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
|
/*========================== begin_copyright_notice ============================
Copyright (C) 2019-2021 Intel Corporation
SPDX-License-Identifier: MIT
============================= end_copyright_notice ===========================*/
//===----------------------------------------------------------------------===//
///
/// This is really just a stripped down version of the CSE pass. If we do CSE
/// in general prior to continuation splitting, the increased live ranges of
/// CSE'd values can create spills that we don't want. Here, we CSE just the
/// values that we had to hoist to minimize duplicated spills.
///
//===----------------------------------------------------------------------===//
#include "RTBuilder.h"
#include "Compiler/IGCPassSupport.h"
#include "iStdLib/utility.h"
#include "common/LLVMUtils.h"
#include "common/LLVMWarningsPush.hpp"
#include "llvm/ADT/DenseMap.h"
#include "common/LLVMWarningsPop.hpp"
using namespace llvm;
using namespace IGC;
#define DEBUG_TYPE "rayinfo-cse"
class RayInfoCSEPass : public FunctionPass
{
public:
RayInfoCSEPass() : FunctionPass(ID)
{
initializeRayInfoCSEPassPass(*PassRegistry::getPassRegistry());
}
bool runOnFunction(Function &F) override;
StringRef getPassName() const override
{
return "RayInfoCSEPass";
}
void getAnalysisUsage(llvm::AnalysisUsage &AU) const override
{
AU.setPreservesCFG();
}
static char ID;
private:
bool Changed;
};
char RayInfoCSEPass::ID = 0;
struct CSEVal
{
Instruction* Inst;
CSEVal(Instruction* I) : Inst(I) { }
bool isSentinel() const
{
return Inst == DenseMapInfo<Instruction*>::getEmptyKey() ||
Inst == DenseMapInfo<Instruction*>::getTombstoneKey();
}
};
namespace llvm {
// Tell DenseMap how to compare instructions to cheaply tell whether we
// can merge an instruction with one we've already seen.
template <> struct DenseMapInfo<CSEVal> {
static inline CSEVal getEmptyKey() {
return DenseMapInfo<Instruction*>::getEmptyKey();
}
static inline CSEVal getTombstoneKey() {
return DenseMapInfo<Instruction*>::getTombstoneKey();
}
static unsigned getHashValue(CSEVal Val);
static bool isEqual(CSEVal LHS, CSEVal RHS);
};
} // end namespace llvm
unsigned DenseMapInfo<CSEVal>::getHashValue(CSEVal Val) {
Instruction *Inst = Val.Inst;
// Mix in the opcode.
return hash_combine(
Inst->getOpcode(),
hash_combine_range(Inst->value_op_begin(), Inst->value_op_end()));
}
bool DenseMapInfo<CSEVal>::isEqual(CSEVal LHS, CSEVal RHS)
{
Instruction* LHSI = LHS.Inst, * RHSI = RHS.Inst;
if (LHS.isSentinel() || RHS.isSentinel())
return LHSI == RHSI;
if (LHSI->getOpcode() != RHSI->getOpcode())
return false;
if (LHSI->isIdenticalToWhenDefined(RHSI))
return true;
return false;
}
// Register pass to igc-opt
#define PASS_FLAG "rayinfo-cse"
#define PASS_DESCRIPTION "Lightweight CSE on entry block rayinfo intrinsics"
#define PASS_CFG_ONLY false
#define PASS_ANALYSIS false
IGC_INITIALIZE_PASS_BEGIN(RayInfoCSEPass, PASS_FLAG, PASS_DESCRIPTION, PASS_CFG_ONLY, PASS_ANALYSIS)
IGC_INITIALIZE_PASS_END(RayInfoCSEPass, PASS_FLAG, PASS_DESCRIPTION, PASS_CFG_ONLY, PASS_ANALYSIS)
bool RayInfoCSEPass::runOnFunction(Function &F)
{
Changed = false;
auto& EntryBB = F.getEntryBlock();
uint32_t MemGeneration = 0;
DenseMap<CSEVal, Instruction*> AvailableValues;
DenseMap<CSEVal, std::pair<Instruction*, uint32_t>> AvailableCalls;
for (auto II = EntryBB.begin(); II != EntryBB.end(); /* empty */)
{
auto* I = &*II++;
if (I->mayWriteToMemory())
{
// Conservatively, don't CSE values that read from memory that
// have a write in between them.
MemGeneration++;
continue;
}
auto* GII = dyn_cast<GenIntrinsicInst>(I);
if (!GII)
continue;
switch (GII->getIntrinsicID())
{
case GenISAIntrinsic::GenISA_RayInfo:
case GenISAIntrinsic::GenISA_RayTCurrent:
case GenISAIntrinsic::GenISA_LocalBufferPointer:
break;
default:
continue;
}
if (GII->doesNotAccessMemory())
{
// If this is a simple instruction that we can value number, process it.
// See if the instruction has an available value. If so, use it.
if (Value *V = AvailableValues.lookup(I)) {
I->replaceAllUsesWith(V);
I->eraseFromParent();
Changed = true;
continue;
}
// Otherwise, just remember that this value is available.
AvailableValues.insert(std::make_pair(I, I));
continue;
}
else if (GII->onlyReadsMemory())
{
// If we have an available version of this call, and if it is the right
// generation, replace this instruction.
std::pair<Instruction*, unsigned> InVal = AvailableCalls.lookup(I);
if (InVal.first != nullptr && (MemGeneration == InVal.second))
{
I->replaceAllUsesWith(InVal.first);
I->eraseFromParent();
Changed = true;
continue;
}
// Otherwise, remember that we have this instruction.
AvailableCalls.insert(
std::make_pair(I, std::make_pair(I, MemGeneration)));
continue;
}
}
return Changed;
}
namespace IGC
{
Pass* createRayInfoCSEPass(void)
{
return new RayInfoCSEPass();
}
} // namespace IGC
|
