Files
gem5/src/python/m5/simulate.py
Andreas Sandberg 6f38428abb stats: Add support for hierarchical stats
This change makes the stat system aware of the hierarchical nature of
stats. The aim is to achieve the following goals:

  * Make the SimObject hierarchy explicit in the stat system (i.e.,
    get rid of name() + ".foo"). This makes stat naming less fragile
    and makes it possible to implement hierarchical formats like
    XML/HDF5/JSON in a clean way.

  * Make it more convenient to split stats into a separate
    struct/class that can be bound to a SimObject. This makes the
    namespace cleaner and makes stat accesses a bit more obvious.

  * Make it possible to build groups of stats in C++ that can be used
    in subcomponents in a SimObject (similar to what we do for
    checkpoint sections). This makes it easier to structure large
    components.

  * Enable partial stat dumps. Some of our internal users have been
    asking for this since a full stat dump can be large.

  * Enable better stat access from Python.

This changeset implements solves the first three points by introducing
a class (Stats::Group) that owns statistics belonging to the same
object. SimObjects inherit from Stats::Group since they typically have
statistics.

New-style statistics need to be associated with a parent group at
instantiation time. Instantiation typically sets the name and the
description, other parameters need to be set by overriding
Group::regStats() just like with legacy stats. Simple objects with
scalar stats can typically avoid implementing regStats() altogether
since the stat name and description are both specified in the
constructor.

For convenience reasons, statistics groups can be merged into other
groups. This means that a SimObject can create a stat struct that
inherits from Stats::Group and merge it into the parent group
(SimObject). This can make the code cleaner since statistics tracking
gets grouped into a single object.

Stat visitors have a new API to expose the group structure. The
Output::beginGroup(name) method is called at the beginning of a group
and the Output::endGroup() method is called when all stats, and
sub-groups, have been visited. Flat formats (e.g., the text format)
typically need to maintain a stack to track the full path to a stat.

Legacy, flat, statistics are still supported after applying this
change. These stats don't belong to any group and stat visitors will
not see a Output::beginGroup(name) call before their corresponding
Output::visit() methods are called.

Change-Id: I9025d61dfadeabcc8ecf30813ab2060def455648
Signed-off-by: Andreas Sandberg <andreas.sandberg@arm.com>
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5/+/19368
Tested-by: kokoro <noreply+kokoro@google.com>
Reviewed-by: Jason Lowe-Power <jason@lowepower.com>
Reviewed-by: Daniel Carvalho <odanrc@yahoo.com.br>
2019-08-29 09:01:38 +00:00

377 lines
13 KiB
Python

# Copyright (c) 2012,2019 ARM Limited
# All rights reserved.
#
# The license below extends only to copyright in the software and shall
# not be construed as granting a license to any other intellectual
# property including but not limited to intellectual property relating
# to a hardware implementation of the functionality of the software
# licensed hereunder. You may use the software subject to the license
# terms below provided that you ensure that this notice is replicated
# unmodified and in its entirety in all distributions of the software,
# modified or unmodified, in source code or in binary form.
#
# Copyright (c) 2005 The Regents of The University of Michigan
# Copyright (c) 2010 Advanced Micro Devices, Inc.
# All rights reserved.
#
# Redistribution and use in source and binary forms, with or without
# modification, are permitted provided that the following conditions are
# met: redistributions of source code must retain the above copyright
# notice, this list of conditions and the following disclaimer;
# redistributions in binary form must reproduce the above copyright
# notice, this list of conditions and the following disclaimer in the
# documentation and/or other materials provided with the distribution;
# neither the name of the copyright holders nor the names of its
# contributors may be used to endorse or promote products derived from
# this software without specific prior written permission.
#
# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
# "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
# LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
# A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
# OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
# SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
# LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
# DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
# THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
# (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
# OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
#
# Authors: Nathan Binkert
# Steve Reinhardt
from __future__ import print_function
import atexit
import os
import sys
# import the wrapped C++ functions
import _m5.drain
import _m5.core
from _m5.stats import updateEvents as updateStatEvents
from . import stats
from . import SimObject
from . import ticks
from . import objects
from m5.util.dot_writer import do_dot, do_dvfs_dot
from m5.util.dot_writer_ruby import do_ruby_dot
from .util import fatal
from .util import attrdict
# define a MaxTick parameter, unsigned 64 bit
MaxTick = 2**64 - 1
_memory_modes = {
"atomic" : objects.params.atomic,
"timing" : objects.params.timing,
"atomic_noncaching" : objects.params.atomic_noncaching,
}
_drain_manager = _m5.drain.DrainManager.instance()
# The final hook to generate .ini files. Called from the user script
# once the config is built.
def instantiate(ckpt_dir=None):
from m5 import options
root = objects.Root.getInstance()
if not root:
fatal("Need to instantiate Root() before calling instantiate()")
# we need to fix the global frequency
ticks.fixGlobalFrequency()
# Make sure SimObject-valued params are in the configuration
# hierarchy so we catch them with future descendants() walks
for obj in root.descendants(): obj.adoptOrphanParams()
# Unproxy in sorted order for determinism
for obj in root.descendants(): obj.unproxyParams()
if options.dump_config:
ini_file = open(os.path.join(options.outdir, options.dump_config), 'w')
# Print ini sections in sorted order for easier diffing
for obj in sorted(root.descendants(), key=lambda o: o.path()):
obj.print_ini(ini_file)
ini_file.close()
if options.json_config:
try:
import json
json_file = open(
os.path.join(options.outdir, options.json_config), 'w')
d = root.get_config_as_dict()
json.dump(d, json_file, indent=4)
json_file.close()
except ImportError:
pass
do_dot(root, options.outdir, options.dot_config)
do_ruby_dot(root, options.outdir, options.dot_config)
# Initialize the global statistics
stats.initSimStats()
# Create the C++ sim objects and connect ports
for obj in root.descendants(): obj.createCCObject()
for obj in root.descendants(): obj.connectPorts()
# Do a second pass to finish initializing the sim objects
for obj in root.descendants(): obj.init()
# Do a third pass to initialize statistics
stats._bindStatHierarchy(root)
root.regStats()
# Do a fourth pass to initialize probe points
for obj in root.descendants(): obj.regProbePoints()
# Do a fifth pass to connect probe listeners
for obj in root.descendants(): obj.regProbeListeners()
# We want to generate the DVFS diagram for the system. This can only be
# done once all of the CPP objects have been created and initialised so
# that we are able to figure out which object belongs to which domain.
if options.dot_dvfs_config:
do_dvfs_dot(root, options.outdir, options.dot_dvfs_config)
# We're done registering statistics. Enable the stats package now.
stats.enable()
# Restore checkpoint (if any)
if ckpt_dir:
_drain_manager.preCheckpointRestore()
ckpt = _m5.core.getCheckpoint(ckpt_dir)
_m5.core.unserializeGlobals(ckpt);
for obj in root.descendants(): obj.loadState(ckpt)
else:
for obj in root.descendants(): obj.initState()
# Check to see if any of the stat events are in the past after resuming from
# a checkpoint, If so, this call will shift them to be at a valid time.
updateStatEvents()
need_startup = True
def simulate(*args, **kwargs):
global need_startup
if need_startup:
root = objects.Root.getInstance()
for obj in root.descendants(): obj.startup()
need_startup = False
# Python exit handlers happen in reverse order.
# We want to dump stats last.
atexit.register(stats.dump)
# register our C++ exit callback function with Python
atexit.register(_m5.core.doExitCleanup)
# Reset to put the stats in a consistent state.
stats.reset()
if _drain_manager.isDrained():
_drain_manager.resume()
return _m5.event.simulate(*args, **kwargs)
def drain():
"""Drain the simulator in preparation of a checkpoint or memory mode
switch.
This operation is a no-op if the simulator is already in the
Drained state.
"""
# Try to drain all objects. Draining might not be completed unless
# all objects return that they are drained on the first call. This
# is because as objects drain they may cause other objects to no
# longer be drained.
def _drain():
# Try to drain the system. The drain is successful if all
# objects are done without simulation. We need to simulate
# more if not.
if _drain_manager.tryDrain():
return True
# WARNING: if a valid exit event occurs while draining, it
# will not get returned to the user script
exit_event = _m5.event.simulate()
while exit_event.getCause() != 'Finished drain':
exit_event = simulate()
return False
# Don't try to drain a system that is already drained
is_drained = _drain_manager.isDrained()
while not is_drained:
is_drained = _drain()
assert _drain_manager.isDrained(), "Drain state inconsistent"
def memWriteback(root):
for obj in root.descendants():
obj.memWriteback()
def memInvalidate(root):
for obj in root.descendants():
obj.memInvalidate()
def checkpoint(dir):
root = objects.Root.getInstance()
if not isinstance(root, objects.Root):
raise TypeError("Checkpoint must be called on a root object.")
drain()
memWriteback(root)
print("Writing checkpoint")
_m5.core.serializeAll(dir)
def _changeMemoryMode(system, mode):
if not isinstance(system, (objects.Root, objects.System)):
raise TypeError("Parameter of type '%s'. Must be type %s or %s." % \
(type(system), objects.Root, objects.System))
if system.getMemoryMode() != mode:
system.setMemoryMode(mode)
else:
print("System already in target mode. Memory mode unchanged.")
def switchCpus(system, cpuList, verbose=True):
"""Switch CPUs in a system.
Note: This method may switch the memory mode of the system if that
is required by the CPUs. It may also flush all caches in the
system.
Arguments:
system -- Simulated system.
cpuList -- (old_cpu, new_cpu) tuples
"""
if verbose:
print("switching cpus")
if not isinstance(cpuList, list):
raise RuntimeError("Must pass a list to this function")
for item in cpuList:
if not isinstance(item, tuple) or len(item) != 2:
raise RuntimeError("List must have tuples of (oldCPU,newCPU)")
old_cpus = [old_cpu for old_cpu, new_cpu in cpuList]
new_cpus = [new_cpu for old_cpu, new_cpu in cpuList]
old_cpu_set = set(old_cpus)
memory_mode_name = new_cpus[0].memory_mode()
for old_cpu, new_cpu in cpuList:
if not isinstance(old_cpu, objects.BaseCPU):
raise TypeError("%s is not of type BaseCPU" % old_cpu)
if not isinstance(new_cpu, objects.BaseCPU):
raise TypeError("%s is not of type BaseCPU" % new_cpu)
if new_cpu in old_cpu_set:
raise RuntimeError(
"New CPU (%s) is in the list of old CPUs." % (old_cpu,))
if not new_cpu.switchedOut():
raise RuntimeError("New CPU (%s) is already active." % (new_cpu,))
if not new_cpu.support_take_over():
raise RuntimeError(
"New CPU (%s) does not support CPU handover." % (old_cpu,))
if new_cpu.memory_mode() != memory_mode_name:
raise RuntimeError(
"%s and %s require different memory modes." % (new_cpu,
new_cpus[0]))
if old_cpu.switchedOut():
raise RuntimeError("Old CPU (%s) is inactive." % (new_cpu,))
if not old_cpu.support_take_over():
raise RuntimeError(
"Old CPU (%s) does not support CPU handover." % (old_cpu,))
try:
memory_mode = _memory_modes[memory_mode_name]
except KeyError:
raise RuntimeError("Invalid memory mode (%s)" % memory_mode_name)
drain()
# Now all of the CPUs are ready to be switched out
for old_cpu, new_cpu in cpuList:
old_cpu.switchOut()
# Change the memory mode if required. We check if this is needed
# to avoid printing a warning if no switch was performed.
if system.getMemoryMode() != memory_mode:
# Flush the memory system if we are switching to a memory mode
# that disables caches. This typically happens when switching to a
# hardware virtualized CPU.
if memory_mode == objects.params.atomic_noncaching:
memWriteback(system)
memInvalidate(system)
_changeMemoryMode(system, memory_mode)
for old_cpu, new_cpu in cpuList:
new_cpu.takeOverFrom(old_cpu)
def notifyFork(root):
for obj in root.descendants():
obj.notifyFork()
fork_count = 0
def fork(simout="%(parent)s.f%(fork_seq)i"):
"""Fork the simulator.
This function forks the simulator. After forking the simulator,
the child process gets its output files redirected to a new output
directory. The default name of the output directory is the same as
the parent with the suffix ".fN" added where N is the fork
sequence number. The name of the output directory can be
overridden using the simout keyword argument.
Output file formatting dictionary:
parent -- Path to the parent process's output directory.
fork_seq -- Fork sequence number.
pid -- PID of the child process.
Keyword Arguments:
simout -- New simulation output directory.
Return Value:
pid of the child process or 0 if running in the child.
"""
from m5 import options
global fork_count
if not _m5.core.listenersDisabled():
raise RuntimeError("Can not fork a simulator with listeners enabled")
drain()
try:
pid = os.fork()
except OSError as e:
raise e
if pid == 0:
# In child, notify objects of the fork
root = objects.Root.getInstance()
notifyFork(root)
# Setup a new output directory
parent = options.outdir
options.outdir = simout % {
"parent" : parent,
"fork_seq" : fork_count,
"pid" : os.getpid(),
}
_m5.core.setOutputDir(options.outdir)
else:
fork_count += 1
return pid
from _m5.core import disableAllListeners, listenersDisabled
from _m5.core import listenersLoopbackOnly
from _m5.core import curTick