gem5/src/cpu/base.cc
Andreas Hansson d53d04473e Clock: Rework clocks to avoid tick-to-cycle transformations
This patch introduces the notion of a clock update function that aims
to avoid costly divisions when turning the current tick into a
cycle. Each clocked object advances a private (hidden) cycle member
and a tick member and uses these to implement functions for getting
the tick of the next cycle, or the tick of a cycle some time in the
future.

In the different modules using the clocks, changes are made to avoid
counting in ticks only to later translate to cycles. There are a few
oddities in how the O3 and inorder CPU count idle cycles, as seen by a
few locations where a cycle is subtracted in the calculation. This is
done such that the regression does not change any stats, but should be
revisited in a future patch.

Another, much needed, change that is not done as part of this patch is
to introduce a new typedef uint64_t Cycle to be able to at least hint
at the unit of the variables counting Ticks vs Cycles. This will be
done as a follow-up patch.

As an additional follow up, the thread context still uses ticks for
the book keeping of last activate and last suspend and this should
probably also be changed into cycles as well.
2012-08-28 14:30:31 -04:00

547 lines
18 KiB
C++

/*
* Copyright (c) 2011-2012 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) 2002-2005 The Regents of The University of Michigan
* Copyright (c) 2011 Regents of the University of California
* 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: Steve Reinhardt
* Nathan Binkert
* Rick Strong
*/
#include <iostream>
#include <sstream>
#include <string>
#include "arch/tlb.hh"
#include "base/loader/symtab.hh"
#include "base/cprintf.hh"
#include "base/misc.hh"
#include "base/output.hh"
#include "base/trace.hh"
#include "cpu/base.hh"
#include "cpu/checker/cpu.hh"
#include "cpu/cpuevent.hh"
#include "cpu/profile.hh"
#include "cpu/thread_context.hh"
#include "debug/SyscallVerbose.hh"
#include "params/BaseCPU.hh"
#include "sim/full_system.hh"
#include "sim/process.hh"
#include "sim/sim_events.hh"
#include "sim/sim_exit.hh"
#include "sim/system.hh"
// Hack
#include "sim/stat_control.hh"
using namespace std;
vector<BaseCPU *> BaseCPU::cpuList;
// This variable reflects the max number of threads in any CPU. Be
// careful to only use it once all the CPUs that you care about have
// been initialized
int maxThreadsPerCPU = 1;
CPUProgressEvent::CPUProgressEvent(BaseCPU *_cpu, Tick ival)
: Event(Event::Progress_Event_Pri), _interval(ival), lastNumInst(0),
cpu(_cpu), _repeatEvent(true)
{
if (_interval)
cpu->schedule(this, curTick() + _interval);
}
void
CPUProgressEvent::process()
{
Counter temp = cpu->totalOps();
#ifndef NDEBUG
double ipc = double(temp - lastNumInst) / (_interval / cpu->clockPeriod());
DPRINTFN("%s progress event, total committed:%i, progress insts committed: "
"%lli, IPC: %0.8d\n", cpu->name(), temp, temp - lastNumInst,
ipc);
ipc = 0.0;
#else
cprintf("%lli: %s progress event, total committed:%i, progress insts "
"committed: %lli\n", curTick(), cpu->name(), temp,
temp - lastNumInst);
#endif
lastNumInst = temp;
if (_repeatEvent)
cpu->schedule(this, curTick() + _interval);
}
const char *
CPUProgressEvent::description() const
{
return "CPU Progress";
}
BaseCPU::BaseCPU(Params *p, bool is_checker)
: MemObject(p), instCnt(0), _cpuId(p->cpu_id),
_instMasterId(p->system->getMasterId(name() + ".inst")),
_dataMasterId(p->system->getMasterId(name() + ".data")),
interrupts(p->interrupts),
numThreads(p->numThreads), system(p->system)
{
// if Python did not provide a valid ID, do it here
if (_cpuId == -1 ) {
_cpuId = cpuList.size();
}
// add self to global list of CPUs
cpuList.push_back(this);
DPRINTF(SyscallVerbose, "Constructing CPU with id %d\n", _cpuId);
if (numThreads > maxThreadsPerCPU)
maxThreadsPerCPU = numThreads;
// allocate per-thread instruction-based event queues
comInstEventQueue = new EventQueue *[numThreads];
for (ThreadID tid = 0; tid < numThreads; ++tid)
comInstEventQueue[tid] =
new EventQueue("instruction-based event queue");
//
// set up instruction-count-based termination events, if any
//
if (p->max_insts_any_thread != 0) {
const char *cause = "a thread reached the max instruction count";
for (ThreadID tid = 0; tid < numThreads; ++tid) {
Event *event = new SimLoopExitEvent(cause, 0);
comInstEventQueue[tid]->schedule(event, p->max_insts_any_thread);
}
}
if (p->max_insts_all_threads != 0) {
const char *cause = "all threads reached the max instruction count";
// allocate & initialize shared downcounter: each event will
// decrement this when triggered; simulation will terminate
// when counter reaches 0
int *counter = new int;
*counter = numThreads;
for (ThreadID tid = 0; tid < numThreads; ++tid) {
Event *event = new CountedExitEvent(cause, *counter);
comInstEventQueue[tid]->schedule(event, p->max_insts_all_threads);
}
}
// allocate per-thread load-based event queues
comLoadEventQueue = new EventQueue *[numThreads];
for (ThreadID tid = 0; tid < numThreads; ++tid)
comLoadEventQueue[tid] = new EventQueue("load-based event queue");
//
// set up instruction-count-based termination events, if any
//
if (p->max_loads_any_thread != 0) {
const char *cause = "a thread reached the max load count";
for (ThreadID tid = 0; tid < numThreads; ++tid) {
Event *event = new SimLoopExitEvent(cause, 0);
comLoadEventQueue[tid]->schedule(event, p->max_loads_any_thread);
}
}
if (p->max_loads_all_threads != 0) {
const char *cause = "all threads reached the max load count";
// allocate & initialize shared downcounter: each event will
// decrement this when triggered; simulation will terminate
// when counter reaches 0
int *counter = new int;
*counter = numThreads;
for (ThreadID tid = 0; tid < numThreads; ++tid) {
Event *event = new CountedExitEvent(cause, *counter);
comLoadEventQueue[tid]->schedule(event, p->max_loads_all_threads);
}
}
functionTracingEnabled = false;
if (p->function_trace) {
const string fname = csprintf("ftrace.%s", name());
functionTraceStream = simout.find(fname);
if (!functionTraceStream)
functionTraceStream = simout.create(fname);
currentFunctionStart = currentFunctionEnd = 0;
functionEntryTick = p->function_trace_start;
if (p->function_trace_start == 0) {
functionTracingEnabled = true;
} else {
typedef EventWrapper<BaseCPU, &BaseCPU::enableFunctionTrace> wrap;
Event *event = new wrap(this, true);
schedule(event, p->function_trace_start);
}
}
// The interrupts should always be present unless this CPU is
// switched in later or in case it is a checker CPU
if (!params()->defer_registration && !is_checker) {
if (interrupts) {
interrupts->setCPU(this);
} else {
fatal("CPU %s has no interrupt controller.\n"
"Ensure createInterruptController() is called.\n", name());
}
}
if (FullSystem) {
profileEvent = NULL;
if (params()->profile)
profileEvent = new ProfileEvent(this, params()->profile);
}
tracer = params()->tracer;
}
void
BaseCPU::enableFunctionTrace()
{
functionTracingEnabled = true;
}
BaseCPU::~BaseCPU()
{
delete profileEvent;
delete[] comLoadEventQueue;
delete[] comInstEventQueue;
}
void
BaseCPU::init()
{
if (!params()->defer_registration)
registerThreadContexts();
}
void
BaseCPU::startup()
{
if (FullSystem) {
if (!params()->defer_registration && profileEvent)
schedule(profileEvent, curTick());
}
if (params()->progress_interval) {
new CPUProgressEvent(this, params()->progress_interval);
}
}
void
BaseCPU::regStats()
{
using namespace Stats;
numCycles
.name(name() + ".numCycles")
.desc("number of cpu cycles simulated")
;
numWorkItemsStarted
.name(name() + ".numWorkItemsStarted")
.desc("number of work items this cpu started")
;
numWorkItemsCompleted
.name(name() + ".numWorkItemsCompleted")
.desc("number of work items this cpu completed")
;
int size = threadContexts.size();
if (size > 1) {
for (int i = 0; i < size; ++i) {
stringstream namestr;
ccprintf(namestr, "%s.ctx%d", name(), i);
threadContexts[i]->regStats(namestr.str());
}
} else if (size == 1)
threadContexts[0]->regStats(name());
}
MasterPort &
BaseCPU::getMasterPort(const string &if_name, int idx)
{
// Get the right port based on name. This applies to all the
// subclasses of the base CPU and relies on their implementation
// of getDataPort and getInstPort. In all cases there methods
// return a CpuPort pointer.
if (if_name == "dcache_port")
return getDataPort();
else if (if_name == "icache_port")
return getInstPort();
else
return MemObject::getMasterPort(if_name, idx);
}
void
BaseCPU::registerThreadContexts()
{
ThreadID size = threadContexts.size();
for (ThreadID tid = 0; tid < size; ++tid) {
ThreadContext *tc = threadContexts[tid];
/** This is so that contextId and cpuId match where there is a
* 1cpu:1context relationship. Otherwise, the order of registration
* could affect the assignment and cpu 1 could have context id 3, for
* example. We may even want to do something like this for SMT so that
* cpu 0 has the lowest thread contexts and cpu N has the highest, but
* I'll just do this for now
*/
if (numThreads == 1)
tc->setContextId(system->registerThreadContext(tc, _cpuId));
else
tc->setContextId(system->registerThreadContext(tc));
if (!FullSystem)
tc->getProcessPtr()->assignThreadContext(tc->contextId());
}
}
int
BaseCPU::findContext(ThreadContext *tc)
{
ThreadID size = threadContexts.size();
for (ThreadID tid = 0; tid < size; ++tid) {
if (tc == threadContexts[tid])
return tid;
}
return 0;
}
void
BaseCPU::switchOut()
{
if (profileEvent && profileEvent->scheduled())
deschedule(profileEvent);
}
void
BaseCPU::takeOverFrom(BaseCPU *oldCPU)
{
assert(threadContexts.size() == oldCPU->threadContexts.size());
assert(_cpuId == oldCPU->cpuId());
ThreadID size = threadContexts.size();
for (ThreadID i = 0; i < size; ++i) {
ThreadContext *newTC = threadContexts[i];
ThreadContext *oldTC = oldCPU->threadContexts[i];
newTC->takeOverFrom(oldTC);
CpuEvent::replaceThreadContext(oldTC, newTC);
assert(newTC->contextId() == oldTC->contextId());
assert(newTC->threadId() == oldTC->threadId());
system->replaceThreadContext(newTC, newTC->contextId());
/* This code no longer works since the zero register (e.g.,
* r31 on Alpha) doesn't necessarily contain zero at this
* point.
if (DTRACE(Context))
ThreadContext::compare(oldTC, newTC);
*/
MasterPort *old_itb_port = oldTC->getITBPtr()->getMasterPort();
MasterPort *old_dtb_port = oldTC->getDTBPtr()->getMasterPort();
MasterPort *new_itb_port = newTC->getITBPtr()->getMasterPort();
MasterPort *new_dtb_port = newTC->getDTBPtr()->getMasterPort();
// Move over any table walker ports if they exist
if (new_itb_port) {
assert(!new_itb_port->isConnected());
assert(old_itb_port);
assert(old_itb_port->isConnected());
SlavePort &slavePort = old_itb_port->getSlavePort();
old_itb_port->unbind();
new_itb_port->bind(slavePort);
}
if (new_dtb_port) {
assert(!new_dtb_port->isConnected());
assert(old_dtb_port);
assert(old_dtb_port->isConnected());
SlavePort &slavePort = old_dtb_port->getSlavePort();
old_dtb_port->unbind();
new_dtb_port->bind(slavePort);
}
// Checker whether or not we have to transfer CheckerCPU
// objects over in the switch
CheckerCPU *oldChecker = oldTC->getCheckerCpuPtr();
CheckerCPU *newChecker = newTC->getCheckerCpuPtr();
if (oldChecker && newChecker) {
MasterPort *old_checker_itb_port =
oldChecker->getITBPtr()->getMasterPort();
MasterPort *old_checker_dtb_port =
oldChecker->getDTBPtr()->getMasterPort();
MasterPort *new_checker_itb_port =
newChecker->getITBPtr()->getMasterPort();
MasterPort *new_checker_dtb_port =
newChecker->getDTBPtr()->getMasterPort();
// Move over any table walker ports if they exist for checker
if (new_checker_itb_port) {
assert(!new_checker_itb_port->isConnected());
assert(old_checker_itb_port);
assert(old_checker_itb_port->isConnected());
SlavePort &slavePort = old_checker_itb_port->getSlavePort();
old_checker_itb_port->unbind();
new_checker_itb_port->bind(slavePort);
}
if (new_checker_dtb_port) {
assert(!new_checker_dtb_port->isConnected());
assert(old_checker_dtb_port);
assert(old_checker_dtb_port->isConnected());
SlavePort &slavePort = old_checker_dtb_port->getSlavePort();
old_checker_dtb_port->unbind();
new_checker_dtb_port->bind(slavePort);
}
}
}
interrupts = oldCPU->interrupts;
interrupts->setCPU(this);
oldCPU->interrupts = NULL;
if (FullSystem) {
for (ThreadID i = 0; i < size; ++i)
threadContexts[i]->profileClear();
if (profileEvent)
schedule(profileEvent, curTick());
}
// All CPUs have an instruction and a data port, and the new CPU's
// ports are dangling while the old CPU has its ports connected
// already. Unbind the old CPU and then bind the ports of the one
// we are switching to.
assert(!getInstPort().isConnected());
assert(oldCPU->getInstPort().isConnected());
SlavePort &inst_peer_port = oldCPU->getInstPort().getSlavePort();
oldCPU->getInstPort().unbind();
getInstPort().bind(inst_peer_port);
assert(!getDataPort().isConnected());
assert(oldCPU->getDataPort().isConnected());
SlavePort &data_peer_port = oldCPU->getDataPort().getSlavePort();
oldCPU->getDataPort().unbind();
getDataPort().bind(data_peer_port);
}
BaseCPU::ProfileEvent::ProfileEvent(BaseCPU *_cpu, Tick _interval)
: cpu(_cpu), interval(_interval)
{ }
void
BaseCPU::ProfileEvent::process()
{
ThreadID size = cpu->threadContexts.size();
for (ThreadID i = 0; i < size; ++i) {
ThreadContext *tc = cpu->threadContexts[i];
tc->profileSample();
}
cpu->schedule(this, curTick() + interval);
}
void
BaseCPU::serialize(std::ostream &os)
{
SERIALIZE_SCALAR(instCnt);
interrupts->serialize(os);
}
void
BaseCPU::unserialize(Checkpoint *cp, const std::string &section)
{
UNSERIALIZE_SCALAR(instCnt);
interrupts->unserialize(cp, section);
}
void
BaseCPU::traceFunctionsInternal(Addr pc)
{
if (!debugSymbolTable)
return;
// if pc enters different function, print new function symbol and
// update saved range. Otherwise do nothing.
if (pc < currentFunctionStart || pc >= currentFunctionEnd) {
string sym_str;
bool found = debugSymbolTable->findNearestSymbol(pc, sym_str,
currentFunctionStart,
currentFunctionEnd);
if (!found) {
// no symbol found: use addr as label
sym_str = csprintf("0x%x", pc);
currentFunctionStart = pc;
currentFunctionEnd = pc + 1;
}
ccprintf(*functionTraceStream, " (%d)\n%d: %s",
curTick() - functionEntryTick, curTick(), sym_str);
functionEntryTick = curTick();
}
}
bool
BaseCPU::CpuPort::recvTimingResp(PacketPtr pkt)
{
panic("BaseCPU doesn't expect recvTiming!\n");
return true;
}
void
BaseCPU::CpuPort::recvRetry()
{
panic("BaseCPU doesn't expect recvRetry!\n");
}
void
BaseCPU::CpuPort::recvFunctionalSnoop(PacketPtr pkt)
{
// No internal storage to update (in the general case). A CPU with
// internal storage, e.g. an LSQ that should be part of the
// coherent memory has to check against stored data.
}