• Home
  • Line#
  • Scopes#
  • Navigate#
  • Raw
  • Download
1========================
2LLVM Programmer's Manual
3========================
4
5.. contents::
6   :local:
7
8.. warning::
9   This is always a work in progress.
10
11.. _introduction:
12
13Introduction
14============
15
16This document is meant to highlight some of the important classes and interfaces
17available in the LLVM source-base.  This manual is not intended to explain what
18LLVM is, how it works, and what LLVM code looks like.  It assumes that you know
19the basics of LLVM and are interested in writing transformations or otherwise
20analyzing or manipulating the code.
21
22This document should get you oriented so that you can find your way in the
23continuously growing source code that makes up the LLVM infrastructure.  Note
24that this manual is not intended to serve as a replacement for reading the
25source code, so if you think there should be a method in one of these classes to
26do something, but it's not listed, check the source.  Links to the `doxygen
27<http://llvm.org/doxygen/>`__ sources are provided to make this as easy as
28possible.
29
30The first section of this document describes general information that is useful
31to know when working in the LLVM infrastructure, and the second describes the
32Core LLVM classes.  In the future this manual will be extended with information
33describing how to use extension libraries, such as dominator information, CFG
34traversal routines, and useful utilities like the ``InstVisitor`` (`doxygen
35<http://llvm.org/doxygen/InstVisitor_8h-source.html>`__) template.
36
37.. _general:
38
39General Information
40===================
41
42This section contains general information that is useful if you are working in
43the LLVM source-base, but that isn't specific to any particular API.
44
45.. _stl:
46
47The C++ Standard Template Library
48---------------------------------
49
50LLVM makes heavy use of the C++ Standard Template Library (STL), perhaps much
51more than you are used to, or have seen before.  Because of this, you might want
52to do a little background reading in the techniques used and capabilities of the
53library.  There are many good pages that discuss the STL, and several books on
54the subject that you can get, so it will not be discussed in this document.
55
56Here are some useful links:
57
58#. `cppreference.com
59   <http://en.cppreference.com/w/>`_ - an excellent
60   reference for the STL and other parts of the standard C++ library.
61
62#. `C++ In a Nutshell <http://www.tempest-sw.com/cpp/>`_ - This is an O'Reilly
63   book in the making.  It has a decent Standard Library Reference that rivals
64   Dinkumware's, and is unfortunately no longer free since the book has been
65   published.
66
67#. `C++ Frequently Asked Questions <http://www.parashift.com/c++-faq-lite/>`_.
68
69#. `SGI's STL Programmer's Guide <http://www.sgi.com/tech/stl/>`_ - Contains a
70   useful `Introduction to the STL
71   <http://www.sgi.com/tech/stl/stl_introduction.html>`_.
72
73#. `Bjarne Stroustrup's C++ Page
74   <http://www.research.att.com/%7Ebs/C++.html>`_.
75
76#. `Bruce Eckel's Thinking in C++, 2nd ed. Volume 2 Revision 4.0
77   (even better, get the book)
78   <http://www.mindview.net/Books/TICPP/ThinkingInCPP2e.html>`_.
79
80You are also encouraged to take a look at the :doc:`LLVM Coding Standards
81<CodingStandards>` guide which focuses on how to write maintainable code more
82than where to put your curly braces.
83
84.. _resources:
85
86Other useful references
87-----------------------
88
89#. `Using static and shared libraries across platforms
90   <http://www.fortran-2000.com/ArnaudRecipes/sharedlib.html>`_
91
92.. _apis:
93
94Important and useful LLVM APIs
95==============================
96
97Here we highlight some LLVM APIs that are generally useful and good to know
98about when writing transformations.
99
100.. _isa:
101
102The ``isa<>``, ``cast<>`` and ``dyn_cast<>`` templates
103------------------------------------------------------
104
105The LLVM source-base makes extensive use of a custom form of RTTI.  These
106templates have many similarities to the C++ ``dynamic_cast<>`` operator, but
107they don't have some drawbacks (primarily stemming from the fact that
108``dynamic_cast<>`` only works on classes that have a v-table).  Because they are
109used so often, you must know what they do and how they work.  All of these
110templates are defined in the ``llvm/Support/Casting.h`` (`doxygen
111<http://llvm.org/doxygen/Casting_8h-source.html>`__) file (note that you very
112rarely have to include this file directly).
113
114``isa<>``:
115  The ``isa<>`` operator works exactly like the Java "``instanceof``" operator.
116  It returns true or false depending on whether a reference or pointer points to
117  an instance of the specified class.  This can be very useful for constraint
118  checking of various sorts (example below).
119
120``cast<>``:
121  The ``cast<>`` operator is a "checked cast" operation.  It converts a pointer
122  or reference from a base class to a derived class, causing an assertion
123  failure if it is not really an instance of the right type.  This should be
124  used in cases where you have some information that makes you believe that
125  something is of the right type.  An example of the ``isa<>`` and ``cast<>``
126  template is:
127
128  .. code-block:: c++
129
130    static bool isLoopInvariant(const Value *V, const Loop *L) {
131      if (isa<Constant>(V) || isa<Argument>(V) || isa<GlobalValue>(V))
132        return true;
133
134      // Otherwise, it must be an instruction...
135      return !L->contains(cast<Instruction>(V)->getParent());
136    }
137
138  Note that you should **not** use an ``isa<>`` test followed by a ``cast<>``,
139  for that use the ``dyn_cast<>`` operator.
140
141``dyn_cast<>``:
142  The ``dyn_cast<>`` operator is a "checking cast" operation.  It checks to see
143  if the operand is of the specified type, and if so, returns a pointer to it
144  (this operator does not work with references).  If the operand is not of the
145  correct type, a null pointer is returned.  Thus, this works very much like
146  the ``dynamic_cast<>`` operator in C++, and should be used in the same
147  circumstances.  Typically, the ``dyn_cast<>`` operator is used in an ``if``
148  statement or some other flow control statement like this:
149
150  .. code-block:: c++
151
152    if (AllocationInst *AI = dyn_cast<AllocationInst>(Val)) {
153      // ...
154    }
155
156  This form of the ``if`` statement effectively combines together a call to
157  ``isa<>`` and a call to ``cast<>`` into one statement, which is very
158  convenient.
159
160  Note that the ``dyn_cast<>`` operator, like C++'s ``dynamic_cast<>`` or Java's
161  ``instanceof`` operator, can be abused.  In particular, you should not use big
162  chained ``if/then/else`` blocks to check for lots of different variants of
163  classes.  If you find yourself wanting to do this, it is much cleaner and more
164  efficient to use the ``InstVisitor`` class to dispatch over the instruction
165  type directly.
166
167``cast_or_null<>``:
168  The ``cast_or_null<>`` operator works just like the ``cast<>`` operator,
169  except that it allows for a null pointer as an argument (which it then
170  propagates).  This can sometimes be useful, allowing you to combine several
171  null checks into one.
172
173``dyn_cast_or_null<>``:
174  The ``dyn_cast_or_null<>`` operator works just like the ``dyn_cast<>``
175  operator, except that it allows for a null pointer as an argument (which it
176  then propagates).  This can sometimes be useful, allowing you to combine
177  several null checks into one.
178
179These five templates can be used with any classes, whether they have a v-table
180or not.  If you want to add support for these templates, see the document
181:doc:`How to set up LLVM-style RTTI for your class hierarchy
182<HowToSetUpLLVMStyleRTTI>`
183
184.. _string_apis:
185
186Passing strings (the ``StringRef`` and ``Twine`` classes)
187---------------------------------------------------------
188
189Although LLVM generally does not do much string manipulation, we do have several
190important APIs which take strings.  Two important examples are the Value class
191-- which has names for instructions, functions, etc. -- and the ``StringMap``
192class which is used extensively in LLVM and Clang.
193
194These are generic classes, and they need to be able to accept strings which may
195have embedded null characters.  Therefore, they cannot simply take a ``const
196char *``, and taking a ``const std::string&`` requires clients to perform a heap
197allocation which is usually unnecessary.  Instead, many LLVM APIs use a
198``StringRef`` or a ``const Twine&`` for passing strings efficiently.
199
200.. _StringRef:
201
202The ``StringRef`` class
203^^^^^^^^^^^^^^^^^^^^^^^^^^^^
204
205The ``StringRef`` data type represents a reference to a constant string (a
206character array and a length) and supports the common operations available on
207``std::string``, but does not require heap allocation.
208
209It can be implicitly constructed using a C style null-terminated string, an
210``std::string``, or explicitly with a character pointer and length.  For
211example, the ``StringRef`` find function is declared as:
212
213.. code-block:: c++
214
215  iterator find(StringRef Key);
216
217and clients can call it using any one of:
218
219.. code-block:: c++
220
221  Map.find("foo");                 // Lookup "foo"
222  Map.find(std::string("bar"));    // Lookup "bar"
223  Map.find(StringRef("\0baz", 4)); // Lookup "\0baz"
224
225Similarly, APIs which need to return a string may return a ``StringRef``
226instance, which can be used directly or converted to an ``std::string`` using
227the ``str`` member function.  See ``llvm/ADT/StringRef.h`` (`doxygen
228<http://llvm.org/doxygen/classllvm_1_1StringRef_8h-source.html>`__) for more
229information.
230
231You should rarely use the ``StringRef`` class directly, because it contains
232pointers to external memory it is not generally safe to store an instance of the
233class (unless you know that the external storage will not be freed).
234``StringRef`` is small and pervasive enough in LLVM that it should always be
235passed by value.
236
237The ``Twine`` class
238^^^^^^^^^^^^^^^^^^^
239
240The ``Twine`` (`doxygen <http://llvm.org/doxygen/classllvm_1_1Twine.html>`__)
241class is an efficient way for APIs to accept concatenated strings.  For example,
242a common LLVM paradigm is to name one instruction based on the name of another
243instruction with a suffix, for example:
244
245.. code-block:: c++
246
247    New = CmpInst::Create(..., SO->getName() + ".cmp");
248
249The ``Twine`` class is effectively a lightweight `rope
250<http://en.wikipedia.org/wiki/Rope_(computer_science)>`_ which points to
251temporary (stack allocated) objects.  Twines can be implicitly constructed as
252the result of the plus operator applied to strings (i.e., a C strings, an
253``std::string``, or a ``StringRef``).  The twine delays the actual concatenation
254of strings until it is actually required, at which point it can be efficiently
255rendered directly into a character array.  This avoids unnecessary heap
256allocation involved in constructing the temporary results of string
257concatenation.  See ``llvm/ADT/Twine.h`` (`doxygen
258<http://llvm.org/doxygen/Twine_8h_source.html>`__) and :ref:`here <dss_twine>`
259for more information.
260
261As with a ``StringRef``, ``Twine`` objects point to external memory and should
262almost never be stored or mentioned directly.  They are intended solely for use
263when defining a function which should be able to efficiently accept concatenated
264strings.
265
266.. _error_apis:
267
268Error handling
269--------------
270
271Proper error handling helps us identify bugs in our code, and helps end-users
272understand errors in their tool usage. Errors fall into two broad categories:
273*programmatic* and *recoverable*, with different strategies for handling and
274reporting.
275
276Programmatic Errors
277^^^^^^^^^^^^^^^^^^^
278
279Programmatic errors are violations of program invariants or API contracts, and
280represent bugs within the program itself. Our aim is to document invariants, and
281to abort quickly at the point of failure (providing some basic diagnostic) when
282invariants are broken at runtime.
283
284The fundamental tools for handling programmatic errors are assertions and the
285llvm_unreachable function. Assertions are used to express invariant conditions,
286and should include a message describing the invariant:
287
288.. code-block:: c++
289
290  assert(isPhysReg(R) && "All virt regs should have been allocated already.");
291
292The llvm_unreachable function can be used to document areas of control flow
293that should never be entered if the program invariants hold:
294
295.. code-block:: c++
296
297  enum { Foo, Bar, Baz } X = foo();
298
299  switch (X) {
300    case Foo: /* Handle Foo */; break;
301    case Bar: /* Handle Bar */; break;
302    default:
303      llvm_unreachable("X should be Foo or Bar here");
304  }
305
306Recoverable Errors
307^^^^^^^^^^^^^^^^^^
308
309Recoverable errors represent an error in the program's environment, for example
310a resource failure (a missing file, a dropped network connection, etc.), or
311malformed input. These errors should be detected and communicated to a level of
312the program where they can be handled appropriately. Handling the error may be
313as simple as reporting the issue to the user, or it may involve attempts at
314recovery.
315
316Recoverable errors are modeled using LLVM's ``Error`` scheme. This scheme
317represents errors using function return values, similar to classic C integer
318error codes, or C++'s ``std::error_code``. However, the ``Error`` class is
319actually a lightweight wrapper for user-defined error types, allowing arbitrary
320information to be attached to describe the error. This is similar to the way C++
321exceptions allow throwing of user-defined types.
322
323Success values are created by calling ``Error::success()``:
324
325.. code-block:: c++
326
327  Error foo() {
328    // Do something.
329    // Return success.
330    return Error::success();
331  }
332
333Success values are very cheap to construct and return - they have minimal
334impact on program performance.
335
336Failure values are constructed using ``make_error<T>``, where ``T`` is any class
337that inherits from the ErrorInfo utility:
338
339.. code-block:: c++
340
341  class MyError : public ErrorInfo<MyError> {
342  public:
343    MyError(std::string Msg) : Msg(Msg) {}
344    void log(OStream &OS) const override { OS << "MyError - " << Msg; }
345    static char ID;
346  private:
347    std::string Msg;
348  };
349
350  char MyError::ID = 0; // In MyError.cpp
351
352  Error bar() {
353    if (checkErrorCondition)
354      return make_error<MyError>("Error condition detected");
355
356    // No error - proceed with bar.
357
358    // Return success value.
359    return Error::success();
360  }
361
362Error values can be implicitly converted to bool: true for error, false for
363success, enabling the following idiom:
364
365.. code-block:: c++
366
367  Error mayFail();
368
369  Error foo() {
370    if (auto Err = mayFail())
371      return Err;
372    // Success! We can proceed.
373    ...
374
375For functions that can fail but need to return a value the ``Expected<T>``
376utility can be used. Values of this type can be constructed with either a
377``T``, or a ``Error``. Expected<T> values are also implicitly convertible to
378boolean, but with the opposite convention to Error: true for success, false for
379error. If success, the ``T`` value can be accessed via the dereference operator.
380If failure, the ``Error`` value can be extracted using the ``takeError()``
381method. Idiomatic usage looks like:
382
383.. code-block:: c++
384
385  Expected<float> parseAndSquareRoot(IStream &IS) {
386    float f;
387    OS >> f;
388    if (f < 0)
389      return make_error<FloatingPointError>(...);
390    return sqrt(f);
391  }
392
393  Error foo(IStream &IS) {
394    if (auto SqrtOrErr = parseAndSquartRoot(IS)) {
395      float Sqrt = *SqrtOrErr;
396      // ...
397    } else
398      return SqrtOrErr.takeError();
399  }
400
401All Error instances, whether success or failure, must be either checked or
402moved from (via std::move or a return) before they are destructed. Accidentally
403discarding an unchecked error will cause a program abort at the point where the
404unchecked value's destructor is run, making it easy to identify and fix
405violations of this rule.
406
407Success values are considered checked once they have been tested (by invoking
408the boolean conversion operator):
409
410.. code-block:: c++
411
412  if (auto Err = canFail(...))
413    return Err; // Failure value - move error to caller.
414
415  // Safe to continue: Err was checked.
416
417In contrast, the following code will always cause an abort, regardless of the
418return value of ``foo``:
419
420.. code-block:: c++
421
422    canFail();
423    // Program will always abort here, even if canFail() returns Success, since
424    // the value is not checked.
425
426Failure values are considered checked once a handler for the error type has
427been activated:
428
429.. code-block:: c++
430
431  auto Err = canFail(...);
432  if (auto Err2 =
433       handleErrors(std::move(Err),
434         [](std::unique_ptr<MyError> M) {
435           // Try to handle 'M'. If successful, return a success value from
436           // the handler.
437           if (tryToHandle(M))
438             return Error::success();
439
440           // We failed to handle 'M' - return it from the handler.
441           // This value will be passed back from catchErrors and
442           // wind up in Err2, where it will be returned from this function.
443           return Error(std::move(M));
444         })))
445    return Err2;
446
447
448.. _function_apis:
449
450More information on Error and its related utilities can be found in the
451Error.h header file.
452
453Passing functions and other callable objects
454--------------------------------------------
455
456Sometimes you may want a function to be passed a callback object. In order to
457support lambda expressions and other function objects, you should not use the
458traditional C approach of taking a function pointer and an opaque cookie:
459
460.. code-block:: c++
461
462    void takeCallback(bool (*Callback)(Function *, void *), void *Cookie);
463
464Instead, use one of the following approaches:
465
466Function template
467^^^^^^^^^^^^^^^^^
468
469If you don't mind putting the definition of your function into a header file,
470make it a function template that is templated on the callable type.
471
472.. code-block:: c++
473
474    template<typename Callable>
475    void takeCallback(Callable Callback) {
476      Callback(1, 2, 3);
477    }
478
479The ``function_ref`` class template
480^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
481
482The ``function_ref``
483(`doxygen <http://llvm.org/docs/doxygen/html/classllvm_1_1function__ref_3_01Ret_07Params_8_8_8_08_4.html>`__) class
484template represents a reference to a callable object, templated over the type
485of the callable. This is a good choice for passing a callback to a function,
486if you don't need to hold onto the callback after the function returns. In this
487way, ``function_ref`` is to ``std::function`` as ``StringRef`` is to
488``std::string``.
489
490``function_ref<Ret(Param1, Param2, ...)>`` can be implicitly constructed from
491any callable object that can be called with arguments of type ``Param1``,
492``Param2``, ..., and returns a value that can be converted to type ``Ret``.
493For example:
494
495.. code-block:: c++
496
497    void visitBasicBlocks(Function *F, function_ref<bool (BasicBlock*)> Callback) {
498      for (BasicBlock &BB : *F)
499        if (Callback(&BB))
500          return;
501    }
502
503can be called using:
504
505.. code-block:: c++
506
507    visitBasicBlocks(F, [&](BasicBlock *BB) {
508      if (process(BB))
509        return isEmpty(BB);
510      return false;
511    });
512
513Note that a ``function_ref`` object contains pointers to external memory, so it
514is not generally safe to store an instance of the class (unless you know that
515the external storage will not be freed). If you need this ability, consider
516using ``std::function``. ``function_ref`` is small enough that it should always
517be passed by value.
518
519.. _DEBUG:
520
521The ``DEBUG()`` macro and ``-debug`` option
522-------------------------------------------
523
524Often when working on your pass you will put a bunch of debugging printouts and
525other code into your pass.  After you get it working, you want to remove it, but
526you may need it again in the future (to work out new bugs that you run across).
527
528Naturally, because of this, you don't want to delete the debug printouts, but
529you don't want them to always be noisy.  A standard compromise is to comment
530them out, allowing you to enable them if you need them in the future.
531
532The ``llvm/Support/Debug.h`` (`doxygen
533<http://llvm.org/doxygen/Debug_8h-source.html>`__) file provides a macro named
534``DEBUG()`` that is a much nicer solution to this problem.  Basically, you can
535put arbitrary code into the argument of the ``DEBUG`` macro, and it is only
536executed if '``opt``' (or any other tool) is run with the '``-debug``' command
537line argument:
538
539.. code-block:: c++
540
541  DEBUG(errs() << "I am here!\n");
542
543Then you can run your pass like this:
544
545.. code-block:: none
546
547  $ opt < a.bc > /dev/null -mypass
548  <no output>
549  $ opt < a.bc > /dev/null -mypass -debug
550  I am here!
551
552Using the ``DEBUG()`` macro instead of a home-brewed solution allows you to not
553have to create "yet another" command line option for the debug output for your
554pass.  Note that ``DEBUG()`` macros are disabled for non-asserts builds, so they
555do not cause a performance impact at all (for the same reason, they should also
556not contain side-effects!).
557
558One additional nice thing about the ``DEBUG()`` macro is that you can enable or
559disable it directly in gdb.  Just use "``set DebugFlag=0``" or "``set
560DebugFlag=1``" from the gdb if the program is running.  If the program hasn't
561been started yet, you can always just run it with ``-debug``.
562
563.. _DEBUG_TYPE:
564
565Fine grained debug info with ``DEBUG_TYPE`` and the ``-debug-only`` option
566^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
567
568Sometimes you may find yourself in a situation where enabling ``-debug`` just
569turns on **too much** information (such as when working on the code generator).
570If you want to enable debug information with more fine-grained control, you
571should define the ``DEBUG_TYPE`` macro and use the ``-debug-only`` option as
572follows:
573
574.. code-block:: c++
575
576  #define DEBUG_TYPE "foo"
577  DEBUG(errs() << "'foo' debug type\n");
578  #undef  DEBUG_TYPE
579  #define DEBUG_TYPE "bar"
580  DEBUG(errs() << "'bar' debug type\n"));
581  #undef  DEBUG_TYPE
582
583Then you can run your pass like this:
584
585.. code-block:: none
586
587  $ opt < a.bc > /dev/null -mypass
588  <no output>
589  $ opt < a.bc > /dev/null -mypass -debug
590  'foo' debug type
591  'bar' debug type
592  $ opt < a.bc > /dev/null -mypass -debug-only=foo
593  'foo' debug type
594  $ opt < a.bc > /dev/null -mypass -debug-only=bar
595  'bar' debug type
596  $ opt < a.bc > /dev/null -mypass -debug-only=foo,bar
597  'foo' debug type
598  'bar' debug type
599
600Of course, in practice, you should only set ``DEBUG_TYPE`` at the top of a file,
601to specify the debug type for the entire module. Be careful that you only do
602this after including Debug.h and not around any #include of headers. Also, you
603should use names more meaningful than "foo" and "bar", because there is no
604system in place to ensure that names do not conflict. If two different modules
605use the same string, they will all be turned on when the name is specified.
606This allows, for example, all debug information for instruction scheduling to be
607enabled with ``-debug-only=InstrSched``, even if the source lives in multiple
608files. The name must not include a comma (,) as that is used to separate the
609arguments of the ``-debug-only`` option.
610
611For performance reasons, -debug-only is not available in optimized build
612(``--enable-optimized``) of LLVM.
613
614The ``DEBUG_WITH_TYPE`` macro is also available for situations where you would
615like to set ``DEBUG_TYPE``, but only for one specific ``DEBUG`` statement.  It
616takes an additional first parameter, which is the type to use.  For example, the
617preceding example could be written as:
618
619.. code-block:: c++
620
621  DEBUG_WITH_TYPE("foo", errs() << "'foo' debug type\n");
622  DEBUG_WITH_TYPE("bar", errs() << "'bar' debug type\n"));
623
624.. _Statistic:
625
626The ``Statistic`` class & ``-stats`` option
627-------------------------------------------
628
629The ``llvm/ADT/Statistic.h`` (`doxygen
630<http://llvm.org/doxygen/Statistic_8h-source.html>`__) file provides a class
631named ``Statistic`` that is used as a unified way to keep track of what the LLVM
632compiler is doing and how effective various optimizations are.  It is useful to
633see what optimizations are contributing to making a particular program run
634faster.
635
636Often you may run your pass on some big program, and you're interested to see
637how many times it makes a certain transformation.  Although you can do this with
638hand inspection, or some ad-hoc method, this is a real pain and not very useful
639for big programs.  Using the ``Statistic`` class makes it very easy to keep
640track of this information, and the calculated information is presented in a
641uniform manner with the rest of the passes being executed.
642
643There are many examples of ``Statistic`` uses, but the basics of using it are as
644follows:
645
646#. Define your statistic like this:
647
648  .. code-block:: c++
649
650    #define DEBUG_TYPE "mypassname"   // This goes before any #includes.
651    STATISTIC(NumXForms, "The # of times I did stuff");
652
653  The ``STATISTIC`` macro defines a static variable, whose name is specified by
654  the first argument.  The pass name is taken from the ``DEBUG_TYPE`` macro, and
655  the description is taken from the second argument.  The variable defined
656  ("NumXForms" in this case) acts like an unsigned integer.
657
658#. Whenever you make a transformation, bump the counter:
659
660  .. code-block:: c++
661
662    ++NumXForms;   // I did stuff!
663
664That's all you have to do.  To get '``opt``' to print out the statistics
665gathered, use the '``-stats``' option:
666
667.. code-block:: none
668
669  $ opt -stats -mypassname < program.bc > /dev/null
670  ... statistics output ...
671
672Note that in order to use the '``-stats``' option, LLVM must be
673compiled with assertions enabled.
674
675When running ``opt`` on a C file from the SPEC benchmark suite, it gives a
676report that looks like this:
677
678.. code-block:: none
679
680   7646 bitcodewriter   - Number of normal instructions
681    725 bitcodewriter   - Number of oversized instructions
682 129996 bitcodewriter   - Number of bitcode bytes written
683   2817 raise           - Number of insts DCEd or constprop'd
684   3213 raise           - Number of cast-of-self removed
685   5046 raise           - Number of expression trees converted
686     75 raise           - Number of other getelementptr's formed
687    138 raise           - Number of load/store peepholes
688     42 deadtypeelim    - Number of unused typenames removed from symtab
689    392 funcresolve     - Number of varargs functions resolved
690     27 globaldce       - Number of global variables removed
691      2 adce            - Number of basic blocks removed
692    134 cee             - Number of branches revectored
693     49 cee             - Number of setcc instruction eliminated
694    532 gcse            - Number of loads removed
695   2919 gcse            - Number of instructions removed
696     86 indvars         - Number of canonical indvars added
697     87 indvars         - Number of aux indvars removed
698     25 instcombine     - Number of dead inst eliminate
699    434 instcombine     - Number of insts combined
700    248 licm            - Number of load insts hoisted
701   1298 licm            - Number of insts hoisted to a loop pre-header
702      3 licm            - Number of insts hoisted to multiple loop preds (bad, no loop pre-header)
703     75 mem2reg         - Number of alloca's promoted
704   1444 cfgsimplify     - Number of blocks simplified
705
706Obviously, with so many optimizations, having a unified framework for this stuff
707is very nice.  Making your pass fit well into the framework makes it more
708maintainable and useful.
709
710.. _ViewGraph:
711
712Viewing graphs while debugging code
713-----------------------------------
714
715Several of the important data structures in LLVM are graphs: for example CFGs
716made out of LLVM :ref:`BasicBlocks <BasicBlock>`, CFGs made out of LLVM
717:ref:`MachineBasicBlocks <MachineBasicBlock>`, and :ref:`Instruction Selection
718DAGs <SelectionDAG>`.  In many cases, while debugging various parts of the
719compiler, it is nice to instantly visualize these graphs.
720
721LLVM provides several callbacks that are available in a debug build to do
722exactly that.  If you call the ``Function::viewCFG()`` method, for example, the
723current LLVM tool will pop up a window containing the CFG for the function where
724each basic block is a node in the graph, and each node contains the instructions
725in the block.  Similarly, there also exists ``Function::viewCFGOnly()`` (does
726not include the instructions), the ``MachineFunction::viewCFG()`` and
727``MachineFunction::viewCFGOnly()``, and the ``SelectionDAG::viewGraph()``
728methods.  Within GDB, for example, you can usually use something like ``call
729DAG.viewGraph()`` to pop up a window.  Alternatively, you can sprinkle calls to
730these functions in your code in places you want to debug.
731
732Getting this to work requires a small amount of setup.  On Unix systems
733with X11, install the `graphviz <http://www.graphviz.org>`_ toolkit, and make
734sure 'dot' and 'gv' are in your path.  If you are running on Mac OS X, download
735and install the Mac OS X `Graphviz program
736<http://www.pixelglow.com/graphviz/>`_ and add
737``/Applications/Graphviz.app/Contents/MacOS/`` (or wherever you install it) to
738your path. The programs need not be present when configuring, building or
739running LLVM and can simply be installed when needed during an active debug
740session.
741
742``SelectionDAG`` has been extended to make it easier to locate *interesting*
743nodes in large complex graphs.  From gdb, if you ``call DAG.setGraphColor(node,
744"color")``, then the next ``call DAG.viewGraph()`` would highlight the node in
745the specified color (choices of colors can be found at `colors
746<http://www.graphviz.org/doc/info/colors.html>`_.) More complex node attributes
747can be provided with ``call DAG.setGraphAttrs(node, "attributes")`` (choices can
748be found at `Graph attributes <http://www.graphviz.org/doc/info/attrs.html>`_.)
749If you want to restart and clear all the current graph attributes, then you can
750``call DAG.clearGraphAttrs()``.
751
752Note that graph visualization features are compiled out of Release builds to
753reduce file size.  This means that you need a Debug+Asserts or Release+Asserts
754build to use these features.
755
756.. _datastructure:
757
758Picking the Right Data Structure for a Task
759===========================================
760
761LLVM has a plethora of data structures in the ``llvm/ADT/`` directory, and we
762commonly use STL data structures.  This section describes the trade-offs you
763should consider when you pick one.
764
765The first step is a choose your own adventure: do you want a sequential
766container, a set-like container, or a map-like container?  The most important
767thing when choosing a container is the algorithmic properties of how you plan to
768access the container.  Based on that, you should use:
769
770
771* a :ref:`map-like <ds_map>` container if you need efficient look-up of a
772  value based on another value.  Map-like containers also support efficient
773  queries for containment (whether a key is in the map).  Map-like containers
774  generally do not support efficient reverse mapping (values to keys).  If you
775  need that, use two maps.  Some map-like containers also support efficient
776  iteration through the keys in sorted order.  Map-like containers are the most
777  expensive sort, only use them if you need one of these capabilities.
778
779* a :ref:`set-like <ds_set>` container if you need to put a bunch of stuff into
780  a container that automatically eliminates duplicates.  Some set-like
781  containers support efficient iteration through the elements in sorted order.
782  Set-like containers are more expensive than sequential containers.
783
784* a :ref:`sequential <ds_sequential>` container provides the most efficient way
785  to add elements and keeps track of the order they are added to the collection.
786  They permit duplicates and support efficient iteration, but do not support
787  efficient look-up based on a key.
788
789* a :ref:`string <ds_string>` container is a specialized sequential container or
790  reference structure that is used for character or byte arrays.
791
792* a :ref:`bit <ds_bit>` container provides an efficient way to store and
793  perform set operations on sets of numeric id's, while automatically
794  eliminating duplicates.  Bit containers require a maximum of 1 bit for each
795  identifier you want to store.
796
797Once the proper category of container is determined, you can fine tune the
798memory use, constant factors, and cache behaviors of access by intelligently
799picking a member of the category.  Note that constant factors and cache behavior
800can be a big deal.  If you have a vector that usually only contains a few
801elements (but could contain many), for example, it's much better to use
802:ref:`SmallVector <dss_smallvector>` than :ref:`vector <dss_vector>`.  Doing so
803avoids (relatively) expensive malloc/free calls, which dwarf the cost of adding
804the elements to the container.
805
806.. _ds_sequential:
807
808Sequential Containers (std::vector, std::list, etc)
809---------------------------------------------------
810
811There are a variety of sequential containers available for you, based on your
812needs.  Pick the first in this section that will do what you want.
813
814.. _dss_arrayref:
815
816llvm/ADT/ArrayRef.h
817^^^^^^^^^^^^^^^^^^^
818
819The ``llvm::ArrayRef`` class is the preferred class to use in an interface that
820accepts a sequential list of elements in memory and just reads from them.  By
821taking an ``ArrayRef``, the API can be passed a fixed size array, an
822``std::vector``, an ``llvm::SmallVector`` and anything else that is contiguous
823in memory.
824
825.. _dss_fixedarrays:
826
827Fixed Size Arrays
828^^^^^^^^^^^^^^^^^
829
830Fixed size arrays are very simple and very fast.  They are good if you know
831exactly how many elements you have, or you have a (low) upper bound on how many
832you have.
833
834.. _dss_heaparrays:
835
836Heap Allocated Arrays
837^^^^^^^^^^^^^^^^^^^^^
838
839Heap allocated arrays (``new[]`` + ``delete[]``) are also simple.  They are good
840if the number of elements is variable, if you know how many elements you will
841need before the array is allocated, and if the array is usually large (if not,
842consider a :ref:`SmallVector <dss_smallvector>`).  The cost of a heap allocated
843array is the cost of the new/delete (aka malloc/free).  Also note that if you
844are allocating an array of a type with a constructor, the constructor and
845destructors will be run for every element in the array (re-sizable vectors only
846construct those elements actually used).
847
848.. _dss_tinyptrvector:
849
850llvm/ADT/TinyPtrVector.h
851^^^^^^^^^^^^^^^^^^^^^^^^
852
853``TinyPtrVector<Type>`` is a highly specialized collection class that is
854optimized to avoid allocation in the case when a vector has zero or one
855elements.  It has two major restrictions: 1) it can only hold values of pointer
856type, and 2) it cannot hold a null pointer.
857
858Since this container is highly specialized, it is rarely used.
859
860.. _dss_smallvector:
861
862llvm/ADT/SmallVector.h
863^^^^^^^^^^^^^^^^^^^^^^
864
865``SmallVector<Type, N>`` is a simple class that looks and smells just like
866``vector<Type>``: it supports efficient iteration, lays out elements in memory
867order (so you can do pointer arithmetic between elements), supports efficient
868push_back/pop_back operations, supports efficient random access to its elements,
869etc.
870
871The advantage of SmallVector is that it allocates space for some number of
872elements (N) **in the object itself**.  Because of this, if the SmallVector is
873dynamically smaller than N, no malloc is performed.  This can be a big win in
874cases where the malloc/free call is far more expensive than the code that
875fiddles around with the elements.
876
877This is good for vectors that are "usually small" (e.g. the number of
878predecessors/successors of a block is usually less than 8).  On the other hand,
879this makes the size of the SmallVector itself large, so you don't want to
880allocate lots of them (doing so will waste a lot of space).  As such,
881SmallVectors are most useful when on the stack.
882
883SmallVector also provides a nice portable and efficient replacement for
884``alloca``.
885
886.. note::
887
888   Prefer to use ``SmallVectorImpl<T>`` as a parameter type.
889
890   In APIs that don't care about the "small size" (most?), prefer to use
891   the ``SmallVectorImpl<T>`` class, which is basically just the "vector
892   header" (and methods) without the elements allocated after it. Note that
893   ``SmallVector<T, N>`` inherits from ``SmallVectorImpl<T>`` so the
894   conversion is implicit and costs nothing. E.g.
895
896   .. code-block:: c++
897
898      // BAD: Clients cannot pass e.g. SmallVector<Foo, 4>.
899      hardcodedSmallSize(SmallVector<Foo, 2> &Out);
900      // GOOD: Clients can pass any SmallVector<Foo, N>.
901      allowsAnySmallSize(SmallVectorImpl<Foo> &Out);
902
903      void someFunc() {
904        SmallVector<Foo, 8> Vec;
905        hardcodedSmallSize(Vec); // Error.
906        allowsAnySmallSize(Vec); // Works.
907      }
908
909   Even though it has "``Impl``" in the name, this is so widely used that
910   it really isn't "private to the implementation" anymore. A name like
911   ``SmallVectorHeader`` would be more appropriate.
912
913.. _dss_vector:
914
915<vector>
916^^^^^^^^
917
918``std::vector`` is well loved and respected.  It is useful when SmallVector
919isn't: when the size of the vector is often large (thus the small optimization
920will rarely be a benefit) or if you will be allocating many instances of the
921vector itself (which would waste space for elements that aren't in the
922container).  vector is also useful when interfacing with code that expects
923vectors :).
924
925One worthwhile note about std::vector: avoid code like this:
926
927.. code-block:: c++
928
929  for ( ... ) {
930     std::vector<foo> V;
931     // make use of V.
932  }
933
934Instead, write this as:
935
936.. code-block:: c++
937
938  std::vector<foo> V;
939  for ( ... ) {
940     // make use of V.
941     V.clear();
942  }
943
944Doing so will save (at least) one heap allocation and free per iteration of the
945loop.
946
947.. _dss_deque:
948
949<deque>
950^^^^^^^
951
952``std::deque`` is, in some senses, a generalized version of ``std::vector``.
953Like ``std::vector``, it provides constant time random access and other similar
954properties, but it also provides efficient access to the front of the list.  It
955does not guarantee continuity of elements within memory.
956
957In exchange for this extra flexibility, ``std::deque`` has significantly higher
958constant factor costs than ``std::vector``.  If possible, use ``std::vector`` or
959something cheaper.
960
961.. _dss_list:
962
963<list>
964^^^^^^
965
966``std::list`` is an extremely inefficient class that is rarely useful.  It
967performs a heap allocation for every element inserted into it, thus having an
968extremely high constant factor, particularly for small data types.
969``std::list`` also only supports bidirectional iteration, not random access
970iteration.
971
972In exchange for this high cost, std::list supports efficient access to both ends
973of the list (like ``std::deque``, but unlike ``std::vector`` or
974``SmallVector``).  In addition, the iterator invalidation characteristics of
975std::list are stronger than that of a vector class: inserting or removing an
976element into the list does not invalidate iterator or pointers to other elements
977in the list.
978
979.. _dss_ilist:
980
981llvm/ADT/ilist.h
982^^^^^^^^^^^^^^^^
983
984``ilist<T>`` implements an 'intrusive' doubly-linked list.  It is intrusive,
985because it requires the element to store and provide access to the prev/next
986pointers for the list.
987
988``ilist`` has the same drawbacks as ``std::list``, and additionally requires an
989``ilist_traits`` implementation for the element type, but it provides some novel
990characteristics.  In particular, it can efficiently store polymorphic objects,
991the traits class is informed when an element is inserted or removed from the
992list, and ``ilist``\ s are guaranteed to support a constant-time splice
993operation.
994
995These properties are exactly what we want for things like ``Instruction``\ s and
996basic blocks, which is why these are implemented with ``ilist``\ s.
997
998Related classes of interest are explained in the following subsections:
999
1000* :ref:`ilist_traits <dss_ilist_traits>`
1001
1002* :ref:`iplist <dss_iplist>`
1003
1004* :ref:`llvm/ADT/ilist_node.h <dss_ilist_node>`
1005
1006* :ref:`Sentinels <dss_ilist_sentinel>`
1007
1008.. _dss_packedvector:
1009
1010llvm/ADT/PackedVector.h
1011^^^^^^^^^^^^^^^^^^^^^^^
1012
1013Useful for storing a vector of values using only a few number of bits for each
1014value.  Apart from the standard operations of a vector-like container, it can
1015also perform an 'or' set operation.
1016
1017For example:
1018
1019.. code-block:: c++
1020
1021  enum State {
1022      None = 0x0,
1023      FirstCondition = 0x1,
1024      SecondCondition = 0x2,
1025      Both = 0x3
1026  };
1027
1028  State get() {
1029      PackedVector<State, 2> Vec1;
1030      Vec1.push_back(FirstCondition);
1031
1032      PackedVector<State, 2> Vec2;
1033      Vec2.push_back(SecondCondition);
1034
1035      Vec1 |= Vec2;
1036      return Vec1[0]; // returns 'Both'.
1037  }
1038
1039.. _dss_ilist_traits:
1040
1041ilist_traits
1042^^^^^^^^^^^^
1043
1044``ilist_traits<T>`` is ``ilist<T>``'s customization mechanism. ``iplist<T>``
1045(and consequently ``ilist<T>``) publicly derive from this traits class.
1046
1047.. _dss_iplist:
1048
1049iplist
1050^^^^^^
1051
1052``iplist<T>`` is ``ilist<T>``'s base and as such supports a slightly narrower
1053interface.  Notably, inserters from ``T&`` are absent.
1054
1055``ilist_traits<T>`` is a public base of this class and can be used for a wide
1056variety of customizations.
1057
1058.. _dss_ilist_node:
1059
1060llvm/ADT/ilist_node.h
1061^^^^^^^^^^^^^^^^^^^^^
1062
1063``ilist_node<T>`` implements the forward and backward links that are expected
1064by the ``ilist<T>`` (and analogous containers) in the default manner.
1065
1066``ilist_node<T>``\ s are meant to be embedded in the node type ``T``, usually
1067``T`` publicly derives from ``ilist_node<T>``.
1068
1069.. _dss_ilist_sentinel:
1070
1071Sentinels
1072^^^^^^^^^
1073
1074``ilist``\ s have another specialty that must be considered.  To be a good
1075citizen in the C++ ecosystem, it needs to support the standard container
1076operations, such as ``begin`` and ``end`` iterators, etc.  Also, the
1077``operator--`` must work correctly on the ``end`` iterator in the case of
1078non-empty ``ilist``\ s.
1079
1080The only sensible solution to this problem is to allocate a so-called *sentinel*
1081along with the intrusive list, which serves as the ``end`` iterator, providing
1082the back-link to the last element.  However conforming to the C++ convention it
1083is illegal to ``operator++`` beyond the sentinel and it also must not be
1084dereferenced.
1085
1086These constraints allow for some implementation freedom to the ``ilist`` how to
1087allocate and store the sentinel.  The corresponding policy is dictated by
1088``ilist_traits<T>``.  By default a ``T`` gets heap-allocated whenever the need
1089for a sentinel arises.
1090
1091While the default policy is sufficient in most cases, it may break down when
1092``T`` does not provide a default constructor.  Also, in the case of many
1093instances of ``ilist``\ s, the memory overhead of the associated sentinels is
1094wasted.  To alleviate the situation with numerous and voluminous
1095``T``-sentinels, sometimes a trick is employed, leading to *ghostly sentinels*.
1096
1097Ghostly sentinels are obtained by specially-crafted ``ilist_traits<T>`` which
1098superpose the sentinel with the ``ilist`` instance in memory.  Pointer
1099arithmetic is used to obtain the sentinel, which is relative to the ``ilist``'s
1100``this`` pointer.  The ``ilist`` is augmented by an extra pointer, which serves
1101as the back-link of the sentinel.  This is the only field in the ghostly
1102sentinel which can be legally accessed.
1103
1104.. _dss_other:
1105
1106Other Sequential Container options
1107^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
1108
1109Other STL containers are available, such as ``std::string``.
1110
1111There are also various STL adapter classes such as ``std::queue``,
1112``std::priority_queue``, ``std::stack``, etc.  These provide simplified access
1113to an underlying container but don't affect the cost of the container itself.
1114
1115.. _ds_string:
1116
1117String-like containers
1118----------------------
1119
1120There are a variety of ways to pass around and use strings in C and C++, and
1121LLVM adds a few new options to choose from.  Pick the first option on this list
1122that will do what you need, they are ordered according to their relative cost.
1123
1124Note that it is generally preferred to *not* pass strings around as ``const
1125char*``'s.  These have a number of problems, including the fact that they
1126cannot represent embedded nul ("\0") characters, and do not have a length
1127available efficiently.  The general replacement for '``const char*``' is
1128StringRef.
1129
1130For more information on choosing string containers for APIs, please see
1131:ref:`Passing Strings <string_apis>`.
1132
1133.. _dss_stringref:
1134
1135llvm/ADT/StringRef.h
1136^^^^^^^^^^^^^^^^^^^^
1137
1138The StringRef class is a simple value class that contains a pointer to a
1139character and a length, and is quite related to the :ref:`ArrayRef
1140<dss_arrayref>` class (but specialized for arrays of characters).  Because
1141StringRef carries a length with it, it safely handles strings with embedded nul
1142characters in it, getting the length does not require a strlen call, and it even
1143has very convenient APIs for slicing and dicing the character range that it
1144represents.
1145
1146StringRef is ideal for passing simple strings around that are known to be live,
1147either because they are C string literals, std::string, a C array, or a
1148SmallVector.  Each of these cases has an efficient implicit conversion to
1149StringRef, which doesn't result in a dynamic strlen being executed.
1150
1151StringRef has a few major limitations which make more powerful string containers
1152useful:
1153
1154#. You cannot directly convert a StringRef to a 'const char*' because there is
1155   no way to add a trailing nul (unlike the .c_str() method on various stronger
1156   classes).
1157
1158#. StringRef doesn't own or keep alive the underlying string bytes.
1159   As such it can easily lead to dangling pointers, and is not suitable for
1160   embedding in datastructures in most cases (instead, use an std::string or
1161   something like that).
1162
1163#. For the same reason, StringRef cannot be used as the return value of a
1164   method if the method "computes" the result string.  Instead, use std::string.
1165
1166#. StringRef's do not allow you to mutate the pointed-to string bytes and it
1167   doesn't allow you to insert or remove bytes from the range.  For editing
1168   operations like this, it interoperates with the :ref:`Twine <dss_twine>`
1169   class.
1170
1171Because of its strengths and limitations, it is very common for a function to
1172take a StringRef and for a method on an object to return a StringRef that points
1173into some string that it owns.
1174
1175.. _dss_twine:
1176
1177llvm/ADT/Twine.h
1178^^^^^^^^^^^^^^^^
1179
1180The Twine class is used as an intermediary datatype for APIs that want to take a
1181string that can be constructed inline with a series of concatenations.  Twine
1182works by forming recursive instances of the Twine datatype (a simple value
1183object) on the stack as temporary objects, linking them together into a tree
1184which is then linearized when the Twine is consumed.  Twine is only safe to use
1185as the argument to a function, and should always be a const reference, e.g.:
1186
1187.. code-block:: c++
1188
1189  void foo(const Twine &T);
1190  ...
1191  StringRef X = ...
1192  unsigned i = ...
1193  foo(X + "." + Twine(i));
1194
1195This example forms a string like "blarg.42" by concatenating the values
1196together, and does not form intermediate strings containing "blarg" or "blarg.".
1197
1198Because Twine is constructed with temporary objects on the stack, and because
1199these instances are destroyed at the end of the current statement, it is an
1200inherently dangerous API.  For example, this simple variant contains undefined
1201behavior and will probably crash:
1202
1203.. code-block:: c++
1204
1205  void foo(const Twine &T);
1206  ...
1207  StringRef X = ...
1208  unsigned i = ...
1209  const Twine &Tmp = X + "." + Twine(i);
1210  foo(Tmp);
1211
1212... because the temporaries are destroyed before the call.  That said, Twine's
1213are much more efficient than intermediate std::string temporaries, and they work
1214really well with StringRef.  Just be aware of their limitations.
1215
1216.. _dss_smallstring:
1217
1218llvm/ADT/SmallString.h
1219^^^^^^^^^^^^^^^^^^^^^^
1220
1221SmallString is a subclass of :ref:`SmallVector <dss_smallvector>` that adds some
1222convenience APIs like += that takes StringRef's.  SmallString avoids allocating
1223memory in the case when the preallocated space is enough to hold its data, and
1224it calls back to general heap allocation when required.  Since it owns its data,
1225it is very safe to use and supports full mutation of the string.
1226
1227Like SmallVector's, the big downside to SmallString is their sizeof.  While they
1228are optimized for small strings, they themselves are not particularly small.
1229This means that they work great for temporary scratch buffers on the stack, but
1230should not generally be put into the heap: it is very rare to see a SmallString
1231as the member of a frequently-allocated heap data structure or returned
1232by-value.
1233
1234.. _dss_stdstring:
1235
1236std::string
1237^^^^^^^^^^^
1238
1239The standard C++ std::string class is a very general class that (like
1240SmallString) owns its underlying data.  sizeof(std::string) is very reasonable
1241so it can be embedded into heap data structures and returned by-value.  On the
1242other hand, std::string is highly inefficient for inline editing (e.g.
1243concatenating a bunch of stuff together) and because it is provided by the
1244standard library, its performance characteristics depend a lot of the host
1245standard library (e.g. libc++ and MSVC provide a highly optimized string class,
1246GCC contains a really slow implementation).
1247
1248The major disadvantage of std::string is that almost every operation that makes
1249them larger can allocate memory, which is slow.  As such, it is better to use
1250SmallVector or Twine as a scratch buffer, but then use std::string to persist
1251the result.
1252
1253.. _ds_set:
1254
1255Set-Like Containers (std::set, SmallSet, SetVector, etc)
1256--------------------------------------------------------
1257
1258Set-like containers are useful when you need to canonicalize multiple values
1259into a single representation.  There are several different choices for how to do
1260this, providing various trade-offs.
1261
1262.. _dss_sortedvectorset:
1263
1264A sorted 'vector'
1265^^^^^^^^^^^^^^^^^
1266
1267If you intend to insert a lot of elements, then do a lot of queries, a great
1268approach is to use a vector (or other sequential container) with
1269std::sort+std::unique to remove duplicates.  This approach works really well if
1270your usage pattern has these two distinct phases (insert then query), and can be
1271coupled with a good choice of :ref:`sequential container <ds_sequential>`.
1272
1273This combination provides the several nice properties: the result data is
1274contiguous in memory (good for cache locality), has few allocations, is easy to
1275address (iterators in the final vector are just indices or pointers), and can be
1276efficiently queried with a standard binary search (e.g.
1277``std::lower_bound``; if you want the whole range of elements comparing
1278equal, use ``std::equal_range``).
1279
1280.. _dss_smallset:
1281
1282llvm/ADT/SmallSet.h
1283^^^^^^^^^^^^^^^^^^^
1284
1285If you have a set-like data structure that is usually small and whose elements
1286are reasonably small, a ``SmallSet<Type, N>`` is a good choice.  This set has
1287space for N elements in place (thus, if the set is dynamically smaller than N,
1288no malloc traffic is required) and accesses them with a simple linear search.
1289When the set grows beyond N elements, it allocates a more expensive
1290representation that guarantees efficient access (for most types, it falls back
1291to :ref:`std::set <dss_set>`, but for pointers it uses something far better,
1292:ref:`SmallPtrSet <dss_smallptrset>`.
1293
1294The magic of this class is that it handles small sets extremely efficiently, but
1295gracefully handles extremely large sets without loss of efficiency.  The
1296drawback is that the interface is quite small: it supports insertion, queries
1297and erasing, but does not support iteration.
1298
1299.. _dss_smallptrset:
1300
1301llvm/ADT/SmallPtrSet.h
1302^^^^^^^^^^^^^^^^^^^^^^
1303
1304``SmallPtrSet`` has all the advantages of ``SmallSet`` (and a ``SmallSet`` of
1305pointers is transparently implemented with a ``SmallPtrSet``), but also supports
1306iterators.  If more than N insertions are performed, a single quadratically
1307probed hash table is allocated and grows as needed, providing extremely
1308efficient access (constant time insertion/deleting/queries with low constant
1309factors) and is very stingy with malloc traffic.
1310
1311Note that, unlike :ref:`std::set <dss_set>`, the iterators of ``SmallPtrSet``
1312are invalidated whenever an insertion occurs.  Also, the values visited by the
1313iterators are not visited in sorted order.
1314
1315.. _dss_stringset:
1316
1317llvm/ADT/StringSet.h
1318^^^^^^^^^^^^^^^^^^^^
1319
1320``StringSet`` is a thin wrapper around :ref:`StringMap\<char\> <dss_stringmap>`,
1321and it allows efficient storage and retrieval of unique strings.
1322
1323Functionally analogous to ``SmallSet<StringRef>``, ``StringSet`` also supports
1324iteration. (The iterator dereferences to a ``StringMapEntry<char>``, so you
1325need to call ``i->getKey()`` to access the item of the StringSet.)  On the
1326other hand, ``StringSet`` doesn't support range-insertion and
1327copy-construction, which :ref:`SmallSet <dss_smallset>` and :ref:`SmallPtrSet
1328<dss_smallptrset>` do support.
1329
1330.. _dss_denseset:
1331
1332llvm/ADT/DenseSet.h
1333^^^^^^^^^^^^^^^^^^^
1334
1335DenseSet is a simple quadratically probed hash table.  It excels at supporting
1336small values: it uses a single allocation to hold all of the pairs that are
1337currently inserted in the set.  DenseSet is a great way to unique small values
1338that are not simple pointers (use :ref:`SmallPtrSet <dss_smallptrset>` for
1339pointers).  Note that DenseSet has the same requirements for the value type that
1340:ref:`DenseMap <dss_densemap>` has.
1341
1342.. _dss_sparseset:
1343
1344llvm/ADT/SparseSet.h
1345^^^^^^^^^^^^^^^^^^^^
1346
1347SparseSet holds a small number of objects identified by unsigned keys of
1348moderate size.  It uses a lot of memory, but provides operations that are almost
1349as fast as a vector.  Typical keys are physical registers, virtual registers, or
1350numbered basic blocks.
1351
1352SparseSet is useful for algorithms that need very fast clear/find/insert/erase
1353and fast iteration over small sets.  It is not intended for building composite
1354data structures.
1355
1356.. _dss_sparsemultiset:
1357
1358llvm/ADT/SparseMultiSet.h
1359^^^^^^^^^^^^^^^^^^^^^^^^^^^^
1360
1361SparseMultiSet adds multiset behavior to SparseSet, while retaining SparseSet's
1362desirable attributes. Like SparseSet, it typically uses a lot of memory, but
1363provides operations that are almost as fast as a vector.  Typical keys are
1364physical registers, virtual registers, or numbered basic blocks.
1365
1366SparseMultiSet is useful for algorithms that need very fast
1367clear/find/insert/erase of the entire collection, and iteration over sets of
1368elements sharing a key. It is often a more efficient choice than using composite
1369data structures (e.g. vector-of-vectors, map-of-vectors). It is not intended for
1370building composite data structures.
1371
1372.. _dss_FoldingSet:
1373
1374llvm/ADT/FoldingSet.h
1375^^^^^^^^^^^^^^^^^^^^^
1376
1377FoldingSet is an aggregate class that is really good at uniquing
1378expensive-to-create or polymorphic objects.  It is a combination of a chained
1379hash table with intrusive links (uniqued objects are required to inherit from
1380FoldingSetNode) that uses :ref:`SmallVector <dss_smallvector>` as part of its ID
1381process.
1382
1383Consider a case where you want to implement a "getOrCreateFoo" method for a
1384complex object (for example, a node in the code generator).  The client has a
1385description of **what** it wants to generate (it knows the opcode and all the
1386operands), but we don't want to 'new' a node, then try inserting it into a set
1387only to find out it already exists, at which point we would have to delete it
1388and return the node that already exists.
1389
1390To support this style of client, FoldingSet perform a query with a
1391FoldingSetNodeID (which wraps SmallVector) that can be used to describe the
1392element that we want to query for.  The query either returns the element
1393matching the ID or it returns an opaque ID that indicates where insertion should
1394take place.  Construction of the ID usually does not require heap traffic.
1395
1396Because FoldingSet uses intrusive links, it can support polymorphic objects in
1397the set (for example, you can have SDNode instances mixed with LoadSDNodes).
1398Because the elements are individually allocated, pointers to the elements are
1399stable: inserting or removing elements does not invalidate any pointers to other
1400elements.
1401
1402.. _dss_set:
1403
1404<set>
1405^^^^^
1406
1407``std::set`` is a reasonable all-around set class, which is decent at many
1408things but great at nothing.  std::set allocates memory for each element
1409inserted (thus it is very malloc intensive) and typically stores three pointers
1410per element in the set (thus adding a large amount of per-element space
1411overhead).  It offers guaranteed log(n) performance, which is not particularly
1412fast from a complexity standpoint (particularly if the elements of the set are
1413expensive to compare, like strings), and has extremely high constant factors for
1414lookup, insertion and removal.
1415
1416The advantages of std::set are that its iterators are stable (deleting or
1417inserting an element from the set does not affect iterators or pointers to other
1418elements) and that iteration over the set is guaranteed to be in sorted order.
1419If the elements in the set are large, then the relative overhead of the pointers
1420and malloc traffic is not a big deal, but if the elements of the set are small,
1421std::set is almost never a good choice.
1422
1423.. _dss_setvector:
1424
1425llvm/ADT/SetVector.h
1426^^^^^^^^^^^^^^^^^^^^
1427
1428LLVM's ``SetVector<Type>`` is an adapter class that combines your choice of a
1429set-like container along with a :ref:`Sequential Container <ds_sequential>` The
1430important property that this provides is efficient insertion with uniquing
1431(duplicate elements are ignored) with iteration support.  It implements this by
1432inserting elements into both a set-like container and the sequential container,
1433using the set-like container for uniquing and the sequential container for
1434iteration.
1435
1436The difference between SetVector and other sets is that the order of iteration
1437is guaranteed to match the order of insertion into the SetVector.  This property
1438is really important for things like sets of pointers.  Because pointer values
1439are non-deterministic (e.g. vary across runs of the program on different
1440machines), iterating over the pointers in the set will not be in a well-defined
1441order.
1442
1443The drawback of SetVector is that it requires twice as much space as a normal
1444set and has the sum of constant factors from the set-like container and the
1445sequential container that it uses.  Use it **only** if you need to iterate over
1446the elements in a deterministic order.  SetVector is also expensive to delete
1447elements out of (linear time), unless you use its "pop_back" method, which is
1448faster.
1449
1450``SetVector`` is an adapter class that defaults to using ``std::vector`` and a
1451size 16 ``SmallSet`` for the underlying containers, so it is quite expensive.
1452However, ``"llvm/ADT/SetVector.h"`` also provides a ``SmallSetVector`` class,
1453which defaults to using a ``SmallVector`` and ``SmallSet`` of a specified size.
1454If you use this, and if your sets are dynamically smaller than ``N``, you will
1455save a lot of heap traffic.
1456
1457.. _dss_uniquevector:
1458
1459llvm/ADT/UniqueVector.h
1460^^^^^^^^^^^^^^^^^^^^^^^
1461
1462UniqueVector is similar to :ref:`SetVector <dss_setvector>` but it retains a
1463unique ID for each element inserted into the set.  It internally contains a map
1464and a vector, and it assigns a unique ID for each value inserted into the set.
1465
1466UniqueVector is very expensive: its cost is the sum of the cost of maintaining
1467both the map and vector, it has high complexity, high constant factors, and
1468produces a lot of malloc traffic.  It should be avoided.
1469
1470.. _dss_immutableset:
1471
1472llvm/ADT/ImmutableSet.h
1473^^^^^^^^^^^^^^^^^^^^^^^
1474
1475ImmutableSet is an immutable (functional) set implementation based on an AVL
1476tree.  Adding or removing elements is done through a Factory object and results
1477in the creation of a new ImmutableSet object.  If an ImmutableSet already exists
1478with the given contents, then the existing one is returned; equality is compared
1479with a FoldingSetNodeID.  The time and space complexity of add or remove
1480operations is logarithmic in the size of the original set.
1481
1482There is no method for returning an element of the set, you can only check for
1483membership.
1484
1485.. _dss_otherset:
1486
1487Other Set-Like Container Options
1488^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
1489
1490The STL provides several other options, such as std::multiset and the various
1491"hash_set" like containers (whether from C++ TR1 or from the SGI library).  We
1492never use hash_set and unordered_set because they are generally very expensive
1493(each insertion requires a malloc) and very non-portable.
1494
1495std::multiset is useful if you're not interested in elimination of duplicates,
1496but has all the drawbacks of :ref:`std::set <dss_set>`.  A sorted vector
1497(where you don't delete duplicate entries) or some other approach is almost
1498always better.
1499
1500.. _ds_map:
1501
1502Map-Like Containers (std::map, DenseMap, etc)
1503---------------------------------------------
1504
1505Map-like containers are useful when you want to associate data to a key.  As
1506usual, there are a lot of different ways to do this. :)
1507
1508.. _dss_sortedvectormap:
1509
1510A sorted 'vector'
1511^^^^^^^^^^^^^^^^^
1512
1513If your usage pattern follows a strict insert-then-query approach, you can
1514trivially use the same approach as :ref:`sorted vectors for set-like containers
1515<dss_sortedvectorset>`.  The only difference is that your query function (which
1516uses std::lower_bound to get efficient log(n) lookup) should only compare the
1517key, not both the key and value.  This yields the same advantages as sorted
1518vectors for sets.
1519
1520.. _dss_stringmap:
1521
1522llvm/ADT/StringMap.h
1523^^^^^^^^^^^^^^^^^^^^
1524
1525Strings are commonly used as keys in maps, and they are difficult to support
1526efficiently: they are variable length, inefficient to hash and compare when
1527long, expensive to copy, etc.  StringMap is a specialized container designed to
1528cope with these issues.  It supports mapping an arbitrary range of bytes to an
1529arbitrary other object.
1530
1531The StringMap implementation uses a quadratically-probed hash table, where the
1532buckets store a pointer to the heap allocated entries (and some other stuff).
1533The entries in the map must be heap allocated because the strings are variable
1534length.  The string data (key) and the element object (value) are stored in the
1535same allocation with the string data immediately after the element object.
1536This container guarantees the "``(char*)(&Value+1)``" points to the key string
1537for a value.
1538
1539The StringMap is very fast for several reasons: quadratic probing is very cache
1540efficient for lookups, the hash value of strings in buckets is not recomputed
1541when looking up an element, StringMap rarely has to touch the memory for
1542unrelated objects when looking up a value (even when hash collisions happen),
1543hash table growth does not recompute the hash values for strings already in the
1544table, and each pair in the map is store in a single allocation (the string data
1545is stored in the same allocation as the Value of a pair).
1546
1547StringMap also provides query methods that take byte ranges, so it only ever
1548copies a string if a value is inserted into the table.
1549
1550StringMap iteratation order, however, is not guaranteed to be deterministic, so
1551any uses which require that should instead use a std::map.
1552
1553.. _dss_indexmap:
1554
1555llvm/ADT/IndexedMap.h
1556^^^^^^^^^^^^^^^^^^^^^
1557
1558IndexedMap is a specialized container for mapping small dense integers (or
1559values that can be mapped to small dense integers) to some other type.  It is
1560internally implemented as a vector with a mapping function that maps the keys
1561to the dense integer range.
1562
1563This is useful for cases like virtual registers in the LLVM code generator: they
1564have a dense mapping that is offset by a compile-time constant (the first
1565virtual register ID).
1566
1567.. _dss_densemap:
1568
1569llvm/ADT/DenseMap.h
1570^^^^^^^^^^^^^^^^^^^
1571
1572DenseMap is a simple quadratically probed hash table.  It excels at supporting
1573small keys and values: it uses a single allocation to hold all of the pairs
1574that are currently inserted in the map.  DenseMap is a great way to map
1575pointers to pointers, or map other small types to each other.
1576
1577There are several aspects of DenseMap that you should be aware of, however.
1578The iterators in a DenseMap are invalidated whenever an insertion occurs,
1579unlike map.  Also, because DenseMap allocates space for a large number of
1580key/value pairs (it starts with 64 by default), it will waste a lot of space if
1581your keys or values are large.  Finally, you must implement a partial
1582specialization of DenseMapInfo for the key that you want, if it isn't already
1583supported.  This is required to tell DenseMap about two special marker values
1584(which can never be inserted into the map) that it needs internally.
1585
1586DenseMap's find_as() method supports lookup operations using an alternate key
1587type.  This is useful in cases where the normal key type is expensive to
1588construct, but cheap to compare against.  The DenseMapInfo is responsible for
1589defining the appropriate comparison and hashing methods for each alternate key
1590type used.
1591
1592.. _dss_valuemap:
1593
1594llvm/IR/ValueMap.h
1595^^^^^^^^^^^^^^^^^^^
1596
1597ValueMap is a wrapper around a :ref:`DenseMap <dss_densemap>` mapping
1598``Value*``\ s (or subclasses) to another type.  When a Value is deleted or
1599RAUW'ed, ValueMap will update itself so the new version of the key is mapped to
1600the same value, just as if the key were a WeakVH.  You can configure exactly how
1601this happens, and what else happens on these two events, by passing a ``Config``
1602parameter to the ValueMap template.
1603
1604.. _dss_intervalmap:
1605
1606llvm/ADT/IntervalMap.h
1607^^^^^^^^^^^^^^^^^^^^^^
1608
1609IntervalMap is a compact map for small keys and values.  It maps key intervals
1610instead of single keys, and it will automatically coalesce adjacent intervals.
1611When the map only contains a few intervals, they are stored in the map object
1612itself to avoid allocations.
1613
1614The IntervalMap iterators are quite big, so they should not be passed around as
1615STL iterators.  The heavyweight iterators allow a smaller data structure.
1616
1617.. _dss_map:
1618
1619<map>
1620^^^^^
1621
1622std::map has similar characteristics to :ref:`std::set <dss_set>`: it uses a
1623single allocation per pair inserted into the map, it offers log(n) lookup with
1624an extremely large constant factor, imposes a space penalty of 3 pointers per
1625pair in the map, etc.
1626
1627std::map is most useful when your keys or values are very large, if you need to
1628iterate over the collection in sorted order, or if you need stable iterators
1629into the map (i.e. they don't get invalidated if an insertion or deletion of
1630another element takes place).
1631
1632.. _dss_mapvector:
1633
1634llvm/ADT/MapVector.h
1635^^^^^^^^^^^^^^^^^^^^
1636
1637``MapVector<KeyT,ValueT>`` provides a subset of the DenseMap interface.  The
1638main difference is that the iteration order is guaranteed to be the insertion
1639order, making it an easy (but somewhat expensive) solution for non-deterministic
1640iteration over maps of pointers.
1641
1642It is implemented by mapping from key to an index in a vector of key,value
1643pairs.  This provides fast lookup and iteration, but has two main drawbacks:
1644the key is stored twice and removing elements takes linear time.  If it is
1645necessary to remove elements, it's best to remove them in bulk using
1646``remove_if()``.
1647
1648.. _dss_inteqclasses:
1649
1650llvm/ADT/IntEqClasses.h
1651^^^^^^^^^^^^^^^^^^^^^^^
1652
1653IntEqClasses provides a compact representation of equivalence classes of small
1654integers.  Initially, each integer in the range 0..n-1 has its own equivalence
1655class.  Classes can be joined by passing two class representatives to the
1656join(a, b) method.  Two integers are in the same class when findLeader() returns
1657the same representative.
1658
1659Once all equivalence classes are formed, the map can be compressed so each
1660integer 0..n-1 maps to an equivalence class number in the range 0..m-1, where m
1661is the total number of equivalence classes.  The map must be uncompressed before
1662it can be edited again.
1663
1664.. _dss_immutablemap:
1665
1666llvm/ADT/ImmutableMap.h
1667^^^^^^^^^^^^^^^^^^^^^^^
1668
1669ImmutableMap is an immutable (functional) map implementation based on an AVL
1670tree.  Adding or removing elements is done through a Factory object and results
1671in the creation of a new ImmutableMap object.  If an ImmutableMap already exists
1672with the given key set, then the existing one is returned; equality is compared
1673with a FoldingSetNodeID.  The time and space complexity of add or remove
1674operations is logarithmic in the size of the original map.
1675
1676.. _dss_othermap:
1677
1678Other Map-Like Container Options
1679^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
1680
1681The STL provides several other options, such as std::multimap and the various
1682"hash_map" like containers (whether from C++ TR1 or from the SGI library).  We
1683never use hash_set and unordered_set because they are generally very expensive
1684(each insertion requires a malloc) and very non-portable.
1685
1686std::multimap is useful if you want to map a key to multiple values, but has all
1687the drawbacks of std::map.  A sorted vector or some other approach is almost
1688always better.
1689
1690.. _ds_bit:
1691
1692Bit storage containers (BitVector, SparseBitVector)
1693---------------------------------------------------
1694
1695Unlike the other containers, there are only two bit storage containers, and
1696choosing when to use each is relatively straightforward.
1697
1698One additional option is ``std::vector<bool>``: we discourage its use for two
1699reasons 1) the implementation in many common compilers (e.g.  commonly
1700available versions of GCC) is extremely inefficient and 2) the C++ standards
1701committee is likely to deprecate this container and/or change it significantly
1702somehow.  In any case, please don't use it.
1703
1704.. _dss_bitvector:
1705
1706BitVector
1707^^^^^^^^^
1708
1709The BitVector container provides a dynamic size set of bits for manipulation.
1710It supports individual bit setting/testing, as well as set operations.  The set
1711operations take time O(size of bitvector), but operations are performed one word
1712at a time, instead of one bit at a time.  This makes the BitVector very fast for
1713set operations compared to other containers.  Use the BitVector when you expect
1714the number of set bits to be high (i.e. a dense set).
1715
1716.. _dss_smallbitvector:
1717
1718SmallBitVector
1719^^^^^^^^^^^^^^
1720
1721The SmallBitVector container provides the same interface as BitVector, but it is
1722optimized for the case where only a small number of bits, less than 25 or so,
1723are needed.  It also transparently supports larger bit counts, but slightly less
1724efficiently than a plain BitVector, so SmallBitVector should only be used when
1725larger counts are rare.
1726
1727At this time, SmallBitVector does not support set operations (and, or, xor), and
1728its operator[] does not provide an assignable lvalue.
1729
1730.. _dss_sparsebitvector:
1731
1732SparseBitVector
1733^^^^^^^^^^^^^^^
1734
1735The SparseBitVector container is much like BitVector, with one major difference:
1736Only the bits that are set, are stored.  This makes the SparseBitVector much
1737more space efficient than BitVector when the set is sparse, as well as making
1738set operations O(number of set bits) instead of O(size of universe).  The
1739downside to the SparseBitVector is that setting and testing of random bits is
1740O(N), and on large SparseBitVectors, this can be slower than BitVector.  In our
1741implementation, setting or testing bits in sorted order (either forwards or
1742reverse) is O(1) worst case.  Testing and setting bits within 128 bits (depends
1743on size) of the current bit is also O(1).  As a general statement,
1744testing/setting bits in a SparseBitVector is O(distance away from last set bit).
1745
1746.. _common:
1747
1748Helpful Hints for Common Operations
1749===================================
1750
1751This section describes how to perform some very simple transformations of LLVM
1752code.  This is meant to give examples of common idioms used, showing the
1753practical side of LLVM transformations.
1754
1755Because this is a "how-to" section, you should also read about the main classes
1756that you will be working with.  The :ref:`Core LLVM Class Hierarchy Reference
1757<coreclasses>` contains details and descriptions of the main classes that you
1758should know about.
1759
1760.. _inspection:
1761
1762Basic Inspection and Traversal Routines
1763---------------------------------------
1764
1765The LLVM compiler infrastructure have many different data structures that may be
1766traversed.  Following the example of the C++ standard template library, the
1767techniques used to traverse these various data structures are all basically the
1768same.  For a enumerable sequence of values, the ``XXXbegin()`` function (or
1769method) returns an iterator to the start of the sequence, the ``XXXend()``
1770function returns an iterator pointing to one past the last valid element of the
1771sequence, and there is some ``XXXiterator`` data type that is common between the
1772two operations.
1773
1774Because the pattern for iteration is common across many different aspects of the
1775program representation, the standard template library algorithms may be used on
1776them, and it is easier to remember how to iterate.  First we show a few common
1777examples of the data structures that need to be traversed.  Other data
1778structures are traversed in very similar ways.
1779
1780.. _iterate_function:
1781
1782Iterating over the ``BasicBlock`` in a ``Function``
1783^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
1784
1785It's quite common to have a ``Function`` instance that you'd like to transform
1786in some way; in particular, you'd like to manipulate its ``BasicBlock``\ s.  To
1787facilitate this, you'll need to iterate over all of the ``BasicBlock``\ s that
1788constitute the ``Function``.  The following is an example that prints the name
1789of a ``BasicBlock`` and the number of ``Instruction``\ s it contains:
1790
1791.. code-block:: c++
1792
1793  // func is a pointer to a Function instance
1794  for (Function::iterator i = func->begin(), e = func->end(); i != e; ++i)
1795    // Print out the name of the basic block if it has one, and then the
1796    // number of instructions that it contains
1797    errs() << "Basic block (name=" << i->getName() << ") has "
1798               << i->size() << " instructions.\n";
1799
1800Note that i can be used as if it were a pointer for the purposes of invoking
1801member functions of the ``Instruction`` class.  This is because the indirection
1802operator is overloaded for the iterator classes.  In the above code, the
1803expression ``i->size()`` is exactly equivalent to ``(*i).size()`` just like
1804you'd expect.
1805
1806.. _iterate_basicblock:
1807
1808Iterating over the ``Instruction`` in a ``BasicBlock``
1809^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
1810
1811Just like when dealing with ``BasicBlock``\ s in ``Function``\ s, it's easy to
1812iterate over the individual instructions that make up ``BasicBlock``\ s.  Here's
1813a code snippet that prints out each instruction in a ``BasicBlock``:
1814
1815.. code-block:: c++
1816
1817  // blk is a pointer to a BasicBlock instance
1818  for (BasicBlock::iterator i = blk->begin(), e = blk->end(); i != e; ++i)
1819     // The next statement works since operator<<(ostream&,...)
1820     // is overloaded for Instruction&
1821     errs() << *i << "\n";
1822
1823
1824However, this isn't really the best way to print out the contents of a
1825``BasicBlock``!  Since the ostream operators are overloaded for virtually
1826anything you'll care about, you could have just invoked the print routine on the
1827basic block itself: ``errs() << *blk << "\n";``.
1828
1829.. _iterate_insiter:
1830
1831Iterating over the ``Instruction`` in a ``Function``
1832^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
1833
1834If you're finding that you commonly iterate over a ``Function``'s
1835``BasicBlock``\ s and then that ``BasicBlock``'s ``Instruction``\ s,
1836``InstIterator`` should be used instead.  You'll need to include
1837``llvm/IR/InstIterator.h`` (`doxygen
1838<http://llvm.org/doxygen/InstIterator_8h.html>`__) and then instantiate
1839``InstIterator``\ s explicitly in your code.  Here's a small example that shows
1840how to dump all instructions in a function to the standard error stream:
1841
1842.. code-block:: c++
1843
1844  #include "llvm/IR/InstIterator.h"
1845
1846  // F is a pointer to a Function instance
1847  for (inst_iterator I = inst_begin(F), E = inst_end(F); I != E; ++I)
1848    errs() << *I << "\n";
1849
1850Easy, isn't it?  You can also use ``InstIterator``\ s to fill a work list with
1851its initial contents.  For example, if you wanted to initialize a work list to
1852contain all instructions in a ``Function`` F, all you would need to do is
1853something like:
1854
1855.. code-block:: c++
1856
1857  std::set<Instruction*> worklist;
1858  // or better yet, SmallPtrSet<Instruction*, 64> worklist;
1859
1860  for (inst_iterator I = inst_begin(F), E = inst_end(F); I != E; ++I)
1861    worklist.insert(&*I);
1862
1863The STL set ``worklist`` would now contain all instructions in the ``Function``
1864pointed to by F.
1865
1866.. _iterate_convert:
1867
1868Turning an iterator into a class pointer (and vice-versa)
1869^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
1870
1871Sometimes, it'll be useful to grab a reference (or pointer) to a class instance
1872when all you've got at hand is an iterator.  Well, extracting a reference or a
1873pointer from an iterator is very straight-forward.  Assuming that ``i`` is a
1874``BasicBlock::iterator`` and ``j`` is a ``BasicBlock::const_iterator``:
1875
1876.. code-block:: c++
1877
1878  Instruction& inst = *i;   // Grab reference to instruction reference
1879  Instruction* pinst = &*i; // Grab pointer to instruction reference
1880  const Instruction& inst = *j;
1881
1882However, the iterators you'll be working with in the LLVM framework are special:
1883they will automatically convert to a ptr-to-instance type whenever they need to.
1884Instead of dereferencing the iterator and then taking the address of the result,
1885you can simply assign the iterator to the proper pointer type and you get the
1886dereference and address-of operation as a result of the assignment (behind the
1887scenes, this is a result of overloading casting mechanisms).  Thus the second
1888line of the last example,
1889
1890.. code-block:: c++
1891
1892  Instruction *pinst = &*i;
1893
1894is semantically equivalent to
1895
1896.. code-block:: c++
1897
1898  Instruction *pinst = i;
1899
1900It's also possible to turn a class pointer into the corresponding iterator, and
1901this is a constant time operation (very efficient).  The following code snippet
1902illustrates use of the conversion constructors provided by LLVM iterators.  By
1903using these, you can explicitly grab the iterator of something without actually
1904obtaining it via iteration over some structure:
1905
1906.. code-block:: c++
1907
1908  void printNextInstruction(Instruction* inst) {
1909    BasicBlock::iterator it(inst);
1910    ++it; // After this line, it refers to the instruction after *inst
1911    if (it != inst->getParent()->end()) errs() << *it << "\n";
1912  }
1913
1914Unfortunately, these implicit conversions come at a cost; they prevent these
1915iterators from conforming to standard iterator conventions, and thus from being
1916usable with standard algorithms and containers.  For example, they prevent the
1917following code, where ``B`` is a ``BasicBlock``, from compiling:
1918
1919.. code-block:: c++
1920
1921  llvm::SmallVector<llvm::Instruction *, 16>(B->begin(), B->end());
1922
1923Because of this, these implicit conversions may be removed some day, and
1924``operator*`` changed to return a pointer instead of a reference.
1925
1926.. _iterate_complex:
1927
1928Finding call sites: a slightly more complex example
1929^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
1930
1931Say that you're writing a FunctionPass and would like to count all the locations
1932in the entire module (that is, across every ``Function``) where a certain
1933function (i.e., some ``Function *``) is already in scope.  As you'll learn
1934later, you may want to use an ``InstVisitor`` to accomplish this in a much more
1935straight-forward manner, but this example will allow us to explore how you'd do
1936it if you didn't have ``InstVisitor`` around.  In pseudo-code, this is what we
1937want to do:
1938
1939.. code-block:: none
1940
1941  initialize callCounter to zero
1942  for each Function f in the Module
1943    for each BasicBlock b in f
1944      for each Instruction i in b
1945        if (i is a CallInst and calls the given function)
1946          increment callCounter
1947
1948And the actual code is (remember, because we're writing a ``FunctionPass``, our
1949``FunctionPass``-derived class simply has to override the ``runOnFunction``
1950method):
1951
1952.. code-block:: c++
1953
1954  Function* targetFunc = ...;
1955
1956  class OurFunctionPass : public FunctionPass {
1957    public:
1958      OurFunctionPass(): callCounter(0) { }
1959
1960      virtual runOnFunction(Function& F) {
1961        for (Function::iterator b = F.begin(), be = F.end(); b != be; ++b) {
1962          for (BasicBlock::iterator i = b->begin(), ie = b->end(); i != ie; ++i) {
1963            if (CallInst* callInst = dyn_cast<CallInst>(&*i)) {
1964              // We know we've encountered a call instruction, so we
1965              // need to determine if it's a call to the
1966              // function pointed to by m_func or not.
1967              if (callInst->getCalledFunction() == targetFunc)
1968                ++callCounter;
1969            }
1970          }
1971        }
1972      }
1973
1974    private:
1975      unsigned callCounter;
1976  };
1977
1978.. _calls_and_invokes:
1979
1980Treating calls and invokes the same way
1981^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
1982
1983You may have noticed that the previous example was a bit oversimplified in that
1984it did not deal with call sites generated by 'invoke' instructions.  In this,
1985and in other situations, you may find that you want to treat ``CallInst``\ s and
1986``InvokeInst``\ s the same way, even though their most-specific common base
1987class is ``Instruction``, which includes lots of less closely-related things.
1988For these cases, LLVM provides a handy wrapper class called ``CallSite``
1989(`doxygen <http://llvm.org/doxygen/classllvm_1_1CallSite.html>`__) It is
1990essentially a wrapper around an ``Instruction`` pointer, with some methods that
1991provide functionality common to ``CallInst``\ s and ``InvokeInst``\ s.
1992
1993This class has "value semantics": it should be passed by value, not by reference
1994and it should not be dynamically allocated or deallocated using ``operator new``
1995or ``operator delete``.  It is efficiently copyable, assignable and
1996constructable, with costs equivalents to that of a bare pointer.  If you look at
1997its definition, it has only a single pointer member.
1998
1999.. _iterate_chains:
2000
2001Iterating over def-use & use-def chains
2002^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
2003
2004Frequently, we might have an instance of the ``Value`` class (`doxygen
2005<http://llvm.org/doxygen/classllvm_1_1Value.html>`__) and we want to determine
2006which ``User`` s use the ``Value``.  The list of all ``User``\ s of a particular
2007``Value`` is called a *def-use* chain.  For example, let's say we have a
2008``Function*`` named ``F`` to a particular function ``foo``.  Finding all of the
2009instructions that *use* ``foo`` is as simple as iterating over the *def-use*
2010chain of ``F``:
2011
2012.. code-block:: c++
2013
2014  Function *F = ...;
2015
2016  for (User *U : F->users()) {
2017    if (Instruction *Inst = dyn_cast<Instruction>(U)) {
2018      errs() << "F is used in instruction:\n";
2019      errs() << *Inst << "\n";
2020    }
2021
2022Alternatively, it's common to have an instance of the ``User`` Class (`doxygen
2023<http://llvm.org/doxygen/classllvm_1_1User.html>`__) and need to know what
2024``Value``\ s are used by it.  The list of all ``Value``\ s used by a ``User`` is
2025known as a *use-def* chain.  Instances of class ``Instruction`` are common
2026``User`` s, so we might want to iterate over all of the values that a particular
2027instruction uses (that is, the operands of the particular ``Instruction``):
2028
2029.. code-block:: c++
2030
2031  Instruction *pi = ...;
2032
2033  for (Use &U : pi->operands()) {
2034    Value *v = U.get();
2035    // ...
2036  }
2037
2038Declaring objects as ``const`` is an important tool of enforcing mutation free
2039algorithms (such as analyses, etc.).  For this purpose above iterators come in
2040constant flavors as ``Value::const_use_iterator`` and
2041``Value::const_op_iterator``.  They automatically arise when calling
2042``use/op_begin()`` on ``const Value*``\ s or ``const User*``\ s respectively.
2043Upon dereferencing, they return ``const Use*``\ s.  Otherwise the above patterns
2044remain unchanged.
2045
2046.. _iterate_preds:
2047
2048Iterating over predecessors & successors of blocks
2049^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
2050
2051Iterating over the predecessors and successors of a block is quite easy with the
2052routines defined in ``"llvm/IR/CFG.h"``.  Just use code like this to
2053iterate over all predecessors of BB:
2054
2055.. code-block:: c++
2056
2057  #include "llvm/Support/CFG.h"
2058  BasicBlock *BB = ...;
2059
2060  for (pred_iterator PI = pred_begin(BB), E = pred_end(BB); PI != E; ++PI) {
2061    BasicBlock *Pred = *PI;
2062    // ...
2063  }
2064
2065Similarly, to iterate over successors use ``succ_iterator/succ_begin/succ_end``.
2066
2067.. _simplechanges:
2068
2069Making simple changes
2070---------------------
2071
2072There are some primitive transformation operations present in the LLVM
2073infrastructure that are worth knowing about.  When performing transformations,
2074it's fairly common to manipulate the contents of basic blocks.  This section
2075describes some of the common methods for doing so and gives example code.
2076
2077.. _schanges_creating:
2078
2079Creating and inserting new ``Instruction``\ s
2080^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
2081
2082*Instantiating Instructions*
2083
2084Creation of ``Instruction``\ s is straight-forward: simply call the constructor
2085for the kind of instruction to instantiate and provide the necessary parameters.
2086For example, an ``AllocaInst`` only *requires* a (const-ptr-to) ``Type``.  Thus:
2087
2088.. code-block:: c++
2089
2090  AllocaInst* ai = new AllocaInst(Type::Int32Ty);
2091
2092will create an ``AllocaInst`` instance that represents the allocation of one
2093integer in the current stack frame, at run time.  Each ``Instruction`` subclass
2094is likely to have varying default parameters which change the semantics of the
2095instruction, so refer to the `doxygen documentation for the subclass of
2096Instruction <http://llvm.org/doxygen/classllvm_1_1Instruction.html>`_ that
2097you're interested in instantiating.
2098
2099*Naming values*
2100
2101It is very useful to name the values of instructions when you're able to, as
2102this facilitates the debugging of your transformations.  If you end up looking
2103at generated LLVM machine code, you definitely want to have logical names
2104associated with the results of instructions!  By supplying a value for the
2105``Name`` (default) parameter of the ``Instruction`` constructor, you associate a
2106logical name with the result of the instruction's execution at run time.  For
2107example, say that I'm writing a transformation that dynamically allocates space
2108for an integer on the stack, and that integer is going to be used as some kind
2109of index by some other code.  To accomplish this, I place an ``AllocaInst`` at
2110the first point in the first ``BasicBlock`` of some ``Function``, and I'm
2111intending to use it within the same ``Function``.  I might do:
2112
2113.. code-block:: c++
2114
2115  AllocaInst* pa = new AllocaInst(Type::Int32Ty, 0, "indexLoc");
2116
2117where ``indexLoc`` is now the logical name of the instruction's execution value,
2118which is a pointer to an integer on the run time stack.
2119
2120*Inserting instructions*
2121
2122There are essentially three ways to insert an ``Instruction`` into an existing
2123sequence of instructions that form a ``BasicBlock``:
2124
2125* Insertion into an explicit instruction list
2126
2127  Given a ``BasicBlock* pb``, an ``Instruction* pi`` within that ``BasicBlock``,
2128  and a newly-created instruction we wish to insert before ``*pi``, we do the
2129  following:
2130
2131  .. code-block:: c++
2132
2133      BasicBlock *pb = ...;
2134      Instruction *pi = ...;
2135      Instruction *newInst = new Instruction(...);
2136
2137      pb->getInstList().insert(pi, newInst); // Inserts newInst before pi in pb
2138
2139  Appending to the end of a ``BasicBlock`` is so common that the ``Instruction``
2140  class and ``Instruction``-derived classes provide constructors which take a
2141  pointer to a ``BasicBlock`` to be appended to.  For example code that looked
2142  like:
2143
2144  .. code-block:: c++
2145
2146    BasicBlock *pb = ...;
2147    Instruction *newInst = new Instruction(...);
2148
2149    pb->getInstList().push_back(newInst); // Appends newInst to pb
2150
2151  becomes:
2152
2153  .. code-block:: c++
2154
2155    BasicBlock *pb = ...;
2156    Instruction *newInst = new Instruction(..., pb);
2157
2158  which is much cleaner, especially if you are creating long instruction
2159  streams.
2160
2161* Insertion into an implicit instruction list
2162
2163  ``Instruction`` instances that are already in ``BasicBlock``\ s are implicitly
2164  associated with an existing instruction list: the instruction list of the
2165  enclosing basic block.  Thus, we could have accomplished the same thing as the
2166  above code without being given a ``BasicBlock`` by doing:
2167
2168  .. code-block:: c++
2169
2170    Instruction *pi = ...;
2171    Instruction *newInst = new Instruction(...);
2172
2173    pi->getParent()->getInstList().insert(pi, newInst);
2174
2175  In fact, this sequence of steps occurs so frequently that the ``Instruction``
2176  class and ``Instruction``-derived classes provide constructors which take (as
2177  a default parameter) a pointer to an ``Instruction`` which the newly-created
2178  ``Instruction`` should precede.  That is, ``Instruction`` constructors are
2179  capable of inserting the newly-created instance into the ``BasicBlock`` of a
2180  provided instruction, immediately before that instruction.  Using an
2181  ``Instruction`` constructor with a ``insertBefore`` (default) parameter, the
2182  above code becomes:
2183
2184  .. code-block:: c++
2185
2186    Instruction* pi = ...;
2187    Instruction* newInst = new Instruction(..., pi);
2188
2189  which is much cleaner, especially if you're creating a lot of instructions and
2190  adding them to ``BasicBlock``\ s.
2191
2192* Insertion using an instance of ``IRBuilder``
2193
2194  Inserting several ``Instruction``\ s can be quite laborious using the previous
2195  methods. The ``IRBuilder`` is a convenience class that can be used to add
2196  several instructions to the end of a ``BasicBlock`` or before a particular
2197  ``Instruction``. It also supports constant folding and renaming named
2198  registers (see ``IRBuilder``'s template arguments).
2199
2200  The example below demonstrates a very simple use of the ``IRBuilder`` where
2201  three instructions are inserted before the instruction ``pi``. The first two
2202  instructions are Call instructions and third instruction multiplies the return
2203  value of the two calls.
2204
2205  .. code-block:: c++
2206
2207    Instruction *pi = ...;
2208    IRBuilder<> Builder(pi);
2209    CallInst* callOne = Builder.CreateCall(...);
2210    CallInst* callTwo = Builder.CreateCall(...);
2211    Value* result = Builder.CreateMul(callOne, callTwo);
2212
2213  The example below is similar to the above example except that the created
2214  ``IRBuilder`` inserts instructions at the end of the ``BasicBlock`` ``pb``.
2215
2216  .. code-block:: c++
2217
2218    BasicBlock *pb = ...;
2219    IRBuilder<> Builder(pb);
2220    CallInst* callOne = Builder.CreateCall(...);
2221    CallInst* callTwo = Builder.CreateCall(...);
2222    Value* result = Builder.CreateMul(callOne, callTwo);
2223
2224  See :doc:`tutorial/LangImpl03` for a practical use of the ``IRBuilder``.
2225
2226
2227.. _schanges_deleting:
2228
2229Deleting Instructions
2230^^^^^^^^^^^^^^^^^^^^^
2231
2232Deleting an instruction from an existing sequence of instructions that form a
2233BasicBlock_ is very straight-forward: just call the instruction's
2234``eraseFromParent()`` method.  For example:
2235
2236.. code-block:: c++
2237
2238  Instruction *I = .. ;
2239  I->eraseFromParent();
2240
2241This unlinks the instruction from its containing basic block and deletes it.  If
2242you'd just like to unlink the instruction from its containing basic block but
2243not delete it, you can use the ``removeFromParent()`` method.
2244
2245.. _schanges_replacing:
2246
2247Replacing an Instruction with another Value
2248^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
2249
2250Replacing individual instructions
2251"""""""""""""""""""""""""""""""""
2252
2253Including "`llvm/Transforms/Utils/BasicBlockUtils.h
2254<http://llvm.org/doxygen/BasicBlockUtils_8h-source.html>`_" permits use of two
2255very useful replace functions: ``ReplaceInstWithValue`` and
2256``ReplaceInstWithInst``.
2257
2258.. _schanges_deleting_sub:
2259
2260Deleting Instructions
2261"""""""""""""""""""""
2262
2263* ``ReplaceInstWithValue``
2264
2265  This function replaces all uses of a given instruction with a value, and then
2266  removes the original instruction.  The following example illustrates the
2267  replacement of the result of a particular ``AllocaInst`` that allocates memory
2268  for a single integer with a null pointer to an integer.
2269
2270  .. code-block:: c++
2271
2272    AllocaInst* instToReplace = ...;
2273    BasicBlock::iterator ii(instToReplace);
2274
2275    ReplaceInstWithValue(instToReplace->getParent()->getInstList(), ii,
2276                         Constant::getNullValue(PointerType::getUnqual(Type::Int32Ty)));
2277
2278* ``ReplaceInstWithInst``
2279
2280  This function replaces a particular instruction with another instruction,
2281  inserting the new instruction into the basic block at the location where the
2282  old instruction was, and replacing any uses of the old instruction with the
2283  new instruction.  The following example illustrates the replacement of one
2284  ``AllocaInst`` with another.
2285
2286  .. code-block:: c++
2287
2288    AllocaInst* instToReplace = ...;
2289    BasicBlock::iterator ii(instToReplace);
2290
2291    ReplaceInstWithInst(instToReplace->getParent()->getInstList(), ii,
2292                        new AllocaInst(Type::Int32Ty, 0, "ptrToReplacedInt"));
2293
2294
2295Replacing multiple uses of Users and Values
2296"""""""""""""""""""""""""""""""""""""""""""
2297
2298You can use ``Value::replaceAllUsesWith`` and ``User::replaceUsesOfWith`` to
2299change more than one use at a time.  See the doxygen documentation for the
2300`Value Class <http://llvm.org/doxygen/classllvm_1_1Value.html>`_ and `User Class
2301<http://llvm.org/doxygen/classllvm_1_1User.html>`_, respectively, for more
2302information.
2303
2304.. _schanges_deletingGV:
2305
2306Deleting GlobalVariables
2307^^^^^^^^^^^^^^^^^^^^^^^^
2308
2309Deleting a global variable from a module is just as easy as deleting an
2310Instruction.  First, you must have a pointer to the global variable that you
2311wish to delete.  You use this pointer to erase it from its parent, the module.
2312For example:
2313
2314.. code-block:: c++
2315
2316  GlobalVariable *GV = .. ;
2317
2318  GV->eraseFromParent();
2319
2320
2321.. _create_types:
2322
2323How to Create Types
2324-------------------
2325
2326In generating IR, you may need some complex types.  If you know these types
2327statically, you can use ``TypeBuilder<...>::get()``, defined in
2328``llvm/Support/TypeBuilder.h``, to retrieve them.  ``TypeBuilder`` has two forms
2329depending on whether you're building types for cross-compilation or native
2330library use.  ``TypeBuilder<T, true>`` requires that ``T`` be independent of the
2331host environment, meaning that it's built out of types from the ``llvm::types``
2332(`doxygen <http://llvm.org/doxygen/namespacellvm_1_1types.html>`__) namespace
2333and pointers, functions, arrays, etc. built of those.  ``TypeBuilder<T, false>``
2334additionally allows native C types whose size may depend on the host compiler.
2335For example,
2336
2337.. code-block:: c++
2338
2339  FunctionType *ft = TypeBuilder<types::i<8>(types::i<32>*), true>::get();
2340
2341is easier to read and write than the equivalent
2342
2343.. code-block:: c++
2344
2345  std::vector<const Type*> params;
2346  params.push_back(PointerType::getUnqual(Type::Int32Ty));
2347  FunctionType *ft = FunctionType::get(Type::Int8Ty, params, false);
2348
2349See the `class comment
2350<http://llvm.org/doxygen/TypeBuilder_8h-source.html#l00001>`_ for more details.
2351
2352.. _threading:
2353
2354Threads and LLVM
2355================
2356
2357This section describes the interaction of the LLVM APIs with multithreading,
2358both on the part of client applications, and in the JIT, in the hosted
2359application.
2360
2361Note that LLVM's support for multithreading is still relatively young.  Up
2362through version 2.5, the execution of threaded hosted applications was
2363supported, but not threaded client access to the APIs.  While this use case is
2364now supported, clients *must* adhere to the guidelines specified below to ensure
2365proper operation in multithreaded mode.
2366
2367Note that, on Unix-like platforms, LLVM requires the presence of GCC's atomic
2368intrinsics in order to support threaded operation.  If you need a
2369multhreading-capable LLVM on a platform without a suitably modern system
2370compiler, consider compiling LLVM and LLVM-GCC in single-threaded mode, and
2371using the resultant compiler to build a copy of LLVM with multithreading
2372support.
2373
2374.. _shutdown:
2375
2376Ending Execution with ``llvm_shutdown()``
2377-----------------------------------------
2378
2379When you are done using the LLVM APIs, you should call ``llvm_shutdown()`` to
2380deallocate memory used for internal structures.
2381
2382.. _managedstatic:
2383
2384Lazy Initialization with ``ManagedStatic``
2385------------------------------------------
2386
2387``ManagedStatic`` is a utility class in LLVM used to implement static
2388initialization of static resources, such as the global type tables.  In a
2389single-threaded environment, it implements a simple lazy initialization scheme.
2390When LLVM is compiled with support for multi-threading, however, it uses
2391double-checked locking to implement thread-safe lazy initialization.
2392
2393.. _llvmcontext:
2394
2395Achieving Isolation with ``LLVMContext``
2396----------------------------------------
2397
2398``LLVMContext`` is an opaque class in the LLVM API which clients can use to
2399operate multiple, isolated instances of LLVM concurrently within the same
2400address space.  For instance, in a hypothetical compile-server, the compilation
2401of an individual translation unit is conceptually independent from all the
2402others, and it would be desirable to be able to compile incoming translation
2403units concurrently on independent server threads.  Fortunately, ``LLVMContext``
2404exists to enable just this kind of scenario!
2405
2406Conceptually, ``LLVMContext`` provides isolation.  Every LLVM entity
2407(``Module``\ s, ``Value``\ s, ``Type``\ s, ``Constant``\ s, etc.) in LLVM's
2408in-memory IR belongs to an ``LLVMContext``.  Entities in different contexts
2409*cannot* interact with each other: ``Module``\ s in different contexts cannot be
2410linked together, ``Function``\ s cannot be added to ``Module``\ s in different
2411contexts, etc.  What this means is that is is safe to compile on multiple
2412threads simultaneously, as long as no two threads operate on entities within the
2413same context.
2414
2415In practice, very few places in the API require the explicit specification of a
2416``LLVMContext``, other than the ``Type`` creation/lookup APIs.  Because every
2417``Type`` carries a reference to its owning context, most other entities can
2418determine what context they belong to by looking at their own ``Type``.  If you
2419are adding new entities to LLVM IR, please try to maintain this interface
2420design.
2421
2422.. _jitthreading:
2423
2424Threads and the JIT
2425-------------------
2426
2427LLVM's "eager" JIT compiler is safe to use in threaded programs.  Multiple
2428threads can call ``ExecutionEngine::getPointerToFunction()`` or
2429``ExecutionEngine::runFunction()`` concurrently, and multiple threads can run
2430code output by the JIT concurrently.  The user must still ensure that only one
2431thread accesses IR in a given ``LLVMContext`` while another thread might be
2432modifying it.  One way to do that is to always hold the JIT lock while accessing
2433IR outside the JIT (the JIT *modifies* the IR by adding ``CallbackVH``\ s).
2434Another way is to only call ``getPointerToFunction()`` from the
2435``LLVMContext``'s thread.
2436
2437When the JIT is configured to compile lazily (using
2438``ExecutionEngine::DisableLazyCompilation(false)``), there is currently a `race
2439condition <http://llvm.org/bugs/show_bug.cgi?id=5184>`_ in updating call sites
2440after a function is lazily-jitted.  It's still possible to use the lazy JIT in a
2441threaded program if you ensure that only one thread at a time can call any
2442particular lazy stub and that the JIT lock guards any IR access, but we suggest
2443using only the eager JIT in threaded programs.
2444
2445.. _advanced:
2446
2447Advanced Topics
2448===============
2449
2450This section describes some of the advanced or obscure API's that most clients
2451do not need to be aware of.  These API's tend manage the inner workings of the
2452LLVM system, and only need to be accessed in unusual circumstances.
2453
2454.. _SymbolTable:
2455
2456The ``ValueSymbolTable`` class
2457------------------------------
2458
2459The ``ValueSymbolTable`` (`doxygen
2460<http://llvm.org/doxygen/classllvm_1_1ValueSymbolTable.html>`__) class provides
2461a symbol table that the :ref:`Function <c_Function>` and Module_ classes use for
2462naming value definitions.  The symbol table can provide a name for any Value_.
2463
2464Note that the ``SymbolTable`` class should not be directly accessed by most
2465clients.  It should only be used when iteration over the symbol table names
2466themselves are required, which is very special purpose.  Note that not all LLVM
2467Value_\ s have names, and those without names (i.e. they have an empty name) do
2468not exist in the symbol table.
2469
2470Symbol tables support iteration over the values in the symbol table with
2471``begin/end/iterator`` and supports querying to see if a specific name is in the
2472symbol table (with ``lookup``).  The ``ValueSymbolTable`` class exposes no
2473public mutator methods, instead, simply call ``setName`` on a value, which will
2474autoinsert it into the appropriate symbol table.
2475
2476.. _UserLayout:
2477
2478The ``User`` and owned ``Use`` classes' memory layout
2479-----------------------------------------------------
2480
2481The ``User`` (`doxygen <http://llvm.org/doxygen/classllvm_1_1User.html>`__)
2482class provides a basis for expressing the ownership of ``User`` towards other
2483`Value instance <http://llvm.org/doxygen/classllvm_1_1Value.html>`_\ s.  The
2484``Use`` (`doxygen <http://llvm.org/doxygen/classllvm_1_1Use.html>`__) helper
2485class is employed to do the bookkeeping and to facilitate *O(1)* addition and
2486removal.
2487
2488.. _Use2User:
2489
2490Interaction and relationship between ``User`` and ``Use`` objects
2491^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
2492
2493A subclass of ``User`` can choose between incorporating its ``Use`` objects or
2494refer to them out-of-line by means of a pointer.  A mixed variant (some ``Use``
2495s inline others hung off) is impractical and breaks the invariant that the
2496``Use`` objects belonging to the same ``User`` form a contiguous array.
2497
2498We have 2 different layouts in the ``User`` (sub)classes:
2499
2500* Layout a)
2501
2502  The ``Use`` object(s) are inside (resp. at fixed offset) of the ``User``
2503  object and there are a fixed number of them.
2504
2505* Layout b)
2506
2507  The ``Use`` object(s) are referenced by a pointer to an array from the
2508  ``User`` object and there may be a variable number of them.
2509
2510As of v2.4 each layout still possesses a direct pointer to the start of the
2511array of ``Use``\ s.  Though not mandatory for layout a), we stick to this
2512redundancy for the sake of simplicity.  The ``User`` object also stores the
2513number of ``Use`` objects it has. (Theoretically this information can also be
2514calculated given the scheme presented below.)
2515
2516Special forms of allocation operators (``operator new``) enforce the following
2517memory layouts:
2518
2519* Layout a) is modelled by prepending the ``User`` object by the ``Use[]``
2520  array.
2521
2522  .. code-block:: none
2523
2524    ...---.---.---.---.-------...
2525      | P | P | P | P | User
2526    '''---'---'---'---'-------'''
2527
2528* Layout b) is modelled by pointing at the ``Use[]`` array.
2529
2530  .. code-block:: none
2531
2532    .-------...
2533    | User
2534    '-------'''
2535        |
2536        v
2537        .---.---.---.---...
2538        | P | P | P | P |
2539        '---'---'---'---'''
2540
2541*(In the above figures* '``P``' *stands for the* ``Use**`` *that is stored in
2542each* ``Use`` *object in the member* ``Use::Prev`` *)*
2543
2544.. _Waymarking:
2545
2546The waymarking algorithm
2547^^^^^^^^^^^^^^^^^^^^^^^^
2548
2549Since the ``Use`` objects are deprived of the direct (back)pointer to their
2550``User`` objects, there must be a fast and exact method to recover it.  This is
2551accomplished by the following scheme:
2552
2553A bit-encoding in the 2 LSBits (least significant bits) of the ``Use::Prev``
2554allows to find the start of the ``User`` object:
2555
2556* ``00`` --- binary digit 0
2557
2558* ``01`` --- binary digit 1
2559
2560* ``10`` --- stop and calculate (``s``)
2561
2562* ``11`` --- full stop (``S``)
2563
2564Given a ``Use*``, all we have to do is to walk till we get a stop and we either
2565have a ``User`` immediately behind or we have to walk to the next stop picking
2566up digits and calculating the offset:
2567
2568.. code-block:: none
2569
2570  .---.---.---.---.---.---.---.---.---.---.---.---.---.---.---.---.----------------
2571  | 1 | s | 1 | 0 | 1 | 0 | s | 1 | 1 | 0 | s | 1 | 1 | s | 1 | S | User (or User*)
2572  '---'---'---'---'---'---'---'---'---'---'---'---'---'---'---'---'----------------
2573      |+15                |+10            |+6         |+3     |+1
2574      |                   |               |           |       | __>
2575      |                   |               |           | __________>
2576      |                   |               | ______________________>
2577      |                   | ______________________________________>
2578      | __________________________________________________________>
2579
2580Only the significant number of bits need to be stored between the stops, so that
2581the *worst case is 20 memory accesses* when there are 1000 ``Use`` objects
2582associated with a ``User``.
2583
2584.. _ReferenceImpl:
2585
2586Reference implementation
2587^^^^^^^^^^^^^^^^^^^^^^^^
2588
2589The following literate Haskell fragment demonstrates the concept:
2590
2591.. code-block:: haskell
2592
2593  > import Test.QuickCheck
2594  >
2595  > digits :: Int -> [Char] -> [Char]
2596  > digits 0 acc = '0' : acc
2597  > digits 1 acc = '1' : acc
2598  > digits n acc = digits (n `div` 2) $ digits (n `mod` 2) acc
2599  >
2600  > dist :: Int -> [Char] -> [Char]
2601  > dist 0 [] = ['S']
2602  > dist 0 acc = acc
2603  > dist 1 acc = let r = dist 0 acc in 's' : digits (length r) r
2604  > dist n acc = dist (n - 1) $ dist 1 acc
2605  >
2606  > takeLast n ss = reverse $ take n $ reverse ss
2607  >
2608  > test = takeLast 40 $ dist 20 []
2609  >
2610
2611Printing <test> gives: ``"1s100000s11010s10100s1111s1010s110s11s1S"``
2612
2613The reverse algorithm computes the length of the string just by examining a
2614certain prefix:
2615
2616.. code-block:: haskell
2617
2618  > pref :: [Char] -> Int
2619  > pref "S" = 1
2620  > pref ('s':'1':rest) = decode 2 1 rest
2621  > pref (_:rest) = 1 + pref rest
2622  >
2623  > decode walk acc ('0':rest) = decode (walk + 1) (acc * 2) rest
2624  > decode walk acc ('1':rest) = decode (walk + 1) (acc * 2 + 1) rest
2625  > decode walk acc _ = walk + acc
2626  >
2627
2628Now, as expected, printing <pref test> gives ``40``.
2629
2630We can *quickCheck* this with following property:
2631
2632.. code-block:: haskell
2633
2634  > testcase = dist 2000 []
2635  > testcaseLength = length testcase
2636  >
2637  > identityProp n = n > 0 && n <= testcaseLength ==> length arr == pref arr
2638  >     where arr = takeLast n testcase
2639  >
2640
2641As expected <quickCheck identityProp> gives:
2642
2643::
2644
2645  *Main> quickCheck identityProp
2646  OK, passed 100 tests.
2647
2648Let's be a bit more exhaustive:
2649
2650.. code-block:: haskell
2651
2652  >
2653  > deepCheck p = check (defaultConfig { configMaxTest = 500 }) p
2654  >
2655
2656And here is the result of <deepCheck identityProp>:
2657
2658::
2659
2660  *Main> deepCheck identityProp
2661  OK, passed 500 tests.
2662
2663.. _Tagging:
2664
2665Tagging considerations
2666^^^^^^^^^^^^^^^^^^^^^^
2667
2668To maintain the invariant that the 2 LSBits of each ``Use**`` in ``Use`` never
2669change after being set up, setters of ``Use::Prev`` must re-tag the new
2670``Use**`` on every modification.  Accordingly getters must strip the tag bits.
2671
2672For layout b) instead of the ``User`` we find a pointer (``User*`` with LSBit
2673set).  Following this pointer brings us to the ``User``.  A portable trick
2674ensures that the first bytes of ``User`` (if interpreted as a pointer) never has
2675the LSBit set. (Portability is relying on the fact that all known compilers
2676place the ``vptr`` in the first word of the instances.)
2677
2678.. _polymorphism:
2679
2680Designing Type Hiercharies and Polymorphic Interfaces
2681-----------------------------------------------------
2682
2683There are two different design patterns that tend to result in the use of
2684virtual dispatch for methods in a type hierarchy in C++ programs. The first is
2685a genuine type hierarchy where different types in the hierarchy model
2686a specific subset of the functionality and semantics, and these types nest
2687strictly within each other. Good examples of this can be seen in the ``Value``
2688or ``Type`` type hierarchies.
2689
2690A second is the desire to dispatch dynamically across a collection of
2691polymorphic interface implementations. This latter use case can be modeled with
2692virtual dispatch and inheritance by defining an abstract interface base class
2693which all implementations derive from and override. However, this
2694implementation strategy forces an **"is-a"** relationship to exist that is not
2695actually meaningful. There is often not some nested hierarchy of useful
2696generalizations which code might interact with and move up and down. Instead,
2697there is a singular interface which is dispatched across a range of
2698implementations.
2699
2700The preferred implementation strategy for the second use case is that of
2701generic programming (sometimes called "compile-time duck typing" or "static
2702polymorphism"). For example, a template over some type parameter ``T`` can be
2703instantiated across any particular implementation that conforms to the
2704interface or *concept*. A good example here is the highly generic properties of
2705any type which models a node in a directed graph. LLVM models these primarily
2706through templates and generic programming. Such templates include the
2707``LoopInfoBase`` and ``DominatorTreeBase``. When this type of polymorphism
2708truly needs **dynamic** dispatch you can generalize it using a technique
2709called *concept-based polymorphism*. This pattern emulates the interfaces and
2710behaviors of templates using a very limited form of virtual dispatch for type
2711erasure inside its implementation. You can find examples of this technique in
2712the ``PassManager.h`` system, and there is a more detailed introduction to it
2713by Sean Parent in several of his talks and papers:
2714
2715#. `Inheritance Is The Base Class of Evil
2716   <http://channel9.msdn.com/Events/GoingNative/2013/Inheritance-Is-The-Base-Class-of-Evil>`_
2717   - The GoingNative 2013 talk describing this technique, and probably the best
2718   place to start.
2719#. `Value Semantics and Concepts-based Polymorphism
2720   <http://www.youtube.com/watch?v=_BpMYeUFXv8>`_ - The C++Now! 2012 talk
2721   describing this technique in more detail.
2722#. `Sean Parent's Papers and Presentations
2723   <http://github.com/sean-parent/sean-parent.github.com/wiki/Papers-and-Presentations>`_
2724   - A Github project full of links to slides, video, and sometimes code.
2725
2726When deciding between creating a type hierarchy (with either tagged or virtual
2727dispatch) and using templates or concepts-based polymorphism, consider whether
2728there is some refinement of an abstract base class which is a semantically
2729meaningful type on an interface boundary. If anything more refined than the
2730root abstract interface is meaningless to talk about as a partial extension of
2731the semantic model, then your use case likely fits better with polymorphism and
2732you should avoid using virtual dispatch. However, there may be some exigent
2733circumstances that require one technique or the other to be used.
2734
2735If you do need to introduce a type hierarchy, we prefer to use explicitly
2736closed type hierarchies with manual tagged dispatch and/or RTTI rather than the
2737open inheritance model and virtual dispatch that is more common in C++ code.
2738This is because LLVM rarely encourages library consumers to extend its core
2739types, and leverages the closed and tag-dispatched nature of its hierarchies to
2740generate significantly more efficient code. We have also found that a large
2741amount of our usage of type hierarchies fits better with tag-based pattern
2742matching rather than dynamic dispatch across a common interface. Within LLVM we
2743have built custom helpers to facilitate this design. See this document's
2744section on :ref:`isa and dyn_cast <isa>` and our :doc:`detailed document
2745<HowToSetUpLLVMStyleRTTI>` which describes how you can implement this
2746pattern for use with the LLVM helpers.
2747
2748.. _abi_breaking_checks:
2749
2750ABI Breaking Checks
2751-------------------
2752
2753Checks and asserts that alter the LLVM C++ ABI are predicated on the
2754preprocessor symbol `LLVM_ENABLE_ABI_BREAKING_CHECKS` -- LLVM
2755libraries built with `LLVM_ENABLE_ABI_BREAKING_CHECKS` are not ABI
2756compatible LLVM libraries built without it defined.  By default,
2757turning on assertions also turns on `LLVM_ENABLE_ABI_BREAKING_CHECKS`
2758so a default +Asserts build is not ABI compatible with a
2759default -Asserts build.  Clients that want ABI compatibility
2760between +Asserts and -Asserts builds should use the CMake or autoconf
2761build systems to set `LLVM_ENABLE_ABI_BREAKING_CHECKS` independently
2762of `LLVM_ENABLE_ASSERTIONS`.
2763
2764.. _coreclasses:
2765
2766The Core LLVM Class Hierarchy Reference
2767=======================================
2768
2769``#include "llvm/IR/Type.h"``
2770
2771header source: `Type.h <http://llvm.org/doxygen/Type_8h-source.html>`_
2772
2773doxygen info: `Type Clases <http://llvm.org/doxygen/classllvm_1_1Type.html>`_
2774
2775The Core LLVM classes are the primary means of representing the program being
2776inspected or transformed.  The core LLVM classes are defined in header files in
2777the ``include/llvm/IR`` directory, and implemented in the ``lib/IR``
2778directory. It's worth noting that, for historical reasons, this library is
2779called ``libLLVMCore.so``, not ``libLLVMIR.so`` as you might expect.
2780
2781.. _Type:
2782
2783The Type class and Derived Types
2784--------------------------------
2785
2786``Type`` is a superclass of all type classes.  Every ``Value`` has a ``Type``.
2787``Type`` cannot be instantiated directly but only through its subclasses.
2788Certain primitive types (``VoidType``, ``LabelType``, ``FloatType`` and
2789``DoubleType``) have hidden subclasses.  They are hidden because they offer no
2790useful functionality beyond what the ``Type`` class offers except to distinguish
2791themselves from other subclasses of ``Type``.
2792
2793All other types are subclasses of ``DerivedType``.  Types can be named, but this
2794is not a requirement.  There exists exactly one instance of a given shape at any
2795one time.  This allows type equality to be performed with address equality of
2796the Type Instance.  That is, given two ``Type*`` values, the types are identical
2797if the pointers are identical.
2798
2799.. _m_Type:
2800
2801Important Public Methods
2802^^^^^^^^^^^^^^^^^^^^^^^^
2803
2804* ``bool isIntegerTy() const``: Returns true for any integer type.
2805
2806* ``bool isFloatingPointTy()``: Return true if this is one of the five
2807  floating point types.
2808
2809* ``bool isSized()``: Return true if the type has known size.  Things
2810  that don't have a size are abstract types, labels and void.
2811
2812.. _derivedtypes:
2813
2814Important Derived Types
2815^^^^^^^^^^^^^^^^^^^^^^^
2816
2817``IntegerType``
2818  Subclass of DerivedType that represents integer types of any bit width.  Any
2819  bit width between ``IntegerType::MIN_INT_BITS`` (1) and
2820  ``IntegerType::MAX_INT_BITS`` (~8 million) can be represented.
2821
2822  * ``static const IntegerType* get(unsigned NumBits)``: get an integer
2823    type of a specific bit width.
2824
2825  * ``unsigned getBitWidth() const``: Get the bit width of an integer type.
2826
2827``SequentialType``
2828  This is subclassed by ArrayType, PointerType and VectorType.
2829
2830  * ``const Type * getElementType() const``: Returns the type of each
2831    of the elements in the sequential type.
2832
2833``ArrayType``
2834  This is a subclass of SequentialType and defines the interface for array
2835  types.
2836
2837  * ``unsigned getNumElements() const``: Returns the number of elements
2838    in the array.
2839
2840``PointerType``
2841  Subclass of SequentialType for pointer types.
2842
2843``VectorType``
2844  Subclass of SequentialType for vector types.  A vector type is similar to an
2845  ArrayType but is distinguished because it is a first class type whereas
2846  ArrayType is not.  Vector types are used for vector operations and are usually
2847  small vectors of an integer or floating point type.
2848
2849``StructType``
2850  Subclass of DerivedTypes for struct types.
2851
2852.. _FunctionType:
2853
2854``FunctionType``
2855  Subclass of DerivedTypes for function types.
2856
2857  * ``bool isVarArg() const``: Returns true if it's a vararg function.
2858
2859  * ``const Type * getReturnType() const``: Returns the return type of the
2860    function.
2861
2862  * ``const Type * getParamType (unsigned i)``: Returns the type of the ith
2863    parameter.
2864
2865  * ``const unsigned getNumParams() const``: Returns the number of formal
2866    parameters.
2867
2868.. _Module:
2869
2870The ``Module`` class
2871--------------------
2872
2873``#include "llvm/IR/Module.h"``
2874
2875header source: `Module.h <http://llvm.org/doxygen/Module_8h-source.html>`_
2876
2877doxygen info: `Module Class <http://llvm.org/doxygen/classllvm_1_1Module.html>`_
2878
2879The ``Module`` class represents the top level structure present in LLVM
2880programs.  An LLVM module is effectively either a translation unit of the
2881original program or a combination of several translation units merged by the
2882linker.  The ``Module`` class keeps track of a list of :ref:`Function
2883<c_Function>`\ s, a list of GlobalVariable_\ s, and a SymbolTable_.
2884Additionally, it contains a few helpful member functions that try to make common
2885operations easy.
2886
2887.. _m_Module:
2888
2889Important Public Members of the ``Module`` class
2890^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
2891
2892* ``Module::Module(std::string name = "")``
2893
2894  Constructing a Module_ is easy.  You can optionally provide a name for it
2895  (probably based on the name of the translation unit).
2896
2897* | ``Module::iterator`` - Typedef for function list iterator
2898  | ``Module::const_iterator`` - Typedef for const_iterator.
2899  | ``begin()``, ``end()``, ``size()``, ``empty()``
2900
2901  These are forwarding methods that make it easy to access the contents of a
2902  ``Module`` object's :ref:`Function <c_Function>` list.
2903
2904* ``Module::FunctionListType &getFunctionList()``
2905
2906  Returns the list of :ref:`Function <c_Function>`\ s.  This is necessary to use
2907  when you need to update the list or perform a complex action that doesn't have
2908  a forwarding method.
2909
2910----------------
2911
2912* | ``Module::global_iterator`` - Typedef for global variable list iterator
2913  | ``Module::const_global_iterator`` - Typedef for const_iterator.
2914  | ``global_begin()``, ``global_end()``, ``global_size()``, ``global_empty()``
2915
2916  These are forwarding methods that make it easy to access the contents of a
2917  ``Module`` object's GlobalVariable_ list.
2918
2919* ``Module::GlobalListType &getGlobalList()``
2920
2921  Returns the list of GlobalVariable_\ s.  This is necessary to use when you
2922  need to update the list or perform a complex action that doesn't have a
2923  forwarding method.
2924
2925----------------
2926
2927* ``SymbolTable *getSymbolTable()``
2928
2929  Return a reference to the SymbolTable_ for this ``Module``.
2930
2931----------------
2932
2933* ``Function *getFunction(StringRef Name) const``
2934
2935  Look up the specified function in the ``Module`` SymbolTable_.  If it does not
2936  exist, return ``null``.
2937
2938* ``Function *getOrInsertFunction(const std::string &Name, const FunctionType
2939  *T)``
2940
2941  Look up the specified function in the ``Module`` SymbolTable_.  If it does not
2942  exist, add an external declaration for the function and return it.
2943
2944* ``std::string getTypeName(const Type *Ty)``
2945
2946  If there is at least one entry in the SymbolTable_ for the specified Type_,
2947  return it.  Otherwise return the empty string.
2948
2949* ``bool addTypeName(const std::string &Name, const Type *Ty)``
2950
2951  Insert an entry in the SymbolTable_ mapping ``Name`` to ``Ty``.  If there is
2952  already an entry for this name, true is returned and the SymbolTable_ is not
2953  modified.
2954
2955.. _Value:
2956
2957The ``Value`` class
2958-------------------
2959
2960``#include "llvm/IR/Value.h"``
2961
2962header source: `Value.h <http://llvm.org/doxygen/Value_8h-source.html>`_
2963
2964doxygen info: `Value Class <http://llvm.org/doxygen/classllvm_1_1Value.html>`_
2965
2966The ``Value`` class is the most important class in the LLVM Source base.  It
2967represents a typed value that may be used (among other things) as an operand to
2968an instruction.  There are many different types of ``Value``\ s, such as
2969Constant_\ s, Argument_\ s.  Even Instruction_\ s and :ref:`Function
2970<c_Function>`\ s are ``Value``\ s.
2971
2972A particular ``Value`` may be used many times in the LLVM representation for a
2973program.  For example, an incoming argument to a function (represented with an
2974instance of the Argument_ class) is "used" by every instruction in the function
2975that references the argument.  To keep track of this relationship, the ``Value``
2976class keeps a list of all of the ``User``\ s that is using it (the User_ class
2977is a base class for all nodes in the LLVM graph that can refer to ``Value``\ s).
2978This use list is how LLVM represents def-use information in the program, and is
2979accessible through the ``use_*`` methods, shown below.
2980
2981Because LLVM is a typed representation, every LLVM ``Value`` is typed, and this
2982Type_ is available through the ``getType()`` method.  In addition, all LLVM
2983values can be named.  The "name" of the ``Value`` is a symbolic string printed
2984in the LLVM code:
2985
2986.. code-block:: llvm
2987
2988  %foo = add i32 1, 2
2989
2990.. _nameWarning:
2991
2992The name of this instruction is "foo". **NOTE** that the name of any value may
2993be missing (an empty string), so names should **ONLY** be used for debugging
2994(making the source code easier to read, debugging printouts), they should not be
2995used to keep track of values or map between them.  For this purpose, use a
2996``std::map`` of pointers to the ``Value`` itself instead.
2997
2998One important aspect of LLVM is that there is no distinction between an SSA
2999variable and the operation that produces it.  Because of this, any reference to
3000the value produced by an instruction (or the value available as an incoming
3001argument, for example) is represented as a direct pointer to the instance of the
3002class that represents this value.  Although this may take some getting used to,
3003it simplifies the representation and makes it easier to manipulate.
3004
3005.. _m_Value:
3006
3007Important Public Members of the ``Value`` class
3008^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
3009
3010* | ``Value::use_iterator`` - Typedef for iterator over the use-list
3011  | ``Value::const_use_iterator`` - Typedef for const_iterator over the
3012    use-list
3013  | ``unsigned use_size()`` - Returns the number of users of the value.
3014  | ``bool use_empty()`` - Returns true if there are no users.
3015  | ``use_iterator use_begin()`` - Get an iterator to the start of the
3016    use-list.
3017  | ``use_iterator use_end()`` - Get an iterator to the end of the use-list.
3018  | ``User *use_back()`` - Returns the last element in the list.
3019
3020  These methods are the interface to access the def-use information in LLVM.
3021  As with all other iterators in LLVM, the naming conventions follow the
3022  conventions defined by the STL_.
3023
3024* ``Type *getType() const``
3025  This method returns the Type of the Value.
3026
3027* | ``bool hasName() const``
3028  | ``std::string getName() const``
3029  | ``void setName(const std::string &Name)``
3030
3031  This family of methods is used to access and assign a name to a ``Value``, be
3032  aware of the :ref:`precaution above <nameWarning>`.
3033
3034* ``void replaceAllUsesWith(Value *V)``
3035
3036  This method traverses the use list of a ``Value`` changing all User_\ s of the
3037  current value to refer to "``V``" instead.  For example, if you detect that an
3038  instruction always produces a constant value (for example through constant
3039  folding), you can replace all uses of the instruction with the constant like
3040  this:
3041
3042  .. code-block:: c++
3043
3044    Inst->replaceAllUsesWith(ConstVal);
3045
3046.. _User:
3047
3048The ``User`` class
3049------------------
3050
3051``#include "llvm/IR/User.h"``
3052
3053header source: `User.h <http://llvm.org/doxygen/User_8h-source.html>`_
3054
3055doxygen info: `User Class <http://llvm.org/doxygen/classllvm_1_1User.html>`_
3056
3057Superclass: Value_
3058
3059The ``User`` class is the common base class of all LLVM nodes that may refer to
3060``Value``\ s.  It exposes a list of "Operands" that are all of the ``Value``\ s
3061that the User is referring to.  The ``User`` class itself is a subclass of
3062``Value``.
3063
3064The operands of a ``User`` point directly to the LLVM ``Value`` that it refers
3065to.  Because LLVM uses Static Single Assignment (SSA) form, there can only be
3066one definition referred to, allowing this direct connection.  This connection
3067provides the use-def information in LLVM.
3068
3069.. _m_User:
3070
3071Important Public Members of the ``User`` class
3072^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
3073
3074The ``User`` class exposes the operand list in two ways: through an index access
3075interface and through an iterator based interface.
3076
3077* | ``Value *getOperand(unsigned i)``
3078  | ``unsigned getNumOperands()``
3079
3080  These two methods expose the operands of the ``User`` in a convenient form for
3081  direct access.
3082
3083* | ``User::op_iterator`` - Typedef for iterator over the operand list
3084  | ``op_iterator op_begin()`` - Get an iterator to the start of the operand
3085    list.
3086  | ``op_iterator op_end()`` - Get an iterator to the end of the operand list.
3087
3088  Together, these methods make up the iterator based interface to the operands
3089  of a ``User``.
3090
3091
3092.. _Instruction:
3093
3094The ``Instruction`` class
3095-------------------------
3096
3097``#include "llvm/IR/Instruction.h"``
3098
3099header source: `Instruction.h
3100<http://llvm.org/doxygen/Instruction_8h-source.html>`_
3101
3102doxygen info: `Instruction Class
3103<http://llvm.org/doxygen/classllvm_1_1Instruction.html>`_
3104
3105Superclasses: User_, Value_
3106
3107The ``Instruction`` class is the common base class for all LLVM instructions.
3108It provides only a few methods, but is a very commonly used class.  The primary
3109data tracked by the ``Instruction`` class itself is the opcode (instruction
3110type) and the parent BasicBlock_ the ``Instruction`` is embedded into.  To
3111represent a specific type of instruction, one of many subclasses of
3112``Instruction`` are used.
3113
3114Because the ``Instruction`` class subclasses the User_ class, its operands can
3115be accessed in the same way as for other ``User``\ s (with the
3116``getOperand()``/``getNumOperands()`` and ``op_begin()``/``op_end()`` methods).
3117An important file for the ``Instruction`` class is the ``llvm/Instruction.def``
3118file.  This file contains some meta-data about the various different types of
3119instructions in LLVM.  It describes the enum values that are used as opcodes
3120(for example ``Instruction::Add`` and ``Instruction::ICmp``), as well as the
3121concrete sub-classes of ``Instruction`` that implement the instruction (for
3122example BinaryOperator_ and CmpInst_).  Unfortunately, the use of macros in this
3123file confuses doxygen, so these enum values don't show up correctly in the
3124`doxygen output <http://llvm.org/doxygen/classllvm_1_1Instruction.html>`_.
3125
3126.. _s_Instruction:
3127
3128Important Subclasses of the ``Instruction`` class
3129^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
3130
3131.. _BinaryOperator:
3132
3133* ``BinaryOperator``
3134
3135  This subclasses represents all two operand instructions whose operands must be
3136  the same type, except for the comparison instructions.
3137
3138.. _CastInst:
3139
3140* ``CastInst``
3141  This subclass is the parent of the 12 casting instructions.  It provides
3142  common operations on cast instructions.
3143
3144.. _CmpInst:
3145
3146* ``CmpInst``
3147
3148  This subclass respresents the two comparison instructions,
3149  `ICmpInst <LangRef.html#i_icmp>`_ (integer opreands), and
3150  `FCmpInst <LangRef.html#i_fcmp>`_ (floating point operands).
3151
3152.. _TerminatorInst:
3153
3154* ``TerminatorInst``
3155
3156  This subclass is the parent of all terminator instructions (those which can
3157  terminate a block).
3158
3159.. _m_Instruction:
3160
3161Important Public Members of the ``Instruction`` class
3162^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
3163
3164* ``BasicBlock *getParent()``
3165
3166  Returns the BasicBlock_ that this
3167  ``Instruction`` is embedded into.
3168
3169* ``bool mayWriteToMemory()``
3170
3171  Returns true if the instruction writes to memory, i.e. it is a ``call``,
3172  ``free``, ``invoke``, or ``store``.
3173
3174* ``unsigned getOpcode()``
3175
3176  Returns the opcode for the ``Instruction``.
3177
3178* ``Instruction *clone() const``
3179
3180  Returns another instance of the specified instruction, identical in all ways
3181  to the original except that the instruction has no parent (i.e. it's not
3182  embedded into a BasicBlock_), and it has no name.
3183
3184.. _Constant:
3185
3186The ``Constant`` class and subclasses
3187-------------------------------------
3188
3189Constant represents a base class for different types of constants.  It is
3190subclassed by ConstantInt, ConstantArray, etc. for representing the various
3191types of Constants.  GlobalValue_ is also a subclass, which represents the
3192address of a global variable or function.
3193
3194.. _s_Constant:
3195
3196Important Subclasses of Constant
3197^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
3198
3199* ConstantInt : This subclass of Constant represents an integer constant of
3200  any width.
3201
3202  * ``const APInt& getValue() const``: Returns the underlying
3203    value of this constant, an APInt value.
3204
3205  * ``int64_t getSExtValue() const``: Converts the underlying APInt value to an
3206    int64_t via sign extension.  If the value (not the bit width) of the APInt
3207    is too large to fit in an int64_t, an assertion will result.  For this
3208    reason, use of this method is discouraged.
3209
3210  * ``uint64_t getZExtValue() const``: Converts the underlying APInt value
3211    to a uint64_t via zero extension.  IF the value (not the bit width) of the
3212    APInt is too large to fit in a uint64_t, an assertion will result.  For this
3213    reason, use of this method is discouraged.
3214
3215  * ``static ConstantInt* get(const APInt& Val)``: Returns the ConstantInt
3216    object that represents the value provided by ``Val``.  The type is implied
3217    as the IntegerType that corresponds to the bit width of ``Val``.
3218
3219  * ``static ConstantInt* get(const Type *Ty, uint64_t Val)``: Returns the
3220    ConstantInt object that represents the value provided by ``Val`` for integer
3221    type ``Ty``.
3222
3223* ConstantFP : This class represents a floating point constant.
3224
3225  * ``double getValue() const``: Returns the underlying value of this constant.
3226
3227* ConstantArray : This represents a constant array.
3228
3229  * ``const std::vector<Use> &getValues() const``: Returns a vector of
3230    component constants that makeup this array.
3231
3232* ConstantStruct : This represents a constant struct.
3233
3234  * ``const std::vector<Use> &getValues() const``: Returns a vector of
3235    component constants that makeup this array.
3236
3237* GlobalValue : This represents either a global variable or a function.  In
3238  either case, the value is a constant fixed address (after linking).
3239
3240.. _GlobalValue:
3241
3242The ``GlobalValue`` class
3243-------------------------
3244
3245``#include "llvm/IR/GlobalValue.h"``
3246
3247header source: `GlobalValue.h
3248<http://llvm.org/doxygen/GlobalValue_8h-source.html>`_
3249
3250doxygen info: `GlobalValue Class
3251<http://llvm.org/doxygen/classllvm_1_1GlobalValue.html>`_
3252
3253Superclasses: Constant_, User_, Value_
3254
3255Global values ( GlobalVariable_\ s or :ref:`Function <c_Function>`\ s) are the
3256only LLVM values that are visible in the bodies of all :ref:`Function
3257<c_Function>`\ s.  Because they are visible at global scope, they are also
3258subject to linking with other globals defined in different translation units.
3259To control the linking process, ``GlobalValue``\ s know their linkage rules.
3260Specifically, ``GlobalValue``\ s know whether they have internal or external
3261linkage, as defined by the ``LinkageTypes`` enumeration.
3262
3263If a ``GlobalValue`` has internal linkage (equivalent to being ``static`` in C),
3264it is not visible to code outside the current translation unit, and does not
3265participate in linking.  If it has external linkage, it is visible to external
3266code, and does participate in linking.  In addition to linkage information,
3267``GlobalValue``\ s keep track of which Module_ they are currently part of.
3268
3269Because ``GlobalValue``\ s are memory objects, they are always referred to by
3270their **address**.  As such, the Type_ of a global is always a pointer to its
3271contents.  It is important to remember this when using the ``GetElementPtrInst``
3272instruction because this pointer must be dereferenced first.  For example, if
3273you have a ``GlobalVariable`` (a subclass of ``GlobalValue)`` that is an array
3274of 24 ints, type ``[24 x i32]``, then the ``GlobalVariable`` is a pointer to
3275that array.  Although the address of the first element of this array and the
3276value of the ``GlobalVariable`` are the same, they have different types.  The
3277``GlobalVariable``'s type is ``[24 x i32]``.  The first element's type is
3278``i32.`` Because of this, accessing a global value requires you to dereference
3279the pointer with ``GetElementPtrInst`` first, then its elements can be accessed.
3280This is explained in the `LLVM Language Reference Manual
3281<LangRef.html#globalvars>`_.
3282
3283.. _m_GlobalValue:
3284
3285Important Public Members of the ``GlobalValue`` class
3286^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
3287
3288* | ``bool hasInternalLinkage() const``
3289  | ``bool hasExternalLinkage() const``
3290  | ``void setInternalLinkage(bool HasInternalLinkage)``
3291
3292  These methods manipulate the linkage characteristics of the ``GlobalValue``.
3293
3294* ``Module *getParent()``
3295
3296  This returns the Module_ that the
3297  GlobalValue is currently embedded into.
3298
3299.. _c_Function:
3300
3301The ``Function`` class
3302----------------------
3303
3304``#include "llvm/IR/Function.h"``
3305
3306header source: `Function.h <http://llvm.org/doxygen/Function_8h-source.html>`_
3307
3308doxygen info: `Function Class
3309<http://llvm.org/doxygen/classllvm_1_1Function.html>`_
3310
3311Superclasses: GlobalValue_, Constant_, User_, Value_
3312
3313The ``Function`` class represents a single procedure in LLVM.  It is actually
3314one of the more complex classes in the LLVM hierarchy because it must keep track
3315of a large amount of data.  The ``Function`` class keeps track of a list of
3316BasicBlock_\ s, a list of formal Argument_\ s, and a SymbolTable_.
3317
3318The list of BasicBlock_\ s is the most commonly used part of ``Function``
3319objects.  The list imposes an implicit ordering of the blocks in the function,
3320which indicate how the code will be laid out by the backend.  Additionally, the
3321first BasicBlock_ is the implicit entry node for the ``Function``.  It is not
3322legal in LLVM to explicitly branch to this initial block.  There are no implicit
3323exit nodes, and in fact there may be multiple exit nodes from a single
3324``Function``.  If the BasicBlock_ list is empty, this indicates that the
3325``Function`` is actually a function declaration: the actual body of the function
3326hasn't been linked in yet.
3327
3328In addition to a list of BasicBlock_\ s, the ``Function`` class also keeps track
3329of the list of formal Argument_\ s that the function receives.  This container
3330manages the lifetime of the Argument_ nodes, just like the BasicBlock_ list does
3331for the BasicBlock_\ s.
3332
3333The SymbolTable_ is a very rarely used LLVM feature that is only used when you
3334have to look up a value by name.  Aside from that, the SymbolTable_ is used
3335internally to make sure that there are not conflicts between the names of
3336Instruction_\ s, BasicBlock_\ s, or Argument_\ s in the function body.
3337
3338Note that ``Function`` is a GlobalValue_ and therefore also a Constant_.  The
3339value of the function is its address (after linking) which is guaranteed to be
3340constant.
3341
3342.. _m_Function:
3343
3344Important Public Members of the ``Function``
3345^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
3346
3347* ``Function(const FunctionType *Ty, LinkageTypes Linkage,
3348  const std::string &N = "", Module* Parent = 0)``
3349
3350  Constructor used when you need to create new ``Function``\ s to add the
3351  program.  The constructor must specify the type of the function to create and
3352  what type of linkage the function should have.  The FunctionType_ argument
3353  specifies the formal arguments and return value for the function.  The same
3354  FunctionType_ value can be used to create multiple functions.  The ``Parent``
3355  argument specifies the Module in which the function is defined.  If this
3356  argument is provided, the function will automatically be inserted into that
3357  module's list of functions.
3358
3359* ``bool isDeclaration()``
3360
3361  Return whether or not the ``Function`` has a body defined.  If the function is
3362  "external", it does not have a body, and thus must be resolved by linking with
3363  a function defined in a different translation unit.
3364
3365* | ``Function::iterator`` - Typedef for basic block list iterator
3366  | ``Function::const_iterator`` - Typedef for const_iterator.
3367  | ``begin()``, ``end()``, ``size()``, ``empty()``
3368
3369  These are forwarding methods that make it easy to access the contents of a
3370  ``Function`` object's BasicBlock_ list.
3371
3372* ``Function::BasicBlockListType &getBasicBlockList()``
3373
3374  Returns the list of BasicBlock_\ s.  This is necessary to use when you need to
3375  update the list or perform a complex action that doesn't have a forwarding
3376  method.
3377
3378* | ``Function::arg_iterator`` - Typedef for the argument list iterator
3379  | ``Function::const_arg_iterator`` - Typedef for const_iterator.
3380  | ``arg_begin()``, ``arg_end()``, ``arg_size()``, ``arg_empty()``
3381
3382  These are forwarding methods that make it easy to access the contents of a
3383  ``Function`` object's Argument_ list.
3384
3385* ``Function::ArgumentListType &getArgumentList()``
3386
3387  Returns the list of Argument_.  This is necessary to use when you need to
3388  update the list or perform a complex action that doesn't have a forwarding
3389  method.
3390
3391* ``BasicBlock &getEntryBlock()``
3392
3393  Returns the entry ``BasicBlock`` for the function.  Because the entry block
3394  for the function is always the first block, this returns the first block of
3395  the ``Function``.
3396
3397* | ``Type *getReturnType()``
3398  | ``FunctionType *getFunctionType()``
3399
3400  This traverses the Type_ of the ``Function`` and returns the return type of
3401  the function, or the FunctionType_ of the actual function.
3402
3403* ``SymbolTable *getSymbolTable()``
3404
3405  Return a pointer to the SymbolTable_ for this ``Function``.
3406
3407.. _GlobalVariable:
3408
3409The ``GlobalVariable`` class
3410----------------------------
3411
3412``#include "llvm/IR/GlobalVariable.h"``
3413
3414header source: `GlobalVariable.h
3415<http://llvm.org/doxygen/GlobalVariable_8h-source.html>`_
3416
3417doxygen info: `GlobalVariable Class
3418<http://llvm.org/doxygen/classllvm_1_1GlobalVariable.html>`_
3419
3420Superclasses: GlobalValue_, Constant_, User_, Value_
3421
3422Global variables are represented with the (surprise surprise) ``GlobalVariable``
3423class.  Like functions, ``GlobalVariable``\ s are also subclasses of
3424GlobalValue_, and as such are always referenced by their address (global values
3425must live in memory, so their "name" refers to their constant address).  See
3426GlobalValue_ for more on this.  Global variables may have an initial value
3427(which must be a Constant_), and if they have an initializer, they may be marked
3428as "constant" themselves (indicating that their contents never change at
3429runtime).
3430
3431.. _m_GlobalVariable:
3432
3433Important Public Members of the ``GlobalVariable`` class
3434^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
3435
3436* ``GlobalVariable(const Type *Ty, bool isConstant, LinkageTypes &Linkage,
3437  Constant *Initializer = 0, const std::string &Name = "", Module* Parent = 0)``
3438
3439  Create a new global variable of the specified type.  If ``isConstant`` is true
3440  then the global variable will be marked as unchanging for the program.  The
3441  Linkage parameter specifies the type of linkage (internal, external, weak,
3442  linkonce, appending) for the variable.  If the linkage is InternalLinkage,
3443  WeakAnyLinkage, WeakODRLinkage, LinkOnceAnyLinkage or LinkOnceODRLinkage, then
3444  the resultant global variable will have internal linkage.  AppendingLinkage
3445  concatenates together all instances (in different translation units) of the
3446  variable into a single variable but is only applicable to arrays.  See the
3447  `LLVM Language Reference <LangRef.html#modulestructure>`_ for further details
3448  on linkage types.  Optionally an initializer, a name, and the module to put
3449  the variable into may be specified for the global variable as well.
3450
3451* ``bool isConstant() const``
3452
3453  Returns true if this is a global variable that is known not to be modified at
3454  runtime.
3455
3456* ``bool hasInitializer()``
3457
3458  Returns true if this ``GlobalVariable`` has an intializer.
3459
3460* ``Constant *getInitializer()``
3461
3462  Returns the initial value for a ``GlobalVariable``.  It is not legal to call
3463  this method if there is no initializer.
3464
3465.. _BasicBlock:
3466
3467The ``BasicBlock`` class
3468------------------------
3469
3470``#include "llvm/IR/BasicBlock.h"``
3471
3472header source: `BasicBlock.h
3473<http://llvm.org/doxygen/BasicBlock_8h-source.html>`_
3474
3475doxygen info: `BasicBlock Class
3476<http://llvm.org/doxygen/classllvm_1_1BasicBlock.html>`_
3477
3478Superclass: Value_
3479
3480This class represents a single entry single exit section of the code, commonly
3481known as a basic block by the compiler community.  The ``BasicBlock`` class
3482maintains a list of Instruction_\ s, which form the body of the block.  Matching
3483the language definition, the last element of this list of instructions is always
3484a terminator instruction (a subclass of the TerminatorInst_ class).
3485
3486In addition to tracking the list of instructions that make up the block, the
3487``BasicBlock`` class also keeps track of the :ref:`Function <c_Function>` that
3488it is embedded into.
3489
3490Note that ``BasicBlock``\ s themselves are Value_\ s, because they are
3491referenced by instructions like branches and can go in the switch tables.
3492``BasicBlock``\ s have type ``label``.
3493
3494.. _m_BasicBlock:
3495
3496Important Public Members of the ``BasicBlock`` class
3497^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
3498
3499* ``BasicBlock(const std::string &Name = "", Function *Parent = 0)``
3500
3501  The ``BasicBlock`` constructor is used to create new basic blocks for
3502  insertion into a function.  The constructor optionally takes a name for the
3503  new block, and a :ref:`Function <c_Function>` to insert it into.  If the
3504  ``Parent`` parameter is specified, the new ``BasicBlock`` is automatically
3505  inserted at the end of the specified :ref:`Function <c_Function>`, if not
3506  specified, the BasicBlock must be manually inserted into the :ref:`Function
3507  <c_Function>`.
3508
3509* | ``BasicBlock::iterator`` - Typedef for instruction list iterator
3510  | ``BasicBlock::const_iterator`` - Typedef for const_iterator.
3511  | ``begin()``, ``end()``, ``front()``, ``back()``,
3512    ``size()``, ``empty()``
3513    STL-style functions for accessing the instruction list.
3514
3515  These methods and typedefs are forwarding functions that have the same
3516  semantics as the standard library methods of the same names.  These methods
3517  expose the underlying instruction list of a basic block in a way that is easy
3518  to manipulate.  To get the full complement of container operations (including
3519  operations to update the list), you must use the ``getInstList()`` method.
3520
3521* ``BasicBlock::InstListType &getInstList()``
3522
3523  This method is used to get access to the underlying container that actually
3524  holds the Instructions.  This method must be used when there isn't a
3525  forwarding function in the ``BasicBlock`` class for the operation that you
3526  would like to perform.  Because there are no forwarding functions for
3527  "updating" operations, you need to use this if you want to update the contents
3528  of a ``BasicBlock``.
3529
3530* ``Function *getParent()``
3531
3532  Returns a pointer to :ref:`Function <c_Function>` the block is embedded into,
3533  or a null pointer if it is homeless.
3534
3535* ``TerminatorInst *getTerminator()``
3536
3537  Returns a pointer to the terminator instruction that appears at the end of the
3538  ``BasicBlock``.  If there is no terminator instruction, or if the last
3539  instruction in the block is not a terminator, then a null pointer is returned.
3540
3541.. _Argument:
3542
3543The ``Argument`` class
3544----------------------
3545
3546This subclass of Value defines the interface for incoming formal arguments to a
3547function.  A Function maintains a list of its formal arguments.  An argument has
3548a pointer to the parent Function.
3549
3550
3551