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 that 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. The ``"\01"`` prefix
83 can be used on global variables to suppress mangling.
84 #. Unnamed values are represented as an unsigned numeric value with
85 their prefix. For example, ``%12``, ``@2``, ``%44``.
86 #. Constants, which are described in the section Constants_ below.
88 LLVM requires that values start with a prefix for two reasons: Compilers
89 don't need to worry about name clashes with reserved words, and the set
90 of reserved words may be expanded in the future without penalty.
91 Additionally, unnamed identifiers allow a compiler to quickly come up
92 with a temporary variable without having to avoid symbol table
95 Reserved words in LLVM are very similar to reserved words in other
96 languages. There are keywords for different opcodes ('``add``',
97 '``bitcast``', '``ret``', etc...), for primitive type names ('``void``',
98 '``i32``', etc...), and others. These reserved words cannot conflict
99 with variable names, because none of them start with a prefix character
100 (``'%'`` or ``'@'``).
102 Here is an example of LLVM code to multiply the integer variable
109 %result = mul i32 %X, 8
111 After strength reduction:
115 %result = shl i32 %X, 3
121 %0 = add i32 %X, %X ; yields i32:%0
122 %1 = add i32 %0, %0 ; yields i32:%1
123 %result = add i32 %1, %1
125 This last way of multiplying ``%X`` by 8 illustrates several important
126 lexical features of LLVM:
128 #. Comments are delimited with a '``;``' and go until the end of line.
129 #. Unnamed temporaries are created when the result of a computation is
130 not assigned to a named value.
131 #. Unnamed temporaries are numbered sequentially (using a per-function
132 incrementing counter, starting with 0). Note that basic blocks and unnamed
133 function parameters are included in this numbering. For example, if the
134 entry basic block is not given a label name and all function parameters are
135 named, then it will get number 0.
137 It also shows a convention that we follow in this document. When
138 demonstrating instructions, we will follow an instruction with a comment
139 that defines the type and name of value produced.
147 LLVM programs are composed of ``Module``'s, each of which is a
148 translation unit of the input programs. Each module consists of
149 functions, global variables, and symbol table entries. Modules may be
150 combined together with the LLVM linker, which merges function (and
151 global variable) definitions, resolves forward declarations, and merges
152 symbol table entries. Here is an example of the "hello world" module:
156 ; Declare the string constant as a global constant.
157 @.str = private unnamed_addr constant [13 x i8] c"hello world\0A\00"
159 ; External declaration of the puts function
160 declare i32 @puts(i8* nocapture) nounwind
162 ; Definition of main function
163 define i32 @main() { ; i32()*
164 ; Convert [13 x i8]* to i8 *...
165 %cast210 = getelementptr [13 x i8], [13 x i8]* @.str, i64 0, i64 0
167 ; Call puts function to write out the string to stdout.
168 call i32 @puts(i8* %cast210)
173 !0 = !{i32 42, null, !"string"}
176 This example is made up of a :ref:`global variable <globalvars>` named
177 "``.str``", an external declaration of the "``puts``" function, a
178 :ref:`function definition <functionstructure>` for "``main``" and
179 :ref:`named metadata <namedmetadatastructure>` "``foo``".
181 In general, a module is made up of a list of global values (where both
182 functions and global variables are global values). Global values are
183 represented by a pointer to a memory location (in this case, a pointer
184 to an array of char, and a pointer to a function), and have one of the
185 following :ref:`linkage types <linkage>`.
192 All Global Variables and Functions have one of the following types of
196 Global values with "``private``" linkage are only directly
197 accessible by objects in the current module. In particular, linking
198 code into a module with an private global value may cause the
199 private to be renamed as necessary to avoid collisions. Because the
200 symbol is private to the module, all references can be updated. This
201 doesn't show up in any symbol table in the object file.
203 Similar to private, but the value shows as a local symbol
204 (``STB_LOCAL`` in the case of ELF) in the object file. This
205 corresponds to the notion of the '``static``' keyword in C.
206 ``available_externally``
207 Globals with "``available_externally``" linkage are never emitted
208 into the object file corresponding to the LLVM module. They exist to
209 allow inlining and other optimizations to take place given knowledge
210 of the definition of the global, which is known to be somewhere
211 outside the module. Globals with ``available_externally`` linkage
212 are allowed to be discarded at will, and are otherwise the same as
213 ``linkonce_odr``. This linkage type is only allowed on definitions,
216 Globals with "``linkonce``" linkage are merged with other globals of
217 the same name when linkage occurs. This can be used to implement
218 some forms of inline functions, templates, or other code which must
219 be generated in each translation unit that uses it, but where the
220 body may be overridden with a more definitive definition later.
221 Unreferenced ``linkonce`` globals are allowed to be discarded. Note
222 that ``linkonce`` linkage does not actually allow the optimizer to
223 inline the body of this function into callers because it doesn't
224 know if this definition of the function is the definitive definition
225 within the program or whether it will be overridden by a stronger
226 definition. To enable inlining and other optimizations, use
227 "``linkonce_odr``" linkage.
229 "``weak``" linkage has the same merging semantics as ``linkonce``
230 linkage, except that unreferenced globals with ``weak`` linkage may
231 not be discarded. This is used for globals that are declared "weak"
234 "``common``" linkage is most similar to "``weak``" linkage, but they
235 are used for tentative definitions in C, such as "``int X;``" at
236 global scope. Symbols with "``common``" linkage are merged in the
237 same way as ``weak symbols``, and they may not be deleted if
238 unreferenced. ``common`` symbols may not have an explicit section,
239 must have a zero initializer, and may not be marked
240 ':ref:`constant <globalvars>`'. Functions and aliases may not have
243 .. _linkage_appending:
246 "``appending``" linkage may only be applied to global variables of
247 pointer to array type. When two global variables with appending
248 linkage are linked together, the two global arrays are appended
249 together. This is the LLVM, typesafe, equivalent of having the
250 system linker append together "sections" with identical names when
253 The semantics of this linkage follow the ELF object file model: the
254 symbol is weak until linked, if not linked, the symbol becomes null
255 instead of being an undefined reference.
256 ``linkonce_odr``, ``weak_odr``
257 Some languages allow differing globals to be merged, such as two
258 functions with different semantics. Other languages, such as
259 ``C++``, ensure that only equivalent globals are ever merged (the
260 "one definition rule" --- "ODR"). Such languages can use the
261 ``linkonce_odr`` and ``weak_odr`` linkage types to indicate that the
262 global will only be merged with equivalent globals. These linkage
263 types are otherwise the same as their non-``odr`` versions.
265 If none of the above identifiers are used, the global is externally
266 visible, meaning that it participates in linkage and can be used to
267 resolve external symbol references.
269 It is illegal for a function *declaration* to have any linkage type
270 other than ``external`` or ``extern_weak``.
277 LLVM :ref:`functions <functionstructure>`, :ref:`calls <i_call>` and
278 :ref:`invokes <i_invoke>` can all have an optional calling convention
279 specified for the call. The calling convention of any pair of dynamic
280 caller/callee must match, or the behavior of the program is undefined.
281 The following calling conventions are supported by LLVM, and more may be
284 "``ccc``" - The C calling convention
285 This calling convention (the default if no other calling convention
286 is specified) matches the target C calling conventions. This calling
287 convention supports varargs function calls and tolerates some
288 mismatch in the declared prototype and implemented declaration of
289 the function (as does normal C).
290 "``fastcc``" - The fast calling convention
291 This calling convention attempts to make calls as fast as possible
292 (e.g. by passing things in registers). This calling convention
293 allows the target to use whatever tricks it wants to produce fast
294 code for the target, without having to conform to an externally
295 specified ABI (Application Binary Interface). `Tail calls can only
296 be optimized when this, the GHC or the HiPE convention is
297 used. <CodeGenerator.html#id80>`_ This calling convention does not
298 support varargs and requires the prototype of all callees to exactly
299 match the prototype of the function definition.
300 "``coldcc``" - The cold calling convention
301 This calling convention attempts to make code in the caller as
302 efficient as possible under the assumption that the call is not
303 commonly executed. As such, these calls often preserve all registers
304 so that the call does not break any live ranges in the caller side.
305 This calling convention does not support varargs and requires the
306 prototype of all callees to exactly match the prototype of the
307 function definition. Furthermore the inliner doesn't consider such function
309 "``cc 10``" - GHC convention
310 This calling convention has been implemented specifically for use by
311 the `Glasgow Haskell Compiler (GHC) <http://www.haskell.org/ghc>`_.
312 It passes everything in registers, going to extremes to achieve this
313 by disabling callee save registers. This calling convention should
314 not be used lightly but only for specific situations such as an
315 alternative to the *register pinning* performance technique often
316 used when implementing functional programming languages. At the
317 moment only X86 supports this convention and it has the following
320 - On *X86-32* only supports up to 4 bit type parameters. No
321 floating point types are supported.
322 - On *X86-64* only supports up to 10 bit type parameters and 6
323 floating point parameters.
325 This calling convention supports `tail call
326 optimization <CodeGenerator.html#id80>`_ but requires both the
327 caller and callee are using it.
328 "``cc 11``" - The HiPE calling convention
329 This calling convention has been implemented specifically for use by
330 the `High-Performance Erlang
331 (HiPE) <http://www.it.uu.se/research/group/hipe/>`_ compiler, *the*
332 native code compiler of the `Ericsson's Open Source Erlang/OTP
333 system <http://www.erlang.org/download.shtml>`_. It uses more
334 registers for argument passing than the ordinary C calling
335 convention and defines no callee-saved registers. The calling
336 convention properly supports `tail call
337 optimization <CodeGenerator.html#id80>`_ but requires that both the
338 caller and the callee use it. It uses a *register pinning*
339 mechanism, similar to GHC's convention, for keeping frequently
340 accessed runtime components pinned to specific hardware registers.
341 At the moment only X86 supports this convention (both 32 and 64
343 "``webkit_jscc``" - WebKit's JavaScript calling convention
344 This calling convention has been implemented for `WebKit FTL JIT
345 <https://trac.webkit.org/wiki/FTLJIT>`_. It passes arguments on the
346 stack right to left (as cdecl does), and returns a value in the
347 platform's customary return register.
348 "``anyregcc``" - Dynamic calling convention for code patching
349 This is a special convention that supports patching an arbitrary code
350 sequence in place of a call site. This convention forces the call
351 arguments into registers but allows them to be dynamically
352 allocated. This can currently only be used with calls to
353 llvm.experimental.patchpoint because only this intrinsic records
354 the location of its arguments in a side table. See :doc:`StackMaps`.
355 "``preserve_mostcc``" - The `PreserveMost` calling convention
356 This calling convention attempts to make the code in the caller as
357 unintrusive as possible. This convention behaves identically to the `C`
358 calling convention on how arguments and return values are passed, but it
359 uses a different set of caller/callee-saved registers. This alleviates the
360 burden of saving and recovering a large register set before and after the
361 call in the caller. If the arguments are passed in callee-saved registers,
362 then they will be preserved by the callee across the call. This doesn't
363 apply for values returned in callee-saved registers.
365 - On X86-64 the callee preserves all general purpose registers, except for
366 R11. R11 can be used as a scratch register. Floating-point registers
367 (XMMs/YMMs) are not preserved and need to be saved by the caller.
369 The idea behind this convention is to support calls to runtime functions
370 that have a hot path and a cold path. The hot path is usually a small piece
371 of code that doesn't use many registers. The cold path might need to call out to
372 another function and therefore only needs to preserve the caller-saved
373 registers, which haven't already been saved by the caller. The
374 `PreserveMost` calling convention is very similar to the `cold` calling
375 convention in terms of caller/callee-saved registers, but they are used for
376 different types of function calls. `coldcc` is for function calls that are
377 rarely executed, whereas `preserve_mostcc` function calls are intended to be
378 on the hot path and definitely executed a lot. Furthermore `preserve_mostcc`
379 doesn't prevent the inliner from inlining the function call.
381 This calling convention will be used by a future version of the ObjectiveC
382 runtime and should therefore still be considered experimental at this time.
383 Although this convention was created to optimize certain runtime calls to
384 the ObjectiveC runtime, it is not limited to this runtime and might be used
385 by other runtimes in the future too. The current implementation only
386 supports X86-64, but the intention is to support more architectures in the
388 "``preserve_allcc``" - The `PreserveAll` calling convention
389 This calling convention attempts to make the code in the caller even less
390 intrusive than the `PreserveMost` calling convention. This calling
391 convention also behaves identical to the `C` calling convention on how
392 arguments and return values are passed, but it uses a different set of
393 caller/callee-saved registers. This removes the burden of saving and
394 recovering a large register set before and after the call in the caller. If
395 the arguments are passed in callee-saved registers, then they will be
396 preserved by the callee across the call. This doesn't apply for values
397 returned in callee-saved registers.
399 - On X86-64 the callee preserves all general purpose registers, except for
400 R11. R11 can be used as a scratch register. Furthermore it also preserves
401 all floating-point registers (XMMs/YMMs).
403 The idea behind this convention is to support calls to runtime functions
404 that don't need to call out to any other functions.
406 This calling convention, like the `PreserveMost` calling convention, will be
407 used by a future version of the ObjectiveC runtime and should be considered
408 experimental at this time.
409 "``cc <n>``" - Numbered convention
410 Any calling convention may be specified by number, allowing
411 target-specific calling conventions to be used. Target specific
412 calling conventions start at 64.
414 More calling conventions can be added/defined on an as-needed basis, to
415 support Pascal conventions or any other well-known target-independent
418 .. _visibilitystyles:
423 All Global Variables and Functions have one of the following visibility
426 "``default``" - Default style
427 On targets that use the ELF object file format, default visibility
428 means that the declaration is visible to other modules and, in
429 shared libraries, means that the declared entity may be overridden.
430 On Darwin, default visibility means that the declaration is visible
431 to other modules. Default visibility corresponds to "external
432 linkage" in the language.
433 "``hidden``" - Hidden style
434 Two declarations of an object with hidden visibility refer to the
435 same object if they are in the same shared object. Usually, hidden
436 visibility indicates that the symbol will not be placed into the
437 dynamic symbol table, so no other module (executable or shared
438 library) can reference it directly.
439 "``protected``" - Protected style
440 On ELF, protected visibility indicates that the symbol will be
441 placed in the dynamic symbol table, but that references within the
442 defining module will bind to the local symbol. That is, the symbol
443 cannot be overridden by another module.
445 A symbol with ``internal`` or ``private`` linkage must have ``default``
453 All Global Variables, Functions and Aliases can have one of the following
457 "``dllimport``" causes the compiler to reference a function or variable via
458 a global pointer to a pointer that is set up by the DLL exporting the
459 symbol. On Microsoft Windows targets, the pointer name is formed by
460 combining ``__imp_`` and the function or variable name.
462 "``dllexport``" causes the compiler to provide a global pointer to a pointer
463 in a DLL, so that it can be referenced with the ``dllimport`` attribute. On
464 Microsoft Windows targets, the pointer name is formed by combining
465 ``__imp_`` and the function or variable name. Since this storage class
466 exists for defining a dll interface, the compiler, assembler and linker know
467 it is externally referenced and must refrain from deleting the symbol.
471 Thread Local Storage Models
472 ---------------------------
474 A variable may be defined as ``thread_local``, which means that it will
475 not be shared by threads (each thread will have a separated copy of the
476 variable). Not all targets support thread-local variables. Optionally, a
477 TLS model may be specified:
480 For variables that are only used within the current shared library.
482 For variables in modules that will not be loaded dynamically.
484 For variables defined in the executable and only used within it.
486 If no explicit model is given, the "general dynamic" model is used.
488 The models correspond to the ELF TLS models; see `ELF Handling For
489 Thread-Local Storage <http://people.redhat.com/drepper/tls.pdf>`_ for
490 more information on under which circumstances the different models may
491 be used. The target may choose a different TLS model if the specified
492 model is not supported, or if a better choice of model can be made.
494 A model can also be specified in an alias, but then it only governs how
495 the alias is accessed. It will not have any effect in the aliasee.
497 For platforms without linker support of ELF TLS model, the -femulated-tls
498 flag can be used to generate GCC compatible emulated TLS code.
505 LLVM IR allows you to specify both "identified" and "literal" :ref:`structure
506 types <t_struct>`. Literal types are uniqued structurally, but identified types
507 are never uniqued. An :ref:`opaque structural type <t_opaque>` can also be used
508 to forward declare a type that is not yet available.
510 An example of an identified structure specification is:
514 %mytype = type { %mytype*, i32 }
516 Prior to the LLVM 3.0 release, identified types were structurally uniqued. Only
517 literal types are uniqued in recent versions of LLVM.
524 Global variables define regions of memory allocated at compilation time
527 Global variable definitions must be initialized.
529 Global variables in other translation units can also be declared, in which
530 case they don't have an initializer.
532 Either global variable definitions or declarations may have an explicit section
533 to be placed in and may have an optional explicit alignment specified.
535 A variable may be defined as a global ``constant``, which indicates that
536 the contents of the variable will **never** be modified (enabling better
537 optimization, allowing the global data to be placed in the read-only
538 section of an executable, etc). Note that variables that need runtime
539 initialization cannot be marked ``constant`` as there is a store to the
542 LLVM explicitly allows *declarations* of global variables to be marked
543 constant, even if the final definition of the global is not. This
544 capability can be used to enable slightly better optimization of the
545 program, but requires the language definition to guarantee that
546 optimizations based on the 'constantness' are valid for the translation
547 units that do not include the definition.
549 As SSA values, global variables define pointer values that are in scope
550 (i.e. they dominate) all basic blocks in the program. Global variables
551 always define a pointer to their "content" type because they describe a
552 region of memory, and all memory objects in LLVM are accessed through
555 Global variables can be marked with ``unnamed_addr`` which indicates
556 that the address is not significant, only the content. Constants marked
557 like this can be merged with other constants if they have the same
558 initializer. Note that a constant with significant address *can* be
559 merged with a ``unnamed_addr`` constant, the result being a constant
560 whose address is significant.
562 A global variable may be declared to reside in a target-specific
563 numbered address space. For targets that support them, address spaces
564 may affect how optimizations are performed and/or what target
565 instructions are used to access the variable. The default address space
566 is zero. The address space qualifier must precede any other attributes.
568 LLVM allows an explicit section to be specified for globals. If the
569 target supports it, it will emit globals to the section specified.
570 Additionally, the global can placed in a comdat if the target has the necessary
573 By default, global initializers are optimized by assuming that global
574 variables defined within the module are not modified from their
575 initial values before the start of the global initializer. This is
576 true even for variables potentially accessible from outside the
577 module, including those with external linkage or appearing in
578 ``@llvm.used`` or dllexported variables. This assumption may be suppressed
579 by marking the variable with ``externally_initialized``.
581 An explicit alignment may be specified for a global, which must be a
582 power of 2. If not present, or if the alignment is set to zero, the
583 alignment of the global is set by the target to whatever it feels
584 convenient. If an explicit alignment is specified, the global is forced
585 to have exactly that alignment. Targets and optimizers are not allowed
586 to over-align the global if the global has an assigned section. In this
587 case, the extra alignment could be observable: for example, code could
588 assume that the globals are densely packed in their section and try to
589 iterate over them as an array, alignment padding would break this
590 iteration. The maximum alignment is ``1 << 29``.
592 Globals can also have a :ref:`DLL storage class <dllstorageclass>`.
594 Variables and aliases can have a
595 :ref:`Thread Local Storage Model <tls_model>`.
599 [@<GlobalVarName> =] [Linkage] [Visibility] [DLLStorageClass] [ThreadLocal]
600 [unnamed_addr] [AddrSpace] [ExternallyInitialized]
601 <global | constant> <Type> [<InitializerConstant>]
602 [, section "name"] [, comdat [($name)]]
603 [, align <Alignment>]
605 For example, the following defines a global in a numbered address space
606 with an initializer, section, and alignment:
610 @G = addrspace(5) constant float 1.0, section "foo", align 4
612 The following example just declares a global variable
616 @G = external global i32
618 The following example defines a thread-local global with the
619 ``initialexec`` TLS model:
623 @G = thread_local(initialexec) global i32 0, align 4
625 .. _functionstructure:
630 LLVM function definitions consist of the "``define``" keyword, an
631 optional :ref:`linkage type <linkage>`, an optional :ref:`visibility
632 style <visibility>`, an optional :ref:`DLL storage class <dllstorageclass>`,
633 an optional :ref:`calling convention <callingconv>`,
634 an optional ``unnamed_addr`` attribute, a return type, an optional
635 :ref:`parameter attribute <paramattrs>` for the return type, a function
636 name, a (possibly empty) argument list (each with optional :ref:`parameter
637 attributes <paramattrs>`), optional :ref:`function attributes <fnattrs>`,
638 an optional section, an optional alignment,
639 an optional :ref:`comdat <langref_comdats>`,
640 an optional :ref:`garbage collector name <gc>`, an optional :ref:`prefix <prefixdata>`,
641 an optional :ref:`prologue <prologuedata>`,
642 an optional :ref:`personality <personalityfn>`,
643 an opening curly brace, a list of basic blocks, and a closing curly brace.
645 LLVM function declarations consist of the "``declare``" keyword, an
646 optional :ref:`linkage type <linkage>`, an optional :ref:`visibility
647 style <visibility>`, an optional :ref:`DLL storage class <dllstorageclass>`,
648 an optional :ref:`calling convention <callingconv>`,
649 an optional ``unnamed_addr`` attribute, a return type, an optional
650 :ref:`parameter attribute <paramattrs>` for the return type, a function
651 name, a possibly empty list of arguments, an optional alignment, an optional
652 :ref:`garbage collector name <gc>`, an optional :ref:`prefix <prefixdata>`,
653 and an optional :ref:`prologue <prologuedata>`.
655 A function definition contains a list of basic blocks, forming the CFG (Control
656 Flow Graph) for the function. Each basic block may optionally start with a label
657 (giving the basic block a symbol table entry), contains a list of instructions,
658 and ends with a :ref:`terminator <terminators>` instruction (such as a branch or
659 function return). If an explicit label is not provided, a block is assigned an
660 implicit numbered label, using the next value from the same counter as used for
661 unnamed temporaries (:ref:`see above<identifiers>`). For example, if a function
662 entry block does not have an explicit label, it will be assigned label "%0",
663 then the first unnamed temporary in that block will be "%1", etc.
665 The first basic block in a function is special in two ways: it is
666 immediately executed on entrance to the function, and it is not allowed
667 to have predecessor basic blocks (i.e. there can not be any branches to
668 the entry block of a function). Because the block can have no
669 predecessors, it also cannot have any :ref:`PHI nodes <i_phi>`.
671 LLVM allows an explicit section to be specified for functions. If the
672 target supports it, it will emit functions to the section specified.
673 Additionally, the function can be placed in a COMDAT.
675 An explicit alignment may be specified for a function. If not present,
676 or if the alignment is set to zero, the alignment of the function is set
677 by the target to whatever it feels convenient. If an explicit alignment
678 is specified, the function is forced to have at least that much
679 alignment. All alignments must be a power of 2.
681 If the ``unnamed_addr`` attribute is given, the address is known to not
682 be significant and two identical functions can be merged.
686 define [linkage] [visibility] [DLLStorageClass]
688 <ResultType> @<FunctionName> ([argument list])
689 [unnamed_addr] [fn Attrs] [section "name"] [comdat [($name)]]
690 [align N] [gc] [prefix Constant] [prologue Constant]
691 [personality Constant] { ... }
693 The argument list is a comma separated sequence of arguments where each
694 argument is of the following form:
698 <type> [parameter Attrs] [name]
706 Aliases, unlike function or variables, don't create any new data. They
707 are just a new symbol and metadata for an existing position.
709 Aliases have a name and an aliasee that is either a global value or a
712 Aliases may have an optional :ref:`linkage type <linkage>`, an optional
713 :ref:`visibility style <visibility>`, an optional :ref:`DLL storage class
714 <dllstorageclass>` and an optional :ref:`tls model <tls_model>`.
718 @<Name> = [Linkage] [Visibility] [DLLStorageClass] [ThreadLocal] [unnamed_addr] alias <AliaseeTy> @<Aliasee>
720 The linkage must be one of ``private``, ``internal``, ``linkonce``, ``weak``,
721 ``linkonce_odr``, ``weak_odr``, ``external``. Note that some system linkers
722 might not correctly handle dropping a weak symbol that is aliased.
724 Aliases that are not ``unnamed_addr`` are guaranteed to have the same address as
725 the aliasee expression. ``unnamed_addr`` ones are only guaranteed to point
728 Since aliases are only a second name, some restrictions apply, of which
729 some can only be checked when producing an object file:
731 * The expression defining the aliasee must be computable at assembly
732 time. Since it is just a name, no relocations can be used.
734 * No alias in the expression can be weak as the possibility of the
735 intermediate alias being overridden cannot be represented in an
738 * No global value in the expression can be a declaration, since that
739 would require a relocation, which is not possible.
746 Comdat IR provides access to COFF and ELF object file COMDAT functionality.
748 Comdats have a name which represents the COMDAT key. All global objects that
749 specify this key will only end up in the final object file if the linker chooses
750 that key over some other key. Aliases are placed in the same COMDAT that their
751 aliasee computes to, if any.
753 Comdats have a selection kind to provide input on how the linker should
754 choose between keys in two different object files.
758 $<Name> = comdat SelectionKind
760 The selection kind must be one of the following:
763 The linker may choose any COMDAT key, the choice is arbitrary.
765 The linker may choose any COMDAT key but the sections must contain the
768 The linker will choose the section containing the largest COMDAT key.
770 The linker requires that only section with this COMDAT key exist.
772 The linker may choose any COMDAT key but the sections must contain the
775 Note that the Mach-O platform doesn't support COMDATs and ELF only supports
776 ``any`` as a selection kind.
778 Here is an example of a COMDAT group where a function will only be selected if
779 the COMDAT key's section is the largest:
783 $foo = comdat largest
784 @foo = global i32 2, comdat($foo)
786 define void @bar() comdat($foo) {
790 As a syntactic sugar the ``$name`` can be omitted if the name is the same as
796 @foo = global i32 2, comdat
799 In a COFF object file, this will create a COMDAT section with selection kind
800 ``IMAGE_COMDAT_SELECT_LARGEST`` containing the contents of the ``@foo`` symbol
801 and another COMDAT section with selection kind
802 ``IMAGE_COMDAT_SELECT_ASSOCIATIVE`` which is associated with the first COMDAT
803 section and contains the contents of the ``@bar`` symbol.
805 There are some restrictions on the properties of the global object.
806 It, or an alias to it, must have the same name as the COMDAT group when
808 The contents and size of this object may be used during link-time to determine
809 which COMDAT groups get selected depending on the selection kind.
810 Because the name of the object must match the name of the COMDAT group, the
811 linkage of the global object must not be local; local symbols can get renamed
812 if a collision occurs in the symbol table.
814 The combined use of COMDATS and section attributes may yield surprising results.
821 @g1 = global i32 42, section "sec", comdat($foo)
822 @g2 = global i32 42, section "sec", comdat($bar)
824 From the object file perspective, this requires the creation of two sections
825 with the same name. This is necessary because both globals belong to different
826 COMDAT groups and COMDATs, at the object file level, are represented by
829 Note that certain IR constructs like global variables and functions may
830 create COMDATs in the object file in addition to any which are specified using
831 COMDAT IR. This arises when the code generator is configured to emit globals
832 in individual sections (e.g. when `-data-sections` or `-function-sections`
833 is supplied to `llc`).
835 .. _namedmetadatastructure:
840 Named metadata is a collection of metadata. :ref:`Metadata
841 nodes <metadata>` (but not metadata strings) are the only valid
842 operands for a named metadata.
844 #. Named metadata are represented as a string of characters with the
845 metadata prefix. The rules for metadata names are the same as for
846 identifiers, but quoted names are not allowed. ``"\xx"`` type escapes
847 are still valid, which allows any character to be part of a name.
851 ; Some unnamed metadata nodes, which are referenced by the named metadata.
856 !name = !{!0, !1, !2}
863 The return type and each parameter of a function type may have a set of
864 *parameter attributes* associated with them. Parameter attributes are
865 used to communicate additional information about the result or
866 parameters of a function. Parameter attributes are considered to be part
867 of the function, not of the function type, so functions with different
868 parameter attributes can have the same function type.
870 Parameter attributes are simple keywords that follow the type specified.
871 If multiple parameter attributes are needed, they are space separated.
876 declare i32 @printf(i8* noalias nocapture, ...)
877 declare i32 @atoi(i8 zeroext)
878 declare signext i8 @returns_signed_char()
880 Note that any attributes for the function result (``nounwind``,
881 ``readonly``) come immediately after the argument list.
883 Currently, only the following parameter attributes are defined:
886 This indicates to the code generator that the parameter or return
887 value should be zero-extended to the extent required by the target's
888 ABI (which is usually 32-bits, but is 8-bits for a i1 on x86-64) by
889 the caller (for a parameter) or the callee (for a return value).
891 This indicates to the code generator that the parameter or return
892 value should be sign-extended to the extent required by the target's
893 ABI (which is usually 32-bits) by the caller (for a parameter) or
894 the callee (for a return value).
896 This indicates that this parameter or return value should be treated
897 in a special target-dependent fashion while emitting code for
898 a function call or return (usually, by putting it in a register as
899 opposed to memory, though some targets use it to distinguish between
900 two different kinds of registers). Use of this attribute is
903 This indicates that the pointer parameter should really be passed by
904 value to the function. The attribute implies that a hidden copy of
905 the pointee is made between the caller and the callee, so the callee
906 is unable to modify the value in the caller. This attribute is only
907 valid on LLVM pointer arguments. It is generally used to pass
908 structs and arrays by value, but is also valid on pointers to
909 scalars. The copy is considered to belong to the caller not the
910 callee (for example, ``readonly`` functions should not write to
911 ``byval`` parameters). This is not a valid attribute for return
914 The byval attribute also supports specifying an alignment with the
915 align attribute. It indicates the alignment of the stack slot to
916 form and the known alignment of the pointer specified to the call
917 site. If the alignment is not specified, then the code generator
918 makes a target-specific assumption.
924 The ``inalloca`` argument attribute allows the caller to take the
925 address of outgoing stack arguments. An ``inalloca`` argument must
926 be a pointer to stack memory produced by an ``alloca`` instruction.
927 The alloca, or argument allocation, must also be tagged with the
928 inalloca keyword. Only the last argument may have the ``inalloca``
929 attribute, and that argument is guaranteed to be passed in memory.
931 An argument allocation may be used by a call at most once because
932 the call may deallocate it. The ``inalloca`` attribute cannot be
933 used in conjunction with other attributes that affect argument
934 storage, like ``inreg``, ``nest``, ``sret``, or ``byval``. The
935 ``inalloca`` attribute also disables LLVM's implicit lowering of
936 large aggregate return values, which means that frontend authors
937 must lower them with ``sret`` pointers.
939 When the call site is reached, the argument allocation must have
940 been the most recent stack allocation that is still live, or the
941 results are undefined. It is possible to allocate additional stack
942 space after an argument allocation and before its call site, but it
943 must be cleared off with :ref:`llvm.stackrestore
946 See :doc:`InAlloca` for more information on how to use this
950 This indicates that the pointer parameter specifies the address of a
951 structure that is the return value of the function in the source
952 program. This pointer must be guaranteed by the caller to be valid:
953 loads and stores to the structure may be assumed by the callee
954 not to trap and to be properly aligned. This may only be applied to
955 the first parameter. This is not a valid attribute for return
959 This indicates that the pointer value may be assumed by the optimizer to
960 have the specified alignment.
962 Note that this attribute has additional semantics when combined with the
968 This indicates that objects accessed via pointer values
969 :ref:`based <pointeraliasing>` on the argument or return value are not also
970 accessed, during the execution of the function, via pointer values not
971 *based* on the argument or return value. The attribute on a return value
972 also has additional semantics described below. The caller shares the
973 responsibility with the callee for ensuring that these requirements are met.
974 For further details, please see the discussion of the NoAlias response in
975 :ref:`alias analysis <Must, May, or No>`.
977 Note that this definition of ``noalias`` is intentionally similar
978 to the definition of ``restrict`` in C99 for function arguments.
980 For function return values, C99's ``restrict`` is not meaningful,
981 while LLVM's ``noalias`` is. Furthermore, the semantics of the ``noalias``
982 attribute on return values are stronger than the semantics of the attribute
983 when used on function arguments. On function return values, the ``noalias``
984 attribute indicates that the function acts like a system memory allocation
985 function, returning a pointer to allocated storage disjoint from the
986 storage for any other object accessible to the caller.
989 This indicates that the callee does not make any copies of the
990 pointer that outlive the callee itself. This is not a valid
991 attribute for return values.
996 This indicates that the pointer parameter can be excised using the
997 :ref:`trampoline intrinsics <int_trampoline>`. This is not a valid
998 attribute for return values and can only be applied to one parameter.
1001 This indicates that the function always returns the argument as its return
1002 value. This is an optimization hint to the code generator when generating
1003 the caller, allowing tail call optimization and omission of register saves
1004 and restores in some cases; it is not checked or enforced when generating
1005 the callee. The parameter and the function return type must be valid
1006 operands for the :ref:`bitcast instruction <i_bitcast>`. This is not a
1007 valid attribute for return values and can only be applied to one parameter.
1010 This indicates that the parameter or return pointer is not null. This
1011 attribute may only be applied to pointer typed parameters. This is not
1012 checked or enforced by LLVM, the caller must ensure that the pointer
1013 passed in is non-null, or the callee must ensure that the returned pointer
1016 ``dereferenceable(<n>)``
1017 This indicates that the parameter or return pointer is dereferenceable. This
1018 attribute may only be applied to pointer typed parameters. A pointer that
1019 is dereferenceable can be loaded from speculatively without a risk of
1020 trapping. The number of bytes known to be dereferenceable must be provided
1021 in parentheses. It is legal for the number of bytes to be less than the
1022 size of the pointee type. The ``nonnull`` attribute does not imply
1023 dereferenceability (consider a pointer to one element past the end of an
1024 array), however ``dereferenceable(<n>)`` does imply ``nonnull`` in
1025 ``addrspace(0)`` (which is the default address space).
1027 ``dereferenceable_or_null(<n>)``
1028 This indicates that the parameter or return value isn't both
1029 non-null and non-dereferenceable (up to ``<n>`` bytes) at the same
1030 time. All non-null pointers tagged with
1031 ``dereferenceable_or_null(<n>)`` are ``dereferenceable(<n>)``.
1032 For address space 0 ``dereferenceable_or_null(<n>)`` implies that
1033 a pointer is exactly one of ``dereferenceable(<n>)`` or ``null``,
1034 and in other address spaces ``dereferenceable_or_null(<n>)``
1035 implies that a pointer is at least one of ``dereferenceable(<n>)``
1036 or ``null`` (i.e. it may be both ``null`` and
1037 ``dereferenceable(<n>)``). This attribute may only be applied to
1038 pointer typed parameters.
1042 Garbage Collector Strategy Names
1043 --------------------------------
1045 Each function may specify a garbage collector strategy name, which is simply a
1048 .. code-block:: llvm
1050 define void @f() gc "name" { ... }
1052 The supported values of *name* includes those :ref:`built in to LLVM
1053 <builtin-gc-strategies>` and any provided by loaded plugins. Specifying a GC
1054 strategy will cause the compiler to alter its output in order to support the
1055 named garbage collection algorithm. Note that LLVM itself does not contain a
1056 garbage collector, this functionality is restricted to generating machine code
1057 which can interoperate with a collector provided externally.
1064 Prefix data is data associated with a function which the code
1065 generator will emit immediately before the function's entrypoint.
1066 The purpose of this feature is to allow frontends to associate
1067 language-specific runtime metadata with specific functions and make it
1068 available through the function pointer while still allowing the
1069 function pointer to be called.
1071 To access the data for a given function, a program may bitcast the
1072 function pointer to a pointer to the constant's type and dereference
1073 index -1. This implies that the IR symbol points just past the end of
1074 the prefix data. For instance, take the example of a function annotated
1075 with a single ``i32``,
1077 .. code-block:: llvm
1079 define void @f() prefix i32 123 { ... }
1081 The prefix data can be referenced as,
1083 .. code-block:: llvm
1085 %0 = bitcast void* () @f to i32*
1086 %a = getelementptr inbounds i32, i32* %0, i32 -1
1087 %b = load i32, i32* %a
1089 Prefix data is laid out as if it were an initializer for a global variable
1090 of the prefix data's type. The function will be placed such that the
1091 beginning of the prefix data is aligned. This means that if the size
1092 of the prefix data is not a multiple of the alignment size, the
1093 function's entrypoint will not be aligned. If alignment of the
1094 function's entrypoint is desired, padding must be added to the prefix
1097 A function may have prefix data but no body. This has similar semantics
1098 to the ``available_externally`` linkage in that the data may be used by the
1099 optimizers but will not be emitted in the object file.
1106 The ``prologue`` attribute allows arbitrary code (encoded as bytes) to
1107 be inserted prior to the function body. This can be used for enabling
1108 function hot-patching and instrumentation.
1110 To maintain the semantics of ordinary function calls, the prologue data must
1111 have a particular format. Specifically, it must begin with a sequence of
1112 bytes which decode to a sequence of machine instructions, valid for the
1113 module's target, which transfer control to the point immediately succeeding
1114 the prologue data, without performing any other visible action. This allows
1115 the inliner and other passes to reason about the semantics of the function
1116 definition without needing to reason about the prologue data. Obviously this
1117 makes the format of the prologue data highly target dependent.
1119 A trivial example of valid prologue data for the x86 architecture is ``i8 144``,
1120 which encodes the ``nop`` instruction:
1122 .. code-block:: llvm
1124 define void @f() prologue i8 144 { ... }
1126 Generally prologue data can be formed by encoding a relative branch instruction
1127 which skips the metadata, as in this example of valid prologue data for the
1128 x86_64 architecture, where the first two bytes encode ``jmp .+10``:
1130 .. code-block:: llvm
1132 %0 = type <{ i8, i8, i8* }>
1134 define void @f() prologue %0 <{ i8 235, i8 8, i8* @md}> { ... }
1136 A function may have prologue data but no body. This has similar semantics
1137 to the ``available_externally`` linkage in that the data may be used by the
1138 optimizers but will not be emitted in the object file.
1142 Personality Function
1143 --------------------
1145 The ``personality`` attribute permits functions to specify what function
1146 to use for exception handling.
1153 Attribute groups are groups of attributes that are referenced by objects within
1154 the IR. They are important for keeping ``.ll`` files readable, because a lot of
1155 functions will use the same set of attributes. In the degenerative case of a
1156 ``.ll`` file that corresponds to a single ``.c`` file, the single attribute
1157 group will capture the important command line flags used to build that file.
1159 An attribute group is a module-level object. To use an attribute group, an
1160 object references the attribute group's ID (e.g. ``#37``). An object may refer
1161 to more than one attribute group. In that situation, the attributes from the
1162 different groups are merged.
1164 Here is an example of attribute groups for a function that should always be
1165 inlined, has a stack alignment of 4, and which shouldn't use SSE instructions:
1167 .. code-block:: llvm
1169 ; Target-independent attributes:
1170 attributes #0 = { alwaysinline alignstack=4 }
1172 ; Target-dependent attributes:
1173 attributes #1 = { "no-sse" }
1175 ; Function @f has attributes: alwaysinline, alignstack=4, and "no-sse".
1176 define void @f() #0 #1 { ... }
1183 Function attributes are set to communicate additional information about
1184 a function. Function attributes are considered to be part of the
1185 function, not of the function type, so functions with different function
1186 attributes can have the same function type.
1188 Function attributes are simple keywords that follow the type specified.
1189 If multiple attributes are needed, they are space separated. For
1192 .. code-block:: llvm
1194 define void @f() noinline { ... }
1195 define void @f() alwaysinline { ... }
1196 define void @f() alwaysinline optsize { ... }
1197 define void @f() optsize { ... }
1200 This attribute indicates that, when emitting the prologue and
1201 epilogue, the backend should forcibly align the stack pointer.
1202 Specify the desired alignment, which must be a power of two, in
1205 This attribute indicates that the inliner should attempt to inline
1206 this function into callers whenever possible, ignoring any active
1207 inlining size threshold for this caller.
1209 This indicates that the callee function at a call site should be
1210 recognized as a built-in function, even though the function's declaration
1211 uses the ``nobuiltin`` attribute. This is only valid at call sites for
1212 direct calls to functions that are declared with the ``nobuiltin``
1215 This attribute indicates that this function is rarely called. When
1216 computing edge weights, basic blocks post-dominated by a cold
1217 function call are also considered to be cold; and, thus, given low
1220 This attribute indicates that the callee is dependent on a convergent
1221 thread execution pattern under certain parallel execution models.
1222 Transformations that are execution model agnostic may only move or
1223 tranform this call if the final location is control equivalent to its
1224 original position in the program, where control equivalence is defined as
1225 A dominates B and B post-dominates A, or vice versa.
1227 This attribute indicates that the source code contained a hint that
1228 inlining this function is desirable (such as the "inline" keyword in
1229 C/C++). It is just a hint; it imposes no requirements on the
1232 This attribute indicates that the function should be added to a
1233 jump-instruction table at code-generation time, and that all address-taken
1234 references to this function should be replaced with a reference to the
1235 appropriate jump-instruction-table function pointer. Note that this creates
1236 a new pointer for the original function, which means that code that depends
1237 on function-pointer identity can break. So, any function annotated with
1238 ``jumptable`` must also be ``unnamed_addr``.
1240 This attribute suggests that optimization passes and code generator
1241 passes make choices that keep the code size of this function as small
1242 as possible and perform optimizations that may sacrifice runtime
1243 performance in order to minimize the size of the generated code.
1245 This attribute disables prologue / epilogue emission for the
1246 function. This can have very system-specific consequences.
1248 This indicates that the callee function at a call site is not recognized as
1249 a built-in function. LLVM will retain the original call and not replace it
1250 with equivalent code based on the semantics of the built-in function, unless
1251 the call site uses the ``builtin`` attribute. This is valid at call sites
1252 and on function declarations and definitions.
1254 This attribute indicates that calls to the function cannot be
1255 duplicated. A call to a ``noduplicate`` function may be moved
1256 within its parent function, but may not be duplicated within
1257 its parent function.
1259 A function containing a ``noduplicate`` call may still
1260 be an inlining candidate, provided that the call is not
1261 duplicated by inlining. That implies that the function has
1262 internal linkage and only has one call site, so the original
1263 call is dead after inlining.
1265 This attributes disables implicit floating point instructions.
1267 This attribute indicates that the inliner should never inline this
1268 function in any situation. This attribute may not be used together
1269 with the ``alwaysinline`` attribute.
1271 This attribute suppresses lazy symbol binding for the function. This
1272 may make calls to the function faster, at the cost of extra program
1273 startup time if the function is not called during program startup.
1275 This attribute indicates that the code generator should not use a
1276 red zone, even if the target-specific ABI normally permits it.
1278 This function attribute indicates that the function never returns
1279 normally. This produces undefined behavior at runtime if the
1280 function ever does dynamically return.
1282 This function attribute indicates that the function never raises an
1283 exception. If the function does raise an exception, its runtime
1284 behavior is undefined. However, functions marked nounwind may still
1285 trap or generate asynchronous exceptions. Exception handling schemes
1286 that are recognized by LLVM to handle asynchronous exceptions, such
1287 as SEH, will still provide their implementation defined semantics.
1289 This function attribute indicates that the function is not optimized
1290 by any optimization or code generator passes with the
1291 exception of interprocedural optimization passes.
1292 This attribute cannot be used together with the ``alwaysinline``
1293 attribute; this attribute is also incompatible
1294 with the ``minsize`` attribute and the ``optsize`` attribute.
1296 This attribute requires the ``noinline`` attribute to be specified on
1297 the function as well, so the function is never inlined into any caller.
1298 Only functions with the ``alwaysinline`` attribute are valid
1299 candidates for inlining into the body of this function.
1301 This attribute suggests that optimization passes and code generator
1302 passes make choices that keep the code size of this function low,
1303 and otherwise do optimizations specifically to reduce code size as
1304 long as they do not significantly impact runtime performance.
1306 On a function, this attribute indicates that the function computes its
1307 result (or decides to unwind an exception) based strictly on its arguments,
1308 without dereferencing any pointer arguments or otherwise accessing
1309 any mutable state (e.g. memory, control registers, etc) visible to
1310 caller functions. It does not write through any pointer arguments
1311 (including ``byval`` arguments) and never changes any state visible
1312 to callers. This means that it cannot unwind exceptions by calling
1313 the ``C++`` exception throwing methods.
1315 On an argument, this attribute indicates that the function does not
1316 dereference that pointer argument, even though it may read or write the
1317 memory that the pointer points to if accessed through other pointers.
1319 On a function, this attribute indicates that the function does not write
1320 through any pointer arguments (including ``byval`` arguments) or otherwise
1321 modify any state (e.g. memory, control registers, etc) visible to
1322 caller functions. It may dereference pointer arguments and read
1323 state that may be set in the caller. A readonly function always
1324 returns the same value (or unwinds an exception identically) when
1325 called with the same set of arguments and global state. It cannot
1326 unwind an exception by calling the ``C++`` exception throwing
1329 On an argument, this attribute indicates that the function does not write
1330 through this pointer argument, even though it may write to the memory that
1331 the pointer points to.
1333 This attribute indicates that the only memory accesses inside function are
1334 loads and stores from objects pointed to by its pointer-typed arguments,
1335 with arbitrary offsets. Or in other words, all memory operations in the
1336 function can refer to memory only using pointers based on its function
1338 Note that ``argmemonly`` can be used together with ``readonly`` attribute
1339 in order to specify that function reads only from its arguments.
1341 This attribute indicates that this function can return twice. The C
1342 ``setjmp`` is an example of such a function. The compiler disables
1343 some optimizations (like tail calls) in the caller of these
1346 This attribute indicates that
1347 `SafeStack <http://clang.llvm.org/docs/SafeStack.html>`_
1348 protection is enabled for this function.
1350 If a function that has a ``safestack`` attribute is inlined into a
1351 function that doesn't have a ``safestack`` attribute or which has an
1352 ``ssp``, ``sspstrong`` or ``sspreq`` attribute, then the resulting
1353 function will have a ``safestack`` attribute.
1354 ``sanitize_address``
1355 This attribute indicates that AddressSanitizer checks
1356 (dynamic address safety analysis) are enabled for this function.
1358 This attribute indicates that MemorySanitizer checks (dynamic detection
1359 of accesses to uninitialized memory) are enabled for this function.
1361 This attribute indicates that ThreadSanitizer checks
1362 (dynamic thread safety analysis) are enabled for this function.
1364 This attribute indicates that the function should emit a stack
1365 smashing protector. It is in the form of a "canary" --- a random value
1366 placed on the stack before the local variables that's checked upon
1367 return from the function to see if it has been overwritten. A
1368 heuristic is used to determine if a function needs stack protectors
1369 or not. The heuristic used will enable protectors for functions with:
1371 - Character arrays larger than ``ssp-buffer-size`` (default 8).
1372 - Aggregates containing character arrays larger than ``ssp-buffer-size``.
1373 - Calls to alloca() with variable sizes or constant sizes greater than
1374 ``ssp-buffer-size``.
1376 Variables that are identified as requiring a protector will be arranged
1377 on the stack such that they are adjacent to the stack protector guard.
1379 If a function that has an ``ssp`` attribute is inlined into a
1380 function that doesn't have an ``ssp`` attribute, then the resulting
1381 function will have an ``ssp`` attribute.
1383 This attribute indicates that the function should *always* emit a
1384 stack smashing protector. This overrides the ``ssp`` function
1387 Variables that are identified as requiring a protector will be arranged
1388 on the stack such that they are adjacent to the stack protector guard.
1389 The specific layout rules are:
1391 #. Large arrays and structures containing large arrays
1392 (``>= ssp-buffer-size``) are closest to the stack protector.
1393 #. Small arrays and structures containing small arrays
1394 (``< ssp-buffer-size``) are 2nd closest to the protector.
1395 #. Variables that have had their address taken are 3rd closest to the
1398 If a function that has an ``sspreq`` attribute is inlined into a
1399 function that doesn't have an ``sspreq`` attribute or which has an
1400 ``ssp`` or ``sspstrong`` attribute, then the resulting function will have
1401 an ``sspreq`` attribute.
1403 This attribute indicates that the function should emit a stack smashing
1404 protector. This attribute causes a strong heuristic to be used when
1405 determining if a function needs stack protectors. The strong heuristic
1406 will enable protectors for functions with:
1408 - Arrays of any size and type
1409 - Aggregates containing an array of any size and type.
1410 - Calls to alloca().
1411 - Local variables that have had their address taken.
1413 Variables that are identified as requiring a protector will be arranged
1414 on the stack such that they are adjacent to the stack protector guard.
1415 The specific layout rules are:
1417 #. Large arrays and structures containing large arrays
1418 (``>= ssp-buffer-size``) are closest to the stack protector.
1419 #. Small arrays and structures containing small arrays
1420 (``< ssp-buffer-size``) are 2nd closest to the protector.
1421 #. Variables that have had their address taken are 3rd closest to the
1424 This overrides the ``ssp`` function attribute.
1426 If a function that has an ``sspstrong`` attribute is inlined into a
1427 function that doesn't have an ``sspstrong`` attribute, then the
1428 resulting function will have an ``sspstrong`` attribute.
1430 This attribute indicates that the function will delegate to some other
1431 function with a tail call. The prototype of a thunk should not be used for
1432 optimization purposes. The caller is expected to cast the thunk prototype to
1433 match the thunk target prototype.
1435 This attribute indicates that the ABI being targeted requires that
1436 an unwind table entry be produced for this function even if we can
1437 show that no exceptions passes by it. This is normally the case for
1438 the ELF x86-64 abi, but it can be disabled for some compilation
1443 Module-Level Inline Assembly
1444 ----------------------------
1446 Modules may contain "module-level inline asm" blocks, which corresponds
1447 to the GCC "file scope inline asm" blocks. These blocks are internally
1448 concatenated by LLVM and treated as a single unit, but may be separated
1449 in the ``.ll`` file if desired. The syntax is very simple:
1451 .. code-block:: llvm
1453 module asm "inline asm code goes here"
1454 module asm "more can go here"
1456 The strings can contain any character by escaping non-printable
1457 characters. The escape sequence used is simply "\\xx" where "xx" is the
1458 two digit hex code for the number.
1460 Note that the assembly string *must* be parseable by LLVM's integrated assembler
1461 (unless it is disabled), even when emitting a ``.s`` file.
1463 .. _langref_datalayout:
1468 A module may specify a target specific data layout string that specifies
1469 how data is to be laid out in memory. The syntax for the data layout is
1472 .. code-block:: llvm
1474 target datalayout = "layout specification"
1476 The *layout specification* consists of a list of specifications
1477 separated by the minus sign character ('-'). Each specification starts
1478 with a letter and may include other information after the letter to
1479 define some aspect of the data layout. The specifications accepted are
1483 Specifies that the target lays out data in big-endian form. That is,
1484 the bits with the most significance have the lowest address
1487 Specifies that the target lays out data in little-endian form. That
1488 is, the bits with the least significance have the lowest address
1491 Specifies the natural alignment of the stack in bits. Alignment
1492 promotion of stack variables is limited to the natural stack
1493 alignment to avoid dynamic stack realignment. The stack alignment
1494 must be a multiple of 8-bits. If omitted, the natural stack
1495 alignment defaults to "unspecified", which does not prevent any
1496 alignment promotions.
1497 ``p[n]:<size>:<abi>:<pref>``
1498 This specifies the *size* of a pointer and its ``<abi>`` and
1499 ``<pref>``\erred alignments for address space ``n``. All sizes are in
1500 bits. The address space, ``n``, is optional, and if not specified,
1501 denotes the default address space 0. The value of ``n`` must be
1502 in the range [1,2^23).
1503 ``i<size>:<abi>:<pref>``
1504 This specifies the alignment for an integer type of a given bit
1505 ``<size>``. The value of ``<size>`` must be in the range [1,2^23).
1506 ``v<size>:<abi>:<pref>``
1507 This specifies the alignment for a vector type of a given bit
1509 ``f<size>:<abi>:<pref>``
1510 This specifies the alignment for a floating point type of a given bit
1511 ``<size>``. Only values of ``<size>`` that are supported by the target
1512 will work. 32 (float) and 64 (double) are supported on all targets; 80
1513 or 128 (different flavors of long double) are also supported on some
1516 This specifies the alignment for an object of aggregate type.
1518 If present, specifies that llvm names are mangled in the output. The
1521 * ``e``: ELF mangling: Private symbols get a ``.L`` prefix.
1522 * ``m``: Mips mangling: Private symbols get a ``$`` prefix.
1523 * ``o``: Mach-O mangling: Private symbols get ``L`` prefix. Other
1524 symbols get a ``_`` prefix.
1525 * ``w``: Windows COFF prefix: Similar to Mach-O, but stdcall and fastcall
1526 functions also get a suffix based on the frame size.
1527 ``n<size1>:<size2>:<size3>...``
1528 This specifies a set of native integer widths for the target CPU in
1529 bits. For example, it might contain ``n32`` for 32-bit PowerPC,
1530 ``n32:64`` for PowerPC 64, or ``n8:16:32:64`` for X86-64. Elements of
1531 this set are considered to support most general arithmetic operations
1534 On every specification that takes a ``<abi>:<pref>``, specifying the
1535 ``<pref>`` alignment is optional. If omitted, the preceding ``:``
1536 should be omitted too and ``<pref>`` will be equal to ``<abi>``.
1538 When constructing the data layout for a given target, LLVM starts with a
1539 default set of specifications which are then (possibly) overridden by
1540 the specifications in the ``datalayout`` keyword. The default
1541 specifications are given in this list:
1543 - ``E`` - big endian
1544 - ``p:64:64:64`` - 64-bit pointers with 64-bit alignment.
1545 - ``p[n]:64:64:64`` - Other address spaces are assumed to be the
1546 same as the default address space.
1547 - ``S0`` - natural stack alignment is unspecified
1548 - ``i1:8:8`` - i1 is 8-bit (byte) aligned
1549 - ``i8:8:8`` - i8 is 8-bit (byte) aligned
1550 - ``i16:16:16`` - i16 is 16-bit aligned
1551 - ``i32:32:32`` - i32 is 32-bit aligned
1552 - ``i64:32:64`` - i64 has ABI alignment of 32-bits but preferred
1553 alignment of 64-bits
1554 - ``f16:16:16`` - half is 16-bit aligned
1555 - ``f32:32:32`` - float is 32-bit aligned
1556 - ``f64:64:64`` - double is 64-bit aligned
1557 - ``f128:128:128`` - quad is 128-bit aligned
1558 - ``v64:64:64`` - 64-bit vector is 64-bit aligned
1559 - ``v128:128:128`` - 128-bit vector is 128-bit aligned
1560 - ``a:0:64`` - aggregates are 64-bit aligned
1562 When LLVM is determining the alignment for a given type, it uses the
1565 #. If the type sought is an exact match for one of the specifications,
1566 that specification is used.
1567 #. If no match is found, and the type sought is an integer type, then
1568 the smallest integer type that is larger than the bitwidth of the
1569 sought type is used. If none of the specifications are larger than
1570 the bitwidth then the largest integer type is used. For example,
1571 given the default specifications above, the i7 type will use the
1572 alignment of i8 (next largest) while both i65 and i256 will use the
1573 alignment of i64 (largest specified).
1574 #. If no match is found, and the type sought is a vector type, then the
1575 largest vector type that is smaller than the sought vector type will
1576 be used as a fall back. This happens because <128 x double> can be
1577 implemented in terms of 64 <2 x double>, for example.
1579 The function of the data layout string may not be what you expect.
1580 Notably, this is not a specification from the frontend of what alignment
1581 the code generator should use.
1583 Instead, if specified, the target data layout is required to match what
1584 the ultimate *code generator* expects. This string is used by the
1585 mid-level optimizers to improve code, and this only works if it matches
1586 what the ultimate code generator uses. There is no way to generate IR
1587 that does not embed this target-specific detail into the IR. If you
1588 don't specify the string, the default specifications will be used to
1589 generate a Data Layout and the optimization phases will operate
1590 accordingly and introduce target specificity into the IR with respect to
1591 these default specifications.
1598 A module may specify a target triple string that describes the target
1599 host. The syntax for the target triple is simply:
1601 .. code-block:: llvm
1603 target triple = "x86_64-apple-macosx10.7.0"
1605 The *target triple* string consists of a series of identifiers delimited
1606 by the minus sign character ('-'). The canonical forms are:
1610 ARCHITECTURE-VENDOR-OPERATING_SYSTEM
1611 ARCHITECTURE-VENDOR-OPERATING_SYSTEM-ENVIRONMENT
1613 This information is passed along to the backend so that it generates
1614 code for the proper architecture. It's possible to override this on the
1615 command line with the ``-mtriple`` command line option.
1617 .. _pointeraliasing:
1619 Pointer Aliasing Rules
1620 ----------------------
1622 Any memory access must be done through a pointer value associated with
1623 an address range of the memory access, otherwise the behavior is
1624 undefined. Pointer values are associated with address ranges according
1625 to the following rules:
1627 - A pointer value is associated with the addresses associated with any
1628 value it is *based* on.
1629 - An address of a global variable is associated with the address range
1630 of the variable's storage.
1631 - The result value of an allocation instruction is associated with the
1632 address range of the allocated storage.
1633 - A null pointer in the default address-space is associated with no
1635 - An integer constant other than zero or a pointer value returned from
1636 a function not defined within LLVM may be associated with address
1637 ranges allocated through mechanisms other than those provided by
1638 LLVM. Such ranges shall not overlap with any ranges of addresses
1639 allocated by mechanisms provided by LLVM.
1641 A pointer value is *based* on another pointer value according to the
1644 - A pointer value formed from a ``getelementptr`` operation is *based*
1645 on the first value operand of the ``getelementptr``.
1646 - The result value of a ``bitcast`` is *based* on the operand of the
1648 - A pointer value formed by an ``inttoptr`` is *based* on all pointer
1649 values that contribute (directly or indirectly) to the computation of
1650 the pointer's value.
1651 - The "*based* on" relationship is transitive.
1653 Note that this definition of *"based"* is intentionally similar to the
1654 definition of *"based"* in C99, though it is slightly weaker.
1656 LLVM IR does not associate types with memory. The result type of a
1657 ``load`` merely indicates the size and alignment of the memory from
1658 which to load, as well as the interpretation of the value. The first
1659 operand type of a ``store`` similarly only indicates the size and
1660 alignment of the store.
1662 Consequently, type-based alias analysis, aka TBAA, aka
1663 ``-fstrict-aliasing``, is not applicable to general unadorned LLVM IR.
1664 :ref:`Metadata <metadata>` may be used to encode additional information
1665 which specialized optimization passes may use to implement type-based
1670 Volatile Memory Accesses
1671 ------------------------
1673 Certain memory accesses, such as :ref:`load <i_load>`'s,
1674 :ref:`store <i_store>`'s, and :ref:`llvm.memcpy <int_memcpy>`'s may be
1675 marked ``volatile``. The optimizers must not change the number of
1676 volatile operations or change their order of execution relative to other
1677 volatile operations. The optimizers *may* change the order of volatile
1678 operations relative to non-volatile operations. This is not Java's
1679 "volatile" and has no cross-thread synchronization behavior.
1681 IR-level volatile loads and stores cannot safely be optimized into
1682 llvm.memcpy or llvm.memmove intrinsics even when those intrinsics are
1683 flagged volatile. Likewise, the backend should never split or merge
1684 target-legal volatile load/store instructions.
1686 .. admonition:: Rationale
1688 Platforms may rely on volatile loads and stores of natively supported
1689 data width to be executed as single instruction. For example, in C
1690 this holds for an l-value of volatile primitive type with native
1691 hardware support, but not necessarily for aggregate types. The
1692 frontend upholds these expectations, which are intentionally
1693 unspecified in the IR. The rules above ensure that IR transformations
1694 do not violate the frontend's contract with the language.
1698 Memory Model for Concurrent Operations
1699 --------------------------------------
1701 The LLVM IR does not define any way to start parallel threads of
1702 execution or to register signal handlers. Nonetheless, there are
1703 platform-specific ways to create them, and we define LLVM IR's behavior
1704 in their presence. This model is inspired by the C++0x memory model.
1706 For a more informal introduction to this model, see the :doc:`Atomics`.
1708 We define a *happens-before* partial order as the least partial order
1711 - Is a superset of single-thread program order, and
1712 - When a *synchronizes-with* ``b``, includes an edge from ``a`` to
1713 ``b``. *Synchronizes-with* pairs are introduced by platform-specific
1714 techniques, like pthread locks, thread creation, thread joining,
1715 etc., and by atomic instructions. (See also :ref:`Atomic Memory Ordering
1716 Constraints <ordering>`).
1718 Note that program order does not introduce *happens-before* edges
1719 between a thread and signals executing inside that thread.
1721 Every (defined) read operation (load instructions, memcpy, atomic
1722 loads/read-modify-writes, etc.) R reads a series of bytes written by
1723 (defined) write operations (store instructions, atomic
1724 stores/read-modify-writes, memcpy, etc.). For the purposes of this
1725 section, initialized globals are considered to have a write of the
1726 initializer which is atomic and happens before any other read or write
1727 of the memory in question. For each byte of a read R, R\ :sub:`byte`
1728 may see any write to the same byte, except:
1730 - If write\ :sub:`1` happens before write\ :sub:`2`, and
1731 write\ :sub:`2` happens before R\ :sub:`byte`, then
1732 R\ :sub:`byte` does not see write\ :sub:`1`.
1733 - If R\ :sub:`byte` happens before write\ :sub:`3`, then
1734 R\ :sub:`byte` does not see write\ :sub:`3`.
1736 Given that definition, R\ :sub:`byte` is defined as follows:
1738 - If R is volatile, the result is target-dependent. (Volatile is
1739 supposed to give guarantees which can support ``sig_atomic_t`` in
1740 C/C++, and may be used for accesses to addresses that do not behave
1741 like normal memory. It does not generally provide cross-thread
1743 - Otherwise, if there is no write to the same byte that happens before
1744 R\ :sub:`byte`, R\ :sub:`byte` returns ``undef`` for that byte.
1745 - Otherwise, if R\ :sub:`byte` may see exactly one write,
1746 R\ :sub:`byte` returns the value written by that write.
1747 - Otherwise, if R is atomic, and all the writes R\ :sub:`byte` may
1748 see are atomic, it chooses one of the values written. See the :ref:`Atomic
1749 Memory Ordering Constraints <ordering>` section for additional
1750 constraints on how the choice is made.
1751 - Otherwise R\ :sub:`byte` returns ``undef``.
1753 R returns the value composed of the series of bytes it read. This
1754 implies that some bytes within the value may be ``undef`` **without**
1755 the entire value being ``undef``. Note that this only defines the
1756 semantics of the operation; it doesn't mean that targets will emit more
1757 than one instruction to read the series of bytes.
1759 Note that in cases where none of the atomic intrinsics are used, this
1760 model places only one restriction on IR transformations on top of what
1761 is required for single-threaded execution: introducing a store to a byte
1762 which might not otherwise be stored is not allowed in general.
1763 (Specifically, in the case where another thread might write to and read
1764 from an address, introducing a store can change a load that may see
1765 exactly one write into a load that may see multiple writes.)
1769 Atomic Memory Ordering Constraints
1770 ----------------------------------
1772 Atomic instructions (:ref:`cmpxchg <i_cmpxchg>`,
1773 :ref:`atomicrmw <i_atomicrmw>`, :ref:`fence <i_fence>`,
1774 :ref:`atomic load <i_load>`, and :ref:`atomic store <i_store>`) take
1775 ordering parameters that determine which other atomic instructions on
1776 the same address they *synchronize with*. These semantics are borrowed
1777 from Java and C++0x, but are somewhat more colloquial. If these
1778 descriptions aren't precise enough, check those specs (see spec
1779 references in the :doc:`atomics guide <Atomics>`).
1780 :ref:`fence <i_fence>` instructions treat these orderings somewhat
1781 differently since they don't take an address. See that instruction's
1782 documentation for details.
1784 For a simpler introduction to the ordering constraints, see the
1788 The set of values that can be read is governed by the happens-before
1789 partial order. A value cannot be read unless some operation wrote
1790 it. This is intended to provide a guarantee strong enough to model
1791 Java's non-volatile shared variables. This ordering cannot be
1792 specified for read-modify-write operations; it is not strong enough
1793 to make them atomic in any interesting way.
1795 In addition to the guarantees of ``unordered``, there is a single
1796 total order for modifications by ``monotonic`` operations on each
1797 address. All modification orders must be compatible with the
1798 happens-before order. There is no guarantee that the modification
1799 orders can be combined to a global total order for the whole program
1800 (and this often will not be possible). The read in an atomic
1801 read-modify-write operation (:ref:`cmpxchg <i_cmpxchg>` and
1802 :ref:`atomicrmw <i_atomicrmw>`) reads the value in the modification
1803 order immediately before the value it writes. If one atomic read
1804 happens before another atomic read of the same address, the later
1805 read must see the same value or a later value in the address's
1806 modification order. This disallows reordering of ``monotonic`` (or
1807 stronger) operations on the same address. If an address is written
1808 ``monotonic``-ally by one thread, and other threads ``monotonic``-ally
1809 read that address repeatedly, the other threads must eventually see
1810 the write. This corresponds to the C++0x/C1x
1811 ``memory_order_relaxed``.
1813 In addition to the guarantees of ``monotonic``, a
1814 *synchronizes-with* edge may be formed with a ``release`` operation.
1815 This is intended to model C++'s ``memory_order_acquire``.
1817 In addition to the guarantees of ``monotonic``, if this operation
1818 writes a value which is subsequently read by an ``acquire``
1819 operation, it *synchronizes-with* that operation. (This isn't a
1820 complete description; see the C++0x definition of a release
1821 sequence.) This corresponds to the C++0x/C1x
1822 ``memory_order_release``.
1823 ``acq_rel`` (acquire+release)
1824 Acts as both an ``acquire`` and ``release`` operation on its
1825 address. This corresponds to the C++0x/C1x ``memory_order_acq_rel``.
1826 ``seq_cst`` (sequentially consistent)
1827 In addition to the guarantees of ``acq_rel`` (``acquire`` for an
1828 operation that only reads, ``release`` for an operation that only
1829 writes), there is a global total order on all
1830 sequentially-consistent operations on all addresses, which is
1831 consistent with the *happens-before* partial order and with the
1832 modification orders of all the affected addresses. Each
1833 sequentially-consistent read sees the last preceding write to the
1834 same address in this global order. This corresponds to the C++0x/C1x
1835 ``memory_order_seq_cst`` and Java volatile.
1839 If an atomic operation is marked ``singlethread``, it only *synchronizes
1840 with* or participates in modification and seq\_cst total orderings with
1841 other operations running in the same thread (for example, in signal
1849 LLVM IR floating-point binary ops (:ref:`fadd <i_fadd>`,
1850 :ref:`fsub <i_fsub>`, :ref:`fmul <i_fmul>`, :ref:`fdiv <i_fdiv>`,
1851 :ref:`frem <i_frem>`, :ref:`fcmp <i_fcmp>`) have the following flags that can
1852 be set to enable otherwise unsafe floating point operations
1855 No NaNs - Allow optimizations to assume the arguments and result are not
1856 NaN. Such optimizations are required to retain defined behavior over
1857 NaNs, but the value of the result is undefined.
1860 No Infs - Allow optimizations to assume the arguments and result are not
1861 +/-Inf. Such optimizations are required to retain defined behavior over
1862 +/-Inf, but the value of the result is undefined.
1865 No Signed Zeros - Allow optimizations to treat the sign of a zero
1866 argument or result as insignificant.
1869 Allow Reciprocal - Allow optimizations to use the reciprocal of an
1870 argument rather than perform division.
1873 Fast - Allow algebraically equivalent transformations that may
1874 dramatically change results in floating point (e.g. reassociate). This
1875 flag implies all the others.
1879 Use-list Order Directives
1880 -------------------------
1882 Use-list directives encode the in-memory order of each use-list, allowing the
1883 order to be recreated. ``<order-indexes>`` is a comma-separated list of
1884 indexes that are assigned to the referenced value's uses. The referenced
1885 value's use-list is immediately sorted by these indexes.
1887 Use-list directives may appear at function scope or global scope. They are not
1888 instructions, and have no effect on the semantics of the IR. When they're at
1889 function scope, they must appear after the terminator of the final basic block.
1891 If basic blocks have their address taken via ``blockaddress()`` expressions,
1892 ``uselistorder_bb`` can be used to reorder their use-lists from outside their
1899 uselistorder <ty> <value>, { <order-indexes> }
1900 uselistorder_bb @function, %block { <order-indexes> }
1906 define void @foo(i32 %arg1, i32 %arg2) {
1908 ; ... instructions ...
1910 ; ... instructions ...
1912 ; At function scope.
1913 uselistorder i32 %arg1, { 1, 0, 2 }
1914 uselistorder label %bb, { 1, 0 }
1918 uselistorder i32* @global, { 1, 2, 0 }
1919 uselistorder i32 7, { 1, 0 }
1920 uselistorder i32 (i32) @bar, { 1, 0 }
1921 uselistorder_bb @foo, %bb, { 5, 1, 3, 2, 0, 4 }
1928 The LLVM type system is one of the most important features of the
1929 intermediate representation. Being typed enables a number of
1930 optimizations to be performed on the intermediate representation
1931 directly, without having to do extra analyses on the side before the
1932 transformation. A strong type system makes it easier to read the
1933 generated code and enables novel analyses and transformations that are
1934 not feasible to perform on normal three address code representations.
1944 The void type does not represent any value and has no size.
1962 The function type can be thought of as a function signature. It consists of a
1963 return type and a list of formal parameter types. The return type of a function
1964 type is a void type or first class type --- except for :ref:`label <t_label>`
1965 and :ref:`metadata <t_metadata>` types.
1971 <returntype> (<parameter list>)
1973 ...where '``<parameter list>``' is a comma-separated list of type
1974 specifiers. Optionally, the parameter list may include a type ``...``, which
1975 indicates that the function takes a variable number of arguments. Variable
1976 argument functions can access their arguments with the :ref:`variable argument
1977 handling intrinsic <int_varargs>` functions. '``<returntype>``' is any type
1978 except :ref:`label <t_label>` and :ref:`metadata <t_metadata>`.
1982 +---------------------------------+---------------------------------------------------------------------------------------------------------------------------------------------------------------------+
1983 | ``i32 (i32)`` | function taking an ``i32``, returning an ``i32`` |
1984 +---------------------------------+---------------------------------------------------------------------------------------------------------------------------------------------------------------------+
1985 | ``float (i16, i32 *) *`` | :ref:`Pointer <t_pointer>` to a function that takes an ``i16`` and a :ref:`pointer <t_pointer>` to ``i32``, returning ``float``. |
1986 +---------------------------------+---------------------------------------------------------------------------------------------------------------------------------------------------------------------+
1987 | ``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. |
1988 +---------------------------------+---------------------------------------------------------------------------------------------------------------------------------------------------------------------+
1989 | ``{i32, i32} (i32)`` | A function taking an ``i32``, returning a :ref:`structure <t_struct>` containing two ``i32`` values |
1990 +---------------------------------+---------------------------------------------------------------------------------------------------------------------------------------------------------------------+
1997 The :ref:`first class <t_firstclass>` types are perhaps the most important.
1998 Values of these types are the only ones which can be produced by
2006 These are the types that are valid in registers from CodeGen's perspective.
2015 The integer type is a very simple type that simply specifies an
2016 arbitrary bit width for the integer type desired. Any bit width from 1
2017 bit to 2\ :sup:`23`\ -1 (about 8 million) can be specified.
2025 The number of bits the integer will occupy is specified by the ``N``
2031 +----------------+------------------------------------------------+
2032 | ``i1`` | a single-bit integer. |
2033 +----------------+------------------------------------------------+
2034 | ``i32`` | a 32-bit integer. |
2035 +----------------+------------------------------------------------+
2036 | ``i1942652`` | a really big integer of over 1 million bits. |
2037 +----------------+------------------------------------------------+
2041 Floating Point Types
2042 """"""""""""""""""""
2051 - 16-bit floating point value
2054 - 32-bit floating point value
2057 - 64-bit floating point value
2060 - 128-bit floating point value (112-bit mantissa)
2063 - 80-bit floating point value (X87)
2066 - 128-bit floating point value (two 64-bits)
2073 The x86_mmx type represents a value held in an MMX register on an x86
2074 machine. The operations allowed on it are quite limited: parameters and
2075 return values, load and store, and bitcast. User-specified MMX
2076 instructions are represented as intrinsic or asm calls with arguments
2077 and/or results of this type. There are no arrays, vectors or constants
2094 The pointer type is used to specify memory locations. Pointers are
2095 commonly used to reference objects in memory.
2097 Pointer types may have an optional address space attribute defining the
2098 numbered address space where the pointed-to object resides. The default
2099 address space is number zero. The semantics of non-zero address spaces
2100 are target-specific.
2102 Note that LLVM does not permit pointers to void (``void*``) nor does it
2103 permit pointers to labels (``label*``). Use ``i8*`` instead.
2113 +-------------------------+--------------------------------------------------------------------------------------------------------------+
2114 | ``[4 x i32]*`` | A :ref:`pointer <t_pointer>` to :ref:`array <t_array>` of four ``i32`` values. |
2115 +-------------------------+--------------------------------------------------------------------------------------------------------------+
2116 | ``i32 (i32*) *`` | A :ref:`pointer <t_pointer>` to a :ref:`function <t_function>` that takes an ``i32*``, returning an ``i32``. |
2117 +-------------------------+--------------------------------------------------------------------------------------------------------------+
2118 | ``i32 addrspace(5)*`` | A :ref:`pointer <t_pointer>` to an ``i32`` value that resides in address space #5. |
2119 +-------------------------+--------------------------------------------------------------------------------------------------------------+
2128 A vector type is a simple derived type that represents a vector of
2129 elements. Vector types are used when multiple primitive data are
2130 operated in parallel using a single instruction (SIMD). A vector type
2131 requires a size (number of elements) and an underlying primitive data
2132 type. Vector types are considered :ref:`first class <t_firstclass>`.
2138 < <# elements> x <elementtype> >
2140 The number of elements is a constant integer value larger than 0;
2141 elementtype may be any integer, floating point or pointer type. Vectors
2142 of size zero are not allowed.
2146 +-------------------+--------------------------------------------------+
2147 | ``<4 x i32>`` | Vector of 4 32-bit integer values. |
2148 +-------------------+--------------------------------------------------+
2149 | ``<8 x float>`` | Vector of 8 32-bit floating-point values. |
2150 +-------------------+--------------------------------------------------+
2151 | ``<2 x i64>`` | Vector of 2 64-bit integer values. |
2152 +-------------------+--------------------------------------------------+
2153 | ``<4 x i64*>`` | Vector of 4 pointers to 64-bit integer values. |
2154 +-------------------+--------------------------------------------------+
2163 The label type represents code labels.
2178 The token type is used when a value is associated with an instruction
2179 but all uses of the value must not attempt to introspect or obscure it.
2180 As such, it is not appropriate to have a :ref:`phi <i_phi>` or
2181 :ref:`select <i_select>` of type token.
2198 The metadata type represents embedded metadata. No derived types may be
2199 created from metadata except for :ref:`function <t_function>` arguments.
2212 Aggregate Types are a subset of derived types that can contain multiple
2213 member types. :ref:`Arrays <t_array>` and :ref:`structs <t_struct>` are
2214 aggregate types. :ref:`Vectors <t_vector>` are not considered to be
2224 The array type is a very simple derived type that arranges elements
2225 sequentially in memory. The array type requires a size (number of
2226 elements) and an underlying data type.
2232 [<# elements> x <elementtype>]
2234 The number of elements is a constant integer value; ``elementtype`` may
2235 be any type with a size.
2239 +------------------+--------------------------------------+
2240 | ``[40 x i32]`` | Array of 40 32-bit integer values. |
2241 +------------------+--------------------------------------+
2242 | ``[41 x i32]`` | Array of 41 32-bit integer values. |
2243 +------------------+--------------------------------------+
2244 | ``[4 x i8]`` | Array of 4 8-bit integer values. |
2245 +------------------+--------------------------------------+
2247 Here are some examples of multidimensional arrays:
2249 +-----------------------------+----------------------------------------------------------+
2250 | ``[3 x [4 x i32]]`` | 3x4 array of 32-bit integer values. |
2251 +-----------------------------+----------------------------------------------------------+
2252 | ``[12 x [10 x float]]`` | 12x10 array of single precision floating point values. |
2253 +-----------------------------+----------------------------------------------------------+
2254 | ``[2 x [3 x [4 x i16]]]`` | 2x3x4 array of 16-bit integer values. |
2255 +-----------------------------+----------------------------------------------------------+
2257 There is no restriction on indexing beyond the end of the array implied
2258 by a static type (though there are restrictions on indexing beyond the
2259 bounds of an allocated object in some cases). This means that
2260 single-dimension 'variable sized array' addressing can be implemented in
2261 LLVM with a zero length array type. An implementation of 'pascal style
2262 arrays' in LLVM could use the type "``{ i32, [0 x float]}``", for
2272 The structure type is used to represent a collection of data members
2273 together in memory. The elements of a structure may be any type that has
2276 Structures in memory are accessed using '``load``' and '``store``' by
2277 getting a pointer to a field with the '``getelementptr``' instruction.
2278 Structures in registers are accessed using the '``extractvalue``' and
2279 '``insertvalue``' instructions.
2281 Structures may optionally be "packed" structures, which indicate that
2282 the alignment of the struct is one byte, and that there is no padding
2283 between the elements. In non-packed structs, padding between field types
2284 is inserted as defined by the DataLayout string in the module, which is
2285 required to match what the underlying code generator expects.
2287 Structures can either be "literal" or "identified". A literal structure
2288 is defined inline with other types (e.g. ``{i32, i32}*``) whereas
2289 identified types are always defined at the top level with a name.
2290 Literal types are uniqued by their contents and can never be recursive
2291 or opaque since there is no way to write one. Identified types can be
2292 recursive, can be opaqued, and are never uniqued.
2298 %T1 = type { <type list> } ; Identified normal struct type
2299 %T2 = type <{ <type list> }> ; Identified packed struct type
2303 +------------------------------+---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
2304 | ``{ i32, i32, i32 }`` | A triple of three ``i32`` values |
2305 +------------------------------+---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
2306 | ``{ 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``. |
2307 +------------------------------+---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
2308 | ``<{ i8, i32 }>`` | A packed struct known to be 5 bytes in size. |
2309 +------------------------------+---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
2313 Opaque Structure Types
2314 """"""""""""""""""""""
2318 Opaque structure types are used to represent named structure types that
2319 do not have a body specified. This corresponds (for example) to the C
2320 notion of a forward declared structure.
2331 +--------------+-------------------+
2332 | ``opaque`` | An opaque type. |
2333 +--------------+-------------------+
2340 LLVM has several different basic types of constants. This section
2341 describes them all and their syntax.
2346 **Boolean constants**
2347 The two strings '``true``' and '``false``' are both valid constants
2349 **Integer constants**
2350 Standard integers (such as '4') are constants of the
2351 :ref:`integer <t_integer>` type. Negative numbers may be used with
2353 **Floating point constants**
2354 Floating point constants use standard decimal notation (e.g.
2355 123.421), exponential notation (e.g. 1.23421e+2), or a more precise
2356 hexadecimal notation (see below). The assembler requires the exact
2357 decimal value of a floating-point constant. For example, the
2358 assembler accepts 1.25 but rejects 1.3 because 1.3 is a repeating
2359 decimal in binary. Floating point constants must have a :ref:`floating
2360 point <t_floating>` type.
2361 **Null pointer constants**
2362 The identifier '``null``' is recognized as a null pointer constant
2363 and must be of :ref:`pointer type <t_pointer>`.
2365 The one non-intuitive notation for constants is the hexadecimal form of
2366 floating point constants. For example, the form
2367 '``double 0x432ff973cafa8000``' is equivalent to (but harder to read
2368 than) '``double 4.5e+15``'. The only time hexadecimal floating point
2369 constants are required (and the only time that they are generated by the
2370 disassembler) is when a floating point constant must be emitted but it
2371 cannot be represented as a decimal floating point number in a reasonable
2372 number of digits. For example, NaN's, infinities, and other special
2373 values are represented in their IEEE hexadecimal format so that assembly
2374 and disassembly do not cause any bits to change in the constants.
2376 When using the hexadecimal form, constants of types half, float, and
2377 double are represented using the 16-digit form shown above (which
2378 matches the IEEE754 representation for double); half and float values
2379 must, however, be exactly representable as IEEE 754 half and single
2380 precision, respectively. Hexadecimal format is always used for long
2381 double, and there are three forms of long double. The 80-bit format used
2382 by x86 is represented as ``0xK`` followed by 20 hexadecimal digits. The
2383 128-bit format used by PowerPC (two adjacent doubles) is represented by
2384 ``0xM`` followed by 32 hexadecimal digits. The IEEE 128-bit format is
2385 represented by ``0xL`` followed by 32 hexadecimal digits. Long doubles
2386 will only work if they match the long double format on your target.
2387 The IEEE 16-bit format (half precision) is represented by ``0xH``
2388 followed by 4 hexadecimal digits. All hexadecimal formats are big-endian
2389 (sign bit at the left).
2391 There are no constants of type x86_mmx.
2393 .. _complexconstants:
2398 Complex constants are a (potentially recursive) combination of simple
2399 constants and smaller complex constants.
2401 **Structure constants**
2402 Structure constants are represented with notation similar to
2403 structure type definitions (a comma separated list of elements,
2404 surrounded by braces (``{}``)). For example:
2405 "``{ i32 4, float 17.0, i32* @G }``", where "``@G``" is declared as
2406 "``@G = external global i32``". Structure constants must have
2407 :ref:`structure type <t_struct>`, and the number and types of elements
2408 must match those specified by the type.
2410 Array constants are represented with notation similar to array type
2411 definitions (a comma separated list of elements, surrounded by
2412 square brackets (``[]``)). For example:
2413 "``[ i32 42, i32 11, i32 74 ]``". Array constants must have
2414 :ref:`array type <t_array>`, and the number and types of elements must
2415 match those specified by the type. As a special case, character array
2416 constants may also be represented as a double-quoted string using the ``c``
2417 prefix. For example: "``c"Hello World\0A\00"``".
2418 **Vector constants**
2419 Vector constants are represented with notation similar to vector
2420 type definitions (a comma separated list of elements, surrounded by
2421 less-than/greater-than's (``<>``)). For example:
2422 "``< i32 42, i32 11, i32 74, i32 100 >``". Vector constants
2423 must have :ref:`vector type <t_vector>`, and the number and types of
2424 elements must match those specified by the type.
2425 **Zero initialization**
2426 The string '``zeroinitializer``' can be used to zero initialize a
2427 value to zero of *any* type, including scalar and
2428 :ref:`aggregate <t_aggregate>` types. This is often used to avoid
2429 having to print large zero initializers (e.g. for large arrays) and
2430 is always exactly equivalent to using explicit zero initializers.
2432 A metadata node is a constant tuple without types. For example:
2433 "``!{!0, !{!2, !0}, !"test"}``". Metadata can reference constant values,
2434 for example: "``!{!0, i32 0, i8* @global, i64 (i64)* @function, !"str"}``".
2435 Unlike other typed constants that are meant to be interpreted as part of
2436 the instruction stream, metadata is a place to attach additional
2437 information such as debug info.
2439 Global Variable and Function Addresses
2440 --------------------------------------
2442 The addresses of :ref:`global variables <globalvars>` and
2443 :ref:`functions <functionstructure>` are always implicitly valid
2444 (link-time) constants. These constants are explicitly referenced when
2445 the :ref:`identifier for the global <identifiers>` is used and always have
2446 :ref:`pointer <t_pointer>` type. For example, the following is a legal LLVM
2449 .. code-block:: llvm
2453 @Z = global [2 x i32*] [ i32* @X, i32* @Y ]
2460 The string '``undef``' can be used anywhere a constant is expected, and
2461 indicates that the user of the value may receive an unspecified
2462 bit-pattern. Undefined values may be of any type (other than '``label``'
2463 or '``void``') and be used anywhere a constant is permitted.
2465 Undefined values are useful because they indicate to the compiler that
2466 the program is well defined no matter what value is used. This gives the
2467 compiler more freedom to optimize. Here are some examples of
2468 (potentially surprising) transformations that are valid (in pseudo IR):
2470 .. code-block:: llvm
2480 This is safe because all of the output bits are affected by the undef
2481 bits. Any output bit can have a zero or one depending on the input bits.
2483 .. code-block:: llvm
2494 These logical operations have bits that are not always affected by the
2495 input. For example, if ``%X`` has a zero bit, then the output of the
2496 '``and``' operation will always be a zero for that bit, no matter what
2497 the corresponding bit from the '``undef``' is. As such, it is unsafe to
2498 optimize or assume that the result of the '``and``' is '``undef``'.
2499 However, it is safe to assume that all bits of the '``undef``' could be
2500 0, and optimize the '``and``' to 0. Likewise, it is safe to assume that
2501 all the bits of the '``undef``' operand to the '``or``' could be set,
2502 allowing the '``or``' to be folded to -1.
2504 .. code-block:: llvm
2506 %A = select undef, %X, %Y
2507 %B = select undef, 42, %Y
2508 %C = select %X, %Y, undef
2518 This set of examples shows that undefined '``select``' (and conditional
2519 branch) conditions can go *either way*, but they have to come from one
2520 of the two operands. In the ``%A`` example, if ``%X`` and ``%Y`` were
2521 both known to have a clear low bit, then ``%A`` would have to have a
2522 cleared low bit. However, in the ``%C`` example, the optimizer is
2523 allowed to assume that the '``undef``' operand could be the same as
2524 ``%Y``, allowing the whole '``select``' to be eliminated.
2526 .. code-block:: llvm
2528 %A = xor undef, undef
2545 This example points out that two '``undef``' operands are not
2546 necessarily the same. This can be surprising to people (and also matches
2547 C semantics) where they assume that "``X^X``" is always zero, even if
2548 ``X`` is undefined. This isn't true for a number of reasons, but the
2549 short answer is that an '``undef``' "variable" can arbitrarily change
2550 its value over its "live range". This is true because the variable
2551 doesn't actually *have a live range*. Instead, the value is logically
2552 read from arbitrary registers that happen to be around when needed, so
2553 the value is not necessarily consistent over time. In fact, ``%A`` and
2554 ``%C`` need to have the same semantics or the core LLVM "replace all
2555 uses with" concept would not hold.
2557 .. code-block:: llvm
2565 These examples show the crucial difference between an *undefined value*
2566 and *undefined behavior*. An undefined value (like '``undef``') is
2567 allowed to have an arbitrary bit-pattern. This means that the ``%A``
2568 operation can be constant folded to '``undef``', because the '``undef``'
2569 could be an SNaN, and ``fdiv`` is not (currently) defined on SNaN's.
2570 However, in the second example, we can make a more aggressive
2571 assumption: because the ``undef`` is allowed to be an arbitrary value,
2572 we are allowed to assume that it could be zero. Since a divide by zero
2573 has *undefined behavior*, we are allowed to assume that the operation
2574 does not execute at all. This allows us to delete the divide and all
2575 code after it. Because the undefined operation "can't happen", the
2576 optimizer can assume that it occurs in dead code.
2578 .. code-block:: llvm
2580 a: store undef -> %X
2581 b: store %X -> undef
2586 These examples reiterate the ``fdiv`` example: a store *of* an undefined
2587 value can be assumed to not have any effect; we can assume that the
2588 value is overwritten with bits that happen to match what was already
2589 there. However, a store *to* an undefined location could clobber
2590 arbitrary memory, therefore, it has undefined behavior.
2597 Poison values are similar to :ref:`undef values <undefvalues>`, however
2598 they also represent the fact that an instruction or constant expression
2599 that cannot evoke side effects has nevertheless detected a condition
2600 that results in undefined behavior.
2602 There is currently no way of representing a poison value in the IR; they
2603 only exist when produced by operations such as :ref:`add <i_add>` with
2606 Poison value behavior is defined in terms of value *dependence*:
2608 - Values other than :ref:`phi <i_phi>` nodes depend on their operands.
2609 - :ref:`Phi <i_phi>` nodes depend on the operand corresponding to
2610 their dynamic predecessor basic block.
2611 - Function arguments depend on the corresponding actual argument values
2612 in the dynamic callers of their functions.
2613 - :ref:`Call <i_call>` instructions depend on the :ref:`ret <i_ret>`
2614 instructions that dynamically transfer control back to them.
2615 - :ref:`Invoke <i_invoke>` instructions depend on the
2616 :ref:`ret <i_ret>`, :ref:`resume <i_resume>`, or exception-throwing
2617 call instructions that dynamically transfer control back to them.
2618 - Non-volatile loads and stores depend on the most recent stores to all
2619 of the referenced memory addresses, following the order in the IR
2620 (including loads and stores implied by intrinsics such as
2621 :ref:`@llvm.memcpy <int_memcpy>`.)
2622 - An instruction with externally visible side effects depends on the
2623 most recent preceding instruction with externally visible side
2624 effects, following the order in the IR. (This includes :ref:`volatile
2625 operations <volatile>`.)
2626 - An instruction *control-depends* on a :ref:`terminator
2627 instruction <terminators>` if the terminator instruction has
2628 multiple successors and the instruction is always executed when
2629 control transfers to one of the successors, and may not be executed
2630 when control is transferred to another.
2631 - Additionally, an instruction also *control-depends* on a terminator
2632 instruction if the set of instructions it otherwise depends on would
2633 be different if the terminator had transferred control to a different
2635 - Dependence is transitive.
2637 Poison values have the same behavior as :ref:`undef values <undefvalues>`,
2638 with the additional effect that any instruction that has a *dependence*
2639 on a poison value has undefined behavior.
2641 Here are some examples:
2643 .. code-block:: llvm
2646 %poison = sub nuw i32 0, 1 ; Results in a poison value.
2647 %still_poison = and i32 %poison, 0 ; 0, but also poison.
2648 %poison_yet_again = getelementptr i32, i32* @h, i32 %still_poison
2649 store i32 0, i32* %poison_yet_again ; memory at @h[0] is poisoned
2651 store i32 %poison, i32* @g ; Poison value stored to memory.
2652 %poison2 = load i32, i32* @g ; Poison value loaded back from memory.
2654 store volatile i32 %poison, i32* @g ; External observation; undefined behavior.
2656 %narrowaddr = bitcast i32* @g to i16*
2657 %wideaddr = bitcast i32* @g to i64*
2658 %poison3 = load i16, i16* %narrowaddr ; Returns a poison value.
2659 %poison4 = load i64, i64* %wideaddr ; Returns a poison value.
2661 %cmp = icmp slt i32 %poison, 0 ; Returns a poison value.
2662 br i1 %cmp, label %true, label %end ; Branch to either destination.
2665 store volatile i32 0, i32* @g ; This is control-dependent on %cmp, so
2666 ; it has undefined behavior.
2670 %p = phi i32 [ 0, %entry ], [ 1, %true ]
2671 ; Both edges into this PHI are
2672 ; control-dependent on %cmp, so this
2673 ; always results in a poison value.
2675 store volatile i32 0, i32* @g ; This would depend on the store in %true
2676 ; if %cmp is true, or the store in %entry
2677 ; otherwise, so this is undefined behavior.
2679 br i1 %cmp, label %second_true, label %second_end
2680 ; The same branch again, but this time the
2681 ; true block doesn't have side effects.
2688 store volatile i32 0, i32* @g ; This time, the instruction always depends
2689 ; on the store in %end. Also, it is
2690 ; control-equivalent to %end, so this is
2691 ; well-defined (ignoring earlier undefined
2692 ; behavior in this example).
2696 Addresses of Basic Blocks
2697 -------------------------
2699 ``blockaddress(@function, %block)``
2701 The '``blockaddress``' constant computes the address of the specified
2702 basic block in the specified function, and always has an ``i8*`` type.
2703 Taking the address of the entry block is illegal.
2705 This value only has defined behavior when used as an operand to the
2706 ':ref:`indirectbr <i_indirectbr>`' instruction, or for comparisons
2707 against null. Pointer equality tests between labels addresses results in
2708 undefined behavior --- though, again, comparison against null is ok, and
2709 no label is equal to the null pointer. This may be passed around as an
2710 opaque pointer sized value as long as the bits are not inspected. This
2711 allows ``ptrtoint`` and arithmetic to be performed on these values so
2712 long as the original value is reconstituted before the ``indirectbr``
2715 Finally, some targets may provide defined semantics when using the value
2716 as the operand to an inline assembly, but that is target specific.
2720 Constant Expressions
2721 --------------------
2723 Constant expressions are used to allow expressions involving other
2724 constants to be used as constants. Constant expressions may be of any
2725 :ref:`first class <t_firstclass>` type and may involve any LLVM operation
2726 that does not have side effects (e.g. load and call are not supported).
2727 The following is the syntax for constant expressions:
2729 ``trunc (CST to TYPE)``
2730 Truncate a constant to another type. The bit size of CST must be
2731 larger than the bit size of TYPE. Both types must be integers.
2732 ``zext (CST to TYPE)``
2733 Zero extend a constant to another type. The bit size of CST must be
2734 smaller than the bit size of TYPE. Both types must be integers.
2735 ``sext (CST to TYPE)``
2736 Sign extend a constant to another type. The bit size of CST must be
2737 smaller than the bit size of TYPE. Both types must be integers.
2738 ``fptrunc (CST to TYPE)``
2739 Truncate a floating point constant to another floating point type.
2740 The size of CST must be larger than the size of TYPE. Both types
2741 must be floating point.
2742 ``fpext (CST to TYPE)``
2743 Floating point extend a constant to another type. The size of CST
2744 must be smaller or equal to the size of TYPE. Both types must be
2746 ``fptoui (CST to TYPE)``
2747 Convert a floating point constant to the corresponding unsigned
2748 integer constant. TYPE must be a scalar or vector integer type. CST
2749 must be of scalar or vector floating point type. Both CST and TYPE
2750 must be scalars, or vectors of the same number of elements. If the
2751 value won't fit in the integer type, the results are undefined.
2752 ``fptosi (CST to TYPE)``
2753 Convert a floating point constant to the corresponding signed
2754 integer constant. TYPE must be a scalar or vector integer type. CST
2755 must be of scalar or vector floating point type. Both CST and TYPE
2756 must be scalars, or vectors of the same number of elements. If the
2757 value won't fit in the integer type, the results are undefined.
2758 ``uitofp (CST to TYPE)``
2759 Convert an unsigned integer constant to the corresponding floating
2760 point constant. TYPE must be a scalar or vector floating point type.
2761 CST must be of scalar or vector integer type. Both CST and TYPE must
2762 be scalars, or vectors of the same number of elements. If the value
2763 won't fit in the floating point type, the results are undefined.
2764 ``sitofp (CST to TYPE)``
2765 Convert a signed integer constant to the corresponding floating
2766 point constant. TYPE must be a scalar or vector floating point type.
2767 CST must be of scalar or vector integer type. Both CST and TYPE must
2768 be scalars, or vectors of the same number of elements. If the value
2769 won't fit in the floating point type, the results are undefined.
2770 ``ptrtoint (CST to TYPE)``
2771 Convert a pointer typed constant to the corresponding integer
2772 constant. ``TYPE`` must be an integer type. ``CST`` must be of
2773 pointer type. The ``CST`` value is zero extended, truncated, or
2774 unchanged to make it fit in ``TYPE``.
2775 ``inttoptr (CST to TYPE)``
2776 Convert an integer constant to a pointer constant. TYPE must be a
2777 pointer type. CST must be of integer type. The CST value is zero
2778 extended, truncated, or unchanged to make it fit in a pointer size.
2779 This one is *really* dangerous!
2780 ``bitcast (CST to TYPE)``
2781 Convert a constant, CST, to another TYPE. The constraints of the
2782 operands are the same as those for the :ref:`bitcast
2783 instruction <i_bitcast>`.
2784 ``addrspacecast (CST to TYPE)``
2785 Convert a constant pointer or constant vector of pointer, CST, to another
2786 TYPE in a different address space. The constraints of the operands are the
2787 same as those for the :ref:`addrspacecast instruction <i_addrspacecast>`.
2788 ``getelementptr (TY, CSTPTR, IDX0, IDX1, ...)``, ``getelementptr inbounds (TY, CSTPTR, IDX0, IDX1, ...)``
2789 Perform the :ref:`getelementptr operation <i_getelementptr>` on
2790 constants. As with the :ref:`getelementptr <i_getelementptr>`
2791 instruction, the index list may have zero or more indexes, which are
2792 required to make sense for the type of "pointer to TY".
2793 ``select (COND, VAL1, VAL2)``
2794 Perform the :ref:`select operation <i_select>` on constants.
2795 ``icmp COND (VAL1, VAL2)``
2796 Performs the :ref:`icmp operation <i_icmp>` on constants.
2797 ``fcmp COND (VAL1, VAL2)``
2798 Performs the :ref:`fcmp operation <i_fcmp>` on constants.
2799 ``extractelement (VAL, IDX)``
2800 Perform the :ref:`extractelement operation <i_extractelement>` on
2802 ``insertelement (VAL, ELT, IDX)``
2803 Perform the :ref:`insertelement operation <i_insertelement>` on
2805 ``shufflevector (VEC1, VEC2, IDXMASK)``
2806 Perform the :ref:`shufflevector operation <i_shufflevector>` on
2808 ``extractvalue (VAL, IDX0, IDX1, ...)``
2809 Perform the :ref:`extractvalue operation <i_extractvalue>` on
2810 constants. The index list is interpreted in a similar manner as
2811 indices in a ':ref:`getelementptr <i_getelementptr>`' operation. At
2812 least one index value must be specified.
2813 ``insertvalue (VAL, ELT, IDX0, IDX1, ...)``
2814 Perform the :ref:`insertvalue operation <i_insertvalue>` on constants.
2815 The index list is interpreted in a similar manner as indices in a
2816 ':ref:`getelementptr <i_getelementptr>`' operation. At least one index
2817 value must be specified.
2818 ``OPCODE (LHS, RHS)``
2819 Perform the specified operation of the LHS and RHS constants. OPCODE
2820 may be any of the :ref:`binary <binaryops>` or :ref:`bitwise
2821 binary <bitwiseops>` operations. The constraints on operands are
2822 the same as those for the corresponding instruction (e.g. no bitwise
2823 operations on floating point values are allowed).
2830 Inline Assembler Expressions
2831 ----------------------------
2833 LLVM supports inline assembler expressions (as opposed to :ref:`Module-Level
2834 Inline Assembly <moduleasm>`) through the use of a special value. This value
2835 represents the inline assembler as a template string (containing the
2836 instructions to emit), a list of operand constraints (stored as a string), a
2837 flag that indicates whether or not the inline asm expression has side effects,
2838 and a flag indicating whether the function containing the asm needs to align its
2839 stack conservatively.
2841 The template string supports argument substitution of the operands using "``$``"
2842 followed by a number, to indicate substitution of the given register/memory
2843 location, as specified by the constraint string. "``${NUM:MODIFIER}``" may also
2844 be used, where ``MODIFIER`` is a target-specific annotation for how to print the
2845 operand (See :ref:`inline-asm-modifiers`).
2847 A literal "``$``" may be included by using "``$$``" in the template. To include
2848 other special characters into the output, the usual "``\XX``" escapes may be
2849 used, just as in other strings. Note that after template substitution, the
2850 resulting assembly string is parsed by LLVM's integrated assembler unless it is
2851 disabled -- even when emitting a ``.s`` file -- and thus must contain assembly
2852 syntax known to LLVM.
2854 LLVM's support for inline asm is modeled closely on the requirements of Clang's
2855 GCC-compatible inline-asm support. Thus, the feature-set and the constraint and
2856 modifier codes listed here are similar or identical to those in GCC's inline asm
2857 support. However, to be clear, the syntax of the template and constraint strings
2858 described here is *not* the same as the syntax accepted by GCC and Clang, and,
2859 while most constraint letters are passed through as-is by Clang, some get
2860 translated to other codes when converting from the C source to the LLVM
2863 An example inline assembler expression is:
2865 .. code-block:: llvm
2867 i32 (i32) asm "bswap $0", "=r,r"
2869 Inline assembler expressions may **only** be used as the callee operand
2870 of a :ref:`call <i_call>` or an :ref:`invoke <i_invoke>` instruction.
2871 Thus, typically we have:
2873 .. code-block:: llvm
2875 %X = call i32 asm "bswap $0", "=r,r"(i32 %Y)
2877 Inline asms with side effects not visible in the constraint list must be
2878 marked as having side effects. This is done through the use of the
2879 '``sideeffect``' keyword, like so:
2881 .. code-block:: llvm
2883 call void asm sideeffect "eieio", ""()
2885 In some cases inline asms will contain code that will not work unless
2886 the stack is aligned in some way, such as calls or SSE instructions on
2887 x86, yet will not contain code that does that alignment within the asm.
2888 The compiler should make conservative assumptions about what the asm
2889 might contain and should generate its usual stack alignment code in the
2890 prologue if the '``alignstack``' keyword is present:
2892 .. code-block:: llvm
2894 call void asm alignstack "eieio", ""()
2896 Inline asms also support using non-standard assembly dialects. The
2897 assumed dialect is ATT. When the '``inteldialect``' keyword is present,
2898 the inline asm is using the Intel dialect. Currently, ATT and Intel are
2899 the only supported dialects. An example is:
2901 .. code-block:: llvm
2903 call void asm inteldialect "eieio", ""()
2905 If multiple keywords appear the '``sideeffect``' keyword must come
2906 first, the '``alignstack``' keyword second and the '``inteldialect``'
2909 Inline Asm Constraint String
2910 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
2912 The constraint list is a comma-separated string, each element containing one or
2913 more constraint codes.
2915 For each element in the constraint list an appropriate register or memory
2916 operand will be chosen, and it will be made available to assembly template
2917 string expansion as ``$0`` for the first constraint in the list, ``$1`` for the
2920 There are three different types of constraints, which are distinguished by a
2921 prefix symbol in front of the constraint code: Output, Input, and Clobber. The
2922 constraints must always be given in that order: outputs first, then inputs, then
2923 clobbers. They cannot be intermingled.
2925 There are also three different categories of constraint codes:
2927 - Register constraint. This is either a register class, or a fixed physical
2928 register. This kind of constraint will allocate a register, and if necessary,
2929 bitcast the argument or result to the appropriate type.
2930 - Memory constraint. This kind of constraint is for use with an instruction
2931 taking a memory operand. Different constraints allow for different addressing
2932 modes used by the target.
2933 - Immediate value constraint. This kind of constraint is for an integer or other
2934 immediate value which can be rendered directly into an instruction. The
2935 various target-specific constraints allow the selection of a value in the
2936 proper range for the instruction you wish to use it with.
2941 Output constraints are specified by an "``=``" prefix (e.g. "``=r``"). This
2942 indicates that the assembly will write to this operand, and the operand will
2943 then be made available as a return value of the ``asm`` expression. Output
2944 constraints do not consume an argument from the call instruction. (Except, see
2945 below about indirect outputs).
2947 Normally, it is expected that no output locations are written to by the assembly
2948 expression until *all* of the inputs have been read. As such, LLVM may assign
2949 the same register to an output and an input. If this is not safe (e.g. if the
2950 assembly contains two instructions, where the first writes to one output, and
2951 the second reads an input and writes to a second output), then the "``&``"
2952 modifier must be used (e.g. "``=&r``") to specify that the output is an
2953 "early-clobber" output. Marking an ouput as "early-clobber" ensures that LLVM
2954 will not use the same register for any inputs (other than an input tied to this
2960 Input constraints do not have a prefix -- just the constraint codes. Each input
2961 constraint will consume one argument from the call instruction. It is not
2962 permitted for the asm to write to any input register or memory location (unless
2963 that input is tied to an output). Note also that multiple inputs may all be
2964 assigned to the same register, if LLVM can determine that they necessarily all
2965 contain the same value.
2967 Instead of providing a Constraint Code, input constraints may also "tie"
2968 themselves to an output constraint, by providing an integer as the constraint
2969 string. Tied inputs still consume an argument from the call instruction, and
2970 take up a position in the asm template numbering as is usual -- they will simply
2971 be constrained to always use the same register as the output they've been tied
2972 to. For example, a constraint string of "``=r,0``" says to assign a register for
2973 output, and use that register as an input as well (it being the 0'th
2976 It is permitted to tie an input to an "early-clobber" output. In that case, no
2977 *other* input may share the same register as the input tied to the early-clobber
2978 (even when the other input has the same value).
2980 You may only tie an input to an output which has a register constraint, not a
2981 memory constraint. Only a single input may be tied to an output.
2983 There is also an "interesting" feature which deserves a bit of explanation: if a
2984 register class constraint allocates a register which is too small for the value
2985 type operand provided as input, the input value will be split into multiple
2986 registers, and all of them passed to the inline asm.
2988 However, this feature is often not as useful as you might think.
2990 Firstly, the registers are *not* guaranteed to be consecutive. So, on those
2991 architectures that have instructions which operate on multiple consecutive
2992 instructions, this is not an appropriate way to support them. (e.g. the 32-bit
2993 SparcV8 has a 64-bit load, which instruction takes a single 32-bit register. The
2994 hardware then loads into both the named register, and the next register. This
2995 feature of inline asm would not be useful to support that.)
2997 A few of the targets provide a template string modifier allowing explicit access
2998 to the second register of a two-register operand (e.g. MIPS ``L``, ``M``, and
2999 ``D``). On such an architecture, you can actually access the second allocated
3000 register (yet, still, not any subsequent ones). But, in that case, you're still
3001 probably better off simply splitting the value into two separate operands, for
3002 clarity. (e.g. see the description of the ``A`` constraint on X86, which,
3003 despite existing only for use with this feature, is not really a good idea to
3006 Indirect inputs and outputs
3007 """""""""""""""""""""""""""
3009 Indirect output or input constraints can be specified by the "``*``" modifier
3010 (which goes after the "``=``" in case of an output). This indicates that the asm
3011 will write to or read from the contents of an *address* provided as an input
3012 argument. (Note that in this way, indirect outputs act more like an *input* than
3013 an output: just like an input, they consume an argument of the call expression,
3014 rather than producing a return value. An indirect output constraint is an
3015 "output" only in that the asm is expected to write to the contents of the input
3016 memory location, instead of just read from it).
3018 This is most typically used for memory constraint, e.g. "``=*m``", to pass the
3019 address of a variable as a value.
3021 It is also possible to use an indirect *register* constraint, but only on output
3022 (e.g. "``=*r``"). This will cause LLVM to allocate a register for an output
3023 value normally, and then, separately emit a store to the address provided as
3024 input, after the provided inline asm. (It's not clear what value this
3025 functionality provides, compared to writing the store explicitly after the asm
3026 statement, and it can only produce worse code, since it bypasses many
3027 optimization passes. I would recommend not using it.)
3033 A clobber constraint is indicated by a "``~``" prefix. A clobber does not
3034 consume an input operand, nor generate an output. Clobbers cannot use any of the
3035 general constraint code letters -- they may use only explicit register
3036 constraints, e.g. "``~{eax}``". The one exception is that a clobber string of
3037 "``~{memory}``" indicates that the assembly writes to arbitrary undeclared
3038 memory locations -- not only the memory pointed to by a declared indirect
3044 After a potential prefix comes constraint code, or codes.
3046 A Constraint Code is either a single letter (e.g. "``r``"), a "``^``" character
3047 followed by two letters (e.g. "``^wc``"), or "``{``" register-name "``}``"
3050 The one and two letter constraint codes are typically chosen to be the same as
3051 GCC's constraint codes.
3053 A single constraint may include one or more than constraint code in it, leaving
3054 it up to LLVM to choose which one to use. This is included mainly for
3055 compatibility with the translation of GCC inline asm coming from clang.
3057 There are two ways to specify alternatives, and either or both may be used in an
3058 inline asm constraint list:
3060 1) Append the codes to each other, making a constraint code set. E.g. "``im``"
3061 or "``{eax}m``". This means "choose any of the options in the set". The
3062 choice of constraint is made independently for each constraint in the
3065 2) Use "``|``" between constraint code sets, creating alternatives. Every
3066 constraint in the constraint list must have the same number of alternative
3067 sets. With this syntax, the same alternative in *all* of the items in the
3068 constraint list will be chosen together.
3070 Putting those together, you might have a two operand constraint string like
3071 ``"rm|r,ri|rm"``. This indicates that if operand 0 is ``r`` or ``m``, then
3072 operand 1 may be one of ``r`` or ``i``. If operand 0 is ``r``, then operand 1
3073 may be one of ``r`` or ``m``. But, operand 0 and 1 cannot both be of type m.
3075 However, the use of either of the alternatives features is *NOT* recommended, as
3076 LLVM is not able to make an intelligent choice about which one to use. (At the
3077 point it currently needs to choose, not enough information is available to do so
3078 in a smart way.) Thus, it simply tries to make a choice that's most likely to
3079 compile, not one that will be optimal performance. (e.g., given "``rm``", it'll
3080 always choose to use memory, not registers). And, if given multiple registers,
3081 or multiple register classes, it will simply choose the first one. (In fact, it
3082 doesn't currently even ensure explicitly specified physical registers are
3083 unique, so specifying multiple physical registers as alternatives, like
3084 ``{r11}{r12},{r11}{r12}``, will assign r11 to both operands, not at all what was
3087 Supported Constraint Code List
3088 """"""""""""""""""""""""""""""
3090 The constraint codes are, in general, expected to behave the same way they do in
3091 GCC. LLVM's support is often implemented on an 'as-needed' basis, to support C
3092 inline asm code which was supported by GCC. A mismatch in behavior between LLVM
3093 and GCC likely indicates a bug in LLVM.
3095 Some constraint codes are typically supported by all targets:
3097 - ``r``: A register in the target's general purpose register class.
3098 - ``m``: A memory address operand. It is target-specific what addressing modes
3099 are supported, typical examples are register, or register + register offset,
3100 or register + immediate offset (of some target-specific size).
3101 - ``i``: An integer constant (of target-specific width). Allows either a simple
3102 immediate, or a relocatable value.
3103 - ``n``: An integer constant -- *not* including relocatable values.
3104 - ``s``: An integer constant, but allowing *only* relocatable values.
3105 - ``X``: Allows an operand of any kind, no constraint whatsoever. Typically
3106 useful to pass a label for an asm branch or call.
3108 .. FIXME: but that surely isn't actually okay to jump out of an asm
3109 block without telling llvm about the control transfer???)
3111 - ``{register-name}``: Requires exactly the named physical register.
3113 Other constraints are target-specific:
3117 - ``z``: An immediate integer 0. Outputs ``WZR`` or ``XZR``, as appropriate.
3118 - ``I``: An immediate integer valid for an ``ADD`` or ``SUB`` instruction,
3119 i.e. 0 to 4095 with optional shift by 12.
3120 - ``J``: An immediate integer that, when negated, is valid for an ``ADD`` or
3121 ``SUB`` instruction, i.e. -1 to -4095 with optional left shift by 12.
3122 - ``K``: An immediate integer that is valid for the 'bitmask immediate 32' of a
3123 logical instruction like ``AND``, ``EOR``, or ``ORR`` with a 32-bit register.
3124 - ``L``: An immediate integer that is valid for the 'bitmask immediate 64' of a
3125 logical instruction like ``AND``, ``EOR``, or ``ORR`` with a 64-bit register.
3126 - ``M``: An immediate integer for use with the ``MOV`` assembly alias on a
3127 32-bit register. This is a superset of ``K``: in addition to the bitmask
3128 immediate, also allows immediate integers which can be loaded with a single
3129 ``MOVZ`` or ``MOVL`` instruction.
3130 - ``N``: An immediate integer for use with the ``MOV`` assembly alias on a
3131 64-bit register. This is a superset of ``L``.
3132 - ``Q``: Memory address operand must be in a single register (no
3133 offsets). (However, LLVM currently does this for the ``m`` constraint as
3135 - ``r``: A 32 or 64-bit integer register (W* or X*).
3136 - ``w``: A 32, 64, or 128-bit floating-point/SIMD register.
3137 - ``x``: A lower 128-bit floating-point/SIMD register (``V0`` to ``V15``).
3141 - ``r``: A 32 or 64-bit integer register.
3142 - ``[0-9]v``: The 32-bit VGPR register, number 0-9.
3143 - ``[0-9]s``: The 32-bit SGPR register, number 0-9.
3148 - ``Q``, ``Um``, ``Un``, ``Uq``, ``Us``, ``Ut``, ``Uv``, ``Uy``: Memory address
3149 operand. Treated the same as operand ``m``, at the moment.
3151 ARM and ARM's Thumb2 mode:
3153 - ``j``: An immediate integer between 0 and 65535 (valid for ``MOVW``)
3154 - ``I``: An immediate integer valid for a data-processing instruction.
3155 - ``J``: An immediate integer between -4095 and 4095.
3156 - ``K``: An immediate integer whose bitwise inverse is valid for a
3157 data-processing instruction. (Can be used with template modifier "``B``" to
3158 print the inverted value).
3159 - ``L``: An immediate integer whose negation is valid for a data-processing
3160 instruction. (Can be used with template modifier "``n``" to print the negated
3162 - ``M``: A power of two or a integer between 0 and 32.
3163 - ``N``: Invalid immediate constraint.
3164 - ``O``: Invalid immediate constraint.
3165 - ``r``: A general-purpose 32-bit integer register (``r0-r15``).
3166 - ``l``: In Thumb2 mode, low 32-bit GPR registers (``r0-r7``). In ARM mode, same
3168 - ``h``: In Thumb2 mode, a high 32-bit GPR register (``r8-r15``). In ARM mode,
3170 - ``w``: A 32, 64, or 128-bit floating-point/SIMD register: ``s0-s31``,
3171 ``d0-d31``, or ``q0-q15``.
3172 - ``x``: A 32, 64, or 128-bit floating-point/SIMD register: ``s0-s15``,
3173 ``d0-d7``, or ``q0-q3``.
3174 - ``t``: A floating-point/SIMD register, only supports 32-bit values:
3179 - ``I``: An immediate integer between 0 and 255.
3180 - ``J``: An immediate integer between -255 and -1.
3181 - ``K``: An immediate integer between 0 and 255, with optional left-shift by
3183 - ``L``: An immediate integer between -7 and 7.
3184 - ``M``: An immediate integer which is a multiple of 4 between 0 and 1020.
3185 - ``N``: An immediate integer between 0 and 31.
3186 - ``O``: An immediate integer which is a multiple of 4 between -508 and 508.
3187 - ``r``: A low 32-bit GPR register (``r0-r7``).
3188 - ``l``: A low 32-bit GPR register (``r0-r7``).
3189 - ``h``: A high GPR register (``r0-r7``).
3190 - ``w``: A 32, 64, or 128-bit floating-point/SIMD register: ``s0-s31``,
3191 ``d0-d31``, or ``q0-q15``.
3192 - ``x``: A 32, 64, or 128-bit floating-point/SIMD register: ``s0-s15``,
3193 ``d0-d7``, or ``q0-q3``.
3194 - ``t``: A floating-point/SIMD register, only supports 32-bit values:
3200 - ``o``, ``v``: A memory address operand, treated the same as constraint ``m``,
3202 - ``r``: A 32 or 64-bit register.
3206 - ``r``: An 8 or 16-bit register.
3210 - ``I``: An immediate signed 16-bit integer.
3211 - ``J``: An immediate integer zero.
3212 - ``K``: An immediate unsigned 16-bit integer.
3213 - ``L``: An immediate 32-bit integer, where the lower 16 bits are 0.
3214 - ``N``: An immediate integer between -65535 and -1.
3215 - ``O``: An immediate signed 15-bit integer.
3216 - ``P``: An immediate integer between 1 and 65535.
3217 - ``m``: A memory address operand. In MIPS-SE mode, allows a base address
3218 register plus 16-bit immediate offset. In MIPS mode, just a base register.
3219 - ``R``: A memory address operand. In MIPS-SE mode, allows a base address
3220 register plus a 9-bit signed offset. In MIPS mode, the same as constraint
3222 - ``ZC``: A memory address operand, suitable for use in a ``pref``, ``ll``, or
3223 ``sc`` instruction on the given subtarget (details vary).
3224 - ``r``, ``d``, ``y``: A 32 or 64-bit GPR register.
3225 - ``f``: A 32 or 64-bit FPU register (``F0-F31``), or a 128-bit MSA register
3226 (``W0-W31``). In the case of MSA registers, it is recommended to use the ``w``
3227 argument modifier for compatibility with GCC.
3228 - ``c``: A 32-bit or 64-bit GPR register suitable for indirect jump (always
3230 - ``l``: The ``lo`` register, 32 or 64-bit.
3235 - ``b``: A 1-bit integer register.
3236 - ``c`` or ``h``: A 16-bit integer register.
3237 - ``r``: A 32-bit integer register.
3238 - ``l`` or ``N``: A 64-bit integer register.
3239 - ``f``: A 32-bit float register.
3240 - ``d``: A 64-bit float register.
3245 - ``I``: An immediate signed 16-bit integer.
3246 - ``J``: An immediate unsigned 16-bit integer, shifted left 16 bits.
3247 - ``K``: An immediate unsigned 16-bit integer.
3248 - ``L``: An immediate signed 16-bit integer, shifted left 16 bits.
3249 - ``M``: An immediate integer greater than 31.
3250 - ``N``: An immediate integer that is an exact power of 2.
3251 - ``O``: The immediate integer constant 0.
3252 - ``P``: An immediate integer constant whose negation is a signed 16-bit
3254 - ``es``, ``o``, ``Q``, ``Z``, ``Zy``: A memory address operand, currently
3255 treated the same as ``m``.
3256 - ``r``: A 32 or 64-bit integer register.
3257 - ``b``: A 32 or 64-bit integer register, excluding ``R0`` (that is:
3259 - ``f``: A 32 or 64-bit float register (``F0-F31``), or when QPX is enabled, a
3260 128 or 256-bit QPX register (``Q0-Q31``; aliases the ``F`` registers).
3261 - ``v``: For ``4 x f32`` or ``4 x f64`` types, when QPX is enabled, a
3262 128 or 256-bit QPX register (``Q0-Q31``), otherwise a 128-bit
3263 altivec vector register (``V0-V31``).
3265 .. FIXME: is this a bug that v accepts QPX registers? I think this
3266 is supposed to only use the altivec vector registers?
3268 - ``y``: Condition register (``CR0-CR7``).
3269 - ``wc``: An individual CR bit in a CR register.
3270 - ``wa``, ``wd``, ``wf``: Any 128-bit VSX vector register, from the full VSX
3271 register set (overlapping both the floating-point and vector register files).
3272 - ``ws``: A 32 or 64-bit floating point register, from the full VSX register
3277 - ``I``: An immediate 13-bit signed integer.
3278 - ``r``: A 32-bit integer register.
3282 - ``I``: An immediate unsigned 8-bit integer.
3283 - ``J``: An immediate unsigned 12-bit integer.
3284 - ``K``: An immediate signed 16-bit integer.
3285 - ``L``: An immediate signed 20-bit integer.
3286 - ``M``: An immediate integer 0x7fffffff.
3287 - ``Q``, ``R``, ``S``, ``T``: A memory address operand, treated the same as
3288 ``m``, at the moment.
3289 - ``r`` or ``d``: A 32, 64, or 128-bit integer register.
3290 - ``a``: A 32, 64, or 128-bit integer address register (excludes R0, which in an
3291 address context evaluates as zero).
3292 - ``h``: A 32-bit value in the high part of a 64bit data register
3294 - ``f``: A 32, 64, or 128-bit floating point register.
3298 - ``I``: An immediate integer between 0 and 31.
3299 - ``J``: An immediate integer between 0 and 64.
3300 - ``K``: An immediate signed 8-bit integer.
3301 - ``L``: An immediate integer, 0xff or 0xffff or (in 64-bit mode only)
3303 - ``M``: An immediate integer between 0 and 3.
3304 - ``N``: An immediate unsigned 8-bit integer.
3305 - ``O``: An immediate integer between 0 and 127.
3306 - ``e``: An immediate 32-bit signed integer.
3307 - ``Z``: An immediate 32-bit unsigned integer.
3308 - ``o``, ``v``: Treated the same as ``m``, at the moment.
3309 - ``q``: An 8, 16, 32, or 64-bit register which can be accessed as an 8-bit
3310 ``l`` integer register. On X86-32, this is the ``a``, ``b``, ``c``, and ``d``
3311 registers, and on X86-64, it is all of the integer registers.
3312 - ``Q``: An 8, 16, 32, or 64-bit register which can be accessed as an 8-bit
3313 ``h`` integer register. This is the ``a``, ``b``, ``c``, and ``d`` registers.
3314 - ``r`` or ``l``: An 8, 16, 32, or 64-bit integer register.
3315 - ``R``: An 8, 16, 32, or 64-bit "legacy" integer register -- one which has
3316 existed since i386, and can be accessed without the REX prefix.
3317 - ``f``: A 32, 64, or 80-bit '387 FPU stack pseudo-register.
3318 - ``y``: A 64-bit MMX register, if MMX is enabled.
3319 - ``x``: If SSE is enabled: a 32 or 64-bit scalar operand, or 128-bit vector
3320 operand in a SSE register. If AVX is also enabled, can also be a 256-bit
3321 vector operand in an AVX register. If AVX-512 is also enabled, can also be a
3322 512-bit vector operand in an AVX512 register, Otherwise, an error.
3323 - ``Y``: The same as ``x``, if *SSE2* is enabled, otherwise an error.
3324 - ``A``: Special case: allocates EAX first, then EDX, for a single operand (in
3325 32-bit mode, a 64-bit integer operand will get split into two registers). It
3326 is not recommended to use this constraint, as in 64-bit mode, the 64-bit
3327 operand will get allocated only to RAX -- if two 32-bit operands are needed,
3328 you're better off splitting it yourself, before passing it to the asm
3333 - ``r``: A 32-bit integer register.
3336 .. _inline-asm-modifiers:
3338 Asm template argument modifiers
3339 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
3341 In the asm template string, modifiers can be used on the operand reference, like
3344 The modifiers are, in general, expected to behave the same way they do in
3345 GCC. LLVM's support is often implemented on an 'as-needed' basis, to support C
3346 inline asm code which was supported by GCC. A mismatch in behavior between LLVM
3347 and GCC likely indicates a bug in LLVM.
3351 - ``c``: Print an immediate integer constant unadorned, without
3352 the target-specific immediate punctuation (e.g. no ``$`` prefix).
3353 - ``n``: Negate and print immediate integer constant unadorned, without the
3354 target-specific immediate punctuation (e.g. no ``$`` prefix).
3355 - ``l``: Print as an unadorned label, without the target-specific label
3356 punctuation (e.g. no ``$`` prefix).
3360 - ``w``: Print a GPR register with a ``w*`` name instead of ``x*`` name. E.g.,
3361 instead of ``x30``, print ``w30``.
3362 - ``x``: Print a GPR register with a ``x*`` name. (this is the default, anyhow).
3363 - ``b``, ``h``, ``s``, ``d``, ``q``: Print a floating-point/SIMD register with a
3364 ``b*``, ``h*``, ``s*``, ``d*``, or ``q*`` name, rather than the default of
3373 - ``a``: Print an operand as an address (with ``[`` and ``]`` surrounding a
3377 - ``y``: Print a VFP single-precision register as an indexed double (e.g. print
3378 as ``d4[1]`` instead of ``s9``)
3379 - ``B``: Bitwise invert and print an immediate integer constant without ``#``
3381 - ``L``: Print the low 16-bits of an immediate integer constant.
3382 - ``M``: Print as a register set suitable for ldm/stm. Also prints *all*
3383 register operands subsequent to the specified one (!), so use carefully.
3384 - ``Q``: Print the low-order register of a register-pair, or the low-order
3385 register of a two-register operand.
3386 - ``R``: Print the high-order register of a register-pair, or the high-order
3387 register of a two-register operand.
3388 - ``H``: Print the second register of a register-pair. (On a big-endian system,
3389 ``H`` is equivalent to ``Q``, and on little-endian system, ``H`` is equivalent
3392 .. FIXME: H doesn't currently support printing the second register
3393 of a two-register operand.
3395 - ``e``: Print the low doubleword register of a NEON quad register.
3396 - ``f``: Print the high doubleword register of a NEON quad register.
3397 - ``m``: Print the base register of a memory operand without the ``[`` and ``]``
3402 - ``L``: Print the second register of a two-register operand. Requires that it
3403 has been allocated consecutively to the first.
3405 .. FIXME: why is it restricted to consecutive ones? And there's
3406 nothing that ensures that happens, is there?
3408 - ``I``: Print the letter 'i' if the operand is an integer constant, otherwise
3409 nothing. Used to print 'addi' vs 'add' instructions.
3413 No additional modifiers.
3417 - ``X``: Print an immediate integer as hexadecimal
3418 - ``x``: Print the low 16 bits of an immediate integer as hexadecimal.
3419 - ``d``: Print an immediate integer as decimal.
3420 - ``m``: Subtract one and print an immediate integer as decimal.
3421 - ``z``: Print $0 if an immediate zero, otherwise print normally.
3422 - ``L``: Print the low-order register of a two-register operand, or prints the
3423 address of the low-order word of a double-word memory operand.
3425 .. FIXME: L seems to be missing memory operand support.
3427 - ``M``: Print the high-order register of a two-register operand, or prints the
3428 address of the high-order word of a double-word memory operand.
3430 .. FIXME: M seems to be missing memory operand support.
3432 - ``D``: Print the second register of a two-register operand, or prints the
3433 second word of a double-word memory operand. (On a big-endian system, ``D`` is
3434 equivalent to ``L``, and on little-endian system, ``D`` is equivalent to
3436 - ``w``: No effect. Provided for compatibility with GCC which requires this
3437 modifier in order to print MSA registers (``W0-W31``) with the ``f``
3446 - ``L``: Print the second register of a two-register operand. Requires that it
3447 has been allocated consecutively to the first.
3449 .. FIXME: why is it restricted to consecutive ones? And there's
3450 nothing that ensures that happens, is there?
3452 - ``I``: Print the letter 'i' if the operand is an integer constant, otherwise
3453 nothing. Used to print 'addi' vs 'add' instructions.
3454 - ``y``: For a memory operand, prints formatter for a two-register X-form
3455 instruction. (Currently always prints ``r0,OPERAND``).
3456 - ``U``: Prints 'u' if the memory operand is an update form, and nothing
3457 otherwise. (NOTE: LLVM does not support update form, so this will currently
3458 always print nothing)
3459 - ``X``: Prints 'x' if the memory operand is an indexed form. (NOTE: LLVM does
3460 not support indexed form, so this will currently always print nothing)
3468 SystemZ implements only ``n``, and does *not* support any of the other
3469 target-independent modifiers.
3473 - ``c``: Print an unadorned integer or symbol name. (The latter is
3474 target-specific behavior for this typically target-independent modifier).
3475 - ``A``: Print a register name with a '``*``' before it.
3476 - ``b``: Print an 8-bit register name (e.g. ``al``); do nothing on a memory
3478 - ``h``: Print the upper 8-bit register name (e.g. ``ah``); do nothing on a
3480 - ``w``: Print the 16-bit register name (e.g. ``ax``); do nothing on a memory
3482 - ``k``: Print the 32-bit register name (e.g. ``eax``); do nothing on a memory
3484 - ``q``: Print the 64-bit register name (e.g. ``rax``), if 64-bit registers are
3485 available, otherwise the 32-bit register name; do nothing on a memory operand.
3486 - ``n``: Negate and print an unadorned integer, or, for operands other than an
3487 immediate integer (e.g. a relocatable symbol expression), print a '-' before
3488 the operand. (The behavior for relocatable symbol expressions is a
3489 target-specific behavior for this typically target-independent modifier)
3490 - ``H``: Print a memory reference with additional offset +8.
3491 - ``P``: Print a memory reference or operand for use as the argument of a call
3492 instruction. (E.g. omit ``(rip)``, even though it's PC-relative.)
3496 No additional modifiers.
3502 The call instructions that wrap inline asm nodes may have a
3503 "``!srcloc``" MDNode attached to it that contains a list of constant
3504 integers. If present, the code generator will use the integer as the
3505 location cookie value when report errors through the ``LLVMContext``
3506 error reporting mechanisms. This allows a front-end to correlate backend
3507 errors that occur with inline asm back to the source code that produced
3510 .. code-block:: llvm
3512 call void asm sideeffect "something bad", ""(), !srcloc !42
3514 !42 = !{ i32 1234567 }
3516 It is up to the front-end to make sense of the magic numbers it places
3517 in the IR. If the MDNode contains multiple constants, the code generator
3518 will use the one that corresponds to the line of the asm that the error
3526 LLVM IR allows metadata to be attached to instructions in the program
3527 that can convey extra information about the code to the optimizers and
3528 code generator. One example application of metadata is source-level
3529 debug information. There are two metadata primitives: strings and nodes.
3531 Metadata does not have a type, and is not a value. If referenced from a
3532 ``call`` instruction, it uses the ``metadata`` type.
3534 All metadata are identified in syntax by a exclamation point ('``!``').
3536 .. _metadata-string:
3538 Metadata Nodes and Metadata Strings
3539 -----------------------------------
3541 A metadata string is a string surrounded by double quotes. It can
3542 contain any character by escaping non-printable characters with
3543 "``\xx``" where "``xx``" is the two digit hex code. For example:
3546 Metadata nodes are represented with notation similar to structure
3547 constants (a comma separated list of elements, surrounded by braces and
3548 preceded by an exclamation point). Metadata nodes can have any values as
3549 their operand. For example:
3551 .. code-block:: llvm
3553 !{ !"test\00", i32 10}
3555 Metadata nodes that aren't uniqued use the ``distinct`` keyword. For example:
3557 .. code-block:: llvm
3559 !0 = distinct !{!"test\00", i32 10}
3561 ``distinct`` nodes are useful when nodes shouldn't be merged based on their
3562 content. They can also occur when transformations cause uniquing collisions
3563 when metadata operands change.
3565 A :ref:`named metadata <namedmetadatastructure>` is a collection of
3566 metadata nodes, which can be looked up in the module symbol table. For
3569 .. code-block:: llvm
3573 Metadata can be used as function arguments. Here ``llvm.dbg.value``
3574 function is using two metadata arguments:
3576 .. code-block:: llvm
3578 call void @llvm.dbg.value(metadata !24, i64 0, metadata !25)
3580 Metadata can be attached with an instruction. Here metadata ``!21`` is
3581 attached to the ``add`` instruction using the ``!dbg`` identifier:
3583 .. code-block:: llvm
3585 %indvar.next = add i64 %indvar, 1, !dbg !21
3587 More information about specific metadata nodes recognized by the
3588 optimizers and code generator is found below.
3590 .. _specialized-metadata:
3592 Specialized Metadata Nodes
3593 ^^^^^^^^^^^^^^^^^^^^^^^^^^
3595 Specialized metadata nodes are custom data structures in metadata (as opposed
3596 to generic tuples). Their fields are labelled, and can be specified in any
3599 These aren't inherently debug info centric, but currently all the specialized
3600 metadata nodes are related to debug info.
3607 ``DICompileUnit`` nodes represent a compile unit. The ``enums:``,
3608 ``retainedTypes:``, ``subprograms:``, ``globals:`` and ``imports:`` fields are
3609 tuples containing the debug info to be emitted along with the compile unit,
3610 regardless of code optimizations (some nodes are only emitted if there are
3611 references to them from instructions).
3613 .. code-block:: llvm
3615 !0 = !DICompileUnit(language: DW_LANG_C99, file: !1, producer: "clang",
3616 isOptimized: true, flags: "-O2", runtimeVersion: 2,
3617 splitDebugFilename: "abc.debug", emissionKind: 1,
3618 enums: !2, retainedTypes: !3, subprograms: !4,
3619 globals: !5, imports: !6)
3621 Compile unit descriptors provide the root scope for objects declared in a
3622 specific compilation unit. File descriptors are defined using this scope.
3623 These descriptors are collected by a named metadata ``!llvm.dbg.cu``. They
3624 keep track of subprograms, global variables, type information, and imported
3625 entities (declarations and namespaces).
3632 ``DIFile`` nodes represent files. The ``filename:`` can include slashes.
3634 .. code-block:: llvm
3636 !0 = !DIFile(filename: "path/to/file", directory: "/path/to/dir")
3638 Files are sometimes used in ``scope:`` fields, and are the only valid target
3639 for ``file:`` fields.
3646 ``DIBasicType`` nodes represent primitive types, such as ``int``, ``bool`` and
3647 ``float``. ``tag:`` defaults to ``DW_TAG_base_type``.
3649 .. code-block:: llvm
3651 !0 = !DIBasicType(name: "unsigned char", size: 8, align: 8,
3652 encoding: DW_ATE_unsigned_char)
3653 !1 = !DIBasicType(tag: DW_TAG_unspecified_type, name: "decltype(nullptr)")
3655 The ``encoding:`` describes the details of the type. Usually it's one of the
3658 .. code-block:: llvm
3664 DW_ATE_signed_char = 6
3666 DW_ATE_unsigned_char = 8
3668 .. _DISubroutineType:
3673 ``DISubroutineType`` nodes represent subroutine types. Their ``types:`` field
3674 refers to a tuple; the first operand is the return type, while the rest are the
3675 types of the formal arguments in order. If the first operand is ``null``, that
3676 represents a function with no return value (such as ``void foo() {}`` in C++).
3678 .. code-block:: llvm
3680 !0 = !BasicType(name: "int", size: 32, align: 32, DW_ATE_signed)
3681 !1 = !BasicType(name: "char", size: 8, align: 8, DW_ATE_signed_char)
3682 !2 = !DISubroutineType(types: !{null, !0, !1}) ; void (int, char)
3689 ``DIDerivedType`` nodes represent types derived from other types, such as
3692 .. code-block:: llvm
3694 !0 = !DIBasicType(name: "unsigned char", size: 8, align: 8,
3695 encoding: DW_ATE_unsigned_char)
3696 !1 = !DIDerivedType(tag: DW_TAG_pointer_type, baseType: !0, size: 32,
3699 The following ``tag:`` values are valid:
3701 .. code-block:: llvm
3703 DW_TAG_formal_parameter = 5
3705 DW_TAG_pointer_type = 15
3706 DW_TAG_reference_type = 16
3708 DW_TAG_ptr_to_member_type = 31
3709 DW_TAG_const_type = 38
3710 DW_TAG_volatile_type = 53
3711 DW_TAG_restrict_type = 55
3713 ``DW_TAG_member`` is used to define a member of a :ref:`composite type
3714 <DICompositeType>` or :ref:`subprogram <DISubprogram>`. The type of the member
3715 is the ``baseType:``. The ``offset:`` is the member's bit offset.
3716 ``DW_TAG_formal_parameter`` is used to define a member which is a formal
3717 argument of a subprogram.
3719 ``DW_TAG_typedef`` is used to provide a name for the ``baseType:``.
3721 ``DW_TAG_pointer_type``, ``DW_TAG_reference_type``, ``DW_TAG_const_type``,
3722 ``DW_TAG_volatile_type`` and ``DW_TAG_restrict_type`` are used to qualify the
3725 Note that the ``void *`` type is expressed as a type derived from NULL.
3727 .. _DICompositeType:
3732 ``DICompositeType`` nodes represent types composed of other types, like
3733 structures and unions. ``elements:`` points to a tuple of the composed types.
3735 If the source language supports ODR, the ``identifier:`` field gives the unique
3736 identifier used for type merging between modules. When specified, other types
3737 can refer to composite types indirectly via a :ref:`metadata string
3738 <metadata-string>` that matches their identifier.
3740 .. code-block:: llvm
3742 !0 = !DIEnumerator(name: "SixKind", value: 7)
3743 !1 = !DIEnumerator(name: "SevenKind", value: 7)
3744 !2 = !DIEnumerator(name: "NegEightKind", value: -8)
3745 !3 = !DICompositeType(tag: DW_TAG_enumeration_type, name: "Enum", file: !12,
3746 line: 2, size: 32, align: 32, identifier: "_M4Enum",
3747 elements: !{!0, !1, !2})
3749 The following ``tag:`` values are valid:
3751 .. code-block:: llvm
3753 DW_TAG_array_type = 1
3754 DW_TAG_class_type = 2
3755 DW_TAG_enumeration_type = 4
3756 DW_TAG_structure_type = 19
3757 DW_TAG_union_type = 23
3758 DW_TAG_subroutine_type = 21
3759 DW_TAG_inheritance = 28
3762 For ``DW_TAG_array_type``, the ``elements:`` should be :ref:`subrange
3763 descriptors <DISubrange>`, each representing the range of subscripts at that
3764 level of indexing. The ``DIFlagVector`` flag to ``flags:`` indicates that an
3765 array type is a native packed vector.
3767 For ``DW_TAG_enumeration_type``, the ``elements:`` should be :ref:`enumerator
3768 descriptors <DIEnumerator>`, each representing the definition of an enumeration
3769 value for the set. All enumeration type descriptors are collected in the
3770 ``enums:`` field of the :ref:`compile unit <DICompileUnit>`.
3772 For ``DW_TAG_structure_type``, ``DW_TAG_class_type``, and
3773 ``DW_TAG_union_type``, the ``elements:`` should be :ref:`derived types
3774 <DIDerivedType>` with ``tag: DW_TAG_member`` or ``tag: DW_TAG_inheritance``.
3781 ``DISubrange`` nodes are the elements for ``DW_TAG_array_type`` variants of
3782 :ref:`DICompositeType`. ``count: -1`` indicates an empty array.
3784 .. code-block:: llvm
3786 !0 = !DISubrange(count: 5, lowerBound: 0) ; array counting from 0
3787 !1 = !DISubrange(count: 5, lowerBound: 1) ; array counting from 1
3788 !2 = !DISubrange(count: -1) ; empty array.
3795 ``DIEnumerator`` nodes are the elements for ``DW_TAG_enumeration_type``
3796 variants of :ref:`DICompositeType`.
3798 .. code-block:: llvm
3800 !0 = !DIEnumerator(name: "SixKind", value: 7)
3801 !1 = !DIEnumerator(name: "SevenKind", value: 7)
3802 !2 = !DIEnumerator(name: "NegEightKind", value: -8)
3804 DITemplateTypeParameter
3805 """""""""""""""""""""""
3807 ``DITemplateTypeParameter`` nodes represent type parameters to generic source
3808 language constructs. They are used (optionally) in :ref:`DICompositeType` and
3809 :ref:`DISubprogram` ``templateParams:`` fields.
3811 .. code-block:: llvm
3813 !0 = !DITemplateTypeParameter(name: "Ty", type: !1)
3815 DITemplateValueParameter
3816 """"""""""""""""""""""""
3818 ``DITemplateValueParameter`` nodes represent value parameters to generic source
3819 language constructs. ``tag:`` defaults to ``DW_TAG_template_value_parameter``,
3820 but if specified can also be set to ``DW_TAG_GNU_template_template_param`` or
3821 ``DW_TAG_GNU_template_param_pack``. They are used (optionally) in
3822 :ref:`DICompositeType` and :ref:`DISubprogram` ``templateParams:`` fields.
3824 .. code-block:: llvm
3826 !0 = !DITemplateValueParameter(name: "Ty", type: !1, value: i32 7)
3831 ``DINamespace`` nodes represent namespaces in the source language.
3833 .. code-block:: llvm
3835 !0 = !DINamespace(name: "myawesomeproject", scope: !1, file: !2, line: 7)
3840 ``DIGlobalVariable`` nodes represent global variables in the source language.
3842 .. code-block:: llvm
3844 !0 = !DIGlobalVariable(name: "foo", linkageName: "foo", scope: !1,
3845 file: !2, line: 7, type: !3, isLocal: true,
3846 isDefinition: false, variable: i32* @foo,
3849 All global variables should be referenced by the `globals:` field of a
3850 :ref:`compile unit <DICompileUnit>`.
3857 ``DISubprogram`` nodes represent functions from the source language. The
3858 ``variables:`` field points at :ref:`variables <DILocalVariable>` that must be
3859 retained, even if their IR counterparts are optimized out of the IR. The
3860 ``type:`` field must point at an :ref:`DISubroutineType`.
3862 .. code-block:: llvm
3864 !0 = !DISubprogram(name: "foo", linkageName: "_Zfoov", scope: !1,
3865 file: !2, line: 7, type: !3, isLocal: true,
3866 isDefinition: false, scopeLine: 8, containingType: !4,
3867 virtuality: DW_VIRTUALITY_pure_virtual, virtualIndex: 10,
3868 flags: DIFlagPrototyped, isOptimized: true,
3869 function: void ()* @_Z3foov,
3870 templateParams: !5, declaration: !6, variables: !7)
3877 ``DILexicalBlock`` nodes describe nested blocks within a :ref:`subprogram
3878 <DISubprogram>`. The line number and column numbers are used to dinstinguish
3879 two lexical blocks at same depth. They are valid targets for ``scope:``
3882 .. code-block:: llvm
3884 !0 = distinct !DILexicalBlock(scope: !1, file: !2, line: 7, column: 35)
3886 Usually lexical blocks are ``distinct`` to prevent node merging based on
3889 .. _DILexicalBlockFile:
3894 ``DILexicalBlockFile`` nodes are used to discriminate between sections of a
3895 :ref:`lexical block <DILexicalBlock>`. The ``file:`` field can be changed to
3896 indicate textual inclusion, or the ``discriminator:`` field can be used to
3897 discriminate between control flow within a single block in the source language.
3899 .. code-block:: llvm
3901 !0 = !DILexicalBlock(scope: !3, file: !4, line: 7, column: 35)
3902 !1 = !DILexicalBlockFile(scope: !0, file: !4, discriminator: 0)
3903 !2 = !DILexicalBlockFile(scope: !0, file: !4, discriminator: 1)
3910 ``DILocation`` nodes represent source debug locations. The ``scope:`` field is
3911 mandatory, and points at an :ref:`DILexicalBlockFile`, an
3912 :ref:`DILexicalBlock`, or an :ref:`DISubprogram`.
3914 .. code-block:: llvm
3916 !0 = !DILocation(line: 2900, column: 42, scope: !1, inlinedAt: !2)
3918 .. _DILocalVariable:
3923 ``DILocalVariable`` nodes represent local variables in the source language. If
3924 the ``arg:`` field is set to non-zero, then this variable is a subprogram
3925 parameter, and it will be included in the ``variables:`` field of its
3926 :ref:`DISubprogram`.
3928 .. code-block:: llvm
3930 !0 = !DILocalVariable(name: "this", arg: 1, scope: !3, file: !2, line: 7,
3931 type: !3, flags: DIFlagArtificial)
3932 !1 = !DILocalVariable(name: "x", arg: 2, scope: !4, file: !2, line: 7,
3934 !2 = !DILocalVariable(name: "y", scope: !5, file: !2, line: 7, type: !3)
3939 ``DIExpression`` nodes represent DWARF expression sequences. They are used in
3940 :ref:`debug intrinsics<dbg_intrinsics>` (such as ``llvm.dbg.declare``) to
3941 describe how the referenced LLVM variable relates to the source language
3944 The current supported vocabulary is limited:
3946 - ``DW_OP_deref`` dereferences the working expression.
3947 - ``DW_OP_plus, 93`` adds ``93`` to the working expression.
3948 - ``DW_OP_bit_piece, 16, 8`` specifies the offset and size (``16`` and ``8``
3949 here, respectively) of the variable piece from the working expression.
3951 .. code-block:: llvm
3953 !0 = !DIExpression(DW_OP_deref)
3954 !1 = !DIExpression(DW_OP_plus, 3)
3955 !2 = !DIExpression(DW_OP_bit_piece, 3, 7)
3956 !3 = !DIExpression(DW_OP_deref, DW_OP_plus, 3, DW_OP_bit_piece, 3, 7)
3961 ``DIObjCProperty`` nodes represent Objective-C property nodes.
3963 .. code-block:: llvm
3965 !3 = !DIObjCProperty(name: "foo", file: !1, line: 7, setter: "setFoo",
3966 getter: "getFoo", attributes: 7, type: !2)
3971 ``DIImportedEntity`` nodes represent entities (such as modules) imported into a
3974 .. code-block:: llvm
3976 !2 = !DIImportedEntity(tag: DW_TAG_imported_module, name: "foo", scope: !0,
3977 entity: !1, line: 7)
3982 In LLVM IR, memory does not have types, so LLVM's own type system is not
3983 suitable for doing TBAA. Instead, metadata is added to the IR to
3984 describe a type system of a higher level language. This can be used to
3985 implement typical C/C++ TBAA, but it can also be used to implement
3986 custom alias analysis behavior for other languages.
3988 The current metadata format is very simple. TBAA metadata nodes have up
3989 to three fields, e.g.:
3991 .. code-block:: llvm
3993 !0 = !{ !"an example type tree" }
3994 !1 = !{ !"int", !0 }
3995 !2 = !{ !"float", !0 }
3996 !3 = !{ !"const float", !2, i64 1 }
3998 The first field is an identity field. It can be any value, usually a
3999 metadata string, which uniquely identifies the type. The most important
4000 name in the tree is the name of the root node. Two trees with different
4001 root node names are entirely disjoint, even if they have leaves with
4004 The second field identifies the type's parent node in the tree, or is
4005 null or omitted for a root node. A type is considered to alias all of
4006 its descendants and all of its ancestors in the tree. Also, a type is
4007 considered to alias all types in other trees, so that bitcode produced
4008 from multiple front-ends is handled conservatively.
4010 If the third field is present, it's an integer which if equal to 1
4011 indicates that the type is "constant" (meaning
4012 ``pointsToConstantMemory`` should return true; see `other useful
4013 AliasAnalysis methods <AliasAnalysis.html#OtherItfs>`_).
4015 '``tbaa.struct``' Metadata
4016 ^^^^^^^^^^^^^^^^^^^^^^^^^^
4018 The :ref:`llvm.memcpy <int_memcpy>` is often used to implement
4019 aggregate assignment operations in C and similar languages, however it
4020 is defined to copy a contiguous region of memory, which is more than
4021 strictly necessary for aggregate types which contain holes due to
4022 padding. Also, it doesn't contain any TBAA information about the fields
4025 ``!tbaa.struct`` metadata can describe which memory subregions in a
4026 memcpy are padding and what the TBAA tags of the struct are.
4028 The current metadata format is very simple. ``!tbaa.struct`` metadata
4029 nodes are a list of operands which are in conceptual groups of three.
4030 For each group of three, the first operand gives the byte offset of a
4031 field in bytes, the second gives its size in bytes, and the third gives
4034 .. code-block:: llvm
4036 !4 = !{ i64 0, i64 4, !1, i64 8, i64 4, !2 }
4038 This describes a struct with two fields. The first is at offset 0 bytes
4039 with size 4 bytes, and has tbaa tag !1. The second is at offset 8 bytes
4040 and has size 4 bytes and has tbaa tag !2.
4042 Note that the fields need not be contiguous. In this example, there is a
4043 4 byte gap between the two fields. This gap represents padding which
4044 does not carry useful data and need not be preserved.
4046 '``noalias``' and '``alias.scope``' Metadata
4047 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
4049 ``noalias`` and ``alias.scope`` metadata provide the ability to specify generic
4050 noalias memory-access sets. This means that some collection of memory access
4051 instructions (loads, stores, memory-accessing calls, etc.) that carry
4052 ``noalias`` metadata can specifically be specified not to alias with some other
4053 collection of memory access instructions that carry ``alias.scope`` metadata.
4054 Each type of metadata specifies a list of scopes where each scope has an id and
4055 a domain. When evaluating an aliasing query, if for some domain, the set
4056 of scopes with that domain in one instruction's ``alias.scope`` list is a
4057 subset of (or equal to) the set of scopes for that domain in another
4058 instruction's ``noalias`` list, then the two memory accesses are assumed not to
4061 The metadata identifying each domain is itself a list containing one or two
4062 entries. The first entry is the name of the domain. Note that if the name is a
4063 string then it can be combined accross functions and translation units. A
4064 self-reference can be used to create globally unique domain names. A
4065 descriptive string may optionally be provided as a second list entry.
4067 The metadata identifying each scope is also itself a list containing two or
4068 three entries. The first entry is the name of the scope. Note that if the name
4069 is a string then it can be combined accross functions and translation units. A
4070 self-reference can be used to create globally unique scope names. A metadata
4071 reference to the scope's domain is the second entry. A descriptive string may
4072 optionally be provided as a third list entry.
4076 .. code-block:: llvm
4078 ; Two scope domains:
4082 ; Some scopes in these domains:
4088 !5 = !{!4} ; A list containing only scope !4
4092 ; These two instructions don't alias:
4093 %0 = load float, float* %c, align 4, !alias.scope !5
4094 store float %0, float* %arrayidx.i, align 4, !noalias !5
4096 ; These two instructions also don't alias (for domain !1, the set of scopes
4097 ; in the !alias.scope equals that in the !noalias list):
4098 %2 = load float, float* %c, align 4, !alias.scope !5
4099 store float %2, float* %arrayidx.i2, align 4, !noalias !6
4101 ; These two instructions may alias (for domain !0, the set of scopes in
4102 ; the !noalias list is not a superset of, or equal to, the scopes in the
4103 ; !alias.scope list):
4104 %2 = load float, float* %c, align 4, !alias.scope !6
4105 store float %0, float* %arrayidx.i, align 4, !noalias !7
4107 '``fpmath``' Metadata
4108 ^^^^^^^^^^^^^^^^^^^^^
4110 ``fpmath`` metadata may be attached to any instruction of floating point
4111 type. It can be used to express the maximum acceptable error in the
4112 result of that instruction, in ULPs, thus potentially allowing the
4113 compiler to use a more efficient but less accurate method of computing
4114 it. ULP is defined as follows:
4116 If ``x`` is a real number that lies between two finite consecutive
4117 floating-point numbers ``a`` and ``b``, without being equal to one
4118 of them, then ``ulp(x) = |b - a|``, otherwise ``ulp(x)`` is the
4119 distance between the two non-equal finite floating-point numbers
4120 nearest ``x``. Moreover, ``ulp(NaN)`` is ``NaN``.
4122 The metadata node shall consist of a single positive floating point
4123 number representing the maximum relative error, for example:
4125 .. code-block:: llvm
4127 !0 = !{ float 2.5 } ; maximum acceptable inaccuracy is 2.5 ULPs
4131 '``range``' Metadata
4132 ^^^^^^^^^^^^^^^^^^^^
4134 ``range`` metadata may be attached only to ``load``, ``call`` and ``invoke`` of
4135 integer types. It expresses the possible ranges the loaded value or the value
4136 returned by the called function at this call site is in. The ranges are
4137 represented with a flattened list of integers. The loaded value or the value
4138 returned is known to be in the union of the ranges defined by each consecutive
4139 pair. Each pair has the following properties:
4141 - The type must match the type loaded by the instruction.
4142 - The pair ``a,b`` represents the range ``[a,b)``.
4143 - Both ``a`` and ``b`` are constants.
4144 - The range is allowed to wrap.
4145 - The range should not represent the full or empty set. That is,
4148 In addition, the pairs must be in signed order of the lower bound and
4149 they must be non-contiguous.
4153 .. code-block:: llvm
4155 %a = load i8, i8* %x, align 1, !range !0 ; Can only be 0 or 1
4156 %b = load i8, i8* %y, align 1, !range !1 ; Can only be 255 (-1), 0 or 1
4157 %c = call i8 @foo(), !range !2 ; Can only be 0, 1, 3, 4 or 5
4158 %d = invoke i8 @bar() to label %cont
4159 unwind label %lpad, !range !3 ; Can only be -2, -1, 3, 4 or 5
4161 !0 = !{ i8 0, i8 2 }
4162 !1 = !{ i8 255, i8 2 }
4163 !2 = !{ i8 0, i8 2, i8 3, i8 6 }
4164 !3 = !{ i8 -2, i8 0, i8 3, i8 6 }
4166 '``unpredictable``' Metadata
4167 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
4169 ``unpredictable`` metadata may be attached to any branch or switch
4170 instruction. It can be used to express the unpredictability of control
4171 flow. Similar to the llvm.expect intrinsic, it may be used to alter
4172 optimizations related to compare and branch instructions. The metadata
4173 is treated as a boolean value; if it exists, it signals that the branch
4174 or switch that it is attached to is completely unpredictable.
4179 It is sometimes useful to attach information to loop constructs. Currently,
4180 loop metadata is implemented as metadata attached to the branch instruction
4181 in the loop latch block. This type of metadata refer to a metadata node that is
4182 guaranteed to be separate for each loop. The loop identifier metadata is
4183 specified with the name ``llvm.loop``.
4185 The loop identifier metadata is implemented using a metadata that refers to
4186 itself to avoid merging it with any other identifier metadata, e.g.,
4187 during module linkage or function inlining. That is, each loop should refer
4188 to their own identification metadata even if they reside in separate functions.
4189 The following example contains loop identifier metadata for two separate loop
4192 .. code-block:: llvm
4197 The loop identifier metadata can be used to specify additional
4198 per-loop metadata. Any operands after the first operand can be treated
4199 as user-defined metadata. For example the ``llvm.loop.unroll.count``
4200 suggests an unroll factor to the loop unroller:
4202 .. code-block:: llvm
4204 br i1 %exitcond, label %._crit_edge, label %.lr.ph, !llvm.loop !0
4207 !1 = !{!"llvm.loop.unroll.count", i32 4}
4209 '``llvm.loop.vectorize``' and '``llvm.loop.interleave``'
4210 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
4212 Metadata prefixed with ``llvm.loop.vectorize`` or ``llvm.loop.interleave`` are
4213 used to control per-loop vectorization and interleaving parameters such as
4214 vectorization width and interleave count. These metadata should be used in
4215 conjunction with ``llvm.loop`` loop identification metadata. The
4216 ``llvm.loop.vectorize`` and ``llvm.loop.interleave`` metadata are only
4217 optimization hints and the optimizer will only interleave and vectorize loops if
4218 it believes it is safe to do so. The ``llvm.mem.parallel_loop_access`` metadata
4219 which contains information about loop-carried memory dependencies can be helpful
4220 in determining the safety of these transformations.
4222 '``llvm.loop.interleave.count``' Metadata
4223 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
4225 This metadata suggests an interleave count to the loop interleaver.
4226 The first operand is the string ``llvm.loop.interleave.count`` and the
4227 second operand is an integer specifying the interleave count. For
4230 .. code-block:: llvm
4232 !0 = !{!"llvm.loop.interleave.count", i32 4}
4234 Note that setting ``llvm.loop.interleave.count`` to 1 disables interleaving
4235 multiple iterations of the loop. If ``llvm.loop.interleave.count`` is set to 0
4236 then the interleave count will be determined automatically.
4238 '``llvm.loop.vectorize.enable``' Metadata
4239 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
4241 This metadata selectively enables or disables vectorization for the loop. The
4242 first operand is the string ``llvm.loop.vectorize.enable`` and the second operand
4243 is a bit. If the bit operand value is 1 vectorization is enabled. A value of
4244 0 disables vectorization:
4246 .. code-block:: llvm
4248 !0 = !{!"llvm.loop.vectorize.enable", i1 0}
4249 !1 = !{!"llvm.loop.vectorize.enable", i1 1}
4251 '``llvm.loop.vectorize.width``' Metadata
4252 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
4254 This metadata sets the target width of the vectorizer. The first
4255 operand is the string ``llvm.loop.vectorize.width`` and the second
4256 operand is an integer specifying the width. For example:
4258 .. code-block:: llvm
4260 !0 = !{!"llvm.loop.vectorize.width", i32 4}
4262 Note that setting ``llvm.loop.vectorize.width`` to 1 disables
4263 vectorization of the loop. If ``llvm.loop.vectorize.width`` is set to
4264 0 or if the loop does not have this metadata the width will be
4265 determined automatically.
4267 '``llvm.loop.unroll``'
4268 ^^^^^^^^^^^^^^^^^^^^^^
4270 Metadata prefixed with ``llvm.loop.unroll`` are loop unrolling
4271 optimization hints such as the unroll factor. ``llvm.loop.unroll``
4272 metadata should be used in conjunction with ``llvm.loop`` loop
4273 identification metadata. The ``llvm.loop.unroll`` metadata are only
4274 optimization hints and the unrolling will only be performed if the
4275 optimizer believes it is safe to do so.
4277 '``llvm.loop.unroll.count``' Metadata
4278 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
4280 This metadata suggests an unroll factor to the loop unroller. The
4281 first operand is the string ``llvm.loop.unroll.count`` and the second
4282 operand is a positive integer specifying the unroll factor. For
4285 .. code-block:: llvm
4287 !0 = !{!"llvm.loop.unroll.count", i32 4}
4289 If the trip count of the loop is less than the unroll count the loop
4290 will be partially unrolled.
4292 '``llvm.loop.unroll.disable``' Metadata
4293 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
4295 This metadata disables loop unrolling. The metadata has a single operand
4296 which is the string ``llvm.loop.unroll.disable``. For example:
4298 .. code-block:: llvm
4300 !0 = !{!"llvm.loop.unroll.disable"}
4302 '``llvm.loop.unroll.runtime.disable``' Metadata
4303 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
4305 This metadata disables runtime loop unrolling. The metadata has a single
4306 operand which is the string ``llvm.loop.unroll.runtime.disable``. For example:
4308 .. code-block:: llvm
4310 !0 = !{!"llvm.loop.unroll.runtime.disable"}
4312 '``llvm.loop.unroll.enable``' Metadata
4313 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
4315 This metadata suggests that the loop should be fully unrolled if the trip count
4316 is known at compile time and partially unrolled if the trip count is not known
4317 at compile time. The metadata has a single operand which is the string
4318 ``llvm.loop.unroll.enable``. For example:
4320 .. code-block:: llvm
4322 !0 = !{!"llvm.loop.unroll.enable"}
4324 '``llvm.loop.unroll.full``' Metadata
4325 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
4327 This metadata suggests that the loop should be unrolled fully. The
4328 metadata has a single operand which is the string ``llvm.loop.unroll.full``.
4331 .. code-block:: llvm
4333 !0 = !{!"llvm.loop.unroll.full"}
4338 Metadata types used to annotate memory accesses with information helpful
4339 for optimizations are prefixed with ``llvm.mem``.
4341 '``llvm.mem.parallel_loop_access``' Metadata
4342 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
4344 The ``llvm.mem.parallel_loop_access`` metadata refers to a loop identifier,
4345 or metadata containing a list of loop identifiers for nested loops.
4346 The metadata is attached to memory accessing instructions and denotes that
4347 no loop carried memory dependence exist between it and other instructions denoted
4348 with the same loop identifier.
4350 Precisely, given two instructions ``m1`` and ``m2`` that both have the
4351 ``llvm.mem.parallel_loop_access`` metadata, with ``L1`` and ``L2`` being the
4352 set of loops associated with that metadata, respectively, then there is no loop
4353 carried dependence between ``m1`` and ``m2`` for loops in both ``L1`` and
4356 As a special case, if all memory accessing instructions in a loop have
4357 ``llvm.mem.parallel_loop_access`` metadata that refers to that loop, then the
4358 loop has no loop carried memory dependences and is considered to be a parallel
4361 Note that if not all memory access instructions have such metadata referring to
4362 the loop, then the loop is considered not being trivially parallel. Additional
4363 memory dependence analysis is required to make that determination. As a fail
4364 safe mechanism, this causes loops that were originally parallel to be considered
4365 sequential (if optimization passes that are unaware of the parallel semantics
4366 insert new memory instructions into the loop body).
4368 Example of a loop that is considered parallel due to its correct use of
4369 both ``llvm.loop`` and ``llvm.mem.parallel_loop_access``
4370 metadata types that refer to the same loop identifier metadata.
4372 .. code-block:: llvm
4376 %val0 = load i32, i32* %arrayidx, !llvm.mem.parallel_loop_access !0
4378 store i32 %val0, i32* %arrayidx1, !llvm.mem.parallel_loop_access !0
4380 br i1 %exitcond, label %for.end, label %for.body, !llvm.loop !0
4386 It is also possible to have nested parallel loops. In that case the
4387 memory accesses refer to a list of loop identifier metadata nodes instead of
4388 the loop identifier metadata node directly:
4390 .. code-block:: llvm
4394 %val1 = load i32, i32* %arrayidx3, !llvm.mem.parallel_loop_access !2
4396 br label %inner.for.body
4400 %val0 = load i32, i32* %arrayidx1, !llvm.mem.parallel_loop_access !0
4402 store i32 %val0, i32* %arrayidx2, !llvm.mem.parallel_loop_access !0
4404 br i1 %exitcond, label %inner.for.end, label %inner.for.body, !llvm.loop !1
4408 store i32 %val1, i32* %arrayidx4, !llvm.mem.parallel_loop_access !2
4410 br i1 %exitcond, label %outer.for.end, label %outer.for.body, !llvm.loop !2
4412 outer.for.end: ; preds = %for.body
4414 !0 = !{!1, !2} ; a list of loop identifiers
4415 !1 = !{!1} ; an identifier for the inner loop
4416 !2 = !{!2} ; an identifier for the outer loop
4421 The ``llvm.bitsets`` global metadata is used to implement
4422 :doc:`bitsets <BitSets>`.
4424 Module Flags Metadata
4425 =====================
4427 Information about the module as a whole is difficult to convey to LLVM's
4428 subsystems. The LLVM IR isn't sufficient to transmit this information.
4429 The ``llvm.module.flags`` named metadata exists in order to facilitate
4430 this. These flags are in the form of key / value pairs --- much like a
4431 dictionary --- making it easy for any subsystem who cares about a flag to
4434 The ``llvm.module.flags`` metadata contains a list of metadata triplets.
4435 Each triplet has the following form:
4437 - The first element is a *behavior* flag, which specifies the behavior
4438 when two (or more) modules are merged together, and it encounters two
4439 (or more) metadata with the same ID. The supported behaviors are
4441 - The second element is a metadata string that is a unique ID for the
4442 metadata. Each module may only have one flag entry for each unique ID (not
4443 including entries with the **Require** behavior).
4444 - The third element is the value of the flag.
4446 When two (or more) modules are merged together, the resulting
4447 ``llvm.module.flags`` metadata is the union of the modules' flags. That is, for
4448 each unique metadata ID string, there will be exactly one entry in the merged
4449 modules ``llvm.module.flags`` metadata table, and the value for that entry will
4450 be determined by the merge behavior flag, as described below. The only exception
4451 is that entries with the *Require* behavior are always preserved.
4453 The following behaviors are supported:
4464 Emits an error if two values disagree, otherwise the resulting value
4465 is that of the operands.
4469 Emits a warning if two values disagree. The result value will be the
4470 operand for the flag from the first module being linked.
4474 Adds a requirement that another module flag be present and have a
4475 specified value after linking is performed. The value must be a
4476 metadata pair, where the first element of the pair is the ID of the
4477 module flag to be restricted, and the second element of the pair is
4478 the value the module flag should be restricted to. This behavior can
4479 be used to restrict the allowable results (via triggering of an
4480 error) of linking IDs with the **Override** behavior.
4484 Uses the specified value, regardless of the behavior or value of the
4485 other module. If both modules specify **Override**, but the values
4486 differ, an error will be emitted.
4490 Appends the two values, which are required to be metadata nodes.
4494 Appends the two values, which are required to be metadata
4495 nodes. However, duplicate entries in the second list are dropped
4496 during the append operation.
4498 It is an error for a particular unique flag ID to have multiple behaviors,
4499 except in the case of **Require** (which adds restrictions on another metadata
4500 value) or **Override**.
4502 An example of module flags:
4504 .. code-block:: llvm
4506 !0 = !{ i32 1, !"foo", i32 1 }
4507 !1 = !{ i32 4, !"bar", i32 37 }
4508 !2 = !{ i32 2, !"qux", i32 42 }
4509 !3 = !{ i32 3, !"qux",
4514 !llvm.module.flags = !{ !0, !1, !2, !3 }
4516 - Metadata ``!0`` has the ID ``!"foo"`` and the value '1'. The behavior
4517 if two or more ``!"foo"`` flags are seen is to emit an error if their
4518 values are not equal.
4520 - Metadata ``!1`` has the ID ``!"bar"`` and the value '37'. The
4521 behavior if two or more ``!"bar"`` flags are seen is to use the value
4524 - Metadata ``!2`` has the ID ``!"qux"`` and the value '42'. The
4525 behavior if two or more ``!"qux"`` flags are seen is to emit a
4526 warning if their values are not equal.
4528 - Metadata ``!3`` has the ID ``!"qux"`` and the value:
4534 The behavior is to emit an error if the ``llvm.module.flags`` does not
4535 contain a flag with the ID ``!"foo"`` that has the value '1' after linking is
4538 Objective-C Garbage Collection Module Flags Metadata
4539 ----------------------------------------------------
4541 On the Mach-O platform, Objective-C stores metadata about garbage
4542 collection in a special section called "image info". The metadata
4543 consists of a version number and a bitmask specifying what types of
4544 garbage collection are supported (if any) by the file. If two or more
4545 modules are linked together their garbage collection metadata needs to
4546 be merged rather than appended together.
4548 The Objective-C garbage collection module flags metadata consists of the
4549 following key-value pairs:
4558 * - ``Objective-C Version``
4559 - **[Required]** --- The Objective-C ABI version. Valid values are 1 and 2.
4561 * - ``Objective-C Image Info Version``
4562 - **[Required]** --- The version of the image info section. Currently
4565 * - ``Objective-C Image Info Section``
4566 - **[Required]** --- The section to place the metadata. Valid values are
4567 ``"__OBJC, __image_info, regular"`` for Objective-C ABI version 1, and
4568 ``"__DATA,__objc_imageinfo, regular, no_dead_strip"`` for
4569 Objective-C ABI version 2.
4571 * - ``Objective-C Garbage Collection``
4572 - **[Required]** --- Specifies whether garbage collection is supported or
4573 not. Valid values are 0, for no garbage collection, and 2, for garbage
4574 collection supported.
4576 * - ``Objective-C GC Only``
4577 - **[Optional]** --- Specifies that only garbage collection is supported.
4578 If present, its value must be 6. This flag requires that the
4579 ``Objective-C Garbage Collection`` flag have the value 2.
4581 Some important flag interactions:
4583 - If a module with ``Objective-C Garbage Collection`` set to 0 is
4584 merged with a module with ``Objective-C Garbage Collection`` set to
4585 2, then the resulting module has the
4586 ``Objective-C Garbage Collection`` flag set to 0.
4587 - A module with ``Objective-C Garbage Collection`` set to 0 cannot be
4588 merged with a module with ``Objective-C GC Only`` set to 6.
4590 Automatic Linker Flags Module Flags Metadata
4591 --------------------------------------------
4593 Some targets support embedding flags to the linker inside individual object
4594 files. Typically this is used in conjunction with language extensions which
4595 allow source files to explicitly declare the libraries they depend on, and have
4596 these automatically be transmitted to the linker via object files.
4598 These flags are encoded in the IR using metadata in the module flags section,
4599 using the ``Linker Options`` key. The merge behavior for this flag is required
4600 to be ``AppendUnique``, and the value for the key is expected to be a metadata
4601 node which should be a list of other metadata nodes, each of which should be a
4602 list of metadata strings defining linker options.
4604 For example, the following metadata section specifies two separate sets of
4605 linker options, presumably to link against ``libz`` and the ``Cocoa``
4608 !0 = !{ i32 6, !"Linker Options",
4611 !{ !"-framework", !"Cocoa" } } }
4612 !llvm.module.flags = !{ !0 }
4614 The metadata encoding as lists of lists of options, as opposed to a collapsed
4615 list of options, is chosen so that the IR encoding can use multiple option
4616 strings to specify e.g., a single library, while still having that specifier be
4617 preserved as an atomic element that can be recognized by a target specific
4618 assembly writer or object file emitter.
4620 Each individual option is required to be either a valid option for the target's
4621 linker, or an option that is reserved by the target specific assembly writer or
4622 object file emitter. No other aspect of these options is defined by the IR.
4624 C type width Module Flags Metadata
4625 ----------------------------------
4627 The ARM backend emits a section into each generated object file describing the
4628 options that it was compiled with (in a compiler-independent way) to prevent
4629 linking incompatible objects, and to allow automatic library selection. Some
4630 of these options are not visible at the IR level, namely wchar_t width and enum
4633 To pass this information to the backend, these options are encoded in module
4634 flags metadata, using the following key-value pairs:
4644 - * 0 --- sizeof(wchar_t) == 4
4645 * 1 --- sizeof(wchar_t) == 2
4648 - * 0 --- Enums are at least as large as an ``int``.
4649 * 1 --- Enums are stored in the smallest integer type which can
4650 represent all of its values.
4652 For example, the following metadata section specifies that the module was
4653 compiled with a ``wchar_t`` width of 4 bytes, and the underlying type of an
4654 enum is the smallest type which can represent all of its values::
4656 !llvm.module.flags = !{!0, !1}
4657 !0 = !{i32 1, !"short_wchar", i32 1}
4658 !1 = !{i32 1, !"short_enum", i32 0}
4660 .. _intrinsicglobalvariables:
4662 Intrinsic Global Variables
4663 ==========================
4665 LLVM has a number of "magic" global variables that contain data that
4666 affect code generation or other IR semantics. These are documented here.
4667 All globals of this sort should have a section specified as
4668 "``llvm.metadata``". This section and all globals that start with
4669 "``llvm.``" are reserved for use by LLVM.
4673 The '``llvm.used``' Global Variable
4674 -----------------------------------
4676 The ``@llvm.used`` global is an array which has
4677 :ref:`appending linkage <linkage_appending>`. This array contains a list of
4678 pointers to named global variables, functions and aliases which may optionally
4679 have a pointer cast formed of bitcast or getelementptr. For example, a legal
4682 .. code-block:: llvm
4687 @llvm.used = appending global [2 x i8*] [
4689 i8* bitcast (i32* @Y to i8*)
4690 ], section "llvm.metadata"
4692 If a symbol appears in the ``@llvm.used`` list, then the compiler, assembler,
4693 and linker are required to treat the symbol as if there is a reference to the
4694 symbol that it cannot see (which is why they have to be named). For example, if
4695 a variable has internal linkage and no references other than that from the
4696 ``@llvm.used`` list, it cannot be deleted. This is commonly used to represent
4697 references from inline asms and other things the compiler cannot "see", and
4698 corresponds to "``attribute((used))``" in GNU C.
4700 On some targets, the code generator must emit a directive to the
4701 assembler or object file to prevent the assembler and linker from
4702 molesting the symbol.
4704 .. _gv_llvmcompilerused:
4706 The '``llvm.compiler.used``' Global Variable
4707 --------------------------------------------
4709 The ``@llvm.compiler.used`` directive is the same as the ``@llvm.used``
4710 directive, except that it only prevents the compiler from touching the
4711 symbol. On targets that support it, this allows an intelligent linker to
4712 optimize references to the symbol without being impeded as it would be
4715 This is a rare construct that should only be used in rare circumstances,
4716 and should not be exposed to source languages.
4718 .. _gv_llvmglobalctors:
4720 The '``llvm.global_ctors``' Global Variable
4721 -------------------------------------------
4723 .. code-block:: llvm
4725 %0 = type { i32, void ()*, i8* }
4726 @llvm.global_ctors = appending global [1 x %0] [%0 { i32 65535, void ()* @ctor, i8* @data }]
4728 The ``@llvm.global_ctors`` array contains a list of constructor
4729 functions, priorities, and an optional associated global or function.
4730 The functions referenced by this array will be called in ascending order
4731 of priority (i.e. lowest first) when the module is loaded. The order of
4732 functions with the same priority is not defined.
4734 If the third field is present, non-null, and points to a global variable
4735 or function, the initializer function will only run if the associated
4736 data from the current module is not discarded.
4738 .. _llvmglobaldtors:
4740 The '``llvm.global_dtors``' Global Variable
4741 -------------------------------------------
4743 .. code-block:: llvm
4745 %0 = type { i32, void ()*, i8* }
4746 @llvm.global_dtors = appending global [1 x %0] [%0 { i32 65535, void ()* @dtor, i8* @data }]
4748 The ``@llvm.global_dtors`` array contains a list of destructor
4749 functions, priorities, and an optional associated global or function.
4750 The functions referenced by this array will be called in descending
4751 order of priority (i.e. highest first) when the module is unloaded. The
4752 order of functions with the same priority is not defined.
4754 If the third field is present, non-null, and points to a global variable
4755 or function, the destructor function will only run if the associated
4756 data from the current module is not discarded.
4758 Instruction Reference
4759 =====================
4761 The LLVM instruction set consists of several different classifications
4762 of instructions: :ref:`terminator instructions <terminators>`, :ref:`binary
4763 instructions <binaryops>`, :ref:`bitwise binary
4764 instructions <bitwiseops>`, :ref:`memory instructions <memoryops>`, and
4765 :ref:`other instructions <otherops>`.
4769 Terminator Instructions
4770 -----------------------
4772 As mentioned :ref:`previously <functionstructure>`, every basic block in a
4773 program ends with a "Terminator" instruction, which indicates which
4774 block should be executed after the current block is finished. These
4775 terminator instructions typically yield a '``void``' value: they produce
4776 control flow, not values (the one exception being the
4777 ':ref:`invoke <i_invoke>`' instruction).
4779 The terminator instructions are: ':ref:`ret <i_ret>`',
4780 ':ref:`br <i_br>`', ':ref:`switch <i_switch>`',
4781 ':ref:`indirectbr <i_indirectbr>`', ':ref:`invoke <i_invoke>`',
4782 ':ref:`resume <i_resume>`', ':ref:`catchpad <i_catchpad>`',
4783 ':ref:`catchendpad <i_catchendpad>`',
4784 ':ref:`catchret <i_catchret>`',
4785 ':ref:`cleanupendpad <i_cleanupendpad>`',
4786 ':ref:`cleanupret <i_cleanupret>`',
4787 ':ref:`terminatepad <i_terminatepad>`',
4788 and ':ref:`unreachable <i_unreachable>`'.
4792 '``ret``' Instruction
4793 ^^^^^^^^^^^^^^^^^^^^^
4800 ret <type> <value> ; Return a value from a non-void function
4801 ret void ; Return from void function
4806 The '``ret``' instruction is used to return control flow (and optionally
4807 a value) from a function back to the caller.
4809 There are two forms of the '``ret``' instruction: one that returns a
4810 value and then causes control flow, and one that just causes control
4816 The '``ret``' instruction optionally accepts a single argument, the
4817 return value. The type of the return value must be a ':ref:`first
4818 class <t_firstclass>`' type.
4820 A function is not :ref:`well formed <wellformed>` if it it has a non-void
4821 return type and contains a '``ret``' instruction with no return value or
4822 a return value with a type that does not match its type, or if it has a
4823 void return type and contains a '``ret``' instruction with a return
4829 When the '``ret``' instruction is executed, control flow returns back to
4830 the calling function's context. If the caller is a
4831 ":ref:`call <i_call>`" instruction, execution continues at the
4832 instruction after the call. If the caller was an
4833 ":ref:`invoke <i_invoke>`" instruction, execution continues at the
4834 beginning of the "normal" destination block. If the instruction returns
4835 a value, that value shall set the call or invoke instruction's return
4841 .. code-block:: llvm
4843 ret i32 5 ; Return an integer value of 5
4844 ret void ; Return from a void function
4845 ret { i32, i8 } { i32 4, i8 2 } ; Return a struct of values 4 and 2
4849 '``br``' Instruction
4850 ^^^^^^^^^^^^^^^^^^^^
4857 br i1 <cond>, label <iftrue>, label <iffalse>
4858 br label <dest> ; Unconditional branch
4863 The '``br``' instruction is used to cause control flow to transfer to a
4864 different basic block in the current function. There are two forms of
4865 this instruction, corresponding to a conditional branch and an
4866 unconditional branch.
4871 The conditional branch form of the '``br``' instruction takes a single
4872 '``i1``' value and two '``label``' values. The unconditional form of the
4873 '``br``' instruction takes a single '``label``' value as a target.
4878 Upon execution of a conditional '``br``' instruction, the '``i1``'
4879 argument is evaluated. If the value is ``true``, control flows to the
4880 '``iftrue``' ``label`` argument. If "cond" is ``false``, control flows
4881 to the '``iffalse``' ``label`` argument.
4886 .. code-block:: llvm
4889 %cond = icmp eq i32 %a, %b
4890 br i1 %cond, label %IfEqual, label %IfUnequal
4898 '``switch``' Instruction
4899 ^^^^^^^^^^^^^^^^^^^^^^^^
4906 switch <intty> <value>, label <defaultdest> [ <intty> <val>, label <dest> ... ]
4911 The '``switch``' instruction is used to transfer control flow to one of
4912 several different places. It is a generalization of the '``br``'
4913 instruction, allowing a branch to occur to one of many possible
4919 The '``switch``' instruction uses three parameters: an integer
4920 comparison value '``value``', a default '``label``' destination, and an
4921 array of pairs of comparison value constants and '``label``'s. The table
4922 is not allowed to contain duplicate constant entries.
4927 The ``switch`` instruction specifies a table of values and destinations.
4928 When the '``switch``' instruction is executed, this table is searched
4929 for the given value. If the value is found, control flow is transferred
4930 to the corresponding destination; otherwise, control flow is transferred
4931 to the default destination.
4936 Depending on properties of the target machine and the particular
4937 ``switch`` instruction, this instruction may be code generated in
4938 different ways. For example, it could be generated as a series of
4939 chained conditional branches or with a lookup table.
4944 .. code-block:: llvm
4946 ; Emulate a conditional br instruction
4947 %Val = zext i1 %value to i32
4948 switch i32 %Val, label %truedest [ i32 0, label %falsedest ]
4950 ; Emulate an unconditional br instruction
4951 switch i32 0, label %dest [ ]
4953 ; Implement a jump table:
4954 switch i32 %val, label %otherwise [ i32 0, label %onzero
4956 i32 2, label %ontwo ]
4960 '``indirectbr``' Instruction
4961 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
4968 indirectbr <somety>* <address>, [ label <dest1>, label <dest2>, ... ]
4973 The '``indirectbr``' instruction implements an indirect branch to a
4974 label within the current function, whose address is specified by
4975 "``address``". Address must be derived from a
4976 :ref:`blockaddress <blockaddress>` constant.
4981 The '``address``' argument is the address of the label to jump to. The
4982 rest of the arguments indicate the full set of possible destinations
4983 that the address may point to. Blocks are allowed to occur multiple
4984 times in the destination list, though this isn't particularly useful.
4986 This destination list is required so that dataflow analysis has an
4987 accurate understanding of the CFG.
4992 Control transfers to the block specified in the address argument. All
4993 possible destination blocks must be listed in the label list, otherwise
4994 this instruction has undefined behavior. This implies that jumps to
4995 labels defined in other functions have undefined behavior as well.
5000 This is typically implemented with a jump through a register.
5005 .. code-block:: llvm
5007 indirectbr i8* %Addr, [ label %bb1, label %bb2, label %bb3 ]
5011 '``invoke``' Instruction
5012 ^^^^^^^^^^^^^^^^^^^^^^^^
5019 <result> = invoke [cconv] [ret attrs] <ptr to function ty> <function ptr val>(<function args>) [fn attrs]
5020 to label <normal label> unwind label <exception label>
5025 The '``invoke``' instruction causes control to transfer to a specified
5026 function, with the possibility of control flow transfer to either the
5027 '``normal``' label or the '``exception``' label. If the callee function
5028 returns with the "``ret``" instruction, control flow will return to the
5029 "normal" label. If the callee (or any indirect callees) returns via the
5030 ":ref:`resume <i_resume>`" instruction or other exception handling
5031 mechanism, control is interrupted and continued at the dynamically
5032 nearest "exception" label.
5034 The '``exception``' label is a `landing
5035 pad <ExceptionHandling.html#overview>`_ for the exception. As such,
5036 '``exception``' label is required to have the
5037 ":ref:`landingpad <i_landingpad>`" instruction, which contains the
5038 information about the behavior of the program after unwinding happens,
5039 as its first non-PHI instruction. The restrictions on the
5040 "``landingpad``" instruction's tightly couples it to the "``invoke``"
5041 instruction, so that the important information contained within the
5042 "``landingpad``" instruction can't be lost through normal code motion.
5047 This instruction requires several arguments:
5049 #. The optional "cconv" marker indicates which :ref:`calling
5050 convention <callingconv>` the call should use. If none is
5051 specified, the call defaults to using C calling conventions.
5052 #. The optional :ref:`Parameter Attributes <paramattrs>` list for return
5053 values. Only '``zeroext``', '``signext``', and '``inreg``' attributes
5055 #. '``ptr to function ty``': shall be the signature of the pointer to
5056 function value being invoked. In most cases, this is a direct
5057 function invocation, but indirect ``invoke``'s are just as possible,
5058 branching off an arbitrary pointer to function value.
5059 #. '``function ptr val``': An LLVM value containing a pointer to a
5060 function to be invoked.
5061 #. '``function args``': argument list whose types match the function
5062 signature argument types and parameter attributes. All arguments must
5063 be of :ref:`first class <t_firstclass>` type. If the function signature
5064 indicates the function accepts a variable number of arguments, the
5065 extra arguments can be specified.
5066 #. '``normal label``': the label reached when the called function
5067 executes a '``ret``' instruction.
5068 #. '``exception label``': the label reached when a callee returns via
5069 the :ref:`resume <i_resume>` instruction or other exception handling
5071 #. The optional :ref:`function attributes <fnattrs>` list. Only
5072 '``noreturn``', '``nounwind``', '``readonly``' and '``readnone``'
5073 attributes are valid here.
5078 This instruction is designed to operate as a standard '``call``'
5079 instruction in most regards. The primary difference is that it
5080 establishes an association with a label, which is used by the runtime
5081 library to unwind the stack.
5083 This instruction is used in languages with destructors to ensure that
5084 proper cleanup is performed in the case of either a ``longjmp`` or a
5085 thrown exception. Additionally, this is important for implementation of
5086 '``catch``' clauses in high-level languages that support them.
5088 For the purposes of the SSA form, the definition of the value returned
5089 by the '``invoke``' instruction is deemed to occur on the edge from the
5090 current block to the "normal" label. If the callee unwinds then no
5091 return value is available.
5096 .. code-block:: llvm
5098 %retval = invoke i32 @Test(i32 15) to label %Continue
5099 unwind label %TestCleanup ; i32:retval set
5100 %retval = invoke coldcc i32 %Testfnptr(i32 15) to label %Continue
5101 unwind label %TestCleanup ; i32:retval set
5105 '``resume``' Instruction
5106 ^^^^^^^^^^^^^^^^^^^^^^^^
5113 resume <type> <value>
5118 The '``resume``' instruction is a terminator instruction that has no
5124 The '``resume``' instruction requires one argument, which must have the
5125 same type as the result of any '``landingpad``' instruction in the same
5131 The '``resume``' instruction resumes propagation of an existing
5132 (in-flight) exception whose unwinding was interrupted with a
5133 :ref:`landingpad <i_landingpad>` instruction.
5138 .. code-block:: llvm
5140 resume { i8*, i32 } %exn
5144 '``catchpad``' Instruction
5145 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
5152 <resultval> = catchpad [<args>*]
5153 to label <normal label> unwind label <exception label>
5158 The '``catchpad``' instruction is used by `LLVM's exception handling
5159 system <ExceptionHandling.html#overview>`_ to specify that a basic block
5160 is a catch block --- one where a personality routine attempts to transfer
5161 control to catch an exception.
5162 The ``args`` correspond to whatever information the personality
5163 routine requires to know if this is an appropriate place to catch the
5164 exception. Control is tranfered to the ``exception`` label if the
5165 ``catchpad`` is not an appropriate handler for the in-flight exception.
5166 The ``normal`` label should contain the code found in the ``catch``
5167 portion of a ``try``/``catch`` sequence. The ``resultval`` has the type
5168 :ref:`token <t_token>` and is used to match the ``catchpad`` to
5169 corresponding :ref:`catchrets <i_catchret>`.
5174 The instruction takes a list of arbitrary values which are interpreted
5175 by the :ref:`personality function <personalityfn>`.
5177 The ``catchpad`` must be provided a ``normal`` label to transfer control
5178 to if the ``catchpad`` matches the exception and an ``exception``
5179 label to transfer control to if it doesn't.
5184 When the call stack is being unwound due to an exception being thrown,
5185 the exception is compared against the ``args``. If it doesn't match,
5186 then control is transfered to the ``exception`` basic block.
5187 As with calling conventions, how the personality function results are
5188 represented in LLVM IR is target specific.
5190 The ``catchpad`` instruction has several restrictions:
5192 - A catch block is a basic block which is the unwind destination of
5193 an exceptional instruction.
5194 - A catch block must have a '``catchpad``' instruction as its
5195 first non-PHI instruction.
5196 - A catch block's ``exception`` edge must refer to a catch block or a
5198 - There can be only one '``catchpad``' instruction within the
5200 - A basic block that is not a catch block may not include a
5201 '``catchpad``' instruction.
5202 - A catch block which has another catch block as a predecessor may not have
5203 any other predecessors.
5204 - It is undefined behavior for control to transfer from a ``catchpad`` to a
5205 ``ret`` without first executing a ``catchret`` that consumes the
5206 ``catchpad`` or unwinding through its ``catchendpad``.
5207 - It is undefined behavior for control to transfer from a ``catchpad`` to
5208 itself without first executing a ``catchret`` that consumes the
5209 ``catchpad`` or unwinding through its ``catchendpad``.
5214 .. code-block:: llvm
5216 ;; A catch block which can catch an integer.
5217 %tok = catchpad [i8** @_ZTIi]
5218 to label %int.handler unwind label %terminate
5222 '``catchendpad``' Instruction
5223 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
5230 catchendpad unwind label <nextaction>
5231 catchendpad unwind to caller
5236 The '``catchendpad``' instruction is used by `LLVM's exception handling
5237 system <ExceptionHandling.html#overview>`_ to communicate to the
5238 :ref:`personality function <personalityfn>` which invokes are associated
5239 with a chain of :ref:`catchpad <i_catchpad>` instructions; propagating an
5240 exception out of a catch handler is represented by unwinding through its
5241 ``catchendpad``. Unwinding to the outer scope when a chain of catch handlers
5242 do not handle an exception is also represented by unwinding through their
5245 The ``nextaction`` label indicates where control should transfer to if
5246 none of the ``catchpad`` instructions are suitable for catching the
5247 in-flight exception.
5249 If a ``nextaction`` label is not present, the instruction unwinds out of
5250 its parent function. The
5251 :ref:`personality function <personalityfn>` will continue processing
5252 exception handling actions in the caller.
5257 The instruction optionally takes a label, ``nextaction``, indicating
5258 where control should transfer to if none of the preceding
5259 ``catchpad`` instructions are suitable for the in-flight exception.
5264 When the call stack is being unwound due to an exception being thrown
5265 and none of the constituent ``catchpad`` instructions match, then
5266 control is transfered to ``nextaction`` if it is present. If it is not
5267 present, control is transfered to the caller.
5269 The ``catchendpad`` instruction has several restrictions:
5271 - A catch-end block is a basic block which is the unwind destination of
5272 an exceptional instruction.
5273 - A catch-end block must have a '``catchendpad``' instruction as its
5274 first non-PHI instruction.
5275 - There can be only one '``catchendpad``' instruction within the
5277 - A basic block that is not a catch-end block may not include a
5278 '``catchendpad``' instruction.
5279 - Exactly one catch block may unwind to a ``catchendpad``.
5280 - It is undefined behavior to execute a ``catchendpad`` if none of the
5281 '``catchpad``'s chained to it have been executed.
5282 - It is undefined behavior to execute a ``catchendpad`` twice without an
5283 intervening execution of one or more of the '``catchpad``'s chained to it.
5284 - It is undefined behavior to execute a ``catchendpad`` if, after the most
5285 recent execution of the normal successor edge of any ``catchpad`` chained
5286 to it, some ``catchret`` consuming that ``catchpad`` has already been
5288 - It is undefined behavior to execute a ``catchendpad`` if, after the most
5289 recent execution of the normal successor edge of any ``catchpad`` chained
5290 to it, any other ``catchpad`` or ``cleanuppad`` has been executed but has
5291 not had a corresponding
5292 ``catchret``/``cleanupret``/``catchendpad``/``cleanupendpad`` executed.
5297 .. code-block:: llvm
5299 catchendpad unwind label %terminate
5300 catchendpad unwind to caller
5304 '``catchret``' Instruction
5305 ^^^^^^^^^^^^^^^^^^^^^^^^^^
5312 catchret <value> to label <normal>
5317 The '``catchret``' instruction is a terminator instruction that has a
5324 The first argument to a '``catchret``' indicates which ``catchpad`` it
5325 exits. It must be a :ref:`catchpad <i_catchpad>`.
5326 The second argument to a '``catchret``' specifies where control will
5332 The '``catchret``' instruction ends the existing (in-flight) exception
5333 whose unwinding was interrupted with a
5334 :ref:`catchpad <i_catchpad>` instruction.
5335 The :ref:`personality function <personalityfn>` gets a chance to execute
5336 arbitrary code to, for example, run a C++ destructor.
5337 Control then transfers to ``normal``.
5338 It may be passed an optional, personality specific, value.
5340 It is undefined behavior to execute a ``catchret`` whose ``catchpad`` has
5343 It is undefined behavior to execute a ``catchret`` if, after the most recent
5344 execution of its ``catchpad``, some ``catchret`` or ``catchendpad`` linked
5345 to the same ``catchpad`` has already been executed.
5347 It is undefined behavior to execute a ``catchret`` if, after the most recent
5348 execution of its ``catchpad``, any other ``catchpad`` or ``cleanuppad`` has
5349 been executed but has not had a corresponding
5350 ``catchret``/``cleanupret``/``catchendpad``/``cleanupendpad`` executed.
5355 .. code-block:: llvm
5357 catchret %catch label %continue
5359 .. _i_cleanupendpad:
5361 '``cleanupendpad``' Instruction
5362 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
5369 cleanupendpad <value> unwind label <nextaction>
5370 cleanupendpad <value> unwind to caller
5375 The '``cleanupendpad``' instruction is used by `LLVM's exception handling
5376 system <ExceptionHandling.html#overview>`_ to communicate to the
5377 :ref:`personality function <personalityfn>` which invokes are associated
5378 with a :ref:`cleanuppad <i_cleanuppad>` instructions; propagating an exception
5379 out of a cleanup is represented by unwinding through its ``cleanupendpad``.
5381 The ``nextaction`` label indicates where control should unwind to next, in the
5382 event that a cleanup is exited by means of an(other) exception being raised.
5384 If a ``nextaction`` label is not present, the instruction unwinds out of
5385 its parent function. The
5386 :ref:`personality function <personalityfn>` will continue processing
5387 exception handling actions in the caller.
5392 The '``cleanupendpad``' instruction requires one argument, which indicates
5393 which ``cleanuppad`` it exits, and must be a :ref:`cleanuppad <i_cleanuppad>`.
5394 It also has an optional successor, ``nextaction``, indicating where control
5400 When and exception propagates to a ``cleanupendpad``, control is transfered to
5401 ``nextaction`` if it is present. If it is not present, control is transfered to
5404 The ``cleanupendpad`` instruction has several restrictions:
5406 - A cleanup-end block is a basic block which is the unwind destination of
5407 an exceptional instruction.
5408 - A cleanup-end block must have a '``cleanupendpad``' instruction as its
5409 first non-PHI instruction.
5410 - There can be only one '``cleanupendpad``' instruction within the
5412 - A basic block that is not a cleanup-end block may not include a
5413 '``cleanupendpad``' instruction.
5414 - It is undefined behavior to execute a ``cleanupendpad`` whose ``cleanuppad``
5415 has not been executed.
5416 - It is undefined behavior to execute a ``cleanupendpad`` if, after the most
5417 recent execution of its ``cleanuppad``, some ``cleanupret`` or ``cleanupendpad``
5418 consuming the same ``cleanuppad`` has already been executed.
5419 - It is undefined behavior to execute a ``cleanupendpad`` if, after the most
5420 recent execution of its ``cleanuppad``, any other ``cleanuppad`` or
5421 ``catchpad`` has been executed but has not had a corresponding
5422 ``cleanupret``/``catchret``/``cleanupendpad``/``catchendpad`` executed.
5427 .. code-block:: llvm
5429 cleanupendpad %cleanup unwind label %terminate
5430 cleanupendpad %cleanup unwind to caller
5434 '``cleanupret``' Instruction
5435 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
5442 cleanupret <value> unwind label <continue>
5443 cleanupret <value> unwind to caller
5448 The '``cleanupret``' instruction is a terminator instruction that has
5449 an optional successor.
5455 The '``cleanupret``' instruction requires one argument, which indicates
5456 which ``cleanuppad`` it exits, and must be a :ref:`cleanuppad <i_cleanuppad>`.
5457 It also has an optional successor, ``continue``.
5462 The '``cleanupret``' instruction indicates to the
5463 :ref:`personality function <personalityfn>` that one
5464 :ref:`cleanuppad <i_cleanuppad>` it transferred control to has ended.
5465 It transfers control to ``continue`` or unwinds out of the function.
5467 It is undefined behavior to execute a ``cleanupret`` whose ``cleanuppad`` has
5470 It is undefined behavior to execute a ``cleanupret`` if, after the most recent
5471 execution of its ``cleanuppad``, some ``cleanupret`` or ``cleanupendpad``
5472 consuming the same ``cleanuppad`` has already been executed.
5474 It is undefined behavior to execute a ``cleanupret`` if, after the most recent
5475 execution of its ``cleanuppad``, any other ``cleanuppad`` or ``catchpad`` has
5476 been executed but has not had a corresponding
5477 ``cleanupret``/``catchret``/``cleanupendpad``/``catchendpad`` executed.
5482 .. code-block:: llvm
5484 cleanupret %cleanup unwind to caller
5485 cleanupret %cleanup unwind label %continue
5489 '``terminatepad``' Instruction
5490 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
5497 terminatepad [<args>*] unwind label <exception label>
5498 terminatepad [<args>*] unwind to caller
5503 The '``terminatepad``' instruction is used by `LLVM's exception handling
5504 system <ExceptionHandling.html#overview>`_ to specify that a basic block
5505 is a terminate block --- one where a personality routine may decide to
5506 terminate the program.
5507 The ``args`` correspond to whatever information the personality
5508 routine requires to know if this is an appropriate place to terminate the
5509 program. Control is transferred to the ``exception`` label if the
5510 personality routine decides not to terminate the program for the
5511 in-flight exception.
5516 The instruction takes a list of arbitrary values which are interpreted
5517 by the :ref:`personality function <personalityfn>`.
5519 The ``terminatepad`` may be given an ``exception`` label to
5520 transfer control to if the in-flight exception matches the ``args``.
5525 When the call stack is being unwound due to an exception being thrown,
5526 the exception is compared against the ``args``. If it matches,
5527 then control is transfered to the ``exception`` basic block. Otherwise,
5528 the program is terminated via personality-specific means. Typically,
5529 the first argument to ``terminatepad`` specifies what function the
5530 personality should defer to in order to terminate the program.
5532 The ``terminatepad`` instruction has several restrictions:
5534 - A terminate block is a basic block which is the unwind destination of
5535 an exceptional instruction.
5536 - A terminate block must have a '``terminatepad``' instruction as its
5537 first non-PHI instruction.
5538 - There can be only one '``terminatepad``' instruction within the
5540 - A basic block that is not a terminate block may not include a
5541 '``terminatepad``' instruction.
5546 .. code-block:: llvm
5548 ;; A terminate block which only permits integers.
5549 terminatepad [i8** @_ZTIi] unwind label %continue
5553 '``unreachable``' Instruction
5554 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
5566 The '``unreachable``' instruction has no defined semantics. This
5567 instruction is used to inform the optimizer that a particular portion of
5568 the code is not reachable. This can be used to indicate that the code
5569 after a no-return function cannot be reached, and other facts.
5574 The '``unreachable``' instruction has no defined semantics.
5581 Binary operators are used to do most of the computation in a program.
5582 They require two operands of the same type, execute an operation on
5583 them, and produce a single value. The operands might represent multiple
5584 data, as is the case with the :ref:`vector <t_vector>` data type. The
5585 result value has the same type as its operands.
5587 There are several different binary operators:
5591 '``add``' Instruction
5592 ^^^^^^^^^^^^^^^^^^^^^
5599 <result> = add <ty> <op1>, <op2> ; yields ty:result
5600 <result> = add nuw <ty> <op1>, <op2> ; yields ty:result
5601 <result> = add nsw <ty> <op1>, <op2> ; yields ty:result
5602 <result> = add nuw nsw <ty> <op1>, <op2> ; yields ty:result
5607 The '``add``' instruction returns the sum of its two operands.
5612 The two arguments to the '``add``' instruction must be
5613 :ref:`integer <t_integer>` or :ref:`vector <t_vector>` of integer values. Both
5614 arguments must have identical types.
5619 The value produced is the integer sum of the two operands.
5621 If the sum has unsigned overflow, the result returned is the
5622 mathematical result modulo 2\ :sup:`n`\ , where n is the bit width of
5625 Because LLVM integers use a two's complement representation, this
5626 instruction is appropriate for both signed and unsigned integers.
5628 ``nuw`` and ``nsw`` stand for "No Unsigned Wrap" and "No Signed Wrap",
5629 respectively. If the ``nuw`` and/or ``nsw`` keywords are present, the
5630 result value of the ``add`` is a :ref:`poison value <poisonvalues>` if
5631 unsigned and/or signed overflow, respectively, occurs.
5636 .. code-block:: llvm
5638 <result> = add i32 4, %var ; yields i32:result = 4 + %var
5642 '``fadd``' Instruction
5643 ^^^^^^^^^^^^^^^^^^^^^^
5650 <result> = fadd [fast-math flags]* <ty> <op1>, <op2> ; yields ty:result
5655 The '``fadd``' instruction returns the sum of its two operands.
5660 The two arguments to the '``fadd``' instruction must be :ref:`floating
5661 point <t_floating>` or :ref:`vector <t_vector>` of floating point values.
5662 Both arguments must have identical types.
5667 The value produced is the floating point sum of the two operands. This
5668 instruction can also take any number of :ref:`fast-math flags <fastmath>`,
5669 which are optimization hints to enable otherwise unsafe floating point
5675 .. code-block:: llvm
5677 <result> = fadd float 4.0, %var ; yields float:result = 4.0 + %var
5679 '``sub``' Instruction
5680 ^^^^^^^^^^^^^^^^^^^^^
5687 <result> = sub <ty> <op1>, <op2> ; yields ty:result
5688 <result> = sub nuw <ty> <op1>, <op2> ; yields ty:result
5689 <result> = sub nsw <ty> <op1>, <op2> ; yields ty:result
5690 <result> = sub nuw nsw <ty> <op1>, <op2> ; yields ty:result
5695 The '``sub``' instruction returns the difference of its two operands.
5697 Note that the '``sub``' instruction is used to represent the '``neg``'
5698 instruction present in most other intermediate representations.
5703 The two arguments to the '``sub``' instruction must be
5704 :ref:`integer <t_integer>` or :ref:`vector <t_vector>` of integer values. Both
5705 arguments must have identical types.
5710 The value produced is the integer difference of the two operands.
5712 If the difference has unsigned overflow, the result returned is the
5713 mathematical result modulo 2\ :sup:`n`\ , where n is the bit width of
5716 Because LLVM integers use a two's complement representation, this
5717 instruction is appropriate for both signed and unsigned integers.
5719 ``nuw`` and ``nsw`` stand for "No Unsigned Wrap" and "No Signed Wrap",
5720 respectively. If the ``nuw`` and/or ``nsw`` keywords are present, the
5721 result value of the ``sub`` is a :ref:`poison value <poisonvalues>` if
5722 unsigned and/or signed overflow, respectively, occurs.
5727 .. code-block:: llvm
5729 <result> = sub i32 4, %var ; yields i32:result = 4 - %var
5730 <result> = sub i32 0, %val ; yields i32:result = -%var
5734 '``fsub``' Instruction
5735 ^^^^^^^^^^^^^^^^^^^^^^
5742 <result> = fsub [fast-math flags]* <ty> <op1>, <op2> ; yields ty:result
5747 The '``fsub``' instruction returns the difference of its two operands.
5749 Note that the '``fsub``' instruction is used to represent the '``fneg``'
5750 instruction present in most other intermediate representations.
5755 The two arguments to the '``fsub``' instruction must be :ref:`floating
5756 point <t_floating>` or :ref:`vector <t_vector>` of floating point values.
5757 Both arguments must have identical types.
5762 The value produced is the floating point difference of the two operands.
5763 This instruction can also take any number of :ref:`fast-math
5764 flags <fastmath>`, which are optimization hints to enable otherwise
5765 unsafe floating point optimizations:
5770 .. code-block:: llvm
5772 <result> = fsub float 4.0, %var ; yields float:result = 4.0 - %var
5773 <result> = fsub float -0.0, %val ; yields float:result = -%var
5775 '``mul``' Instruction
5776 ^^^^^^^^^^^^^^^^^^^^^
5783 <result> = mul <ty> <op1>, <op2> ; yields ty:result
5784 <result> = mul nuw <ty> <op1>, <op2> ; yields ty:result
5785 <result> = mul nsw <ty> <op1>, <op2> ; yields ty:result
5786 <result> = mul nuw nsw <ty> <op1>, <op2> ; yields ty:result
5791 The '``mul``' instruction returns the product of its two operands.
5796 The two arguments to the '``mul``' instruction must be
5797 :ref:`integer <t_integer>` or :ref:`vector <t_vector>` of integer values. Both
5798 arguments must have identical types.
5803 The value produced is the integer product of the two operands.
5805 If the result of the multiplication has unsigned overflow, the result
5806 returned is the mathematical result modulo 2\ :sup:`n`\ , where n is the
5807 bit width of the result.
5809 Because LLVM integers use a two's complement representation, and the
5810 result is the same width as the operands, this instruction returns the
5811 correct result for both signed and unsigned integers. If a full product
5812 (e.g. ``i32`` * ``i32`` -> ``i64``) is needed, the operands should be
5813 sign-extended or zero-extended as appropriate to the width of the full
5816 ``nuw`` and ``nsw`` stand for "No Unsigned Wrap" and "No Signed Wrap",
5817 respectively. If the ``nuw`` and/or ``nsw`` keywords are present, the
5818 result value of the ``mul`` is a :ref:`poison value <poisonvalues>` if
5819 unsigned and/or signed overflow, respectively, occurs.
5824 .. code-block:: llvm
5826 <result> = mul i32 4, %var ; yields i32:result = 4 * %var
5830 '``fmul``' Instruction
5831 ^^^^^^^^^^^^^^^^^^^^^^
5838 <result> = fmul [fast-math flags]* <ty> <op1>, <op2> ; yields ty:result
5843 The '``fmul``' instruction returns the product of its two operands.
5848 The two arguments to the '``fmul``' instruction must be :ref:`floating
5849 point <t_floating>` or :ref:`vector <t_vector>` of floating point values.
5850 Both arguments must have identical types.
5855 The value produced is the floating point product of the two operands.
5856 This instruction can also take any number of :ref:`fast-math
5857 flags <fastmath>`, which are optimization hints to enable otherwise
5858 unsafe floating point optimizations:
5863 .. code-block:: llvm
5865 <result> = fmul float 4.0, %var ; yields float:result = 4.0 * %var
5867 '``udiv``' Instruction
5868 ^^^^^^^^^^^^^^^^^^^^^^
5875 <result> = udiv <ty> <op1>, <op2> ; yields ty:result
5876 <result> = udiv exact <ty> <op1>, <op2> ; yields ty:result
5881 The '``udiv``' instruction returns the quotient of its two operands.
5886 The two arguments to the '``udiv``' instruction must be
5887 :ref:`integer <t_integer>` or :ref:`vector <t_vector>` of integer values. Both
5888 arguments must have identical types.
5893 The value produced is the unsigned integer quotient of the two operands.
5895 Note that unsigned integer division and signed integer division are
5896 distinct operations; for signed integer division, use '``sdiv``'.
5898 Division by zero leads to undefined behavior.
5900 If the ``exact`` keyword is present, the result value of the ``udiv`` is
5901 a :ref:`poison value <poisonvalues>` if %op1 is not a multiple of %op2 (as
5902 such, "((a udiv exact b) mul b) == a").
5907 .. code-block:: llvm
5909 <result> = udiv i32 4, %var ; yields i32:result = 4 / %var
5911 '``sdiv``' Instruction
5912 ^^^^^^^^^^^^^^^^^^^^^^
5919 <result> = sdiv <ty> <op1>, <op2> ; yields ty:result
5920 <result> = sdiv exact <ty> <op1>, <op2> ; yields ty:result
5925 The '``sdiv``' instruction returns the quotient of its two operands.
5930 The two arguments to the '``sdiv``' instruction must be
5931 :ref:`integer <t_integer>` or :ref:`vector <t_vector>` of integer values. Both
5932 arguments must have identical types.
5937 The value produced is the signed integer quotient of the two operands
5938 rounded towards zero.
5940 Note that signed integer division and unsigned integer division are
5941 distinct operations; for unsigned integer division, use '``udiv``'.
5943 Division by zero leads to undefined behavior. Overflow also leads to
5944 undefined behavior; this is a rare case, but can occur, for example, by
5945 doing a 32-bit division of -2147483648 by -1.
5947 If the ``exact`` keyword is present, the result value of the ``sdiv`` is
5948 a :ref:`poison value <poisonvalues>` if the result would be rounded.
5953 .. code-block:: llvm
5955 <result> = sdiv i32 4, %var ; yields i32:result = 4 / %var
5959 '``fdiv``' Instruction
5960 ^^^^^^^^^^^^^^^^^^^^^^
5967 <result> = fdiv [fast-math flags]* <ty> <op1>, <op2> ; yields ty:result
5972 The '``fdiv``' instruction returns the quotient of its two operands.
5977 The two arguments to the '``fdiv``' instruction must be :ref:`floating
5978 point <t_floating>` or :ref:`vector <t_vector>` of floating point values.
5979 Both arguments must have identical types.
5984 The value produced is the floating point quotient of the two operands.
5985 This instruction can also take any number of :ref:`fast-math
5986 flags <fastmath>`, which are optimization hints to enable otherwise
5987 unsafe floating point optimizations:
5992 .. code-block:: llvm
5994 <result> = fdiv float 4.0, %var ; yields float:result = 4.0 / %var
5996 '``urem``' Instruction
5997 ^^^^^^^^^^^^^^^^^^^^^^
6004 <result> = urem <ty> <op1>, <op2> ; yields ty:result
6009 The '``urem``' instruction returns the remainder from the unsigned
6010 division of its two arguments.
6015 The two arguments to the '``urem``' instruction must be
6016 :ref:`integer <t_integer>` or :ref:`vector <t_vector>` of integer values. Both
6017 arguments must have identical types.
6022 This instruction returns the unsigned integer *remainder* of a division.
6023 This instruction always performs an unsigned division to get the
6026 Note that unsigned integer remainder and signed integer remainder are
6027 distinct operations; for signed integer remainder, use '``srem``'.
6029 Taking the remainder of a division by zero leads to undefined behavior.
6034 .. code-block:: llvm
6036 <result> = urem i32 4, %var ; yields i32:result = 4 % %var
6038 '``srem``' Instruction
6039 ^^^^^^^^^^^^^^^^^^^^^^
6046 <result> = srem <ty> <op1>, <op2> ; yields ty:result
6051 The '``srem``' instruction returns the remainder from the signed
6052 division of its two operands. This instruction can also take
6053 :ref:`vector <t_vector>` versions of the values in which case the elements
6059 The two arguments to the '``srem``' instruction must be
6060 :ref:`integer <t_integer>` or :ref:`vector <t_vector>` of integer values. Both
6061 arguments must have identical types.
6066 This instruction returns the *remainder* of a division (where the result
6067 is either zero or has the same sign as the dividend, ``op1``), not the
6068 *modulo* operator (where the result is either zero or has the same sign
6069 as the divisor, ``op2``) of a value. For more information about the
6070 difference, see `The Math
6071 Forum <http://mathforum.org/dr.math/problems/anne.4.28.99.html>`_. For a
6072 table of how this is implemented in various languages, please see
6074 operation <http://en.wikipedia.org/wiki/Modulo_operation>`_.
6076 Note that signed integer remainder and unsigned integer remainder are
6077 distinct operations; for unsigned integer remainder, use '``urem``'.
6079 Taking the remainder of a division by zero leads to undefined behavior.
6080 Overflow also leads to undefined behavior; this is a rare case, but can
6081 occur, for example, by taking the remainder of a 32-bit division of
6082 -2147483648 by -1. (The remainder doesn't actually overflow, but this
6083 rule lets srem be implemented using instructions that return both the
6084 result of the division and the remainder.)
6089 .. code-block:: llvm
6091 <result> = srem i32 4, %var ; yields i32:result = 4 % %var
6095 '``frem``' Instruction
6096 ^^^^^^^^^^^^^^^^^^^^^^
6103 <result> = frem [fast-math flags]* <ty> <op1>, <op2> ; yields ty:result
6108 The '``frem``' instruction returns the remainder from the division of
6114 The two arguments to the '``frem``' instruction must be :ref:`floating
6115 point <t_floating>` or :ref:`vector <t_vector>` of floating point values.
6116 Both arguments must have identical types.
6121 This instruction returns the *remainder* of a division. The remainder
6122 has the same sign as the dividend. This instruction can also take any
6123 number of :ref:`fast-math flags <fastmath>`, which are optimization hints
6124 to enable otherwise unsafe floating point optimizations:
6129 .. code-block:: llvm
6131 <result> = frem float 4.0, %var ; yields float:result = 4.0 % %var
6135 Bitwise Binary Operations
6136 -------------------------
6138 Bitwise binary operators are used to do various forms of bit-twiddling
6139 in a program. They are generally very efficient instructions and can
6140 commonly be strength reduced from other instructions. They require two
6141 operands of the same type, execute an operation on them, and produce a
6142 single value. The resulting value is the same type as its operands.
6144 '``shl``' Instruction
6145 ^^^^^^^^^^^^^^^^^^^^^
6152 <result> = shl <ty> <op1>, <op2> ; yields ty:result
6153 <result> = shl nuw <ty> <op1>, <op2> ; yields ty:result
6154 <result> = shl nsw <ty> <op1>, <op2> ; yields ty:result
6155 <result> = shl nuw nsw <ty> <op1>, <op2> ; yields ty:result
6160 The '``shl``' instruction returns the first operand shifted to the left
6161 a specified number of bits.
6166 Both arguments to the '``shl``' instruction must be the same
6167 :ref:`integer <t_integer>` or :ref:`vector <t_vector>` of integer type.
6168 '``op2``' is treated as an unsigned value.
6173 The value produced is ``op1`` \* 2\ :sup:`op2` mod 2\ :sup:`n`,
6174 where ``n`` is the width of the result. If ``op2`` is (statically or
6175 dynamically) equal to or larger than the number of bits in
6176 ``op1``, the result is undefined. If the arguments are vectors, each
6177 vector element of ``op1`` is shifted by the corresponding shift amount
6180 If the ``nuw`` keyword is present, then the shift produces a :ref:`poison
6181 value <poisonvalues>` if it shifts out any non-zero bits. If the
6182 ``nsw`` keyword is present, then the shift produces a :ref:`poison
6183 value <poisonvalues>` if it shifts out any bits that disagree with the
6184 resultant sign bit. As such, NUW/NSW have the same semantics as they
6185 would if the shift were expressed as a mul instruction with the same
6186 nsw/nuw bits in (mul %op1, (shl 1, %op2)).
6191 .. code-block:: llvm
6193 <result> = shl i32 4, %var ; yields i32: 4 << %var
6194 <result> = shl i32 4, 2 ; yields i32: 16
6195 <result> = shl i32 1, 10 ; yields i32: 1024
6196 <result> = shl i32 1, 32 ; undefined
6197 <result> = shl <2 x i32> < i32 1, i32 1>, < i32 1, i32 2> ; yields: result=<2 x i32> < i32 2, i32 4>
6199 '``lshr``' Instruction
6200 ^^^^^^^^^^^^^^^^^^^^^^
6207 <result> = lshr <ty> <op1>, <op2> ; yields ty:result
6208 <result> = lshr exact <ty> <op1>, <op2> ; yields ty:result
6213 The '``lshr``' instruction (logical shift right) returns the first
6214 operand shifted to the right a specified number of bits with zero fill.
6219 Both arguments to the '``lshr``' instruction must be the same
6220 :ref:`integer <t_integer>` or :ref:`vector <t_vector>` of integer type.
6221 '``op2``' is treated as an unsigned value.
6226 This instruction always performs a logical shift right operation. The
6227 most significant bits of the result will be filled with zero bits after
6228 the shift. If ``op2`` is (statically or dynamically) equal to or larger
6229 than the number of bits in ``op1``, the result is undefined. If the
6230 arguments are vectors, each vector element of ``op1`` is shifted by the
6231 corresponding shift amount in ``op2``.
6233 If the ``exact`` keyword is present, the result value of the ``lshr`` is
6234 a :ref:`poison value <poisonvalues>` if any of the bits shifted out are
6240 .. code-block:: llvm
6242 <result> = lshr i32 4, 1 ; yields i32:result = 2
6243 <result> = lshr i32 4, 2 ; yields i32:result = 1
6244 <result> = lshr i8 4, 3 ; yields i8:result = 0
6245 <result> = lshr i8 -2, 1 ; yields i8:result = 0x7F
6246 <result> = lshr i32 1, 32 ; undefined
6247 <result> = lshr <2 x i32> < i32 -2, i32 4>, < i32 1, i32 2> ; yields: result=<2 x i32> < i32 0x7FFFFFFF, i32 1>
6249 '``ashr``' Instruction
6250 ^^^^^^^^^^^^^^^^^^^^^^
6257 <result> = ashr <ty> <op1>, <op2> ; yields ty:result
6258 <result> = ashr exact <ty> <op1>, <op2> ; yields ty:result
6263 The '``ashr``' instruction (arithmetic shift right) returns the first
6264 operand shifted to the right a specified number of bits with sign
6270 Both arguments to the '``ashr``' instruction must be the same
6271 :ref:`integer <t_integer>` or :ref:`vector <t_vector>` of integer type.
6272 '``op2``' is treated as an unsigned value.
6277 This instruction always performs an arithmetic shift right operation,
6278 The most significant bits of the result will be filled with the sign bit
6279 of ``op1``. If ``op2`` is (statically or dynamically) equal to or larger
6280 than the number of bits in ``op1``, the result is undefined. If the
6281 arguments are vectors, each vector element of ``op1`` is shifted by the
6282 corresponding shift amount in ``op2``.
6284 If the ``exact`` keyword is present, the result value of the ``ashr`` is
6285 a :ref:`poison value <poisonvalues>` if any of the bits shifted out are
6291 .. code-block:: llvm
6293 <result> = ashr i32 4, 1 ; yields i32:result = 2
6294 <result> = ashr i32 4, 2 ; yields i32:result = 1
6295 <result> = ashr i8 4, 3 ; yields i8:result = 0
6296 <result> = ashr i8 -2, 1 ; yields i8:result = -1
6297 <result> = ashr i32 1, 32 ; undefined
6298 <result> = ashr <2 x i32> < i32 -2, i32 4>, < i32 1, i32 3> ; yields: result=<2 x i32> < i32 -1, i32 0>
6300 '``and``' Instruction
6301 ^^^^^^^^^^^^^^^^^^^^^
6308 <result> = and <ty> <op1>, <op2> ; yields ty:result
6313 The '``and``' instruction returns the bitwise logical and of its two
6319 The two arguments to the '``and``' instruction must be
6320 :ref:`integer <t_integer>` or :ref:`vector <t_vector>` of integer values. Both
6321 arguments must have identical types.
6326 The truth table used for the '``and``' instruction is:
6343 .. code-block:: llvm
6345 <result> = and i32 4, %var ; yields i32:result = 4 & %var
6346 <result> = and i32 15, 40 ; yields i32:result = 8
6347 <result> = and i32 4, 8 ; yields i32:result = 0
6349 '``or``' Instruction
6350 ^^^^^^^^^^^^^^^^^^^^
6357 <result> = or <ty> <op1>, <op2> ; yields ty:result
6362 The '``or``' instruction returns the bitwise logical inclusive or of its
6368 The two arguments to the '``or``' instruction must be
6369 :ref:`integer <t_integer>` or :ref:`vector <t_vector>` of integer values. Both
6370 arguments must have identical types.
6375 The truth table used for the '``or``' instruction is:
6394 <result> = or i32 4, %var ; yields i32:result = 4 | %var
6395 <result> = or i32 15, 40 ; yields i32:result = 47
6396 <result> = or i32 4, 8 ; yields i32:result = 12
6398 '``xor``' Instruction
6399 ^^^^^^^^^^^^^^^^^^^^^
6406 <result> = xor <ty> <op1>, <op2> ; yields ty:result
6411 The '``xor``' instruction returns the bitwise logical exclusive or of
6412 its two operands. The ``xor`` is used to implement the "one's
6413 complement" operation, which is the "~" operator in C.
6418 The two arguments to the '``xor``' instruction must be
6419 :ref:`integer <t_integer>` or :ref:`vector <t_vector>` of integer values. Both
6420 arguments must have identical types.
6425 The truth table used for the '``xor``' instruction is:
6442 .. code-block:: llvm
6444 <result> = xor i32 4, %var ; yields i32:result = 4 ^ %var
6445 <result> = xor i32 15, 40 ; yields i32:result = 39
6446 <result> = xor i32 4, 8 ; yields i32:result = 12
6447 <result> = xor i32 %V, -1 ; yields i32:result = ~%V
6452 LLVM supports several instructions to represent vector operations in a
6453 target-independent manner. These instructions cover the element-access
6454 and vector-specific operations needed to process vectors effectively.
6455 While LLVM does directly support these vector operations, many
6456 sophisticated algorithms will want to use target-specific intrinsics to
6457 take full advantage of a specific target.
6459 .. _i_extractelement:
6461 '``extractelement``' Instruction
6462 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
6469 <result> = extractelement <n x <ty>> <val>, <ty2> <idx> ; yields <ty>
6474 The '``extractelement``' instruction extracts a single scalar element
6475 from a vector at a specified index.
6480 The first operand of an '``extractelement``' instruction is a value of
6481 :ref:`vector <t_vector>` type. The second operand is an index indicating
6482 the position from which to extract the element. The index may be a
6483 variable of any integer type.
6488 The result is a scalar of the same type as the element type of ``val``.
6489 Its value is the value at position ``idx`` of ``val``. If ``idx``
6490 exceeds the length of ``val``, the results are undefined.
6495 .. code-block:: llvm
6497 <result> = extractelement <4 x i32> %vec, i32 0 ; yields i32
6499 .. _i_insertelement:
6501 '``insertelement``' Instruction
6502 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
6509 <result> = insertelement <n x <ty>> <val>, <ty> <elt>, <ty2> <idx> ; yields <n x <ty>>
6514 The '``insertelement``' instruction inserts a scalar element into a
6515 vector at a specified index.
6520 The first operand of an '``insertelement``' instruction is a value of
6521 :ref:`vector <t_vector>` type. The second operand is a scalar value whose
6522 type must equal the element type of the first operand. The third operand
6523 is an index indicating the position at which to insert the value. The
6524 index may be a variable of any integer type.
6529 The result is a vector of the same type as ``val``. Its element values
6530 are those of ``val`` except at position ``idx``, where it gets the value
6531 ``elt``. If ``idx`` exceeds the length of ``val``, the results are
6537 .. code-block:: llvm
6539 <result> = insertelement <4 x i32> %vec, i32 1, i32 0 ; yields <4 x i32>
6541 .. _i_shufflevector:
6543 '``shufflevector``' Instruction
6544 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
6551 <result> = shufflevector <n x <ty>> <v1>, <n x <ty>> <v2>, <m x i32> <mask> ; yields <m x <ty>>
6556 The '``shufflevector``' instruction constructs a permutation of elements
6557 from two input vectors, returning a vector with the same element type as
6558 the input and length that is the same as the shuffle mask.
6563 The first two operands of a '``shufflevector``' instruction are vectors
6564 with the same type. The third argument is a shuffle mask whose element
6565 type is always 'i32'. The result of the instruction is a vector whose
6566 length is the same as the shuffle mask and whose element type is the
6567 same as the element type of the first two operands.
6569 The shuffle mask operand is required to be a constant vector with either
6570 constant integer or undef values.
6575 The elements of the two input vectors are numbered from left to right
6576 across both of the vectors. The shuffle mask operand specifies, for each
6577 element of the result vector, which element of the two input vectors the
6578 result element gets. The element selector may be undef (meaning "don't
6579 care") and the second operand may be undef if performing a shuffle from
6585 .. code-block:: llvm
6587 <result> = shufflevector <4 x i32> %v1, <4 x i32> %v2,
6588 <4 x i32> <i32 0, i32 4, i32 1, i32 5> ; yields <4 x i32>
6589 <result> = shufflevector <4 x i32> %v1, <4 x i32> undef,
6590 <4 x i32> <i32 0, i32 1, i32 2, i32 3> ; yields <4 x i32> - Identity shuffle.
6591 <result> = shufflevector <8 x i32> %v1, <8 x i32> undef,
6592 <4 x i32> <i32 0, i32 1, i32 2, i32 3> ; yields <4 x i32>
6593 <result> = shufflevector <4 x i32> %v1, <4 x i32> %v2,
6594 <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7 > ; yields <8 x i32>
6596 Aggregate Operations
6597 --------------------
6599 LLVM supports several instructions for working with
6600 :ref:`aggregate <t_aggregate>` values.
6604 '``extractvalue``' Instruction
6605 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
6612 <result> = extractvalue <aggregate type> <val>, <idx>{, <idx>}*
6617 The '``extractvalue``' instruction extracts the value of a member field
6618 from an :ref:`aggregate <t_aggregate>` value.
6623 The first operand of an '``extractvalue``' instruction is a value of
6624 :ref:`struct <t_struct>` or :ref:`array <t_array>` type. The operands are
6625 constant indices to specify which value to extract in a similar manner
6626 as indices in a '``getelementptr``' instruction.
6628 The major differences to ``getelementptr`` indexing are:
6630 - Since the value being indexed is not a pointer, the first index is
6631 omitted and assumed to be zero.
6632 - At least one index must be specified.
6633 - Not only struct indices but also array indices must be in bounds.
6638 The result is the value at the position in the aggregate specified by
6644 .. code-block:: llvm
6646 <result> = extractvalue {i32, float} %agg, 0 ; yields i32
6650 '``insertvalue``' Instruction
6651 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
6658 <result> = insertvalue <aggregate type> <val>, <ty> <elt>, <idx>{, <idx>}* ; yields <aggregate type>
6663 The '``insertvalue``' instruction inserts a value into a member field in
6664 an :ref:`aggregate <t_aggregate>` value.
6669 The first operand of an '``insertvalue``' instruction is a value of
6670 :ref:`struct <t_struct>` or :ref:`array <t_array>` type. The second operand is
6671 a first-class value to insert. The following operands are constant
6672 indices indicating the position at which to insert the value in a
6673 similar manner as indices in a '``extractvalue``' instruction. The value
6674 to insert must have the same type as the value identified by the
6680 The result is an aggregate of the same type as ``val``. Its value is
6681 that of ``val`` except that the value at the position specified by the
6682 indices is that of ``elt``.
6687 .. code-block:: llvm
6689 %agg1 = insertvalue {i32, float} undef, i32 1, 0 ; yields {i32 1, float undef}
6690 %agg2 = insertvalue {i32, float} %agg1, float %val, 1 ; yields {i32 1, float %val}
6691 %agg3 = insertvalue {i32, {float}} undef, float %val, 1, 0 ; yields {i32 undef, {float %val}}
6695 Memory Access and Addressing Operations
6696 ---------------------------------------
6698 A key design point of an SSA-based representation is how it represents
6699 memory. In LLVM, no memory locations are in SSA form, which makes things
6700 very simple. This section describes how to read, write, and allocate
6705 '``alloca``' Instruction
6706 ^^^^^^^^^^^^^^^^^^^^^^^^
6713 <result> = alloca [inalloca] <type> [, <ty> <NumElements>] [, align <alignment>] ; yields type*:result
6718 The '``alloca``' instruction allocates memory on the stack frame of the
6719 currently executing function, to be automatically released when this
6720 function returns to its caller. The object is always allocated in the
6721 generic address space (address space zero).
6726 The '``alloca``' instruction allocates ``sizeof(<type>)*NumElements``
6727 bytes of memory on the runtime stack, returning a pointer of the
6728 appropriate type to the program. If "NumElements" is specified, it is
6729 the number of elements allocated, otherwise "NumElements" is defaulted
6730 to be one. If a constant alignment is specified, the value result of the
6731 allocation is guaranteed to be aligned to at least that boundary. The
6732 alignment may not be greater than ``1 << 29``. If not specified, or if
6733 zero, the target can choose to align the allocation on any convenient
6734 boundary compatible with the type.
6736 '``type``' may be any sized type.
6741 Memory is allocated; a pointer is returned. The operation is undefined
6742 if there is insufficient stack space for the allocation. '``alloca``'d
6743 memory is automatically released when the function returns. The
6744 '``alloca``' instruction is commonly used to represent automatic
6745 variables that must have an address available. When the function returns
6746 (either with the ``ret`` or ``resume`` instructions), the memory is
6747 reclaimed. Allocating zero bytes is legal, but the result is undefined.
6748 The order in which memory is allocated (ie., which way the stack grows)
6754 .. code-block:: llvm
6756 %ptr = alloca i32 ; yields i32*:ptr
6757 %ptr = alloca i32, i32 4 ; yields i32*:ptr
6758 %ptr = alloca i32, i32 4, align 1024 ; yields i32*:ptr
6759 %ptr = alloca i32, align 1024 ; yields i32*:ptr
6763 '``load``' Instruction
6764 ^^^^^^^^^^^^^^^^^^^^^^
6771 <result> = load [volatile] <ty>, <ty>* <pointer>[, align <alignment>][, !nontemporal !<index>][, !invariant.load !<index>][, !nonnull !<index>][, !dereferenceable !<index>][, !dereferenceable_or_null !<index>]
6772 <result> = load atomic [volatile] <ty>* <pointer> [singlethread] <ordering>, align <alignment>
6773 !<index> = !{ i32 1 }
6778 The '``load``' instruction is used to read from memory.
6783 The argument to the ``load`` instruction specifies the memory address
6784 from which to load. The type specified must be a :ref:`first
6785 class <t_firstclass>` type. If the ``load`` is marked as ``volatile``,
6786 then the optimizer is not allowed to modify the number or order of
6787 execution of this ``load`` with other :ref:`volatile
6788 operations <volatile>`.
6790 If the ``load`` is marked as ``atomic``, it takes an extra
6791 :ref:`ordering <ordering>` and optional ``singlethread`` argument. The
6792 ``release`` and ``acq_rel`` orderings are not valid on ``load``
6793 instructions. Atomic loads produce :ref:`defined <memmodel>` results
6794 when they may see multiple atomic stores. The type of the pointee must
6795 be an integer type whose bit width is a power of two greater than or
6796 equal to eight and less than or equal to a target-specific size limit.
6797 ``align`` must be explicitly specified on atomic loads, and the load has
6798 undefined behavior if the alignment is not set to a value which is at
6799 least the size in bytes of the pointee. ``!nontemporal`` does not have
6800 any defined semantics for atomic loads.
6802 The optional constant ``align`` argument specifies the alignment of the
6803 operation (that is, the alignment of the memory address). A value of 0
6804 or an omitted ``align`` argument means that the operation has the ABI
6805 alignment for the target. It is the responsibility of the code emitter
6806 to ensure that the alignment information is correct. Overestimating the
6807 alignment results in undefined behavior. Underestimating the alignment
6808 may produce less efficient code. An alignment of 1 is always safe. The
6809 maximum possible alignment is ``1 << 29``.
6811 The optional ``!nontemporal`` metadata must reference a single
6812 metadata name ``<index>`` corresponding to a metadata node with one
6813 ``i32`` entry of value 1. The existence of the ``!nontemporal``
6814 metadata on the instruction tells the optimizer and code generator
6815 that this load is not expected to be reused in the cache. The code
6816 generator may select special instructions to save cache bandwidth, such
6817 as the ``MOVNT`` instruction on x86.
6819 The optional ``!invariant.load`` metadata must reference a single
6820 metadata name ``<index>`` corresponding to a metadata node with no
6821 entries. The existence of the ``!invariant.load`` metadata on the
6822 instruction tells the optimizer and code generator that the address
6823 operand to this load points to memory which can be assumed unchanged.
6824 Being invariant does not imply that a location is dereferenceable,
6825 but it does imply that once the location is known dereferenceable
6826 its value is henceforth unchanging.
6828 The optional ``!nonnull`` metadata must reference a single
6829 metadata name ``<index>`` corresponding to a metadata node with no
6830 entries. The existence of the ``!nonnull`` metadata on the
6831 instruction tells the optimizer that the value loaded is known to
6832 never be null. This is analogous to the ``nonnull`` attribute
6833 on parameters and return values. This metadata can only be applied
6834 to loads of a pointer type.
6836 The optional ``!dereferenceable`` metadata must reference a single
6837 metadata name ``<index>`` corresponding to a metadata node with one ``i64``
6838 entry. The existence of the ``!dereferenceable`` metadata on the instruction
6839 tells the optimizer that the value loaded is known to be dereferenceable.
6840 The number of bytes known to be dereferenceable is specified by the integer
6841 value in the metadata node. This is analogous to the ''dereferenceable''
6842 attribute on parameters and return values. This metadata can only be applied
6843 to loads of a pointer type.
6845 The optional ``!dereferenceable_or_null`` metadata must reference a single
6846 metadata name ``<index>`` corresponding to a metadata node with one ``i64``
6847 entry. The existence of the ``!dereferenceable_or_null`` metadata on the
6848 instruction tells the optimizer that the value loaded is known to be either
6849 dereferenceable or null.
6850 The number of bytes known to be dereferenceable is specified by the integer
6851 value in the metadata node. This is analogous to the ''dereferenceable_or_null''
6852 attribute on parameters and return values. This metadata can only be applied
6853 to loads of a pointer type.
6858 The location of memory pointed to is loaded. If the value being loaded
6859 is of scalar type then the number of bytes read does not exceed the
6860 minimum number of bytes needed to hold all bits of the type. For
6861 example, loading an ``i24`` reads at most three bytes. When loading a
6862 value of a type like ``i20`` with a size that is not an integral number
6863 of bytes, the result is undefined if the value was not originally
6864 written using a store of the same type.
6869 .. code-block:: llvm
6871 %ptr = alloca i32 ; yields i32*:ptr
6872 store i32 3, i32* %ptr ; yields void
6873 %val = load i32, i32* %ptr ; yields i32:val = i32 3
6877 '``store``' Instruction
6878 ^^^^^^^^^^^^^^^^^^^^^^^
6885 store [volatile] <ty> <value>, <ty>* <pointer>[, align <alignment>][, !nontemporal !<index>] ; yields void
6886 store atomic [volatile] <ty> <value>, <ty>* <pointer> [singlethread] <ordering>, align <alignment> ; yields void
6891 The '``store``' instruction is used to write to memory.
6896 There are two arguments to the ``store`` instruction: a value to store
6897 and an address at which to store it. The type of the ``<pointer>``
6898 operand must be a pointer to the :ref:`first class <t_firstclass>` type of
6899 the ``<value>`` operand. If the ``store`` is marked as ``volatile``,
6900 then the optimizer is not allowed to modify the number or order of
6901 execution of this ``store`` with other :ref:`volatile
6902 operations <volatile>`.
6904 If the ``store`` is marked as ``atomic``, it takes an extra
6905 :ref:`ordering <ordering>` and optional ``singlethread`` argument. The
6906 ``acquire`` and ``acq_rel`` orderings aren't valid on ``store``
6907 instructions. Atomic loads produce :ref:`defined <memmodel>` results
6908 when they may see multiple atomic stores. The type of the pointee must
6909 be an integer type whose bit width is a power of two greater than or
6910 equal to eight and less than or equal to a target-specific size limit.
6911 ``align`` must be explicitly specified on atomic stores, and the store
6912 has undefined behavior if the alignment is not set to a value which is
6913 at least the size in bytes of the pointee. ``!nontemporal`` does not
6914 have any defined semantics for atomic stores.
6916 The optional constant ``align`` argument specifies the alignment of the
6917 operation (that is, the alignment of the memory address). A value of 0
6918 or an omitted ``align`` argument means that the operation has the ABI
6919 alignment for the target. It is the responsibility of the code emitter
6920 to ensure that the alignment information is correct. Overestimating the
6921 alignment results in undefined behavior. Underestimating the
6922 alignment may produce less efficient code. An alignment of 1 is always
6923 safe. The maximum possible alignment is ``1 << 29``.
6925 The optional ``!nontemporal`` metadata must reference a single metadata
6926 name ``<index>`` corresponding to a metadata node with one ``i32`` entry of
6927 value 1. The existence of the ``!nontemporal`` metadata on the instruction
6928 tells the optimizer and code generator that this load is not expected to
6929 be reused in the cache. The code generator may select special
6930 instructions to save cache bandwidth, such as the MOVNT instruction on
6936 The contents of memory are updated to contain ``<value>`` at the
6937 location specified by the ``<pointer>`` operand. If ``<value>`` is
6938 of scalar type then the number of bytes written does not exceed the
6939 minimum number of bytes needed to hold all bits of the type. For
6940 example, storing an ``i24`` writes at most three bytes. When writing a
6941 value of a type like ``i20`` with a size that is not an integral number
6942 of bytes, it is unspecified what happens to the extra bits that do not
6943 belong to the type, but they will typically be overwritten.
6948 .. code-block:: llvm
6950 %ptr = alloca i32 ; yields i32*:ptr
6951 store i32 3, i32* %ptr ; yields void
6952 %val = load i32, i32* %ptr ; yields i32:val = i32 3
6956 '``fence``' Instruction
6957 ^^^^^^^^^^^^^^^^^^^^^^^
6964 fence [singlethread] <ordering> ; yields void
6969 The '``fence``' instruction is used to introduce happens-before edges
6975 '``fence``' instructions take an :ref:`ordering <ordering>` argument which
6976 defines what *synchronizes-with* edges they add. They can only be given
6977 ``acquire``, ``release``, ``acq_rel``, and ``seq_cst`` orderings.
6982 A fence A which has (at least) ``release`` ordering semantics
6983 *synchronizes with* a fence B with (at least) ``acquire`` ordering
6984 semantics if and only if there exist atomic operations X and Y, both
6985 operating on some atomic object M, such that A is sequenced before X, X
6986 modifies M (either directly or through some side effect of a sequence
6987 headed by X), Y is sequenced before B, and Y observes M. This provides a
6988 *happens-before* dependency between A and B. Rather than an explicit
6989 ``fence``, one (but not both) of the atomic operations X or Y might
6990 provide a ``release`` or ``acquire`` (resp.) ordering constraint and
6991 still *synchronize-with* the explicit ``fence`` and establish the
6992 *happens-before* edge.
6994 A ``fence`` which has ``seq_cst`` ordering, in addition to having both
6995 ``acquire`` and ``release`` semantics specified above, participates in
6996 the global program order of other ``seq_cst`` operations and/or fences.
6998 The optional ":ref:`singlethread <singlethread>`" argument specifies
6999 that the fence only synchronizes with other fences in the same thread.
7000 (This is useful for interacting with signal handlers.)
7005 .. code-block:: llvm
7007 fence acquire ; yields void
7008 fence singlethread seq_cst ; yields void
7012 '``cmpxchg``' Instruction
7013 ^^^^^^^^^^^^^^^^^^^^^^^^^
7020 cmpxchg [weak] [volatile] <ty>* <pointer>, <ty> <cmp>, <ty> <new> [singlethread] <success ordering> <failure ordering> ; yields { ty, i1 }
7025 The '``cmpxchg``' instruction is used to atomically modify memory. It
7026 loads a value in memory and compares it to a given value. If they are
7027 equal, it tries to store a new value into the memory.
7032 There are three arguments to the '``cmpxchg``' instruction: an address
7033 to operate on, a value to compare to the value currently be at that
7034 address, and a new value to place at that address if the compared values
7035 are equal. The type of '<cmp>' must be an integer type whose bit width
7036 is a power of two greater than or equal to eight and less than or equal
7037 to a target-specific size limit. '<cmp>' and '<new>' must have the same
7038 type, and the type of '<pointer>' must be a pointer to that type. If the
7039 ``cmpxchg`` is marked as ``volatile``, then the optimizer is not allowed
7040 to modify the number or order of execution of this ``cmpxchg`` with
7041 other :ref:`volatile operations <volatile>`.
7043 The success and failure :ref:`ordering <ordering>` arguments specify how this
7044 ``cmpxchg`` synchronizes with other atomic operations. Both ordering parameters
7045 must be at least ``monotonic``, the ordering constraint on failure must be no
7046 stronger than that on success, and the failure ordering cannot be either
7047 ``release`` or ``acq_rel``.
7049 The optional "``singlethread``" argument declares that the ``cmpxchg``
7050 is only atomic with respect to code (usually signal handlers) running in
7051 the same thread as the ``cmpxchg``. Otherwise the cmpxchg is atomic with
7052 respect to all other code in the system.
7054 The pointer passed into cmpxchg must have alignment greater than or
7055 equal to the size in memory of the operand.
7060 The contents of memory at the location specified by the '``<pointer>``' operand
7061 is read and compared to '``<cmp>``'; if the read value is the equal, the
7062 '``<new>``' is written. The original value at the location is returned, together
7063 with a flag indicating success (true) or failure (false).
7065 If the cmpxchg operation is marked as ``weak`` then a spurious failure is
7066 permitted: the operation may not write ``<new>`` even if the comparison
7069 If the cmpxchg operation is strong (the default), the i1 value is 1 if and only
7070 if the value loaded equals ``cmp``.
7072 A successful ``cmpxchg`` is a read-modify-write instruction for the purpose of
7073 identifying release sequences. A failed ``cmpxchg`` is equivalent to an atomic
7074 load with an ordering parameter determined the second ordering parameter.
7079 .. code-block:: llvm
7082 %orig = atomic load i32, i32* %ptr unordered ; yields i32
7086 %cmp = phi i32 [ %orig, %entry ], [%old, %loop]
7087 %squared = mul i32 %cmp, %cmp
7088 %val_success = cmpxchg i32* %ptr, i32 %cmp, i32 %squared acq_rel monotonic ; yields { i32, i1 }
7089 %value_loaded = extractvalue { i32, i1 } %val_success, 0
7090 %success = extractvalue { i32, i1 } %val_success, 1
7091 br i1 %success, label %done, label %loop
7098 '``atomicrmw``' Instruction
7099 ^^^^^^^^^^^^^^^^^^^^^^^^^^^
7106 atomicrmw [volatile] <operation> <ty>* <pointer>, <ty> <value> [singlethread] <ordering> ; yields ty
7111 The '``atomicrmw``' instruction is used to atomically modify memory.
7116 There are three arguments to the '``atomicrmw``' instruction: an
7117 operation to apply, an address whose value to modify, an argument to the
7118 operation. The operation must be one of the following keywords:
7132 The type of '<value>' must be an integer type whose bit width is a power
7133 of two greater than or equal to eight and less than or equal to a
7134 target-specific size limit. The type of the '``<pointer>``' operand must
7135 be a pointer to that type. If the ``atomicrmw`` is marked as
7136 ``volatile``, then the optimizer is not allowed to modify the number or
7137 order of execution of this ``atomicrmw`` with other :ref:`volatile
7138 operations <volatile>`.
7143 The contents of memory at the location specified by the '``<pointer>``'
7144 operand are atomically read, modified, and written back. The original
7145 value at the location is returned. The modification is specified by the
7148 - xchg: ``*ptr = val``
7149 - add: ``*ptr = *ptr + val``
7150 - sub: ``*ptr = *ptr - val``
7151 - and: ``*ptr = *ptr & val``
7152 - nand: ``*ptr = ~(*ptr & val)``
7153 - or: ``*ptr = *ptr | val``
7154 - xor: ``*ptr = *ptr ^ val``
7155 - max: ``*ptr = *ptr > val ? *ptr : val`` (using a signed comparison)
7156 - min: ``*ptr = *ptr < val ? *ptr : val`` (using a signed comparison)
7157 - umax: ``*ptr = *ptr > val ? *ptr : val`` (using an unsigned
7159 - umin: ``*ptr = *ptr < val ? *ptr : val`` (using an unsigned
7165 .. code-block:: llvm
7167 %old = atomicrmw add i32* %ptr, i32 1 acquire ; yields i32
7169 .. _i_getelementptr:
7171 '``getelementptr``' Instruction
7172 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
7179 <result> = getelementptr <ty>, <ty>* <ptrval>{, <ty> <idx>}*
7180 <result> = getelementptr inbounds <ty>, <ty>* <ptrval>{, <ty> <idx>}*
7181 <result> = getelementptr <ty>, <ptr vector> <ptrval>, <vector index type> <idx>
7186 The '``getelementptr``' instruction is used to get the address of a
7187 subelement of an :ref:`aggregate <t_aggregate>` data structure. It performs
7188 address calculation only and does not access memory. The instruction can also
7189 be used to calculate a vector of such addresses.
7194 The first argument is always a type used as the basis for the calculations.
7195 The second argument is always a pointer or a vector of pointers, and is the
7196 base address to start from. The remaining arguments are indices
7197 that indicate which of the elements of the aggregate object are indexed.
7198 The interpretation of each index is dependent on the type being indexed
7199 into. The first index always indexes the pointer value given as the
7200 first argument, the second index indexes a value of the type pointed to
7201 (not necessarily the value directly pointed to, since the first index
7202 can be non-zero), etc. The first type indexed into must be a pointer
7203 value, subsequent types can be arrays, vectors, and structs. Note that
7204 subsequent types being indexed into can never be pointers, since that
7205 would require loading the pointer before continuing calculation.
7207 The type of each index argument depends on the type it is indexing into.
7208 When indexing into a (optionally packed) structure, only ``i32`` integer
7209 **constants** are allowed (when using a vector of indices they must all
7210 be the **same** ``i32`` integer constant). When indexing into an array,
7211 pointer or vector, integers of any width are allowed, and they are not
7212 required to be constant. These integers are treated as signed values
7215 For example, let's consider a C code fragment and how it gets compiled
7231 int *foo(struct ST *s) {
7232 return &s[1].Z.B[5][13];
7235 The LLVM code generated by Clang is:
7237 .. code-block:: llvm
7239 %struct.RT = type { i8, [10 x [20 x i32]], i8 }
7240 %struct.ST = type { i32, double, %struct.RT }
7242 define i32* @foo(%struct.ST* %s) nounwind uwtable readnone optsize ssp {
7244 %arrayidx = getelementptr inbounds %struct.ST, %struct.ST* %s, i64 1, i32 2, i32 1, i64 5, i64 13
7251 In the example above, the first index is indexing into the
7252 '``%struct.ST*``' type, which is a pointer, yielding a '``%struct.ST``'
7253 = '``{ i32, double, %struct.RT }``' type, a structure. The second index
7254 indexes into the third element of the structure, yielding a
7255 '``%struct.RT``' = '``{ i8 , [10 x [20 x i32]], i8 }``' type, another
7256 structure. The third index indexes into the second element of the
7257 structure, yielding a '``[10 x [20 x i32]]``' type, an array. The two
7258 dimensions of the array are subscripted into, yielding an '``i32``'
7259 type. The '``getelementptr``' instruction returns a pointer to this
7260 element, thus computing a value of '``i32*``' type.
7262 Note that it is perfectly legal to index partially through a structure,
7263 returning a pointer to an inner element. Because of this, the LLVM code
7264 for the given testcase is equivalent to:
7266 .. code-block:: llvm
7268 define i32* @foo(%struct.ST* %s) {
7269 %t1 = getelementptr %struct.ST, %struct.ST* %s, i32 1 ; yields %struct.ST*:%t1
7270 %t2 = getelementptr %struct.ST, %struct.ST* %t1, i32 0, i32 2 ; yields %struct.RT*:%t2
7271 %t3 = getelementptr %struct.RT, %struct.RT* %t2, i32 0, i32 1 ; yields [10 x [20 x i32]]*:%t3
7272 %t4 = getelementptr [10 x [20 x i32]], [10 x [20 x i32]]* %t3, i32 0, i32 5 ; yields [20 x i32]*:%t4
7273 %t5 = getelementptr [20 x i32], [20 x i32]* %t4, i32 0, i32 13 ; yields i32*:%t5
7277 If the ``inbounds`` keyword is present, the result value of the
7278 ``getelementptr`` is a :ref:`poison value <poisonvalues>` if the base
7279 pointer is not an *in bounds* address of an allocated object, or if any
7280 of the addresses that would be formed by successive addition of the
7281 offsets implied by the indices to the base address with infinitely
7282 precise signed arithmetic are not an *in bounds* address of that
7283 allocated object. The *in bounds* addresses for an allocated object are
7284 all the addresses that point into the object, plus the address one byte
7285 past the end. In cases where the base is a vector of pointers the
7286 ``inbounds`` keyword applies to each of the computations element-wise.
7288 If the ``inbounds`` keyword is not present, the offsets are added to the
7289 base address with silently-wrapping two's complement arithmetic. If the
7290 offsets have a different width from the pointer, they are sign-extended
7291 or truncated to the width of the pointer. The result value of the
7292 ``getelementptr`` may be outside the object pointed to by the base
7293 pointer. The result value may not necessarily be used to access memory
7294 though, even if it happens to point into allocated storage. See the
7295 :ref:`Pointer Aliasing Rules <pointeraliasing>` section for more
7298 The getelementptr instruction is often confusing. For some more insight
7299 into how it works, see :doc:`the getelementptr FAQ <GetElementPtr>`.
7304 .. code-block:: llvm
7306 ; yields [12 x i8]*:aptr
7307 %aptr = getelementptr {i32, [12 x i8]}, {i32, [12 x i8]}* %saptr, i64 0, i32 1
7309 %vptr = getelementptr {i32, <2 x i8>}, {i32, <2 x i8>}* %svptr, i64 0, i32 1, i32 1
7311 %eptr = getelementptr [12 x i8], [12 x i8]* %aptr, i64 0, i32 1
7313 %iptr = getelementptr [10 x i32], [10 x i32]* @arr, i16 0, i16 0
7318 The ``getelementptr`` returns a vector of pointers, instead of a single address,
7319 when one or more of its arguments is a vector. In such cases, all vector
7320 arguments should have the same number of elements, and every scalar argument
7321 will be effectively broadcast into a vector during address calculation.
7323 .. code-block:: llvm
7325 ; All arguments are vectors:
7326 ; A[i] = ptrs[i] + offsets[i]*sizeof(i8)
7327 %A = getelementptr i8, <4 x i8*> %ptrs, <4 x i64> %offsets
7329 ; Add the same scalar offset to each pointer of a vector:
7330 ; A[i] = ptrs[i] + offset*sizeof(i8)
7331 %A = getelementptr i8, <4 x i8*> %ptrs, i64 %offset
7333 ; Add distinct offsets to the same pointer:
7334 ; A[i] = ptr + offsets[i]*sizeof(i8)
7335 %A = getelementptr i8, i8* %ptr, <4 x i64> %offsets
7337 ; In all cases described above the type of the result is <4 x i8*>
7339 The two following instructions are equivalent:
7341 .. code-block:: llvm
7343 getelementptr %struct.ST, <4 x %struct.ST*> %s, <4 x i64> %ind1,
7344 <4 x i32> <i32 2, i32 2, i32 2, i32 2>,
7345 <4 x i32> <i32 1, i32 1, i32 1, i32 1>,
7347 <4 x i64> <i64 13, i64 13, i64 13, i64 13>
7349 getelementptr %struct.ST, <4 x %struct.ST*> %s, <4 x i64> %ind1,
7350 i32 2, i32 1, <4 x i32> %ind4, i64 13
7352 Let's look at the C code, where the vector version of ``getelementptr``
7357 // Let's assume that we vectorize the following loop:
7358 double *A, B; int *C;
7359 for (int i = 0; i < size; ++i) {
7363 .. code-block:: llvm
7365 ; get pointers for 8 elements from array B
7366 %ptrs = getelementptr double, double* %B, <8 x i32> %C
7367 ; load 8 elements from array B into A
7368 %A = call <8 x double> @llvm.masked.gather.v8f64(<8 x double*> %ptrs,
7369 i32 8, <8 x i1> %mask, <8 x double> %passthru)
7371 Conversion Operations
7372 ---------------------
7374 The instructions in this category are the conversion instructions
7375 (casting) which all take a single operand and a type. They perform
7376 various bit conversions on the operand.
7378 '``trunc .. to``' Instruction
7379 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
7386 <result> = trunc <ty> <value> to <ty2> ; yields ty2
7391 The '``trunc``' instruction truncates its operand to the type ``ty2``.
7396 The '``trunc``' instruction takes a value to trunc, and a type to trunc
7397 it to. Both types must be of :ref:`integer <t_integer>` types, or vectors
7398 of the same number of integers. The bit size of the ``value`` must be
7399 larger than the bit size of the destination type, ``ty2``. Equal sized
7400 types are not allowed.
7405 The '``trunc``' instruction truncates the high order bits in ``value``
7406 and converts the remaining bits to ``ty2``. Since the source size must
7407 be larger than the destination size, ``trunc`` cannot be a *no-op cast*.
7408 It will always truncate bits.
7413 .. code-block:: llvm
7415 %X = trunc i32 257 to i8 ; yields i8:1
7416 %Y = trunc i32 123 to i1 ; yields i1:true
7417 %Z = trunc i32 122 to i1 ; yields i1:false
7418 %W = trunc <2 x i16> <i16 8, i16 7> to <2 x i8> ; yields <i8 8, i8 7>
7420 '``zext .. to``' Instruction
7421 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
7428 <result> = zext <ty> <value> to <ty2> ; yields ty2
7433 The '``zext``' instruction zero extends its operand to type ``ty2``.
7438 The '``zext``' instruction takes a value to cast, and a type to cast it
7439 to. Both types must be of :ref:`integer <t_integer>` types, or vectors of
7440 the same number of integers. The bit size of the ``value`` must be
7441 smaller than the bit size of the destination type, ``ty2``.
7446 The ``zext`` fills the high order bits of the ``value`` with zero bits
7447 until it reaches the size of the destination type, ``ty2``.
7449 When zero extending from i1, the result will always be either 0 or 1.
7454 .. code-block:: llvm
7456 %X = zext i32 257 to i64 ; yields i64:257
7457 %Y = zext i1 true to i32 ; yields i32:1
7458 %Z = zext <2 x i16> <i16 8, i16 7> to <2 x i32> ; yields <i32 8, i32 7>
7460 '``sext .. to``' Instruction
7461 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
7468 <result> = sext <ty> <value> to <ty2> ; yields ty2
7473 The '``sext``' sign extends ``value`` to the type ``ty2``.
7478 The '``sext``' instruction takes a value to cast, and a type to cast it
7479 to. Both types must be of :ref:`integer <t_integer>` types, or vectors of
7480 the same number of integers. The bit size of the ``value`` must be
7481 smaller than the bit size of the destination type, ``ty2``.
7486 The '``sext``' instruction performs a sign extension by copying the sign
7487 bit (highest order bit) of the ``value`` until it reaches the bit size
7488 of the type ``ty2``.
7490 When sign extending from i1, the extension always results in -1 or 0.
7495 .. code-block:: llvm
7497 %X = sext i8 -1 to i16 ; yields i16 :65535
7498 %Y = sext i1 true to i32 ; yields i32:-1
7499 %Z = sext <2 x i16> <i16 8, i16 7> to <2 x i32> ; yields <i32 8, i32 7>
7501 '``fptrunc .. to``' Instruction
7502 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
7509 <result> = fptrunc <ty> <value> to <ty2> ; yields ty2
7514 The '``fptrunc``' instruction truncates ``value`` to type ``ty2``.
7519 The '``fptrunc``' instruction takes a :ref:`floating point <t_floating>`
7520 value to cast and a :ref:`floating point <t_floating>` type to cast it to.
7521 The size of ``value`` must be larger than the size of ``ty2``. This
7522 implies that ``fptrunc`` cannot be used to make a *no-op cast*.
7527 The '``fptrunc``' instruction casts a ``value`` from a larger
7528 :ref:`floating point <t_floating>` type to a smaller :ref:`floating
7529 point <t_floating>` type. If the value cannot fit (i.e. overflows) within the
7530 destination type, ``ty2``, then the results are undefined. If the cast produces
7531 an inexact result, how rounding is performed (e.g. truncation, also known as
7532 round to zero) is undefined.
7537 .. code-block:: llvm
7539 %X = fptrunc double 123.0 to float ; yields float:123.0
7540 %Y = fptrunc double 1.0E+300 to float ; yields undefined
7542 '``fpext .. to``' Instruction
7543 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
7550 <result> = fpext <ty> <value> to <ty2> ; yields ty2
7555 The '``fpext``' extends a floating point ``value`` to a larger floating
7561 The '``fpext``' instruction takes a :ref:`floating point <t_floating>`
7562 ``value`` to cast, and a :ref:`floating point <t_floating>` type to cast it
7563 to. The source type must be smaller than the destination type.
7568 The '``fpext``' instruction extends the ``value`` from a smaller
7569 :ref:`floating point <t_floating>` type to a larger :ref:`floating
7570 point <t_floating>` type. The ``fpext`` cannot be used to make a
7571 *no-op cast* because it always changes bits. Use ``bitcast`` to make a
7572 *no-op cast* for a floating point cast.
7577 .. code-block:: llvm
7579 %X = fpext float 3.125 to double ; yields double:3.125000e+00
7580 %Y = fpext double %X to fp128 ; yields fp128:0xL00000000000000004000900000000000
7582 '``fptoui .. to``' Instruction
7583 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
7590 <result> = fptoui <ty> <value> to <ty2> ; yields ty2
7595 The '``fptoui``' converts a floating point ``value`` to its unsigned
7596 integer equivalent of type ``ty2``.
7601 The '``fptoui``' instruction takes a value to cast, which must be a
7602 scalar or vector :ref:`floating point <t_floating>` value, and a type to
7603 cast it to ``ty2``, which must be an :ref:`integer <t_integer>` type. If
7604 ``ty`` is a vector floating point type, ``ty2`` must be a vector integer
7605 type with the same number of elements as ``ty``
7610 The '``fptoui``' instruction converts its :ref:`floating
7611 point <t_floating>` operand into the nearest (rounding towards zero)
7612 unsigned integer value. If the value cannot fit in ``ty2``, the results
7618 .. code-block:: llvm
7620 %X = fptoui double 123.0 to i32 ; yields i32:123
7621 %Y = fptoui float 1.0E+300 to i1 ; yields undefined:1
7622 %Z = fptoui float 1.04E+17 to i8 ; yields undefined:1
7624 '``fptosi .. to``' Instruction
7625 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
7632 <result> = fptosi <ty> <value> to <ty2> ; yields ty2
7637 The '``fptosi``' instruction converts :ref:`floating point <t_floating>`
7638 ``value`` to type ``ty2``.
7643 The '``fptosi``' instruction takes a value to cast, which must be a
7644 scalar or vector :ref:`floating point <t_floating>` value, and a type to
7645 cast it to ``ty2``, which must be an :ref:`integer <t_integer>` type. If
7646 ``ty`` is a vector floating point type, ``ty2`` must be a vector integer
7647 type with the same number of elements as ``ty``
7652 The '``fptosi``' instruction converts its :ref:`floating
7653 point <t_floating>` operand into the nearest (rounding towards zero)
7654 signed integer value. If the value cannot fit in ``ty2``, the results
7660 .. code-block:: llvm
7662 %X = fptosi double -123.0 to i32 ; yields i32:-123
7663 %Y = fptosi float 1.0E-247 to i1 ; yields undefined:1
7664 %Z = fptosi float 1.04E+17 to i8 ; yields undefined:1
7666 '``uitofp .. to``' Instruction
7667 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
7674 <result> = uitofp <ty> <value> to <ty2> ; yields ty2
7679 The '``uitofp``' instruction regards ``value`` as an unsigned integer
7680 and converts that value to the ``ty2`` type.
7685 The '``uitofp``' instruction takes a value to cast, which must be a
7686 scalar or vector :ref:`integer <t_integer>` value, and a type to cast it to
7687 ``ty2``, which must be an :ref:`floating point <t_floating>` type. If
7688 ``ty`` is a vector integer type, ``ty2`` must be a vector floating point
7689 type with the same number of elements as ``ty``
7694 The '``uitofp``' instruction interprets its operand as an unsigned
7695 integer quantity and converts it to the corresponding floating point
7696 value. If the value cannot fit in the floating point value, the results
7702 .. code-block:: llvm
7704 %X = uitofp i32 257 to float ; yields float:257.0
7705 %Y = uitofp i8 -1 to double ; yields double:255.0
7707 '``sitofp .. to``' Instruction
7708 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
7715 <result> = sitofp <ty> <value> to <ty2> ; yields ty2
7720 The '``sitofp``' instruction regards ``value`` as a signed integer and
7721 converts that value to the ``ty2`` type.
7726 The '``sitofp``' instruction takes a value to cast, which must be a
7727 scalar or vector :ref:`integer <t_integer>` value, and a type to cast it to
7728 ``ty2``, which must be an :ref:`floating point <t_floating>` type. If
7729 ``ty`` is a vector integer type, ``ty2`` must be a vector floating point
7730 type with the same number of elements as ``ty``
7735 The '``sitofp``' instruction interprets its operand as a signed integer
7736 quantity and converts it to the corresponding floating point value. If
7737 the value cannot fit in the floating point value, the results are
7743 .. code-block:: llvm
7745 %X = sitofp i32 257 to float ; yields float:257.0
7746 %Y = sitofp i8 -1 to double ; yields double:-1.0
7750 '``ptrtoint .. to``' Instruction
7751 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
7758 <result> = ptrtoint <ty> <value> to <ty2> ; yields ty2
7763 The '``ptrtoint``' instruction converts the pointer or a vector of
7764 pointers ``value`` to the integer (or vector of integers) type ``ty2``.
7769 The '``ptrtoint``' instruction takes a ``value`` to cast, which must be
7770 a value of type :ref:`pointer <t_pointer>` or a vector of pointers, and a
7771 type to cast it to ``ty2``, which must be an :ref:`integer <t_integer>` or
7772 a vector of integers type.
7777 The '``ptrtoint``' instruction converts ``value`` to integer type
7778 ``ty2`` by interpreting the pointer value as an integer and either
7779 truncating or zero extending that value to the size of the integer type.
7780 If ``value`` is smaller than ``ty2`` then a zero extension is done. If
7781 ``value`` is larger than ``ty2`` then a truncation is done. If they are
7782 the same size, then nothing is done (*no-op cast*) other than a type
7788 .. code-block:: llvm
7790 %X = ptrtoint i32* %P to i8 ; yields truncation on 32-bit architecture
7791 %Y = ptrtoint i32* %P to i64 ; yields zero extension on 32-bit architecture
7792 %Z = ptrtoint <4 x i32*> %P to <4 x i64>; yields vector zero extension for a vector of addresses on 32-bit architecture
7796 '``inttoptr .. to``' Instruction
7797 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
7804 <result> = inttoptr <ty> <value> to <ty2> ; yields ty2
7809 The '``inttoptr``' instruction converts an integer ``value`` to a
7810 pointer type, ``ty2``.
7815 The '``inttoptr``' instruction takes an :ref:`integer <t_integer>` value to
7816 cast, and a type to cast it to, which must be a :ref:`pointer <t_pointer>`
7822 The '``inttoptr``' instruction converts ``value`` to type ``ty2`` by
7823 applying either a zero extension or a truncation depending on the size
7824 of the integer ``value``. If ``value`` is larger than the size of a
7825 pointer then a truncation is done. If ``value`` is smaller than the size
7826 of a pointer then a zero extension is done. If they are the same size,
7827 nothing is done (*no-op cast*).
7832 .. code-block:: llvm
7834 %X = inttoptr i32 255 to i32* ; yields zero extension on 64-bit architecture
7835 %Y = inttoptr i32 255 to i32* ; yields no-op on 32-bit architecture
7836 %Z = inttoptr i64 0 to i32* ; yields truncation on 32-bit architecture
7837 %Z = inttoptr <4 x i32> %G to <4 x i8*>; yields truncation of vector G to four pointers
7841 '``bitcast .. to``' Instruction
7842 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
7849 <result> = bitcast <ty> <value> to <ty2> ; yields ty2
7854 The '``bitcast``' instruction converts ``value`` to type ``ty2`` without
7860 The '``bitcast``' instruction takes a value to cast, which must be a
7861 non-aggregate first class value, and a type to cast it to, which must
7862 also be a non-aggregate :ref:`first class <t_firstclass>` type. The
7863 bit sizes of ``value`` and the destination type, ``ty2``, must be
7864 identical. If the source type is a pointer, the destination type must
7865 also be a pointer of the same size. This instruction supports bitwise
7866 conversion of vectors to integers and to vectors of other types (as
7867 long as they have the same size).
7872 The '``bitcast``' instruction converts ``value`` to type ``ty2``. It
7873 is always a *no-op cast* because no bits change with this
7874 conversion. The conversion is done as if the ``value`` had been stored
7875 to memory and read back as type ``ty2``. Pointer (or vector of
7876 pointers) types may only be converted to other pointer (or vector of
7877 pointers) types with the same address space through this instruction.
7878 To convert pointers to other types, use the :ref:`inttoptr <i_inttoptr>`
7879 or :ref:`ptrtoint <i_ptrtoint>` instructions first.
7884 .. code-block:: llvm
7886 %X = bitcast i8 255 to i8 ; yields i8 :-1
7887 %Y = bitcast i32* %x to sint* ; yields sint*:%x
7888 %Z = bitcast <2 x int> %V to i64; ; yields i64: %V
7889 %Z = bitcast <2 x i32*> %V to <2 x i64*> ; yields <2 x i64*>
7891 .. _i_addrspacecast:
7893 '``addrspacecast .. to``' Instruction
7894 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
7901 <result> = addrspacecast <pty> <ptrval> to <pty2> ; yields pty2
7906 The '``addrspacecast``' instruction converts ``ptrval`` from ``pty`` in
7907 address space ``n`` to type ``pty2`` in address space ``m``.
7912 The '``addrspacecast``' instruction takes a pointer or vector of pointer value
7913 to cast and a pointer type to cast it to, which must have a different
7919 The '``addrspacecast``' instruction converts the pointer value
7920 ``ptrval`` to type ``pty2``. It can be a *no-op cast* or a complex
7921 value modification, depending on the target and the address space
7922 pair. Pointer conversions within the same address space must be
7923 performed with the ``bitcast`` instruction. Note that if the address space
7924 conversion is legal then both result and operand refer to the same memory
7930 .. code-block:: llvm
7932 %X = addrspacecast i32* %x to i32 addrspace(1)* ; yields i32 addrspace(1)*:%x
7933 %Y = addrspacecast i32 addrspace(1)* %y to i64 addrspace(2)* ; yields i64 addrspace(2)*:%y
7934 %Z = addrspacecast <4 x i32*> %z to <4 x float addrspace(3)*> ; yields <4 x float addrspace(3)*>:%z
7941 The instructions in this category are the "miscellaneous" instructions,
7942 which defy better classification.
7946 '``icmp``' Instruction
7947 ^^^^^^^^^^^^^^^^^^^^^^
7954 <result> = icmp <cond> <ty> <op1>, <op2> ; yields i1 or <N x i1>:result
7959 The '``icmp``' instruction returns a boolean value or a vector of
7960 boolean values based on comparison of its two integer, integer vector,
7961 pointer, or pointer vector operands.
7966 The '``icmp``' instruction takes three operands. The first operand is
7967 the condition code indicating the kind of comparison to perform. It is
7968 not a value, just a keyword. The possible condition code are:
7971 #. ``ne``: not equal
7972 #. ``ugt``: unsigned greater than
7973 #. ``uge``: unsigned greater or equal
7974 #. ``ult``: unsigned less than
7975 #. ``ule``: unsigned less or equal
7976 #. ``sgt``: signed greater than
7977 #. ``sge``: signed greater or equal
7978 #. ``slt``: signed less than
7979 #. ``sle``: signed less or equal
7981 The remaining two arguments must be :ref:`integer <t_integer>` or
7982 :ref:`pointer <t_pointer>` or integer :ref:`vector <t_vector>` typed. They
7983 must also be identical types.
7988 The '``icmp``' compares ``op1`` and ``op2`` according to the condition
7989 code given as ``cond``. The comparison performed always yields either an
7990 :ref:`i1 <t_integer>` or vector of ``i1`` result, as follows:
7992 #. ``eq``: yields ``true`` if the operands are equal, ``false``
7993 otherwise. No sign interpretation is necessary or performed.
7994 #. ``ne``: yields ``true`` if the operands are unequal, ``false``
7995 otherwise. No sign interpretation is necessary or performed.
7996 #. ``ugt``: interprets the operands as unsigned values and yields
7997 ``true`` if ``op1`` is greater than ``op2``.
7998 #. ``uge``: interprets the operands as unsigned values and yields
7999 ``true`` if ``op1`` is greater than or equal to ``op2``.
8000 #. ``ult``: interprets the operands as unsigned values and yields
8001 ``true`` if ``op1`` is less than ``op2``.
8002 #. ``ule``: interprets the operands as unsigned values and yields
8003 ``true`` if ``op1`` is less than or equal to ``op2``.
8004 #. ``sgt``: interprets the operands as signed values and yields ``true``
8005 if ``op1`` is greater than ``op2``.
8006 #. ``sge``: interprets the operands as signed values and yields ``true``
8007 if ``op1`` is greater than or equal to ``op2``.
8008 #. ``slt``: interprets the operands as signed values and yields ``true``
8009 if ``op1`` is less than ``op2``.
8010 #. ``sle``: interprets the operands as signed values and yields ``true``
8011 if ``op1`` is less than or equal to ``op2``.
8013 If the operands are :ref:`pointer <t_pointer>` typed, the pointer values
8014 are compared as if they were integers.
8016 If the operands are integer vectors, then they are compared element by
8017 element. The result is an ``i1`` vector with the same number of elements
8018 as the values being compared. Otherwise, the result is an ``i1``.
8023 .. code-block:: llvm
8025 <result> = icmp eq i32 4, 5 ; yields: result=false
8026 <result> = icmp ne float* %X, %X ; yields: result=false
8027 <result> = icmp ult i16 4, 5 ; yields: result=true
8028 <result> = icmp sgt i16 4, 5 ; yields: result=false
8029 <result> = icmp ule i16 -4, 5 ; yields: result=false
8030 <result> = icmp sge i16 4, 5 ; yields: result=false
8032 Note that the code generator does not yet support vector types with the
8033 ``icmp`` instruction.
8037 '``fcmp``' Instruction
8038 ^^^^^^^^^^^^^^^^^^^^^^
8045 <result> = fcmp [fast-math flags]* <cond> <ty> <op1>, <op2> ; yields i1 or <N x i1>:result
8050 The '``fcmp``' instruction returns a boolean value or vector of boolean
8051 values based on comparison of its operands.
8053 If the operands are floating point scalars, then the result type is a
8054 boolean (:ref:`i1 <t_integer>`).
8056 If the operands are floating point vectors, then the result type is a
8057 vector of boolean with the same number of elements as the operands being
8063 The '``fcmp``' instruction takes three operands. The first operand is
8064 the condition code indicating the kind of comparison to perform. It is
8065 not a value, just a keyword. The possible condition code are:
8067 #. ``false``: no comparison, always returns false
8068 #. ``oeq``: ordered and equal
8069 #. ``ogt``: ordered and greater than
8070 #. ``oge``: ordered and greater than or equal
8071 #. ``olt``: ordered and less than
8072 #. ``ole``: ordered and less than or equal
8073 #. ``one``: ordered and not equal
8074 #. ``ord``: ordered (no nans)
8075 #. ``ueq``: unordered or equal
8076 #. ``ugt``: unordered or greater than
8077 #. ``uge``: unordered or greater than or equal
8078 #. ``ult``: unordered or less than
8079 #. ``ule``: unordered or less than or equal
8080 #. ``une``: unordered or not equal
8081 #. ``uno``: unordered (either nans)
8082 #. ``true``: no comparison, always returns true
8084 *Ordered* means that neither operand is a QNAN while *unordered* means
8085 that either operand may be a QNAN.
8087 Each of ``val1`` and ``val2`` arguments must be either a :ref:`floating
8088 point <t_floating>` type or a :ref:`vector <t_vector>` of floating point
8089 type. They must have identical types.
8094 The '``fcmp``' instruction compares ``op1`` and ``op2`` according to the
8095 condition code given as ``cond``. If the operands are vectors, then the
8096 vectors are compared element by element. Each comparison performed
8097 always yields an :ref:`i1 <t_integer>` result, as follows:
8099 #. ``false``: always yields ``false``, regardless of operands.
8100 #. ``oeq``: yields ``true`` if both operands are not a QNAN and ``op1``
8101 is equal to ``op2``.
8102 #. ``ogt``: yields ``true`` if both operands are not a QNAN and ``op1``
8103 is greater than ``op2``.
8104 #. ``oge``: yields ``true`` if both operands are not a QNAN and ``op1``
8105 is greater than or equal to ``op2``.
8106 #. ``olt``: yields ``true`` if both operands are not a QNAN and ``op1``
8107 is less than ``op2``.
8108 #. ``ole``: yields ``true`` if both operands are not a QNAN and ``op1``
8109 is less than or equal to ``op2``.
8110 #. ``one``: yields ``true`` if both operands are not a QNAN and ``op1``
8111 is not equal to ``op2``.
8112 #. ``ord``: yields ``true`` if both operands are not a QNAN.
8113 #. ``ueq``: yields ``true`` if either operand is a QNAN or ``op1`` is
8115 #. ``ugt``: yields ``true`` if either operand is a QNAN or ``op1`` is
8116 greater than ``op2``.
8117 #. ``uge``: yields ``true`` if either operand is a QNAN or ``op1`` is
8118 greater than or equal to ``op2``.
8119 #. ``ult``: yields ``true`` if either operand is a QNAN or ``op1`` is
8121 #. ``ule``: yields ``true`` if either operand is a QNAN or ``op1`` is
8122 less than or equal to ``op2``.
8123 #. ``une``: yields ``true`` if either operand is a QNAN or ``op1`` is
8124 not equal to ``op2``.
8125 #. ``uno``: yields ``true`` if either operand is a QNAN.
8126 #. ``true``: always yields ``true``, regardless of operands.
8128 The ``fcmp`` instruction can also optionally take any number of
8129 :ref:`fast-math flags <fastmath>`, which are optimization hints to enable
8130 otherwise unsafe floating point optimizations.
8132 Any set of fast-math flags are legal on an ``fcmp`` instruction, but the
8133 only flags that have any effect on its semantics are those that allow
8134 assumptions to be made about the values of input arguments; namely
8135 ``nnan``, ``ninf``, and ``nsz``. See :ref:`fastmath` for more information.
8140 .. code-block:: llvm
8142 <result> = fcmp oeq float 4.0, 5.0 ; yields: result=false
8143 <result> = fcmp one float 4.0, 5.0 ; yields: result=true
8144 <result> = fcmp olt float 4.0, 5.0 ; yields: result=true
8145 <result> = fcmp ueq double 1.0, 2.0 ; yields: result=false
8147 Note that the code generator does not yet support vector types with the
8148 ``fcmp`` instruction.
8152 '``phi``' Instruction
8153 ^^^^^^^^^^^^^^^^^^^^^
8160 <result> = phi <ty> [ <val0>, <label0>], ...
8165 The '``phi``' instruction is used to implement the φ node in the SSA
8166 graph representing the function.
8171 The type of the incoming values is specified with the first type field.
8172 After this, the '``phi``' instruction takes a list of pairs as
8173 arguments, with one pair for each predecessor basic block of the current
8174 block. Only values of :ref:`first class <t_firstclass>` type may be used as
8175 the value arguments to the PHI node. Only labels may be used as the
8178 There must be no non-phi instructions between the start of a basic block
8179 and the PHI instructions: i.e. PHI instructions must be first in a basic
8182 For the purposes of the SSA form, the use of each incoming value is
8183 deemed to occur on the edge from the corresponding predecessor block to
8184 the current block (but after any definition of an '``invoke``'
8185 instruction's return value on the same edge).
8190 At runtime, the '``phi``' instruction logically takes on the value
8191 specified by the pair corresponding to the predecessor basic block that
8192 executed just prior to the current block.
8197 .. code-block:: llvm
8199 Loop: ; Infinite loop that counts from 0 on up...
8200 %indvar = phi i32 [ 0, %LoopHeader ], [ %nextindvar, %Loop ]
8201 %nextindvar = add i32 %indvar, 1
8206 '``select``' Instruction
8207 ^^^^^^^^^^^^^^^^^^^^^^^^
8214 <result> = select selty <cond>, <ty> <val1>, <ty> <val2> ; yields ty
8216 selty is either i1 or {<N x i1>}
8221 The '``select``' instruction is used to choose one value based on a
8222 condition, without IR-level branching.
8227 The '``select``' instruction requires an 'i1' value or a vector of 'i1'
8228 values indicating the condition, and two values of the same :ref:`first
8229 class <t_firstclass>` type.
8234 If the condition is an i1 and it evaluates to 1, the instruction returns
8235 the first value argument; otherwise, it returns the second value
8238 If the condition is a vector of i1, then the value arguments must be
8239 vectors of the same size, and the selection is done element by element.
8241 If the condition is an i1 and the value arguments are vectors of the
8242 same size, then an entire vector is selected.
8247 .. code-block:: llvm
8249 %X = select i1 true, i8 17, i8 42 ; yields i8:17
8253 '``call``' Instruction
8254 ^^^^^^^^^^^^^^^^^^^^^^
8261 <result> = [tail | musttail] call [cconv] [ret attrs] <ty> [<fnty>*] <fnptrval>(<function args>) [fn attrs]
8266 The '``call``' instruction represents a simple function call.
8271 This instruction requires several arguments:
8273 #. The optional ``tail`` and ``musttail`` markers indicate that the optimizers
8274 should perform tail call optimization. The ``tail`` marker is a hint that
8275 `can be ignored <CodeGenerator.html#sibcallopt>`_. The ``musttail`` marker
8276 means that the call must be tail call optimized in order for the program to
8277 be correct. The ``musttail`` marker provides these guarantees:
8279 #. The call will not cause unbounded stack growth if it is part of a
8280 recursive cycle in the call graph.
8281 #. Arguments with the :ref:`inalloca <attr_inalloca>` attribute are
8284 Both markers imply that the callee does not access allocas or varargs from
8285 the caller. Calls marked ``musttail`` must obey the following additional
8288 - The call must immediately precede a :ref:`ret <i_ret>` instruction,
8289 or a pointer bitcast followed by a ret instruction.
8290 - The ret instruction must return the (possibly bitcasted) value
8291 produced by the call or void.
8292 - The caller and callee prototypes must match. Pointer types of
8293 parameters or return types may differ in pointee type, but not
8295 - The calling conventions of the caller and callee must match.
8296 - All ABI-impacting function attributes, such as sret, byval, inreg,
8297 returned, and inalloca, must match.
8298 - The callee must be varargs iff the caller is varargs. Bitcasting a
8299 non-varargs function to the appropriate varargs type is legal so
8300 long as the non-varargs prefixes obey the other rules.
8302 Tail call optimization for calls marked ``tail`` is guaranteed to occur if
8303 the following conditions are met:
8305 - Caller and callee both have the calling convention ``fastcc``.
8306 - The call is in tail position (ret immediately follows call and ret
8307 uses value of call or is void).
8308 - Option ``-tailcallopt`` is enabled, or
8309 ``llvm::GuaranteedTailCallOpt`` is ``true``.
8310 - `Platform-specific constraints are
8311 met. <CodeGenerator.html#tailcallopt>`_
8313 #. The optional "cconv" marker indicates which :ref:`calling
8314 convention <callingconv>` the call should use. If none is
8315 specified, the call defaults to using C calling conventions. The
8316 calling convention of the call must match the calling convention of
8317 the target function, or else the behavior is undefined.
8318 #. The optional :ref:`Parameter Attributes <paramattrs>` list for return
8319 values. Only '``zeroext``', '``signext``', and '``inreg``' attributes
8321 #. '``ty``': the type of the call instruction itself which is also the
8322 type of the return value. Functions that return no value are marked
8324 #. '``fnty``': shall be the signature of the pointer to function value
8325 being invoked. The argument types must match the types implied by
8326 this signature. This type can be omitted if the function is not
8327 varargs and if the function type does not return a pointer to a
8329 #. '``fnptrval``': An LLVM value containing a pointer to a function to
8330 be invoked. In most cases, this is a direct function invocation, but
8331 indirect ``call``'s are just as possible, calling an arbitrary pointer
8333 #. '``function args``': argument list whose types match the function
8334 signature argument types and parameter attributes. All arguments must
8335 be of :ref:`first class <t_firstclass>` type. If the function signature
8336 indicates the function accepts a variable number of arguments, the
8337 extra arguments can be specified.
8338 #. The optional :ref:`function attributes <fnattrs>` list. Only
8339 '``noreturn``', '``nounwind``', '``readonly``' and '``readnone``'
8340 attributes are valid here.
8345 The '``call``' instruction is used to cause control flow to transfer to
8346 a specified function, with its incoming arguments bound to the specified
8347 values. Upon a '``ret``' instruction in the called function, control
8348 flow continues with the instruction after the function call, and the
8349 return value of the function is bound to the result argument.
8354 .. code-block:: llvm
8356 %retval = call i32 @test(i32 %argc)
8357 call i32 (i8*, ...)* @printf(i8* %msg, i32 12, i8 42) ; yields i32
8358 %X = tail call i32 @foo() ; yields i32
8359 %Y = tail call fastcc i32 @foo() ; yields i32
8360 call void %foo(i8 97 signext)
8362 %struct.A = type { i32, i8 }
8363 %r = call %struct.A @foo() ; yields { i32, i8 }
8364 %gr = extractvalue %struct.A %r, 0 ; yields i32
8365 %gr1 = extractvalue %struct.A %r, 1 ; yields i8
8366 %Z = call void @foo() noreturn ; indicates that %foo never returns normally
8367 %ZZ = call zeroext i32 @bar() ; Return value is %zero extended
8369 llvm treats calls to some functions with names and arguments that match
8370 the standard C99 library as being the C99 library functions, and may
8371 perform optimizations or generate code for them under that assumption.
8372 This is something we'd like to change in the future to provide better
8373 support for freestanding environments and non-C-based languages.
8377 '``va_arg``' Instruction
8378 ^^^^^^^^^^^^^^^^^^^^^^^^
8385 <resultval> = va_arg <va_list*> <arglist>, <argty>
8390 The '``va_arg``' instruction is used to access arguments passed through
8391 the "variable argument" area of a function call. It is used to implement
8392 the ``va_arg`` macro in C.
8397 This instruction takes a ``va_list*`` value and the type of the
8398 argument. It returns a value of the specified argument type and
8399 increments the ``va_list`` to point to the next argument. The actual
8400 type of ``va_list`` is target specific.
8405 The '``va_arg``' instruction loads an argument of the specified type
8406 from the specified ``va_list`` and causes the ``va_list`` to point to
8407 the next argument. For more information, see the variable argument
8408 handling :ref:`Intrinsic Functions <int_varargs>`.
8410 It is legal for this instruction to be called in a function which does
8411 not take a variable number of arguments, for example, the ``vfprintf``
8414 ``va_arg`` is an LLVM instruction instead of an :ref:`intrinsic
8415 function <intrinsics>` because it takes a type as an argument.
8420 See the :ref:`variable argument processing <int_varargs>` section.
8422 Note that the code generator does not yet fully support va\_arg on many
8423 targets. Also, it does not currently support va\_arg with aggregate
8424 types on any target.
8428 '``landingpad``' Instruction
8429 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
8436 <resultval> = landingpad <resultty> <clause>+
8437 <resultval> = landingpad <resultty> cleanup <clause>*
8439 <clause> := catch <type> <value>
8440 <clause> := filter <array constant type> <array constant>
8445 The '``landingpad``' instruction is used by `LLVM's exception handling
8446 system <ExceptionHandling.html#overview>`_ to specify that a basic block
8447 is a landing pad --- one where the exception lands, and corresponds to the
8448 code found in the ``catch`` portion of a ``try``/``catch`` sequence. It
8449 defines values supplied by the :ref:`personality function <personalityfn>` upon
8450 re-entry to the function. The ``resultval`` has the type ``resultty``.
8456 ``cleanup`` flag indicates that the landing pad block is a cleanup.
8458 A ``clause`` begins with the clause type --- ``catch`` or ``filter`` --- and
8459 contains the global variable representing the "type" that may be caught
8460 or filtered respectively. Unlike the ``catch`` clause, the ``filter``
8461 clause takes an array constant as its argument. Use
8462 "``[0 x i8**] undef``" for a filter which cannot throw. The
8463 '``landingpad``' instruction must contain *at least* one ``clause`` or
8464 the ``cleanup`` flag.
8469 The '``landingpad``' instruction defines the values which are set by the
8470 :ref:`personality function <personalityfn>` upon re-entry to the function, and
8471 therefore the "result type" of the ``landingpad`` instruction. As with
8472 calling conventions, how the personality function results are
8473 represented in LLVM IR is target specific.
8475 The clauses are applied in order from top to bottom. If two
8476 ``landingpad`` instructions are merged together through inlining, the
8477 clauses from the calling function are appended to the list of clauses.
8478 When the call stack is being unwound due to an exception being thrown,
8479 the exception is compared against each ``clause`` in turn. If it doesn't
8480 match any of the clauses, and the ``cleanup`` flag is not set, then
8481 unwinding continues further up the call stack.
8483 The ``landingpad`` instruction has several restrictions:
8485 - A landing pad block is a basic block which is the unwind destination
8486 of an '``invoke``' instruction.
8487 - A landing pad block must have a '``landingpad``' instruction as its
8488 first non-PHI instruction.
8489 - There can be only one '``landingpad``' instruction within the landing
8491 - A basic block that is not a landing pad block may not include a
8492 '``landingpad``' instruction.
8497 .. code-block:: llvm
8499 ;; A landing pad which can catch an integer.
8500 %res = landingpad { i8*, i32 }
8502 ;; A landing pad that is a cleanup.
8503 %res = landingpad { i8*, i32 }
8505 ;; A landing pad which can catch an integer and can only throw a double.
8506 %res = landingpad { i8*, i32 }
8508 filter [1 x i8**] [@_ZTId]
8512 '``cleanuppad``' Instruction
8513 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
8520 <resultval> = cleanuppad [<args>*]
8525 The '``cleanuppad``' instruction is used by `LLVM's exception handling
8526 system <ExceptionHandling.html#overview>`_ to specify that a basic block
8527 is a cleanup block --- one where a personality routine attempts to
8528 transfer control to run cleanup actions.
8529 The ``args`` correspond to whatever additional
8530 information the :ref:`personality function <personalityfn>` requires to
8531 execute the cleanup.
8532 The ``resultval`` has the type :ref:`token <t_token>` and is used to
8533 match the ``cleanuppad`` to corresponding :ref:`cleanuprets <i_cleanupret>`
8534 and :ref:`cleanupendpads <i_cleanupendpad>`.
8539 The instruction takes a list of arbitrary values which are interpreted
8540 by the :ref:`personality function <personalityfn>`.
8545 When the call stack is being unwound due to an exception being thrown,
8546 the :ref:`personality function <personalityfn>` transfers control to the
8547 ``cleanuppad`` with the aid of the personality-specific arguments.
8548 As with calling conventions, how the personality function results are
8549 represented in LLVM IR is target specific.
8551 The ``cleanuppad`` instruction has several restrictions:
8553 - A cleanup block is a basic block which is the unwind destination of
8554 an exceptional instruction.
8555 - A cleanup block must have a '``cleanuppad``' instruction as its
8556 first non-PHI instruction.
8557 - There can be only one '``cleanuppad``' instruction within the
8559 - A basic block that is not a cleanup block may not include a
8560 '``cleanuppad``' instruction.
8561 - All '``cleanupret``'s and '``cleanupendpad``'s which consume a ``cleanuppad``
8562 must have the same exceptional successor.
8563 - It is undefined behavior for control to transfer from a ``cleanuppad`` to a
8564 ``ret`` without first executing a ``cleanupret`` or ``cleanupendpad`` that
8565 consumes the ``cleanuppad``.
8566 - It is undefined behavior for control to transfer from a ``cleanuppad`` to
8567 itself without first executing a ``cleanupret`` or ``cleanupendpad`` that
8568 consumes the ``cleanuppad``.
8573 .. code-block:: llvm
8575 %tok = cleanuppad []
8582 LLVM supports the notion of an "intrinsic function". These functions
8583 have well known names and semantics and are required to follow certain
8584 restrictions. Overall, these intrinsics represent an extension mechanism
8585 for the LLVM language that does not require changing all of the
8586 transformations in LLVM when adding to the language (or the bitcode
8587 reader/writer, the parser, etc...).
8589 Intrinsic function names must all start with an "``llvm.``" prefix. This
8590 prefix is reserved in LLVM for intrinsic names; thus, function names may
8591 not begin with this prefix. Intrinsic functions must always be external
8592 functions: you cannot define the body of intrinsic functions. Intrinsic
8593 functions may only be used in call or invoke instructions: it is illegal
8594 to take the address of an intrinsic function. Additionally, because
8595 intrinsic functions are part of the LLVM language, it is required if any
8596 are added that they be documented here.
8598 Some intrinsic functions can be overloaded, i.e., the intrinsic
8599 represents a family of functions that perform the same operation but on
8600 different data types. Because LLVM can represent over 8 million
8601 different integer types, overloading is used commonly to allow an
8602 intrinsic function to operate on any integer type. One or more of the
8603 argument types or the result type can be overloaded to accept any
8604 integer type. Argument types may also be defined as exactly matching a
8605 previous argument's type or the result type. This allows an intrinsic
8606 function which accepts multiple arguments, but needs all of them to be
8607 of the same type, to only be overloaded with respect to a single
8608 argument or the result.
8610 Overloaded intrinsics will have the names of its overloaded argument
8611 types encoded into its function name, each preceded by a period. Only
8612 those types which are overloaded result in a name suffix. Arguments
8613 whose type is matched against another type do not. For example, the
8614 ``llvm.ctpop`` function can take an integer of any width and returns an
8615 integer of exactly the same integer width. This leads to a family of
8616 functions such as ``i8 @llvm.ctpop.i8(i8 %val)`` and
8617 ``i29 @llvm.ctpop.i29(i29 %val)``. Only one type, the return type, is
8618 overloaded, and only one type suffix is required. Because the argument's
8619 type is matched against the return type, it does not require its own
8622 To learn how to add an intrinsic function, please see the `Extending
8623 LLVM Guide <ExtendingLLVM.html>`_.
8627 Variable Argument Handling Intrinsics
8628 -------------------------------------
8630 Variable argument support is defined in LLVM with the
8631 :ref:`va_arg <i_va_arg>` instruction and these three intrinsic
8632 functions. These functions are related to the similarly named macros
8633 defined in the ``<stdarg.h>`` header file.
8635 All of these functions operate on arguments that use a target-specific
8636 value type "``va_list``". The LLVM assembly language reference manual
8637 does not define what this type is, so all transformations should be
8638 prepared to handle these functions regardless of the type used.
8640 This example shows how the :ref:`va_arg <i_va_arg>` instruction and the
8641 variable argument handling intrinsic functions are used.
8643 .. code-block:: llvm
8645 ; This struct is different for every platform. For most platforms,
8646 ; it is merely an i8*.
8647 %struct.va_list = type { i8* }
8649 ; For Unix x86_64 platforms, va_list is the following struct:
8650 ; %struct.va_list = type { i32, i32, i8*, i8* }
8652 define i32 @test(i32 %X, ...) {
8653 ; Initialize variable argument processing
8654 %ap = alloca %struct.va_list
8655 %ap2 = bitcast %struct.va_list* %ap to i8*
8656 call void @llvm.va_start(i8* %ap2)
8658 ; Read a single integer argument
8659 %tmp = va_arg i8* %ap2, i32
8661 ; Demonstrate usage of llvm.va_copy and llvm.va_end
8663 %aq2 = bitcast i8** %aq to i8*
8664 call void @llvm.va_copy(i8* %aq2, i8* %ap2)
8665 call void @llvm.va_end(i8* %aq2)
8667 ; Stop processing of arguments.
8668 call void @llvm.va_end(i8* %ap2)
8672 declare void @llvm.va_start(i8*)
8673 declare void @llvm.va_copy(i8*, i8*)
8674 declare void @llvm.va_end(i8*)
8678 '``llvm.va_start``' Intrinsic
8679 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
8686 declare void @llvm.va_start(i8* <arglist>)
8691 The '``llvm.va_start``' intrinsic initializes ``*<arglist>`` for
8692 subsequent use by ``va_arg``.
8697 The argument is a pointer to a ``va_list`` element to initialize.
8702 The '``llvm.va_start``' intrinsic works just like the ``va_start`` macro
8703 available in C. In a target-dependent way, it initializes the
8704 ``va_list`` element to which the argument points, so that the next call
8705 to ``va_arg`` will produce the first variable argument passed to the
8706 function. Unlike the C ``va_start`` macro, this intrinsic does not need
8707 to know the last argument of the function as the compiler can figure
8710 '``llvm.va_end``' Intrinsic
8711 ^^^^^^^^^^^^^^^^^^^^^^^^^^^
8718 declare void @llvm.va_end(i8* <arglist>)
8723 The '``llvm.va_end``' intrinsic destroys ``*<arglist>``, which has been
8724 initialized previously with ``llvm.va_start`` or ``llvm.va_copy``.
8729 The argument is a pointer to a ``va_list`` to destroy.
8734 The '``llvm.va_end``' intrinsic works just like the ``va_end`` macro
8735 available in C. In a target-dependent way, it destroys the ``va_list``
8736 element to which the argument points. Calls to
8737 :ref:`llvm.va_start <int_va_start>` and
8738 :ref:`llvm.va_copy <int_va_copy>` must be matched exactly with calls to
8743 '``llvm.va_copy``' Intrinsic
8744 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
8751 declare void @llvm.va_copy(i8* <destarglist>, i8* <srcarglist>)
8756 The '``llvm.va_copy``' intrinsic copies the current argument position
8757 from the source argument list to the destination argument list.
8762 The first argument is a pointer to a ``va_list`` element to initialize.
8763 The second argument is a pointer to a ``va_list`` element to copy from.
8768 The '``llvm.va_copy``' intrinsic works just like the ``va_copy`` macro
8769 available in C. In a target-dependent way, it copies the source
8770 ``va_list`` element into the destination ``va_list`` element. This
8771 intrinsic is necessary because the `` llvm.va_start`` intrinsic may be
8772 arbitrarily complex and require, for example, memory allocation.
8774 Accurate Garbage Collection Intrinsics
8775 --------------------------------------
8777 LLVM's support for `Accurate Garbage Collection <GarbageCollection.html>`_
8778 (GC) requires the frontend to generate code containing appropriate intrinsic
8779 calls and select an appropriate GC strategy which knows how to lower these
8780 intrinsics in a manner which is appropriate for the target collector.
8782 These intrinsics allow identification of :ref:`GC roots on the
8783 stack <int_gcroot>`, as well as garbage collector implementations that
8784 require :ref:`read <int_gcread>` and :ref:`write <int_gcwrite>` barriers.
8785 Frontends for type-safe garbage collected languages should generate
8786 these intrinsics to make use of the LLVM garbage collectors. For more
8787 details, see `Garbage Collection with LLVM <GarbageCollection.html>`_.
8789 Experimental Statepoint Intrinsics
8790 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
8792 LLVM provides an second experimental set of intrinsics for describing garbage
8793 collection safepoints in compiled code. These intrinsics are an alternative
8794 to the ``llvm.gcroot`` intrinsics, but are compatible with the ones for
8795 :ref:`read <int_gcread>` and :ref:`write <int_gcwrite>` barriers. The
8796 differences in approach are covered in the `Garbage Collection with LLVM
8797 <GarbageCollection.html>`_ documentation. The intrinsics themselves are
8798 described in :doc:`Statepoints`.
8802 '``llvm.gcroot``' Intrinsic
8803 ^^^^^^^^^^^^^^^^^^^^^^^^^^^
8810 declare void @llvm.gcroot(i8** %ptrloc, i8* %metadata)
8815 The '``llvm.gcroot``' intrinsic declares the existence of a GC root to
8816 the code generator, and allows some metadata to be associated with it.
8821 The first argument specifies the address of a stack object that contains
8822 the root pointer. The second pointer (which must be either a constant or
8823 a global value address) contains the meta-data to be associated with the
8829 At runtime, a call to this intrinsic stores a null pointer into the
8830 "ptrloc" location. At compile-time, the code generator generates
8831 information to allow the runtime to find the pointer at GC safe points.
8832 The '``llvm.gcroot``' intrinsic may only be used in a function which
8833 :ref:`specifies a GC algorithm <gc>`.
8837 '``llvm.gcread``' Intrinsic
8838 ^^^^^^^^^^^^^^^^^^^^^^^^^^^
8845 declare i8* @llvm.gcread(i8* %ObjPtr, i8** %Ptr)
8850 The '``llvm.gcread``' intrinsic identifies reads of references from heap
8851 locations, allowing garbage collector implementations that require read
8857 The second argument is the address to read from, which should be an
8858 address allocated from the garbage collector. The first object is a
8859 pointer to the start of the referenced object, if needed by the language
8860 runtime (otherwise null).
8865 The '``llvm.gcread``' intrinsic has the same semantics as a load
8866 instruction, but may be replaced with substantially more complex code by
8867 the garbage collector runtime, as needed. The '``llvm.gcread``'
8868 intrinsic may only be used in a function which :ref:`specifies a GC
8873 '``llvm.gcwrite``' Intrinsic
8874 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
8881 declare void @llvm.gcwrite(i8* %P1, i8* %Obj, i8** %P2)
8886 The '``llvm.gcwrite``' intrinsic identifies writes of references to heap
8887 locations, allowing garbage collector implementations that require write
8888 barriers (such as generational or reference counting collectors).
8893 The first argument is the reference to store, the second is the start of
8894 the object to store it to, and the third is the address of the field of
8895 Obj to store to. If the runtime does not require a pointer to the
8896 object, Obj may be null.
8901 The '``llvm.gcwrite``' intrinsic has the same semantics as a store
8902 instruction, but may be replaced with substantially more complex code by
8903 the garbage collector runtime, as needed. The '``llvm.gcwrite``'
8904 intrinsic may only be used in a function which :ref:`specifies a GC
8907 Code Generator Intrinsics
8908 -------------------------
8910 These intrinsics are provided by LLVM to expose special features that
8911 may only be implemented with code generator support.
8913 '``llvm.returnaddress``' Intrinsic
8914 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
8921 declare i8 *@llvm.returnaddress(i32 <level>)
8926 The '``llvm.returnaddress``' intrinsic attempts to compute a
8927 target-specific value indicating the return address of the current
8928 function or one of its callers.
8933 The argument to this intrinsic indicates which function to return the
8934 address for. Zero indicates the calling function, one indicates its
8935 caller, etc. The argument is **required** to be a constant integer
8941 The '``llvm.returnaddress``' intrinsic either returns a pointer
8942 indicating the return address of the specified call frame, or zero if it
8943 cannot be identified. The value returned by this intrinsic is likely to
8944 be incorrect or 0 for arguments other than zero, so it should only be
8945 used for debugging purposes.
8947 Note that calling this intrinsic does not prevent function inlining or
8948 other aggressive transformations, so the value returned may not be that
8949 of the obvious source-language caller.
8951 '``llvm.frameaddress``' Intrinsic
8952 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
8959 declare i8* @llvm.frameaddress(i32 <level>)
8964 The '``llvm.frameaddress``' intrinsic attempts to return the
8965 target-specific frame pointer value for the specified stack frame.
8970 The argument to this intrinsic indicates which function to return the
8971 frame pointer for. Zero indicates the calling function, one indicates
8972 its caller, etc. The argument is **required** to be a constant integer
8978 The '``llvm.frameaddress``' intrinsic either returns a pointer
8979 indicating the frame address of the specified call frame, or zero if it
8980 cannot be identified. The value returned by this intrinsic is likely to
8981 be incorrect or 0 for arguments other than zero, so it should only be
8982 used for debugging purposes.
8984 Note that calling this intrinsic does not prevent function inlining or
8985 other aggressive transformations, so the value returned may not be that
8986 of the obvious source-language caller.
8988 '``llvm.localescape``' and '``llvm.localrecover``' Intrinsics
8989 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
8996 declare void @llvm.localescape(...)
8997 declare i8* @llvm.localrecover(i8* %func, i8* %fp, i32 %idx)
9002 The '``llvm.localescape``' intrinsic escapes offsets of a collection of static
9003 allocas, and the '``llvm.localrecover``' intrinsic applies those offsets to a
9004 live frame pointer to recover the address of the allocation. The offset is
9005 computed during frame layout of the caller of ``llvm.localescape``.
9010 All arguments to '``llvm.localescape``' must be pointers to static allocas or
9011 casts of static allocas. Each function can only call '``llvm.localescape``'
9012 once, and it can only do so from the entry block.
9014 The ``func`` argument to '``llvm.localrecover``' must be a constant
9015 bitcasted pointer to a function defined in the current module. The code
9016 generator cannot determine the frame allocation offset of functions defined in
9019 The ``fp`` argument to '``llvm.localrecover``' must be a frame pointer of a
9020 call frame that is currently live. The return value of '``llvm.localaddress``'
9021 is one way to produce such a value, but various runtimes also expose a suitable
9022 pointer in platform-specific ways.
9024 The ``idx`` argument to '``llvm.localrecover``' indicates which alloca passed to
9025 '``llvm.localescape``' to recover. It is zero-indexed.
9030 These intrinsics allow a group of functions to share access to a set of local
9031 stack allocations of a one parent function. The parent function may call the
9032 '``llvm.localescape``' intrinsic once from the function entry block, and the
9033 child functions can use '``llvm.localrecover``' to access the escaped allocas.
9034 The '``llvm.localescape``' intrinsic blocks inlining, as inlining changes where
9035 the escaped allocas are allocated, which would break attempts to use
9036 '``llvm.localrecover``'.
9038 .. _int_read_register:
9039 .. _int_write_register:
9041 '``llvm.read_register``' and '``llvm.write_register``' Intrinsics
9042 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
9049 declare i32 @llvm.read_register.i32(metadata)
9050 declare i64 @llvm.read_register.i64(metadata)
9051 declare void @llvm.write_register.i32(metadata, i32 @value)
9052 declare void @llvm.write_register.i64(metadata, i64 @value)
9058 The '``llvm.read_register``' and '``llvm.write_register``' intrinsics
9059 provides access to the named register. The register must be valid on
9060 the architecture being compiled to. The type needs to be compatible
9061 with the register being read.
9066 The '``llvm.read_register``' intrinsic returns the current value of the
9067 register, where possible. The '``llvm.write_register``' intrinsic sets
9068 the current value of the register, where possible.
9070 This is useful to implement named register global variables that need
9071 to always be mapped to a specific register, as is common practice on
9072 bare-metal programs including OS kernels.
9074 The compiler doesn't check for register availability or use of the used
9075 register in surrounding code, including inline assembly. Because of that,
9076 allocatable registers are not supported.
9078 Warning: So far it only works with the stack pointer on selected
9079 architectures (ARM, AArch64, PowerPC and x86_64). Significant amount of
9080 work is needed to support other registers and even more so, allocatable
9085 '``llvm.stacksave``' Intrinsic
9086 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
9093 declare i8* @llvm.stacksave()
9098 The '``llvm.stacksave``' intrinsic is used to remember the current state
9099 of the function stack, for use with
9100 :ref:`llvm.stackrestore <int_stackrestore>`. This is useful for
9101 implementing language features like scoped automatic variable sized
9107 This intrinsic returns a opaque pointer value that can be passed to
9108 :ref:`llvm.stackrestore <int_stackrestore>`. When an
9109 ``llvm.stackrestore`` intrinsic is executed with a value saved from
9110 ``llvm.stacksave``, it effectively restores the state of the stack to
9111 the state it was in when the ``llvm.stacksave`` intrinsic executed. In
9112 practice, this pops any :ref:`alloca <i_alloca>` blocks from the stack that
9113 were allocated after the ``llvm.stacksave`` was executed.
9115 .. _int_stackrestore:
9117 '``llvm.stackrestore``' Intrinsic
9118 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
9125 declare void @llvm.stackrestore(i8* %ptr)
9130 The '``llvm.stackrestore``' intrinsic is used to restore the state of
9131 the function stack to the state it was in when the corresponding
9132 :ref:`llvm.stacksave <int_stacksave>` intrinsic executed. This is
9133 useful for implementing language features like scoped automatic variable
9134 sized arrays in C99.
9139 See the description for :ref:`llvm.stacksave <int_stacksave>`.
9141 '``llvm.prefetch``' Intrinsic
9142 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
9149 declare void @llvm.prefetch(i8* <address>, i32 <rw>, i32 <locality>, i32 <cache type>)
9154 The '``llvm.prefetch``' intrinsic is a hint to the code generator to
9155 insert a prefetch instruction if supported; otherwise, it is a noop.
9156 Prefetches have no effect on the behavior of the program but can change
9157 its performance characteristics.
9162 ``address`` is the address to be prefetched, ``rw`` is the specifier
9163 determining if the fetch should be for a read (0) or write (1), and
9164 ``locality`` is a temporal locality specifier ranging from (0) - no
9165 locality, to (3) - extremely local keep in cache. The ``cache type``
9166 specifies whether the prefetch is performed on the data (1) or
9167 instruction (0) cache. The ``rw``, ``locality`` and ``cache type``
9168 arguments must be constant integers.
9173 This intrinsic does not modify the behavior of the program. In
9174 particular, prefetches cannot trap and do not produce a value. On
9175 targets that support this intrinsic, the prefetch can provide hints to
9176 the processor cache for better performance.
9178 '``llvm.pcmarker``' Intrinsic
9179 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
9186 declare void @llvm.pcmarker(i32 <id>)
9191 The '``llvm.pcmarker``' intrinsic is a method to export a Program
9192 Counter (PC) in a region of code to simulators and other tools. The
9193 method is target specific, but it is expected that the marker will use
9194 exported symbols to transmit the PC of the marker. The marker makes no
9195 guarantees that it will remain with any specific instruction after
9196 optimizations. It is possible that the presence of a marker will inhibit
9197 optimizations. The intended use is to be inserted after optimizations to
9198 allow correlations of simulation runs.
9203 ``id`` is a numerical id identifying the marker.
9208 This intrinsic does not modify the behavior of the program. Backends
9209 that do not support this intrinsic may ignore it.
9211 '``llvm.readcyclecounter``' Intrinsic
9212 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
9219 declare i64 @llvm.readcyclecounter()
9224 The '``llvm.readcyclecounter``' intrinsic provides access to the cycle
9225 counter register (or similar low latency, high accuracy clocks) on those
9226 targets that support it. On X86, it should map to RDTSC. On Alpha, it
9227 should map to RPCC. As the backing counters overflow quickly (on the
9228 order of 9 seconds on alpha), this should only be used for small
9234 When directly supported, reading the cycle counter should not modify any
9235 memory. Implementations are allowed to either return a application
9236 specific value or a system wide value. On backends without support, this
9237 is lowered to a constant 0.
9239 Note that runtime support may be conditional on the privilege-level code is
9240 running at and the host platform.
9242 '``llvm.clear_cache``' Intrinsic
9243 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
9250 declare void @llvm.clear_cache(i8*, i8*)
9255 The '``llvm.clear_cache``' intrinsic ensures visibility of modifications
9256 in the specified range to the execution unit of the processor. On
9257 targets with non-unified instruction and data cache, the implementation
9258 flushes the instruction cache.
9263 On platforms with coherent instruction and data caches (e.g. x86), this
9264 intrinsic is a nop. On platforms with non-coherent instruction and data
9265 cache (e.g. ARM, MIPS), the intrinsic is lowered either to appropriate
9266 instructions or a system call, if cache flushing requires special
9269 The default behavior is to emit a call to ``__clear_cache`` from the run
9272 This instrinsic does *not* empty the instruction pipeline. Modifications
9273 of the current function are outside the scope of the intrinsic.
9275 '``llvm.instrprof_increment``' Intrinsic
9276 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
9283 declare void @llvm.instrprof_increment(i8* <name>, i64 <hash>,
9284 i32 <num-counters>, i32 <index>)
9289 The '``llvm.instrprof_increment``' intrinsic can be emitted by a
9290 frontend for use with instrumentation based profiling. These will be
9291 lowered by the ``-instrprof`` pass to generate execution counts of a
9297 The first argument is a pointer to a global variable containing the
9298 name of the entity being instrumented. This should generally be the
9299 (mangled) function name for a set of counters.
9301 The second argument is a hash value that can be used by the consumer
9302 of the profile data to detect changes to the instrumented source, and
9303 the third is the number of counters associated with ``name``. It is an
9304 error if ``hash`` or ``num-counters`` differ between two instances of
9305 ``instrprof_increment`` that refer to the same name.
9307 The last argument refers to which of the counters for ``name`` should
9308 be incremented. It should be a value between 0 and ``num-counters``.
9313 This intrinsic represents an increment of a profiling counter. It will
9314 cause the ``-instrprof`` pass to generate the appropriate data
9315 structures and the code to increment the appropriate value, in a
9316 format that can be written out by a compiler runtime and consumed via
9317 the ``llvm-profdata`` tool.
9319 Standard C Library Intrinsics
9320 -----------------------------
9322 LLVM provides intrinsics for a few important standard C library
9323 functions. These intrinsics allow source-language front-ends to pass
9324 information about the alignment of the pointer arguments to the code
9325 generator, providing opportunity for more efficient code generation.
9329 '``llvm.memcpy``' Intrinsic
9330 ^^^^^^^^^^^^^^^^^^^^^^^^^^^
9335 This is an overloaded intrinsic. You can use ``llvm.memcpy`` on any
9336 integer bit width and for different address spaces. Not all targets
9337 support all bit widths however.
9341 declare void @llvm.memcpy.p0i8.p0i8.i32(i8* <dest>, i8* <src>,
9342 i32 <len>, i32 <align>, i1 <isvolatile>)
9343 declare void @llvm.memcpy.p0i8.p0i8.i64(i8* <dest>, i8* <src>,
9344 i64 <len>, i32 <align>, i1 <isvolatile>)
9349 The '``llvm.memcpy.*``' intrinsics copy a block of memory from the
9350 source location to the destination location.
9352 Note that, unlike the standard libc function, the ``llvm.memcpy.*``
9353 intrinsics do not return a value, takes extra alignment/isvolatile
9354 arguments and the pointers can be in specified address spaces.
9359 The first argument is a pointer to the destination, the second is a
9360 pointer to the source. The third argument is an integer argument
9361 specifying the number of bytes to copy, the fourth argument is the
9362 alignment of the source and destination locations, and the fifth is a
9363 boolean indicating a volatile access.
9365 If the call to this intrinsic has an alignment value that is not 0 or 1,
9366 then the caller guarantees that both the source and destination pointers
9367 are aligned to that boundary.
9369 If the ``isvolatile`` parameter is ``true``, the ``llvm.memcpy`` call is
9370 a :ref:`volatile operation <volatile>`. The detailed access behavior is not
9371 very cleanly specified and it is unwise to depend on it.
9376 The '``llvm.memcpy.*``' intrinsics copy a block of memory from the
9377 source location to the destination location, which are not allowed to
9378 overlap. It copies "len" bytes of memory over. If the argument is known
9379 to be aligned to some boundary, this can be specified as the fourth
9380 argument, otherwise it should be set to 0 or 1 (both meaning no alignment).
9382 '``llvm.memmove``' Intrinsic
9383 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
9388 This is an overloaded intrinsic. You can use llvm.memmove on any integer
9389 bit width and for different address space. Not all targets support all
9394 declare void @llvm.memmove.p0i8.p0i8.i32(i8* <dest>, i8* <src>,
9395 i32 <len>, i32 <align>, i1 <isvolatile>)
9396 declare void @llvm.memmove.p0i8.p0i8.i64(i8* <dest>, i8* <src>,
9397 i64 <len>, i32 <align>, i1 <isvolatile>)
9402 The '``llvm.memmove.*``' intrinsics move a block of memory from the
9403 source location to the destination location. It is similar to the
9404 '``llvm.memcpy``' intrinsic but allows the two memory locations to
9407 Note that, unlike the standard libc function, the ``llvm.memmove.*``
9408 intrinsics do not return a value, takes extra alignment/isvolatile
9409 arguments and the pointers can be in specified address spaces.
9414 The first argument is a pointer to the destination, the second is a
9415 pointer to the source. The third argument is an integer argument
9416 specifying the number of bytes to copy, the fourth argument is the
9417 alignment of the source and destination locations, and the fifth is a
9418 boolean indicating a volatile access.
9420 If the call to this intrinsic has an alignment value that is not 0 or 1,
9421 then the caller guarantees that the source and destination pointers are
9422 aligned to that boundary.
9424 If the ``isvolatile`` parameter is ``true``, the ``llvm.memmove`` call
9425 is a :ref:`volatile operation <volatile>`. The detailed access behavior is
9426 not very cleanly specified and it is unwise to depend on it.
9431 The '``llvm.memmove.*``' intrinsics copy a block of memory from the
9432 source location to the destination location, which may overlap. It
9433 copies "len" bytes of memory over. If the argument is known to be
9434 aligned to some boundary, this can be specified as the fourth argument,
9435 otherwise it should be set to 0 or 1 (both meaning no alignment).
9437 '``llvm.memset.*``' Intrinsics
9438 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
9443 This is an overloaded intrinsic. You can use llvm.memset on any integer
9444 bit width and for different address spaces. However, not all targets
9445 support all bit widths.
9449 declare void @llvm.memset.p0i8.i32(i8* <dest>, i8 <val>,
9450 i32 <len>, i32 <align>, i1 <isvolatile>)
9451 declare void @llvm.memset.p0i8.i64(i8* <dest>, i8 <val>,
9452 i64 <len>, i32 <align>, i1 <isvolatile>)
9457 The '``llvm.memset.*``' intrinsics fill a block of memory with a
9458 particular byte value.
9460 Note that, unlike the standard libc function, the ``llvm.memset``
9461 intrinsic does not return a value and takes extra alignment/volatile
9462 arguments. Also, the destination can be in an arbitrary address space.
9467 The first argument is a pointer to the destination to fill, the second
9468 is the byte value with which to fill it, the third argument is an
9469 integer argument specifying the number of bytes to fill, and the fourth
9470 argument is the known alignment of the destination location.
9472 If the call to this intrinsic has an alignment value that is not 0 or 1,
9473 then the caller guarantees that the destination pointer is aligned to
9476 If the ``isvolatile`` parameter is ``true``, the ``llvm.memset`` call is
9477 a :ref:`volatile operation <volatile>`. The detailed access behavior is not
9478 very cleanly specified and it is unwise to depend on it.
9483 The '``llvm.memset.*``' intrinsics fill "len" bytes of memory starting
9484 at the destination location. If the argument is known to be aligned to
9485 some boundary, this can be specified as the fourth argument, otherwise
9486 it should be set to 0 or 1 (both meaning no alignment).
9488 '``llvm.sqrt.*``' Intrinsic
9489 ^^^^^^^^^^^^^^^^^^^^^^^^^^^
9494 This is an overloaded intrinsic. You can use ``llvm.sqrt`` on any
9495 floating point or vector of floating point type. Not all targets support
9500 declare float @llvm.sqrt.f32(float %Val)
9501 declare double @llvm.sqrt.f64(double %Val)
9502 declare x86_fp80 @llvm.sqrt.f80(x86_fp80 %Val)
9503 declare fp128 @llvm.sqrt.f128(fp128 %Val)
9504 declare ppc_fp128 @llvm.sqrt.ppcf128(ppc_fp128 %Val)
9509 The '``llvm.sqrt``' intrinsics return the sqrt of the specified operand,
9510 returning the same value as the libm '``sqrt``' functions would. Unlike
9511 ``sqrt`` in libm, however, ``llvm.sqrt`` has undefined behavior for
9512 negative numbers other than -0.0 (which allows for better optimization,
9513 because there is no need to worry about errno being set).
9514 ``llvm.sqrt(-0.0)`` is defined to return -0.0 like IEEE sqrt.
9519 The argument and return value are floating point numbers of the same
9525 This function returns the sqrt of the specified operand if it is a
9526 nonnegative floating point number.
9528 '``llvm.powi.*``' Intrinsic
9529 ^^^^^^^^^^^^^^^^^^^^^^^^^^^
9534 This is an overloaded intrinsic. You can use ``llvm.powi`` on any
9535 floating point or vector of floating point type. Not all targets support
9540 declare float @llvm.powi.f32(float %Val, i32 %power)
9541 declare double @llvm.powi.f64(double %Val, i32 %power)
9542 declare x86_fp80 @llvm.powi.f80(x86_fp80 %Val, i32 %power)
9543 declare fp128 @llvm.powi.f128(fp128 %Val, i32 %power)
9544 declare ppc_fp128 @llvm.powi.ppcf128(ppc_fp128 %Val, i32 %power)
9549 The '``llvm.powi.*``' intrinsics return the first operand raised to the
9550 specified (positive or negative) power. The order of evaluation of
9551 multiplications is not defined. When a vector of floating point type is
9552 used, the second argument remains a scalar integer value.
9557 The second argument is an integer power, and the first is a value to
9558 raise to that power.
9563 This function returns the first value raised to the second power with an
9564 unspecified sequence of rounding operations.
9566 '``llvm.sin.*``' Intrinsic
9567 ^^^^^^^^^^^^^^^^^^^^^^^^^^
9572 This is an overloaded intrinsic. You can use ``llvm.sin`` on any
9573 floating point or vector of floating point type. Not all targets support
9578 declare float @llvm.sin.f32(float %Val)
9579 declare double @llvm.sin.f64(double %Val)
9580 declare x86_fp80 @llvm.sin.f80(x86_fp80 %Val)
9581 declare fp128 @llvm.sin.f128(fp128 %Val)
9582 declare ppc_fp128 @llvm.sin.ppcf128(ppc_fp128 %Val)
9587 The '``llvm.sin.*``' intrinsics return the sine of the operand.
9592 The argument and return value are floating point numbers of the same
9598 This function returns the sine of the specified operand, returning the
9599 same values as the libm ``sin`` functions would, and handles error
9600 conditions in the same way.
9602 '``llvm.cos.*``' Intrinsic
9603 ^^^^^^^^^^^^^^^^^^^^^^^^^^
9608 This is an overloaded intrinsic. You can use ``llvm.cos`` on any
9609 floating point or vector of floating point type. Not all targets support
9614 declare float @llvm.cos.f32(float %Val)
9615 declare double @llvm.cos.f64(double %Val)
9616 declare x86_fp80 @llvm.cos.f80(x86_fp80 %Val)
9617 declare fp128 @llvm.cos.f128(fp128 %Val)
9618 declare ppc_fp128 @llvm.cos.ppcf128(ppc_fp128 %Val)
9623 The '``llvm.cos.*``' intrinsics return the cosine of the operand.
9628 The argument and return value are floating point numbers of the same
9634 This function returns the cosine of the specified operand, returning the
9635 same values as the libm ``cos`` functions would, and handles error
9636 conditions in the same way.
9638 '``llvm.pow.*``' Intrinsic
9639 ^^^^^^^^^^^^^^^^^^^^^^^^^^
9644 This is an overloaded intrinsic. You can use ``llvm.pow`` on any
9645 floating point or vector of floating point type. Not all targets support
9650 declare float @llvm.pow.f32(float %Val, float %Power)
9651 declare double @llvm.pow.f64(double %Val, double %Power)
9652 declare x86_fp80 @llvm.pow.f80(x86_fp80 %Val, x86_fp80 %Power)
9653 declare fp128 @llvm.pow.f128(fp128 %Val, fp128 %Power)
9654 declare ppc_fp128 @llvm.pow.ppcf128(ppc_fp128 %Val, ppc_fp128 Power)
9659 The '``llvm.pow.*``' intrinsics return the first operand raised to the
9660 specified (positive or negative) power.
9665 The second argument is a floating point power, and the first is a value
9666 to raise to that power.
9671 This function returns the first value raised to the second power,
9672 returning the same values as the libm ``pow`` functions would, and
9673 handles error conditions in the same way.
9675 '``llvm.exp.*``' Intrinsic
9676 ^^^^^^^^^^^^^^^^^^^^^^^^^^
9681 This is an overloaded intrinsic. You can use ``llvm.exp`` on any
9682 floating point or vector of floating point type. Not all targets support
9687 declare float @llvm.exp.f32(float %Val)
9688 declare double @llvm.exp.f64(double %Val)
9689 declare x86_fp80 @llvm.exp.f80(x86_fp80 %Val)
9690 declare fp128 @llvm.exp.f128(fp128 %Val)
9691 declare ppc_fp128 @llvm.exp.ppcf128(ppc_fp128 %Val)
9696 The '``llvm.exp.*``' intrinsics perform the exp function.
9701 The argument and return value are floating point numbers of the same
9707 This function returns the same values as the libm ``exp`` functions
9708 would, and handles error conditions in the same way.
9710 '``llvm.exp2.*``' Intrinsic
9711 ^^^^^^^^^^^^^^^^^^^^^^^^^^^
9716 This is an overloaded intrinsic. You can use ``llvm.exp2`` on any
9717 floating point or vector of floating point type. Not all targets support
9722 declare float @llvm.exp2.f32(float %Val)
9723 declare double @llvm.exp2.f64(double %Val)
9724 declare x86_fp80 @llvm.exp2.f80(x86_fp80 %Val)
9725 declare fp128 @llvm.exp2.f128(fp128 %Val)
9726 declare ppc_fp128 @llvm.exp2.ppcf128(ppc_fp128 %Val)
9731 The '``llvm.exp2.*``' intrinsics perform the exp2 function.
9736 The argument and return value are floating point numbers of the same
9742 This function returns the same values as the libm ``exp2`` functions
9743 would, and handles error conditions in the same way.
9745 '``llvm.log.*``' Intrinsic
9746 ^^^^^^^^^^^^^^^^^^^^^^^^^^
9751 This is an overloaded intrinsic. You can use ``llvm.log`` on any
9752 floating point or vector of floating point type. Not all targets support
9757 declare float @llvm.log.f32(float %Val)
9758 declare double @llvm.log.f64(double %Val)
9759 declare x86_fp80 @llvm.log.f80(x86_fp80 %Val)
9760 declare fp128 @llvm.log.f128(fp128 %Val)
9761 declare ppc_fp128 @llvm.log.ppcf128(ppc_fp128 %Val)
9766 The '``llvm.log.*``' intrinsics perform the log function.
9771 The argument and return value are floating point numbers of the same
9777 This function returns the same values as the libm ``log`` functions
9778 would, and handles error conditions in the same way.
9780 '``llvm.log10.*``' Intrinsic
9781 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
9786 This is an overloaded intrinsic. You can use ``llvm.log10`` on any
9787 floating point or vector of floating point type. Not all targets support
9792 declare float @llvm.log10.f32(float %Val)
9793 declare double @llvm.log10.f64(double %Val)
9794 declare x86_fp80 @llvm.log10.f80(x86_fp80 %Val)
9795 declare fp128 @llvm.log10.f128(fp128 %Val)
9796 declare ppc_fp128 @llvm.log10.ppcf128(ppc_fp128 %Val)
9801 The '``llvm.log10.*``' intrinsics perform the log10 function.
9806 The argument and return value are floating point numbers of the same
9812 This function returns the same values as the libm ``log10`` functions
9813 would, and handles error conditions in the same way.
9815 '``llvm.log2.*``' Intrinsic
9816 ^^^^^^^^^^^^^^^^^^^^^^^^^^^
9821 This is an overloaded intrinsic. You can use ``llvm.log2`` on any
9822 floating point or vector of floating point type. Not all targets support
9827 declare float @llvm.log2.f32(float %Val)
9828 declare double @llvm.log2.f64(double %Val)
9829 declare x86_fp80 @llvm.log2.f80(x86_fp80 %Val)
9830 declare fp128 @llvm.log2.f128(fp128 %Val)
9831 declare ppc_fp128 @llvm.log2.ppcf128(ppc_fp128 %Val)
9836 The '``llvm.log2.*``' intrinsics perform the log2 function.
9841 The argument and return value are floating point numbers of the same
9847 This function returns the same values as the libm ``log2`` functions
9848 would, and handles error conditions in the same way.
9850 '``llvm.fma.*``' Intrinsic
9851 ^^^^^^^^^^^^^^^^^^^^^^^^^^
9856 This is an overloaded intrinsic. You can use ``llvm.fma`` on any
9857 floating point or vector of floating point type. Not all targets support
9862 declare float @llvm.fma.f32(float %a, float %b, float %c)
9863 declare double @llvm.fma.f64(double %a, double %b, double %c)
9864 declare x86_fp80 @llvm.fma.f80(x86_fp80 %a, x86_fp80 %b, x86_fp80 %c)
9865 declare fp128 @llvm.fma.f128(fp128 %a, fp128 %b, fp128 %c)
9866 declare ppc_fp128 @llvm.fma.ppcf128(ppc_fp128 %a, ppc_fp128 %b, ppc_fp128 %c)
9871 The '``llvm.fma.*``' intrinsics perform the fused multiply-add
9877 The argument and return value are floating point numbers of the same
9883 This function returns the same values as the libm ``fma`` functions
9884 would, and does not set errno.
9886 '``llvm.fabs.*``' Intrinsic
9887 ^^^^^^^^^^^^^^^^^^^^^^^^^^^
9892 This is an overloaded intrinsic. You can use ``llvm.fabs`` on any
9893 floating point or vector of floating point type. Not all targets support
9898 declare float @llvm.fabs.f32(float %Val)
9899 declare double @llvm.fabs.f64(double %Val)
9900 declare x86_fp80 @llvm.fabs.f80(x86_fp80 %Val)
9901 declare fp128 @llvm.fabs.f128(fp128 %Val)
9902 declare ppc_fp128 @llvm.fabs.ppcf128(ppc_fp128 %Val)
9907 The '``llvm.fabs.*``' intrinsics return the absolute value of the
9913 The argument and return value are floating point numbers of the same
9919 This function returns the same values as the libm ``fabs`` functions
9920 would, and handles error conditions in the same way.
9922 '``llvm.minnum.*``' Intrinsic
9923 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
9928 This is an overloaded intrinsic. You can use ``llvm.minnum`` on any
9929 floating point or vector of floating point type. Not all targets support
9934 declare float @llvm.minnum.f32(float %Val0, float %Val1)
9935 declare double @llvm.minnum.f64(double %Val0, double %Val1)
9936 declare x86_fp80 @llvm.minnum.f80(x86_fp80 %Val0, x86_fp80 %Val1)
9937 declare fp128 @llvm.minnum.f128(fp128 %Val0, fp128 %Val1)
9938 declare ppc_fp128 @llvm.minnum.ppcf128(ppc_fp128 %Val0, ppc_fp128 %Val1)
9943 The '``llvm.minnum.*``' intrinsics return the minimum of the two
9950 The arguments and return value are floating point numbers of the same
9956 Follows the IEEE-754 semantics for minNum, which also match for libm's
9959 If either operand is a NaN, returns the other non-NaN operand. Returns
9960 NaN only if both operands are NaN. If the operands compare equal,
9961 returns a value that compares equal to both operands. This means that
9962 fmin(+/-0.0, +/-0.0) could return either -0.0 or 0.0.
9964 '``llvm.maxnum.*``' Intrinsic
9965 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
9970 This is an overloaded intrinsic. You can use ``llvm.maxnum`` on any
9971 floating point or vector of floating point type. Not all targets support
9976 declare float @llvm.maxnum.f32(float %Val0, float %Val1l)
9977 declare double @llvm.maxnum.f64(double %Val0, double %Val1)
9978 declare x86_fp80 @llvm.maxnum.f80(x86_fp80 %Val0, x86_fp80 %Val1)
9979 declare fp128 @llvm.maxnum.f128(fp128 %Val0, fp128 %Val1)
9980 declare ppc_fp128 @llvm.maxnum.ppcf128(ppc_fp128 %Val0, ppc_fp128 %Val1)
9985 The '``llvm.maxnum.*``' intrinsics return the maximum of the two
9992 The arguments and return value are floating point numbers of the same
9997 Follows the IEEE-754 semantics for maxNum, which also match for libm's
10000 If either operand is a NaN, returns the other non-NaN operand. Returns
10001 NaN only if both operands are NaN. If the operands compare equal,
10002 returns a value that compares equal to both operands. This means that
10003 fmax(+/-0.0, +/-0.0) could return either -0.0 or 0.0.
10005 '``llvm.copysign.*``' Intrinsic
10006 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
10011 This is an overloaded intrinsic. You can use ``llvm.copysign`` on any
10012 floating point or vector of floating point type. Not all targets support
10017 declare float @llvm.copysign.f32(float %Mag, float %Sgn)
10018 declare double @llvm.copysign.f64(double %Mag, double %Sgn)
10019 declare x86_fp80 @llvm.copysign.f80(x86_fp80 %Mag, x86_fp80 %Sgn)
10020 declare fp128 @llvm.copysign.f128(fp128 %Mag, fp128 %Sgn)
10021 declare ppc_fp128 @llvm.copysign.ppcf128(ppc_fp128 %Mag, ppc_fp128 %Sgn)
10026 The '``llvm.copysign.*``' intrinsics return a value with the magnitude of the
10027 first operand and the sign of the second operand.
10032 The arguments and return value are floating point numbers of the same
10038 This function returns the same values as the libm ``copysign``
10039 functions would, and handles error conditions in the same way.
10041 '``llvm.floor.*``' Intrinsic
10042 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
10047 This is an overloaded intrinsic. You can use ``llvm.floor`` on any
10048 floating point or vector of floating point type. Not all targets support
10053 declare float @llvm.floor.f32(float %Val)
10054 declare double @llvm.floor.f64(double %Val)
10055 declare x86_fp80 @llvm.floor.f80(x86_fp80 %Val)
10056 declare fp128 @llvm.floor.f128(fp128 %Val)
10057 declare ppc_fp128 @llvm.floor.ppcf128(ppc_fp128 %Val)
10062 The '``llvm.floor.*``' intrinsics return the floor of the operand.
10067 The argument and return value are floating point numbers of the same
10073 This function returns the same values as the libm ``floor`` functions
10074 would, and handles error conditions in the same way.
10076 '``llvm.ceil.*``' Intrinsic
10077 ^^^^^^^^^^^^^^^^^^^^^^^^^^^
10082 This is an overloaded intrinsic. You can use ``llvm.ceil`` on any
10083 floating point or vector of floating point type. Not all targets support
10088 declare float @llvm.ceil.f32(float %Val)
10089 declare double @llvm.ceil.f64(double %Val)
10090 declare x86_fp80 @llvm.ceil.f80(x86_fp80 %Val)
10091 declare fp128 @llvm.ceil.f128(fp128 %Val)
10092 declare ppc_fp128 @llvm.ceil.ppcf128(ppc_fp128 %Val)
10097 The '``llvm.ceil.*``' intrinsics return the ceiling of the operand.
10102 The argument and return value are floating point numbers of the same
10108 This function returns the same values as the libm ``ceil`` functions
10109 would, and handles error conditions in the same way.
10111 '``llvm.trunc.*``' Intrinsic
10112 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
10117 This is an overloaded intrinsic. You can use ``llvm.trunc`` on any
10118 floating point or vector of floating point type. Not all targets support
10123 declare float @llvm.trunc.f32(float %Val)
10124 declare double @llvm.trunc.f64(double %Val)
10125 declare x86_fp80 @llvm.trunc.f80(x86_fp80 %Val)
10126 declare fp128 @llvm.trunc.f128(fp128 %Val)
10127 declare ppc_fp128 @llvm.trunc.ppcf128(ppc_fp128 %Val)
10132 The '``llvm.trunc.*``' intrinsics returns the operand rounded to the
10133 nearest integer not larger in magnitude than the operand.
10138 The argument and return value are floating point numbers of the same
10144 This function returns the same values as the libm ``trunc`` functions
10145 would, and handles error conditions in the same way.
10147 '``llvm.rint.*``' Intrinsic
10148 ^^^^^^^^^^^^^^^^^^^^^^^^^^^
10153 This is an overloaded intrinsic. You can use ``llvm.rint`` on any
10154 floating point or vector of floating point type. Not all targets support
10159 declare float @llvm.rint.f32(float %Val)
10160 declare double @llvm.rint.f64(double %Val)
10161 declare x86_fp80 @llvm.rint.f80(x86_fp80 %Val)
10162 declare fp128 @llvm.rint.f128(fp128 %Val)
10163 declare ppc_fp128 @llvm.rint.ppcf128(ppc_fp128 %Val)
10168 The '``llvm.rint.*``' intrinsics returns the operand rounded to the
10169 nearest integer. It may raise an inexact floating-point exception if the
10170 operand isn't an integer.
10175 The argument and return value are floating point numbers of the same
10181 This function returns the same values as the libm ``rint`` functions
10182 would, and handles error conditions in the same way.
10184 '``llvm.nearbyint.*``' Intrinsic
10185 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
10190 This is an overloaded intrinsic. You can use ``llvm.nearbyint`` on any
10191 floating point or vector of floating point type. Not all targets support
10196 declare float @llvm.nearbyint.f32(float %Val)
10197 declare double @llvm.nearbyint.f64(double %Val)
10198 declare x86_fp80 @llvm.nearbyint.f80(x86_fp80 %Val)
10199 declare fp128 @llvm.nearbyint.f128(fp128 %Val)
10200 declare ppc_fp128 @llvm.nearbyint.ppcf128(ppc_fp128 %Val)
10205 The '``llvm.nearbyint.*``' intrinsics returns the operand rounded to the
10211 The argument and return value are floating point numbers of the same
10217 This function returns the same values as the libm ``nearbyint``
10218 functions would, and handles error conditions in the same way.
10220 '``llvm.round.*``' Intrinsic
10221 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
10226 This is an overloaded intrinsic. You can use ``llvm.round`` on any
10227 floating point or vector of floating point type. Not all targets support
10232 declare float @llvm.round.f32(float %Val)
10233 declare double @llvm.round.f64(double %Val)
10234 declare x86_fp80 @llvm.round.f80(x86_fp80 %Val)
10235 declare fp128 @llvm.round.f128(fp128 %Val)
10236 declare ppc_fp128 @llvm.round.ppcf128(ppc_fp128 %Val)
10241 The '``llvm.round.*``' intrinsics returns the operand rounded to the
10247 The argument and return value are floating point numbers of the same
10253 This function returns the same values as the libm ``round``
10254 functions would, and handles error conditions in the same way.
10256 Bit Manipulation Intrinsics
10257 ---------------------------
10259 LLVM provides intrinsics for a few important bit manipulation
10260 operations. These allow efficient code generation for some algorithms.
10262 '``llvm.bswap.*``' Intrinsics
10263 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
10268 This is an overloaded intrinsic function. You can use bswap on any
10269 integer type that is an even number of bytes (i.e. BitWidth % 16 == 0).
10273 declare i16 @llvm.bswap.i16(i16 <id>)
10274 declare i32 @llvm.bswap.i32(i32 <id>)
10275 declare i64 @llvm.bswap.i64(i64 <id>)
10280 The '``llvm.bswap``' family of intrinsics is used to byte swap integer
10281 values with an even number of bytes (positive multiple of 16 bits).
10282 These are useful for performing operations on data that is not in the
10283 target's native byte order.
10288 The ``llvm.bswap.i16`` intrinsic returns an i16 value that has the high
10289 and low byte of the input i16 swapped. Similarly, the ``llvm.bswap.i32``
10290 intrinsic returns an i32 value that has the four bytes of the input i32
10291 swapped, so that if the input bytes are numbered 0, 1, 2, 3 then the
10292 returned i32 will have its bytes in 3, 2, 1, 0 order. The
10293 ``llvm.bswap.i48``, ``llvm.bswap.i64`` and other intrinsics extend this
10294 concept to additional even-byte lengths (6 bytes, 8 bytes and more,
10297 '``llvm.ctpop.*``' Intrinsic
10298 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
10303 This is an overloaded intrinsic. You can use llvm.ctpop on any integer
10304 bit width, or on any vector with integer elements. Not all targets
10305 support all bit widths or vector types, however.
10309 declare i8 @llvm.ctpop.i8(i8 <src>)
10310 declare i16 @llvm.ctpop.i16(i16 <src>)
10311 declare i32 @llvm.ctpop.i32(i32 <src>)
10312 declare i64 @llvm.ctpop.i64(i64 <src>)
10313 declare i256 @llvm.ctpop.i256(i256 <src>)
10314 declare <2 x i32> @llvm.ctpop.v2i32(<2 x i32> <src>)
10319 The '``llvm.ctpop``' family of intrinsics counts the number of bits set
10325 The only argument is the value to be counted. The argument may be of any
10326 integer type, or a vector with integer elements. The return type must
10327 match the argument type.
10332 The '``llvm.ctpop``' intrinsic counts the 1's in a variable, or within
10333 each element of a vector.
10335 '``llvm.ctlz.*``' Intrinsic
10336 ^^^^^^^^^^^^^^^^^^^^^^^^^^^
10341 This is an overloaded intrinsic. You can use ``llvm.ctlz`` on any
10342 integer bit width, or any vector whose elements are integers. Not all
10343 targets support all bit widths or vector types, however.
10347 declare i8 @llvm.ctlz.i8 (i8 <src>, i1 <is_zero_undef>)
10348 declare i16 @llvm.ctlz.i16 (i16 <src>, i1 <is_zero_undef>)
10349 declare i32 @llvm.ctlz.i32 (i32 <src>, i1 <is_zero_undef>)
10350 declare i64 @llvm.ctlz.i64 (i64 <src>, i1 <is_zero_undef>)
10351 declare i256 @llvm.ctlz.i256(i256 <src>, i1 <is_zero_undef>)
10352 declase <2 x i32> @llvm.ctlz.v2i32(<2 x i32> <src>, i1 <is_zero_undef>)
10357 The '``llvm.ctlz``' family of intrinsic functions counts the number of
10358 leading zeros in a variable.
10363 The first argument is the value to be counted. This argument may be of
10364 any integer type, or a vector with integer element type. The return
10365 type must match the first argument type.
10367 The second argument must be a constant and is a flag to indicate whether
10368 the intrinsic should ensure that a zero as the first argument produces a
10369 defined result. Historically some architectures did not provide a
10370 defined result for zero values as efficiently, and many algorithms are
10371 now predicated on avoiding zero-value inputs.
10376 The '``llvm.ctlz``' intrinsic counts the leading (most significant)
10377 zeros in a variable, or within each element of the vector. If
10378 ``src == 0`` then the result is the size in bits of the type of ``src``
10379 if ``is_zero_undef == 0`` and ``undef`` otherwise. For example,
10380 ``llvm.ctlz(i32 2) = 30``.
10382 '``llvm.cttz.*``' Intrinsic
10383 ^^^^^^^^^^^^^^^^^^^^^^^^^^^
10388 This is an overloaded intrinsic. You can use ``llvm.cttz`` on any
10389 integer bit width, or any vector of integer elements. Not all targets
10390 support all bit widths or vector types, however.
10394 declare i8 @llvm.cttz.i8 (i8 <src>, i1 <is_zero_undef>)
10395 declare i16 @llvm.cttz.i16 (i16 <src>, i1 <is_zero_undef>)
10396 declare i32 @llvm.cttz.i32 (i32 <src>, i1 <is_zero_undef>)
10397 declare i64 @llvm.cttz.i64 (i64 <src>, i1 <is_zero_undef>)
10398 declare i256 @llvm.cttz.i256(i256 <src>, i1 <is_zero_undef>)
10399 declase <2 x i32> @llvm.cttz.v2i32(<2 x i32> <src>, i1 <is_zero_undef>)
10404 The '``llvm.cttz``' family of intrinsic functions counts the number of
10410 The first argument is the value to be counted. This argument may be of
10411 any integer type, or a vector with integer element type. The return
10412 type must match the first argument type.
10414 The second argument must be a constant and is a flag to indicate whether
10415 the intrinsic should ensure that a zero as the first argument produces a
10416 defined result. Historically some architectures did not provide a
10417 defined result for zero values as efficiently, and many algorithms are
10418 now predicated on avoiding zero-value inputs.
10423 The '``llvm.cttz``' intrinsic counts the trailing (least significant)
10424 zeros in a variable, or within each element of a vector. If ``src == 0``
10425 then the result is the size in bits of the type of ``src`` if
10426 ``is_zero_undef == 0`` and ``undef`` otherwise. For example,
10427 ``llvm.cttz(2) = 1``.
10431 Arithmetic with Overflow Intrinsics
10432 -----------------------------------
10434 LLVM provides intrinsics for some arithmetic with overflow operations.
10436 '``llvm.sadd.with.overflow.*``' Intrinsics
10437 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
10442 This is an overloaded intrinsic. You can use ``llvm.sadd.with.overflow``
10443 on any integer bit width.
10447 declare {i16, i1} @llvm.sadd.with.overflow.i16(i16 %a, i16 %b)
10448 declare {i32, i1} @llvm.sadd.with.overflow.i32(i32 %a, i32 %b)
10449 declare {i64, i1} @llvm.sadd.with.overflow.i64(i64 %a, i64 %b)
10454 The '``llvm.sadd.with.overflow``' family of intrinsic functions perform
10455 a signed addition of the two arguments, and indicate whether an overflow
10456 occurred during the signed summation.
10461 The arguments (%a and %b) and the first element of the result structure
10462 may be of integer types of any bit width, but they must have the same
10463 bit width. The second element of the result structure must be of type
10464 ``i1``. ``%a`` and ``%b`` are the two values that will undergo signed
10470 The '``llvm.sadd.with.overflow``' family of intrinsic functions perform
10471 a signed addition of the two variables. They return a structure --- the
10472 first element of which is the signed summation, and the second element
10473 of which is a bit specifying if the signed summation resulted in an
10479 .. code-block:: llvm
10481 %res = call {i32, i1} @llvm.sadd.with.overflow.i32(i32 %a, i32 %b)
10482 %sum = extractvalue {i32, i1} %res, 0
10483 %obit = extractvalue {i32, i1} %res, 1
10484 br i1 %obit, label %overflow, label %normal
10486 '``llvm.uadd.with.overflow.*``' Intrinsics
10487 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
10492 This is an overloaded intrinsic. You can use ``llvm.uadd.with.overflow``
10493 on any integer bit width.
10497 declare {i16, i1} @llvm.uadd.with.overflow.i16(i16 %a, i16 %b)
10498 declare {i32, i1} @llvm.uadd.with.overflow.i32(i32 %a, i32 %b)
10499 declare {i64, i1} @llvm.uadd.with.overflow.i64(i64 %a, i64 %b)
10504 The '``llvm.uadd.with.overflow``' family of intrinsic functions perform
10505 an unsigned addition of the two arguments, and indicate whether a carry
10506 occurred during the unsigned summation.
10511 The arguments (%a and %b) and the first element of the result structure
10512 may be of integer types of any bit width, but they must have the same
10513 bit width. The second element of the result structure must be of type
10514 ``i1``. ``%a`` and ``%b`` are the two values that will undergo unsigned
10520 The '``llvm.uadd.with.overflow``' family of intrinsic functions perform
10521 an unsigned addition of the two arguments. They return a structure --- the
10522 first element of which is the sum, and the second element of which is a
10523 bit specifying if the unsigned summation resulted in a carry.
10528 .. code-block:: llvm
10530 %res = call {i32, i1} @llvm.uadd.with.overflow.i32(i32 %a, i32 %b)
10531 %sum = extractvalue {i32, i1} %res, 0
10532 %obit = extractvalue {i32, i1} %res, 1
10533 br i1 %obit, label %carry, label %normal
10535 '``llvm.ssub.with.overflow.*``' Intrinsics
10536 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
10541 This is an overloaded intrinsic. You can use ``llvm.ssub.with.overflow``
10542 on any integer bit width.
10546 declare {i16, i1} @llvm.ssub.with.overflow.i16(i16 %a, i16 %b)
10547 declare {i32, i1} @llvm.ssub.with.overflow.i32(i32 %a, i32 %b)
10548 declare {i64, i1} @llvm.ssub.with.overflow.i64(i64 %a, i64 %b)
10553 The '``llvm.ssub.with.overflow``' family of intrinsic functions perform
10554 a signed subtraction of the two arguments, and indicate whether an
10555 overflow occurred during the signed subtraction.
10560 The arguments (%a and %b) and the first element of the result structure
10561 may be of integer types of any bit width, but they must have the same
10562 bit width. The second element of the result structure must be of type
10563 ``i1``. ``%a`` and ``%b`` are the two values that will undergo signed
10569 The '``llvm.ssub.with.overflow``' family of intrinsic functions perform
10570 a signed subtraction of the two arguments. They return a structure --- the
10571 first element of which is the subtraction, and the second element of
10572 which is a bit specifying if the signed subtraction resulted in an
10578 .. code-block:: llvm
10580 %res = call {i32, i1} @llvm.ssub.with.overflow.i32(i32 %a, i32 %b)
10581 %sum = extractvalue {i32, i1} %res, 0
10582 %obit = extractvalue {i32, i1} %res, 1
10583 br i1 %obit, label %overflow, label %normal
10585 '``llvm.usub.with.overflow.*``' Intrinsics
10586 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
10591 This is an overloaded intrinsic. You can use ``llvm.usub.with.overflow``
10592 on any integer bit width.
10596 declare {i16, i1} @llvm.usub.with.overflow.i16(i16 %a, i16 %b)
10597 declare {i32, i1} @llvm.usub.with.overflow.i32(i32 %a, i32 %b)
10598 declare {i64, i1} @llvm.usub.with.overflow.i64(i64 %a, i64 %b)
10603 The '``llvm.usub.with.overflow``' family of intrinsic functions perform
10604 an unsigned subtraction of the two arguments, and indicate whether an
10605 overflow occurred during the unsigned subtraction.
10610 The arguments (%a and %b) and the first element of the result structure
10611 may be of integer types of any bit width, but they must have the same
10612 bit width. The second element of the result structure must be of type
10613 ``i1``. ``%a`` and ``%b`` are the two values that will undergo unsigned
10619 The '``llvm.usub.with.overflow``' family of intrinsic functions perform
10620 an unsigned subtraction of the two arguments. They return a structure ---
10621 the first element of which is the subtraction, and the second element of
10622 which is a bit specifying if the unsigned subtraction resulted in an
10628 .. code-block:: llvm
10630 %res = call {i32, i1} @llvm.usub.with.overflow.i32(i32 %a, i32 %b)
10631 %sum = extractvalue {i32, i1} %res, 0
10632 %obit = extractvalue {i32, i1} %res, 1
10633 br i1 %obit, label %overflow, label %normal
10635 '``llvm.smul.with.overflow.*``' Intrinsics
10636 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
10641 This is an overloaded intrinsic. You can use ``llvm.smul.with.overflow``
10642 on any integer bit width.
10646 declare {i16, i1} @llvm.smul.with.overflow.i16(i16 %a, i16 %b)
10647 declare {i32, i1} @llvm.smul.with.overflow.i32(i32 %a, i32 %b)
10648 declare {i64, i1} @llvm.smul.with.overflow.i64(i64 %a, i64 %b)
10653 The '``llvm.smul.with.overflow``' family of intrinsic functions perform
10654 a signed multiplication of the two arguments, and indicate whether an
10655 overflow occurred during the signed multiplication.
10660 The arguments (%a and %b) and the first element of the result structure
10661 may be of integer types of any bit width, but they must have the same
10662 bit width. The second element of the result structure must be of type
10663 ``i1``. ``%a`` and ``%b`` are the two values that will undergo signed
10669 The '``llvm.smul.with.overflow``' family of intrinsic functions perform
10670 a signed multiplication of the two arguments. They return a structure ---
10671 the first element of which is the multiplication, and the second element
10672 of which is a bit specifying if the signed multiplication resulted in an
10678 .. code-block:: llvm
10680 %res = call {i32, i1} @llvm.smul.with.overflow.i32(i32 %a, i32 %b)
10681 %sum = extractvalue {i32, i1} %res, 0
10682 %obit = extractvalue {i32, i1} %res, 1
10683 br i1 %obit, label %overflow, label %normal
10685 '``llvm.umul.with.overflow.*``' Intrinsics
10686 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
10691 This is an overloaded intrinsic. You can use ``llvm.umul.with.overflow``
10692 on any integer bit width.
10696 declare {i16, i1} @llvm.umul.with.overflow.i16(i16 %a, i16 %b)
10697 declare {i32, i1} @llvm.umul.with.overflow.i32(i32 %a, i32 %b)
10698 declare {i64, i1} @llvm.umul.with.overflow.i64(i64 %a, i64 %b)
10703 The '``llvm.umul.with.overflow``' family of intrinsic functions perform
10704 a unsigned multiplication of the two arguments, and indicate whether an
10705 overflow occurred during the unsigned multiplication.
10710 The arguments (%a and %b) and the first element of the result structure
10711 may be of integer types of any bit width, but they must have the same
10712 bit width. The second element of the result structure must be of type
10713 ``i1``. ``%a`` and ``%b`` are the two values that will undergo unsigned
10719 The '``llvm.umul.with.overflow``' family of intrinsic functions perform
10720 an unsigned multiplication of the two arguments. They return a structure ---
10721 the first element of which is the multiplication, and the second
10722 element of which is a bit specifying if the unsigned multiplication
10723 resulted in an overflow.
10728 .. code-block:: llvm
10730 %res = call {i32, i1} @llvm.umul.with.overflow.i32(i32 %a, i32 %b)
10731 %sum = extractvalue {i32, i1} %res, 0
10732 %obit = extractvalue {i32, i1} %res, 1
10733 br i1 %obit, label %overflow, label %normal
10735 Specialised Arithmetic Intrinsics
10736 ---------------------------------
10738 '``llvm.canonicalize.*``' Intrinsic
10739 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
10746 declare float @llvm.canonicalize.f32(float %a)
10747 declare double @llvm.canonicalize.f64(double %b)
10752 The '``llvm.canonicalize.*``' intrinsic returns the platform specific canonical
10753 encoding of a floating point number. This canonicalization is useful for
10754 implementing certain numeric primitives such as frexp. The canonical encoding is
10755 defined by IEEE-754-2008 to be:
10759 2.1.8 canonical encoding: The preferred encoding of a floating-point
10760 representation in a format. Applied to declets, significands of finite
10761 numbers, infinities, and NaNs, especially in decimal formats.
10763 This operation can also be considered equivalent to the IEEE-754-2008
10764 conversion of a floating-point value to the same format. NaNs are handled
10765 according to section 6.2.
10767 Examples of non-canonical encodings:
10769 - x87 pseudo denormals, pseudo NaNs, pseudo Infinity, Unnormals. These are
10770 converted to a canonical representation per hardware-specific protocol.
10771 - Many normal decimal floating point numbers have non-canonical alternative
10773 - Some machines, like GPUs or ARMv7 NEON, do not support subnormal values.
10774 These are treated as non-canonical encodings of zero and with be flushed to
10775 a zero of the same sign by this operation.
10777 Note that per IEEE-754-2008 6.2, systems that support signaling NaNs with
10778 default exception handling must signal an invalid exception, and produce a
10781 This function should always be implementable as multiplication by 1.0, provided
10782 that the compiler does not constant fold the operation. Likewise, division by
10783 1.0 and ``llvm.minnum(x, x)`` are possible implementations. Addition with
10784 -0.0 is also sufficient provided that the rounding mode is not -Infinity.
10786 ``@llvm.canonicalize`` must preserve the equality relation. That is:
10788 - ``(@llvm.canonicalize(x) == x)`` is equivalent to ``(x == x)``
10789 - ``(@llvm.canonicalize(x) == @llvm.canonicalize(y))`` is equivalent to
10792 Additionally, the sign of zero must be conserved:
10793 ``@llvm.canonicalize(-0.0) = -0.0`` and ``@llvm.canonicalize(+0.0) = +0.0``
10795 The payload bits of a NaN must be conserved, with two exceptions.
10796 First, environments which use only a single canonical representation of NaN
10797 must perform said canonicalization. Second, SNaNs must be quieted per the
10800 The canonicalization operation may be optimized away if:
10802 - The input is known to be canonical. For example, it was produced by a
10803 floating-point operation that is required by the standard to be canonical.
10804 - The result is consumed only by (or fused with) other floating-point
10805 operations. That is, the bits of the floating point value are not examined.
10807 '``llvm.fmuladd.*``' Intrinsic
10808 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
10815 declare float @llvm.fmuladd.f32(float %a, float %b, float %c)
10816 declare double @llvm.fmuladd.f64(double %a, double %b, double %c)
10821 The '``llvm.fmuladd.*``' intrinsic functions represent multiply-add
10822 expressions that can be fused if the code generator determines that (a) the
10823 target instruction set has support for a fused operation, and (b) that the
10824 fused operation is more efficient than the equivalent, separate pair of mul
10825 and add instructions.
10830 The '``llvm.fmuladd.*``' intrinsics each take three arguments: two
10831 multiplicands, a and b, and an addend c.
10840 %0 = call float @llvm.fmuladd.f32(%a, %b, %c)
10842 is equivalent to the expression a \* b + c, except that rounding will
10843 not be performed between the multiplication and addition steps if the
10844 code generator fuses the operations. Fusion is not guaranteed, even if
10845 the target platform supports it. If a fused multiply-add is required the
10846 corresponding llvm.fma.\* intrinsic function should be used
10847 instead. This never sets errno, just as '``llvm.fma.*``'.
10852 .. code-block:: llvm
10854 %r2 = call float @llvm.fmuladd.f32(float %a, float %b, float %c) ; yields float:r2 = (a * b) + c
10857 '``llvm.uabsdiff.*``' and '``llvm.sabsdiff.*``' Intrinsics
10858 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
10862 This is an overloaded intrinsic. The loaded data is a vector of any integer bit width.
10864 .. code-block:: llvm
10866 declare <4 x integer> @llvm.uabsdiff.v4i32(<4 x integer> %a, <4 x integer> %b)
10872 The ``llvm.uabsdiff`` intrinsic returns a vector result of the absolute difference of
10873 the two operands, treating them both as unsigned integers.
10875 The ``llvm.sabsdiff`` intrinsic returns a vector result of the absolute difference of
10876 the two operands, treating them both as signed integers.
10880 These intrinsics are primarily used during the code generation stage of compilation.
10881 They are generated by compiler passes such as the Loop and SLP vectorizers.it is not
10882 recommended for users to create them manually.
10887 Both intrinsics take two integer of the same bitwidth.
10894 call <4 x i32> @llvm.uabsdiff.v4i32(<4 x i32> %a, <4 x i32> %b)
10898 %sub = sub <4 x i32> %a, %b
10899 %ispos = icmp ugt <4 x i32> %sub, <i32 -1, i32 -1, i32 -1, i32 -1>
10900 %neg = sub <4 x i32> zeroinitializer, %sub
10901 %1 = select <4 x i1> %ispos, <4 x i32> %sub, <4 x i32> %neg
10903 Similarly the expression::
10905 call <4 x i32> @llvm.sabsdiff.v4i32(<4 x i32> %a, <4 x i32> %b)
10909 %sub = sub nsw <4 x i32> %a, %b
10910 %ispos = icmp sgt <4 x i32> %sub, <i32 -1, i32 -1, i32 -1, i32 -1>
10911 %neg = sub nsw <4 x i32> zeroinitializer, %sub
10912 %1 = select <4 x i1> %ispos, <4 x i32> %sub, <4 x i32> %neg
10915 Half Precision Floating Point Intrinsics
10916 ----------------------------------------
10918 For most target platforms, half precision floating point is a
10919 storage-only format. This means that it is a dense encoding (in memory)
10920 but does not support computation in the format.
10922 This means that code must first load the half-precision floating point
10923 value as an i16, then convert it to float with
10924 :ref:`llvm.convert.from.fp16 <int_convert_from_fp16>`. Computation can
10925 then be performed on the float value (including extending to double
10926 etc). To store the value back to memory, it is first converted to float
10927 if needed, then converted to i16 with
10928 :ref:`llvm.convert.to.fp16 <int_convert_to_fp16>`, then storing as an
10931 .. _int_convert_to_fp16:
10933 '``llvm.convert.to.fp16``' Intrinsic
10934 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
10941 declare i16 @llvm.convert.to.fp16.f32(float %a)
10942 declare i16 @llvm.convert.to.fp16.f64(double %a)
10947 The '``llvm.convert.to.fp16``' intrinsic function performs a conversion from a
10948 conventional floating point type to half precision floating point format.
10953 The intrinsic function contains single argument - the value to be
10959 The '``llvm.convert.to.fp16``' intrinsic function performs a conversion from a
10960 conventional floating point format to half precision floating point format. The
10961 return value is an ``i16`` which contains the converted number.
10966 .. code-block:: llvm
10968 %res = call i16 @llvm.convert.to.fp16.f32(float %a)
10969 store i16 %res, i16* @x, align 2
10971 .. _int_convert_from_fp16:
10973 '``llvm.convert.from.fp16``' Intrinsic
10974 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
10981 declare float @llvm.convert.from.fp16.f32(i16 %a)
10982 declare double @llvm.convert.from.fp16.f64(i16 %a)
10987 The '``llvm.convert.from.fp16``' intrinsic function performs a
10988 conversion from half precision floating point format to single precision
10989 floating point format.
10994 The intrinsic function contains single argument - the value to be
11000 The '``llvm.convert.from.fp16``' intrinsic function performs a
11001 conversion from half single precision floating point format to single
11002 precision floating point format. The input half-float value is
11003 represented by an ``i16`` value.
11008 .. code-block:: llvm
11010 %a = load i16, i16* @x, align 2
11011 %res = call float @llvm.convert.from.fp16(i16 %a)
11013 .. _dbg_intrinsics:
11015 Debugger Intrinsics
11016 -------------------
11018 The LLVM debugger intrinsics (which all start with ``llvm.dbg.``
11019 prefix), are described in the `LLVM Source Level
11020 Debugging <SourceLevelDebugging.html#format_common_intrinsics>`_
11023 Exception Handling Intrinsics
11024 -----------------------------
11026 The LLVM exception handling intrinsics (which all start with
11027 ``llvm.eh.`` prefix), are described in the `LLVM Exception
11028 Handling <ExceptionHandling.html#format_common_intrinsics>`_ document.
11030 .. _int_trampoline:
11032 Trampoline Intrinsics
11033 ---------------------
11035 These intrinsics make it possible to excise one parameter, marked with
11036 the :ref:`nest <nest>` attribute, from a function. The result is a
11037 callable function pointer lacking the nest parameter - the caller does
11038 not need to provide a value for it. Instead, the value to use is stored
11039 in advance in a "trampoline", a block of memory usually allocated on the
11040 stack, which also contains code to splice the nest value into the
11041 argument list. This is used to implement the GCC nested function address
11044 For example, if the function is ``i32 f(i8* nest %c, i32 %x, i32 %y)``
11045 then the resulting function pointer has signature ``i32 (i32, i32)*``.
11046 It can be created as follows:
11048 .. code-block:: llvm
11050 %tramp = alloca [10 x i8], align 4 ; size and alignment only correct for X86
11051 %tramp1 = getelementptr [10 x i8], [10 x i8]* %tramp, i32 0, i32 0
11052 call i8* @llvm.init.trampoline(i8* %tramp1, i8* bitcast (i32 (i8*, i32, i32)* @f to i8*), i8* %nval)
11053 %p = call i8* @llvm.adjust.trampoline(i8* %tramp1)
11054 %fp = bitcast i8* %p to i32 (i32, i32)*
11056 The call ``%val = call i32 %fp(i32 %x, i32 %y)`` is then equivalent to
11057 ``%val = call i32 %f(i8* %nval, i32 %x, i32 %y)``.
11061 '``llvm.init.trampoline``' Intrinsic
11062 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
11069 declare void @llvm.init.trampoline(i8* <tramp>, i8* <func>, i8* <nval>)
11074 This fills the memory pointed to by ``tramp`` with executable code,
11075 turning it into a trampoline.
11080 The ``llvm.init.trampoline`` intrinsic takes three arguments, all
11081 pointers. The ``tramp`` argument must point to a sufficiently large and
11082 sufficiently aligned block of memory; this memory is written to by the
11083 intrinsic. Note that the size and the alignment are target-specific -
11084 LLVM currently provides no portable way of determining them, so a
11085 front-end that generates this intrinsic needs to have some
11086 target-specific knowledge. The ``func`` argument must hold a function
11087 bitcast to an ``i8*``.
11092 The block of memory pointed to by ``tramp`` is filled with target
11093 dependent code, turning it into a function. Then ``tramp`` needs to be
11094 passed to :ref:`llvm.adjust.trampoline <int_at>` to get a pointer which can
11095 be :ref:`bitcast (to a new function) and called <int_trampoline>`. The new
11096 function's signature is the same as that of ``func`` with any arguments
11097 marked with the ``nest`` attribute removed. At most one such ``nest``
11098 argument is allowed, and it must be of pointer type. Calling the new
11099 function is equivalent to calling ``func`` with the same argument list,
11100 but with ``nval`` used for the missing ``nest`` argument. If, after
11101 calling ``llvm.init.trampoline``, the memory pointed to by ``tramp`` is
11102 modified, then the effect of any later call to the returned function
11103 pointer is undefined.
11107 '``llvm.adjust.trampoline``' Intrinsic
11108 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
11115 declare i8* @llvm.adjust.trampoline(i8* <tramp>)
11120 This performs any required machine-specific adjustment to the address of
11121 a trampoline (passed as ``tramp``).
11126 ``tramp`` must point to a block of memory which already has trampoline
11127 code filled in by a previous call to
11128 :ref:`llvm.init.trampoline <int_it>`.
11133 On some architectures the address of the code to be executed needs to be
11134 different than the address where the trampoline is actually stored. This
11135 intrinsic returns the executable address corresponding to ``tramp``
11136 after performing the required machine specific adjustments. The pointer
11137 returned can then be :ref:`bitcast and executed <int_trampoline>`.
11139 .. _int_mload_mstore:
11141 Masked Vector Load and Store Intrinsics
11142 ---------------------------------------
11144 LLVM provides intrinsics for predicated vector load and store operations. The predicate is specified by a mask operand, which holds one bit per vector element, switching the associated vector lane on or off. The memory addresses corresponding to the "off" lanes are not accessed. When all bits of the mask are on, the intrinsic is identical to a regular vector load or store. When all bits are off, no memory is accessed.
11148 '``llvm.masked.load.*``' Intrinsics
11149 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
11153 This is an overloaded intrinsic. The loaded data is a vector of any integer or floating point data type.
11157 declare <16 x float> @llvm.masked.load.v16f32 (<16 x float>* <ptr>, i32 <alignment>, <16 x i1> <mask>, <16 x float> <passthru>)
11158 declare <2 x double> @llvm.masked.load.v2f64 (<2 x double>* <ptr>, i32 <alignment>, <2 x i1> <mask>, <2 x double> <passthru>)
11163 Reads a vector from memory according to the provided mask. The mask holds a bit for each vector lane, and is used to prevent memory accesses to the masked-off lanes. The masked-off lanes in the result vector are taken from the corresponding lanes of the '``passthru``' operand.
11169 The first operand is the base pointer for the load. The second operand is the alignment of the source location. It must be a constant integer value. The third operand, mask, is a vector of boolean values with the same number of elements as the return type. The fourth is a pass-through value that is used to fill the masked-off lanes of the result. The return type, underlying type of the base pointer and the type of the '``passthru``' operand are the same vector types.
11175 The '``llvm.masked.load``' intrinsic is designed for conditional reading of selected vector elements in a single IR operation. It is useful for targets that support vector masked loads and allows vectorizing predicated basic blocks on these targets. Other targets may support this intrinsic differently, for example by lowering it into a sequence of branches that guard scalar load operations.
11176 The result of this operation is equivalent to a regular vector load instruction followed by a 'select' between the loaded and the passthru values, predicated on the same mask. However, using this intrinsic prevents exceptions on memory access to masked-off lanes.
11181 %res = call <16 x float> @llvm.masked.load.v16f32 (<16 x float>* %ptr, i32 4, <16 x i1>%mask, <16 x float> %passthru)
11183 ;; The result of the two following instructions is identical aside from potential memory access exception
11184 %loadlal = load <16 x float>, <16 x float>* %ptr, align 4
11185 %res = select <16 x i1> %mask, <16 x float> %loadlal, <16 x float> %passthru
11189 '``llvm.masked.store.*``' Intrinsics
11190 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
11194 This is an overloaded intrinsic. The data stored in memory is a vector of any integer or floating point data type.
11198 declare void @llvm.masked.store.v8i32 (<8 x i32> <value>, <8 x i32> * <ptr>, i32 <alignment>, <8 x i1> <mask>)
11199 declare void @llvm.masked.store.v16f32(<16 x i32> <value>, <16 x i32>* <ptr>, i32 <alignment>, <16 x i1> <mask>)
11204 Writes a vector to memory according to the provided mask. The mask holds a bit for each vector lane, and is used to prevent memory accesses to the masked-off lanes.
11209 The first operand is the vector value to be written to memory. The second operand is the base pointer for the store, it has the same underlying type as the value operand. The third operand is the alignment of the destination location. The fourth operand, mask, is a vector of boolean values. The types of the mask and the value operand must have the same number of vector elements.
11215 The '``llvm.masked.store``' intrinsics is designed for conditional writing of selected vector elements in a single IR operation. It is useful for targets that support vector masked store and allows vectorizing predicated basic blocks on these targets. Other targets may support this intrinsic differently, for example by lowering it into a sequence of branches that guard scalar store operations.
11216 The result of this operation is equivalent to a load-modify-store sequence. However, using this intrinsic prevents exceptions and data races on memory access to masked-off lanes.
11220 call void @llvm.masked.store.v16f32(<16 x float> %value, <16 x float>* %ptr, i32 4, <16 x i1> %mask)
11222 ;; The result of the following instructions is identical aside from potential data races and memory access exceptions
11223 %oldval = load <16 x float>, <16 x float>* %ptr, align 4
11224 %res = select <16 x i1> %mask, <16 x float> %value, <16 x float> %oldval
11225 store <16 x float> %res, <16 x float>* %ptr, align 4
11228 Masked Vector Gather and Scatter Intrinsics
11229 -------------------------------------------
11231 LLVM provides intrinsics for vector gather and scatter operations. They are similar to :ref:`Masked Vector Load and Store <int_mload_mstore>`, except they are designed for arbitrary memory accesses, rather than sequential memory accesses. Gather and scatter also employ a mask operand, which holds one bit per vector element, switching the associated vector lane on or off. The memory addresses corresponding to the "off" lanes are not accessed. When all bits are off, no memory is accessed.
11235 '``llvm.masked.gather.*``' Intrinsics
11236 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
11240 This is an overloaded intrinsic. The loaded data are multiple scalar values of any integer or floating point data type gathered together into one vector.
11244 declare <16 x float> @llvm.masked.gather.v16f32 (<16 x float*> <ptrs>, i32 <alignment>, <16 x i1> <mask>, <16 x float> <passthru>)
11245 declare <2 x double> @llvm.masked.gather.v2f64 (<2 x double*> <ptrs>, i32 <alignment>, <2 x i1> <mask>, <2 x double> <passthru>)
11250 Reads scalar values from arbitrary memory locations and gathers them into one vector. The memory locations are provided in the vector of pointers '``ptrs``'. The memory is accessed according to the provided mask. The mask holds a bit for each vector lane, and is used to prevent memory accesses to the masked-off lanes. The masked-off lanes in the result vector are taken from the corresponding lanes of the '``passthru``' operand.
11256 The first operand is a vector of pointers which holds all memory addresses to read. The second operand is an alignment of the source addresses. It must be a constant integer value. The third operand, mask, is a vector of boolean values with the same number of elements as the return type. The fourth is a pass-through value that is used to fill the masked-off lanes of the result. The return type, underlying type of the vector of pointers and the type of the '``passthru``' operand are the same vector types.
11262 The '``llvm.masked.gather``' intrinsic is designed for conditional reading of multiple scalar values from arbitrary memory locations in a single IR operation. It is useful for targets that support vector masked gathers and allows vectorizing basic blocks with data and control divergence. Other targets may support this intrinsic differently, for example by lowering it into a sequence of scalar load operations.
11263 The semantics of this operation are equivalent to a sequence of conditional scalar loads with subsequent gathering all loaded values into a single vector. The mask restricts memory access to certain lanes and facilitates vectorization of predicated basic blocks.
11268 %res = call <4 x double> @llvm.masked.gather.v4f64 (<4 x double*> %ptrs, i32 8, <4 x i1>%mask, <4 x double> <true, true, true, true>)
11270 ;; The gather with all-true mask is equivalent to the following instruction sequence
11271 %ptr0 = extractelement <4 x double*> %ptrs, i32 0
11272 %ptr1 = extractelement <4 x double*> %ptrs, i32 1
11273 %ptr2 = extractelement <4 x double*> %ptrs, i32 2
11274 %ptr3 = extractelement <4 x double*> %ptrs, i32 3
11276 %val0 = load double, double* %ptr0, align 8
11277 %val1 = load double, double* %ptr1, align 8
11278 %val2 = load double, double* %ptr2, align 8
11279 %val3 = load double, double* %ptr3, align 8
11281 %vec0 = insertelement <4 x double>undef, %val0, 0
11282 %vec01 = insertelement <4 x double>%vec0, %val1, 1
11283 %vec012 = insertelement <4 x double>%vec01, %val2, 2
11284 %vec0123 = insertelement <4 x double>%vec012, %val3, 3
11288 '``llvm.masked.scatter.*``' Intrinsics
11289 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
11293 This is an overloaded intrinsic. The data stored in memory is a vector of any integer or floating point data type. Each vector element is stored in an arbitrary memory addresses. Scatter with overlapping addresses is guaranteed to be ordered from least-significant to most-significant element.
11297 declare void @llvm.masked.scatter.v8i32 (<8 x i32> <value>, <8 x i32*> <ptrs>, i32 <alignment>, <8 x i1> <mask>)
11298 declare void @llvm.masked.scatter.v16f32(<16 x i32> <value>, <16 x i32*> <ptrs>, i32 <alignment>, <16 x i1> <mask>)
11303 Writes each element from the value vector to the corresponding memory address. The memory addresses are represented as a vector of pointers. Writing is done according to the provided mask. The mask holds a bit for each vector lane, and is used to prevent memory accesses to the masked-off lanes.
11308 The first operand is a vector value to be written to memory. The second operand is a vector of pointers, pointing to where the value elements should be stored. It has the same underlying type as the value operand. The third operand is an alignment of the destination addresses. The fourth operand, mask, is a vector of boolean values. The types of the mask and the value operand must have the same number of vector elements.
11314 The '``llvm.masked.scatter``' intrinsics is designed for writing selected vector elements to arbitrary memory addresses in a single IR operation. The operation may be conditional, when not all bits in the mask are switched on. It is useful for targets that support vector masked scatter and allows vectorizing basic blocks with data and control divergency. Other targets may support this intrinsic differently, for example by lowering it into a sequence of branches that guard scalar store operations.
11318 ;; This instruction unconditionaly stores data vector in multiple addresses
11319 call @llvm.masked.scatter.v8i32 (<8 x i32> %value, <8 x i32*> %ptrs, i32 4, <8 x i1> <true, true, .. true>)
11321 ;; It is equivalent to a list of scalar stores
11322 %val0 = extractelement <8 x i32> %value, i32 0
11323 %val1 = extractelement <8 x i32> %value, i32 1
11325 %val7 = extractelement <8 x i32> %value, i32 7
11326 %ptr0 = extractelement <8 x i32*> %ptrs, i32 0
11327 %ptr1 = extractelement <8 x i32*> %ptrs, i32 1
11329 %ptr7 = extractelement <8 x i32*> %ptrs, i32 7
11330 ;; Note: the order of the following stores is important when they overlap:
11331 store i32 %val0, i32* %ptr0, align 4
11332 store i32 %val1, i32* %ptr1, align 4
11334 store i32 %val7, i32* %ptr7, align 4
11340 This class of intrinsics provides information about the lifetime of
11341 memory objects and ranges where variables are immutable.
11345 '``llvm.lifetime.start``' Intrinsic
11346 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
11353 declare void @llvm.lifetime.start(i64 <size>, i8* nocapture <ptr>)
11358 The '``llvm.lifetime.start``' intrinsic specifies the start of a memory
11364 The first argument is a constant integer representing the size of the
11365 object, or -1 if it is variable sized. The second argument is a pointer
11371 This intrinsic indicates that before this point in the code, the value
11372 of the memory pointed to by ``ptr`` is dead. This means that it is known
11373 to never be used and has an undefined value. A load from the pointer
11374 that precedes this intrinsic can be replaced with ``'undef'``.
11378 '``llvm.lifetime.end``' Intrinsic
11379 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
11386 declare void @llvm.lifetime.end(i64 <size>, i8* nocapture <ptr>)
11391 The '``llvm.lifetime.end``' intrinsic specifies the end of a memory
11397 The first argument is a constant integer representing the size of the
11398 object, or -1 if it is variable sized. The second argument is a pointer
11404 This intrinsic indicates that after this point in the code, the value of
11405 the memory pointed to by ``ptr`` is dead. This means that it is known to
11406 never be used and has an undefined value. Any stores into the memory
11407 object following this intrinsic may be removed as dead.
11409 '``llvm.invariant.start``' Intrinsic
11410 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
11417 declare {}* @llvm.invariant.start(i64 <size>, i8* nocapture <ptr>)
11422 The '``llvm.invariant.start``' intrinsic specifies that the contents of
11423 a memory object will not change.
11428 The first argument is a constant integer representing the size of the
11429 object, or -1 if it is variable sized. The second argument is a pointer
11435 This intrinsic indicates that until an ``llvm.invariant.end`` that uses
11436 the return value, the referenced memory location is constant and
11439 '``llvm.invariant.end``' Intrinsic
11440 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
11447 declare void @llvm.invariant.end({}* <start>, i64 <size>, i8* nocapture <ptr>)
11452 The '``llvm.invariant.end``' intrinsic specifies that the contents of a
11453 memory object are mutable.
11458 The first argument is the matching ``llvm.invariant.start`` intrinsic.
11459 The second argument is a constant integer representing the size of the
11460 object, or -1 if it is variable sized and the third argument is a
11461 pointer to the object.
11466 This intrinsic indicates that the memory is mutable again.
11471 This class of intrinsics is designed to be generic and has no specific
11474 '``llvm.var.annotation``' Intrinsic
11475 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
11482 declare void @llvm.var.annotation(i8* <val>, i8* <str>, i8* <str>, i32 <int>)
11487 The '``llvm.var.annotation``' intrinsic.
11492 The first argument is a pointer to a value, the second is a pointer to a
11493 global string, the third is a pointer to a global string which is the
11494 source file name, and the last argument is the line number.
11499 This intrinsic allows annotation of local variables with arbitrary
11500 strings. This can be useful for special purpose optimizations that want
11501 to look for these annotations. These have no other defined use; they are
11502 ignored by code generation and optimization.
11504 '``llvm.ptr.annotation.*``' Intrinsic
11505 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
11510 This is an overloaded intrinsic. You can use '``llvm.ptr.annotation``' on a
11511 pointer to an integer of any width. *NOTE* you must specify an address space for
11512 the pointer. The identifier for the default address space is the integer
11517 declare i8* @llvm.ptr.annotation.p<address space>i8(i8* <val>, i8* <str>, i8* <str>, i32 <int>)
11518 declare i16* @llvm.ptr.annotation.p<address space>i16(i16* <val>, i8* <str>, i8* <str>, i32 <int>)
11519 declare i32* @llvm.ptr.annotation.p<address space>i32(i32* <val>, i8* <str>, i8* <str>, i32 <int>)
11520 declare i64* @llvm.ptr.annotation.p<address space>i64(i64* <val>, i8* <str>, i8* <str>, i32 <int>)
11521 declare i256* @llvm.ptr.annotation.p<address space>i256(i256* <val>, i8* <str>, i8* <str>, i32 <int>)
11526 The '``llvm.ptr.annotation``' intrinsic.
11531 The first argument is a pointer to an integer value of arbitrary bitwidth
11532 (result of some expression), the second is a pointer to a global string, the
11533 third is a pointer to a global string which is the source file name, and the
11534 last argument is the line number. It returns the value of the first argument.
11539 This intrinsic allows annotation of a pointer to an integer with arbitrary
11540 strings. This can be useful for special purpose optimizations that want to look
11541 for these annotations. These have no other defined use; they are ignored by code
11542 generation and optimization.
11544 '``llvm.annotation.*``' Intrinsic
11545 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
11550 This is an overloaded intrinsic. You can use '``llvm.annotation``' on
11551 any integer bit width.
11555 declare i8 @llvm.annotation.i8(i8 <val>, i8* <str>, i8* <str>, i32 <int>)
11556 declare i16 @llvm.annotation.i16(i16 <val>, i8* <str>, i8* <str>, i32 <int>)
11557 declare i32 @llvm.annotation.i32(i32 <val>, i8* <str>, i8* <str>, i32 <int>)
11558 declare i64 @llvm.annotation.i64(i64 <val>, i8* <str>, i8* <str>, i32 <int>)
11559 declare i256 @llvm.annotation.i256(i256 <val>, i8* <str>, i8* <str>, i32 <int>)
11564 The '``llvm.annotation``' intrinsic.
11569 The first argument is an integer value (result of some expression), the
11570 second is a pointer to a global string, the third is a pointer to a
11571 global string which is the source file name, and the last argument is
11572 the line number. It returns the value of the first argument.
11577 This intrinsic allows annotations to be put on arbitrary expressions
11578 with arbitrary strings. This can be useful for special purpose
11579 optimizations that want to look for these annotations. These have no
11580 other defined use; they are ignored by code generation and optimization.
11582 '``llvm.trap``' Intrinsic
11583 ^^^^^^^^^^^^^^^^^^^^^^^^^
11590 declare void @llvm.trap() noreturn nounwind
11595 The '``llvm.trap``' intrinsic.
11605 This intrinsic is lowered to the target dependent trap instruction. If
11606 the target does not have a trap instruction, this intrinsic will be
11607 lowered to a call of the ``abort()`` function.
11609 '``llvm.debugtrap``' Intrinsic
11610 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
11617 declare void @llvm.debugtrap() nounwind
11622 The '``llvm.debugtrap``' intrinsic.
11632 This intrinsic is lowered to code which is intended to cause an
11633 execution trap with the intention of requesting the attention of a
11636 '``llvm.stackprotector``' Intrinsic
11637 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
11644 declare void @llvm.stackprotector(i8* <guard>, i8** <slot>)
11649 The ``llvm.stackprotector`` intrinsic takes the ``guard`` and stores it
11650 onto the stack at ``slot``. The stack slot is adjusted to ensure that it
11651 is placed on the stack before local variables.
11656 The ``llvm.stackprotector`` intrinsic requires two pointer arguments.
11657 The first argument is the value loaded from the stack guard
11658 ``@__stack_chk_guard``. The second variable is an ``alloca`` that has
11659 enough space to hold the value of the guard.
11664 This intrinsic causes the prologue/epilogue inserter to force the position of
11665 the ``AllocaInst`` stack slot to be before local variables on the stack. This is
11666 to ensure that if a local variable on the stack is overwritten, it will destroy
11667 the value of the guard. When the function exits, the guard on the stack is
11668 checked against the original guard by ``llvm.stackprotectorcheck``. If they are
11669 different, then ``llvm.stackprotectorcheck`` causes the program to abort by
11670 calling the ``__stack_chk_fail()`` function.
11672 '``llvm.stackprotectorcheck``' Intrinsic
11673 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
11680 declare void @llvm.stackprotectorcheck(i8** <guard>)
11685 The ``llvm.stackprotectorcheck`` intrinsic compares ``guard`` against an already
11686 created stack protector and if they are not equal calls the
11687 ``__stack_chk_fail()`` function.
11692 The ``llvm.stackprotectorcheck`` intrinsic requires one pointer argument, the
11693 the variable ``@__stack_chk_guard``.
11698 This intrinsic is provided to perform the stack protector check by comparing
11699 ``guard`` with the stack slot created by ``llvm.stackprotector`` and if the
11700 values do not match call the ``__stack_chk_fail()`` function.
11702 The reason to provide this as an IR level intrinsic instead of implementing it
11703 via other IR operations is that in order to perform this operation at the IR
11704 level without an intrinsic, one would need to create additional basic blocks to
11705 handle the success/failure cases. This makes it difficult to stop the stack
11706 protector check from disrupting sibling tail calls in Codegen. With this
11707 intrinsic, we are able to generate the stack protector basic blocks late in
11708 codegen after the tail call decision has occurred.
11710 '``llvm.objectsize``' Intrinsic
11711 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
11718 declare i32 @llvm.objectsize.i32(i8* <object>, i1 <min>)
11719 declare i64 @llvm.objectsize.i64(i8* <object>, i1 <min>)
11724 The ``llvm.objectsize`` intrinsic is designed to provide information to
11725 the optimizers to determine at compile time whether a) an operation
11726 (like memcpy) will overflow a buffer that corresponds to an object, or
11727 b) that a runtime check for overflow isn't necessary. An object in this
11728 context means an allocation of a specific class, structure, array, or
11734 The ``llvm.objectsize`` intrinsic takes two arguments. The first
11735 argument is a pointer to or into the ``object``. The second argument is
11736 a boolean and determines whether ``llvm.objectsize`` returns 0 (if true)
11737 or -1 (if false) when the object size is unknown. The second argument
11738 only accepts constants.
11743 The ``llvm.objectsize`` intrinsic is lowered to a constant representing
11744 the size of the object concerned. If the size cannot be determined at
11745 compile time, ``llvm.objectsize`` returns ``i32/i64 -1 or 0`` (depending
11746 on the ``min`` argument).
11748 '``llvm.expect``' Intrinsic
11749 ^^^^^^^^^^^^^^^^^^^^^^^^^^^
11754 This is an overloaded intrinsic. You can use ``llvm.expect`` on any
11759 declare i1 @llvm.expect.i1(i1 <val>, i1 <expected_val>)
11760 declare i32 @llvm.expect.i32(i32 <val>, i32 <expected_val>)
11761 declare i64 @llvm.expect.i64(i64 <val>, i64 <expected_val>)
11766 The ``llvm.expect`` intrinsic provides information about expected (the
11767 most probable) value of ``val``, which can be used by optimizers.
11772 The ``llvm.expect`` intrinsic takes two arguments. The first argument is
11773 a value. The second argument is an expected value, this needs to be a
11774 constant value, variables are not allowed.
11779 This intrinsic is lowered to the ``val``.
11783 '``llvm.assume``' Intrinsic
11784 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
11791 declare void @llvm.assume(i1 %cond)
11796 The ``llvm.assume`` allows the optimizer to assume that the provided
11797 condition is true. This information can then be used in simplifying other parts
11803 The condition which the optimizer may assume is always true.
11808 The intrinsic allows the optimizer to assume that the provided condition is
11809 always true whenever the control flow reaches the intrinsic call. No code is
11810 generated for this intrinsic, and instructions that contribute only to the
11811 provided condition are not used for code generation. If the condition is
11812 violated during execution, the behavior is undefined.
11814 Note that the optimizer might limit the transformations performed on values
11815 used by the ``llvm.assume`` intrinsic in order to preserve the instructions
11816 only used to form the intrinsic's input argument. This might prove undesirable
11817 if the extra information provided by the ``llvm.assume`` intrinsic does not cause
11818 sufficient overall improvement in code quality. For this reason,
11819 ``llvm.assume`` should not be used to document basic mathematical invariants
11820 that the optimizer can otherwise deduce or facts that are of little use to the
11825 '``llvm.bitset.test``' Intrinsic
11826 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
11833 declare i1 @llvm.bitset.test(i8* %ptr, metadata %bitset) nounwind readnone
11839 The first argument is a pointer to be tested. The second argument is a
11840 metadata string containing the name of a :doc:`bitset <BitSets>`.
11845 The ``llvm.bitset.test`` intrinsic tests whether the given pointer is a
11846 member of the given bitset.
11848 '``llvm.donothing``' Intrinsic
11849 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
11856 declare void @llvm.donothing() nounwind readnone
11861 The ``llvm.donothing`` intrinsic doesn't perform any operation. It's one of only
11862 two intrinsics (besides ``llvm.experimental.patchpoint``) that can be called
11863 with an invoke instruction.
11873 This intrinsic does nothing, and it's removed by optimizers and ignored
11876 Stack Map Intrinsics
11877 --------------------
11879 LLVM provides experimental intrinsics to support runtime patching
11880 mechanisms commonly desired in dynamic language JITs. These intrinsics
11881 are described in :doc:`StackMaps`.