1//===-------------------------- CodeRegion.h -------------------*- C++ -* -===//
2//
3// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
4// See https://llvm.org/LICENSE.txt for license information.
5// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
6//
7//===----------------------------------------------------------------------===//
8/// \file
9///
10/// This file implements class CodeRegion and CodeRegions, InstrumentRegion,
11/// AnalysisRegions, and InstrumentRegions.
12///
13/// A CodeRegion describes a region of assembly code guarded by special LLVM-MCA
14/// comment directives.
15///
16/// # LLVM-MCA-BEGIN foo
17/// ... ## asm
18/// # LLVM-MCA-END
19///
20/// A comment starting with substring LLVM-MCA-BEGIN marks the beginning of a
21/// new region of code.
22/// A comment starting with substring LLVM-MCA-END marks the end of the
23/// last-seen region of code.
24///
25/// Code regions are not allowed to overlap. Each region can have a optional
26/// description; internally, regions are described by a range of source
27/// locations (SMLoc objects).
28///
29/// An instruction (a MCInst) is added to a CodeRegion R only if its
30/// location is in range [R.RangeStart, R.RangeEnd].
31///
32/// A InstrumentRegion describes a region of assembly code guarded by
33/// special LLVM-MCA comment directives.
34///
35/// # LLVM-MCA-<INSTRUMENTATION_TYPE> <data>
36/// ... ## asm
37///
38/// where INSTRUMENTATION_TYPE is a type defined in llvm and expects to use
39/// data.
40///
41/// A comment starting with substring LLVM-MCA-<INSTRUMENTATION_TYPE>
42/// brings data into scope for llvm-mca to use in its analysis for
43/// all following instructions.
44///
45/// If the same INSTRUMENTATION_TYPE is found later in the instruction list,
46/// then the original InstrumentRegion will be automatically ended,
47/// and a new InstrumentRegion will begin.
48///
49/// If there are comments containing the different INSTRUMENTATION_TYPEs,
50/// then both data sets remain available. In contrast with a CodeRegion,
51/// an InstrumentRegion does not need a comment to end the region.
52//
53// An instruction (a MCInst) is added to an InstrumentRegion R only
54// if its location is in range [R.RangeStart, R.RangeEnd].
55//
56//===----------------------------------------------------------------------===//
57
58#ifndef LLVM_TOOLS_LLVM_MCA_CODEREGION_H
59#define LLVM_TOOLS_LLVM_MCA_CODEREGION_H
60
61#include "llvm/ADT/ArrayRef.h"
62#include "llvm/ADT/SmallPtrSet.h"
63#include "llvm/ADT/SmallVector.h"
64#include "llvm/ADT/StringMap.h"
65#include "llvm/ADT/StringRef.h"
66#include "llvm/MC/MCInst.h"
67#include "llvm/MCA/CustomBehaviour.h"
68#include "llvm/Support/Error.h"
69#include "llvm/Support/SMLoc.h"
70#include "llvm/Support/SourceMgr.h"
71#include <vector>
72
73namespace llvm {
74namespace mca {
75
76/// A region of assembly code.
77///
78/// It identifies a sequence of machine instructions.
79class CodeRegion {
80 // An optional descriptor for this region.
81 llvm::StringRef Description;
82 // Instructions that form this region.
83 llvm::SmallVector<llvm::MCInst, 16> Instructions;
84 // Source location range.
85 llvm::SMLoc RangeStart;
86 llvm::SMLoc RangeEnd;
87
88 CodeRegion(const CodeRegion &) = delete;
89 CodeRegion &operator=(const CodeRegion &) = delete;
90
91public:
92 CodeRegion(llvm::StringRef Desc, llvm::SMLoc Start)
93 : Description(Desc), RangeStart(Start) {}
94
95 virtual ~CodeRegion() = default;
96
97 void addInstruction(const llvm::MCInst &Instruction) {
98 Instructions.emplace_back(Args: Instruction);
99 }
100
101 // Remove the given instructions from the set, for unsupported instructions
102 // being skipped. Returns an ArrayRef for the updated vector of Instructions.
103 [[nodiscard]] llvm::ArrayRef<llvm::MCInst>
104 dropInstructions(const llvm::SmallPtrSetImpl<const llvm::MCInst *> &Insts) {
105 if (Insts.empty())
106 return Instructions;
107 llvm::erase_if(C&: Instructions, P: [&Insts](const llvm::MCInst &Inst) {
108 return Insts.contains(Ptr: &Inst);
109 });
110 return Instructions;
111 }
112
113 llvm::SMLoc startLoc() const { return RangeStart; }
114 llvm::SMLoc endLoc() const { return RangeEnd; }
115
116 void setEndLocation(llvm::SMLoc End) { RangeEnd = End; }
117 bool empty() const { return Instructions.empty(); }
118 bool isLocInRange(llvm::SMLoc Loc) const;
119
120 llvm::ArrayRef<llvm::MCInst> getInstructions() const { return Instructions; }
121
122 llvm::StringRef getDescription() const { return Description; }
123};
124
125/// Alias AnalysisRegion with CodeRegion since CodeRegionGenerator
126/// is absract and AnalysisRegionGenerator operates on AnalysisRegions
127using AnalysisRegion = CodeRegion;
128
129/// A CodeRegion that contains instrumentation that can be used
130/// in analysis of the region.
131class InstrumentRegion : public CodeRegion {
132 /// Instrument for this region.
133 UniqueInstrument I;
134
135public:
136 InstrumentRegion(llvm::StringRef Desc, llvm::SMLoc Start, UniqueInstrument I)
137 : CodeRegion(Desc, Start), I(std::move(I)) {}
138
139public:
140 Instrument *getInstrument() const { return I.get(); }
141};
142
143class CodeRegionParseError final : public Error {};
144
145class CodeRegions {
146 CodeRegions(const CodeRegions &) = delete;
147 CodeRegions &operator=(const CodeRegions &) = delete;
148
149protected:
150 // A source manager. Used by the tool to generate meaningful warnings.
151 llvm::SourceMgr &SM;
152
153 using UniqueCodeRegion = std::unique_ptr<CodeRegion>;
154 std::vector<UniqueCodeRegion> Regions;
155 llvm::StringMap<unsigned> ActiveRegions;
156 bool FoundErrors;
157
158public:
159 CodeRegions(llvm::SourceMgr &S) : SM(S), FoundErrors(false) {}
160 virtual ~CodeRegions() = default;
161
162 typedef std::vector<UniqueCodeRegion>::iterator iterator;
163 typedef std::vector<UniqueCodeRegion>::const_iterator const_iterator;
164
165 iterator begin() { return Regions.begin(); }
166 iterator end() { return Regions.end(); }
167 const_iterator begin() const { return Regions.cbegin(); }
168 const_iterator end() const { return Regions.cend(); }
169
170 void addInstruction(const llvm::MCInst &Instruction);
171 llvm::SourceMgr &getSourceMgr() const { return SM; }
172
173 llvm::ArrayRef<llvm::MCInst> getInstructionSequence(unsigned Idx) const {
174 return Regions[Idx]->getInstructions();
175 }
176
177 bool empty() const {
178 return llvm::all_of(Range: Regions, P: [](const UniqueCodeRegion &Region) {
179 return Region->empty();
180 });
181 }
182
183 bool isValid() const { return !FoundErrors; }
184
185 bool isRegionActive(llvm::StringRef Description) const {
186 return ActiveRegions.contains(Key: Description);
187 }
188
189 virtual void beginRegion(llvm::StringRef Description, llvm::SMLoc Loc) = 0;
190 virtual void beginRegion(llvm::StringRef Description, llvm::SMLoc Loc,
191 UniqueInstrument Instrument) = 0;
192 virtual void endRegion(llvm::StringRef Description, llvm::SMLoc Loc) = 0;
193};
194
195struct AnalysisRegions : public CodeRegions {
196 AnalysisRegions(llvm::SourceMgr &S);
197
198 void beginRegion(llvm::StringRef Description, llvm::SMLoc Loc) override;
199 void beginRegion(llvm::StringRef Description, llvm::SMLoc Loc,
200 UniqueInstrument Instrument) override {}
201 void endRegion(llvm::StringRef Description, llvm::SMLoc Loc) override;
202};
203
204struct InstrumentRegions : public CodeRegions {
205
206 InstrumentRegions(llvm::SourceMgr &S);
207
208 void beginRegion(llvm::StringRef Description, llvm::SMLoc Loc) override{};
209 void beginRegion(llvm::StringRef Description, llvm::SMLoc Loc,
210 UniqueInstrument Instrument) override;
211 void endRegion(llvm::StringRef Description, llvm::SMLoc Loc) override;
212
213 const SmallVector<Instrument *> getActiveInstruments(llvm::SMLoc Loc) const;
214};
215
216} // namespace mca
217} // namespace llvm
218
219#endif
220