setdiscovery: precompute children revisions to avoid quadratic lookup

Moving away from dagutil a few commits ago introduced quadratic
behavior when resolving children revisions during discovery.

This commit introduces a precompute step of the children revisions
to avoid the bad behavior.

I believe the new code should have near identical performance to
what dagutil was doing before. Behavior is still slightly different
because we take into account filtered revisions. But this change was
made when we moved off dagutil.

I added a comment about multiple invocations of this function
redundantly calculating the children revisions. I believe this
potentially undesirable behavior was present when we used dagutil,
as the call to inverse() previously in this function created a new
object and required computing children on every invocation. I thought
we should document the potential for a performance issue rather than
let it go undocumented.

Differential Revision: https://phab.mercurial-scm.org/D4326

mercurial/setdiscovery.py

@@ -1,293 +1,313 b''
 # setdiscovery.py - improved discovery of common nodeset for mercurial
 #
 # Copyright 2010 Benoit Boissinot <[e-mail]>
 # and Peter Arrenbrecht <[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.
 """
 Algorithm works in the following way. You have two repository: local and
 remote. They both contains a DAG of changelists.
 
 The goal of the discovery protocol is to find one set of node *common*,
 the set of nodes shared by local and remote.
 
 One of the issue with the original protocol was latency, it could
 potentially require lots of roundtrips to discover that the local repo was a
 subset of remote (which is a very common case, you usually have few changes
 compared to upstream, while upstream probably had lots of development).
 
 The new protocol only requires one interface for the remote repo: `known()`,
 which given a set of changelists tells you if they are present in the DAG.
 
 The algorithm then works as follow:
 
  - We will be using three sets, `common`, `missing`, `unknown`. Originally
  all nodes are in `unknown`.
  - Take a sample from `unknown`, call `remote.known(sample)`
    - For each node that remote knows, move it and all its ancestors to `common`
    - For each node that remote doesn't know, move it and all its descendants
    to `missing`
  - Iterate until `unknown` is empty
 
 There are a couple optimizations, first is instead of starting with a random
 sample of missing, start by sending all heads, in the case where the local
 repo is a subset, you computed the answer in one round trip.
 
 Then you can do something similar to the bisecting strategy used when
 finding faulty changesets. Instead of random samples, you can try picking
 nodes that will maximize the number of nodes that will be
 classified with it (since all ancestors or descendants will be marked as well).
 """
 
 from __future__ import absolute_import
 
 import collections
 import random
 
 from .i18n import _
 from .node import (
     nullid,
     nullrev,
 )
 from . import (
     error,
     util,
 )
 
 def _updatesample(revs, heads, sample, parentfn, quicksamplesize=0):
     """update an existing sample to match the expected size
 
     The sample is updated with revs exponentially distant from each head of the
     <revs> set. (H~1, H~2, H~4, H~8, etc).
 
     If a target size is specified, the sampling will stop once this size is
     reached. Otherwise sampling will happen until roots of the <revs> set are
     reached.
 
     :revs:  set of revs we want to discover (if None, assume the whole dag)
     :heads: set of DAG head revs
     :sample: a sample to update
     :parentfn: a callable to resolve parents for a revision
     :quicksamplesize: optional target size of the sample"""
     dist = {}
     visit = collections.deque(heads)
     seen = set()
     factor = 1
     while visit:
         curr = visit.popleft()
         if curr in seen:
             continue
         d = dist.setdefault(curr, 1)
         if d > factor:
             factor *= 2
         if d == factor:
             sample.add(curr)
             if quicksamplesize and (len(sample) >= quicksamplesize):
                 return
         seen.add(curr)
 
         for p in parentfn(curr):
             if p != nullrev and (not revs or p in revs):
                 dist.setdefault(p, d + 1)
                 visit.append(p)
 
 def _takequicksample(repo, headrevs, revs, size):
     """takes a quick sample of size <size>
 
     It is meant for initial sampling and focuses on querying heads and close
     ancestors of heads.
 
     :dag: a dag object
     :headrevs: set of head revisions in local DAG to consider
     :revs: set of revs to discover
     :size: the maximum size of the sample"""
     sample = set(repo.revs('heads(%ld)', revs))
 
     if len(sample) >= size:
         return _limitsample(sample, size)
 
     _updatesample(None, headrevs, sample, repo.changelog.parentrevs,
                   quicksamplesize=size)
     return sample
 
 def _takefullsample(repo, headrevs, revs, size):
     sample = set(repo.revs('heads(%ld)', revs))
 
     # update from heads
     revsheads = set(repo.revs('heads(%ld)', revs))
     _updatesample(revs, revsheads, sample, repo.changelog.parentrevs)
+
     # update from roots
     revsroots = set(repo.revs('roots(%ld)', revs))
 
-    # TODO this is quadratic
-    parentfn = lambda rev: repo.changelog.children(repo.changelog.node(rev))
+    # _updatesample() essentially does interaction over revisions to look up
+    # their children. This lookup is expensive and doing it in a loop is
+    # quadratic. We precompute the children for all relevant revisions and
+    # make the lookup in _updatesample() a simple dict lookup.
+    #
+    # Because this function can be called multiple times during discovery, we
+    # may still perform redundant work and there is room to optimize this by
+    # keeping a persistent cache of children across invocations.
+    children = {}
 
-    _updatesample(revs, revsroots, sample, parentfn)
+    parentrevs = repo.changelog.parentrevs
+    for rev in repo.changelog.revs(start=min(revsroots)):
+        # Always ensure revision has an entry so we don't need to worry about
+        # missing keys.
+        children.setdefault(rev, [])
+
+        for prev in parentrevs(rev):
+            if prev == nullrev:
+                continue
+
+            children.setdefault(prev, []).append(rev)
+
+    _updatesample(revs, revsroots, sample, children.__getitem__)
     assert sample
     sample = _limitsample(sample, size)
     if len(sample) < size:
         more = size - len(sample)
         sample.update(random.sample(list(revs - sample), more))
     return sample
 
 def _limitsample(sample, desiredlen):
     """return a random subset of sample of at most desiredlen item"""
     if len(sample) > desiredlen:
         sample = set(random.sample(sample, desiredlen))
     return sample
 
 def findcommonheads(ui, local, remote,
                     initialsamplesize=100,
                     fullsamplesize=200,
                     abortwhenunrelated=True,
                     ancestorsof=None):
     '''Return a tuple (common, anyincoming, remoteheads) used to identify
     missing nodes from or in remote.
     '''
     start = util.timer()
 
     roundtrips = 0
     cl = local.changelog
     clnode = cl.node
     clrev = cl.rev
 
     if ancestorsof is not None:
         ownheads = [clrev(n) for n in ancestorsof]
     else:
         ownheads = [rev for rev in cl.headrevs() if rev != nullrev]
 
     # early exit if we know all the specified remote heads already
     ui.debug("query 1; heads\n")
     roundtrips += 1
     sample = _limitsample(ownheads, initialsamplesize)
     # indices between sample and externalized version must match
     sample = list(sample)
 
     with remote.commandexecutor() as e:
         fheads = e.callcommand('heads', {})
         fknown = e.callcommand('known', {
             'nodes': [clnode(r) for r in sample],
         })
 
     srvheadhashes, yesno = fheads.result(), fknown.result()
 
     if cl.tip() == nullid:
         if srvheadhashes != [nullid]:
             return [nullid], True, srvheadhashes
         return [nullid], False, []
 
     # start actual discovery (we note this before the next "if" for
     # compatibility reasons)
     ui.status(_("searching for changes\n"))
 
     srvheads = []
     for node in srvheadhashes:
         if node == nullid:
             continue
 
         try:
             srvheads.append(clrev(node))
         # Catches unknown and filtered nodes.
         except error.LookupError:
             continue
 
     if len(srvheads) == len(srvheadhashes):
         ui.debug("all remote heads known locally\n")
         return srvheadhashes, False, srvheadhashes
 
     if len(sample) == len(ownheads) and all(yesno):
         ui.note(_("all local heads known remotely\n"))
         ownheadhashes = [clnode(r) for r in ownheads]
         return ownheadhashes, True, srvheadhashes
 
     # full blown discovery
 
     # own nodes I know we both know
     # treat remote heads (and maybe own heads) as a first implicit sample
     # response
     common = cl.incrementalmissingrevs(srvheads)
     commoninsample = set(n for i, n in enumerate(sample) if yesno[i])
     common.addbases(commoninsample)
     # own nodes where I don't know if remote knows them
     undecided = set(common.missingancestors(ownheads))
     # own nodes I know remote lacks
     missing = set()
 
     full = False
     progress = ui.makeprogress(_('searching'), unit=_('queries'))
     while undecided:
 
         if sample:
             missinginsample = [n for i, n in enumerate(sample) if not yesno[i]]
 
             if missing:
                 missing.update(local.revs('descendants(%ld) - descendants(%ld)',
                                           missinginsample, missing))
             else:
                 missing.update(local.revs('descendants(%ld)', missinginsample))
 
             undecided.difference_update(missing)
 
         if not undecided:
             break
 
         if full or common.hasbases():
             if full:
                 ui.note(_("sampling from both directions\n"))
             else:
                 ui.debug("taking initial sample\n")
             samplefunc = _takefullsample
             targetsize = fullsamplesize
         else:
             # use even cheaper initial sample
             ui.debug("taking quick initial sample\n")
             samplefunc = _takequicksample
             targetsize = initialsamplesize
         if len(undecided) < targetsize:
             sample = list(undecided)
         else:
             sample = samplefunc(local, ownheads, undecided, targetsize)
 
         roundtrips += 1
         progress.update(roundtrips)
         ui.debug("query %i; still undecided: %i, sample size is: %i\n"
                  % (roundtrips, len(undecided), len(sample)))
         # indices between sample and externalized version must match
         sample = list(sample)
 
         with remote.commandexecutor() as e:
             yesno = e.callcommand('known', {
                 'nodes': [clnode(r) for r in sample],
             }).result()
 
         full = True
 
         if sample:
             commoninsample = set(n for i, n in enumerate(sample) if yesno[i])
             common.addbases(commoninsample)
             common.removeancestorsfrom(undecided)
 
     # heads(common) == heads(common.bases) since common represents common.bases
     # and all its ancestors
     # The presence of nullrev will confuse heads(). So filter it out.
     result = set(local.revs('heads(%ld)', common.bases - {nullrev}))
     elapsed = util.timer() - start
     progress.complete()
     ui.debug("%d total queries in %.4fs\n" % (roundtrips, elapsed))
     msg = ('found %d common and %d unknown server heads,'
            ' %d roundtrips in %.4fs\n')
     missing = set(result) - set(srvheads)
     ui.log('discovery', msg, len(result), len(missing), roundtrips,
            elapsed)
 
     if not result and srvheadhashes != [nullid]:
         if abortwhenunrelated:
             raise error.Abort(_("repository is unrelated"))
         else:
             ui.warn(_("warning: repository is unrelated\n"))
         return ({nullid}, True, srvheadhashes,)
 
     anyincoming = (srvheadhashes != [nullid])
     result = {clnode(r) for r in result}
     return result, anyincoming, srvheadhashes