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="#moduleasm">Module-Level Inline Assembly</a></li>
30 <li><a href="#typesystem">Type System</a>
32 <li><a href="#t_primitive">Primitive Types</a>
34 <li><a href="#t_classifications">Type Classifications</a></li>
37 <li><a href="#t_derived">Derived Types</a>
39 <li><a href="#t_array">Array Type</a></li>
40 <li><a href="#t_function">Function Type</a></li>
41 <li><a href="#t_pointer">Pointer Type</a></li>
42 <li><a href="#t_struct">Structure Type</a></li>
43 <li><a href="#t_pstruct">Packed Structure Type</a></li>
44 <li><a href="#t_packed">Packed Type</a></li>
45 <li><a href="#t_opaque">Opaque Type</a></li>
50 <li><a href="#constants">Constants</a>
52 <li><a href="#simpleconstants">Simple Constants</a>
53 <li><a href="#aggregateconstants">Aggregate Constants</a>
54 <li><a href="#globalconstants">Global Variable and Function Addresses</a>
55 <li><a href="#undefvalues">Undefined Values</a>
56 <li><a href="#constantexprs">Constant Expressions</a>
59 <li><a href="#othervalues">Other Values</a>
61 <li><a href="#inlineasm">Inline Assembler Expressions</a>
64 <li><a href="#instref">Instruction Reference</a>
66 <li><a href="#terminators">Terminator Instructions</a>
68 <li><a href="#i_ret">'<tt>ret</tt>' Instruction</a></li>
69 <li><a href="#i_br">'<tt>br</tt>' Instruction</a></li>
70 <li><a href="#i_switch">'<tt>switch</tt>' Instruction</a></li>
71 <li><a href="#i_invoke">'<tt>invoke</tt>' Instruction</a></li>
72 <li><a href="#i_unwind">'<tt>unwind</tt>' Instruction</a></li>
73 <li><a href="#i_unreachable">'<tt>unreachable</tt>' Instruction</a></li>
76 <li><a href="#binaryops">Binary Operations</a>
78 <li><a href="#i_add">'<tt>add</tt>' Instruction</a></li>
79 <li><a href="#i_sub">'<tt>sub</tt>' Instruction</a></li>
80 <li><a href="#i_mul">'<tt>mul</tt>' Instruction</a></li>
81 <li><a href="#i_udiv">'<tt>udiv</tt>' Instruction</a></li>
82 <li><a href="#i_sdiv">'<tt>sdiv</tt>' Instruction</a></li>
83 <li><a href="#i_fdiv">'<tt>fdiv</tt>' Instruction</a></li>
84 <li><a href="#i_urem">'<tt>urem</tt>' Instruction</a></li>
85 <li><a href="#i_srem">'<tt>srem</tt>' Instruction</a></li>
86 <li><a href="#i_frem">'<tt>frem</tt>' Instruction</a></li>
89 <li><a href="#bitwiseops">Bitwise Binary Operations</a>
91 <li><a href="#i_and">'<tt>and</tt>' Instruction</a></li>
92 <li><a href="#i_or">'<tt>or</tt>' Instruction</a></li>
93 <li><a href="#i_xor">'<tt>xor</tt>' Instruction</a></li>
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>
99 <li><a href="#vectorops">Vector Operations</a>
101 <li><a href="#i_extractelement">'<tt>extractelement</tt>' Instruction</a></li>
102 <li><a href="#i_insertelement">'<tt>insertelement</tt>' Instruction</a></li>
103 <li><a href="#i_shufflevector">'<tt>shufflevector</tt>' Instruction</a></li>
106 <li><a href="#memoryops">Memory Access and Addressing Operations</a>
108 <li><a href="#i_malloc">'<tt>malloc</tt>' Instruction</a></li>
109 <li><a href="#i_free">'<tt>free</tt>' Instruction</a></li>
110 <li><a href="#i_alloca">'<tt>alloca</tt>' Instruction</a></li>
111 <li><a href="#i_load">'<tt>load</tt>' Instruction</a></li>
112 <li><a href="#i_store">'<tt>store</tt>' Instruction</a></li>
113 <li><a href="#i_getelementptr">'<tt>getelementptr</tt>' Instruction</a></li>
116 <li><a href="#convertops">Conversion Operations</a>
118 <li><a href="#i_trunc">'<tt>trunc .. to</tt>' Instruction</a></li>
119 <li><a href="#i_zext">'<tt>zext .. to</tt>' Instruction</a></li>
120 <li><a href="#i_sext">'<tt>sext .. to</tt>' Instruction</a></li>
121 <li><a href="#i_fptrunc">'<tt>fptrunc .. to</tt>' Instruction</a></li>
122 <li><a href="#i_fpext">'<tt>fpext .. to</tt>' Instruction</a></li>
123 <li><a href="#i_fptoui">'<tt>fptoui .. to</tt>' Instruction</a></li>
124 <li><a href="#i_fptosi">'<tt>fptosi .. to</tt>' Instruction</a></li>
125 <li><a href="#i_uitofp">'<tt>uitofp .. to</tt>' Instruction</a></li>
126 <li><a href="#i_sitofp">'<tt>sitofp .. to</tt>' Instruction</a></li>
127 <li><a href="#i_ptrtoint">'<tt>ptrtoint .. to</tt>' Instruction</a></li>
128 <li><a href="#i_inttoptr">'<tt>inttoptr .. to</tt>' Instruction</a></li>
129 <li><a href="#i_bitcast">'<tt>bitcast .. to</tt>' Instruction</a></li>
131 <li><a href="#otherops">Other Operations</a>
133 <li><a href="#i_icmp">'<tt>icmp</tt>' Instruction</a></li>
134 <li><a href="#i_fcmp">'<tt>fcmp</tt>' Instruction</a></li>
135 <li><a href="#i_phi">'<tt>phi</tt>' Instruction</a></li>
136 <li><a href="#i_select">'<tt>select</tt>' Instruction</a></li>
137 <li><a href="#i_call">'<tt>call</tt>' Instruction</a></li>
138 <li><a href="#i_va_arg">'<tt>va_arg</tt>' Instruction</a></li>
143 <li><a href="#intrinsics">Intrinsic Functions</a>
145 <li><a href="#int_varargs">Variable Argument Handling Intrinsics</a>
147 <li><a href="#i_va_start">'<tt>llvm.va_start</tt>' Intrinsic</a></li>
148 <li><a href="#i_va_end">'<tt>llvm.va_end</tt>' Intrinsic</a></li>
149 <li><a href="#i_va_copy">'<tt>llvm.va_copy</tt>' Intrinsic</a></li>
152 <li><a href="#int_gc">Accurate Garbage Collection Intrinsics</a>
154 <li><a href="#i_gcroot">'<tt>llvm.gcroot</tt>' Intrinsic</a></li>
155 <li><a href="#i_gcread">'<tt>llvm.gcread</tt>' Intrinsic</a></li>
156 <li><a href="#i_gcwrite">'<tt>llvm.gcwrite</tt>' Intrinsic</a></li>
159 <li><a href="#int_codegen">Code Generator Intrinsics</a>
161 <li><a href="#i_returnaddress">'<tt>llvm.returnaddress</tt>' Intrinsic</a></li>
162 <li><a href="#i_frameaddress">'<tt>llvm.frameaddress</tt>' Intrinsic</a></li>
163 <li><a href="#i_stacksave">'<tt>llvm.stacksave</tt>' Intrinsic</a></li>
164 <li><a href="#i_stackrestore">'<tt>llvm.stackrestore</tt>' Intrinsic</a></li>
165 <li><a href="#i_prefetch">'<tt>llvm.prefetch</tt>' Intrinsic</a></li>
166 <li><a href="#i_pcmarker">'<tt>llvm.pcmarker</tt>' Intrinsic</a></li>
167 <li><a href="#i_readcyclecounter"><tt>llvm.readcyclecounter</tt>' Intrinsic</a></li>
170 <li><a href="#int_libc">Standard C Library Intrinsics</a>
172 <li><a href="#i_memcpy">'<tt>llvm.memcpy.*</tt>' Intrinsic</a></li>
173 <li><a href="#i_memmove">'<tt>llvm.memmove.*</tt>' Intrinsic</a></li>
174 <li><a href="#i_memset">'<tt>llvm.memset.*</tt>' Intrinsic</a></li>
175 <li><a href="#i_isunordered">'<tt>llvm.isunordered.*</tt>' Intrinsic</a></li>
176 <li><a href="#i_sqrt">'<tt>llvm.sqrt.*</tt>' Intrinsic</a></li>
177 <li><a href="#i_powi">'<tt>llvm.powi.*</tt>' Intrinsic</a></li>
180 <li><a href="#int_manip">Bit Manipulation Intrinsics</a>
182 <li><a href="#i_bswap">'<tt>llvm.bswap.*</tt>' Intrinsics</a></li>
183 <li><a href="#int_ctpop">'<tt>llvm.ctpop.*</tt>' Intrinsic </a></li>
184 <li><a href="#int_ctlz">'<tt>llvm.ctlz.*</tt>' Intrinsic </a></li>
185 <li><a href="#int_cttz">'<tt>llvm.cttz.*</tt>' Intrinsic </a></li>
188 <li><a href="#int_debugger">Debugger intrinsics</a></li>
193 <div class="doc_author">
194 <p>Written by <a href="mailto:sabre@nondot.org">Chris Lattner</a>
195 and <a href="mailto:vadve@cs.uiuc.edu">Vikram Adve</a></p>
198 <!-- *********************************************************************** -->
199 <div class="doc_section"> <a name="abstract">Abstract </a></div>
200 <!-- *********************************************************************** -->
202 <div class="doc_text">
203 <p>This document is a reference manual for the LLVM assembly language.
204 LLVM is an SSA based representation that provides type safety,
205 low-level operations, flexibility, and the capability of representing
206 'all' high-level languages cleanly. It is the common code
207 representation used throughout all phases of the LLVM compilation
211 <!-- *********************************************************************** -->
212 <div class="doc_section"> <a name="introduction">Introduction</a> </div>
213 <!-- *********************************************************************** -->
215 <div class="doc_text">
217 <p>The LLVM code representation is designed to be used in three
218 different forms: as an in-memory compiler IR, as an on-disk bytecode
219 representation (suitable for fast loading by a Just-In-Time compiler),
220 and as a human readable assembly language representation. This allows
221 LLVM to provide a powerful intermediate representation for efficient
222 compiler transformations and analysis, while providing a natural means
223 to debug and visualize the transformations. The three different forms
224 of LLVM are all equivalent. This document describes the human readable
225 representation and notation.</p>
227 <p>The LLVM representation aims to be light-weight and low-level
228 while being expressive, typed, and extensible at the same time. It
229 aims to be a "universal IR" of sorts, by being at a low enough level
230 that high-level ideas may be cleanly mapped to it (similar to how
231 microprocessors are "universal IR's", allowing many source languages to
232 be mapped to them). By providing type information, LLVM can be used as
233 the target of optimizations: for example, through pointer analysis, it
234 can be proven that a C automatic variable is never accessed outside of
235 the current function... allowing it to be promoted to a simple SSA
236 value instead of a memory location.</p>
240 <!-- _______________________________________________________________________ -->
241 <div class="doc_subsubsection"> <a name="wellformed">Well-Formedness</a> </div>
243 <div class="doc_text">
245 <p>It is important to note that this document describes 'well formed'
246 LLVM assembly language. There is a difference between what the parser
247 accepts and what is considered 'well formed'. For example, the
248 following instruction is syntactically okay, but not well formed:</p>
251 %x = <a href="#i_add">add</a> int 1, %x
254 <p>...because the definition of <tt>%x</tt> does not dominate all of
255 its uses. The LLVM infrastructure provides a verification pass that may
256 be used to verify that an LLVM module is well formed. This pass is
257 automatically run by the parser after parsing input assembly and by
258 the optimizer before it outputs bytecode. The violations pointed out
259 by the verifier pass indicate bugs in transformation passes or input to
262 <!-- Describe the typesetting conventions here. --> </div>
264 <!-- *********************************************************************** -->
265 <div class="doc_section"> <a name="identifiers">Identifiers</a> </div>
266 <!-- *********************************************************************** -->
268 <div class="doc_text">
270 <p>LLVM uses three different forms of identifiers, for different
274 <li>Named values are represented as a string of characters with a '%' prefix.
275 For example, %foo, %DivisionByZero, %a.really.long.identifier. The actual
276 regular expression used is '<tt>%[a-zA-Z$._][a-zA-Z$._0-9]*</tt>'.
277 Identifiers which require other characters in their names can be surrounded
278 with quotes. In this way, anything except a <tt>"</tt> character can be used
281 <li>Unnamed values are represented as an unsigned numeric value with a '%'
282 prefix. For example, %12, %2, %44.</li>
284 <li>Constants, which are described in a <a href="#constants">section about
285 constants</a>, below.</li>
288 <p>LLVM requires that values start with a '%' sign for two reasons: Compilers
289 don't need to worry about name clashes with reserved words, and the set of
290 reserved words may be expanded in the future without penalty. Additionally,
291 unnamed identifiers allow a compiler to quickly come up with a temporary
292 variable without having to avoid symbol table conflicts.</p>
294 <p>Reserved words in LLVM are very similar to reserved words in other
295 languages. There are keywords for different opcodes
296 ('<tt><a href="#i_add">add</a></tt>',
297 '<tt><a href="#i_bitcast">bitcast</a></tt>',
298 '<tt><a href="#i_ret">ret</a></tt>', etc...), for primitive type names ('<tt><a
299 href="#t_void">void</a></tt>', '<tt><a href="#t_uint">uint</a></tt>', etc...),
300 and others. These reserved words cannot conflict with variable names, because
301 none of them start with a '%' character.</p>
303 <p>Here is an example of LLVM code to multiply the integer variable
304 '<tt>%X</tt>' by 8:</p>
309 %result = <a href="#i_mul">mul</a> uint %X, 8
312 <p>After strength reduction:</p>
315 %result = <a href="#i_shl">shl</a> uint %X, ubyte 3
318 <p>And the hard way:</p>
321 <a href="#i_add">add</a> uint %X, %X <i>; yields {uint}:%0</i>
322 <a href="#i_add">add</a> uint %0, %0 <i>; yields {uint}:%1</i>
323 %result = <a href="#i_add">add</a> uint %1, %1
326 <p>This last way of multiplying <tt>%X</tt> by 8 illustrates several
327 important lexical features of LLVM:</p>
331 <li>Comments are delimited with a '<tt>;</tt>' and go until the end of
334 <li>Unnamed temporaries are created when the result of a computation is not
335 assigned to a named value.</li>
337 <li>Unnamed temporaries are numbered sequentially</li>
341 <p>...and it also shows a convention that we follow in this document. When
342 demonstrating instructions, we will follow an instruction with a comment that
343 defines the type and name of value produced. Comments are shown in italic
348 <!-- *********************************************************************** -->
349 <div class="doc_section"> <a name="highlevel">High Level Structure</a> </div>
350 <!-- *********************************************************************** -->
352 <!-- ======================================================================= -->
353 <div class="doc_subsection"> <a name="modulestructure">Module Structure</a>
356 <div class="doc_text">
358 <p>LLVM programs are composed of "Module"s, each of which is a
359 translation unit of the input programs. Each module consists of
360 functions, global variables, and symbol table entries. Modules may be
361 combined together with the LLVM linker, which merges function (and
362 global variable) definitions, resolves forward declarations, and merges
363 symbol table entries. Here is an example of the "hello world" module:</p>
365 <pre><i>; Declare the string constant as a global constant...</i>
366 <a href="#identifiers">%.LC0</a> = <a href="#linkage_internal">internal</a> <a
367 href="#globalvars">constant</a> <a href="#t_array">[13 x sbyte]</a> c"hello world\0A\00" <i>; [13 x sbyte]*</i>
369 <i>; External declaration of the puts function</i>
370 <a href="#functionstructure">declare</a> int %puts(sbyte*) <i>; int(sbyte*)* </i>
372 <i>; Global variable / Function body section separator</i>
375 <i>; Definition of main function</i>
376 int %main() { <i>; int()* </i>
377 <i>; Convert [13x sbyte]* to sbyte *...</i>
379 href="#i_getelementptr">getelementptr</a> [13 x sbyte]* %.LC0, long 0, long 0 <i>; sbyte*</i>
381 <i>; Call puts function to write out the string to stdout...</i>
383 href="#i_call">call</a> int %puts(sbyte* %cast210) <i>; int</i>
385 href="#i_ret">ret</a> int 0<br>}<br></pre>
387 <p>This example is made up of a <a href="#globalvars">global variable</a>
388 named "<tt>.LC0</tt>", an external declaration of the "<tt>puts</tt>"
389 function, and a <a href="#functionstructure">function definition</a>
390 for "<tt>main</tt>".</p>
392 <p>In general, a module is made up of a list of global values,
393 where both functions and global variables are global values. Global values are
394 represented by a pointer to a memory location (in this case, a pointer to an
395 array of char, and a pointer to a function), and have one of the following <a
396 href="#linkage">linkage types</a>.</p>
398 <p>Due to a limitation in the current LLVM assembly parser (it is limited by
399 one-token lookahead), modules are split into two pieces by the "implementation"
400 keyword. Global variable prototypes and definitions must occur before the
401 keyword, and function definitions must occur after it. Function prototypes may
402 occur either before or after it. In the future, the implementation keyword may
403 become a noop, if the parser gets smarter.</p>
407 <!-- ======================================================================= -->
408 <div class="doc_subsection">
409 <a name="linkage">Linkage Types</a>
412 <div class="doc_text">
415 All Global Variables and Functions have one of the following types of linkage:
420 <dt><tt><b><a name="linkage_internal">internal</a></b></tt> </dt>
422 <dd>Global values with internal linkage are only directly accessible by
423 objects in the current module. In particular, linking code into a module with
424 an internal global value may cause the internal to be renamed as necessary to
425 avoid collisions. Because the symbol is internal to the module, all
426 references can be updated. This corresponds to the notion of the
427 '<tt>static</tt>' keyword in C, or the idea of "anonymous namespaces" in C++.
430 <dt><tt><b><a name="linkage_linkonce">linkonce</a></b></tt>: </dt>
432 <dd>"<tt>linkonce</tt>" linkage is similar to <tt>internal</tt> linkage, with
433 the twist that linking together two modules defining the same
434 <tt>linkonce</tt> globals will cause one of the globals to be discarded. This
435 is typically used to implement inline functions. Unreferenced
436 <tt>linkonce</tt> globals are allowed to be discarded.
439 <dt><tt><b><a name="linkage_weak">weak</a></b></tt>: </dt>
441 <dd>"<tt>weak</tt>" linkage is exactly the same as <tt>linkonce</tt> linkage,
442 except that unreferenced <tt>weak</tt> globals may not be discarded. This is
443 used to implement constructs in C such as "<tt>int X;</tt>" at global scope.
446 <dt><tt><b><a name="linkage_appending">appending</a></b></tt>: </dt>
448 <dd>"<tt>appending</tt>" linkage may only be applied to global variables of
449 pointer to array type. When two global variables with appending linkage are
450 linked together, the two global arrays are appended together. This is the
451 LLVM, typesafe, equivalent of having the system linker append together
452 "sections" with identical names when .o files are linked.
455 <dt><tt><b><a name="linkage_external">externally visible</a></b></tt>:</dt>
457 <dd>If none of the above identifiers are used, the global is externally
458 visible, meaning that it participates in linkage and can be used to resolve
459 external symbol references.
462 <dt><tt><b><a name="linkage_externweak">extern_weak</a></b></tt>: </dt>
464 <dd>"<tt>extern_weak</tt>" TBD
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)
473 <dt><tt><b><a name="linkage_dllimport">dllimport</a></b></tt>: </dt>
475 <dd>"<tt>dllimport</tt>" linkage causes the compiler to reference a function
476 or variable via a global pointer to a pointer that is set up by the DLL
477 exporting the symbol. On Microsoft Windows targets, the pointer name is
478 formed by combining <code>_imp__</code> and the function or variable name.
481 <dt><tt><b><a name="linkage_dllexport">dllexport</a></b></tt>: </dt>
483 <dd>"<tt>dllexport</tt>" linkage causes the compiler to provide a global
484 pointer to a pointer in a DLL, so that it can be referenced with the
485 <tt>dllimport</tt> attribute. On Microsoft Windows targets, the pointer
486 name is formed by combining <code>_imp__</code> and the function or variable
492 <p><a name="linkage_external">For example, since the "<tt>.LC0</tt>"
493 variable is defined to be internal, if another module defined a "<tt>.LC0</tt>"
494 variable and was linked with this one, one of the two would be renamed,
495 preventing a collision. Since "<tt>main</tt>" and "<tt>puts</tt>" are
496 external (i.e., lacking any linkage declarations), they are accessible
497 outside of the current module. It is illegal for a function <i>declaration</i>
498 to have any linkage type other than "externally visible".</a></p>
502 <!-- ======================================================================= -->
503 <div class="doc_subsection">
504 <a name="callingconv">Calling Conventions</a>
507 <div class="doc_text">
509 <p>LLVM <a href="#functionstructure">functions</a>, <a href="#i_call">calls</a>
510 and <a href="#i_invoke">invokes</a> can all have an optional calling convention
511 specified for the call. The calling convention of any pair of dynamic
512 caller/callee must match, or the behavior of the program is undefined. The
513 following calling conventions are supported by LLVM, and more may be added in
517 <dt><b>"<tt>ccc</tt>" - The C calling convention</b>:</dt>
519 <dd>This calling convention (the default if no other calling convention is
520 specified) matches the target C calling conventions. This calling convention
521 supports varargs function calls and tolerates some mismatch in the declared
522 prototype and implemented declaration of the function (as does normal C).
525 <dt><b>"<tt>csretcc</tt>" - The C struct return calling convention</b>:</dt>
527 <dd>This calling convention matches the target C calling conventions, except
528 that functions with this convention are required to take a pointer as their
529 first argument, and the return type of the function must be void. This is
530 used for C functions that return aggregates by-value. In this case, the
531 function has been transformed to take a pointer to the struct as the first
532 argument to the function. For targets where the ABI specifies specific
533 behavior for structure-return calls, the calling convention can be used to
534 distinguish between struct return functions and other functions that take a
535 pointer to a struct as the first argument.
538 <dt><b>"<tt>fastcc</tt>" - The fast calling convention</b>:</dt>
540 <dd>This calling convention attempts to make calls as fast as possible
541 (e.g. by passing things in registers). This calling convention allows the
542 target to use whatever tricks it wants to produce fast code for the target,
543 without having to conform to an externally specified ABI. Implementations of
544 this convention should allow arbitrary tail call optimization to be supported.
545 This calling convention does not support varargs and requires the prototype of
546 all callees to exactly match the prototype of the function definition.
549 <dt><b>"<tt>coldcc</tt>" - The cold calling convention</b>:</dt>
551 <dd>This calling convention attempts to make code in the caller as efficient
552 as possible under the assumption that the call is not commonly executed. As
553 such, these calls often preserve all registers so that the call does not break
554 any live ranges in the caller side. This calling convention does not support
555 varargs and requires the prototype of all callees to exactly match the
556 prototype of the function definition.
559 <dt><b>"<tt>cc <<em>n</em>></tt>" - Numbered convention</b>:</dt>
561 <dd>Any calling convention may be specified by number, allowing
562 target-specific calling conventions to be used. Target specific calling
563 conventions start at 64.
567 <p>More calling conventions can be added/defined on an as-needed basis, to
568 support pascal conventions or any other well-known target-independent
573 <!-- ======================================================================= -->
574 <div class="doc_subsection">
575 <a name="globalvars">Global Variables</a>
578 <div class="doc_text">
580 <p>Global variables define regions of memory allocated at compilation time
581 instead of run-time. Global variables may optionally be initialized, may have
582 an explicit section to be placed in, and may
583 have an optional explicit alignment specified. A
584 variable may be defined as a global "constant," which indicates that the
585 contents of the variable will <b>never</b> be modified (enabling better
586 optimization, allowing the global data to be placed in the read-only section of
587 an executable, etc). Note that variables that need runtime initialization
588 cannot be marked "constant" as there is a store to the variable.</p>
591 LLVM explicitly allows <em>declarations</em> of global variables to be marked
592 constant, even if the final definition of the global is not. This capability
593 can be used to enable slightly better optimization of the program, but requires
594 the language definition to guarantee that optimizations based on the
595 'constantness' are valid for the translation units that do not include the
599 <p>As SSA values, global variables define pointer values that are in
600 scope (i.e. they dominate) all basic blocks in the program. Global
601 variables always define a pointer to their "content" type because they
602 describe a region of memory, and all memory objects in LLVM are
603 accessed through pointers.</p>
605 <p>LLVM allows an explicit section to be specified for globals. If the target
606 supports it, it will emit globals to the section specified.</p>
608 <p>An explicit alignment may be specified for a global. If not present, or if
609 the alignment is set to zero, the alignment of the global is set by the target
610 to whatever it feels convenient. If an explicit alignment is specified, the
611 global is forced to have at least that much alignment. All alignments must be
617 <!-- ======================================================================= -->
618 <div class="doc_subsection">
619 <a name="functionstructure">Functions</a>
622 <div class="doc_text">
624 <p>LLVM function definitions consist of an optional <a href="#linkage">linkage
625 type</a>, an optional <a href="#callingconv">calling convention</a>, a return
626 type, a function name, a (possibly empty) argument list, an optional section,
627 an optional alignment, an opening curly brace,
628 a list of basic blocks, and a closing curly brace. LLVM function declarations
629 are defined with the "<tt>declare</tt>" keyword, an optional <a
630 href="#callingconv">calling convention</a>, a return type, a function name,
631 a possibly empty list of arguments, and an optional alignment.</p>
633 <p>A function definition contains a list of basic blocks, forming the CFG for
634 the function. Each basic block may optionally start with a label (giving the
635 basic block a symbol table entry), contains a list of instructions, and ends
636 with a <a href="#terminators">terminator</a> instruction (such as a branch or
637 function return).</p>
639 <p>The first basic block in a program is special in two ways: it is immediately
640 executed on entrance to the function, and it is not allowed to have predecessor
641 basic blocks (i.e. there can not be any branches to the entry block of a
642 function). Because the block can have no predecessors, it also cannot have any
643 <a href="#i_phi">PHI nodes</a>.</p>
645 <p>LLVM functions are identified by their name and type signature. Hence, two
646 functions with the same name but different parameter lists or return values are
647 considered different functions, and LLVM will resolve references to each
650 <p>LLVM allows an explicit section to be specified for functions. If the target
651 supports it, it will emit functions to the section specified.</p>
653 <p>An explicit alignment may be specified for a function. If not present, or if
654 the alignment is set to zero, the alignment of the function is set by the target
655 to whatever it feels convenient. If an explicit alignment is specified, the
656 function is forced to have at least that much alignment. All alignments must be
661 <!-- ======================================================================= -->
662 <div class="doc_subsection">
663 <a name="moduleasm">Module-Level Inline Assembly</a>
666 <div class="doc_text">
668 Modules may contain "module-level inline asm" blocks, which corresponds to the
669 GCC "file scope inline asm" blocks. These blocks are internally concatenated by
670 LLVM and treated as a single unit, but may be separated in the .ll file if
671 desired. The syntax is very simple:
674 <div class="doc_code"><pre>
675 module asm "inline asm code goes here"
676 module asm "more can go here"
679 <p>The strings can contain any character by escaping non-printable characters.
680 The escape sequence used is simply "\xx" where "xx" is the two digit hex code
685 The inline asm code is simply printed to the machine code .s file when
686 assembly code is generated.
691 <!-- *********************************************************************** -->
692 <div class="doc_section"> <a name="typesystem">Type System</a> </div>
693 <!-- *********************************************************************** -->
695 <div class="doc_text">
697 <p>The LLVM type system is one of the most important features of the
698 intermediate representation. Being typed enables a number of
699 optimizations to be performed on the IR directly, without having to do
700 extra analyses on the side before the transformation. A strong type
701 system makes it easier to read the generated code and enables novel
702 analyses and transformations that are not feasible to perform on normal
703 three address code representations.</p>
707 <!-- ======================================================================= -->
708 <div class="doc_subsection"> <a name="t_primitive">Primitive Types</a> </div>
709 <div class="doc_text">
710 <p>The primitive types are the fundamental building blocks of the LLVM
711 system. The current set of primitive types is as follows:</p>
713 <table class="layout">
718 <tr><th>Type</th><th>Description</th></tr>
719 <tr><td><tt>void</tt></td><td>No value</td></tr>
720 <tr><td><tt>ubyte</tt></td><td>Unsigned 8-bit value</td></tr>
721 <tr><td><tt>ushort</tt></td><td>Unsigned 16-bit value</td></tr>
722 <tr><td><tt>uint</tt></td><td>Unsigned 32-bit value</td></tr>
723 <tr><td><tt>ulong</tt></td><td>Unsigned 64-bit value</td></tr>
724 <tr><td><tt>float</tt></td><td>32-bit floating point value</td></tr>
725 <tr><td><tt>label</tt></td><td>Branch destination</td></tr>
732 <tr><th>Type</th><th>Description</th></tr>
733 <tr><td><tt>bool</tt></td><td>True or False value</td></tr>
734 <tr><td><tt>sbyte</tt></td><td>Signed 8-bit value</td></tr>
735 <tr><td><tt>short</tt></td><td>Signed 16-bit value</td></tr>
736 <tr><td><tt>int</tt></td><td>Signed 32-bit value</td></tr>
737 <tr><td><tt>long</tt></td><td>Signed 64-bit value</td></tr>
738 <tr><td><tt>double</tt></td><td>64-bit floating point value</td></tr>
746 <!-- _______________________________________________________________________ -->
747 <div class="doc_subsubsection"> <a name="t_classifications">Type
748 Classifications</a> </div>
749 <div class="doc_text">
750 <p>These different primitive types fall into a few useful
753 <table border="1" cellspacing="0" cellpadding="4">
755 <tr><th>Classification</th><th>Types</th></tr>
757 <td><a name="t_signed">signed</a></td>
758 <td><tt>sbyte, short, int, long, float, double</tt></td>
761 <td><a name="t_unsigned">unsigned</a></td>
762 <td><tt>ubyte, ushort, uint, ulong</tt></td>
765 <td><a name="t_integer">integer</a></td>
766 <td><tt>ubyte, sbyte, ushort, short, uint, int, ulong, long</tt></td>
769 <td><a name="t_integral">integral</a></td>
770 <td><tt>bool, ubyte, sbyte, ushort, short, uint, int, ulong, long</tt>
774 <td><a name="t_floating">floating point</a></td>
775 <td><tt>float, double</tt></td>
778 <td><a name="t_firstclass">first class</a></td>
779 <td><tt>bool, ubyte, sbyte, ushort, short, uint, int, ulong, long,<br>
780 float, double, <a href="#t_pointer">pointer</a>,
781 <a href="#t_packed">packed</a></tt></td>
786 <p>The <a href="#t_firstclass">first class</a> types are perhaps the
787 most important. Values of these types are the only ones which can be
788 produced by instructions, passed as arguments, or used as operands to
789 instructions. This means that all structures and arrays must be
790 manipulated either by pointer or by component.</p>
793 <!-- ======================================================================= -->
794 <div class="doc_subsection"> <a name="t_derived">Derived Types</a> </div>
796 <div class="doc_text">
798 <p>The real power in LLVM comes from the derived types in the system.
799 This is what allows a programmer to represent arrays, functions,
800 pointers, and other useful types. Note that these derived types may be
801 recursive: For example, it is possible to have a two dimensional array.</p>
805 <!-- _______________________________________________________________________ -->
806 <div class="doc_subsubsection"> <a name="t_array">Array Type</a> </div>
808 <div class="doc_text">
812 <p>The array type is a very simple derived type that arranges elements
813 sequentially in memory. The array type requires a size (number of
814 elements) and an underlying data type.</p>
819 [<# elements> x <elementtype>]
822 <p>The number of elements is a constant integer value; elementtype may
823 be any type with a size.</p>
826 <table class="layout">
829 <tt>[40 x int ]</tt><br/>
830 <tt>[41 x int ]</tt><br/>
831 <tt>[40 x uint]</tt><br/>
834 Array of 40 integer values.<br/>
835 Array of 41 integer values.<br/>
836 Array of 40 unsigned integer values.<br/>
840 <p>Here are some examples of multidimensional arrays:</p>
841 <table class="layout">
844 <tt>[3 x [4 x int]]</tt><br/>
845 <tt>[12 x [10 x float]]</tt><br/>
846 <tt>[2 x [3 x [4 x uint]]]</tt><br/>
849 3x4 array of integer values.<br/>
850 12x10 array of single precision floating point values.<br/>
851 2x3x4 array of unsigned integer values.<br/>
856 <p>Note that 'variable sized arrays' can be implemented in LLVM with a zero
857 length array. Normally, accesses past the end of an array are undefined in
858 LLVM (e.g. it is illegal to access the 5th element of a 3 element array).
859 As a special case, however, zero length arrays are recognized to be variable
860 length. This allows implementation of 'pascal style arrays' with the LLVM
861 type "{ int, [0 x float]}", for example.</p>
865 <!-- _______________________________________________________________________ -->
866 <div class="doc_subsubsection"> <a name="t_function">Function Type</a> </div>
867 <div class="doc_text">
869 <p>The function type can be thought of as a function signature. It
870 consists of a return type and a list of formal parameter types.
871 Function types are usually used to build virtual function tables
872 (which are structures of pointers to functions), for indirect function
873 calls, and when defining a function.</p>
875 The return type of a function type cannot be an aggregate type.
878 <pre> <returntype> (<parameter list>)<br></pre>
879 <p>...where '<tt><parameter list></tt>' is a comma-separated list of type
880 specifiers. Optionally, the parameter list may include a type <tt>...</tt>,
881 which indicates that the function takes a variable number of arguments.
882 Variable argument functions can access their arguments with the <a
883 href="#int_varargs">variable argument handling intrinsic</a> functions.</p>
885 <table class="layout">
888 <tt>int (int)</tt> <br/>
889 <tt>float (int, int *) *</tt><br/>
890 <tt>int (sbyte *, ...)</tt><br/>
893 function taking an <tt>int</tt>, returning an <tt>int</tt><br/>
894 <a href="#t_pointer">Pointer</a> to a function that takes an
895 <tt>int</tt> and a <a href="#t_pointer">pointer</a> to <tt>int</tt>,
896 returning <tt>float</tt>.<br/>
897 A vararg function that takes at least one <a href="#t_pointer">pointer</a>
898 to <tt>sbyte</tt> (signed char in C), which returns an integer. This is
899 the signature for <tt>printf</tt> in LLVM.<br/>
905 <!-- _______________________________________________________________________ -->
906 <div class="doc_subsubsection"> <a name="t_struct">Structure Type</a> </div>
907 <div class="doc_text">
909 <p>The structure type is used to represent a collection of data members
910 together in memory. The packing of the field types is defined to match
911 the ABI of the underlying processor. The elements of a structure may
912 be any type that has a size.</p>
913 <p>Structures are accessed using '<tt><a href="#i_load">load</a></tt>
914 and '<tt><a href="#i_store">store</a></tt>' by getting a pointer to a
915 field with the '<tt><a href="#i_getelementptr">getelementptr</a></tt>'
918 <pre> { <type list> }<br></pre>
920 <table class="layout">
923 <tt>{ int, int, int }</tt><br/>
924 <tt>{ float, int (int) * }</tt><br/>
927 a triple of three <tt>int</tt> values<br/>
928 A pair, where the first element is a <tt>float</tt> and the second element
929 is a <a href="#t_pointer">pointer</a> to a <a href="#t_function">function</a>
930 that takes an <tt>int</tt>, returning an <tt>int</tt>.<br/>
936 <!-- _______________________________________________________________________ -->
937 <div class="doc_subsubsection"> <a name="t_pstruct">Packed Structure Type</a>
939 <div class="doc_text">
941 <p>The packed structure type is used to represent a collection of data members
942 together in memory. There is no padding between fields. Further, the alignment
943 of a packed structure is 1 byte. The elements of a packed structure may
944 be any type that has a size.</p>
945 <p>Structures are accessed using '<tt><a href="#i_load">load</a></tt>
946 and '<tt><a href="#i_store">store</a></tt>' by getting a pointer to a
947 field with the '<tt><a href="#i_getelementptr">getelementptr</a></tt>'
950 <pre> < { <type list> } > <br></pre>
952 <table class="layout">
955 <tt> < { int, int, int } > </tt><br/>
956 <tt> < { float, int (int) * } > </tt><br/>
959 a triple of three <tt>int</tt> values<br/>
960 A pair, where the first element is a <tt>float</tt> and the second element
961 is a <a href="#t_pointer">pointer</a> to a <a href="#t_function">function</a>
962 that takes an <tt>int</tt>, returning an <tt>int</tt>.<br/>
968 <!-- _______________________________________________________________________ -->
969 <div class="doc_subsubsection"> <a name="t_pointer">Pointer Type</a> </div>
970 <div class="doc_text">
972 <p>As in many languages, the pointer type represents a pointer or
973 reference to another object, which must live in memory.</p>
975 <pre> <type> *<br></pre>
977 <table class="layout">
980 <tt>[4x int]*</tt><br/>
981 <tt>int (int *) *</tt><br/>
984 A <a href="#t_pointer">pointer</a> to <a href="#t_array">array</a> of
985 four <tt>int</tt> values<br/>
986 A <a href="#t_pointer">pointer</a> to a <a
987 href="#t_function">function</a> that takes an <tt>int*</tt>, returning an
994 <!-- _______________________________________________________________________ -->
995 <div class="doc_subsubsection"> <a name="t_packed">Packed Type</a> </div>
996 <div class="doc_text">
1000 <p>A packed type is a simple derived type that represents a vector
1001 of elements. Packed types are used when multiple primitive data
1002 are operated in parallel using a single instruction (SIMD).
1003 A packed type requires a size (number of
1004 elements) and an underlying primitive data type. Vectors must have a power
1005 of two length (1, 2, 4, 8, 16 ...). Packed types are
1006 considered <a href="#t_firstclass">first class</a>.</p>
1011 < <# elements> x <elementtype> >
1014 <p>The number of elements is a constant integer value; elementtype may
1015 be any integral or floating point type.</p>
1019 <table class="layout">
1022 <tt><4 x int></tt><br/>
1023 <tt><8 x float></tt><br/>
1024 <tt><2 x uint></tt><br/>
1027 Packed vector of 4 integer values.<br/>
1028 Packed vector of 8 floating-point values.<br/>
1029 Packed vector of 2 unsigned integer values.<br/>
1035 <!-- _______________________________________________________________________ -->
1036 <div class="doc_subsubsection"> <a name="t_opaque">Opaque Type</a> </div>
1037 <div class="doc_text">
1041 <p>Opaque types are used to represent unknown types in the system. This
1042 corresponds (for example) to the C notion of a foward declared structure type.
1043 In LLVM, opaque types can eventually be resolved to any type (not just a
1044 structure type).</p>
1054 <table class="layout">
1060 An opaque type.<br/>
1067 <!-- *********************************************************************** -->
1068 <div class="doc_section"> <a name="constants">Constants</a> </div>
1069 <!-- *********************************************************************** -->
1071 <div class="doc_text">
1073 <p>LLVM has several different basic types of constants. This section describes
1074 them all and their syntax.</p>
1078 <!-- ======================================================================= -->
1079 <div class="doc_subsection"><a name="simpleconstants">Simple Constants</a></div>
1081 <div class="doc_text">
1084 <dt><b>Boolean constants</b></dt>
1086 <dd>The two strings '<tt>true</tt>' and '<tt>false</tt>' are both valid
1087 constants of the <tt><a href="#t_primitive">bool</a></tt> type.
1090 <dt><b>Integer constants</b></dt>
1092 <dd>Standard integers (such as '4') are constants of the <a
1093 href="#t_integer">integer</a> type. Negative numbers may be used with signed
1097 <dt><b>Floating point constants</b></dt>
1099 <dd>Floating point constants use standard decimal notation (e.g. 123.421),
1100 exponential notation (e.g. 1.23421e+2), or a more precise hexadecimal
1101 notation (see below). Floating point constants must have a <a
1102 href="#t_floating">floating point</a> type. </dd>
1104 <dt><b>Null pointer constants</b></dt>
1106 <dd>The identifier '<tt>null</tt>' is recognized as a null pointer constant
1107 and must be of <a href="#t_pointer">pointer type</a>.</dd>
1111 <p>The one non-intuitive notation for constants is the optional hexadecimal form
1112 of floating point constants. For example, the form '<tt>double
1113 0x432ff973cafa8000</tt>' is equivalent to (but harder to read than) '<tt>double
1114 4.5e+15</tt>'. The only time hexadecimal floating point constants are required
1115 (and the only time that they are generated by the disassembler) is when a
1116 floating point constant must be emitted but it cannot be represented as a
1117 decimal floating point number. For example, NaN's, infinities, and other
1118 special values are represented in their IEEE hexadecimal format so that
1119 assembly and disassembly do not cause any bits to change in the constants.</p>
1123 <!-- ======================================================================= -->
1124 <div class="doc_subsection"><a name="aggregateconstants">Aggregate Constants</a>
1127 <div class="doc_text">
1128 <p>Aggregate constants arise from aggregation of simple constants
1129 and smaller aggregate constants.</p>
1132 <dt><b>Structure constants</b></dt>
1134 <dd>Structure constants are represented with notation similar to structure
1135 type definitions (a comma separated list of elements, surrounded by braces
1136 (<tt>{}</tt>)). For example: "<tt>{ int 4, float 17.0, int* %G }</tt>",
1137 where "<tt>%G</tt>" is declared as "<tt>%G = external global int</tt>". Structure constants
1138 must have <a href="#t_struct">structure type</a>, and the number and
1139 types of elements must match those specified by the type.
1142 <dt><b>Array constants</b></dt>
1144 <dd>Array constants are represented with notation similar to array type
1145 definitions (a comma separated list of elements, surrounded by square brackets
1146 (<tt>[]</tt>)). For example: "<tt>[ int 42, int 11, int 74 ]</tt>". Array
1147 constants must have <a href="#t_array">array type</a>, and the number and
1148 types of elements must match those specified by the type.
1151 <dt><b>Packed constants</b></dt>
1153 <dd>Packed constants are represented with notation similar to packed type
1154 definitions (a comma separated list of elements, surrounded by
1155 less-than/greater-than's (<tt><></tt>)). For example: "<tt>< int 42,
1156 int 11, int 74, int 100 ></tt>". Packed constants must have <a
1157 href="#t_packed">packed type</a>, and the number and types of elements must
1158 match those specified by the type.
1161 <dt><b>Zero initialization</b></dt>
1163 <dd>The string '<tt>zeroinitializer</tt>' can be used to zero initialize a
1164 value to zero of <em>any</em> type, including scalar and aggregate types.
1165 This is often used to avoid having to print large zero initializers (e.g. for
1166 large arrays) and is always exactly equivalent to using explicit zero
1173 <!-- ======================================================================= -->
1174 <div class="doc_subsection">
1175 <a name="globalconstants">Global Variable and Function Addresses</a>
1178 <div class="doc_text">
1180 <p>The addresses of <a href="#globalvars">global variables</a> and <a
1181 href="#functionstructure">functions</a> are always implicitly valid (link-time)
1182 constants. These constants are explicitly referenced when the <a
1183 href="#identifiers">identifier for the global</a> is used and always have <a
1184 href="#t_pointer">pointer</a> type. For example, the following is a legal LLVM
1190 %Z = global [2 x int*] [ int* %X, int* %Y ]
1195 <!-- ======================================================================= -->
1196 <div class="doc_subsection"><a name="undefvalues">Undefined Values</a></div>
1197 <div class="doc_text">
1198 <p>The string '<tt>undef</tt>' is recognized as a type-less constant that has
1199 no specific value. Undefined values may be of any type and be used anywhere
1200 a constant is permitted.</p>
1202 <p>Undefined values indicate to the compiler that the program is well defined
1203 no matter what value is used, giving the compiler more freedom to optimize.
1207 <!-- ======================================================================= -->
1208 <div class="doc_subsection"><a name="constantexprs">Constant Expressions</a>
1211 <div class="doc_text">
1213 <p>Constant expressions are used to allow expressions involving other constants
1214 to be used as constants. Constant expressions may be of any <a
1215 href="#t_firstclass">first class</a> type and may involve any LLVM operation
1216 that does not have side effects (e.g. load and call are not supported). The
1217 following is the syntax for constant expressions:</p>
1220 <dt><b><tt>trunc ( CST to TYPE )</tt></b></dt>
1221 <dd>Truncate a constant to another type. The bit size of CST must be larger
1222 than the bit size of TYPE. Both types must be integral.</dd>
1224 <dt><b><tt>zext ( CST to TYPE )</tt></b></dt>
1225 <dd>Zero extend a constant to another type. The bit size of CST must be
1226 smaller or equal to the bit size of TYPE. Both types must be integral.</dd>
1228 <dt><b><tt>sext ( CST to TYPE )</tt></b></dt>
1229 <dd>Sign extend a constant to another type. The bit size of CST must be
1230 smaller or equal to the bit size of TYPE. Both types must be integral.</dd>
1232 <dt><b><tt>fptrunc ( CST to TYPE )</tt></b></dt>
1233 <dd>Truncate a floating point constant to another floating point type. The
1234 size of CST must be larger than the size of TYPE. Both types must be
1235 floating point.</dd>
1237 <dt><b><tt>fpext ( CST to TYPE )</tt></b></dt>
1238 <dd>Floating point extend a constant to another type. The size of CST must be
1239 smaller or equal to the size of TYPE. Both types must be floating point.</dd>
1241 <dt><b><tt>fp2uint ( CST to TYPE )</tt></b></dt>
1242 <dd>Convert a floating point constant to the corresponding unsigned integer
1243 constant. TYPE must be an integer type. CST must be floating point. If the
1244 value won't fit in the integer type, the results are undefined.</dd>
1246 <dt><b><tt>fptosi ( CST to TYPE )</tt></b></dt>
1247 <dd>Convert a floating point constant to the corresponding signed integer
1248 constant. TYPE must be an integer type. CST must be floating point. If the
1249 value won't fit in the integer type, the results are undefined.</dd>
1251 <dt><b><tt>uitofp ( CST to TYPE )</tt></b></dt>
1252 <dd>Convert an unsigned integer constant to the corresponding floating point
1253 constant. TYPE must be floating point. CST must be of integer type. If the
1254 value won't fit in the floating point type, the results are undefined.</dd>
1256 <dt><b><tt>sitofp ( CST to TYPE )</tt></b></dt>
1257 <dd>Convert a signed integer constant to the corresponding floating point
1258 constant. TYPE must be floating point. CST must be of integer type. If the
1259 value won't fit in the floating point type, the results are undefined.</dd>
1261 <dt><b><tt>ptrtoint ( CST to TYPE )</tt></b></dt>
1262 <dd>Convert a pointer typed constant to the corresponding integer constant
1263 TYPE must be an integer type. CST must be of pointer type. The CST value is
1264 zero extended, truncated, or unchanged to make it fit in TYPE.</dd>
1266 <dt><b><tt>inttoptr ( CST to TYPE )</tt></b></dt>
1267 <dd>Convert a integer constant to a pointer constant. TYPE must be a
1268 pointer type. CST must be of integer type. The CST value is zero extended,
1269 truncated, or unchanged to make it fit in a pointer size. This one is
1270 <i>really</i> dangerous!</dd>
1272 <dt><b><tt>bitcast ( CST to TYPE )</tt></b></dt>
1273 <dd>Convert a constant, CST, to another TYPE. The size of CST and TYPE must be
1274 identical (same number of bits). The conversion is done as if the CST value
1275 was stored to memory and read back as TYPE. In other words, no bits change
1276 with this operator, just the type. This can be used for conversion of
1277 packed types to any other type, as long as they have the same bit width. For
1278 pointers it is only valid to cast to another pointer type.
1281 <dt><b><tt>getelementptr ( CSTPTR, IDX0, IDX1, ... )</tt></b></dt>
1283 <dd>Perform the <a href="#i_getelementptr">getelementptr operation</a> on
1284 constants. As with the <a href="#i_getelementptr">getelementptr</a>
1285 instruction, the index list may have zero or more indexes, which are required
1286 to make sense for the type of "CSTPTR".</dd>
1288 <dt><b><tt>select ( COND, VAL1, VAL2 )</tt></b></dt>
1290 <dd>Perform the <a href="#i_select">select operation</a> on
1293 <dt><b><tt>icmp COND ( VAL1, VAL2 )</tt></b></dt>
1294 <dd>Performs the <a href="#i_icmp">icmp operation</a> on constants.</dd>
1296 <dt><b><tt>fcmp COND ( VAL1, VAL2 )</tt></b></dt>
1297 <dd>Performs the <a href="#i_fcmp">fcmp operation</a> on constants.</dd>
1299 <dt><b><tt>extractelement ( VAL, IDX )</tt></b></dt>
1301 <dd>Perform the <a href="#i_extractelement">extractelement
1302 operation</a> on constants.
1304 <dt><b><tt>insertelement ( VAL, ELT, IDX )</tt></b></dt>
1306 <dd>Perform the <a href="#i_insertelement">insertelement
1307 operation</a> on constants.</dd>
1310 <dt><b><tt>shufflevector ( VEC1, VEC2, IDXMASK )</tt></b></dt>
1312 <dd>Perform the <a href="#i_shufflevector">shufflevector
1313 operation</a> on constants.</dd>
1315 <dt><b><tt>OPCODE ( LHS, RHS )</tt></b></dt>
1317 <dd>Perform the specified operation of the LHS and RHS constants. OPCODE may
1318 be any of the <a href="#binaryops">binary</a> or <a href="#bitwiseops">bitwise
1319 binary</a> operations. The constraints on operands are the same as those for
1320 the corresponding instruction (e.g. no bitwise operations on floating point
1321 values are allowed).</dd>
1325 <!-- *********************************************************************** -->
1326 <div class="doc_section"> <a name="othervalues">Other Values</a> </div>
1327 <!-- *********************************************************************** -->
1329 <!-- ======================================================================= -->
1330 <div class="doc_subsection">
1331 <a name="inlineasm">Inline Assembler Expressions</a>
1334 <div class="doc_text">
1337 LLVM supports inline assembler expressions (as opposed to <a href="#moduleasm">
1338 Module-Level Inline Assembly</a>) through the use of a special value. This
1339 value represents the inline assembler as a string (containing the instructions
1340 to emit), a list of operand constraints (stored as a string), and a flag that
1341 indicates whether or not the inline asm expression has side effects. An example
1342 inline assembler expression is:
1346 int(int) asm "bswap $0", "=r,r"
1350 Inline assembler expressions may <b>only</b> be used as the callee operand of
1351 a <a href="#i_call"><tt>call</tt> instruction</a>. Thus, typically we have:
1355 %X = call int asm "<a href="#i_bswap">bswap</a> $0", "=r,r"(int %Y)
1359 Inline asms with side effects not visible in the constraint list must be marked
1360 as having side effects. This is done through the use of the
1361 '<tt>sideeffect</tt>' keyword, like so:
1365 call void asm sideeffect "eieio", ""()
1368 <p>TODO: The format of the asm and constraints string still need to be
1369 documented here. Constraints on what can be done (e.g. duplication, moving, etc
1370 need to be documented).
1375 <!-- *********************************************************************** -->
1376 <div class="doc_section"> <a name="instref">Instruction Reference</a> </div>
1377 <!-- *********************************************************************** -->
1379 <div class="doc_text">
1381 <p>The LLVM instruction set consists of several different
1382 classifications of instructions: <a href="#terminators">terminator
1383 instructions</a>, <a href="#binaryops">binary instructions</a>,
1384 <a href="#bitwiseops">bitwise binary instructions</a>, <a
1385 href="#memoryops">memory instructions</a>, and <a href="#otherops">other
1386 instructions</a>.</p>
1390 <!-- ======================================================================= -->
1391 <div class="doc_subsection"> <a name="terminators">Terminator
1392 Instructions</a> </div>
1394 <div class="doc_text">
1396 <p>As mentioned <a href="#functionstructure">previously</a>, every
1397 basic block in a program ends with a "Terminator" instruction, which
1398 indicates which block should be executed after the current block is
1399 finished. These terminator instructions typically yield a '<tt>void</tt>'
1400 value: they produce control flow, not values (the one exception being
1401 the '<a href="#i_invoke"><tt>invoke</tt></a>' instruction).</p>
1402 <p>There are six different terminator instructions: the '<a
1403 href="#i_ret"><tt>ret</tt></a>' instruction, the '<a href="#i_br"><tt>br</tt></a>'
1404 instruction, the '<a href="#i_switch"><tt>switch</tt></a>' instruction,
1405 the '<a href="#i_invoke"><tt>invoke</tt></a>' instruction, the '<a
1406 href="#i_unwind"><tt>unwind</tt></a>' instruction, and the '<a
1407 href="#i_unreachable"><tt>unreachable</tt></a>' instruction.</p>
1411 <!-- _______________________________________________________________________ -->
1412 <div class="doc_subsubsection"> <a name="i_ret">'<tt>ret</tt>'
1413 Instruction</a> </div>
1414 <div class="doc_text">
1416 <pre> ret <type> <value> <i>; Return a value from a non-void function</i>
1417 ret void <i>; Return from void function</i>
1420 <p>The '<tt>ret</tt>' instruction is used to return control flow (and a
1421 value) from a function back to the caller.</p>
1422 <p>There are two forms of the '<tt>ret</tt>' instruction: one that
1423 returns a value and then causes control flow, and one that just causes
1424 control flow to occur.</p>
1426 <p>The '<tt>ret</tt>' instruction may return any '<a
1427 href="#t_firstclass">first class</a>' type. Notice that a function is
1428 not <a href="#wellformed">well formed</a> if there exists a '<tt>ret</tt>'
1429 instruction inside of the function that returns a value that does not
1430 match the return type of the function.</p>
1432 <p>When the '<tt>ret</tt>' instruction is executed, control flow
1433 returns back to the calling function's context. If the caller is a "<a
1434 href="#i_call"><tt>call</tt></a>" instruction, execution continues at
1435 the instruction after the call. If the caller was an "<a
1436 href="#i_invoke"><tt>invoke</tt></a>" instruction, execution continues
1437 at the beginning of the "normal" destination block. If the instruction
1438 returns a value, that value shall set the call or invoke instruction's
1441 <pre> ret int 5 <i>; Return an integer value of 5</i>
1442 ret void <i>; Return from a void function</i>
1445 <!-- _______________________________________________________________________ -->
1446 <div class="doc_subsubsection"> <a name="i_br">'<tt>br</tt>' Instruction</a> </div>
1447 <div class="doc_text">
1449 <pre> br bool <cond>, label <iftrue>, label <iffalse><br> br label <dest> <i>; Unconditional branch</i>
1452 <p>The '<tt>br</tt>' instruction is used to cause control flow to
1453 transfer to a different basic block in the current function. There are
1454 two forms of this instruction, corresponding to a conditional branch
1455 and an unconditional branch.</p>
1457 <p>The conditional branch form of the '<tt>br</tt>' instruction takes a
1458 single '<tt>bool</tt>' value and two '<tt>label</tt>' values. The
1459 unconditional form of the '<tt>br</tt>' instruction takes a single '<tt>label</tt>'
1460 value as a target.</p>
1462 <p>Upon execution of a conditional '<tt>br</tt>' instruction, the '<tt>bool</tt>'
1463 argument is evaluated. If the value is <tt>true</tt>, control flows
1464 to the '<tt>iftrue</tt>' <tt>label</tt> argument. If "cond" is <tt>false</tt>,
1465 control flows to the '<tt>iffalse</tt>' <tt>label</tt> argument.</p>
1467 <pre>Test:<br> %cond = <a href="#i_icmp">icmp</a> eq, int %a, %b<br> br bool %cond, label %IfEqual, label %IfUnequal<br>IfEqual:<br> <a
1468 href="#i_ret">ret</a> int 1<br>IfUnequal:<br> <a href="#i_ret">ret</a> int 0<br></pre>
1470 <!-- _______________________________________________________________________ -->
1471 <div class="doc_subsubsection">
1472 <a name="i_switch">'<tt>switch</tt>' Instruction</a>
1475 <div class="doc_text">
1479 switch <intty> <value>, label <defaultdest> [ <intty> <val>, label <dest> ... ]
1484 <p>The '<tt>switch</tt>' instruction is used to transfer control flow to one of
1485 several different places. It is a generalization of the '<tt>br</tt>'
1486 instruction, allowing a branch to occur to one of many possible
1492 <p>The '<tt>switch</tt>' instruction uses three parameters: an integer
1493 comparison value '<tt>value</tt>', a default '<tt>label</tt>' destination, and
1494 an array of pairs of comparison value constants and '<tt>label</tt>'s. The
1495 table is not allowed to contain duplicate constant entries.</p>
1499 <p>The <tt>switch</tt> instruction specifies a table of values and
1500 destinations. When the '<tt>switch</tt>' instruction is executed, this
1501 table is searched for the given value. If the value is found, control flow is
1502 transfered to the corresponding destination; otherwise, control flow is
1503 transfered to the default destination.</p>
1505 <h5>Implementation:</h5>
1507 <p>Depending on properties of the target machine and the particular
1508 <tt>switch</tt> instruction, this instruction may be code generated in different
1509 ways. For example, it could be generated as a series of chained conditional
1510 branches or with a lookup table.</p>
1515 <i>; Emulate a conditional br instruction</i>
1516 %Val = <a href="#i_zext">zext</a> bool %value to int
1517 switch int %Val, label %truedest [int 0, label %falsedest ]
1519 <i>; Emulate an unconditional br instruction</i>
1520 switch uint 0, label %dest [ ]
1522 <i>; Implement a jump table:</i>
1523 switch uint %val, label %otherwise [ uint 0, label %onzero
1524 uint 1, label %onone
1525 uint 2, label %ontwo ]
1529 <!-- _______________________________________________________________________ -->
1530 <div class="doc_subsubsection">
1531 <a name="i_invoke">'<tt>invoke</tt>' Instruction</a>
1534 <div class="doc_text">
1539 <result> = invoke [<a href="#callingconv">cconv</a>] <ptr to function ty> %<function ptr val>(<function args>)
1540 to label <normal label> unwind label <exception label>
1545 <p>The '<tt>invoke</tt>' instruction causes control to transfer to a specified
1546 function, with the possibility of control flow transfer to either the
1547 '<tt>normal</tt>' label or the
1548 '<tt>exception</tt>' label. If the callee function returns with the
1549 "<tt><a href="#i_ret">ret</a></tt>" instruction, control flow will return to the
1550 "normal" label. If the callee (or any indirect callees) returns with the "<a
1551 href="#i_unwind"><tt>unwind</tt></a>" instruction, control is interrupted and
1552 continued at the dynamically nearest "exception" label.</p>
1556 <p>This instruction requires several arguments:</p>
1560 The optional "cconv" marker indicates which <a href="callingconv">calling
1561 convention</a> the call should use. If none is specified, the call defaults
1562 to using C calling conventions.
1564 <li>'<tt>ptr to function ty</tt>': shall be the signature of the pointer to
1565 function value being invoked. In most cases, this is a direct function
1566 invocation, but indirect <tt>invoke</tt>s are just as possible, branching off
1567 an arbitrary pointer to function value.
1570 <li>'<tt>function ptr val</tt>': An LLVM value containing a pointer to a
1571 function to be invoked. </li>
1573 <li>'<tt>function args</tt>': argument list whose types match the function
1574 signature argument types. If the function signature indicates the function
1575 accepts a variable number of arguments, the extra arguments can be
1578 <li>'<tt>normal label</tt>': the label reached when the called function
1579 executes a '<tt><a href="#i_ret">ret</a></tt>' instruction. </li>
1581 <li>'<tt>exception label</tt>': the label reached when a callee returns with
1582 the <a href="#i_unwind"><tt>unwind</tt></a> instruction. </li>
1588 <p>This instruction is designed to operate as a standard '<tt><a
1589 href="#i_call">call</a></tt>' instruction in most regards. The primary
1590 difference is that it establishes an association with a label, which is used by
1591 the runtime library to unwind the stack.</p>
1593 <p>This instruction is used in languages with destructors to ensure that proper
1594 cleanup is performed in the case of either a <tt>longjmp</tt> or a thrown
1595 exception. Additionally, this is important for implementation of
1596 '<tt>catch</tt>' clauses in high-level languages that support them.</p>
1600 %retval = invoke int %Test(int 15) to label %Continue
1601 unwind label %TestCleanup <i>; {int}:retval set</i>
1602 %retval = invoke <a href="#callingconv">coldcc</a> int %Test(int 15) to label %Continue
1603 unwind label %TestCleanup <i>; {int}:retval set</i>
1608 <!-- _______________________________________________________________________ -->
1610 <div class="doc_subsubsection"> <a name="i_unwind">'<tt>unwind</tt>'
1611 Instruction</a> </div>
1613 <div class="doc_text">
1622 <p>The '<tt>unwind</tt>' instruction unwinds the stack, continuing control flow
1623 at the first callee in the dynamic call stack which used an <a
1624 href="#i_invoke"><tt>invoke</tt></a> instruction to perform the call. This is
1625 primarily used to implement exception handling.</p>
1629 <p>The '<tt>unwind</tt>' intrinsic causes execution of the current function to
1630 immediately halt. The dynamic call stack is then searched for the first <a
1631 href="#i_invoke"><tt>invoke</tt></a> instruction on the call stack. Once found,
1632 execution continues at the "exceptional" destination block specified by the
1633 <tt>invoke</tt> instruction. If there is no <tt>invoke</tt> instruction in the
1634 dynamic call chain, undefined behavior results.</p>
1637 <!-- _______________________________________________________________________ -->
1639 <div class="doc_subsubsection"> <a name="i_unreachable">'<tt>unreachable</tt>'
1640 Instruction</a> </div>
1642 <div class="doc_text">
1651 <p>The '<tt>unreachable</tt>' instruction has no defined semantics. This
1652 instruction is used to inform the optimizer that a particular portion of the
1653 code is not reachable. This can be used to indicate that the code after a
1654 no-return function cannot be reached, and other facts.</p>
1658 <p>The '<tt>unreachable</tt>' instruction has no defined semantics.</p>
1663 <!-- ======================================================================= -->
1664 <div class="doc_subsection"> <a name="binaryops">Binary Operations</a> </div>
1665 <div class="doc_text">
1666 <p>Binary operators are used to do most of the computation in a
1667 program. They require two operands, execute an operation on them, and
1668 produce a single value. The operands might represent
1669 multiple data, as is the case with the <a href="#t_packed">packed</a> data type.
1670 The result value of a binary operator is not
1671 necessarily the same type as its operands.</p>
1672 <p>There are several different binary operators:</p>
1674 <!-- _______________________________________________________________________ -->
1675 <div class="doc_subsubsection"> <a name="i_add">'<tt>add</tt>'
1676 Instruction</a> </div>
1677 <div class="doc_text">
1679 <pre> <result> = add <ty> <var1>, <var2> <i>; yields {ty}:result</i>
1682 <p>The '<tt>add</tt>' instruction returns the sum of its two operands.</p>
1684 <p>The two arguments to the '<tt>add</tt>' instruction must be either <a
1685 href="#t_integer">integer</a> or <a href="#t_floating">floating point</a> values.
1686 This instruction can also take <a href="#t_packed">packed</a> versions of the values.
1687 Both arguments must have identical types.</p>
1689 <p>The value produced is the integer or floating point sum of the two
1692 <pre> <result> = add int 4, %var <i>; yields {int}:result = 4 + %var</i>
1695 <!-- _______________________________________________________________________ -->
1696 <div class="doc_subsubsection"> <a name="i_sub">'<tt>sub</tt>'
1697 Instruction</a> </div>
1698 <div class="doc_text">
1700 <pre> <result> = sub <ty> <var1>, <var2> <i>; yields {ty}:result</i>
1703 <p>The '<tt>sub</tt>' instruction returns the difference of its two
1705 <p>Note that the '<tt>sub</tt>' instruction is used to represent the '<tt>neg</tt>'
1706 instruction present in most other intermediate representations.</p>
1708 <p>The two arguments to the '<tt>sub</tt>' instruction must be either <a
1709 href="#t_integer">integer</a> or <a href="#t_floating">floating point</a>
1711 This instruction can also take <a href="#t_packed">packed</a> versions of the values.
1712 Both arguments must have identical types.</p>
1714 <p>The value produced is the integer or floating point difference of
1715 the two operands.</p>
1717 <pre> <result> = sub int 4, %var <i>; yields {int}:result = 4 - %var</i>
1718 <result> = sub int 0, %val <i>; yields {int}:result = -%var</i>
1721 <!-- _______________________________________________________________________ -->
1722 <div class="doc_subsubsection"> <a name="i_mul">'<tt>mul</tt>'
1723 Instruction</a> </div>
1724 <div class="doc_text">
1726 <pre> <result> = mul <ty> <var1>, <var2> <i>; yields {ty}:result</i>
1729 <p>The '<tt>mul</tt>' instruction returns the product of its two
1732 <p>The two arguments to the '<tt>mul</tt>' instruction must be either <a
1733 href="#t_integer">integer</a> or <a href="#t_floating">floating point</a>
1735 This instruction can also take <a href="#t_packed">packed</a> versions of the values.
1736 Both arguments must have identical types.</p>
1738 <p>The value produced is the integer or floating point product of the
1740 <p>There is no signed vs unsigned multiplication. The appropriate
1741 action is taken based on the type of the operand.</p>
1743 <pre> <result> = mul int 4, %var <i>; yields {int}:result = 4 * %var</i>
1746 <!-- _______________________________________________________________________ -->
1747 <div class="doc_subsubsection"> <a name="i_udiv">'<tt>udiv</tt>' Instruction
1749 <div class="doc_text">
1751 <pre> <result> = udiv <ty> <var1>, <var2> <i>; yields {ty}:result</i>
1754 <p>The '<tt>udiv</tt>' instruction returns the quotient of its two
1757 <p>The two arguments to the '<tt>udiv</tt>' instruction must be
1758 <a href="#t_integer">integer</a> values. Both arguments must have identical
1759 types. This instruction can also take <a href="#t_packed">packed</a> versions
1760 of the values in which case the elements must be integers.</p>
1762 <p>The value produced is the unsigned integer quotient of the two operands. This
1763 instruction always performs an unsigned division operation, regardless of
1764 whether the arguments are unsigned or not.</p>
1766 <pre> <result> = udiv uint 4, %var <i>; yields {uint}:result = 4 / %var</i>
1769 <!-- _______________________________________________________________________ -->
1770 <div class="doc_subsubsection"> <a name="i_sdiv">'<tt>sdiv</tt>' Instruction
1772 <div class="doc_text">
1774 <pre> <result> = sdiv <ty> <var1>, <var2> <i>; yields {ty}:result</i>
1777 <p>The '<tt>sdiv</tt>' instruction returns the quotient of its two
1780 <p>The two arguments to the '<tt>sdiv</tt>' instruction must be
1781 <a href="#t_integer">integer</a> values. Both arguments must have identical
1782 types. This instruction can also take <a href="#t_packed">packed</a> versions
1783 of the values in which case the elements must be integers.</p>
1785 <p>The value produced is the signed integer quotient of the two operands. This
1786 instruction always performs a signed division operation, regardless of whether
1787 the arguments are signed or not.</p>
1789 <pre> <result> = sdiv int 4, %var <i>; yields {int}:result = 4 / %var</i>
1792 <!-- _______________________________________________________________________ -->
1793 <div class="doc_subsubsection"> <a name="i_fdiv">'<tt>fdiv</tt>'
1794 Instruction</a> </div>
1795 <div class="doc_text">
1797 <pre> <result> = fdiv <ty> <var1>, <var2> <i>; yields {ty}:result</i>
1800 <p>The '<tt>fdiv</tt>' instruction returns the quotient of its two
1803 <p>The two arguments to the '<tt>div</tt>' instruction must be
1804 <a href="#t_floating">floating point</a> values. Both arguments must have
1805 identical types. This instruction can also take <a href="#t_packed">packed</a>
1806 versions of the values in which case the elements must be floating point.</p>
1808 <p>The value produced is the floating point quotient of the two operands.</p>
1810 <pre> <result> = fdiv float 4.0, %var <i>; yields {float}:result = 4.0 / %var</i>
1813 <!-- _______________________________________________________________________ -->
1814 <div class="doc_subsubsection"> <a name="i_urem">'<tt>urem</tt>' Instruction</a>
1816 <div class="doc_text">
1818 <pre> <result> = urem <ty> <var1>, <var2> <i>; yields {ty}:result</i>
1821 <p>The '<tt>urem</tt>' instruction returns the remainder from the
1822 unsigned division of its two arguments.</p>
1824 <p>The two arguments to the '<tt>urem</tt>' instruction must be
1825 <a href="#t_integer">integer</a> values. Both arguments must have identical
1828 <p>This instruction returns the unsigned integer <i>remainder</i> of a division.
1829 This instruction always performs an unsigned division to get the remainder,
1830 regardless of whether the arguments are unsigned or not.</p>
1832 <pre> <result> = urem uint 4, %var <i>; yields {uint}:result = 4 % %var</i>
1836 <!-- _______________________________________________________________________ -->
1837 <div class="doc_subsubsection"> <a name="i_srem">'<tt>srem</tt>'
1838 Instruction</a> </div>
1839 <div class="doc_text">
1841 <pre> <result> = srem <ty> <var1>, <var2> <i>; yields {ty}:result</i>
1844 <p>The '<tt>srem</tt>' instruction returns the remainder from the
1845 signed division of its two operands.</p>
1847 <p>The two arguments to the '<tt>srem</tt>' instruction must be
1848 <a href="#t_integer">integer</a> values. Both arguments must have identical
1851 <p>This instruction returns the <i>remainder</i> of a division (where the result
1852 has the same sign as the divisor), not the <i>modulus</i> (where the
1853 result has the same sign as the dividend) of a value. For more
1854 information about the difference, see <a
1855 href="http://mathforum.org/dr.math/problems/anne.4.28.99.html">The
1858 <pre> <result> = srem int 4, %var <i>; yields {int}:result = 4 % %var</i>
1862 <!-- _______________________________________________________________________ -->
1863 <div class="doc_subsubsection"> <a name="i_frem">'<tt>frem</tt>'
1864 Instruction</a> </div>
1865 <div class="doc_text">
1867 <pre> <result> = frem <ty> <var1>, <var2> <i>; yields {ty}:result</i>
1870 <p>The '<tt>frem</tt>' instruction returns the remainder from the
1871 division of its two operands.</p>
1873 <p>The two arguments to the '<tt>frem</tt>' instruction must be
1874 <a href="#t_floating">floating point</a> values. Both arguments must have
1875 identical types.</p>
1877 <p>This instruction returns the <i>remainder</i> of a division.</p>
1879 <pre> <result> = frem float 4.0, %var <i>; yields {float}:result = 4.0 % %var</i>
1883 <!-- ======================================================================= -->
1884 <div class="doc_subsection"> <a name="bitwiseops">Bitwise Binary
1885 Operations</a> </div>
1886 <div class="doc_text">
1887 <p>Bitwise binary operators are used to do various forms of
1888 bit-twiddling in a program. They are generally very efficient
1889 instructions and can commonly be strength reduced from other
1890 instructions. They require two operands, execute an operation on them,
1891 and produce a single value. The resulting value of the bitwise binary
1892 operators is always the same type as its first operand.</p>
1894 <!-- _______________________________________________________________________ -->
1895 <div class="doc_subsubsection"> <a name="i_and">'<tt>and</tt>'
1896 Instruction</a> </div>
1897 <div class="doc_text">
1899 <pre> <result> = and <ty> <var1>, <var2> <i>; yields {ty}:result</i>
1902 <p>The '<tt>and</tt>' instruction returns the bitwise logical and of
1903 its two operands.</p>
1905 <p>The two arguments to the '<tt>and</tt>' instruction must be <a
1906 href="#t_integral">integral</a> values. Both arguments must have
1907 identical types.</p>
1909 <p>The truth table used for the '<tt>and</tt>' instruction is:</p>
1911 <div style="align: center">
1912 <table border="1" cellspacing="0" cellpadding="4">
1943 <pre> <result> = and int 4, %var <i>; yields {int}:result = 4 & %var</i>
1944 <result> = and int 15, 40 <i>; yields {int}:result = 8</i>
1945 <result> = and int 4, 8 <i>; yields {int}:result = 0</i>
1948 <!-- _______________________________________________________________________ -->
1949 <div class="doc_subsubsection"> <a name="i_or">'<tt>or</tt>' Instruction</a> </div>
1950 <div class="doc_text">
1952 <pre> <result> = or <ty> <var1>, <var2> <i>; yields {ty}:result</i>
1955 <p>The '<tt>or</tt>' instruction returns the bitwise logical inclusive
1956 or of its two operands.</p>
1958 <p>The two arguments to the '<tt>or</tt>' instruction must be <a
1959 href="#t_integral">integral</a> values. Both arguments must have
1960 identical types.</p>
1962 <p>The truth table used for the '<tt>or</tt>' instruction is:</p>
1964 <div style="align: center">
1965 <table border="1" cellspacing="0" cellpadding="4">
1996 <pre> <result> = or int 4, %var <i>; yields {int}:result = 4 | %var</i>
1997 <result> = or int 15, 40 <i>; yields {int}:result = 47</i>
1998 <result> = or int 4, 8 <i>; yields {int}:result = 12</i>
2001 <!-- _______________________________________________________________________ -->
2002 <div class="doc_subsubsection"> <a name="i_xor">'<tt>xor</tt>'
2003 Instruction</a> </div>
2004 <div class="doc_text">
2006 <pre> <result> = xor <ty> <var1>, <var2> <i>; yields {ty}:result</i>
2009 <p>The '<tt>xor</tt>' instruction returns the bitwise logical exclusive
2010 or of its two operands. The <tt>xor</tt> is used to implement the
2011 "one's complement" operation, which is the "~" operator in C.</p>
2013 <p>The two arguments to the '<tt>xor</tt>' instruction must be <a
2014 href="#t_integral">integral</a> values. Both arguments must have
2015 identical types.</p>
2017 <p>The truth table used for the '<tt>xor</tt>' instruction is:</p>
2019 <div style="align: center">
2020 <table border="1" cellspacing="0" cellpadding="4">
2052 <pre> <result> = xor int 4, %var <i>; yields {int}:result = 4 ^ %var</i>
2053 <result> = xor int 15, 40 <i>; yields {int}:result = 39</i>
2054 <result> = xor int 4, 8 <i>; yields {int}:result = 12</i>
2055 <result> = xor int %V, -1 <i>; yields {int}:result = ~%V</i>
2058 <!-- _______________________________________________________________________ -->
2059 <div class="doc_subsubsection"> <a name="i_shl">'<tt>shl</tt>'
2060 Instruction</a> </div>
2061 <div class="doc_text">
2063 <pre> <result> = shl <ty> <var1>, ubyte <var2> <i>; yields {ty}:result</i>
2066 <p>The '<tt>shl</tt>' instruction returns the first operand shifted to
2067 the left a specified number of bits.</p>
2069 <p>The first argument to the '<tt>shl</tt>' instruction must be an <a
2070 href="#t_integer">integer</a> type. The second argument must be an '<tt>ubyte</tt>'
2073 <p>The value produced is <tt>var1</tt> * 2<sup><tt>var2</tt></sup>.</p>
2075 <pre> <result> = shl int 4, ubyte %var <i>; yields {int}:result = 4 << %var</i>
2076 <result> = shl int 4, ubyte 2 <i>; yields {int}:result = 16</i>
2077 <result> = shl int 1, ubyte 10 <i>; yields {int}:result = 1024</i>
2080 <!-- _______________________________________________________________________ -->
2081 <div class="doc_subsubsection"> <a name="i_lshr">'<tt>lshr</tt>'
2082 Instruction</a> </div>
2083 <div class="doc_text">
2085 <pre> <result> = lshr <ty> <var1>, ubyte <var2> <i>; yields {ty}:result</i>
2089 <p>The '<tt>lshr</tt>' instruction (logical shift right) returns the first
2090 operand shifted to the right a specified number of bits.</p>
2093 <p>The first argument to the '<tt>lshr</tt>' instruction must be an <a
2094 href="#t_integer">integer</a> type. The second argument must be an '<tt>ubyte</tt>' type.</p>
2097 <p>This instruction always performs a logical shift right operation, regardless
2098 of whether the arguments are unsigned or not. The <tt>var2</tt> most significant
2099 bits will be filled with zero bits after the shift.</p>
2103 <result> = lshr uint 4, ubyte 1 <i>; yields {uint}:result = 2</i>
2104 <result> = lshr int 4, ubyte 2 <i>; yields {uint}:result = 1</i>
2105 <result> = lshr sbyte 4, ubyte 3 <i>; yields {sbyte}:result = 0</i>
2106 <result> = lshr sbyte -2, ubyte 1 <i>; yields {sbyte}:result = 0x7FFFFFFF </i>
2110 <!-- ======================================================================= -->
2111 <div class="doc_subsubsection"> <a name="i_ashr">'<tt>ashr</tt>'
2112 Instruction</a> </div>
2113 <div class="doc_text">
2116 <pre> <result> = ashr <ty> <var1>, ubyte <var2> <i>; yields {ty}:result</i>
2120 <p>The '<tt>ashr</tt>' instruction (arithmetic shift right) returns the first
2121 operand shifted to the right a specified number of bits.</p>
2124 <p>The first argument to the '<tt>ashr</tt>' instruction must be an
2125 <a href="#t_integer">integer</a> type. The second argument must be an
2126 '<tt>ubyte</tt>' type.</p>
2129 <p>This instruction always performs an arithmetic shift right operation,
2130 regardless of whether the arguments are signed or not. The <tt>var2</tt> most
2131 significant bits will be filled with the sign bit of <tt>var1</tt>.</p>
2135 <result> = ashr uint 4, ubyte 1 <i>; yields {uint}:result = 2</i>
2136 <result> = ashr int 4, ubyte 2 <i>; yields {int}:result = 1</i>
2137 <result> = ashr ubyte 4, ubyte 3 <i>; yields {ubyte}:result = 0</i>
2138 <result> = ashr sbyte -2, ubyte 1 <i>; yields {sbyte}:result = -1</i>
2142 <!-- ======================================================================= -->
2143 <div class="doc_subsection">
2144 <a name="vectorops">Vector Operations</a>
2147 <div class="doc_text">
2149 <p>LLVM supports several instructions to represent vector operations in a
2150 target-independent manner. This instructions cover the element-access and
2151 vector-specific operations needed to process vectors effectively. While LLVM
2152 does directly support these vector operations, many sophisticated algorithms
2153 will want to use target-specific intrinsics to take full advantage of a specific
2158 <!-- _______________________________________________________________________ -->
2159 <div class="doc_subsubsection">
2160 <a name="i_extractelement">'<tt>extractelement</tt>' Instruction</a>
2163 <div class="doc_text">
2168 <result> = extractelement <n x <ty>> <val>, uint <idx> <i>; yields <ty></i>
2174 The '<tt>extractelement</tt>' instruction extracts a single scalar
2175 element from a packed vector at a specified index.
2182 The first operand of an '<tt>extractelement</tt>' instruction is a
2183 value of <a href="#t_packed">packed</a> type. The second operand is
2184 an index indicating the position from which to extract the element.
2185 The index may be a variable.</p>
2190 The result is a scalar of the same type as the element type of
2191 <tt>val</tt>. Its value is the value at position <tt>idx</tt> of
2192 <tt>val</tt>. If <tt>idx</tt> exceeds the length of <tt>val</tt>, the
2193 results are undefined.
2199 %result = extractelement <4 x int> %vec, uint 0 <i>; yields int</i>
2204 <!-- _______________________________________________________________________ -->
2205 <div class="doc_subsubsection">
2206 <a name="i_insertelement">'<tt>insertelement</tt>' Instruction</a>
2209 <div class="doc_text">
2214 <result> = insertelement <n x <ty>> <val>, <ty> <elt>, uint <idx> <i>; yields <n x <ty>></i>
2220 The '<tt>insertelement</tt>' instruction inserts a scalar
2221 element into a packed vector at a specified index.
2228 The first operand of an '<tt>insertelement</tt>' instruction is a
2229 value of <a href="#t_packed">packed</a> type. The second operand is a
2230 scalar value whose type must equal the element type of the first
2231 operand. The third operand is an index indicating the position at
2232 which to insert the value. The index may be a variable.</p>
2237 The result is a packed vector of the same type as <tt>val</tt>. Its
2238 element values are those of <tt>val</tt> except at position
2239 <tt>idx</tt>, where it gets the value <tt>elt</tt>. If <tt>idx</tt>
2240 exceeds the length of <tt>val</tt>, the results are undefined.
2246 %result = insertelement <4 x int> %vec, int 1, uint 0 <i>; yields <4 x int></i>
2250 <!-- _______________________________________________________________________ -->
2251 <div class="doc_subsubsection">
2252 <a name="i_shufflevector">'<tt>shufflevector</tt>' Instruction</a>
2255 <div class="doc_text">
2260 <result> = shufflevector <n x <ty>> <v1>, <n x <ty>> <v2>, <n x uint> <mask> <i>; yields <n x <ty>></i>
2266 The '<tt>shufflevector</tt>' instruction constructs a permutation of elements
2267 from two input vectors, returning a vector of the same type.
2273 The first two operands of a '<tt>shufflevector</tt>' instruction are vectors
2274 with types that match each other and types that match the result of the
2275 instruction. The third argument is a shuffle mask, which has the same number
2276 of elements as the other vector type, but whose element type is always 'uint'.
2280 The shuffle mask operand is required to be a constant vector with either
2281 constant integer or undef values.
2287 The elements of the two input vectors are numbered from left to right across
2288 both of the vectors. The shuffle mask operand specifies, for each element of
2289 the result vector, which element of the two input registers the result element
2290 gets. The element selector may be undef (meaning "don't care") and the second
2291 operand may be undef if performing a shuffle from only one vector.
2297 %result = shufflevector <4 x int> %v1, <4 x int> %v2,
2298 <4 x uint> <uint 0, uint 4, uint 1, uint 5> <i>; yields <4 x int></i>
2299 %result = shufflevector <4 x int> %v1, <4 x int> undef,
2300 <4 x uint> <uint 0, uint 1, uint 2, uint 3> <i>; yields <4 x int></i> - Identity shuffle.
2305 <!-- ======================================================================= -->
2306 <div class="doc_subsection">
2307 <a name="memoryops">Memory Access and Addressing Operations</a>
2310 <div class="doc_text">
2312 <p>A key design point of an SSA-based representation is how it
2313 represents memory. In LLVM, no memory locations are in SSA form, which
2314 makes things very simple. This section describes how to read, write,
2315 allocate, and free memory in LLVM.</p>
2319 <!-- _______________________________________________________________________ -->
2320 <div class="doc_subsubsection">
2321 <a name="i_malloc">'<tt>malloc</tt>' Instruction</a>
2324 <div class="doc_text">
2329 <result> = malloc <type>[, uint <NumElements>][, align <alignment>] <i>; yields {type*}:result</i>
2334 <p>The '<tt>malloc</tt>' instruction allocates memory from the system
2335 heap and returns a pointer to it.</p>
2339 <p>The '<tt>malloc</tt>' instruction allocates
2340 <tt>sizeof(<type>)*NumElements</tt>
2341 bytes of memory from the operating system and returns a pointer of the
2342 appropriate type to the program. If "NumElements" is specified, it is the
2343 number of elements allocated. If an alignment is specified, the value result
2344 of the allocation is guaranteed to be aligned to at least that boundary. If
2345 not specified, or if zero, the target can choose to align the allocation on any
2346 convenient boundary.</p>
2348 <p>'<tt>type</tt>' must be a sized type.</p>
2352 <p>Memory is allocated using the system "<tt>malloc</tt>" function, and
2353 a pointer is returned.</p>
2358 %array = malloc [4 x ubyte ] <i>; yields {[%4 x ubyte]*}:array</i>
2360 %size = <a href="#i_add">add</a> uint 2, 2 <i>; yields {uint}:size = uint 4</i>
2361 %array1 = malloc ubyte, uint 4 <i>; yields {ubyte*}:array1</i>
2362 %array2 = malloc [12 x ubyte], uint %size <i>; yields {[12 x ubyte]*}:array2</i>
2363 %array3 = malloc int, uint 4, align 1024 <i>; yields {int*}:array3</i>
2364 %array4 = malloc int, align 1024 <i>; yields {int*}:array4</i>
2368 <!-- _______________________________________________________________________ -->
2369 <div class="doc_subsubsection">
2370 <a name="i_free">'<tt>free</tt>' Instruction</a>
2373 <div class="doc_text">
2378 free <type> <value> <i>; yields {void}</i>
2383 <p>The '<tt>free</tt>' instruction returns memory back to the unused
2384 memory heap to be reallocated in the future.</p>
2388 <p>'<tt>value</tt>' shall be a pointer value that points to a value
2389 that was allocated with the '<tt><a href="#i_malloc">malloc</a></tt>'
2394 <p>Access to the memory pointed to by the pointer is no longer defined
2395 after this instruction executes.</p>
2400 %array = <a href="#i_malloc">malloc</a> [4 x ubyte] <i>; yields {[4 x ubyte]*}:array</i>
2401 free [4 x ubyte]* %array
2405 <!-- _______________________________________________________________________ -->
2406 <div class="doc_subsubsection">
2407 <a name="i_alloca">'<tt>alloca</tt>' Instruction</a>
2410 <div class="doc_text">
2415 <result> = alloca <type>[, uint <NumElements>][, align <alignment>] <i>; yields {type*}:result</i>
2420 <p>The '<tt>alloca</tt>' instruction allocates memory on the current
2421 stack frame of the procedure that is live until the current function
2422 returns to its caller.</p>
2426 <p>The '<tt>alloca</tt>' instruction allocates <tt>sizeof(<type>)*NumElements</tt>
2427 bytes of memory on the runtime stack, returning a pointer of the
2428 appropriate type to the program. If "NumElements" is specified, it is the
2429 number of elements allocated. If an alignment is specified, the value result
2430 of the allocation is guaranteed to be aligned to at least that boundary. If
2431 not specified, or if zero, the target can choose to align the allocation on any
2432 convenient boundary.</p>
2434 <p>'<tt>type</tt>' may be any sized type.</p>
2438 <p>Memory is allocated; a pointer is returned. '<tt>alloca</tt>'d
2439 memory is automatically released when the function returns. The '<tt>alloca</tt>'
2440 instruction is commonly used to represent automatic variables that must
2441 have an address available. When the function returns (either with the <tt><a
2442 href="#i_ret">ret</a></tt> or <tt><a href="#i_unwind">unwind</a></tt>
2443 instructions), the memory is reclaimed.</p>
2448 %ptr = alloca int <i>; yields {int*}:ptr</i>
2449 %ptr = alloca int, uint 4 <i>; yields {int*}:ptr</i>
2450 %ptr = alloca int, uint 4, align 1024 <i>; yields {int*}:ptr</i>
2451 %ptr = alloca int, align 1024 <i>; yields {int*}:ptr</i>
2455 <!-- _______________________________________________________________________ -->
2456 <div class="doc_subsubsection"> <a name="i_load">'<tt>load</tt>'
2457 Instruction</a> </div>
2458 <div class="doc_text">
2460 <pre> <result> = load <ty>* <pointer><br> <result> = volatile load <ty>* <pointer><br></pre>
2462 <p>The '<tt>load</tt>' instruction is used to read from memory.</p>
2464 <p>The argument to the '<tt>load</tt>' instruction specifies the memory
2465 address from which to load. The pointer must point to a <a
2466 href="#t_firstclass">first class</a> type. If the <tt>load</tt> is
2467 marked as <tt>volatile</tt>, then the optimizer is not allowed to modify
2468 the number or order of execution of this <tt>load</tt> with other
2469 volatile <tt>load</tt> and <tt><a href="#i_store">store</a></tt>
2472 <p>The location of memory pointed to is loaded.</p>
2474 <pre> %ptr = <a href="#i_alloca">alloca</a> int <i>; yields {int*}:ptr</i>
2476 href="#i_store">store</a> int 3, int* %ptr <i>; yields {void}</i>
2477 %val = load int* %ptr <i>; yields {int}:val = int 3</i>
2480 <!-- _______________________________________________________________________ -->
2481 <div class="doc_subsubsection"> <a name="i_store">'<tt>store</tt>'
2482 Instruction</a> </div>
2483 <div class="doc_text">
2485 <pre> store <ty> <value>, <ty>* <pointer> <i>; yields {void}</i>
2486 volatile store <ty> <value>, <ty>* <pointer> <i>; yields {void}</i>
2489 <p>The '<tt>store</tt>' instruction is used to write to memory.</p>
2491 <p>There are two arguments to the '<tt>store</tt>' instruction: a value
2492 to store and an address in which to store it. The type of the '<tt><pointer></tt>'
2493 operand must be a pointer to the type of the '<tt><value></tt>'
2494 operand. If the <tt>store</tt> is marked as <tt>volatile</tt>, then the
2495 optimizer is not allowed to modify the number or order of execution of
2496 this <tt>store</tt> with other volatile <tt>load</tt> and <tt><a
2497 href="#i_store">store</a></tt> instructions.</p>
2499 <p>The contents of memory are updated to contain '<tt><value></tt>'
2500 at the location specified by the '<tt><pointer></tt>' operand.</p>
2502 <pre> %ptr = <a href="#i_alloca">alloca</a> int <i>; yields {int*}:ptr</i>
2504 href="#i_store">store</a> int 3, int* %ptr <i>; yields {void}</i>
2505 %val = load int* %ptr <i>; yields {int}:val = int 3</i>
2509 <!-- _______________________________________________________________________ -->
2510 <div class="doc_subsubsection">
2511 <a name="i_getelementptr">'<tt>getelementptr</tt>' Instruction</a>
2514 <div class="doc_text">
2517 <result> = getelementptr <ty>* <ptrval>{, <ty> <idx>}*
2523 The '<tt>getelementptr</tt>' instruction is used to get the address of a
2524 subelement of an aggregate data structure.</p>
2528 <p>This instruction takes a list of integer operands that indicate what
2529 elements of the aggregate object to index to. The actual types of the arguments
2530 provided depend on the type of the first pointer argument. The
2531 '<tt>getelementptr</tt>' instruction is used to index down through the type
2532 levels of a structure or to a specific index in an array. When indexing into a
2533 structure, only <tt>uint</tt> integer constants are allowed. When indexing
2534 into an array or pointer, only integers of 32 or 64 bits are allowed, and will
2535 be sign extended to 64-bit values.</p>
2537 <p>For example, let's consider a C code fragment and how it gets
2538 compiled to LLVM:</p>
2552 int *foo(struct ST *s) {
2553 return &s[1].Z.B[5][13];
2557 <p>The LLVM code generated by the GCC frontend is:</p>
2560 %RT = type { sbyte, [10 x [20 x int]], sbyte }
2561 %ST = type { int, double, %RT }
2565 int* %foo(%ST* %s) {
2567 %reg = getelementptr %ST* %s, int 1, uint 2, uint 1, int 5, int 13
2574 <p>The index types specified for the '<tt>getelementptr</tt>' instruction depend
2575 on the pointer type that is being indexed into. <a href="#t_pointer">Pointer</a>
2576 and <a href="#t_array">array</a> types can use a 32-bit or 64-bit
2577 <a href="#t_integer">integer</a> type but the value will always be sign extended
2578 to 64-bits. <a href="#t_struct">Structure</a> types, require <tt>uint</tt>
2579 <b>constants</b>.</p>
2581 <p>In the example above, the first index is indexing into the '<tt>%ST*</tt>'
2582 type, which is a pointer, yielding a '<tt>%ST</tt>' = '<tt>{ int, double, %RT
2583 }</tt>' type, a structure. The second index indexes into the third element of
2584 the structure, yielding a '<tt>%RT</tt>' = '<tt>{ sbyte, [10 x [20 x int]],
2585 sbyte }</tt>' type, another structure. The third index indexes into the second
2586 element of the structure, yielding a '<tt>[10 x [20 x int]]</tt>' type, an
2587 array. The two dimensions of the array are subscripted into, yielding an
2588 '<tt>int</tt>' type. The '<tt>getelementptr</tt>' instruction returns a pointer
2589 to this element, thus computing a value of '<tt>int*</tt>' type.</p>
2591 <p>Note that it is perfectly legal to index partially through a
2592 structure, returning a pointer to an inner element. Because of this,
2593 the LLVM code for the given testcase is equivalent to:</p>
2596 int* %foo(%ST* %s) {
2597 %t1 = getelementptr %ST* %s, int 1 <i>; yields %ST*:%t1</i>
2598 %t2 = getelementptr %ST* %t1, int 0, uint 2 <i>; yields %RT*:%t2</i>
2599 %t3 = getelementptr %RT* %t2, int 0, uint 1 <i>; yields [10 x [20 x int]]*:%t3</i>
2600 %t4 = getelementptr [10 x [20 x int]]* %t3, int 0, int 5 <i>; yields [20 x int]*:%t4</i>
2601 %t5 = getelementptr [20 x int]* %t4, int 0, int 13 <i>; yields int*:%t5</i>
2606 <p>Note that it is undefined to access an array out of bounds: array and
2607 pointer indexes must always be within the defined bounds of the array type.
2608 The one exception for this rules is zero length arrays. These arrays are
2609 defined to be accessible as variable length arrays, which requires access
2610 beyond the zero'th element.</p>
2612 <p>The getelementptr instruction is often confusing. For some more insight
2613 into how it works, see <a href="GetElementPtr.html">the getelementptr
2619 <i>; yields [12 x ubyte]*:aptr</i>
2620 %aptr = getelementptr {int, [12 x ubyte]}* %sptr, long 0, uint 1
2624 <!-- ======================================================================= -->
2625 <div class="doc_subsection"> <a name="convertops">Conversion Operations</a>
2627 <div class="doc_text">
2628 <p>The instructions in this category are the conversion instructions (casting)
2629 which all take a single operand and a type. They perform various bit conversions
2633 <!-- _______________________________________________________________________ -->
2634 <div class="doc_subsubsection">
2635 <a name="i_trunc">'<tt>trunc .. to</tt>' Instruction</a>
2637 <div class="doc_text">
2641 <result> = trunc <ty> <value> to <ty2> <i>; yields ty2</i>
2646 The '<tt>trunc</tt>' instruction truncates its operand to the type <tt>ty2</tt>.
2651 The '<tt>trunc</tt>' instruction takes a <tt>value</tt> to trunc, which must
2652 be an <a href="#t_integer">integer</a> type, and a type that specifies the size
2653 and type of the result, which must be an <a href="#t_integral">integral</a>
2654 type. The bit size of <tt>value</tt> must be larger than the bit size of
2655 <tt>ty2</tt>. Equal sized types are not allowed.</p>
2659 The '<tt>trunc</tt>' instruction truncates the high order bits in <tt>value</tt>
2660 and converts the remaining bits to <tt>ty2</tt>. Since the source size must be
2661 larger than the destination size, <tt>trunc</tt> cannot be a <i>no-op cast</i>.
2662 It will always truncate bits.</p>
2666 %X = trunc int 257 to ubyte <i>; yields ubyte:1</i>
2667 %Y = trunc int 123 to bool <i>; yields bool:true</i>
2671 <!-- _______________________________________________________________________ -->
2672 <div class="doc_subsubsection">
2673 <a name="i_zext">'<tt>zext .. to</tt>' Instruction</a>
2675 <div class="doc_text">
2679 <result> = zext <ty> <value> to <ty2> <i>; yields ty2</i>
2683 <p>The '<tt>zext</tt>' instruction zero extends its operand to type
2688 <p>The '<tt>zext</tt>' instruction takes a value to cast, which must be of
2689 <a href="#t_integral">integral</a> type, and a type to cast it to, which must
2690 also be of <a href="#t_integral">integral</a> type. The bit size of the
2691 <tt>value</tt> must be smaller than the bit size of the destination type,
2695 <p>The <tt>zext</tt> fills the high order bits of the <tt>value</tt> with zero
2696 bits until it reaches the size of the destination type, <tt>ty2</tt>. When the
2697 the operand and the type are the same size, no bit filling is done and the
2698 cast is considered a <i>no-op cast</i> because no bits change (only the type
2701 <p>When zero extending from bool, the result will alwasy be either 0 or 1.</p>
2705 %X = zext int 257 to ulong <i>; yields ulong:257</i>
2706 %Y = zext bool true to int <i>; yields int:1</i>
2710 <!-- _______________________________________________________________________ -->
2711 <div class="doc_subsubsection">
2712 <a name="i_sext">'<tt>sext .. to</tt>' Instruction</a>
2714 <div class="doc_text">
2718 <result> = sext <ty> <value> to <ty2> <i>; yields ty2</i>
2722 <p>The '<tt>sext</tt>' sign extends <tt>value</tt> to the type <tt>ty2</tt>.</p>
2726 The '<tt>sext</tt>' instruction takes a value to cast, which must be of
2727 <a href="#t_integral">integral</a> type, and a type to cast it to, which must
2728 also be of <a href="#t_integral">integral</a> type. The bit size of the
2729 <tt>value</tt> must be smaller than the bit size of the destination type,
2734 The '<tt>sext</tt>' instruction performs a sign extension by copying the sign
2735 bit (highest order bit) of the <tt>value</tt> until it reaches the bit size of
2736 the type <tt>ty2</tt>. When the the operand and the type are the same size,
2737 no bit filling is done and the cast is considered a <i>no-op cast</i> because
2738 no bits change (only the type changes).</p>
2740 <p>When sign extending from bool, the extension always results in -1 or 0.</p>
2744 %X = sext sbyte -1 to ushort <i>; yields ushort:65535</i>
2745 %Y = sext bool true to int <i>; yields int:-1</i>
2749 <!-- _______________________________________________________________________ -->
2750 <div class="doc_subsubsection">
2751 <a name="i_fptrunc">'<tt>fptrunc .. to</tt>' Instruction</a>
2754 <div class="doc_text">
2759 <result> = fptrunc <ty> <value> to <ty2> <i>; yields ty2</i>
2763 <p>The '<tt>fptrunc</tt>' instruction truncates <tt>value</tt> to type
2768 <p>The '<tt>fptrunc</tt>' instruction takes a <a href="#t_floating">floating
2769 point</a> value to cast and a <a href="#t_floating">floating point</a> type to
2770 cast it to. The size of <tt>value</tt> must be larger than the size of
2771 <tt>ty2</tt>. This implies that <tt>fptrunc</tt> cannot be used to make a
2772 <i>no-op cast</i>.</p>
2775 <p> The '<tt>fptrunc</tt>' instruction truncates a <tt>value</tt> from a larger
2776 <a href="#t_floating">floating point</a> type to a smaller
2777 <a href="#t_floating">floating point</a> type. If the value cannot fit within
2778 the destination type, <tt>ty2</tt>, then the results are undefined.</p>
2782 %X = fptrunc double 123.0 to float <i>; yields float:123.0</i>
2783 %Y = fptrunc double 1.0E+300 to float <i>; yields undefined</i>
2787 <!-- _______________________________________________________________________ -->
2788 <div class="doc_subsubsection">
2789 <a name="i_fpext">'<tt>fpext .. to</tt>' Instruction</a>
2791 <div class="doc_text">
2795 <result> = fpext <ty> <value> to <ty2> <i>; yields ty2</i>
2799 <p>The '<tt>fpext</tt>' extends a floating point <tt>value</tt> to a larger
2800 floating point value.</p>
2803 <p>The '<tt>fpext</tt>' instruction takes a
2804 <a href="#t_floating">floating point</a> <tt>value</tt> to cast,
2805 and a <a href="#t_floating">floating point</a> type to cast it to. The source
2806 type must be smaller than the destination type.</p>
2809 <p>The '<tt>fpext</tt>' instruction extends the <tt>value</tt> from a smaller
2810 <a href="t_floating">floating point</a> type to a larger
2811 <a href="t_floating">floating point</a> type. The <tt>fpext</tt> cannot be
2812 used to make a <i>no-op cast</i> because it always changes bits. Use
2813 <tt>bitcast</tt> to make a <i>no-op cast</i> for a floating point cast.</p>
2817 %X = fpext float 3.1415 to double <i>; yields double:3.1415</i>
2818 %Y = fpext float 1.0 to float <i>; yields float:1.0 (no-op)</i>
2822 <!-- _______________________________________________________________________ -->
2823 <div class="doc_subsubsection">
2824 <a name="i_fp2uint">'<tt>fptoui .. to</tt>' Instruction</a>
2826 <div class="doc_text">
2830 <result> = fp2uint <ty> <value> to <ty2> <i>; yields ty2</i>
2834 <p>The '<tt>fp2uint</tt>' converts a floating point <tt>value</tt> to its
2835 unsigned integer equivalent of type <tt>ty2</tt>.
2839 <p>The '<tt>fp2uint</tt>' instruction takes a value to cast, which must be a
2840 <a href="#t_floating">floating point</a> value, and a type to cast it to, which
2841 must be an <a href="#t_integral">integral</a> type.</p>
2844 <p> The '<tt>fp2uint</tt>' instruction converts its
2845 <a href="#t_floating">floating point</a> operand into the nearest (rounding
2846 towards zero) unsigned integer value. If the value cannot fit in <tt>ty2</tt>,
2847 the results are undefined.</p>
2849 <p>When converting to bool, the conversion is done as a comparison against
2850 zero. If the <tt>value</tt> was zero, the bool result will be <tt>false</tt>.
2851 If the <tt>value</tt> was non-zero, the bool result will be <tt>true</tt>.</p>
2855 %X = fp2uint double 123.0 to int <i>; yields int:123</i>
2856 %Y = fp2uint float 1.0E+300 to bool <i>; yields bool:true</i>
2857 %X = fp2uint float 1.04E+17 to ubyte <i>; yields undefined:1</i>
2861 <!-- _______________________________________________________________________ -->
2862 <div class="doc_subsubsection">
2863 <a name="i_fptosi">'<tt>fptosi .. to</tt>' Instruction</a>
2865 <div class="doc_text">
2869 <result> = fptosi <ty> <value> to <ty2> <i>; yields ty2</i>
2873 <p>The '<tt>fptosi</tt>' instruction converts
2874 <a href="#t_floating">floating point</a> <tt>value</tt> to type <tt>ty2</tt>.
2879 <p> The '<tt>fptosi</tt>' instruction takes a value to cast, which must be a
2880 <a href="#t_floating">floating point</a> value, and a type to cast it to, which
2881 must also be an <a href="#t_integral">integral</a> type.</p>
2884 <p>The '<tt>fptosi</tt>' instruction converts its
2885 <a href="#t_floating">floating point</a> operand into the nearest (rounding
2886 towards zero) signed integer value. If the value cannot fit in <tt>ty2</tt>,
2887 the results are undefined.</p>
2889 <p>When converting to bool, the conversion is done as a comparison against
2890 zero. If the <tt>value</tt> was zero, the bool result will be <tt>false</tt>.
2891 If the <tt>value</tt> was non-zero, the bool result will be <tt>true</tt>.</p>
2895 %X = fptosi double -123.0 to int <i>; yields int:-123</i>
2896 %Y = fptosi float 1.0E-247 to bool <i>; yields bool:true</i>
2897 %X = fptosi float 1.04E+17 to sbyte <i>; yields undefined:1</i>
2901 <!-- _______________________________________________________________________ -->
2902 <div class="doc_subsubsection">
2903 <a name="i_uitofp">'<tt>uitofp .. to</tt>' Instruction</a>
2905 <div class="doc_text">
2909 <result> = uitofp <ty> <value> to <ty2> <i>; yields ty2</i>
2913 <p>The '<tt>uitofp</tt>' instruction regards <tt>value</tt> as an unsigned
2914 integer and converts that value to the <tt>ty2</tt> type.</p>
2918 <p>The '<tt>uitofp</tt>' instruction takes a value to cast, which must be an
2919 <a href="#t_integral">integral</a> value, and a type to cast it to, which must
2920 be a <a href="#t_floating">floating point</a> type.</p>
2923 <p>The '<tt>uitofp</tt>' instruction interprets its operand as an unsigned
2924 integer quantity and converts it to the corresponding floating point value. If
2925 the value cannot fit in the floating point value, the results are undefined.</p>
2930 %X = uitofp int 257 to float <i>; yields float:257.0</i>
2931 %Y = uitofp sbyte -1 to double <i>; yields double:255.0</i>
2935 <!-- _______________________________________________________________________ -->
2936 <div class="doc_subsubsection">
2937 <a name="i_sitofp">'<tt>sitofp .. to</tt>' Instruction</a>
2939 <div class="doc_text">
2943 <result> = sitofp <ty> <value> to <ty2> <i>; yields ty2</i>
2947 <p>The '<tt>sitofp</tt>' instruction regards <tt>value</tt> as a signed
2948 integer and converts that value to the <tt>ty2</tt> type.</p>
2951 <p>The '<tt>sitofp</tt>' instruction takes a value to cast, which must be an
2952 <a href="#t_integral">integral</a> value, and a type to cast it to, which must be
2953 a <a href="#t_floating">floating point</a> type.</p>
2956 <p>The '<tt>sitofp</tt>' instruction interprets its operand as a signed
2957 integer quantity and converts it to the corresponding floating point value. If
2958 the value cannot fit in the floating point value, the results are undefined.</p>
2962 %X = sitofp int 257 to float <i>; yields float:257.0</i>
2963 %Y = sitofp sbyte -1 to double <i>; yields double:-1.0</i>
2967 <!-- _______________________________________________________________________ -->
2968 <div class="doc_subsubsection">
2969 <a name="i_ptrtoint">'<tt>ptrtoint .. to</tt>' Instruction</a>
2971 <div class="doc_text">
2975 <result> = ptrtoint <ty> <value> to <ty2> <i>; yields ty2</i>
2979 <p>The '<tt>ptrtoint</tt>' instruction converts the pointer <tt>value</tt> to
2980 the integer type <tt>ty2</tt>.</p>
2983 <p>The '<tt>ptrtoint</tt>' instruction takes a <tt>value</tt> to cast, which
2984 must be a <a href="t_pointer">pointer</a> value, and a type to cast it to
2985 <tt>ty2</tt>, which must be an <a href="#t_integer">integer</a> type.
2988 <p>The '<tt>ptrtoint</tt>' instruction converts <tt>value</tt> to integer type
2989 <tt>ty2</tt> by interpreting the pointer value as an integer and either
2990 truncating or zero extending that value to the size of the integer type. If
2991 <tt>value</tt> is smaller than <tt>ty2</tt> then a zero extension is done. If
2992 <tt>value</tt> is larger than <tt>ty2</tt> then a truncation is done. If they
2993 are the same size, then nothing is done (<i>no-op cast</i>).</p>
2997 %X = ptrtoint int* %X to sbyte <i>; yields truncation on 32-bit</i>
2998 %Y = ptrtoint int* %x to ulong <i>; yields zero extend on 32-bit</i>
3002 <!-- _______________________________________________________________________ -->
3003 <div class="doc_subsubsection">
3004 <a name="i_inttoptr">'<tt>inttoptr .. to</tt>' Instruction</a>
3006 <div class="doc_text">
3010 <result> = inttoptr <ty> <value> to <ty2> <i>; yields ty2</i>
3014 <p>The '<tt>inttoptr</tt>' instruction converts an integer <tt>value</tt> to
3015 a pointer type, <tt>ty2</tt>.</p>
3018 <p>The '<tt>inttoptr</tt>' instruction takes an <a href="i_integer">integer</a>
3019 value to cast, and a type to cast it to, which must be a
3020 <a href="#t_pointer">pointer</a> type. </tt>
3023 <p>The '<tt>inttoptr</tt>' instruction converts <tt>value</tt> to type
3024 <tt>ty2</tt> by applying either a zero extension or a truncation depending on
3025 the size of the integer <tt>value</tt>. If <tt>value</tt> is larger than the
3026 size of a pointer then a truncation is done. If <tt>value</tt> is smaller than
3027 the size of a pointer then a zero extension is done. If they are the same size,
3028 nothing is done (<i>no-op cast</i>).</p>
3032 %X = inttoptr int 255 to int* <i>; yields zero extend on 64-bit</i>
3033 %X = inttoptr int 255 to int* <i>; yields no-op on 32-bit </i>
3034 %Y = inttoptr short 0 to int* <i>; yields zero extend on 32-bit</i>
3038 <!-- _______________________________________________________________________ -->
3039 <div class="doc_subsubsection">
3040 <a name="i_bitcast">'<tt>bitcast .. to</tt>' Instruction</a>
3042 <div class="doc_text">
3046 <result> = bitcast <ty> <value> to <ty2> <i>; yields ty2</i>
3050 <p>The '<tt>bitcast</tt>' instruction converts <tt>value</tt> to type
3051 <tt>ty2</tt> without changing any bits.</p>
3054 <p>The '<tt>bitcast</tt>' instruction takes a value to cast, which must be
3055 a first class value, and a type to cast it to, which must also be a <a
3056 href="#t_firstclass">first class</a> type. The bit sizes of <tt>value</tt>
3057 and the destination type, <tt>ty2</tt>, must be identical.</p>
3060 <p>The '<tt>bitcast</tt>' instruction converts <tt>value</tt> to type
3061 <tt>ty2</tt>. It is always a <i>no-op cast</i> because no bits change with
3062 this conversion. The conversion is done as if the <tt>value</tt> had been
3063 stored to memory and read back as type <tt>ty2</tt>. Pointer types may only be
3064 converted to other pointer types with this instruction. To convert pointers to
3065 other types, use the <a href="#i_inttoptr">inttoptr</a> or
3066 <a href="#i_ptrtoint">ptrtoint</a> instructions first.</p>
3070 %X = bitcast ubyte 255 to sbyte <i>; yields sbyte:-1</i>
3071 %Y = bitcast uint* %x to sint* <i>; yields sint*:%x</i>
3072 %Z = bitcast <2xint> %V to long; <i>; yields long: %V</i>
3076 <!-- ======================================================================= -->
3077 <div class="doc_subsection"> <a name="otherops">Other Operations</a> </div>
3078 <div class="doc_text">
3079 <p>The instructions in this category are the "miscellaneous"
3080 instructions, which defy better classification.</p>
3083 <!-- _______________________________________________________________________ -->
3084 <div class="doc_subsubsection"><a name="i_icmp">'<tt>icmp</tt>' Instruction</a>
3086 <div class="doc_text">
3088 <pre> <result> = icmp <cond> <ty> <var1>, <var2> <i>; yields {bool}:result</i>
3091 <p>The '<tt>icmp</tt>' instruction returns a boolean value based on comparison
3092 of its two integer operands.</p>
3094 <p>The '<tt>icmp</tt>' instruction takes three operands. The first operand is
3095 the condition code which indicates the kind of comparison to perform. It is not
3096 a value, just a keyword. The possibilities for the condition code are:
3098 <li><tt>eq</tt>: equal</li>
3099 <li><tt>ne</tt>: not equal </li>
3100 <li><tt>ugt</tt>: unsigned greater than</li>
3101 <li><tt>uge</tt>: unsigned greater or equal</li>
3102 <li><tt>ult</tt>: unsigned less than</li>
3103 <li><tt>ule</tt>: unsigned less or equal</li>
3104 <li><tt>sgt</tt>: signed greater than</li>
3105 <li><tt>sge</tt>: signed greater or equal</li>
3106 <li><tt>slt</tt>: signed less than</li>
3107 <li><tt>sle</tt>: signed less or equal</li>
3109 <p>The remaining two arguments must be of <a href="#t_integral">integral</a>,
3110 <a href="#t_pointer">pointer</a> or a <a href="#t_packed">packed</a> integral
3111 type. They must have identical types.</p>
3113 <p>The '<tt>icmp</tt>' compares <tt>var1</tt> and <tt>var2</tt> according to
3114 the condition code given as <tt>cond</tt>. The comparison performed always
3115 yields a <a href="#t_bool">bool</a> result, as follows:
3117 <li><tt>eq</tt>: yields <tt>true</tt> if the operands are equal,
3118 <tt>false</tt> otherwise. No sign interpretation is necessary or performed.
3120 <li><tt>ne</tt>: yields <tt>true</tt> if the operands are unequal,
3121 <tt>false</tt> otherwise. No sign interpretation is necessary or performed.
3122 <li><tt>ugt</tt>: interprets the operands as unsigned values and yields
3123 <tt>true</tt> if <tt>var1</tt> is greater than <tt>var2</tt>.</li>
3124 <li><tt>uge</tt>: interprets the operands as unsigned values and yields
3125 <tt>true</tt> if <tt>var1</tt> is greater than or equal to <tt>var2</tt>.</li>
3126 <li><tt>ult</tt>: interprets the operands as unsigned values and yields
3127 <tt>true</tt> if <tt>var1</tt> is less than <tt>var2</tt>.</li>
3128 <li><tt>ule</tt>: interprets the operands as unsigned values and yields
3129 <tt>true</tt> if <tt>var1</tt> is less than or equal to <tt>var2</tt>.</li>
3130 <li><tt>sgt</tt>: interprets the operands as signed values and yields
3131 <tt>true</tt> if <tt>var1</tt> is greater than <tt>var2</tt>.</li>
3132 <li><tt>sge</tt>: interprets the operands as signed values and yields
3133 <tt>true</tt> if <tt>var1</tt> is greater than or equal to <tt>var2</tt>.</li>
3134 <li><tt>slt</tt>: interprets the operands as signed values and yields
3135 <tt>true</tt> if <tt>var1</tt> is less than <tt>var2</tt>.</li>
3136 <li><tt>sle</tt>: interprets the operands as signed values and yields
3137 <tt>true</tt> if <tt>var1</tt> is less than or equal to <tt>var2</tt>.</li>
3140 <p>If the operands are <a href="#t_pointer">pointer</a> typed, the pointer
3141 values are treated as integers and then compared.</p>
3142 <p>If the operands are <a href="#t_packed">packed</a> typed, the elements of
3143 the vector are compared in turn and the predicate must hold for all
3147 <pre> <result> = icmp eq int 4, 5 <i>; yields: result=false</i>
3148 <result> = icmp ne float* %X, %X <i>; yields: result=false</i>
3149 <result> = icmp ult short 4, 5 <i>; yields: result=true</i>
3150 <result> = icmp sgt sbyte 4, 5 <i>; yields: result=false</i>
3151 <result> = icmp ule sbyte -4, 5 <i>; yields: result=false</i>
3152 <result> = icmp sge sbyte 4, 5 <i>; yields: result=false</i>
3156 <!-- _______________________________________________________________________ -->
3157 <div class="doc_subsubsection"><a name="i_fcmp">'<tt>fcmp</tt>' Instruction</a>
3159 <div class="doc_text">
3161 <pre> <result> = fcmp <cond> <ty> <var1>, <var2> <i>; yields {bool}:result</i>
3164 <p>The '<tt>fcmp</tt>' instruction returns a boolean value based on comparison
3165 of its floating point operands.</p>
3167 <p>The '<tt>fcmp</tt>' instruction takes three operands. The first operand is
3168 the condition code which indicates the kind of comparison to perform. It is not
3169 a value, just a keyword. The possibilities for the condition code are:
3171 <li><tt>false</tt>: no comparison, always returns false</li>
3172 <li><tt>oeq</tt>: ordered and equal</li>
3173 <li><tt>ogt</tt>: ordered and greater than </li>
3174 <li><tt>oge</tt>: ordered and greater than or equal</li>
3175 <li><tt>olt</tt>: ordered and less than </li>
3176 <li><tt>ole</tt>: ordered and less than or equal</li>
3177 <li><tt>one</tt>: ordered and not equal</li>
3178 <li><tt>ord</tt>: ordered (no nans)</li>
3179 <li><tt>ueq</tt>: unordered or equal</li>
3180 <li><tt>ugt</tt>: unordered or greater than </li>
3181 <li><tt>uge</tt>: unordered or greater than or equal</li>
3182 <li><tt>ult</tt>: unordered or less than </li>
3183 <li><tt>ule</tt>: unordered or less than or equal</li>
3184 <li><tt>une</tt>: unordered or not equal</li>
3185 <li><tt>uno</tt>: unordered (either nans)</li>
3186 <li><tt>true</tt>: no comparison, always returns true</li>
3188 <p>In the preceding, <i>ordered</i> means that neither operand is a QNAN while
3189 <i>unordered</i> means that either operand may be a QNAN.</p>
3190 <p>The <tt>val1</tt> and <tt>val2</tt> arguments must be of
3191 <a href="#t_floating">floating point</a>, or a <a href="#t_packed">packed</a>
3192 floating point type. They must have identical types.</p>
3193 <p>In the foregoing, <i>ordered</i> means that neither operand is a QNAN and
3194 <i>unordered</i> means that either operand is a QNAN.</p>
3196 <p>The '<tt>fcmp</tt>' compares <tt>var1</tt> and <tt>var2</tt> according to
3197 the condition code given as <tt>cond</tt>. The comparison performed always
3198 yields a <a href="#t_bool">bool</a> result, as follows:
3200 <li><tt>false</tt>: always yields <tt>false</tt>, regardless of operands.</li>
3201 <li><tt>oeq</tt>: yields <tt>true</tt> if both operands are not a QNAN and
3202 <tt>var1</tt> is equal to <tt>var2</tt>.</li>
3203 <li><tt>ogt</tt>: yields <tt>true</tt> if both operands are not a QNAN and
3204 <tt>var1</tt> is greather than <tt>var2</tt>.</li>
3205 <li><tt>oge</tt>: yields <tt>true</tt> if both operands are not a QNAN and
3206 <tt>var1</tt> is greater than or equal to <tt>var2</tt>.</li>
3207 <li><tt>olt</tt>: yields <tt>true</tt> if both operands are not a QNAN and
3208 <tt>var1</tt> is less than <tt>var2</tt>.</li>
3209 <li><tt>ole</tt>: yields <tt>true</tt> if both operands are not a QNAN and
3210 <tt>var1</tt> is less than or equal to <tt>var2</tt>.</li>
3211 <li><tt>one</tt>: yields <tt>true</tt> if both operands are not a QNAN and
3212 <tt>var1</tt> is not equal to <tt>var2</tt>.</li>
3213 <li><tt>ord</tt>: yields <tt>true</tt> if both operands are not a QNAN.</li>
3214 <li><tt>ueq</tt>: yields <tt>true</tt> if either operand is a QNAN or
3215 <tt>var1</tt> is equal to <tt>var2</tt>.</li>
3216 <li><tt>ugt</tt>: yields <tt>true</tt> if either operand is a QNAN or
3217 <tt>var1</tt> is greater than <tt>var2</tt>.</li>
3218 <li><tt>uge</tt>: yields <tt>true</tt> if either operand is a QNAN or
3219 <tt>var1</tt> is greater than or equal to <tt>var2</tt>.</li>
3220 <li><tt>ult</tt>: yields <tt>true</tt> if either operand is a QNAN or
3221 <tt>var1</tt> is less than <tt>var2</tt>.</li>
3222 <li><tt>ule</tt>: yields <tt>true</tt> if either operand is a QNAN or
3223 <tt>var1</tt> is less than or equal to <tt>var2</tt>.</li>
3224 <li><tt>une</tt>: yields <tt>true</tt> if either operand is a QNAN or
3225 <tt>var1</tt> is not equal to <tt>var2</tt>.</li>
3226 <li><tt>uno</tt>: yields <tt>true</tt> if either operand is a QNAN.</li>
3227 <li><tt>true</tt>: always yields <tt>true</tt>, regardless of operands.</li>
3229 <p>If the operands are <a href="#t_packed">packed</a> typed, the elements of
3230 the vector are compared in turn and the predicate must hold for all elements.
3234 <pre> <result> = fcmp oeq float 4.0, 5.0 <i>; yields: result=false</i>
3235 <result> = icmp one float 4.0, 5.0 <i>; yields: result=true</i>
3236 <result> = icmp olt float 4.0, 5.0 <i>; yields: result=true</i>
3237 <result> = icmp ueq double 1.0, 2.0 <i>; yields: result=false</i>
3241 <!-- _______________________________________________________________________ -->
3242 <div class="doc_subsubsection"> <a name="i_phi">'<tt>phi</tt>'
3243 Instruction</a> </div>
3244 <div class="doc_text">
3246 <pre> <result> = phi <ty> [ <val0>, <label0>], ...<br></pre>
3248 <p>The '<tt>phi</tt>' instruction is used to implement the φ node in
3249 the SSA graph representing the function.</p>
3251 <p>The type of the incoming values are specified with the first type
3252 field. After this, the '<tt>phi</tt>' instruction takes a list of pairs
3253 as arguments, with one pair for each predecessor basic block of the
3254 current block. Only values of <a href="#t_firstclass">first class</a>
3255 type may be used as the value arguments to the PHI node. Only labels
3256 may be used as the label arguments.</p>
3257 <p>There must be no non-phi instructions between the start of a basic
3258 block and the PHI instructions: i.e. PHI instructions must be first in
3261 <p>At runtime, the '<tt>phi</tt>' instruction logically takes on the
3262 value specified by the parameter, depending on which basic block we
3263 came from in the last <a href="#terminators">terminator</a> instruction.</p>
3265 <pre>Loop: ; Infinite loop that counts from 0 on up...<br> %indvar = phi uint [ 0, %LoopHeader ], [ %nextindvar, %Loop ]<br> %nextindvar = add uint %indvar, 1<br> br label %Loop<br></pre>
3268 <!-- _______________________________________________________________________ -->
3269 <div class="doc_subsubsection">
3270 <a name="i_select">'<tt>select</tt>' Instruction</a>
3273 <div class="doc_text">
3278 <result> = select bool <cond>, <ty> <val1>, <ty> <val2> <i>; yields ty</i>
3284 The '<tt>select</tt>' instruction is used to choose one value based on a
3285 condition, without branching.
3292 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.
3298 If the boolean condition evaluates to true, the instruction returns the first
3299 value argument; otherwise, it returns the second value argument.
3305 %X = select bool true, ubyte 17, ubyte 42 <i>; yields ubyte:17</i>
3310 <!-- _______________________________________________________________________ -->
3311 <div class="doc_subsubsection">
3312 <a name="i_call">'<tt>call</tt>' Instruction</a>
3315 <div class="doc_text">
3319 <result> = [tail] call [<a href="#callingconv">cconv</a>] <ty>* <fnptrval>(<param list>)
3324 <p>The '<tt>call</tt>' instruction represents a simple function call.</p>
3328 <p>This instruction requires several arguments:</p>
3332 <p>The optional "tail" marker indicates whether the callee function accesses
3333 any allocas or varargs in the caller. If the "tail" marker is present, the
3334 function call is eligible for tail call optimization. Note that calls may
3335 be marked "tail" even if they do not occur before a <a
3336 href="#i_ret"><tt>ret</tt></a> instruction.
3339 <p>The optional "cconv" marker indicates which <a href="callingconv">calling
3340 convention</a> the call should use. If none is specified, the call defaults
3341 to using C calling conventions.
3344 <p>'<tt>ty</tt>': shall be the signature of the pointer to function value
3345 being invoked. The argument types must match the types implied by this
3346 signature. This type can be omitted if the function is not varargs and
3347 if the function type does not return a pointer to a function.</p>
3350 <p>'<tt>fnptrval</tt>': An LLVM value containing a pointer to a function to
3351 be invoked. In most cases, this is a direct function invocation, but
3352 indirect <tt>call</tt>s are just as possible, calling an arbitrary pointer
3353 to function value.</p>
3356 <p>'<tt>function args</tt>': argument list whose types match the
3357 function signature argument types. All arguments must be of
3358 <a href="#t_firstclass">first class</a> type. If the function signature
3359 indicates the function accepts a variable number of arguments, the extra
3360 arguments can be specified.</p>
3366 <p>The '<tt>call</tt>' instruction is used to cause control flow to
3367 transfer to a specified function, with its incoming arguments bound to
3368 the specified values. Upon a '<tt><a href="#i_ret">ret</a></tt>'
3369 instruction in the called function, control flow continues with the
3370 instruction after the function call, and the return value of the
3371 function is bound to the result argument. This is a simpler case of
3372 the <a href="#i_invoke">invoke</a> instruction.</p>
3377 %retval = call int %test(int %argc)
3378 call int(sbyte*, ...) *%printf(sbyte* %msg, int 12, sbyte 42);
3379 %X = tail call int %foo()
3380 %Y = tail call <a href="#callingconv">fastcc</a> int %foo()
3385 <!-- _______________________________________________________________________ -->
3386 <div class="doc_subsubsection">
3387 <a name="i_va_arg">'<tt>va_arg</tt>' Instruction</a>
3390 <div class="doc_text">
3395 <resultval> = va_arg <va_list*> <arglist>, <argty>
3400 <p>The '<tt>va_arg</tt>' instruction is used to access arguments passed through
3401 the "variable argument" area of a function call. It is used to implement the
3402 <tt>va_arg</tt> macro in C.</p>
3406 <p>This instruction takes a <tt>va_list*</tt> value and the type of
3407 the argument. It returns a value of the specified argument type and
3408 increments the <tt>va_list</tt> to point to the next argument. Again, the
3409 actual type of <tt>va_list</tt> is target specific.</p>
3413 <p>The '<tt>va_arg</tt>' instruction loads an argument of the specified
3414 type from the specified <tt>va_list</tt> and causes the
3415 <tt>va_list</tt> to point to the next argument. For more information,
3416 see the variable argument handling <a href="#int_varargs">Intrinsic
3419 <p>It is legal for this instruction to be called in a function which does not
3420 take a variable number of arguments, for example, the <tt>vfprintf</tt>
3423 <p><tt>va_arg</tt> is an LLVM instruction instead of an <a
3424 href="#intrinsics">intrinsic function</a> because it takes a type as an
3429 <p>See the <a href="#int_varargs">variable argument processing</a> section.</p>
3433 <!-- *********************************************************************** -->
3434 <div class="doc_section"> <a name="intrinsics">Intrinsic Functions</a> </div>
3435 <!-- *********************************************************************** -->
3437 <div class="doc_text">
3439 <p>LLVM supports the notion of an "intrinsic function". These functions have
3440 well known names and semantics and are required to follow certain
3441 restrictions. Overall, these instructions represent an extension mechanism for
3442 the LLVM language that does not require changing all of the transformations in
3443 LLVM to add to the language (or the bytecode reader/writer, the parser,
3446 <p>Intrinsic function names must all start with an "<tt>llvm.</tt>" prefix. This
3447 prefix is reserved in LLVM for intrinsic names; thus, functions may not be named
3448 this. Intrinsic functions must always be external functions: you cannot define
3449 the body of intrinsic functions. Intrinsic functions may only be used in call
3450 or invoke instructions: it is illegal to take the address of an intrinsic
3451 function. Additionally, because intrinsic functions are part of the LLVM
3452 language, it is required that they all be documented here if any are added.</p>
3455 <p>To learn how to add an intrinsic function, please see the <a
3456 href="ExtendingLLVM.html">Extending LLVM Guide</a>.
3461 <!-- ======================================================================= -->
3462 <div class="doc_subsection">
3463 <a name="int_varargs">Variable Argument Handling Intrinsics</a>
3466 <div class="doc_text">
3468 <p>Variable argument support is defined in LLVM with the <a
3469 href="#i_va_arg"><tt>va_arg</tt></a> instruction and these three
3470 intrinsic functions. These functions are related to the similarly
3471 named macros defined in the <tt><stdarg.h></tt> header file.</p>
3473 <p>All of these functions operate on arguments that use a
3474 target-specific value type "<tt>va_list</tt>". The LLVM assembly
3475 language reference manual does not define what this type is, so all
3476 transformations should be prepared to handle intrinsics with any type
3479 <p>This example shows how the <a href="#i_va_arg"><tt>va_arg</tt></a>
3480 instruction and the variable argument handling intrinsic functions are
3484 int %test(int %X, ...) {
3485 ; Initialize variable argument processing
3487 call void %<a href="#i_va_start">llvm.va_start</a>(sbyte** %ap)
3489 ; Read a single integer argument
3490 %tmp = va_arg sbyte** %ap, int
3492 ; Demonstrate usage of llvm.va_copy and llvm.va_end
3494 call void %<a href="#i_va_copy">llvm.va_copy</a>(sbyte** %aq, sbyte** %ap)
3495 call void %<a href="#i_va_end">llvm.va_end</a>(sbyte** %aq)
3497 ; Stop processing of arguments.
3498 call void %<a href="#i_va_end">llvm.va_end</a>(sbyte** %ap)
3504 <!-- _______________________________________________________________________ -->
3505 <div class="doc_subsubsection">
3506 <a name="i_va_start">'<tt>llvm.va_start</tt>' Intrinsic</a>
3510 <div class="doc_text">
3512 <pre> declare void %llvm.va_start(<va_list>* <arglist>)<br></pre>
3514 <P>The '<tt>llvm.va_start</tt>' intrinsic initializes
3515 <tt>*<arglist></tt> for subsequent use by <tt><a
3516 href="#i_va_arg">va_arg</a></tt>.</p>
3520 <P>The argument is a pointer to a <tt>va_list</tt> element to initialize.</p>
3524 <P>The '<tt>llvm.va_start</tt>' intrinsic works just like the <tt>va_start</tt>
3525 macro available in C. In a target-dependent way, it initializes the
3526 <tt>va_list</tt> element the argument points to, so that the next call to
3527 <tt>va_arg</tt> will produce the first variable argument passed to the function.
3528 Unlike the C <tt>va_start</tt> macro, this intrinsic does not need to know the
3529 last argument of the function, the compiler can figure that out.</p>
3533 <!-- _______________________________________________________________________ -->
3534 <div class="doc_subsubsection">
3535 <a name="i_va_end">'<tt>llvm.va_end</tt>' Intrinsic</a>
3538 <div class="doc_text">
3540 <pre> declare void %llvm.va_end(<va_list*> <arglist>)<br></pre>
3542 <p>The '<tt>llvm.va_end</tt>' intrinsic destroys <tt><arglist></tt>
3543 which has been initialized previously with <tt><a href="#i_va_start">llvm.va_start</a></tt>
3544 or <tt><a href="#i_va_copy">llvm.va_copy</a></tt>.</p>
3546 <p>The argument is a <tt>va_list</tt> to destroy.</p>
3548 <p>The '<tt>llvm.va_end</tt>' intrinsic works just like the <tt>va_end</tt>
3549 macro available in C. In a target-dependent way, it destroys the <tt>va_list</tt>.
3550 Calls to <a href="#i_va_start"><tt>llvm.va_start</tt></a> and <a
3551 href="#i_va_copy"><tt>llvm.va_copy</tt></a> must be matched exactly
3552 with calls to <tt>llvm.va_end</tt>.</p>
3555 <!-- _______________________________________________________________________ -->
3556 <div class="doc_subsubsection">
3557 <a name="i_va_copy">'<tt>llvm.va_copy</tt>' Intrinsic</a>
3560 <div class="doc_text">
3565 declare void %llvm.va_copy(<va_list>* <destarglist>,
3566 <va_list>* <srcarglist>)
3571 <p>The '<tt>llvm.va_copy</tt>' intrinsic copies the current argument position from
3572 the source argument list to the destination argument list.</p>
3576 <p>The first argument is a pointer to a <tt>va_list</tt> element to initialize.
3577 The second argument is a pointer to a <tt>va_list</tt> element to copy from.</p>
3582 <p>The '<tt>llvm.va_copy</tt>' intrinsic works just like the <tt>va_copy</tt> macro
3583 available in C. In a target-dependent way, it copies the source
3584 <tt>va_list</tt> element into the destination list. This intrinsic is necessary
3585 because the <tt><a href="i_va_begin">llvm.va_begin</a></tt> intrinsic may be
3586 arbitrarily complex and require memory allocation, for example.</p>
3590 <!-- ======================================================================= -->
3591 <div class="doc_subsection">
3592 <a name="int_gc">Accurate Garbage Collection Intrinsics</a>
3595 <div class="doc_text">
3598 LLVM support for <a href="GarbageCollection.html">Accurate Garbage
3599 Collection</a> requires the implementation and generation of these intrinsics.
3600 These intrinsics allow identification of <a href="#i_gcroot">GC roots on the
3601 stack</a>, as well as garbage collector implementations that require <a
3602 href="#i_gcread">read</a> and <a href="#i_gcwrite">write</a> barriers.
3603 Front-ends for type-safe garbage collected languages should generate these
3604 intrinsics to make use of the LLVM garbage collectors. For more details, see <a
3605 href="GarbageCollection.html">Accurate Garbage Collection with LLVM</a>.
3609 <!-- _______________________________________________________________________ -->
3610 <div class="doc_subsubsection">
3611 <a name="i_gcroot">'<tt>llvm.gcroot</tt>' Intrinsic</a>
3614 <div class="doc_text">
3619 declare void %llvm.gcroot(<ty>** %ptrloc, <ty2>* %metadata)
3624 <p>The '<tt>llvm.gcroot</tt>' intrinsic declares the existence of a GC root to
3625 the code generator, and allows some metadata to be associated with it.</p>
3629 <p>The first argument specifies the address of a stack object that contains the
3630 root pointer. The second pointer (which must be either a constant or a global
3631 value address) contains the meta-data to be associated with the root.</p>
3635 <p>At runtime, a call to this intrinsics stores a null pointer into the "ptrloc"
3636 location. At compile-time, the code generator generates information to allow
3637 the runtime to find the pointer at GC safe points.
3643 <!-- _______________________________________________________________________ -->
3644 <div class="doc_subsubsection">
3645 <a name="i_gcread">'<tt>llvm.gcread</tt>' Intrinsic</a>
3648 <div class="doc_text">
3653 declare sbyte* %llvm.gcread(sbyte* %ObjPtr, sbyte** %Ptr)
3658 <p>The '<tt>llvm.gcread</tt>' intrinsic identifies reads of references from heap
3659 locations, allowing garbage collector implementations that require read
3664 <p>The second argument is the address to read from, which should be an address
3665 allocated from the garbage collector. The first object is a pointer to the
3666 start of the referenced object, if needed by the language runtime (otherwise
3671 <p>The '<tt>llvm.gcread</tt>' intrinsic has the same semantics as a load
3672 instruction, but may be replaced with substantially more complex code by the
3673 garbage collector runtime, as needed.</p>
3678 <!-- _______________________________________________________________________ -->
3679 <div class="doc_subsubsection">
3680 <a name="i_gcwrite">'<tt>llvm.gcwrite</tt>' Intrinsic</a>
3683 <div class="doc_text">
3688 declare void %llvm.gcwrite(sbyte* %P1, sbyte* %Obj, sbyte** %P2)
3693 <p>The '<tt>llvm.gcwrite</tt>' intrinsic identifies writes of references to heap
3694 locations, allowing garbage collector implementations that require write
3695 barriers (such as generational or reference counting collectors).</p>
3699 <p>The first argument is the reference to store, the second is the start of the
3700 object to store it to, and the third is the address of the field of Obj to
3701 store to. If the runtime does not require a pointer to the object, Obj may be
3706 <p>The '<tt>llvm.gcwrite</tt>' intrinsic has the same semantics as a store
3707 instruction, but may be replaced with substantially more complex code by the
3708 garbage collector runtime, as needed.</p>
3714 <!-- ======================================================================= -->
3715 <div class="doc_subsection">
3716 <a name="int_codegen">Code Generator Intrinsics</a>
3719 <div class="doc_text">
3721 These intrinsics are provided by LLVM to expose special features that may only
3722 be implemented with code generator support.
3727 <!-- _______________________________________________________________________ -->
3728 <div class="doc_subsubsection">
3729 <a name="i_returnaddress">'<tt>llvm.returnaddress</tt>' Intrinsic</a>
3732 <div class="doc_text">
3736 declare sbyte *%llvm.returnaddress(uint <level>)
3742 The '<tt>llvm.returnaddress</tt>' intrinsic attempts to compute a
3743 target-specific value indicating the return address of the current function
3744 or one of its callers.
3750 The argument to this intrinsic indicates which function to return the address
3751 for. Zero indicates the calling function, one indicates its caller, etc. The
3752 argument is <b>required</b> to be a constant integer value.
3758 The '<tt>llvm.returnaddress</tt>' intrinsic either returns a pointer indicating
3759 the return address of the specified call frame, or zero if it cannot be
3760 identified. The value returned by this intrinsic is likely to be incorrect or 0
3761 for arguments other than zero, so it should only be used for debugging purposes.
3765 Note that calling this intrinsic does not prevent function inlining or other
3766 aggressive transformations, so the value returned may not be that of the obvious
3767 source-language caller.
3772 <!-- _______________________________________________________________________ -->
3773 <div class="doc_subsubsection">
3774 <a name="i_frameaddress">'<tt>llvm.frameaddress</tt>' Intrinsic</a>
3777 <div class="doc_text">
3781 declare sbyte *%llvm.frameaddress(uint <level>)
3787 The '<tt>llvm.frameaddress</tt>' intrinsic attempts to return the
3788 target-specific frame pointer value for the specified stack frame.
3794 The argument to this intrinsic indicates which function to return the frame
3795 pointer for. Zero indicates the calling function, one indicates its caller,
3796 etc. The argument is <b>required</b> to be a constant integer value.
3802 The '<tt>llvm.frameaddress</tt>' intrinsic either returns a pointer indicating
3803 the frame address of the specified call frame, or zero if it cannot be
3804 identified. The value returned by this intrinsic is likely to be incorrect or 0
3805 for arguments other than zero, so it should only be used for debugging purposes.
3809 Note that calling this intrinsic does not prevent function inlining or other
3810 aggressive transformations, so the value returned may not be that of the obvious
3811 source-language caller.
3815 <!-- _______________________________________________________________________ -->
3816 <div class="doc_subsubsection">
3817 <a name="i_stacksave">'<tt>llvm.stacksave</tt>' Intrinsic</a>
3820 <div class="doc_text">
3824 declare sbyte *%llvm.stacksave()
3830 The '<tt>llvm.stacksave</tt>' intrinsic is used to remember the current state of
3831 the function stack, for use with <a href="#i_stackrestore">
3832 <tt>llvm.stackrestore</tt></a>. This is useful for implementing language
3833 features like scoped automatic variable sized arrays in C99.
3839 This intrinsic returns a opaque pointer value that can be passed to <a
3840 href="#i_stackrestore"><tt>llvm.stackrestore</tt></a>. When an
3841 <tt>llvm.stackrestore</tt> intrinsic is executed with a value saved from
3842 <tt>llvm.stacksave</tt>, it effectively restores the state of the stack to the
3843 state it was in when the <tt>llvm.stacksave</tt> intrinsic executed. In
3844 practice, this pops any <a href="#i_alloca">alloca</a> blocks from the stack
3845 that were allocated after the <tt>llvm.stacksave</tt> was executed.
3850 <!-- _______________________________________________________________________ -->
3851 <div class="doc_subsubsection">
3852 <a name="i_stackrestore">'<tt>llvm.stackrestore</tt>' Intrinsic</a>
3855 <div class="doc_text">
3859 declare void %llvm.stackrestore(sbyte* %ptr)
3865 The '<tt>llvm.stackrestore</tt>' intrinsic is used to restore the state of
3866 the function stack to the state it was in when the corresponding <a
3867 href="#llvm.stacksave"><tt>llvm.stacksave</tt></a> intrinsic executed. This is
3868 useful for implementing language features like scoped automatic variable sized
3875 See the description for <a href="#i_stacksave"><tt>llvm.stacksave</tt></a>.
3881 <!-- _______________________________________________________________________ -->
3882 <div class="doc_subsubsection">
3883 <a name="i_prefetch">'<tt>llvm.prefetch</tt>' Intrinsic</a>
3886 <div class="doc_text">
3890 declare void %llvm.prefetch(sbyte * <address>,
3891 uint <rw>, uint <locality>)
3898 The '<tt>llvm.prefetch</tt>' intrinsic is a hint to the code generator to insert
3899 a prefetch instruction if supported; otherwise, it is a noop. Prefetches have
3901 effect on the behavior of the program but can change its performance
3908 <tt>address</tt> is the address to be prefetched, <tt>rw</tt> is the specifier
3909 determining if the fetch should be for a read (0) or write (1), and
3910 <tt>locality</tt> is a temporal locality specifier ranging from (0) - no
3911 locality, to (3) - extremely local keep in cache. The <tt>rw</tt> and
3912 <tt>locality</tt> arguments must be constant integers.
3918 This intrinsic does not modify the behavior of the program. In particular,
3919 prefetches cannot trap and do not produce a value. On targets that support this
3920 intrinsic, the prefetch can provide hints to the processor cache for better
3926 <!-- _______________________________________________________________________ -->
3927 <div class="doc_subsubsection">
3928 <a name="i_pcmarker">'<tt>llvm.pcmarker</tt>' Intrinsic</a>
3931 <div class="doc_text">
3935 declare void %llvm.pcmarker( uint <id> )
3942 The '<tt>llvm.pcmarker</tt>' intrinsic is a method to export a Program Counter
3944 code to simulators and other tools. The method is target specific, but it is
3945 expected that the marker will use exported symbols to transmit the PC of the marker.
3946 The marker makes no guarantees that it will remain with any specific instruction
3947 after optimizations. It is possible that the presence of a marker will inhibit
3948 optimizations. The intended use is to be inserted after optimizations to allow
3949 correlations of simulation runs.
3955 <tt>id</tt> is a numerical id identifying the marker.
3961 This intrinsic does not modify the behavior of the program. Backends that do not
3962 support this intrinisic may ignore it.
3967 <!-- _______________________________________________________________________ -->
3968 <div class="doc_subsubsection">
3969 <a name="i_readcyclecounter">'<tt>llvm.readcyclecounter</tt>' Intrinsic</a>
3972 <div class="doc_text">
3976 declare ulong %llvm.readcyclecounter( )
3983 The '<tt>llvm.readcyclecounter</tt>' intrinsic provides access to the cycle
3984 counter register (or similar low latency, high accuracy clocks) on those targets
3985 that support it. On X86, it should map to RDTSC. On Alpha, it should map to RPCC.
3986 As the backing counters overflow quickly (on the order of 9 seconds on alpha), this
3987 should only be used for small timings.
3993 When directly supported, reading the cycle counter should not modify any memory.
3994 Implementations are allowed to either return a application specific value or a
3995 system wide value. On backends without support, this is lowered to a constant 0.
4000 <!-- ======================================================================= -->
4001 <div class="doc_subsection">
4002 <a name="int_libc">Standard C Library Intrinsics</a>
4005 <div class="doc_text">
4007 LLVM provides intrinsics for a few important standard C library functions.
4008 These intrinsics allow source-language front-ends to pass information about the
4009 alignment of the pointer arguments to the code generator, providing opportunity
4010 for more efficient code generation.
4015 <!-- _______________________________________________________________________ -->
4016 <div class="doc_subsubsection">
4017 <a name="i_memcpy">'<tt>llvm.memcpy</tt>' Intrinsic</a>
4020 <div class="doc_text">
4024 declare void %llvm.memcpy.i32(sbyte* <dest>, sbyte* <src>,
4025 uint <len>, uint <align>)
4026 declare void %llvm.memcpy.i64(sbyte* <dest>, sbyte* <src>,
4027 ulong <len>, uint <align>)
4033 The '<tt>llvm.memcpy.*</tt>' intrinsics copy a block of memory from the source
4034 location to the destination location.
4038 Note that, unlike the standard libc function, the <tt>llvm.memcpy.*</tt>
4039 intrinsics do not return a value, and takes an extra alignment argument.
4045 The first argument is a pointer to the destination, the second is a pointer to
4046 the source. The third argument is an integer argument
4047 specifying the number of bytes to copy, and the fourth argument is the alignment
4048 of the source and destination locations.
4052 If the call to this intrinisic has an alignment value that is not 0 or 1, then
4053 the caller guarantees that both the source and destination pointers are aligned
4060 The '<tt>llvm.memcpy.*</tt>' intrinsics copy a block of memory from the source
4061 location to the destination location, which are not allowed to overlap. It
4062 copies "len" bytes of memory over. If the argument is known to be aligned to
4063 some boundary, this can be specified as the fourth argument, otherwise it should
4069 <!-- _______________________________________________________________________ -->
4070 <div class="doc_subsubsection">
4071 <a name="i_memmove">'<tt>llvm.memmove</tt>' Intrinsic</a>
4074 <div class="doc_text">
4078 declare void %llvm.memmove.i32(sbyte* <dest>, sbyte* <src>,
4079 uint <len>, uint <align>)
4080 declare void %llvm.memmove.i64(sbyte* <dest>, sbyte* <src>,
4081 ulong <len>, uint <align>)
4087 The '<tt>llvm.memmove.*</tt>' intrinsics move a block of memory from the source
4088 location to the destination location. It is similar to the
4089 '<tt>llvm.memcmp</tt>' intrinsic but allows the two memory locations to overlap.
4093 Note that, unlike the standard libc function, the <tt>llvm.memmove.*</tt>
4094 intrinsics do not return a value, and takes an extra alignment argument.
4100 The first argument is a pointer to the destination, the second is a pointer to
4101 the source. The third argument is an integer argument
4102 specifying the number of bytes to copy, and the fourth argument is the alignment
4103 of the source and destination locations.
4107 If the call to this intrinisic has an alignment value that is not 0 or 1, then
4108 the caller guarantees that the source and destination pointers are aligned to
4115 The '<tt>llvm.memmove.*</tt>' intrinsics copy a block of memory from the source
4116 location to the destination location, which may overlap. It
4117 copies "len" bytes of memory over. If the argument is known to be aligned to
4118 some boundary, this can be specified as the fourth argument, otherwise it should
4124 <!-- _______________________________________________________________________ -->
4125 <div class="doc_subsubsection">
4126 <a name="i_memset">'<tt>llvm.memset.*</tt>' Intrinsics</a>
4129 <div class="doc_text">
4133 declare void %llvm.memset.i32(sbyte* <dest>, ubyte <val>,
4134 uint <len>, uint <align>)
4135 declare void %llvm.memset.i64(sbyte* <dest>, ubyte <val>,
4136 ulong <len>, uint <align>)
4142 The '<tt>llvm.memset.*</tt>' intrinsics fill a block of memory with a particular
4147 Note that, unlike the standard libc function, the <tt>llvm.memset</tt> intrinsic
4148 does not return a value, and takes an extra alignment argument.
4154 The first argument is a pointer to the destination to fill, the second is the
4155 byte value to fill it with, the third argument is an integer
4156 argument specifying the number of bytes to fill, and the fourth argument is the
4157 known alignment of destination location.
4161 If the call to this intrinisic has an alignment value that is not 0 or 1, then
4162 the caller guarantees that the destination pointer is aligned to that boundary.
4168 The '<tt>llvm.memset.*</tt>' intrinsics fill "len" bytes of memory starting at
4170 destination location. If the argument is known to be aligned to some boundary,
4171 this can be specified as the fourth argument, otherwise it should be set to 0 or
4177 <!-- _______________________________________________________________________ -->
4178 <div class="doc_subsubsection">
4179 <a name="i_isunordered">'<tt>llvm.isunordered.*</tt>' Intrinsic</a>
4182 <div class="doc_text">
4186 declare bool %llvm.isunordered.f32(float Val1, float Val2)
4187 declare bool %llvm.isunordered.f64(double Val1, double Val2)
4193 The '<tt>llvm.isunordered</tt>' intrinsics return true if either or both of the
4194 specified floating point values is a NAN.
4200 The arguments are floating point numbers of the same type.
4206 If either or both of the arguments is a SNAN or QNAN, it returns true, otherwise
4212 <!-- _______________________________________________________________________ -->
4213 <div class="doc_subsubsection">
4214 <a name="i_sqrt">'<tt>llvm.sqrt.*</tt>' Intrinsic</a>
4217 <div class="doc_text">
4221 declare float %llvm.sqrt.f32(float %Val)
4222 declare double %llvm.sqrt.f64(double %Val)
4228 The '<tt>llvm.sqrt</tt>' intrinsics return the sqrt of the specified operand,
4229 returning the same value as the libm '<tt>sqrt</tt>' function would. Unlike
4230 <tt>sqrt</tt> in libm, however, <tt>llvm.sqrt</tt> has undefined behavior for
4231 negative numbers (which allows for better optimization).
4237 The argument and return value are floating point numbers of the same type.
4243 This function returns the sqrt of the specified operand if it is a positive
4244 floating point number.
4248 <!-- _______________________________________________________________________ -->
4249 <div class="doc_subsubsection">
4250 <a name="i_powi">'<tt>llvm.powi.*</tt>' Intrinsic</a>
4253 <div class="doc_text">
4257 declare float %llvm.powi.f32(float %Val, int %power)
4258 declare double %llvm.powi.f64(double %Val, int %power)
4264 The '<tt>llvm.powi.*</tt>' intrinsics return the first operand raised to the
4265 specified (positive or negative) power. The order of evaluation of
4266 multiplications is not defined.
4272 The second argument is an integer power, and the first is a value to raise to
4279 This function returns the first value raised to the second power with an
4280 unspecified sequence of rounding operations.</p>
4284 <!-- ======================================================================= -->
4285 <div class="doc_subsection">
4286 <a name="int_manip">Bit Manipulation Intrinsics</a>
4289 <div class="doc_text">
4291 LLVM provides intrinsics for a few important bit manipulation operations.
4292 These allow efficient code generation for some algorithms.
4297 <!-- _______________________________________________________________________ -->
4298 <div class="doc_subsubsection">
4299 <a name="i_bswap">'<tt>llvm.bswap.*</tt>' Intrinsics</a>
4302 <div class="doc_text">
4306 declare ushort %llvm.bswap.i16(ushort <id>)
4307 declare uint %llvm.bswap.i32(uint <id>)
4308 declare ulong %llvm.bswap.i64(ulong <id>)
4314 The '<tt>llvm.bwsap</tt>' family of intrinsics is used to byteswap a 16, 32 or
4315 64 bit quantity. These are useful for performing operations on data that is not
4316 in the target's native byte order.
4322 The <tt>llvm.bswap.16</tt> intrinsic returns a ushort value that has the high and low
4323 byte of the input ushort swapped. Similarly, the <tt>llvm.bswap.i32</tt> intrinsic
4324 returns a uint value that has the four bytes of the input uint swapped, so that
4325 if the input bytes are numbered 0, 1, 2, 3 then the returned uint will have its
4326 bytes in 3, 2, 1, 0 order. The <tt>llvm.bswap.i64</tt> intrinsic extends this concept
4332 <!-- _______________________________________________________________________ -->
4333 <div class="doc_subsubsection">
4334 <a name="int_ctpop">'<tt>llvm.ctpop.*</tt>' Intrinsic</a>
4337 <div class="doc_text">
4341 declare ubyte %llvm.ctpop.i8 (ubyte <src>)
4342 declare ushort %llvm.ctpop.i16(ushort <src>)
4343 declare uint %llvm.ctpop.i32(uint <src>)
4344 declare ulong %llvm.ctpop.i64(ulong <src>)
4350 The '<tt>llvm.ctpop</tt>' family of intrinsics counts the number of bits set in a
4357 The only argument is the value to be counted. The argument may be of any
4358 unsigned integer type. The return type must match the argument type.
4364 The '<tt>llvm.ctpop</tt>' intrinsic counts the 1's in a variable.
4368 <!-- _______________________________________________________________________ -->
4369 <div class="doc_subsubsection">
4370 <a name="int_ctlz">'<tt>llvm.ctlz.*</tt>' Intrinsic</a>
4373 <div class="doc_text">
4377 declare ubyte %llvm.ctlz.i8 (ubyte <src>)
4378 declare ushort %llvm.ctlz.i16(ushort <src>)
4379 declare uint %llvm.ctlz.i32(uint <src>)
4380 declare ulong %llvm.ctlz.i64(ulong <src>)
4386 The '<tt>llvm.ctlz</tt>' family of intrinsic functions counts the number of
4387 leading zeros in a variable.
4393 The only argument is the value to be counted. The argument may be of any
4394 unsigned integer type. The return type must match the argument type.
4400 The '<tt>llvm.ctlz</tt>' intrinsic counts the leading (most significant) zeros
4401 in a variable. If the src == 0 then the result is the size in bits of the type
4402 of src. For example, <tt>llvm.ctlz(int 2) = 30</tt>.
4408 <!-- _______________________________________________________________________ -->
4409 <div class="doc_subsubsection">
4410 <a name="int_cttz">'<tt>llvm.cttz.*</tt>' Intrinsic</a>
4413 <div class="doc_text">
4417 declare ubyte %llvm.cttz.i8 (ubyte <src>)
4418 declare ushort %llvm.cttz.i16(ushort <src>)
4419 declare uint %llvm.cttz.i32(uint <src>)
4420 declare ulong %llvm.cttz.i64(ulong <src>)
4426 The '<tt>llvm.cttz</tt>' family of intrinsic functions counts the number of
4433 The only argument is the value to be counted. The argument may be of any
4434 unsigned integer type. The return type must match the argument type.
4440 The '<tt>llvm.cttz</tt>' intrinsic counts the trailing (least significant) zeros
4441 in a variable. If the src == 0 then the result is the size in bits of the type
4442 of src. For example, <tt>llvm.cttz(2) = 1</tt>.
4446 <!-- ======================================================================= -->
4447 <div class="doc_subsection">
4448 <a name="int_debugger">Debugger Intrinsics</a>
4451 <div class="doc_text">
4453 The LLVM debugger intrinsics (which all start with <tt>llvm.dbg.</tt> prefix),
4454 are described in the <a
4455 href="SourceLevelDebugging.html#format_common_intrinsics">LLVM Source Level
4456 Debugging</a> document.
4461 <!-- *********************************************************************** -->
4464 <a href="http://jigsaw.w3.org/css-validator/check/referer"><img
4465 src="http://jigsaw.w3.org/css-validator/images/vcss" alt="Valid CSS!"></a>
4466 <a href="http://validator.w3.org/check/referer"><img
4467 src="http://www.w3.org/Icons/valid-html401" alt="Valid HTML 4.01!" /></a>
4469 <a href="mailto:sabre@nondot.org">Chris Lattner</a><br>
4470 <a href="http://llvm.org">The LLVM Compiler Infrastructure</a><br>
4471 Last modified: $Date$