LLVM 20.0.0git
GlobalsStream.cpp
Go to the documentation of this file.
1//===- GlobalsStream.cpp - PDB Index of Symbols by Name ---------*- 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//
9// The on-disk structores used in this file are based on the reference
10// implementation which is available at
11// https://github.com/Microsoft/microsoft-pdb/blob/master/PDB/dbi/gsi.h
12//
13// When you are reading the reference source code, you'd find the
14// information below useful.
15//
16// - ppdb1->m_fMinimalDbgInfo seems to be always true.
17// - SMALLBUCKETS macro is defined.
18//
19//===----------------------------------------------------------------------===//
20
22
29#include "llvm/Support/Error.h"
30#include <algorithm>
31
32using namespace llvm;
33using namespace llvm::msf;
34using namespace llvm::pdb;
35
36GlobalsStream::GlobalsStream(std::unique_ptr<MappedBlockStream> Stream)
37 : Stream(std::move(Stream)) {}
38
40
42 BinaryStreamReader Reader(*Stream);
43 if (auto E = GlobalsTable.read(Reader))
44 return E;
45 return Error::success();
46}
47
48std::vector<std::pair<uint32_t, codeview::CVSymbol>>
50 const SymbolStream &Symbols) const {
51 std::vector<std::pair<uint32_t, codeview::CVSymbol>> Result;
52
53 // Hash the name to figure out which bucket this goes into.
54 size_t ExpandedBucketIndex = hashStringV1(Name) % IPHR_HASH;
55 int32_t CompressedBucketIndex = GlobalsTable.BucketMap[ExpandedBucketIndex];
56 if (CompressedBucketIndex == -1)
57 return Result;
58
59 uint32_t LastBucketIndex = GlobalsTable.HashBuckets.size() - 1;
60 uint32_t StartRecordIndex =
61 GlobalsTable.HashBuckets[CompressedBucketIndex] / 12;
62 uint32_t EndRecordIndex = 0;
63 if (LLVM_LIKELY(uint32_t(CompressedBucketIndex) < LastBucketIndex)) {
64 EndRecordIndex = GlobalsTable.HashBuckets[CompressedBucketIndex + 1];
65 } else {
66 // If this is the last bucket, it consists of all hash records until the end
67 // of the HashRecords array.
68 EndRecordIndex = GlobalsTable.HashRecords.size() * 12;
69 }
70
71 EndRecordIndex /= 12;
72
73 assert(EndRecordIndex <= GlobalsTable.HashRecords.size());
74 while (StartRecordIndex < EndRecordIndex) {
75 PSHashRecord PSH = GlobalsTable.HashRecords[StartRecordIndex];
76 uint32_t Off = PSH.Off - 1;
77 codeview::CVSymbol Record = Symbols.readRecord(Off);
79 Result.push_back(std::make_pair(Off, std::move(Record)));
80 ++StartRecordIndex;
81 }
82 return Result;
83}
84
85static Error checkHashHdrVersion(const GSIHashHeader *HashHdr) {
86 if (HashHdr->VerHdr != GSIHashHeader::HdrVersion)
87 return make_error<RawError>(
89 "Encountered unsupported globals stream version.");
90
91 return Error::success();
92}
93
94static Error readGSIHashHeader(const GSIHashHeader *&HashHdr,
95 BinaryStreamReader &Reader) {
96 if (Reader.readObject(HashHdr))
97 return make_error<RawError>(raw_error_code::corrupt_file,
98 "Stream does not contain a GSIHashHeader.");
99
101 return make_error<RawError>(
103 "GSIHashHeader signature (0xffffffff) not found.");
104
105 return Error::success();
106}
107
109 const GSIHashHeader *HashHdr,
110 BinaryStreamReader &Reader) {
111 if (auto EC = checkHashHdrVersion(HashHdr))
112 return EC;
113
114 // HashHdr->HrSize specifies the number of bytes of PSHashRecords we have.
115 // Verify that we can read them all.
116 if (HashHdr->HrSize % sizeof(PSHashRecord))
117 return make_error<RawError>(raw_error_code::corrupt_file,
118 "Invalid HR array size.");
119 uint32_t NumHashRecords = HashHdr->HrSize / sizeof(PSHashRecord);
120 if (auto EC = Reader.readArray(HashRecords, NumHashRecords))
121 return joinErrors(std::move(EC),
122 make_error<RawError>(raw_error_code::corrupt_file,
123 "Error reading hash records."));
124
125 return Error::success();
126}
127
128static Error
131 const GSIHashHeader *HashHdr,
132 MutableArrayRef<int32_t> BucketMap,
133 BinaryStreamReader &Reader) {
134 if (auto EC = checkHashHdrVersion(HashHdr))
135 return EC;
136
137 // Before the actual hash buckets, there is a bitmap of length determined by
138 // IPHR_HASH.
139 size_t BitmapSizeInBits = alignTo(IPHR_HASH + 1, 32);
140 uint32_t NumBitmapEntries = BitmapSizeInBits / 32;
141 if (auto EC = Reader.readArray(HashBitmap, NumBitmapEntries))
142 return joinErrors(std::move(EC),
143 make_error<RawError>(raw_error_code::corrupt_file,
144 "Could not read a bitmap."));
145 uint32_t CompressedBucketIdx = 0;
146 for (uint32_t I = 0; I <= IPHR_HASH; ++I) {
147 uint8_t WordIdx = I / 32;
148 uint8_t BitIdx = I % 32;
149 bool IsSet = HashBitmap[WordIdx] & (1U << BitIdx);
150 if (IsSet) {
151 BucketMap[I] = CompressedBucketIdx++;
152 } else {
153 BucketMap[I] = -1;
154 }
155 }
156
157 uint32_t NumBuckets = 0;
158 for (uint32_t B : HashBitmap)
159 NumBuckets += llvm::popcount(B);
160
161 // Hash buckets follow.
162 if (auto EC = Reader.readArray(HashBuckets, NumBuckets))
163 return joinErrors(std::move(EC),
164 make_error<RawError>(raw_error_code::corrupt_file,
165 "Hash buckets corrupted."));
166
167 return Error::success();
168}
169
171 if (auto EC = readGSIHashHeader(HashHdr, Reader))
172 return EC;
173 if (auto EC = readGSIHashRecords(HashRecords, HashHdr, Reader))
174 return EC;
175 if (HashHdr->HrSize > 0)
177 BucketMap, Reader))
178 return EC;
179 return Error::success();
180}
static GCRegistry::Add< OcamlGC > B("ocaml", "ocaml 3.10-compatible GC")
#define LLVM_LIKELY(EXPR)
Definition: Compiler.h:240
std::string Name
static Error readGSIHashHeader(const GSIHashHeader *&HashHdr, BinaryStreamReader &Reader)
static Error readGSIHashBuckets(FixedStreamArray< support::ulittle32_t > &HashBuckets, FixedStreamArray< support::ulittle32_t > &HashBitmap, const GSIHashHeader *HashHdr, MutableArrayRef< int32_t > BucketMap, BinaryStreamReader &Reader)
static Error readGSIHashRecords(FixedStreamArray< PSHashRecord > &HashRecords, const GSIHashHeader *HashHdr, BinaryStreamReader &Reader)
static Error checkHashHdrVersion(const GSIHashHeader *HashHdr)
#define I(x, y, z)
Definition: MD5.cpp:58
assert(ImpDefSCC.getReg()==AMDGPU::SCC &&ImpDefSCC.isDef())
Provides read only access to a subclass of BinaryStream.
Error readObject(const T *&Dest)
Get a pointer to an object of type T from the underlying stream, as if by memcpy, and store the resul...
Error readArray(ArrayRef< T > &Array, uint32_t NumElements)
Get a reference to a NumElements element array of objects of type T from the underlying stream as if ...
Lightweight error class with error context and mandatory checking.
Definition: Error.h:160
static ErrorSuccess success()
Create a success value.
Definition: Error.h:337
FixedStreamArray is similar to VarStreamArray, except with each record having a fixed-length.
MutableArrayRef - Represent a mutable reference to an array (0 or more elements consecutively in memo...
Definition: ArrayRef.h:307
StringRef - Represent a constant reference to a string, i.e.
Definition: StringRef.h:50
CVRecord is a fat pointer (base + size pair) to a symbol or type record.
Definition: CVRecord.h:29
const GSIHashHeader * HashHdr
Definition: GlobalsStream.h:52
Error read(BinaryStreamReader &Reader)
std::array< int32_t, IPHR_HASH+1 > BucketMap
Definition: GlobalsStream.h:56
FixedStreamArray< support::ulittle32_t > HashBuckets
Definition: GlobalsStream.h:55
FixedStreamArray< support::ulittle32_t > HashBitmap
Definition: GlobalsStream.h:54
FixedStreamArray< PSHashRecord > HashRecords
Definition: GlobalsStream.h:53
std::vector< std::pair< uint32_t, codeview::CVSymbol > > findRecordsByName(StringRef Name, const SymbolStream &Symbols) const
GlobalsStream(std::unique_ptr< msf::MappedBlockStream > Stream)
StringRef getSymbolName(CVSymbol Sym)
Definition: RecordName.cpp:323
uint32_t hashStringV1(StringRef Str)
Definition: Hash.cpp:20
This is an optimization pass for GlobalISel generic memory operations.
Definition: AddressRanges.h:18
int popcount(T Value) noexcept
Count the number of set bits in a value.
Definition: bit.h:385
Error joinErrors(Error E1, Error E2)
Concatenate errors.
Definition: Error.h:438
uint64_t alignTo(uint64_t Size, Align A)
Returns a multiple of A needed to store Size bytes.
Definition: Alignment.h:155
OutputIt move(R &&Range, OutputIt Out)
Provide wrappers to std::move which take ranges instead of having to pass begin/end explicitly.
Definition: STLExtras.h:1849
Implement std::hash so that hash_code can be used in STL containers.
Definition: BitVector.h:858
Header of the hash tables found in the globals and publics sections.
Definition: RawTypes.h:28
support::ulittle32_t HrSize
Definition: RawTypes.h:35
support::ulittle32_t VerHdr
Definition: RawTypes.h:34
support::ulittle32_t VerSignature
Definition: RawTypes.h:33
support::ulittle32_t Off
Definition: RawTypes.h:41