mercurial/worker.py

worker: use os._exit for posix worker in all cases

Like commandserver, the worker should never run other resource cleanup logic.

Previously this is not true for workers if they have exceptions other than
KeyboardInterrupt.

This actually caused a real-world deadlock with remotefilelog:

1. remotefilelog/fileserverclient creates a sshpeer. pipei/o/e get created.
2. worker inherits that sshpeer's pipei/o/e.
3. worker runs sshpeer.cleanup (only happens without os._exit)
4. worker closes pipeo/i, which will normally make the sshpeer read EOF from
   its stdin and exit. But the master process still have pipeo, so no EOF.
5. worker reads pipee (stderr of sshpeer), which never completes because
   the ssh process does not exit, does not close its stderr.
6. master waits for all workers, which never completes because they never
   complete sshpeer.cleanup.

This could also be addressed by closing these fds after fork, which is not
easy because Python 2.x does not have an official "afterfork" hook. Hacking
os.fork is also ugly. Besides, sshpeer is probably not the only troublemarker.

The patch changes _posixworker so all its code paths will use os._exit to
avoid running unwanted resource clean-ups.

# worker.py - master-slave parallelism support
#
# 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.

from __future__ import absolute_import

import errno
import os
import signal
import sys

from .i18n import _
from . import (
    error,
    scmutil,
    util,
)

def countcpus():
    '''try to count the number of CPUs on the system'''

    # posix
    try:
        n = int(os.sysconf('SC_NPROCESSORS_ONLN'))
        if n > 0:
            return n
    except (AttributeError, ValueError):
        pass

    # windows
    try:
        n = int(os.environ['NUMBER_OF_PROCESSORS'])
        if n > 0:
            return n
    except (KeyError, ValueError):
        pass

    return 1

def _numworkers(ui):
    s = ui.config('worker', 'numcpus')
    if s:
        try:
            n = int(s)
            if n >= 1:
                return n
        except ValueError:
            raise error.Abort(_('number of cpus must be an integer'))
    return min(max(countcpus(), 4), 32)

if os.name == 'posix':
    _startupcost = 0.01
else:
    _startupcost = 1e30

def worthwhile(ui, costperop, nops):
    '''try to determine whether the benefit of multiple processes can
    outweigh the cost of starting them'''
    linear = costperop * nops
    workers = _numworkers(ui)
    benefit = linear - (_startupcost * workers + linear / workers)
    return benefit >= 0.15

def worker(ui, costperarg, func, staticargs, args):
    '''run a function, possibly in parallel in multiple worker
    processes.

    returns a progress iterator

    costperarg - cost of a single task

    func - function to run

    staticargs - arguments to pass to every invocation of the function

    args - arguments to split into chunks, to pass to individual
    workers
    '''
    if worthwhile(ui, costperarg, len(args)):
        return _platformworker(ui, func, staticargs, args)
    return func(*staticargs + (args,))

def _posixworker(ui, func, staticargs, args):
    rfd, wfd = os.pipe()
    workers = _numworkers(ui)
    oldhandler = signal.getsignal(signal.SIGINT)
    signal.signal(signal.SIGINT, signal.SIG_IGN)
    pids, problem = set(), [0]
    def killworkers():
        # unregister SIGCHLD handler as all children will be killed. This
        # function shouldn't be interrupted by another SIGCHLD; otherwise pids
        # could be updated while iterating, which would cause inconsistency.
        signal.signal(signal.SIGCHLD, oldchldhandler)
        # if one worker bails, there's no good reason to wait for the rest
        for p in pids:
            try:
                os.kill(p, signal.SIGTERM)
            except OSError as err:
                if err.errno != errno.ESRCH:
                    raise
    def waitforworkers(blocking=True):
        for pid in pids.copy():
            p = st = 0
            while True:
                try:
                    p, st = os.waitpid(pid, (0 if blocking else os.WNOHANG))
                    break
                except OSError as e:
                    if e.errno == errno.EINTR:
                        continue
                    elif e.errno == errno.ECHILD:
                        # child would already be reaped, but pids yet been
                        # updated (maybe interrupted just after waitpid)
                        pids.discard(pid)
                        break
                    else:
                        raise
            if p:
                pids.discard(p)
                st = _exitstatus(st)
            if st and not problem[0]:
                problem[0] = st
    def sigchldhandler(signum, frame):
        waitforworkers(blocking=False)
        if problem[0]:
            killworkers()
    oldchldhandler = signal.signal(signal.SIGCHLD, sigchldhandler)
    for pargs in partition(args, workers):
        pid = os.fork()
        if pid == 0:
            signal.signal(signal.SIGINT, oldhandler)
            signal.signal(signal.SIGCHLD, oldchldhandler)

            def workerfunc():
                os.close(rfd)
                for i, item in func(*(staticargs + (pargs,))):
                    os.write(wfd, '%d %s\n' % (i, item))

            # make sure we use os._exit in all code paths. otherwise the worker
            # may do some clean-ups which could cause surprises like deadlock.
            # see sshpeer.cleanup for example.
            try:
                scmutil.callcatch(ui, workerfunc)
            except KeyboardInterrupt:
                os._exit(255)
            except: # never return, therefore no re-raises
                try:
                    ui.traceback()
                finally:
                    os._exit(255)
            else:
                os._exit(0)
        pids.add(pid)
    os.close(wfd)
    fp = os.fdopen(rfd, 'rb', 0)
    def cleanup():
        signal.signal(signal.SIGINT, oldhandler)
        waitforworkers()
        signal.signal(signal.SIGCHLD, oldchldhandler)
        status = problem[0]
        if status:
            if status < 0:
                os.kill(os.getpid(), -status)
            sys.exit(status)
    try:
        for line in util.iterfile(fp):
            l = line.split(' ', 1)
            yield int(l[0]), l[1][:-1]
    except: # re-raises
        killworkers()
        cleanup()
        raise
    cleanup()

def _posixexitstatus(code):
    '''convert a posix exit status into the same form returned by
    os.spawnv

    returns None if the process was stopped instead of exiting'''
    if os.WIFEXITED(code):
        return os.WEXITSTATUS(code)
    elif os.WIFSIGNALED(code):
        return -os.WTERMSIG(code)

if os.name != 'nt':
    _platformworker = _posixworker
    _exitstatus = _posixexitstatus

def partition(lst, nslices):
    '''partition a list into N slices of roughly equal size

    The current strategy takes every Nth element from the input. If
    we ever write workers that need to preserve grouping in input
    we should consider allowing callers to specify a partition strategy.

    mpm is not a fan of this partitioning strategy when files are involved.
    In his words:

        Single-threaded Mercurial makes a point of creating and visiting
        files in a fixed order (alphabetical). When creating files in order,
        a typical filesystem is likely to allocate them on nearby regions on
        disk. Thus, when revisiting in the same order, locality is maximized
        and various forms of OS and disk-level caching and read-ahead get a
        chance to work.

        This effect can be quite significant on spinning disks. I discovered it
        circa Mercurial v0.4 when revlogs were named by hashes of filenames.
        Tarring a repo and copying it to another disk effectively randomized
        the revlog ordering on disk by sorting the revlogs by hash and suddenly
        performance of my kernel checkout benchmark dropped by ~10x because the
        "working set" of sectors visited no longer fit in the drive's cache and
        the workload switched from streaming to random I/O.

        What we should really be doing is have workers read filenames from a
        ordered queue. This preserves locality and also keeps any worker from
        getting more than one file out of balance.
    '''
    for i in range(nslices):
        yield lst[i::nslices]