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 
27 using namespace llvm;
28 
29 /// \brief Grabs the string in whichever attribute is passed in and returns
30 /// a reference to it.
31 static 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.
50 void 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.
60 void 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..
72 void 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 // Collect all of the attributes for a particular DIE in single structure.
106 void DIEHash::collectAttributes(DIE *Die, DIEAttrs &Attrs) {
107   const SmallVectorImpl<DIEValue *> &Values = Die->getValues();
108   const DIEAbbrev &Abbrevs = Die->getAbbrev();
109 
110 #define COLLECT_ATTR(NAME)                                                     \
111   Attrs.NAME.Val = Values[i];                                                  \
112   Attrs.NAME.Desc = &Abbrevs.getData()[i];
113 
114   for (size_t i = 0, e = Values.size(); i != e; ++i) {
115     DEBUG(dbgs() << "Attribute: "
116                  << dwarf::AttributeString(Abbrevs.getData()[i].getAttribute())
117                  << " added.\n");
118     switch (Abbrevs.getData()[i].getAttribute()) {
119     case dwarf::DW_AT_name:
120       COLLECT_ATTR(DW_AT_name);
121       break;
122     default:
123       break;
124     }
125   }
126 }
127 
128 // Hash an individual attribute \param Attr based on the type of attribute and
129 // the form.
130 void DIEHash::hashAttribute(AttrEntry Attr) {
131   const DIEValue *Value = Attr.Val;
132   const DIEAbbrevData *Desc = Attr.Desc;
133 
134   // TODO: Add support for types.
135 
136   // Add the letter A to the hash.
137   addULEB128('A');
138 
139   // Then the attribute code and form.
140   addULEB128(Desc->getAttribute());
141   addULEB128(Desc->getForm());
142 
143   // TODO: Add support for additional forms.
144   switch (Desc->getForm()) {
145   case dwarf::DW_FORM_strp:
146     addString(cast<DIEString>(Value)->getString());
147     break;
148   }
149 }
150 
151 // Go through the attributes from \param Attrs in the order specified in 7.27.4
152 // and hash them.
153 void DIEHash::hashAttributes(const DIEAttrs &Attrs) {
154 #define ADD_ATTR(ATTR)                                                         \
155   {                                                                            \
156     if (ATTR.Val != 0)                                                         \
157       hashAttribute(ATTR);                                                     \
158   }
159 
160   // FIXME: Add the rest.
161   ADD_ATTR(Attrs.DW_AT_name);
162 }
163 
164 // Add all of the attributes for \param Die to the hash.
165 void DIEHash::addAttributes(DIE *Die) {
166   DIEAttrs Attrs;
167   memset(&Attrs, 0, sizeof(Attrs));
168   collectAttributes(Die, Attrs);
169   hashAttributes(Attrs);
170 }
171 
172 // Compute the hash of a DIE. This is based on the type signature computation
173 // given in section 7.27 of the DWARF4 standard. It is the md5 hash of a
174 // flattened description of the DIE.
175 void DIEHash::computeHash(DIE *Die) {
176 
177   // Append the letter 'D', followed by the DWARF tag of the DIE.
178   addULEB128('D');
179   addULEB128(Die->getTag());
180 
181   // Add each of the attributes of the DIE.
182   addAttributes(Die);
183 
184   // Then hash each of the children of the DIE.
185   for (std::vector<DIE *>::const_iterator I = Die->getChildren().begin(),
186                                           E = Die->getChildren().end();
187        I != E; ++I)
188     computeHash(*I);
189 }
190 
191 /// This is based on the type signature computation given in section 7.27 of the
192 /// DWARF4 standard. It is the md5 hash of a flattened description of the DIE
193 /// with the exception that we are hashing only the context and the name of the
194 /// type.
195 uint64_t DIEHash::computeDIEODRSignature(DIE *Die) {
196 
197   // Add the contexts to the hash. We won't be computing the ODR hash for
198   // function local types so it's safe to use the generic context hashing
199   // algorithm here.
200   // FIXME: If we figure out how to account for linkage in some way we could
201   // actually do this with a slight modification to the parent hash algorithm.
202   DIE *Parent = Die->getParent();
203   if (Parent)
204     addParentContext(Parent);
205 
206   // Add the current DIE information.
207 
208   // Add the DWARF tag of the DIE.
209   addULEB128(Die->getTag());
210 
211   // Add the name of the type to the hash.
212   addString(getDIEStringAttr(Die, dwarf::DW_AT_name));
213 
214   // Now get the result.
215   MD5::MD5Result Result;
216   Hash.final(Result);
217 
218   // ... take the least significant 8 bytes and return those. Our MD5
219   // implementation always returns its results in little endian, swap bytes
220   // appropriately.
221   return *reinterpret_cast<support::ulittle64_t *>(Result + 8);
222 }
223 
224 /// This is based on the type signature computation given in section 7.27 of the
225 /// DWARF4 standard. It is an md5 hash of the flattened description of the DIE
226 /// with the inclusion of the full CU and all top level CU entities.
227 uint64_t DIEHash::computeCUSignature(DIE *Die) {
228 
229   // Hash the DIE.
230   computeHash(Die);
231 
232   // Now return the result.
233   MD5::MD5Result Result;
234   Hash.final(Result);
235 
236   // ... take the least significant 8 bytes and return those. Our MD5
237   // implementation always returns its results in little endian, swap bytes
238   // appropriately.
239   return *reinterpret_cast<support::ulittle64_t *>(Result + 8);
240 }
241