-
-
Notifications
You must be signed in to change notification settings - Fork 5.5k
/
llvm-simdloop.cpp
264 lines (225 loc) · 8.67 KB
/
llvm-simdloop.cpp
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
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
// This file is a part of Julia. License is MIT: https://julialang.org/license
#include "llvm-version.h"
#define DEBUG_TYPE "lower_simd_loop"
// This file defines a LLVM pass that:
// 1. Set's loop information in form of metadata
// 2. If the metadata contains `julia.simdloop` finds reduction chains and marks
// floating-point operations as fast-math. `See enableUnsafeAlgebraIfReduction`.
// 3. If the metadata contains `julia.ivdep` marks all memory accesses in the loop
// as independent of each other.
//
// The pass hinges on a call to a marker function that has metadata attached to it.
// To construct the pass call `createLowerSimdLoopPass`.
#include "support/dtypes.h"
#include <llvm-c/Core.h>
#include <llvm-c/Types.h>
#include <llvm/Analysis/LoopPass.h>
#include <llvm/IR/LegacyPassManager.h>
#include <llvm/IR/Instructions.h>
#include <llvm/IR/LLVMContext.h>
#include <llvm/IR/Metadata.h>
#include <llvm/Support/Debug.h>
#include <cstdio>
#include "julia_assert.h"
namespace llvm {
/// This pass should run after reduction variables have been converted to phi nodes,
/// otherwise floating-point reductions might not be recognized as such and
/// prevent SIMDization.
struct LowerSIMDLoop : public ModulePass {
static char ID;
LowerSIMDLoop() : ModulePass(ID)
{
}
protected:
void getAnalysisUsage(AnalysisUsage &AU) const override
{
ModulePass::getAnalysisUsage(AU);
AU.addRequired<LoopInfoWrapperPass>();
AU.addPreserved<LoopInfoWrapperPass>();
AU.setPreservesCFG();
}
private:
bool runOnModule(Module &M) override;
bool markLoopInfo(Module &M, Function *marker);
/// If Phi is part of a reduction cycle of FAdd, FSub, FMul or FDiv,
/// mark the ops as permitting reassociation/commuting.
/// As of LLVM 4.0, FDiv is not handled by the loop vectorizer
void enableUnsafeAlgebraIfReduction(PHINode *Phi, Loop *L) const;
};
static unsigned getReduceOpcode(Instruction *J, Instruction *operand)
{
switch (J->getOpcode()) {
case Instruction::FSub:
if (J->getOperand(0) != operand)
return 0;
JL_FALLTHROUGH;
case Instruction::FAdd:
return Instruction::FAdd;
case Instruction::FDiv:
if (J->getOperand(0) != operand)
return 0;
JL_FALLTHROUGH;
case Instruction::FMul:
return Instruction::FMul;
default:
return 0;
}
}
void LowerSIMDLoop::enableUnsafeAlgebraIfReduction(PHINode *Phi, Loop *L) const
{
typedef SmallVector<Instruction*, 8> chainVector;
chainVector chain;
Instruction *J;
unsigned opcode = 0;
for (Instruction *I = Phi; ; I=J) {
J = NULL;
// Find the user of instruction I that is within loop L.
for (User *UI : I->users()) { /*}*/
Instruction *U = cast<Instruction>(UI);
if (L->contains(U)) {
if (J) {
LLVM_DEBUG(dbgs() << "LSL: not a reduction var because op has two internal uses: " << *I << "\n");
return;
}
J = U;
}
}
if (!J) {
LLVM_DEBUG(dbgs() << "LSL: chain prematurely terminated at " << *I << "\n");
return;
}
if (J == Phi) {
// Found the entire chain.
break;
}
if (opcode) {
// Check that arithmetic op matches prior arithmetic ops in the chain.
if (getReduceOpcode(J, I) != opcode) {
LLVM_DEBUG(dbgs() << "LSL: chain broke at " << *J << " because of wrong opcode\n");
return;
}
}
else {
// First arithmetic op in the chain.
opcode = getReduceOpcode(J, I);
if (!opcode) {
LLVM_DEBUG(dbgs() << "LSL: first arithmetic op in chain is uninteresting" << *J << "\n");
return;
}
}
chain.push_back(J);
}
for (chainVector::const_iterator K=chain.begin(); K!=chain.end(); ++K) {
LLVM_DEBUG(dbgs() << "LSL: marking " << **K << "\n");
(*K)->setFast(true);
}
}
bool LowerSIMDLoop::runOnModule(Module &M)
{
Function *loopinfo_marker = M.getFunction("julia.loopinfo_marker");
bool Changed = false;
if (loopinfo_marker)
Changed |= markLoopInfo(M, loopinfo_marker);
return Changed;
}
bool LowerSIMDLoop::markLoopInfo(Module &M, Function *marker)
{
bool Changed = false;
std::vector<Instruction*> ToDelete;
for (User *U : marker->users()) {
Instruction *I = cast<Instruction>(U);
ToDelete.push_back(I);
LoopInfo &LI = getAnalysis<LoopInfoWrapperPass>(*I->getParent()->getParent()).getLoopInfo();
Loop *L = LI.getLoopFor(I->getParent());
I->removeFromParent();
if (!L)
continue;
LLVM_DEBUG(dbgs() << "LSL: loopinfo marker found\n");
bool simd = false;
bool ivdep = false;
SmallVector<Metadata *, 8> MDs;
BasicBlock *Lh = L->getHeader();
LLVM_DEBUG(dbgs() << "LSL: loop header: " << *Lh << "\n");
// Reserve first location for self reference to the LoopID metadata node.
TempMDTuple TempNode = MDNode::getTemporary(Lh->getContext(), None);
MDs.push_back(TempNode.get());
// Walk `julia.loopinfo` metadata and filter out `julia.simdloop` and `julia.ivdep`
if (I->hasMetadataOtherThanDebugLoc()) {
MDNode *JLMD= I->getMetadata("julia.loopinfo");
if (JLMD) {
LLVM_DEBUG(dbgs() << "LSL: has julia.loopinfo metadata with " << JLMD->getNumOperands() <<" operands\n");
for (unsigned i = 0, ie = JLMD->getNumOperands(); i < ie; ++i) {
Metadata *Op = JLMD->getOperand(i);
const MDString *S = dyn_cast<MDString>(Op);
if (S) {
LLVM_DEBUG(dbgs() << "LSL: found " << S->getString() << "\n");
if (S->getString().startswith("julia")) {
if (S->getString().equals("julia.simdloop"))
simd = true;
if (S->getString().equals("julia.ivdep"))
ivdep = true;
continue;
}
}
MDs.push_back(Op);
}
}
}
LLVM_DEBUG(dbgs() << "LSL: simd: " << simd << " ivdep: " << ivdep << "\n");
MDNode *n = L->getLoopID();
if (n) {
// Loop already has a LoopID so copy over Metadata
// original loop id is operand 0
for (unsigned i = 1, ie = n->getNumOperands(); i < ie; ++i) {
Metadata *Op = n->getOperand(i);
MDs.push_back(Op);
}
}
MDNode *LoopID = MDNode::getDistinct(Lh->getContext(), MDs);
// Replace the temporary node with a self-reference.
LoopID->replaceOperandWith(0, LoopID);
L->setLoopID(LoopID);
assert(L->getLoopID());
MDNode *m = MDNode::get(Lh->getContext(), ArrayRef<Metadata *>(LoopID));
// If ivdep is true we assume that there is no memory dependency between loop iterations
// This is a fairly strong assumption and does often not hold true for generic code.
if (ivdep) {
// Mark memory references so that Loop::isAnnotatedParallel will return true for this loop.
for (BasicBlock *BB : L->blocks()) {
for (Instruction &I : *BB) {
if (I.mayReadOrWriteMemory()) {
I.setMetadata(LLVMContext::MD_mem_parallel_loop_access, m);
}
}
}
assert(L->isAnnotatedParallel());
}
if (simd) {
// Mark floating-point reductions as okay to reassociate/commute.
for (BasicBlock::iterator I = Lh->begin(), E = Lh->end(); I != E; ++I) {
if (PHINode *Phi = dyn_cast<PHINode>(I))
enableUnsafeAlgebraIfReduction(Phi, L);
else
break;
}
}
Changed = true;
}
for (Instruction *I : ToDelete)
I->deleteValue();
marker->eraseFromParent();
return Changed;
}
char LowerSIMDLoop::ID = 0;
static RegisterPass<LowerSIMDLoop> X("LowerSIMDLoop", "LowerSIMDLoop Pass",
false /* Only looks at CFG */,
false /* Analysis Pass */);
JL_DLLEXPORT Pass *createLowerSimdLoopPass()
{
return new LowerSIMDLoop();
}
extern "C" JL_DLLEXPORT void LLVMExtraAddLowerSimdLoopPass(LLVMPassManagerRef PM)
{
unwrap(PM)->add(createLowerSimdLoopPass());
}
} // namespace llvm