summaryrefslogtreecommitdiff
path: root/src/cpu/base.cc
blob: d174995a98d189c42cb230315b18178956911ecf (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
/*
 * Copyright (c) 2011 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/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->totalInstructions();
#ifndef NDEBUG
    double ipc = double(temp - lastNumInst) / (_interval / cpu->ticks(1));

    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)
    : MemObject(p), clock(p->clock), instCnt(0), _cpuId(p->cpu_id),
      interrupts(p->interrupts),
      numThreads(p->numThreads), system(p->system),
      phase(p->phase)
{
//    currentTick = curTick();

    // 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);
        }
    }
    interrupts->setCPU(this);

    if (FullSystem) {
        profileEvent = NULL;
        if (params()->profile)
            profileEvent = new ProfileEvent(this, params()->profile);
    }
    tracer = params()->tracer;
}

void
BaseCPU::enableFunctionTrace()
{
    functionTracingEnabled = true;
}

BaseCPU::~BaseCPU()
{
}

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) {
        Tick num_ticks = ticks(params()->progress_interval);

        new CPUProgressEvent(this, num_ticks);
    }
}


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());
}

Tick
BaseCPU::nextCycle()
{
    Tick next_tick = curTick() - phase + clock - 1;
    next_tick -= (next_tick % clock);
    next_tick += phase;
    return next_tick;
}

Tick
BaseCPU::nextCycle(Tick begin_tick)
{
    Tick next_tick = begin_tick;
    if (next_tick % clock != 0)
        next_tick = next_tick - (next_tick % clock) + clock;
    next_tick += phase;

    assert(next_tick >= curTick());
    return next_tick;
}

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, Port *ic, Port *dc)
{
    assert(threadContexts.size() == oldCPU->threadContexts.size());

    _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);
        */

        Port  *old_itb_port, *old_dtb_port, *new_itb_port, *new_dtb_port;
        old_itb_port = oldTC->getITBPtr()->getPort();
        old_dtb_port = oldTC->getDTBPtr()->getPort();
        new_itb_port = newTC->getITBPtr()->getPort();
        new_dtb_port = newTC->getDTBPtr()->getPort();

        // Move over any table walker ports if they exist
        if (new_itb_port && !new_itb_port->isConnected()) {
            assert(old_itb_port);
            Port *peer = old_itb_port->getPeer();;
            new_itb_port->setPeer(peer);
            peer->setPeer(new_itb_port);
        }
        if (new_dtb_port && !new_dtb_port->isConnected()) {
            assert(old_dtb_port);
            Port *peer = old_dtb_port->getPeer();;
            new_dtb_port->setPeer(peer);
            peer->setPeer(new_dtb_port);
        }
    }

    interrupts = oldCPU->interrupts;
    interrupts->setCPU(this);

    if (FullSystem) {
        for (ThreadID i = 0; i < size; ++i)
            threadContexts[i]->profileClear();

        if (profileEvent)
            schedule(profileEvent, curTick());
    }

    // Connect new CPU to old CPU's memory only if new CPU isn't
    // connected to anything.  Also connect old CPU's memory to new
    // CPU.
    if (!ic->isConnected()) {
        Port *peer = oldCPU->getPort("icache_port")->getPeer();
        ic->setPeer(peer);
        peer->setPeer(ic);
    }

    if (!dc->isConnected()) {
        Port *peer = oldCPU->getPort("dcache_port")->getPeer();
        dc->setPeer(peer);
        peer->setPeer(dc);
    }
}


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::recvTiming(PacketPtr pkt)
{
    panic("BaseCPU doesn't expect recvTiming callback!");
    return true;
}

void
BaseCPU::CpuPort::recvRetry()
{
    panic("BaseCPU doesn't expect recvRetry callback!");
}

Tick
BaseCPU::CpuPort::recvAtomic(PacketPtr pkt)
{
    panic("BaseCPU doesn't expect recvAtomic callback!");
    return curTick();
}

void
BaseCPU::CpuPort::recvFunctional(PacketPtr pkt)
{
    // No internal storage to update (in the general case). In the
    // long term this should never be called, but that assumed a split
    // into master/slave and request/response.
}

void
BaseCPU::CpuPort::recvRangeChange()
{
}