tests/bruterebase.py

cborutil: change buffering strategy

Profiling revealed that we were spending a lot of time on the
line that was concatenating the old buffer with the incoming data
when attempting to decode long byte strings, such as manifest
revisions.

Essentially, we were feeding N chunks of size len(X) << len(Y) into
decode() and continuously allocating a new, larger buffer to hold
the undecoded input. This created substantial memory churn and
slowed down execution.

Changing the code to aggregate pending chunks in a list until we
have enough data to fully decode the next atom makes things much
more efficient.

I don't have exact data, but I recall the old code spending >1s
on manifest fulltexts from the mozilla-unified repo. The new code
doesn't significantly appear in profile output.

Differential Revision: https://phab.mercurial-scm.org/D4854

# bruterebase.py - brute force rebase testing
#
# Copyright 2017 Facebook, Inc.
#
# 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

from mercurial import (
    error,
    registrar,
    revsetlang,
)

from hgext import rebase

try:
    xrange
except NameError:
    xrange = range

cmdtable = {}
command = registrar.command(cmdtable)

@command(b'debugbruterebase')
def debugbruterebase(ui, repo, source, dest):
    """for every non-empty subset of source, run rebase -r subset -d dest

    Print one line summary for each subset. Assume obsstore is enabled.
    """
    srevs = list(repo.revs(source))

    with repo.wlock(), repo.lock():
        repolen = len(repo)
        cl = repo.changelog

        def getdesc(rev):
            result = cl.changelogrevision(rev).description
            if rev >= repolen:
                result += b"'"
            return result

        for i in xrange(1, 2 ** len(srevs)):
            subset = [rev for j, rev in enumerate(srevs) if i & (1 << j) != 0]
            spec = revsetlang.formatspec(b'%ld', subset)
            tr = repo.transaction(b'rebase')
            tr._report = lambda x: 0 # hide "transaction abort"

            ui.pushbuffer()
            try:
                rebase.rebase(ui, repo, dest=dest, rev=[spec])
            except error.Abort as ex:
                summary = b'ABORT: %s' % ex
            except Exception as ex:
                summary = b'CRASH: %s' % ex
            else:
                # short summary about new nodes
                cl = repo.changelog
                descs = []
                for rev in xrange(repolen, len(repo)):
                    desc = b'%s:' % getdesc(rev)
                    for prev in cl.parentrevs(rev):
                        if prev > -1:
                            desc += getdesc(prev)
                    descs.append(desc)
                descs.sort()
                summary = b' '.join(descs)
            ui.popbuffer()
            repo.vfs.tryunlink(b'rebasestate')

            subsetdesc = b''.join(getdesc(rev) for rev in subset)
            ui.write((b'%s: %s\n') % (subsetdesc.rjust(len(srevs)), summary))
            tr.abort()