mirrors/cosmopolitan
1
0
Fork 0
mirror of https://github.com/jart/cosmopolitan.git synced 2025-01-31 03:27:39 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions
cosmopolitan/third_party/chibicc/help.txt
Justine Tunney a6baba1b07
Stop using .com extension in monorepo 
The WIN32 CreateProcess() function does not require an .exe or .com
suffix in order to spawn an executable. Now that we have Cosmo bash
we're no longer so dependent on the cmd.exe prompt.
2024-03-03 03:12:19 -08:00

662 lines
18 KiB
Text
Raw Permalink Blame History

This file contains invisible Unicode characters

This file contains invisible Unicode characters that are indistinguishable to humans but may be processed differently by a computer. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.

SYNOPSIS
chibicc [FLAGS] INPUTS
DESCRIPTION
chibicc - A Small GNU-Style ISO/IEC 9899:2011 Standard Compiler
OVERVIEW
chibicc is the simplest/tiniest/hackable/readable c11 compiler in the
world that can compile code quickly and consistently across platforms
FLAGS
-o PATH
Specifies path of output.
-c
Objectify the source file, but do not link.
-S
Compile the source file, but do not objectify.
-E
Preprocess the source file, but do not compile.
Output defaults to stdout.
-D TOKEN[=VALUE]
Defines preprocessor token.
-U TOKEN
Undefines preprocessor token.
-include DIR
Add include.
-I DIR
-iquote DIR
-isystem DIR
Adds include directory.
-x c (default for .c files)
-x assembler (default for .s files)
-x assembler-with-cpp (default for .S files)
Explicitly specifies programming language.
-Wa arg1[,arg2...]
-Xassembler arg1[,arg2...]
Appends opaque arguments passed along to assembler.
-Wl arg1[,arg2...]
-Xlinker arg1[,arg2...]
Appends opaque arguments passed along to linker.
-v
Enables verbose mode.
Lines with subprocess commands start with a space.
-###
Implies -v and enables shell command argument quoting.
--version
Shows compiler version.
--help
Shows this information.
CODEGEN
-pg
-mfentry
-mnop-mcount
-mrecord-mcount
Controls output of profiling hooks in function prologues.
-fdata-sections
-ffunction-sections
Controls granularity of code visible to the linker.
-msse3
-msse4
-msse4.1
-msse4.2
-mpopcnt
Specifies microarchitectural features. Default is K8.
-fpie
-fpic
-fPIC
Controls output of position independent code.
-fcommon
-fno-common
Controls usage of traditional STT_COMMON objects.
MAKEFILE
-M
Generate makefile header dependency code.
Output defaults to stdout.
-MD
Generate makefile header dependency code, and compile.
Output defaults to output path with .d extension.
-MMD
Generate makefile header dependency code, and compile.
Default include roots are excluded as dependencies.
Output defaults to output path with .d extension.
-MF PATH
Specifies output path of header dependency code.
-MT NAME
Specifies name of target in generated makefile code.
-MQ NAME
Specifies name of target in makefile code w/ quoting.
INTERNALS
-P
Generate Python bindings.
-A
Print abstract syntax tree.
-J
Generate HTML documentation for public APIs based on Javadoc
comments containing Markdown.
INTEGRAL TYPES
_Bool
char
short
int
long
long long
__int128
signed char
unsigned char
unsigned short
unsigned int
unsigned long
unsigned long long
unsigned __int128
FLOATING POINT TYPES
float
double
long double
BUILTIN FUNCTIONS
T __builtin_expect(T, int)
unsigned long __builtin_offsetof(typename, token)
int __builtin_reg_class(typename)
num __builtin_constant_p(expr)
int __builtin_unreachable()
void * __builtin_frame_address(int)
_Bool __builtin_types_compatible_p(typename, typename)
T __builtin_atomic_exchange(T *addr, T neu)
T * __builtin_assume_aligned(T *addr)
void * __builtin_alloca(unsigned long)
void __builtin_trap()
int __builtin_clz(int)
int __builtin_clzl(long)
int __builtin_clzll(long long)
int __builtin_ctz(int)
int __builtin_ctzl(long)
int __builtin_ctzll(long long)
int __builtin_ffs(int)
int __builtin_ffsl(long)
int __builtin_ffsll(long long)
int __builtin_popcount(unsigned int)
long __builtin_popcountl(unsigned long)
long __builtin_popcountll(unsigned long)
unsigned long __builtin_strlen(char *)
char * __builtin_strstr(char *, char *)
char * __builtin_strchr(char *, int)
void * __builtin_memcpy(void *, void *, unsigned long)
char * __builtin_strpbrk(char *, char *)
unsigned short __builtin_bswap16(unsigned short)
unsigned int __builtin_bswap32(unsigned int)
unsigned long __builtin_bswap64(unsigned long)
int __builtin_isnan(flonum)
int __builtin_isinf(flonum)
int __builtin_isfinite(flonum)
int __builtin_fpclassify(flonum)
int __builtin_isless(flonum, flonum)
int __builtin_isgreater(flonum, flonum)
int __builtin_isunordered(flonum, flonum)
int __builtin_islessequal(flonum, flonum)
int __builtin_islessgreater(flonum, flonum)
int __builtin_isgreaterequal(flonum, flonum)
double __builtin_nan(char *)
float __builtin_nanf(char *)
long double __builtin_nanl(char *)
long __builtin_signbit(double)
int __builtin_signbitf(float)
int __builtin_signbitl(long double)
double __builtin_huge_val()
float __builtin_huge_valf()
long double __builtin_huge_vall()
double __builtin_fabs(double)
float __builtin_fabsf(float)
long double __builtin_fabsl(long double)
double __builtin_logb(double)
float __builtin_logbf(float)
long double __builtin_logbl(long double)
double __builtin_fmax(double, double)
float __builtin_fmaxf(float, float)
long double __builtin_fmaxl(long double, long double)
double __builtin_fmin(double, double)
float __builtin_fminf(float, float)
long double __builtin_fminl(long double, long double)
double __builtin_copysign(double, double)
float __builtin_copysignf(float, float)
long double __builtin_copysignl(long double, long double)
void __builtin_ia32_movntq(di *, di)
int __builtin_ia32_pmovmskb128(v16qi)
T __atomic_load_n(T *addr, int memorder)
void __atomic_load(T *addr, T *res, int memorder)
void __atomic_store_n(T *addr, T val, int memorder)
void __atomic_store(T *addr, T *val, int memorder)
void __atomic_clear(_Bool *addr, int memorder)
_Bool __atomic_test_and_set(void *addr, int memorder)
T __atomic_fetch_add(T *addr, T amt, int memorder)
T __atomic_fetch_sub(T *addr, T amt, int memorder)
T __atomic_fetch_xor(T *addr, T amt, int memorder)
T __atomic_fetch_and(T *addr, T amt, int memorder)
T __atomic_fetch_or(T *addr, T amt, int memorder)
_Bool __atomic_compare_exchange_n(T *addr, T *old, T neu, int weak, int goodorder, int failorder)
T __sync_lock_test_and_set(T *addr, T value, ...)
void __sync_lock_release(T *addr, ...)
BUILTIN OBJECTS
__func__
__FUNCTION__
__va_area__
__alloca_size__
BUILTIN MACROS
__FILE__
__LINE__
__DATE__
__TIME__
__COUNTER__
__TIMESTAMP__
__BASE_FILE__
__chibicc__
__cosmopolitan__
__GNUC__
__GNUC_MINOR__
__GNUC_PATCHLEVEL__
__NO_INLINE__
__GNUC_STDC_INLINE__
__BIGGEST_ALIGNMENT__
__C99_MACRO_WITH_VA_ARGS
__GCC_ASM_FLAG_OUTPUTS__
__ELF__
__LP64__
_LP64
__STDC__
__STDC_HOSTED__
__STDC_NO_COMPLEX__
__STDC_UTF_16__
__STDC_UTF_32__
__STDC_VERSION__
__USER_LABEL_PREFIX__
__alignof__
__const__
__inline__
__signed__
__typeof__
__volatile__
__unix
__unix__
__linux
__linux__
__gnu_linux__
__BYTE_ORDER__
__FLOAT_WORD_ORDER__
__ORDER_BIG_ENDIAN__
__ORDER_LITTLE_ENDIAN__
__INT8_MAX__
__UINT8_MAX__
__INT16_MAX__
__UINT16_MAX__
__SHRT_MAX__
__INT_MAX__
__INT32_MAX__
__UINT32_MAX__
__INT64_MAX__
__LONG_MAX__
__LONG_LONG_MAX__
__UINT64_MAX__
__SIZE_MAX__
__INTPTR_MAX__
__UINTPTR_MAX__
__WINT_MAX__
__CHAR_BIT__
__SIZEOF_SHORT__
__SIZEOF_INT__
__SIZEOF_LONG__
__SIZEOF_LONG_LONG__
__SIZEOF_POINTER__
__SIZEOF_PTRDIFF_T__
__SIZEOF_SIZE_T__
__SIZEOF_WCHAR_T__
__SIZEOF_WINT_T__
__SIZEOF_FLOAT__
__SIZEOF_FLOAT128__
__SIZEOF_DOUBLE__
__SIZEOF_FLOAT80__
__SIZEOF_LONG_DOUBLE__
__INT8_TYPE__
__UINT8_TYPE__
__INT16_TYPE__
__UINT16_TYPE__
__INT32_TYPE__
__UINT32_TYPE__
__INT64_TYPE__
__UINT64_TYPE__
__INTPTR_TYPE__
__UINTPTR_TYPE__
__PTRDIFF_TYPE__
__SIZE_TYPE__
__WCHAR_TYPE__
__CHAR16_TYPE__
__CHAR32_TYPE__
__WINT_TYPE__
__CHAR16_TYPE__
__WCHAR_TYPE__
__CHAR32_TYPE__
__INT_LEAST8_TYPE__
__UINT_LEAST8_TYPE__
__INT_LEAST16_TYPE__
__UINT_LEAST16_TYPE__
__INT_LEAST32_TYPE__
__UINT_LEAST32_TYPE__
__INT_LEAST64_TYPE__
__UINT_LEAST64_TYPE__
__INT_FAST8_TYPE__
__UINT_FAST8_TYPE__
__INT_FAST16_TYPE__
__UINT_FAST16_TYPE__
__INT_FAST32_TYPE__
__UINT_FAST32_TYPE__
__INT_FAST64_TYPE__
__UINT_FAST64_TYPE__
__DBL_DECIMAL_DIG__
__DBL_DENORM_MIN__
__DBL_DIG__
__DBL_EPSILON__
__DBL_HAS_DENORM__
__DBL_HAS_INFINITY__
__DBL_HAS_QUIET_NAN__
__DBL_MANT_DIG__
__DBL_MAX_10_EXP__
__DBL_MAX_EXP__
__DBL_MAX__
__DBL_MIN_10_EXP__
__DBL_MIN_EXP__
__DBL_MIN__
__FLT_DECIMAL_DIG__
__FLT_DENORM_MIN__
__FLT_DIG__
__FLT_EPSILON__
__FLT_EVAL_METHOD_TS_18661_3__
__FLT_EVAL_METHOD__
__FLT_HAS_DENORM__
__FLT_HAS_INFINITY__
__FLT_HAS_QUIET_NAN__
__FLT_MANT_DIG__
__FLT_MAX_10_EXP__
__FLT_MAX_EXP__
__FLT_MAX__
__FLT_MIN_10_EXP__
__FLT_MIN_EXP__
__FLT_MIN__
__FLT_RADIX__
__LDBL_DECIMAL_DIG__
__LDBL_DENORM_MIN__
__LDBL_DIG__
__LDBL_EPSILON__
__LDBL_HAS_DENORM__
__LDBL_HAS_INFINITY__
__LDBL_HAS_QUIET_NAN__
__LDBL_MANT_DIG__
__LDBL_MAX_10_EXP__
__LDBL_MAX_EXP__
__LDBL_MAX__
__LDBL_MIN_10_EXP__
__LDBL_MIN_EXP__
__LDBL_MIN__
__x86_64
__x86_64__
__amd64
__amd64__
__MMX__
__SSE__
__SSE_MATH__
__SSE2__
__SSE2_MATH__
__SSE3__ [conditional]
__SSE4__ [conditional]
__POPCNT__ [conditional]
__PG__ [conditional]
__PIC__ [conditional]
__MFENTRY__ [conditional]
ASSEMBLER
That process is normally an implementation detail of your compiler,
which can embed this program or launch it as a subprocess. Much GNU
style syntax is supported. Your code that gets embedded in an asm()
statement will ultimately end up here. This implementation, has the
advantage of behaving the same across platforms, in a simple single
file implementation that compiles down to a 100kilo ape executable.
Your assembler supports the following flags:
-o FILE output path [default: a.out]
-I DIR append include path [default: .]
-W inhibit .warning
-Z inhibit .error and .err
Your assembler supports the following directives:
.zero N emits 0's
.byte INT... emits int8
.word INT... emits int16
.long INT... emits int32
.quad INT... emits int64
.octa INT... emits int128
.ascii STR... emits string
.asciz STR... emits string and 0 byte
.ident STR emits string to .comment section
.float NUMBER... emits binary32
.double NUMBER... emits binary64
.float80 NUMBER... emits x86 float (10 bytes)
.ldbl NUMBER... emits x86 float (16 bytes)
.zleb128 NUMBER... emits LEB-128 zigzag varint
.sleb128 NUMBER... emits LEB-128 signed varint
.uleb128 NUMBER... emits LEB-128 unsigned varint
.align BYTES [FILL [MAXSKIP]] emits fill bytes to boundary
.end halts tokenization
.abort crashes assembler
.err aborts (ignorable w/ -Z)
.error STR aborts (ignorable w/ -Z)
.warning STR whines (ignorable w/ -W)
.text enters text section (default)
.data enters data section
.bss enters bss section
.section NAME [SFLG SHT] enters section
.previous enters previous section
.pushsection NAME [SFLG SHT] pushes section
.popsection pops section
.type SYM TYPE sets type of symbol
.size SYM EXPR sets size of symbol
.internal SYM... marks symbol STV_INTERNAL
.hidden SYM... marks symbol STV_HIDDEN
.protected SYM... marks symbol STV_PROTECTED
.globl SYM... marks symbol STB_GLOBAL
.local SYM... marks symbol STB_LOCAL
.weak SYM... marks symbol STB_WEAK
.include FILE assembles file source
.incbin FILE emits file content
.file FILENO PATH dwarf file define
.loc FILENO LINENO dwarf source line
TYPE can be one of the following:
- @notype STT_NOTYPE (default)
- @object STT_OBJECT
- @function STT_FUNC
- @common STT_COMMON
- @tls_object STT_TLS
SHT can be one of the following:
- @progbits SHT_PROGBITS
- @note SHT_NOTE
- @nobits SHT_NOBITS
- @preinit_array SHT_PREINIT_ARRAY
- @init_array SHT_INIT_ARRAY
- @fini_array SHT_FINI_ARRAY
SFLG is a string which may have the following characters:
- a SHF_ALLOC
- w SHF_WRITE
- x SHF_EXECINSTR
- g SHF_GROUP
- M SHF_MERGE
- S SHF_STRINGS
- T SHF_TLS
PREFIXES
addr32 cs data16 ds es fs
gs lock rep repe repne repnz
repz rex rex.b rex.r rex.rb rex.rx
rex.rxb rex.w rex.wb rex.wr rex.wrb rex.wrx
rex.wrxb rex.wx rex.wxb rex.x rex.xb ss
REGISTERS
64-bit 32-bit 16-bit lo byte hi byte │ sse mmx │ fpu
─────── ─────── ─────── ─────── ─────── │ ─────── ─────── │ ───────
%rax %eax %ax %al %ah │ %xmm0 %mm0 │ %st(0)
%rcx %ecx %cx %cl %ch │ %xmm1 %mm1 │ %st(1)
%rdx %edx %dx %dl %dh │ %xmm2 %mm2 │ %st(2)
%rbx %ebx %bx %bl %bh │ %xmm3 %mm3 │ %st(3)
%rsp %esp %sp %spl │ %xmm4 %mm4 │ %st(4)
%rbp %ebp %bp %bpl │ %xmm5 %mm5 │ %st(5)
%rsi %esi %si %sil │ %xmm6 %mm6 │ %st(6)
%rdi %edi %di %dil │ %xmm7 %mm7 │ %st(7)
%r8 %r8d %r8w %r8b │ %xmm8
%r9 %r9d %r9w %r9b │ %xmm9
%r10 %r10d %r10w %r10b │ %xmm10
%r11 %r11d %r11w %r11b │ %xmm11
%r12 %r12d %r12w %r12b │ %xmm12
%r13 %r13d %r13w %r13b │ %xmm13
%r14 %r14d %r14w %r14b │ %xmm14
%r15 %r15d %r15w %r15b │ %xmm15
%riz %eiz
RICHARD STALLMAN MATH55 ASM() NOTATION
BEHAVIOR
=: write-only
+: read/writeable
SELECTION
Autonomous
a: ax/eax/rax
b: bx/ebx/rbx
c: bx/ebx/rbx
d: dx/edx/rdx
S: si/esi/rsi
D: di/edi/rdi
Algorithmic
r: pick one of a,b,c,d,D,S,bp,sp,r8-15 registers, referenced as %0,etc.
l: pick one of a,b,c,d,D,S,bp,r8-15 for indexing, referenced as %0,etc.
q: pick one of a,b,c,d,r8-r15 for lo-byte access, e.g. %b0,%w0,%k0,etc.
Q: pick one of a,b,c,d for hi-byte access, e.g. %h0,etc.
U: pick one of a,c,d,D,S,r8-11 (call-clobbered)
R: pick one of a,b,c,d,D,S,bp,sp (all models)
y: pick mmx register
x: pick sse register
X: pick anything
m: memory
o: memory offsetable by an immediate, referenced as %0,2+%0,etc.
p: memory, intended for load/push address and segments (movl %@:%p1, %0)
g: probably shorthand for "rmi" combo
X: allow anything
Combos
rm: pick register or memory address (converting immediates)
rmi: pick register or memory address (allowing immediates)
Immediates
i: integer literal or compiler/assembler constexpr or linker embedding
e: i∊[-2^31,2^31) for sign-extending immediates
Z: i∊[0,2^32) for zero-extending immediates
I: i∊[0,31] (5 bits for 32-bit shifts)
J: i∊[0,63] (6 bits for 64-bit shifts)
K: i∊[-128,127]
L: permit uncasted char/short literal as zero-extended operand to andl?
M: i∊[0,3] (intended for index scaling, e.g. "mov\t(?,?,1<<%0),?")
N: i∊[0,255] (for in & out)
M: 2-bit integer constant (shifts for index scaling)
I: 5-bit integer constant (for 32-bit shifts)
J: 6-bit integer constant (for 64-bit shifts)
Transcendentals
f: any stack slot
t: top of stack (%st(0)) and possibly converts xmm to ymm
u: second top of stack (%st(1))
Specials
%= generates number unique to each instance
%%REG explicitly-supplied register (used w/ clobbers)
AUGMENTATION
%pN print raw
%PN print w/ @plt
%aN print address
%zN print only opcode suffix for operand type
%lN print label without punctuation, e.g. jumps
%cN print immediate w/o punctuation, e.g. lea %c0(%1),%2
%bN print lo-byte form, e.g. xchgb %b0,%%al (QImode 8-bit)
%hN print hi-byte form, e.g. xchgb %h0,%%ah (QImode 8-bit)
%wN print word form, e.g. xchgw %w0,%%ax (HImode 16-bit)
%kN print dword form, e.g. xchgl %k0,%%eax (SImode 32-bit)
%qN print qword form, e.g. xchgq %q0,%%rax (DImode 64-bit)
%HN access high 8 bytes of SSE register, or +8 displacement
%nN negated literal, e.g. lea %n0(%1),%2
%VN print register name without %, e.g. call foo%V0
EXAMPLE
static inline void MixAudio(short a[static 8], const short b[static 8]) {
asm("paddsw\t%1,%0"
: "+x"(a)
: "x"(b)
: "memory");
}
Reference in a new issue View git blame Copy permalink