Leahnaya/SDL
1
0
Fork 0
mirror of https://github.com/yawut/SDL.git synced 2026-06-16 21:49:33 -05:00
Code Issues Actions Packages Projects Releases Wiki Activity
cc1860f252
SDL/src/stdlib/SDL_stdlib.c
Ryan C. Gordon 1fa34e72be Improvements to stdlib. 
All SDL_* functions are always available as real symbols, so you can always
link against them as a stable ABI. By default, however, all the things that
might have dithered down to macros in your application are now force-inlined,
to give you the same effect as before and theoretically better performance,
but still solve the classic macro problems.

Elsewhere, we provide real functions for these things that simply wrap the
inline functions, in case one needs to have a real function available.

Also: this exposed bugs: SDL_abs() does something different if you had the
macro vs the libc function, SDL_memcpy() returns a void* in the function
but not the macro, etc.
2013-03-15 01:01:20 -04:00

762 lines
19 KiB
C
Raw Blame History

/*
Simple DirectMedia Layer
Copyright (C) 1997-2013 Sam Lantinga <slouken@libsdl.org>
This software is provided 'as-is', without any express or implied
warranty. In no event will the authors be held liable for any damages
arising from the use of this software.
Permission is granted to anyone to use this software for any purpose,
including commercial applications, and to alter it and redistribute it
freely, subject to the following restrictions:
1. The origin of this software must not be misrepresented; you must not
claim that you wrote the original software. If you use this software
in a product, an acknowledgment in the product documentation would be
appreciated but is not required.
2. Altered source versions must be plainly marked as such, and must not be
misrepresented as being the original software.
3. This notice may not be removed or altered from any source distribution.
*/
#include "SDL_config.h"
/* This file contains portable stdlib functions for SDL */
#include "SDL_stdinc.h"
/* these are always #defined, make them real symbol in the library, too... */
#undef SDL_ceil
#undef SDL_abs
#undef SDL_sinf
#undef SDL_cosf
#undef SDL_isdigit
#undef SDL_isspace
#undef SDL_toupper
#undef SDL_tolower
double SDL_ceil(double x) { return SDL_ceil_inline(x); }
float SDL_cosf(float x) { return SDL_cosf_inline(x); }
float SDL_sinf(float x) { return SDL_sinf_inline(x); }
int SDL_abs(int x) { return SDL_abs_inline(x); }
int SDL_isdigit(int x) { return SDL_isdigit_inline(x); }
int SDL_isspace(int x) { return SDL_isspace_inline(x); }
int SDL_toupper(int x) { return SDL_toupper_inline(x); }
int SDL_tolower(int x) { return SDL_tolower_inline(x); }
#ifndef HAVE_LIBC
/* These are some C runtime intrinsics that need to be defined */
#if defined(_MSC_VER)
#ifndef __FLTUSED__
#define __FLTUSED__
__declspec(selectany) int _fltused = 1;
#endif
/* The optimizer on Visual Studio 2010/2012 generates memcpy() calls */
#if _MSC_VER >= 1600 && defined(_WIN64) && !defined(_DEBUG)
#include <intrin.h>
#pragma function(memcpy)
void * memcpy ( void * destination, const void * source, size_t num )
{
const Uint8 *src = (const Uint8 *)source;
Uint8 *dst = (Uint8 *)destination;
size_t i;
/* All WIN64 architectures have SSE, right? */
if (!((uintptr_t) src & 15) && !((uintptr_t) dst & 15)) {
__m128 values[4];
for (i = num / 64; i--;) {
_mm_prefetch(src, _MM_HINT_NTA);
values[0] = *(__m128 *) (src + 0);
values[1] = *(__m128 *) (src + 16);
values[2] = *(__m128 *) (src + 32);
values[3] = *(__m128 *) (src + 48);
_mm_stream_ps((float *) (dst + 0), values[0]);
_mm_stream_ps((float *) (dst + 16), values[1]);
_mm_stream_ps((float *) (dst + 32), values[2]);
_mm_stream_ps((float *) (dst + 48), values[3]);
src += 64;
dst += 64;
}
num &= 63;
}
while (num--) {
*dst++ = *src++;
}
return destination;
}
#endif /* _MSC_VER == 1600 && defined(_WIN64) && !defined(_DEBUG) */
#ifdef _M_IX86
void
__declspec(naked)
_chkstk()
{
}
/* Float to long */
void
__declspec(naked)
_ftol()
{
/* *INDENT-OFF* */
__asm {
push ebp
mov ebp,esp
sub esp,20h
and esp,0FFFFFFF0h
fld st(0)
fst dword ptr [esp+18h]
fistp qword ptr [esp+10h]
fild qword ptr [esp+10h]
mov edx,dword ptr [esp+18h]
mov eax,dword ptr [esp+10h]
test eax,eax
je integer_QnaN_or_zero
arg_is_not_integer_QnaN:
fsubp st(1),st
test edx,edx
jns positive
fstp dword ptr [esp]
mov ecx,dword ptr [esp]
xor ecx,80000000h
add ecx,7FFFFFFFh
adc eax,0
mov edx,dword ptr [esp+14h]
adc edx,0
jmp localexit
positive:
fstp dword ptr [esp]
mov ecx,dword ptr [esp]
add ecx,7FFFFFFFh
sbb eax,0
mov edx,dword ptr [esp+14h]
sbb edx,0
jmp localexit
integer_QnaN_or_zero:
mov edx,dword ptr [esp+14h]
test edx,7FFFFFFFh
jne arg_is_not_integer_QnaN
fstp dword ptr [esp+18h]
fstp dword ptr [esp+18h]
localexit:
leave
ret
}
/* *INDENT-ON* */
}
void
_ftol2_sse()
{
_ftol();
}
/* 64-bit math operators for 32-bit systems */
void
__declspec(naked)
_allmul()
{
/* *INDENT-OFF* */
__asm {
push ebp
mov ebp,esp
push edi
push esi
push ebx
sub esp,0Ch
mov eax,dword ptr [ebp+10h]
mov edi,dword ptr [ebp+8]
mov ebx,eax
mov esi,eax
sar esi,1Fh
mov eax,dword ptr [ebp+8]
mul ebx
imul edi,esi
mov ecx,edx
mov dword ptr [ebp-18h],eax
mov edx,dword ptr [ebp+0Ch]
add ecx,edi
imul ebx,edx
mov eax,dword ptr [ebp-18h]
lea ebx,[ebx+ecx]
mov dword ptr [ebp-14h],ebx
mov edx,dword ptr [ebp-14h]
add esp,0Ch
pop ebx
pop esi
pop edi
pop ebp
ret
}
/* *INDENT-ON* */
}
void
__declspec(naked)
_alldiv()
{
/* *INDENT-OFF* */
__asm {
push edi
push esi
push ebx
xor edi,edi
mov eax,dword ptr [esp+14h]
or eax,eax
jge L1
inc edi
mov edx,dword ptr [esp+10h]
neg eax
neg edx
sbb eax,0
mov dword ptr [esp+14h],eax
mov dword ptr [esp+10h],edx
L1:
mov eax,dword ptr [esp+1Ch]
or eax,eax
jge L2
inc edi
mov edx,dword ptr [esp+18h]
neg eax
neg edx
sbb eax,0
mov dword ptr [esp+1Ch],eax
mov dword ptr [esp+18h],edx
L2:
or eax,eax
jne L3
mov ecx,dword ptr [esp+18h]
mov eax,dword ptr [esp+14h]
xor edx,edx
div ecx
mov ebx,eax
mov eax,dword ptr [esp+10h]
div ecx
mov edx,ebx
jmp L4
L3:
mov ebx,eax
mov ecx,dword ptr [esp+18h]
mov edx,dword ptr [esp+14h]
mov eax,dword ptr [esp+10h]
L5:
shr ebx,1
rcr ecx,1
shr edx,1
rcr eax,1
or ebx,ebx
jne L5
div ecx
mov esi,eax
mul dword ptr [esp+1Ch]
mov ecx,eax
mov eax,dword ptr [esp+18h]
mul esi
add edx,ecx
jb L6
cmp edx,dword ptr [esp+14h]
ja L6
jb L7
cmp eax,dword ptr [esp+10h]
jbe L7
L6:
dec esi
L7:
xor edx,edx
mov eax,esi
L4:
dec edi
jne L8
neg edx
neg eax
sbb edx,0
L8:
pop ebx
pop esi
pop edi
ret 10h
}
/* *INDENT-ON* */
}
void
__declspec(naked)
_aulldiv()
{
/* *INDENT-OFF* */
__asm {
push ebx
push esi
mov eax,dword ptr [esp+18h]
or eax,eax
jne L1
mov ecx,dword ptr [esp+14h]
mov eax,dword ptr [esp+10h]
xor edx,edx
div ecx
mov ebx,eax
mov eax,dword ptr [esp+0Ch]
div ecx
mov edx,ebx
jmp L2
L1:
mov ecx,eax
mov ebx,dword ptr [esp+14h]
mov edx,dword ptr [esp+10h]
mov eax,dword ptr [esp+0Ch]
L3:
shr ecx,1
rcr ebx,1
shr edx,1
rcr eax,1
or ecx,ecx
jne L3
div ebx
mov esi,eax
mul dword ptr [esp+18h]
mov ecx,eax
mov eax,dword ptr [esp+14h]
mul esi
add edx,ecx
jb L4
cmp edx,dword ptr [esp+10h]
ja L4
jb L5
cmp eax,dword ptr [esp+0Ch]
jbe L5
L4:
dec esi
L5:
xor edx,edx
mov eax,esi
L2:
pop esi
pop ebx
ret 10h
}
/* *INDENT-ON* */
}
void
__declspec(naked)
_allrem()
{
/* *INDENT-OFF* */
__asm {
push ebx
push edi
xor edi,edi
mov eax,dword ptr [esp+10h]
or eax,eax
jge L1
inc edi
mov edx,dword ptr [esp+0Ch]
neg eax
neg edx
sbb eax,0
mov dword ptr [esp+10h],eax
mov dword ptr [esp+0Ch],edx
L1:
mov eax,dword ptr [esp+18h]
or eax,eax
jge L2
mov edx,dword ptr [esp+14h]
neg eax
neg edx
sbb eax,0
mov dword ptr [esp+18h],eax
mov dword ptr [esp+14h],edx
L2:
or eax,eax
jne L3
mov ecx,dword ptr [esp+14h]
mov eax,dword ptr [esp+10h]
xor edx,edx
div ecx
mov eax,dword ptr [esp+0Ch]
div ecx
mov eax,edx
xor edx,edx
dec edi
jns L4
jmp L8
L3:
mov ebx,eax
mov ecx,dword ptr [esp+14h]
mov edx,dword ptr [esp+10h]
mov eax,dword ptr [esp+0Ch]
L5:
shr ebx,1
rcr ecx,1
shr edx,1
rcr eax,1
or ebx,ebx
jne L5
div ecx
mov ecx,eax
mul dword ptr [esp+18h]
xchg eax,ecx
mul dword ptr [esp+14h]
add edx,ecx
jb L6
cmp edx,dword ptr [esp+10h]
ja L6
jb L7
cmp eax,dword ptr [esp+0Ch]
jbe L7
L6:
sub eax,dword ptr [esp+14h]
sbb edx,dword ptr [esp+18h]
L7:
sub eax,dword ptr [esp+0Ch]
sbb edx,dword ptr [esp+10h]
dec edi
jns L8
L4:
neg edx
neg eax
sbb edx,0
L8:
pop edi
pop ebx
ret 10h
}
/* *INDENT-ON* */
}
void
__declspec(naked)
_aullrem()
{
/* *INDENT-OFF* */
__asm {
push ebx
mov eax,dword ptr [esp+14h]
or eax,eax
jne L1
mov ecx,dword ptr [esp+10h]
mov eax,dword ptr [esp+0Ch]
xor edx,edx
div ecx
mov eax,dword ptr [esp+8]
div ecx
mov eax,edx
xor edx,edx
jmp L2
L1:
mov ecx,eax
mov ebx,dword ptr [esp+10h]
mov edx,dword ptr [esp+0Ch]
mov eax,dword ptr [esp+8]
L3:
shr ecx,1
rcr ebx,1
shr edx,1
rcr eax,1
or ecx,ecx
jne L3
div ebx
mov ecx,eax
mul dword ptr [esp+14h]
xchg eax,ecx
mul dword ptr [esp+10h]
add edx,ecx
jb L4
cmp edx,dword ptr [esp+0Ch]
ja L4
jb L5
cmp eax,dword ptr [esp+8]
jbe L5
L4:
sub eax,dword ptr [esp+10h]
sbb edx,dword ptr [esp+14h]
L5:
sub eax,dword ptr [esp+8]
sbb edx,dword ptr [esp+0Ch]
neg edx
neg eax
sbb edx,0
L2:
pop ebx
ret 10h
}
/* *INDENT-ON* */
}
void
__declspec(naked)
_alldvrm()
{
/* *INDENT-OFF* */
__asm {
push edi
push esi
push ebp
xor edi,edi
xor ebp,ebp
mov eax,dword ptr [esp+14h]
or eax,eax
jge L1
inc edi
inc ebp
mov edx,dword ptr [esp+10h]
neg eax
neg edx
sbb eax,0
mov dword ptr [esp+14h],eax
mov dword ptr [esp+10h],edx
L1:
mov eax,dword ptr [esp+1Ch]
or eax,eax
jge L2
inc edi
mov edx,dword ptr [esp+18h]
neg eax
neg edx
sbb eax,0
mov dword ptr [esp+1Ch],eax
mov dword ptr [esp+18h],edx
L2:
or eax,eax
jne L3
mov ecx,dword ptr [esp+18h]
mov eax,dword ptr [esp+14h]
xor edx,edx
div ecx
mov ebx,eax
mov eax,dword ptr [esp+10h]
div ecx
mov esi,eax
mov eax,ebx
mul dword ptr [esp+18h]
mov ecx,eax
mov eax,esi
mul dword ptr [esp+18h]
add edx,ecx
jmp L4
L3:
mov ebx,eax
mov ecx,dword ptr [esp+18h]
mov edx,dword ptr [esp+14h]
mov eax,dword ptr [esp+10h]
L5:
shr ebx,1
rcr ecx,1
shr edx,1
rcr eax,1
or ebx,ebx
jne L5
div ecx
mov esi,eax
mul dword ptr [esp+1Ch]
mov ecx,eax
mov eax,dword ptr [esp+18h]
mul esi
add edx,ecx
jb L6
cmp edx,dword ptr [esp+14h]
ja L6
jb L7
cmp eax,dword ptr [esp+10h]
jbe L7
L6:
dec esi
sub eax,dword ptr [esp+18h]
sbb edx,dword ptr [esp+1Ch]
L7:
xor ebx,ebx
L4:
sub eax,dword ptr [esp+10h]
sbb edx,dword ptr [esp+14h]
dec ebp
jns L9
neg edx
neg eax
sbb edx,0
L9:
mov ecx,edx
mov edx,ebx
mov ebx,ecx
mov ecx,eax
mov eax,esi
dec edi
jne L8
neg edx
neg eax
sbb edx,0
L8:
pop ebp
pop esi
pop edi
ret 10h
}
/* *INDENT-ON* */
}
void
__declspec(naked)
_aulldvrm()
{
/* *INDENT-OFF* */
__asm {
push esi
mov eax,dword ptr [esp+14h]
or eax,eax
jne L1
mov ecx,dword ptr [esp+10h]
mov eax,dword ptr [esp+0Ch]
xor edx,edx
div ecx
mov ebx,eax
mov eax,dword ptr [esp+8]
div ecx
mov esi,eax
mov eax,ebx
mul dword ptr [esp+10h]
mov ecx,eax
mov eax,esi
mul dword ptr [esp+10h]
add edx,ecx
jmp L2
L1:
mov ecx,eax
mov ebx,dword ptr [esp+10h]
mov edx,dword ptr [esp+0Ch]
mov eax,dword ptr [esp+8]
L3:
shr ecx,1
rcr ebx,1
shr edx,1
rcr eax,1
or ecx,ecx
jne L3
div ebx
mov esi,eax
mul dword ptr [esp+14h]
mov ecx,eax
mov eax,dword ptr [esp+10h]
mul esi
add edx,ecx
jb L4
cmp edx,dword ptr [esp+0Ch]
ja L4
jb L5
cmp eax,dword ptr [esp+8]
jbe L5
L4:
dec esi
sub eax,dword ptr [esp+10h]
sbb edx,dword ptr [esp+14h]
L5:
xor ebx,ebx
L2:
sub eax,dword ptr [esp+8]
sbb edx,dword ptr [esp+0Ch]
neg edx
neg eax
sbb edx,0
mov ecx,edx
mov edx,ebx
mov ebx,ecx
mov ecx,eax
mov eax,esi
pop esi
ret 10h
}
/* *INDENT-ON* */
}
void
__declspec(naked)
_allshl()
{
/* *INDENT-OFF* */
__asm {
cmp cl,40h
jae RETZERO
cmp cl,20h
jae MORE32
shld edx,eax,cl
shl eax,cl
ret
MORE32:
mov edx,eax
xor eax,eax
and cl,1Fh
shl edx,cl
ret
RETZERO:
xor eax,eax
xor edx,edx
ret
}
/* *INDENT-ON* */
}
void
__declspec(naked)
_allshr()
{
/* *INDENT-OFF* */
__asm {
cmp cl,40h
jae RETZERO
cmp cl,20h
jae MORE32
shrd eax,edx,cl
sar edx,cl
ret
MORE32:
mov eax,edx
xor edx,edx
and cl,1Fh
sar eax,cl
ret
RETZERO:
xor eax,eax
xor edx,edx
ret
}
/* *INDENT-ON* */
}
void
__declspec(naked)
_aullshr()
{
/* *INDENT-OFF* */
__asm {
cmp cl,40h
jae RETZERO
cmp cl,20h
jae MORE32
shrd eax,edx,cl
shr edx,cl
ret
MORE32:
mov eax,edx
xor edx,edx
and cl,1Fh
shr eax,cl
ret
RETZERO:
xor eax,eax
xor edx,edx
ret
}
/* *INDENT-ON* */
}
#endif /* _M_IX86 */
#endif /* MSC_VER */
#endif /* !HAVE_LIBC */
/* vi: set ts=4 sw=4 expandtab: */
Reference in New Issue View Git Blame Copy Permalink