gem5/src/cpu/memtest/memtest.cc
Ron Dreslinski 567afbf6ce More cache fixes. Atomic coherence now works as well.
src/cpu/memtest/memtest.cc:
src/cpu/memtest/memtest.hh:
    Make Memtester able to test atomic as well
src/mem/bus.cc:
src/mem/bus.hh:
    Handle atomic snoops properly for cache->cache transfers
src/mem/cache/cache_impl.hh:
    Debug output.
    Clean up memleak in atomic mode.
    Set hitLatency.
    Still need to send back reasonable number for atomic return value.
src/mem/packet.cc:
    Add command strings for new commands
src/python/m5/objects/MemTest.py:
    Add param to test atomic memory.

--HG--
extra : convert_revision : 43f880e29215776167c16ea90793ebf8122c785b
2006-10-11 18:28:33 -04:00

536 lines
16 KiB
C++

/*
* Copyright (c) 2002-2005 The Regents of The University of Michigan
* 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: Erik Hallnor
* Steve Reinhardt
*/
// FIX ME: make trackBlkAddr use blocksize from actual cache, not hard coded
#include <iomanip>
#include <set>
#include <string>
#include <vector>
#include "base/misc.hh"
#include "base/statistics.hh"
//#include "cpu/simple_thread.hh"
#include "cpu/memtest/memtest.hh"
//#include "mem/cache/base_cache.hh"
//#include "mem/physical.hh"
#include "sim/builder.hh"
#include "sim/sim_events.hh"
#include "sim/stats.hh"
#include "mem/packet.hh"
#include "mem/request.hh"
#include "mem/port.hh"
#include "mem/mem_object.hh"
using namespace std;
int TESTER_ALLOCATOR=0;
bool
MemTest::CpuPort::recvTiming(Packet *pkt)
{
memtest->completeRequest(pkt);
return true;
}
Tick
MemTest::CpuPort::recvAtomic(Packet *pkt)
{
panic("MemTest doesn't expect recvAtomic callback!");
return curTick;
}
void
MemTest::CpuPort::recvFunctional(Packet *pkt)
{
//Do nothing if we see one come through
if (curTick != 0)//Supress warning durring initialization
warn("Functional Writes not implemented in MemTester\n");
//Need to find any response values that intersect and update
return;
}
void
MemTest::CpuPort::recvStatusChange(Status status)
{
if (status == RangeChange)
return;
panic("MemTest doesn't expect recvStatusChange callback!");
}
void
MemTest::CpuPort::recvRetry()
{
memtest->doRetry();
}
void
MemTest::sendPkt(Packet *pkt) {
if (atomic) {
cachePort.sendAtomic(pkt);
pkt->makeAtomicResponse();
completeRequest(pkt);
}
else if (!cachePort.sendTiming(pkt)) {
accessRetry = true;
retryPkt = pkt;
}
}
MemTest::MemTest(const string &name,
// MemInterface *_cache_interface,
// PhysicalMemory *main_mem,
// PhysicalMemory *check_mem,
unsigned _memorySize,
unsigned _percentReads,
// unsigned _percentCopies,
unsigned _percentUncacheable,
unsigned _progressInterval,
unsigned _percentSourceUnaligned,
unsigned _percentDestUnaligned,
Addr _traceAddr,
Counter _max_loads,
bool _atomic)
: MemObject(name),
tickEvent(this),
cachePort("test", this),
funcPort("functional", this),
retryPkt(NULL),
// mainMem(main_mem),
// checkMem(check_mem),
size(_memorySize),
percentReads(_percentReads),
// percentCopies(_percentCopies),
percentUncacheable(_percentUncacheable),
progressInterval(_progressInterval),
nextProgressMessage(_progressInterval),
percentSourceUnaligned(_percentSourceUnaligned),
percentDestUnaligned(percentDestUnaligned),
maxLoads(_max_loads),
atomic(_atomic)
{
vector<string> cmd;
cmd.push_back("/bin/ls");
vector<string> null_vec;
// thread = new SimpleThread(NULL, 0, NULL, 0, mainMem);
curTick = 0;
// Needs to be masked off once we know the block size.
traceBlockAddr = _traceAddr;
baseAddr1 = 0x100000;
baseAddr2 = 0x400000;
uncacheAddr = 0x800000;
// set up counters
noResponseCycles = 0;
numReads = 0;
tickEvent.schedule(0);
id = TESTER_ALLOCATOR++;
if (TESTER_ALLOCATOR > 8)
panic("False sharing memtester only allows up to 8 testers");
accessRetry = false;
}
Port *
MemTest::getPort(const std::string &if_name, int idx)
{
if (if_name == "functional")
return &funcPort;
else if (if_name == "test")
return &cachePort;
else
panic("No Such Port\n");
}
void
MemTest::init()
{
// By the time init() is called, the ports should be hooked up.
blockSize = cachePort.peerBlockSize();
blockAddrMask = blockSize - 1;
traceBlockAddr = blockAddr(traceBlockAddr);
// set up intial memory contents here
cachePort.memsetBlob(baseAddr1, 1, size);
funcPort.memsetBlob(baseAddr1, 1, size);
cachePort.memsetBlob(baseAddr2, 2, size);
funcPort.memsetBlob(baseAddr2, 2, size);
cachePort.memsetBlob(uncacheAddr, 3, size);
funcPort.memsetBlob(uncacheAddr, 3, size);
}
static void
printData(ostream &os, uint8_t *data, int nbytes)
{
os << hex << setfill('0');
// assume little-endian: print bytes from highest address to lowest
for (uint8_t *dp = data + nbytes - 1; dp >= data; --dp) {
os << setw(2) << (unsigned)*dp;
}
os << dec;
}
void
MemTest::completeRequest(Packet *pkt)
{
MemTestSenderState *state =
dynamic_cast<MemTestSenderState *>(pkt->senderState);
uint8_t *data = state->data;
uint8_t *pkt_data = pkt->getPtr<uint8_t>();
Request *req = pkt->req;
//Remove the address from the list of outstanding
std::set<unsigned>::iterator removeAddr = outstandingAddrs.find(req->getPaddr());
assert(removeAddr != outstandingAddrs.end());
outstandingAddrs.erase(removeAddr);
switch (pkt->cmd) {
case Packet::ReadResp:
if (memcmp(pkt_data, data, pkt->getSize()) != 0) {
cerr << name() << ": on read of 0x" << hex << req->getPaddr()
<< " (0x" << hex << blockAddr(req->getPaddr()) << ")"
<< "@ cycle " << dec << curTick
<< ", cache returns 0x";
printData(cerr, pkt_data, pkt->getSize());
cerr << ", expected 0x";
printData(cerr, data, pkt->getSize());
cerr << endl;
fatal("");
}
numReads++;
numReadsStat++;
if (numReads == nextProgressMessage) {
ccprintf(cerr, "%s: completed %d read accesses @%d\n",
name(), numReads, curTick);
nextProgressMessage += progressInterval;
}
if (numReads >= maxLoads)
exitSimLoop("Maximum number of loads reached!");
break;
case Packet::WriteResp:
numWritesStat++;
break;
/*
case Copy:
//Also remove dest from outstanding list
removeAddr = outstandingAddrs.find(req->dest);
assert(removeAddr != outstandingAddrs.end());
outstandingAddrs.erase(removeAddr);
numCopiesStat++;
break;
*/
default:
panic("invalid command");
}
if (blockAddr(req->getPaddr()) == traceBlockAddr) {
cerr << name() << ": completed "
<< (pkt->isWrite() ? "write" : "read")
<< " access of "
<< dec << pkt->getSize() << " bytes at address 0x"
<< hex << req->getPaddr()
<< " (0x" << hex << blockAddr(req->getPaddr()) << ")"
<< ", value = 0x";
printData(cerr, pkt_data, pkt->getSize());
cerr << " @ cycle " << dec << curTick;
cerr << endl;
}
noResponseCycles = 0;
delete state;
delete [] data;
delete pkt->req;
delete pkt;
}
void
MemTest::regStats()
{
using namespace Stats;
numReadsStat
.name(name() + ".num_reads")
.desc("number of read accesses completed")
;
numWritesStat
.name(name() + ".num_writes")
.desc("number of write accesses completed")
;
numCopiesStat
.name(name() + ".num_copies")
.desc("number of copy accesses completed")
;
}
void
MemTest::tick()
{
if (!tickEvent.scheduled())
tickEvent.schedule(curTick + cycles(1));
if (++noResponseCycles >= 500000) {
cerr << name() << ": deadlocked at cycle " << curTick << endl;
fatal("");
}
if (accessRetry) {
return;
}
//make new request
unsigned cmd = random() % 100;
unsigned offset = random() % size;
unsigned base = random() % 2;
uint64_t data = random();
unsigned access_size = random() % 4;
unsigned cacheable = random() % 100;
//If we aren't doing copies, use id as offset, and do a false sharing
//mem tester
//We can eliminate the lower bits of the offset, and then use the id
//to offset within the blks
offset &= ~63; //Not the low order bits
offset += id;
access_size = 0;
Request *req = new Request();
uint32_t flags = 0;
Addr paddr;
if (cacheable < percentUncacheable) {
flags |= UNCACHEABLE;
paddr = uncacheAddr + offset;
} else {
paddr = ((base) ? baseAddr1 : baseAddr2) + offset;
}
//bool probe = (random() % 2 == 1) && !req->isUncacheable();
bool probe = false;
paddr &= ~((1 << access_size) - 1);
req->setPhys(paddr, 1 << access_size, flags);
req->setThreadContext(id,0);
uint8_t *result = new uint8_t[8];
if (cmd < percentReads) {
// read
//For now we only allow one outstanding request per addreess per tester
//This means we assume CPU does write forwarding to reads that alias something
//in the cpu store buffer.
if (outstandingAddrs.find(paddr) != outstandingAddrs.end()) return;
else outstandingAddrs.insert(paddr);
// ***** NOTE FOR RON: I'm not sure how to access checkMem. - Kevin
funcPort.readBlob(req->getPaddr(), result, req->getSize());
if (blockAddr(paddr) == traceBlockAddr) {
cerr << name()
<< ": initiating read "
<< ((probe) ? "probe of " : "access of ")
<< dec << req->getSize() << " bytes from addr 0x"
<< hex << paddr
<< " (0x" << hex << blockAddr(paddr) << ")"
<< " at cycle "
<< dec << curTick << endl;
}
Packet *pkt = new Packet(req, Packet::ReadReq, Packet::Broadcast);
pkt->dataDynamicArray(new uint8_t[req->getSize()]);
MemTestSenderState *state = new MemTestSenderState(result);
pkt->senderState = state;
if (probe) {
cachePort.sendFunctional(pkt);
completeRequest(pkt);
} else {
// req->completionEvent = new MemCompleteEvent(req, result, this);
sendPkt(pkt);
}
} else {
// write
//For now we only allow one outstanding request per addreess per tester
//This means we assume CPU does write forwarding to reads that alias something
//in the cpu store buffer.
if (outstandingAddrs.find(paddr) != outstandingAddrs.end()) return;
else outstandingAddrs.insert(paddr);
/*
if (blockAddr(req->getPaddr()) == traceBlockAddr) {
cerr << name() << ": initiating write "
<< ((probe)?"probe of ":"access of ")
<< dec << req->getSize() << " bytes (value = 0x";
printData(cerr, data_pkt->getPtr(), req->getSize());
cerr << ") to addr 0x"
<< hex << req->getPaddr()
<< " (0x" << hex << blockAddr(req->getPaddr()) << ")"
<< " at cycle "
<< dec << curTick << endl;
}
*/
Packet *pkt = new Packet(req, Packet::WriteReq, Packet::Broadcast);
uint8_t *pkt_data = new uint8_t[req->getSize()];
pkt->dataDynamicArray(pkt_data);
memcpy(pkt_data, &data, req->getSize());
MemTestSenderState *state = new MemTestSenderState(result);
pkt->senderState = state;
funcPort.writeBlob(req->getPaddr(), pkt_data, req->getSize());
if (probe) {
cachePort.sendFunctional(pkt);
completeRequest(pkt);
} else {
// req->completionEvent = new MemCompleteEvent(req, NULL, this);
sendPkt(pkt);
}
}
/* else {
// copy
unsigned source_align = random() % 100;
unsigned dest_align = random() % 100;
unsigned offset2 = random() % size;
Addr source = ((base) ? baseAddr1 : baseAddr2) + offset;
Addr dest = ((base) ? baseAddr2 : baseAddr1) + offset2;
if (outstandingAddrs.find(source) != outstandingAddrs.end()) return;
else outstandingAddrs.insert(source);
if (outstandingAddrs.find(dest) != outstandingAddrs.end()) return;
else outstandingAddrs.insert(dest);
if (source_align >= percentSourceUnaligned) {
source = blockAddr(source);
}
if (dest_align >= percentDestUnaligned) {
dest = blockAddr(dest);
}
req->cmd = Copy;
req->flags &= ~UNCACHEABLE;
req->paddr = source;
req->dest = dest;
delete [] req->data;
req->data = new uint8_t[blockSize];
req->size = blockSize;
if (source == traceBlockAddr || dest == traceBlockAddr) {
cerr << name()
<< ": initiating copy of "
<< dec << req->size << " bytes from addr 0x"
<< hex << source
<< " (0x" << hex << blockAddr(source) << ")"
<< " to addr 0x"
<< hex << dest
<< " (0x" << hex << blockAddr(dest) << ")"
<< " at cycle "
<< dec << curTick << endl;
}*
cacheInterface->access(req);
uint8_t result[blockSize];
checkMem->access(Read, source, &result, blockSize);
checkMem->access(Write, dest, &result, blockSize);
}
*/
}
void
MemTest::doRetry()
{
if (cachePort.sendTiming(retryPkt)) {
accessRetry = false;
retryPkt = NULL;
}
}
BEGIN_DECLARE_SIM_OBJECT_PARAMS(MemTest)
// SimObjectParam<BaseCache *> cache;
// SimObjectParam<PhysicalMemory *> main_mem;
// SimObjectParam<PhysicalMemory *> check_mem;
Param<unsigned> memory_size;
Param<unsigned> percent_reads;
// Param<unsigned> percent_copies;
Param<unsigned> percent_uncacheable;
Param<unsigned> progress_interval;
Param<unsigned> percent_source_unaligned;
Param<unsigned> percent_dest_unaligned;
Param<Addr> trace_addr;
Param<Counter> max_loads;
Param<bool> atomic;
END_DECLARE_SIM_OBJECT_PARAMS(MemTest)
BEGIN_INIT_SIM_OBJECT_PARAMS(MemTest)
// INIT_PARAM(cache, "L1 cache"),
// INIT_PARAM(main_mem, "hierarchical memory"),
// INIT_PARAM(check_mem, "check memory"),
INIT_PARAM(memory_size, "memory size"),
INIT_PARAM(percent_reads, "target read percentage"),
// INIT_PARAM(percent_copies, "target copy percentage"),
INIT_PARAM(percent_uncacheable, "target uncacheable percentage"),
INIT_PARAM(progress_interval, "progress report interval (in accesses)"),
INIT_PARAM(percent_source_unaligned,
"percent of copy source address that are unaligned"),
INIT_PARAM(percent_dest_unaligned,
"percent of copy dest address that are unaligned"),
INIT_PARAM(trace_addr, "address to trace"),
INIT_PARAM(max_loads, "terminate when we have reached this load count"),
INIT_PARAM(atomic, "Is the tester testing atomic mode (or timing)")
END_INIT_SIM_OBJECT_PARAMS(MemTest)
CREATE_SIM_OBJECT(MemTest)
{
return new MemTest(getInstanceName(), /*cache->getInterface(),*/ /*main_mem,*/
/*check_mem,*/ memory_size, percent_reads, /*percent_copies,*/
percent_uncacheable, progress_interval,
percent_source_unaligned, percent_dest_unaligned,
trace_addr, max_loads, atomic);
}
REGISTER_SIM_OBJECT("MemTest", MemTest)