bisect: avoid copying ancestor list for non-merge commits
During a bisection, hg needs to compute a list of all ancestors for every
candidate commit. This is accomplished via a bottom-up traversal of the set of
candidates, during which each revision's ancestor list is populated using the
ancestor list of its parent(s). Previously, this involved copying the entire
list, which could be very long in if the bisection range was large.
To help improve this, we can observe that each candidate commit is visited
exactly once, at which point its ancestor list is copied into its children's
lists and then dropped. In the case of non-merge commits, a commit's ancestor
list consists exactly of its parent's list plus itself. This means that we can
trivially reuse the parent's existing list for one of its non-merge children,
which avoids copying entirely if that commit is the parent's only child. This
makes bisections over linear ranges of commits much faster.
During some informal testing in the large publicly-available `mozilla-central`
repository, this noticeably sped up bisections over large ranges of history:
Setup:
$ cd mozilla-central
$ hg bisect --reset
$ hg bisect --good 0
$ hg log -r tip -T '{rev}\n'
628417
Test:
$ time hg bisect --bad tip --noupdate
Before:
real 3m35.927s
user 3m35.553s
sys 0m0.319s
After:
real 1m41.142s
user 1m40.810s
sys 0m0.285s
# coding: utf8
# ext-sidedata-5.py - small extension to test (differently still) the sidedata
# logic
#
# Simulates a server for a simple sidedata exchange.
#
# Copyright 2021 Raphaël Gomès <rgomes@octobus.net>
#
# This software may be used and distributed according to the terms of the
# GNU General Public License version 2 or any later version.
import hashlib
import struct
from mercurial import (
extensions,
revlog,
)
from mercurial.revlogutils import sidedata as sidedatamod
from mercurial.revlogutils import constants
NO_FLAGS = (0, 0)
def compute_sidedata_1(repo, revlog, rev, sidedata, text=None):
sidedata = sidedata.copy()
if text is None:
text = revlog.revision(rev)
sidedata[sidedatamod.SD_TEST1] = struct.pack('>I', len(text))
return sidedata, NO_FLAGS
def compute_sidedata_2(repo, revlog, rev, sidedata, text=None):
sidedata = sidedata.copy()
if text is None:
text = revlog.revision(rev)
sha256 = hashlib.sha256(text).digest()
sidedata[sidedatamod.SD_TEST2] = struct.pack('>32s', sha256)
return sidedata, NO_FLAGS
def reposetup(ui, repo):
# Sidedata keys happen to be the same as the categories, easier for testing.
for kind in constants.ALL_KINDS:
repo.register_sidedata_computer(
kind,
sidedatamod.SD_TEST1,
(sidedatamod.SD_TEST1,),
compute_sidedata_1,
0,
)
repo.register_sidedata_computer(
kind,
sidedatamod.SD_TEST2,
(sidedatamod.SD_TEST2,),
compute_sidedata_2,
0,
)
# We don't register sidedata computers because we don't care within these
# tests
repo.register_wanted_sidedata(sidedatamod.SD_TEST1)
repo.register_wanted_sidedata(sidedatamod.SD_TEST2)
def wrapaddrevision(
orig, self, text, transaction, link, p1, p2, *args, **kwargs
):
if kwargs.get('sidedata') is None:
kwargs['sidedata'] = {}
sd = kwargs['sidedata']
## let's store some arbitrary data just for testing
# text length
sd[sidedatamod.SD_TEST1] = struct.pack('>I', len(text))
# and sha2 hashes
sha256 = hashlib.sha256(text).digest()
sd[sidedatamod.SD_TEST2] = struct.pack('>32s', sha256)
return orig(self, text, transaction, link, p1, p2, *args, **kwargs)
def extsetup(ui):
extensions.wrapfunction(revlog.revlog, 'addrevision', wrapaddrevision)