1 <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01//EN"
2 "http://www.w3.org/TR/html4/strict.dtd">
5 <title>LLVM Assembly Language Reference Manual</title>
6 <meta http-equiv="Content-Type" content="text/html; charset=utf-8">
7 <meta name="author" content="Chris Lattner">
8 <meta name="description"
9 content="LLVM Assembly Language Reference Manual.">
10 <link rel="stylesheet" href="llvm.css" type="text/css">
15 <div class="doc_title"> LLVM Language Reference Manual </div>
17 <li><a href="#abstract">Abstract</a></li>
18 <li><a href="#introduction">Introduction</a></li>
19 <li><a href="#identifiers">Identifiers</a></li>
20 <li><a href="#highlevel">High Level Structure</a>
22 <li><a href="#modulestructure">Module Structure</a></li>
23 <li><a href="#linkage">Linkage Types</a></li>
24 <li><a href="#callingconv">Calling Conventions</a></li>
25 <li><a href="#globalvars">Global Variables</a></li>
26 <li><a href="#functionstructure">Functions</a></li>
27 <li><a href="aliasstructure">Aliases</a>
28 <li><a href="#paramattrs">Parameter Attributes</a></li>
29 <li><a href="#moduleasm">Module-Level Inline Assembly</a></li>
30 <li><a href="#datalayout">Data Layout</a></li>
33 <li><a href="#typesystem">Type System</a>
35 <li><a href="#t_primitive">Primitive Types</a>
37 <li><a href="#t_classifications">Type Classifications</a></li>
40 <li><a href="#t_derived">Derived Types</a>
42 <li><a href="#t_array">Array Type</a></li>
43 <li><a href="#t_function">Function Type</a></li>
44 <li><a href="#t_pointer">Pointer Type</a></li>
45 <li><a href="#t_struct">Structure Type</a></li>
46 <li><a href="#t_pstruct">Packed Structure Type</a></li>
47 <li><a href="#t_vector">Vector Type</a></li>
48 <li><a href="#t_opaque">Opaque Type</a></li>
53 <li><a href="#constants">Constants</a>
55 <li><a href="#simpleconstants">Simple Constants</a>
56 <li><a href="#aggregateconstants">Aggregate Constants</a>
57 <li><a href="#globalconstants">Global Variable and Function Addresses</a>
58 <li><a href="#undefvalues">Undefined Values</a>
59 <li><a href="#constantexprs">Constant Expressions</a>
62 <li><a href="#othervalues">Other Values</a>
64 <li><a href="#inlineasm">Inline Assembler Expressions</a>
67 <li><a href="#instref">Instruction Reference</a>
69 <li><a href="#terminators">Terminator Instructions</a>
71 <li><a href="#i_ret">'<tt>ret</tt>' Instruction</a></li>
72 <li><a href="#i_br">'<tt>br</tt>' Instruction</a></li>
73 <li><a href="#i_switch">'<tt>switch</tt>' Instruction</a></li>
74 <li><a href="#i_invoke">'<tt>invoke</tt>' Instruction</a></li>
75 <li><a href="#i_unwind">'<tt>unwind</tt>' Instruction</a></li>
76 <li><a href="#i_unreachable">'<tt>unreachable</tt>' Instruction</a></li>
79 <li><a href="#binaryops">Binary Operations</a>
81 <li><a href="#i_add">'<tt>add</tt>' Instruction</a></li>
82 <li><a href="#i_sub">'<tt>sub</tt>' Instruction</a></li>
83 <li><a href="#i_mul">'<tt>mul</tt>' Instruction</a></li>
84 <li><a href="#i_udiv">'<tt>udiv</tt>' Instruction</a></li>
85 <li><a href="#i_sdiv">'<tt>sdiv</tt>' Instruction</a></li>
86 <li><a href="#i_fdiv">'<tt>fdiv</tt>' Instruction</a></li>
87 <li><a href="#i_urem">'<tt>urem</tt>' Instruction</a></li>
88 <li><a href="#i_srem">'<tt>srem</tt>' Instruction</a></li>
89 <li><a href="#i_frem">'<tt>frem</tt>' Instruction</a></li>
92 <li><a href="#bitwiseops">Bitwise Binary Operations</a>
94 <li><a href="#i_shl">'<tt>shl</tt>' Instruction</a></li>
95 <li><a href="#i_lshr">'<tt>lshr</tt>' Instruction</a></li>
96 <li><a href="#i_ashr">'<tt>ashr</tt>' Instruction</a></li>
97 <li><a href="#i_and">'<tt>and</tt>' Instruction</a></li>
98 <li><a href="#i_or">'<tt>or</tt>' Instruction</a></li>
99 <li><a href="#i_xor">'<tt>xor</tt>' Instruction</a></li>
102 <li><a href="#vectorops">Vector Operations</a>
104 <li><a href="#i_extractelement">'<tt>extractelement</tt>' Instruction</a></li>
105 <li><a href="#i_insertelement">'<tt>insertelement</tt>' Instruction</a></li>
106 <li><a href="#i_shufflevector">'<tt>shufflevector</tt>' Instruction</a></li>
109 <li><a href="#memoryops">Memory Access and Addressing Operations</a>
111 <li><a href="#i_malloc">'<tt>malloc</tt>' Instruction</a></li>
112 <li><a href="#i_free">'<tt>free</tt>' Instruction</a></li>
113 <li><a href="#i_alloca">'<tt>alloca</tt>' Instruction</a></li>
114 <li><a href="#i_load">'<tt>load</tt>' Instruction</a></li>
115 <li><a href="#i_store">'<tt>store</tt>' Instruction</a></li>
116 <li><a href="#i_getelementptr">'<tt>getelementptr</tt>' Instruction</a></li>
119 <li><a href="#convertops">Conversion Operations</a>
121 <li><a href="#i_trunc">'<tt>trunc .. to</tt>' Instruction</a></li>
122 <li><a href="#i_zext">'<tt>zext .. to</tt>' Instruction</a></li>
123 <li><a href="#i_sext">'<tt>sext .. to</tt>' Instruction</a></li>
124 <li><a href="#i_fptrunc">'<tt>fptrunc .. to</tt>' Instruction</a></li>
125 <li><a href="#i_fpext">'<tt>fpext .. to</tt>' Instruction</a></li>
126 <li><a href="#i_fptoui">'<tt>fptoui .. to</tt>' Instruction</a></li>
127 <li><a href="#i_fptosi">'<tt>fptosi .. to</tt>' Instruction</a></li>
128 <li><a href="#i_uitofp">'<tt>uitofp .. to</tt>' Instruction</a></li>
129 <li><a href="#i_sitofp">'<tt>sitofp .. to</tt>' Instruction</a></li>
130 <li><a href="#i_ptrtoint">'<tt>ptrtoint .. to</tt>' Instruction</a></li>
131 <li><a href="#i_inttoptr">'<tt>inttoptr .. to</tt>' Instruction</a></li>
132 <li><a href="#i_bitcast">'<tt>bitcast .. to</tt>' Instruction</a></li>
134 <li><a href="#otherops">Other Operations</a>
136 <li><a href="#i_icmp">'<tt>icmp</tt>' Instruction</a></li>
137 <li><a href="#i_fcmp">'<tt>fcmp</tt>' Instruction</a></li>
138 <li><a href="#i_phi">'<tt>phi</tt>' Instruction</a></li>
139 <li><a href="#i_select">'<tt>select</tt>' Instruction</a></li>
140 <li><a href="#i_call">'<tt>call</tt>' Instruction</a></li>
141 <li><a href="#i_va_arg">'<tt>va_arg</tt>' Instruction</a></li>
146 <li><a href="#intrinsics">Intrinsic Functions</a>
148 <li><a href="#int_varargs">Variable Argument Handling Intrinsics</a>
150 <li><a href="#int_va_start">'<tt>llvm.va_start</tt>' Intrinsic</a></li>
151 <li><a href="#int_va_end">'<tt>llvm.va_end</tt>' Intrinsic</a></li>
152 <li><a href="#int_va_copy">'<tt>llvm.va_copy</tt>' Intrinsic</a></li>
155 <li><a href="#int_gc">Accurate Garbage Collection Intrinsics</a>
157 <li><a href="#int_gcroot">'<tt>llvm.gcroot</tt>' Intrinsic</a></li>
158 <li><a href="#int_gcread">'<tt>llvm.gcread</tt>' Intrinsic</a></li>
159 <li><a href="#int_gcwrite">'<tt>llvm.gcwrite</tt>' Intrinsic</a></li>
162 <li><a href="#int_codegen">Code Generator Intrinsics</a>
164 <li><a href="#int_returnaddress">'<tt>llvm.returnaddress</tt>' Intrinsic</a></li>
165 <li><a href="#int_frameaddress">'<tt>llvm.frameaddress</tt>' Intrinsic</a></li>
166 <li><a href="#int_stacksave">'<tt>llvm.stacksave</tt>' Intrinsic</a></li>
167 <li><a href="#int_stackrestore">'<tt>llvm.stackrestore</tt>' Intrinsic</a></li>
168 <li><a href="#int_prefetch">'<tt>llvm.prefetch</tt>' Intrinsic</a></li>
169 <li><a href="#int_pcmarker">'<tt>llvm.pcmarker</tt>' Intrinsic</a></li>
170 <li><a href="#int_readcyclecounter"><tt>llvm.readcyclecounter</tt>' Intrinsic</a></li>
173 <li><a href="#int_libc">Standard C Library Intrinsics</a>
175 <li><a href="#int_memcpy">'<tt>llvm.memcpy.*</tt>' Intrinsic</a></li>
176 <li><a href="#int_memmove">'<tt>llvm.memmove.*</tt>' Intrinsic</a></li>
177 <li><a href="#int_memset">'<tt>llvm.memset.*</tt>' Intrinsic</a></li>
178 <li><a href="#int_sqrt">'<tt>llvm.sqrt.*</tt>' Intrinsic</a></li>
179 <li><a href="#int_powi">'<tt>llvm.powi.*</tt>' Intrinsic</a></li>
182 <li><a href="#int_manip">Bit Manipulation Intrinsics</a>
184 <li><a href="#int_bswap">'<tt>llvm.bswap.*</tt>' Intrinsics</a></li>
185 <li><a href="#int_ctpop">'<tt>llvm.ctpop.*</tt>' Intrinsic </a></li>
186 <li><a href="#int_ctlz">'<tt>llvm.ctlz.*</tt>' Intrinsic </a></li>
187 <li><a href="#int_cttz">'<tt>llvm.cttz.*</tt>' Intrinsic </a></li>
188 <li><a href="#int_part_select">'<tt>llvm.part.select.*</tt>' Intrinsic </a></li>
189 <li><a href="#int_part_set">'<tt>llvm.part.set.*</tt>' Intrinsic </a></li>
192 <li><a href="#int_debugger">Debugger intrinsics</a></li>
193 <li><a href="#int_eh">Exception Handling intrinsics</a></li>
198 <div class="doc_author">
199 <p>Written by <a href="mailto:sabre@nondot.org">Chris Lattner</a>
200 and <a href="mailto:vadve@cs.uiuc.edu">Vikram Adve</a></p>
203 <!-- *********************************************************************** -->
204 <div class="doc_section"> <a name="abstract">Abstract </a></div>
205 <!-- *********************************************************************** -->
207 <div class="doc_text">
208 <p>This document is a reference manual for the LLVM assembly language.
209 LLVM is an SSA based representation that provides type safety,
210 low-level operations, flexibility, and the capability of representing
211 'all' high-level languages cleanly. It is the common code
212 representation used throughout all phases of the LLVM compilation
216 <!-- *********************************************************************** -->
217 <div class="doc_section"> <a name="introduction">Introduction</a> </div>
218 <!-- *********************************************************************** -->
220 <div class="doc_text">
222 <p>The LLVM code representation is designed to be used in three
223 different forms: as an in-memory compiler IR, as an on-disk bytecode
224 representation (suitable for fast loading by a Just-In-Time compiler),
225 and as a human readable assembly language representation. This allows
226 LLVM to provide a powerful intermediate representation for efficient
227 compiler transformations and analysis, while providing a natural means
228 to debug and visualize the transformations. The three different forms
229 of LLVM are all equivalent. This document describes the human readable
230 representation and notation.</p>
232 <p>The LLVM representation aims to be light-weight and low-level
233 while being expressive, typed, and extensible at the same time. It
234 aims to be a "universal IR" of sorts, by being at a low enough level
235 that high-level ideas may be cleanly mapped to it (similar to how
236 microprocessors are "universal IR's", allowing many source languages to
237 be mapped to them). By providing type information, LLVM can be used as
238 the target of optimizations: for example, through pointer analysis, it
239 can be proven that a C automatic variable is never accessed outside of
240 the current function... allowing it to be promoted to a simple SSA
241 value instead of a memory location.</p>
245 <!-- _______________________________________________________________________ -->
246 <div class="doc_subsubsection"> <a name="wellformed">Well-Formedness</a> </div>
248 <div class="doc_text">
250 <p>It is important to note that this document describes 'well formed'
251 LLVM assembly language. There is a difference between what the parser
252 accepts and what is considered 'well formed'. For example, the
253 following instruction is syntactically okay, but not well formed:</p>
256 %x = <a href="#i_add">add</a> i32 1, %x
259 <p>...because the definition of <tt>%x</tt> does not dominate all of
260 its uses. The LLVM infrastructure provides a verification pass that may
261 be used to verify that an LLVM module is well formed. This pass is
262 automatically run by the parser after parsing input assembly and by
263 the optimizer before it outputs bytecode. The violations pointed out
264 by the verifier pass indicate bugs in transformation passes or input to
267 <!-- Describe the typesetting conventions here. --> </div>
269 <!-- *********************************************************************** -->
270 <div class="doc_section"> <a name="identifiers">Identifiers</a> </div>
271 <!-- *********************************************************************** -->
273 <div class="doc_text">
275 <p>LLVM uses three different forms of identifiers, for different
279 <li>Named values are represented as a string of characters with a '%' prefix.
280 For example, %foo, %DivisionByZero, %a.really.long.identifier. The actual
281 regular expression used is '<tt>%[a-zA-Z$._][a-zA-Z$._0-9]*</tt>'.
282 Identifiers which require other characters in their names can be surrounded
283 with quotes. In this way, anything except a <tt>"</tt> character can be used
286 <li>Unnamed values are represented as an unsigned numeric value with a '%'
287 prefix. For example, %12, %2, %44.</li>
289 <li>Constants, which are described in a <a href="#constants">section about
290 constants</a>, below.</li>
293 <p>LLVM requires that values start with a '%' sign for two reasons: Compilers
294 don't need to worry about name clashes with reserved words, and the set of
295 reserved words may be expanded in the future without penalty. Additionally,
296 unnamed identifiers allow a compiler to quickly come up with a temporary
297 variable without having to avoid symbol table conflicts.</p>
299 <p>Reserved words in LLVM are very similar to reserved words in other
300 languages. There are keywords for different opcodes
301 ('<tt><a href="#i_add">add</a></tt>',
302 '<tt><a href="#i_bitcast">bitcast</a></tt>',
303 '<tt><a href="#i_ret">ret</a></tt>', etc...), for primitive type names ('<tt><a
304 href="#t_void">void</a></tt>', '<tt><a href="#t_primitive">i32</a></tt>', etc...),
305 and others. These reserved words cannot conflict with variable names, because
306 none of them start with a '%' character.</p>
308 <p>Here is an example of LLVM code to multiply the integer variable
309 '<tt>%X</tt>' by 8:</p>
314 %result = <a href="#i_mul">mul</a> i32 %X, 8
317 <p>After strength reduction:</p>
320 %result = <a href="#i_shl">shl</a> i32 %X, i8 3
323 <p>And the hard way:</p>
326 <a href="#i_add">add</a> i32 %X, %X <i>; yields {i32}:%0</i>
327 <a href="#i_add">add</a> i32 %0, %0 <i>; yields {i32}:%1</i>
328 %result = <a href="#i_add">add</a> i32 %1, %1
331 <p>This last way of multiplying <tt>%X</tt> by 8 illustrates several
332 important lexical features of LLVM:</p>
336 <li>Comments are delimited with a '<tt>;</tt>' and go until the end of
339 <li>Unnamed temporaries are created when the result of a computation is not
340 assigned to a named value.</li>
342 <li>Unnamed temporaries are numbered sequentially</li>
346 <p>...and it also shows a convention that we follow in this document. When
347 demonstrating instructions, we will follow an instruction with a comment that
348 defines the type and name of value produced. Comments are shown in italic
353 <!-- *********************************************************************** -->
354 <div class="doc_section"> <a name="highlevel">High Level Structure</a> </div>
355 <!-- *********************************************************************** -->
357 <!-- ======================================================================= -->
358 <div class="doc_subsection"> <a name="modulestructure">Module Structure</a>
361 <div class="doc_text">
363 <p>LLVM programs are composed of "Module"s, each of which is a
364 translation unit of the input programs. Each module consists of
365 functions, global variables, and symbol table entries. Modules may be
366 combined together with the LLVM linker, which merges function (and
367 global variable) definitions, resolves forward declarations, and merges
368 symbol table entries. Here is an example of the "hello world" module:</p>
370 <pre><i>; Declare the string constant as a global constant...</i>
371 <a href="#identifiers">%.LC0</a> = <a href="#linkage_internal">internal</a> <a
372 href="#globalvars">constant</a> <a href="#t_array">[13 x i8 ]</a> c"hello world\0A\00" <i>; [13 x i8 ]*</i>
374 <i>; External declaration of the puts function</i>
375 <a href="#functionstructure">declare</a> i32 %puts(i8 *) <i>; i32(i8 *)* </i>
377 <i>; Definition of main function</i>
378 define i32 %main() { <i>; i32()* </i>
379 <i>; Convert [13x i8 ]* to i8 *...</i>
381 href="#i_getelementptr">getelementptr</a> [13 x i8 ]* %.LC0, i64 0, i64 0 <i>; i8 *</i>
383 <i>; Call puts function to write out the string to stdout...</i>
385 href="#i_call">call</a> i32 %puts(i8 * %cast210) <i>; i32</i>
387 href="#i_ret">ret</a> i32 0<br>}<br></pre>
389 <p>This example is made up of a <a href="#globalvars">global variable</a>
390 named "<tt>.LC0</tt>", an external declaration of the "<tt>puts</tt>"
391 function, and a <a href="#functionstructure">function definition</a>
392 for "<tt>main</tt>".</p>
394 <p>In general, a module is made up of a list of global values,
395 where both functions and global variables are global values. Global values are
396 represented by a pointer to a memory location (in this case, a pointer to an
397 array of char, and a pointer to a function), and have one of the following <a
398 href="#linkage">linkage types</a>.</p>
402 <!-- ======================================================================= -->
403 <div class="doc_subsection">
404 <a name="linkage">Linkage Types</a>
407 <div class="doc_text">
410 All Global Variables and Functions have one of the following types of linkage:
415 <dt><tt><b><a name="linkage_internal">internal</a></b></tt> </dt>
417 <dd>Global values with internal linkage are only directly accessible by
418 objects in the current module. In particular, linking code into a module with
419 an internal global value may cause the internal to be renamed as necessary to
420 avoid collisions. Because the symbol is internal to the module, all
421 references can be updated. This corresponds to the notion of the
422 '<tt>static</tt>' keyword in C.
425 <dt><tt><b><a name="linkage_linkonce">linkonce</a></b></tt>: </dt>
427 <dd>Globals with "<tt>linkonce</tt>" linkage are merged with other globals of
428 the same name when linkage occurs. This is typically used to implement
429 inline functions, templates, or other code which must be generated in each
430 translation unit that uses it. Unreferenced <tt>linkonce</tt> globals are
431 allowed to be discarded.
434 <dt><tt><b><a name="linkage_weak">weak</a></b></tt>: </dt>
436 <dd>"<tt>weak</tt>" linkage is exactly the same as <tt>linkonce</tt> linkage,
437 except that unreferenced <tt>weak</tt> globals may not be discarded. This is
438 used for globals that may be emitted in multiple translation units, but that
439 are not guaranteed to be emitted into every translation unit that uses them.
440 One example of this are common globals in C, such as "<tt>int X;</tt>" at
444 <dt><tt><b><a name="linkage_appending">appending</a></b></tt>: </dt>
446 <dd>"<tt>appending</tt>" linkage may only be applied to global variables of
447 pointer to array type. When two global variables with appending linkage are
448 linked together, the two global arrays are appended together. This is the
449 LLVM, typesafe, equivalent of having the system linker append together
450 "sections" with identical names when .o files are linked.
453 <dt><tt><b><a name="linkage_externweak">extern_weak</a></b></tt>: </dt>
454 <dd>The semantics of this linkage follow the ELF model: the symbol is weak
455 until linked, if not linked, the symbol becomes null instead of being an
459 <dt><tt><b><a name="linkage_external">externally visible</a></b></tt>:</dt>
461 <dd>If none of the above identifiers are used, the global is externally
462 visible, meaning that it participates in linkage and can be used to resolve
463 external symbol references.
468 The next two types of linkage are targeted for Microsoft Windows platform
469 only. They are designed to support importing (exporting) symbols from (to)
474 <dt><tt><b><a name="linkage_dllimport">dllimport</a></b></tt>: </dt>
476 <dd>"<tt>dllimport</tt>" linkage causes the compiler to reference a function
477 or variable via a global pointer to a pointer that is set up by the DLL
478 exporting the symbol. On Microsoft Windows targets, the pointer name is
479 formed by combining <code>_imp__</code> and the function or variable name.
482 <dt><tt><b><a name="linkage_dllexport">dllexport</a></b></tt>: </dt>
484 <dd>"<tt>dllexport</tt>" linkage causes the compiler to provide a global
485 pointer to a pointer in a DLL, so that it can be referenced with the
486 <tt>dllimport</tt> attribute. On Microsoft Windows targets, the pointer
487 name is formed by combining <code>_imp__</code> and the function or variable
493 <p><a name="linkage_external"></a>For example, since the "<tt>.LC0</tt>"
494 variable is defined to be internal, if another module defined a "<tt>.LC0</tt>"
495 variable and was linked with this one, one of the two would be renamed,
496 preventing a collision. Since "<tt>main</tt>" and "<tt>puts</tt>" are
497 external (i.e., lacking any linkage declarations), they are accessible
498 outside of the current module.</p>
499 <p>It is illegal for a function <i>declaration</i>
500 to have any linkage type other than "externally visible", <tt>dllimport</tt>,
501 or <tt>extern_weak</tt>.</p>
502 <p>Aliases can have only <tt>external</tt>, <tt>internal</tt> and <tt>weak</tt>
506 <!-- ======================================================================= -->
507 <div class="doc_subsection">
508 <a name="callingconv">Calling Conventions</a>
511 <div class="doc_text">
513 <p>LLVM <a href="#functionstructure">functions</a>, <a href="#i_call">calls</a>
514 and <a href="#i_invoke">invokes</a> can all have an optional calling convention
515 specified for the call. The calling convention of any pair of dynamic
516 caller/callee must match, or the behavior of the program is undefined. The
517 following calling conventions are supported by LLVM, and more may be added in
521 <dt><b>"<tt>ccc</tt>" - The C calling convention</b>:</dt>
523 <dd>This calling convention (the default if no other calling convention is
524 specified) matches the target C calling conventions. This calling convention
525 supports varargs function calls and tolerates some mismatch in the declared
526 prototype and implemented declaration of the function (as does normal C).
529 <dt><b>"<tt>fastcc</tt>" - The fast calling convention</b>:</dt>
531 <dd>This calling convention attempts to make calls as fast as possible
532 (e.g. by passing things in registers). This calling convention allows the
533 target to use whatever tricks it wants to produce fast code for the target,
534 without having to conform to an externally specified ABI. Implementations of
535 this convention should allow arbitrary tail call optimization to be supported.
536 This calling convention does not support varargs and requires the prototype of
537 all callees to exactly match the prototype of the function definition.
540 <dt><b>"<tt>coldcc</tt>" - The cold calling convention</b>:</dt>
542 <dd>This calling convention attempts to make code in the caller as efficient
543 as possible under the assumption that the call is not commonly executed. As
544 such, these calls often preserve all registers so that the call does not break
545 any live ranges in the caller side. This calling convention does not support
546 varargs and requires the prototype of all callees to exactly match the
547 prototype of the function definition.
550 <dt><b>"<tt>cc <<em>n</em>></tt>" - Numbered convention</b>:</dt>
552 <dd>Any calling convention may be specified by number, allowing
553 target-specific calling conventions to be used. Target specific calling
554 conventions start at 64.
558 <p>More calling conventions can be added/defined on an as-needed basis, to
559 support pascal conventions or any other well-known target-independent
564 <!-- ======================================================================= -->
565 <div class="doc_subsection">
566 <a name="visibility">Visibility Styles</a>
569 <div class="doc_text">
572 All Global Variables and Functions have one of the following visibility styles:
576 <dt><b>"<tt>default</tt>" - Default style</b>:</dt>
578 <dd>On ELF, default visibility means that the declaration is visible to other
579 modules and, in shared libraries, means that the declared entity may be
580 overridden. On Darwin, default visibility means that the declaration is
581 visible to other modules. Default visibility corresponds to "external
582 linkage" in the language.
585 <dt><b>"<tt>hidden</tt>" - Hidden style</b>:</dt>
587 <dd>Two declarations of an object with hidden visibility refer to the same
588 object if they are in the same shared object. Usually, hidden visibility
589 indicates that the symbol will not be placed into the dynamic symbol table,
590 so no other module (executable or shared library) can reference it
598 <!-- ======================================================================= -->
599 <div class="doc_subsection">
600 <a name="globalvars">Global Variables</a>
603 <div class="doc_text">
605 <p>Global variables define regions of memory allocated at compilation time
606 instead of run-time. Global variables may optionally be initialized, may have
607 an explicit section to be placed in, and may have an optional explicit alignment
608 specified. A variable may be defined as "thread_local", which means that it
609 will not be shared by threads (each thread will have a separated copy of the
610 variable). A variable may be defined as a global "constant," which indicates
611 that the contents of the variable will <b>never</b> be modified (enabling better
612 optimization, allowing the global data to be placed in the read-only section of
613 an executable, etc). Note that variables that need runtime initialization
614 cannot be marked "constant" as there is a store to the variable.</p>
617 LLVM explicitly allows <em>declarations</em> of global variables to be marked
618 constant, even if the final definition of the global is not. This capability
619 can be used to enable slightly better optimization of the program, but requires
620 the language definition to guarantee that optimizations based on the
621 'constantness' are valid for the translation units that do not include the
625 <p>As SSA values, global variables define pointer values that are in
626 scope (i.e. they dominate) all basic blocks in the program. Global
627 variables always define a pointer to their "content" type because they
628 describe a region of memory, and all memory objects in LLVM are
629 accessed through pointers.</p>
631 <p>LLVM allows an explicit section to be specified for globals. If the target
632 supports it, it will emit globals to the section specified.</p>
634 <p>An explicit alignment may be specified for a global. If not present, or if
635 the alignment is set to zero, the alignment of the global is set by the target
636 to whatever it feels convenient. If an explicit alignment is specified, the
637 global is forced to have at least that much alignment. All alignments must be
640 <p>For example, the following defines a global with an initializer, section,
644 %G = constant float 1.0, section "foo", align 4
650 <!-- ======================================================================= -->
651 <div class="doc_subsection">
652 <a name="functionstructure">Functions</a>
655 <div class="doc_text">
657 <p>LLVM function definitions consist of the "<tt>define</tt>" keyord,
658 an optional <a href="#linkage">linkage type</a>, an optional
659 <a href="#visibility">visibility style</a>, an optional
660 <a href="#callingconv">calling convention</a>, a return type, an optional
661 <a href="#paramattrs">parameter attribute</a> for the return type, a function
662 name, a (possibly empty) argument list (each with optional
663 <a href="#paramattrs">parameter attributes</a>), an optional section, an
664 optional alignment, an opening curly brace, a list of basic blocks, and a
667 LLVM function declarations consist of the "<tt>declare</tt>" keyword, an
668 optional <a href="#linkage">linkage type</a>, an optional
669 <a href="#visibility">visibility style</a>, an optional
670 <a href="#callingconv">calling convention</a>, a return type, an optional
671 <a href="#paramattrs">parameter attribute</a> for the return type, a function
672 name, a possibly empty list of arguments, and an optional alignment.</p>
674 <p>A function definition contains a list of basic blocks, forming the CFG for
675 the function. Each basic block may optionally start with a label (giving the
676 basic block a symbol table entry), contains a list of instructions, and ends
677 with a <a href="#terminators">terminator</a> instruction (such as a branch or
678 function return).</p>
680 <p>The first basic block in a program is special in two ways: it is immediately
681 executed on entrance to the function, and it is not allowed to have predecessor
682 basic blocks (i.e. there can not be any branches to the entry block of a
683 function). Because the block can have no predecessors, it also cannot have any
684 <a href="#i_phi">PHI nodes</a>.</p>
686 <p>LLVM functions are identified by their name and type signature. Hence, two
687 functions with the same name but different parameter lists or return values are
688 considered different functions, and LLVM will resolve references to each
691 <p>LLVM allows an explicit section to be specified for functions. If the target
692 supports it, it will emit functions to the section specified.</p>
694 <p>An explicit alignment may be specified for a function. If not present, or if
695 the alignment is set to zero, the alignment of the function is set by the target
696 to whatever it feels convenient. If an explicit alignment is specified, the
697 function is forced to have at least that much alignment. All alignments must be
703 <!-- ======================================================================= -->
704 <div class="doc_subsection">
705 <a name="aliasstructure">Aliases</a>
707 <div class="doc_text">
708 <p>Aliases act as "second name" for the aliasee value (which can be either
709 function or global variable or bitcast of global value). Aliases may have an
710 optional <a href="#linkage">linkage type</a>, and an
711 optional <a href="#visibility">visibility style</a>.</p>
716 @<Name> = [Linkage] [Visibility] alias <AliaseeTy> @<Aliasee>
723 <!-- ======================================================================= -->
724 <div class="doc_subsection"><a name="paramattrs">Parameter Attributes</a></div>
725 <div class="doc_text">
726 <p>The return type and each parameter of a function type may have a set of
727 <i>parameter attributes</i> associated with them. Parameter attributes are
728 used to communicate additional information about the result or parameters of
729 a function. Parameter attributes are considered to be part of the function
730 type so two functions types that differ only by the parameter attributes
731 are different function types.</p>
733 <p>Parameter attributes are simple keywords that follow the type specified. If
734 multiple parameter attributes are needed, they are space separated. For
736 %someFunc = i16 (i8 sext %someParam) zext
737 %someFunc = i16 (i8 zext %someParam) zext</pre>
738 <p>Note that the two function types above are unique because the parameter has
739 a different attribute (sext in the first one, zext in the second). Also note
740 that the attribute for the function result (zext) comes immediately after the
743 <p>Currently, only the following parameter attributes are defined:</p>
745 <dt><tt>zext</tt></dt>
746 <dd>This indicates that the parameter should be zero extended just before
747 a call to this function.</dd>
748 <dt><tt>sext</tt></dt>
749 <dd>This indicates that the parameter should be sign extended just before
750 a call to this function.</dd>
751 <dt><tt>inreg</tt></dt>
752 <dd>This indicates that the parameter should be placed in register (if
753 possible) during assembling function call. Support for this attribute is
755 <dt><tt>sret</tt></dt>
756 <dd>This indicates that the parameter specifies the address of a structure
757 that is the return value of the function in the source program.</dd>
758 <dt><tt>noreturn</tt></dt>
759 <dd>This function attribute indicates that the function never returns. This
760 indicates to LLVM that every call to this function should be treated as if
761 an <tt>unreachable</tt> instruction immediately followed the call.</dd>
762 <dt><tt>nounwind</tt></dt>
763 <dd>This function attribute indicates that the function type does not use
764 the unwind instruction and does not allow stack unwinding to propagate
770 <!-- ======================================================================= -->
771 <div class="doc_subsection">
772 <a name="moduleasm">Module-Level Inline Assembly</a>
775 <div class="doc_text">
777 Modules may contain "module-level inline asm" blocks, which corresponds to the
778 GCC "file scope inline asm" blocks. These blocks are internally concatenated by
779 LLVM and treated as a single unit, but may be separated in the .ll file if
780 desired. The syntax is very simple:
783 <div class="doc_code"><pre>
784 module asm "inline asm code goes here"
785 module asm "more can go here"
788 <p>The strings can contain any character by escaping non-printable characters.
789 The escape sequence used is simply "\xx" where "xx" is the two digit hex code
794 The inline asm code is simply printed to the machine code .s file when
795 assembly code is generated.
799 <!-- ======================================================================= -->
800 <div class="doc_subsection">
801 <a name="datalayout">Data Layout</a>
804 <div class="doc_text">
805 <p>A module may specify a target specific data layout string that specifies how
806 data is to be laid out in memory. The syntax for the data layout is simply:</p>
807 <pre> target datalayout = "<i>layout specification</i>"</pre>
808 <p>The <i>layout specification</i> consists of a list of specifications
809 separated by the minus sign character ('-'). Each specification starts with a
810 letter and may include other information after the letter to define some
811 aspect of the data layout. The specifications accepted are as follows: </p>
814 <dd>Specifies that the target lays out data in big-endian form. That is, the
815 bits with the most significance have the lowest address location.</dd>
817 <dd>Specifies that hte target lays out data in little-endian form. That is,
818 the bits with the least significance have the lowest address location.</dd>
819 <dt><tt>p:<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
820 <dd>This specifies the <i>size</i> of a pointer and its <i>abi</i> and
821 <i>preferred</i> alignments. All sizes are in bits. Specifying the <i>pref</i>
822 alignment is optional. If omitted, the preceding <tt>:</tt> should be omitted
824 <dt><tt>i<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
825 <dd>This specifies the alignment for an integer type of a given bit
826 <i>size</i>. The value of <i>size</i> must be in the range [1,2^23).</dd>
827 <dt><tt>v<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
828 <dd>This specifies the alignment for a vector type of a given bit
830 <dt><tt>f<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
831 <dd>This specifies the alignment for a floating point type of a given bit
832 <i>size</i>. The value of <i>size</i> must be either 32 (float) or 64
834 <dt><tt>a<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
835 <dd>This specifies the alignment for an aggregate type of a given bit
838 <p>When constructing the data layout for a given target, LLVM starts with a
839 default set of specifications which are then (possibly) overriden by the
840 specifications in the <tt>datalayout</tt> keyword. The default specifications
841 are given in this list:</p>
843 <li><tt>E</tt> - big endian</li>
844 <li><tt>p:32:64:64</tt> - 32-bit pointers with 64-bit alignment</li>
845 <li><tt>i1:8:8</tt> - i1 is 8-bit (byte) aligned</li>
846 <li><tt>i8:8:8</tt> - i8 is 8-bit (byte) aligned</li>
847 <li><tt>i16:16:16</tt> - i16 is 16-bit aligned</li>
848 <li><tt>i32:32:32</tt> - i32 is 32-bit aligned</li>
849 <li><tt>i64:32:64</tt> - i64 has abi alignment of 32-bits but preferred
850 alignment of 64-bits</li>
851 <li><tt>f32:32:32</tt> - float is 32-bit aligned</li>
852 <li><tt>f64:64:64</tt> - double is 64-bit aligned</li>
853 <li><tt>v64:64:64</tt> - 64-bit vector is 64-bit aligned</li>
854 <li><tt>v128:128:128</tt> - 128-bit vector is 128-bit aligned</li>
855 <li><tt>a0:0:1</tt> - aggregates are 8-bit aligned</li>
857 <p>When llvm is determining the alignment for a given type, it uses the
860 <li>If the type sought is an exact match for one of the specifications, that
861 specification is used.</li>
862 <li>If no match is found, and the type sought is an integer type, then the
863 smallest integer type that is larger than the bitwidth of the sought type is
864 used. If none of the specifications are larger than the bitwidth then the the
865 largest integer type is used. For example, given the default specifications
866 above, the i7 type will use the alignment of i8 (next largest) while both
867 i65 and i256 will use the alignment of i64 (largest specified).</li>
868 <li>If no match is found, and the type sought is a vector type, then the
869 largest vector type that is smaller than the sought vector type will be used
870 as a fall back. This happens because <128 x double> can be implemented in
871 terms of 64 <2 x double>, for example.</li>
875 <!-- *********************************************************************** -->
876 <div class="doc_section"> <a name="typesystem">Type System</a> </div>
877 <!-- *********************************************************************** -->
879 <div class="doc_text">
881 <p>The LLVM type system is one of the most important features of the
882 intermediate representation. Being typed enables a number of
883 optimizations to be performed on the IR directly, without having to do
884 extra analyses on the side before the transformation. A strong type
885 system makes it easier to read the generated code and enables novel
886 analyses and transformations that are not feasible to perform on normal
887 three address code representations.</p>
891 <!-- ======================================================================= -->
892 <div class="doc_subsection"> <a name="t_primitive">Primitive Types</a> </div>
893 <div class="doc_text">
894 <p>The primitive types are the fundamental building blocks of the LLVM
895 system. The current set of primitive types is as follows:</p>
897 <table class="layout">
902 <tr><th>Type</th><th>Description</th></tr>
903 <tr><td><tt><a name="t_void">void</a></tt></td><td>No value</td></tr>
904 <tr><td><tt>i8</tt></td><td>8-bit value</td></tr>
905 <tr><td><tt>i32</tt></td><td>32-bit value</td></tr>
906 <tr><td><tt>float</tt></td><td>32-bit floating point value</td></tr>
907 <tr><td><tt>label</tt></td><td>Branch destination</td></tr>
914 <tr><th>Type</th><th>Description</th></tr>
915 <tr><td><tt>i1</tt></td><td>True or False value</td></tr>
916 <tr><td><tt>i16</tt></td><td>16-bit value</td></tr>
917 <tr><td><tt>i64</tt></td><td>64-bit value</td></tr>
918 <tr><td><tt>double</tt></td><td>64-bit floating point value</td></tr>
926 <!-- _______________________________________________________________________ -->
927 <div class="doc_subsubsection"> <a name="t_classifications">Type
928 Classifications</a> </div>
929 <div class="doc_text">
930 <p>These different primitive types fall into a few useful
933 <table border="1" cellspacing="0" cellpadding="4">
935 <tr><th>Classification</th><th>Types</th></tr>
937 <td><a name="t_integer">integer</a></td>
938 <td><tt>i1, i8, i16, i32, i64</tt></td>
941 <td><a name="t_floating">floating point</a></td>
942 <td><tt>float, double</tt></td>
945 <td><a name="t_firstclass">first class</a></td>
946 <td><tt>i1, i8, i16, i32, i64, float, double, <br/>
947 <a href="#t_pointer">pointer</a>,<a href="#t_vector">vector</a></tt>
953 <p>The <a href="#t_firstclass">first class</a> types are perhaps the
954 most important. Values of these types are the only ones which can be
955 produced by instructions, passed as arguments, or used as operands to
956 instructions. This means that all structures and arrays must be
957 manipulated either by pointer or by component.</p>
960 <!-- ======================================================================= -->
961 <div class="doc_subsection"> <a name="t_derived">Derived Types</a> </div>
963 <div class="doc_text">
965 <p>The real power in LLVM comes from the derived types in the system.
966 This is what allows a programmer to represent arrays, functions,
967 pointers, and other useful types. Note that these derived types may be
968 recursive: For example, it is possible to have a two dimensional array.</p>
972 <!-- _______________________________________________________________________ -->
973 <div class="doc_subsubsection"> <a name="t_array">Array Type</a> </div>
975 <div class="doc_text">
979 <p>The array type is a very simple derived type that arranges elements
980 sequentially in memory. The array type requires a size (number of
981 elements) and an underlying data type.</p>
986 [<# elements> x <elementtype>]
989 <p>The number of elements is a constant integer value; elementtype may
990 be any type with a size.</p>
993 <table class="layout">
996 <tt>[40 x i32 ]</tt><br/>
997 <tt>[41 x i32 ]</tt><br/>
998 <tt>[40 x i8]</tt><br/>
1001 Array of 40 32-bit integer values.<br/>
1002 Array of 41 32-bit integer values.<br/>
1003 Array of 40 8-bit integer values.<br/>
1007 <p>Here are some examples of multidimensional arrays:</p>
1008 <table class="layout">
1011 <tt>[3 x [4 x i32]]</tt><br/>
1012 <tt>[12 x [10 x float]]</tt><br/>
1013 <tt>[2 x [3 x [4 x i16]]]</tt><br/>
1016 3x4 array of 32-bit integer values.<br/>
1017 12x10 array of single precision floating point values.<br/>
1018 2x3x4 array of 16-bit integer values.<br/>
1023 <p>Note that 'variable sized arrays' can be implemented in LLVM with a zero
1024 length array. Normally, accesses past the end of an array are undefined in
1025 LLVM (e.g. it is illegal to access the 5th element of a 3 element array).
1026 As a special case, however, zero length arrays are recognized to be variable
1027 length. This allows implementation of 'pascal style arrays' with the LLVM
1028 type "{ i32, [0 x float]}", for example.</p>
1032 <!-- _______________________________________________________________________ -->
1033 <div class="doc_subsubsection"> <a name="t_function">Function Type</a> </div>
1034 <div class="doc_text">
1036 <p>The function type can be thought of as a function signature. It
1037 consists of a return type and a list of formal parameter types.
1038 Function types are usually used to build virtual function tables
1039 (which are structures of pointers to functions), for indirect function
1040 calls, and when defining a function.</p>
1042 The return type of a function type cannot be an aggregate type.
1045 <pre> <returntype> (<parameter list>)<br></pre>
1046 <p>...where '<tt><parameter list></tt>' is a comma-separated list of type
1047 specifiers. Optionally, the parameter list may include a type <tt>...</tt>,
1048 which indicates that the function takes a variable number of arguments.
1049 Variable argument functions can access their arguments with the <a
1050 href="#int_varargs">variable argument handling intrinsic</a> functions.</p>
1052 <table class="layout">
1054 <td class="left"><tt>i32 (i32)</tt></td>
1055 <td class="left">function taking an <tt>i32</tt>, returning an <tt>i32</tt>
1057 </tr><tr class="layout">
1058 <td class="left"><tt>float (i16 sext, i32 *) *
1060 <td class="left"><a href="#t_pointer">Pointer</a> to a function that takes
1061 an <tt>i16</tt> that should be sign extended and a
1062 <a href="#t_pointer">pointer</a> to <tt>i32</tt>, returning
1065 </tr><tr class="layout">
1066 <td class="left"><tt>i32 (i8*, ...)</tt></td>
1067 <td class="left">A vararg function that takes at least one
1068 <a href="#t_pointer">pointer</a> to <tt>i8 </tt> (char in C),
1069 which returns an integer. This is the signature for <tt>printf</tt> in
1076 <!-- _______________________________________________________________________ -->
1077 <div class="doc_subsubsection"> <a name="t_struct">Structure Type</a> </div>
1078 <div class="doc_text">
1080 <p>The structure type is used to represent a collection of data members
1081 together in memory. The packing of the field types is defined to match
1082 the ABI of the underlying processor. The elements of a structure may
1083 be any type that has a size.</p>
1084 <p>Structures are accessed using '<tt><a href="#i_load">load</a></tt>
1085 and '<tt><a href="#i_store">store</a></tt>' by getting a pointer to a
1086 field with the '<tt><a href="#i_getelementptr">getelementptr</a></tt>'
1089 <pre> { <type list> }<br></pre>
1091 <table class="layout">
1093 <td class="left"><tt>{ i32, i32, i32 }</tt></td>
1094 <td class="left">A triple of three <tt>i32</tt> values</td>
1095 </tr><tr class="layout">
1096 <td class="left"><tt>{ float, i32 (i32) * }</tt></td>
1097 <td class="left">A pair, where the first element is a <tt>float</tt> and the
1098 second element is a <a href="#t_pointer">pointer</a> to a
1099 <a href="#t_function">function</a> that takes an <tt>i32</tt>, returning
1100 an <tt>i32</tt>.</td>
1105 <!-- _______________________________________________________________________ -->
1106 <div class="doc_subsubsection"> <a name="t_pstruct">Packed Structure Type</a>
1108 <div class="doc_text">
1110 <p>The packed structure type is used to represent a collection of data members
1111 together in memory. There is no padding between fields. Further, the alignment
1112 of a packed structure is 1 byte. The elements of a packed structure may
1113 be any type that has a size.</p>
1114 <p>Structures are accessed using '<tt><a href="#i_load">load</a></tt>
1115 and '<tt><a href="#i_store">store</a></tt>' by getting a pointer to a
1116 field with the '<tt><a href="#i_getelementptr">getelementptr</a></tt>'
1119 <pre> < { <type list> } > <br></pre>
1121 <table class="layout">
1123 <td class="left"><tt>< { i32, i32, i32 } ></tt></td>
1124 <td class="left">A triple of three <tt>i32</tt> values</td>
1125 </tr><tr class="layout">
1126 <td class="left"><tt>< { float, i32 (i32) * } ></tt></td>
1127 <td class="left">A pair, where the first element is a <tt>float</tt> and the
1128 second element is a <a href="#t_pointer">pointer</a> to a
1129 <a href="#t_function">function</a> that takes an <tt>i32</tt>, returning
1130 an <tt>i32</tt>.</td>
1135 <!-- _______________________________________________________________________ -->
1136 <div class="doc_subsubsection"> <a name="t_pointer">Pointer Type</a> </div>
1137 <div class="doc_text">
1139 <p>As in many languages, the pointer type represents a pointer or
1140 reference to another object, which must live in memory.</p>
1142 <pre> <type> *<br></pre>
1144 <table class="layout">
1147 <tt>[4x i32]*</tt><br/>
1148 <tt>i32 (i32 *) *</tt><br/>
1151 A <a href="#t_pointer">pointer</a> to <a href="#t_array">array</a> of
1152 four <tt>i32</tt> values<br/>
1153 A <a href="#t_pointer">pointer</a> to a <a
1154 href="#t_function">function</a> that takes an <tt>i32*</tt>, returning an
1161 <!-- _______________________________________________________________________ -->
1162 <div class="doc_subsubsection"> <a name="t_vector">Vector Type</a> </div>
1163 <div class="doc_text">
1167 <p>A vector type is a simple derived type that represents a vector
1168 of elements. Vector types are used when multiple primitive data
1169 are operated in parallel using a single instruction (SIMD).
1170 A vector type requires a size (number of
1171 elements) and an underlying primitive data type. Vectors must have a power
1172 of two length (1, 2, 4, 8, 16 ...). Vector types are
1173 considered <a href="#t_firstclass">first class</a>.</p>
1178 < <# elements> x <elementtype> >
1181 <p>The number of elements is a constant integer value; elementtype may
1182 be any integer or floating point type.</p>
1186 <table class="layout">
1189 <tt><4 x i32></tt><br/>
1190 <tt><8 x float></tt><br/>
1191 <tt><2 x i64></tt><br/>
1194 Vector of 4 32-bit integer values.<br/>
1195 Vector of 8 floating-point values.<br/>
1196 Vector of 2 64-bit integer values.<br/>
1202 <!-- _______________________________________________________________________ -->
1203 <div class="doc_subsubsection"> <a name="t_opaque">Opaque Type</a> </div>
1204 <div class="doc_text">
1208 <p>Opaque types are used to represent unknown types in the system. This
1209 corresponds (for example) to the C notion of a foward declared structure type.
1210 In LLVM, opaque types can eventually be resolved to any type (not just a
1211 structure type).</p>
1221 <table class="layout">
1227 An opaque type.<br/>
1234 <!-- *********************************************************************** -->
1235 <div class="doc_section"> <a name="constants">Constants</a> </div>
1236 <!-- *********************************************************************** -->
1238 <div class="doc_text">
1240 <p>LLVM has several different basic types of constants. This section describes
1241 them all and their syntax.</p>
1245 <!-- ======================================================================= -->
1246 <div class="doc_subsection"><a name="simpleconstants">Simple Constants</a></div>
1248 <div class="doc_text">
1251 <dt><b>Boolean constants</b></dt>
1253 <dd>The two strings '<tt>true</tt>' and '<tt>false</tt>' are both valid
1254 constants of the <tt><a href="#t_primitive">i1</a></tt> type.
1257 <dt><b>Integer constants</b></dt>
1259 <dd>Standard integers (such as '4') are constants of the <a
1260 href="#t_integer">integer</a> type. Negative numbers may be used with
1264 <dt><b>Floating point constants</b></dt>
1266 <dd>Floating point constants use standard decimal notation (e.g. 123.421),
1267 exponential notation (e.g. 1.23421e+2), or a more precise hexadecimal
1268 notation (see below). Floating point constants must have a <a
1269 href="#t_floating">floating point</a> type. </dd>
1271 <dt><b>Null pointer constants</b></dt>
1273 <dd>The identifier '<tt>null</tt>' is recognized as a null pointer constant
1274 and must be of <a href="#t_pointer">pointer type</a>.</dd>
1278 <p>The one non-intuitive notation for constants is the optional hexadecimal form
1279 of floating point constants. For example, the form '<tt>double
1280 0x432ff973cafa8000</tt>' is equivalent to (but harder to read than) '<tt>double
1281 4.5e+15</tt>'. The only time hexadecimal floating point constants are required
1282 (and the only time that they are generated by the disassembler) is when a
1283 floating point constant must be emitted but it cannot be represented as a
1284 decimal floating point number. For example, NaN's, infinities, and other
1285 special values are represented in their IEEE hexadecimal format so that
1286 assembly and disassembly do not cause any bits to change in the constants.</p>
1290 <!-- ======================================================================= -->
1291 <div class="doc_subsection"><a name="aggregateconstants">Aggregate Constants</a>
1294 <div class="doc_text">
1295 <p>Aggregate constants arise from aggregation of simple constants
1296 and smaller aggregate constants.</p>
1299 <dt><b>Structure constants</b></dt>
1301 <dd>Structure constants are represented with notation similar to structure
1302 type definitions (a comma separated list of elements, surrounded by braces
1303 (<tt>{}</tt>)). For example: "<tt>{ i32 4, float 17.0, i32* %G }</tt>",
1304 where "<tt>%G</tt>" is declared as "<tt>%G = external global i32</tt>". Structure constants
1305 must have <a href="#t_struct">structure type</a>, and the number and
1306 types of elements must match those specified by the type.
1309 <dt><b>Array constants</b></dt>
1311 <dd>Array constants are represented with notation similar to array type
1312 definitions (a comma separated list of elements, surrounded by square brackets
1313 (<tt>[]</tt>)). For example: "<tt>[ i32 42, i32 11, i32 74 ]</tt>". Array
1314 constants must have <a href="#t_array">array type</a>, and the number and
1315 types of elements must match those specified by the type.
1318 <dt><b>Vector constants</b></dt>
1320 <dd>Vector constants are represented with notation similar to vector type
1321 definitions (a comma separated list of elements, surrounded by
1322 less-than/greater-than's (<tt><></tt>)). For example: "<tt>< i32 42,
1323 i32 11, i32 74, i32 100 ></tt>". Vector constants must have <a
1324 href="#t_vector">vector type</a>, and the number and types of elements must
1325 match those specified by the type.
1328 <dt><b>Zero initialization</b></dt>
1330 <dd>The string '<tt>zeroinitializer</tt>' can be used to zero initialize a
1331 value to zero of <em>any</em> type, including scalar and aggregate types.
1332 This is often used to avoid having to print large zero initializers (e.g. for
1333 large arrays) and is always exactly equivalent to using explicit zero
1340 <!-- ======================================================================= -->
1341 <div class="doc_subsection">
1342 <a name="globalconstants">Global Variable and Function Addresses</a>
1345 <div class="doc_text">
1347 <p>The addresses of <a href="#globalvars">global variables</a> and <a
1348 href="#functionstructure">functions</a> are always implicitly valid (link-time)
1349 constants. These constants are explicitly referenced when the <a
1350 href="#identifiers">identifier for the global</a> is used and always have <a
1351 href="#t_pointer">pointer</a> type. For example, the following is a legal LLVM
1357 %Z = global [2 x i32*] [ i32* %X, i32* %Y ]
1362 <!-- ======================================================================= -->
1363 <div class="doc_subsection"><a name="undefvalues">Undefined Values</a></div>
1364 <div class="doc_text">
1365 <p>The string '<tt>undef</tt>' is recognized as a type-less constant that has
1366 no specific value. Undefined values may be of any type and be used anywhere
1367 a constant is permitted.</p>
1369 <p>Undefined values indicate to the compiler that the program is well defined
1370 no matter what value is used, giving the compiler more freedom to optimize.
1374 <!-- ======================================================================= -->
1375 <div class="doc_subsection"><a name="constantexprs">Constant Expressions</a>
1378 <div class="doc_text">
1380 <p>Constant expressions are used to allow expressions involving other constants
1381 to be used as constants. Constant expressions may be of any <a
1382 href="#t_firstclass">first class</a> type and may involve any LLVM operation
1383 that does not have side effects (e.g. load and call are not supported). The
1384 following is the syntax for constant expressions:</p>
1387 <dt><b><tt>trunc ( CST to TYPE )</tt></b></dt>
1388 <dd>Truncate a constant to another type. The bit size of CST must be larger
1389 than the bit size of TYPE. Both types must be integers.</dd>
1391 <dt><b><tt>zext ( CST to TYPE )</tt></b></dt>
1392 <dd>Zero extend a constant to another type. The bit size of CST must be
1393 smaller or equal to the bit size of TYPE. Both types must be integers.</dd>
1395 <dt><b><tt>sext ( CST to TYPE )</tt></b></dt>
1396 <dd>Sign extend a constant to another type. The bit size of CST must be
1397 smaller or equal to the bit size of TYPE. Both types must be integers.</dd>
1399 <dt><b><tt>fptrunc ( CST to TYPE )</tt></b></dt>
1400 <dd>Truncate a floating point constant to another floating point type. The
1401 size of CST must be larger than the size of TYPE. Both types must be
1402 floating point.</dd>
1404 <dt><b><tt>fpext ( CST to TYPE )</tt></b></dt>
1405 <dd>Floating point extend a constant to another type. The size of CST must be
1406 smaller or equal to the size of TYPE. Both types must be floating point.</dd>
1408 <dt><b><tt>fp2uint ( CST to TYPE )</tt></b></dt>
1409 <dd>Convert a floating point constant to the corresponding unsigned integer
1410 constant. TYPE must be an integer type. CST must be floating point. If the
1411 value won't fit in the integer type, the results are undefined.</dd>
1413 <dt><b><tt>fptosi ( CST to TYPE )</tt></b></dt>
1414 <dd>Convert a floating point constant to the corresponding signed integer
1415 constant. TYPE must be an integer type. CST must be floating point. If the
1416 value won't fit in the integer type, the results are undefined.</dd>
1418 <dt><b><tt>uitofp ( CST to TYPE )</tt></b></dt>
1419 <dd>Convert an unsigned integer constant to the corresponding floating point
1420 constant. TYPE must be floating point. CST must be of integer type. If the
1421 value won't fit in the floating point type, the results are undefined.</dd>
1423 <dt><b><tt>sitofp ( CST to TYPE )</tt></b></dt>
1424 <dd>Convert a signed integer constant to the corresponding floating point
1425 constant. TYPE must be floating point. CST must be of integer type. If the
1426 value won't fit in the floating point type, the results are undefined.</dd>
1428 <dt><b><tt>ptrtoint ( CST to TYPE )</tt></b></dt>
1429 <dd>Convert a pointer typed constant to the corresponding integer constant
1430 TYPE must be an integer type. CST must be of pointer type. The CST value is
1431 zero extended, truncated, or unchanged to make it fit in TYPE.</dd>
1433 <dt><b><tt>inttoptr ( CST to TYPE )</tt></b></dt>
1434 <dd>Convert a integer constant to a pointer constant. TYPE must be a
1435 pointer type. CST must be of integer type. The CST value is zero extended,
1436 truncated, or unchanged to make it fit in a pointer size. This one is
1437 <i>really</i> dangerous!</dd>
1439 <dt><b><tt>bitcast ( CST to TYPE )</tt></b></dt>
1440 <dd>Convert a constant, CST, to another TYPE. The size of CST and TYPE must be
1441 identical (same number of bits). The conversion is done as if the CST value
1442 was stored to memory and read back as TYPE. In other words, no bits change
1443 with this operator, just the type. This can be used for conversion of
1444 vector types to any other type, as long as they have the same bit width. For
1445 pointers it is only valid to cast to another pointer type.
1448 <dt><b><tt>getelementptr ( CSTPTR, IDX0, IDX1, ... )</tt></b></dt>
1450 <dd>Perform the <a href="#i_getelementptr">getelementptr operation</a> on
1451 constants. As with the <a href="#i_getelementptr">getelementptr</a>
1452 instruction, the index list may have zero or more indexes, which are required
1453 to make sense for the type of "CSTPTR".</dd>
1455 <dt><b><tt>select ( COND, VAL1, VAL2 )</tt></b></dt>
1457 <dd>Perform the <a href="#i_select">select operation</a> on
1460 <dt><b><tt>icmp COND ( VAL1, VAL2 )</tt></b></dt>
1461 <dd>Performs the <a href="#i_icmp">icmp operation</a> on constants.</dd>
1463 <dt><b><tt>fcmp COND ( VAL1, VAL2 )</tt></b></dt>
1464 <dd>Performs the <a href="#i_fcmp">fcmp operation</a> on constants.</dd>
1466 <dt><b><tt>extractelement ( VAL, IDX )</tt></b></dt>
1468 <dd>Perform the <a href="#i_extractelement">extractelement
1469 operation</a> on constants.
1471 <dt><b><tt>insertelement ( VAL, ELT, IDX )</tt></b></dt>
1473 <dd>Perform the <a href="#i_insertelement">insertelement
1474 operation</a> on constants.</dd>
1477 <dt><b><tt>shufflevector ( VEC1, VEC2, IDXMASK )</tt></b></dt>
1479 <dd>Perform the <a href="#i_shufflevector">shufflevector
1480 operation</a> on constants.</dd>
1482 <dt><b><tt>OPCODE ( LHS, RHS )</tt></b></dt>
1484 <dd>Perform the specified operation of the LHS and RHS constants. OPCODE may
1485 be any of the <a href="#binaryops">binary</a> or <a href="#bitwiseops">bitwise
1486 binary</a> operations. The constraints on operands are the same as those for
1487 the corresponding instruction (e.g. no bitwise operations on floating point
1488 values are allowed).</dd>
1492 <!-- *********************************************************************** -->
1493 <div class="doc_section"> <a name="othervalues">Other Values</a> </div>
1494 <!-- *********************************************************************** -->
1496 <!-- ======================================================================= -->
1497 <div class="doc_subsection">
1498 <a name="inlineasm">Inline Assembler Expressions</a>
1501 <div class="doc_text">
1504 LLVM supports inline assembler expressions (as opposed to <a href="#moduleasm">
1505 Module-Level Inline Assembly</a>) through the use of a special value. This
1506 value represents the inline assembler as a string (containing the instructions
1507 to emit), a list of operand constraints (stored as a string), and a flag that
1508 indicates whether or not the inline asm expression has side effects. An example
1509 inline assembler expression is:
1513 i32 (i32) asm "bswap $0", "=r,r"
1517 Inline assembler expressions may <b>only</b> be used as the callee operand of
1518 a <a href="#i_call"><tt>call</tt> instruction</a>. Thus, typically we have:
1522 %X = call i32 asm "<a href="#int_bswap">bswap</a> $0", "=r,r"(i32 %Y)
1526 Inline asms with side effects not visible in the constraint list must be marked
1527 as having side effects. This is done through the use of the
1528 '<tt>sideeffect</tt>' keyword, like so:
1532 call void asm sideeffect "eieio", ""()
1535 <p>TODO: The format of the asm and constraints string still need to be
1536 documented here. Constraints on what can be done (e.g. duplication, moving, etc
1537 need to be documented).
1542 <!-- *********************************************************************** -->
1543 <div class="doc_section"> <a name="instref">Instruction Reference</a> </div>
1544 <!-- *********************************************************************** -->
1546 <div class="doc_text">
1548 <p>The LLVM instruction set consists of several different
1549 classifications of instructions: <a href="#terminators">terminator
1550 instructions</a>, <a href="#binaryops">binary instructions</a>,
1551 <a href="#bitwiseops">bitwise binary instructions</a>, <a
1552 href="#memoryops">memory instructions</a>, and <a href="#otherops">other
1553 instructions</a>.</p>
1557 <!-- ======================================================================= -->
1558 <div class="doc_subsection"> <a name="terminators">Terminator
1559 Instructions</a> </div>
1561 <div class="doc_text">
1563 <p>As mentioned <a href="#functionstructure">previously</a>, every
1564 basic block in a program ends with a "Terminator" instruction, which
1565 indicates which block should be executed after the current block is
1566 finished. These terminator instructions typically yield a '<tt>void</tt>'
1567 value: they produce control flow, not values (the one exception being
1568 the '<a href="#i_invoke"><tt>invoke</tt></a>' instruction).</p>
1569 <p>There are six different terminator instructions: the '<a
1570 href="#i_ret"><tt>ret</tt></a>' instruction, the '<a href="#i_br"><tt>br</tt></a>'
1571 instruction, the '<a href="#i_switch"><tt>switch</tt></a>' instruction,
1572 the '<a href="#i_invoke"><tt>invoke</tt></a>' instruction, the '<a
1573 href="#i_unwind"><tt>unwind</tt></a>' instruction, and the '<a
1574 href="#i_unreachable"><tt>unreachable</tt></a>' instruction.</p>
1578 <!-- _______________________________________________________________________ -->
1579 <div class="doc_subsubsection"> <a name="i_ret">'<tt>ret</tt>'
1580 Instruction</a> </div>
1581 <div class="doc_text">
1583 <pre> ret <type> <value> <i>; Return a value from a non-void function</i>
1584 ret void <i>; Return from void function</i>
1587 <p>The '<tt>ret</tt>' instruction is used to return control flow (and a
1588 value) from a function back to the caller.</p>
1589 <p>There are two forms of the '<tt>ret</tt>' instruction: one that
1590 returns a value and then causes control flow, and one that just causes
1591 control flow to occur.</p>
1593 <p>The '<tt>ret</tt>' instruction may return any '<a
1594 href="#t_firstclass">first class</a>' type. Notice that a function is
1595 not <a href="#wellformed">well formed</a> if there exists a '<tt>ret</tt>'
1596 instruction inside of the function that returns a value that does not
1597 match the return type of the function.</p>
1599 <p>When the '<tt>ret</tt>' instruction is executed, control flow
1600 returns back to the calling function's context. If the caller is a "<a
1601 href="#i_call"><tt>call</tt></a>" instruction, execution continues at
1602 the instruction after the call. If the caller was an "<a
1603 href="#i_invoke"><tt>invoke</tt></a>" instruction, execution continues
1604 at the beginning of the "normal" destination block. If the instruction
1605 returns a value, that value shall set the call or invoke instruction's
1608 <pre> ret i32 5 <i>; Return an integer value of 5</i>
1609 ret void <i>; Return from a void function</i>
1612 <!-- _______________________________________________________________________ -->
1613 <div class="doc_subsubsection"> <a name="i_br">'<tt>br</tt>' Instruction</a> </div>
1614 <div class="doc_text">
1616 <pre> br i1 <cond>, label <iftrue>, label <iffalse><br> br label <dest> <i>; Unconditional branch</i>
1619 <p>The '<tt>br</tt>' instruction is used to cause control flow to
1620 transfer to a different basic block in the current function. There are
1621 two forms of this instruction, corresponding to a conditional branch
1622 and an unconditional branch.</p>
1624 <p>The conditional branch form of the '<tt>br</tt>' instruction takes a
1625 single '<tt>i1</tt>' value and two '<tt>label</tt>' values. The
1626 unconditional form of the '<tt>br</tt>' instruction takes a single
1627 '<tt>label</tt>' value as a target.</p>
1629 <p>Upon execution of a conditional '<tt>br</tt>' instruction, the '<tt>i1</tt>'
1630 argument is evaluated. If the value is <tt>true</tt>, control flows
1631 to the '<tt>iftrue</tt>' <tt>label</tt> argument. If "cond" is <tt>false</tt>,
1632 control flows to the '<tt>iffalse</tt>' <tt>label</tt> argument.</p>
1634 <pre>Test:<br> %cond = <a href="#i_icmp">icmp</a> eq, i32 %a, %b<br> br i1 %cond, label %IfEqual, label %IfUnequal<br>IfEqual:<br> <a
1635 href="#i_ret">ret</a> i32 1<br>IfUnequal:<br> <a href="#i_ret">ret</a> i32 0<br></pre>
1637 <!-- _______________________________________________________________________ -->
1638 <div class="doc_subsubsection">
1639 <a name="i_switch">'<tt>switch</tt>' Instruction</a>
1642 <div class="doc_text">
1646 switch <intty> <value>, label <defaultdest> [ <intty> <val>, label <dest> ... ]
1651 <p>The '<tt>switch</tt>' instruction is used to transfer control flow to one of
1652 several different places. It is a generalization of the '<tt>br</tt>'
1653 instruction, allowing a branch to occur to one of many possible
1659 <p>The '<tt>switch</tt>' instruction uses three parameters: an integer
1660 comparison value '<tt>value</tt>', a default '<tt>label</tt>' destination, and
1661 an array of pairs of comparison value constants and '<tt>label</tt>'s. The
1662 table is not allowed to contain duplicate constant entries.</p>
1666 <p>The <tt>switch</tt> instruction specifies a table of values and
1667 destinations. When the '<tt>switch</tt>' instruction is executed, this
1668 table is searched for the given value. If the value is found, control flow is
1669 transfered to the corresponding destination; otherwise, control flow is
1670 transfered to the default destination.</p>
1672 <h5>Implementation:</h5>
1674 <p>Depending on properties of the target machine and the particular
1675 <tt>switch</tt> instruction, this instruction may be code generated in different
1676 ways. For example, it could be generated as a series of chained conditional
1677 branches or with a lookup table.</p>
1682 <i>; Emulate a conditional br instruction</i>
1683 %Val = <a href="#i_zext">zext</a> i1 %value to i32
1684 switch i32 %Val, label %truedest [i32 0, label %falsedest ]
1686 <i>; Emulate an unconditional br instruction</i>
1687 switch i32 0, label %dest [ ]
1689 <i>; Implement a jump table:</i>
1690 switch i32 %val, label %otherwise [ i32 0, label %onzero
1692 i32 2, label %ontwo ]
1696 <!-- _______________________________________________________________________ -->
1697 <div class="doc_subsubsection">
1698 <a name="i_invoke">'<tt>invoke</tt>' Instruction</a>
1701 <div class="doc_text">
1706 <result> = invoke [<a href="#callingconv">cconv</a>] <ptr to function ty> %<function ptr val>(<function args>)
1707 to label <normal label> unwind label <exception label>
1712 <p>The '<tt>invoke</tt>' instruction causes control to transfer to a specified
1713 function, with the possibility of control flow transfer to either the
1714 '<tt>normal</tt>' label or the
1715 '<tt>exception</tt>' label. If the callee function returns with the
1716 "<tt><a href="#i_ret">ret</a></tt>" instruction, control flow will return to the
1717 "normal" label. If the callee (or any indirect callees) returns with the "<a
1718 href="#i_unwind"><tt>unwind</tt></a>" instruction, control is interrupted and
1719 continued at the dynamically nearest "exception" label.</p>
1723 <p>This instruction requires several arguments:</p>
1727 The optional "cconv" marker indicates which <a href="#callingconv">calling
1728 convention</a> the call should use. If none is specified, the call defaults
1729 to using C calling conventions.
1731 <li>'<tt>ptr to function ty</tt>': shall be the signature of the pointer to
1732 function value being invoked. In most cases, this is a direct function
1733 invocation, but indirect <tt>invoke</tt>s are just as possible, branching off
1734 an arbitrary pointer to function value.
1737 <li>'<tt>function ptr val</tt>': An LLVM value containing a pointer to a
1738 function to be invoked. </li>
1740 <li>'<tt>function args</tt>': argument list whose types match the function
1741 signature argument types. If the function signature indicates the function
1742 accepts a variable number of arguments, the extra arguments can be
1745 <li>'<tt>normal label</tt>': the label reached when the called function
1746 executes a '<tt><a href="#i_ret">ret</a></tt>' instruction. </li>
1748 <li>'<tt>exception label</tt>': the label reached when a callee returns with
1749 the <a href="#i_unwind"><tt>unwind</tt></a> instruction. </li>
1755 <p>This instruction is designed to operate as a standard '<tt><a
1756 href="#i_call">call</a></tt>' instruction in most regards. The primary
1757 difference is that it establishes an association with a label, which is used by
1758 the runtime library to unwind the stack.</p>
1760 <p>This instruction is used in languages with destructors to ensure that proper
1761 cleanup is performed in the case of either a <tt>longjmp</tt> or a thrown
1762 exception. Additionally, this is important for implementation of
1763 '<tt>catch</tt>' clauses in high-level languages that support them.</p>
1767 %retval = invoke i32 %Test(i32 15) to label %Continue
1768 unwind label %TestCleanup <i>; {i32}:retval set</i>
1769 %retval = invoke <a href="#callingconv">coldcc</a> i32 %Test(i32 15) to label %Continue
1770 unwind label %TestCleanup <i>; {i32}:retval set</i>
1775 <!-- _______________________________________________________________________ -->
1777 <div class="doc_subsubsection"> <a name="i_unwind">'<tt>unwind</tt>'
1778 Instruction</a> </div>
1780 <div class="doc_text">
1789 <p>The '<tt>unwind</tt>' instruction unwinds the stack, continuing control flow
1790 at the first callee in the dynamic call stack which used an <a
1791 href="#i_invoke"><tt>invoke</tt></a> instruction to perform the call. This is
1792 primarily used to implement exception handling.</p>
1796 <p>The '<tt>unwind</tt>' intrinsic causes execution of the current function to
1797 immediately halt. The dynamic call stack is then searched for the first <a
1798 href="#i_invoke"><tt>invoke</tt></a> instruction on the call stack. Once found,
1799 execution continues at the "exceptional" destination block specified by the
1800 <tt>invoke</tt> instruction. If there is no <tt>invoke</tt> instruction in the
1801 dynamic call chain, undefined behavior results.</p>
1804 <!-- _______________________________________________________________________ -->
1806 <div class="doc_subsubsection"> <a name="i_unreachable">'<tt>unreachable</tt>'
1807 Instruction</a> </div>
1809 <div class="doc_text">
1818 <p>The '<tt>unreachable</tt>' instruction has no defined semantics. This
1819 instruction is used to inform the optimizer that a particular portion of the
1820 code is not reachable. This can be used to indicate that the code after a
1821 no-return function cannot be reached, and other facts.</p>
1825 <p>The '<tt>unreachable</tt>' instruction has no defined semantics.</p>
1830 <!-- ======================================================================= -->
1831 <div class="doc_subsection"> <a name="binaryops">Binary Operations</a> </div>
1832 <div class="doc_text">
1833 <p>Binary operators are used to do most of the computation in a
1834 program. They require two operands, execute an operation on them, and
1835 produce a single value. The operands might represent
1836 multiple data, as is the case with the <a href="#t_vector">vector</a> data type.
1837 The result value of a binary operator is not
1838 necessarily the same type as its operands.</p>
1839 <p>There are several different binary operators:</p>
1841 <!-- _______________________________________________________________________ -->
1842 <div class="doc_subsubsection"> <a name="i_add">'<tt>add</tt>'
1843 Instruction</a> </div>
1844 <div class="doc_text">
1846 <pre> <result> = add <ty> <var1>, <var2> <i>; yields {ty}:result</i>
1849 <p>The '<tt>add</tt>' instruction returns the sum of its two operands.</p>
1851 <p>The two arguments to the '<tt>add</tt>' instruction must be either <a
1852 href="#t_integer">integer</a> or <a href="#t_floating">floating point</a> values.
1853 This instruction can also take <a href="#t_vector">vector</a> versions of the values.
1854 Both arguments must have identical types.</p>
1856 <p>The value produced is the integer or floating point sum of the two
1859 <pre> <result> = add i32 4, %var <i>; yields {i32}:result = 4 + %var</i>
1862 <!-- _______________________________________________________________________ -->
1863 <div class="doc_subsubsection"> <a name="i_sub">'<tt>sub</tt>'
1864 Instruction</a> </div>
1865 <div class="doc_text">
1867 <pre> <result> = sub <ty> <var1>, <var2> <i>; yields {ty}:result</i>
1870 <p>The '<tt>sub</tt>' instruction returns the difference of its two
1872 <p>Note that the '<tt>sub</tt>' instruction is used to represent the '<tt>neg</tt>'
1873 instruction present in most other intermediate representations.</p>
1875 <p>The two arguments to the '<tt>sub</tt>' instruction must be either <a
1876 href="#t_integer">integer</a> or <a href="#t_floating">floating point</a>
1878 This instruction can also take <a href="#t_vector">vector</a> versions of the values.
1879 Both arguments must have identical types.</p>
1881 <p>The value produced is the integer or floating point difference of
1882 the two operands.</p>
1884 <pre> <result> = sub i32 4, %var <i>; yields {i32}:result = 4 - %var</i>
1885 <result> = sub i32 0, %val <i>; yields {i32}:result = -%var</i>
1888 <!-- _______________________________________________________________________ -->
1889 <div class="doc_subsubsection"> <a name="i_mul">'<tt>mul</tt>'
1890 Instruction</a> </div>
1891 <div class="doc_text">
1893 <pre> <result> = mul <ty> <var1>, <var2> <i>; yields {ty}:result</i>
1896 <p>The '<tt>mul</tt>' instruction returns the product of its two
1899 <p>The two arguments to the '<tt>mul</tt>' instruction must be either <a
1900 href="#t_integer">integer</a> or <a href="#t_floating">floating point</a>
1902 This instruction can also take <a href="#t_vector">vector</a> versions of the values.
1903 Both arguments must have identical types.</p>
1905 <p>The value produced is the integer or floating point product of the
1907 <p>Because the operands are the same width, the result of an integer
1908 multiplication is the same whether the operands should be deemed unsigned or
1911 <pre> <result> = mul i32 4, %var <i>; yields {i32}:result = 4 * %var</i>
1914 <!-- _______________________________________________________________________ -->
1915 <div class="doc_subsubsection"> <a name="i_udiv">'<tt>udiv</tt>' Instruction
1917 <div class="doc_text">
1919 <pre> <result> = udiv <ty> <var1>, <var2> <i>; yields {ty}:result</i>
1922 <p>The '<tt>udiv</tt>' instruction returns the quotient of its two
1925 <p>The two arguments to the '<tt>udiv</tt>' instruction must be
1926 <a href="#t_integer">integer</a> values. Both arguments must have identical
1927 types. This instruction can also take <a href="#t_vector">vector</a> versions
1928 of the values in which case the elements must be integers.</p>
1930 <p>The value produced is the unsigned integer quotient of the two operands. This
1931 instruction always performs an unsigned division operation, regardless of
1932 whether the arguments are unsigned or not.</p>
1934 <pre> <result> = udiv i32 4, %var <i>; yields {i32}:result = 4 / %var</i>
1937 <!-- _______________________________________________________________________ -->
1938 <div class="doc_subsubsection"> <a name="i_sdiv">'<tt>sdiv</tt>' Instruction
1940 <div class="doc_text">
1942 <pre> <result> = sdiv <ty> <var1>, <var2> <i>; yields {ty}:result</i>
1945 <p>The '<tt>sdiv</tt>' instruction returns the quotient of its two
1948 <p>The two arguments to the '<tt>sdiv</tt>' instruction must be
1949 <a href="#t_integer">integer</a> values. Both arguments must have identical
1950 types. This instruction can also take <a href="#t_vector">vector</a> versions
1951 of the values in which case the elements must be integers.</p>
1953 <p>The value produced is the signed integer quotient of the two operands. This
1954 instruction always performs a signed division operation, regardless of whether
1955 the arguments are signed or not.</p>
1957 <pre> <result> = sdiv i32 4, %var <i>; yields {i32}:result = 4 / %var</i>
1960 <!-- _______________________________________________________________________ -->
1961 <div class="doc_subsubsection"> <a name="i_fdiv">'<tt>fdiv</tt>'
1962 Instruction</a> </div>
1963 <div class="doc_text">
1965 <pre> <result> = fdiv <ty> <var1>, <var2> <i>; yields {ty}:result</i>
1968 <p>The '<tt>fdiv</tt>' instruction returns the quotient of its two
1971 <p>The two arguments to the '<tt>fdiv</tt>' instruction must be
1972 <a href="#t_floating">floating point</a> values. Both arguments must have
1973 identical types. This instruction can also take <a href="#t_vector">vector</a>
1974 versions of floating point values.</p>
1976 <p>The value produced is the floating point quotient of the two operands.</p>
1978 <pre> <result> = fdiv float 4.0, %var <i>; yields {float}:result = 4.0 / %var</i>
1981 <!-- _______________________________________________________________________ -->
1982 <div class="doc_subsubsection"> <a name="i_urem">'<tt>urem</tt>' Instruction</a>
1984 <div class="doc_text">
1986 <pre> <result> = urem <ty> <var1>, <var2> <i>; yields {ty}:result</i>
1989 <p>The '<tt>urem</tt>' instruction returns the remainder from the
1990 unsigned division of its two arguments.</p>
1992 <p>The two arguments to the '<tt>urem</tt>' instruction must be
1993 <a href="#t_integer">integer</a> values. Both arguments must have identical
1996 <p>This instruction returns the unsigned integer <i>remainder</i> of a division.
1997 This instruction always performs an unsigned division to get the remainder,
1998 regardless of whether the arguments are unsigned or not.</p>
2000 <pre> <result> = urem i32 4, %var <i>; yields {i32}:result = 4 % %var</i>
2004 <!-- _______________________________________________________________________ -->
2005 <div class="doc_subsubsection"> <a name="i_srem">'<tt>srem</tt>'
2006 Instruction</a> </div>
2007 <div class="doc_text">
2009 <pre> <result> = srem <ty> <var1>, <var2> <i>; yields {ty}:result</i>
2012 <p>The '<tt>srem</tt>' instruction returns the remainder from the
2013 signed division of its two operands.</p>
2015 <p>The two arguments to the '<tt>srem</tt>' instruction must be
2016 <a href="#t_integer">integer</a> values. Both arguments must have identical
2019 <p>This instruction returns the <i>remainder</i> of a division (where the result
2020 has the same sign as the dividend, <tt>var1</tt>), not the <i>modulo</i>
2021 operator (where the result has the same sign as the divisor, <tt>var2</tt>) of
2022 a value. For more information about the difference, see <a
2023 href="http://mathforum.org/dr.math/problems/anne.4.28.99.html">The
2024 Math Forum</a>. For a table of how this is implemented in various languages,
2025 please see <a href="http://en.wikipedia.org/wiki/Modulo_operation">
2026 Wikipedia: modulo operation</a>.</p>
2028 <pre> <result> = srem i32 4, %var <i>; yields {i32}:result = 4 % %var</i>
2032 <!-- _______________________________________________________________________ -->
2033 <div class="doc_subsubsection"> <a name="i_frem">'<tt>frem</tt>'
2034 Instruction</a> </div>
2035 <div class="doc_text">
2037 <pre> <result> = frem <ty> <var1>, <var2> <i>; yields {ty}:result</i>
2040 <p>The '<tt>frem</tt>' instruction returns the remainder from the
2041 division of its two operands.</p>
2043 <p>The two arguments to the '<tt>frem</tt>' instruction must be
2044 <a href="#t_floating">floating point</a> values. Both arguments must have
2045 identical types.</p>
2047 <p>This instruction returns the <i>remainder</i> of a division.</p>
2049 <pre> <result> = frem float 4.0, %var <i>; yields {float}:result = 4.0 % %var</i>
2053 <!-- ======================================================================= -->
2054 <div class="doc_subsection"> <a name="bitwiseops">Bitwise Binary
2055 Operations</a> </div>
2056 <div class="doc_text">
2057 <p>Bitwise binary operators are used to do various forms of
2058 bit-twiddling in a program. They are generally very efficient
2059 instructions and can commonly be strength reduced from other
2060 instructions. They require two operands, execute an operation on them,
2061 and produce a single value. The resulting value of the bitwise binary
2062 operators is always the same type as its first operand.</p>
2065 <!-- _______________________________________________________________________ -->
2066 <div class="doc_subsubsection"> <a name="i_shl">'<tt>shl</tt>'
2067 Instruction</a> </div>
2068 <div class="doc_text">
2070 <pre> <result> = shl <ty> <var1>, <var2> <i>; yields {ty}:result</i>
2073 <p>The '<tt>shl</tt>' instruction returns the first operand shifted to
2074 the left a specified number of bits.</p>
2076 <p>Both arguments to the '<tt>shl</tt>' instruction must be the same <a
2077 href="#t_integer">integer</a> type.</p>
2079 <p>The value produced is <tt>var1</tt> * 2<sup><tt>var2</tt></sup>.</p>
2080 <h5>Example:</h5><pre>
2081 <result> = shl i32 4, %var <i>; yields {i32}: 4 << %var</i>
2082 <result> = shl i32 4, 2 <i>; yields {i32}: 16</i>
2083 <result> = shl i32 1, 10 <i>; yields {i32}: 1024</i>
2086 <!-- _______________________________________________________________________ -->
2087 <div class="doc_subsubsection"> <a name="i_lshr">'<tt>lshr</tt>'
2088 Instruction</a> </div>
2089 <div class="doc_text">
2091 <pre> <result> = lshr <ty> <var1>, <var2> <i>; yields {ty}:result</i>
2095 <p>The '<tt>lshr</tt>' instruction (logical shift right) returns the first
2096 operand shifted to the right a specified number of bits with zero fill.</p>
2099 <p>Both arguments to the '<tt>lshr</tt>' instruction must be the same
2100 <a href="#t_integer">integer</a> type.</p>
2103 <p>This instruction always performs a logical shift right operation. The most
2104 significant bits of the result will be filled with zero bits after the
2109 <result> = lshr i32 4, 1 <i>; yields {i32}:result = 2</i>
2110 <result> = lshr i32 4, 2 <i>; yields {i32}:result = 1</i>
2111 <result> = lshr i8 4, 3 <i>; yields {i8}:result = 0</i>
2112 <result> = lshr i8 -2, 1 <i>; yields {i8}:result = 0x7FFFFFFF </i>
2116 <!-- _______________________________________________________________________ -->
2117 <div class="doc_subsubsection"> <a name="i_ashr">'<tt>ashr</tt>'
2118 Instruction</a> </div>
2119 <div class="doc_text">
2122 <pre> <result> = ashr <ty> <var1>, <var2> <i>; yields {ty}:result</i>
2126 <p>The '<tt>ashr</tt>' instruction (arithmetic shift right) returns the first
2127 operand shifted to the right a specified number of bits with sign extension.</p>
2130 <p>Both arguments to the '<tt>ashr</tt>' instruction must be the same
2131 <a href="#t_integer">integer</a> type.</p>
2134 <p>This instruction always performs an arithmetic shift right operation,
2135 The most significant bits of the result will be filled with the sign bit
2136 of <tt>var1</tt>.</p>
2140 <result> = ashr i32 4, 1 <i>; yields {i32}:result = 2</i>
2141 <result> = ashr i32 4, 2 <i>; yields {i32}:result = 1</i>
2142 <result> = ashr i8 4, 3 <i>; yields {i8}:result = 0</i>
2143 <result> = ashr i8 -2, 1 <i>; yields {i8}:result = -1</i>
2147 <!-- _______________________________________________________________________ -->
2148 <div class="doc_subsubsection"> <a name="i_and">'<tt>and</tt>'
2149 Instruction</a> </div>
2150 <div class="doc_text">
2152 <pre> <result> = and <ty> <var1>, <var2> <i>; yields {ty}:result</i>
2155 <p>The '<tt>and</tt>' instruction returns the bitwise logical and of
2156 its two operands.</p>
2158 <p>The two arguments to the '<tt>and</tt>' instruction must be <a
2159 href="#t_integer">integer</a> values. Both arguments must have
2160 identical types.</p>
2162 <p>The truth table used for the '<tt>and</tt>' instruction is:</p>
2164 <div style="align: center">
2165 <table border="1" cellspacing="0" cellpadding="4">
2196 <pre> <result> = and i32 4, %var <i>; yields {i32}:result = 4 & %var</i>
2197 <result> = and i32 15, 40 <i>; yields {i32}:result = 8</i>
2198 <result> = and i32 4, 8 <i>; yields {i32}:result = 0</i>
2201 <!-- _______________________________________________________________________ -->
2202 <div class="doc_subsubsection"> <a name="i_or">'<tt>or</tt>' Instruction</a> </div>
2203 <div class="doc_text">
2205 <pre> <result> = or <ty> <var1>, <var2> <i>; yields {ty}:result</i>
2208 <p>The '<tt>or</tt>' instruction returns the bitwise logical inclusive
2209 or of its two operands.</p>
2211 <p>The two arguments to the '<tt>or</tt>' instruction must be <a
2212 href="#t_integer">integer</a> values. Both arguments must have
2213 identical types.</p>
2215 <p>The truth table used for the '<tt>or</tt>' instruction is:</p>
2217 <div style="align: center">
2218 <table border="1" cellspacing="0" cellpadding="4">
2249 <pre> <result> = or i32 4, %var <i>; yields {i32}:result = 4 | %var</i>
2250 <result> = or i32 15, 40 <i>; yields {i32}:result = 47</i>
2251 <result> = or i32 4, 8 <i>; yields {i32}:result = 12</i>
2254 <!-- _______________________________________________________________________ -->
2255 <div class="doc_subsubsection"> <a name="i_xor">'<tt>xor</tt>'
2256 Instruction</a> </div>
2257 <div class="doc_text">
2259 <pre> <result> = xor <ty> <var1>, <var2> <i>; yields {ty}:result</i>
2262 <p>The '<tt>xor</tt>' instruction returns the bitwise logical exclusive
2263 or of its two operands. The <tt>xor</tt> is used to implement the
2264 "one's complement" operation, which is the "~" operator in C.</p>
2266 <p>The two arguments to the '<tt>xor</tt>' instruction must be <a
2267 href="#t_integer">integer</a> values. Both arguments must have
2268 identical types.</p>
2270 <p>The truth table used for the '<tt>xor</tt>' instruction is:</p>
2272 <div style="align: center">
2273 <table border="1" cellspacing="0" cellpadding="4">
2305 <pre> <result> = xor i32 4, %var <i>; yields {i32}:result = 4 ^ %var</i>
2306 <result> = xor i32 15, 40 <i>; yields {i32}:result = 39</i>
2307 <result> = xor i32 4, 8 <i>; yields {i32}:result = 12</i>
2308 <result> = xor i32 %V, -1 <i>; yields {i32}:result = ~%V</i>
2312 <!-- ======================================================================= -->
2313 <div class="doc_subsection">
2314 <a name="vectorops">Vector Operations</a>
2317 <div class="doc_text">
2319 <p>LLVM supports several instructions to represent vector operations in a
2320 target-independent manner. These instructions cover the element-access and
2321 vector-specific operations needed to process vectors effectively. While LLVM
2322 does directly support these vector operations, many sophisticated algorithms
2323 will want to use target-specific intrinsics to take full advantage of a specific
2328 <!-- _______________________________________________________________________ -->
2329 <div class="doc_subsubsection">
2330 <a name="i_extractelement">'<tt>extractelement</tt>' Instruction</a>
2333 <div class="doc_text">
2338 <result> = extractelement <n x <ty>> <val>, i32 <idx> <i>; yields <ty></i>
2344 The '<tt>extractelement</tt>' instruction extracts a single scalar
2345 element from a vector at a specified index.
2352 The first operand of an '<tt>extractelement</tt>' instruction is a
2353 value of <a href="#t_vector">vector</a> type. The second operand is
2354 an index indicating the position from which to extract the element.
2355 The index may be a variable.</p>
2360 The result is a scalar of the same type as the element type of
2361 <tt>val</tt>. Its value is the value at position <tt>idx</tt> of
2362 <tt>val</tt>. If <tt>idx</tt> exceeds the length of <tt>val</tt>, the
2363 results are undefined.
2369 %result = extractelement <4 x i32> %vec, i32 0 <i>; yields i32</i>
2374 <!-- _______________________________________________________________________ -->
2375 <div class="doc_subsubsection">
2376 <a name="i_insertelement">'<tt>insertelement</tt>' Instruction</a>
2379 <div class="doc_text">
2384 <result> = insertelement <n x <ty>> <val>, <ty> <elt>, i32 <idx> <i>; yields <n x <ty>></i>
2390 The '<tt>insertelement</tt>' instruction inserts a scalar
2391 element into a vector at a specified index.
2398 The first operand of an '<tt>insertelement</tt>' instruction is a
2399 value of <a href="#t_vector">vector</a> type. The second operand is a
2400 scalar value whose type must equal the element type of the first
2401 operand. The third operand is an index indicating the position at
2402 which to insert the value. The index may be a variable.</p>
2407 The result is a vector of the same type as <tt>val</tt>. Its
2408 element values are those of <tt>val</tt> except at position
2409 <tt>idx</tt>, where it gets the value <tt>elt</tt>. If <tt>idx</tt>
2410 exceeds the length of <tt>val</tt>, the results are undefined.
2416 %result = insertelement <4 x i32> %vec, i32 1, i32 0 <i>; yields <4 x i32></i>
2420 <!-- _______________________________________________________________________ -->
2421 <div class="doc_subsubsection">
2422 <a name="i_shufflevector">'<tt>shufflevector</tt>' Instruction</a>
2425 <div class="doc_text">
2430 <result> = shufflevector <n x <ty>> <v1>, <n x <ty>> <v2>, <n x i32> <mask> <i>; yields <n x <ty>></i>
2436 The '<tt>shufflevector</tt>' instruction constructs a permutation of elements
2437 from two input vectors, returning a vector of the same type.
2443 The first two operands of a '<tt>shufflevector</tt>' instruction are vectors
2444 with types that match each other and types that match the result of the
2445 instruction. The third argument is a shuffle mask, which has the same number
2446 of elements as the other vector type, but whose element type is always 'i32'.
2450 The shuffle mask operand is required to be a constant vector with either
2451 constant integer or undef values.
2457 The elements of the two input vectors are numbered from left to right across
2458 both of the vectors. The shuffle mask operand specifies, for each element of
2459 the result vector, which element of the two input registers the result element
2460 gets. The element selector may be undef (meaning "don't care") and the second
2461 operand may be undef if performing a shuffle from only one vector.
2467 %result = shufflevector <4 x i32> %v1, <4 x i32> %v2,
2468 <4 x i32> <i32 0, i32 4, i32 1, i32 5> <i>; yields <4 x i32></i>
2469 %result = shufflevector <4 x i32> %v1, <4 x i32> undef,
2470 <4 x i32> <i32 0, i32 1, i32 2, i32 3> <i>; yields <4 x i32></i> - Identity shuffle.
2475 <!-- ======================================================================= -->
2476 <div class="doc_subsection">
2477 <a name="memoryops">Memory Access and Addressing Operations</a>
2480 <div class="doc_text">
2482 <p>A key design point of an SSA-based representation is how it
2483 represents memory. In LLVM, no memory locations are in SSA form, which
2484 makes things very simple. This section describes how to read, write,
2485 allocate, and free memory in LLVM.</p>
2489 <!-- _______________________________________________________________________ -->
2490 <div class="doc_subsubsection">
2491 <a name="i_malloc">'<tt>malloc</tt>' Instruction</a>
2494 <div class="doc_text">
2499 <result> = malloc <type>[, i32 <NumElements>][, align <alignment>] <i>; yields {type*}:result</i>
2504 <p>The '<tt>malloc</tt>' instruction allocates memory from the system
2505 heap and returns a pointer to it.</p>
2509 <p>The '<tt>malloc</tt>' instruction allocates
2510 <tt>sizeof(<type>)*NumElements</tt>
2511 bytes of memory from the operating system and returns a pointer of the
2512 appropriate type to the program. If "NumElements" is specified, it is the
2513 number of elements allocated. If an alignment is specified, the value result
2514 of the allocation is guaranteed to be aligned to at least that boundary. If
2515 not specified, or if zero, the target can choose to align the allocation on any
2516 convenient boundary.</p>
2518 <p>'<tt>type</tt>' must be a sized type.</p>
2522 <p>Memory is allocated using the system "<tt>malloc</tt>" function, and
2523 a pointer is returned.</p>
2528 %array = malloc [4 x i8 ] <i>; yields {[%4 x i8]*}:array</i>
2530 %size = <a href="#i_add">add</a> i32 2, 2 <i>; yields {i32}:size = i32 4</i>
2531 %array1 = malloc i8, i32 4 <i>; yields {i8*}:array1</i>
2532 %array2 = malloc [12 x i8], i32 %size <i>; yields {[12 x i8]*}:array2</i>
2533 %array3 = malloc i32, i32 4, align 1024 <i>; yields {i32*}:array3</i>
2534 %array4 = malloc i32, align 1024 <i>; yields {i32*}:array4</i>
2538 <!-- _______________________________________________________________________ -->
2539 <div class="doc_subsubsection">
2540 <a name="i_free">'<tt>free</tt>' Instruction</a>
2543 <div class="doc_text">
2548 free <type> <value> <i>; yields {void}</i>
2553 <p>The '<tt>free</tt>' instruction returns memory back to the unused
2554 memory heap to be reallocated in the future.</p>
2558 <p>'<tt>value</tt>' shall be a pointer value that points to a value
2559 that was allocated with the '<tt><a href="#i_malloc">malloc</a></tt>'
2564 <p>Access to the memory pointed to by the pointer is no longer defined
2565 after this instruction executes.</p>
2570 %array = <a href="#i_malloc">malloc</a> [4 x i8] <i>; yields {[4 x i8]*}:array</i>
2571 free [4 x i8]* %array
2575 <!-- _______________________________________________________________________ -->
2576 <div class="doc_subsubsection">
2577 <a name="i_alloca">'<tt>alloca</tt>' Instruction</a>
2580 <div class="doc_text">
2585 <result> = alloca <type>[, i32 <NumElements>][, align <alignment>] <i>; yields {type*}:result</i>
2590 <p>The '<tt>alloca</tt>' instruction allocates memory on the stack frame of the
2591 currently executing function, to be automatically released when this function
2592 returns to its caller.</p>
2596 <p>The '<tt>alloca</tt>' instruction allocates <tt>sizeof(<type>)*NumElements</tt>
2597 bytes of memory on the runtime stack, returning a pointer of the
2598 appropriate type to the program. If "NumElements" is specified, it is the
2599 number of elements allocated. If an alignment is specified, the value result
2600 of the allocation is guaranteed to be aligned to at least that boundary. If
2601 not specified, or if zero, the target can choose to align the allocation on any
2602 convenient boundary.</p>
2604 <p>'<tt>type</tt>' may be any sized type.</p>
2608 <p>Memory is allocated; a pointer is returned. '<tt>alloca</tt>'d
2609 memory is automatically released when the function returns. The '<tt>alloca</tt>'
2610 instruction is commonly used to represent automatic variables that must
2611 have an address available. When the function returns (either with the <tt><a
2612 href="#i_ret">ret</a></tt> or <tt><a href="#i_unwind">unwind</a></tt>
2613 instructions), the memory is reclaimed.</p>
2618 %ptr = alloca i32 <i>; yields {i32*}:ptr</i>
2619 %ptr = alloca i32, i32 4 <i>; yields {i32*}:ptr</i>
2620 %ptr = alloca i32, i32 4, align 1024 <i>; yields {i32*}:ptr</i>
2621 %ptr = alloca i32, align 1024 <i>; yields {i32*}:ptr</i>
2625 <!-- _______________________________________________________________________ -->
2626 <div class="doc_subsubsection"> <a name="i_load">'<tt>load</tt>'
2627 Instruction</a> </div>
2628 <div class="doc_text">
2630 <pre> <result> = load <ty>* <pointer>[, align <alignment>]<br> <result> = volatile load <ty>* <pointer>[, align <alignment>]<br></pre>
2632 <p>The '<tt>load</tt>' instruction is used to read from memory.</p>
2634 <p>The argument to the '<tt>load</tt>' instruction specifies the memory
2635 address from which to load. The pointer must point to a <a
2636 href="#t_firstclass">first class</a> type. If the <tt>load</tt> is
2637 marked as <tt>volatile</tt>, then the optimizer is not allowed to modify
2638 the number or order of execution of this <tt>load</tt> with other
2639 volatile <tt>load</tt> and <tt><a href="#i_store">store</a></tt>
2642 <p>The location of memory pointed to is loaded.</p>
2644 <pre> %ptr = <a href="#i_alloca">alloca</a> i32 <i>; yields {i32*}:ptr</i>
2646 href="#i_store">store</a> i32 3, i32* %ptr <i>; yields {void}</i>
2647 %val = load i32* %ptr <i>; yields {i32}:val = i32 3</i>
2650 <!-- _______________________________________________________________________ -->
2651 <div class="doc_subsubsection"> <a name="i_store">'<tt>store</tt>'
2652 Instruction</a> </div>
2653 <div class="doc_text">
2655 <pre> store <ty> <value>, <ty>* <pointer>[, align <alignment>] <i>; yields {void}</i>
2656 volatile store <ty> <value>, <ty>* <pointer>[, align <alignment>] <i>; yields {void}</i>
2659 <p>The '<tt>store</tt>' instruction is used to write to memory.</p>
2661 <p>There are two arguments to the '<tt>store</tt>' instruction: a value
2662 to store and an address at which to store it. The type of the '<tt><pointer></tt>'
2663 operand must be a pointer to the type of the '<tt><value></tt>'
2664 operand. If the <tt>store</tt> is marked as <tt>volatile</tt>, then the
2665 optimizer is not allowed to modify the number or order of execution of
2666 this <tt>store</tt> with other volatile <tt>load</tt> and <tt><a
2667 href="#i_store">store</a></tt> instructions.</p>
2669 <p>The contents of memory are updated to contain '<tt><value></tt>'
2670 at the location specified by the '<tt><pointer></tt>' operand.</p>
2672 <pre> %ptr = <a href="#i_alloca">alloca</a> i32 <i>; yields {i32*}:ptr</i>
2674 href="#i_store">store</a> i32 3, i32* %ptr <i>; yields {void}</i>
2675 %val = load i32* %ptr <i>; yields {i32}:val = i32 3</i>
2679 <!-- _______________________________________________________________________ -->
2680 <div class="doc_subsubsection">
2681 <a name="i_getelementptr">'<tt>getelementptr</tt>' Instruction</a>
2684 <div class="doc_text">
2687 <result> = getelementptr <ty>* <ptrval>{, <ty> <idx>}*
2693 The '<tt>getelementptr</tt>' instruction is used to get the address of a
2694 subelement of an aggregate data structure.</p>
2698 <p>This instruction takes a list of integer operands that indicate what
2699 elements of the aggregate object to index to. The actual types of the arguments
2700 provided depend on the type of the first pointer argument. The
2701 '<tt>getelementptr</tt>' instruction is used to index down through the type
2702 levels of a structure or to a specific index in an array. When indexing into a
2703 structure, only <tt>i32</tt> integer constants are allowed. When indexing
2704 into an array or pointer, only integers of 32 or 64 bits are allowed, and will
2705 be sign extended to 64-bit values.</p>
2707 <p>For example, let's consider a C code fragment and how it gets
2708 compiled to LLVM:</p>
2722 define i32 *foo(struct ST *s) {
2723 return &s[1].Z.B[5][13];
2727 <p>The LLVM code generated by the GCC frontend is:</p>
2730 %RT = type { i8 , [10 x [20 x i32]], i8 }
2731 %ST = type { i32, double, %RT }
2733 define i32* %foo(%ST* %s) {
2735 %reg = getelementptr %ST* %s, i32 1, i32 2, i32 1, i32 5, i32 13
2742 <p>The index types specified for the '<tt>getelementptr</tt>' instruction depend
2743 on the pointer type that is being indexed into. <a href="#t_pointer">Pointer</a>
2744 and <a href="#t_array">array</a> types can use a 32-bit or 64-bit
2745 <a href="#t_integer">integer</a> type but the value will always be sign extended
2746 to 64-bits. <a href="#t_struct">Structure</a> types require <tt>i32</tt>
2747 <b>constants</b>.</p>
2749 <p>In the example above, the first index is indexing into the '<tt>%ST*</tt>'
2750 type, which is a pointer, yielding a '<tt>%ST</tt>' = '<tt>{ i32, double, %RT
2751 }</tt>' type, a structure. The second index indexes into the third element of
2752 the structure, yielding a '<tt>%RT</tt>' = '<tt>{ i8 , [10 x [20 x i32]],
2753 i8 }</tt>' type, another structure. The third index indexes into the second
2754 element of the structure, yielding a '<tt>[10 x [20 x i32]]</tt>' type, an
2755 array. The two dimensions of the array are subscripted into, yielding an
2756 '<tt>i32</tt>' type. The '<tt>getelementptr</tt>' instruction returns a pointer
2757 to this element, thus computing a value of '<tt>i32*</tt>' type.</p>
2759 <p>Note that it is perfectly legal to index partially through a
2760 structure, returning a pointer to an inner element. Because of this,
2761 the LLVM code for the given testcase is equivalent to:</p>
2764 define i32* %foo(%ST* %s) {
2765 %t1 = getelementptr %ST* %s, i32 1 <i>; yields %ST*:%t1</i>
2766 %t2 = getelementptr %ST* %t1, i32 0, i32 2 <i>; yields %RT*:%t2</i>
2767 %t3 = getelementptr %RT* %t2, i32 0, i32 1 <i>; yields [10 x [20 x i32]]*:%t3</i>
2768 %t4 = getelementptr [10 x [20 x i32]]* %t3, i32 0, i32 5 <i>; yields [20 x i32]*:%t4</i>
2769 %t5 = getelementptr [20 x i32]* %t4, i32 0, i32 13 <i>; yields i32*:%t5</i>
2774 <p>Note that it is undefined to access an array out of bounds: array and
2775 pointer indexes must always be within the defined bounds of the array type.
2776 The one exception for this rules is zero length arrays. These arrays are
2777 defined to be accessible as variable length arrays, which requires access
2778 beyond the zero'th element.</p>
2780 <p>The getelementptr instruction is often confusing. For some more insight
2781 into how it works, see <a href="GetElementPtr.html">the getelementptr
2787 <i>; yields [12 x i8]*:aptr</i>
2788 %aptr = getelementptr {i32, [12 x i8]}* %sptr, i64 0, i32 1
2792 <!-- ======================================================================= -->
2793 <div class="doc_subsection"> <a name="convertops">Conversion Operations</a>
2795 <div class="doc_text">
2796 <p>The instructions in this category are the conversion instructions (casting)
2797 which all take a single operand and a type. They perform various bit conversions
2801 <!-- _______________________________________________________________________ -->
2802 <div class="doc_subsubsection">
2803 <a name="i_trunc">'<tt>trunc .. to</tt>' Instruction</a>
2805 <div class="doc_text">
2809 <result> = trunc <ty> <value> to <ty2> <i>; yields ty2</i>
2814 The '<tt>trunc</tt>' instruction truncates its operand to the type <tt>ty2</tt>.
2819 The '<tt>trunc</tt>' instruction takes a <tt>value</tt> to trunc, which must
2820 be an <a href="#t_integer">integer</a> type, and a type that specifies the size
2821 and type of the result, which must be an <a href="#t_integer">integer</a>
2822 type. The bit size of <tt>value</tt> must be larger than the bit size of
2823 <tt>ty2</tt>. Equal sized types are not allowed.</p>
2827 The '<tt>trunc</tt>' instruction truncates the high order bits in <tt>value</tt>
2828 and converts the remaining bits to <tt>ty2</tt>. Since the source size must be
2829 larger than the destination size, <tt>trunc</tt> cannot be a <i>no-op cast</i>.
2830 It will always truncate bits.</p>
2834 %X = trunc i32 257 to i8 <i>; yields i8:1</i>
2835 %Y = trunc i32 123 to i1 <i>; yields i1:true</i>
2836 %Y = trunc i32 122 to i1 <i>; yields i1:false</i>
2840 <!-- _______________________________________________________________________ -->
2841 <div class="doc_subsubsection">
2842 <a name="i_zext">'<tt>zext .. to</tt>' Instruction</a>
2844 <div class="doc_text">
2848 <result> = zext <ty> <value> to <ty2> <i>; yields ty2</i>
2852 <p>The '<tt>zext</tt>' instruction zero extends its operand to type
2857 <p>The '<tt>zext</tt>' instruction takes a value to cast, which must be of
2858 <a href="#t_integer">integer</a> type, and a type to cast it to, which must
2859 also be of <a href="#t_integer">integer</a> type. The bit size of the
2860 <tt>value</tt> must be smaller than the bit size of the destination type,
2864 <p>The <tt>zext</tt> fills the high order bits of the <tt>value</tt> with zero
2865 bits until it reaches the size of the destination type, <tt>ty2</tt>. When the
2866 the operand and the type are the same size, no bit filling is done and the
2867 cast is considered a <i>no-op cast</i> because no bits change (only the type
2870 <p>When zero extending from i1, the result will always be either 0 or 1.</p>
2874 %X = zext i32 257 to i64 <i>; yields i64:257</i>
2875 %Y = zext i1 true to i32 <i>; yields i32:1</i>
2879 <!-- _______________________________________________________________________ -->
2880 <div class="doc_subsubsection">
2881 <a name="i_sext">'<tt>sext .. to</tt>' Instruction</a>
2883 <div class="doc_text">
2887 <result> = sext <ty> <value> to <ty2> <i>; yields ty2</i>
2891 <p>The '<tt>sext</tt>' sign extends <tt>value</tt> to the type <tt>ty2</tt>.</p>
2895 The '<tt>sext</tt>' instruction takes a value to cast, which must be of
2896 <a href="#t_integer">integer</a> type, and a type to cast it to, which must
2897 also be of <a href="#t_integer">integer</a> type. The bit size of the
2898 <tt>value</tt> must be smaller than the bit size of the destination type,
2903 The '<tt>sext</tt>' instruction performs a sign extension by copying the sign
2904 bit (highest order bit) of the <tt>value</tt> until it reaches the bit size of
2905 the type <tt>ty2</tt>. When the the operand and the type are the same size,
2906 no bit filling is done and the cast is considered a <i>no-op cast</i> because
2907 no bits change (only the type changes).</p>
2909 <p>When sign extending from i1, the extension always results in -1 or 0.</p>
2913 %X = sext i8 -1 to i16 <i>; yields i16 :65535</i>
2914 %Y = sext i1 true to i32 <i>; yields i32:-1</i>
2918 <!-- _______________________________________________________________________ -->
2919 <div class="doc_subsubsection">
2920 <a name="i_fptrunc">'<tt>fptrunc .. to</tt>' Instruction</a>
2923 <div class="doc_text">
2928 <result> = fptrunc <ty> <value> to <ty2> <i>; yields ty2</i>
2932 <p>The '<tt>fptrunc</tt>' instruction truncates <tt>value</tt> to type
2937 <p>The '<tt>fptrunc</tt>' instruction takes a <a href="#t_floating">floating
2938 point</a> value to cast and a <a href="#t_floating">floating point</a> type to
2939 cast it to. The size of <tt>value</tt> must be larger than the size of
2940 <tt>ty2</tt>. This implies that <tt>fptrunc</tt> cannot be used to make a
2941 <i>no-op cast</i>.</p>
2944 <p> The '<tt>fptrunc</tt>' instruction truncates a <tt>value</tt> from a larger
2945 <a href="#t_floating">floating point</a> type to a smaller
2946 <a href="#t_floating">floating point</a> type. If the value cannot fit within
2947 the destination type, <tt>ty2</tt>, then the results are undefined.</p>
2951 %X = fptrunc double 123.0 to float <i>; yields float:123.0</i>
2952 %Y = fptrunc double 1.0E+300 to float <i>; yields undefined</i>
2956 <!-- _______________________________________________________________________ -->
2957 <div class="doc_subsubsection">
2958 <a name="i_fpext">'<tt>fpext .. to</tt>' Instruction</a>
2960 <div class="doc_text">
2964 <result> = fpext <ty> <value> to <ty2> <i>; yields ty2</i>
2968 <p>The '<tt>fpext</tt>' extends a floating point <tt>value</tt> to a larger
2969 floating point value.</p>
2972 <p>The '<tt>fpext</tt>' instruction takes a
2973 <a href="#t_floating">floating point</a> <tt>value</tt> to cast,
2974 and a <a href="#t_floating">floating point</a> type to cast it to. The source
2975 type must be smaller than the destination type.</p>
2978 <p>The '<tt>fpext</tt>' instruction extends the <tt>value</tt> from a smaller
2979 <a href="#t_floating">floating point</a> type to a larger
2980 <a href="#t_floating">floating point</a> type. The <tt>fpext</tt> cannot be
2981 used to make a <i>no-op cast</i> because it always changes bits. Use
2982 <tt>bitcast</tt> to make a <i>no-op cast</i> for a floating point cast.</p>
2986 %X = fpext float 3.1415 to double <i>; yields double:3.1415</i>
2987 %Y = fpext float 1.0 to float <i>; yields float:1.0 (no-op)</i>
2991 <!-- _______________________________________________________________________ -->
2992 <div class="doc_subsubsection">
2993 <a name="i_fptoui">'<tt>fptoui .. to</tt>' Instruction</a>
2995 <div class="doc_text">
2999 <result> = fp2uint <ty> <value> to <ty2> <i>; yields ty2</i>
3003 <p>The '<tt>fp2uint</tt>' converts a floating point <tt>value</tt> to its
3004 unsigned integer equivalent of type <tt>ty2</tt>.
3008 <p>The '<tt>fp2uint</tt>' instruction takes a value to cast, which must be a
3009 <a href="#t_floating">floating point</a> value, and a type to cast it to, which
3010 must be an <a href="#t_integer">integer</a> type.</p>
3013 <p> The '<tt>fp2uint</tt>' instruction converts its
3014 <a href="#t_floating">floating point</a> operand into the nearest (rounding
3015 towards zero) unsigned integer value. If the value cannot fit in <tt>ty2</tt>,
3016 the results are undefined.</p>
3018 <p>When converting to i1, the conversion is done as a comparison against
3019 zero. If the <tt>value</tt> was zero, the i1 result will be <tt>false</tt>.
3020 If the <tt>value</tt> was non-zero, the i1 result will be <tt>true</tt>.</p>
3024 %X = fp2uint double 123.0 to i32 <i>; yields i32:123</i>
3025 %Y = fp2uint float 1.0E+300 to i1 <i>; yields i1:true</i>
3026 %X = fp2uint float 1.04E+17 to i8 <i>; yields undefined:1</i>
3030 <!-- _______________________________________________________________________ -->
3031 <div class="doc_subsubsection">
3032 <a name="i_fptosi">'<tt>fptosi .. to</tt>' Instruction</a>
3034 <div class="doc_text">
3038 <result> = fptosi <ty> <value> to <ty2> <i>; yields ty2</i>
3042 <p>The '<tt>fptosi</tt>' instruction converts
3043 <a href="#t_floating">floating point</a> <tt>value</tt> to type <tt>ty2</tt>.
3048 <p> The '<tt>fptosi</tt>' instruction takes a value to cast, which must be a
3049 <a href="#t_floating">floating point</a> value, and a type to cast it to, which
3050 must also be an <a href="#t_integer">integer</a> type.</p>
3053 <p>The '<tt>fptosi</tt>' instruction converts its
3054 <a href="#t_floating">floating point</a> operand into the nearest (rounding
3055 towards zero) signed integer value. If the value cannot fit in <tt>ty2</tt>,
3056 the results are undefined.</p>
3058 <p>When converting to i1, the conversion is done as a comparison against
3059 zero. If the <tt>value</tt> was zero, the i1 result will be <tt>false</tt>.
3060 If the <tt>value</tt> was non-zero, the i1 result will be <tt>true</tt>.</p>
3064 %X = fptosi double -123.0 to i32 <i>; yields i32:-123</i>
3065 %Y = fptosi float 1.0E-247 to i1 <i>; yields i1:true</i>
3066 %X = fptosi float 1.04E+17 to i8 <i>; yields undefined:1</i>
3070 <!-- _______________________________________________________________________ -->
3071 <div class="doc_subsubsection">
3072 <a name="i_uitofp">'<tt>uitofp .. to</tt>' Instruction</a>
3074 <div class="doc_text">
3078 <result> = uitofp <ty> <value> to <ty2> <i>; yields ty2</i>
3082 <p>The '<tt>uitofp</tt>' instruction regards <tt>value</tt> as an unsigned
3083 integer and converts that value to the <tt>ty2</tt> type.</p>
3087 <p>The '<tt>uitofp</tt>' instruction takes a value to cast, which must be an
3088 <a href="#t_integer">integer</a> value, and a type to cast it to, which must
3089 be a <a href="#t_floating">floating point</a> type.</p>
3092 <p>The '<tt>uitofp</tt>' instruction interprets its operand as an unsigned
3093 integer quantity and converts it to the corresponding floating point value. If
3094 the value cannot fit in the floating point value, the results are undefined.</p>
3099 %X = uitofp i32 257 to float <i>; yields float:257.0</i>
3100 %Y = uitofp i8 -1 to double <i>; yields double:255.0</i>
3104 <!-- _______________________________________________________________________ -->
3105 <div class="doc_subsubsection">
3106 <a name="i_sitofp">'<tt>sitofp .. to</tt>' Instruction</a>
3108 <div class="doc_text">
3112 <result> = sitofp <ty> <value> to <ty2> <i>; yields ty2</i>
3116 <p>The '<tt>sitofp</tt>' instruction regards <tt>value</tt> as a signed
3117 integer and converts that value to the <tt>ty2</tt> type.</p>
3120 <p>The '<tt>sitofp</tt>' instruction takes a value to cast, which must be an
3121 <a href="#t_integer">integer</a> value, and a type to cast it to, which must be
3122 a <a href="#t_floating">floating point</a> type.</p>
3125 <p>The '<tt>sitofp</tt>' instruction interprets its operand as a signed
3126 integer quantity and converts it to the corresponding floating point value. If
3127 the value cannot fit in the floating point value, the results are undefined.</p>
3131 %X = sitofp i32 257 to float <i>; yields float:257.0</i>
3132 %Y = sitofp i8 -1 to double <i>; yields double:-1.0</i>
3136 <!-- _______________________________________________________________________ -->
3137 <div class="doc_subsubsection">
3138 <a name="i_ptrtoint">'<tt>ptrtoint .. to</tt>' Instruction</a>
3140 <div class="doc_text">
3144 <result> = ptrtoint <ty> <value> to <ty2> <i>; yields ty2</i>
3148 <p>The '<tt>ptrtoint</tt>' instruction converts the pointer <tt>value</tt> to
3149 the integer type <tt>ty2</tt>.</p>
3152 <p>The '<tt>ptrtoint</tt>' instruction takes a <tt>value</tt> to cast, which
3153 must be a <a href="#t_pointer">pointer</a> value, and a type to cast it to
3154 <tt>ty2</tt>, which must be an <a href="#t_integer">integer</a> type.
3157 <p>The '<tt>ptrtoint</tt>' instruction converts <tt>value</tt> to integer type
3158 <tt>ty2</tt> by interpreting the pointer value as an integer and either
3159 truncating or zero extending that value to the size of the integer type. If
3160 <tt>value</tt> is smaller than <tt>ty2</tt> then a zero extension is done. If
3161 <tt>value</tt> is larger than <tt>ty2</tt> then a truncation is done. If they
3162 are the same size, then nothing is done (<i>no-op cast</i>) other than a type
3167 %X = ptrtoint i32* %X to i8 <i>; yields truncation on 32-bit architecture</i>
3168 %Y = ptrtoint i32* %x to i64 <i>; yields zero extension on 32-bit architecture</i>
3172 <!-- _______________________________________________________________________ -->
3173 <div class="doc_subsubsection">
3174 <a name="i_inttoptr">'<tt>inttoptr .. to</tt>' Instruction</a>
3176 <div class="doc_text">
3180 <result> = inttoptr <ty> <value> to <ty2> <i>; yields ty2</i>
3184 <p>The '<tt>inttoptr</tt>' instruction converts an integer <tt>value</tt> to
3185 a pointer type, <tt>ty2</tt>.</p>
3188 <p>The '<tt>inttoptr</tt>' instruction takes an <a href="#t_integer">integer</a>
3189 value to cast, and a type to cast it to, which must be a
3190 <a href="#t_pointer">pointer</a> type.
3193 <p>The '<tt>inttoptr</tt>' instruction converts <tt>value</tt> to type
3194 <tt>ty2</tt> by applying either a zero extension or a truncation depending on
3195 the size of the integer <tt>value</tt>. If <tt>value</tt> is larger than the
3196 size of a pointer then a truncation is done. If <tt>value</tt> is smaller than
3197 the size of a pointer then a zero extension is done. If they are the same size,
3198 nothing is done (<i>no-op cast</i>).</p>
3202 %X = inttoptr i32 255 to i32* <i>; yields zero extension on 64-bit architecture</i>
3203 %X = inttoptr i32 255 to i32* <i>; yields no-op on 32-bit architecture</i>
3204 %Y = inttoptr i64 0 to i32* <i>; yields truncation on 32-bit architecture</i>
3208 <!-- _______________________________________________________________________ -->
3209 <div class="doc_subsubsection">
3210 <a name="i_bitcast">'<tt>bitcast .. to</tt>' Instruction</a>
3212 <div class="doc_text">
3216 <result> = bitcast <ty> <value> to <ty2> <i>; yields ty2</i>
3220 <p>The '<tt>bitcast</tt>' instruction converts <tt>value</tt> to type
3221 <tt>ty2</tt> without changing any bits.</p>
3224 <p>The '<tt>bitcast</tt>' instruction takes a value to cast, which must be
3225 a first class value, and a type to cast it to, which must also be a <a
3226 href="#t_firstclass">first class</a> type. The bit sizes of <tt>value</tt>
3227 and the destination type, <tt>ty2</tt>, must be identical. If the source
3228 type is a pointer, the destination type must also be a pointer.</p>
3231 <p>The '<tt>bitcast</tt>' instruction converts <tt>value</tt> to type
3232 <tt>ty2</tt>. It is always a <i>no-op cast</i> because no bits change with
3233 this conversion. The conversion is done as if the <tt>value</tt> had been
3234 stored to memory and read back as type <tt>ty2</tt>. Pointer types may only be
3235 converted to other pointer types with this instruction. To convert pointers to
3236 other types, use the <a href="#i_inttoptr">inttoptr</a> or
3237 <a href="#i_ptrtoint">ptrtoint</a> instructions first.</p>
3241 %X = bitcast i8 255 to i8 <i>; yields i8 :-1</i>
3242 %Y = bitcast i32* %x to sint* <i>; yields sint*:%x</i>
3243 %Z = bitcast <2xint> %V to i64; <i>; yields i64: %V</i>
3247 <!-- ======================================================================= -->
3248 <div class="doc_subsection"> <a name="otherops">Other Operations</a> </div>
3249 <div class="doc_text">
3250 <p>The instructions in this category are the "miscellaneous"
3251 instructions, which defy better classification.</p>
3254 <!-- _______________________________________________________________________ -->
3255 <div class="doc_subsubsection"><a name="i_icmp">'<tt>icmp</tt>' Instruction</a>
3257 <div class="doc_text">
3259 <pre> <result> = icmp <cond> <ty> <var1>, <var2> <i>; yields {i1}:result</i>
3262 <p>The '<tt>icmp</tt>' instruction returns a boolean value based on comparison
3263 of its two integer operands.</p>
3265 <p>The '<tt>icmp</tt>' instruction takes three operands. The first operand is
3266 the condition code indicating the kind of comparison to perform. It is not
3267 a value, just a keyword. The possible condition code are:
3269 <li><tt>eq</tt>: equal</li>
3270 <li><tt>ne</tt>: not equal </li>
3271 <li><tt>ugt</tt>: unsigned greater than</li>
3272 <li><tt>uge</tt>: unsigned greater or equal</li>
3273 <li><tt>ult</tt>: unsigned less than</li>
3274 <li><tt>ule</tt>: unsigned less or equal</li>
3275 <li><tt>sgt</tt>: signed greater than</li>
3276 <li><tt>sge</tt>: signed greater or equal</li>
3277 <li><tt>slt</tt>: signed less than</li>
3278 <li><tt>sle</tt>: signed less or equal</li>
3280 <p>The remaining two arguments must be <a href="#t_integer">integer</a> or
3281 <a href="#t_pointer">pointer</a> typed. They must also be identical types.</p>
3283 <p>The '<tt>icmp</tt>' compares <tt>var1</tt> and <tt>var2</tt> according to
3284 the condition code given as <tt>cond</tt>. The comparison performed always
3285 yields a <a href="#t_primitive">i1</a> result, as follows:
3287 <li><tt>eq</tt>: yields <tt>true</tt> if the operands are equal,
3288 <tt>false</tt> otherwise. No sign interpretation is necessary or performed.
3290 <li><tt>ne</tt>: yields <tt>true</tt> if the operands are unequal,
3291 <tt>false</tt> otherwise. No sign interpretation is necessary or performed.
3292 <li><tt>ugt</tt>: interprets the operands as unsigned values and yields
3293 <tt>true</tt> if <tt>var1</tt> is greater than <tt>var2</tt>.</li>
3294 <li><tt>uge</tt>: interprets the operands as unsigned values and yields
3295 <tt>true</tt> if <tt>var1</tt> is greater than or equal to <tt>var2</tt>.</li>
3296 <li><tt>ult</tt>: interprets the operands as unsigned values and yields
3297 <tt>true</tt> if <tt>var1</tt> is less than <tt>var2</tt>.</li>
3298 <li><tt>ule</tt>: interprets the operands as unsigned values and yields
3299 <tt>true</tt> if <tt>var1</tt> is less than or equal to <tt>var2</tt>.</li>
3300 <li><tt>sgt</tt>: interprets the operands as signed values and yields
3301 <tt>true</tt> if <tt>var1</tt> is greater than <tt>var2</tt>.</li>
3302 <li><tt>sge</tt>: interprets the operands as signed values and yields
3303 <tt>true</tt> if <tt>var1</tt> is greater than or equal to <tt>var2</tt>.</li>
3304 <li><tt>slt</tt>: interprets the operands as signed values and yields
3305 <tt>true</tt> if <tt>var1</tt> is less than <tt>var2</tt>.</li>
3306 <li><tt>sle</tt>: interprets the operands as signed values and yields
3307 <tt>true</tt> if <tt>var1</tt> is less than or equal to <tt>var2</tt>.</li>
3309 <p>If the operands are <a href="#t_pointer">pointer</a> typed, the pointer
3310 values are compared as if they were integers.</p>
3313 <pre> <result> = icmp eq i32 4, 5 <i>; yields: result=false</i>
3314 <result> = icmp ne float* %X, %X <i>; yields: result=false</i>
3315 <result> = icmp ult i16 4, 5 <i>; yields: result=true</i>
3316 <result> = icmp sgt i16 4, 5 <i>; yields: result=false</i>
3317 <result> = icmp ule i16 -4, 5 <i>; yields: result=false</i>
3318 <result> = icmp sge i16 4, 5 <i>; yields: result=false</i>
3322 <!-- _______________________________________________________________________ -->
3323 <div class="doc_subsubsection"><a name="i_fcmp">'<tt>fcmp</tt>' Instruction</a>
3325 <div class="doc_text">
3327 <pre> <result> = fcmp <cond> <ty> <var1>, <var2> <i>; yields {i1}:result</i>
3330 <p>The '<tt>fcmp</tt>' instruction returns a boolean value based on comparison
3331 of its floating point operands.</p>
3333 <p>The '<tt>fcmp</tt>' instruction takes three operands. The first operand is
3334 the condition code indicating the kind of comparison to perform. It is not
3335 a value, just a keyword. The possible condition code are:
3337 <li><tt>false</tt>: no comparison, always returns false</li>
3338 <li><tt>oeq</tt>: ordered and equal</li>
3339 <li><tt>ogt</tt>: ordered and greater than </li>
3340 <li><tt>oge</tt>: ordered and greater than or equal</li>
3341 <li><tt>olt</tt>: ordered and less than </li>
3342 <li><tt>ole</tt>: ordered and less than or equal</li>
3343 <li><tt>one</tt>: ordered and not equal</li>
3344 <li><tt>ord</tt>: ordered (no nans)</li>
3345 <li><tt>ueq</tt>: unordered or equal</li>
3346 <li><tt>ugt</tt>: unordered or greater than </li>
3347 <li><tt>uge</tt>: unordered or greater than or equal</li>
3348 <li><tt>ult</tt>: unordered or less than </li>
3349 <li><tt>ule</tt>: unordered or less than or equal</li>
3350 <li><tt>une</tt>: unordered or not equal</li>
3351 <li><tt>uno</tt>: unordered (either nans)</li>
3352 <li><tt>true</tt>: no comparison, always returns true</li>
3354 <p><i>Ordered</i> means that neither operand is a QNAN while
3355 <i>unordered</i> means that either operand may be a QNAN.</p>
3356 <p>The <tt>val1</tt> and <tt>val2</tt> arguments must be
3357 <a href="#t_floating">floating point</a> typed. They must have identical
3360 <p>The '<tt>fcmp</tt>' compares <tt>var1</tt> and <tt>var2</tt> according to
3361 the condition code given as <tt>cond</tt>. The comparison performed always
3362 yields a <a href="#t_primitive">i1</a> result, as follows:
3364 <li><tt>false</tt>: always yields <tt>false</tt>, regardless of operands.</li>
3365 <li><tt>oeq</tt>: yields <tt>true</tt> if both operands are not a QNAN and
3366 <tt>var1</tt> is equal to <tt>var2</tt>.</li>
3367 <li><tt>ogt</tt>: yields <tt>true</tt> if both operands are not a QNAN and
3368 <tt>var1</tt> is greather than <tt>var2</tt>.</li>
3369 <li><tt>oge</tt>: yields <tt>true</tt> if both operands are not a QNAN and
3370 <tt>var1</tt> is greater than or equal to <tt>var2</tt>.</li>
3371 <li><tt>olt</tt>: yields <tt>true</tt> if both operands are not a QNAN and
3372 <tt>var1</tt> is less than <tt>var2</tt>.</li>
3373 <li><tt>ole</tt>: yields <tt>true</tt> if both operands are not a QNAN and
3374 <tt>var1</tt> is less than or equal to <tt>var2</tt>.</li>
3375 <li><tt>one</tt>: yields <tt>true</tt> if both operands are not a QNAN and
3376 <tt>var1</tt> is not equal to <tt>var2</tt>.</li>
3377 <li><tt>ord</tt>: yields <tt>true</tt> if both operands are not a QNAN.</li>
3378 <li><tt>ueq</tt>: yields <tt>true</tt> if either operand is a QNAN or
3379 <tt>var1</tt> is equal to <tt>var2</tt>.</li>
3380 <li><tt>ugt</tt>: yields <tt>true</tt> if either operand is a QNAN or
3381 <tt>var1</tt> is greater than <tt>var2</tt>.</li>
3382 <li><tt>uge</tt>: yields <tt>true</tt> if either operand is a QNAN or
3383 <tt>var1</tt> is greater than or equal to <tt>var2</tt>.</li>
3384 <li><tt>ult</tt>: yields <tt>true</tt> if either operand is a QNAN or
3385 <tt>var1</tt> is less than <tt>var2</tt>.</li>
3386 <li><tt>ule</tt>: yields <tt>true</tt> if either operand is a QNAN or
3387 <tt>var1</tt> is less than or equal to <tt>var2</tt>.</li>
3388 <li><tt>une</tt>: yields <tt>true</tt> if either operand is a QNAN or
3389 <tt>var1</tt> is not equal to <tt>var2</tt>.</li>
3390 <li><tt>uno</tt>: yields <tt>true</tt> if either operand is a QNAN.</li>
3391 <li><tt>true</tt>: always yields <tt>true</tt>, regardless of operands.</li>
3395 <pre> <result> = fcmp oeq float 4.0, 5.0 <i>; yields: result=false</i>
3396 <result> = icmp one float 4.0, 5.0 <i>; yields: result=true</i>
3397 <result> = icmp olt float 4.0, 5.0 <i>; yields: result=true</i>
3398 <result> = icmp ueq double 1.0, 2.0 <i>; yields: result=false</i>
3402 <!-- _______________________________________________________________________ -->
3403 <div class="doc_subsubsection"> <a name="i_phi">'<tt>phi</tt>'
3404 Instruction</a> </div>
3405 <div class="doc_text">
3407 <pre> <result> = phi <ty> [ <val0>, <label0>], ...<br></pre>
3409 <p>The '<tt>phi</tt>' instruction is used to implement the φ node in
3410 the SSA graph representing the function.</p>
3412 <p>The type of the incoming values is specified with the first type
3413 field. After this, the '<tt>phi</tt>' instruction takes a list of pairs
3414 as arguments, with one pair for each predecessor basic block of the
3415 current block. Only values of <a href="#t_firstclass">first class</a>
3416 type may be used as the value arguments to the PHI node. Only labels
3417 may be used as the label arguments.</p>
3418 <p>There must be no non-phi instructions between the start of a basic
3419 block and the PHI instructions: i.e. PHI instructions must be first in
3422 <p>At runtime, the '<tt>phi</tt>' instruction logically takes on the value
3423 specified by the pair corresponding to the predecessor basic block that executed
3424 just prior to the current block.</p>
3426 <pre>Loop: ; Infinite loop that counts from 0 on up...<br> %indvar = phi i32 [ 0, %LoopHeader ], [ %nextindvar, %Loop ]<br> %nextindvar = add i32 %indvar, 1<br> br label %Loop<br></pre>
3429 <!-- _______________________________________________________________________ -->
3430 <div class="doc_subsubsection">
3431 <a name="i_select">'<tt>select</tt>' Instruction</a>
3434 <div class="doc_text">
3439 <result> = select i1 <cond>, <ty> <val1>, <ty> <val2> <i>; yields ty</i>
3445 The '<tt>select</tt>' instruction is used to choose one value based on a
3446 condition, without branching.
3453 The '<tt>select</tt>' instruction requires a boolean value indicating the condition, and two values of the same <a href="#t_firstclass">first class</a> type.
3459 If the boolean condition evaluates to true, the instruction returns the first
3460 value argument; otherwise, it returns the second value argument.
3466 %X = select i1 true, i8 17, i8 42 <i>; yields i8:17</i>
3471 <!-- _______________________________________________________________________ -->
3472 <div class="doc_subsubsection">
3473 <a name="i_call">'<tt>call</tt>' Instruction</a>
3476 <div class="doc_text">
3480 <result> = [tail] call [<a href="#callingconv">cconv</a>] <ty>* <fnptrval>(<param list>)
3485 <p>The '<tt>call</tt>' instruction represents a simple function call.</p>
3489 <p>This instruction requires several arguments:</p>
3493 <p>The optional "tail" marker indicates whether the callee function accesses
3494 any allocas or varargs in the caller. If the "tail" marker is present, the
3495 function call is eligible for tail call optimization. Note that calls may
3496 be marked "tail" even if they do not occur before a <a
3497 href="#i_ret"><tt>ret</tt></a> instruction.
3500 <p>The optional "cconv" marker indicates which <a href="#callingconv">calling
3501 convention</a> the call should use. If none is specified, the call defaults
3502 to using C calling conventions.
3505 <p>'<tt>ty</tt>': shall be the signature of the pointer to function value
3506 being invoked. The argument types must match the types implied by this
3507 signature. This type can be omitted if the function is not varargs and
3508 if the function type does not return a pointer to a function.</p>
3511 <p>'<tt>fnptrval</tt>': An LLVM value containing a pointer to a function to
3512 be invoked. In most cases, this is a direct function invocation, but
3513 indirect <tt>call</tt>s are just as possible, calling an arbitrary pointer
3514 to function value.</p>
3517 <p>'<tt>function args</tt>': argument list whose types match the
3518 function signature argument types. All arguments must be of
3519 <a href="#t_firstclass">first class</a> type. If the function signature
3520 indicates the function accepts a variable number of arguments, the extra
3521 arguments can be specified.</p>
3527 <p>The '<tt>call</tt>' instruction is used to cause control flow to
3528 transfer to a specified function, with its incoming arguments bound to
3529 the specified values. Upon a '<tt><a href="#i_ret">ret</a></tt>'
3530 instruction in the called function, control flow continues with the
3531 instruction after the function call, and the return value of the
3532 function is bound to the result argument. This is a simpler case of
3533 the <a href="#i_invoke">invoke</a> instruction.</p>
3538 %retval = call i32 %test(i32 %argc)
3539 call i32(i8 *, ...) *%printf(i8 * %msg, i32 12, i8 42);
3540 %X = tail call i32 %foo()
3541 %Y = tail call <a href="#callingconv">fastcc</a> i32 %foo()
3546 <!-- _______________________________________________________________________ -->
3547 <div class="doc_subsubsection">
3548 <a name="i_va_arg">'<tt>va_arg</tt>' Instruction</a>
3551 <div class="doc_text">
3556 <resultval> = va_arg <va_list*> <arglist>, <argty>
3561 <p>The '<tt>va_arg</tt>' instruction is used to access arguments passed through
3562 the "variable argument" area of a function call. It is used to implement the
3563 <tt>va_arg</tt> macro in C.</p>
3567 <p>This instruction takes a <tt>va_list*</tt> value and the type of
3568 the argument. It returns a value of the specified argument type and
3569 increments the <tt>va_list</tt> to point to the next argument. The
3570 actual type of <tt>va_list</tt> is target specific.</p>
3574 <p>The '<tt>va_arg</tt>' instruction loads an argument of the specified
3575 type from the specified <tt>va_list</tt> and causes the
3576 <tt>va_list</tt> to point to the next argument. For more information,
3577 see the variable argument handling <a href="#int_varargs">Intrinsic
3580 <p>It is legal for this instruction to be called in a function which does not
3581 take a variable number of arguments, for example, the <tt>vfprintf</tt>
3584 <p><tt>va_arg</tt> is an LLVM instruction instead of an <a
3585 href="#intrinsics">intrinsic function</a> because it takes a type as an
3590 <p>See the <a href="#int_varargs">variable argument processing</a> section.</p>
3594 <!-- *********************************************************************** -->
3595 <div class="doc_section"> <a name="intrinsics">Intrinsic Functions</a> </div>
3596 <!-- *********************************************************************** -->
3598 <div class="doc_text">
3600 <p>LLVM supports the notion of an "intrinsic function". These functions have
3601 well known names and semantics and are required to follow certain restrictions.
3602 Overall, these intrinsics represent an extension mechanism for the LLVM
3603 language that does not require changing all of the transformations in LLVM when
3604 adding to the language (or the bytecode reader/writer, the parser, etc...).</p>
3606 <p>Intrinsic function names must all start with an "<tt>llvm.</tt>" prefix. This
3607 prefix is reserved in LLVM for intrinsic names; thus, function names may not
3608 begin with this prefix. Intrinsic functions must always be external functions:
3609 you cannot define the body of intrinsic functions. Intrinsic functions may
3610 only be used in call or invoke instructions: it is illegal to take the address
3611 of an intrinsic function. Additionally, because intrinsic functions are part
3612 of the LLVM language, it is required if any are added that they be documented
3615 <p>Some intrinsic functions can be overloaded, i.e., the intrinsic represents
3616 a family of functions that perform the same operation but on different data
3617 types. This is most frequent with the integer types. Since LLVM can represent
3618 over 8 million different integer types, there is a way to declare an intrinsic
3619 that can be overloaded based on its arguments. Such an intrinsic will have the
3620 names of its argument types encoded into its function name, each
3621 preceded by a period. For example, the <tt>llvm.ctpop</tt> function can take an
3622 integer of any width. This leads to a family of functions such as
3623 <tt>i32 @llvm.ctpop.i8(i8 %val)</tt> and <tt>i32 @llvm.ctpop.i29(i29 %val)</tt>.
3627 <p>To learn how to add an intrinsic function, please see the
3628 <a href="ExtendingLLVM.html">Extending LLVM Guide</a>.
3633 <!-- ======================================================================= -->
3634 <div class="doc_subsection">
3635 <a name="int_varargs">Variable Argument Handling Intrinsics</a>
3638 <div class="doc_text">
3640 <p>Variable argument support is defined in LLVM with the <a
3641 href="#i_va_arg"><tt>va_arg</tt></a> instruction and these three
3642 intrinsic functions. These functions are related to the similarly
3643 named macros defined in the <tt><stdarg.h></tt> header file.</p>
3645 <p>All of these functions operate on arguments that use a
3646 target-specific value type "<tt>va_list</tt>". The LLVM assembly
3647 language reference manual does not define what this type is, so all
3648 transformations should be prepared to handle these functions regardless of
3651 <p>This example shows how the <a href="#i_va_arg"><tt>va_arg</tt></a>
3652 instruction and the variable argument handling intrinsic functions are
3656 define i32 @test(i32 %X, ...) {
3657 ; Initialize variable argument processing
3659 %ap2 = bitcast i8** %ap to i8*
3660 call void @llvm.va_start(i8* %ap2)
3662 ; Read a single integer argument
3663 %tmp = va_arg i8** %ap, i32
3665 ; Demonstrate usage of llvm.va_copy and llvm.va_end
3667 %aq2 = bitcast i8** %aq to i8*
3668 call void @llvm.va_copy(i8* %aq2, i8* %ap2)
3669 call void @llvm.va_end(i8* %aq2)
3671 ; Stop processing of arguments.
3672 call void @llvm.va_end(i8* %ap2)
3676 declare void @llvm.va_start(i8*)
3677 declare void @llvm.va_copy(i8*, i8*)
3678 declare void @llvm.va_end(i8*)
3682 <!-- _______________________________________________________________________ -->
3683 <div class="doc_subsubsection">
3684 <a name="int_va_start">'<tt>llvm.va_start</tt>' Intrinsic</a>
3688 <div class="doc_text">
3690 <pre> declare void %llvm.va_start(i8* <arglist>)<br></pre>
3692 <P>The '<tt>llvm.va_start</tt>' intrinsic initializes
3693 <tt>*<arglist></tt> for subsequent use by <tt><a
3694 href="#i_va_arg">va_arg</a></tt>.</p>
3698 <P>The argument is a pointer to a <tt>va_list</tt> element to initialize.</p>
3702 <P>The '<tt>llvm.va_start</tt>' intrinsic works just like the <tt>va_start</tt>
3703 macro available in C. In a target-dependent way, it initializes the
3704 <tt>va_list</tt> element to which the argument points, so that the next call to
3705 <tt>va_arg</tt> will produce the first variable argument passed to the function.
3706 Unlike the C <tt>va_start</tt> macro, this intrinsic does not need to know the
3707 last argument of the function as the compiler can figure that out.</p>
3711 <!-- _______________________________________________________________________ -->
3712 <div class="doc_subsubsection">
3713 <a name="int_va_end">'<tt>llvm.va_end</tt>' Intrinsic</a>
3716 <div class="doc_text">
3718 <pre> declare void @llvm.va_end(i8* <arglist>)<br></pre>
3721 <p>The '<tt>llvm.va_end</tt>' intrinsic destroys <tt>*<arglist></tt>,
3722 which has been initialized previously with <tt><a href="#int_va_start">llvm.va_start</a></tt>
3723 or <tt><a href="#i_va_copy">llvm.va_copy</a></tt>.</p>
3727 <p>The argument is a pointer to a <tt>va_list</tt> to destroy.</p>
3731 <p>The '<tt>llvm.va_end</tt>' intrinsic works just like the <tt>va_end</tt>
3732 macro available in C. In a target-dependent way, it destroys the
3733 <tt>va_list</tt> element to which the argument points. Calls to <a
3734 href="#int_va_start"><tt>llvm.va_start</tt></a> and <a href="#int_va_copy">
3735 <tt>llvm.va_copy</tt></a> must be matched exactly with calls to
3736 <tt>llvm.va_end</tt>.</p>
3740 <!-- _______________________________________________________________________ -->
3741 <div class="doc_subsubsection">
3742 <a name="int_va_copy">'<tt>llvm.va_copy</tt>' Intrinsic</a>
3745 <div class="doc_text">
3750 declare void @llvm.va_copy(i8* <destarglist>, i8* <srcarglist>)
3755 <p>The '<tt>llvm.va_copy</tt>' intrinsic copies the current argument position
3756 from the source argument list to the destination argument list.</p>
3760 <p>The first argument is a pointer to a <tt>va_list</tt> element to initialize.
3761 The second argument is a pointer to a <tt>va_list</tt> element to copy from.</p>
3766 <p>The '<tt>llvm.va_copy</tt>' intrinsic works just like the <tt>va_copy</tt>
3767 macro available in C. In a target-dependent way, it copies the source
3768 <tt>va_list</tt> element into the destination <tt>va_list</tt> element. This
3769 intrinsic is necessary because the <tt><a href="#int_va_start">
3770 llvm.va_start</a></tt> intrinsic may be arbitrarily complex and require, for
3771 example, memory allocation.</p>
3775 <!-- ======================================================================= -->
3776 <div class="doc_subsection">
3777 <a name="int_gc">Accurate Garbage Collection Intrinsics</a>
3780 <div class="doc_text">
3783 LLVM support for <a href="GarbageCollection.html">Accurate Garbage
3784 Collection</a> requires the implementation and generation of these intrinsics.
3785 These intrinsics allow identification of <a href="#int_gcroot">GC roots on the
3786 stack</a>, as well as garbage collector implementations that require <a
3787 href="#int_gcread">read</a> and <a href="#int_gcwrite">write</a> barriers.
3788 Front-ends for type-safe garbage collected languages should generate these
3789 intrinsics to make use of the LLVM garbage collectors. For more details, see <a
3790 href="GarbageCollection.html">Accurate Garbage Collection with LLVM</a>.
3794 <!-- _______________________________________________________________________ -->
3795 <div class="doc_subsubsection">
3796 <a name="int_gcroot">'<tt>llvm.gcroot</tt>' Intrinsic</a>
3799 <div class="doc_text">
3804 declare void @llvm.gcroot(<ty>** %ptrloc, <ty2>* %metadata)
3809 <p>The '<tt>llvm.gcroot</tt>' intrinsic declares the existence of a GC root to
3810 the code generator, and allows some metadata to be associated with it.</p>
3814 <p>The first argument specifies the address of a stack object that contains the
3815 root pointer. The second pointer (which must be either a constant or a global
3816 value address) contains the meta-data to be associated with the root.</p>
3820 <p>At runtime, a call to this intrinsics stores a null pointer into the "ptrloc"
3821 location. At compile-time, the code generator generates information to allow
3822 the runtime to find the pointer at GC safe points.
3828 <!-- _______________________________________________________________________ -->
3829 <div class="doc_subsubsection">
3830 <a name="int_gcread">'<tt>llvm.gcread</tt>' Intrinsic</a>
3833 <div class="doc_text">
3838 declare i8 * @llvm.gcread(i8 * %ObjPtr, i8 ** %Ptr)
3843 <p>The '<tt>llvm.gcread</tt>' intrinsic identifies reads of references from heap
3844 locations, allowing garbage collector implementations that require read
3849 <p>The second argument is the address to read from, which should be an address
3850 allocated from the garbage collector. The first object is a pointer to the
3851 start of the referenced object, if needed by the language runtime (otherwise
3856 <p>The '<tt>llvm.gcread</tt>' intrinsic has the same semantics as a load
3857 instruction, but may be replaced with substantially more complex code by the
3858 garbage collector runtime, as needed.</p>
3863 <!-- _______________________________________________________________________ -->
3864 <div class="doc_subsubsection">
3865 <a name="int_gcwrite">'<tt>llvm.gcwrite</tt>' Intrinsic</a>
3868 <div class="doc_text">
3873 declare void @llvm.gcwrite(i8 * %P1, i8 * %Obj, i8 ** %P2)
3878 <p>The '<tt>llvm.gcwrite</tt>' intrinsic identifies writes of references to heap
3879 locations, allowing garbage collector implementations that require write
3880 barriers (such as generational or reference counting collectors).</p>
3884 <p>The first argument is the reference to store, the second is the start of the
3885 object to store it to, and the third is the address of the field of Obj to
3886 store to. If the runtime does not require a pointer to the object, Obj may be
3891 <p>The '<tt>llvm.gcwrite</tt>' intrinsic has the same semantics as a store
3892 instruction, but may be replaced with substantially more complex code by the
3893 garbage collector runtime, as needed.</p>
3899 <!-- ======================================================================= -->
3900 <div class="doc_subsection">
3901 <a name="int_codegen">Code Generator Intrinsics</a>
3904 <div class="doc_text">
3906 These intrinsics are provided by LLVM to expose special features that may only
3907 be implemented with code generator support.
3912 <!-- _______________________________________________________________________ -->
3913 <div class="doc_subsubsection">
3914 <a name="int_returnaddress">'<tt>llvm.returnaddress</tt>' Intrinsic</a>
3917 <div class="doc_text">
3921 declare i8 *@llvm.returnaddress(i32 <level>)
3927 The '<tt>llvm.returnaddress</tt>' intrinsic attempts to compute a
3928 target-specific value indicating the return address of the current function
3929 or one of its callers.
3935 The argument to this intrinsic indicates which function to return the address
3936 for. Zero indicates the calling function, one indicates its caller, etc. The
3937 argument is <b>required</b> to be a constant integer value.
3943 The '<tt>llvm.returnaddress</tt>' intrinsic either returns a pointer indicating
3944 the return address of the specified call frame, or zero if it cannot be
3945 identified. The value returned by this intrinsic is likely to be incorrect or 0
3946 for arguments other than zero, so it should only be used for debugging purposes.
3950 Note that calling this intrinsic does not prevent function inlining or other
3951 aggressive transformations, so the value returned may not be that of the obvious
3952 source-language caller.
3957 <!-- _______________________________________________________________________ -->
3958 <div class="doc_subsubsection">
3959 <a name="int_frameaddress">'<tt>llvm.frameaddress</tt>' Intrinsic</a>
3962 <div class="doc_text">
3966 declare i8 *@llvm.frameaddress(i32 <level>)
3972 The '<tt>llvm.frameaddress</tt>' intrinsic attempts to return the
3973 target-specific frame pointer value for the specified stack frame.
3979 The argument to this intrinsic indicates which function to return the frame
3980 pointer for. Zero indicates the calling function, one indicates its caller,
3981 etc. The argument is <b>required</b> to be a constant integer value.
3987 The '<tt>llvm.frameaddress</tt>' intrinsic either returns a pointer indicating
3988 the frame address of the specified call frame, or zero if it cannot be
3989 identified. The value returned by this intrinsic is likely to be incorrect or 0
3990 for arguments other than zero, so it should only be used for debugging purposes.
3994 Note that calling this intrinsic does not prevent function inlining or other
3995 aggressive transformations, so the value returned may not be that of the obvious
3996 source-language caller.
4000 <!-- _______________________________________________________________________ -->
4001 <div class="doc_subsubsection">
4002 <a name="int_stacksave">'<tt>llvm.stacksave</tt>' Intrinsic</a>
4005 <div class="doc_text">
4009 declare i8 *@llvm.stacksave()
4015 The '<tt>llvm.stacksave</tt>' intrinsic is used to remember the current state of
4016 the function stack, for use with <a href="#int_stackrestore">
4017 <tt>llvm.stackrestore</tt></a>. This is useful for implementing language
4018 features like scoped automatic variable sized arrays in C99.
4024 This intrinsic returns a opaque pointer value that can be passed to <a
4025 href="#int_stackrestore"><tt>llvm.stackrestore</tt></a>. When an
4026 <tt>llvm.stackrestore</tt> intrinsic is executed with a value saved from
4027 <tt>llvm.stacksave</tt>, it effectively restores the state of the stack to the
4028 state it was in when the <tt>llvm.stacksave</tt> intrinsic executed. In
4029 practice, this pops any <a href="#i_alloca">alloca</a> blocks from the stack
4030 that were allocated after the <tt>llvm.stacksave</tt> was executed.
4035 <!-- _______________________________________________________________________ -->
4036 <div class="doc_subsubsection">
4037 <a name="int_stackrestore">'<tt>llvm.stackrestore</tt>' Intrinsic</a>
4040 <div class="doc_text">
4044 declare void @llvm.stackrestore(i8 * %ptr)
4050 The '<tt>llvm.stackrestore</tt>' intrinsic is used to restore the state of
4051 the function stack to the state it was in when the corresponding <a
4052 href="#int_stacksave"><tt>llvm.stacksave</tt></a> intrinsic executed. This is
4053 useful for implementing language features like scoped automatic variable sized
4060 See the description for <a href="#int_stacksave"><tt>llvm.stacksave</tt></a>.
4066 <!-- _______________________________________________________________________ -->
4067 <div class="doc_subsubsection">
4068 <a name="int_prefetch">'<tt>llvm.prefetch</tt>' Intrinsic</a>
4071 <div class="doc_text">
4075 declare void @llvm.prefetch(i8 * <address>,
4076 i32 <rw>, i32 <locality>)
4083 The '<tt>llvm.prefetch</tt>' intrinsic is a hint to the code generator to insert
4084 a prefetch instruction if supported; otherwise, it is a noop. Prefetches have
4086 effect on the behavior of the program but can change its performance
4093 <tt>address</tt> is the address to be prefetched, <tt>rw</tt> is the specifier
4094 determining if the fetch should be for a read (0) or write (1), and
4095 <tt>locality</tt> is a temporal locality specifier ranging from (0) - no
4096 locality, to (3) - extremely local keep in cache. The <tt>rw</tt> and
4097 <tt>locality</tt> arguments must be constant integers.
4103 This intrinsic does not modify the behavior of the program. In particular,
4104 prefetches cannot trap and do not produce a value. On targets that support this
4105 intrinsic, the prefetch can provide hints to the processor cache for better
4111 <!-- _______________________________________________________________________ -->
4112 <div class="doc_subsubsection">
4113 <a name="int_pcmarker">'<tt>llvm.pcmarker</tt>' Intrinsic</a>
4116 <div class="doc_text">
4120 declare void @llvm.pcmarker( i32 <id> )
4127 The '<tt>llvm.pcmarker</tt>' intrinsic is a method to export a Program Counter
4129 code to simulators and other tools. The method is target specific, but it is
4130 expected that the marker will use exported symbols to transmit the PC of the marker.
4131 The marker makes no guarantees that it will remain with any specific instruction
4132 after optimizations. It is possible that the presence of a marker will inhibit
4133 optimizations. The intended use is to be inserted after optimizations to allow
4134 correlations of simulation runs.
4140 <tt>id</tt> is a numerical id identifying the marker.
4146 This intrinsic does not modify the behavior of the program. Backends that do not
4147 support this intrinisic may ignore it.
4152 <!-- _______________________________________________________________________ -->
4153 <div class="doc_subsubsection">
4154 <a name="int_readcyclecounter">'<tt>llvm.readcyclecounter</tt>' Intrinsic</a>
4157 <div class="doc_text">
4161 declare i64 @llvm.readcyclecounter( )
4168 The '<tt>llvm.readcyclecounter</tt>' intrinsic provides access to the cycle
4169 counter register (or similar low latency, high accuracy clocks) on those targets
4170 that support it. On X86, it should map to RDTSC. On Alpha, it should map to RPCC.
4171 As the backing counters overflow quickly (on the order of 9 seconds on alpha), this
4172 should only be used for small timings.
4178 When directly supported, reading the cycle counter should not modify any memory.
4179 Implementations are allowed to either return a application specific value or a
4180 system wide value. On backends without support, this is lowered to a constant 0.
4185 <!-- ======================================================================= -->
4186 <div class="doc_subsection">
4187 <a name="int_libc">Standard C Library Intrinsics</a>
4190 <div class="doc_text">
4192 LLVM provides intrinsics for a few important standard C library functions.
4193 These intrinsics allow source-language front-ends to pass information about the
4194 alignment of the pointer arguments to the code generator, providing opportunity
4195 for more efficient code generation.
4200 <!-- _______________________________________________________________________ -->
4201 <div class="doc_subsubsection">
4202 <a name="int_memcpy">'<tt>llvm.memcpy</tt>' Intrinsic</a>
4205 <div class="doc_text">
4209 declare void @llvm.memcpy.i32(i8 * <dest>, i8 * <src>,
4210 i32 <len>, i32 <align>)
4211 declare void @llvm.memcpy.i64(i8 * <dest>, i8 * <src>,
4212 i64 <len>, i32 <align>)
4218 The '<tt>llvm.memcpy.*</tt>' intrinsics copy a block of memory from the source
4219 location to the destination location.
4223 Note that, unlike the standard libc function, the <tt>llvm.memcpy.*</tt>
4224 intrinsics do not return a value, and takes an extra alignment argument.
4230 The first argument is a pointer to the destination, the second is a pointer to
4231 the source. The third argument is an integer argument
4232 specifying the number of bytes to copy, and the fourth argument is the alignment
4233 of the source and destination locations.
4237 If the call to this intrinisic has an alignment value that is not 0 or 1, then
4238 the caller guarantees that both the source and destination pointers are aligned
4245 The '<tt>llvm.memcpy.*</tt>' intrinsics copy a block of memory from the source
4246 location to the destination location, which are not allowed to overlap. It
4247 copies "len" bytes of memory over. If the argument is known to be aligned to
4248 some boundary, this can be specified as the fourth argument, otherwise it should
4254 <!-- _______________________________________________________________________ -->
4255 <div class="doc_subsubsection">
4256 <a name="int_memmove">'<tt>llvm.memmove</tt>' Intrinsic</a>
4259 <div class="doc_text">
4263 declare void @llvm.memmove.i32(i8 * <dest>, i8 * <src>,
4264 i32 <len>, i32 <align>)
4265 declare void @llvm.memmove.i64(i8 * <dest>, i8 * <src>,
4266 i64 <len>, i32 <align>)
4272 The '<tt>llvm.memmove.*</tt>' intrinsics move a block of memory from the source
4273 location to the destination location. It is similar to the
4274 '<tt>llvm.memcmp</tt>' intrinsic but allows the two memory locations to overlap.
4278 Note that, unlike the standard libc function, the <tt>llvm.memmove.*</tt>
4279 intrinsics do not return a value, and takes an extra alignment argument.
4285 The first argument is a pointer to the destination, the second is a pointer to
4286 the source. The third argument is an integer argument
4287 specifying the number of bytes to copy, and the fourth argument is the alignment
4288 of the source and destination locations.
4292 If the call to this intrinisic has an alignment value that is not 0 or 1, then
4293 the caller guarantees that the source and destination pointers are aligned to
4300 The '<tt>llvm.memmove.*</tt>' intrinsics copy a block of memory from the source
4301 location to the destination location, which may overlap. It
4302 copies "len" bytes of memory over. If the argument is known to be aligned to
4303 some boundary, this can be specified as the fourth argument, otherwise it should
4309 <!-- _______________________________________________________________________ -->
4310 <div class="doc_subsubsection">
4311 <a name="int_memset">'<tt>llvm.memset.*</tt>' Intrinsics</a>
4314 <div class="doc_text">
4318 declare void @llvm.memset.i32(i8 * <dest>, i8 <val>,
4319 i32 <len>, i32 <align>)
4320 declare void @llvm.memset.i64(i8 * <dest>, i8 <val>,
4321 i64 <len>, i32 <align>)
4327 The '<tt>llvm.memset.*</tt>' intrinsics fill a block of memory with a particular
4332 Note that, unlike the standard libc function, the <tt>llvm.memset</tt> intrinsic
4333 does not return a value, and takes an extra alignment argument.
4339 The first argument is a pointer to the destination to fill, the second is the
4340 byte value to fill it with, the third argument is an integer
4341 argument specifying the number of bytes to fill, and the fourth argument is the
4342 known alignment of destination location.
4346 If the call to this intrinisic has an alignment value that is not 0 or 1, then
4347 the caller guarantees that the destination pointer is aligned to that boundary.
4353 The '<tt>llvm.memset.*</tt>' intrinsics fill "len" bytes of memory starting at
4355 destination location. If the argument is known to be aligned to some boundary,
4356 this can be specified as the fourth argument, otherwise it should be set to 0 or
4362 <!-- _______________________________________________________________________ -->
4363 <div class="doc_subsubsection">
4364 <a name="int_sqrt">'<tt>llvm.sqrt.*</tt>' Intrinsic</a>
4367 <div class="doc_text">
4371 declare float @llvm.sqrt.f32(float %Val)
4372 declare double @llvm.sqrt.f64(double %Val)
4378 The '<tt>llvm.sqrt</tt>' intrinsics return the sqrt of the specified operand,
4379 returning the same value as the libm '<tt>sqrt</tt>' function would. Unlike
4380 <tt>sqrt</tt> in libm, however, <tt>llvm.sqrt</tt> has undefined behavior for
4381 negative numbers (which allows for better optimization).
4387 The argument and return value are floating point numbers of the same type.
4393 This function returns the sqrt of the specified operand if it is a positive
4394 floating point number.
4398 <!-- _______________________________________________________________________ -->
4399 <div class="doc_subsubsection">
4400 <a name="int_powi">'<tt>llvm.powi.*</tt>' Intrinsic</a>
4403 <div class="doc_text">
4407 declare float @llvm.powi.f32(float %Val, i32 %power)
4408 declare double @llvm.powi.f64(double %Val, i32 %power)
4414 The '<tt>llvm.powi.*</tt>' intrinsics return the first operand raised to the
4415 specified (positive or negative) power. The order of evaluation of
4416 multiplications is not defined.
4422 The second argument is an integer power, and the first is a value to raise to
4429 This function returns the first value raised to the second power with an
4430 unspecified sequence of rounding operations.</p>
4434 <!-- ======================================================================= -->
4435 <div class="doc_subsection">
4436 <a name="int_manip">Bit Manipulation Intrinsics</a>
4439 <div class="doc_text">
4441 LLVM provides intrinsics for a few important bit manipulation operations.
4442 These allow efficient code generation for some algorithms.
4447 <!-- _______________________________________________________________________ -->
4448 <div class="doc_subsubsection">
4449 <a name="int_bswap">'<tt>llvm.bswap.*</tt>' Intrinsics</a>
4452 <div class="doc_text">
4455 <p>This is an overloaded intrinsic function. You can use bswap on any integer
4456 type that is an even number of bytes (i.e. BitWidth % 16 == 0). Note the suffix
4457 that includes the type for the result and the operand.
4459 declare i16 @llvm.bswap.i16.i16(i16 <id>)
4460 declare i32 @llvm.bswap.i32.i32(i32 <id>)
4461 declare i64 @llvm.bswap.i64.i64(i64 <id>)
4467 The '<tt>llvm.bswap</tt>' family of intrinsics is used to byte swap integer
4468 values with an even number of bytes (positive multiple of 16 bits). These are
4469 useful for performing operations on data that is not in the target's native
4476 The <tt>llvm.bswap.16.i16</tt> intrinsic returns an i16 value that has the high
4477 and low byte of the input i16 swapped. Similarly, the <tt>llvm.bswap.i32</tt>
4478 intrinsic returns an i32 value that has the four bytes of the input i32
4479 swapped, so that if the input bytes are numbered 0, 1, 2, 3 then the returned
4480 i32 will have its bytes in 3, 2, 1, 0 order. The <tt>llvm.bswap.i48.i48</tt>,
4481 <tt>llvm.bswap.i64.i64</tt> and other intrinsics extend this concept to
4482 additional even-byte lengths (6 bytes, 8 bytes and more, respectively).
4487 <!-- _______________________________________________________________________ -->
4488 <div class="doc_subsubsection">
4489 <a name="int_ctpop">'<tt>llvm.ctpop.*</tt>' Intrinsic</a>
4492 <div class="doc_text">
4495 <p>This is an overloaded intrinsic. You can use llvm.ctpop on any integer bit
4496 width. Not all targets support all bit widths however.
4498 declare i32 @llvm.ctpop.i8 (i8 <src>)
4499 declare i32 @llvm.ctpop.i16(i16 <src>)
4500 declare i32 @llvm.ctpop.i32(i32 <src>)
4501 declare i32 @llvm.ctpop.i64(i64 <src>)
4502 declare i32 @llvm.ctpop.i256(i256 <src>)
4508 The '<tt>llvm.ctpop</tt>' family of intrinsics counts the number of bits set in a
4515 The only argument is the value to be counted. The argument may be of any
4516 integer type. The return type must match the argument type.
4522 The '<tt>llvm.ctpop</tt>' intrinsic counts the 1's in a variable.
4526 <!-- _______________________________________________________________________ -->
4527 <div class="doc_subsubsection">
4528 <a name="int_ctlz">'<tt>llvm.ctlz.*</tt>' Intrinsic</a>
4531 <div class="doc_text">
4534 <p>This is an overloaded intrinsic. You can use <tt>llvm.ctlz</tt> on any
4535 integer bit width. Not all targets support all bit widths however.
4537 declare i32 @llvm.ctlz.i8 (i8 <src>)
4538 declare i32 @llvm.ctlz.i16(i16 <src>)
4539 declare i32 @llvm.ctlz.i32(i32 <src>)
4540 declare i32 @llvm.ctlz.i64(i64 <src>)
4541 declare i32 @llvm.ctlz.i256(i256 <src>)
4547 The '<tt>llvm.ctlz</tt>' family of intrinsic functions counts the number of
4548 leading zeros in a variable.
4554 The only argument is the value to be counted. The argument may be of any
4555 integer type. The return type must match the argument type.
4561 The '<tt>llvm.ctlz</tt>' intrinsic counts the leading (most significant) zeros
4562 in a variable. If the src == 0 then the result is the size in bits of the type
4563 of src. For example, <tt>llvm.ctlz(i32 2) = 30</tt>.
4569 <!-- _______________________________________________________________________ -->
4570 <div class="doc_subsubsection">
4571 <a name="int_cttz">'<tt>llvm.cttz.*</tt>' Intrinsic</a>
4574 <div class="doc_text">
4577 <p>This is an overloaded intrinsic. You can use <tt>llvm.cttz</tt> on any
4578 integer bit width. Not all targets support all bit widths however.
4580 declare i32 @llvm.cttz.i8 (i8 <src>)
4581 declare i32 @llvm.cttz.i16(i16 <src>)
4582 declare i32 @llvm.cttz.i32(i32 <src>)
4583 declare i32 @llvm.cttz.i64(i64 <src>)
4584 declare i32 @llvm.cttz.i256(i256 <src>)
4590 The '<tt>llvm.cttz</tt>' family of intrinsic functions counts the number of
4597 The only argument is the value to be counted. The argument may be of any
4598 integer type. The return type must match the argument type.
4604 The '<tt>llvm.cttz</tt>' intrinsic counts the trailing (least significant) zeros
4605 in a variable. If the src == 0 then the result is the size in bits of the type
4606 of src. For example, <tt>llvm.cttz(2) = 1</tt>.
4610 <!-- _______________________________________________________________________ -->
4611 <div class="doc_subsubsection">
4612 <a name="int_part_select">'<tt>llvm.part.select.*</tt>' Intrinsic</a>
4615 <div class="doc_text">
4618 <p>This is an overloaded intrinsic. You can use <tt>llvm.part.select</tt>
4619 on any integer bit width.
4621 declare i17 @llvm.part.select.i17.i17 (i17 %val, i32 %loBit, i32 %hiBit)
4622 declare i29 @llvm.part.select.i29.i29 (i29 %val, i32 %loBit, i32 %hiBit)
4626 <p>The '<tt>llvm.part.select</tt>' family of intrinsic functions selects a
4627 range of bits from an integer value and returns them in the same bit width as
4628 the original value.</p>
4631 <p>The first argument, <tt>%val</tt> and the result may be integer types of
4632 any bit width but they must have the same bit width. The second and third
4633 arguments must be <tt>i32</tt> type since they specify only a bit index.</p>
4636 <p>The operation of the '<tt>llvm.part.select</tt>' intrinsic has two modes
4637 of operation: forwards and reverse. If <tt>%loBit</tt> is greater than
4638 <tt>%hiBits</tt> then the intrinsic operates in reverse mode. Otherwise it
4639 operates in forward mode.</p>
4640 <p>In forward mode, this intrinsic is the equivalent of shifting <tt>%val</tt>
4641 right by <tt>%loBit</tt> bits and then ANDing it with a mask with
4642 only the <tt>%hiBit - %loBit</tt> bits set, as follows:</p>
4644 <li>The <tt>%val</tt> is shifted right (LSHR) by the number of bits specified
4645 by <tt>%loBits</tt>. This normalizes the value to the low order bits.</li>
4646 <li>The <tt>%loBits</tt> value is subtracted from the <tt>%hiBits</tt> value
4647 to determine the number of bits to retain.</li>
4648 <li>A mask of the retained bits is created by shifting a -1 value.</li>
4649 <li>The mask is ANDed with <tt>%val</tt> to produce the result.
4651 <p>In reverse mode, a similar computation is made except that:</p>
4653 <li>The bits selected wrap around to include both the highest and lowest bits.
4654 For example, part.select(i16 X, 4, 7) selects bits from X with a mask of
4655 0x00F0 (forwards case) while part.select(i16 X, 8, 3) selects bits from X
4656 with a mask of 0xFF0F.</li>
4657 <li>The bits returned in the reverse case are reversed. So, if X has the value
4658 0x6ACF and we apply part.select(i16 X, 8, 3) to it, we get back the value
4663 <div class="doc_subsubsection">
4664 <a name="int_part_set">'<tt>llvm.part.set.*</tt>' Intrinsic</a>
4667 <div class="doc_text">
4670 <p>This is an overloaded intrinsic. You can use <tt>llvm.part.set</tt>
4671 on any integer bit width.
4673 declare i17 @llvm.part.set.i17.i17.i9 (i17 %val, i9 %repl, i32 %lo, i32 %hi)
4674 declare i29 @llvm.part.set.i29.i29.i9 (i29 %val, i9 %repl, i32 %lo, i32 %hi)
4678 <p>The '<tt>llvm.part.set</tt>' family of intrinsic functions replaces a range
4679 of bits in an integer value with another integer value. It returns the integer
4680 with the replaced bits.</p>
4683 <p>The first argument, <tt>%val</tt> and the result may be integer types of
4684 any bit width but they must have the same bit width. <tt>%val</tt> is the value
4685 whose bits will be replaced. The second argument, <tt>%repl</tt> may be an
4686 integer of any bit width. The third and fourth arguments must be <tt>i32</tt>
4687 type since they specify only a bit index.</p>
4690 <p>The operation of the '<tt>llvm.part.set</tt>' intrinsic has two modes
4691 of operation: forwards and reverse. If <tt>%lo</tt> is greater than
4692 <tt>%hi</tt> then the intrinsic operates in reverse mode. Otherwise it
4693 operates in forward mode.</p>
4694 <p>For both modes, the <tt>%repl</tt> value is prepared for use by either
4695 truncating it down to the size of the replacement area or zero extending it
4696 up to that size.</p>
4697 <p>In forward mode, the bits between <tt>%lo</tt> and <tt>%hi</tt> (inclusive)
4698 are replaced with corresponding bits from <tt>%repl</tt>. That is the 0th bit
4699 in <tt>%repl</tt> replaces the <tt>%lo</tt>th bit in <tt>%val</tt> and etc. up
4700 to the <tt>%hi</tt>th bit.
4701 <p>In reverse mode, a similar computation is made except that the bits replaced
4702 wrap around to include both the highest and lowest bits. For example, if a
4703 16 bit value is being replaced then <tt>%lo=8</tt> and <tt>%hi=4</tt> would
4704 cause these bits to be set: <tt>0xFF1F</tt>.</p>
4707 llvm.part.set(0xFFFF, 0, 4, 7) -> 0xFF0F
4708 llvm.part.set(0xFFFF, 0, 7, 4) -> 0x0060
4709 llvm.part.set(0xFFFF, 0, 8, 3) -> 0x00F0
4710 llvm.part.set(0xFFFF, 0, 3, 8) -> 0xFE07
4714 <!-- ======================================================================= -->
4715 <div class="doc_subsection">
4716 <a name="int_debugger">Debugger Intrinsics</a>
4719 <div class="doc_text">
4721 The LLVM debugger intrinsics (which all start with <tt>llvm.dbg.</tt> prefix),
4722 are described in the <a
4723 href="SourceLevelDebugging.html#format_common_intrinsics">LLVM Source Level
4724 Debugging</a> document.
4729 <!-- ======================================================================= -->
4730 <div class="doc_subsection">
4731 <a name="int_eh">Exception Handling Intrinsics</a>
4734 <div class="doc_text">
4735 <p> The LLVM exception handling intrinsics (which all start with
4736 <tt>llvm.eh.</tt> prefix), are described in the <a
4737 href="ExceptionHandling.html#format_common_intrinsics">LLVM Exception
4738 Handling</a> document. </p>
4742 <!-- *********************************************************************** -->
4745 <a href="http://jigsaw.w3.org/css-validator/check/referer"><img
4746 src="http://jigsaw.w3.org/css-validator/images/vcss" alt="Valid CSS!"></a>
4747 <a href="http://validator.w3.org/check/referer"><img
4748 src="http://www.w3.org/Icons/valid-html401" alt="Valid HTML 4.01!" /></a>
4750 <a href="mailto:sabre@nondot.org">Chris Lattner</a><br>
4751 <a href="http://llvm.org">The LLVM Compiler Infrastructure</a><br>
4752 Last modified: $Date$