1 <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01//EN"
2 "http://www.w3.org/TR/html4/strict.dtd">
5 <title>Extending LLVM: Adding instructions, intrinsics, types, etc.</title>
6 <link rel="stylesheet" href="llvm.css" type="text/css">
11 <div class="doc_title">
12 Extending LLVM: Adding instructions, intrinsics, types, etc.
16 <li><a href="#introduction">Introduction and Warning</a></li>
17 <li><a href="#intrinsic">Adding a new intrinsic function</a></li>
18 <li><a href="#instruction">Adding a new instruction</a></li>
19 <li><a href="#type">Adding a new type</a>
21 <li><a href="#fund_type">Adding a new fundamental type</a></li>
22 <li><a href="#derived_type">Adding a new derived type</a></li>
26 <div class="doc_author">
27 <p>Written by <a href="http://misha.brukman.net">Misha Brukman</a> and
31 <!-- *********************************************************************** -->
32 <div class="doc_section">
33 <a name="introduction">Introduction and Warning</a>
35 <!-- *********************************************************************** -->
37 <div class="doc_text">
39 <p>During the course of using LLVM, you may wish to customize it for your
40 research project or for experimentation. At this point, you may realize that
41 you need to add something to LLVM, whether it be a new fundamental type, a new
42 intrinsic function, or a whole new instruction.</p>
44 <p>When you come to this realization, stop and think. Do you really need to
45 extend LLVM? Is it a new fundamental capability that LLVM does not support at
46 its current incarnation or can it be synthesized from already pre-existing LLVM
47 elements? If you are not sure, ask on the <a
48 href="http://mail.cs.uiuc.edu/mailman/listinfo/llvmdev">LLVM-dev</a> list. The
49 reason is that extending LLVM will get involved as you need to update all the
50 different passes that you intend to use with your extension, and there are
51 <em>many</em> LLVM analyses and transformations, so it may be quite a bit of
54 <p>Adding an <a href="#intrinsic">intrinsic function</a> is easier than adding
55 an instruction, and is transparent to optimization passes which treat it as an
56 unanalyzable function. If your added functionality can be expressed as a
57 function call, an intrinsic function is the method of choice for LLVM
60 <p>Before you invest a significant amount of effort into a non-trivial
61 extension, <span class="doc_warning">ask on the list</span> if what you are
62 looking to do can be done with already-existing infrastructure, or if maybe
63 someone else is already working on it. You will save yourself a lot of time and
64 effort by doing so.</p>
68 <!-- *********************************************************************** -->
69 <div class="doc_section">
70 <a name="intrinsic">Adding a new intrinsic function</a>
72 <!-- *********************************************************************** -->
74 <div class="doc_text">
76 <p>Adding a new intrinsic function to LLVM is much easier than adding a new
77 instruction. Almost all extensions to LLVM should start as an intrinsic
78 function and then be turned into an instruction if warranted.</p>
81 <li><tt>llvm/docs/LangRef.html</tt>:
82 Document the intrinsic. Decide whether it is code generator specific and
83 what the restrictions are. Talk to other people about it so that you are
84 sure it's a good idea.</li>
86 <li><tt>llvm/include/llvm/Intrinsics.h</tt>:
87 add an enum in the <tt>llvm::Intrinsic</tt> namespace</li>
89 <li><tt>llvm/lib/CodeGen/IntrinsicLowering.cpp</tt>:
90 implement the lowering for this intrinsic</li>
92 <li><tt>llvm/lib/VMCore/Verifier.cpp</tt>:
93 Add code to check the invariants of the intrinsic are respected.</li>
95 <li><tt>llvm/lib/VMCore/Function.cpp (<tt>Function::getIntrinsicID()</tt>)</tt>:
96 Identify the new intrinsic function, returning the enum for the intrinsic
99 <li><tt>llvm/lib/Analysis/BasicAliasAnalysis.cpp</tt>: If the new intrinsic does
100 not access memory or does not write to memory, add it to the relevant list
103 <li><tt>llvm/lib/Transforms/Utils/Local.cpp</tt>: If it is possible to constant
104 propagate your intrinsic, add support to it in the
105 <tt>canConstantFoldCallTo</tt> and <tt>ConstantFoldCall</tt> functions.</li>
107 <li>Test your intrinsic</li>
108 <li><tt>llvm/test/Regression/*</tt>: add your test cases to the test suite.</li>
111 <p>If this intrinsic requires code generator support (i.e., it cannot be
112 lowered), you should also add support to the code generator in question.</p>
116 <!-- *********************************************************************** -->
117 <div class="doc_section">
118 <a name="instruction">Adding a new instruction</a>
120 <!-- *********************************************************************** -->
122 <div class="doc_text">
124 <p><span class="doc_warning">WARNING: adding instructions changes the bytecode
125 format, and it will take some effort to maintain compatibility with
126 the previous version.</span> Only add an instruction if it is absolutely
131 <li><tt>llvm/include/llvm/Instruction.def</tt>:
132 add a number for your instruction and an enum name</li>
134 <li><tt>llvm/include/llvm/Instructions.h</tt>:
135 add a definition for the class that will represent your instruction</li>
137 <li><tt>llvm/include/llvm/Support/InstVisitor.h</tt>:
138 add a prototype for a visitor to your new instruction type</li>
140 <li><tt>llvm/lib/AsmParser/Lexer.l</tt>:
141 add a new token to parse your instruction from assembly text file</li>
143 <li><tt>llvm/lib/AsmParser/llvmAsmParser.y</tt>:
144 add the grammar on how your instruction can be read and what it will
145 construct as a result</li>
147 <li><tt>llvm/lib/Bytecode/Reader/InstructionReader.cpp</tt>:
148 add a case for your instruction and how it will be parsed from bytecode</li>
150 <li><tt>llvm/lib/VMCore/Instruction.cpp</tt>:
151 add a case for how your instruction will be printed out to assembly</li>
153 <li><tt>llvm/lib/VMCore/Instructions.cpp</tt>:
154 implement the class you defined in <tt>llvm/include/llvm/Instructions.h</tt></li>
158 <p>Also, you need to implement (or modify) any analyses or passes that you want
159 to understand this new instruction.</p>
164 <!-- *********************************************************************** -->
165 <div class="doc_section">
166 <a name="type">Adding a new type</a>
168 <!-- *********************************************************************** -->
170 <div class="doc_text">
172 <p><span class="doc_warning">WARNING: adding new types changes the bytecode
173 format, and will break compatibility with currently-existing LLVM
174 installations.</span> Only add new types if it is absolutely necessary.</p>
178 <!-- ======================================================================= -->
179 <div class="doc_subsection">
180 <a name="fund_type">Adding a fundamental type</a>
183 <div class="doc_text">
187 <li><tt>llvm/include/llvm/Type.def</tt>:
188 add enum for the type</li>
190 <li><tt>llvm/include/llvm/Type.h</tt>:
191 add ID number for the new type; add static <tt>Type*</tt> for this type</li>
193 <li><tt>llvm/lib/VMCore/Type.cpp</tt>:
194 add mapping from <tt>TypeID</tt> => <tt>Type*</tt>;
195 initialize the static <tt>Type*</tt></li>
197 <li><tt>llvm/lib/AsmReader/Lexer.l</tt>:
198 add ability to parse in the type from text assembly</li>
200 <li><tt>llvm/lib/AsmReader/llvmAsmParser.y</tt>:
201 add a token for that type</li>
207 <!-- ======================================================================= -->
208 <div class="doc_subsection">
209 <a name="derived_type">Adding a derived type</a>
212 <div class="doc_text">
215 <li><tt>llvm/include/llvm/Type.def</tt>:
216 add enum for the type</li>
218 <li><tt>llvm/include/llvm/Type.h</tt>:
219 add ID number for the new type; add a forward declaration of the type
222 <li><tt>llvm/include/llvm/DerivedType.h</tt>:
223 add new class to represent new class in the hierarchy; add forward
224 declaration to the TypeMap value type</li>
226 <li><tt>llvm/lib/VMCore/Type.cpp</tt>:
227 add support for derived type to:
228 <div class="doc_code">
230 std::string getTypeDescription(const Type &Ty,
231 std::vector<const Type*> &TypeStack)
232 bool TypesEqual(const Type *Ty, const Type *Ty2,
233 std::map<const Type*, const Type*> & EqTypes)
236 add necessary member functions for type, and factory methods</li>
238 <li><tt>llvm/lib/AsmReader/Lexer.l</tt>:
239 add ability to parse in the type from text assembly</li>
241 <li><tt>llvm/lib/ByteCode/Writer/Writer.cpp</tt>:
242 modify <tt>void BytecodeWriter::outputType(const Type *T)</tt> to serialize
245 <li><tt>llvm/lib/ByteCode/Reader/Reader.cpp</tt>:
246 modify <tt>const Type *BytecodeReader::ParseType()</tt> to read your data
249 <li><tt>llvm/lib/VMCore/AsmWriter.cpp</tt>:
251 <div class="doc_code">
253 void calcTypeName(const Type *Ty,
254 std::vector<const Type*> &TypeStack,
255 std::map<const Type*,std::string> &TypeNames,
256 std::string & Result)
259 to output the new derived type
267 <!-- *********************************************************************** -->
271 <a href="http://jigsaw.w3.org/css-validator/check/referer"><img
272 src="http://jigsaw.w3.org/css-validator/images/vcss" alt="Valid CSS!"></a>
273 <a href="http://validator.w3.org/check/referer"><img
274 src="http://www.w3.org/Icons/valid-html401" alt="Valid HTML 4.01!" /></a>
276 <a href="http://misha.brukman.net">Misha Brukman</a><br>
277 <a href="http://llvm.cs.uiuc.edu">The LLVM Compiler Infrastructure</a>
279 Last modified: $Date$