1 //===-- SpecialCaseList.h - special case list for sanitizers ----*- 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 // This is a utility class used to parse user-provided text files with
9 // "special case lists" for code sanitizers. Such files are used to
10 // define an "ABI list" for DataFlowSanitizer and allow/exclusion lists for
11 // sanitizers like AddressSanitizer or UndefinedBehaviorSanitizer.
12 //
13 // Empty lines and lines starting with "#" are ignored. Sections are defined
14 // using a '[section_name]' header and can be used to specify sanitizers the
15 // entries below it apply to. Section names are regular expressions, and
16 // entries without a section header match all sections (e.g. an '[*]' header
17 // is assumed.)
18 // The remaining lines should have the form:
19 //   prefix:wildcard_expression[=category]
20 // If category is not specified, it is assumed to be empty string.
21 // Definitions of "prefix" and "category" are sanitizer-specific. For example,
22 // sanitizer exclusion support prefixes "src", "fun" and "global".
23 // Wildcard expressions define, respectively, source files, functions or
24 // globals which shouldn't be instrumented.
25 // Examples of categories:
26 //   "functional": used in DFSan to list functions with pure functional
27 //                 semantics.
28 //   "init": used in ASan exclusion list to disable initialization-order bugs
29 //           detection for certain globals or source files.
30 // Full special case list file example:
31 // ---
32 // [address]
33 // # Excluded items:
34 // fun:*_ZN4base6subtle*
35 // global:*global_with_bad_access_or_initialization*
36 // global:*global_with_initialization_issues*=init
37 // type:*Namespace::ClassName*=init
38 // src:file_with_tricky_code.cc
39 // src:ignore-global-initializers-issues.cc=init
40 //
41 // [dataflow]
42 // # Functions with pure functional semantics:
43 // fun:cos=functional
44 // fun:sin=functional
45 // ---
46 // Note that the wild card is in fact an llvm::Regex, but * is automatically
47 // replaced with .*
48 //
49 //===----------------------------------------------------------------------===//
50 
51 #ifndef LLVM_SUPPORT_SPECIALCASELIST_H
52 #define LLVM_SUPPORT_SPECIALCASELIST_H
53 
54 #include "llvm/ADT/StringMap.h"
55 #include "llvm/Support/Regex.h"
56 #include "llvm/Support/TrigramIndex.h"
57 #include <memory>
58 #include <string>
59 #include <vector>
60 
61 namespace llvm {
62 class MemoryBuffer;
63 class StringRef;
64 
65 namespace vfs {
66 class FileSystem;
67 }
68 
69 class SpecialCaseList {
70 public:
71   /// Parses the special case list entries from files. On failure, returns
72   /// 0 and writes an error message to string.
73   static std::unique_ptr<SpecialCaseList>
74   create(const std::vector<std::string> &Paths, llvm::vfs::FileSystem &FS,
75          std::string &Error);
76   /// Parses the special case list from a memory buffer. On failure, returns
77   /// 0 and writes an error message to string.
78   static std::unique_ptr<SpecialCaseList> create(const MemoryBuffer *MB,
79                                                  std::string &Error);
80   /// Parses the special case list entries from files. On failure, reports a
81   /// fatal error.
82   static std::unique_ptr<SpecialCaseList>
83   createOrDie(const std::vector<std::string> &Paths, llvm::vfs::FileSystem &FS);
84 
85   ~SpecialCaseList();
86 
87   /// Returns true, if special case list contains a line
88   /// \code
89   ///   @Prefix:<E>=@Category
90   /// \endcode
91   /// where @Query satisfies wildcard expression <E> in a given @Section.
92   bool inSection(StringRef Section, StringRef Prefix, StringRef Query,
93                  StringRef Category = StringRef()) const;
94 
95   /// Returns the line number corresponding to the special case list entry if
96   /// the special case list contains a line
97   /// \code
98   ///   @Prefix:<E>=@Category
99   /// \endcode
100   /// where @Query satisfies wildcard expression <E> in a given @Section.
101   /// Returns zero if there is no exclusion entry corresponding to this
102   /// expression.
103   unsigned inSectionBlame(StringRef Section, StringRef Prefix, StringRef Query,
104                           StringRef Category = StringRef()) const;
105 
106 protected:
107   // Implementations of the create*() functions that can also be used by derived
108   // classes.
109   bool createInternal(const std::vector<std::string> &Paths,
110                       vfs::FileSystem &VFS, std::string &Error);
111   bool createInternal(const MemoryBuffer *MB, std::string &Error);
112 
113   SpecialCaseList() = default;
114   SpecialCaseList(SpecialCaseList const &) = delete;
115   SpecialCaseList &operator=(SpecialCaseList const &) = delete;
116 
117   /// Represents a set of regular expressions.  Regular expressions which are
118   /// "literal" (i.e. no regex metacharacters) are stored in Strings.  The
119   /// reason for doing so is efficiency; StringMap is much faster at matching
120   /// literal strings than Regex.
121   class Matcher {
122   public:
123     bool insert(std::string Regexp, unsigned LineNumber, std::string &REError);
124     // Returns the line number in the source file that this query matches to.
125     // Returns zero if no match is found.
126     unsigned match(StringRef Query) const;
127 
128   private:
129     StringMap<unsigned> Strings;
130     TrigramIndex Trigrams;
131     std::vector<std::pair<std::unique_ptr<Regex>, unsigned>> RegExes;
132   };
133 
134   using SectionEntries = StringMap<StringMap<Matcher>>;
135 
136   struct Section {
137     Section(std::unique_ptr<Matcher> M) : SectionMatcher(std::move(M)){};
138 
139     std::unique_ptr<Matcher> SectionMatcher;
140     SectionEntries Entries;
141   };
142 
143   std::vector<Section> Sections;
144 
145   /// Parses just-constructed SpecialCaseList entries from a memory buffer.
146   bool parse(const MemoryBuffer *MB, StringMap<size_t> &SectionsMap,
147              std::string &Error);
148 
149   // Helper method for derived classes to search by Prefix, Query, and Category
150   // once they have already resolved a section entry.
151   unsigned inSectionBlame(const SectionEntries &Entries, StringRef Prefix,
152                           StringRef Query, StringRef Category) const;
153 };
154 
155 }  // namespace llvm
156 
157 #endif  // LLVM_SUPPORT_SPECIALCASELIST_H
158 
159