ae5a51994c
Previously, the cache recorder used the Sequencer to issue flush requests and cache warmup requests. The GPU however uses GPUCoalescer to access the cache, and not the Sequencer. This commit adds a GPUCoalescer map to the cache recorder and uses it to send flushes and cache warmup requests to any GPU caches in the system Change-Id: I10490cf5e561c8559a98d4eb0550c62eefe769c9
244 lines
8.8 KiB
C++
244 lines
8.8 KiB
C++
/*
|
|
* Copyright (c) 1999-2012 Mark D. Hill and David A. Wood
|
|
* 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.
|
|
*/
|
|
|
|
#include "mem/ruby/system/CacheRecorder.hh"
|
|
|
|
#include "debug/RubyCacheTrace.hh"
|
|
#include "mem/packet.hh"
|
|
#include "mem/ruby/system/GPUCoalescer.hh"
|
|
#include "mem/ruby/system/RubySystem.hh"
|
|
#include "mem/ruby/system/Sequencer.hh"
|
|
#include "sim/sim_exit.hh"
|
|
|
|
namespace gem5
|
|
{
|
|
|
|
namespace ruby
|
|
{
|
|
|
|
void
|
|
TraceRecord::print(std::ostream& out) const
|
|
{
|
|
out << "[TraceRecord: Node, " << m_cntrl_id << ", "
|
|
<< m_data_address << ", " << m_pc_address << ", "
|
|
<< m_type << ", Time: " << m_time << "]";
|
|
}
|
|
|
|
CacheRecorder::CacheRecorder()
|
|
: m_uncompressed_trace(NULL),
|
|
m_uncompressed_trace_size(0),
|
|
m_block_size_bytes(RubySystem::getBlockSizeBytes())
|
|
{
|
|
}
|
|
|
|
CacheRecorder::CacheRecorder(uint8_t* uncompressed_trace,
|
|
uint64_t uncompressed_trace_size,
|
|
std::vector<Sequencer*>& seq_map,
|
|
std::vector<GPUCoalescer*>& coal_map,
|
|
uint64_t block_size_bytes)
|
|
: m_uncompressed_trace(uncompressed_trace),
|
|
m_uncompressed_trace_size(uncompressed_trace_size),
|
|
m_seq_map(seq_map), m_coalescer_map(coal_map), m_bytes_read(0),
|
|
m_records_read(0), m_records_flushed(0),
|
|
m_block_size_bytes(block_size_bytes)
|
|
{
|
|
if (m_uncompressed_trace != NULL) {
|
|
if (m_block_size_bytes < RubySystem::getBlockSizeBytes()) {
|
|
// Block sizes larger than when the trace was recorded are not
|
|
// supported, as we cannot reliably turn accesses to smaller blocks
|
|
// into larger ones.
|
|
panic("Recorded cache block size (%d) < current block size (%d) !!",
|
|
m_block_size_bytes, RubySystem::getBlockSizeBytes());
|
|
}
|
|
}
|
|
}
|
|
|
|
CacheRecorder::~CacheRecorder()
|
|
{
|
|
if (m_uncompressed_trace != NULL) {
|
|
delete [] m_uncompressed_trace;
|
|
m_uncompressed_trace = NULL;
|
|
}
|
|
m_seq_map.clear();
|
|
m_coalescer_map.clear();
|
|
}
|
|
|
|
void
|
|
CacheRecorder::enqueueNextFlushRequest()
|
|
{
|
|
if (m_records_flushed < m_records.size()) {
|
|
TraceRecord* rec = m_records[m_records_flushed];
|
|
m_records_flushed++;
|
|
auto req = std::make_shared<Request>(rec->m_data_address,
|
|
m_block_size_bytes, 0,
|
|
Request::funcRequestorId);
|
|
MemCmd::Command requestType = MemCmd::FlushReq;
|
|
Packet *pkt = new Packet(req, requestType);
|
|
|
|
Sequencer* m_sequencer_ptr = m_seq_map[rec->m_cntrl_id];
|
|
GPUCoalescer* m_coal_ptr = m_coalescer_map[rec->m_cntrl_id];
|
|
assert(m_sequencer_ptr != NULL);
|
|
if (m_coal_ptr == NULL)
|
|
m_sequencer_ptr->makeRequest(pkt);
|
|
else {
|
|
pkt->req->setReqInstSeqNum(m_records_flushed - 1);
|
|
m_coal_ptr->makeRequest(pkt);
|
|
}
|
|
|
|
DPRINTF(RubyCacheTrace, "Flushing %s\n", *rec);
|
|
|
|
} else {
|
|
if (m_records_flushed > 0) {
|
|
exitSimLoop("Finished Drain", 0);
|
|
}
|
|
DPRINTF(RubyCacheTrace, "Flushed all %d records\n", m_records_flushed);
|
|
}
|
|
}
|
|
|
|
void
|
|
CacheRecorder::enqueueNextFetchRequest()
|
|
{
|
|
if (m_bytes_read < m_uncompressed_trace_size) {
|
|
TraceRecord* traceRecord = (TraceRecord*) (m_uncompressed_trace +
|
|
m_bytes_read);
|
|
|
|
DPRINTF(RubyCacheTrace, "Issuing %s\n", *traceRecord);
|
|
|
|
for (int rec_bytes_read = 0; rec_bytes_read < m_block_size_bytes;
|
|
rec_bytes_read += RubySystem::getBlockSizeBytes()) {
|
|
RequestPtr req;
|
|
MemCmd::Command requestType;
|
|
|
|
if (traceRecord->m_type == RubyRequestType_LD) {
|
|
requestType = MemCmd::ReadReq;
|
|
req = std::make_shared<Request>(
|
|
traceRecord->m_data_address + rec_bytes_read,
|
|
RubySystem::getBlockSizeBytes(), 0,
|
|
Request::funcRequestorId);
|
|
} else if (traceRecord->m_type == RubyRequestType_IFETCH) {
|
|
requestType = MemCmd::ReadReq;
|
|
req = std::make_shared<Request>(
|
|
traceRecord->m_data_address + rec_bytes_read,
|
|
RubySystem::getBlockSizeBytes(),
|
|
Request::INST_FETCH, Request::funcRequestorId);
|
|
} else {
|
|
requestType = MemCmd::WriteReq;
|
|
req = std::make_shared<Request>(
|
|
traceRecord->m_data_address + rec_bytes_read,
|
|
RubySystem::getBlockSizeBytes(), 0,
|
|
Request::funcRequestorId);
|
|
}
|
|
|
|
Packet *pkt = new Packet(req, requestType);
|
|
pkt->dataStatic(traceRecord->m_data + rec_bytes_read);
|
|
|
|
Sequencer* m_sequencer_ptr = m_seq_map[traceRecord->m_cntrl_id];
|
|
GPUCoalescer* m_coal_ptr;
|
|
m_coal_ptr = m_coalescer_map[traceRecord->m_cntrl_id];
|
|
assert(m_sequencer_ptr != NULL);
|
|
if (m_coal_ptr == NULL)
|
|
m_sequencer_ptr->makeRequest(pkt);
|
|
else {
|
|
pkt->req->setReqInstSeqNum(m_records_read);
|
|
m_coal_ptr->makeRequest(pkt);
|
|
}
|
|
}
|
|
|
|
m_bytes_read += (sizeof(TraceRecord) + m_block_size_bytes);
|
|
m_records_read++;
|
|
} else {
|
|
exitSimLoop("Finished Warmup", 0);
|
|
DPRINTF(RubyCacheTrace, "Fetched all %d records\n", m_records_read);
|
|
}
|
|
}
|
|
|
|
void
|
|
CacheRecorder::addRecord(int cntrl, Addr data_addr, Addr pc_addr,
|
|
RubyRequestType type, Tick time, DataBlock& data)
|
|
{
|
|
TraceRecord* rec = (TraceRecord*)malloc(sizeof(TraceRecord) +
|
|
m_block_size_bytes);
|
|
rec->m_cntrl_id = cntrl;
|
|
rec->m_time = time;
|
|
rec->m_data_address = data_addr;
|
|
rec->m_pc_address = pc_addr;
|
|
rec->m_type = type;
|
|
memcpy(rec->m_data, data.getData(0, m_block_size_bytes),
|
|
m_block_size_bytes);
|
|
|
|
DPRINTF(RubyCacheTrace, "Inside addRecord with cntrl id %d and type %d\n",
|
|
cntrl, type);
|
|
m_records.push_back(rec);
|
|
}
|
|
|
|
uint64_t
|
|
CacheRecorder::aggregateRecords(uint8_t **buf, uint64_t total_size)
|
|
{
|
|
std::sort(m_records.begin(), m_records.end(), compareTraceRecords);
|
|
|
|
int size = m_records.size();
|
|
uint64_t current_size = 0;
|
|
int record_size = sizeof(TraceRecord) + m_block_size_bytes;
|
|
|
|
for (int i = 0; i < size; ++i) {
|
|
// Determine if we need to expand the buffer size
|
|
if (current_size + record_size > total_size) {
|
|
uint8_t* new_buf = new (std::nothrow) uint8_t[total_size * 2];
|
|
if (new_buf == NULL) {
|
|
fatal("Unable to allocate buffer of size %s\n",
|
|
total_size * 2);
|
|
}
|
|
total_size = total_size * 2;
|
|
uint8_t* old_buf = *buf;
|
|
memcpy(new_buf, old_buf, current_size);
|
|
*buf = new_buf;
|
|
delete [] old_buf;
|
|
}
|
|
|
|
// Copy the current record into the buffer
|
|
memcpy(&((*buf)[current_size]), m_records[i], record_size);
|
|
current_size += record_size;
|
|
|
|
free(m_records[i]);
|
|
m_records[i] = NULL;
|
|
}
|
|
|
|
m_records.clear();
|
|
return current_size;
|
|
}
|
|
|
|
uint64_t
|
|
CacheRecorder::getNumRecords() const
|
|
{
|
|
return m_records.size();
|
|
}
|
|
|
|
} // namespace ruby
|
|
} // namespace gem5
|