1 //===--- HeaderMap.cpp - A file that acts like dir of symlinks ------------===//
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 implements the HeaderMap interface.
11 //
12 //===----------------------------------------------------------------------===//
13 
14 #include "clang/Lex/HeaderMap.h"
15 #include "clang/Lex/HeaderMapTypes.h"
16 #include "clang/Basic/CharInfo.h"
17 #include "clang/Basic/FileManager.h"
18 #include "llvm/ADT/SmallString.h"
19 #include "llvm/Support/DataTypes.h"
20 #include "llvm/Support/MathExtras.h"
21 #include "llvm/Support/MemoryBuffer.h"
22 #include "llvm/Support/SwapByteOrder.h"
23 #include <cstdio>
24 #include <memory>
25 using namespace clang;
26 
27 /// HashHMapKey - This is the 'well known' hash function required by the file
28 /// format, used to look up keys in the hash table.  The hash table uses simple
29 /// linear probing based on this function.
30 static inline unsigned HashHMapKey(StringRef Str) {
31   unsigned Result = 0;
32   const char *S = Str.begin(), *End = Str.end();
33 
34   for (; S != End; S++)
35     Result += toLowercase(*S) * 13;
36   return Result;
37 }
38 
39 
40 
41 //===----------------------------------------------------------------------===//
42 // Verification and Construction
43 //===----------------------------------------------------------------------===//
44 
45 /// HeaderMap::Create - This attempts to load the specified file as a header
46 /// map.  If it doesn't look like a HeaderMap, it gives up and returns null.
47 /// If it looks like a HeaderMap but is obviously corrupted, it puts a reason
48 /// into the string error argument and returns null.
49 const HeaderMap *HeaderMap::Create(const FileEntry *FE, FileManager &FM) {
50   // If the file is too small to be a header map, ignore it.
51   unsigned FileSize = FE->getSize();
52   if (FileSize <= sizeof(HMapHeader)) return nullptr;
53 
54   auto FileBuffer = FM.getBufferForFile(FE);
55   if (!FileBuffer || !*FileBuffer)
56     return nullptr;
57   bool NeedsByteSwap;
58   if (!checkHeader(**FileBuffer, NeedsByteSwap))
59     return nullptr;
60   return new HeaderMap(std::move(*FileBuffer), NeedsByteSwap);
61 }
62 
63 bool HeaderMapImpl::checkHeader(const llvm::MemoryBuffer &File,
64                                 bool &NeedsByteSwap) {
65   if (File.getBufferSize() <= sizeof(HMapHeader))
66     return false;
67   const char *FileStart = File.getBufferStart();
68 
69   // We know the file is at least as big as the header, check it now.
70   const HMapHeader *Header = reinterpret_cast<const HMapHeader*>(FileStart);
71 
72   // Sniff it to see if it's a headermap by checking the magic number and
73   // version.
74   if (Header->Magic == HMAP_HeaderMagicNumber &&
75       Header->Version == HMAP_HeaderVersion)
76     NeedsByteSwap = false;
77   else if (Header->Magic == llvm::ByteSwap_32(HMAP_HeaderMagicNumber) &&
78            Header->Version == llvm::ByteSwap_16(HMAP_HeaderVersion))
79     NeedsByteSwap = true;  // Mixed endianness headermap.
80   else
81     return false;  // Not a header map.
82 
83   if (Header->Reserved != 0)
84     return false;
85 
86   // Check the number of buckets.
87   auto NumBuckets = NeedsByteSwap
88                         ? llvm::sys::getSwappedBytes(Header->NumBuckets)
89                         : Header->NumBuckets;
90 
91   // If the number of buckets is not a power of two, the headermap is corrupt.
92   if (NumBuckets & (NumBuckets - 1))
93     return false;
94 
95   // Okay, everything looks good.
96   return true;
97 }
98 
99 //===----------------------------------------------------------------------===//
100 //  Utility Methods
101 //===----------------------------------------------------------------------===//
102 
103 
104 /// getFileName - Return the filename of the headermap.
105 const char *HeaderMapImpl::getFileName() const {
106   return FileBuffer->getBufferIdentifier();
107 }
108 
109 unsigned HeaderMapImpl::getEndianAdjustedWord(unsigned X) const {
110   if (!NeedsBSwap) return X;
111   return llvm::ByteSwap_32(X);
112 }
113 
114 /// getHeader - Return a reference to the file header, in unbyte-swapped form.
115 /// This method cannot fail.
116 const HMapHeader &HeaderMapImpl::getHeader() const {
117   // We know the file is at least as big as the header.  Return it.
118   return *reinterpret_cast<const HMapHeader*>(FileBuffer->getBufferStart());
119 }
120 
121 /// getBucket - Return the specified hash table bucket from the header map,
122 /// bswap'ing its fields as appropriate.  If the bucket number is not valid,
123 /// this return a bucket with an empty key (0).
124 HMapBucket HeaderMapImpl::getBucket(unsigned BucketNo) const {
125   HMapBucket Result;
126   Result.Key = HMAP_EmptyBucketKey;
127 
128   const HMapBucket *BucketArray =
129     reinterpret_cast<const HMapBucket*>(FileBuffer->getBufferStart() +
130                                         sizeof(HMapHeader));
131 
132   const HMapBucket *BucketPtr = BucketArray+BucketNo;
133   if ((const char*)(BucketPtr+1) > FileBuffer->getBufferEnd()) {
134     Result.Prefix = 0;
135     Result.Suffix = 0;
136     return Result;  // Invalid buffer, corrupt hmap.
137   }
138 
139   // Otherwise, the bucket is valid.  Load the values, bswapping as needed.
140   Result.Key    = getEndianAdjustedWord(BucketPtr->Key);
141   Result.Prefix = getEndianAdjustedWord(BucketPtr->Prefix);
142   Result.Suffix = getEndianAdjustedWord(BucketPtr->Suffix);
143   return Result;
144 }
145 
146 /// getString - Look up the specified string in the string table.  If the string
147 /// index is not valid, it returns an empty string.
148 const char *HeaderMapImpl::getString(unsigned StrTabIdx) const {
149   // Add the start of the string table to the idx.
150   StrTabIdx += getEndianAdjustedWord(getHeader().StringsOffset);
151 
152   // Check for invalid index.
153   if (StrTabIdx >= FileBuffer->getBufferSize())
154     return nullptr;
155 
156   // Otherwise, we have a valid pointer into the file.  Just return it.  We know
157   // that the "string" can not overrun the end of the file, because the buffer
158   // is nul terminated by virtue of being a MemoryBuffer.
159   return FileBuffer->getBufferStart()+StrTabIdx;
160 }
161 
162 //===----------------------------------------------------------------------===//
163 // The Main Drivers
164 //===----------------------------------------------------------------------===//
165 
166 /// dump - Print the contents of this headermap to stderr.
167 LLVM_DUMP_METHOD void HeaderMapImpl::dump() const {
168   const HMapHeader &Hdr = getHeader();
169   unsigned NumBuckets = getEndianAdjustedWord(Hdr.NumBuckets);
170 
171   fprintf(stderr, "Header Map %s:\n  %d buckets, %d entries\n",
172           getFileName(), NumBuckets,
173           getEndianAdjustedWord(Hdr.NumEntries));
174 
175   for (unsigned i = 0; i != NumBuckets; ++i) {
176     HMapBucket B = getBucket(i);
177     if (B.Key == HMAP_EmptyBucketKey) continue;
178 
179     const char *Key    = getString(B.Key);
180     const char *Prefix = getString(B.Prefix);
181     const char *Suffix = getString(B.Suffix);
182     fprintf(stderr, "  %d. %s -> '%s' '%s'\n", i, Key, Prefix, Suffix);
183   }
184 }
185 
186 /// LookupFile - Check to see if the specified relative filename is located in
187 /// this HeaderMap.  If so, open it and return its FileEntry.
188 const FileEntry *HeaderMap::LookupFile(
189     StringRef Filename, FileManager &FM) const {
190 
191   SmallString<1024> Path;
192   StringRef Dest = HeaderMapImpl::lookupFilename(Filename, Path);
193   if (Dest.empty())
194     return nullptr;
195 
196   return FM.getFile(Dest);
197 }
198 
199 StringRef HeaderMapImpl::lookupFilename(StringRef Filename,
200                                         SmallVectorImpl<char> &DestPath) const {
201   const HMapHeader &Hdr = getHeader();
202   unsigned NumBuckets = getEndianAdjustedWord(Hdr.NumBuckets);
203 
204   // Don't probe infinitely.  This should be checked before constructing.
205   assert(!(NumBuckets & (NumBuckets - 1)) && "Expected power of 2");
206 
207   // Linearly probe the hash table.
208   for (unsigned Bucket = HashHMapKey(Filename);; ++Bucket) {
209     HMapBucket B = getBucket(Bucket & (NumBuckets-1));
210     if (B.Key == HMAP_EmptyBucketKey) return StringRef(); // Hash miss.
211 
212     // See if the key matches.  If not, probe on.
213     if (!Filename.equals_lower(getString(B.Key)))
214       continue;
215 
216     // If so, we have a match in the hash table.  Construct the destination
217     // path.
218     StringRef Prefix = getString(B.Prefix);
219     StringRef Suffix = getString(B.Suffix);
220     DestPath.clear();
221     DestPath.append(Prefix.begin(), Prefix.end());
222     DestPath.append(Suffix.begin(), Suffix.end());
223     return StringRef(DestPath.begin(), DestPath.size());
224   }
225 }
226