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
|
//===-- AMDGPURemoveIncompatibleFunctions.cpp -----------------------------===//
//
// 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
//
//===----------------------------------------------------------------------===//
//
/// \file
/// This pass replaces all uses of functions that use GPU features
/// incompatible with the current GPU with null then deletes the function.
//
//===----------------------------------------------------------------------===//
#include "AMDGPURemoveIncompatibleFunctions.h"
#include "AMDGPU.h"
#include "GCNSubtarget.h"
#include "llvm/Analysis/OptimizationRemarkEmitter.h"
#include "llvm/IR/Function.h"
#include "llvm/IR/Module.h"
#include "llvm/Pass.h"
#include "llvm/Target/TargetMachine.h"
#define DEBUG_TYPE "amdgpu-remove-incompatible-functions"
using namespace llvm;
namespace llvm {
extern const SubtargetFeatureKV
AMDGPUFeatureKV[AMDGPU::NumSubtargetFeatures - 1];
} // namespace llvm
namespace {
using Generation = AMDGPUSubtarget::Generation;
class AMDGPURemoveIncompatibleFunctions {
public:
AMDGPURemoveIncompatibleFunctions(const TargetMachine *TM = nullptr)
: TM(TM) {
assert(TM && "No TargetMachine!");
}
/// Checks a single function, returns true if the function must be deleted.
bool checkFunction(Function &F);
bool run(Module &M) {
assert(TM->getTargetTriple().isAMDGCN());
SmallVector<Function *, 4> FnsToDelete;
for (Function &F : M) {
if (checkFunction(F))
FnsToDelete.push_back(&F);
}
for (Function *F : FnsToDelete) {
F->replaceAllUsesWith(ConstantPointerNull::get(F->getType()));
F->eraseFromParent();
}
return !FnsToDelete.empty();
}
private:
const TargetMachine *TM = nullptr;
};
class AMDGPURemoveIncompatibleFunctionsLegacy : public ModulePass {
public:
static char ID;
AMDGPURemoveIncompatibleFunctionsLegacy(const TargetMachine *TM)
: ModulePass(ID), TM(TM) {}
bool runOnModule(Module &M) override {
AMDGPURemoveIncompatibleFunctions Pass(TM);
return Pass.run(M);
}
StringRef getPassName() const override {
return "AMDGPU Remove Incompatible Functions";
}
void getAnalysisUsage(AnalysisUsage &AU) const override {}
private:
const TargetMachine *TM = nullptr;
};
StringRef getFeatureName(unsigned Feature) {
for (const SubtargetFeatureKV &KV : AMDGPUFeatureKV)
if (Feature == KV.Value)
return KV.Key;
llvm_unreachable("Unknown Target feature");
}
const SubtargetSubTypeKV *getGPUInfo(const GCNSubtarget &ST,
StringRef GPUName) {
for (const SubtargetSubTypeKV &KV : ST.getAllProcessorDescriptions())
if (StringRef(KV.Key) == GPUName)
return &KV;
return nullptr;
}
constexpr unsigned FeaturesToCheck[] = {AMDGPU::FeatureGFX11Insts,
AMDGPU::FeatureGFX10Insts,
AMDGPU::FeatureGFX9Insts,
AMDGPU::FeatureGFX8Insts,
AMDGPU::FeatureDPP,
AMDGPU::Feature16BitInsts,
AMDGPU::FeatureDot1Insts,
AMDGPU::FeatureDot2Insts,
AMDGPU::FeatureDot3Insts,
AMDGPU::FeatureDot4Insts,
AMDGPU::FeatureDot5Insts,
AMDGPU::FeatureDot6Insts,
AMDGPU::FeatureDot7Insts,
AMDGPU::FeatureDot8Insts,
AMDGPU::FeatureExtendedImageInsts,
AMDGPU::FeatureSMemRealTime,
AMDGPU::FeatureSMemTimeInst,
AMDGPU::FeatureGWS};
FeatureBitset expandImpliedFeatures(const FeatureBitset &Features) {
FeatureBitset Result = Features;
for (const SubtargetFeatureKV &FE : AMDGPUFeatureKV) {
if (Features.test(FE.Value) && FE.Implies.any())
Result |= expandImpliedFeatures(FE.Implies.getAsBitset());
}
return Result;
}
void reportFunctionRemoved(Function &F, unsigned Feature) {
OptimizationRemarkEmitter ORE(&F);
ORE.emit([&]() {
// Note: we print the function name as part of the diagnostic because if
// debug info is not present, users get "<unknown>:0:0" as the debug
// loc. If we didn't print the function name there would be no way to
// tell which function got removed.
return OptimizationRemark(DEBUG_TYPE, "AMDGPUIncompatibleFnRemoved", &F)
<< "removing function '" << F.getName() << "': +"
<< getFeatureName(Feature)
<< " is not supported on the current target";
});
}
} // end anonymous namespace
PreservedAnalyses
AMDGPURemoveIncompatibleFunctionsPass::run(Module &M,
ModuleAnalysisManager &MAM) {
AMDGPURemoveIncompatibleFunctions Impl(TM);
if (Impl.run(M))
return PreservedAnalyses::none();
return PreservedAnalyses::all();
}
bool AMDGPURemoveIncompatibleFunctions::checkFunction(Function &F) {
if (F.isDeclaration())
return false;
const GCNSubtarget *ST =
static_cast<const GCNSubtarget *>(TM->getSubtargetImpl(F));
// Check the GPU isn't generic or generic-hsa. Generic is used for testing
// only and we don't want this pass to interfere with it.
StringRef GPUName = ST->getCPU();
if (GPUName.empty() || GPUName.starts_with("generic"))
return false;
// Try to fetch the GPU's info. If we can't, it's likely an unknown processor
// so just bail out.
const SubtargetSubTypeKV *GPUInfo = getGPUInfo(*ST, GPUName);
if (!GPUInfo)
return false;
// Get all the features implied by the current GPU, and recursively expand
// the features that imply other features.
//
// e.g. GFX90A implies FeatureGFX9, and FeatureGFX9 implies a whole set of
// other features.
const FeatureBitset GPUFeatureBits =
expandImpliedFeatures(GPUInfo->Implies.getAsBitset());
// Now that the have a FeatureBitset containing all possible features for
// the chosen GPU, check our list of "suspicious" features.
// Check that the user didn't enable any features that aren't part of that
// GPU's feature set. We only check a predetermined set of features.
for (unsigned Feature : FeaturesToCheck) {
if (ST->hasFeature(Feature) && !GPUFeatureBits.test(Feature)) {
reportFunctionRemoved(F, Feature);
return true;
}
}
// Delete FeatureWavefrontSize32 functions for
// gfx9 and below targets that don't support the mode.
// gfx10+ is implied to support both wave32 and 64 features.
// They are not in the feature set. So, we need a separate check
if (ST->getGeneration() < AMDGPUSubtarget::GFX10 &&
ST->hasFeature(AMDGPU::FeatureWavefrontSize32)) {
reportFunctionRemoved(F, AMDGPU::FeatureWavefrontSize32);
return true;
}
return false;
}
INITIALIZE_PASS(AMDGPURemoveIncompatibleFunctionsLegacy, DEBUG_TYPE,
"AMDGPU Remove Incompatible Functions", false, false)
char AMDGPURemoveIncompatibleFunctionsLegacy::ID = 0;
ModulePass *
llvm::createAMDGPURemoveIncompatibleFunctionsPass(const TargetMachine *TM) {
return new AMDGPURemoveIncompatibleFunctionsLegacy(TM);
}
|
