DIEHash.cpp revision dce4a407a24b04eebc6a376f8e62b41aaa7b071f 
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#include "ByteStreamer.h"
15#include "DIEHash.h"
16#include "DIE.h"
17#include "DwarfDebug.h"
18#include "llvm/ADT/ArrayRef.h"
19#include "llvm/ADT/StringRef.h"
20#include "llvm/CodeGen/AsmPrinter.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#define DEBUG_TYPE "dwarfdebug"
30
31/// \brief Grabs the string in whichever attribute is passed in and returns
32/// a reference to it.
33static StringRef getDIEStringAttr(const DIE &Die, uint16_t Attr) {
34  const SmallVectorImpl<DIEValue *> &Values = Die.getValues();
35  const DIEAbbrev &Abbrevs = Die.getAbbrev();
36
37  // Iterate through all the attributes until we find the one we're
38  // looking for, if we can't find it return an empty string.
39  for (size_t i = 0; i < Values.size(); ++i) {
40    if (Abbrevs.getData()[i].getAttribute() == Attr) {
41      DIEValue *V = Values[i];
42      assert(isa<DIEString>(V) && "String requested. Not a string.");
43      DIEString *S = cast<DIEString>(V);
44      return S->getString();
45    }
46  }
47  return StringRef("");
48}
49
50/// \brief Adds the string in \p Str to the hash. This also hashes
51/// a trailing NULL with the string.
52void DIEHash::addString(StringRef Str) {
53  DEBUG(dbgs() << "Adding string " << Str << " to hash.\n");
54  Hash.update(Str);
55  Hash.update(makeArrayRef((uint8_t)'\0'));
56}
57
58// FIXME: The LEB128 routines are copied and only slightly modified out of
59// LEB128.h.
60
61/// \brief Adds the unsigned in \p Value to the hash encoded as a ULEB128.
62void DIEHash::addULEB128(uint64_t Value) {
63  DEBUG(dbgs() << "Adding ULEB128 " << Value << " to hash.\n");
64  do {
65    uint8_t Byte = Value & 0x7f;
66    Value >>= 7;
67    if (Value != 0)
68      Byte |= 0x80; // Mark this byte to show that more bytes will follow.
69    Hash.update(Byte);
70  } while (Value != 0);
71}
72
73void DIEHash::addSLEB128(int64_t Value) {
74  DEBUG(dbgs() << "Adding ULEB128 " << Value << " to hash.\n");
75  bool More;
76  do {
77    uint8_t Byte = Value & 0x7f;
78    Value >>= 7;
79    More = !((((Value == 0) && ((Byte & 0x40) == 0)) ||
80              ((Value == -1) && ((Byte & 0x40) != 0))));
81    if (More)
82      Byte |= 0x80; // Mark this byte to show that more bytes will follow.
83    Hash.update(Byte);
84  } while (More);
85}
86
87/// \brief Including \p Parent adds the context of Parent to the hash..
88void DIEHash::addParentContext(const DIE &Parent) {
89
90  DEBUG(dbgs() << "Adding parent context to hash...\n");
91
92  // [7.27.2] For each surrounding type or namespace beginning with the
93  // outermost such construct...
94  SmallVector<const DIE *, 1> Parents;
95  const DIE *Cur = &Parent;
96  while (Cur->getParent()) {
97    Parents.push_back(Cur);
98    Cur = Cur->getParent();
99  }
100  assert(Cur->getTag() == dwarf::DW_TAG_compile_unit ||
101         Cur->getTag() == dwarf::DW_TAG_type_unit);
102
103  // Reverse iterate over our list to go from the outermost construct to the
104  // innermost.
105  for (SmallVectorImpl<const DIE *>::reverse_iterator I = Parents.rbegin(),
106                                                      E = Parents.rend();
107       I != E; ++I) {
108    const DIE &Die = **I;
109
110    // ... Append the letter "C" to the sequence...
111    addULEB128('C');
112
113    // ... Followed by the DWARF tag of the construct...
114    addULEB128(Die.getTag());
115
116    // ... Then the name, taken from the DW_AT_name attribute.
117    StringRef Name = getDIEStringAttr(Die, dwarf::DW_AT_name);
118    DEBUG(dbgs() << "... adding context: " << Name << "\n");
119    if (!Name.empty())
120      addString(Name);
121  }
122}
123
124// Collect all of the attributes for a particular DIE in single structure.
125void DIEHash::collectAttributes(const DIE &Die, DIEAttrs &Attrs) {
126  const SmallVectorImpl<DIEValue *> &Values = Die.getValues();
127  const DIEAbbrev &Abbrevs = Die.getAbbrev();
128
129#define COLLECT_ATTR(NAME)                                                     \
130  case dwarf::NAME:                                                            \
131    Attrs.NAME.Val = Values[i];                                                \
132    Attrs.NAME.Desc = &Abbrevs.getData()[i];                                   \
133    break
134
135  for (size_t i = 0, e = Values.size(); i != e; ++i) {
136    DEBUG(dbgs() << "Attribute: "
137                 << dwarf::AttributeString(Abbrevs.getData()[i].getAttribute())
138                 << " added.\n");
139    switch (Abbrevs.getData()[i].getAttribute()) {
140      COLLECT_ATTR(DW_AT_name);
141      COLLECT_ATTR(DW_AT_accessibility);
142      COLLECT_ATTR(DW_AT_address_class);
143      COLLECT_ATTR(DW_AT_allocated);
144      COLLECT_ATTR(DW_AT_artificial);
145      COLLECT_ATTR(DW_AT_associated);
146      COLLECT_ATTR(DW_AT_binary_scale);
147      COLLECT_ATTR(DW_AT_bit_offset);
148      COLLECT_ATTR(DW_AT_bit_size);
149      COLLECT_ATTR(DW_AT_bit_stride);
150      COLLECT_ATTR(DW_AT_byte_size);
151      COLLECT_ATTR(DW_AT_byte_stride);
152      COLLECT_ATTR(DW_AT_const_expr);
153      COLLECT_ATTR(DW_AT_const_value);
154      COLLECT_ATTR(DW_AT_containing_type);
155      COLLECT_ATTR(DW_AT_count);
156      COLLECT_ATTR(DW_AT_data_bit_offset);
157      COLLECT_ATTR(DW_AT_data_location);
158      COLLECT_ATTR(DW_AT_data_member_location);
159      COLLECT_ATTR(DW_AT_decimal_scale);
160      COLLECT_ATTR(DW_AT_decimal_sign);
161      COLLECT_ATTR(DW_AT_default_value);
162      COLLECT_ATTR(DW_AT_digit_count);
163      COLLECT_ATTR(DW_AT_discr);
164      COLLECT_ATTR(DW_AT_discr_list);
165      COLLECT_ATTR(DW_AT_discr_value);
166      COLLECT_ATTR(DW_AT_encoding);
167      COLLECT_ATTR(DW_AT_enum_class);
168      COLLECT_ATTR(DW_AT_endianity);
169      COLLECT_ATTR(DW_AT_explicit);
170      COLLECT_ATTR(DW_AT_is_optional);
171      COLLECT_ATTR(DW_AT_location);
172      COLLECT_ATTR(DW_AT_lower_bound);
173      COLLECT_ATTR(DW_AT_mutable);
174      COLLECT_ATTR(DW_AT_ordering);
175      COLLECT_ATTR(DW_AT_picture_string);
176      COLLECT_ATTR(DW_AT_prototyped);
177      COLLECT_ATTR(DW_AT_small);
178      COLLECT_ATTR(DW_AT_segment);
179      COLLECT_ATTR(DW_AT_string_length);
180      COLLECT_ATTR(DW_AT_threads_scaled);
181      COLLECT_ATTR(DW_AT_upper_bound);
182      COLLECT_ATTR(DW_AT_use_location);
183      COLLECT_ATTR(DW_AT_use_UTF8);
184      COLLECT_ATTR(DW_AT_variable_parameter);
185      COLLECT_ATTR(DW_AT_virtuality);
186      COLLECT_ATTR(DW_AT_visibility);
187      COLLECT_ATTR(DW_AT_vtable_elem_location);
188      COLLECT_ATTR(DW_AT_type);
189    default:
190      break;
191    }
192  }
193}
194
195void DIEHash::hashShallowTypeReference(dwarf::Attribute Attribute,
196                                       const DIE &Entry, StringRef Name) {
197  // append the letter 'N'
198  addULEB128('N');
199
200  // the DWARF attribute code (DW_AT_type or DW_AT_friend),
201  addULEB128(Attribute);
202
203  // the context of the tag,
204  if (const DIE *Parent = Entry.getParent())
205    addParentContext(*Parent);
206
207  // the letter 'E',
208  addULEB128('E');
209
210  // and the name of the type.
211  addString(Name);
212
213  // Currently DW_TAG_friends are not used by Clang, but if they do become so,
214  // here's the relevant spec text to implement:
215  //
216  // For DW_TAG_friend, if the referenced entry is the DW_TAG_subprogram,
217  // the context is omitted and the name to be used is the ABI-specific name
218  // of the subprogram (e.g., the mangled linker name).
219}
220
221void DIEHash::hashRepeatedTypeReference(dwarf::Attribute Attribute,
222                                        unsigned DieNumber) {
223  // a) If T is in the list of [previously hashed types], use the letter
224  // 'R' as the marker
225  addULEB128('R');
226
227  addULEB128(Attribute);
228
229  // and use the unsigned LEB128 encoding of [the index of T in the
230  // list] as the attribute value;
231  addULEB128(DieNumber);
232}
233
234void DIEHash::hashDIEEntry(dwarf::Attribute Attribute, dwarf::Tag Tag,
235                           const DIE &Entry) {
236  assert(Tag != dwarf::DW_TAG_friend && "No current LLVM clients emit friend "
237                                        "tags. Add support here when there's "
238                                        "a use case");
239  // Step 5
240  // If the tag in Step 3 is one of [the below tags]
241  if ((Tag == dwarf::DW_TAG_pointer_type ||
242       Tag == dwarf::DW_TAG_reference_type ||
243       Tag == dwarf::DW_TAG_rvalue_reference_type ||
244       Tag == dwarf::DW_TAG_ptr_to_member_type) &&
245      // and the referenced type (via the [below attributes])
246      // FIXME: This seems overly restrictive, and causes hash mismatches
247      // there's a decl/def difference in the containing type of a
248      // ptr_to_member_type, but it's what DWARF says, for some reason.
249      Attribute == dwarf::DW_AT_type) {
250    // ... has a DW_AT_name attribute,
251    StringRef Name = getDIEStringAttr(Entry, dwarf::DW_AT_name);
252    if (!Name.empty()) {
253      hashShallowTypeReference(Attribute, Entry, Name);
254      return;
255    }
256  }
257
258  unsigned &DieNumber = Numbering[&Entry];
259  if (DieNumber) {
260    hashRepeatedTypeReference(Attribute, DieNumber);
261    return;
262  }
263
264  // otherwise, b) use the letter 'T' as a the marker, ...
265  addULEB128('T');
266
267  addULEB128(Attribute);
268
269  // ... process the type T recursively by performing Steps 2 through 7, and
270  // use the result as the attribute value.
271  DieNumber = Numbering.size();
272  computeHash(Entry);
273}
274
275// Hash all of the values in a block like set of values. This assumes that
276// all of the data is going to be added as integers.
277void DIEHash::hashBlockData(const SmallVectorImpl<DIEValue *> &Values) {
278  for (SmallVectorImpl<DIEValue *>::const_iterator I = Values.begin(),
279                                                   E = Values.end();
280       I != E; ++I)
281    Hash.update((uint64_t)cast<DIEInteger>(*I)->getValue());
282}
283
284// Hash the contents of a loclistptr class.
285void DIEHash::hashLocList(const DIELocList &LocList) {
286  HashingByteStreamer Streamer(*this);
287  DwarfDebug &DD = *AP->getDwarfDebug();
288  for (const auto &Entry :
289       DD.getDebugLocEntries()[LocList.getValue()].List)
290    DD.emitDebugLocEntry(Streamer, Entry);
291}
292
293// Hash an individual attribute \param Attr based on the type of attribute and
294// the form.
295void DIEHash::hashAttribute(AttrEntry Attr, dwarf::Tag Tag) {
296  const DIEValue *Value = Attr.Val;
297  const DIEAbbrevData *Desc = Attr.Desc;
298  dwarf::Attribute Attribute = Desc->getAttribute();
299
300  // Other attribute values use the letter 'A' as the marker, and the value
301  // consists of the form code (encoded as an unsigned LEB128 value) followed by
302  // the encoding of the value according to the form code. To ensure
303  // reproducibility of the signature, the set of forms used in the signature
304  // computation is limited to the following: DW_FORM_sdata, DW_FORM_flag,
305  // DW_FORM_string, and DW_FORM_block.
306
307  switch (Value->getType()) {
308    // 7.27 Step 3
309    // ... An attribute that refers to another type entry T is processed as
310    // follows:
311  case DIEValue::isEntry:
312    hashDIEEntry(Attribute, Tag, cast<DIEEntry>(Value)->getEntry());
313    break;
314  case DIEValue::isInteger: {
315    addULEB128('A');
316    addULEB128(Attribute);
317    switch (Desc->getForm()) {
318    case dwarf::DW_FORM_data1:
319    case dwarf::DW_FORM_data2:
320    case dwarf::DW_FORM_data4:
321    case dwarf::DW_FORM_data8:
322    case dwarf::DW_FORM_udata:
323    case dwarf::DW_FORM_sdata:
324      addULEB128(dwarf::DW_FORM_sdata);
325      addSLEB128((int64_t)cast<DIEInteger>(Value)->getValue());
326      break;
327    // DW_FORM_flag_present is just flag with a value of one. We still give it a
328    // value so just use the value.
329    case dwarf::DW_FORM_flag_present:
330    case dwarf::DW_FORM_flag:
331      addULEB128(dwarf::DW_FORM_flag);
332      addULEB128((int64_t)cast<DIEInteger>(Value)->getValue());
333      break;
334    default:
335      llvm_unreachable("Unknown integer form!");
336    }
337    break;
338  }
339  case DIEValue::isString:
340    addULEB128('A');
341    addULEB128(Attribute);
342    addULEB128(dwarf::DW_FORM_string);
343    addString(cast<DIEString>(Value)->getString());
344    break;
345  case DIEValue::isBlock:
346  case DIEValue::isLoc:
347  case DIEValue::isLocList:
348    addULEB128('A');
349    addULEB128(Attribute);
350    addULEB128(dwarf::DW_FORM_block);
351    if (isa<DIEBlock>(Value)) {
352      addULEB128(cast<DIEBlock>(Value)->ComputeSize(AP));
353      hashBlockData(cast<DIEBlock>(Value)->getValues());
354    } else if (isa<DIELoc>(Value)) {
355      addULEB128(cast<DIELoc>(Value)->ComputeSize(AP));
356      hashBlockData(cast<DIELoc>(Value)->getValues());
357    } else {
358      // We could add the block length, but that would take
359      // a bit of work and not add a lot of uniqueness
360      // to the hash in some way we could test.
361      hashLocList(*cast<DIELocList>(Value));
362    }
363    break;
364    // FIXME: It's uncertain whether or not we should handle this at the moment.
365  case DIEValue::isExpr:
366  case DIEValue::isLabel:
367  case DIEValue::isDelta:
368  case DIEValue::isTypeSignature:
369    llvm_unreachable("Add support for additional value types.");
370  }
371}
372
373// Go through the attributes from \param Attrs in the order specified in 7.27.4
374// and hash them.
375void DIEHash::hashAttributes(const DIEAttrs &Attrs, dwarf::Tag Tag) {
376#define ADD_ATTR(ATTR)                                                         \
377  {                                                                            \
378    if (ATTR.Val != 0)                                                         \
379      hashAttribute(ATTR, Tag);                                                \
380  }
381
382  ADD_ATTR(Attrs.DW_AT_name);
383  ADD_ATTR(Attrs.DW_AT_accessibility);
384  ADD_ATTR(Attrs.DW_AT_address_class);
385  ADD_ATTR(Attrs.DW_AT_allocated);
386  ADD_ATTR(Attrs.DW_AT_artificial);
387  ADD_ATTR(Attrs.DW_AT_associated);
388  ADD_ATTR(Attrs.DW_AT_binary_scale);
389  ADD_ATTR(Attrs.DW_AT_bit_offset);
390  ADD_ATTR(Attrs.DW_AT_bit_size);
391  ADD_ATTR(Attrs.DW_AT_bit_stride);
392  ADD_ATTR(Attrs.DW_AT_byte_size);
393  ADD_ATTR(Attrs.DW_AT_byte_stride);
394  ADD_ATTR(Attrs.DW_AT_const_expr);
395  ADD_ATTR(Attrs.DW_AT_const_value);
396  ADD_ATTR(Attrs.DW_AT_containing_type);
397  ADD_ATTR(Attrs.DW_AT_count);
398  ADD_ATTR(Attrs.DW_AT_data_bit_offset);
399  ADD_ATTR(Attrs.DW_AT_data_location);
400  ADD_ATTR(Attrs.DW_AT_data_member_location);
401  ADD_ATTR(Attrs.DW_AT_decimal_scale);
402  ADD_ATTR(Attrs.DW_AT_decimal_sign);
403  ADD_ATTR(Attrs.DW_AT_default_value);
404  ADD_ATTR(Attrs.DW_AT_digit_count);
405  ADD_ATTR(Attrs.DW_AT_discr);
406  ADD_ATTR(Attrs.DW_AT_discr_list);
407  ADD_ATTR(Attrs.DW_AT_discr_value);
408  ADD_ATTR(Attrs.DW_AT_encoding);
409  ADD_ATTR(Attrs.DW_AT_enum_class);
410  ADD_ATTR(Attrs.DW_AT_endianity);
411  ADD_ATTR(Attrs.DW_AT_explicit);
412  ADD_ATTR(Attrs.DW_AT_is_optional);
413  ADD_ATTR(Attrs.DW_AT_location);
414  ADD_ATTR(Attrs.DW_AT_lower_bound);
415  ADD_ATTR(Attrs.DW_AT_mutable);
416  ADD_ATTR(Attrs.DW_AT_ordering);
417  ADD_ATTR(Attrs.DW_AT_picture_string);
418  ADD_ATTR(Attrs.DW_AT_prototyped);
419  ADD_ATTR(Attrs.DW_AT_small);
420  ADD_ATTR(Attrs.DW_AT_segment);
421  ADD_ATTR(Attrs.DW_AT_string_length);
422  ADD_ATTR(Attrs.DW_AT_threads_scaled);
423  ADD_ATTR(Attrs.DW_AT_upper_bound);
424  ADD_ATTR(Attrs.DW_AT_use_location);
425  ADD_ATTR(Attrs.DW_AT_use_UTF8);
426  ADD_ATTR(Attrs.DW_AT_variable_parameter);
427  ADD_ATTR(Attrs.DW_AT_virtuality);
428  ADD_ATTR(Attrs.DW_AT_visibility);
429  ADD_ATTR(Attrs.DW_AT_vtable_elem_location);
430  ADD_ATTR(Attrs.DW_AT_type);
431
432  // FIXME: Add the extended attributes.
433}
434
435// Add all of the attributes for \param Die to the hash.
436void DIEHash::addAttributes(const DIE &Die) {
437  DIEAttrs Attrs = {};
438  collectAttributes(Die, Attrs);
439  hashAttributes(Attrs, Die.getTag());
440}
441
442void DIEHash::hashNestedType(const DIE &Die, StringRef Name) {
443  // 7.27 Step 7
444  // ... append the letter 'S',
445  addULEB128('S');
446
447  // the tag of C,
448  addULEB128(Die.getTag());
449
450  // and the name.
451  addString(Name);
452}
453
454// Compute the hash of a DIE. This is based on the type signature computation
455// given in section 7.27 of the DWARF4 standard. It is the md5 hash of a
456// flattened description of the DIE.
457void DIEHash::computeHash(const DIE &Die) {
458  // Append the letter 'D', followed by the DWARF tag of the DIE.
459  addULEB128('D');
460  addULEB128(Die.getTag());
461
462  // Add each of the attributes of the DIE.
463  addAttributes(Die);
464
465  // Then hash each of the children of the DIE.
466  for (auto &C : Die.getChildren()) {
467    // 7.27 Step 7
468    // If C is a nested type entry or a member function entry, ...
469    if (isType(C->getTag()) || C->getTag() == dwarf::DW_TAG_subprogram) {
470      StringRef Name = getDIEStringAttr(*C, dwarf::DW_AT_name);
471      // ... and has a DW_AT_name attribute
472      if (!Name.empty()) {
473        hashNestedType(*C, Name);
474        continue;
475      }
476    }
477    computeHash(*C);
478  }
479
480  // Following the last (or if there are no children), append a zero byte.
481  Hash.update(makeArrayRef((uint8_t)'\0'));
482}
483
484/// This is based on the type signature computation given in section 7.27 of the
485/// DWARF4 standard. It is the md5 hash of a flattened description of the DIE
486/// with the exception that we are hashing only the context and the name of the
487/// type.
488uint64_t DIEHash::computeDIEODRSignature(const DIE &Die) {
489
490  // Add the contexts to the hash. We won't be computing the ODR hash for
491  // function local types so it's safe to use the generic context hashing
492  // algorithm here.
493  // FIXME: If we figure out how to account for linkage in some way we could
494  // actually do this with a slight modification to the parent hash algorithm.
495  if (const DIE *Parent = Die.getParent())
496    addParentContext(*Parent);
497
498  // Add the current DIE information.
499
500  // Add the DWARF tag of the DIE.
501  addULEB128(Die.getTag());
502
503  // Add the name of the type to the hash.
504  addString(getDIEStringAttr(Die, dwarf::DW_AT_name));
505
506  // Now get the result.
507  MD5::MD5Result Result;
508  Hash.final(Result);
509
510  // ... take the least significant 8 bytes and return those. Our MD5
511  // implementation always returns its results in little endian, swap bytes
512  // appropriately.
513  return *reinterpret_cast<support::ulittle64_t *>(Result + 8);
514}
515
516/// This is based on the type signature computation given in section 7.27 of the
517/// DWARF4 standard. It is an md5 hash of the flattened description of the DIE
518/// with the inclusion of the full CU and all top level CU entities.
519// TODO: Initialize the type chain at 0 instead of 1 for CU signatures.
520uint64_t DIEHash::computeCUSignature(const DIE &Die) {
521  Numbering.clear();
522  Numbering[&Die] = 1;
523
524  // Hash the DIE.
525  computeHash(Die);
526
527  // Now return the result.
528  MD5::MD5Result Result;
529  Hash.final(Result);
530
531  // ... take the least significant 8 bytes and return those. Our MD5
532  // implementation always returns its results in little endian, swap bytes
533  // appropriately.
534  return *reinterpret_cast<support::ulittle64_t *>(Result + 8);
535}
536
537/// This is based on the type signature computation given in section 7.27 of the
538/// DWARF4 standard. It is an md5 hash of the flattened description of the DIE
539/// with the inclusion of additional forms not specifically called out in the
540/// standard.
541uint64_t DIEHash::computeTypeSignature(const DIE &Die) {
542  Numbering.clear();
543  Numbering[&Die] = 1;
544
545  if (const DIE *Parent = Die.getParent())
546    addParentContext(*Parent);
547
548  // Hash the DIE.
549  computeHash(Die);
550
551  // Now return the result.
552  MD5::MD5Result Result;
553  Hash.final(Result);
554
555  // ... take the least significant 8 bytes and return those. Our MD5
556  // implementation always returns its results in little endian, swap bytes
557  // appropriately.
558  return *reinterpret_cast<support::ulittle64_t *>(Result + 8);
559}
560