mirror of
https://github.com/DrBeef/Raze.git
synced 2024-11-15 17:01:51 +00:00
6299d2fe86
Since no division function was handled here this is totally pointless on modern systems and will even on 32 bit cause performance degradation due to lack of optimizability. The rest of that code in pragmas.h and pragmas.cpp should probably also be put under review. Much of it made sense 20 years ago but not anymore.
270 lines
8.5 KiB
C++
270 lines
8.5 KiB
C++
// "Build Engine & Tools" Copyright (c) 1993-1997 Ken Silverman
|
|
// Ken Silverman's official web site: "http://www.advsys.net/ken"
|
|
// See the included license file "BUILDLIC.TXT" for license info.
|
|
//
|
|
// This file has been modified from Ken Silverman's original release
|
|
// by Jonathon Fowler (jf@jonof.id.au)
|
|
// by the EDuke32 team (development@voidpoint.com)
|
|
|
|
#ifndef pragmas_h_
|
|
#define pragmas_h_
|
|
|
|
#ifdef __cplusplus
|
|
extern "C" {
|
|
#endif
|
|
|
|
#define EDUKE32_GENERATE_PRAGMAS \
|
|
EDUKE32_SCALER_PRAGMA(1) EDUKE32_SCALER_PRAGMA(2) EDUKE32_SCALER_PRAGMA(3) EDUKE32_SCALER_PRAGMA(4) \
|
|
EDUKE32_SCALER_PRAGMA(5) EDUKE32_SCALER_PRAGMA(6) EDUKE32_SCALER_PRAGMA(7) EDUKE32_SCALER_PRAGMA(8) \
|
|
EDUKE32_SCALER_PRAGMA(9) EDUKE32_SCALER_PRAGMA(10) EDUKE32_SCALER_PRAGMA(11) EDUKE32_SCALER_PRAGMA(12) \
|
|
EDUKE32_SCALER_PRAGMA(13) EDUKE32_SCALER_PRAGMA(14) EDUKE32_SCALER_PRAGMA(15) EDUKE32_SCALER_PRAGMA(16) \
|
|
EDUKE32_SCALER_PRAGMA(17) EDUKE32_SCALER_PRAGMA(18) EDUKE32_SCALER_PRAGMA(19) EDUKE32_SCALER_PRAGMA(20) \
|
|
EDUKE32_SCALER_PRAGMA(21) EDUKE32_SCALER_PRAGMA(22) EDUKE32_SCALER_PRAGMA(23) EDUKE32_SCALER_PRAGMA(24) \
|
|
EDUKE32_SCALER_PRAGMA(25) EDUKE32_SCALER_PRAGMA(26) EDUKE32_SCALER_PRAGMA(27) EDUKE32_SCALER_PRAGMA(28) \
|
|
EDUKE32_SCALER_PRAGMA(29) EDUKE32_SCALER_PRAGMA(30) EDUKE32_SCALER_PRAGMA(31)
|
|
|
|
#if !defined(NOASM) && defined __cplusplus
|
|
extern "C" {
|
|
#endif
|
|
extern int32_t reciptable[2048], fpuasm;
|
|
#if !defined(NOASM) && defined __cplusplus
|
|
}
|
|
#endif
|
|
|
|
// break the C version of divscale out from the others
|
|
// because asm version overflows in drawmapview()
|
|
|
|
#define qw(x) ((int64_t)(x)) // quadword cast
|
|
#define dw(x) ((int32_t)(x)) // doubleword cast
|
|
#define wo(x) ((int16_t)(x)) // word cast
|
|
#define by(x) ((uint8_t)(x)) // byte cast
|
|
|
|
#define DIVTABLESIZE 16384
|
|
|
|
extern libdivide::libdivide_s64_t divtable64[DIVTABLESIZE];
|
|
extern libdivide::libdivide_s32_t divtable32[DIVTABLESIZE];
|
|
extern void initdivtables(void);
|
|
|
|
static inline uint32_t divideu32(uint32_t const n, uint32_t const d)
|
|
{
|
|
static libdivide::libdivide_u32_t udiv;
|
|
static uint32_t lastd;
|
|
|
|
if (d == lastd)
|
|
goto skip;
|
|
|
|
udiv = libdivide::libdivide_u32_gen((lastd = d));
|
|
skip:
|
|
return libdivide::libdivide_u32_do(n, &udiv);
|
|
}
|
|
|
|
static inline int64_t tabledivide64(int64_t const n, int64_t const d)
|
|
{
|
|
static libdivide::libdivide_s64_t sdiv;
|
|
static int32_t lastd;
|
|
auto const dptr = ((unsigned)d < DIVTABLESIZE) ? &divtable64[d] : &sdiv;
|
|
|
|
if (d == lastd || dptr != &sdiv)
|
|
goto skip;
|
|
|
|
sdiv = libdivide::libdivide_s64_gen((lastd = d));
|
|
skip:
|
|
return libdivide::libdivide_s64_do(n, dptr);
|
|
}
|
|
|
|
static inline int32_t tabledivide32(int32_t const n, int32_t const d)
|
|
{
|
|
static libdivide::libdivide_s32_t sdiv;
|
|
static int32_t lastd;
|
|
auto const dptr = ((unsigned)d < DIVTABLESIZE) ? &divtable32[d] : &sdiv;
|
|
|
|
if (d == lastd || dptr != &sdiv)
|
|
goto skip;
|
|
|
|
sdiv = libdivide::libdivide_s32_gen((lastd = d));
|
|
skip:
|
|
return libdivide::libdivide_s32_do(n, dptr);
|
|
}
|
|
|
|
extern uint32_t divideu32_noinline(uint32_t n, uint32_t d);
|
|
extern int32_t tabledivide32_noinline(int32_t n, int32_t d);
|
|
extern int64_t tabledivide64_noinline(int64_t n, int64_t d);
|
|
|
|
#ifdef GEKKO
|
|
static inline int32_t divscale(int32_t eax, int32_t ebx, int32_t ecx) { return dw(tabledivide64(ldexp(eax, ecx), ebx)); }
|
|
#else
|
|
static inline int32_t divscale(int32_t eax, int32_t ebx, int32_t ecx) { return dw(tabledivide64(qw(eax) << by(ecx), ebx)); }
|
|
#endif
|
|
|
|
static inline int64_t divscale64(int64_t eax, int64_t ebx, int64_t ecx) { return tabledivide64(eax << ecx, ebx); }
|
|
|
|
#define EDUKE32_SCALER_PRAGMA(a) \
|
|
static FORCE_INLINE int32_t divscale##a(int32_t eax, int32_t ebx) { return divscale(eax, ebx, a); }
|
|
EDUKE32_GENERATE_PRAGMAS EDUKE32_SCALER_PRAGMA(32)
|
|
#undef EDUKE32_SCALER_PRAGMA
|
|
|
|
static inline int32_t scale(int32_t eax, int32_t edx, int32_t ecx)
|
|
{
|
|
return dw(tabledivide64(qw(eax) * edx, ecx));
|
|
}
|
|
|
|
static FORCE_INLINE int32_t scaleadd(int32_t eax, int32_t edx, int32_t addend, int32_t ecx)
|
|
{
|
|
return dw(tabledivide64(qw(eax) * edx + addend, ecx));
|
|
}
|
|
|
|
static inline int32_t roundscale(int32_t eax, int32_t edx, int32_t ecx)
|
|
{
|
|
return scaleadd(eax, edx, ecx / 2, ecx);
|
|
}
|
|
|
|
|
|
//
|
|
// Generic C
|
|
//
|
|
|
|
#ifndef pragmas_have_mulscale
|
|
|
|
#define EDUKE32_SCALER_PRAGMA(a) \
|
|
static FORCE_INLINE CONSTEXPR int32_t mulscale##a(int32_t eax, int32_t edx) { return dw((qw(eax) * edx) >> by(a)); } \
|
|
static FORCE_INLINE CONSTEXPR int32_t dmulscale##a(int32_t eax, int32_t edx, int32_t esi, int32_t edi) \
|
|
{ \
|
|
return dw(((qw(eax) * edx) + (qw(esi) * edi)) >> by(a)); \
|
|
} \
|
|
static FORCE_INLINE CONSTEXPR int32_t tmulscale##a(int32_t eax, int32_t edx, int32_t ebx, int32_t ecx, int32_t esi, int32_t edi) \
|
|
{ \
|
|
return dw(((qw(eax) * edx) + (qw(ebx) * ecx) + (qw(esi) * edi)) >> by(a)); \
|
|
}
|
|
|
|
EDUKE32_GENERATE_PRAGMAS EDUKE32_SCALER_PRAGMA(32)
|
|
|
|
#undef EDUKE32_SCALER_PRAGMA
|
|
|
|
#endif
|
|
|
|
|
|
#ifdef __cplusplus
|
|
}
|
|
template <typename T>
|
|
static FORCE_INLINE void swap(T * const a, T * const b)
|
|
{
|
|
T const t = *a;
|
|
*a = *b;
|
|
*b = t;
|
|
}
|
|
#define swapptr swap
|
|
extern "C" {
|
|
#else
|
|
static FORCE_INLINE void swapptr(void *a, void *b)
|
|
{
|
|
intptr_t const t = *(intptr_t*) a;
|
|
*(intptr_t*) a = *(intptr_t*) b;
|
|
*(intptr_t*) b = t;
|
|
}
|
|
#endif
|
|
|
|
#ifndef pragmas_have_swaps
|
|
#define swapchar swap
|
|
#define swapshort swap
|
|
#define swaplong swap
|
|
#define swapfloat swap
|
|
#define swapdouble swap
|
|
#define swap64bit swap
|
|
|
|
static FORCE_INLINE void swapchar2(void *a, void *b, int32_t s)
|
|
{
|
|
swapchar((char *)a, (char *)b);
|
|
swapchar((char *)a + 1, (char *)b + s);
|
|
}
|
|
#endif
|
|
|
|
static FORCE_INLINE CONSTEXPR char readpixel(void *s) { return *(char *)s; }
|
|
static FORCE_INLINE void drawpixel(void *s, char a) { *(char *)s = a; }
|
|
|
|
#ifndef pragmas_have_klabs
|
|
#if 0
|
|
static FORCE_INLINE int32_t klabs(int32_t const a)
|
|
{
|
|
uint32_t const m = a >> (sizeof(uint32_t) * CHAR_BIT - 1);
|
|
return (a ^ m) - m;
|
|
}
|
|
#else
|
|
#define klabs(x) abs(x)
|
|
#endif
|
|
#endif
|
|
#ifndef pragmas_have_ksgn
|
|
static FORCE_INLINE CONSTEXPR int ksgn(int32_t a) { return (a > 0) - (a < 0); }
|
|
#endif
|
|
|
|
#ifndef pragmas_have_mulscale
|
|
static FORCE_INLINE CONSTEXPR int32_t mulscale(int32_t eax, int32_t edx, int32_t ecx) { return dw((qw(eax) * edx) >> by(ecx)); }
|
|
static FORCE_INLINE CONSTEXPR int32_t dmulscale(int32_t eax, int32_t edx, int32_t esi, int32_t edi, int32_t ecx)
|
|
{
|
|
return dw(((qw(eax) * edx) + (qw(esi) * edi)) >> by(ecx));
|
|
}
|
|
#endif
|
|
|
|
#ifndef pragmas_have_qinterpolatedown16
|
|
void qinterpolatedown16(intptr_t bufptr, int32_t num, int32_t val, int32_t add);
|
|
void qinterpolatedown16short(intptr_t bufptr, int32_t num, int32_t val, int32_t add);
|
|
#endif
|
|
|
|
#ifndef pragmas_have_clearbuf
|
|
void clearbuf(void *d, int32_t c, int32_t a);
|
|
#endif
|
|
#ifndef pragmas_have_copybuf
|
|
void copybuf(const void *s, void *d, int32_t c);
|
|
#endif
|
|
#ifndef pragmas_have_swaps
|
|
void swapbuf4(void *a, void *b, int32_t c);
|
|
#endif
|
|
|
|
#ifndef pragmas_have_clearbufbyte
|
|
void clearbufbyte(void *D, int32_t c, int32_t a);
|
|
#endif
|
|
#ifndef pragmas_have_copybufbyte
|
|
void copybufbyte(const void *S, void *D, int32_t c);
|
|
#endif
|
|
#ifndef pragmas_have_copybufreverse
|
|
void copybufreverse(const void *S, void *D, int32_t c);
|
|
#endif
|
|
|
|
#ifndef pragmas_have_krecipasm
|
|
static inline int32_t krecipasm(int32_t i)
|
|
{
|
|
// Ken did this
|
|
union { int32_t i; float f; } x;
|
|
x.f = (float)i;
|
|
i = x.i;
|
|
return ((reciptable[(i >> 12) & 2047] >> (((i - 0x3f800000) >> 23) & 31)) ^ (i >> 31));
|
|
}
|
|
#endif
|
|
|
|
#undef qw
|
|
#undef dw
|
|
#undef wo
|
|
#undef by
|
|
|
|
static inline void swapbufreverse(void *s, void *d, int32_t c)
|
|
{
|
|
uint8_t *src = (uint8_t *)s, *dst = (uint8_t *)d;
|
|
Bassert(c >= 4);
|
|
|
|
do
|
|
{
|
|
swapchar(dst, src);
|
|
swapchar(dst + 1, src - 1);
|
|
swapchar(dst + 2, src - 2);
|
|
swapchar(dst + 3, src - 3);
|
|
dst += 4, src -= 4;
|
|
} while ((c -= 4) > 4);
|
|
|
|
while (c--)
|
|
swapchar(dst++, src--);
|
|
}
|
|
|
|
#ifdef __cplusplus
|
|
}
|
|
#endif
|
|
|
|
#endif // pragmas_h_
|