xref: /openbsd-src/gnu/llvm/llvm/include/llvm/ProfileData/RawMemProfReader.h (revision d415bd752c734aee168c4ee86ff32e8cc249eb16)
1 #ifndef LLVM_PROFILEDATA_RAWMEMPROFREADER_H_
2 #define LLVM_PROFILEDATA_RAWMEMPROFREADER_H_
3 //===- MemProfReader.h - Instrumented memory profiling reader ---*- C++ -*-===//
4 //
5 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
6 // See https://llvm.org/LICENSE.txt for license information.
7 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
8 //
9 //===----------------------------------------------------------------------===//
10 //
11 // This file contains support for reading MemProf profiling data.
12 //
13 //===----------------------------------------------------------------------===//
14 
15 #include "llvm/ADT/DenseMap.h"
16 #include "llvm/ADT/MapVector.h"
17 #include "llvm/ADT/StringRef.h"
18 #include "llvm/DebugInfo/Symbolize/SymbolizableModule.h"
19 #include "llvm/DebugInfo/Symbolize/Symbolize.h"
20 #include "llvm/IR/GlobalValue.h"
21 #include "llvm/Object/Binary.h"
22 #include "llvm/Object/ObjectFile.h"
23 #include "llvm/ProfileData/InstrProfReader.h"
24 #include "llvm/ProfileData/MemProf.h"
25 #include "llvm/ProfileData/MemProfData.inc"
26 #include "llvm/Support/Error.h"
27 #include "llvm/Support/MemoryBuffer.h"
28 
29 #include <cstddef>
30 
31 namespace llvm {
32 namespace memprof {
33 
34 // Map from id (recorded from sanitizer stack depot) to virtual addresses for
35 // each program counter address in the callstack.
36 using CallStackMap = llvm::DenseMap<uint64_t, llvm::SmallVector<uint64_t>>;
37 
38 class RawMemProfReader {
39 public:
40   RawMemProfReader(const RawMemProfReader &) = delete;
41   RawMemProfReader &operator=(const RawMemProfReader &) = delete;
42 
43   // Prints the contents of the profile in YAML format.
44   void printYAML(raw_ostream &OS);
45 
46   // Return true if the \p DataBuffer starts with magic bytes indicating it is
47   // a raw binary memprof profile.
48   static bool hasFormat(const MemoryBuffer &DataBuffer);
49   // Return true if the file at \p Path starts with magic bytes indicating it is
50   // a raw binary memprof profile.
51   static bool hasFormat(const StringRef Path);
52 
53   // Create a RawMemProfReader after sanity checking the contents of the file at
54   // \p Path. The binary from which the profile has been collected is specified
55   // via a path in \p ProfiledBinary.
56   static Expected<std::unique_ptr<RawMemProfReader>>
57   create(const Twine &Path, const StringRef ProfiledBinary,
58          bool KeepName = false);
59 
60   using GuidMemProfRecordPair = std::pair<GlobalValue::GUID, MemProfRecord>;
61   using Iterator = InstrProfIterator<GuidMemProfRecordPair, RawMemProfReader>;
end()62   Iterator end() { return Iterator(); }
begin()63   Iterator begin() {
64     Iter = FunctionProfileData.begin();
65     return Iterator(this);
66   }
67 
68   Error readNextRecord(GuidMemProfRecordPair &GuidRecord);
69 
70   // The RawMemProfReader only holds memory profile information.
getProfileKind()71   InstrProfKind getProfileKind() const { return InstrProfKind::MemProf; }
72 
73   // Constructor for unittests only.
74   RawMemProfReader(std::unique_ptr<llvm::symbolize::SymbolizableModule> Sym,
75                    llvm::SmallVectorImpl<SegmentEntry> &Seg,
76                    llvm::MapVector<uint64_t, MemInfoBlock> &Prof,
77                    CallStackMap &SM, bool KeepName = false)
Symbolizer(std::move (Sym))78       : Symbolizer(std::move(Sym)), SegmentInfo(Seg.begin(), Seg.end()),
79         CallstackProfileData(Prof), StackMap(SM), KeepSymbolName(KeepName) {
80     // We don't call initialize here since there is no raw profile to read. The
81     // test should pass in the raw profile as structured data.
82 
83     // If there is an error here then the mock symbolizer has not been
84     // initialized properly.
85     if (Error E = symbolizeAndFilterStackFrames())
86       report_fatal_error(std::move(E));
87     if (Error E = mapRawProfileToRecords())
88       report_fatal_error(std::move(E));
89   }
90 
91   // Return a const reference to the internal Id to Frame mappings.
getFrameMapping()92   const llvm::DenseMap<FrameId, Frame> &getFrameMapping() const {
93     return IdToFrame;
94   }
95 
96   // Return a const reference to the internal function profile data.
97   const llvm::MapVector<GlobalValue::GUID, IndexedMemProfRecord> &
getProfileData()98   getProfileData() const {
99     return FunctionProfileData;
100   }
101 
102 private:
RawMemProfReader(object::OwningBinary<object::Binary> && Bin,bool KeepName)103   RawMemProfReader(object::OwningBinary<object::Binary> &&Bin, bool KeepName)
104       : Binary(std::move(Bin)), KeepSymbolName(KeepName) {}
105   // Initializes the RawMemProfReader with the contents in `DataBuffer`.
106   Error initialize(std::unique_ptr<MemoryBuffer> DataBuffer);
107   // Read and parse the contents of the `DataBuffer` as a binary format profile.
108   Error readRawProfile(std::unique_ptr<MemoryBuffer> DataBuffer);
109   // Symbolize and cache all the virtual addresses we encounter in the
110   // callstacks from the raw profile. Also prune callstack frames which we can't
111   // symbolize or those that belong to the runtime. For profile entries where
112   // the entire callstack is pruned, we drop the entry from the profile.
113   Error symbolizeAndFilterStackFrames();
114   // Construct memprof records for each function and store it in the
115   // `FunctionProfileData` map. A function may have allocation profile data or
116   // callsite data or both.
117   Error mapRawProfileToRecords();
118 
119   // A helper method to extract the frame from the IdToFrame map.
idToFrame(const FrameId Id)120   const Frame &idToFrame(const FrameId Id) const {
121     auto It = IdToFrame.find(Id);
122     assert(It != IdToFrame.end() && "Id not found in map.");
123     return It->getSecond();
124   }
125 
126   object::SectionedAddress getModuleOffset(uint64_t VirtualAddress);
127 
128   object::OwningBinary<object::Binary> Binary;
129   std::unique_ptr<llvm::symbolize::SymbolizableModule> Symbolizer;
130 
131   // The contents of the raw profile.
132   llvm::SmallVector<SegmentEntry, 16> SegmentInfo;
133   // A map from callstack id (same as key in CallStackMap below) to the heap
134   // information recorded for that allocation context.
135   llvm::MapVector<uint64_t, MemInfoBlock> CallstackProfileData;
136   CallStackMap StackMap;
137 
138   // Cached symbolization from PC to Frame.
139   llvm::DenseMap<uint64_t, llvm::SmallVector<FrameId>> SymbolizedFrame;
140   llvm::DenseMap<FrameId, Frame> IdToFrame;
141 
142   llvm::MapVector<GlobalValue::GUID, IndexedMemProfRecord> FunctionProfileData;
143   llvm::MapVector<GlobalValue::GUID, IndexedMemProfRecord>::iterator Iter;
144 
145   // Whether to keep the symbol name for each frame after hashing.
146   bool KeepSymbolName = false;
147   // A mapping of the hash to symbol name, only used if KeepSymbolName is true.
148   llvm::DenseMap<uint64_t, std::string> GuidToSymbolName;
149 };
150 } // namespace memprof
151 } // namespace llvm
152 
153 #endif // LLVM_PROFILEDATA_RAWMEMPROFREADER_H_
154