DIEHash.cpp revision dd0cd3c4f468b55651bf33240667881bbed8d04c 
1//===-- llvm/CodeGen/DIEHash.cpp - Dwarf Hashing Framework ----------------===//
2//
3//                     The LLVM Compiler Infrastructure
4//
5// This file is distributed under the University of Illinois Open Source
6// License. See LICENSE.TXT for details.
7//
8//===----------------------------------------------------------------------===//
9//
10// This file contains support for DWARF4 hashing of DIEs.
11//
12//===----------------------------------------------------------------------===//
13
14#define DEBUG_TYPE "dwarfdebug"
15
16#include "DIE.h"
17#include "DIEHash.h"
18#include "DwarfCompileUnit.h"
19#include "llvm/ADT/ArrayRef.h"
20#include "llvm/ADT/StringRef.h"
21#include "llvm/Support/Debug.h"
22#include "llvm/Support/Dwarf.h"
23#include "llvm/Support/Endian.h"
24#include "llvm/Support/MD5.h"
25#include "llvm/Support/raw_ostream.h"
26
27using namespace llvm;
28
29/// \brief Grabs the string in whichever attribute is passed in and returns
30/// a reference to it.
31static StringRef getDIEStringAttr(DIE *Die, uint16_t Attr) {
32  const SmallVectorImpl<DIEValue *> &Values = Die->getValues();
33  const DIEAbbrev &Abbrevs = Die->getAbbrev();
34
35  // Iterate through all the attributes until we find the one we're
36  // looking for, if we can't find it return an empty string.
37  for (size_t i = 0; i < Values.size(); ++i) {
38    if (Abbrevs.getData()[i].getAttribute() == Attr) {
39      DIEValue *V = Values[i];
40      assert(isa<DIEString>(V) && "String requested. Not a string.");
41      DIEString *S = cast<DIEString>(V);
42      return S->getString();
43    }
44  }
45  return StringRef("");
46}
47
48/// \brief Adds the string in \p Str to the hash. This also hashes
49/// a trailing NULL with the string.
50void DIEHash::addString(StringRef Str) {
51  DEBUG(dbgs() << "Adding string " << Str << " to hash.\n");
52  Hash.update(Str);
53  Hash.update(makeArrayRef((uint8_t)'\0'));
54}
55
56// FIXME: The LEB128 routines are copied and only slightly modified out of
57// LEB128.h.
58
59/// \brief Adds the unsigned in \p Value to the hash encoded as a ULEB128.
60void DIEHash::addULEB128(uint64_t Value) {
61  DEBUG(dbgs() << "Adding ULEB128 " << Value << " to hash.\n");
62  do {
63    uint8_t Byte = Value & 0x7f;
64    Value >>= 7;
65    if (Value != 0)
66      Byte |= 0x80; // Mark this byte to show that more bytes will follow.
67    Hash.update(Byte);
68  } while (Value != 0);
69}
70
71/// \brief Including \p Parent adds the context of Parent to the hash..
72void DIEHash::addParentContext(DIE *Parent) {
73
74  DEBUG(dbgs() << "Adding parent context to hash...\n");
75
76  // [7.27.2] For each surrounding type or namespace beginning with the
77  // outermost such construct...
78  SmallVector<DIE *, 1> Parents;
79  while (Parent->getTag() != dwarf::DW_TAG_compile_unit) {
80    Parents.push_back(Parent);
81    Parent = Parent->getParent();
82  }
83
84  // Reverse iterate over our list to go from the outermost construct to the
85  // innermost.
86  for (SmallVectorImpl<DIE *>::reverse_iterator I = Parents.rbegin(),
87                                                E = Parents.rend();
88       I != E; ++I) {
89    DIE *Die = *I;
90
91    // ... Append the letter "C" to the sequence...
92    addULEB128('C');
93
94    // ... Followed by the DWARF tag of the construct...
95    addULEB128(Die->getTag());
96
97    // ... Then the name, taken from the DW_AT_name attribute.
98    StringRef Name = getDIEStringAttr(Die, dwarf::DW_AT_name);
99    DEBUG(dbgs() << "... adding context: " << Name << "\n");
100    if (!Name.empty())
101      addString(Name);
102  }
103}
104
105/// This is based on the type signature computation given in section 7.27 of the
106/// DWARF4 standard. It is the md5 hash of a flattened description of the DIE
107/// with the exception that we are hashing only the context and the name of the
108/// type.
109uint64_t DIEHash::computeDIEODRSignature(DIE *Die) {
110
111  // Add the contexts to the hash. We won't be computing the ODR hash for
112  // function local types so it's safe to use the generic context hashing
113  // algorithm here.
114  // FIXME: If we figure out how to account for linkage in some way we could
115  // actually do this with a slight modification to the parent hash algorithm.
116  DIE *Parent = Die->getParent();
117  if (Parent)
118    addParentContext(Parent);
119
120  // Add the current DIE information.
121
122  // Add the DWARF tag of the DIE.
123  addULEB128(Die->getTag());
124
125  // Add the name of the type to the hash.
126  addString(getDIEStringAttr(Die, dwarf::DW_AT_name));
127
128  // Now get the result.
129  MD5::MD5Result Result;
130  Hash.final(Result);
131
132  // ... take the least significant 8 bytes and return those. Our MD5
133  // implementation always returns its results in little endian, swap bytes
134  // appropriately.
135  return *reinterpret_cast<support::ulittle64_t *>(Result + 8);
136}
137