sdl

FORK: Simple Directmedia Layer
git clone https://git.neptards.moe/neptards/sdl.git
Log | Files | Refs

SDL_blit_copy.c (4208B)

      1 /*
      2   Simple DirectMedia Layer
      3   Copyright (C) 1997-2020 Sam Lantinga <slouken@libsdl.org>
      4 
      5   This software is provided 'as-is', without any express or implied
      6   warranty.  In no event will the authors be held liable for any damages
      7   arising from the use of this software.
      8 
      9   Permission is granted to anyone to use this software for any purpose,
     10   including commercial applications, and to alter it and redistribute it
     11   freely, subject to the following restrictions:
     12 
     13   1. The origin of this software must not be misrepresented; you must not
     14      claim that you wrote the original software. If you use this software
     15      in a product, an acknowledgment in the product documentation would be
     16      appreciated but is not required.
     17   2. Altered source versions must be plainly marked as such, and must not be
     18      misrepresented as being the original software.
     19   3. This notice may not be removed or altered from any source distribution.
     20 */
     21 #include "../SDL_internal.h"
     22 
     23 #include "SDL_video.h"
     24 #include "SDL_blit.h"
     25 #include "SDL_blit_copy.h"
     26 
     27 
     28 #ifdef __SSE__
     29 /* This assumes 16-byte aligned src and dst */
     30 static SDL_INLINE void
     31 SDL_memcpySSE(Uint8 * dst, const Uint8 * src, int len)
     32 {
     33     int i;
     34 
     35     __m128 values[4];
     36     for (i = len / 64; i--;) {
     37         _mm_prefetch(src, _MM_HINT_NTA);
     38         values[0] = *(__m128 *) (src + 0);
     39         values[1] = *(__m128 *) (src + 16);
     40         values[2] = *(__m128 *) (src + 32);
     41         values[3] = *(__m128 *) (src + 48);
     42         _mm_stream_ps((float *) (dst + 0), values[0]);
     43         _mm_stream_ps((float *) (dst + 16), values[1]);
     44         _mm_stream_ps((float *) (dst + 32), values[2]);
     45         _mm_stream_ps((float *) (dst + 48), values[3]);
     46         src += 64;
     47         dst += 64;
     48     }
     49 
     50     if (len & 63)
     51         SDL_memcpy(dst, src, len & 63);
     52 }
     53 #endif /* __SSE__ */
     54 
     55 #ifdef __MMX__
     56 #ifdef _MSC_VER
     57 #pragma warning(disable:4799)
     58 #endif
     59 static SDL_INLINE void
     60 SDL_memcpyMMX(Uint8 * dst, const Uint8 * src, int len)
     61 {
     62     const int remain = (len & 63);
     63     int i;
     64 
     65     __m64* d64 = (__m64*)dst;
     66     __m64* s64 = (__m64*)src;
     67 
     68     for(i= len / 64; i--;) {
     69         d64[0] = s64[0];
     70         d64[1] = s64[1];
     71         d64[2] = s64[2];
     72         d64[3] = s64[3];
     73         d64[4] = s64[4];
     74         d64[5] = s64[5];
     75         d64[6] = s64[6];
     76         d64[7] = s64[7];
     77 
     78         d64 += 8;
     79         s64 += 8;
     80     }
     81 
     82     if (remain)
     83     {
     84         const int skip = len - remain;
     85         SDL_memcpy(dst + skip, src + skip, remain);
     86     }
     87 }
     88 #endif /* __MMX__ */
     89 
     90 void
     91 SDL_BlitCopy(SDL_BlitInfo * info)
     92 {
     93     SDL_bool overlap;
     94     Uint8 *src, *dst;
     95     int w, h;
     96     int srcskip, dstskip;
     97 
     98     w = info->dst_w * info->dst_fmt->BytesPerPixel;
     99     h = info->dst_h;
    100     src = info->src;
    101     dst = info->dst;
    102     srcskip = info->src_pitch;
    103     dstskip = info->dst_pitch;
    104 
    105     /* Properly handle overlapping blits */
    106     if (src < dst) {
    107         overlap = (dst < (src + h*srcskip));
    108     } else {
    109         overlap = (src < (dst + h*dstskip));
    110     }
    111     if (overlap) {
    112         if ( dst < src ) {
    113                 while ( h-- ) {
    114                         SDL_memmove(dst, src, w);
    115                         src += srcskip;
    116                         dst += dstskip;
    117                 }
    118         } else {
    119                 src += ((h-1) * srcskip);
    120                 dst += ((h-1) * dstskip);
    121                 while ( h-- ) {
    122                         SDL_memmove(dst, src, w);
    123                         src -= srcskip;
    124                         dst -= dstskip;
    125                 }
    126         }
    127         return;
    128     }
    129 
    130 #ifdef __SSE__
    131     if (SDL_HasSSE() &&
    132         !((uintptr_t) src & 15) && !(srcskip & 15) &&
    133         !((uintptr_t) dst & 15) && !(dstskip & 15)) {
    134         while (h--) {
    135             SDL_memcpySSE(dst, src, w);
    136             src += srcskip;
    137             dst += dstskip;
    138         }
    139         return;
    140     }
    141 #endif
    142 
    143 #ifdef __MMX__
    144     if (SDL_HasMMX() && !(srcskip & 7) && !(dstskip & 7)) {
    145         while (h--) {
    146             SDL_memcpyMMX(dst, src, w);
    147             src += srcskip;
    148             dst += dstskip;
    149         }
    150         _mm_empty();
    151         return;
    152     }
    153 #endif
    154 
    155     while (h--) {
    156         SDL_memcpy(dst, src, w);
    157         src += srcskip;
    158         dst += dstskip;
    159     }
    160 }
    161 
    162 /* vi: set ts=4 sw=4 expandtab: */