view mercurial/hbisect.py @ 45962:a66568f20ddc

copies: use the rust code for `combine_changeset_copies` Changeset centric copy tracing now use the rust code. The rust code focussed on simplicity and will be optimised later. So the performance is not great yet. Now that all the pieces are in place we can start working on performance in the coming changesets. Below is a table that summarize how slower we got: Repo Cases Source-Rev Dest-Rev Py-time Rust-time Difference Factor ------------------------------------------------------------------------------------------------------------------------------------ mercurial x_revs_x_added_0_copies ad6b123de1c7 39cfcef4f463 : 0.000049 s, 0.000046 s, -0.000003 s, × 0.9388 mercurial x_revs_x_added_x_copies 2b1c78674230 0c1d10351869 : 0.000112 s, 0.000173 s, +0.000061 s, × 1.5446 mercurial x000_revs_x000_added_x_copies 81f8ff2a9bf2 dd3267698d84 : 0.004216 s, 0.006303 s, +0.002087 s, × 1.4950 pypy x_revs_x_added_0_copies aed021ee8ae8 099ed31b181b : 0.000204 s, 0.000229 s, +0.000025 s, × 1.1225 pypy x_revs_x000_added_0_copies 4aa4e1f8e19a 359343b9ac0e : 0.000058 s, 0.000056 s, -0.000002 s, × 0.9655 pypy x_revs_x_added_x_copies ac52eb7bbbb0 72e022663155 : 0.000112 s, 0.000143 s, +0.000031 s, × 1.2768 pypy x_revs_x00_added_x_copies c3b14617fbd7 ace7255d9a26 : 0.000339 s, 0.001166 s, +0.000827 s, × 3.4395 pypy x_revs_x000_added_x000_copies df6f7a526b60 a83dc6a2d56f : 0.010214 s, 0.022931 s, +0.012717 s, × 2.2451 pypy x000_revs_xx00_added_0_copies 89a76aede314 2f22446ff07e : 0.047497 s, 0.852446 s, +0.804949 s, × 17.9474 pypy x000_revs_x000_added_x_copies 8a3b5bfd266e 2c68e87c3efe : 0.075297 s, 2.221824 s, +2.146527 s, × 29.5075 pypy x000_revs_x000_added_x000_copies 89a76aede314 7b3dda341c84 : 0.057322 s, 1.194162 s, +1.136840 s, × 20.8325 pypy x0000_revs_x_added_0_copies d1defd0dc478 c9cb1334cc78 : 0.796264 s, 62.468362 s, +61.672098 s, × 78.4518 pypy x0000_revs_xx000_added_0_copies bf2c629d0071 4ffed77c095c : 0.020491 s, 0.022116 s, +0.001625 s, × 1.0793 pypy x0000_revs_xx000_added_x000_copies 08ea3258278e d9fa043f30c0 : 0.121612 s, 2.972788 s, +2.851176 s, × 24.4449 netbeans x_revs_x_added_0_copies fb0955ffcbcd a01e9239f9e7 : 0.000143 s, 0.000180 s, +0.000037 s, × 1.2587 netbeans x_revs_x000_added_0_copies 6f360122949f 20eb231cc7d0 : 0.000112 s, 0.000123 s, +0.000011 s, × 1.0982 netbeans x_revs_x_added_x_copies 1ada3faf6fb6 5a39d12eecf4 : 0.000232 s, 0.000315 s, +0.000083 s, × 1.3578 netbeans x_revs_x00_added_x_copies 35be93ba1e2c 9eec5e90c05f : 0.000721 s, 0.001297 s, +0.000576 s, × 1.7989 netbeans x000_revs_xx00_added_0_copies eac3045b4fdd 51d4ae7f1290 : 0.010115 s, 0.024884 s, +0.014769 s, × 2.4601 netbeans x000_revs_x000_added_x_copies e2063d266acd 6081d72689dc : 0.015461 s, 0.032653 s, +0.017192 s, × 2.1120 netbeans x000_revs_x000_added_x000_copies ff453e9fee32 411350406ec2 : 0.060756 s, 4.230118 s, +4.169362 s, × 69.6247 netbeans x0000_revs_xx000_added_x000_copies 588c2d1ced70 1aad62e59ddd : 0.605842 s, killed mozilla-central x_revs_x_added_0_copies 3697f962bb7b 7015fcdd43a2 : 0.000164 s, 0.000197 s, +0.000033 s, × 1.2012 mozilla-central x_revs_x000_added_0_copies dd390860c6c9 40d0c5bed75d : 0.000331 s, 0.000622 s, +0.000291 s, × 1.8792 mozilla-central x_revs_x_added_x_copies 8d198483ae3b 14207ffc2b2f : 0.000249 s, 0.000296 s, +0.000047 s, × 1.1888 mozilla-central x_revs_x00_added_x_copies 98cbc58cc6bc 446a150332c3 : 0.000711 s, 0.001626 s, +0.000915 s, × 2.2869 mozilla-central x_revs_x000_added_x000_copies 3c684b4b8f68 0a5e72d1b479 : 0.003438 s, 0.006218 s, +0.002780 s, × 1.8086 mozilla-central x_revs_x0000_added_x0000_copies effb563bb7e5 c07a39dc4e80 : 0.069869 s, 0.132760 s, +0.062891 s, × 1.9001 mozilla-central x000_revs_xx00_added_0_copies 6100d773079a 04a55431795e : 0.005701 s, 0.029001 s, +0.023300 s, × 5.0870 mozilla-central x000_revs_x000_added_x_copies 9f17a6fc04f9 2d37b966abed : 0.005757 s, 0.005886 s, +0.000129 s, × 1.0224 mozilla-central x000_revs_x000_added_x000_copies 7c97034feb78 4407bd0c6330 : 0.061826 s, 3.619850 s, +3.558024 s, × 58.5490 mozilla-central x0000_revs_xx000_added_0_copies 9eec5917337d 67118cc6dcad : 0.043354 s, 0.058678 s, +0.015324 s, × 1.3535 mozilla-central x0000_revs_xx000_added_x000_copies f78c615a656c 96a38b690156 : 0.198979 s, 11.926587 s, +11.727608 s, × 59.9389 mozilla-central x00000_revs_x0000_added_x0000_copies 6832ae71433c 4c222a1d9a00 : 2.067096 s, killed mozilla-central x00000_revs_x00000_added_x000_copies 76caed42cf7c 1daa622bbe42 : 3.102616 s, killed mozilla-try x_revs_x_added_0_copies aaf6dde0deb8 9790f499805a : 0.001212 s, 0.001204 s, -0.000008 s, × 0.9934 mozilla-try x_revs_x000_added_0_copies d8d0222927b4 5bb8ce8c7450 : 0.001237 s, 0.001217 s, -0.000020 s, × 0.9838 mozilla-try x_revs_x_added_x_copies 092fcca11bdb 936255a0384a : 0.000557 s, 0.000605 s, +0.000048 s, × 1.0862 mozilla-try x_revs_x00_added_x_copies b53d2fadbdb5 017afae788ec : 0.001532 s, 0.001876 s, +0.000344 s, × 1.2245 mozilla-try x_revs_x000_added_x000_copies 20408ad61ce5 6f0ee96e21ad : 0.035166 s, 0.078190 s, +0.043024 s, × 2.2235 mozilla-try x_revs_x0000_added_x0000_copies effb563bb7e5 c07a39dc4e80 : 0.070336 s, 0.135428 s, +0.065092 s, × 1.9254 mozilla-try x000_revs_xx00_added_0_copies 6100d773079a 04a55431795e : 0.006080 s, 0.029123 s, +0.023043 s, × 4.7900 mozilla-try x000_revs_x000_added_x_copies 9f17a6fc04f9 2d37b966abed : 0.006099 s, 0.006141 s, +0.000042 s, × 1.0069 mozilla-try x000_revs_x000_added_x000_copies 1346fd0130e4 4c65cbdabc1f : 0.064317 s, 4.857827 s, +4.793510 s, × 75.5294 mozilla-try x0000_revs_x_added_0_copies 63519bfd42ee a36a2a865d92 : 0.303263 s, 10.674920 s, +10.371657 s, × 35.2002 mozilla-try x0000_revs_x_added_x_copies 9fe69ff0762d bcabf2a78927 : 0.292804 s, 9.789462 s, +9.496658 s, × 33.4335 mozilla-try x0000_revs_xx000_added_x_copies 156f6e2674f2 4d0f2c178e66 : 0.107594 s, 1.087890 s, +0.980296 s, × 10.1111 mozilla-try x0000_revs_xx000_added_0_copies 9eec5917337d 67118cc6dcad : 0.045202 s, 0.060556 s, +0.015354 s, × 1.3397 mozilla-try x0000_revs_xx000_added_x000_copies 89294cd501d9 7ccb2fc7ccb5 : 1.926277 s, killed mozilla-try x0000_revs_x0000_added_x0000_copies e928c65095ed e951f4ad123a : 0.794492 s, killed mozilla-try x00000_revs_x_added_0_copies 6a320851d377 1ebb79acd503 : 84.521497 s, killed mozilla-try x00000_revs_x00000_added_0_copies dc8a3ca7010e d16fde900c9c : 0.965937 s, 19.647038 s, +18.681101 s, × 20.3399 mozilla-try x00000_revs_x_added_x_copies 5173c4b6f97c 95d83ee7242d : 83.367146 s, killed mozilla-try x00000_revs_x000_added_x_copies 9126823d0e9c ca82787bb23c : 84.260895 s, killed mozilla-try x00000_revs_x0000_added_x0000_copies 8d3fafa80d4b eb884023b810 : 3.274537 s, killed mozilla-try x00000_revs_x00000_added_x0000_copies 1b661134e2ca 1ae03d022d6d : 42.235843 s, killed mozilla-try x00000_revs_x00000_added_x000_copies 9b2a99adc05e 8e29777b48e6 : 49.872829 s, killed Differential Revision: https://phab.mercurial-scm.org/D9299
author Pierre-Yves David <pierre-yves.david@octobus.net>
date Thu, 01 Oct 2020 18:52:13 +0200
parents e641bb2a6159
children d4ba4d51f85f
line wrap: on
line source

# changelog bisection for mercurial
#
# Copyright 2007 Matt Mackall
# Copyright 2005, 2006 Benoit Boissinot <benoit.boissinot@ens-lyon.org>
#
# Inspired by git bisect, extension skeleton taken from mq.py.
#
# This software may be used and distributed according to the terms of the
# GNU General Public License version 2 or any later version.

from __future__ import absolute_import

import collections
import contextlib

from .i18n import _
from .node import (
    hex,
    short,
)
from . import error


def bisect(repo, state):
    """find the next node (if any) for testing during a bisect search.
    returns a (nodes, number, good) tuple.

    'nodes' is the final result of the bisect if 'number' is 0.
    Otherwise 'number' indicates the remaining possible candidates for
    the search and 'nodes' contains the next bisect target.
    'good' is True if bisect is searching for a first good changeset, False
    if searching for a first bad one.
    """

    repo = repo.unfiltered()
    changelog = repo.changelog
    clparents = changelog.parentrevs
    skip = {changelog.rev(n) for n in state[b'skip']}

    def buildancestors(bad, good):
        badrev = min([changelog.rev(n) for n in bad])
        ancestors = collections.defaultdict(lambda: None)
        for rev in repo.revs(b"descendants(%ln) - ancestors(%ln)", good, good):
            ancestors[rev] = []
        if ancestors[badrev] is None:
            return badrev, None
        return badrev, ancestors

    good = False
    badrev, ancestors = buildancestors(state[b'bad'], state[b'good'])
    if not ancestors:  # looking for bad to good transition?
        good = True
        badrev, ancestors = buildancestors(state[b'good'], state[b'bad'])
    bad = changelog.node(badrev)
    if not ancestors:  # now we're confused
        if (
            len(state[b'bad']) == 1
            and len(state[b'good']) == 1
            and state[b'bad'] != state[b'good']
        ):
            raise error.Abort(_(b"starting revisions are not directly related"))
        raise error.Abort(
            _(b"inconsistent state, %d:%s is good and bad")
            % (badrev, short(bad))
        )

    # build children dict
    children = {}
    visit = collections.deque([badrev])
    candidates = []
    while visit:
        rev = visit.popleft()
        if ancestors[rev] == []:
            candidates.append(rev)
            for prev in clparents(rev):
                if prev != -1:
                    if prev in children:
                        children[prev].append(rev)
                    else:
                        children[prev] = [rev]
                        visit.append(prev)

    candidates.sort()
    # have we narrowed it down to one entry?
    # or have all other possible candidates besides 'bad' have been skipped?
    tot = len(candidates)
    unskipped = [c for c in candidates if (c not in skip) and (c != badrev)]
    if tot == 1 or not unskipped:
        return ([changelog.node(c) for c in candidates], 0, good)
    perfect = tot // 2

    # find the best node to test
    best_rev = None
    best_len = -1
    poison = set()
    for rev in candidates:
        if rev in poison:
            # poison children
            poison.update(children.get(rev, []))
            continue

        a = ancestors[rev] or [rev]
        ancestors[rev] = None

        x = len(a)  # number of ancestors
        y = tot - x  # number of non-ancestors
        value = min(x, y)  # how good is this test?
        if value > best_len and rev not in skip:
            best_len = value
            best_rev = rev
            if value == perfect:  # found a perfect candidate? quit early
                break

        if y < perfect and rev not in skip:  # all downhill from here?
            # poison children
            poison.update(children.get(rev, []))
            continue

        for c in children.get(rev, []):
            if ancestors[c]:
                ancestors[c] = list(set(ancestors[c] + a))
            else:
                ancestors[c] = a + [c]

    assert best_rev is not None
    best_node = changelog.node(best_rev)

    return ([best_node], tot, good)


def extendrange(repo, state, nodes, good):
    # bisect is incomplete when it ends on a merge node and
    # one of the parent was not checked.
    parents = repo[nodes[0]].parents()
    if len(parents) > 1:
        if good:
            side = state[b'bad']
        else:
            side = state[b'good']
        num = len({i.node() for i in parents} & set(side))
        if num == 1:
            return parents[0].ancestor(parents[1])
    return None


def load_state(repo):
    state = {b'current': [], b'good': [], b'bad': [], b'skip': []}
    for l in repo.vfs.tryreadlines(b"bisect.state"):
        kind, node = l[:-1].split()
        node = repo.unfiltered().lookup(node)
        if kind not in state:
            raise error.Abort(_(b"unknown bisect kind %s") % kind)
        state[kind].append(node)
    return state


def save_state(repo, state):
    f = repo.vfs(b"bisect.state", b"w", atomictemp=True)
    with repo.wlock():
        for kind in sorted(state):
            for node in state[kind]:
                f.write(b"%s %s\n" % (kind, hex(node)))
        f.close()


def resetstate(repo):
    """remove any bisect state from the repository"""
    if repo.vfs.exists(b"bisect.state"):
        repo.vfs.unlink(b"bisect.state")


def checkstate(state):
    """check we have both 'good' and 'bad' to define a range

    Raise StateError exception otherwise."""
    if state[b'good'] and state[b'bad']:
        return True
    if not state[b'good']:
        raise error.StateError(_(b'cannot bisect (no known good revisions)'))
    else:
        raise error.StateError(_(b'cannot bisect (no known bad revisions)'))


@contextlib.contextmanager
def restore_state(repo, state, node):
    try:
        yield
    finally:
        state[b'current'] = [node]
        save_state(repo, state)


def get(repo, status):
    """
    Return a list of revision(s) that match the given status:

    - ``good``, ``bad``, ``skip``: csets explicitly marked as good/bad/skip
    - ``goods``, ``bads``      : csets topologically good/bad
    - ``range``              : csets taking part in the bisection
    - ``pruned``             : csets that are goods, bads or skipped
    - ``untested``           : csets whose fate is yet unknown
    - ``ignored``            : csets ignored due to DAG topology
    - ``current``            : the cset currently being bisected
    """
    state = load_state(repo)
    if status in (b'good', b'bad', b'skip', b'current'):
        return map(repo.unfiltered().changelog.rev, state[status])
    else:
        # In the following sets, we do *not* call 'bisect()' with more
        # than one level of recursion, because that can be very, very
        # time consuming. Instead, we always develop the expression as
        # much as possible.

        # 'range' is all csets that make the bisection:
        #   - have a good ancestor and a bad descendant, or conversely
        # that's because the bisection can go either way
        range = b'( bisect(bad)::bisect(good) | bisect(good)::bisect(bad) )'

        _t = repo.revs(b'bisect(good)::bisect(bad)')
        # The sets of topologically good or bad csets
        if len(_t) == 0:
            # Goods are topologically after bads
            goods = b'bisect(good)::'  # Pruned good csets
            bads = b'::bisect(bad)'  # Pruned bad csets
        else:
            # Goods are topologically before bads
            goods = b'::bisect(good)'  # Pruned good csets
            bads = b'bisect(bad)::'  # Pruned bad csets

        # 'pruned' is all csets whose fate is already known: good, bad, skip
        skips = b'bisect(skip)'  # Pruned skipped csets
        pruned = b'( (%s) | (%s) | (%s) )' % (goods, bads, skips)

        # 'untested' is all cset that are- in 'range', but not in 'pruned'
        untested = b'( (%s) - (%s) )' % (range, pruned)

        # 'ignored' is all csets that were not used during the bisection
        # due to DAG topology, but may however have had an impact.
        # E.g., a branch merged between bads and goods, but whose branch-
        # point is out-side of the range.
        iba = b'::bisect(bad) - ::bisect(good)'  # Ignored bads' ancestors
        iga = b'::bisect(good) - ::bisect(bad)'  # Ignored goods' ancestors
        ignored = b'( ( (%s) | (%s) ) - (%s) )' % (iba, iga, range)

        if status == b'range':
            return repo.revs(range)
        elif status == b'pruned':
            return repo.revs(pruned)
        elif status == b'untested':
            return repo.revs(untested)
        elif status == b'ignored':
            return repo.revs(ignored)
        elif status == b"goods":
            return repo.revs(goods)
        elif status == b"bads":
            return repo.revs(bads)
        else:
            raise error.ParseError(_(b'invalid bisect state'))


def label(repo, node):
    rev = repo.changelog.rev(node)

    # Try explicit sets
    if rev in get(repo, b'good'):
        # i18n: bisect changeset status
        return _(b'good')
    if rev in get(repo, b'bad'):
        # i18n: bisect changeset status
        return _(b'bad')
    if rev in get(repo, b'skip'):
        # i18n: bisect changeset status
        return _(b'skipped')
    if rev in get(repo, b'untested') or rev in get(repo, b'current'):
        # i18n: bisect changeset status
        return _(b'untested')
    if rev in get(repo, b'ignored'):
        # i18n: bisect changeset status
        return _(b'ignored')

    # Try implicit sets
    if rev in get(repo, b'goods'):
        # i18n: bisect changeset status
        return _(b'good (implicit)')
    if rev in get(repo, b'bads'):
        # i18n: bisect changeset status
        return _(b'bad (implicit)')

    return None


def printresult(ui, repo, state, displayer, nodes, good):
    repo = repo.unfiltered()
    if len(nodes) == 1:
        # narrowed it down to a single revision
        if good:
            ui.write(_(b"The first good revision is:\n"))
        else:
            ui.write(_(b"The first bad revision is:\n"))
        displayer.show(repo[nodes[0]])
        extendnode = extendrange(repo, state, nodes, good)
        if extendnode is not None:
            ui.write(
                _(
                    b'Not all ancestors of this changeset have been'
                    b' checked.\nUse bisect --extend to continue the '
                    b'bisection from\nthe common ancestor, %s.\n'
                )
                % extendnode
            )
    else:
        # multiple possible revisions
        if good:
            ui.write(
                _(
                    b"Due to skipped revisions, the first "
                    b"good revision could be any of:\n"
                )
            )
        else:
            ui.write(
                _(
                    b"Due to skipped revisions, the first "
                    b"bad revision could be any of:\n"
                )
            )
        for n in nodes:
            displayer.show(repo[n])
    displayer.close()