # systems with a granularity of 1 sec). This commonly happens
                 # for at least a couple of files on 'update'.
                 # The user could change the file without changing its size
                 # within the same second. Invalidate the file's stat data in
                 # dirstate, forcing future 'status' calls to compare the
                 # contents of the file. This prevents mistakenly treating such
                 # files as clean.
                 e = (e[0], 0, -1, -1)   # mark entry as 'unset'
                 self._map[f] = e
 
             if f in copymap:
                 f = "%s\0%s" % (f, copymap[f])
             e = pack(_format, e[0], e[1], e[2], e[3], len(f))
             write(e)
             write(f)
         st.write(cs.getvalue())
         st.rename()
         self._dirty = self._dirtypl = False
 
     def _dirignore(self, f):
         if f == '.':
             return False
         if self._ignore(f):
             return True
         for p in _finddirs(f):
             if self._ignore(p):
                 return True
         return False
 
     def walk(self, match, subrepos, unknown, ignored):
         '''
         Walk recursively through the directory tree, finding all files
         matched by match.
 
         Return a dict mapping filename to stat-like object (either
         mercurial.osutil.stat instance or return value of os.stat()).
         '''
 
         def fwarn(f, msg):
             self._ui.warn('%s: %s\n' % (self.pathto(f), msg))
             return False
 
         def badtype(mode):
             kind = _('unknown')
             if stat.S_ISCHR(mode):
                 kind = _('character device')
             elif stat.S_ISBLK(mode):
                 kind = _('block device')
             elif stat.S_ISFIFO(mode):
                 kind = _('fifo')
             elif stat.S_ISSOCK(mode):
                 kind = _('socket')
             elif stat.S_ISDIR(mode):
                 kind = _('directory')
             return _('unsupported file type (type is %s)') % kind
 
         ignore = self._ignore
         dirignore = self._dirignore
         if ignored:
             ignore = util.never
             dirignore = util.never
         elif not unknown:
             # if unknown and ignored are False, skip step 2
             ignore = util.always
             dirignore = util.always
 
         matchfn = match.matchfn
         badfn = match.bad
         dmap = self._map
         normpath = util.normpath
         listdir = osutil.listdir
         lstat = os.lstat
         getkind = stat.S_IFMT
         dirkind = stat.S_IFDIR
         regkind = stat.S_IFREG
         lnkkind = stat.S_IFLNK
         join = self._join
         work = []
         wadd = work.append
 
         exact = skipstep3 = False
         if matchfn == match.exact: # match.exact
             exact = True
             dirignore = util.always # skip step 2
         elif match.files() and not match.anypats(): # match.match, no patterns
             skipstep3 = True
 
         if self._checkcase:
             normalize = self._normalize
             skipstep3 = False
         else:
             normalize = lambda x, y: x
 
         files = sorted(match.files())
         subrepos.sort()
         i, j = 0, 0
         while i < len(files) and j < len(subrepos):
             subpath = subrepos[j] + "/"
             if files[i] < subpath:
                 i += 1
                 continue
             while i < len(files) and files[i].startswith(subpath):
                 del files[i]
             j += 1
 
         if not files or '.' in files:
             files = ['']
         results = dict.fromkeys(subrepos)
         results['.hg'] = None
 
         # step 1: find all explicit files
         for ff in files:
             nf = normalize(normpath(ff), False)
             if nf in results:
                 continue
 
             try:
                 st = lstat(join(nf))
                 kind = getkind(st.st_mode)
                 if kind == dirkind:
                     skipstep3 = False
                     if nf in dmap:
                         #file deleted on disk but still in dirstate
                         results[nf] = None
                     match.dir(nf)
                     if not dirignore(nf):
                         wadd(nf)
                 elif kind == regkind or kind == lnkkind:
                     results[nf] = st
                 else:
                     badfn(ff, badtype(kind))
                     if nf in dmap:
                         results[nf] = None
             except OSError, inst:
                 if nf in dmap: # does it exactly match a file?
                     results[nf] = None
                 else: # does it match a directory?
                     prefix = nf + "/"
                     for fn in dmap:
                         if fn.startswith(prefix):
                             match.dir(nf)
                             skipstep3 = False
                             break
                     else:
                         badfn(ff, inst.strerror)
 
         # step 2: visit subdirectories
         while work:
             nd = work.pop()
             skip = None
             if nd == '.':
                 nd = ''
             else:
                 skip = '.hg'
             try:
                 entries = listdir(join(nd), stat=True, skip=skip)
             except OSError, inst:
                 if inst.errno == errno.EACCES:
                     fwarn(nd, inst.strerror)
                     continue
                 raise
             for f, kind, st in entries:
                 nf = normalize(nd and (nd + "/" + f) or f, True)
                 if nf not in results:
                     if kind == dirkind:
                         if not ignore(nf):
                             match.dir(nf)
                             wadd(nf)
                         if nf in dmap and matchfn(nf):
                             results[nf] = None
                     elif kind == regkind or kind == lnkkind:
                         if nf in dmap:
                             if matchfn(nf):
                                 results[nf] = st
                         elif matchfn(nf) and not ignore(nf):
                             results[nf] = st
                     elif nf in dmap and matchfn(nf):
                         results[nf] = None
 
         # step 3: report unseen items in the dmap hash
         if not skipstep3 and not exact:
             visit = sorted([f for f in dmap if f not in results and matchfn(f)])
             for nf, st in zip(visit, util.statfiles([join(i) for i in visit])):
                 if not st is None and not getkind(st.st_mode) in (regkind, lnkkind):
                     st = None
                 results[nf] = st
         for s in subrepos:
             del results[s]
         del results['.hg']
         return results
 
     def status(self, match, subrepos, ignored, clean, unknown):
         '''Determine the status of the working copy relative to the
         dirstate and return a tuple of lists (unsure, modified, added,
         removed, deleted, unknown, ignored, clean), where:
 
           unsure:
             files that might have been modified since the dirstate was
             written, but need to be read to be sure (size is the same
             but mtime differs)
           modified:
             files that have definitely been modified since the dirstate
             was written (different size or mode)
           added:
             files that have been explicitly added with hg add
           removed:
             files that have been explicitly removed with hg remove
           deleted:
             files that have been deleted through other means ("missing")
           unknown:
             files not in the dirstate that are not ignored
           ignored:
             files not in the dirstate that are ignored
             (by _dirignore())
           clean:
             files that have definitely not been modified since the
             dirstate was written
         '''
         listignored, listclean, listunknown = ignored, clean, unknown
         lookup, modified, added, unknown, ignored = [], [], [], [], []
         removed, deleted, clean = [], [], []
 
         dmap = self._map
         ladd = lookup.append            # aka "unsure"
         madd = modified.append
         aadd = added.append
         uadd = unknown.append
         iadd = ignored.append
         radd = removed.append
         dadd = deleted.append
         cadd = clean.append
+        lastnormal = self._lastnormal.__contains__
 
         lnkkind = stat.S_IFLNK
 
         for fn, st in self.walk(match, subrepos, listunknown,
                                 listignored).iteritems():
             if fn not in dmap:
                 if (listignored or match.exact(fn)) and self._dirignore(fn):
                     if listignored:
                         iadd(fn)
                 elif listunknown:
                     uadd(fn)
                 continue
 
             state, mode, size, time = dmap[fn]
 
             if not st and state in "nma":
                 dadd(fn)
             elif state == 'n':
                 # The "mode & lnkkind != lnkkind or self._checklink"
                 # lines are an expansion of "islink => checklink"
                 # where islink means "is this a link?" and checklink
                 # means "can we check links?".
                 if (size >= 0 and
                     (size != st.st_size
                      or ((mode ^ st.st_mode) & 0100 and self._checkexec))
                     and (mode & lnkkind != lnkkind or self._checklink)
                     or size == -2 # other parent
                     or fn in self._copymap):
                     madd(fn)
                 elif (time != int(st.st_mtime)
                       and (mode & lnkkind != lnkkind or self._checklink)):
                     ladd(fn)
+                elif lastnormal(fn):
+                    # If previously in this process we recorded that
+                    # this file is clean, think twice: intervening code
+                    # may have modified the file in the same second
+                    # without changing its size. So force caller to
+                    # check file contents. Because we're not updating
+                    # self._map, this only affects the current process.
+                    # That should be OK because this mainly affects
+                    # multiple commits in the same process, and each
+                    # commit by definition makes the committed files
+                    # clean.
+                    ladd(fn)
                 elif listclean:
                     cadd(fn)
             elif state == 'm':
                 madd(fn)
             elif state == 'a':
                 aadd(fn)
             elif state == 'r':
                 radd(fn)
 
         return (lookup, modified, added, removed, deleted, unknown, ignored,
                 clean)