2
0
mirror of https://opendev.org/x/pyghmi synced 2025-07-11 06:51:32 +00:00
Files
pyghmi/pyghmi/ipmi/private/session.py
Jarrod Johnson c0f4959ed7 Change long timeout to select() from poll()
Change from poll() to select() for the potentially long wait in ipmi
event loop.  This is to facilitate interop with being monkey patched
with eventlet, as eventlet does not currently monkey patch poll(), but
does monkey patch select()

Change-Id: I8d6f42193ee82e2c5de7ad112232c2c77ef78825
2013-08-08 14:19:09 -04:00

1109 lines
48 KiB
Python

# vim: tabstop=4 shiftwidth=4 softtabstop=4
# Copyright 2013 IBM Corporation
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# This represents the low layer message framing portion of IPMI
import atexit
import collections
import hashlib
import os
import random
import select
import socket
import struct
import time
from Crypto.Cipher import AES
from Crypto.Hash import HMAC
from Crypto.Hash import SHA
from pyghmi.ipmi.private import constants
initialtimeout = 0.5 # minimum timeout for first packet to retry in any given
# session. This will be randomized to stagger out retries
# in case of congestion
def _monotonic_time():
"""Provides a monotonic timer
This code is concerned with relative, not absolute time.
This function facilitates that prior to python 3.3
"""
# Python does not provide one until 3.3, so we make do
# for most OSes, os.times()[4] works well.
# for microsoft, GetTickCount64
if (os.name == "posix"):
return os.times()[4]
else: # last resort, non monotonic time
return time.time()
#TODO(jbjohnso): Windows variant
def _aespad(data):
"""ipmi demands a certain pad scheme,
per table 13-20 AES-CBC encrypted payload fields.
"""
newdata = list(data)
currlen = len(data) + 1 # need to count the pad length field as well
neededpad = currlen % 16
if neededpad: # if it happens to be zero, hurray, but otherwise invert the
# sense of the padding
neededpad = 16 - neededpad
padval = 1
while padval <= neededpad:
newdata.append(padval)
padval += 1
newdata.append(neededpad)
return newdata
def call_with_optional_args(callback, *args):
"""In order to simplify things, in a number of places there is a callback
facility and optional arguments to pass in. An object-oriented caller may
find the additional argument needless. Allow them to ignore it by skipping
the argument if None.
"""
newargs = []
for arg in args:
if arg is not None:
newargs.append(arg)
callback(*newargs)
def get_ipmi_error(response, suffix=""):
if 'error' in response:
return response['error'] + suffix
code = response['code']
if code == 0:
return False
command = response['command']
netfn = response['netfn']
if ((netfn, command) in constants.command_completion_codes
and code in constants.command_completion_codes[(netfn, command)]):
res = constants.command_completion_codes[(netfn, command)][code]
res += suffix
elif code in constants.ipmi_completion_codes:
res = constants.ipmi_completion_codes[code] + suffix
else:
res = "Unknown code " + code + " encountered"
return res
class Session:
"""A class to manage common IPMI session logistics
Almost all developers should not worry about this class and instead be
looking toward ipmi.Command and ipmi.Console.
For those that do have to worry, the main interesting thing is that the
event loop can go one of two ways. Either a larger manager can query using
class methods
the soonest timeout deadline and the filehandles to poll and assume
responsibility for the polling, or it can register filehandles to be
watched. This is primarily of interest to Console class, which may have an
input filehandle to watch and can pass it to Session.
:param bmc: hostname or ip address of the BMC
:param userid: username to use to connect
:param password: password to connect to the BMC
:param kg: optional parameter if BMC requires Kg be set
:param port: UDP port to communicate with, pretty much always 623
:param onlogon: callback to receive notification of login completion
"""
poller = select.poll()
ipmipoller = select.poll()
_external_handlers = {}
bmc_handlers = {}
waiting_sessions = {}
keepalive_sessions = {}
peeraddr_to_nodes = {}
# Upon exit of python, make sure we play nice with BMCs by assuring closed
# sessions for all that we tracked
@classmethod
def _cleanup(cls):
for session in cls.bmc_handlers.itervalues():
session.logout()
@classmethod
def _createsocket(cls):
atexit.register(cls._cleanup)
cls.socket = socket.socket(socket.AF_INET6, socket.SOCK_DGRAM) # INET6
# can do IPv4 if you are nice to it
try: # we will try to fixup our receive buffer size if we are smaller
# than allowed.
maxmf = open("/proc/sys/net/core/rmem_max")
rmemmax = int(maxmf.read())
rmemmax = rmemmax / 2
curmax = cls.socket.getsockopt(socket.SOL_SOCKET, socket.SO_RCVBUF)
curmax = curmax / 2
if (rmemmax > curmax):
cls.socket.setsockopt(socket.SOL_SOCKET,
socket.SO_RCVBUF,
rmemmax)
except Exception:
# FIXME: be more selective in catching exceptions
pass
curmax = cls.socket.getsockopt(socket.SOL_SOCKET, socket.SO_RCVBUF)
cls.poller.register(cls.socket, select.POLLIN)
cls.readersockets = [cls.socket]
cls.ipmipoller.register(cls.socket, select.POLLIN)
curmax = curmax / 2
# we throttle such that we never have no more outstanding packets than
# our receive buffer should be able to handle
cls.pending = 0
cls.maxpending = curmax / 1000
# pessimistically assume 1 kilobyte messages,
# which is way larger than almost all ipmi datagrams.
# For faster performance, sysadmins may want to examine and tune
# /proc/sys/net/core/rmem_max up. This allows the module to request
# more, but does not increase buffers for applications that do less
# creative things
# TODO(jbjohnso): perhaps spread sessions across a socket pool when
# rmem_max is small, still get ~65/socket, but avoid long queues that
# might happen with low rmem_max and putting thousands of nodes in line
def _sync_login(self, response):
"""Handle synchronous callers in liue of
a client-provided callback.
"""
if 'error' in response:
raise Exception(response['error'])
def __init__(self,
bmc,
userid,
password,
port=623,
kg=None,
onlogon=None):
self.lastpayload = None
self.bmc = bmc
self.userid = userid
self.password = password
self.nowait = False
self.pendingpayloads = collections.deque([])
if kg is not None:
self.kg = kg
else:
self.kg = password
self.port = port
if (onlogon is None):
self.async = False
self.onlogon = self._sync_login
else:
self.async = True
self.onlogon = onlogon
if not hasattr(Session, 'socket'):
self._createsocket()
self.login()
if not self.async:
while not self.logged:
Session.wait_for_rsp()
def _initsession(self):
# NOTE(jbjohnso): this number can be whatever we want.
# I picked 'xCAT' minus 1 so that a hexdump of packet
# would show xCAT
self.localsid = 2017673555
# NOTE(jbjohnso): for the moment, assume admin access
# TODO(jbjohnso): make flexible
self.privlevel = 4
self.confalgo = 0
self.aeskey = None
self.integrityalgo = 0
self.k1 = None
self.rmcptag = 1
self.ipmicallback = None
self.ipmicallbackargs = None
self.sessioncontext = None
self.sequencenumber = 0
self.sessionid = 0
self.authtype = 0
self.ipmiversion = 1.5
self.timeout = initialtimeout + (0.5 * random.random())
self.seqlun = 0
# NOTE(jbjohnso): per IPMI table 5-4, software ids in the ipmi spec may
# be 0x81 through 0x8d. We'll stick with 0x81 for now,
# do not forsee a reason to adjust
self.rqaddr = 0x81
self.logged = 0
# NOTE(jbjohnso): when we confirm a working sockaddr, put it here to
# skip getaddrinfo
self.sockaddr = None
# NOTE(jbjohnso): this tracks netfn,command,seqlun combinations that
# were retried so that we don't loop around and reuse
# the same request data and cause potential ambiguity
# in return
self.tabooseq = {}
# NOTE(jbjohnso): default to supporting ipmi 2.0. Strictly by spec,
# this should gracefully be backwards compat, but some
# 1.5 implementations checked reserved bits
self.ipmi15only = 0
self.sol_handler = None
# NOTE(jbjohnso): This is the callback handler for any SOL payload
def _checksum(self, *data): # Two's complement over the data
csum = sum(data)
csum = csum ^ 0xff
csum += 1
csum &= 0xff
return csum
def _make_ipmi_payload(self, netfn, command, data=()):
"""This function generates the core ipmi payload that would be
applicable for any channel (including KCS)
"""
self.expectedcmd = command
self.expectednetfn = netfn + \
1 # in ipmi, the response netfn is always one
# higher than the request payload, we assume
# we are always the requestor for now
seqincrement = 7 # IPMI spec forbids gaps bigger then 7 in seq number.
# Risk the taboo rather than violate the rules
while ((netfn, command, self.seqlun) in self.tabooseq and
self.tabooseq[(netfn, command, self.seqlun)] and seqincrement):
self.tabooseq[(self.expectednetfn, command, self.seqlun)] -= 1
# Allow taboo to eventually expire after a few rounds
self.seqlun += 4 # the last two bits are lun, so add 4 to add 1
self.seqlun &= 0xff # we only have one byte, wrap when exceeded
seqincrement -= 1
header = [0x20, netfn << 2]
#figure 13-4, first two bytes are rsaddr and
# netfn, rsaddr is always 0x20 since we are
# addressing BMC
reqbody = [self.rqaddr, self.seqlun, command] + list(data)
headsum = self._checksum(*header)
bodysum = self._checksum(*reqbody)
payload = header + [headsum] + reqbody + [bodysum]
return payload
def _generic_callback(self, response):
errorstr = get_ipmi_error(response)
if errorstr:
response['error'] = errorstr
self.lastresponse = response
def raw_command(self,
netfn,
command,
data=[],
retry=True,
callback=None,
callback_args=None):
self.ipmicallbackargs = callback_args
if callback is None:
self.lastresponse = None
self.ipmicallback = self._generic_callback
else:
self.ipmicallback = callback
self._send_ipmi_net_payload(netfn, command, data, retry=retry)
if retry: # in retry case, let the retry timers indicate wait time
timeout = None
else: # if not retry, give it a second before surrending
timeout = 1
#In the synchronous case, wrap the event loop in this call
#The event loop is shared amongst python-ipmi session instances
#within a process. In this way, synchronous usage of the interface
#plays well with asynchronous use. In fact, this produces the behavior
#of only the constructor *really* needing a callback. From then on,
#synchronous usage of the class acts in a greenthread style governed by
#order of data on the network
if callback is None:
while self.lastresponse is None:
Session.wait_for_rsp(timeout=timeout)
return self.lastresponse
def _send_ipmi_net_payload(self, netfn, command, data, retry=True):
ipmipayload = self._make_ipmi_payload(netfn, command, data)
payload_type = constants.payload_types['ipmi']
self.send_payload(payload=ipmipayload, payload_type=payload_type,
retry=retry)
def send_payload(self, payload=None, payload_type=None, retry=True):
if self.lastpayload is not None:
#we already have a packet outgoing, make this
# a pending payload
# this way a simplistic BMC won't get confused
# and we also avoid having to do more complicated
# retry mechanism where each payload is
# retried separately
self.pendingpayloads.append((payload, payload_type, retry))
return
if payload_type is None:
payload_type = self.last_payload_type
if payload is None:
payload = self.lastpayload
message = [0x6, 0, 0xff, 0x07] # constant RMCP header for IPMI
if retry:
self.lastpayload = payload
self.last_payload_type = payload_type
message.append(self.authtype)
baretype = payload_type
if self.integrityalgo:
payload_type |= 0b01000000
if self.confalgo:
payload_type |= 0b10000000
if (self.ipmiversion == 2.0):
message.append(payload_type)
if (baretype == 2):
raise Exception("TODO(jbjohnso): OEM Payloads")
elif baretype not in constants.payload_types.values():
raise Exception("Unrecognized payload type %d" % baretype)
message += struct.unpack("!4B", struct.pack("<I", self.sessionid))
message += struct.unpack("!4B", struct.pack("<I", self.sequencenumber))
if (self.ipmiversion == 1.5):
message += struct.unpack("!4B", struct.pack("<I", self.sessionid))
if not self.authtype == 0:
message += self._ipmi15authcode(payload)
message.append(len(payload))
message += payload
totlen = 34 + \
len(message) # Guessing the ipmi spec means the whole
# packet and assume no tag in old 1.5 world
if (totlen in (56, 84, 112, 128, 156)):
message.append(0) # Legacy pad as mandated by ipmi spec
elif self.ipmiversion == 2.0:
psize = len(payload)
if self.confalgo:
pad = (
psize + 1) % 16 # pad has to cope with one byte field like
# the _aespad function
if pad: # if no pad needed, then we take no more action
pad = 16 - pad
newpsize = psize + pad + \
17 # new payload size grew according to pad
# size, plus pad length, plus 16 byte IV
#(Table 13-20)
message.append(newpsize & 0xff)
message.append(newpsize >> 8)
iv = os.urandom(16)
message += list(struct.unpack("16B", iv))
payloadtocrypt = _aespad(payload)
crypter = AES.new(self.aeskey, AES.MODE_CBC, iv)
crypted = crypter.encrypt(struct.pack("%dB" %
len(payloadtocrypt),
*payloadtocrypt))
crypted = list(struct.unpack("%dB" % len(crypted), crypted))
message += crypted
else: # no confidetiality algorithm
message.append(psize & 0xff)
message.append(psize >> 8)
message += list(payload)
if self.integrityalgo: # see table 13-8,
# RMCP+ packet format
# TODO(jbjohnso): SHA256 which is now
# allowed
neededpad = (len(message) - 2) % 4
if neededpad:
neededpad = 4 - neededpad
message += [0xff] * neededpad
message.append(neededpad)
message.append(7) # reserved, 7 is the required value for the
# specification followed
integdata = message[4:]
authcode = HMAC.new(self.k1,
struct.pack("%dB" % len(integdata),
*integdata),
SHA).digest()[:12] # SHA1-96
# per RFC2404 truncates to 96 bits
message += struct.unpack("12B", authcode)
self.netpacket = struct.pack("!%dB" % len(message), *message)
#advance idle timer since we don't need keepalive while sending packets
#out naturally
if self in Session.keepalive_sessions:
Session.keepalive_sessions[self]['timeout'] = _monotonic_time() + \
25 + (random.random() * 4.9)
self._xmit_packet(retry)
def _ipmi15authcode(self, payload, checkremotecode=False):
if self.authtype == 0: # Only for things before auth in ipmi 1.5, not
# like 2.0 cipher suite 0
return ()
password = self.password
padneeded = 16 - len(password)
if padneeded < 0:
raise Exception("Password is too long for ipmi 1.5")
password += '\x00' * padneeded
passdata = struct.unpack("16B", password)
if checkremotecode:
seqbytes = struct.unpack("!4B",
struct.pack("<I", self.remsequencenumber))
else:
seqbytes = struct.unpack("!4B",
struct.pack("<I", self.sequencenumber))
sessdata = struct.unpack("!4B", struct.pack("<I", self.sessionid))
bodydata = passdata + sessdata + tuple(payload) + seqbytes + passdata
dgst = hashlib.md5(
struct.pack("%dB" % len(bodydata), *bodydata)).digest()
hashdata = struct.unpack("!%dB" % len(dgst), dgst)
return hashdata
def _got_channel_auth_cap(self, response):
if 'error' in response:
self.onlogon(response)
return
if response['code'] == 0xcc and self.ipmi15only is not None:
# tried ipmi 2.0 against a 1.5 which should work, but some bmcs
# thought 'reserved' meant 'must be zero'
self.ipmi15only = 1
return self._get_channel_auth_cap()
mysuffix = " while trying to get channel authentication capabalities"
errstr = get_ipmi_error(response, suffix=mysuffix)
if errstr:
self.onlogon({'error': errstr})
return
data = response['data']
self.currentchannel = data[0]
if data[1] & 0b10000000 and data[3] & 0b10: # ipmi 2.0 support
self.ipmiversion = 2.0
if self.ipmiversion == 1.5:
if not (data[1] & 0b100):
self.onlogon(
{'error':
"MD5 required but not enabled/available on target BMC"})
return
self._get_session_challenge()
elif self.ipmiversion == 2.0:
self._open_rmcpplus_request()
def _got_session_challenge(self, response):
errstr = get_ipmi_error(response,
suffix=" while getting session challenge")
if errstr:
self.onlogon({'error': errstr})
return
data = response['data']
self.sessionid = struct.unpack("<I", struct.pack("4B", *data[0:4]))[0]
self.authtype = 2
self._activate_session(data[4:])
'''
This sends the activate session payload. We pick '1' as the requested
sequence number without perturbing our real sequence number
'''
def _activate_session(self, data):
rqdata = [2, 4] + list(data) + [1, 0, 0, 0]
# TODO(jbjohnso): this always requests admin level (1.5)
self.ipmicallback = self._activated_session
self._send_ipmi_net_payload(netfn=0x6, command=0x3a, data=rqdata)
def _activated_session(self, response):
errstr = get_ipmi_error(response)
if errstr:
self.onlogon({'error': errstr})
return
data = response['data']
self.sessionid = struct.unpack("<I", struct.pack("4B", *data[1:5]))[0]
self.sequencenumber = struct.unpack("<I",
struct.pack("4B", *data[5:9]))[0]
self._req_priv_level()
def _req_priv_level(self):
self.ipmicallback = self._got_priv_level
self._send_ipmi_net_payload(netfn=0x6,
command=0x3b,
data=[self.privlevel])
def _got_priv_level(self, response):
mysuffix = " while requesting privelege level %d for %s" % (
self.privlevel, self.userid)
errstr = get_ipmi_error(response, suffix=mysuffix)
if errstr:
self.onlogon({'error': errstr})
return
self.logged = 1
Session.keepalive_sessions[self] = {}
Session.keepalive_sessions[self]['ipmisession'] = self
Session.keepalive_sessions[self]['timeout'] = _monotonic_time() + \
25 + (random.random() * 4.9)
self.onlogon({'success': True})
def _get_session_challenge(self):
reqdata = [2]
if len(self.userid) > 16:
raise Exception("Username too long for IPMI, must not exceed 16")
padneeded = 16 - len(self.userid)
userid = self.userid + ('\x00' * padneeded)
reqdata += struct.unpack("!16B", userid)
self.ipmicallback = self._got_session_challenge
self._send_ipmi_net_payload(netfn=0x6, command=0x39, data=reqdata)
def _open_rmcpplus_request(self):
self.authtype = 6
self.localsid += 1 # have unique local session ids to ignore aborted
# login attempts from the past
self.rmcptag += 1
data = [
self.rmcptag,
0, # request as much privilege as the channel will give us
0, 0, # reserved
]
data += list(struct.unpack("4B", struct.pack("<I", self.localsid)))
data += [
0, 0, 0, 8, 1, 0, 0, 0, # table 13-17, SHA-1
1, 0, 0, 8, 1, 0, 0, 0, # SHA-1 integrity
2, 0, 0, 8, 1, 0, 0, 0, # AES privacy
#2,0,0,8,0,0,0,0, #no privacy confalgo
]
self.sessioncontext = 'OPENSESSION'
self.send_payload(
payload=data,
payload_type=constants.payload_types['rmcpplusopenreq'])
def _get_channel_auth_cap(self):
self.ipmicallback = self._got_channel_auth_cap
if (self.ipmi15only):
self._send_ipmi_net_payload(netfn=0x6,
command=0x38,
data=[0x0e, self.privlevel])
else:
self._send_ipmi_net_payload(netfn=0x6,
command=0x38,
data=[0x8e, self.privlevel])
def login(self):
self.logontries = 5
self._initsession()
self._get_channel_auth_cap()
@classmethod
def wait_for_rsp(cls, timeout=None):
"""IPMI Session Event loop iteration
This watches for any activity on IPMI handles and handles registered
by register_handle_callback. Callers are satisfied in the order that
packets return from nework, not in the order of calling.
:param timeout: Maximum time to wait for data to come across. If
unspecified, will autodetect based on earliest timeout
"""
#Assume:
#Instance A sends request to packet B
#Then Instance C sends request to BMC D
#BMC D was faster, so data comes back before BMC B
#Instance C gets to go ahead of Instance A, because
#Instance C can get work done, but instance A cannot
curtime = _monotonic_time()
# There ar a number of parties that each has their own timeout
# The caller can specify a deadline in timeout argument
# each session with active outbound payload has callback to
# handle retry/timout error
# each session that is 'alive' wants to send a keepalive ever so often.
# We want to make sure the most strict request is honored and block for
# no more time than that, so that whatever part(ies) need to service in
# a deadline, will be honored
if timeout != 0:
for session, parms in cls.waiting_sessions.iteritems():
if parms['timeout'] <= curtime:
timeout = 0 # exit after one guaranteed pass
break
if (timeout is not None and
timeout < parms['timeout'] - curtime):
continue # timeout smaller than the current session needs
timeout = parms['timeout'] - curtime # set new timeout value
for session, parms in cls.keepalive_sessions.iteritems():
if parms['timeout'] <= curtime:
timeout = 0
break
if (timeout is not None and
timeout < parms['timeout'] - curtime):
continue
timeout = parms['timeout'] - curtime
# If the loop above found no sessions wanting *and* the caller had no
# timeout, exit function. In this case there is no way a session
# could be waiting so we can always return 0
if timeout is None:
return 0
rdylist, _, _ = select.select(cls.readersockets, (), (), timeout)
if len(rdylist) > 0:
while cls.ipmipoller.poll(0): # if the somewhat lengthy queue
# processing takes long enough for packets to come in,
# be eager
pktqueue = collections.deque([])
while cls.ipmipoller.poll(0): # looks rendundant, but want to
#queue and process packets to keep things off
#RCVBUF
rdata = cls.socket.recvfrom(3000)
pktqueue.append(rdata)
while len(pktqueue):
(data, sockaddr) = pktqueue.popleft()
cls._route_ipmiresponse(sockaddr, data)
while cls.ipmipoller.poll(0): # seems ridiculous, but
# between every callback, check for packets again
rdata = cls.socket.recvfrom(3000)
pktqueue.append(rdata)
for handlepair in cls.poller.poll(0):
myhandle = handlepair[0]
if myhandle != cls.socket.fileno():
myfile = cls._external_handlers[myhandle][1]
cls._external_handlers[myhandle][0](myfile)
sessionstodel = []
for session, parms in cls.keepalive_sessions.iteritems():
if parms['timeout'] < curtime:
cls.keepalive_sessions[session]['timeout'] = 25 + \
(random.random() * 4.9)
session._keepalive()
for session, parms in cls.waiting_sessions.iteritems():
if parms['timeout'] < curtime: # timeout has expired, time to
# give up on it and trigger timeout
# response in the respective
# session
sessionstodel.append(
session) # defer deletion until after loop
# to avoid confusing the for loop
for session in sessionstodel:
cls.pending -= 1
session.lastpayload = None
cls.waiting_sessions.pop(session, None)
session._timedout()
return len(cls.waiting_sessions)
def _keepalive(self):
"""Performs a keepalive to avoid idle disconnect
"""
self.raw_command(netfn=6, command=1)
@classmethod
def register_handle_callback(cls, handle, callback):
"""Add a handle to be watched by Session's event loop
In the event that an application would like IPMI Session event loop
to drive things while adding their own filehandle to watch for events,
this class method will register that.
:param handle: filehandle too watch for input
:param callback: function to call when input detected on the handle.
will receive the handle as an argument
"""
cls._external_handlers[handle.fileno()] = (callback, handle)
cls.readersockets += [handle]
cls.poller.register(handle, select.POLLIN)
@classmethod
def _route_ipmiresponse(cls, sockaddr, data):
if not (data[0] == '\x06' and data[2:4] == '\xff\x07'): # not ipmi
return
try:
cls.bmc_handlers[sockaddr]._handle_ipmi_packet(data,
sockaddr=sockaddr)
cls.pending -= 1
except KeyError:
pass
def _handle_ipmi_packet(self, data, sockaddr=None):
if self.sockaddr is None and sockaddr is not None:
self.sockaddr = sockaddr
elif (self.sockaddr is not None and
sockaddr is not None and
self.sockaddr != sockaddr):
return # here, we might have sent an ipv4 and ipv6 packet to kick
# things off ignore the second reply since we have one
# satisfactory answer
if data[4] in ('\x00', '\x02'): # This is an ipmi 1.5 paylod
remsequencenumber = struct.unpack('<I', data[5:9])[0]
if (hasattr(self, 'remsequencenumber') and
remsequencenumber < self.remsequencenumber):
return -5 # remote sequence number is too low, reject it
self.remsequencenumber = remsequencenumber
if ord(data[4]) != self.authtype:
return -2 # BMC responded with mismatch authtype, for
# mutual authentication reject it. If this causes
# legitimate issues, it's the vendor's fault
remsessid = struct.unpack("<I", data[9:13])[0]
if remsessid != self.sessionid:
return -1 # does not match our session id, drop it
# now we need a mutable representation of the packet, rather than
# copying pieces of the packet over and over
rsp = list(struct.unpack("!%dB" % len(data), data))
authcode = False
if data[4] == '\x02': # we have authcode in this ipmi 1.5 packet
authcode = data[13:29]
del rsp[13:29]
# this is why we needed a mutable representation
payload = list(rsp[14:14 + rsp[13]])
if authcode:
expectedauthcode = self._ipmi15authcode(payload,
checkremotecode=True)
expectedauthcode = struct.pack("%dB" % len(expectedauthcode),
*expectedauthcode)
if expectedauthcode != authcode:
return
self._parse_ipmi_payload(payload)
elif data[4] == '\x06':
self._handle_ipmi2_packet(data)
else:
return # unrecognized data, assume evil
def _handle_ipmi2_packet(self, rawdata):
data = list(struct.unpack("%dB" % len(rawdata), rawdata))
#now need mutable array
ptype = data[5] & 0b00111111
# the first 16 bytes are header information as can be seen in 13-8 that
# we will toss out
if ptype == 0x11: # rmcp+ response
return self._got_rmcp_response(data[16:])
elif ptype == 0x13:
return self._got_rakp2(data[16:])
elif ptype == 0x15:
return self._got_rakp4(data[16:])
elif ptype == 0 or ptype == 1: # good old ipmi payload or sol
# If endorsing a shared secret scheme, then at the very least it
# needs to do mutual assurance
if not (data[5] & 0b01000000): # This would be the line that might
# trip up some insecure BMC
# implementation
return
encrypted = 0
if data[5] & 0b10000000:
encrypted = 1
authcode = rawdata[-12:]
expectedauthcode = HMAC.new(
self.k1, rawdata[4:-12], SHA).digest()[:12]
if authcode != expectedauthcode:
return # BMC failed to assure integrity to us, drop it
sid = struct.unpack("<I", rawdata[6:10])[0]
if sid != self.localsid: # session id mismatch, drop it
return
remseqnumber = struct.unpack("<I", rawdata[10:14])[0]
if (hasattr(self, 'remseqnumber') and
(remseqnumber < self.remseqnumber) and
(self.remseqnumber != 0xffffffff)):
return
self.remseqnumber = remseqnumber
psize = data[14] + (data[15] << 8)
payload = data[16:16 + psize]
if encrypted:
iv = rawdata[16:32]
decrypter = AES.new(self.aeskey, AES.MODE_CBC, iv)
decrypted = decrypter.decrypt(
struct.pack("%dB" % len(payload[16:]),
*payload[16:]))
payload = struct.unpack("%dB" % len(decrypted), decrypted)
padsize = payload[-1] + 1
payload = list(payload[:-padsize])
if ptype == 0:
self._parse_ipmi_payload(payload)
elif ptype == 1: # There should be no other option
# note that we assume the SOL payload is good enough to avoid
# retry SOL logic is sufficiently different, we just
# defer that call to the sol handler, it can re submit if it
# is unhappy
if self.last_payload_type == 1: # but only if SOL was last tx
self.lastpayload = None
self.last_payload_type = None
Session.waiting_sessions.pop(self, None)
if self.sol_handler:
self.sol_handler(payload)
def _got_rmcp_response(self, data):
# see RMCP+ open session response table
if not (self.sessioncontext and self.sessioncontext != "Established"):
return -9
# ignore payload as we are not in a state valid it
if data[0] != self.rmcptag:
return -9 # use rmcp tag to track and reject stale responses
if data[1] != 0: # response code...
if data[1] in constants.rmcp_codes:
errstr = constants.rmcp_codes[data[1]]
else:
errstr = "Unrecognized RMCP code %d" % data[1]
self.onlogon({'error': errstr})
return -9
self.allowedpriv = data[2]
# TODO(jbjohnso): enable lower priv access (e.g. operator/user)
localsid = struct.unpack("<I", struct.pack("4B", *data[4:8]))[0]
if self.localsid != localsid:
return -9
self.pendingsessionid = struct.unpack(
"<I", struct.pack("4B", *data[8:12]))[0]
# TODO(jbjohnso): currently, we take it for granted that the responder
# accepted our integrity/auth/confidentiality proposal
self.lastpayload = None
self._send_rakp1()
def _send_rakp1(self):
self.rmcptag += 1
self.randombytes = os.urandom(16)
userlen = len(self.userid)
payload = [self.rmcptag, 0, 0, 0] + \
list(struct.unpack("4B",
struct.pack("<I", self.pendingsessionid))) +\
list(struct.unpack("16B", self.randombytes)) +\
[self.privlevel, 0, 0] +\
[userlen] +\
list(struct.unpack("%dB" % userlen, self.userid))
self.sessioncontext = "EXPECTINGRAKP2"
self.send_payload(
payload=payload, payload_type=constants.payload_types['rakp1'])
def _got_rakp2(self, data):
if not (self.sessioncontext in ('EXPECTINGRAKP2', 'EXPECTINGRAKP4')):
return -9 # if we are not expecting rakp2, ignore. In a retry
# scenario, replying from stale RAKP2 after sending
# RAKP3 seems to be best
if data[0] != self.rmcptag: # ignore mismatched tags for retry logic
return -9
if data[1] != 0: # if not successful, consider next move
if data[1] == 2: # invalid sessionid 99% of the time means a retry
# scenario invalidated an in-flight transaction
return
if data[1] in constants.rmcp_codes:
errstr = constants.rmcp_codes[data[1]]
else:
errstr = "Unrecognized RMCP code %d" % data[1]
self.onlogon({'error': errstr + " in RAKP2"})
return -9
localsid = struct.unpack("<I", struct.pack("4B", *data[4:8]))[0]
if localsid != self.localsid:
return -9 # discard mismatch in the session identifier
self.remoterandombytes = struct.pack("16B", *data[8:24])
self.remoteguid = struct.pack("16B", *data[24:40])
userlen = len(self.userid)
hmacdata = struct.pack("<II", localsid, self.pendingsessionid) +\
self.randombytes + self.remoterandombytes + self.remoteguid +\
struct.pack("2B", self.privlevel, userlen) +\
self.userid
expectedhash = HMAC.new(self.password, hmacdata, SHA).digest()
givenhash = struct.pack("%dB" % len(data[40:]), *data[40:])
if givenhash != expectedhash:
self.sessioncontext = "FAILED"
self.onlogon({'error': "Incorrect password provided"})
return -9
# We have now validated that the BMC and client agree on password, time
# to store the keys
self.sik = HMAC.new(self.kg,
self.randombytes + self.remoterandombytes +
struct.pack("2B", self.privlevel, userlen) +
self.userid, SHA).digest()
self.k1 = HMAC.new(self.sik, '\x01' * 20, SHA).digest()
self.k2 = HMAC.new(self.sik, '\x02' * 20, SHA).digest()
self.aeskey = self.k2[0:16]
self.sessioncontext = "EXPECTINGRAKP4"
self.lastpayload = None
self._send_rakp3()
def _send_rakp3(self): # rakp message 3
self.rmcptag += 1
# rmcptag, then status 0, then two reserved 0s
payload = [self.rmcptag, 0, 0, 0] +\
list(struct.unpack("4B", struct.pack("<I", self.pendingsessionid)))
hmacdata = self.remoterandombytes +\
struct.pack("<I", self.localsid) +\
struct.pack("2B", self.privlevel, len(self.userid)) +\
self.userid
authcode = HMAC.new(self.password, hmacdata, SHA).digest()
payload += list(struct.unpack("%dB" % len(authcode), authcode))
self.send_payload(
payload=payload, payload_type=constants.payload_types['rakp3'])
def _relog(self):
self._initsession()
self.logontries -= 1
return self._get_channel_auth_cap()
def _got_rakp4(self, data):
if self.sessioncontext != "EXPECTINGRAKP4" or data[0] != self.rmcptag:
return -9
if data[1] != 0:
if data[1] == 2 and self.logontries: # if we retried RAKP3 because
# RAKP4 got dropped, BMC can consider it done and we must
# restart
self._relog()
if data[1] == 15 and self.logontries: # ignore 15 value if we are
# retrying. xCAT did but I can't recall why exactly
# TODO(jbjohnso) jog my memory to update the comment
return
if data[1] in constants.rmcp_codes:
errstr = constants.rmcp_codes[data[1]]
else:
errstr = "Unrecognized RMCP code %d" % data[1]
self.onlogon({'error': errstr + " reported in RAKP4"})
return -9
localsid = struct.unpack("<I", struct.pack("4B", *data[4:8]))[0]
if localsid != self.localsid: # ignore if wrong session id indicated
return -9
hmacdata = self.randombytes +\
struct.pack("<I", self.pendingsessionid) +\
self.remoteguid
expectedauthcode = HMAC.new(self.sik, hmacdata, SHA).digest()[:12]
authcode = struct.pack("%dB" % len(data[8:]), *data[8:])
if authcode != expectedauthcode:
self.onlogon({'error': "Invalid RAKP4 integrity code (wrong Kg?)"})
return
self.sessionid = self.pendingsessionid
self.integrityalgo = 'sha1'
self.confalgo = 'aes'
self.sequencenumber = 1
self.sessioncontext = 'ESTABLISHED'
self.lastpayload = None
self._req_priv_level()
'''
Internal function to parse IPMI nugget once extracted from its framing
'''
def _parse_ipmi_payload(self, payload):
# For now, skip the checksums since we are in LAN only,
# TODO(jbjohnso): if implementing other channels, add checksum checks
# here
if (payload[4] != self.seqlun or
payload[1] >> 2 != self.expectednetfn or
payload[5] != self.expectedcmd):
return -1 # payload is not a match for our last packet
if hasattr(self, 'hasretried') and self.hasretried:
self.hasretried = 0
self.tabooseq[
(self.expectednetfn, self.expectedcmd, self.seqlun)] = 16
# try to skip it for at most 16 cycles of overflow
# We want to now remember that we do not have an expected packet
self.expectednetfn = 0x1ff # bigger than one byte means it can never
# match the one byte value by mistake
self.expectedcmd = 0x1ff
self.seqlun += 4 # prepare seqlun for next transmit
self.seqlun &= 0xff # when overflowing, wrap around
Session.waiting_sessions.pop(self, None)
self.lastpayload = None # render retry mechanism utterly incapable of
# doing anything, though it shouldn't matter
self.last_payload_type = None
response = {}
response['netfn'] = payload[1] >> 2
del payload[0:5]
# ^^ remove header of rsaddr/netfn/lun/checksum/rq/seq/lun
del payload[-1] # remove the trailing checksum
response['command'] = payload[0]
response['code'] = payload[1]
del payload[0:2]
response['data'] = payload
self.timeout = initialtimeout + (0.5 * random.random())
if len(self.pendingpayloads) > 0:
(nextpayload, nextpayloadtype, retry) = \
self.pendingpayloads.popleft()
self.send_payload(payload=nextpayload,
payload_type=nextpayloadtype,
retry=retry)
call_with_optional_args(self.ipmicallback,
response,
self.ipmicallbackargs)
def _timedout(self):
if not self.lastpayload:
return
self.nowait = True
self.timeout += 1
if self.timeout > 5:
response = {'error': 'timeout'}
call_with_optional_args(self.ipmicallback,
response,
self.ipmicallbackargs)
self.nowait = False
return
elif self.sessioncontext == 'FAILED':
self.nowait = False
return
if self.sessioncontext == 'OPENSESSION':
# In this case, we want to craft a new session request to have
# unambiguous session id regardless of how packet was dropped or
# delayed in this case, it's safe to just redo the request
self._open_rmcpplus_request()
elif (self.sessioncontext == 'EXPECTINGRAKP2' or
self.sessioncontext == 'EXPECTINGRAKP4'):
# If we can't be sure which RAKP was dropped or if RAKP3/4 was just
# delayed, the most reliable thing to do is rewind and start over
# bmcs do not take kindly to receiving RAKP1 or RAKP3 twice
self._relog()
else: # in IPMI case, the only recourse is to act as if the packet is
# idempotent. SOL has more sophisticated retry handling
# the biggest risks are reset sp which is often fruitless to retry
# and chassis reset, which sometimes will shoot itself
# systematically in the head in a shared port case making replies
# impossible
self.hasretried = 1 # remember so that we can track taboo
# combinations
# of sequence number, netfn, and lun due to
# ambiguity on the wire
self.send_payload()
self.nowait = False
def _xmit_packet(self, retry=True):
if not self.nowait: # if we are retrying, we really need to get the
# packet out and get our timeout updated
Session.wait_for_rsp(timeout=0) # take a convenient opportunity
# to drain the socket queue if
# applicable
while Session.pending > Session.maxpending:
Session.wait_for_rsp()
if retry:
Session.waiting_sessions[self] = {}
Session.waiting_sessions[self]['ipmisession'] = self
Session.waiting_sessions[self]['timeout'] = self.timeout + \
_monotonic_time()
Session.pending += 1
if self.sockaddr:
Session.socket.sendto(self.netpacket, self.sockaddr)
else: # he have not yet picked a working sockaddr for this connection,
# try all the candidates that getaddrinfo provides
for res in socket.getaddrinfo(self.bmc,
self.port,
0,
socket.SOCK_DGRAM):
sockaddr = res[4]
if (res[0] == socket.AF_INET): # convert the sockaddr AF_INET6
newhost = '::ffff:' + sockaddr[0]
sockaddr = (newhost, sockaddr[1], 0, 0)
Session.bmc_handlers[sockaddr] = self
Session.socket.sendto(self.netpacket, sockaddr)
if self.sequencenumber: # seq number of zero will be left alone, it is
# special, otherwise increment
self.sequencenumber += 1
def logout(self, callback=None, callback_args=None):
if not self.logged:
if callback is None:
return {'success': True}
callback({'success': True})
return
self.raw_command(command=0x3c,
netfn=6,
data=struct.unpack("4B",
struct.pack("I", self.sessionid)),
retry=False,
callback=callback,
callback_args=callback_args)
self.logged = 0
if callback is None:
return {'success': True}
callback({'success': True})
if __name__ == "__main__":
import sys
ipmis = Session(bmc=sys.argv[1],
userid=sys.argv[2],
password=os.environ['IPMIPASS'])
print ipmis.raw_command(command=2, data=[1], netfn=0)
print get_ipmi_error({'command': 8, 'code': 128, 'netfn': 1})