1 \input texinfo @c -*- texinfo -*-
3 @setfilename tcc-doc.info
4 @settitle Tiny C Compiler Reference Documentation
5 @dircategory Software development
7 * TCC: (tcc-doc). The Tiny C Compiler.
17 @center @titlefont{Tiny C Compiler Reference Documentation}
25 @node Top, Introduction, (dir), (dir)
26 @top Tiny C Compiler Reference Documentation
28 This manual documents version @value{VERSION} of the Tiny C Compiler.
31 * Introduction:: Introduction to tcc.
32 * Invoke:: Invocation of tcc (command line, options).
33 * Clang:: ANSI C and extensions.
34 * asm:: Assembler syntax.
35 * linker:: Output file generation and supported targets.
36 * Bounds:: Automatic bounds-checking of C code.
37 * Libtcc:: The libtcc library.
38 * devel:: Guide for Developers.
45 TinyCC (aka TCC) is a small but hyper fast C compiler. Unlike other C
46 compilers, it is meant to be self-relying: you do not need an
47 external assembler or linker because TCC does that for you.
49 TCC compiles so @emph{fast} that even for big projects @code{Makefile}s may
52 TCC not only supports ANSI C, but also most of the new ISO C99
53 standard and many GNUC extensions including inline assembly.
55 TCC can also be used to make @emph{C scripts}, i.e. pieces of C source
56 that you run as a Perl or Python script. Compilation is so fast that
57 your script will be as fast as if it was an executable.
59 TCC can also automatically generate memory and bound checks
60 (@pxref{Bounds}) while allowing all C pointers operations. TCC can do
61 these checks even if non patched libraries are used.
63 With @code{libtcc}, you can use TCC as a backend for dynamic code
64 generation (@pxref{Libtcc}).
66 TCC mainly supports the i386 target on Linux and Windows. There are alpha
67 ports for the ARM (@code{arm-tcc}) and the TMS320C67xx targets
68 (@code{c67-tcc}). More information about the ARM port is available at
69 @url{http://lists.gnu.org/archive/html/tinycc-devel/2003-10/msg00044.html}.
71 For usage on Windows, see also @url{tcc-win32.txt}.
74 @chapter Command line invocation
80 usage: tcc [options] [@var{infile1} @var{infile2}@dots{}] [@option{-run} @var{infile} @var{args}@dots{}]
85 @c man begin DESCRIPTION
86 TCC options are a very much like gcc options. The main difference is that TCC
87 can also execute directly the resulting program and give it runtime
90 Here are some examples to understand the logic:
93 @item @samp{tcc -run a.c}
94 Compile @file{a.c} and execute it directly
96 @item @samp{tcc -run a.c arg1}
97 Compile a.c and execute it directly. arg1 is given as first argument to
98 the @code{main()} of a.c.
100 @item @samp{tcc a.c -run b.c arg1}
101 Compile @file{a.c} and @file{b.c}, link them together and execute them. arg1 is given
102 as first argument to the @code{main()} of the resulting program.
104 Because multiple C files are specified, @option{--} are necessary to clearly
105 separate the program arguments from the TCC options.
108 @item @samp{tcc -o myprog a.c b.c}
109 Compile @file{a.c} and @file{b.c}, link them and generate the executable @file{myprog}.
111 @item @samp{tcc -o myprog a.o b.o}
112 link @file{a.o} and @file{b.o} together and generate the executable @file{myprog}.
114 @item @samp{tcc -c a.c}
115 Compile @file{a.c} and generate object file @file{a.o}.
117 @item @samp{tcc -c asmfile.S}
118 Preprocess with C preprocess and assemble @file{asmfile.S} and generate
119 object file @file{asmfile.o}.
121 @item @samp{tcc -c asmfile.s}
122 Assemble (but not preprocess) @file{asmfile.s} and generate object file
125 @item @samp{tcc -r -o ab.o a.c b.c}
126 Compile @file{a.c} and @file{b.c}, link them together and generate the object file @file{ab.o}.
132 TCC can be invoked from @emph{scripts}, just as shell scripts. You just
133 need to add @code{#!/usr/local/bin/tcc -run} at the start of your C source:
136 #!/usr/local/bin/tcc -run
141 printf("Hello World\n");
146 TCC can read C source code from @emph{standard input} when @option{-} is used in
147 place of @option{infile}. Example:
150 echo 'main()@{puts("hello");@}' | tcc -run -
154 @section Option summary
161 Generate an object file.
164 Put object file, executable, or dll into output file @file{outfile}.
166 @item -run source [args...]
167 Compile file @var{source} and run it with the command line arguments
168 @var{args}. In order to be able to give more than one argument to a
169 script, several TCC options can be given @emph{after} the
170 @option{-run} option, separated by spaces:
172 tcc "-run -L/usr/X11R6/lib -lX11" ex4.c
174 In a script, it gives the following header:
176 #!/usr/local/bin/tcc -run -L/usr/X11R6/lib -lX11
180 Print only the compiler version and nothing else.
186 Show included files. As sole argument, print search dirs (as below).
189 Display compilation statistics.
191 @item -print-search-dirs
192 Print the configured installation directory and a list of library
193 and include directories tcc will search.
197 Preprocessor options:
201 Specify an additional include path. Include paths are searched in the
202 order they are specified.
204 System include paths are always searched after. The default system
205 include paths are: @file{/usr/local/include}, @file{/usr/include}
206 and @file{PREFIX/lib/tcc/include}. (@file{PREFIX} is usually
207 @file{/usr} or @file{/usr/local}).
210 Define preprocessor symbol @samp{sym} to
211 val. If val is not present, its value is @samp{1}. Function-like macros can
212 also be defined: @option{-DF(a)=a+1}
215 Undefine preprocessor symbol @samp{sym}.
220 Note: each of the following warning options has a negative form beginning with
224 @item -funsigned-char
225 Let the @code{char} type be unsigned.
228 Let the @code{char} type be signed.
231 Do not generate common symbols for uninitialized data.
233 @item -fleading-underscore
234 Add a leading underscore at the beginning of each C symbol.
242 Disable all warnings.
246 Note: each of the following warning options has a negative form beginning with
250 @item -Wimplicit-function-declaration
251 Warn about implicit function declaration.
254 Warn about unsupported GCC features that are ignored by TCC.
256 @item -Wwrite-strings
257 Make string constants be of type @code{const char *} instead of @code{char
261 Abort compilation if warnings are issued.
264 Activate all warnings, except @option{-Werror}, @option{-Wunusupported} and
265 @option{-Wwrite-strings}.
273 Specify an additional static library path for the @option{-l} option. The
274 default library paths are @file{/usr/local/lib}, @file{/usr/lib} and @file{/lib}.
277 Link your program with dynamic library libxxx.so or static library
278 libxxx.a. The library is searched in the paths specified by the
279 @option{-L} option and @env{LIBRARY_PATH} variable.
282 Set the path where the tcc internal libraries (and include files) can be
283 found (default is @file{PREFIX/lib/tcc}).
286 Generate a shared library instead of an executable.
289 set name for shared library to be used at runtime
292 Generate a statically linked executable (default is a shared linked
296 Export global symbols to the dynamic linker. It is useful when a library
297 opened with @code{dlopen()} needs to access executable symbols.
300 Generate an object file combining all input files.
302 @item -Wl,-rpath=path
303 Put custom seatch path for dynamic libraries into executable.
305 @item -Wl,--oformat=fmt
306 Use @var{fmt} as output format. The supported output formats are:
309 ELF output format (default)
311 Binary image (only for executable output)
313 COFF output format (only for executable output for TMS320C67xx target)
316 @item -Wl,-subsystem=console/gui/wince/...
317 Set type for PE (Windows) executables.
319 @item -Wl,-[Ttext=# | section-alignment=# | file-alignment=# | image-base=# | stack=#]
320 Modify executable layout.
331 Generate run time debug information so that you get clear run time
332 error messages: @code{ test.c:68: in function 'test5()': dereferencing
333 invalid pointer} instead of the laconic @code{Segmentation
337 Generate additional support code to check
338 memory allocations and array/pointer bounds. @option{-g} is implied. Note
339 that the generated code is slower and bigger in this case.
341 Note: @option{-b} is only available on i386 for the moment.
344 Display N callers in stack traces. This is useful with @option{-g} or
353 Generate makefile fragment with dependencies.
356 Use @file{depfile} as output for -MD.
360 Note: GCC options @option{-Ox}, @option{-fx} and @option{-mx} are
364 @c man begin ENVIRONMENT
365 Environment variables that affect how tcc operates.
371 A colon-separated list of directories searched for include files,
372 directories given with @option{-I} are searched first.
375 A colon-separated list of directories searched for libraries for the
376 @option{-l} option, directories given with @option{-L} are searched first.
385 @settitle Tiny C Compiler
399 @chapter C language support
403 TCC implements all the ANSI C standard, including structure bit fields
404 and floating point numbers (@code{long double}, @code{double}, and
405 @code{float} fully supported).
407 @section ISOC99 extensions
409 TCC implements many features of the new C standard: ISO C99. Currently
410 missing items are: complex and imaginary numbers.
412 Currently implemented ISOC99 features:
416 @item variable length arrays.
418 @item 64 bit @code{long long} types are fully supported.
420 @item The boolean type @code{_Bool} is supported.
422 @item @code{__func__} is a string variable containing the current
425 @item Variadic macros: @code{__VA_ARGS__} can be used for
426 function-like macros:
428 #define dprintf(level, __VA_ARGS__) printf(__VA_ARGS__)
432 @code{dprintf} can then be used with a variable number of parameters.
434 @item Declarations can appear anywhere in a block (as in C++).
436 @item Array and struct/union elements can be initialized in any order by
439 struct @{ int x, y; @} st[10] = @{ [0].x = 1, [0].y = 2 @};
441 int tab[10] = @{ 1, 2, [5] = 5, [9] = 9@};
444 @item Compound initializers are supported:
446 int *p = (int [])@{ 1, 2, 3 @};
448 to initialize a pointer pointing to an initialized array. The same
449 works for structures and strings.
451 @item Hexadecimal floating point constants are supported:
453 double d = 0x1234p10;
457 is the same as writing
459 double d = 4771840.0;
462 @item @code{inline} keyword is ignored.
464 @item @code{restrict} keyword is ignored.
467 @section GNU C extensions
469 TCC implements some GNU C extensions:
473 @item array designators can be used without '=':
475 int a[10] = @{ [0] 1, [5] 2, 3, 4 @};
478 @item Structure field designators can be a label:
480 struct @{ int x, y; @} st = @{ x: 1, y: 1@};
484 struct @{ int x, y; @} st = @{ .x = 1, .y = 1@};
487 @item @code{\e} is ASCII character 27.
489 @item case ranges : ranges can be used in @code{case}s:
493 printf("range 1 to 9\n");
496 printf("unexpected\n");
501 @cindex aligned attribute
502 @cindex packed attribute
503 @cindex section attribute
504 @cindex unused attribute
505 @cindex cdecl attribute
506 @cindex stdcall attribute
507 @cindex regparm attribute
508 @cindex dllexport attribute
510 @item The keyword @code{__attribute__} is handled to specify variable or
511 function attributes. The following attributes are supported:
514 @item @code{aligned(n)}: align a variable or a structure field to n bytes
515 (must be a power of two).
517 @item @code{packed}: force alignment of a variable or a structure field to
520 @item @code{section(name)}: generate function or data in assembly section
521 name (name is a string containing the section name) instead of the default
524 @item @code{unused}: specify that the variable or the function is unused.
526 @item @code{cdecl}: use standard C calling convention (default).
528 @item @code{stdcall}: use Pascal-like calling convention.
530 @item @code{regparm(n)}: use fast i386 calling convention. @var{n} must be
531 between 1 and 3. The first @var{n} function parameters are respectively put in
532 registers @code{%eax}, @code{%edx} and @code{%ecx}.
534 @item @code{dllexport}: export function from dll/executable (win32 only)
538 Here are some examples:
540 int a __attribute__ ((aligned(8), section(".mysection")));
544 align variable @code{a} to 8 bytes and put it in section @code{.mysection}.
547 int my_add(int a, int b) __attribute__ ((section(".mycodesection")))
554 generate function @code{my_add} in section @code{.mycodesection}.
556 @item GNU style variadic macros:
558 #define dprintf(fmt, args@dots{}) printf(fmt, ## args)
561 dprintf("one arg %d\n", 1);
564 @item @code{__FUNCTION__} is interpreted as C99 @code{__func__}
565 (so it has not exactly the same semantics as string literal GNUC
566 where it is a string literal).
568 @item The @code{__alignof__} keyword can be used as @code{sizeof}
569 to get the alignment of a type or an expression.
571 @item The @code{typeof(x)} returns the type of @code{x}.
572 @code{x} is an expression or a type.
574 @item Computed gotos: @code{&&label} returns a pointer of type
575 @code{void *} on the goto label @code{label}. @code{goto *expr} can be
576 used to jump on the pointer resulting from @code{expr}.
578 @item Inline assembly with asm instruction:
579 @cindex inline assembly
580 @cindex assembly, inline
583 static inline void * my_memcpy(void * to, const void * from, size_t n)
586 __asm__ __volatile__(
591 "1:\ttestb $1,%b4\n\t"
595 : "=&c" (d0), "=&D" (d1), "=&S" (d2)
596 :"0" (n/4), "q" (n),"1" ((long) to),"2" ((long) from)
604 TCC includes its own x86 inline assembler with a @code{gas}-like (GNU
605 assembler) syntax. No intermediate files are generated. GCC 3.x named
606 operands are supported.
608 @item @code{__builtin_types_compatible_p()} and @code{__builtin_constant_p()}
611 @item @code{#pragma pack} is supported for win32 compatibility.
615 @section TinyCC extensions
619 @item @code{__TINYC__} is a predefined macro to indicate that you use TCC.
621 @item @code{#!} at the start of a line is ignored to allow scripting.
623 @item Binary digits can be entered (@code{0b101} instead of
626 @item @code{__BOUNDS_CHECKING_ON} is defined if bound checking is activated.
631 @chapter TinyCC Assembler
633 Since version 0.9.16, TinyCC integrates its own assembler. TinyCC
634 assembler supports a gas-like syntax (GNU assembler). You can
635 desactivate assembler support if you want a smaller TinyCC executable
636 (the C compiler does not rely on the assembler).
638 TinyCC Assembler is used to handle files with @file{.S} (C
639 preprocessed assembler) and @file{.s} extensions. It is also used to
640 handle the GNU inline assembler with the @code{asm} keyword.
644 TinyCC Assembler supports most of the gas syntax. The tokens are the
649 @item C and C++ comments are supported.
651 @item Identifiers are the same as C, so you cannot use '.' or '$'.
653 @item Only 32 bit integer numbers are supported.
661 @item Integers in decimal, octal and hexa are supported.
663 @item Unary operators: +, -, ~.
665 @item Binary operators in decreasing priority order:
673 @item A value is either an absolute number or a label plus an offset.
674 All operators accept absolute values except '+' and '-'. '+' or '-' can be
675 used to add an offset to a label. '-' supports two labels only if they
676 are the same or if they are both defined and in the same section.
684 @item All labels are considered as local, except undefined ones.
686 @item Numeric labels can be used as local @code{gas}-like labels.
687 They can be defined several times in the same source. Use 'b'
688 (backward) or 'f' (forward) as suffix to reference them:
692 jmp 1b /* jump to '1' label before */
693 jmp 1f /* jump to '1' label after */
700 @cindex assembler directives
701 @cindex directives, assembler
702 @cindex align directive
703 @cindex skip directive
704 @cindex space directive
705 @cindex byte directive
706 @cindex word directive
707 @cindex short directive
708 @cindex int directive
709 @cindex long directive
710 @cindex quad directive
711 @cindex globl directive
712 @cindex global directive
713 @cindex section directive
714 @cindex text directive
715 @cindex data directive
716 @cindex bss directive
717 @cindex fill directive
718 @cindex org directive
719 @cindex previous directive
720 @cindex string directive
721 @cindex asciz directive
722 @cindex ascii directive
724 All directives are preceded by a '.'. The following directives are
728 @item .align n[,value]
729 @item .skip n[,value]
730 @item .space n[,value]
731 @item .byte value1[,...]
732 @item .word value1[,...]
733 @item .short value1[,...]
734 @item .int value1[,...]
735 @item .long value1[,...]
736 @item .quad immediate_value1[,...]
739 @item .section section
743 @item .fill repeat[,size[,value]]
746 @item .string string[,...]
747 @item .asciz string[,...]
748 @item .ascii string[,...]
751 @section X86 Assembler
754 All X86 opcodes are supported. Only ATT syntax is supported (source
755 then destination operand order). If no size suffix is given, TinyCC
756 tries to guess it from the operand sizes.
758 Currently, MMX opcodes are supported but not SSE ones.
761 @chapter TinyCC Linker
764 @section ELF file generation
767 TCC can directly output relocatable ELF files (object files),
768 executable ELF files and dynamic ELF libraries without relying on an
771 Dynamic ELF libraries can be output but the C compiler does not generate
772 position independent code (PIC). It means that the dynamic library
773 code generated by TCC cannot be factorized among processes yet.
775 TCC linker eliminates unreferenced object code in libraries. A single pass is
776 done on the object and library list, so the order in which object files and
777 libraries are specified is important (same constraint as GNU ld). No grouping
778 options (@option{--start-group} and @option{--end-group}) are supported.
780 @section ELF file loader
782 TCC can load ELF object files, archives (.a files) and dynamic
785 @section PE-i386 file generation
788 TCC for Windows supports the native Win32 executable file format (PE-i386). It
789 generates EXE files (console and gui) and DLL files.
791 For usage on Windows, see also tcc-win32.txt.
793 @section GNU Linker Scripts
794 @cindex scripts, linker
795 @cindex linker scripts
796 @cindex GROUP, linker command
797 @cindex FILE, linker command
798 @cindex OUTPUT_FORMAT, linker command
799 @cindex TARGET, linker command
801 Because on many Linux systems some dynamic libraries (such as
802 @file{/usr/lib/libc.so}) are in fact GNU ld link scripts (horrible!),
803 the TCC linker also supports a subset of GNU ld scripts.
805 The @code{GROUP} and @code{FILE} commands are supported. @code{OUTPUT_FORMAT}
806 and @code{TARGET} are ignored.
808 Example from @file{/usr/lib/libc.so}:
811 Use the shared library, but some functions are only in
812 the static library, so try that secondarily. */
813 GROUP ( /lib/libc.so.6 /usr/lib/libc_nonshared.a )
817 @chapter TinyCC Memory and Bound checks
819 @cindex memory checks
821 This feature is activated with the @option{-b} (@pxref{Invoke}).
823 Note that pointer size is @emph{unchanged} and that code generated
824 with bound checks is @emph{fully compatible} with unchecked
825 code. When a pointer comes from unchecked code, it is assumed to be
826 valid. Even very obscure C code with casts should work correctly.
828 For more information about the ideas behind this method, see
829 @url{http://www.doc.ic.ac.uk/~phjk/BoundsChecking.html}.
831 Here are some examples of caught errors:
835 @item Invalid range with standard string function:
843 @item Out of bounds-error in global or local arrays:
853 @item Out of bounds-error in malloc'ed data:
857 tab = malloc(20 * sizeof(int));
865 @item Access of freed memory:
869 tab = malloc(20 * sizeof(int));
881 tab = malloc(20 * sizeof(int));
890 @chapter The @code{libtcc} library
892 The @code{libtcc} library enables you to use TCC as a backend for
893 dynamic code generation.
895 Read the @file{libtcc.h} to have an overview of the API. Read
896 @file{libtcc_test.c} to have a very simple example.
898 The idea consists in giving a C string containing the program you want
899 to compile directly to @code{libtcc}. Then you can access to any global
900 symbol (function or variable) defined.
903 @chapter Developer's guide
905 This chapter gives some hints to understand how TCC works. You can skip
906 it if you do not intend to modify the TCC code.
908 @section File reading
910 The @code{BufferedFile} structure contains the context needed to read a
911 file, including the current line number. @code{tcc_open()} opens a new
912 file and @code{tcc_close()} closes it. @code{inp()} returns the next
917 @code{next()} reads the next token in the current
918 file. @code{next_nomacro()} reads the next token without macro
921 @code{tok} contains the current token (see @code{TOK_xxx})
922 constants. Identifiers and keywords are also keywords. @code{tokc}
923 contains additional infos about the token (for example a constant value
924 if number or string token).
928 The parser is hardcoded (yacc is not necessary). It does only one pass,
933 @item For initialized arrays with unknown size, a first pass
934 is done to count the number of elements.
936 @item For architectures where arguments are evaluated in
937 reverse order, a first pass is done to reverse the argument order.
943 The types are stored in a single 'int' variable. It was chosen in the
944 first stages of development when tcc was much simpler. Now, it may not
945 be the best solution.
948 #define VT_INT 0 /* integer type */
949 #define VT_BYTE 1 /* signed byte type */
950 #define VT_SHORT 2 /* short type */
951 #define VT_VOID 3 /* void type */
952 #define VT_PTR 4 /* pointer */
953 #define VT_ENUM 5 /* enum definition */
954 #define VT_FUNC 6 /* function type */
955 #define VT_STRUCT 7 /* struct/union definition */
956 #define VT_FLOAT 8 /* IEEE float */
957 #define VT_DOUBLE 9 /* IEEE double */
958 #define VT_LDOUBLE 10 /* IEEE long double */
959 #define VT_BOOL 11 /* ISOC99 boolean type */
960 #define VT_LLONG 12 /* 64 bit integer */
961 #define VT_LONG 13 /* long integer (NEVER USED as type, only
963 #define VT_BTYPE 0x000f /* mask for basic type */
964 #define VT_UNSIGNED 0x0010 /* unsigned type */
965 #define VT_ARRAY 0x0020 /* array type (also has VT_PTR) */
966 #define VT_VLA 0x20000 /* VLA type (also has VT_PTR and VT_ARRAY) */
967 #define VT_BITFIELD 0x0040 /* bitfield modifier */
968 #define VT_CONSTANT 0x0800 /* const modifier */
969 #define VT_VOLATILE 0x1000 /* volatile modifier */
970 #define VT_SIGNED 0x2000 /* signed type */
972 #define VT_STRUCT_SHIFT 18 /* structure/enum name shift (14 bits left) */
975 When a reference to another type is needed (for pointers, functions and
976 structures), the @code{32 - VT_STRUCT_SHIFT} high order bits are used to
977 store an identifier reference.
979 The @code{VT_UNSIGNED} flag can be set for chars, shorts, ints and long
982 Arrays are considered as pointers @code{VT_PTR} with the flag
983 @code{VT_ARRAY} set. Variable length arrays are considered as special
984 arrays and have flag @code{VT_VLA} set instead of @code{VT_ARRAY}.
986 The @code{VT_BITFIELD} flag can be set for chars, shorts, ints and long
987 longs. If it is set, then the bitfield position is stored from bits
988 VT_STRUCT_SHIFT to VT_STRUCT_SHIFT + 5 and the bit field size is stored
989 from bits VT_STRUCT_SHIFT + 6 to VT_STRUCT_SHIFT + 11.
991 @code{VT_LONG} is never used except during parsing.
993 During parsing, the storage of an object is also stored in the type
997 #define VT_EXTERN 0x00000080 /* extern definition */
998 #define VT_STATIC 0x00000100 /* static variable */
999 #define VT_TYPEDEF 0x00000200 /* typedef definition */
1000 #define VT_INLINE 0x00000400 /* inline definition */
1001 #define VT_IMPORT 0x00004000 /* win32: extern data imported from dll */
1002 #define VT_EXPORT 0x00008000 /* win32: data exported from dll */
1003 #define VT_WEAK 0x00010000 /* win32: data exported from dll */
1008 All symbols are stored in hashed symbol stacks. Each symbol stack
1009 contains @code{Sym} structures.
1011 @code{Sym.v} contains the symbol name (remember
1012 an idenfier is also a token, so a string is never necessary to store
1013 it). @code{Sym.t} gives the type of the symbol. @code{Sym.r} is usually
1014 the register in which the corresponding variable is stored. @code{Sym.c} is
1015 usually a constant associated to the symbol like its address for normal
1016 symbols, and the number of entries for symbols representing arrays.
1017 Variable length array types use @code{Sym.c} as a location on the stack
1018 which holds the runtime sizeof for the type.
1020 Four main symbol stacks are defined:
1025 for the macros (@code{#define}s).
1028 for the global variables, functions and types.
1031 for the local variables, functions and types.
1033 @item global_label_stack
1034 for the local labels (for @code{goto}).
1037 for GCC block local labels (see the @code{__label__} keyword).
1041 @code{sym_push()} is used to add a new symbol in the local symbol
1042 stack. If no local symbol stack is active, it is added in the global
1045 @code{sym_pop(st,b)} pops symbols from the symbol stack @var{st} until
1046 the symbol @var{b} is on the top of stack. If @var{b} is NULL, the stack
1049 @code{sym_find(v)} return the symbol associated to the identifier
1050 @var{v}. The local stack is searched first from top to bottom, then the
1055 The generated code and datas are written in sections. The structure
1056 @code{Section} contains all the necessary information for a given
1057 section. @code{new_section()} creates a new section. ELF file semantics
1058 is assumed for each section.
1060 The following sections are predefined:
1065 is the section containing the generated code. @var{ind} contains the
1066 current position in the code section.
1069 contains initialized data
1072 contains uninitialized data
1074 @item bounds_section
1075 @itemx lbounds_section
1076 are used when bound checking is activated
1079 @itemx stabstr_section
1080 are used when debugging is actived to store debug information
1082 @item symtab_section
1083 @itemx strtab_section
1084 contain the exported symbols (currently only used for debugging).
1088 @section Code generation
1089 @cindex code generation
1091 @subsection Introduction
1093 The TCC code generator directly generates linked binary code in one
1094 pass. It is rather unusual these days (see gcc for example which
1095 generates text assembly), but it can be very fast and surprisingly
1098 The TCC code generator is register based. Optimization is only done at
1099 the expression level. No intermediate representation of expression is
1100 kept except the current values stored in the @emph{value stack}.
1102 On x86, three temporary registers are used. When more registers are
1103 needed, one register is spilled into a new temporary variable on the stack.
1105 @subsection The value stack
1106 @cindex value stack, introduction
1108 When an expression is parsed, its value is pushed on the value stack
1109 (@var{vstack}). The top of the value stack is @var{vtop}. Each value
1110 stack entry is the structure @code{SValue}.
1112 @code{SValue.t} is the type. @code{SValue.r} indicates how the value is
1113 currently stored in the generated code. It is usually a CPU register
1114 index (@code{REG_xxx} constants), but additional values and flags are
1118 #define VT_CONST 0x00f0
1119 #define VT_LLOCAL 0x00f1
1120 #define VT_LOCAL 0x00f2
1121 #define VT_CMP 0x00f3
1122 #define VT_JMP 0x00f4
1123 #define VT_JMPI 0x00f5
1124 #define VT_LVAL 0x0100
1125 #define VT_SYM 0x0200
1126 #define VT_MUSTCAST 0x0400
1127 #define VT_MUSTBOUND 0x0800
1128 #define VT_BOUNDED 0x8000
1129 #define VT_LVAL_BYTE 0x1000
1130 #define VT_LVAL_SHORT 0x2000
1131 #define VT_LVAL_UNSIGNED 0x4000
1132 #define VT_LVAL_TYPE (VT_LVAL_BYTE | VT_LVAL_SHORT | VT_LVAL_UNSIGNED)
1138 indicates that the value is a constant. It is stored in the union
1139 @code{SValue.c}, depending on its type.
1142 indicates a local variable pointer at offset @code{SValue.c.i} in the
1146 indicates that the value is actually stored in the CPU flags (i.e. the
1147 value is the consequence of a test). The value is either 0 or 1. The
1148 actual CPU flags used is indicated in @code{SValue.c.i}.
1150 If any code is generated which destroys the CPU flags, this value MUST be
1151 put in a normal register.
1155 indicates that the value is the consequence of a conditional jump. For VT_JMP,
1156 it is 1 if the jump is taken, 0 otherwise. For VT_JMPI it is inverted.
1158 These values are used to compile the @code{||} and @code{&&} logical
1161 If any code is generated, this value MUST be put in a normal
1162 register. Otherwise, the generated code won't be executed if the jump is
1166 is a flag indicating that the value is actually an lvalue (left value of
1167 an assignment). It means that the value stored is actually a pointer to
1170 Understanding the use @code{VT_LVAL} is very important if you want to
1171 understand how TCC works.
1174 @itemx VT_LVAL_SHORT
1175 @itemx VT_LVAL_UNSIGNED
1176 if the lvalue has an integer type, then these flags give its real
1177 type. The type alone is not enough in case of cast optimisations.
1180 is a saved lvalue on the stack. @code{VT_LLOCAL} should be eliminated
1181 ASAP because its semantics are rather complicated.
1184 indicates that a cast to the value type must be performed if the value
1185 is used (lazy casting).
1188 indicates that the symbol @code{SValue.sym} must be added to the constant.
1192 are only used for optional bound checking.
1196 @subsection Manipulating the value stack
1199 @code{vsetc()} and @code{vset()} pushes a new value on the value
1200 stack. If the previous @var{vtop} was stored in a very unsafe place(for
1201 example in the CPU flags), then some code is generated to put the
1202 previous @var{vtop} in a safe storage.
1204 @code{vpop()} pops @var{vtop}. In some cases, it also generates cleanup
1205 code (for example if stacked floating point registers are used as on
1208 The @code{gv(rc)} function generates code to evaluate @var{vtop} (the
1209 top value of the stack) into registers. @var{rc} selects in which
1210 register class the value should be put. @code{gv()} is the @emph{most
1211 important function} of the code generator.
1213 @code{gv2()} is the same as @code{gv()} but for the top two stack
1216 @subsection CPU dependent code generation
1217 @cindex CPU dependent
1218 See the @file{i386-gen.c} file to have an example.
1223 must generate the code needed to load a stack value into a register.
1226 must generate the code needed to store a register into a stack value
1230 @itemx gfunc_param()
1232 should generate a function call
1234 @item gfunc_prolog()
1235 @itemx gfunc_epilog()
1236 should generate a function prolog/epilog.
1239 must generate the binary integer operation @var{op} on the two top
1240 entries of the stack which are guaranted to contain integer types.
1242 The result value should be put on the stack.
1245 same as @code{gen_opi()} for floating point operations. The two top
1246 entries of the stack are guaranted to contain floating point values of
1249 @item gen_cvt_itof()
1250 integer to floating point conversion.
1252 @item gen_cvt_ftoi()
1253 floating point to integer conversion.
1255 @item gen_cvt_ftof()
1256 floating point to floating point of different size conversion.
1258 @item gen_bounded_ptr_add()
1259 @item gen_bounded_ptr_deref()
1260 are only used for bounds checking.
1264 @section Optimizations done
1265 @cindex optimizations
1266 @cindex constant propagation
1267 @cindex strength reduction
1268 @cindex comparison operators
1269 @cindex caching processor flags
1270 @cindex flags, caching
1271 @cindex jump optimization
1272 Constant propagation is done for all operations. Multiplications and
1273 divisions are optimized to shifts when appropriate. Comparison
1274 operators are optimized by maintaining a special cache for the
1275 processor flags. &&, || and ! are optimized by maintaining a special
1276 'jump target' value. No other jump optimization is currently performed
1277 because it would require to store the code in a more abstract fashion.
1279 @unnumbered Concept Index
1286 @c texinfo-column-for-description: 32