Martin Jansa 7df5f405ce bitbake: prserv: handle PRSERV_HOST = "127.0.0.1:0" the same as "localhost:0"
* When using PRSERV_HOST = "localhost:0" inside
  Docker container (tested with ubuntu 20.04 and 21.04) the
  self.loop.run_until_complete never completed, so self.address
  wasn't ever assigned few lines bellow and then
    self.port = int(self.prserv.address.rsplit(':', 1)[1])
  in lib/prserv/serv.py caused a bit ugly exception:

bitbake@599696cd20aa:~/nodistro/honister$ bitbake -k pkgconfig-native
Traceback (most recent call last):
  File "/OE/nodistro/honister/bitbake/bin/bitbake", line 35, in <module>
    sys.exit(bitbake_main(BitBakeConfigParameters(sys.argv),
  File "/OE/nodistro/honister/bitbake/lib/bb/main.py", line 385, in bitbake_main
    return ui_module.main(server_connection.connection, server_connection.events,
  File "/OE/nodistro/honister/bitbake/lib/bb/ui/knotty.py", line 397, in main
    params.updateToServer(server, os.environ.copy())
  File "/OE/nodistro/honister/bitbake/lib/bb/cookerdata.py", line 75, in updateToServer
    raise Exception("Unable to update the server configuration with local parameters: %s" % error)
Exception: Unable to update the server configuration with local parameters: Traceback (most recent call last):
  File "/OE/nodistro/honister/bitbake/lib/bb/command.py", line 90, in runCommand
    result = command_method(self, commandline)
  File "/OE/nodistro/honister/bitbake/lib/bb/command.py", line 286, in updateConfig
    command.cooker.updateConfigOpts(options, environment, cmdline)
  File "/OE/nodistro/honister/bitbake/lib/bb/cooker.py", line 491, in updateConfigOpts
    self.reset()
  File "/OE/nodistro/honister/bitbake/lib/bb/cooker.py", line 1717, in reset
    self.handlePRServ()
  File "/OE/nodistro/honister/bitbake/lib/bb/cooker.py", line 383, in handlePRServ
    self.prhost = prserv.serv.auto_start(self.data)
  File "/OE/nodistro/honister/bitbake/lib/prserv/serv.py", line 318, in auto_start
    singleton.start()
  File "/OE/nodistro/honister/bitbake/lib/prserv/serv.py", line 133, in start
    self.port = int(self.prserv.address.rsplit(':', 1)[1])
AttributeError: 'NoneType' object has no attribute 'rsplit'

* the issue was caused by "localhost" being resolved as IPv6 address ::1
  and then asyncio failing to bind it, the same is reproducible with hashserv, but
  hashserve at least shows nice error message:
    bitbake$ bitbake-hashserv -l DEBUG -b localhost:0
    Traceback (most recent call last):
      File "/OE/nodistro/honister/bitbake/bin/bitbake-hashserv", line 59, in <module>
        ret = main()
      File "/OE/nodistro/honister/bitbake/bin/bitbake-hashserv", line 53, in main
        server.serve_forever()
      File "/OE/nodistro/honister/bitbake/lib/bb/asyncrpc/serv.py", line 233, in serve_forever
        self.start()
      File "/OE/nodistro/honister/bitbake/lib/bb/asyncrpc/serv.py", line 144, in start_tcp
        self.server = self.loop.run_until_complete(server_coro)
      File "/usr/lib/python3.8/asyncio/base_events.py", line 616, in run_until_complete
        return future.result()
      File "/usr/lib/python3.8/asyncio/streams.py", line 94, in start_server
        return await loop.create_server(factory, host, port, **kwds)
      File "/usr/lib/python3.8/asyncio/base_events.py", line 1463, in create_server
        raise OSError(err.errno, 'error while attempting '
    OSError: [Errno 99] error while attempting to bind on address ('::1', 0, 0, 0): cannot assign requested address

* or by bitbake-prserv in prserv.log:
    bitbake$ bitbake-prserv --start --host=localhost --port=42005
    bitbake$ cat prserv.log
    Traceback (most recent call last):
      File "/OE/nodistro/honister/bitbake/bin/bitbake-prserv", line 55, in <module>
        ret = main()
      File "/OE/nodistro/honister/bitbake/bin/bitbake-prserv", line 46, in main
        ret=prserv.serv.start_daemon(options.dbfile, options.host, options.port,os.path.abspath(options.logfile), options.read_only)
      File "/OE/nodistro/honister/bitbake/lib/prserv/serv.py", line 226, in start_daemon
        run_as_daemon(daemon_main, pidfile, os.path.abspath(logfile))
      File "/OE/nodistro/honister/bitbake/lib/prserv/serv.py", line 202, in run_as_daemon
        func()
      File "/OE/nodistro/honister/bitbake/lib/prserv/serv.py", line 224, in daemon_main
        server.serve_forever()
      File "/OE/nodistro/honister/bitbake/lib/bb/asyncrpc/serv.py", line 233, in serve_forever
        self.start()
      File "/OE/nodistro/honister/bitbake/lib/bb/asyncrpc/serv.py", line 144, in start_tcp
        self.server = self.loop.run_until_complete(server_coro)
      File "/usr/lib/python3.8/asyncio/base_events.py", line 616, in run_until_complete
        return future.result()
      File "/usr/lib/python3.8/asyncio/streams.py", line 94, in start_server
        return await loop.create_server(factory, host, port, **kwds)
      File "/usr/lib/python3.8/asyncio/base_events.py", line 1463, in create_server
        raise OSError(err.errno, 'error while attempting '
    OSError: [Errno 99] error while attempting to bind on address ('::1', 42005, 0, 0): cannot assign requested address

* while 127.0.0.1 works fine:
    bitbake$ bitbake-prserv --start --host=127.0.0.1 --port=42005
    bitbake$ cat prserv.log
    DEBUG: Listening on ('127.0.0.1', 42005)
    2021-08-26 22:28:05,828 Listening on ('127.0.0.1', 42005)
    DEBUG: Opening PRServ database 'file:/OE/nodistro/honister/prserv.sqlite3'
    2021-08-26 22:28:05,829 Opening PRServ database 'file:/OE/nodistro/honister/prserv.sqlite3'
    NOTE: Started PRServer with DBfile: /OE/nodistro/honister/prserv.sqlite3, Address: 127.0.0.1:42005, PID: 39
    2021-08-26 22:28:05,831 Started PRServer with DBfile: /OE/nodistro/honister/prserv.sqlite3, Address: 127.0.0.1:42005, PID: 39

  but 127.0.0.1:0 wasn't handled as "autostart" like localhost:0 is
  update is_local_special to allow that

* /etc/hosts file generated by docker contails localhost for both IPv4 and IPv6:
  $ grep localhost /etc/hosts
  127.0.0.1       localhost
  ::1     localhost ip6-localhost ip6-loopback

  even when ipv6 is disabled in dockerd as reported in:
  https://github.com/docker/for-linux/issues/250

* add a check for self.prserv.address to provide better error message:
  ERROR: Unable to start PR Server, exitting
  when something bad happens, but in this case you still need to read
  bitbake-cookerdaemon.log to see the actuall error, in this case:

90 22:30:39.008441 --- Starting bitbake server pid 90 at 2021-08-26 22:30:39.008419 ---
90 22:30:39.023734 Started bitbake server pid 90
90 22:30:39.024286 Entering server connection loop
90 22:30:39.024753 Accepting [<socket.socket fd=6, family=AddressFamily.AF_UNIX, type=SocketKind.SOCK_STREAM, proto=0, laddr=bitbake.sock>] ([])
90 22:30:39.026314 Processing Client
90 22:30:39.026456 Connecting Client
90 22:30:39.027509 Running command ['setFeatures', [2]]
90 22:30:39.027757 Command Completed
90 22:30:39.028711 Running command ['updateConfig', {'abort': False, 'force': False, 'invalidate_stamp': None, 'dry_run': False, 'dump_signatures': [], 'extra_assume_provided': [], 'profile': False, 'prefile': [], 'postfile': [], 'server_timeout': None, 'nosetscene': False, 'setsceneonly': False, 'skipsetscene': False, 'runall': None, 'runonly': None, 'writeeventlog': None, 'build_verbose_shell': False, 'build_verbose_stdout': False, 'default_loglevel': 20, 'debug_domains': {}}, {'DISTRO': '', 'PWD': '/OE/nodistro/honister', 'HOME': '/OE', 'MACHINE': 'qemux86', 'BB_ENV_EXTRAWHITE': 'MACHINE DISTRO TCMODE TCLIBC http_proxy ftp_proxy https_proxy all_proxy ALL_PROXY no_proxy SSH_AGENT_PID SSH_AUTH_SOCK BB_SRCREV_POLICY SDKMACHINE BB_NUMBER_THREADS PARALLEL_MAKE GIT_PROXY_COMMAND GIT_PROXY_IGNORE SOCKS5_PASSWD SOCKS5_USER WEBOS_DISTRO_BUILD_ID PSEUDO_DISABLED PSEUDO_BUILD', 'PATH': '/OE/nodistro/honister/oe-core/scripts:/OE/nodistro/honister/bitbake/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin', 'LC_ALL': 'en_US.UTF-8', 'MACHINES': 'qemux86', 'HOSTNAME': '6a439759e3c6', 'TOPDIR': '/OE/nodistro/honister', 'LANG': 'en_US.UTF-8', 'TERM': 'xterm', 'SHLVL': '1', 'BITBAKE_HOME': '/OE', 'BUILDDIR': '/OE/nodistro/honister/BUILD', 'OLDPWD': '/OE/nodistro/honister/bitbake', '_': '/OE/nodistro/honister/bitbake/bin/bitbake'}, ['/OE/nodistro/honister/bitbake/bin/bitbake', '-k', 'zlib-native']]
Process Process-1:
Traceback (most recent call last):
  File "/usr/lib/python3.8/multiprocessing/process.py", line 315, in _bootstrap
    self.run()
  File "/usr/lib/python3.8/multiprocessing/process.py", line 108, in run
    self._target(*self._args, **self._kwargs)
  File "/OE/nodistro/honister/bitbake/lib/bb/asyncrpc/serv.py", line 255, in run
    self.start()
  File "/OE/nodistro/honister/bitbake/lib/bb/asyncrpc/serv.py", line 144, in start_tcp
    self.server = self.loop.run_until_complete(server_coro)
  File "/usr/lib/python3.8/asyncio/base_events.py", line 616, in run_until_complete
    return future.result()
  File "/usr/lib/python3.8/asyncio/streams.py", line 94, in start_server
    return await loop.create_server(factory, host, port, **kwds)
  File "/usr/lib/python3.8/asyncio/base_events.py", line 1463, in create_server
    raise OSError(err.errno, 'error while attempting '
OSError: [Errno 99] error while attempting to bind on address ('::1', 0, 0, 0): cannot assign requested address
90 22:30:39.530037 Command Completed
90 22:30:39.530913 Processing Client
90 22:30:39.531023 Disconnecting Client
90 22:30:39.531638 No timeout, exiting.
90 22:30:39.632137 Exiting
90 22:30:39.637562 Original lockfile contents: ['90\n']
90 22:30:39.638107 Exiting as we could obtain the lock

(Bitbake rev: c2cdda0e5fc720c60d3b8537fc702cb118981bd2)

Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
2021-09-01 18:56:25 +01:00

361 lines
10 KiB
Python

#
# SPDX-License-Identifier: GPL-2.0-only
#
import os,sys,logging
import signal, time
import socket
import io
import sqlite3
import prserv
import prserv.db
import errno
import bb.asyncrpc
logger = logging.getLogger("BitBake.PRserv")
PIDPREFIX = "/tmp/PRServer_%s_%s.pid"
singleton = None
class PRServerClient(bb.asyncrpc.AsyncServerConnection):
def __init__(self, reader, writer, table, read_only):
super().__init__(reader, writer, 'PRSERVICE', logger)
self.handlers.update({
'get-pr': self.handle_get_pr,
'import-one': self.handle_import_one,
'export': self.handle_export,
'is-readonly': self.handle_is_readonly,
})
self.table = table
self.read_only = read_only
def validate_proto_version(self):
return (self.proto_version == (1, 0))
async def dispatch_message(self, msg):
try:
await super().dispatch_message(msg)
except:
self.table.sync()
raise
self.table.sync_if_dirty()
async def handle_get_pr(self, request):
version = request['version']
pkgarch = request['pkgarch']
checksum = request['checksum']
response = None
try:
value = self.table.getValue(version, pkgarch, checksum)
response = {'value': value}
except prserv.NotFoundError:
logger.error("can not find value for (%s, %s)",version, checksum)
except sqlite3.Error as exc:
logger.error(str(exc))
self.write_message(response)
async def handle_import_one(self, request):
response = None
if not self.read_only:
version = request['version']
pkgarch = request['pkgarch']
checksum = request['checksum']
value = request['value']
value = self.table.importone(version, pkgarch, checksum, value)
if value is not None:
response = {'value': value}
self.write_message(response)
async def handle_export(self, request):
version = request['version']
pkgarch = request['pkgarch']
checksum = request['checksum']
colinfo = request['colinfo']
try:
(metainfo, datainfo) = self.table.export(version, pkgarch, checksum, colinfo)
except sqlite3.Error as exc:
logger.error(str(exc))
metainfo = datainfo = None
response = {'metainfo': metainfo, 'datainfo': datainfo}
self.write_message(response)
async def handle_is_readonly(self, request):
response = {'readonly': self.read_only}
self.write_message(response)
class PRServer(bb.asyncrpc.AsyncServer):
def __init__(self, dbfile, read_only=False):
super().__init__(logger)
self.dbfile = dbfile
self.table = None
self.read_only = read_only
def accept_client(self, reader, writer):
return PRServerClient(reader, writer, self.table, self.read_only)
def _serve_forever(self):
self.db = prserv.db.PRData(self.dbfile, read_only=self.read_only)
self.table = self.db["PRMAIN"]
logger.info("Started PRServer with DBfile: %s, Address: %s, PID: %s" %
(self.dbfile, self.address, str(os.getpid())))
super()._serve_forever()
self.table.sync_if_dirty()
self.db.disconnect()
def signal_handler(self):
super().signal_handler()
if self.table:
self.table.sync()
class PRServSingleton(object):
def __init__(self, dbfile, logfile, host, port):
self.dbfile = dbfile
self.logfile = logfile
self.host = host
self.port = port
def start(self):
self.prserv = PRServer(self.dbfile)
self.prserv.start_tcp_server(self.host, self.port)
self.process = self.prserv.serve_as_process()
if not self.prserv.address:
raise PRServiceConfigError
if not self.port:
self.port = int(self.prserv.address.rsplit(':', 1)[1])
def run_as_daemon(func, pidfile, logfile):
"""
See Advanced Programming in the UNIX, Sec 13.3
"""
try:
pid = os.fork()
if pid > 0:
os.waitpid(pid, 0)
#parent return instead of exit to give control
return pid
except OSError as e:
raise Exception("%s [%d]" % (e.strerror, e.errno))
os.setsid()
"""
fork again to make sure the daemon is not session leader,
which prevents it from acquiring controlling terminal
"""
try:
pid = os.fork()
if pid > 0: #parent
os._exit(0)
except OSError as e:
raise Exception("%s [%d]" % (e.strerror, e.errno))
os.chdir("/")
sys.stdout.flush()
sys.stderr.flush()
# We could be called from a python thread with io.StringIO as
# stdout/stderr or it could be 'real' unix fd forking where we need
# to physically close the fds to prevent the program launching us from
# potentially hanging on a pipe. Handle both cases.
si = open('/dev/null', 'r')
try:
os.dup2(si.fileno(),sys.stdin.fileno())
except (AttributeError, io.UnsupportedOperation):
sys.stdin = si
so = open(logfile, 'a+')
try:
os.dup2(so.fileno(),sys.stdout.fileno())
except (AttributeError, io.UnsupportedOperation):
sys.stdout = so
try:
os.dup2(so.fileno(),sys.stderr.fileno())
except (AttributeError, io.UnsupportedOperation):
sys.stderr = so
# Clear out all log handlers prior to the fork() to avoid calling
# event handlers not part of the PRserver
for logger_iter in logging.Logger.manager.loggerDict.keys():
logging.getLogger(logger_iter).handlers = []
# Ensure logging makes it to the logfile
streamhandler = logging.StreamHandler()
streamhandler.setLevel(logging.DEBUG)
formatter = bb.msg.BBLogFormatter("%(levelname)s: %(message)s")
streamhandler.setFormatter(formatter)
logger.addHandler(streamhandler)
# write pidfile
pid = str(os.getpid())
with open(pidfile, 'w') as pf:
pf.write("%s\n" % pid)
func()
os.remove(pidfile)
os._exit(0)
def start_daemon(dbfile, host, port, logfile, read_only=False):
ip = socket.gethostbyname(host)
pidfile = PIDPREFIX % (ip, port)
try:
with open(pidfile) as pf:
pid = int(pf.readline().strip())
except IOError:
pid = None
if pid:
sys.stderr.write("pidfile %s already exist. Daemon already running?\n"
% pidfile)
return 1
dbfile = os.path.abspath(dbfile)
def daemon_main():
server = PRServer(dbfile, read_only=read_only)
server.start_tcp_server(host, port)
server.serve_forever()
run_as_daemon(daemon_main, pidfile, os.path.abspath(logfile))
return 0
def stop_daemon(host, port):
import glob
ip = socket.gethostbyname(host)
pidfile = PIDPREFIX % (ip, port)
try:
with open(pidfile) as pf:
pid = int(pf.readline().strip())
except IOError:
pid = None
if not pid:
# when server starts at port=0 (i.e. localhost:0), server actually takes another port,
# so at least advise the user which ports the corresponding server is listening
ports = []
portstr = ""
for pf in glob.glob(PIDPREFIX % (ip,'*')):
bn = os.path.basename(pf)
root, _ = os.path.splitext(bn)
ports.append(root.split('_')[-1])
if len(ports):
portstr = "Wrong port? Other ports listening at %s: %s" % (host, ' '.join(ports))
sys.stderr.write("pidfile %s does not exist. Daemon not running? %s\n"
% (pidfile,portstr))
return 1
try:
if is_running(pid):
print("Sending SIGTERM to pr-server.")
os.kill(pid, signal.SIGTERM)
time.sleep(0.1)
if os.path.exists(pidfile):
os.remove(pidfile)
except OSError as e:
err = str(e)
if err.find("No such process") <= 0:
raise e
return 0
def is_running(pid):
try:
os.kill(pid, 0)
except OSError as err:
if err.errno == errno.ESRCH:
return False
return True
def is_local_special(host, port):
if (host == 'localhost' or host == '127.0.0.1') and not port:
return True
else:
return False
class PRServiceConfigError(Exception):
pass
def auto_start(d):
global singleton
host_params = list(filter(None, (d.getVar('PRSERV_HOST') or '').split(':')))
if not host_params:
# Shutdown any existing PR Server
auto_shutdown()
return None
if len(host_params) != 2:
# Shutdown any existing PR Server
auto_shutdown()
logger.critical('\n'.join(['PRSERV_HOST: incorrect format',
'Usage: PRSERV_HOST = "<hostname>:<port>"']))
raise PRServiceConfigError
host = host_params[0].strip().lower()
port = int(host_params[1])
if is_local_special(host, port):
import bb.utils
cachedir = (d.getVar("PERSISTENT_DIR") or d.getVar("CACHE"))
if not cachedir:
logger.critical("Please set the 'PERSISTENT_DIR' or 'CACHE' variable")
raise PRServiceConfigError
dbfile = os.path.join(cachedir, "prserv.sqlite3")
logfile = os.path.join(cachedir, "prserv.log")
if singleton:
if singleton.dbfile != dbfile:
# Shutdown any existing PR Server as doesn't match config
auto_shutdown()
if not singleton:
bb.utils.mkdirhier(cachedir)
singleton = PRServSingleton(os.path.abspath(dbfile), os.path.abspath(logfile), host, port)
singleton.start()
if singleton:
host = singleton.host
port = singleton.port
try:
ping(host, port)
return str(host) + ":" + str(port)
except Exception:
logger.critical("PRservice %s:%d not available" % (host, port))
raise PRServiceConfigError
def auto_shutdown():
global singleton
if singleton and singleton.process:
singleton.process.terminate()
singleton.process.join()
singleton = None
def ping(host, port):
from . import client
conn = client.PRClient()
conn.connect_tcp(host, port)
return conn.ping()
def connect(host, port):
from . import client
global singleton
if host.strip().lower() == 'localhost' and not port:
host = 'localhost'
port = singleton.port
conn = client.PRClient()
conn.connect_tcp(host, port)
return conn