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
| //===-- Latency.cpp ---------------------------------------------*- C++ -*-===//
//
// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
// See https://llvm.org/LICENSE.txt for license information.
// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
//
//===----------------------------------------------------------------------===//
#include "Latency.h"
#include "Assembler.h"
#include "BenchmarkRunner.h"
#include "MCInstrDescView.h"
#include "PerfHelper.h"
#include "Target.h"
#include "llvm/ADT/STLExtras.h"
#include "llvm/MC/MCInst.h"
#include "llvm/MC/MCInstBuilder.h"
#include "llvm/Support/FormatVariadic.h"
namespace llvm {
namespace exegesis {
struct ExecutionClass {
ExecutionMode Mask;
const char *Description;
} static const kExecutionClasses[] = {
{ExecutionMode::ALWAYS_SERIAL_IMPLICIT_REGS_ALIAS |
ExecutionMode::ALWAYS_SERIAL_TIED_REGS_ALIAS,
"Repeating a single implicitly serial instruction"},
{ExecutionMode::SERIAL_VIA_EXPLICIT_REGS,
"Repeating a single explicitly serial instruction"},
{ExecutionMode::SERIAL_VIA_MEMORY_INSTR |
ExecutionMode::SERIAL_VIA_NON_MEMORY_INSTR,
"Repeating two instructions"},
};
static constexpr size_t kMaxAliasingInstructions = 10;
static std::vector<Instruction>
computeAliasingInstructions(const LLVMState &State, const Instruction &Instr,
size_t MaxAliasingInstructions,
const BitVector &ForbiddenRegisters) {
// Randomly iterate the set of instructions.
std::vector<unsigned> Opcodes;
Opcodes.resize(State.getInstrInfo().getNumOpcodes());
std::iota(Opcodes.begin(), Opcodes.end(), 0U);
std::shuffle(Opcodes.begin(), Opcodes.end(), randomGenerator());
std::vector<Instruction> AliasingInstructions;
for (const unsigned OtherOpcode : Opcodes) {
if (OtherOpcode == Instr.Description->getOpcode())
continue;
const Instruction &OtherInstr = State.getIC().getInstr(OtherOpcode);
if (OtherInstr.hasMemoryOperands())
continue;
if (Instr.hasAliasingRegistersThrough(OtherInstr, ForbiddenRegisters))
AliasingInstructions.push_back(OtherInstr);
if (AliasingInstructions.size() >= MaxAliasingInstructions)
break;
}
return AliasingInstructions;
}
static ExecutionMode getExecutionModes(const Instruction &Instr,
const BitVector &ForbiddenRegisters) {
ExecutionMode EM = ExecutionMode::UNKNOWN;
if (Instr.hasAliasingImplicitRegisters())
EM |= ExecutionMode::ALWAYS_SERIAL_IMPLICIT_REGS_ALIAS;
if (Instr.hasTiedRegisters())
EM |= ExecutionMode::ALWAYS_SERIAL_TIED_REGS_ALIAS;
if (Instr.hasMemoryOperands())
EM |= ExecutionMode::SERIAL_VIA_MEMORY_INSTR;
else {
if (Instr.hasAliasingRegisters(ForbiddenRegisters))
EM |= ExecutionMode::SERIAL_VIA_EXPLICIT_REGS;
if (Instr.hasOneUseOrOneDef())
EM |= ExecutionMode::SERIAL_VIA_NON_MEMORY_INSTR;
}
return EM;
}
static void appendCodeTemplates(const LLVMState &State,
const Instruction &Instr,
const BitVector &ForbiddenRegisters,
ExecutionMode ExecutionModeBit,
StringRef ExecutionClassDescription,
std::vector<CodeTemplate> &CodeTemplates) {
assert(isEnumValue(ExecutionModeBit) && "Bit must be a power of two");
switch (ExecutionModeBit) {
case ExecutionMode::ALWAYS_SERIAL_IMPLICIT_REGS_ALIAS:
// Nothing to do, the instruction is always serial.
LLVM_FALLTHROUGH;
case ExecutionMode::ALWAYS_SERIAL_TIED_REGS_ALIAS: {
// Picking whatever value for the tied variable will make the instruction
// serial.
CodeTemplate CT;
CT.Execution = ExecutionModeBit;
CT.Info = ExecutionClassDescription;
CT.Instructions.push_back(Instr);
CodeTemplates.push_back(std::move(CT));
return;
}
case ExecutionMode::SERIAL_VIA_MEMORY_INSTR: {
// Select back-to-back memory instruction.
// TODO: Implement me.
return;
}
case ExecutionMode::SERIAL_VIA_EXPLICIT_REGS: {
// Making the execution of this instruction serial by selecting one def
// register to alias with one use register.
const AliasingConfigurations SelfAliasing(Instr, Instr);
assert(!SelfAliasing.empty() && !SelfAliasing.hasImplicitAliasing() &&
"Instr must alias itself explicitly");
InstructionTemplate IT(Instr);
// This is a self aliasing instruction so defs and uses are from the same
// instance, hence twice IT in the following call.
setRandomAliasing(SelfAliasing, IT, IT);
CodeTemplate CT;
CT.Execution = ExecutionModeBit;
CT.Info = ExecutionClassDescription;
CT.Instructions.push_back(std::move(IT));
CodeTemplates.push_back(std::move(CT));
return;
}
case ExecutionMode::SERIAL_VIA_NON_MEMORY_INSTR: {
// Select back-to-back non-memory instruction.
for (const auto OtherInstr : computeAliasingInstructions(
State, Instr, kMaxAliasingInstructions, ForbiddenRegisters)) {
const AliasingConfigurations Forward(Instr, OtherInstr);
const AliasingConfigurations Back(OtherInstr, Instr);
InstructionTemplate ThisIT(Instr);
InstructionTemplate OtherIT(OtherInstr);
if (!Forward.hasImplicitAliasing())
setRandomAliasing(Forward, ThisIT, OtherIT);
if (!Back.hasImplicitAliasing())
setRandomAliasing(Back, OtherIT, ThisIT);
CodeTemplate CT;
CT.Execution = ExecutionModeBit;
CT.Info = ExecutionClassDescription;
CT.Instructions.push_back(std::move(ThisIT));
CT.Instructions.push_back(std::move(OtherIT));
CodeTemplates.push_back(std::move(CT));
}
return;
}
default:
llvm_unreachable("Unhandled enum value");
}
}
LatencySnippetGenerator::~LatencySnippetGenerator() = default;
Expected<std::vector<CodeTemplate>>
LatencySnippetGenerator::generateCodeTemplates(
const Instruction &Instr, const BitVector &ForbiddenRegisters) const {
std::vector<CodeTemplate> Results;
const ExecutionMode EM = getExecutionModes(Instr, ForbiddenRegisters);
for (const auto EC : kExecutionClasses) {
for (const auto ExecutionModeBit : getExecutionModeBits(EM & EC.Mask))
appendCodeTemplates(State, Instr, ForbiddenRegisters, ExecutionModeBit,
EC.Description, Results);
if (!Results.empty())
break;
}
if (Results.empty())
return make_error<Failure>(
"No strategy found to make the execution serial");
return std::move(Results);
}
LatencyBenchmarkRunner::LatencyBenchmarkRunner(const LLVMState &State,
InstructionBenchmark::ModeE Mode)
: BenchmarkRunner(State, Mode) {
assert((Mode == InstructionBenchmark::Latency ||
Mode == InstructionBenchmark::InverseThroughput) &&
"invalid mode");
}
LatencyBenchmarkRunner::~LatencyBenchmarkRunner() = default;
Expected<std::vector<BenchmarkMeasure>> LatencyBenchmarkRunner::runMeasurements(
const FunctionExecutor &Executor) const {
// Cycle measurements include some overhead from the kernel. Repeat the
// measure several times and take the minimum value.
constexpr const int NumMeasurements = 30;
int64_t MinValue = std::numeric_limits<int64_t>::max();
const char *CounterName = State.getPfmCounters().CycleCounter;
if (!CounterName)
report_fatal_error("sched model does not define a cycle counter");
for (size_t I = 0; I < NumMeasurements; ++I) {
auto ExpectedCounterValue = Executor.runAndMeasure(CounterName);
if (!ExpectedCounterValue)
return ExpectedCounterValue.takeError();
if (*ExpectedCounterValue < MinValue)
MinValue = *ExpectedCounterValue;
}
std::vector<BenchmarkMeasure> Result;
switch (Mode) {
case InstructionBenchmark::Latency:
Result = {BenchmarkMeasure::Create("latency", MinValue)};
break;
case InstructionBenchmark::InverseThroughput:
Result = {BenchmarkMeasure::Create("inverse_throughput", MinValue)};
break;
default:
break;
}
return std::move(Result);
}
} // namespace exegesis
} // namespace llvm
|