1 ==============================
2 LLVM Language Reference Manual
3 ==============================
12 This document is a reference manual for the LLVM assembly language. LLVM
13 is a Static Single Assignment (SSA) based representation that provides
14 type safety, low-level operations, flexibility, and the capability of
15 representing 'all' high-level languages cleanly. It is the common code
16 representation used throughout all phases of the LLVM compilation
22 The LLVM code representation is designed to be used in three different
23 forms: as an in-memory compiler IR, as an on-disk bitcode representation
24 (suitable for fast loading by a Just-In-Time compiler), and as a human
25 readable assembly language representation. This allows LLVM to provide a
26 powerful intermediate representation for efficient compiler
27 transformations and analysis, while providing a natural means to debug
28 and visualize the transformations. The three different forms of LLVM are
29 all equivalent. This document describes the human readable
30 representation and notation.
32 The LLVM representation aims to be light-weight and low-level while
33 being expressive, typed, and extensible at the same time. It aims to be
34 a "universal IR" of sorts, by being at a low enough level that
35 high-level ideas may be cleanly mapped to it (similar to how
36 microprocessors are "universal IR's", allowing many source languages to
37 be mapped to them). By providing type information, LLVM can be used as
38 the target of optimizations: for example, through pointer analysis, it
39 can be proven that a C automatic variable is never accessed outside of
40 the current function, allowing it to be promoted to a simple SSA value
41 instead of a memory location.
48 It is important to note that this document describes 'well formed' LLVM
49 assembly language. There is a difference between what the parser accepts
50 and what is considered 'well formed'. For example, the following
51 instruction is syntactically okay, but not well formed:
57 because the definition of ``%x`` does not dominate all of its uses. The
58 LLVM infrastructure provides a verification pass that may be used to
59 verify that an LLVM module is well formed. This pass is automatically
60 run by the parser after parsing input assembly and by the optimizer
61 before it outputs bitcode. The violations pointed out by the verifier
62 pass indicate bugs in transformation passes or input to the parser.
69 LLVM identifiers come in two basic types: global and local. Global
70 identifiers (functions, global variables) begin with the ``'@'``
71 character. Local identifiers (register names, types) begin with the
72 ``'%'`` character. Additionally, there are three different formats for
73 identifiers, for different purposes:
75 #. Named values are represented as a string of characters with their
76 prefix. For example, ``%foo``, ``@DivisionByZero``,
77 ``%a.really.long.identifier``. The actual regular expression used is
78 '``[%@][a-zA-Z$._][a-zA-Z$._0-9]*``'. Identifiers which require other
79 characters in their names can be surrounded with quotes. Special
80 characters may be escaped using ``"\xx"`` where ``xx`` is the ASCII
81 code for the character in hexadecimal. In this way, any character can
82 be used in a name value, even quotes themselves.
83 #. Unnamed values are represented as an unsigned numeric value with
84 their prefix. For example, ``%12``, ``@2``, ``%44``.
85 #. Constants, which are described in the section Constants_ below.
87 LLVM requires that values start with a prefix for two reasons: Compilers
88 don't need to worry about name clashes with reserved words, and the set
89 of reserved words may be expanded in the future without penalty.
90 Additionally, unnamed identifiers allow a compiler to quickly come up
91 with a temporary variable without having to avoid symbol table
94 Reserved words in LLVM are very similar to reserved words in other
95 languages. There are keywords for different opcodes ('``add``',
96 '``bitcast``', '``ret``', etc...), for primitive type names ('``void``',
97 '``i32``', etc...), and others. These reserved words cannot conflict
98 with variable names, because none of them start with a prefix character
101 Here is an example of LLVM code to multiply the integer variable
108 %result = mul i32 %X, 8
110 After strength reduction:
114 %result = shl i32 %X, 3
120 %0 = add i32 %X, %X ; yields {i32}:%0
121 %1 = add i32 %0, %0 ; yields {i32}:%1
122 %result = add i32 %1, %1
124 This last way of multiplying ``%X`` by 8 illustrates several important
125 lexical features of LLVM:
127 #. Comments are delimited with a '``;``' and go until the end of line.
128 #. Unnamed temporaries are created when the result of a computation is
129 not assigned to a named value.
130 #. Unnamed temporaries are numbered sequentially (using a per-function
131 incrementing counter, starting with 0). Note that basic blocks are
132 included in this numbering. For example, if the entry basic block is not
133 given a label name, then it will get number 0.
135 It also shows a convention that we follow in this document. When
136 demonstrating instructions, we will follow an instruction with a comment
137 that defines the type and name of value produced.
145 LLVM programs are composed of ``Module``'s, each of which is a
146 translation unit of the input programs. Each module consists of
147 functions, global variables, and symbol table entries. Modules may be
148 combined together with the LLVM linker, which merges function (and
149 global variable) definitions, resolves forward declarations, and merges
150 symbol table entries. Here is an example of the "hello world" module:
154 ; Declare the string constant as a global constant.
155 @.str = private unnamed_addr constant [13 x i8] c"hello world\0A\00"
157 ; External declaration of the puts function
158 declare i32 @puts(i8* nocapture) nounwind
160 ; Definition of main function
161 define i32 @main() { ; i32()*
162 ; Convert [13 x i8]* to i8 *...
163 %cast210 = getelementptr [13 x i8]* @.str, i64 0, i64 0
165 ; Call puts function to write out the string to stdout.
166 call i32 @puts(i8* %cast210)
171 !1 = metadata !{i32 42}
174 This example is made up of a :ref:`global variable <globalvars>` named
175 "``.str``", an external declaration of the "``puts``" function, a
176 :ref:`function definition <functionstructure>` for "``main``" and
177 :ref:`named metadata <namedmetadatastructure>` "``foo``".
179 In general, a module is made up of a list of global values (where both
180 functions and global variables are global values). Global values are
181 represented by a pointer to a memory location (in this case, a pointer
182 to an array of char, and a pointer to a function), and have one of the
183 following :ref:`linkage types <linkage>`.
190 All Global Variables and Functions have one of the following types of
194 Global values with "``private``" linkage are only directly
195 accessible by objects in the current module. In particular, linking
196 code into a module with an private global value may cause the
197 private to be renamed as necessary to avoid collisions. Because the
198 symbol is private to the module, all references can be updated. This
199 doesn't show up in any symbol table in the object file.
201 Similar to private, but the value shows as a local symbol
202 (``STB_LOCAL`` in the case of ELF) in the object file. This
203 corresponds to the notion of the '``static``' keyword in C.
204 ``available_externally``
205 Globals with "``available_externally``" linkage are never emitted
206 into the object file corresponding to the LLVM module. They exist to
207 allow inlining and other optimizations to take place given knowledge
208 of the definition of the global, which is known to be somewhere
209 outside the module. Globals with ``available_externally`` linkage
210 are allowed to be discarded at will, and are otherwise the same as
211 ``linkonce_odr``. This linkage type is only allowed on definitions,
214 Globals with "``linkonce``" linkage are merged with other globals of
215 the same name when linkage occurs. This can be used to implement
216 some forms of inline functions, templates, or other code which must
217 be generated in each translation unit that uses it, but where the
218 body may be overridden with a more definitive definition later.
219 Unreferenced ``linkonce`` globals are allowed to be discarded. Note
220 that ``linkonce`` linkage does not actually allow the optimizer to
221 inline the body of this function into callers because it doesn't
222 know if this definition of the function is the definitive definition
223 within the program or whether it will be overridden by a stronger
224 definition. To enable inlining and other optimizations, use
225 "``linkonce_odr``" linkage.
227 "``weak``" linkage has the same merging semantics as ``linkonce``
228 linkage, except that unreferenced globals with ``weak`` linkage may
229 not be discarded. This is used for globals that are declared "weak"
232 "``common``" linkage is most similar to "``weak``" linkage, but they
233 are used for tentative definitions in C, such as "``int X;``" at
234 global scope. Symbols with "``common``" linkage are merged in the
235 same way as ``weak symbols``, and they may not be deleted if
236 unreferenced. ``common`` symbols may not have an explicit section,
237 must have a zero initializer, and may not be marked
238 ':ref:`constant <globalvars>`'. Functions and aliases may not have
241 .. _linkage_appending:
244 "``appending``" linkage may only be applied to global variables of
245 pointer to array type. When two global variables with appending
246 linkage are linked together, the two global arrays are appended
247 together. This is the LLVM, typesafe, equivalent of having the
248 system linker append together "sections" with identical names when
251 The semantics of this linkage follow the ELF object file model: the
252 symbol is weak until linked, if not linked, the symbol becomes null
253 instead of being an undefined reference.
254 ``linkonce_odr``, ``weak_odr``
255 Some languages allow differing globals to be merged, such as two
256 functions with different semantics. Other languages, such as
257 ``C++``, ensure that only equivalent globals are ever merged (the
258 "one definition rule" --- "ODR"). Such languages can use the
259 ``linkonce_odr`` and ``weak_odr`` linkage types to indicate that the
260 global will only be merged with equivalent globals. These linkage
261 types are otherwise the same as their non-``odr`` versions.
263 If none of the above identifiers are used, the global is externally
264 visible, meaning that it participates in linkage and can be used to
265 resolve external symbol references.
267 It is illegal for a function *declaration* to have any linkage type
268 other than ``external`` or ``extern_weak``.
275 LLVM :ref:`functions <functionstructure>`, :ref:`calls <i_call>` and
276 :ref:`invokes <i_invoke>` can all have an optional calling convention
277 specified for the call. The calling convention of any pair of dynamic
278 caller/callee must match, or the behavior of the program is undefined.
279 The following calling conventions are supported by LLVM, and more may be
282 "``ccc``" - The C calling convention
283 This calling convention (the default if no other calling convention
284 is specified) matches the target C calling conventions. This calling
285 convention supports varargs function calls and tolerates some
286 mismatch in the declared prototype and implemented declaration of
287 the function (as does normal C).
288 "``fastcc``" - The fast calling convention
289 This calling convention attempts to make calls as fast as possible
290 (e.g. by passing things in registers). This calling convention
291 allows the target to use whatever tricks it wants to produce fast
292 code for the target, without having to conform to an externally
293 specified ABI (Application Binary Interface). `Tail calls can only
294 be optimized when this, the GHC or the HiPE convention is
295 used. <CodeGenerator.html#id80>`_ This calling convention does not
296 support varargs and requires the prototype of all callees to exactly
297 match the prototype of the function definition.
298 "``coldcc``" - The cold calling convention
299 This calling convention attempts to make code in the caller as
300 efficient as possible under the assumption that the call is not
301 commonly executed. As such, these calls often preserve all registers
302 so that the call does not break any live ranges in the caller side.
303 This calling convention does not support varargs and requires the
304 prototype of all callees to exactly match the prototype of the
305 function definition. Furthermore the inliner doesn't consider such function
307 "``cc 10``" - GHC convention
308 This calling convention has been implemented specifically for use by
309 the `Glasgow Haskell Compiler (GHC) <http://www.haskell.org/ghc>`_.
310 It passes everything in registers, going to extremes to achieve this
311 by disabling callee save registers. This calling convention should
312 not be used lightly but only for specific situations such as an
313 alternative to the *register pinning* performance technique often
314 used when implementing functional programming languages. At the
315 moment only X86 supports this convention and it has the following
318 - On *X86-32* only supports up to 4 bit type parameters. No
319 floating point types are supported.
320 - On *X86-64* only supports up to 10 bit type parameters and 6
321 floating point parameters.
323 This calling convention supports `tail call
324 optimization <CodeGenerator.html#id80>`_ but requires both the
325 caller and callee are using it.
326 "``cc 11``" - The HiPE calling convention
327 This calling convention has been implemented specifically for use by
328 the `High-Performance Erlang
329 (HiPE) <http://www.it.uu.se/research/group/hipe/>`_ compiler, *the*
330 native code compiler of the `Ericsson's Open Source Erlang/OTP
331 system <http://www.erlang.org/download.shtml>`_. It uses more
332 registers for argument passing than the ordinary C calling
333 convention and defines no callee-saved registers. The calling
334 convention properly supports `tail call
335 optimization <CodeGenerator.html#id80>`_ but requires that both the
336 caller and the callee use it. It uses a *register pinning*
337 mechanism, similar to GHC's convention, for keeping frequently
338 accessed runtime components pinned to specific hardware registers.
339 At the moment only X86 supports this convention (both 32 and 64
341 "``webkit_jscc``" - WebKit's JavaScript calling convention
342 This calling convention has been implemented for `WebKit FTL JIT
343 <https://trac.webkit.org/wiki/FTLJIT>`_. It passes arguments on the
344 stack right to left (as cdecl does), and returns a value in the
345 platform's customary return register.
346 "``anyregcc``" - Dynamic calling convention for code patching
347 This is a special convention that supports patching an arbitrary code
348 sequence in place of a call site. This convention forces the call
349 arguments into registers but allows them to be dynamcially
350 allocated. This can currently only be used with calls to
351 llvm.experimental.patchpoint because only this intrinsic records
352 the location of its arguments in a side table. See :doc:`StackMaps`.
353 "``preserve_mostcc``" - The `PreserveMost` calling convention
354 This calling convention attempts to make the code in the caller as little
355 intrusive as possible. This calling convention behaves identical to the `C`
356 calling convention on how arguments and return values are passed, but it
357 uses a different set of caller/callee-saved registers. This alleviates the
358 burden of saving and recovering a large register set before and after the
359 call in the caller. If the arguments are passed in callee-saved registers,
360 then they will be preserved by the callee across the call. This doesn't
361 apply for values returned in callee-saved registers.
363 - On X86-64 the callee preserves all general purpose registers, except for
364 R11. R11 can be used as a scratch register. Floating-point registers
365 (XMMs/YMMs) are not preserved and need to be saved by the caller.
367 The idea behind this convention is to support calls to runtime functions
368 that have a hot path and a cold path. The hot path is usually a small piece
369 of code that doesn't many registers. The cold path might need to call out to
370 another function and therefore only needs to preserve the caller-saved
371 registers, which haven't already been saved by the caller. The
372 `PreserveMost` calling convention is very similar to the `cold` calling
373 convention in terms of caller/callee-saved registers, but they are used for
374 different types of function calls. `coldcc` is for function calls that are
375 rarely executed, whereas `preserve_mostcc` function calls are intended to be
376 on the hot path and definitely executed a lot. Furthermore `preserve_mostcc`
377 doesn't prevent the inliner from inlining the function call.
379 This calling convention will be used by a future version of the ObjectiveC
380 runtime and should therefore still be considered experimental at this time.
381 Although this convention was created to optimize certain runtime calls to
382 the ObjectiveC runtime, it is not limited to this runtime and might be used
383 by other runtimes in the future too. The current implementation only
384 supports X86-64, but the intention is to support more architectures in the
386 "``preserve_allcc``" - The `PreserveAll` calling convention
387 This calling convention attempts to make the code in the caller even less
388 intrusive than the `PreserveMost` calling convention. This calling
389 convention also behaves identical to the `C` calling convention on how
390 arguments and return values are passed, but it uses a different set of
391 caller/callee-saved registers. This removes the burden of saving and
392 recovering a large register set before and after the call in the caller. If
393 the arguments are passed in callee-saved registers, then they will be
394 preserved by the callee across the call. This doesn't apply for values
395 returned in callee-saved registers.
397 - On X86-64 the callee preserves all general purpose registers, except for
398 R11. R11 can be used as a scratch register. Furthermore it also preserves
399 all floating-point registers (XMMs/YMMs).
401 The idea behind this convention is to support calls to runtime functions
402 that don't need to call out to any other functions.
404 This calling convention, like the `PreserveMost` calling convention, will be
405 used by a future version of the ObjectiveC runtime and should be considered
406 experimental at this time.
407 "``cc <n>``" - Numbered convention
408 Any calling convention may be specified by number, allowing
409 target-specific calling conventions to be used. Target specific
410 calling conventions start at 64.
412 More calling conventions can be added/defined on an as-needed basis, to
413 support Pascal conventions or any other well-known target-independent
416 .. _visibilitystyles:
421 All Global Variables and Functions have one of the following visibility
424 "``default``" - Default style
425 On targets that use the ELF object file format, default visibility
426 means that the declaration is visible to other modules and, in
427 shared libraries, means that the declared entity may be overridden.
428 On Darwin, default visibility means that the declaration is visible
429 to other modules. Default visibility corresponds to "external
430 linkage" in the language.
431 "``hidden``" - Hidden style
432 Two declarations of an object with hidden visibility refer to the
433 same object if they are in the same shared object. Usually, hidden
434 visibility indicates that the symbol will not be placed into the
435 dynamic symbol table, so no other module (executable or shared
436 library) can reference it directly.
437 "``protected``" - Protected style
438 On ELF, protected visibility indicates that the symbol will be
439 placed in the dynamic symbol table, but that references within the
440 defining module will bind to the local symbol. That is, the symbol
441 cannot be overridden by another module.
443 A symbol with ``internal`` or ``private`` linkage must have ``default``
451 All Global Variables, Functions and Aliases can have one of the following
455 "``dllimport``" causes the compiler to reference a function or variable via
456 a global pointer to a pointer that is set up by the DLL exporting the
457 symbol. On Microsoft Windows targets, the pointer name is formed by
458 combining ``__imp_`` and the function or variable name.
460 "``dllexport``" causes the compiler to provide a global pointer to a pointer
461 in a DLL, so that it can be referenced with the ``dllimport`` attribute. On
462 Microsoft Windows targets, the pointer name is formed by combining
463 ``__imp_`` and the function or variable name. Since this storage class
464 exists for defining a dll interface, the compiler, assembler and linker know
465 it is externally referenced and must refrain from deleting the symbol.
470 LLVM IR allows you to specify both "identified" and "literal" :ref:`structure
471 types <t_struct>`. Literal types are uniqued structurally, but identified types
472 are never uniqued. An :ref:`opaque structural type <t_opaque>` can also be used
473 to forward declare a type which is not yet available.
475 An example of a identified structure specification is:
479 %mytype = type { %mytype*, i32 }
481 Prior to the LLVM 3.0 release, identified types were structurally uniqued. Only
482 literal types are uniqued in recent versions of LLVM.
489 Global variables define regions of memory allocated at compilation time
492 Global variables definitions must be initialized, may have an explicit section
493 to be placed in, and may have an optional explicit alignment specified.
495 Global variables in other translation units can also be declared, in which
496 case they don't have an initializer.
498 A variable may be defined as ``thread_local``, which means that it will
499 not be shared by threads (each thread will have a separated copy of the
500 variable). Not all targets support thread-local variables. Optionally, a
501 TLS model may be specified:
504 For variables that are only used within the current shared library.
506 For variables in modules that will not be loaded dynamically.
508 For variables defined in the executable and only used within it.
510 The models correspond to the ELF TLS models; see `ELF Handling For
511 Thread-Local Storage <http://people.redhat.com/drepper/tls.pdf>`_ for
512 more information on under which circumstances the different models may
513 be used. The target may choose a different TLS model if the specified
514 model is not supported, or if a better choice of model can be made.
516 A variable may be defined as a global ``constant``, which indicates that
517 the contents of the variable will **never** be modified (enabling better
518 optimization, allowing the global data to be placed in the read-only
519 section of an executable, etc). Note that variables that need runtime
520 initialization cannot be marked ``constant`` as there is a store to the
523 LLVM explicitly allows *declarations* of global variables to be marked
524 constant, even if the final definition of the global is not. This
525 capability can be used to enable slightly better optimization of the
526 program, but requires the language definition to guarantee that
527 optimizations based on the 'constantness' are valid for the translation
528 units that do not include the definition.
530 As SSA values, global variables define pointer values that are in scope
531 (i.e. they dominate) all basic blocks in the program. Global variables
532 always define a pointer to their "content" type because they describe a
533 region of memory, and all memory objects in LLVM are accessed through
536 Global variables can be marked with ``unnamed_addr`` which indicates
537 that the address is not significant, only the content. Constants marked
538 like this can be merged with other constants if they have the same
539 initializer. Note that a constant with significant address *can* be
540 merged with a ``unnamed_addr`` constant, the result being a constant
541 whose address is significant.
543 A global variable may be declared to reside in a target-specific
544 numbered address space. For targets that support them, address spaces
545 may affect how optimizations are performed and/or what target
546 instructions are used to access the variable. The default address space
547 is zero. The address space qualifier must precede any other attributes.
549 LLVM allows an explicit section to be specified for globals. If the
550 target supports it, it will emit globals to the section specified.
552 By default, global initializers are optimized by assuming that global
553 variables defined within the module are not modified from their
554 initial values before the start of the global initializer. This is
555 true even for variables potentially accessible from outside the
556 module, including those with external linkage or appearing in
557 ``@llvm.used`` or dllexported variables. This assumption may be suppressed
558 by marking the variable with ``externally_initialized``.
560 An explicit alignment may be specified for a global, which must be a
561 power of 2. If not present, or if the alignment is set to zero, the
562 alignment of the global is set by the target to whatever it feels
563 convenient. If an explicit alignment is specified, the global is forced
564 to have exactly that alignment. Targets and optimizers are not allowed
565 to over-align the global if the global has an assigned section. In this
566 case, the extra alignment could be observable: for example, code could
567 assume that the globals are densely packed in their section and try to
568 iterate over them as an array, alignment padding would break this
571 Globals can also have a :ref:`DLL storage class <dllstorageclass>`.
575 [@<GlobalVarName> =] [Linkage] [Visibility] [DLLStorageClass] [ThreadLocal]
576 [AddrSpace] [unnamed_addr] [ExternallyInitialized]
577 <global | constant> <Type>
578 [, section "name"] [, align <Alignment>]
580 For example, the following defines a global in a numbered address space
581 with an initializer, section, and alignment:
585 @G = addrspace(5) constant float 1.0, section "foo", align 4
587 The following example just declares a global variable
591 @G = external global i32
593 The following example defines a thread-local global with the
594 ``initialexec`` TLS model:
598 @G = thread_local(initialexec) global i32 0, align 4
600 .. _functionstructure:
605 LLVM function definitions consist of the "``define``" keyword, an
606 optional :ref:`linkage type <linkage>`, an optional :ref:`visibility
607 style <visibility>`, an optional :ref:`DLL storage class <dllstorageclass>`,
608 an optional :ref:`calling convention <callingconv>`,
609 an optional ``unnamed_addr`` attribute, a return type, an optional
610 :ref:`parameter attribute <paramattrs>` for the return type, a function
611 name, a (possibly empty) argument list (each with optional :ref:`parameter
612 attributes <paramattrs>`), optional :ref:`function attributes <fnattrs>`,
613 an optional section, an optional alignment, an optional :ref:`garbage
614 collector name <gc>`, an optional :ref:`prefix <prefixdata>`, an opening
615 curly brace, a list of basic blocks, and a closing curly brace.
617 LLVM function declarations consist of the "``declare``" keyword, an
618 optional :ref:`linkage type <linkage>`, an optional :ref:`visibility
619 style <visibility>`, an optional :ref:`DLL storage class <dllstorageclass>`,
620 an optional :ref:`calling convention <callingconv>`,
621 an optional ``unnamed_addr`` attribute, a return type, an optional
622 :ref:`parameter attribute <paramattrs>` for the return type, a function
623 name, a possibly empty list of arguments, an optional alignment, an optional
624 :ref:`garbage collector name <gc>` and an optional :ref:`prefix <prefixdata>`.
626 A function definition contains a list of basic blocks, forming the CFG (Control
627 Flow Graph) for the function. Each basic block may optionally start with a label
628 (giving the basic block a symbol table entry), contains a list of instructions,
629 and ends with a :ref:`terminator <terminators>` instruction (such as a branch or
630 function return). If an explicit label is not provided, a block is assigned an
631 implicit numbered label, using the next value from the same counter as used for
632 unnamed temporaries (:ref:`see above<identifiers>`). For example, if a function
633 entry block does not have an explicit label, it will be assigned label "%0",
634 then the first unnamed temporary in that block will be "%1", etc.
636 The first basic block in a function is special in two ways: it is
637 immediately executed on entrance to the function, and it is not allowed
638 to have predecessor basic blocks (i.e. there can not be any branches to
639 the entry block of a function). Because the block can have no
640 predecessors, it also cannot have any :ref:`PHI nodes <i_phi>`.
642 LLVM allows an explicit section to be specified for functions. If the
643 target supports it, it will emit functions to the section specified.
645 An explicit alignment may be specified for a function. If not present,
646 or if the alignment is set to zero, the alignment of the function is set
647 by the target to whatever it feels convenient. If an explicit alignment
648 is specified, the function is forced to have at least that much
649 alignment. All alignments must be a power of 2.
651 If the ``unnamed_addr`` attribute is given, the address is know to not
652 be significant and two identical functions can be merged.
656 define [linkage] [visibility] [DLLStorageClass]
658 <ResultType> @<FunctionName> ([argument list])
659 [unnamed_addr] [fn Attrs] [section "name"] [align N]
660 [gc] [prefix Constant] { ... }
667 Aliases act as "second name" for the aliasee value (which can be either
668 function, global variable, another alias or bitcast of global value).
669 Aliases may have an optional :ref:`linkage type <linkage>`, an optional
670 :ref:`visibility style <visibility>`, and an optional :ref:`DLL storage class
675 @<Name> = [Visibility] [DLLStorageClass] alias [Linkage] <AliaseeTy> @<Aliasee>
677 The linkage must be one of ``private``, ``internal``, ``linkonce``, ``weak``,
678 ``linkonce_odr``, ``weak_odr``, ``external``. Note that some system linkers
679 might not correctly handle dropping a weak symbol that is aliased by a non-weak
682 Alias that are not ``unnamed_addr`` are guaranteed to have the same address as
685 The aliasee must be a definition.
687 Aliases are not allowed to point to aliases with linkages that can be
688 overridden. Since they are only a second name, the possibility of the
689 intermediate alias being overridden cannot be represented in an object file.
691 .. _namedmetadatastructure:
696 Named metadata is a collection of metadata. :ref:`Metadata
697 nodes <metadata>` (but not metadata strings) are the only valid
698 operands for a named metadata.
702 ; Some unnamed metadata nodes, which are referenced by the named metadata.
703 !0 = metadata !{metadata !"zero"}
704 !1 = metadata !{metadata !"one"}
705 !2 = metadata !{metadata !"two"}
707 !name = !{!0, !1, !2}
714 The return type and each parameter of a function type may have a set of
715 *parameter attributes* associated with them. Parameter attributes are
716 used to communicate additional information about the result or
717 parameters of a function. Parameter attributes are considered to be part
718 of the function, not of the function type, so functions with different
719 parameter attributes can have the same function type.
721 Parameter attributes are simple keywords that follow the type specified.
722 If multiple parameter attributes are needed, they are space separated.
727 declare i32 @printf(i8* noalias nocapture, ...)
728 declare i32 @atoi(i8 zeroext)
729 declare signext i8 @returns_signed_char()
731 Note that any attributes for the function result (``nounwind``,
732 ``readonly``) come immediately after the argument list.
734 Currently, only the following parameter attributes are defined:
737 This indicates to the code generator that the parameter or return
738 value should be zero-extended to the extent required by the target's
739 ABI (which is usually 32-bits, but is 8-bits for a i1 on x86-64) by
740 the caller (for a parameter) or the callee (for a return value).
742 This indicates to the code generator that the parameter or return
743 value should be sign-extended to the extent required by the target's
744 ABI (which is usually 32-bits) by the caller (for a parameter) or
745 the callee (for a return value).
747 This indicates that this parameter or return value should be treated
748 in a special target-dependent fashion during while emitting code for
749 a function call or return (usually, by putting it in a register as
750 opposed to memory, though some targets use it to distinguish between
751 two different kinds of registers). Use of this attribute is
754 This indicates that the pointer parameter should really be passed by
755 value to the function. The attribute implies that a hidden copy of
756 the pointee is made between the caller and the callee, so the callee
757 is unable to modify the value in the caller. This attribute is only
758 valid on LLVM pointer arguments. It is generally used to pass
759 structs and arrays by value, but is also valid on pointers to
760 scalars. The copy is considered to belong to the caller not the
761 callee (for example, ``readonly`` functions should not write to
762 ``byval`` parameters). This is not a valid attribute for return
765 The byval attribute also supports specifying an alignment with the
766 align attribute. It indicates the alignment of the stack slot to
767 form and the known alignment of the pointer specified to the call
768 site. If the alignment is not specified, then the code generator
769 makes a target-specific assumption.
775 The ``inalloca`` argument attribute allows the caller to take the
776 address of outgoing stack arguments. An ``inalloca`` argument must
777 be a pointer to stack memory produced by an ``alloca`` instruction.
778 The alloca, or argument allocation, must also be tagged with the
779 inalloca keyword. Only the past argument may have the ``inalloca``
780 attribute, and that argument is guaranteed to be passed in memory.
782 An argument allocation may be used by a call at most once because
783 the call may deallocate it. The ``inalloca`` attribute cannot be
784 used in conjunction with other attributes that affect argument
785 storage, like ``inreg``, ``nest``, ``sret``, or ``byval``. The
786 ``inalloca`` attribute also disables LLVM's implicit lowering of
787 large aggregate return values, which means that frontend authors
788 must lower them with ``sret`` pointers.
790 When the call site is reached, the argument allocation must have
791 been the most recent stack allocation that is still live, or the
792 results are undefined. It is possible to allocate additional stack
793 space after an argument allocation and before its call site, but it
794 must be cleared off with :ref:`llvm.stackrestore
797 See :doc:`InAlloca` for more information on how to use this
801 This indicates that the pointer parameter specifies the address of a
802 structure that is the return value of the function in the source
803 program. This pointer must be guaranteed by the caller to be valid:
804 loads and stores to the structure may be assumed by the callee
805 not to trap and to be properly aligned. This may only be applied to
806 the first parameter. This is not a valid attribute for return
812 This indicates that pointer values :ref:`based <pointeraliasing>` on
813 the argument or return value do not alias pointer values which are
814 not *based* on it, ignoring certain "irrelevant" dependencies. For a
815 call to the parent function, dependencies between memory references
816 from before or after the call and from those during the call are
817 "irrelevant" to the ``noalias`` keyword for the arguments and return
818 value used in that call. The caller shares the responsibility with
819 the callee for ensuring that these requirements are met. For further
820 details, please see the discussion of the NoAlias response in :ref:`alias
821 analysis <Must, May, or No>`.
823 Note that this definition of ``noalias`` is intentionally similar
824 to the definition of ``restrict`` in C99 for function arguments,
825 though it is slightly weaker.
827 For function return values, C99's ``restrict`` is not meaningful,
828 while LLVM's ``noalias`` is.
830 This indicates that the callee does not make any copies of the
831 pointer that outlive the callee itself. This is not a valid
832 attribute for return values.
837 This indicates that the pointer parameter can be excised using the
838 :ref:`trampoline intrinsics <int_trampoline>`. This is not a valid
839 attribute for return values and can only be applied to one parameter.
842 This indicates that the function always returns the argument as its return
843 value. This is an optimization hint to the code generator when generating
844 the caller, allowing tail call optimization and omission of register saves
845 and restores in some cases; it is not checked or enforced when generating
846 the callee. The parameter and the function return type must be valid
847 operands for the :ref:`bitcast instruction <i_bitcast>`. This is not a
848 valid attribute for return values and can only be applied to one parameter.
851 This indicates that the parameter or return pointer is not null. This
852 attribute may only be applied to pointer typed parameters. This is not
853 checked or enforced by LLVM, the caller must ensure that the pointer
854 passed in is non-null, or the callee must ensure that the returned pointer
859 Garbage Collector Names
860 -----------------------
862 Each function may specify a garbage collector name, which is simply a
867 define void @f() gc "name" { ... }
869 The compiler declares the supported values of *name*. Specifying a
870 collector which will cause the compiler to alter its output in order to
871 support the named garbage collection algorithm.
878 Prefix data is data associated with a function which the code generator
879 will emit immediately before the function body. The purpose of this feature
880 is to allow frontends to associate language-specific runtime metadata with
881 specific functions and make it available through the function pointer while
882 still allowing the function pointer to be called. To access the data for a
883 given function, a program may bitcast the function pointer to a pointer to
884 the constant's type. This implies that the IR symbol points to the start
887 To maintain the semantics of ordinary function calls, the prefix data must
888 have a particular format. Specifically, it must begin with a sequence of
889 bytes which decode to a sequence of machine instructions, valid for the
890 module's target, which transfer control to the point immediately succeeding
891 the prefix data, without performing any other visible action. This allows
892 the inliner and other passes to reason about the semantics of the function
893 definition without needing to reason about the prefix data. Obviously this
894 makes the format of the prefix data highly target dependent.
896 Prefix data is laid out as if it were an initializer for a global variable
897 of the prefix data's type. No padding is automatically placed between the
898 prefix data and the function body. If padding is required, it must be part
901 A trivial example of valid prefix data for the x86 architecture is ``i8 144``,
902 which encodes the ``nop`` instruction:
906 define void @f() prefix i8 144 { ... }
908 Generally prefix data can be formed by encoding a relative branch instruction
909 which skips the metadata, as in this example of valid prefix data for the
910 x86_64 architecture, where the first two bytes encode ``jmp .+10``:
914 %0 = type <{ i8, i8, i8* }>
916 define void @f() prefix %0 <{ i8 235, i8 8, i8* @md}> { ... }
918 A function may have prefix data but no body. This has similar semantics
919 to the ``available_externally`` linkage in that the data may be used by the
920 optimizers but will not be emitted in the object file.
927 Attribute groups are groups of attributes that are referenced by objects within
928 the IR. They are important for keeping ``.ll`` files readable, because a lot of
929 functions will use the same set of attributes. In the degenerative case of a
930 ``.ll`` file that corresponds to a single ``.c`` file, the single attribute
931 group will capture the important command line flags used to build that file.
933 An attribute group is a module-level object. To use an attribute group, an
934 object references the attribute group's ID (e.g. ``#37``). An object may refer
935 to more than one attribute group. In that situation, the attributes from the
936 different groups are merged.
938 Here is an example of attribute groups for a function that should always be
939 inlined, has a stack alignment of 4, and which shouldn't use SSE instructions:
943 ; Target-independent attributes:
944 attributes #0 = { alwaysinline alignstack=4 }
946 ; Target-dependent attributes:
947 attributes #1 = { "no-sse" }
949 ; Function @f has attributes: alwaysinline, alignstack=4, and "no-sse".
950 define void @f() #0 #1 { ... }
957 Function attributes are set to communicate additional information about
958 a function. Function attributes are considered to be part of the
959 function, not of the function type, so functions with different function
960 attributes can have the same function type.
962 Function attributes are simple keywords that follow the type specified.
963 If multiple attributes are needed, they are space separated. For
968 define void @f() noinline { ... }
969 define void @f() alwaysinline { ... }
970 define void @f() alwaysinline optsize { ... }
971 define void @f() optsize { ... }
974 This attribute indicates that, when emitting the prologue and
975 epilogue, the backend should forcibly align the stack pointer.
976 Specify the desired alignment, which must be a power of two, in
979 This attribute indicates that the inliner should attempt to inline
980 this function into callers whenever possible, ignoring any active
981 inlining size threshold for this caller.
983 This indicates that the callee function at a call site should be
984 recognized as a built-in function, even though the function's declaration
985 uses the ``nobuiltin`` attribute. This is only valid at call sites for
986 direct calls to functions which are declared with the ``nobuiltin``
989 This attribute indicates that this function is rarely called. When
990 computing edge weights, basic blocks post-dominated by a cold
991 function call are also considered to be cold; and, thus, given low
994 This attribute indicates that the source code contained a hint that
995 inlining this function is desirable (such as the "inline" keyword in
996 C/C++). It is just a hint; it imposes no requirements on the
999 This attribute suggests that optimization passes and code generator
1000 passes make choices that keep the code size of this function as small
1001 as possible and perform optimizations that may sacrifice runtime
1002 performance in order to minimize the size of the generated code.
1004 This attribute disables prologue / epilogue emission for the
1005 function. This can have very system-specific consequences.
1007 This indicates that the callee function at a call site is not recognized as
1008 a built-in function. LLVM will retain the original call and not replace it
1009 with equivalent code based on the semantics of the built-in function, unless
1010 the call site uses the ``builtin`` attribute. This is valid at call sites
1011 and on function declarations and definitions.
1013 This attribute indicates that calls to the function cannot be
1014 duplicated. A call to a ``noduplicate`` function may be moved
1015 within its parent function, but may not be duplicated within
1016 its parent function.
1018 A function containing a ``noduplicate`` call may still
1019 be an inlining candidate, provided that the call is not
1020 duplicated by inlining. That implies that the function has
1021 internal linkage and only has one call site, so the original
1022 call is dead after inlining.
1024 This attributes disables implicit floating point instructions.
1026 This attribute indicates that the inliner should never inline this
1027 function in any situation. This attribute may not be used together
1028 with the ``alwaysinline`` attribute.
1030 This attribute suppresses lazy symbol binding for the function. This
1031 may make calls to the function faster, at the cost of extra program
1032 startup time if the function is not called during program startup.
1034 This attribute indicates that the code generator should not use a
1035 red zone, even if the target-specific ABI normally permits it.
1037 This function attribute indicates that the function never returns
1038 normally. This produces undefined behavior at runtime if the
1039 function ever does dynamically return.
1041 This function attribute indicates that the function never returns
1042 with an unwind or exceptional control flow. If the function does
1043 unwind, its runtime behavior is undefined.
1045 This function attribute indicates that the function is not optimized
1046 by any optimization or code generator passes with the
1047 exception of interprocedural optimization passes.
1048 This attribute cannot be used together with the ``alwaysinline``
1049 attribute; this attribute is also incompatible
1050 with the ``minsize`` attribute and the ``optsize`` attribute.
1052 This attribute requires the ``noinline`` attribute to be specified on
1053 the function as well, so the function is never inlined into any caller.
1054 Only functions with the ``alwaysinline`` attribute are valid
1055 candidates for inlining into the body of this function.
1057 This attribute suggests that optimization passes and code generator
1058 passes make choices that keep the code size of this function low,
1059 and otherwise do optimizations specifically to reduce code size as
1060 long as they do not significantly impact runtime performance.
1062 On a function, this attribute indicates that the function computes its
1063 result (or decides to unwind an exception) based strictly on its arguments,
1064 without dereferencing any pointer arguments or otherwise accessing
1065 any mutable state (e.g. memory, control registers, etc) visible to
1066 caller functions. It does not write through any pointer arguments
1067 (including ``byval`` arguments) and never changes any state visible
1068 to callers. This means that it cannot unwind exceptions by calling
1069 the ``C++`` exception throwing methods.
1071 On an argument, this attribute indicates that the function does not
1072 dereference that pointer argument, even though it may read or write the
1073 memory that the pointer points to if accessed through other pointers.
1075 On a function, this attribute indicates that the function does not write
1076 through any pointer arguments (including ``byval`` arguments) or otherwise
1077 modify any state (e.g. memory, control registers, etc) visible to
1078 caller functions. It may dereference pointer arguments and read
1079 state that may be set in the caller. A readonly function always
1080 returns the same value (or unwinds an exception identically) when
1081 called with the same set of arguments and global state. It cannot
1082 unwind an exception by calling the ``C++`` exception throwing
1085 On an argument, this attribute indicates that the function does not write
1086 through this pointer argument, even though it may write to the memory that
1087 the pointer points to.
1089 This attribute indicates that this function can return twice. The C
1090 ``setjmp`` is an example of such a function. The compiler disables
1091 some optimizations (like tail calls) in the caller of these
1093 ``sanitize_address``
1094 This attribute indicates that AddressSanitizer checks
1095 (dynamic address safety analysis) are enabled for this function.
1097 This attribute indicates that MemorySanitizer checks (dynamic detection
1098 of accesses to uninitialized memory) are enabled for this function.
1100 This attribute indicates that ThreadSanitizer checks
1101 (dynamic thread safety analysis) are enabled for this function.
1103 This attribute indicates that the function should emit a stack
1104 smashing protector. It is in the form of a "canary" --- a random value
1105 placed on the stack before the local variables that's checked upon
1106 return from the function to see if it has been overwritten. A
1107 heuristic is used to determine if a function needs stack protectors
1108 or not. The heuristic used will enable protectors for functions with:
1110 - Character arrays larger than ``ssp-buffer-size`` (default 8).
1111 - Aggregates containing character arrays larger than ``ssp-buffer-size``.
1112 - Calls to alloca() with variable sizes or constant sizes greater than
1113 ``ssp-buffer-size``.
1115 Variables that are identified as requiring a protector will be arranged
1116 on the stack such that they are adjacent to the stack protector guard.
1118 If a function that has an ``ssp`` attribute is inlined into a
1119 function that doesn't have an ``ssp`` attribute, then the resulting
1120 function will have an ``ssp`` attribute.
1122 This attribute indicates that the function should *always* emit a
1123 stack smashing protector. This overrides the ``ssp`` function
1126 Variables that are identified as requiring a protector will be arranged
1127 on the stack such that they are adjacent to the stack protector guard.
1128 The specific layout rules are:
1130 #. Large arrays and structures containing large arrays
1131 (``>= ssp-buffer-size``) are closest to the stack protector.
1132 #. Small arrays and structures containing small arrays
1133 (``< ssp-buffer-size``) are 2nd closest to the protector.
1134 #. Variables that have had their address taken are 3rd closest to the
1137 If a function that has an ``sspreq`` attribute is inlined into a
1138 function that doesn't have an ``sspreq`` attribute or which has an
1139 ``ssp`` or ``sspstrong`` attribute, then the resulting function will have
1140 an ``sspreq`` attribute.
1142 This attribute indicates that the function should emit a stack smashing
1143 protector. This attribute causes a strong heuristic to be used when
1144 determining if a function needs stack protectors. The strong heuristic
1145 will enable protectors for functions with:
1147 - Arrays of any size and type
1148 - Aggregates containing an array of any size and type.
1149 - Calls to alloca().
1150 - Local variables that have had their address taken.
1152 Variables that are identified as requiring a protector will be arranged
1153 on the stack such that they are adjacent to the stack protector guard.
1154 The specific layout rules are:
1156 #. Large arrays and structures containing large arrays
1157 (``>= ssp-buffer-size``) are closest to the stack protector.
1158 #. Small arrays and structures containing small arrays
1159 (``< ssp-buffer-size``) are 2nd closest to the protector.
1160 #. Variables that have had their address taken are 3rd closest to the
1163 This overrides the ``ssp`` function attribute.
1165 If a function that has an ``sspstrong`` attribute is inlined into a
1166 function that doesn't have an ``sspstrong`` attribute, then the
1167 resulting function will have an ``sspstrong`` attribute.
1169 This attribute indicates that the ABI being targeted requires that
1170 an unwind table entry be produce for this function even if we can
1171 show that no exceptions passes by it. This is normally the case for
1172 the ELF x86-64 abi, but it can be disabled for some compilation
1177 Module-Level Inline Assembly
1178 ----------------------------
1180 Modules may contain "module-level inline asm" blocks, which corresponds
1181 to the GCC "file scope inline asm" blocks. These blocks are internally
1182 concatenated by LLVM and treated as a single unit, but may be separated
1183 in the ``.ll`` file if desired. The syntax is very simple:
1185 .. code-block:: llvm
1187 module asm "inline asm code goes here"
1188 module asm "more can go here"
1190 The strings can contain any character by escaping non-printable
1191 characters. The escape sequence used is simply "\\xx" where "xx" is the
1192 two digit hex code for the number.
1194 The inline asm code is simply printed to the machine code .s file when
1195 assembly code is generated.
1197 .. _langref_datalayout:
1202 A module may specify a target specific data layout string that specifies
1203 how data is to be laid out in memory. The syntax for the data layout is
1206 .. code-block:: llvm
1208 target datalayout = "layout specification"
1210 The *layout specification* consists of a list of specifications
1211 separated by the minus sign character ('-'). Each specification starts
1212 with a letter and may include other information after the letter to
1213 define some aspect of the data layout. The specifications accepted are
1217 Specifies that the target lays out data in big-endian form. That is,
1218 the bits with the most significance have the lowest address
1221 Specifies that the target lays out data in little-endian form. That
1222 is, the bits with the least significance have the lowest address
1225 Specifies the natural alignment of the stack in bits. Alignment
1226 promotion of stack variables is limited to the natural stack
1227 alignment to avoid dynamic stack realignment. The stack alignment
1228 must be a multiple of 8-bits. If omitted, the natural stack
1229 alignment defaults to "unspecified", which does not prevent any
1230 alignment promotions.
1231 ``p[n]:<size>:<abi>:<pref>``
1232 This specifies the *size* of a pointer and its ``<abi>`` and
1233 ``<pref>``\erred alignments for address space ``n``. All sizes are in
1234 bits. The address space, ``n`` is optional, and if not specified,
1235 denotes the default address space 0. The value of ``n`` must be
1236 in the range [1,2^23).
1237 ``i<size>:<abi>:<pref>``
1238 This specifies the alignment for an integer type of a given bit
1239 ``<size>``. The value of ``<size>`` must be in the range [1,2^23).
1240 ``v<size>:<abi>:<pref>``
1241 This specifies the alignment for a vector type of a given bit
1243 ``f<size>:<abi>:<pref>``
1244 This specifies the alignment for a floating point type of a given bit
1245 ``<size>``. Only values of ``<size>`` that are supported by the target
1246 will work. 32 (float) and 64 (double) are supported on all targets; 80
1247 or 128 (different flavors of long double) are also supported on some
1250 This specifies the alignment for an object of aggregate type.
1252 If present, specifies that llvm names are mangled in the output. The
1255 * ``e``: ELF mangling: Private symbols get a ``.L`` prefix.
1256 * ``m``: Mips mangling: Private symbols get a ``$`` prefix.
1257 * ``o``: Mach-O mangling: Private symbols get ``L`` prefix. Other
1258 symbols get a ``_`` prefix.
1259 * ``w``: Windows COFF prefix: Similar to Mach-O, but stdcall and fastcall
1260 functions also get a suffix based on the frame size.
1261 ``n<size1>:<size2>:<size3>...``
1262 This specifies a set of native integer widths for the target CPU in
1263 bits. For example, it might contain ``n32`` for 32-bit PowerPC,
1264 ``n32:64`` for PowerPC 64, or ``n8:16:32:64`` for X86-64. Elements of
1265 this set are considered to support most general arithmetic operations
1268 On every specification that takes a ``<abi>:<pref>``, specifying the
1269 ``<pref>`` alignment is optional. If omitted, the preceding ``:``
1270 should be omitted too and ``<pref>`` will be equal to ``<abi>``.
1272 When constructing the data layout for a given target, LLVM starts with a
1273 default set of specifications which are then (possibly) overridden by
1274 the specifications in the ``datalayout`` keyword. The default
1275 specifications are given in this list:
1277 - ``E`` - big endian
1278 - ``p:64:64:64`` - 64-bit pointers with 64-bit alignment.
1279 - ``p[n]:64:64:64`` - Other address spaces are assumed to be the
1280 same as the default address space.
1281 - ``S0`` - natural stack alignment is unspecified
1282 - ``i1:8:8`` - i1 is 8-bit (byte) aligned
1283 - ``i8:8:8`` - i8 is 8-bit (byte) aligned
1284 - ``i16:16:16`` - i16 is 16-bit aligned
1285 - ``i32:32:32`` - i32 is 32-bit aligned
1286 - ``i64:32:64`` - i64 has ABI alignment of 32-bits but preferred
1287 alignment of 64-bits
1288 - ``f16:16:16`` - half is 16-bit aligned
1289 - ``f32:32:32`` - float is 32-bit aligned
1290 - ``f64:64:64`` - double is 64-bit aligned
1291 - ``f128:128:128`` - quad is 128-bit aligned
1292 - ``v64:64:64`` - 64-bit vector is 64-bit aligned
1293 - ``v128:128:128`` - 128-bit vector is 128-bit aligned
1294 - ``a:0:64`` - aggregates are 64-bit aligned
1296 When LLVM is determining the alignment for a given type, it uses the
1299 #. If the type sought is an exact match for one of the specifications,
1300 that specification is used.
1301 #. If no match is found, and the type sought is an integer type, then
1302 the smallest integer type that is larger than the bitwidth of the
1303 sought type is used. If none of the specifications are larger than
1304 the bitwidth then the largest integer type is used. For example,
1305 given the default specifications above, the i7 type will use the
1306 alignment of i8 (next largest) while both i65 and i256 will use the
1307 alignment of i64 (largest specified).
1308 #. If no match is found, and the type sought is a vector type, then the
1309 largest vector type that is smaller than the sought vector type will
1310 be used as a fall back. This happens because <128 x double> can be
1311 implemented in terms of 64 <2 x double>, for example.
1313 The function of the data layout string may not be what you expect.
1314 Notably, this is not a specification from the frontend of what alignment
1315 the code generator should use.
1317 Instead, if specified, the target data layout is required to match what
1318 the ultimate *code generator* expects. This string is used by the
1319 mid-level optimizers to improve code, and this only works if it matches
1320 what the ultimate code generator uses. If you would like to generate IR
1321 that does not embed this target-specific detail into the IR, then you
1322 don't have to specify the string. This will disable some optimizations
1323 that require precise layout information, but this also prevents those
1324 optimizations from introducing target specificity into the IR.
1331 A module may specify a target triple string that describes the target
1332 host. The syntax for the target triple is simply:
1334 .. code-block:: llvm
1336 target triple = "x86_64-apple-macosx10.7.0"
1338 The *target triple* string consists of a series of identifiers delimited
1339 by the minus sign character ('-'). The canonical forms are:
1343 ARCHITECTURE-VENDOR-OPERATING_SYSTEM
1344 ARCHITECTURE-VENDOR-OPERATING_SYSTEM-ENVIRONMENT
1346 This information is passed along to the backend so that it generates
1347 code for the proper architecture. It's possible to override this on the
1348 command line with the ``-mtriple`` command line option.
1350 .. _pointeraliasing:
1352 Pointer Aliasing Rules
1353 ----------------------
1355 Any memory access must be done through a pointer value associated with
1356 an address range of the memory access, otherwise the behavior is
1357 undefined. Pointer values are associated with address ranges according
1358 to the following rules:
1360 - A pointer value is associated with the addresses associated with any
1361 value it is *based* on.
1362 - An address of a global variable is associated with the address range
1363 of the variable's storage.
1364 - The result value of an allocation instruction is associated with the
1365 address range of the allocated storage.
1366 - A null pointer in the default address-space is associated with no
1368 - An integer constant other than zero or a pointer value returned from
1369 a function not defined within LLVM may be associated with address
1370 ranges allocated through mechanisms other than those provided by
1371 LLVM. Such ranges shall not overlap with any ranges of addresses
1372 allocated by mechanisms provided by LLVM.
1374 A pointer value is *based* on another pointer value according to the
1377 - A pointer value formed from a ``getelementptr`` operation is *based*
1378 on the first operand of the ``getelementptr``.
1379 - The result value of a ``bitcast`` is *based* on the operand of the
1381 - A pointer value formed by an ``inttoptr`` is *based* on all pointer
1382 values that contribute (directly or indirectly) to the computation of
1383 the pointer's value.
1384 - The "*based* on" relationship is transitive.
1386 Note that this definition of *"based"* is intentionally similar to the
1387 definition of *"based"* in C99, though it is slightly weaker.
1389 LLVM IR does not associate types with memory. The result type of a
1390 ``load`` merely indicates the size and alignment of the memory from
1391 which to load, as well as the interpretation of the value. The first
1392 operand type of a ``store`` similarly only indicates the size and
1393 alignment of the store.
1395 Consequently, type-based alias analysis, aka TBAA, aka
1396 ``-fstrict-aliasing``, is not applicable to general unadorned LLVM IR.
1397 :ref:`Metadata <metadata>` may be used to encode additional information
1398 which specialized optimization passes may use to implement type-based
1403 Volatile Memory Accesses
1404 ------------------------
1406 Certain memory accesses, such as :ref:`load <i_load>`'s,
1407 :ref:`store <i_store>`'s, and :ref:`llvm.memcpy <int_memcpy>`'s may be
1408 marked ``volatile``. The optimizers must not change the number of
1409 volatile operations or change their order of execution relative to other
1410 volatile operations. The optimizers *may* change the order of volatile
1411 operations relative to non-volatile operations. This is not Java's
1412 "volatile" and has no cross-thread synchronization behavior.
1414 IR-level volatile loads and stores cannot safely be optimized into
1415 llvm.memcpy or llvm.memmove intrinsics even when those intrinsics are
1416 flagged volatile. Likewise, the backend should never split or merge
1417 target-legal volatile load/store instructions.
1419 .. admonition:: Rationale
1421 Platforms may rely on volatile loads and stores of natively supported
1422 data width to be executed as single instruction. For example, in C
1423 this holds for an l-value of volatile primitive type with native
1424 hardware support, but not necessarily for aggregate types. The
1425 frontend upholds these expectations, which are intentionally
1426 unspecified in the IR. The rules above ensure that IR transformation
1427 do not violate the frontend's contract with the language.
1431 Memory Model for Concurrent Operations
1432 --------------------------------------
1434 The LLVM IR does not define any way to start parallel threads of
1435 execution or to register signal handlers. Nonetheless, there are
1436 platform-specific ways to create them, and we define LLVM IR's behavior
1437 in their presence. This model is inspired by the C++0x memory model.
1439 For a more informal introduction to this model, see the :doc:`Atomics`.
1441 We define a *happens-before* partial order as the least partial order
1444 - Is a superset of single-thread program order, and
1445 - When a *synchronizes-with* ``b``, includes an edge from ``a`` to
1446 ``b``. *Synchronizes-with* pairs are introduced by platform-specific
1447 techniques, like pthread locks, thread creation, thread joining,
1448 etc., and by atomic instructions. (See also :ref:`Atomic Memory Ordering
1449 Constraints <ordering>`).
1451 Note that program order does not introduce *happens-before* edges
1452 between a thread and signals executing inside that thread.
1454 Every (defined) read operation (load instructions, memcpy, atomic
1455 loads/read-modify-writes, etc.) R reads a series of bytes written by
1456 (defined) write operations (store instructions, atomic
1457 stores/read-modify-writes, memcpy, etc.). For the purposes of this
1458 section, initialized globals are considered to have a write of the
1459 initializer which is atomic and happens before any other read or write
1460 of the memory in question. For each byte of a read R, R\ :sub:`byte`
1461 may see any write to the same byte, except:
1463 - If write\ :sub:`1` happens before write\ :sub:`2`, and
1464 write\ :sub:`2` happens before R\ :sub:`byte`, then
1465 R\ :sub:`byte` does not see write\ :sub:`1`.
1466 - If R\ :sub:`byte` happens before write\ :sub:`3`, then
1467 R\ :sub:`byte` does not see write\ :sub:`3`.
1469 Given that definition, R\ :sub:`byte` is defined as follows:
1471 - If R is volatile, the result is target-dependent. (Volatile is
1472 supposed to give guarantees which can support ``sig_atomic_t`` in
1473 C/C++, and may be used for accesses to addresses which do not behave
1474 like normal memory. It does not generally provide cross-thread
1476 - Otherwise, if there is no write to the same byte that happens before
1477 R\ :sub:`byte`, R\ :sub:`byte` returns ``undef`` for that byte.
1478 - Otherwise, if R\ :sub:`byte` may see exactly one write,
1479 R\ :sub:`byte` returns the value written by that write.
1480 - Otherwise, if R is atomic, and all the writes R\ :sub:`byte` may
1481 see are atomic, it chooses one of the values written. See the :ref:`Atomic
1482 Memory Ordering Constraints <ordering>` section for additional
1483 constraints on how the choice is made.
1484 - Otherwise R\ :sub:`byte` returns ``undef``.
1486 R returns the value composed of the series of bytes it read. This
1487 implies that some bytes within the value may be ``undef`` **without**
1488 the entire value being ``undef``. Note that this only defines the
1489 semantics of the operation; it doesn't mean that targets will emit more
1490 than one instruction to read the series of bytes.
1492 Note that in cases where none of the atomic intrinsics are used, this
1493 model places only one restriction on IR transformations on top of what
1494 is required for single-threaded execution: introducing a store to a byte
1495 which might not otherwise be stored is not allowed in general.
1496 (Specifically, in the case where another thread might write to and read
1497 from an address, introducing a store can change a load that may see
1498 exactly one write into a load that may see multiple writes.)
1502 Atomic Memory Ordering Constraints
1503 ----------------------------------
1505 Atomic instructions (:ref:`cmpxchg <i_cmpxchg>`,
1506 :ref:`atomicrmw <i_atomicrmw>`, :ref:`fence <i_fence>`,
1507 :ref:`atomic load <i_load>`, and :ref:`atomic store <i_store>`) take
1508 ordering parameters that determine which other atomic instructions on
1509 the same address they *synchronize with*. These semantics are borrowed
1510 from Java and C++0x, but are somewhat more colloquial. If these
1511 descriptions aren't precise enough, check those specs (see spec
1512 references in the :doc:`atomics guide <Atomics>`).
1513 :ref:`fence <i_fence>` instructions treat these orderings somewhat
1514 differently since they don't take an address. See that instruction's
1515 documentation for details.
1517 For a simpler introduction to the ordering constraints, see the
1521 The set of values that can be read is governed by the happens-before
1522 partial order. A value cannot be read unless some operation wrote
1523 it. This is intended to provide a guarantee strong enough to model
1524 Java's non-volatile shared variables. This ordering cannot be
1525 specified for read-modify-write operations; it is not strong enough
1526 to make them atomic in any interesting way.
1528 In addition to the guarantees of ``unordered``, there is a single
1529 total order for modifications by ``monotonic`` operations on each
1530 address. All modification orders must be compatible with the
1531 happens-before order. There is no guarantee that the modification
1532 orders can be combined to a global total order for the whole program
1533 (and this often will not be possible). The read in an atomic
1534 read-modify-write operation (:ref:`cmpxchg <i_cmpxchg>` and
1535 :ref:`atomicrmw <i_atomicrmw>`) reads the value in the modification
1536 order immediately before the value it writes. If one atomic read
1537 happens before another atomic read of the same address, the later
1538 read must see the same value or a later value in the address's
1539 modification order. This disallows reordering of ``monotonic`` (or
1540 stronger) operations on the same address. If an address is written
1541 ``monotonic``-ally by one thread, and other threads ``monotonic``-ally
1542 read that address repeatedly, the other threads must eventually see
1543 the write. This corresponds to the C++0x/C1x
1544 ``memory_order_relaxed``.
1546 In addition to the guarantees of ``monotonic``, a
1547 *synchronizes-with* edge may be formed with a ``release`` operation.
1548 This is intended to model C++'s ``memory_order_acquire``.
1550 In addition to the guarantees of ``monotonic``, if this operation
1551 writes a value which is subsequently read by an ``acquire``
1552 operation, it *synchronizes-with* that operation. (This isn't a
1553 complete description; see the C++0x definition of a release
1554 sequence.) This corresponds to the C++0x/C1x
1555 ``memory_order_release``.
1556 ``acq_rel`` (acquire+release)
1557 Acts as both an ``acquire`` and ``release`` operation on its
1558 address. This corresponds to the C++0x/C1x ``memory_order_acq_rel``.
1559 ``seq_cst`` (sequentially consistent)
1560 In addition to the guarantees of ``acq_rel`` (``acquire`` for an
1561 operation which only reads, ``release`` for an operation which only
1562 writes), there is a global total order on all
1563 sequentially-consistent operations on all addresses, which is
1564 consistent with the *happens-before* partial order and with the
1565 modification orders of all the affected addresses. Each
1566 sequentially-consistent read sees the last preceding write to the
1567 same address in this global order. This corresponds to the C++0x/C1x
1568 ``memory_order_seq_cst`` and Java volatile.
1572 If an atomic operation is marked ``singlethread``, it only *synchronizes
1573 with* or participates in modification and seq\_cst total orderings with
1574 other operations running in the same thread (for example, in signal
1582 LLVM IR floating-point binary ops (:ref:`fadd <i_fadd>`,
1583 :ref:`fsub <i_fsub>`, :ref:`fmul <i_fmul>`, :ref:`fdiv <i_fdiv>`,
1584 :ref:`frem <i_frem>`) have the following flags that can set to enable
1585 otherwise unsafe floating point operations
1588 No NaNs - Allow optimizations to assume the arguments and result are not
1589 NaN. Such optimizations are required to retain defined behavior over
1590 NaNs, but the value of the result is undefined.
1593 No Infs - Allow optimizations to assume the arguments and result are not
1594 +/-Inf. Such optimizations are required to retain defined behavior over
1595 +/-Inf, but the value of the result is undefined.
1598 No Signed Zeros - Allow optimizations to treat the sign of a zero
1599 argument or result as insignificant.
1602 Allow Reciprocal - Allow optimizations to use the reciprocal of an
1603 argument rather than perform division.
1606 Fast - Allow algebraically equivalent transformations that may
1607 dramatically change results in floating point (e.g. reassociate). This
1608 flag implies all the others.
1615 The LLVM type system is one of the most important features of the
1616 intermediate representation. Being typed enables a number of
1617 optimizations to be performed on the intermediate representation
1618 directly, without having to do extra analyses on the side before the
1619 transformation. A strong type system makes it easier to read the
1620 generated code and enables novel analyses and transformations that are
1621 not feasible to perform on normal three address code representations.
1631 The void type does not represent any value and has no size.
1649 The function type can be thought of as a function signature. It consists of a
1650 return type and a list of formal parameter types. The return type of a function
1651 type is a void type or first class type --- except for :ref:`label <t_label>`
1652 and :ref:`metadata <t_metadata>` types.
1658 <returntype> (<parameter list>)
1660 ...where '``<parameter list>``' is a comma-separated list of type
1661 specifiers. Optionally, the parameter list may include a type ``...``, which
1662 indicates that the function takes a variable number of arguments. Variable
1663 argument functions can access their arguments with the :ref:`variable argument
1664 handling intrinsic <int_varargs>` functions. '``<returntype>``' is any type
1665 except :ref:`label <t_label>` and :ref:`metadata <t_metadata>`.
1669 +---------------------------------+---------------------------------------------------------------------------------------------------------------------------------------------------------------------+
1670 | ``i32 (i32)`` | function taking an ``i32``, returning an ``i32`` |
1671 +---------------------------------+---------------------------------------------------------------------------------------------------------------------------------------------------------------------+
1672 | ``float (i16, i32 *) *`` | :ref:`Pointer <t_pointer>` to a function that takes an ``i16`` and a :ref:`pointer <t_pointer>` to ``i32``, returning ``float``. |
1673 +---------------------------------+---------------------------------------------------------------------------------------------------------------------------------------------------------------------+
1674 | ``i32 (i8*, ...)`` | A vararg function that takes at least one :ref:`pointer <t_pointer>` to ``i8`` (char in C), which returns an integer. This is the signature for ``printf`` in LLVM. |
1675 +---------------------------------+---------------------------------------------------------------------------------------------------------------------------------------------------------------------+
1676 | ``{i32, i32} (i32)`` | A function taking an ``i32``, returning a :ref:`structure <t_struct>` containing two ``i32`` values |
1677 +---------------------------------+---------------------------------------------------------------------------------------------------------------------------------------------------------------------+
1684 The :ref:`first class <t_firstclass>` types are perhaps the most important.
1685 Values of these types are the only ones which can be produced by
1693 These are the types that are valid in registers from CodeGen's perspective.
1702 The integer type is a very simple type that simply specifies an
1703 arbitrary bit width for the integer type desired. Any bit width from 1
1704 bit to 2\ :sup:`23`\ -1 (about 8 million) can be specified.
1712 The number of bits the integer will occupy is specified by the ``N``
1718 +----------------+------------------------------------------------+
1719 | ``i1`` | a single-bit integer. |
1720 +----------------+------------------------------------------------+
1721 | ``i32`` | a 32-bit integer. |
1722 +----------------+------------------------------------------------+
1723 | ``i1942652`` | a really big integer of over 1 million bits. |
1724 +----------------+------------------------------------------------+
1728 Floating Point Types
1729 """"""""""""""""""""
1738 - 16-bit floating point value
1741 - 32-bit floating point value
1744 - 64-bit floating point value
1747 - 128-bit floating point value (112-bit mantissa)
1750 - 80-bit floating point value (X87)
1753 - 128-bit floating point value (two 64-bits)
1760 The x86_mmx type represents a value held in an MMX register on an x86
1761 machine. The operations allowed on it are quite limited: parameters and
1762 return values, load and store, and bitcast. User-specified MMX
1763 instructions are represented as intrinsic or asm calls with arguments
1764 and/or results of this type. There are no arrays, vectors or constants
1781 The pointer type is used to specify memory locations. Pointers are
1782 commonly used to reference objects in memory.
1784 Pointer types may have an optional address space attribute defining the
1785 numbered address space where the pointed-to object resides. The default
1786 address space is number zero. The semantics of non-zero address spaces
1787 are target-specific.
1789 Note that LLVM does not permit pointers to void (``void*``) nor does it
1790 permit pointers to labels (``label*``). Use ``i8*`` instead.
1800 +-------------------------+--------------------------------------------------------------------------------------------------------------+
1801 | ``[4 x i32]*`` | A :ref:`pointer <t_pointer>` to :ref:`array <t_array>` of four ``i32`` values. |
1802 +-------------------------+--------------------------------------------------------------------------------------------------------------+
1803 | ``i32 (i32*) *`` | A :ref:`pointer <t_pointer>` to a :ref:`function <t_function>` that takes an ``i32*``, returning an ``i32``. |
1804 +-------------------------+--------------------------------------------------------------------------------------------------------------+
1805 | ``i32 addrspace(5)*`` | A :ref:`pointer <t_pointer>` to an ``i32`` value that resides in address space #5. |
1806 +-------------------------+--------------------------------------------------------------------------------------------------------------+
1815 A vector type is a simple derived type that represents a vector of
1816 elements. Vector types are used when multiple primitive data are
1817 operated in parallel using a single instruction (SIMD). A vector type
1818 requires a size (number of elements) and an underlying primitive data
1819 type. Vector types are considered :ref:`first class <t_firstclass>`.
1825 < <# elements> x <elementtype> >
1827 The number of elements is a constant integer value larger than 0;
1828 elementtype may be any integer or floating point type, or a pointer to
1829 these types. Vectors of size zero are not allowed.
1833 +-------------------+--------------------------------------------------+
1834 | ``<4 x i32>`` | Vector of 4 32-bit integer values. |
1835 +-------------------+--------------------------------------------------+
1836 | ``<8 x float>`` | Vector of 8 32-bit floating-point values. |
1837 +-------------------+--------------------------------------------------+
1838 | ``<2 x i64>`` | Vector of 2 64-bit integer values. |
1839 +-------------------+--------------------------------------------------+
1840 | ``<4 x i64*>`` | Vector of 4 pointers to 64-bit integer values. |
1841 +-------------------+--------------------------------------------------+
1850 The label type represents code labels.
1865 The metadata type represents embedded metadata. No derived types may be
1866 created from metadata except for :ref:`function <t_function>` arguments.
1879 Aggregate Types are a subset of derived types that can contain multiple
1880 member types. :ref:`Arrays <t_array>` and :ref:`structs <t_struct>` are
1881 aggregate types. :ref:`Vectors <t_vector>` are not considered to be
1891 The array type is a very simple derived type that arranges elements
1892 sequentially in memory. The array type requires a size (number of
1893 elements) and an underlying data type.
1899 [<# elements> x <elementtype>]
1901 The number of elements is a constant integer value; ``elementtype`` may
1902 be any type with a size.
1906 +------------------+--------------------------------------+
1907 | ``[40 x i32]`` | Array of 40 32-bit integer values. |
1908 +------------------+--------------------------------------+
1909 | ``[41 x i32]`` | Array of 41 32-bit integer values. |
1910 +------------------+--------------------------------------+
1911 | ``[4 x i8]`` | Array of 4 8-bit integer values. |
1912 +------------------+--------------------------------------+
1914 Here are some examples of multidimensional arrays:
1916 +-----------------------------+----------------------------------------------------------+
1917 | ``[3 x [4 x i32]]`` | 3x4 array of 32-bit integer values. |
1918 +-----------------------------+----------------------------------------------------------+
1919 | ``[12 x [10 x float]]`` | 12x10 array of single precision floating point values. |
1920 +-----------------------------+----------------------------------------------------------+
1921 | ``[2 x [3 x [4 x i16]]]`` | 2x3x4 array of 16-bit integer values. |
1922 +-----------------------------+----------------------------------------------------------+
1924 There is no restriction on indexing beyond the end of the array implied
1925 by a static type (though there are restrictions on indexing beyond the
1926 bounds of an allocated object in some cases). This means that
1927 single-dimension 'variable sized array' addressing can be implemented in
1928 LLVM with a zero length array type. An implementation of 'pascal style
1929 arrays' in LLVM could use the type "``{ i32, [0 x float]}``", for
1939 The structure type is used to represent a collection of data members
1940 together in memory. The elements of a structure may be any type that has
1943 Structures in memory are accessed using '``load``' and '``store``' by
1944 getting a pointer to a field with the '``getelementptr``' instruction.
1945 Structures in registers are accessed using the '``extractvalue``' and
1946 '``insertvalue``' instructions.
1948 Structures may optionally be "packed" structures, which indicate that
1949 the alignment of the struct is one byte, and that there is no padding
1950 between the elements. In non-packed structs, padding between field types
1951 is inserted as defined by the DataLayout string in the module, which is
1952 required to match what the underlying code generator expects.
1954 Structures can either be "literal" or "identified". A literal structure
1955 is defined inline with other types (e.g. ``{i32, i32}*``) whereas
1956 identified types are always defined at the top level with a name.
1957 Literal types are uniqued by their contents and can never be recursive
1958 or opaque since there is no way to write one. Identified types can be
1959 recursive, can be opaqued, and are never uniqued.
1965 %T1 = type { <type list> } ; Identified normal struct type
1966 %T2 = type <{ <type list> }> ; Identified packed struct type
1970 +------------------------------+---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
1971 | ``{ i32, i32, i32 }`` | A triple of three ``i32`` values |
1972 +------------------------------+---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
1973 | ``{ float, i32 (i32) * }`` | A pair, where the first element is a ``float`` and the second element is a :ref:`pointer <t_pointer>` to a :ref:`function <t_function>` that takes an ``i32``, returning an ``i32``. |
1974 +------------------------------+---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
1975 | ``<{ i8, i32 }>`` | A packed struct known to be 5 bytes in size. |
1976 +------------------------------+---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
1980 Opaque Structure Types
1981 """"""""""""""""""""""
1985 Opaque structure types are used to represent named structure types that
1986 do not have a body specified. This corresponds (for example) to the C
1987 notion of a forward declared structure.
1998 +--------------+-------------------+
1999 | ``opaque`` | An opaque type. |
2000 +--------------+-------------------+
2007 LLVM has several different basic types of constants. This section
2008 describes them all and their syntax.
2013 **Boolean constants**
2014 The two strings '``true``' and '``false``' are both valid constants
2016 **Integer constants**
2017 Standard integers (such as '4') are constants of the
2018 :ref:`integer <t_integer>` type. Negative numbers may be used with
2020 **Floating point constants**
2021 Floating point constants use standard decimal notation (e.g.
2022 123.421), exponential notation (e.g. 1.23421e+2), or a more precise
2023 hexadecimal notation (see below). The assembler requires the exact
2024 decimal value of a floating-point constant. For example, the
2025 assembler accepts 1.25 but rejects 1.3 because 1.3 is a repeating
2026 decimal in binary. Floating point constants must have a :ref:`floating
2027 point <t_floating>` type.
2028 **Null pointer constants**
2029 The identifier '``null``' is recognized as a null pointer constant
2030 and must be of :ref:`pointer type <t_pointer>`.
2032 The one non-intuitive notation for constants is the hexadecimal form of
2033 floating point constants. For example, the form
2034 '``double 0x432ff973cafa8000``' is equivalent to (but harder to read
2035 than) '``double 4.5e+15``'. The only time hexadecimal floating point
2036 constants are required (and the only time that they are generated by the
2037 disassembler) is when a floating point constant must be emitted but it
2038 cannot be represented as a decimal floating point number in a reasonable
2039 number of digits. For example, NaN's, infinities, and other special
2040 values are represented in their IEEE hexadecimal format so that assembly
2041 and disassembly do not cause any bits to change in the constants.
2043 When using the hexadecimal form, constants of types half, float, and
2044 double are represented using the 16-digit form shown above (which
2045 matches the IEEE754 representation for double); half and float values
2046 must, however, be exactly representable as IEEE 754 half and single
2047 precision, respectively. Hexadecimal format is always used for long
2048 double, and there are three forms of long double. The 80-bit format used
2049 by x86 is represented as ``0xK`` followed by 20 hexadecimal digits. The
2050 128-bit format used by PowerPC (two adjacent doubles) is represented by
2051 ``0xM`` followed by 32 hexadecimal digits. The IEEE 128-bit format is
2052 represented by ``0xL`` followed by 32 hexadecimal digits. Long doubles
2053 will only work if they match the long double format on your target.
2054 The IEEE 16-bit format (half precision) is represented by ``0xH``
2055 followed by 4 hexadecimal digits. All hexadecimal formats are big-endian
2056 (sign bit at the left).
2058 There are no constants of type x86_mmx.
2060 .. _complexconstants:
2065 Complex constants are a (potentially recursive) combination of simple
2066 constants and smaller complex constants.
2068 **Structure constants**
2069 Structure constants are represented with notation similar to
2070 structure type definitions (a comma separated list of elements,
2071 surrounded by braces (``{}``)). For example:
2072 "``{ i32 4, float 17.0, i32* @G }``", where "``@G``" is declared as
2073 "``@G = external global i32``". Structure constants must have
2074 :ref:`structure type <t_struct>`, and the number and types of elements
2075 must match those specified by the type.
2077 Array constants are represented with notation similar to array type
2078 definitions (a comma separated list of elements, surrounded by
2079 square brackets (``[]``)). For example:
2080 "``[ i32 42, i32 11, i32 74 ]``". Array constants must have
2081 :ref:`array type <t_array>`, and the number and types of elements must
2082 match those specified by the type.
2083 **Vector constants**
2084 Vector constants are represented with notation similar to vector
2085 type definitions (a comma separated list of elements, surrounded by
2086 less-than/greater-than's (``<>``)). For example:
2087 "``< i32 42, i32 11, i32 74, i32 100 >``". Vector constants
2088 must have :ref:`vector type <t_vector>`, and the number and types of
2089 elements must match those specified by the type.
2090 **Zero initialization**
2091 The string '``zeroinitializer``' can be used to zero initialize a
2092 value to zero of *any* type, including scalar and
2093 :ref:`aggregate <t_aggregate>` types. This is often used to avoid
2094 having to print large zero initializers (e.g. for large arrays) and
2095 is always exactly equivalent to using explicit zero initializers.
2097 A metadata node is a structure-like constant with :ref:`metadata
2098 type <t_metadata>`. For example:
2099 "``metadata !{ i32 0, metadata !"test" }``". Unlike other
2100 constants that are meant to be interpreted as part of the
2101 instruction stream, metadata is a place to attach additional
2102 information such as debug info.
2104 Global Variable and Function Addresses
2105 --------------------------------------
2107 The addresses of :ref:`global variables <globalvars>` and
2108 :ref:`functions <functionstructure>` are always implicitly valid
2109 (link-time) constants. These constants are explicitly referenced when
2110 the :ref:`identifier for the global <identifiers>` is used and always have
2111 :ref:`pointer <t_pointer>` type. For example, the following is a legal LLVM
2114 .. code-block:: llvm
2118 @Z = global [2 x i32*] [ i32* @X, i32* @Y ]
2125 The string '``undef``' can be used anywhere a constant is expected, and
2126 indicates that the user of the value may receive an unspecified
2127 bit-pattern. Undefined values may be of any type (other than '``label``'
2128 or '``void``') and be used anywhere a constant is permitted.
2130 Undefined values are useful because they indicate to the compiler that
2131 the program is well defined no matter what value is used. This gives the
2132 compiler more freedom to optimize. Here are some examples of
2133 (potentially surprising) transformations that are valid (in pseudo IR):
2135 .. code-block:: llvm
2145 This is safe because all of the output bits are affected by the undef
2146 bits. Any output bit can have a zero or one depending on the input bits.
2148 .. code-block:: llvm
2159 These logical operations have bits that are not always affected by the
2160 input. For example, if ``%X`` has a zero bit, then the output of the
2161 '``and``' operation will always be a zero for that bit, no matter what
2162 the corresponding bit from the '``undef``' is. As such, it is unsafe to
2163 optimize or assume that the result of the '``and``' is '``undef``'.
2164 However, it is safe to assume that all bits of the '``undef``' could be
2165 0, and optimize the '``and``' to 0. Likewise, it is safe to assume that
2166 all the bits of the '``undef``' operand to the '``or``' could be set,
2167 allowing the '``or``' to be folded to -1.
2169 .. code-block:: llvm
2171 %A = select undef, %X, %Y
2172 %B = select undef, 42, %Y
2173 %C = select %X, %Y, undef
2183 This set of examples shows that undefined '``select``' (and conditional
2184 branch) conditions can go *either way*, but they have to come from one
2185 of the two operands. In the ``%A`` example, if ``%X`` and ``%Y`` were
2186 both known to have a clear low bit, then ``%A`` would have to have a
2187 cleared low bit. However, in the ``%C`` example, the optimizer is
2188 allowed to assume that the '``undef``' operand could be the same as
2189 ``%Y``, allowing the whole '``select``' to be eliminated.
2191 .. code-block:: llvm
2193 %A = xor undef, undef
2210 This example points out that two '``undef``' operands are not
2211 necessarily the same. This can be surprising to people (and also matches
2212 C semantics) where they assume that "``X^X``" is always zero, even if
2213 ``X`` is undefined. This isn't true for a number of reasons, but the
2214 short answer is that an '``undef``' "variable" can arbitrarily change
2215 its value over its "live range". This is true because the variable
2216 doesn't actually *have a live range*. Instead, the value is logically
2217 read from arbitrary registers that happen to be around when needed, so
2218 the value is not necessarily consistent over time. In fact, ``%A`` and
2219 ``%C`` need to have the same semantics or the core LLVM "replace all
2220 uses with" concept would not hold.
2222 .. code-block:: llvm
2230 These examples show the crucial difference between an *undefined value*
2231 and *undefined behavior*. An undefined value (like '``undef``') is
2232 allowed to have an arbitrary bit-pattern. This means that the ``%A``
2233 operation can be constant folded to '``undef``', because the '``undef``'
2234 could be an SNaN, and ``fdiv`` is not (currently) defined on SNaN's.
2235 However, in the second example, we can make a more aggressive
2236 assumption: because the ``undef`` is allowed to be an arbitrary value,
2237 we are allowed to assume that it could be zero. Since a divide by zero
2238 has *undefined behavior*, we are allowed to assume that the operation
2239 does not execute at all. This allows us to delete the divide and all
2240 code after it. Because the undefined operation "can't happen", the
2241 optimizer can assume that it occurs in dead code.
2243 .. code-block:: llvm
2245 a: store undef -> %X
2246 b: store %X -> undef
2251 These examples reiterate the ``fdiv`` example: a store *of* an undefined
2252 value can be assumed to not have any effect; we can assume that the
2253 value is overwritten with bits that happen to match what was already
2254 there. However, a store *to* an undefined location could clobber
2255 arbitrary memory, therefore, it has undefined behavior.
2262 Poison values are similar to :ref:`undef values <undefvalues>`, however
2263 they also represent the fact that an instruction or constant expression
2264 which cannot evoke side effects has nevertheless detected a condition
2265 which results in undefined behavior.
2267 There is currently no way of representing a poison value in the IR; they
2268 only exist when produced by operations such as :ref:`add <i_add>` with
2271 Poison value behavior is defined in terms of value *dependence*:
2273 - Values other than :ref:`phi <i_phi>` nodes depend on their operands.
2274 - :ref:`Phi <i_phi>` nodes depend on the operand corresponding to
2275 their dynamic predecessor basic block.
2276 - Function arguments depend on the corresponding actual argument values
2277 in the dynamic callers of their functions.
2278 - :ref:`Call <i_call>` instructions depend on the :ref:`ret <i_ret>`
2279 instructions that dynamically transfer control back to them.
2280 - :ref:`Invoke <i_invoke>` instructions depend on the
2281 :ref:`ret <i_ret>`, :ref:`resume <i_resume>`, or exception-throwing
2282 call instructions that dynamically transfer control back to them.
2283 - Non-volatile loads and stores depend on the most recent stores to all
2284 of the referenced memory addresses, following the order in the IR
2285 (including loads and stores implied by intrinsics such as
2286 :ref:`@llvm.memcpy <int_memcpy>`.)
2287 - An instruction with externally visible side effects depends on the
2288 most recent preceding instruction with externally visible side
2289 effects, following the order in the IR. (This includes :ref:`volatile
2290 operations <volatile>`.)
2291 - An instruction *control-depends* on a :ref:`terminator
2292 instruction <terminators>` if the terminator instruction has
2293 multiple successors and the instruction is always executed when
2294 control transfers to one of the successors, and may not be executed
2295 when control is transferred to another.
2296 - Additionally, an instruction also *control-depends* on a terminator
2297 instruction if the set of instructions it otherwise depends on would
2298 be different if the terminator had transferred control to a different
2300 - Dependence is transitive.
2302 Poison Values have the same behavior as :ref:`undef values <undefvalues>`,
2303 with the additional affect that any instruction which has a *dependence*
2304 on a poison value has undefined behavior.
2306 Here are some examples:
2308 .. code-block:: llvm
2311 %poison = sub nuw i32 0, 1 ; Results in a poison value.
2312 %still_poison = and i32 %poison, 0 ; 0, but also poison.
2313 %poison_yet_again = getelementptr i32* @h, i32 %still_poison
2314 store i32 0, i32* %poison_yet_again ; memory at @h[0] is poisoned
2316 store i32 %poison, i32* @g ; Poison value stored to memory.
2317 %poison2 = load i32* @g ; Poison value loaded back from memory.
2319 store volatile i32 %poison, i32* @g ; External observation; undefined behavior.
2321 %narrowaddr = bitcast i32* @g to i16*
2322 %wideaddr = bitcast i32* @g to i64*
2323 %poison3 = load i16* %narrowaddr ; Returns a poison value.
2324 %poison4 = load i64* %wideaddr ; Returns a poison value.
2326 %cmp = icmp slt i32 %poison, 0 ; Returns a poison value.
2327 br i1 %cmp, label %true, label %end ; Branch to either destination.
2330 store volatile i32 0, i32* @g ; This is control-dependent on %cmp, so
2331 ; it has undefined behavior.
2335 %p = phi i32 [ 0, %entry ], [ 1, %true ]
2336 ; Both edges into this PHI are
2337 ; control-dependent on %cmp, so this
2338 ; always results in a poison value.
2340 store volatile i32 0, i32* @g ; This would depend on the store in %true
2341 ; if %cmp is true, or the store in %entry
2342 ; otherwise, so this is undefined behavior.
2344 br i1 %cmp, label %second_true, label %second_end
2345 ; The same branch again, but this time the
2346 ; true block doesn't have side effects.
2353 store volatile i32 0, i32* @g ; This time, the instruction always depends
2354 ; on the store in %end. Also, it is
2355 ; control-equivalent to %end, so this is
2356 ; well-defined (ignoring earlier undefined
2357 ; behavior in this example).
2361 Addresses of Basic Blocks
2362 -------------------------
2364 ``blockaddress(@function, %block)``
2366 The '``blockaddress``' constant computes the address of the specified
2367 basic block in the specified function, and always has an ``i8*`` type.
2368 Taking the address of the entry block is illegal.
2370 This value only has defined behavior when used as an operand to the
2371 ':ref:`indirectbr <i_indirectbr>`' instruction, or for comparisons
2372 against null. Pointer equality tests between labels addresses results in
2373 undefined behavior --- though, again, comparison against null is ok, and
2374 no label is equal to the null pointer. This may be passed around as an
2375 opaque pointer sized value as long as the bits are not inspected. This
2376 allows ``ptrtoint`` and arithmetic to be performed on these values so
2377 long as the original value is reconstituted before the ``indirectbr``
2380 Finally, some targets may provide defined semantics when using the value
2381 as the operand to an inline assembly, but that is target specific.
2385 Constant Expressions
2386 --------------------
2388 Constant expressions are used to allow expressions involving other
2389 constants to be used as constants. Constant expressions may be of any
2390 :ref:`first class <t_firstclass>` type and may involve any LLVM operation
2391 that does not have side effects (e.g. load and call are not supported).
2392 The following is the syntax for constant expressions:
2394 ``trunc (CST to TYPE)``
2395 Truncate a constant to another type. The bit size of CST must be
2396 larger than the bit size of TYPE. Both types must be integers.
2397 ``zext (CST to TYPE)``
2398 Zero extend a constant to another type. The bit size of CST must be
2399 smaller than the bit size of TYPE. Both types must be integers.
2400 ``sext (CST to TYPE)``
2401 Sign extend a constant to another type. The bit size of CST must be
2402 smaller than the bit size of TYPE. Both types must be integers.
2403 ``fptrunc (CST to TYPE)``
2404 Truncate a floating point constant to another floating point type.
2405 The size of CST must be larger than the size of TYPE. Both types
2406 must be floating point.
2407 ``fpext (CST to TYPE)``
2408 Floating point extend a constant to another type. The size of CST
2409 must be smaller or equal to the size of TYPE. Both types must be
2411 ``fptoui (CST to TYPE)``
2412 Convert a floating point constant to the corresponding unsigned
2413 integer constant. TYPE must be a scalar or vector integer type. CST
2414 must be of scalar or vector floating point type. Both CST and TYPE
2415 must be scalars, or vectors of the same number of elements. If the
2416 value won't fit in the integer type, the results are undefined.
2417 ``fptosi (CST to TYPE)``
2418 Convert a floating point constant to the corresponding signed
2419 integer constant. TYPE must be a scalar or vector integer type. CST
2420 must be of scalar or vector floating point type. Both CST and TYPE
2421 must be scalars, or vectors of the same number of elements. If the
2422 value won't fit in the integer type, the results are undefined.
2423 ``uitofp (CST to TYPE)``
2424 Convert an unsigned integer constant to the corresponding floating
2425 point constant. TYPE must be a scalar or vector floating point type.
2426 CST must be of scalar or vector integer type. Both CST and TYPE must
2427 be scalars, or vectors of the same number of elements. If the value
2428 won't fit in the floating point type, the results are undefined.
2429 ``sitofp (CST to TYPE)``
2430 Convert a signed integer constant to the corresponding floating
2431 point constant. TYPE must be a scalar or vector floating point type.
2432 CST must be of scalar or vector integer type. Both CST and TYPE must
2433 be scalars, or vectors of the same number of elements. If the value
2434 won't fit in the floating point type, the results are undefined.
2435 ``ptrtoint (CST to TYPE)``
2436 Convert a pointer typed constant to the corresponding integer
2437 constant. ``TYPE`` must be an integer type. ``CST`` must be of
2438 pointer type. The ``CST`` value is zero extended, truncated, or
2439 unchanged to make it fit in ``TYPE``.
2440 ``inttoptr (CST to TYPE)``
2441 Convert an integer constant to a pointer constant. TYPE must be a
2442 pointer type. CST must be of integer type. The CST value is zero
2443 extended, truncated, or unchanged to make it fit in a pointer size.
2444 This one is *really* dangerous!
2445 ``bitcast (CST to TYPE)``
2446 Convert a constant, CST, to another TYPE. The constraints of the
2447 operands are the same as those for the :ref:`bitcast
2448 instruction <i_bitcast>`.
2449 ``addrspacecast (CST to TYPE)``
2450 Convert a constant pointer or constant vector of pointer, CST, to another
2451 TYPE in a different address space. The constraints of the operands are the
2452 same as those for the :ref:`addrspacecast instruction <i_addrspacecast>`.
2453 ``getelementptr (CSTPTR, IDX0, IDX1, ...)``, ``getelementptr inbounds (CSTPTR, IDX0, IDX1, ...)``
2454 Perform the :ref:`getelementptr operation <i_getelementptr>` on
2455 constants. As with the :ref:`getelementptr <i_getelementptr>`
2456 instruction, the index list may have zero or more indexes, which are
2457 required to make sense for the type of "CSTPTR".
2458 ``select (COND, VAL1, VAL2)``
2459 Perform the :ref:`select operation <i_select>` on constants.
2460 ``icmp COND (VAL1, VAL2)``
2461 Performs the :ref:`icmp operation <i_icmp>` on constants.
2462 ``fcmp COND (VAL1, VAL2)``
2463 Performs the :ref:`fcmp operation <i_fcmp>` on constants.
2464 ``extractelement (VAL, IDX)``
2465 Perform the :ref:`extractelement operation <i_extractelement>` on
2467 ``insertelement (VAL, ELT, IDX)``
2468 Perform the :ref:`insertelement operation <i_insertelement>` on
2470 ``shufflevector (VEC1, VEC2, IDXMASK)``
2471 Perform the :ref:`shufflevector operation <i_shufflevector>` on
2473 ``extractvalue (VAL, IDX0, IDX1, ...)``
2474 Perform the :ref:`extractvalue operation <i_extractvalue>` on
2475 constants. The index list is interpreted in a similar manner as
2476 indices in a ':ref:`getelementptr <i_getelementptr>`' operation. At
2477 least one index value must be specified.
2478 ``insertvalue (VAL, ELT, IDX0, IDX1, ...)``
2479 Perform the :ref:`insertvalue operation <i_insertvalue>` on constants.
2480 The index list is interpreted in a similar manner as indices in a
2481 ':ref:`getelementptr <i_getelementptr>`' operation. At least one index
2482 value must be specified.
2483 ``OPCODE (LHS, RHS)``
2484 Perform the specified operation of the LHS and RHS constants. OPCODE
2485 may be any of the :ref:`binary <binaryops>` or :ref:`bitwise
2486 binary <bitwiseops>` operations. The constraints on operands are
2487 the same as those for the corresponding instruction (e.g. no bitwise
2488 operations on floating point values are allowed).
2495 Inline Assembler Expressions
2496 ----------------------------
2498 LLVM supports inline assembler expressions (as opposed to :ref:`Module-Level
2499 Inline Assembly <moduleasm>`) through the use of a special value. This
2500 value represents the inline assembler as a string (containing the
2501 instructions to emit), a list of operand constraints (stored as a
2502 string), a flag that indicates whether or not the inline asm expression
2503 has side effects, and a flag indicating whether the function containing
2504 the asm needs to align its stack conservatively. An example inline
2505 assembler expression is:
2507 .. code-block:: llvm
2509 i32 (i32) asm "bswap $0", "=r,r"
2511 Inline assembler expressions may **only** be used as the callee operand
2512 of a :ref:`call <i_call>` or an :ref:`invoke <i_invoke>` instruction.
2513 Thus, typically we have:
2515 .. code-block:: llvm
2517 %X = call i32 asm "bswap $0", "=r,r"(i32 %Y)
2519 Inline asms with side effects not visible in the constraint list must be
2520 marked as having side effects. This is done through the use of the
2521 '``sideeffect``' keyword, like so:
2523 .. code-block:: llvm
2525 call void asm sideeffect "eieio", ""()
2527 In some cases inline asms will contain code that will not work unless
2528 the stack is aligned in some way, such as calls or SSE instructions on
2529 x86, yet will not contain code that does that alignment within the asm.
2530 The compiler should make conservative assumptions about what the asm
2531 might contain and should generate its usual stack alignment code in the
2532 prologue if the '``alignstack``' keyword is present:
2534 .. code-block:: llvm
2536 call void asm alignstack "eieio", ""()
2538 Inline asms also support using non-standard assembly dialects. The
2539 assumed dialect is ATT. When the '``inteldialect``' keyword is present,
2540 the inline asm is using the Intel dialect. Currently, ATT and Intel are
2541 the only supported dialects. An example is:
2543 .. code-block:: llvm
2545 call void asm inteldialect "eieio", ""()
2547 If multiple keywords appear the '``sideeffect``' keyword must come
2548 first, the '``alignstack``' keyword second and the '``inteldialect``'
2554 The call instructions that wrap inline asm nodes may have a
2555 "``!srcloc``" MDNode attached to it that contains a list of constant
2556 integers. If present, the code generator will use the integer as the
2557 location cookie value when report errors through the ``LLVMContext``
2558 error reporting mechanisms. This allows a front-end to correlate backend
2559 errors that occur with inline asm back to the source code that produced
2562 .. code-block:: llvm
2564 call void asm sideeffect "something bad", ""(), !srcloc !42
2566 !42 = !{ i32 1234567 }
2568 It is up to the front-end to make sense of the magic numbers it places
2569 in the IR. If the MDNode contains multiple constants, the code generator
2570 will use the one that corresponds to the line of the asm that the error
2575 Metadata Nodes and Metadata Strings
2576 -----------------------------------
2578 LLVM IR allows metadata to be attached to instructions in the program
2579 that can convey extra information about the code to the optimizers and
2580 code generator. One example application of metadata is source-level
2581 debug information. There are two metadata primitives: strings and nodes.
2582 All metadata has the ``metadata`` type and is identified in syntax by a
2583 preceding exclamation point ('``!``').
2585 A metadata string is a string surrounded by double quotes. It can
2586 contain any character by escaping non-printable characters with
2587 "``\xx``" where "``xx``" is the two digit hex code. For example:
2590 Metadata nodes are represented with notation similar to structure
2591 constants (a comma separated list of elements, surrounded by braces and
2592 preceded by an exclamation point). Metadata nodes can have any values as
2593 their operand. For example:
2595 .. code-block:: llvm
2597 !{ metadata !"test\00", i32 10}
2599 A :ref:`named metadata <namedmetadatastructure>` is a collection of
2600 metadata nodes, which can be looked up in the module symbol table. For
2603 .. code-block:: llvm
2605 !foo = metadata !{!4, !3}
2607 Metadata can be used as function arguments. Here ``llvm.dbg.value``
2608 function is using two metadata arguments:
2610 .. code-block:: llvm
2612 call void @llvm.dbg.value(metadata !24, i64 0, metadata !25)
2614 Metadata can be attached with an instruction. Here metadata ``!21`` is
2615 attached to the ``add`` instruction using the ``!dbg`` identifier:
2617 .. code-block:: llvm
2619 %indvar.next = add i64 %indvar, 1, !dbg !21
2621 More information about specific metadata nodes recognized by the
2622 optimizers and code generator is found below.
2627 In LLVM IR, memory does not have types, so LLVM's own type system is not
2628 suitable for doing TBAA. Instead, metadata is added to the IR to
2629 describe a type system of a higher level language. This can be used to
2630 implement typical C/C++ TBAA, but it can also be used to implement
2631 custom alias analysis behavior for other languages.
2633 The current metadata format is very simple. TBAA metadata nodes have up
2634 to three fields, e.g.:
2636 .. code-block:: llvm
2638 !0 = metadata !{ metadata !"an example type tree" }
2639 !1 = metadata !{ metadata !"int", metadata !0 }
2640 !2 = metadata !{ metadata !"float", metadata !0 }
2641 !3 = metadata !{ metadata !"const float", metadata !2, i64 1 }
2643 The first field is an identity field. It can be any value, usually a
2644 metadata string, which uniquely identifies the type. The most important
2645 name in the tree is the name of the root node. Two trees with different
2646 root node names are entirely disjoint, even if they have leaves with
2649 The second field identifies the type's parent node in the tree, or is
2650 null or omitted for a root node. A type is considered to alias all of
2651 its descendants and all of its ancestors in the tree. Also, a type is
2652 considered to alias all types in other trees, so that bitcode produced
2653 from multiple front-ends is handled conservatively.
2655 If the third field is present, it's an integer which if equal to 1
2656 indicates that the type is "constant" (meaning
2657 ``pointsToConstantMemory`` should return true; see `other useful
2658 AliasAnalysis methods <AliasAnalysis.html#OtherItfs>`_).
2660 '``tbaa.struct``' Metadata
2661 ^^^^^^^^^^^^^^^^^^^^^^^^^^
2663 The :ref:`llvm.memcpy <int_memcpy>` is often used to implement
2664 aggregate assignment operations in C and similar languages, however it
2665 is defined to copy a contiguous region of memory, which is more than
2666 strictly necessary for aggregate types which contain holes due to
2667 padding. Also, it doesn't contain any TBAA information about the fields
2670 ``!tbaa.struct`` metadata can describe which memory subregions in a
2671 memcpy are padding and what the TBAA tags of the struct are.
2673 The current metadata format is very simple. ``!tbaa.struct`` metadata
2674 nodes are a list of operands which are in conceptual groups of three.
2675 For each group of three, the first operand gives the byte offset of a
2676 field in bytes, the second gives its size in bytes, and the third gives
2679 .. code-block:: llvm
2681 !4 = metadata !{ i64 0, i64 4, metadata !1, i64 8, i64 4, metadata !2 }
2683 This describes a struct with two fields. The first is at offset 0 bytes
2684 with size 4 bytes, and has tbaa tag !1. The second is at offset 8 bytes
2685 and has size 4 bytes and has tbaa tag !2.
2687 Note that the fields need not be contiguous. In this example, there is a
2688 4 byte gap between the two fields. This gap represents padding which
2689 does not carry useful data and need not be preserved.
2691 '``fpmath``' Metadata
2692 ^^^^^^^^^^^^^^^^^^^^^
2694 ``fpmath`` metadata may be attached to any instruction of floating point
2695 type. It can be used to express the maximum acceptable error in the
2696 result of that instruction, in ULPs, thus potentially allowing the
2697 compiler to use a more efficient but less accurate method of computing
2698 it. ULP is defined as follows:
2700 If ``x`` is a real number that lies between two finite consecutive
2701 floating-point numbers ``a`` and ``b``, without being equal to one
2702 of them, then ``ulp(x) = |b - a|``, otherwise ``ulp(x)`` is the
2703 distance between the two non-equal finite floating-point numbers
2704 nearest ``x``. Moreover, ``ulp(NaN)`` is ``NaN``.
2706 The metadata node shall consist of a single positive floating point
2707 number representing the maximum relative error, for example:
2709 .. code-block:: llvm
2711 !0 = metadata !{ float 2.5 } ; maximum acceptable inaccuracy is 2.5 ULPs
2713 '``range``' Metadata
2714 ^^^^^^^^^^^^^^^^^^^^
2716 ``range`` metadata may be attached only to loads of integer types. It
2717 expresses the possible ranges the loaded value is in. The ranges are
2718 represented with a flattened list of integers. The loaded value is known
2719 to be in the union of the ranges defined by each consecutive pair. Each
2720 pair has the following properties:
2722 - The type must match the type loaded by the instruction.
2723 - The pair ``a,b`` represents the range ``[a,b)``.
2724 - Both ``a`` and ``b`` are constants.
2725 - The range is allowed to wrap.
2726 - The range should not represent the full or empty set. That is,
2729 In addition, the pairs must be in signed order of the lower bound and
2730 they must be non-contiguous.
2734 .. code-block:: llvm
2736 %a = load i8* %x, align 1, !range !0 ; Can only be 0 or 1
2737 %b = load i8* %y, align 1, !range !1 ; Can only be 255 (-1), 0 or 1
2738 %c = load i8* %z, align 1, !range !2 ; Can only be 0, 1, 3, 4 or 5
2739 %d = load i8* %z, align 1, !range !3 ; Can only be -2, -1, 3, 4 or 5
2741 !0 = metadata !{ i8 0, i8 2 }
2742 !1 = metadata !{ i8 255, i8 2 }
2743 !2 = metadata !{ i8 0, i8 2, i8 3, i8 6 }
2744 !3 = metadata !{ i8 -2, i8 0, i8 3, i8 6 }
2749 It is sometimes useful to attach information to loop constructs. Currently,
2750 loop metadata is implemented as metadata attached to the branch instruction
2751 in the loop latch block. This type of metadata refer to a metadata node that is
2752 guaranteed to be separate for each loop. The loop identifier metadata is
2753 specified with the name ``llvm.loop``.
2755 The loop identifier metadata is implemented using a metadata that refers to
2756 itself to avoid merging it with any other identifier metadata, e.g.,
2757 during module linkage or function inlining. That is, each loop should refer
2758 to their own identification metadata even if they reside in separate functions.
2759 The following example contains loop identifier metadata for two separate loop
2762 .. code-block:: llvm
2764 !0 = metadata !{ metadata !0 }
2765 !1 = metadata !{ metadata !1 }
2767 The loop identifier metadata can be used to specify additional per-loop
2768 metadata. Any operands after the first operand can be treated as user-defined
2769 metadata. For example the ``llvm.vectorizer.unroll`` metadata is understood
2770 by the loop vectorizer to indicate how many times to unroll the loop:
2772 .. code-block:: llvm
2774 br i1 %exitcond, label %._crit_edge, label %.lr.ph, !llvm.loop !0
2776 !0 = metadata !{ metadata !0, metadata !1 }
2777 !1 = metadata !{ metadata !"llvm.vectorizer.unroll", i32 2 }
2782 Metadata types used to annotate memory accesses with information helpful
2783 for optimizations are prefixed with ``llvm.mem``.
2785 '``llvm.mem.parallel_loop_access``' Metadata
2786 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
2788 For a loop to be parallel, in addition to using
2789 the ``llvm.loop`` metadata to mark the loop latch branch instruction,
2790 also all of the memory accessing instructions in the loop body need to be
2791 marked with the ``llvm.mem.parallel_loop_access`` metadata. If there
2792 is at least one memory accessing instruction not marked with the metadata,
2793 the loop must be considered a sequential loop. This causes parallel loops to be
2794 converted to sequential loops due to optimization passes that are unaware of
2795 the parallel semantics and that insert new memory instructions to the loop
2798 Example of a loop that is considered parallel due to its correct use of
2799 both ``llvm.loop`` and ``llvm.mem.parallel_loop_access``
2800 metadata types that refer to the same loop identifier metadata.
2802 .. code-block:: llvm
2806 %val0 = load i32* %arrayidx, !llvm.mem.parallel_loop_access !0
2808 store i32 %val0, i32* %arrayidx1, !llvm.mem.parallel_loop_access !0
2810 br i1 %exitcond, label %for.end, label %for.body, !llvm.loop !0
2814 !0 = metadata !{ metadata !0 }
2816 It is also possible to have nested parallel loops. In that case the
2817 memory accesses refer to a list of loop identifier metadata nodes instead of
2818 the loop identifier metadata node directly:
2820 .. code-block:: llvm
2824 %val1 = load i32* %arrayidx3, !llvm.mem.parallel_loop_access !2
2826 br label %inner.for.body
2830 %val0 = load i32* %arrayidx1, !llvm.mem.parallel_loop_access !0
2832 store i32 %val0, i32* %arrayidx2, !llvm.mem.parallel_loop_access !0
2834 br i1 %exitcond, label %inner.for.end, label %inner.for.body, !llvm.loop !1
2838 store i32 %val1, i32* %arrayidx4, !llvm.mem.parallel_loop_access !2
2840 br i1 %exitcond, label %outer.for.end, label %outer.for.body, !llvm.loop !2
2842 outer.for.end: ; preds = %for.body
2844 !0 = metadata !{ metadata !1, metadata !2 } ; a list of loop identifiers
2845 !1 = metadata !{ metadata !1 } ; an identifier for the inner loop
2846 !2 = metadata !{ metadata !2 } ; an identifier for the outer loop
2848 '``llvm.vectorizer``'
2849 ^^^^^^^^^^^^^^^^^^^^^
2851 Metadata prefixed with ``llvm.vectorizer`` is used to control per-loop
2852 vectorization parameters such as vectorization factor and unroll factor.
2854 ``llvm.vectorizer`` metadata should be used in conjunction with ``llvm.loop``
2855 loop identification metadata.
2857 '``llvm.vectorizer.unroll``' Metadata
2858 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
2860 This metadata instructs the loop vectorizer to unroll the specified
2861 loop exactly ``N`` times.
2863 The first operand is the string ``llvm.vectorizer.unroll`` and the second
2864 operand is an integer specifying the unroll factor. For example:
2866 .. code-block:: llvm
2868 !0 = metadata !{ metadata !"llvm.vectorizer.unroll", i32 4 }
2870 Note that setting ``llvm.vectorizer.unroll`` to 1 disables unrolling of the
2873 If ``llvm.vectorizer.unroll`` is set to 0 then the amount of unrolling will be
2874 determined automatically.
2876 '``llvm.vectorizer.width``' Metadata
2877 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
2879 This metadata sets the target width of the vectorizer to ``N``. Without
2880 this metadata, the vectorizer will choose a width automatically.
2881 Regardless of this metadata, the vectorizer will only vectorize loops if
2882 it believes it is valid to do so.
2884 The first operand is the string ``llvm.vectorizer.width`` and the second
2885 operand is an integer specifying the width. For example:
2887 .. code-block:: llvm
2889 !0 = metadata !{ metadata !"llvm.vectorizer.width", i32 4 }
2891 Note that setting ``llvm.vectorizer.width`` to 1 disables vectorization of the
2894 If ``llvm.vectorizer.width`` is set to 0 then the width will be determined
2897 Module Flags Metadata
2898 =====================
2900 Information about the module as a whole is difficult to convey to LLVM's
2901 subsystems. The LLVM IR isn't sufficient to transmit this information.
2902 The ``llvm.module.flags`` named metadata exists in order to facilitate
2903 this. These flags are in the form of key / value pairs --- much like a
2904 dictionary --- making it easy for any subsystem who cares about a flag to
2907 The ``llvm.module.flags`` metadata contains a list of metadata triplets.
2908 Each triplet has the following form:
2910 - The first element is a *behavior* flag, which specifies the behavior
2911 when two (or more) modules are merged together, and it encounters two
2912 (or more) metadata with the same ID. The supported behaviors are
2914 - The second element is a metadata string that is a unique ID for the
2915 metadata. Each module may only have one flag entry for each unique ID (not
2916 including entries with the **Require** behavior).
2917 - The third element is the value of the flag.
2919 When two (or more) modules are merged together, the resulting
2920 ``llvm.module.flags`` metadata is the union of the modules' flags. That is, for
2921 each unique metadata ID string, there will be exactly one entry in the merged
2922 modules ``llvm.module.flags`` metadata table, and the value for that entry will
2923 be determined by the merge behavior flag, as described below. The only exception
2924 is that entries with the *Require* behavior are always preserved.
2926 The following behaviors are supported:
2937 Emits an error if two values disagree, otherwise the resulting value
2938 is that of the operands.
2942 Emits a warning if two values disagree. The result value will be the
2943 operand for the flag from the first module being linked.
2947 Adds a requirement that another module flag be present and have a
2948 specified value after linking is performed. The value must be a
2949 metadata pair, where the first element of the pair is the ID of the
2950 module flag to be restricted, and the second element of the pair is
2951 the value the module flag should be restricted to. This behavior can
2952 be used to restrict the allowable results (via triggering of an
2953 error) of linking IDs with the **Override** behavior.
2957 Uses the specified value, regardless of the behavior or value of the
2958 other module. If both modules specify **Override**, but the values
2959 differ, an error will be emitted.
2963 Appends the two values, which are required to be metadata nodes.
2967 Appends the two values, which are required to be metadata
2968 nodes. However, duplicate entries in the second list are dropped
2969 during the append operation.
2971 It is an error for a particular unique flag ID to have multiple behaviors,
2972 except in the case of **Require** (which adds restrictions on another metadata
2973 value) or **Override**.
2975 An example of module flags:
2977 .. code-block:: llvm
2979 !0 = metadata !{ i32 1, metadata !"foo", i32 1 }
2980 !1 = metadata !{ i32 4, metadata !"bar", i32 37 }
2981 !2 = metadata !{ i32 2, metadata !"qux", i32 42 }
2982 !3 = metadata !{ i32 3, metadata !"qux",
2984 metadata !"foo", i32 1
2987 !llvm.module.flags = !{ !0, !1, !2, !3 }
2989 - Metadata ``!0`` has the ID ``!"foo"`` and the value '1'. The behavior
2990 if two or more ``!"foo"`` flags are seen is to emit an error if their
2991 values are not equal.
2993 - Metadata ``!1`` has the ID ``!"bar"`` and the value '37'. The
2994 behavior if two or more ``!"bar"`` flags are seen is to use the value
2997 - Metadata ``!2`` has the ID ``!"qux"`` and the value '42'. The
2998 behavior if two or more ``!"qux"`` flags are seen is to emit a
2999 warning if their values are not equal.
3001 - Metadata ``!3`` has the ID ``!"qux"`` and the value:
3005 metadata !{ metadata !"foo", i32 1 }
3007 The behavior is to emit an error if the ``llvm.module.flags`` does not
3008 contain a flag with the ID ``!"foo"`` that has the value '1' after linking is
3011 Objective-C Garbage Collection Module Flags Metadata
3012 ----------------------------------------------------
3014 On the Mach-O platform, Objective-C stores metadata about garbage
3015 collection in a special section called "image info". The metadata
3016 consists of a version number and a bitmask specifying what types of
3017 garbage collection are supported (if any) by the file. If two or more
3018 modules are linked together their garbage collection metadata needs to
3019 be merged rather than appended together.
3021 The Objective-C garbage collection module flags metadata consists of the
3022 following key-value pairs:
3031 * - ``Objective-C Version``
3032 - **[Required]** --- The Objective-C ABI version. Valid values are 1 and 2.
3034 * - ``Objective-C Image Info Version``
3035 - **[Required]** --- The version of the image info section. Currently
3038 * - ``Objective-C Image Info Section``
3039 - **[Required]** --- The section to place the metadata. Valid values are
3040 ``"__OBJC, __image_info, regular"`` for Objective-C ABI version 1, and
3041 ``"__DATA,__objc_imageinfo, regular, no_dead_strip"`` for
3042 Objective-C ABI version 2.
3044 * - ``Objective-C Garbage Collection``
3045 - **[Required]** --- Specifies whether garbage collection is supported or
3046 not. Valid values are 0, for no garbage collection, and 2, for garbage
3047 collection supported.
3049 * - ``Objective-C GC Only``
3050 - **[Optional]** --- Specifies that only garbage collection is supported.
3051 If present, its value must be 6. This flag requires that the
3052 ``Objective-C Garbage Collection`` flag have the value 2.
3054 Some important flag interactions:
3056 - If a module with ``Objective-C Garbage Collection`` set to 0 is
3057 merged with a module with ``Objective-C Garbage Collection`` set to
3058 2, then the resulting module has the
3059 ``Objective-C Garbage Collection`` flag set to 0.
3060 - A module with ``Objective-C Garbage Collection`` set to 0 cannot be
3061 merged with a module with ``Objective-C GC Only`` set to 6.
3063 Automatic Linker Flags Module Flags Metadata
3064 --------------------------------------------
3066 Some targets support embedding flags to the linker inside individual object
3067 files. Typically this is used in conjunction with language extensions which
3068 allow source files to explicitly declare the libraries they depend on, and have
3069 these automatically be transmitted to the linker via object files.
3071 These flags are encoded in the IR using metadata in the module flags section,
3072 using the ``Linker Options`` key. The merge behavior for this flag is required
3073 to be ``AppendUnique``, and the value for the key is expected to be a metadata
3074 node which should be a list of other metadata nodes, each of which should be a
3075 list of metadata strings defining linker options.
3077 For example, the following metadata section specifies two separate sets of
3078 linker options, presumably to link against ``libz`` and the ``Cocoa``
3081 !0 = metadata !{ i32 6, metadata !"Linker Options",
3083 metadata !{ metadata !"-lz" },
3084 metadata !{ metadata !"-framework", metadata !"Cocoa" } } }
3085 !llvm.module.flags = !{ !0 }
3087 The metadata encoding as lists of lists of options, as opposed to a collapsed
3088 list of options, is chosen so that the IR encoding can use multiple option
3089 strings to specify e.g., a single library, while still having that specifier be
3090 preserved as an atomic element that can be recognized by a target specific
3091 assembly writer or object file emitter.
3093 Each individual option is required to be either a valid option for the target's
3094 linker, or an option that is reserved by the target specific assembly writer or
3095 object file emitter. No other aspect of these options is defined by the IR.
3097 .. _intrinsicglobalvariables:
3099 Intrinsic Global Variables
3100 ==========================
3102 LLVM has a number of "magic" global variables that contain data that
3103 affect code generation or other IR semantics. These are documented here.
3104 All globals of this sort should have a section specified as
3105 "``llvm.metadata``". This section and all globals that start with
3106 "``llvm.``" are reserved for use by LLVM.
3110 The '``llvm.used``' Global Variable
3111 -----------------------------------
3113 The ``@llvm.used`` global is an array which has
3114 :ref:`appending linkage <linkage_appending>`. This array contains a list of
3115 pointers to named global variables, functions and aliases which may optionally
3116 have a pointer cast formed of bitcast or getelementptr. For example, a legal
3119 .. code-block:: llvm
3124 @llvm.used = appending global [2 x i8*] [
3126 i8* bitcast (i32* @Y to i8*)
3127 ], section "llvm.metadata"
3129 If a symbol appears in the ``@llvm.used`` list, then the compiler, assembler,
3130 and linker are required to treat the symbol as if there is a reference to the
3131 symbol that it cannot see (which is why they have to be named). For example, if
3132 a variable has internal linkage and no references other than that from the
3133 ``@llvm.used`` list, it cannot be deleted. This is commonly used to represent
3134 references from inline asms and other things the compiler cannot "see", and
3135 corresponds to "``attribute((used))``" in GNU C.
3137 On some targets, the code generator must emit a directive to the
3138 assembler or object file to prevent the assembler and linker from
3139 molesting the symbol.
3141 .. _gv_llvmcompilerused:
3143 The '``llvm.compiler.used``' Global Variable
3144 --------------------------------------------
3146 The ``@llvm.compiler.used`` directive is the same as the ``@llvm.used``
3147 directive, except that it only prevents the compiler from touching the
3148 symbol. On targets that support it, this allows an intelligent linker to
3149 optimize references to the symbol without being impeded as it would be
3152 This is a rare construct that should only be used in rare circumstances,
3153 and should not be exposed to source languages.
3155 .. _gv_llvmglobalctors:
3157 The '``llvm.global_ctors``' Global Variable
3158 -------------------------------------------
3160 .. code-block:: llvm
3162 %0 = type { i32, void ()*, i8* }
3163 @llvm.global_ctors = appending global [1 x %0] [%0 { i32 65535, void ()* @ctor, i8* @data }]
3165 The ``@llvm.global_ctors`` array contains a list of constructor
3166 functions, priorities, and an optional associated global or function.
3167 The functions referenced by this array will be called in ascending order
3168 of priority (i.e. lowest first) when the module is loaded. The order of
3169 functions with the same priority is not defined.
3171 If the third field is present, non-null, and points to a global variable
3172 or function, the initializer function will only run if the associated
3173 data from the current module is not discarded.
3175 .. _llvmglobaldtors:
3177 The '``llvm.global_dtors``' Global Variable
3178 -------------------------------------------
3180 .. code-block:: llvm
3182 %0 = type { i32, void ()*, i8* }
3183 @llvm.global_dtors = appending global [1 x %0] [%0 { i32 65535, void ()* @dtor, i8* @data }]
3185 The ``@llvm.global_dtors`` array contains a list of destructor
3186 functions, priorities, and an optional associated global or function.
3187 The functions referenced by this array will be called in descending
3188 order of priority (i.e. highest first) when the module is loaded. The
3189 order of functions with the same priority is not defined.
3191 If the third field is present, non-null, and points to a global variable
3192 or function, the destructor function will only run if the associated
3193 data from the current module is not discarded.
3195 Instruction Reference
3196 =====================
3198 The LLVM instruction set consists of several different classifications
3199 of instructions: :ref:`terminator instructions <terminators>`, :ref:`binary
3200 instructions <binaryops>`, :ref:`bitwise binary
3201 instructions <bitwiseops>`, :ref:`memory instructions <memoryops>`, and
3202 :ref:`other instructions <otherops>`.
3206 Terminator Instructions
3207 -----------------------
3209 As mentioned :ref:`previously <functionstructure>`, every basic block in a
3210 program ends with a "Terminator" instruction, which indicates which
3211 block should be executed after the current block is finished. These
3212 terminator instructions typically yield a '``void``' value: they produce
3213 control flow, not values (the one exception being the
3214 ':ref:`invoke <i_invoke>`' instruction).
3216 The terminator instructions are: ':ref:`ret <i_ret>`',
3217 ':ref:`br <i_br>`', ':ref:`switch <i_switch>`',
3218 ':ref:`indirectbr <i_indirectbr>`', ':ref:`invoke <i_invoke>`',
3219 ':ref:`resume <i_resume>`', and ':ref:`unreachable <i_unreachable>`'.
3223 '``ret``' Instruction
3224 ^^^^^^^^^^^^^^^^^^^^^
3231 ret <type> <value> ; Return a value from a non-void function
3232 ret void ; Return from void function
3237 The '``ret``' instruction is used to return control flow (and optionally
3238 a value) from a function back to the caller.
3240 There are two forms of the '``ret``' instruction: one that returns a
3241 value and then causes control flow, and one that just causes control
3247 The '``ret``' instruction optionally accepts a single argument, the
3248 return value. The type of the return value must be a ':ref:`first
3249 class <t_firstclass>`' type.
3251 A function is not :ref:`well formed <wellformed>` if it it has a non-void
3252 return type and contains a '``ret``' instruction with no return value or
3253 a return value with a type that does not match its type, or if it has a
3254 void return type and contains a '``ret``' instruction with a return
3260 When the '``ret``' instruction is executed, control flow returns back to
3261 the calling function's context. If the caller is a
3262 ":ref:`call <i_call>`" instruction, execution continues at the
3263 instruction after the call. If the caller was an
3264 ":ref:`invoke <i_invoke>`" instruction, execution continues at the
3265 beginning of the "normal" destination block. If the instruction returns
3266 a value, that value shall set the call or invoke instruction's return
3272 .. code-block:: llvm
3274 ret i32 5 ; Return an integer value of 5
3275 ret void ; Return from a void function
3276 ret { i32, i8 } { i32 4, i8 2 } ; Return a struct of values 4 and 2
3280 '``br``' Instruction
3281 ^^^^^^^^^^^^^^^^^^^^
3288 br i1 <cond>, label <iftrue>, label <iffalse>
3289 br label <dest> ; Unconditional branch
3294 The '``br``' instruction is used to cause control flow to transfer to a
3295 different basic block in the current function. There are two forms of
3296 this instruction, corresponding to a conditional branch and an
3297 unconditional branch.
3302 The conditional branch form of the '``br``' instruction takes a single
3303 '``i1``' value and two '``label``' values. The unconditional form of the
3304 '``br``' instruction takes a single '``label``' value as a target.
3309 Upon execution of a conditional '``br``' instruction, the '``i1``'
3310 argument is evaluated. If the value is ``true``, control flows to the
3311 '``iftrue``' ``label`` argument. If "cond" is ``false``, control flows
3312 to the '``iffalse``' ``label`` argument.
3317 .. code-block:: llvm
3320 %cond = icmp eq i32 %a, %b
3321 br i1 %cond, label %IfEqual, label %IfUnequal
3329 '``switch``' Instruction
3330 ^^^^^^^^^^^^^^^^^^^^^^^^
3337 switch <intty> <value>, label <defaultdest> [ <intty> <val>, label <dest> ... ]
3342 The '``switch``' instruction is used to transfer control flow to one of
3343 several different places. It is a generalization of the '``br``'
3344 instruction, allowing a branch to occur to one of many possible
3350 The '``switch``' instruction uses three parameters: an integer
3351 comparison value '``value``', a default '``label``' destination, and an
3352 array of pairs of comparison value constants and '``label``'s. The table
3353 is not allowed to contain duplicate constant entries.
3358 The ``switch`` instruction specifies a table of values and destinations.
3359 When the '``switch``' instruction is executed, this table is searched
3360 for the given value. If the value is found, control flow is transferred
3361 to the corresponding destination; otherwise, control flow is transferred
3362 to the default destination.
3367 Depending on properties of the target machine and the particular
3368 ``switch`` instruction, this instruction may be code generated in
3369 different ways. For example, it could be generated as a series of
3370 chained conditional branches or with a lookup table.
3375 .. code-block:: llvm
3377 ; Emulate a conditional br instruction
3378 %Val = zext i1 %value to i32
3379 switch i32 %Val, label %truedest [ i32 0, label %falsedest ]
3381 ; Emulate an unconditional br instruction
3382 switch i32 0, label %dest [ ]
3384 ; Implement a jump table:
3385 switch i32 %val, label %otherwise [ i32 0, label %onzero
3387 i32 2, label %ontwo ]
3391 '``indirectbr``' Instruction
3392 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
3399 indirectbr <somety>* <address>, [ label <dest1>, label <dest2>, ... ]
3404 The '``indirectbr``' instruction implements an indirect branch to a
3405 label within the current function, whose address is specified by
3406 "``address``". Address must be derived from a
3407 :ref:`blockaddress <blockaddress>` constant.
3412 The '``address``' argument is the address of the label to jump to. The
3413 rest of the arguments indicate the full set of possible destinations
3414 that the address may point to. Blocks are allowed to occur multiple
3415 times in the destination list, though this isn't particularly useful.
3417 This destination list is required so that dataflow analysis has an
3418 accurate understanding of the CFG.
3423 Control transfers to the block specified in the address argument. All
3424 possible destination blocks must be listed in the label list, otherwise
3425 this instruction has undefined behavior. This implies that jumps to
3426 labels defined in other functions have undefined behavior as well.
3431 This is typically implemented with a jump through a register.
3436 .. code-block:: llvm
3438 indirectbr i8* %Addr, [ label %bb1, label %bb2, label %bb3 ]
3442 '``invoke``' Instruction
3443 ^^^^^^^^^^^^^^^^^^^^^^^^
3450 <result> = invoke [cconv] [ret attrs] <ptr to function ty> <function ptr val>(<function args>) [fn attrs]
3451 to label <normal label> unwind label <exception label>
3456 The '``invoke``' instruction causes control to transfer to a specified
3457 function, with the possibility of control flow transfer to either the
3458 '``normal``' label or the '``exception``' label. If the callee function
3459 returns with the "``ret``" instruction, control flow will return to the
3460 "normal" label. If the callee (or any indirect callees) returns via the
3461 ":ref:`resume <i_resume>`" instruction or other exception handling
3462 mechanism, control is interrupted and continued at the dynamically
3463 nearest "exception" label.
3465 The '``exception``' label is a `landing
3466 pad <ExceptionHandling.html#overview>`_ for the exception. As such,
3467 '``exception``' label is required to have the
3468 ":ref:`landingpad <i_landingpad>`" instruction, which contains the
3469 information about the behavior of the program after unwinding happens,
3470 as its first non-PHI instruction. The restrictions on the
3471 "``landingpad``" instruction's tightly couples it to the "``invoke``"
3472 instruction, so that the important information contained within the
3473 "``landingpad``" instruction can't be lost through normal code motion.
3478 This instruction requires several arguments:
3480 #. The optional "cconv" marker indicates which :ref:`calling
3481 convention <callingconv>` the call should use. If none is
3482 specified, the call defaults to using C calling conventions.
3483 #. The optional :ref:`Parameter Attributes <paramattrs>` list for return
3484 values. Only '``zeroext``', '``signext``', and '``inreg``' attributes
3486 #. '``ptr to function ty``': shall be the signature of the pointer to
3487 function value being invoked. In most cases, this is a direct
3488 function invocation, but indirect ``invoke``'s are just as possible,
3489 branching off an arbitrary pointer to function value.
3490 #. '``function ptr val``': An LLVM value containing a pointer to a
3491 function to be invoked.
3492 #. '``function args``': argument list whose types match the function
3493 signature argument types and parameter attributes. All arguments must
3494 be of :ref:`first class <t_firstclass>` type. If the function signature
3495 indicates the function accepts a variable number of arguments, the
3496 extra arguments can be specified.
3497 #. '``normal label``': the label reached when the called function
3498 executes a '``ret``' instruction.
3499 #. '``exception label``': the label reached when a callee returns via
3500 the :ref:`resume <i_resume>` instruction or other exception handling
3502 #. The optional :ref:`function attributes <fnattrs>` list. Only
3503 '``noreturn``', '``nounwind``', '``readonly``' and '``readnone``'
3504 attributes are valid here.
3509 This instruction is designed to operate as a standard '``call``'
3510 instruction in most regards. The primary difference is that it
3511 establishes an association with a label, which is used by the runtime
3512 library to unwind the stack.
3514 This instruction is used in languages with destructors to ensure that
3515 proper cleanup is performed in the case of either a ``longjmp`` or a
3516 thrown exception. Additionally, this is important for implementation of
3517 '``catch``' clauses in high-level languages that support them.
3519 For the purposes of the SSA form, the definition of the value returned
3520 by the '``invoke``' instruction is deemed to occur on the edge from the
3521 current block to the "normal" label. If the callee unwinds then no
3522 return value is available.
3527 .. code-block:: llvm
3529 %retval = invoke i32 @Test(i32 15) to label %Continue
3530 unwind label %TestCleanup ; {i32}:retval set
3531 %retval = invoke coldcc i32 %Testfnptr(i32 15) to label %Continue
3532 unwind label %TestCleanup ; {i32}:retval set
3536 '``resume``' Instruction
3537 ^^^^^^^^^^^^^^^^^^^^^^^^
3544 resume <type> <value>
3549 The '``resume``' instruction is a terminator instruction that has no
3555 The '``resume``' instruction requires one argument, which must have the
3556 same type as the result of any '``landingpad``' instruction in the same
3562 The '``resume``' instruction resumes propagation of an existing
3563 (in-flight) exception whose unwinding was interrupted with a
3564 :ref:`landingpad <i_landingpad>` instruction.
3569 .. code-block:: llvm
3571 resume { i8*, i32 } %exn
3575 '``unreachable``' Instruction
3576 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
3588 The '``unreachable``' instruction has no defined semantics. This
3589 instruction is used to inform the optimizer that a particular portion of
3590 the code is not reachable. This can be used to indicate that the code
3591 after a no-return function cannot be reached, and other facts.
3596 The '``unreachable``' instruction has no defined semantics.
3603 Binary operators are used to do most of the computation in a program.
3604 They require two operands of the same type, execute an operation on
3605 them, and produce a single value. The operands might represent multiple
3606 data, as is the case with the :ref:`vector <t_vector>` data type. The
3607 result value has the same type as its operands.
3609 There are several different binary operators:
3613 '``add``' Instruction
3614 ^^^^^^^^^^^^^^^^^^^^^
3621 <result> = add <ty> <op1>, <op2> ; yields {ty}:result
3622 <result> = add nuw <ty> <op1>, <op2> ; yields {ty}:result
3623 <result> = add nsw <ty> <op1>, <op2> ; yields {ty}:result
3624 <result> = add nuw nsw <ty> <op1>, <op2> ; yields {ty}:result
3629 The '``add``' instruction returns the sum of its two operands.
3634 The two arguments to the '``add``' instruction must be
3635 :ref:`integer <t_integer>` or :ref:`vector <t_vector>` of integer values. Both
3636 arguments must have identical types.
3641 The value produced is the integer sum of the two operands.
3643 If the sum has unsigned overflow, the result returned is the
3644 mathematical result modulo 2\ :sup:`n`\ , where n is the bit width of
3647 Because LLVM integers use a two's complement representation, this
3648 instruction is appropriate for both signed and unsigned integers.
3650 ``nuw`` and ``nsw`` stand for "No Unsigned Wrap" and "No Signed Wrap",
3651 respectively. If the ``nuw`` and/or ``nsw`` keywords are present, the
3652 result value of the ``add`` is a :ref:`poison value <poisonvalues>` if
3653 unsigned and/or signed overflow, respectively, occurs.
3658 .. code-block:: llvm
3660 <result> = add i32 4, %var ; yields {i32}:result = 4 + %var
3664 '``fadd``' Instruction
3665 ^^^^^^^^^^^^^^^^^^^^^^
3672 <result> = fadd [fast-math flags]* <ty> <op1>, <op2> ; yields {ty}:result
3677 The '``fadd``' instruction returns the sum of its two operands.
3682 The two arguments to the '``fadd``' instruction must be :ref:`floating
3683 point <t_floating>` or :ref:`vector <t_vector>` of floating point values.
3684 Both arguments must have identical types.
3689 The value produced is the floating point sum of the two operands. This
3690 instruction can also take any number of :ref:`fast-math flags <fastmath>`,
3691 which are optimization hints to enable otherwise unsafe floating point
3697 .. code-block:: llvm
3699 <result> = fadd float 4.0, %var ; yields {float}:result = 4.0 + %var
3701 '``sub``' Instruction
3702 ^^^^^^^^^^^^^^^^^^^^^
3709 <result> = sub <ty> <op1>, <op2> ; yields {ty}:result
3710 <result> = sub nuw <ty> <op1>, <op2> ; yields {ty}:result
3711 <result> = sub nsw <ty> <op1>, <op2> ; yields {ty}:result
3712 <result> = sub nuw nsw <ty> <op1>, <op2> ; yields {ty}:result
3717 The '``sub``' instruction returns the difference of its two operands.
3719 Note that the '``sub``' instruction is used to represent the '``neg``'
3720 instruction present in most other intermediate representations.
3725 The two arguments to the '``sub``' instruction must be
3726 :ref:`integer <t_integer>` or :ref:`vector <t_vector>` of integer values. Both
3727 arguments must have identical types.
3732 The value produced is the integer difference of the two operands.
3734 If the difference has unsigned overflow, the result returned is the
3735 mathematical result modulo 2\ :sup:`n`\ , where n is the bit width of
3738 Because LLVM integers use a two's complement representation, this
3739 instruction is appropriate for both signed and unsigned integers.
3741 ``nuw`` and ``nsw`` stand for "No Unsigned Wrap" and "No Signed Wrap",
3742 respectively. If the ``nuw`` and/or ``nsw`` keywords are present, the
3743 result value of the ``sub`` is a :ref:`poison value <poisonvalues>` if
3744 unsigned and/or signed overflow, respectively, occurs.
3749 .. code-block:: llvm
3751 <result> = sub i32 4, %var ; yields {i32}:result = 4 - %var
3752 <result> = sub i32 0, %val ; yields {i32}:result = -%var
3756 '``fsub``' Instruction
3757 ^^^^^^^^^^^^^^^^^^^^^^
3764 <result> = fsub [fast-math flags]* <ty> <op1>, <op2> ; yields {ty}:result
3769 The '``fsub``' instruction returns the difference of its two operands.
3771 Note that the '``fsub``' instruction is used to represent the '``fneg``'
3772 instruction present in most other intermediate representations.
3777 The two arguments to the '``fsub``' instruction must be :ref:`floating
3778 point <t_floating>` or :ref:`vector <t_vector>` of floating point values.
3779 Both arguments must have identical types.
3784 The value produced is the floating point difference of the two operands.
3785 This instruction can also take any number of :ref:`fast-math
3786 flags <fastmath>`, which are optimization hints to enable otherwise
3787 unsafe floating point optimizations:
3792 .. code-block:: llvm
3794 <result> = fsub float 4.0, %var ; yields {float}:result = 4.0 - %var
3795 <result> = fsub float -0.0, %val ; yields {float}:result = -%var
3797 '``mul``' Instruction
3798 ^^^^^^^^^^^^^^^^^^^^^
3805 <result> = mul <ty> <op1>, <op2> ; yields {ty}:result
3806 <result> = mul nuw <ty> <op1>, <op2> ; yields {ty}:result
3807 <result> = mul nsw <ty> <op1>, <op2> ; yields {ty}:result
3808 <result> = mul nuw nsw <ty> <op1>, <op2> ; yields {ty}:result
3813 The '``mul``' instruction returns the product of its two operands.
3818 The two arguments to the '``mul``' instruction must be
3819 :ref:`integer <t_integer>` or :ref:`vector <t_vector>` of integer values. Both
3820 arguments must have identical types.
3825 The value produced is the integer product of the two operands.
3827 If the result of the multiplication has unsigned overflow, the result
3828 returned is the mathematical result modulo 2\ :sup:`n`\ , where n is the
3829 bit width of the result.
3831 Because LLVM integers use a two's complement representation, and the
3832 result is the same width as the operands, this instruction returns the
3833 correct result for both signed and unsigned integers. If a full product
3834 (e.g. ``i32`` * ``i32`` -> ``i64``) is needed, the operands should be
3835 sign-extended or zero-extended as appropriate to the width of the full
3838 ``nuw`` and ``nsw`` stand for "No Unsigned Wrap" and "No Signed Wrap",
3839 respectively. If the ``nuw`` and/or ``nsw`` keywords are present, the
3840 result value of the ``mul`` is a :ref:`poison value <poisonvalues>` if
3841 unsigned and/or signed overflow, respectively, occurs.
3846 .. code-block:: llvm
3848 <result> = mul i32 4, %var ; yields {i32}:result = 4 * %var
3852 '``fmul``' Instruction
3853 ^^^^^^^^^^^^^^^^^^^^^^
3860 <result> = fmul [fast-math flags]* <ty> <op1>, <op2> ; yields {ty}:result
3865 The '``fmul``' instruction returns the product of its two operands.
3870 The two arguments to the '``fmul``' instruction must be :ref:`floating
3871 point <t_floating>` or :ref:`vector <t_vector>` of floating point values.
3872 Both arguments must have identical types.
3877 The value produced is the floating point product of the two operands.
3878 This instruction can also take any number of :ref:`fast-math
3879 flags <fastmath>`, which are optimization hints to enable otherwise
3880 unsafe floating point optimizations:
3885 .. code-block:: llvm
3887 <result> = fmul float 4.0, %var ; yields {float}:result = 4.0 * %var
3889 '``udiv``' Instruction
3890 ^^^^^^^^^^^^^^^^^^^^^^
3897 <result> = udiv <ty> <op1>, <op2> ; yields {ty}:result
3898 <result> = udiv exact <ty> <op1>, <op2> ; yields {ty}:result
3903 The '``udiv``' instruction returns the quotient of its two operands.
3908 The two arguments to the '``udiv``' instruction must be
3909 :ref:`integer <t_integer>` or :ref:`vector <t_vector>` of integer values. Both
3910 arguments must have identical types.
3915 The value produced is the unsigned integer quotient of the two operands.
3917 Note that unsigned integer division and signed integer division are
3918 distinct operations; for signed integer division, use '``sdiv``'.
3920 Division by zero leads to undefined behavior.
3922 If the ``exact`` keyword is present, the result value of the ``udiv`` is
3923 a :ref:`poison value <poisonvalues>` if %op1 is not a multiple of %op2 (as
3924 such, "((a udiv exact b) mul b) == a").
3929 .. code-block:: llvm
3931 <result> = udiv i32 4, %var ; yields {i32}:result = 4 / %var
3933 '``sdiv``' Instruction
3934 ^^^^^^^^^^^^^^^^^^^^^^
3941 <result> = sdiv <ty> <op1>, <op2> ; yields {ty}:result
3942 <result> = sdiv exact <ty> <op1>, <op2> ; yields {ty}:result
3947 The '``sdiv``' instruction returns the quotient of its two operands.
3952 The two arguments to the '``sdiv``' instruction must be
3953 :ref:`integer <t_integer>` or :ref:`vector <t_vector>` of integer values. Both
3954 arguments must have identical types.
3959 The value produced is the signed integer quotient of the two operands
3960 rounded towards zero.
3962 Note that signed integer division and unsigned integer division are
3963 distinct operations; for unsigned integer division, use '``udiv``'.
3965 Division by zero leads to undefined behavior. Overflow also leads to
3966 undefined behavior; this is a rare case, but can occur, for example, by
3967 doing a 32-bit division of -2147483648 by -1.
3969 If the ``exact`` keyword is present, the result value of the ``sdiv`` is
3970 a :ref:`poison value <poisonvalues>` if the result would be rounded.
3975 .. code-block:: llvm
3977 <result> = sdiv i32 4, %var ; yields {i32}:result = 4 / %var
3981 '``fdiv``' Instruction
3982 ^^^^^^^^^^^^^^^^^^^^^^
3989 <result> = fdiv [fast-math flags]* <ty> <op1>, <op2> ; yields {ty}:result
3994 The '``fdiv``' instruction returns the quotient of its two operands.
3999 The two arguments to the '``fdiv``' instruction must be :ref:`floating
4000 point <t_floating>` or :ref:`vector <t_vector>` of floating point values.
4001 Both arguments must have identical types.
4006 The value produced is the floating point quotient of the two operands.
4007 This instruction can also take any number of :ref:`fast-math
4008 flags <fastmath>`, which are optimization hints to enable otherwise
4009 unsafe floating point optimizations:
4014 .. code-block:: llvm
4016 <result> = fdiv float 4.0, %var ; yields {float}:result = 4.0 / %var
4018 '``urem``' Instruction
4019 ^^^^^^^^^^^^^^^^^^^^^^
4026 <result> = urem <ty> <op1>, <op2> ; yields {ty}:result
4031 The '``urem``' instruction returns the remainder from the unsigned
4032 division of its two arguments.
4037 The two arguments to the '``urem``' instruction must be
4038 :ref:`integer <t_integer>` or :ref:`vector <t_vector>` of integer values. Both
4039 arguments must have identical types.
4044 This instruction returns the unsigned integer *remainder* of a division.
4045 This instruction always performs an unsigned division to get the
4048 Note that unsigned integer remainder and signed integer remainder are
4049 distinct operations; for signed integer remainder, use '``srem``'.
4051 Taking the remainder of a division by zero leads to undefined behavior.
4056 .. code-block:: llvm
4058 <result> = urem i32 4, %var ; yields {i32}:result = 4 % %var
4060 '``srem``' Instruction
4061 ^^^^^^^^^^^^^^^^^^^^^^
4068 <result> = srem <ty> <op1>, <op2> ; yields {ty}:result
4073 The '``srem``' instruction returns the remainder from the signed
4074 division of its two operands. This instruction can also take
4075 :ref:`vector <t_vector>` versions of the values in which case the elements
4081 The two arguments to the '``srem``' instruction must be
4082 :ref:`integer <t_integer>` or :ref:`vector <t_vector>` of integer values. Both
4083 arguments must have identical types.
4088 This instruction returns the *remainder* of a division (where the result
4089 is either zero or has the same sign as the dividend, ``op1``), not the
4090 *modulo* operator (where the result is either zero or has the same sign
4091 as the divisor, ``op2``) of a value. For more information about the
4092 difference, see `The Math
4093 Forum <http://mathforum.org/dr.math/problems/anne.4.28.99.html>`_. For a
4094 table of how this is implemented in various languages, please see
4096 operation <http://en.wikipedia.org/wiki/Modulo_operation>`_.
4098 Note that signed integer remainder and unsigned integer remainder are
4099 distinct operations; for unsigned integer remainder, use '``urem``'.
4101 Taking the remainder of a division by zero leads to undefined behavior.
4102 Overflow also leads to undefined behavior; this is a rare case, but can
4103 occur, for example, by taking the remainder of a 32-bit division of
4104 -2147483648 by -1. (The remainder doesn't actually overflow, but this
4105 rule lets srem be implemented using instructions that return both the
4106 result of the division and the remainder.)
4111 .. code-block:: llvm
4113 <result> = srem i32 4, %var ; yields {i32}:result = 4 % %var
4117 '``frem``' Instruction
4118 ^^^^^^^^^^^^^^^^^^^^^^
4125 <result> = frem [fast-math flags]* <ty> <op1>, <op2> ; yields {ty}:result
4130 The '``frem``' instruction returns the remainder from the division of
4136 The two arguments to the '``frem``' instruction must be :ref:`floating
4137 point <t_floating>` or :ref:`vector <t_vector>` of floating point values.
4138 Both arguments must have identical types.
4143 This instruction returns the *remainder* of a division. The remainder
4144 has the same sign as the dividend. This instruction can also take any
4145 number of :ref:`fast-math flags <fastmath>`, which are optimization hints
4146 to enable otherwise unsafe floating point optimizations:
4151 .. code-block:: llvm
4153 <result> = frem float 4.0, %var ; yields {float}:result = 4.0 % %var
4157 Bitwise Binary Operations
4158 -------------------------
4160 Bitwise binary operators are used to do various forms of bit-twiddling
4161 in a program. They are generally very efficient instructions and can
4162 commonly be strength reduced from other instructions. They require two
4163 operands of the same type, execute an operation on them, and produce a
4164 single value. The resulting value is the same type as its operands.
4166 '``shl``' Instruction
4167 ^^^^^^^^^^^^^^^^^^^^^
4174 <result> = shl <ty> <op1>, <op2> ; yields {ty}:result
4175 <result> = shl nuw <ty> <op1>, <op2> ; yields {ty}:result
4176 <result> = shl nsw <ty> <op1>, <op2> ; yields {ty}:result
4177 <result> = shl nuw nsw <ty> <op1>, <op2> ; yields {ty}:result
4182 The '``shl``' instruction returns the first operand shifted to the left
4183 a specified number of bits.
4188 Both arguments to the '``shl``' instruction must be the same
4189 :ref:`integer <t_integer>` or :ref:`vector <t_vector>` of integer type.
4190 '``op2``' is treated as an unsigned value.
4195 The value produced is ``op1`` \* 2\ :sup:`op2` mod 2\ :sup:`n`,
4196 where ``n`` is the width of the result. If ``op2`` is (statically or
4197 dynamically) negative or equal to or larger than the number of bits in
4198 ``op1``, the result is undefined. If the arguments are vectors, each
4199 vector element of ``op1`` is shifted by the corresponding shift amount
4202 If the ``nuw`` keyword is present, then the shift produces a :ref:`poison
4203 value <poisonvalues>` if it shifts out any non-zero bits. If the
4204 ``nsw`` keyword is present, then the shift produces a :ref:`poison
4205 value <poisonvalues>` if it shifts out any bits that disagree with the
4206 resultant sign bit. As such, NUW/NSW have the same semantics as they
4207 would if the shift were expressed as a mul instruction with the same
4208 nsw/nuw bits in (mul %op1, (shl 1, %op2)).
4213 .. code-block:: llvm
4215 <result> = shl i32 4, %var ; yields {i32}: 4 << %var
4216 <result> = shl i32 4, 2 ; yields {i32}: 16
4217 <result> = shl i32 1, 10 ; yields {i32}: 1024
4218 <result> = shl i32 1, 32 ; undefined
4219 <result> = shl <2 x i32> < i32 1, i32 1>, < i32 1, i32 2> ; yields: result=<2 x i32> < i32 2, i32 4>
4221 '``lshr``' Instruction
4222 ^^^^^^^^^^^^^^^^^^^^^^
4229 <result> = lshr <ty> <op1>, <op2> ; yields {ty}:result
4230 <result> = lshr exact <ty> <op1>, <op2> ; yields {ty}:result
4235 The '``lshr``' instruction (logical shift right) returns the first
4236 operand shifted to the right a specified number of bits with zero fill.
4241 Both arguments to the '``lshr``' instruction must be the same
4242 :ref:`integer <t_integer>` or :ref:`vector <t_vector>` of integer type.
4243 '``op2``' is treated as an unsigned value.
4248 This instruction always performs a logical shift right operation. The
4249 most significant bits of the result will be filled with zero bits after
4250 the shift. If ``op2`` is (statically or dynamically) equal to or larger
4251 than the number of bits in ``op1``, the result is undefined. If the
4252 arguments are vectors, each vector element of ``op1`` is shifted by the
4253 corresponding shift amount in ``op2``.
4255 If the ``exact`` keyword is present, the result value of the ``lshr`` is
4256 a :ref:`poison value <poisonvalues>` if any of the bits shifted out are
4262 .. code-block:: llvm
4264 <result> = lshr i32 4, 1 ; yields {i32}:result = 2
4265 <result> = lshr i32 4, 2 ; yields {i32}:result = 1
4266 <result> = lshr i8 4, 3 ; yields {i8}:result = 0
4267 <result> = lshr i8 -2, 1 ; yields {i8}:result = 0x7F
4268 <result> = lshr i32 1, 32 ; undefined
4269 <result> = lshr <2 x i32> < i32 -2, i32 4>, < i32 1, i32 2> ; yields: result=<2 x i32> < i32 0x7FFFFFFF, i32 1>
4271 '``ashr``' Instruction
4272 ^^^^^^^^^^^^^^^^^^^^^^
4279 <result> = ashr <ty> <op1>, <op2> ; yields {ty}:result
4280 <result> = ashr exact <ty> <op1>, <op2> ; yields {ty}:result
4285 The '``ashr``' instruction (arithmetic shift right) returns the first
4286 operand shifted to the right a specified number of bits with sign
4292 Both arguments to the '``ashr``' instruction must be the same
4293 :ref:`integer <t_integer>` or :ref:`vector <t_vector>` of integer type.
4294 '``op2``' is treated as an unsigned value.
4299 This instruction always performs an arithmetic shift right operation,
4300 The most significant bits of the result will be filled with the sign bit
4301 of ``op1``. If ``op2`` is (statically or dynamically) equal to or larger
4302 than the number of bits in ``op1``, the result is undefined. If the
4303 arguments are vectors, each vector element of ``op1`` is shifted by the
4304 corresponding shift amount in ``op2``.
4306 If the ``exact`` keyword is present, the result value of the ``ashr`` is
4307 a :ref:`poison value <poisonvalues>` if any of the bits shifted out are
4313 .. code-block:: llvm
4315 <result> = ashr i32 4, 1 ; yields {i32}:result = 2
4316 <result> = ashr i32 4, 2 ; yields {i32}:result = 1
4317 <result> = ashr i8 4, 3 ; yields {i8}:result = 0
4318 <result> = ashr i8 -2, 1 ; yields {i8}:result = -1
4319 <result> = ashr i32 1, 32 ; undefined
4320 <result> = ashr <2 x i32> < i32 -2, i32 4>, < i32 1, i32 3> ; yields: result=<2 x i32> < i32 -1, i32 0>
4322 '``and``' Instruction
4323 ^^^^^^^^^^^^^^^^^^^^^
4330 <result> = and <ty> <op1>, <op2> ; yields {ty}:result
4335 The '``and``' instruction returns the bitwise logical and of its two
4341 The two arguments to the '``and``' instruction must be
4342 :ref:`integer <t_integer>` or :ref:`vector <t_vector>` of integer values. Both
4343 arguments must have identical types.
4348 The truth table used for the '``and``' instruction is:
4365 .. code-block:: llvm
4367 <result> = and i32 4, %var ; yields {i32}:result = 4 & %var
4368 <result> = and i32 15, 40 ; yields {i32}:result = 8
4369 <result> = and i32 4, 8 ; yields {i32}:result = 0
4371 '``or``' Instruction
4372 ^^^^^^^^^^^^^^^^^^^^
4379 <result> = or <ty> <op1>, <op2> ; yields {ty}:result
4384 The '``or``' instruction returns the bitwise logical inclusive or of its
4390 The two arguments to the '``or``' instruction must be
4391 :ref:`integer <t_integer>` or :ref:`vector <t_vector>` of integer values. Both
4392 arguments must have identical types.
4397 The truth table used for the '``or``' instruction is:
4416 <result> = or i32 4, %var ; yields {i32}:result = 4 | %var
4417 <result> = or i32 15, 40 ; yields {i32}:result = 47
4418 <result> = or i32 4, 8 ; yields {i32}:result = 12
4420 '``xor``' Instruction
4421 ^^^^^^^^^^^^^^^^^^^^^
4428 <result> = xor <ty> <op1>, <op2> ; yields {ty}:result
4433 The '``xor``' instruction returns the bitwise logical exclusive or of
4434 its two operands. The ``xor`` is used to implement the "one's
4435 complement" operation, which is the "~" operator in C.
4440 The two arguments to the '``xor``' instruction must be
4441 :ref:`integer <t_integer>` or :ref:`vector <t_vector>` of integer values. Both
4442 arguments must have identical types.
4447 The truth table used for the '``xor``' instruction is:
4464 .. code-block:: llvm
4466 <result> = xor i32 4, %var ; yields {i32}:result = 4 ^ %var
4467 <result> = xor i32 15, 40 ; yields {i32}:result = 39
4468 <result> = xor i32 4, 8 ; yields {i32}:result = 12
4469 <result> = xor i32 %V, -1 ; yields {i32}:result = ~%V
4474 LLVM supports several instructions to represent vector operations in a
4475 target-independent manner. These instructions cover the element-access
4476 and vector-specific operations needed to process vectors effectively.
4477 While LLVM does directly support these vector operations, many
4478 sophisticated algorithms will want to use target-specific intrinsics to
4479 take full advantage of a specific target.
4481 .. _i_extractelement:
4483 '``extractelement``' Instruction
4484 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
4491 <result> = extractelement <n x <ty>> <val>, <ty2> <idx> ; yields <ty>
4496 The '``extractelement``' instruction extracts a single scalar element
4497 from a vector at a specified index.
4502 The first operand of an '``extractelement``' instruction is a value of
4503 :ref:`vector <t_vector>` type. The second operand is an index indicating
4504 the position from which to extract the element. The index may be a
4505 variable of any integer type.
4510 The result is a scalar of the same type as the element type of ``val``.
4511 Its value is the value at position ``idx`` of ``val``. If ``idx``
4512 exceeds the length of ``val``, the results are undefined.
4517 .. code-block:: llvm
4519 <result> = extractelement <4 x i32> %vec, i32 0 ; yields i32
4521 .. _i_insertelement:
4523 '``insertelement``' Instruction
4524 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
4531 <result> = insertelement <n x <ty>> <val>, <ty> <elt>, <ty2> <idx> ; yields <n x <ty>>
4536 The '``insertelement``' instruction inserts a scalar element into a
4537 vector at a specified index.
4542 The first operand of an '``insertelement``' instruction is a value of
4543 :ref:`vector <t_vector>` type. The second operand is a scalar value whose
4544 type must equal the element type of the first operand. The third operand
4545 is an index indicating the position at which to insert the value. The
4546 index may be a variable of any integer type.
4551 The result is a vector of the same type as ``val``. Its element values
4552 are those of ``val`` except at position ``idx``, where it gets the value
4553 ``elt``. If ``idx`` exceeds the length of ``val``, the results are
4559 .. code-block:: llvm
4561 <result> = insertelement <4 x i32> %vec, i32 1, i32 0 ; yields <4 x i32>
4563 .. _i_shufflevector:
4565 '``shufflevector``' Instruction
4566 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
4573 <result> = shufflevector <n x <ty>> <v1>, <n x <ty>> <v2>, <m x i32> <mask> ; yields <m x <ty>>
4578 The '``shufflevector``' instruction constructs a permutation of elements
4579 from two input vectors, returning a vector with the same element type as
4580 the input and length that is the same as the shuffle mask.
4585 The first two operands of a '``shufflevector``' instruction are vectors
4586 with the same type. The third argument is a shuffle mask whose element
4587 type is always 'i32'. The result of the instruction is a vector whose
4588 length is the same as the shuffle mask and whose element type is the
4589 same as the element type of the first two operands.
4591 The shuffle mask operand is required to be a constant vector with either
4592 constant integer or undef values.
4597 The elements of the two input vectors are numbered from left to right
4598 across both of the vectors. The shuffle mask operand specifies, for each
4599 element of the result vector, which element of the two input vectors the
4600 result element gets. The element selector may be undef (meaning "don't
4601 care") and the second operand may be undef if performing a shuffle from
4607 .. code-block:: llvm
4609 <result> = shufflevector <4 x i32> %v1, <4 x i32> %v2,
4610 <4 x i32> <i32 0, i32 4, i32 1, i32 5> ; yields <4 x i32>
4611 <result> = shufflevector <4 x i32> %v1, <4 x i32> undef,
4612 <4 x i32> <i32 0, i32 1, i32 2, i32 3> ; yields <4 x i32> - Identity shuffle.
4613 <result> = shufflevector <8 x i32> %v1, <8 x i32> undef,
4614 <4 x i32> <i32 0, i32 1, i32 2, i32 3> ; yields <4 x i32>
4615 <result> = shufflevector <4 x i32> %v1, <4 x i32> %v2,
4616 <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7 > ; yields <8 x i32>
4618 Aggregate Operations
4619 --------------------
4621 LLVM supports several instructions for working with
4622 :ref:`aggregate <t_aggregate>` values.
4626 '``extractvalue``' Instruction
4627 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
4634 <result> = extractvalue <aggregate type> <val>, <idx>{, <idx>}*
4639 The '``extractvalue``' instruction extracts the value of a member field
4640 from an :ref:`aggregate <t_aggregate>` value.
4645 The first operand of an '``extractvalue``' instruction is a value of
4646 :ref:`struct <t_struct>` or :ref:`array <t_array>` type. The operands are
4647 constant indices to specify which value to extract in a similar manner
4648 as indices in a '``getelementptr``' instruction.
4650 The major differences to ``getelementptr`` indexing are:
4652 - Since the value being indexed is not a pointer, the first index is
4653 omitted and assumed to be zero.
4654 - At least one index must be specified.
4655 - Not only struct indices but also array indices must be in bounds.
4660 The result is the value at the position in the aggregate specified by
4666 .. code-block:: llvm
4668 <result> = extractvalue {i32, float} %agg, 0 ; yields i32
4672 '``insertvalue``' Instruction
4673 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
4680 <result> = insertvalue <aggregate type> <val>, <ty> <elt>, <idx>{, <idx>}* ; yields <aggregate type>
4685 The '``insertvalue``' instruction inserts a value into a member field in
4686 an :ref:`aggregate <t_aggregate>` value.
4691 The first operand of an '``insertvalue``' instruction is a value of
4692 :ref:`struct <t_struct>` or :ref:`array <t_array>` type. The second operand is
4693 a first-class value to insert. The following operands are constant
4694 indices indicating the position at which to insert the value in a
4695 similar manner as indices in a '``extractvalue``' instruction. The value
4696 to insert must have the same type as the value identified by the
4702 The result is an aggregate of the same type as ``val``. Its value is
4703 that of ``val`` except that the value at the position specified by the
4704 indices is that of ``elt``.
4709 .. code-block:: llvm
4711 %agg1 = insertvalue {i32, float} undef, i32 1, 0 ; yields {i32 1, float undef}
4712 %agg2 = insertvalue {i32, float} %agg1, float %val, 1 ; yields {i32 1, float %val}
4713 %agg3 = insertvalue {i32, {float}} %agg1, float %val, 1, 0 ; yields {i32 1, float %val}
4717 Memory Access and Addressing Operations
4718 ---------------------------------------
4720 A key design point of an SSA-based representation is how it represents
4721 memory. In LLVM, no memory locations are in SSA form, which makes things
4722 very simple. This section describes how to read, write, and allocate
4727 '``alloca``' Instruction
4728 ^^^^^^^^^^^^^^^^^^^^^^^^
4735 <result> = alloca [inalloca] <type> [, <ty> <NumElements>] [, align <alignment>] ; yields {type*}:result
4740 The '``alloca``' instruction allocates memory on the stack frame of the
4741 currently executing function, to be automatically released when this
4742 function returns to its caller. The object is always allocated in the
4743 generic address space (address space zero).
4748 The '``alloca``' instruction allocates ``sizeof(<type>)*NumElements``
4749 bytes of memory on the runtime stack, returning a pointer of the
4750 appropriate type to the program. If "NumElements" is specified, it is
4751 the number of elements allocated, otherwise "NumElements" is defaulted
4752 to be one. If a constant alignment is specified, the value result of the
4753 allocation is guaranteed to be aligned to at least that boundary. If not
4754 specified, or if zero, the target can choose to align the allocation on
4755 any convenient boundary compatible with the type.
4757 '``type``' may be any sized type.
4762 Memory is allocated; a pointer is returned. The operation is undefined
4763 if there is insufficient stack space for the allocation. '``alloca``'d
4764 memory is automatically released when the function returns. The
4765 '``alloca``' instruction is commonly used to represent automatic
4766 variables that must have an address available. When the function returns
4767 (either with the ``ret`` or ``resume`` instructions), the memory is
4768 reclaimed. Allocating zero bytes is legal, but the result is undefined.
4769 The order in which memory is allocated (ie., which way the stack grows)
4775 .. code-block:: llvm
4777 %ptr = alloca i32 ; yields {i32*}:ptr
4778 %ptr = alloca i32, i32 4 ; yields {i32*}:ptr
4779 %ptr = alloca i32, i32 4, align 1024 ; yields {i32*}:ptr
4780 %ptr = alloca i32, align 1024 ; yields {i32*}:ptr
4784 '``load``' Instruction
4785 ^^^^^^^^^^^^^^^^^^^^^^
4792 <result> = load [volatile] <ty>* <pointer>[, align <alignment>][, !nontemporal !<index>][, !invariant.load !<index>]
4793 <result> = load atomic [volatile] <ty>* <pointer> [singlethread] <ordering>, align <alignment>
4794 !<index> = !{ i32 1 }
4799 The '``load``' instruction is used to read from memory.
4804 The argument to the ``load`` instruction specifies the memory address
4805 from which to load. The pointer must point to a :ref:`first
4806 class <t_firstclass>` type. If the ``load`` is marked as ``volatile``,
4807 then the optimizer is not allowed to modify the number or order of
4808 execution of this ``load`` with other :ref:`volatile
4809 operations <volatile>`.
4811 If the ``load`` is marked as ``atomic``, it takes an extra
4812 :ref:`ordering <ordering>` and optional ``singlethread`` argument. The
4813 ``release`` and ``acq_rel`` orderings are not valid on ``load``
4814 instructions. Atomic loads produce :ref:`defined <memmodel>` results
4815 when they may see multiple atomic stores. The type of the pointee must
4816 be an integer type whose bit width is a power of two greater than or
4817 equal to eight and less than or equal to a target-specific size limit.
4818 ``align`` must be explicitly specified on atomic loads, and the load has
4819 undefined behavior if the alignment is not set to a value which is at
4820 least the size in bytes of the pointee. ``!nontemporal`` does not have
4821 any defined semantics for atomic loads.
4823 The optional constant ``align`` argument specifies the alignment of the
4824 operation (that is, the alignment of the memory address). A value of 0
4825 or an omitted ``align`` argument means that the operation has the ABI
4826 alignment for the target. It is the responsibility of the code emitter
4827 to ensure that the alignment information is correct. Overestimating the
4828 alignment results in undefined behavior. Underestimating the alignment
4829 may produce less efficient code. An alignment of 1 is always safe.
4831 The optional ``!nontemporal`` metadata must reference a single
4832 metadata name ``<index>`` corresponding to a metadata node with one
4833 ``i32`` entry of value 1. The existence of the ``!nontemporal``
4834 metadata on the instruction tells the optimizer and code generator
4835 that this load is not expected to be reused in the cache. The code
4836 generator may select special instructions to save cache bandwidth, such
4837 as the ``MOVNT`` instruction on x86.
4839 The optional ``!invariant.load`` metadata must reference a single
4840 metadata name ``<index>`` corresponding to a metadata node with no
4841 entries. The existence of the ``!invariant.load`` metadata on the
4842 instruction tells the optimizer and code generator that this load
4843 address points to memory which does not change value during program
4844 execution. The optimizer may then move this load around, for example, by
4845 hoisting it out of loops using loop invariant code motion.
4850 The location of memory pointed to is loaded. If the value being loaded
4851 is of scalar type then the number of bytes read does not exceed the
4852 minimum number of bytes needed to hold all bits of the type. For
4853 example, loading an ``i24`` reads at most three bytes. When loading a
4854 value of a type like ``i20`` with a size that is not an integral number
4855 of bytes, the result is undefined if the value was not originally
4856 written using a store of the same type.
4861 .. code-block:: llvm
4863 %ptr = alloca i32 ; yields {i32*}:ptr
4864 store i32 3, i32* %ptr ; yields {void}
4865 %val = load i32* %ptr ; yields {i32}:val = i32 3
4869 '``store``' Instruction
4870 ^^^^^^^^^^^^^^^^^^^^^^^
4877 store [volatile] <ty> <value>, <ty>* <pointer>[, align <alignment>][, !nontemporal !<index>] ; yields {void}
4878 store atomic [volatile] <ty> <value>, <ty>* <pointer> [singlethread] <ordering>, align <alignment> ; yields {void}
4883 The '``store``' instruction is used to write to memory.
4888 There are two arguments to the ``store`` instruction: a value to store
4889 and an address at which to store it. The type of the ``<pointer>``
4890 operand must be a pointer to the :ref:`first class <t_firstclass>` type of
4891 the ``<value>`` operand. If the ``store`` is marked as ``volatile``,
4892 then the optimizer is not allowed to modify the number or order of
4893 execution of this ``store`` with other :ref:`volatile
4894 operations <volatile>`.
4896 If the ``store`` is marked as ``atomic``, it takes an extra
4897 :ref:`ordering <ordering>` and optional ``singlethread`` argument. The
4898 ``acquire`` and ``acq_rel`` orderings aren't valid on ``store``
4899 instructions. Atomic loads produce :ref:`defined <memmodel>` results
4900 when they may see multiple atomic stores. The type of the pointee must
4901 be an integer type whose bit width is a power of two greater than or
4902 equal to eight and less than or equal to a target-specific size limit.
4903 ``align`` must be explicitly specified on atomic stores, and the store
4904 has undefined behavior if the alignment is not set to a value which is
4905 at least the size in bytes of the pointee. ``!nontemporal`` does not
4906 have any defined semantics for atomic stores.
4908 The optional constant ``align`` argument specifies the alignment of the
4909 operation (that is, the alignment of the memory address). A value of 0
4910 or an omitted ``align`` argument means that the operation has the ABI
4911 alignment for the target. It is the responsibility of the code emitter
4912 to ensure that the alignment information is correct. Overestimating the
4913 alignment results in undefined behavior. Underestimating the
4914 alignment may produce less efficient code. An alignment of 1 is always
4917 The optional ``!nontemporal`` metadata must reference a single metadata
4918 name ``<index>`` corresponding to a metadata node with one ``i32`` entry of
4919 value 1. The existence of the ``!nontemporal`` metadata on the instruction
4920 tells the optimizer and code generator that this load is not expected to
4921 be reused in the cache. The code generator may select special
4922 instructions to save cache bandwidth, such as the MOVNT instruction on
4928 The contents of memory are updated to contain ``<value>`` at the
4929 location specified by the ``<pointer>`` operand. If ``<value>`` is
4930 of scalar type then the number of bytes written does not exceed the
4931 minimum number of bytes needed to hold all bits of the type. For
4932 example, storing an ``i24`` writes at most three bytes. When writing a
4933 value of a type like ``i20`` with a size that is not an integral number
4934 of bytes, it is unspecified what happens to the extra bits that do not
4935 belong to the type, but they will typically be overwritten.
4940 .. code-block:: llvm
4942 %ptr = alloca i32 ; yields {i32*}:ptr
4943 store i32 3, i32* %ptr ; yields {void}
4944 %val = load i32* %ptr ; yields {i32}:val = i32 3
4948 '``fence``' Instruction
4949 ^^^^^^^^^^^^^^^^^^^^^^^
4956 fence [singlethread] <ordering> ; yields {void}
4961 The '``fence``' instruction is used to introduce happens-before edges
4967 '``fence``' instructions take an :ref:`ordering <ordering>` argument which
4968 defines what *synchronizes-with* edges they add. They can only be given
4969 ``acquire``, ``release``, ``acq_rel``, and ``seq_cst`` orderings.
4974 A fence A which has (at least) ``release`` ordering semantics
4975 *synchronizes with* a fence B with (at least) ``acquire`` ordering
4976 semantics if and only if there exist atomic operations X and Y, both
4977 operating on some atomic object M, such that A is sequenced before X, X
4978 modifies M (either directly or through some side effect of a sequence
4979 headed by X), Y is sequenced before B, and Y observes M. This provides a
4980 *happens-before* dependency between A and B. Rather than an explicit
4981 ``fence``, one (but not both) of the atomic operations X or Y might
4982 provide a ``release`` or ``acquire`` (resp.) ordering constraint and
4983 still *synchronize-with* the explicit ``fence`` and establish the
4984 *happens-before* edge.
4986 A ``fence`` which has ``seq_cst`` ordering, in addition to having both
4987 ``acquire`` and ``release`` semantics specified above, participates in
4988 the global program order of other ``seq_cst`` operations and/or fences.
4990 The optional ":ref:`singlethread <singlethread>`" argument specifies
4991 that the fence only synchronizes with other fences in the same thread.
4992 (This is useful for interacting with signal handlers.)
4997 .. code-block:: llvm
4999 fence acquire ; yields {void}
5000 fence singlethread seq_cst ; yields {void}
5004 '``cmpxchg``' Instruction
5005 ^^^^^^^^^^^^^^^^^^^^^^^^^
5012 cmpxchg [volatile] <ty>* <pointer>, <ty> <cmp>, <ty> <new> [singlethread] <success ordering> <failure ordering> ; yields {ty}
5017 The '``cmpxchg``' instruction is used to atomically modify memory. It
5018 loads a value in memory and compares it to a given value. If they are
5019 equal, it stores a new value into the memory.
5024 There are three arguments to the '``cmpxchg``' instruction: an address
5025 to operate on, a value to compare to the value currently be at that
5026 address, and a new value to place at that address if the compared values
5027 are equal. The type of '<cmp>' must be an integer type whose bit width
5028 is a power of two greater than or equal to eight and less than or equal
5029 to a target-specific size limit. '<cmp>' and '<new>' must have the same
5030 type, and the type of '<pointer>' must be a pointer to that type. If the
5031 ``cmpxchg`` is marked as ``volatile``, then the optimizer is not allowed
5032 to modify the number or order of execution of this ``cmpxchg`` with
5033 other :ref:`volatile operations <volatile>`.
5035 The success and failure :ref:`ordering <ordering>` arguments specify how this
5036 ``cmpxchg`` synchronizes with other atomic operations. The both ordering
5037 parameters must be at least ``monotonic``, the ordering constraint on failure
5038 must be no stronger than that on success, and the failure ordering cannot be
5039 either ``release`` or ``acq_rel``.
5041 The optional "``singlethread``" argument declares that the ``cmpxchg``
5042 is only atomic with respect to code (usually signal handlers) running in
5043 the same thread as the ``cmpxchg``. Otherwise the cmpxchg is atomic with
5044 respect to all other code in the system.
5046 The pointer passed into cmpxchg must have alignment greater than or
5047 equal to the size in memory of the operand.
5052 The contents of memory at the location specified by the '``<pointer>``'
5053 operand is read and compared to '``<cmp>``'; if the read value is the
5054 equal, '``<new>``' is written. The original value at the location is
5057 A successful ``cmpxchg`` is a read-modify-write instruction for the purpose of
5058 identifying release sequences. A failed ``cmpxchg`` is equivalent to an atomic
5059 load with an ordering parameter determined the second ordering parameter.
5064 .. code-block:: llvm
5067 %orig = atomic load i32* %ptr unordered ; yields {i32}
5071 %cmp = phi i32 [ %orig, %entry ], [%old, %loop]
5072 %squared = mul i32 %cmp, %cmp
5073 %old = cmpxchg i32* %ptr, i32 %cmp, i32 %squared acq_rel monotonic ; yields {i32}
5074 %success = icmp eq i32 %cmp, %old
5075 br i1 %success, label %done, label %loop
5082 '``atomicrmw``' Instruction
5083 ^^^^^^^^^^^^^^^^^^^^^^^^^^^
5090 atomicrmw [volatile] <operation> <ty>* <pointer>, <ty> <value> [singlethread] <ordering> ; yields {ty}
5095 The '``atomicrmw``' instruction is used to atomically modify memory.
5100 There are three arguments to the '``atomicrmw``' instruction: an
5101 operation to apply, an address whose value to modify, an argument to the
5102 operation. The operation must be one of the following keywords:
5116 The type of '<value>' must be an integer type whose bit width is a power
5117 of two greater than or equal to eight and less than or equal to a
5118 target-specific size limit. The type of the '``<pointer>``' operand must
5119 be a pointer to that type. If the ``atomicrmw`` is marked as
5120 ``volatile``, then the optimizer is not allowed to modify the number or
5121 order of execution of this ``atomicrmw`` with other :ref:`volatile
5122 operations <volatile>`.
5127 The contents of memory at the location specified by the '``<pointer>``'
5128 operand are atomically read, modified, and written back. The original
5129 value at the location is returned. The modification is specified by the
5132 - xchg: ``*ptr = val``
5133 - add: ``*ptr = *ptr + val``
5134 - sub: ``*ptr = *ptr - val``
5135 - and: ``*ptr = *ptr & val``
5136 - nand: ``*ptr = ~(*ptr & val)``
5137 - or: ``*ptr = *ptr | val``
5138 - xor: ``*ptr = *ptr ^ val``
5139 - max: ``*ptr = *ptr > val ? *ptr : val`` (using a signed comparison)
5140 - min: ``*ptr = *ptr < val ? *ptr : val`` (using a signed comparison)
5141 - umax: ``*ptr = *ptr > val ? *ptr : val`` (using an unsigned
5143 - umin: ``*ptr = *ptr < val ? *ptr : val`` (using an unsigned
5149 .. code-block:: llvm
5151 %old = atomicrmw add i32* %ptr, i32 1 acquire ; yields {i32}
5153 .. _i_getelementptr:
5155 '``getelementptr``' Instruction
5156 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
5163 <result> = getelementptr <pty>* <ptrval>{, <ty> <idx>}*
5164 <result> = getelementptr inbounds <pty>* <ptrval>{, <ty> <idx>}*
5165 <result> = getelementptr <ptr vector> ptrval, <vector index type> idx
5170 The '``getelementptr``' instruction is used to get the address of a
5171 subelement of an :ref:`aggregate <t_aggregate>` data structure. It performs
5172 address calculation only and does not access memory.
5177 The first argument is always a pointer or a vector of pointers, and
5178 forms the basis of the calculation. The remaining arguments are indices
5179 that indicate which of the elements of the aggregate object are indexed.
5180 The interpretation of each index is dependent on the type being indexed
5181 into. The first index always indexes the pointer value given as the
5182 first argument, the second index indexes a value of the type pointed to
5183 (not necessarily the value directly pointed to, since the first index
5184 can be non-zero), etc. The first type indexed into must be a pointer
5185 value, subsequent types can be arrays, vectors, and structs. Note that
5186 subsequent types being indexed into can never be pointers, since that
5187 would require loading the pointer before continuing calculation.
5189 The type of each index argument depends on the type it is indexing into.
5190 When indexing into a (optionally packed) structure, only ``i32`` integer
5191 **constants** are allowed (when using a vector of indices they must all
5192 be the **same** ``i32`` integer constant). When indexing into an array,
5193 pointer or vector, integers of any width are allowed, and they are not
5194 required to be constant. These integers are treated as signed values
5197 For example, let's consider a C code fragment and how it gets compiled
5213 int *foo(struct ST *s) {
5214 return &s[1].Z.B[5][13];
5217 The LLVM code generated by Clang is:
5219 .. code-block:: llvm
5221 %struct.RT = type { i8, [10 x [20 x i32]], i8 }
5222 %struct.ST = type { i32, double, %struct.RT }
5224 define i32* @foo(%struct.ST* %s) nounwind uwtable readnone optsize ssp {
5226 %arrayidx = getelementptr inbounds %struct.ST* %s, i64 1, i32 2, i32 1, i64 5, i64 13
5233 In the example above, the first index is indexing into the
5234 '``%struct.ST*``' type, which is a pointer, yielding a '``%struct.ST``'
5235 = '``{ i32, double, %struct.RT }``' type, a structure. The second index
5236 indexes into the third element of the structure, yielding a
5237 '``%struct.RT``' = '``{ i8 , [10 x [20 x i32]], i8 }``' type, another
5238 structure. The third index indexes into the second element of the
5239 structure, yielding a '``[10 x [20 x i32]]``' type, an array. The two
5240 dimensions of the array are subscripted into, yielding an '``i32``'
5241 type. The '``getelementptr``' instruction returns a pointer to this
5242 element, thus computing a value of '``i32*``' type.
5244 Note that it is perfectly legal to index partially through a structure,
5245 returning a pointer to an inner element. Because of this, the LLVM code
5246 for the given testcase is equivalent to:
5248 .. code-block:: llvm
5250 define i32* @foo(%struct.ST* %s) {
5251 %t1 = getelementptr %struct.ST* %s, i32 1 ; yields %struct.ST*:%t1
5252 %t2 = getelementptr %struct.ST* %t1, i32 0, i32 2 ; yields %struct.RT*:%t2
5253 %t3 = getelementptr %struct.RT* %t2, i32 0, i32 1 ; yields [10 x [20 x i32]]*:%t3
5254 %t4 = getelementptr [10 x [20 x i32]]* %t3, i32 0, i32 5 ; yields [20 x i32]*:%t4
5255 %t5 = getelementptr [20 x i32]* %t4, i32 0, i32 13 ; yields i32*:%t5
5259 If the ``inbounds`` keyword is present, the result value of the
5260 ``getelementptr`` is a :ref:`poison value <poisonvalues>` if the base
5261 pointer is not an *in bounds* address of an allocated object, or if any
5262 of the addresses that would be formed by successive addition of the
5263 offsets implied by the indices to the base address with infinitely
5264 precise signed arithmetic are not an *in bounds* address of that
5265 allocated object. The *in bounds* addresses for an allocated object are
5266 all the addresses that point into the object, plus the address one byte
5267 past the end. In cases where the base is a vector of pointers the
5268 ``inbounds`` keyword applies to each of the computations element-wise.
5270 If the ``inbounds`` keyword is not present, the offsets are added to the
5271 base address with silently-wrapping two's complement arithmetic. If the
5272 offsets have a different width from the pointer, they are sign-extended
5273 or truncated to the width of the pointer. The result value of the
5274 ``getelementptr`` may be outside the object pointed to by the base
5275 pointer. The result value may not necessarily be used to access memory
5276 though, even if it happens to point into allocated storage. See the
5277 :ref:`Pointer Aliasing Rules <pointeraliasing>` section for more
5280 The getelementptr instruction is often confusing. For some more insight
5281 into how it works, see :doc:`the getelementptr FAQ <GetElementPtr>`.
5286 .. code-block:: llvm
5288 ; yields [12 x i8]*:aptr
5289 %aptr = getelementptr {i32, [12 x i8]}* %saptr, i64 0, i32 1
5291 %vptr = getelementptr {i32, <2 x i8>}* %svptr, i64 0, i32 1, i32 1
5293 %eptr = getelementptr [12 x i8]* %aptr, i64 0, i32 1
5295 %iptr = getelementptr [10 x i32]* @arr, i16 0, i16 0
5297 In cases where the pointer argument is a vector of pointers, each index
5298 must be a vector with the same number of elements. For example:
5300 .. code-block:: llvm
5302 %A = getelementptr <4 x i8*> %ptrs, <4 x i64> %offsets,
5304 Conversion Operations
5305 ---------------------
5307 The instructions in this category are the conversion instructions
5308 (casting) which all take a single operand and a type. They perform
5309 various bit conversions on the operand.
5311 '``trunc .. to``' Instruction
5312 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
5319 <result> = trunc <ty> <value> to <ty2> ; yields ty2
5324 The '``trunc``' instruction truncates its operand to the type ``ty2``.
5329 The '``trunc``' instruction takes a value to trunc, and a type to trunc
5330 it to. Both types must be of :ref:`integer <t_integer>` types, or vectors
5331 of the same number of integers. The bit size of the ``value`` must be
5332 larger than the bit size of the destination type, ``ty2``. Equal sized
5333 types are not allowed.
5338 The '``trunc``' instruction truncates the high order bits in ``value``
5339 and converts the remaining bits to ``ty2``. Since the source size must
5340 be larger than the destination size, ``trunc`` cannot be a *no-op cast*.
5341 It will always truncate bits.
5346 .. code-block:: llvm
5348 %X = trunc i32 257 to i8 ; yields i8:1
5349 %Y = trunc i32 123 to i1 ; yields i1:true
5350 %Z = trunc i32 122 to i1 ; yields i1:false
5351 %W = trunc <2 x i16> <i16 8, i16 7> to <2 x i8> ; yields <i8 8, i8 7>
5353 '``zext .. to``' Instruction
5354 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
5361 <result> = zext <ty> <value> to <ty2> ; yields ty2
5366 The '``zext``' instruction zero extends its operand to type ``ty2``.
5371 The '``zext``' instruction takes a value to cast, and a type to cast it
5372 to. Both types must be of :ref:`integer <t_integer>` types, or vectors of
5373 the same number of integers. The bit size of the ``value`` must be
5374 smaller than the bit size of the destination type, ``ty2``.
5379 The ``zext`` fills the high order bits of the ``value`` with zero bits
5380 until it reaches the size of the destination type, ``ty2``.
5382 When zero extending from i1, the result will always be either 0 or 1.
5387 .. code-block:: llvm
5389 %X = zext i32 257 to i64 ; yields i64:257
5390 %Y = zext i1 true to i32 ; yields i32:1
5391 %Z = zext <2 x i16> <i16 8, i16 7> to <2 x i32> ; yields <i32 8, i32 7>
5393 '``sext .. to``' Instruction
5394 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
5401 <result> = sext <ty> <value> to <ty2> ; yields ty2
5406 The '``sext``' sign extends ``value`` to the type ``ty2``.
5411 The '``sext``' instruction takes a value to cast, and a type to cast it
5412 to. Both types must be of :ref:`integer <t_integer>` types, or vectors of
5413 the same number of integers. The bit size of the ``value`` must be
5414 smaller than the bit size of the destination type, ``ty2``.
5419 The '``sext``' instruction performs a sign extension by copying the sign
5420 bit (highest order bit) of the ``value`` until it reaches the bit size
5421 of the type ``ty2``.
5423 When sign extending from i1, the extension always results in -1 or 0.
5428 .. code-block:: llvm
5430 %X = sext i8 -1 to i16 ; yields i16 :65535
5431 %Y = sext i1 true to i32 ; yields i32:-1
5432 %Z = sext <2 x i16> <i16 8, i16 7> to <2 x i32> ; yields <i32 8, i32 7>
5434 '``fptrunc .. to``' Instruction
5435 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
5442 <result> = fptrunc <ty> <value> to <ty2> ; yields ty2
5447 The '``fptrunc``' instruction truncates ``value`` to type ``ty2``.
5452 The '``fptrunc``' instruction takes a :ref:`floating point <t_floating>`
5453 value to cast and a :ref:`floating point <t_floating>` type to cast it to.
5454 The size of ``value`` must be larger than the size of ``ty2``. This
5455 implies that ``fptrunc`` cannot be used to make a *no-op cast*.
5460 The '``fptrunc``' instruction truncates a ``value`` from a larger
5461 :ref:`floating point <t_floating>` type to a smaller :ref:`floating
5462 point <t_floating>` type. If the value cannot fit within the
5463 destination type, ``ty2``, then the results are undefined.
5468 .. code-block:: llvm
5470 %X = fptrunc double 123.0 to float ; yields float:123.0
5471 %Y = fptrunc double 1.0E+300 to float ; yields undefined
5473 '``fpext .. to``' Instruction
5474 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
5481 <result> = fpext <ty> <value> to <ty2> ; yields ty2
5486 The '``fpext``' extends a floating point ``value`` to a larger floating
5492 The '``fpext``' instruction takes a :ref:`floating point <t_floating>`
5493 ``value`` to cast, and a :ref:`floating point <t_floating>` type to cast it
5494 to. The source type must be smaller than the destination type.
5499 The '``fpext``' instruction extends the ``value`` from a smaller
5500 :ref:`floating point <t_floating>` type to a larger :ref:`floating
5501 point <t_floating>` type. The ``fpext`` cannot be used to make a
5502 *no-op cast* because it always changes bits. Use ``bitcast`` to make a
5503 *no-op cast* for a floating point cast.
5508 .. code-block:: llvm
5510 %X = fpext float 3.125 to double ; yields double:3.125000e+00
5511 %Y = fpext double %X to fp128 ; yields fp128:0xL00000000000000004000900000000000
5513 '``fptoui .. to``' Instruction
5514 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
5521 <result> = fptoui <ty> <value> to <ty2> ; yields ty2
5526 The '``fptoui``' converts a floating point ``value`` to its unsigned
5527 integer equivalent of type ``ty2``.
5532 The '``fptoui``' instruction takes a value to cast, which must be a
5533 scalar or vector :ref:`floating point <t_floating>` value, and a type to
5534 cast it to ``ty2``, which must be an :ref:`integer <t_integer>` type. If
5535 ``ty`` is a vector floating point type, ``ty2`` must be a vector integer
5536 type with the same number of elements as ``ty``
5541 The '``fptoui``' instruction converts its :ref:`floating
5542 point <t_floating>` operand into the nearest (rounding towards zero)
5543 unsigned integer value. If the value cannot fit in ``ty2``, the results
5549 .. code-block:: llvm
5551 %X = fptoui double 123.0 to i32 ; yields i32:123
5552 %Y = fptoui float 1.0E+300 to i1 ; yields undefined:1
5553 %Z = fptoui float 1.04E+17 to i8 ; yields undefined:1
5555 '``fptosi .. to``' Instruction
5556 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
5563 <result> = fptosi <ty> <value> to <ty2> ; yields ty2
5568 The '``fptosi``' instruction converts :ref:`floating point <t_floating>`
5569 ``value`` to type ``ty2``.
5574 The '``fptosi``' instruction takes a value to cast, which must be a
5575 scalar or vector :ref:`floating point <t_floating>` value, and a type to
5576 cast it to ``ty2``, which must be an :ref:`integer <t_integer>` type. If
5577 ``ty`` is a vector floating point type, ``ty2`` must be a vector integer
5578 type with the same number of elements as ``ty``
5583 The '``fptosi``' instruction converts its :ref:`floating
5584 point <t_floating>` operand into the nearest (rounding towards zero)
5585 signed integer value. If the value cannot fit in ``ty2``, the results
5591 .. code-block:: llvm
5593 %X = fptosi double -123.0 to i32 ; yields i32:-123
5594 %Y = fptosi float 1.0E-247 to i1 ; yields undefined:1
5595 %Z = fptosi float 1.04E+17 to i8 ; yields undefined:1
5597 '``uitofp .. to``' Instruction
5598 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
5605 <result> = uitofp <ty> <value> to <ty2> ; yields ty2
5610 The '``uitofp``' instruction regards ``value`` as an unsigned integer
5611 and converts that value to the ``ty2`` type.
5616 The '``uitofp``' instruction takes a value to cast, which must be a
5617 scalar or vector :ref:`integer <t_integer>` value, and a type to cast it to
5618 ``ty2``, which must be an :ref:`floating point <t_floating>` type. If
5619 ``ty`` is a vector integer type, ``ty2`` must be a vector floating point
5620 type with the same number of elements as ``ty``
5625 The '``uitofp``' instruction interprets its operand as an unsigned
5626 integer quantity and converts it to the corresponding floating point
5627 value. If the value cannot fit in the floating point value, the results
5633 .. code-block:: llvm
5635 %X = uitofp i32 257 to float ; yields float:257.0
5636 %Y = uitofp i8 -1 to double ; yields double:255.0
5638 '``sitofp .. to``' Instruction
5639 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
5646 <result> = sitofp <ty> <value> to <ty2> ; yields ty2
5651 The '``sitofp``' instruction regards ``value`` as a signed integer and
5652 converts that value to the ``ty2`` type.
5657 The '``sitofp``' instruction takes a value to cast, which must be a
5658 scalar or vector :ref:`integer <t_integer>` value, and a type to cast it to
5659 ``ty2``, which must be an :ref:`floating point <t_floating>` type. If
5660 ``ty`` is a vector integer type, ``ty2`` must be a vector floating point
5661 type with the same number of elements as ``ty``
5666 The '``sitofp``' instruction interprets its operand as a signed integer
5667 quantity and converts it to the corresponding floating point value. If
5668 the value cannot fit in the floating point value, the results are
5674 .. code-block:: llvm
5676 %X = sitofp i32 257 to float ; yields float:257.0
5677 %Y = sitofp i8 -1 to double ; yields double:-1.0
5681 '``ptrtoint .. to``' Instruction
5682 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
5689 <result> = ptrtoint <ty> <value> to <ty2> ; yields ty2
5694 The '``ptrtoint``' instruction converts the pointer or a vector of
5695 pointers ``value`` to the integer (or vector of integers) type ``ty2``.
5700 The '``ptrtoint``' instruction takes a ``value`` to cast, which must be
5701 a a value of type :ref:`pointer <t_pointer>` or a vector of pointers, and a
5702 type to cast it to ``ty2``, which must be an :ref:`integer <t_integer>` or
5703 a vector of integers type.
5708 The '``ptrtoint``' instruction converts ``value`` to integer type
5709 ``ty2`` by interpreting the pointer value as an integer and either
5710 truncating or zero extending that value to the size of the integer type.
5711 If ``value`` is smaller than ``ty2`` then a zero extension is done. If
5712 ``value`` is larger than ``ty2`` then a truncation is done. If they are
5713 the same size, then nothing is done (*no-op cast*) other than a type
5719 .. code-block:: llvm
5721 %X = ptrtoint i32* %P to i8 ; yields truncation on 32-bit architecture
5722 %Y = ptrtoint i32* %P to i64 ; yields zero extension on 32-bit architecture
5723 %Z = ptrtoint <4 x i32*> %P to <4 x i64>; yields vector zero extension for a vector of addresses on 32-bit architecture
5727 '``inttoptr .. to``' Instruction
5728 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
5735 <result> = inttoptr <ty> <value> to <ty2> ; yields ty2
5740 The '``inttoptr``' instruction converts an integer ``value`` to a
5741 pointer type, ``ty2``.
5746 The '``inttoptr``' instruction takes an :ref:`integer <t_integer>` value to
5747 cast, and a type to cast it to, which must be a :ref:`pointer <t_pointer>`
5753 The '``inttoptr``' instruction converts ``value`` to type ``ty2`` by
5754 applying either a zero extension or a truncation depending on the size
5755 of the integer ``value``. If ``value`` is larger than the size of a
5756 pointer then a truncation is done. If ``value`` is smaller than the size
5757 of a pointer then a zero extension is done. If they are the same size,
5758 nothing is done (*no-op cast*).
5763 .. code-block:: llvm
5765 %X = inttoptr i32 255 to i32* ; yields zero extension on 64-bit architecture
5766 %Y = inttoptr i32 255 to i32* ; yields no-op on 32-bit architecture
5767 %Z = inttoptr i64 0 to i32* ; yields truncation on 32-bit architecture
5768 %Z = inttoptr <4 x i32> %G to <4 x i8*>; yields truncation of vector G to four pointers
5772 '``bitcast .. to``' Instruction
5773 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
5780 <result> = bitcast <ty> <value> to <ty2> ; yields ty2
5785 The '``bitcast``' instruction converts ``value`` to type ``ty2`` without
5791 The '``bitcast``' instruction takes a value to cast, which must be a
5792 non-aggregate first class value, and a type to cast it to, which must
5793 also be a non-aggregate :ref:`first class <t_firstclass>` type. The
5794 bit sizes of ``value`` and the destination type, ``ty2``, must be
5795 identical. If the source type is a pointer, the destination type must
5796 also be a pointer of the same size. This instruction supports bitwise
5797 conversion of vectors to integers and to vectors of other types (as
5798 long as they have the same size).
5803 The '``bitcast``' instruction converts ``value`` to type ``ty2``. It
5804 is always a *no-op cast* because no bits change with this
5805 conversion. The conversion is done as if the ``value`` had been stored
5806 to memory and read back as type ``ty2``. Pointer (or vector of
5807 pointers) types may only be converted to other pointer (or vector of
5808 pointers) types with the same address space through this instruction.
5809 To convert pointers to other types, use the :ref:`inttoptr <i_inttoptr>`
5810 or :ref:`ptrtoint <i_ptrtoint>` instructions first.
5815 .. code-block:: llvm
5817 %X = bitcast i8 255 to i8 ; yields i8 :-1
5818 %Y = bitcast i32* %x to sint* ; yields sint*:%x
5819 %Z = bitcast <2 x int> %V to i64; ; yields i64: %V
5820 %Z = bitcast <2 x i32*> %V to <2 x i64*> ; yields <2 x i64*>
5822 .. _i_addrspacecast:
5824 '``addrspacecast .. to``' Instruction
5825 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
5832 <result> = addrspacecast <pty> <ptrval> to <pty2> ; yields pty2
5837 The '``addrspacecast``' instruction converts ``ptrval`` from ``pty`` in
5838 address space ``n`` to type ``pty2`` in address space ``m``.
5843 The '``addrspacecast``' instruction takes a pointer or vector of pointer value
5844 to cast and a pointer type to cast it to, which must have a different
5850 The '``addrspacecast``' instruction converts the pointer value
5851 ``ptrval`` to type ``pty2``. It can be a *no-op cast* or a complex
5852 value modification, depending on the target and the address space
5853 pair. Pointer conversions within the same address space must be
5854 performed with the ``bitcast`` instruction. Note that if the address space
5855 conversion is legal then both result and operand refer to the same memory
5861 .. code-block:: llvm
5863 %X = addrspacecast i32* %x to i32 addrspace(1)* ; yields i32 addrspace(1)*:%x
5864 %Y = addrspacecast i32 addrspace(1)* %y to i64 addrspace(2)* ; yields i64 addrspace(2)*:%y
5865 %Z = addrspacecast <4 x i32*> %z to <4 x float addrspace(3)*> ; yields <4 x float addrspace(3)*>:%z
5872 The instructions in this category are the "miscellaneous" instructions,
5873 which defy better classification.
5877 '``icmp``' Instruction
5878 ^^^^^^^^^^^^^^^^^^^^^^
5885 <result> = icmp <cond> <ty> <op1>, <op2> ; yields {i1} or {<N x i1>}:result
5890 The '``icmp``' instruction returns a boolean value or a vector of
5891 boolean values based on comparison of its two integer, integer vector,
5892 pointer, or pointer vector operands.
5897 The '``icmp``' instruction takes three operands. The first operand is
5898 the condition code indicating the kind of comparison to perform. It is
5899 not a value, just a keyword. The possible condition code are:
5902 #. ``ne``: not equal
5903 #. ``ugt``: unsigned greater than
5904 #. ``uge``: unsigned greater or equal
5905 #. ``ult``: unsigned less than
5906 #. ``ule``: unsigned less or equal
5907 #. ``sgt``: signed greater than
5908 #. ``sge``: signed greater or equal
5909 #. ``slt``: signed less than
5910 #. ``sle``: signed less or equal
5912 The remaining two arguments must be :ref:`integer <t_integer>` or
5913 :ref:`pointer <t_pointer>` or integer :ref:`vector <t_vector>` typed. They
5914 must also be identical types.
5919 The '``icmp``' compares ``op1`` and ``op2`` according to the condition
5920 code given as ``cond``. The comparison performed always yields either an
5921 :ref:`i1 <t_integer>` or vector of ``i1`` result, as follows:
5923 #. ``eq``: yields ``true`` if the operands are equal, ``false``
5924 otherwise. No sign interpretation is necessary or performed.
5925 #. ``ne``: yields ``true`` if the operands are unequal, ``false``
5926 otherwise. No sign interpretation is necessary or performed.
5927 #. ``ugt``: interprets the operands as unsigned values and yields
5928 ``true`` if ``op1`` is greater than ``op2``.
5929 #. ``uge``: interprets the operands as unsigned values and yields
5930 ``true`` if ``op1`` is greater than or equal to ``op2``.
5931 #. ``ult``: interprets the operands as unsigned values and yields
5932 ``true`` if ``op1`` is less than ``op2``.
5933 #. ``ule``: interprets the operands as unsigned values and yields
5934 ``true`` if ``op1`` is less than or equal to ``op2``.
5935 #. ``sgt``: interprets the operands as signed values and yields ``true``
5936 if ``op1`` is greater than ``op2``.
5937 #. ``sge``: interprets the operands as signed values and yields ``true``
5938 if ``op1`` is greater than or equal to ``op2``.
5939 #. ``slt``: interprets the operands as signed values and yields ``true``
5940 if ``op1`` is less than ``op2``.
5941 #. ``sle``: interprets the operands as signed values and yields ``true``
5942 if ``op1`` is less than or equal to ``op2``.
5944 If the operands are :ref:`pointer <t_pointer>` typed, the pointer values
5945 are compared as if they were integers.
5947 If the operands are integer vectors, then they are compared element by
5948 element. The result is an ``i1`` vector with the same number of elements
5949 as the values being compared. Otherwise, the result is an ``i1``.
5954 .. code-block:: llvm
5956 <result> = icmp eq i32 4, 5 ; yields: result=false
5957 <result> = icmp ne float* %X, %X ; yields: result=false
5958 <result> = icmp ult i16 4, 5 ; yields: result=true
5959 <result> = icmp sgt i16 4, 5 ; yields: result=false
5960 <result> = icmp ule i16 -4, 5 ; yields: result=false
5961 <result> = icmp sge i16 4, 5 ; yields: result=false
5963 Note that the code generator does not yet support vector types with the
5964 ``icmp`` instruction.
5968 '``fcmp``' Instruction
5969 ^^^^^^^^^^^^^^^^^^^^^^
5976 <result> = fcmp <cond> <ty> <op1>, <op2> ; yields {i1} or {<N x i1>}:result
5981 The '``fcmp``' instruction returns a boolean value or vector of boolean
5982 values based on comparison of its operands.
5984 If the operands are floating point scalars, then the result type is a
5985 boolean (:ref:`i1 <t_integer>`).
5987 If the operands are floating point vectors, then the result type is a
5988 vector of boolean with the same number of elements as the operands being
5994 The '``fcmp``' instruction takes three operands. The first operand is
5995 the condition code indicating the kind of comparison to perform. It is
5996 not a value, just a keyword. The possible condition code are:
5998 #. ``false``: no comparison, always returns false
5999 #. ``oeq``: ordered and equal
6000 #. ``ogt``: ordered and greater than
6001 #. ``oge``: ordered and greater than or equal
6002 #. ``olt``: ordered and less than
6003 #. ``ole``: ordered and less than or equal
6004 #. ``one``: ordered and not equal
6005 #. ``ord``: ordered (no nans)
6006 #. ``ueq``: unordered or equal
6007 #. ``ugt``: unordered or greater than
6008 #. ``uge``: unordered or greater than or equal
6009 #. ``ult``: unordered or less than
6010 #. ``ule``: unordered or less than or equal
6011 #. ``une``: unordered or not equal
6012 #. ``uno``: unordered (either nans)
6013 #. ``true``: no comparison, always returns true
6015 *Ordered* means that neither operand is a QNAN while *unordered* means
6016 that either operand may be a QNAN.
6018 Each of ``val1`` and ``val2`` arguments must be either a :ref:`floating
6019 point <t_floating>` type or a :ref:`vector <t_vector>` of floating point
6020 type. They must have identical types.
6025 The '``fcmp``' instruction compares ``op1`` and ``op2`` according to the
6026 condition code given as ``cond``. If the operands are vectors, then the
6027 vectors are compared element by element. Each comparison performed
6028 always yields an :ref:`i1 <t_integer>` result, as follows:
6030 #. ``false``: always yields ``false``, regardless of operands.
6031 #. ``oeq``: yields ``true`` if both operands are not a QNAN and ``op1``
6032 is equal to ``op2``.
6033 #. ``ogt``: yields ``true`` if both operands are not a QNAN and ``op1``
6034 is greater than ``op2``.
6035 #. ``oge``: yields ``true`` if both operands are not a QNAN and ``op1``
6036 is greater than or equal to ``op2``.
6037 #. ``olt``: yields ``true`` if both operands are not a QNAN and ``op1``
6038 is less than ``op2``.
6039 #. ``ole``: yields ``true`` if both operands are not a QNAN and ``op1``
6040 is less than or equal to ``op2``.
6041 #. ``one``: yields ``true`` if both operands are not a QNAN and ``op1``
6042 is not equal to ``op2``.
6043 #. ``ord``: yields ``true`` if both operands are not a QNAN.
6044 #. ``ueq``: yields ``true`` if either operand is a QNAN or ``op1`` is
6046 #. ``ugt``: yields ``true`` if either operand is a QNAN or ``op1`` is
6047 greater than ``op2``.
6048 #. ``uge``: yields ``true`` if either operand is a QNAN or ``op1`` is
6049 greater than or equal to ``op2``.
6050 #. ``ult``: yields ``true`` if either operand is a QNAN or ``op1`` is
6052 #. ``ule``: yields ``true`` if either operand is a QNAN or ``op1`` is
6053 less than or equal to ``op2``.
6054 #. ``une``: yields ``true`` if either operand is a QNAN or ``op1`` is
6055 not equal to ``op2``.
6056 #. ``uno``: yields ``true`` if either operand is a QNAN.
6057 #. ``true``: always yields ``true``, regardless of operands.
6062 .. code-block:: llvm
6064 <result> = fcmp oeq float 4.0, 5.0 ; yields: result=false
6065 <result> = fcmp one float 4.0, 5.0 ; yields: result=true
6066 <result> = fcmp olt float 4.0, 5.0 ; yields: result=true
6067 <result> = fcmp ueq double 1.0, 2.0 ; yields: result=false
6069 Note that the code generator does not yet support vector types with the
6070 ``fcmp`` instruction.
6074 '``phi``' Instruction
6075 ^^^^^^^^^^^^^^^^^^^^^
6082 <result> = phi <ty> [ <val0>, <label0>], ...
6087 The '``phi``' instruction is used to implement the φ node in the SSA
6088 graph representing the function.
6093 The type of the incoming values is specified with the first type field.
6094 After this, the '``phi``' instruction takes a list of pairs as
6095 arguments, with one pair for each predecessor basic block of the current
6096 block. Only values of :ref:`first class <t_firstclass>` type may be used as
6097 the value arguments to the PHI node. Only labels may be used as the
6100 There must be no non-phi instructions between the start of a basic block
6101 and the PHI instructions: i.e. PHI instructions must be first in a basic
6104 For the purposes of the SSA form, the use of each incoming value is
6105 deemed to occur on the edge from the corresponding predecessor block to
6106 the current block (but after any definition of an '``invoke``'
6107 instruction's return value on the same edge).
6112 At runtime, the '``phi``' instruction logically takes on the value
6113 specified by the pair corresponding to the predecessor basic block that
6114 executed just prior to the current block.
6119 .. code-block:: llvm
6121 Loop: ; Infinite loop that counts from 0 on up...
6122 %indvar = phi i32 [ 0, %LoopHeader ], [ %nextindvar, %Loop ]
6123 %nextindvar = add i32 %indvar, 1
6128 '``select``' Instruction
6129 ^^^^^^^^^^^^^^^^^^^^^^^^
6136 <result> = select selty <cond>, <ty> <val1>, <ty> <val2> ; yields ty
6138 selty is either i1 or {<N x i1>}
6143 The '``select``' instruction is used to choose one value based on a
6144 condition, without IR-level branching.
6149 The '``select``' instruction requires an 'i1' value or a vector of 'i1'
6150 values indicating the condition, and two values of the same :ref:`first
6151 class <t_firstclass>` type. If the val1/val2 are vectors and the
6152 condition is a scalar, then entire vectors are selected, not individual
6158 If the condition is an i1 and it evaluates to 1, the instruction returns
6159 the first value argument; otherwise, it returns the second value
6162 If the condition is a vector of i1, then the value arguments must be
6163 vectors of the same size, and the selection is done element by element.
6168 .. code-block:: llvm
6170 %X = select i1 true, i8 17, i8 42 ; yields i8:17
6174 '``call``' Instruction
6175 ^^^^^^^^^^^^^^^^^^^^^^
6182 <result> = [tail | musttail] call [cconv] [ret attrs] <ty> [<fnty>*] <fnptrval>(<function args>) [fn attrs]
6187 The '``call``' instruction represents a simple function call.
6192 This instruction requires several arguments:
6194 #. The optional ``tail`` and ``musttail`` markers indicate that the optimizers
6195 should perform tail call optimization. The ``tail`` marker is a hint that
6196 `can be ignored <CodeGenerator.html#sibcallopt>`_. The ``musttail`` marker
6197 means that the call must be tail call optimized in order for the program to
6198 be correct. The ``musttail`` marker provides these guarantees:
6200 #. The call will not cause unbounded stack growth if it is part of a
6201 recursive cycle in the call graph.
6202 #. Arguments with the :ref:`inalloca <attr_inalloca>` attribute are
6205 Both markers imply that the callee does not access allocas or varargs from
6206 the caller. Calls marked ``musttail`` must obey the following additional
6209 - The call must immediately precede a :ref:`ret <i_ret>` instruction,
6210 or a pointer bitcast followed by a ret instruction.
6211 - The ret instruction must return the (possibly bitcasted) value
6212 produced by the call or void.
6213 - The caller and callee prototypes must match. Pointer types of
6214 parameters or return types may differ in pointee type, but not
6216 - The calling conventions of the caller and callee must match.
6217 - All ABI-impacting function attributes, such as sret, byval, inreg,
6218 returned, and inalloca, must match.
6220 Tail call optimization for calls marked ``tail`` is guaranteed to occur if
6221 the following conditions are met:
6223 - Caller and callee both have the calling convention ``fastcc``.
6224 - The call is in tail position (ret immediately follows call and ret
6225 uses value of call or is void).
6226 - Option ``-tailcallopt`` is enabled, or
6227 ``llvm::GuaranteedTailCallOpt`` is ``true``.
6228 - `Platform specific constraints are
6229 met. <CodeGenerator.html#tailcallopt>`_
6231 #. The optional "cconv" marker indicates which :ref:`calling
6232 convention <callingconv>` the call should use. If none is
6233 specified, the call defaults to using C calling conventions. The
6234 calling convention of the call must match the calling convention of
6235 the target function, or else the behavior is undefined.
6236 #. The optional :ref:`Parameter Attributes <paramattrs>` list for return
6237 values. Only '``zeroext``', '``signext``', and '``inreg``' attributes
6239 #. '``ty``': the type of the call instruction itself which is also the
6240 type of the return value. Functions that return no value are marked
6242 #. '``fnty``': shall be the signature of the pointer to function value
6243 being invoked. The argument types must match the types implied by
6244 this signature. This type can be omitted if the function is not
6245 varargs and if the function type does not return a pointer to a
6247 #. '``fnptrval``': An LLVM value containing a pointer to a function to
6248 be invoked. In most cases, this is a direct function invocation, but
6249 indirect ``call``'s are just as possible, calling an arbitrary pointer
6251 #. '``function args``': argument list whose types match the function
6252 signature argument types and parameter attributes. All arguments must
6253 be of :ref:`first class <t_firstclass>` type. If the function signature
6254 indicates the function accepts a variable number of arguments, the
6255 extra arguments can be specified.
6256 #. The optional :ref:`function attributes <fnattrs>` list. Only
6257 '``noreturn``', '``nounwind``', '``readonly``' and '``readnone``'
6258 attributes are valid here.
6263 The '``call``' instruction is used to cause control flow to transfer to
6264 a specified function, with its incoming arguments bound to the specified
6265 values. Upon a '``ret``' instruction in the called function, control
6266 flow continues with the instruction after the function call, and the
6267 return value of the function is bound to the result argument.
6272 .. code-block:: llvm
6274 %retval = call i32 @test(i32 %argc)
6275 call i32 (i8*, ...)* @printf(i8* %msg, i32 12, i8 42) ; yields i32
6276 %X = tail call i32 @foo() ; yields i32
6277 %Y = tail call fastcc i32 @foo() ; yields i32
6278 call void %foo(i8 97 signext)
6280 %struct.A = type { i32, i8 }
6281 %r = call %struct.A @foo() ; yields { 32, i8 }
6282 %gr = extractvalue %struct.A %r, 0 ; yields i32
6283 %gr1 = extractvalue %struct.A %r, 1 ; yields i8
6284 %Z = call void @foo() noreturn ; indicates that %foo never returns normally
6285 %ZZ = call zeroext i32 @bar() ; Return value is %zero extended
6287 llvm treats calls to some functions with names and arguments that match
6288 the standard C99 library as being the C99 library functions, and may
6289 perform optimizations or generate code for them under that assumption.
6290 This is something we'd like to change in the future to provide better
6291 support for freestanding environments and non-C-based languages.
6295 '``va_arg``' Instruction
6296 ^^^^^^^^^^^^^^^^^^^^^^^^
6303 <resultval> = va_arg <va_list*> <arglist>, <argty>
6308 The '``va_arg``' instruction is used to access arguments passed through
6309 the "variable argument" area of a function call. It is used to implement
6310 the ``va_arg`` macro in C.
6315 This instruction takes a ``va_list*`` value and the type of the
6316 argument. It returns a value of the specified argument type and
6317 increments the ``va_list`` to point to the next argument. The actual
6318 type of ``va_list`` is target specific.
6323 The '``va_arg``' instruction loads an argument of the specified type
6324 from the specified ``va_list`` and causes the ``va_list`` to point to
6325 the next argument. For more information, see the variable argument
6326 handling :ref:`Intrinsic Functions <int_varargs>`.
6328 It is legal for this instruction to be called in a function which does
6329 not take a variable number of arguments, for example, the ``vfprintf``
6332 ``va_arg`` is an LLVM instruction instead of an :ref:`intrinsic
6333 function <intrinsics>` because it takes a type as an argument.
6338 See the :ref:`variable argument processing <int_varargs>` section.
6340 Note that the code generator does not yet fully support va\_arg on many
6341 targets. Also, it does not currently support va\_arg with aggregate
6342 types on any target.
6346 '``landingpad``' Instruction
6347 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
6354 <resultval> = landingpad <resultty> personality <type> <pers_fn> <clause>+
6355 <resultval> = landingpad <resultty> personality <type> <pers_fn> cleanup <clause>*
6357 <clause> := catch <type> <value>
6358 <clause> := filter <array constant type> <array constant>
6363 The '``landingpad``' instruction is used by `LLVM's exception handling
6364 system <ExceptionHandling.html#overview>`_ to specify that a basic block
6365 is a landing pad --- one where the exception lands, and corresponds to the
6366 code found in the ``catch`` portion of a ``try``/``catch`` sequence. It
6367 defines values supplied by the personality function (``pers_fn``) upon
6368 re-entry to the function. The ``resultval`` has the type ``resultty``.
6373 This instruction takes a ``pers_fn`` value. This is the personality
6374 function associated with the unwinding mechanism. The optional
6375 ``cleanup`` flag indicates that the landing pad block is a cleanup.
6377 A ``clause`` begins with the clause type --- ``catch`` or ``filter`` --- and
6378 contains the global variable representing the "type" that may be caught
6379 or filtered respectively. Unlike the ``catch`` clause, the ``filter``
6380 clause takes an array constant as its argument. Use
6381 "``[0 x i8**] undef``" for a filter which cannot throw. The
6382 '``landingpad``' instruction must contain *at least* one ``clause`` or
6383 the ``cleanup`` flag.
6388 The '``landingpad``' instruction defines the values which are set by the
6389 personality function (``pers_fn``) upon re-entry to the function, and
6390 therefore the "result type" of the ``landingpad`` instruction. As with
6391 calling conventions, how the personality function results are
6392 represented in LLVM IR is target specific.
6394 The clauses are applied in order from top to bottom. If two
6395 ``landingpad`` instructions are merged together through inlining, the
6396 clauses from the calling function are appended to the list of clauses.
6397 When the call stack is being unwound due to an exception being thrown,
6398 the exception is compared against each ``clause`` in turn. If it doesn't
6399 match any of the clauses, and the ``cleanup`` flag is not set, then
6400 unwinding continues further up the call stack.
6402 The ``landingpad`` instruction has several restrictions:
6404 - A landing pad block is a basic block which is the unwind destination
6405 of an '``invoke``' instruction.
6406 - A landing pad block must have a '``landingpad``' instruction as its
6407 first non-PHI instruction.
6408 - There can be only one '``landingpad``' instruction within the landing
6410 - A basic block that is not a landing pad block may not include a
6411 '``landingpad``' instruction.
6412 - All '``landingpad``' instructions in a function must have the same
6413 personality function.
6418 .. code-block:: llvm
6420 ;; A landing pad which can catch an integer.
6421 %res = landingpad { i8*, i32 } personality i32 (...)* @__gxx_personality_v0
6423 ;; A landing pad that is a cleanup.
6424 %res = landingpad { i8*, i32 } personality i32 (...)* @__gxx_personality_v0
6426 ;; A landing pad which can catch an integer and can only throw a double.
6427 %res = landingpad { i8*, i32 } personality i32 (...)* @__gxx_personality_v0
6429 filter [1 x i8**] [@_ZTId]
6436 LLVM supports the notion of an "intrinsic function". These functions
6437 have well known names and semantics and are required to follow certain
6438 restrictions. Overall, these intrinsics represent an extension mechanism
6439 for the LLVM language that does not require changing all of the
6440 transformations in LLVM when adding to the language (or the bitcode
6441 reader/writer, the parser, etc...).
6443 Intrinsic function names must all start with an "``llvm.``" prefix. This
6444 prefix is reserved in LLVM for intrinsic names; thus, function names may
6445 not begin with this prefix. Intrinsic functions must always be external
6446 functions: you cannot define the body of intrinsic functions. Intrinsic
6447 functions may only be used in call or invoke instructions: it is illegal
6448 to take the address of an intrinsic function. Additionally, because
6449 intrinsic functions are part of the LLVM language, it is required if any
6450 are added that they be documented here.
6452 Some intrinsic functions can be overloaded, i.e., the intrinsic
6453 represents a family of functions that perform the same operation but on
6454 different data types. Because LLVM can represent over 8 million
6455 different integer types, overloading is used commonly to allow an
6456 intrinsic function to operate on any integer type. One or more of the
6457 argument types or the result type can be overloaded to accept any
6458 integer type. Argument types may also be defined as exactly matching a
6459 previous argument's type or the result type. This allows an intrinsic
6460 function which accepts multiple arguments, but needs all of them to be
6461 of the same type, to only be overloaded with respect to a single
6462 argument or the result.
6464 Overloaded intrinsics will have the names of its overloaded argument
6465 types encoded into its function name, each preceded by a period. Only
6466 those types which are overloaded result in a name suffix. Arguments
6467 whose type is matched against another type do not. For example, the
6468 ``llvm.ctpop`` function can take an integer of any width and returns an
6469 integer of exactly the same integer width. This leads to a family of
6470 functions such as ``i8 @llvm.ctpop.i8(i8 %val)`` and
6471 ``i29 @llvm.ctpop.i29(i29 %val)``. Only one type, the return type, is
6472 overloaded, and only one type suffix is required. Because the argument's
6473 type is matched against the return type, it does not require its own
6476 To learn how to add an intrinsic function, please see the `Extending
6477 LLVM Guide <ExtendingLLVM.html>`_.
6481 Variable Argument Handling Intrinsics
6482 -------------------------------------
6484 Variable argument support is defined in LLVM with the
6485 :ref:`va_arg <i_va_arg>` instruction and these three intrinsic
6486 functions. These functions are related to the similarly named macros
6487 defined in the ``<stdarg.h>`` header file.
6489 All of these functions operate on arguments that use a target-specific
6490 value type "``va_list``". The LLVM assembly language reference manual
6491 does not define what this type is, so all transformations should be
6492 prepared to handle these functions regardless of the type used.
6494 This example shows how the :ref:`va_arg <i_va_arg>` instruction and the
6495 variable argument handling intrinsic functions are used.
6497 .. code-block:: llvm
6499 define i32 @test(i32 %X, ...) {
6500 ; Initialize variable argument processing
6502 %ap2 = bitcast i8** %ap to i8*
6503 call void @llvm.va_start(i8* %ap2)
6505 ; Read a single integer argument
6506 %tmp = va_arg i8** %ap, i32
6508 ; Demonstrate usage of llvm.va_copy and llvm.va_end
6510 %aq2 = bitcast i8** %aq to i8*
6511 call void @llvm.va_copy(i8* %aq2, i8* %ap2)
6512 call void @llvm.va_end(i8* %aq2)
6514 ; Stop processing of arguments.
6515 call void @llvm.va_end(i8* %ap2)
6519 declare void @llvm.va_start(i8*)
6520 declare void @llvm.va_copy(i8*, i8*)
6521 declare void @llvm.va_end(i8*)
6525 '``llvm.va_start``' Intrinsic
6526 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
6533 declare void @llvm.va_start(i8* <arglist>)
6538 The '``llvm.va_start``' intrinsic initializes ``*<arglist>`` for
6539 subsequent use by ``va_arg``.
6544 The argument is a pointer to a ``va_list`` element to initialize.
6549 The '``llvm.va_start``' intrinsic works just like the ``va_start`` macro
6550 available in C. In a target-dependent way, it initializes the
6551 ``va_list`` element to which the argument points, so that the next call
6552 to ``va_arg`` will produce the first variable argument passed to the
6553 function. Unlike the C ``va_start`` macro, this intrinsic does not need
6554 to know the last argument of the function as the compiler can figure
6557 '``llvm.va_end``' Intrinsic
6558 ^^^^^^^^^^^^^^^^^^^^^^^^^^^
6565 declare void @llvm.va_end(i8* <arglist>)
6570 The '``llvm.va_end``' intrinsic destroys ``*<arglist>``, which has been
6571 initialized previously with ``llvm.va_start`` or ``llvm.va_copy``.
6576 The argument is a pointer to a ``va_list`` to destroy.
6581 The '``llvm.va_end``' intrinsic works just like the ``va_end`` macro
6582 available in C. In a target-dependent way, it destroys the ``va_list``
6583 element to which the argument points. Calls to
6584 :ref:`llvm.va_start <int_va_start>` and
6585 :ref:`llvm.va_copy <int_va_copy>` must be matched exactly with calls to
6590 '``llvm.va_copy``' Intrinsic
6591 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
6598 declare void @llvm.va_copy(i8* <destarglist>, i8* <srcarglist>)
6603 The '``llvm.va_copy``' intrinsic copies the current argument position
6604 from the source argument list to the destination argument list.
6609 The first argument is a pointer to a ``va_list`` element to initialize.
6610 The second argument is a pointer to a ``va_list`` element to copy from.
6615 The '``llvm.va_copy``' intrinsic works just like the ``va_copy`` macro
6616 available in C. In a target-dependent way, it copies the source
6617 ``va_list`` element into the destination ``va_list`` element. This
6618 intrinsic is necessary because the `` llvm.va_start`` intrinsic may be
6619 arbitrarily complex and require, for example, memory allocation.
6621 Accurate Garbage Collection Intrinsics
6622 --------------------------------------
6624 LLVM support for `Accurate Garbage Collection <GarbageCollection.html>`_
6625 (GC) requires the implementation and generation of these intrinsics.
6626 These intrinsics allow identification of :ref:`GC roots on the
6627 stack <int_gcroot>`, as well as garbage collector implementations that
6628 require :ref:`read <int_gcread>` and :ref:`write <int_gcwrite>` barriers.
6629 Front-ends for type-safe garbage collected languages should generate
6630 these intrinsics to make use of the LLVM garbage collectors. For more
6631 details, see `Accurate Garbage Collection with
6632 LLVM <GarbageCollection.html>`_.
6634 The garbage collection intrinsics only operate on objects in the generic
6635 address space (address space zero).
6639 '``llvm.gcroot``' Intrinsic
6640 ^^^^^^^^^^^^^^^^^^^^^^^^^^^
6647 declare void @llvm.gcroot(i8** %ptrloc, i8* %metadata)
6652 The '``llvm.gcroot``' intrinsic declares the existence of a GC root to
6653 the code generator, and allows some metadata to be associated with it.
6658 The first argument specifies the address of a stack object that contains
6659 the root pointer. The second pointer (which must be either a constant or
6660 a global value address) contains the meta-data to be associated with the
6666 At runtime, a call to this intrinsic stores a null pointer into the
6667 "ptrloc" location. At compile-time, the code generator generates
6668 information to allow the runtime to find the pointer at GC safe points.
6669 The '``llvm.gcroot``' intrinsic may only be used in a function which
6670 :ref:`specifies a GC algorithm <gc>`.
6674 '``llvm.gcread``' Intrinsic
6675 ^^^^^^^^^^^^^^^^^^^^^^^^^^^
6682 declare i8* @llvm.gcread(i8* %ObjPtr, i8** %Ptr)
6687 The '``llvm.gcread``' intrinsic identifies reads of references from heap
6688 locations, allowing garbage collector implementations that require read
6694 The second argument is the address to read from, which should be an
6695 address allocated from the garbage collector. The first object is a
6696 pointer to the start of the referenced object, if needed by the language
6697 runtime (otherwise null).
6702 The '``llvm.gcread``' intrinsic has the same semantics as a load
6703 instruction, but may be replaced with substantially more complex code by
6704 the garbage collector runtime, as needed. The '``llvm.gcread``'
6705 intrinsic may only be used in a function which :ref:`specifies a GC
6710 '``llvm.gcwrite``' Intrinsic
6711 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
6718 declare void @llvm.gcwrite(i8* %P1, i8* %Obj, i8** %P2)
6723 The '``llvm.gcwrite``' intrinsic identifies writes of references to heap
6724 locations, allowing garbage collector implementations that require write
6725 barriers (such as generational or reference counting collectors).
6730 The first argument is the reference to store, the second is the start of
6731 the object to store it to, and the third is the address of the field of
6732 Obj to store to. If the runtime does not require a pointer to the
6733 object, Obj may be null.
6738 The '``llvm.gcwrite``' intrinsic has the same semantics as a store
6739 instruction, but may be replaced with substantially more complex code by
6740 the garbage collector runtime, as needed. The '``llvm.gcwrite``'
6741 intrinsic may only be used in a function which :ref:`specifies a GC
6744 Code Generator Intrinsics
6745 -------------------------
6747 These intrinsics are provided by LLVM to expose special features that
6748 may only be implemented with code generator support.
6750 '``llvm.returnaddress``' Intrinsic
6751 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
6758 declare i8 *@llvm.returnaddress(i32 <level>)
6763 The '``llvm.returnaddress``' intrinsic attempts to compute a
6764 target-specific value indicating the return address of the current
6765 function or one of its callers.
6770 The argument to this intrinsic indicates which function to return the
6771 address for. Zero indicates the calling function, one indicates its
6772 caller, etc. The argument is **required** to be a constant integer
6778 The '``llvm.returnaddress``' intrinsic either returns a pointer
6779 indicating the return address of the specified call frame, or zero if it
6780 cannot be identified. The value returned by this intrinsic is likely to
6781 be incorrect or 0 for arguments other than zero, so it should only be
6782 used for debugging purposes.
6784 Note that calling this intrinsic does not prevent function inlining or
6785 other aggressive transformations, so the value returned may not be that
6786 of the obvious source-language caller.
6788 '``llvm.frameaddress``' Intrinsic
6789 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
6796 declare i8* @llvm.frameaddress(i32 <level>)
6801 The '``llvm.frameaddress``' intrinsic attempts to return the
6802 target-specific frame pointer value for the specified stack frame.
6807 The argument to this intrinsic indicates which function to return the
6808 frame pointer for. Zero indicates the calling function, one indicates
6809 its caller, etc. The argument is **required** to be a constant integer
6815 The '``llvm.frameaddress``' intrinsic either returns a pointer
6816 indicating the frame address of the specified call frame, or zero if it
6817 cannot be identified. The value returned by this intrinsic is likely to
6818 be incorrect or 0 for arguments other than zero, so it should only be
6819 used for debugging purposes.
6821 Note that calling this intrinsic does not prevent function inlining or
6822 other aggressive transformations, so the value returned may not be that
6823 of the obvious source-language caller.
6825 .. _int_read_register:
6826 .. _int_write_register:
6828 '``llvm.read_register``' and '``llvm.write_register``' Intrinsics
6829 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
6836 declare i32 @llvm.read_register.i32(metadata)
6837 declare i64 @llvm.read_register.i64(metadata)
6838 declare void @llvm.write_register.i32(metadata, i32 @value)
6839 declare void @llvm.write_register.i64(metadata, i64 @value)
6840 !0 = metadata !{metadata !"sp\00"}
6845 The '``llvm.read_register``' and '``llvm.write_register``' intrinsics
6846 provides access to the named register. The register must be valid on
6847 the architecture being compiled to. The type needs to be compatible
6848 with the register being read.
6853 The '``llvm.read_register``' intrinsic returns the current value of the
6854 register, where possible. The '``llvm.write_register``' intrinsic sets
6855 the current value of the register, where possible.
6857 This is useful to implement named register global variables that need
6858 to always be mapped to a specific register, as is common practice on
6859 bare-metal programs including OS kernels.
6861 The compiler doesn't check for register availability or use of the used
6862 register in surrounding code, including inline assembly. Because of that,
6863 allocatable registers are not supported.
6865 Warning: So far it only works with the stack pointer on selected
6866 architectures (ARM, ARM64, AArch64, PowerPC and x86_64). Significant amount of
6867 work is needed to support other registers and even more so, allocatable
6872 '``llvm.stacksave``' Intrinsic
6873 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
6880 declare i8* @llvm.stacksave()
6885 The '``llvm.stacksave``' intrinsic is used to remember the current state
6886 of the function stack, for use with
6887 :ref:`llvm.stackrestore <int_stackrestore>`. This is useful for
6888 implementing language features like scoped automatic variable sized
6894 This intrinsic returns a opaque pointer value that can be passed to
6895 :ref:`llvm.stackrestore <int_stackrestore>`. When an
6896 ``llvm.stackrestore`` intrinsic is executed with a value saved from
6897 ``llvm.stacksave``, it effectively restores the state of the stack to
6898 the state it was in when the ``llvm.stacksave`` intrinsic executed. In
6899 practice, this pops any :ref:`alloca <i_alloca>` blocks from the stack that
6900 were allocated after the ``llvm.stacksave`` was executed.
6902 .. _int_stackrestore:
6904 '``llvm.stackrestore``' Intrinsic
6905 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
6912 declare void @llvm.stackrestore(i8* %ptr)
6917 The '``llvm.stackrestore``' intrinsic is used to restore the state of
6918 the function stack to the state it was in when the corresponding
6919 :ref:`llvm.stacksave <int_stacksave>` intrinsic executed. This is
6920 useful for implementing language features like scoped automatic variable
6921 sized arrays in C99.
6926 See the description for :ref:`llvm.stacksave <int_stacksave>`.
6928 '``llvm.prefetch``' Intrinsic
6929 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
6936 declare void @llvm.prefetch(i8* <address>, i32 <rw>, i32 <locality>, i32 <cache type>)
6941 The '``llvm.prefetch``' intrinsic is a hint to the code generator to
6942 insert a prefetch instruction if supported; otherwise, it is a noop.
6943 Prefetches have no effect on the behavior of the program but can change
6944 its performance characteristics.
6949 ``address`` is the address to be prefetched, ``rw`` is the specifier
6950 determining if the fetch should be for a read (0) or write (1), and
6951 ``locality`` is a temporal locality specifier ranging from (0) - no
6952 locality, to (3) - extremely local keep in cache. The ``cache type``
6953 specifies whether the prefetch is performed on the data (1) or
6954 instruction (0) cache. The ``rw``, ``locality`` and ``cache type``
6955 arguments must be constant integers.
6960 This intrinsic does not modify the behavior of the program. In
6961 particular, prefetches cannot trap and do not produce a value. On
6962 targets that support this intrinsic, the prefetch can provide hints to
6963 the processor cache for better performance.
6965 '``llvm.pcmarker``' Intrinsic
6966 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
6973 declare void @llvm.pcmarker(i32 <id>)
6978 The '``llvm.pcmarker``' intrinsic is a method to export a Program
6979 Counter (PC) in a region of code to simulators and other tools. The
6980 method is target specific, but it is expected that the marker will use
6981 exported symbols to transmit the PC of the marker. The marker makes no
6982 guarantees that it will remain with any specific instruction after
6983 optimizations. It is possible that the presence of a marker will inhibit
6984 optimizations. The intended use is to be inserted after optimizations to
6985 allow correlations of simulation runs.
6990 ``id`` is a numerical id identifying the marker.
6995 This intrinsic does not modify the behavior of the program. Backends
6996 that do not support this intrinsic may ignore it.
6998 '``llvm.readcyclecounter``' Intrinsic
6999 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
7006 declare i64 @llvm.readcyclecounter()
7011 The '``llvm.readcyclecounter``' intrinsic provides access to the cycle
7012 counter register (or similar low latency, high accuracy clocks) on those
7013 targets that support it. On X86, it should map to RDTSC. On Alpha, it
7014 should map to RPCC. As the backing counters overflow quickly (on the
7015 order of 9 seconds on alpha), this should only be used for small
7021 When directly supported, reading the cycle counter should not modify any
7022 memory. Implementations are allowed to either return a application
7023 specific value or a system wide value. On backends without support, this
7024 is lowered to a constant 0.
7026 Note that runtime support may be conditional on the privilege-level code is
7027 running at and the host platform.
7029 '``llvm.clear_cache``' Intrinsic
7030 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
7037 declare void @llvm.clear_cache(i8*, i8*)
7042 The '``llvm.clear_cache``' intrinsic ensures visibility of modifications
7043 in the specified range to the execution unit of the processor. On
7044 targets with non-unified instruction and data cache, the implementation
7045 flushes the instruction cache.
7050 On platforms with coherent instruction and data caches (e.g. x86), this
7051 intrinsic is a nop. On platforms with non-coherent instruction and data
7052 cache (e.g. ARM, MIPS), the intrinsic is lowered either to appropriate
7053 instructions or a system call, if cache flushing requires special
7056 The default behavior is to emit a call to ``__clear_cache`` from the run
7059 This instrinsic does *not* empty the instruction pipeline. Modifications
7060 of the current function are outside the scope of the intrinsic.
7062 Standard C Library Intrinsics
7063 -----------------------------
7065 LLVM provides intrinsics for a few important standard C library
7066 functions. These intrinsics allow source-language front-ends to pass
7067 information about the alignment of the pointer arguments to the code
7068 generator, providing opportunity for more efficient code generation.
7072 '``llvm.memcpy``' Intrinsic
7073 ^^^^^^^^^^^^^^^^^^^^^^^^^^^
7078 This is an overloaded intrinsic. You can use ``llvm.memcpy`` on any
7079 integer bit width and for different address spaces. Not all targets
7080 support all bit widths however.
7084 declare void @llvm.memcpy.p0i8.p0i8.i32(i8* <dest>, i8* <src>,
7085 i32 <len>, i32 <align>, i1 <isvolatile>)
7086 declare void @llvm.memcpy.p0i8.p0i8.i64(i8* <dest>, i8* <src>,
7087 i64 <len>, i32 <align>, i1 <isvolatile>)
7092 The '``llvm.memcpy.*``' intrinsics copy a block of memory from the
7093 source location to the destination location.
7095 Note that, unlike the standard libc function, the ``llvm.memcpy.*``
7096 intrinsics do not return a value, takes extra alignment/isvolatile
7097 arguments and the pointers can be in specified address spaces.
7102 The first argument is a pointer to the destination, the second is a
7103 pointer to the source. The third argument is an integer argument
7104 specifying the number of bytes to copy, the fourth argument is the
7105 alignment of the source and destination locations, and the fifth is a
7106 boolean indicating a volatile access.
7108 If the call to this intrinsic has an alignment value that is not 0 or 1,
7109 then the caller guarantees that both the source and destination pointers
7110 are aligned to that boundary.
7112 If the ``isvolatile`` parameter is ``true``, the ``llvm.memcpy`` call is
7113 a :ref:`volatile operation <volatile>`. The detailed access behavior is not
7114 very cleanly specified and it is unwise to depend on it.
7119 The '``llvm.memcpy.*``' intrinsics copy a block of memory from the
7120 source location to the destination location, which are not allowed to
7121 overlap. It copies "len" bytes of memory over. If the argument is known
7122 to be aligned to some boundary, this can be specified as the fourth
7123 argument, otherwise it should be set to 0 or 1 (both meaning no alignment).
7125 '``llvm.memmove``' Intrinsic
7126 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
7131 This is an overloaded intrinsic. You can use llvm.memmove on any integer
7132 bit width and for different address space. Not all targets support all
7137 declare void @llvm.memmove.p0i8.p0i8.i32(i8* <dest>, i8* <src>,
7138 i32 <len>, i32 <align>, i1 <isvolatile>)
7139 declare void @llvm.memmove.p0i8.p0i8.i64(i8* <dest>, i8* <src>,
7140 i64 <len>, i32 <align>, i1 <isvolatile>)
7145 The '``llvm.memmove.*``' intrinsics move a block of memory from the
7146 source location to the destination location. It is similar to the
7147 '``llvm.memcpy``' intrinsic but allows the two memory locations to
7150 Note that, unlike the standard libc function, the ``llvm.memmove.*``
7151 intrinsics do not return a value, takes extra alignment/isvolatile
7152 arguments and the pointers can be in specified address spaces.
7157 The first argument is a pointer to the destination, the second is a
7158 pointer to the source. The third argument is an integer argument
7159 specifying the number of bytes to copy, the fourth argument is the
7160 alignment of the source and destination locations, and the fifth is a
7161 boolean indicating a volatile access.
7163 If the call to this intrinsic has an alignment value that is not 0 or 1,
7164 then the caller guarantees that the source and destination pointers are
7165 aligned to that boundary.
7167 If the ``isvolatile`` parameter is ``true``, the ``llvm.memmove`` call
7168 is a :ref:`volatile operation <volatile>`. The detailed access behavior is
7169 not very cleanly specified and it is unwise to depend on it.
7174 The '``llvm.memmove.*``' intrinsics copy a block of memory from the
7175 source location to the destination location, which may overlap. It
7176 copies "len" bytes of memory over. If the argument is known to be
7177 aligned to some boundary, this can be specified as the fourth argument,
7178 otherwise it should be set to 0 or 1 (both meaning no alignment).
7180 '``llvm.memset.*``' Intrinsics
7181 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
7186 This is an overloaded intrinsic. You can use llvm.memset on any integer
7187 bit width and for different address spaces. However, not all targets
7188 support all bit widths.
7192 declare void @llvm.memset.p0i8.i32(i8* <dest>, i8 <val>,
7193 i32 <len>, i32 <align>, i1 <isvolatile>)
7194 declare void @llvm.memset.p0i8.i64(i8* <dest>, i8 <val>,
7195 i64 <len>, i32 <align>, i1 <isvolatile>)
7200 The '``llvm.memset.*``' intrinsics fill a block of memory with a
7201 particular byte value.
7203 Note that, unlike the standard libc function, the ``llvm.memset``
7204 intrinsic does not return a value and takes extra alignment/volatile
7205 arguments. Also, the destination can be in an arbitrary address space.
7210 The first argument is a pointer to the destination to fill, the second
7211 is the byte value with which to fill it, the third argument is an
7212 integer argument specifying the number of bytes to fill, and the fourth
7213 argument is the known alignment of the destination location.
7215 If the call to this intrinsic has an alignment value that is not 0 or 1,
7216 then the caller guarantees that the destination pointer is aligned to
7219 If the ``isvolatile`` parameter is ``true``, the ``llvm.memset`` call is
7220 a :ref:`volatile operation <volatile>`. The detailed access behavior is not
7221 very cleanly specified and it is unwise to depend on it.
7226 The '``llvm.memset.*``' intrinsics fill "len" bytes of memory starting
7227 at the destination location. If the argument is known to be aligned to
7228 some boundary, this can be specified as the fourth argument, otherwise
7229 it should be set to 0 or 1 (both meaning no alignment).
7231 '``llvm.sqrt.*``' Intrinsic
7232 ^^^^^^^^^^^^^^^^^^^^^^^^^^^
7237 This is an overloaded intrinsic. You can use ``llvm.sqrt`` on any
7238 floating point or vector of floating point type. Not all targets support
7243 declare float @llvm.sqrt.f32(float %Val)
7244 declare double @llvm.sqrt.f64(double %Val)
7245 declare x86_fp80 @llvm.sqrt.f80(x86_fp80 %Val)
7246 declare fp128 @llvm.sqrt.f128(fp128 %Val)
7247 declare ppc_fp128 @llvm.sqrt.ppcf128(ppc_fp128 %Val)
7252 The '``llvm.sqrt``' intrinsics return the sqrt of the specified operand,
7253 returning the same value as the libm '``sqrt``' functions would. Unlike
7254 ``sqrt`` in libm, however, ``llvm.sqrt`` has undefined behavior for
7255 negative numbers other than -0.0 (which allows for better optimization,
7256 because there is no need to worry about errno being set).
7257 ``llvm.sqrt(-0.0)`` is defined to return -0.0 like IEEE sqrt.
7262 The argument and return value are floating point numbers of the same
7268 This function returns the sqrt of the specified operand if it is a
7269 nonnegative floating point number.
7271 '``llvm.powi.*``' Intrinsic
7272 ^^^^^^^^^^^^^^^^^^^^^^^^^^^
7277 This is an overloaded intrinsic. You can use ``llvm.powi`` on any
7278 floating point or vector of floating point type. Not all targets support
7283 declare float @llvm.powi.f32(float %Val, i32 %power)
7284 declare double @llvm.powi.f64(double %Val, i32 %power)
7285 declare x86_fp80 @llvm.powi.f80(x86_fp80 %Val, i32 %power)
7286 declare fp128 @llvm.powi.f128(fp128 %Val, i32 %power)
7287 declare ppc_fp128 @llvm.powi.ppcf128(ppc_fp128 %Val, i32 %power)
7292 The '``llvm.powi.*``' intrinsics return the first operand raised to the
7293 specified (positive or negative) power. The order of evaluation of
7294 multiplications is not defined. When a vector of floating point type is
7295 used, the second argument remains a scalar integer value.
7300 The second argument is an integer power, and the first is a value to
7301 raise to that power.
7306 This function returns the first value raised to the second power with an
7307 unspecified sequence of rounding operations.
7309 '``llvm.sin.*``' Intrinsic
7310 ^^^^^^^^^^^^^^^^^^^^^^^^^^
7315 This is an overloaded intrinsic. You can use ``llvm.sin`` on any
7316 floating point or vector of floating point type. Not all targets support
7321 declare float @llvm.sin.f32(float %Val)
7322 declare double @llvm.sin.f64(double %Val)
7323 declare x86_fp80 @llvm.sin.f80(x86_fp80 %Val)
7324 declare fp128 @llvm.sin.f128(fp128 %Val)
7325 declare ppc_fp128 @llvm.sin.ppcf128(ppc_fp128 %Val)
7330 The '``llvm.sin.*``' intrinsics return the sine of the operand.
7335 The argument and return value are floating point numbers of the same
7341 This function returns the sine of the specified operand, returning the
7342 same values as the libm ``sin`` functions would, and handles error
7343 conditions in the same way.
7345 '``llvm.cos.*``' Intrinsic
7346 ^^^^^^^^^^^^^^^^^^^^^^^^^^
7351 This is an overloaded intrinsic. You can use ``llvm.cos`` on any
7352 floating point or vector of floating point type. Not all targets support
7357 declare float @llvm.cos.f32(float %Val)
7358 declare double @llvm.cos.f64(double %Val)
7359 declare x86_fp80 @llvm.cos.f80(x86_fp80 %Val)
7360 declare fp128 @llvm.cos.f128(fp128 %Val)
7361 declare ppc_fp128 @llvm.cos.ppcf128(ppc_fp128 %Val)
7366 The '``llvm.cos.*``' intrinsics return the cosine of the operand.
7371 The argument and return value are floating point numbers of the same
7377 This function returns the cosine of the specified operand, returning the
7378 same values as the libm ``cos`` functions would, and handles error
7379 conditions in the same way.
7381 '``llvm.pow.*``' Intrinsic
7382 ^^^^^^^^^^^^^^^^^^^^^^^^^^
7387 This is an overloaded intrinsic. You can use ``llvm.pow`` on any
7388 floating point or vector of floating point type. Not all targets support
7393 declare float @llvm.pow.f32(float %Val, float %Power)
7394 declare double @llvm.pow.f64(double %Val, double %Power)
7395 declare x86_fp80 @llvm.pow.f80(x86_fp80 %Val, x86_fp80 %Power)
7396 declare fp128 @llvm.pow.f128(fp128 %Val, fp128 %Power)
7397 declare ppc_fp128 @llvm.pow.ppcf128(ppc_fp128 %Val, ppc_fp128 Power)
7402 The '``llvm.pow.*``' intrinsics return the first operand raised to the
7403 specified (positive or negative) power.
7408 The second argument is a floating point power, and the first is a value
7409 to raise to that power.
7414 This function returns the first value raised to the second power,
7415 returning the same values as the libm ``pow`` functions would, and
7416 handles error conditions in the same way.
7418 '``llvm.exp.*``' Intrinsic
7419 ^^^^^^^^^^^^^^^^^^^^^^^^^^
7424 This is an overloaded intrinsic. You can use ``llvm.exp`` on any
7425 floating point or vector of floating point type. Not all targets support
7430 declare float @llvm.exp.f32(float %Val)
7431 declare double @llvm.exp.f64(double %Val)
7432 declare x86_fp80 @llvm.exp.f80(x86_fp80 %Val)
7433 declare fp128 @llvm.exp.f128(fp128 %Val)
7434 declare ppc_fp128 @llvm.exp.ppcf128(ppc_fp128 %Val)
7439 The '``llvm.exp.*``' intrinsics perform the exp function.
7444 The argument and return value are floating point numbers of the same
7450 This function returns the same values as the libm ``exp`` functions
7451 would, and handles error conditions in the same way.
7453 '``llvm.exp2.*``' Intrinsic
7454 ^^^^^^^^^^^^^^^^^^^^^^^^^^^
7459 This is an overloaded intrinsic. You can use ``llvm.exp2`` on any
7460 floating point or vector of floating point type. Not all targets support
7465 declare float @llvm.exp2.f32(float %Val)
7466 declare double @llvm.exp2.f64(double %Val)
7467 declare x86_fp80 @llvm.exp2.f80(x86_fp80 %Val)
7468 declare fp128 @llvm.exp2.f128(fp128 %Val)
7469 declare ppc_fp128 @llvm.exp2.ppcf128(ppc_fp128 %Val)
7474 The '``llvm.exp2.*``' intrinsics perform the exp2 function.
7479 The argument and return value are floating point numbers of the same
7485 This function returns the same values as the libm ``exp2`` functions
7486 would, and handles error conditions in the same way.
7488 '``llvm.log.*``' Intrinsic
7489 ^^^^^^^^^^^^^^^^^^^^^^^^^^
7494 This is an overloaded intrinsic. You can use ``llvm.log`` on any
7495 floating point or vector of floating point type. Not all targets support
7500 declare float @llvm.log.f32(float %Val)
7501 declare double @llvm.log.f64(double %Val)
7502 declare x86_fp80 @llvm.log.f80(x86_fp80 %Val)
7503 declare fp128 @llvm.log.f128(fp128 %Val)
7504 declare ppc_fp128 @llvm.log.ppcf128(ppc_fp128 %Val)
7509 The '``llvm.log.*``' intrinsics perform the log function.
7514 The argument and return value are floating point numbers of the same
7520 This function returns the same values as the libm ``log`` functions
7521 would, and handles error conditions in the same way.
7523 '``llvm.log10.*``' Intrinsic
7524 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
7529 This is an overloaded intrinsic. You can use ``llvm.log10`` on any
7530 floating point or vector of floating point type. Not all targets support
7535 declare float @llvm.log10.f32(float %Val)
7536 declare double @llvm.log10.f64(double %Val)
7537 declare x86_fp80 @llvm.log10.f80(x86_fp80 %Val)
7538 declare fp128 @llvm.log10.f128(fp128 %Val)
7539 declare ppc_fp128 @llvm.log10.ppcf128(ppc_fp128 %Val)
7544 The '``llvm.log10.*``' intrinsics perform the log10 function.
7549 The argument and return value are floating point numbers of the same
7555 This function returns the same values as the libm ``log10`` functions
7556 would, and handles error conditions in the same way.
7558 '``llvm.log2.*``' Intrinsic
7559 ^^^^^^^^^^^^^^^^^^^^^^^^^^^
7564 This is an overloaded intrinsic. You can use ``llvm.log2`` on any
7565 floating point or vector of floating point type. Not all targets support
7570 declare float @llvm.log2.f32(float %Val)
7571 declare double @llvm.log2.f64(double %Val)
7572 declare x86_fp80 @llvm.log2.f80(x86_fp80 %Val)
7573 declare fp128 @llvm.log2.f128(fp128 %Val)
7574 declare ppc_fp128 @llvm.log2.ppcf128(ppc_fp128 %Val)
7579 The '``llvm.log2.*``' intrinsics perform the log2 function.
7584 The argument and return value are floating point numbers of the same
7590 This function returns the same values as the libm ``log2`` functions
7591 would, and handles error conditions in the same way.
7593 '``llvm.fma.*``' Intrinsic
7594 ^^^^^^^^^^^^^^^^^^^^^^^^^^
7599 This is an overloaded intrinsic. You can use ``llvm.fma`` on any
7600 floating point or vector of floating point type. Not all targets support
7605 declare float @llvm.fma.f32(float %a, float %b, float %c)
7606 declare double @llvm.fma.f64(double %a, double %b, double %c)
7607 declare x86_fp80 @llvm.fma.f80(x86_fp80 %a, x86_fp80 %b, x86_fp80 %c)
7608 declare fp128 @llvm.fma.f128(fp128 %a, fp128 %b, fp128 %c)
7609 declare ppc_fp128 @llvm.fma.ppcf128(ppc_fp128 %a, ppc_fp128 %b, ppc_fp128 %c)
7614 The '``llvm.fma.*``' intrinsics perform the fused multiply-add
7620 The argument and return value are floating point numbers of the same
7626 This function returns the same values as the libm ``fma`` functions
7627 would, and does not set errno.
7629 '``llvm.fabs.*``' Intrinsic
7630 ^^^^^^^^^^^^^^^^^^^^^^^^^^^
7635 This is an overloaded intrinsic. You can use ``llvm.fabs`` on any
7636 floating point or vector of floating point type. Not all targets support
7641 declare float @llvm.fabs.f32(float %Val)
7642 declare double @llvm.fabs.f64(double %Val)
7643 declare x86_fp80 @llvm.fabs.f80(x86_fp80 %Val)
7644 declare fp128 @llvm.fabs.f128(fp128 %Val)
7645 declare ppc_fp128 @llvm.fabs.ppcf128(ppc_fp128 %Val)
7650 The '``llvm.fabs.*``' intrinsics return the absolute value of the
7656 The argument and return value are floating point numbers of the same
7662 This function returns the same values as the libm ``fabs`` functions
7663 would, and handles error conditions in the same way.
7665 '``llvm.copysign.*``' Intrinsic
7666 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
7671 This is an overloaded intrinsic. You can use ``llvm.copysign`` on any
7672 floating point or vector of floating point type. Not all targets support
7677 declare float @llvm.copysign.f32(float %Mag, float %Sgn)
7678 declare double @llvm.copysign.f64(double %Mag, double %Sgn)
7679 declare x86_fp80 @llvm.copysign.f80(x86_fp80 %Mag, x86_fp80 %Sgn)
7680 declare fp128 @llvm.copysign.f128(fp128 %Mag, fp128 %Sgn)
7681 declare ppc_fp128 @llvm.copysign.ppcf128(ppc_fp128 %Mag, ppc_fp128 %Sgn)
7686 The '``llvm.copysign.*``' intrinsics return a value with the magnitude of the
7687 first operand and the sign of the second operand.
7692 The arguments and return value are floating point numbers of the same
7698 This function returns the same values as the libm ``copysign``
7699 functions would, and handles error conditions in the same way.
7701 '``llvm.floor.*``' Intrinsic
7702 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
7707 This is an overloaded intrinsic. You can use ``llvm.floor`` on any
7708 floating point or vector of floating point type. Not all targets support
7713 declare float @llvm.floor.f32(float %Val)
7714 declare double @llvm.floor.f64(double %Val)
7715 declare x86_fp80 @llvm.floor.f80(x86_fp80 %Val)
7716 declare fp128 @llvm.floor.f128(fp128 %Val)
7717 declare ppc_fp128 @llvm.floor.ppcf128(ppc_fp128 %Val)
7722 The '``llvm.floor.*``' intrinsics return the floor of the operand.
7727 The argument and return value are floating point numbers of the same
7733 This function returns the same values as the libm ``floor`` functions
7734 would, and handles error conditions in the same way.
7736 '``llvm.ceil.*``' Intrinsic
7737 ^^^^^^^^^^^^^^^^^^^^^^^^^^^
7742 This is an overloaded intrinsic. You can use ``llvm.ceil`` on any
7743 floating point or vector of floating point type. Not all targets support
7748 declare float @llvm.ceil.f32(float %Val)
7749 declare double @llvm.ceil.f64(double %Val)
7750 declare x86_fp80 @llvm.ceil.f80(x86_fp80 %Val)
7751 declare fp128 @llvm.ceil.f128(fp128 %Val)
7752 declare ppc_fp128 @llvm.ceil.ppcf128(ppc_fp128 %Val)
7757 The '``llvm.ceil.*``' intrinsics return the ceiling of the operand.
7762 The argument and return value are floating point numbers of the same
7768 This function returns the same values as the libm ``ceil`` functions
7769 would, and handles error conditions in the same way.
7771 '``llvm.trunc.*``' Intrinsic
7772 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
7777 This is an overloaded intrinsic. You can use ``llvm.trunc`` on any
7778 floating point or vector of floating point type. Not all targets support
7783 declare float @llvm.trunc.f32(float %Val)
7784 declare double @llvm.trunc.f64(double %Val)
7785 declare x86_fp80 @llvm.trunc.f80(x86_fp80 %Val)
7786 declare fp128 @llvm.trunc.f128(fp128 %Val)
7787 declare ppc_fp128 @llvm.trunc.ppcf128(ppc_fp128 %Val)
7792 The '``llvm.trunc.*``' intrinsics returns the operand rounded to the
7793 nearest integer not larger in magnitude than the operand.
7798 The argument and return value are floating point numbers of the same
7804 This function returns the same values as the libm ``trunc`` functions
7805 would, and handles error conditions in the same way.
7807 '``llvm.rint.*``' Intrinsic
7808 ^^^^^^^^^^^^^^^^^^^^^^^^^^^
7813 This is an overloaded intrinsic. You can use ``llvm.rint`` on any
7814 floating point or vector of floating point type. Not all targets support
7819 declare float @llvm.rint.f32(float %Val)
7820 declare double @llvm.rint.f64(double %Val)
7821 declare x86_fp80 @llvm.rint.f80(x86_fp80 %Val)
7822 declare fp128 @llvm.rint.f128(fp128 %Val)
7823 declare ppc_fp128 @llvm.rint.ppcf128(ppc_fp128 %Val)
7828 The '``llvm.rint.*``' intrinsics returns the operand rounded to the
7829 nearest integer. It may raise an inexact floating-point exception if the
7830 operand isn't an integer.
7835 The argument and return value are floating point numbers of the same
7841 This function returns the same values as the libm ``rint`` functions
7842 would, and handles error conditions in the same way.
7844 '``llvm.nearbyint.*``' Intrinsic
7845 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
7850 This is an overloaded intrinsic. You can use ``llvm.nearbyint`` on any
7851 floating point or vector of floating point type. Not all targets support
7856 declare float @llvm.nearbyint.f32(float %Val)
7857 declare double @llvm.nearbyint.f64(double %Val)
7858 declare x86_fp80 @llvm.nearbyint.f80(x86_fp80 %Val)
7859 declare fp128 @llvm.nearbyint.f128(fp128 %Val)
7860 declare ppc_fp128 @llvm.nearbyint.ppcf128(ppc_fp128 %Val)
7865 The '``llvm.nearbyint.*``' intrinsics returns the operand rounded to the
7871 The argument and return value are floating point numbers of the same
7877 This function returns the same values as the libm ``nearbyint``
7878 functions would, and handles error conditions in the same way.
7880 '``llvm.round.*``' Intrinsic
7881 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
7886 This is an overloaded intrinsic. You can use ``llvm.round`` on any
7887 floating point or vector of floating point type. Not all targets support
7892 declare float @llvm.round.f32(float %Val)
7893 declare double @llvm.round.f64(double %Val)
7894 declare x86_fp80 @llvm.round.f80(x86_fp80 %Val)
7895 declare fp128 @llvm.round.f128(fp128 %Val)
7896 declare ppc_fp128 @llvm.round.ppcf128(ppc_fp128 %Val)
7901 The '``llvm.round.*``' intrinsics returns the operand rounded to the
7907 The argument and return value are floating point numbers of the same
7913 This function returns the same values as the libm ``round``
7914 functions would, and handles error conditions in the same way.
7916 Bit Manipulation Intrinsics
7917 ---------------------------
7919 LLVM provides intrinsics for a few important bit manipulation
7920 operations. These allow efficient code generation for some algorithms.
7922 '``llvm.bswap.*``' Intrinsics
7923 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
7928 This is an overloaded intrinsic function. You can use bswap on any
7929 integer type that is an even number of bytes (i.e. BitWidth % 16 == 0).
7933 declare i16 @llvm.bswap.i16(i16 <id>)
7934 declare i32 @llvm.bswap.i32(i32 <id>)
7935 declare i64 @llvm.bswap.i64(i64 <id>)
7940 The '``llvm.bswap``' family of intrinsics is used to byte swap integer
7941 values with an even number of bytes (positive multiple of 16 bits).
7942 These are useful for performing operations on data that is not in the
7943 target's native byte order.
7948 The ``llvm.bswap.i16`` intrinsic returns an i16 value that has the high
7949 and low byte of the input i16 swapped. Similarly, the ``llvm.bswap.i32``
7950 intrinsic returns an i32 value that has the four bytes of the input i32
7951 swapped, so that if the input bytes are numbered 0, 1, 2, 3 then the
7952 returned i32 will have its bytes in 3, 2, 1, 0 order. The
7953 ``llvm.bswap.i48``, ``llvm.bswap.i64`` and other intrinsics extend this
7954 concept to additional even-byte lengths (6 bytes, 8 bytes and more,
7957 '``llvm.ctpop.*``' Intrinsic
7958 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
7963 This is an overloaded intrinsic. You can use llvm.ctpop on any integer
7964 bit width, or on any vector with integer elements. Not all targets
7965 support all bit widths or vector types, however.
7969 declare i8 @llvm.ctpop.i8(i8 <src>)
7970 declare i16 @llvm.ctpop.i16(i16 <src>)
7971 declare i32 @llvm.ctpop.i32(i32 <src>)
7972 declare i64 @llvm.ctpop.i64(i64 <src>)
7973 declare i256 @llvm.ctpop.i256(i256 <src>)
7974 declare <2 x i32> @llvm.ctpop.v2i32(<2 x i32> <src>)
7979 The '``llvm.ctpop``' family of intrinsics counts the number of bits set
7985 The only argument is the value to be counted. The argument may be of any
7986 integer type, or a vector with integer elements. The return type must
7987 match the argument type.
7992 The '``llvm.ctpop``' intrinsic counts the 1's in a variable, or within
7993 each element of a vector.
7995 '``llvm.ctlz.*``' Intrinsic
7996 ^^^^^^^^^^^^^^^^^^^^^^^^^^^
8001 This is an overloaded intrinsic. You can use ``llvm.ctlz`` on any
8002 integer bit width, or any vector whose elements are integers. Not all
8003 targets support all bit widths or vector types, however.
8007 declare i8 @llvm.ctlz.i8 (i8 <src>, i1 <is_zero_undef>)
8008 declare i16 @llvm.ctlz.i16 (i16 <src>, i1 <is_zero_undef>)
8009 declare i32 @llvm.ctlz.i32 (i32 <src>, i1 <is_zero_undef>)
8010 declare i64 @llvm.ctlz.i64 (i64 <src>, i1 <is_zero_undef>)
8011 declare i256 @llvm.ctlz.i256(i256 <src>, i1 <is_zero_undef>)
8012 declase <2 x i32> @llvm.ctlz.v2i32(<2 x i32> <src>, i1 <is_zero_undef>)
8017 The '``llvm.ctlz``' family of intrinsic functions counts the number of
8018 leading zeros in a variable.
8023 The first argument is the value to be counted. This argument may be of
8024 any integer type, or a vectory with integer element type. The return
8025 type must match the first argument type.
8027 The second argument must be a constant and is a flag to indicate whether
8028 the intrinsic should ensure that a zero as the first argument produces a
8029 defined result. Historically some architectures did not provide a
8030 defined result for zero values as efficiently, and many algorithms are
8031 now predicated on avoiding zero-value inputs.
8036 The '``llvm.ctlz``' intrinsic counts the leading (most significant)
8037 zeros in a variable, or within each element of the vector. If
8038 ``src == 0`` then the result is the size in bits of the type of ``src``
8039 if ``is_zero_undef == 0`` and ``undef`` otherwise. For example,
8040 ``llvm.ctlz(i32 2) = 30``.
8042 '``llvm.cttz.*``' Intrinsic
8043 ^^^^^^^^^^^^^^^^^^^^^^^^^^^
8048 This is an overloaded intrinsic. You can use ``llvm.cttz`` on any
8049 integer bit width, or any vector of integer elements. Not all targets
8050 support all bit widths or vector types, however.
8054 declare i8 @llvm.cttz.i8 (i8 <src>, i1 <is_zero_undef>)
8055 declare i16 @llvm.cttz.i16 (i16 <src>, i1 <is_zero_undef>)
8056 declare i32 @llvm.cttz.i32 (i32 <src>, i1 <is_zero_undef>)
8057 declare i64 @llvm.cttz.i64 (i64 <src>, i1 <is_zero_undef>)
8058 declare i256 @llvm.cttz.i256(i256 <src>, i1 <is_zero_undef>)
8059 declase <2 x i32> @llvm.cttz.v2i32(<2 x i32> <src>, i1 <is_zero_undef>)
8064 The '``llvm.cttz``' family of intrinsic functions counts the number of
8070 The first argument is the value to be counted. This argument may be of
8071 any integer type, or a vectory with integer element type. The return
8072 type must match the first argument type.
8074 The second argument must be a constant and is a flag to indicate whether
8075 the intrinsic should ensure that a zero as the first argument produces a
8076 defined result. Historically some architectures did not provide a
8077 defined result for zero values as efficiently, and many algorithms are
8078 now predicated on avoiding zero-value inputs.
8083 The '``llvm.cttz``' intrinsic counts the trailing (least significant)
8084 zeros in a variable, or within each element of a vector. If ``src == 0``
8085 then the result is the size in bits of the type of ``src`` if
8086 ``is_zero_undef == 0`` and ``undef`` otherwise. For example,
8087 ``llvm.cttz(2) = 1``.
8089 Arithmetic with Overflow Intrinsics
8090 -----------------------------------
8092 LLVM provides intrinsics for some arithmetic with overflow operations.
8094 '``llvm.sadd.with.overflow.*``' Intrinsics
8095 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
8100 This is an overloaded intrinsic. You can use ``llvm.sadd.with.overflow``
8101 on any integer bit width.
8105 declare {i16, i1} @llvm.sadd.with.overflow.i16(i16 %a, i16 %b)
8106 declare {i32, i1} @llvm.sadd.with.overflow.i32(i32 %a, i32 %b)
8107 declare {i64, i1} @llvm.sadd.with.overflow.i64(i64 %a, i64 %b)
8112 The '``llvm.sadd.with.overflow``' family of intrinsic functions perform
8113 a signed addition of the two arguments, and indicate whether an overflow
8114 occurred during the signed summation.
8119 The arguments (%a and %b) and the first element of the result structure
8120 may be of integer types of any bit width, but they must have the same
8121 bit width. The second element of the result structure must be of type
8122 ``i1``. ``%a`` and ``%b`` are the two values that will undergo signed
8128 The '``llvm.sadd.with.overflow``' family of intrinsic functions perform
8129 a signed addition of the two variables. They return a structure --- the
8130 first element of which is the signed summation, and the second element
8131 of which is a bit specifying if the signed summation resulted in an
8137 .. code-block:: llvm
8139 %res = call {i32, i1} @llvm.sadd.with.overflow.i32(i32 %a, i32 %b)
8140 %sum = extractvalue {i32, i1} %res, 0
8141 %obit = extractvalue {i32, i1} %res, 1
8142 br i1 %obit, label %overflow, label %normal
8144 '``llvm.uadd.with.overflow.*``' Intrinsics
8145 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
8150 This is an overloaded intrinsic. You can use ``llvm.uadd.with.overflow``
8151 on any integer bit width.
8155 declare {i16, i1} @llvm.uadd.with.overflow.i16(i16 %a, i16 %b)
8156 declare {i32, i1} @llvm.uadd.with.overflow.i32(i32 %a, i32 %b)
8157 declare {i64, i1} @llvm.uadd.with.overflow.i64(i64 %a, i64 %b)
8162 The '``llvm.uadd.with.overflow``' family of intrinsic functions perform
8163 an unsigned addition of the two arguments, and indicate whether a carry
8164 occurred during the unsigned summation.
8169 The arguments (%a and %b) and the first element of the result structure
8170 may be of integer types of any bit width, but they must have the same
8171 bit width. The second element of the result structure must be of type
8172 ``i1``. ``%a`` and ``%b`` are the two values that will undergo unsigned
8178 The '``llvm.uadd.with.overflow``' family of intrinsic functions perform
8179 an unsigned addition of the two arguments. They return a structure --- the
8180 first element of which is the sum, and the second element of which is a
8181 bit specifying if the unsigned summation resulted in a carry.
8186 .. code-block:: llvm
8188 %res = call {i32, i1} @llvm.uadd.with.overflow.i32(i32 %a, i32 %b)
8189 %sum = extractvalue {i32, i1} %res, 0
8190 %obit = extractvalue {i32, i1} %res, 1
8191 br i1 %obit, label %carry, label %normal
8193 '``llvm.ssub.with.overflow.*``' Intrinsics
8194 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
8199 This is an overloaded intrinsic. You can use ``llvm.ssub.with.overflow``
8200 on any integer bit width.
8204 declare {i16, i1} @llvm.ssub.with.overflow.i16(i16 %a, i16 %b)
8205 declare {i32, i1} @llvm.ssub.with.overflow.i32(i32 %a, i32 %b)
8206 declare {i64, i1} @llvm.ssub.with.overflow.i64(i64 %a, i64 %b)
8211 The '``llvm.ssub.with.overflow``' family of intrinsic functions perform
8212 a signed subtraction of the two arguments, and indicate whether an
8213 overflow occurred during the signed subtraction.
8218 The arguments (%a and %b) and the first element of the result structure
8219 may be of integer types of any bit width, but they must have the same
8220 bit width. The second element of the result structure must be of type
8221 ``i1``. ``%a`` and ``%b`` are the two values that will undergo signed
8227 The '``llvm.ssub.with.overflow``' family of intrinsic functions perform
8228 a signed subtraction of the two arguments. They return a structure --- the
8229 first element of which is the subtraction, and the second element of
8230 which is a bit specifying if the signed subtraction resulted in an
8236 .. code-block:: llvm
8238 %res = call {i32, i1} @llvm.ssub.with.overflow.i32(i32 %a, i32 %b)
8239 %sum = extractvalue {i32, i1} %res, 0
8240 %obit = extractvalue {i32, i1} %res, 1
8241 br i1 %obit, label %overflow, label %normal
8243 '``llvm.usub.with.overflow.*``' Intrinsics
8244 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
8249 This is an overloaded intrinsic. You can use ``llvm.usub.with.overflow``
8250 on any integer bit width.
8254 declare {i16, i1} @llvm.usub.with.overflow.i16(i16 %a, i16 %b)
8255 declare {i32, i1} @llvm.usub.with.overflow.i32(i32 %a, i32 %b)
8256 declare {i64, i1} @llvm.usub.with.overflow.i64(i64 %a, i64 %b)
8261 The '``llvm.usub.with.overflow``' family of intrinsic functions perform
8262 an unsigned subtraction of the two arguments, and indicate whether an
8263 overflow occurred during the unsigned subtraction.
8268 The arguments (%a and %b) and the first element of the result structure
8269 may be of integer types of any bit width, but they must have the same
8270 bit width. The second element of the result structure must be of type
8271 ``i1``. ``%a`` and ``%b`` are the two values that will undergo unsigned
8277 The '``llvm.usub.with.overflow``' family of intrinsic functions perform
8278 an unsigned subtraction of the two arguments. They return a structure ---
8279 the first element of which is the subtraction, and the second element of
8280 which is a bit specifying if the unsigned subtraction resulted in an
8286 .. code-block:: llvm
8288 %res = call {i32, i1} @llvm.usub.with.overflow.i32(i32 %a, i32 %b)
8289 %sum = extractvalue {i32, i1} %res, 0
8290 %obit = extractvalue {i32, i1} %res, 1
8291 br i1 %obit, label %overflow, label %normal
8293 '``llvm.smul.with.overflow.*``' Intrinsics
8294 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
8299 This is an overloaded intrinsic. You can use ``llvm.smul.with.overflow``
8300 on any integer bit width.
8304 declare {i16, i1} @llvm.smul.with.overflow.i16(i16 %a, i16 %b)
8305 declare {i32, i1} @llvm.smul.with.overflow.i32(i32 %a, i32 %b)
8306 declare {i64, i1} @llvm.smul.with.overflow.i64(i64 %a, i64 %b)
8311 The '``llvm.smul.with.overflow``' family of intrinsic functions perform
8312 a signed multiplication of the two arguments, and indicate whether an
8313 overflow occurred during the signed multiplication.
8318 The arguments (%a and %b) and the first element of the result structure
8319 may be of integer types of any bit width, but they must have the same
8320 bit width. The second element of the result structure must be of type
8321 ``i1``. ``%a`` and ``%b`` are the two values that will undergo signed
8327 The '``llvm.smul.with.overflow``' family of intrinsic functions perform
8328 a signed multiplication of the two arguments. They return a structure ---
8329 the first element of which is the multiplication, and the second element
8330 of which is a bit specifying if the signed multiplication resulted in an
8336 .. code-block:: llvm
8338 %res = call {i32, i1} @llvm.smul.with.overflow.i32(i32 %a, i32 %b)
8339 %sum = extractvalue {i32, i1} %res, 0
8340 %obit = extractvalue {i32, i1} %res, 1
8341 br i1 %obit, label %overflow, label %normal
8343 '``llvm.umul.with.overflow.*``' Intrinsics
8344 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
8349 This is an overloaded intrinsic. You can use ``llvm.umul.with.overflow``
8350 on any integer bit width.
8354 declare {i16, i1} @llvm.umul.with.overflow.i16(i16 %a, i16 %b)
8355 declare {i32, i1} @llvm.umul.with.overflow.i32(i32 %a, i32 %b)
8356 declare {i64, i1} @llvm.umul.with.overflow.i64(i64 %a, i64 %b)
8361 The '``llvm.umul.with.overflow``' family of intrinsic functions perform
8362 a unsigned multiplication of the two arguments, and indicate whether an
8363 overflow occurred during the unsigned multiplication.
8368 The arguments (%a and %b) and the first element of the result structure
8369 may be of integer types of any bit width, but they must have the same
8370 bit width. The second element of the result structure must be of type
8371 ``i1``. ``%a`` and ``%b`` are the two values that will undergo unsigned
8377 The '``llvm.umul.with.overflow``' family of intrinsic functions perform
8378 an unsigned multiplication of the two arguments. They return a structure ---
8379 the first element of which is the multiplication, and the second
8380 element of which is a bit specifying if the unsigned multiplication
8381 resulted in an overflow.
8386 .. code-block:: llvm
8388 %res = call {i32, i1} @llvm.umul.with.overflow.i32(i32 %a, i32 %b)
8389 %sum = extractvalue {i32, i1} %res, 0
8390 %obit = extractvalue {i32, i1} %res, 1
8391 br i1 %obit, label %overflow, label %normal
8393 Specialised Arithmetic Intrinsics
8394 ---------------------------------
8396 '``llvm.fmuladd.*``' Intrinsic
8397 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
8404 declare float @llvm.fmuladd.f32(float %a, float %b, float %c)
8405 declare double @llvm.fmuladd.f64(double %a, double %b, double %c)
8410 The '``llvm.fmuladd.*``' intrinsic functions represent multiply-add
8411 expressions that can be fused if the code generator determines that (a) the
8412 target instruction set has support for a fused operation, and (b) that the
8413 fused operation is more efficient than the equivalent, separate pair of mul
8414 and add instructions.
8419 The '``llvm.fmuladd.*``' intrinsics each take three arguments: two
8420 multiplicands, a and b, and an addend c.
8429 %0 = call float @llvm.fmuladd.f32(%a, %b, %c)
8431 is equivalent to the expression a \* b + c, except that rounding will
8432 not be performed between the multiplication and addition steps if the
8433 code generator fuses the operations. Fusion is not guaranteed, even if
8434 the target platform supports it. If a fused multiply-add is required the
8435 corresponding llvm.fma.\* intrinsic function should be used
8436 instead. This never sets errno, just as '``llvm.fma.*``'.
8441 .. code-block:: llvm
8443 %r2 = call float @llvm.fmuladd.f32(float %a, float %b, float %c) ; yields {float}:r2 = (a * b) + c
8445 Half Precision Floating Point Intrinsics
8446 ----------------------------------------
8448 For most target platforms, half precision floating point is a
8449 storage-only format. This means that it is a dense encoding (in memory)
8450 but does not support computation in the format.
8452 This means that code must first load the half-precision floating point
8453 value as an i16, then convert it to float with
8454 :ref:`llvm.convert.from.fp16 <int_convert_from_fp16>`. Computation can
8455 then be performed on the float value (including extending to double
8456 etc). To store the value back to memory, it is first converted to float
8457 if needed, then converted to i16 with
8458 :ref:`llvm.convert.to.fp16 <int_convert_to_fp16>`, then storing as an
8461 .. _int_convert_to_fp16:
8463 '``llvm.convert.to.fp16``' Intrinsic
8464 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
8471 declare i16 @llvm.convert.to.fp16(f32 %a)
8476 The '``llvm.convert.to.fp16``' intrinsic function performs a conversion
8477 from single precision floating point format to half precision floating
8483 The intrinsic function contains single argument - the value to be
8489 The '``llvm.convert.to.fp16``' intrinsic function performs a conversion
8490 from single precision floating point format to half precision floating
8491 point format. The return value is an ``i16`` which contains the
8497 .. code-block:: llvm
8499 %res = call i16 @llvm.convert.to.fp16(f32 %a)
8500 store i16 %res, i16* @x, align 2
8502 .. _int_convert_from_fp16:
8504 '``llvm.convert.from.fp16``' Intrinsic
8505 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
8512 declare f32 @llvm.convert.from.fp16(i16 %a)
8517 The '``llvm.convert.from.fp16``' intrinsic function performs a
8518 conversion from half precision floating point format to single precision
8519 floating point format.
8524 The intrinsic function contains single argument - the value to be
8530 The '``llvm.convert.from.fp16``' intrinsic function performs a
8531 conversion from half single precision floating point format to single
8532 precision floating point format. The input half-float value is
8533 represented by an ``i16`` value.
8538 .. code-block:: llvm
8540 %a = load i16* @x, align 2
8541 %res = call f32 @llvm.convert.from.fp16(i16 %a)
8546 The LLVM debugger intrinsics (which all start with ``llvm.dbg.``
8547 prefix), are described in the `LLVM Source Level
8548 Debugging <SourceLevelDebugging.html#format_common_intrinsics>`_
8551 Exception Handling Intrinsics
8552 -----------------------------
8554 The LLVM exception handling intrinsics (which all start with
8555 ``llvm.eh.`` prefix), are described in the `LLVM Exception
8556 Handling <ExceptionHandling.html#format_common_intrinsics>`_ document.
8560 Trampoline Intrinsics
8561 ---------------------
8563 These intrinsics make it possible to excise one parameter, marked with
8564 the :ref:`nest <nest>` attribute, from a function. The result is a
8565 callable function pointer lacking the nest parameter - the caller does
8566 not need to provide a value for it. Instead, the value to use is stored
8567 in advance in a "trampoline", a block of memory usually allocated on the
8568 stack, which also contains code to splice the nest value into the
8569 argument list. This is used to implement the GCC nested function address
8572 For example, if the function is ``i32 f(i8* nest %c, i32 %x, i32 %y)``
8573 then the resulting function pointer has signature ``i32 (i32, i32)*``.
8574 It can be created as follows:
8576 .. code-block:: llvm
8578 %tramp = alloca [10 x i8], align 4 ; size and alignment only correct for X86
8579 %tramp1 = getelementptr [10 x i8]* %tramp, i32 0, i32 0
8580 call i8* @llvm.init.trampoline(i8* %tramp1, i8* bitcast (i32 (i8*, i32, i32)* @f to i8*), i8* %nval)
8581 %p = call i8* @llvm.adjust.trampoline(i8* %tramp1)
8582 %fp = bitcast i8* %p to i32 (i32, i32)*
8584 The call ``%val = call i32 %fp(i32 %x, i32 %y)`` is then equivalent to
8585 ``%val = call i32 %f(i8* %nval, i32 %x, i32 %y)``.
8589 '``llvm.init.trampoline``' Intrinsic
8590 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
8597 declare void @llvm.init.trampoline(i8* <tramp>, i8* <func>, i8* <nval>)
8602 This fills the memory pointed to by ``tramp`` with executable code,
8603 turning it into a trampoline.
8608 The ``llvm.init.trampoline`` intrinsic takes three arguments, all
8609 pointers. The ``tramp`` argument must point to a sufficiently large and
8610 sufficiently aligned block of memory; this memory is written to by the
8611 intrinsic. Note that the size and the alignment are target-specific -
8612 LLVM currently provides no portable way of determining them, so a
8613 front-end that generates this intrinsic needs to have some
8614 target-specific knowledge. The ``func`` argument must hold a function
8615 bitcast to an ``i8*``.
8620 The block of memory pointed to by ``tramp`` is filled with target
8621 dependent code, turning it into a function. Then ``tramp`` needs to be
8622 passed to :ref:`llvm.adjust.trampoline <int_at>` to get a pointer which can
8623 be :ref:`bitcast (to a new function) and called <int_trampoline>`. The new
8624 function's signature is the same as that of ``func`` with any arguments
8625 marked with the ``nest`` attribute removed. At most one such ``nest``
8626 argument is allowed, and it must be of pointer type. Calling the new
8627 function is equivalent to calling ``func`` with the same argument list,
8628 but with ``nval`` used for the missing ``nest`` argument. If, after
8629 calling ``llvm.init.trampoline``, the memory pointed to by ``tramp`` is
8630 modified, then the effect of any later call to the returned function
8631 pointer is undefined.
8635 '``llvm.adjust.trampoline``' Intrinsic
8636 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
8643 declare i8* @llvm.adjust.trampoline(i8* <tramp>)
8648 This performs any required machine-specific adjustment to the address of
8649 a trampoline (passed as ``tramp``).
8654 ``tramp`` must point to a block of memory which already has trampoline
8655 code filled in by a previous call to
8656 :ref:`llvm.init.trampoline <int_it>`.
8661 On some architectures the address of the code to be executed needs to be
8662 different to the address where the trampoline is actually stored. This
8663 intrinsic returns the executable address corresponding to ``tramp``
8664 after performing the required machine specific adjustments. The pointer
8665 returned can then be :ref:`bitcast and executed <int_trampoline>`.
8670 This class of intrinsics exists to information about the lifetime of
8671 memory objects and ranges where variables are immutable.
8675 '``llvm.lifetime.start``' Intrinsic
8676 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
8683 declare void @llvm.lifetime.start(i64 <size>, i8* nocapture <ptr>)
8688 The '``llvm.lifetime.start``' intrinsic specifies the start of a memory
8694 The first argument is a constant integer representing the size of the
8695 object, or -1 if it is variable sized. The second argument is a pointer
8701 This intrinsic indicates that before this point in the code, the value
8702 of the memory pointed to by ``ptr`` is dead. This means that it is known
8703 to never be used and has an undefined value. A load from the pointer
8704 that precedes this intrinsic can be replaced with ``'undef'``.
8708 '``llvm.lifetime.end``' Intrinsic
8709 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
8716 declare void @llvm.lifetime.end(i64 <size>, i8* nocapture <ptr>)
8721 The '``llvm.lifetime.end``' intrinsic specifies the end of a memory
8727 The first argument is a constant integer representing the size of the
8728 object, or -1 if it is variable sized. The second argument is a pointer
8734 This intrinsic indicates that after this point in the code, the value of
8735 the memory pointed to by ``ptr`` is dead. This means that it is known to
8736 never be used and has an undefined value. Any stores into the memory
8737 object following this intrinsic may be removed as dead.
8739 '``llvm.invariant.start``' Intrinsic
8740 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
8747 declare {}* @llvm.invariant.start(i64 <size>, i8* nocapture <ptr>)
8752 The '``llvm.invariant.start``' intrinsic specifies that the contents of
8753 a memory object will not change.
8758 The first argument is a constant integer representing the size of the
8759 object, or -1 if it is variable sized. The second argument is a pointer
8765 This intrinsic indicates that until an ``llvm.invariant.end`` that uses
8766 the return value, the referenced memory location is constant and
8769 '``llvm.invariant.end``' Intrinsic
8770 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
8777 declare void @llvm.invariant.end({}* <start>, i64 <size>, i8* nocapture <ptr>)
8782 The '``llvm.invariant.end``' intrinsic specifies that the contents of a
8783 memory object are mutable.
8788 The first argument is the matching ``llvm.invariant.start`` intrinsic.
8789 The second argument is a constant integer representing the size of the
8790 object, or -1 if it is variable sized and the third argument is a
8791 pointer to the object.
8796 This intrinsic indicates that the memory is mutable again.
8801 This class of intrinsics is designed to be generic and has no specific
8804 '``llvm.var.annotation``' Intrinsic
8805 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
8812 declare void @llvm.var.annotation(i8* <val>, i8* <str>, i8* <str>, i32 <int>)
8817 The '``llvm.var.annotation``' intrinsic.
8822 The first argument is a pointer to a value, the second is a pointer to a
8823 global string, the third is a pointer to a global string which is the
8824 source file name, and the last argument is the line number.
8829 This intrinsic allows annotation of local variables with arbitrary
8830 strings. This can be useful for special purpose optimizations that want
8831 to look for these annotations. These have no other defined use; they are
8832 ignored by code generation and optimization.
8834 '``llvm.ptr.annotation.*``' Intrinsic
8835 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
8840 This is an overloaded intrinsic. You can use '``llvm.ptr.annotation``' on a
8841 pointer to an integer of any width. *NOTE* you must specify an address space for
8842 the pointer. The identifier for the default address space is the integer
8847 declare i8* @llvm.ptr.annotation.p<address space>i8(i8* <val>, i8* <str>, i8* <str>, i32 <int>)
8848 declare i16* @llvm.ptr.annotation.p<address space>i16(i16* <val>, i8* <str>, i8* <str>, i32 <int>)
8849 declare i32* @llvm.ptr.annotation.p<address space>i32(i32* <val>, i8* <str>, i8* <str>, i32 <int>)
8850 declare i64* @llvm.ptr.annotation.p<address space>i64(i64* <val>, i8* <str>, i8* <str>, i32 <int>)
8851 declare i256* @llvm.ptr.annotation.p<address space>i256(i256* <val>, i8* <str>, i8* <str>, i32 <int>)
8856 The '``llvm.ptr.annotation``' intrinsic.
8861 The first argument is a pointer to an integer value of arbitrary bitwidth
8862 (result of some expression), the second is a pointer to a global string, the
8863 third is a pointer to a global string which is the source file name, and the
8864 last argument is the line number. It returns the value of the first argument.
8869 This intrinsic allows annotation of a pointer to an integer with arbitrary
8870 strings. This can be useful for special purpose optimizations that want to look
8871 for these annotations. These have no other defined use; they are ignored by code
8872 generation and optimization.
8874 '``llvm.annotation.*``' Intrinsic
8875 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
8880 This is an overloaded intrinsic. You can use '``llvm.annotation``' on
8881 any integer bit width.
8885 declare i8 @llvm.annotation.i8(i8 <val>, i8* <str>, i8* <str>, i32 <int>)
8886 declare i16 @llvm.annotation.i16(i16 <val>, i8* <str>, i8* <str>, i32 <int>)
8887 declare i32 @llvm.annotation.i32(i32 <val>, i8* <str>, i8* <str>, i32 <int>)
8888 declare i64 @llvm.annotation.i64(i64 <val>, i8* <str>, i8* <str>, i32 <int>)
8889 declare i256 @llvm.annotation.i256(i256 <val>, i8* <str>, i8* <str>, i32 <int>)
8894 The '``llvm.annotation``' intrinsic.
8899 The first argument is an integer value (result of some expression), the
8900 second is a pointer to a global string, the third is a pointer to a
8901 global string which is the source file name, and the last argument is
8902 the line number. It returns the value of the first argument.
8907 This intrinsic allows annotations to be put on arbitrary expressions
8908 with arbitrary strings. This can be useful for special purpose
8909 optimizations that want to look for these annotations. These have no
8910 other defined use; they are ignored by code generation and optimization.
8912 '``llvm.trap``' Intrinsic
8913 ^^^^^^^^^^^^^^^^^^^^^^^^^
8920 declare void @llvm.trap() noreturn nounwind
8925 The '``llvm.trap``' intrinsic.
8935 This intrinsic is lowered to the target dependent trap instruction. If
8936 the target does not have a trap instruction, this intrinsic will be
8937 lowered to a call of the ``abort()`` function.
8939 '``llvm.debugtrap``' Intrinsic
8940 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
8947 declare void @llvm.debugtrap() nounwind
8952 The '``llvm.debugtrap``' intrinsic.
8962 This intrinsic is lowered to code which is intended to cause an
8963 execution trap with the intention of requesting the attention of a
8966 '``llvm.stackprotector``' Intrinsic
8967 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
8974 declare void @llvm.stackprotector(i8* <guard>, i8** <slot>)
8979 The ``llvm.stackprotector`` intrinsic takes the ``guard`` and stores it
8980 onto the stack at ``slot``. The stack slot is adjusted to ensure that it
8981 is placed on the stack before local variables.
8986 The ``llvm.stackprotector`` intrinsic requires two pointer arguments.
8987 The first argument is the value loaded from the stack guard
8988 ``@__stack_chk_guard``. The second variable is an ``alloca`` that has
8989 enough space to hold the value of the guard.
8994 This intrinsic causes the prologue/epilogue inserter to force the position of
8995 the ``AllocaInst`` stack slot to be before local variables on the stack. This is
8996 to ensure that if a local variable on the stack is overwritten, it will destroy
8997 the value of the guard. When the function exits, the guard on the stack is
8998 checked against the original guard by ``llvm.stackprotectorcheck``. If they are
8999 different, then ``llvm.stackprotectorcheck`` causes the program to abort by
9000 calling the ``__stack_chk_fail()`` function.
9002 '``llvm.stackprotectorcheck``' Intrinsic
9003 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
9010 declare void @llvm.stackprotectorcheck(i8** <guard>)
9015 The ``llvm.stackprotectorcheck`` intrinsic compares ``guard`` against an already
9016 created stack protector and if they are not equal calls the
9017 ``__stack_chk_fail()`` function.
9022 The ``llvm.stackprotectorcheck`` intrinsic requires one pointer argument, the
9023 the variable ``@__stack_chk_guard``.
9028 This intrinsic is provided to perform the stack protector check by comparing
9029 ``guard`` with the stack slot created by ``llvm.stackprotector`` and if the
9030 values do not match call the ``__stack_chk_fail()`` function.
9032 The reason to provide this as an IR level intrinsic instead of implementing it
9033 via other IR operations is that in order to perform this operation at the IR
9034 level without an intrinsic, one would need to create additional basic blocks to
9035 handle the success/failure cases. This makes it difficult to stop the stack
9036 protector check from disrupting sibling tail calls in Codegen. With this
9037 intrinsic, we are able to generate the stack protector basic blocks late in
9038 codegen after the tail call decision has occurred.
9040 '``llvm.objectsize``' Intrinsic
9041 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
9048 declare i32 @llvm.objectsize.i32(i8* <object>, i1 <min>)
9049 declare i64 @llvm.objectsize.i64(i8* <object>, i1 <min>)
9054 The ``llvm.objectsize`` intrinsic is designed to provide information to
9055 the optimizers to determine at compile time whether a) an operation
9056 (like memcpy) will overflow a buffer that corresponds to an object, or
9057 b) that a runtime check for overflow isn't necessary. An object in this
9058 context means an allocation of a specific class, structure, array, or
9064 The ``llvm.objectsize`` intrinsic takes two arguments. The first
9065 argument is a pointer to or into the ``object``. The second argument is
9066 a boolean and determines whether ``llvm.objectsize`` returns 0 (if true)
9067 or -1 (if false) when the object size is unknown. The second argument
9068 only accepts constants.
9073 The ``llvm.objectsize`` intrinsic is lowered to a constant representing
9074 the size of the object concerned. If the size cannot be determined at
9075 compile time, ``llvm.objectsize`` returns ``i32/i64 -1 or 0`` (depending
9076 on the ``min`` argument).
9078 '``llvm.expect``' Intrinsic
9079 ^^^^^^^^^^^^^^^^^^^^^^^^^^^
9084 This is an overloaded intrinsic. You can use ``llvm.expect`` on any
9089 declare i1 @llvm.expect.i1(i1 <val>, i1 <expected_val>)
9090 declare i32 @llvm.expect.i32(i32 <val>, i32 <expected_val>)
9091 declare i64 @llvm.expect.i64(i64 <val>, i64 <expected_val>)
9096 The ``llvm.expect`` intrinsic provides information about expected (the
9097 most probable) value of ``val``, which can be used by optimizers.
9102 The ``llvm.expect`` intrinsic takes two arguments. The first argument is
9103 a value. The second argument is an expected value, this needs to be a
9104 constant value, variables are not allowed.
9109 This intrinsic is lowered to the ``val``.
9111 '``llvm.donothing``' Intrinsic
9112 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
9119 declare void @llvm.donothing() nounwind readnone
9124 The ``llvm.donothing`` intrinsic doesn't perform any operation. It's the
9125 only intrinsic that can be called with an invoke instruction.
9135 This intrinsic does nothing, and it's removed by optimizers and ignored
9138 Stack Map Intrinsics
9139 --------------------
9141 LLVM provides experimental intrinsics to support runtime patching
9142 mechanisms commonly desired in dynamic language JITs. These intrinsics
9143 are described in :doc:`StackMaps`.