view hgext/blackbox.py @ 42743:8c9a6adec67a

rust-discovery: using the children cache in add_missing

The DAG range computation often needs to get back to very old
revisions, and turns out to be disproportionately long, given
that the end goal is to remove the descendents of the given
missing revisons from the undecided set.

The fast iteration capabilities available in the Rust case make
it possible to avoid the DAG range entirely, at the cost of
precomputing the children cache, and to simply iterate on
children of the given missing revisions.

This is a case where staying on the same side of the interface
between the two languages has clear benefits.

On discoveries with initial undecided sets
small enough to bypass sampling entirely, the total cost of
computing the children cache and the subsequent iteration
becomes better than the Python + C counterpart, which relies on
reachableroots2.

For example, on a repo with more than one million revisions with
an initial undecided set of 11 elements, we get these figures:

Rust version with simple iteration
  addcommons: 57.287us
  first undecided computation: 184.278334ms
  first children cache computation: 131.056us
  addmissings iteration: 42.766us
  first addinfo total: 185.24 ms

Python + C version
  first addcommons: 0.29 ms
  addcommons 0.21 ms
  first undecided computation 191.35 ms
  addmissings 45.75 ms
  first addinfo total: 237.77 ms

On discoveries with large undecided sets, the initial price paid
makes the first addinfo slower than the Python + C version,
but that's more than compensated by the gain in sampling and
subsequent iterations.
Here's an extreme example with an undecided set of a million revisions:

Rust version:
  first undecided computation: 293.842629ms
  first children cache computation: 407.911297ms
  addmissings iteration: 34.312869ms
  first addinfo total: 776.02 ms
  taking initial sample
  query 2: sampling time: 1318.38 ms
  query 2; still undecided: 1005013, sample size is: 200
  addmissings: 143.062us

Python + C version:
  first undecided computation 298.13 ms
  addmissings 80.13 ms
  first addinfo total: 399.62 ms
  taking initial sample
  query 2: sampling time: 3957.23 ms
  query 2; still undecided: 1005013, sample size is: 200
  addmissings 52.88 ms

Differential Revision: https://phab.mercurial-scm.org/D6428

# blackbox.py - log repository events to a file for post-mortem debugging
#
# Copyright 2010 Nicolas Dumazet
# Copyright 2013 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.

"""log repository events to a blackbox for debugging

Logs event information to .hg/blackbox.log to help debug and diagnose problems.
The events that get logged can be configured via the blackbox.track and
blackbox.ignore config keys.

Examples::

  [blackbox]
  track = *
  ignore = pythonhook
  # dirty is *EXPENSIVE* (slow);
  # each log entry indicates `+` if the repository is dirty, like :hg:`id`.
  dirty = True
  # record the source of log messages
  logsource = True

  [blackbox]
  track = command, commandfinish, commandexception, exthook, pythonhook

  [blackbox]
  track = incoming

  [blackbox]
  # limit the size of a log file
  maxsize = 1.5 MB
  # rotate up to N log files when the current one gets too big
  maxfiles = 3

  [blackbox]
  # Include nanoseconds in log entries with %f (see Python function
  # datetime.datetime.strftime)
  date-format = '%Y-%m-%d @ %H:%M:%S.%f'

"""

from __future__ import absolute_import

import re

from mercurial.i18n import _
from mercurial.node import hex

from mercurial import (
    encoding,
    loggingutil,
    registrar,
)
from mercurial.utils import (
    dateutil,
    procutil,
)

# Note for extension authors: ONLY specify testedwith = 'ships-with-hg-core' for
# extensions which SHIP WITH MERCURIAL. Non-mainline extensions should
# be specifying the version(s) of Mercurial they are tested with, or
# leave the attribute unspecified.
testedwith = 'ships-with-hg-core'

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

configtable = {}
configitem = registrar.configitem(configtable)

configitem('blackbox', 'dirty',
    default=False,
)
configitem('blackbox', 'maxsize',
    default='1 MB',
)
configitem('blackbox', 'logsource',
    default=False,
)
configitem('blackbox', 'maxfiles',
    default=7,
)
configitem('blackbox', 'track',
    default=lambda: ['*'],
)
configitem('blackbox', 'ignore',
    default=lambda: ['chgserver', 'cmdserver', 'extension'],
)
configitem('blackbox', 'date-format',
    default='%Y/%m/%d %H:%M:%S',
)

_lastlogger = loggingutil.proxylogger()

class blackboxlogger(object):
    def __init__(self, ui, repo):
        self._repo = repo
        self._trackedevents = set(ui.configlist('blackbox', 'track'))
        self._ignoredevents = set(ui.configlist('blackbox', 'ignore'))
        self._maxfiles = ui.configint('blackbox', 'maxfiles')
        self._maxsize = ui.configbytes('blackbox', 'maxsize')
        self._inlog = False

    def tracked(self, event):
        return ((b'*' in self._trackedevents
                 and event not in self._ignoredevents)
                or event in self._trackedevents)

    def log(self, ui, event, msg, opts):
        # self._log() -> ctx.dirty() may create new subrepo instance, which
        # ui is derived from baseui. So the recursion guard in ui.log()
        # doesn't work as it's local to the ui instance.
        if self._inlog:
            return
        self._inlog = True
        try:
            self._log(ui, event, msg, opts)
        finally:
            self._inlog = False

    def _log(self, ui, event, msg, opts):
        default = ui.configdate('devel', 'default-date')
        date = dateutil.datestr(default, ui.config('blackbox', 'date-format'))
        user = procutil.getuser()
        pid = '%d' % procutil.getpid()
        changed = ''
        ctx = self._repo[None]
        parents = ctx.parents()
        rev = ('+'.join([hex(p.node()) for p in parents]))
        if (ui.configbool('blackbox', 'dirty') and
            ctx.dirty(missing=True, merge=False, branch=False)):
            changed = '+'
        if ui.configbool('blackbox', 'logsource'):
            src = ' [%s]' % event
        else:
            src = ''
        try:
            fmt = '%s %s @%s%s (%s)%s> %s'
            args = (date, user, rev, changed, pid, src, msg)
            with loggingutil.openlogfile(
                    ui, self._repo.vfs, name='blackbox.log',
                    maxfiles=self._maxfiles, maxsize=self._maxsize) as fp:
                fp.write(fmt % args)
        except (IOError, OSError) as err:
            # deactivate this to avoid failed logging again
            self._trackedevents.clear()
            ui.debug('warning: cannot write to blackbox.log: %s\n' %
                     encoding.strtolocal(err.strerror))
            return
        _lastlogger.logger = self

def uipopulate(ui):
    ui.setlogger(b'blackbox', _lastlogger)

def reposetup(ui, repo):
    # During 'hg pull' a httppeer repo is created to represent the remote repo.
    # It doesn't have a .hg directory to put a blackbox in, so we don't do
    # the blackbox setup for it.
    if not repo.local():
        return

    # Since blackbox.log is stored in the repo directory, the logger should be
    # instantiated per repository.
    logger = blackboxlogger(ui, repo)
    ui.setlogger(b'blackbox', logger)

    # Set _lastlogger even if ui.log is not called. This gives blackbox a
    # fallback place to log
    if _lastlogger.logger is None:
        _lastlogger.logger = logger

    repo._wlockfreeprefix.add('blackbox.log')

@command('blackbox',
    [('l', 'limit', 10, _('the number of events to show')),
    ],
    _('hg blackbox [OPTION]...'),
    helpcategory=command.CATEGORY_MAINTENANCE,
    helpbasic=True)
def blackbox(ui, repo, *revs, **opts):
    '''view the recent repository events
    '''

    if not repo.vfs.exists('blackbox.log'):
        return

    limit = opts.get(r'limit')
    fp = repo.vfs('blackbox.log', 'r')
    lines = fp.read().split('\n')

    count = 0
    output = []
    for line in reversed(lines):
        if count >= limit:
            break

        # count the commands by matching lines like: 2013/01/23 19:13:36 root>
        if re.match(br'^\d{4}/\d{2}/\d{2} \d{2}:\d{2}:\d{2} .*> .*', line):
            count += 1
        output.append(line)

    ui.status('\n'.join(reversed(output)))