# Copyright (c) 2017-2019 The University of Manchester
#
# This program is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with this program. If not, see <http://www.gnu.org/licenses/>.
import struct
import logging
from six import iterkeys
from spinn_utilities.progress_bar import ProgressBar
from spinnman.messages.sdp import SDPFlag, SDPHeader, SDPMessage
from spinnman.model.enums import CPUState
from spinn_front_end_common.utilities.constants import (
SDP_PORTS, SDP_RUNNING_MESSAGE_CODES)
from spinn_front_end_common.utilities.exceptions import ConfigurationException
logger = logging.getLogger(__name__)
_ONE_WORD = struct.Struct("<I")
_LIMIT = 10
[docs]class ChipProvenanceUpdater(object):
""" Forces all cores to generate provenance data, and then exit
"""
__slots__ = []
def __call__(self, txrx, app_id, all_core_subsets):
# check that the right number of processors are in sync
processors_completed = txrx.get_core_state_count(
app_id, CPUState.FINISHED)
total_processors = len(all_core_subsets)
left_to_do_cores = total_processors - processors_completed
progress = ProgressBar(
left_to_do_cores,
"Forcing error cores to generate provenance data")
error_cores = txrx.get_cores_in_state(
all_core_subsets, CPUState.RUN_TIME_EXCEPTION)
watchdog_cores = txrx.get_cores_in_state(
all_core_subsets, CPUState.WATCHDOG)
idle_cores = txrx.get_cores_in_state(
all_core_subsets, CPUState.IDLE)
if error_cores or watchdog_cores or idle_cores:
raise ConfigurationException(
"Some cores have crashed. RTE cores {}, watch-dogged cores {},"
" idle cores {}".format(
error_cores.values(), watchdog_cores.values(),
idle_cores.values()))
# check that all cores are in the state FINISHED which shows that
# the core has received the message and done provenance updating
self._update_provenance(txrx, total_processors, processors_completed,
all_core_subsets, app_id, progress)
progress.end()
def _update_provenance(self, txrx, total_processors, processors_completed,
all_core_subsets, app_id, progress):
# pylint: disable=too-many-arguments
left_to_do_cores = total_processors - processors_completed
attempts = 0
while processors_completed != total_processors and attempts < _LIMIT:
attempts += 1
unsuccessful_cores = txrx.get_cores_not_in_state(
all_core_subsets, CPUState.FINISHED)
for (x, y, p) in iterkeys(unsuccessful_cores):
self._send_chip_update_provenance_and_exit(txrx, x, y, p)
processors_completed = txrx.get_core_state_count(
app_id, CPUState.FINISHED)
left_over_now = total_processors - processors_completed
to_update = left_to_do_cores - left_over_now
left_to_do_cores = left_over_now
if to_update != 0:
progress.update(to_update)
if processors_completed != total_processors:
logger.error("Unable to Finish getting provenance data. "
"Abandoned after too many retries. "
"Board may be left in an unstable state!")
@staticmethod
def _send_chip_update_provenance_and_exit(txrx, x, y, p):
cmd = SDP_RUNNING_MESSAGE_CODES.SDP_UPDATE_PROVENCE_REGION_AND_EXIT
port = SDP_PORTS.RUNNING_COMMAND_SDP_PORT
txrx.send_sdp_message(SDPMessage(
SDPHeader(
flags=SDPFlag.REPLY_NOT_EXPECTED,
destination_port=port.value, destination_cpu=p,
destination_chip_x=x, destination_chip_y=y),
data=_ONE_WORD.pack(cmd.value)))