adam/cobm
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

More optimizations to bilinear (now computes two pixels at once in inner

Browse Source 
loop, reducing memory fetches)


git-svn-id: svn://svn.code.sf.net/p/sc2/code/trunk@1063 8092fc87-c524-0410-9efc-e669fe64eaf9
This commit is contained in:
gewlitys
2003-07-18 15:47:52 +00:00
parent 0bca0e6eab
commit 5b78f45bc4
1 changed files with 33 additions and 57 deletions
Download Patch File Download Diff File Expand all files Collapse all files
sc2/src/sc2code/libs/graphics/sdl/2xscalers.c
+32 -56
View File
@@ -2476,7 +2476,7 @@ Scale_BlendPixels_bilinear (BLEND_PARAMS bp, Uint32* p, const Uint32* w)
// bilinear scaling to 2x // bilinear scaling to 2x
void Scale_BilinearFilter (SDL_Surface *src, SDL_Surface *dst, SDL_Rect *r) void Scale_BilinearFilter (SDL_Surface *src, SDL_Surface *dst, SDL_Rect *r)
{ {
int x, y, i = 0, j; int x, y, j = 0;
const int w = dst->w, h = dst->h; const int w = dst->w, h = dst->h;
int drx, dry; int drx, dry;
int xend, yend; int xend, yend;
@@ -2500,7 +2500,7 @@ void Scale_BilinearFilter (SDL_Surface *src, SDL_Surface *dst, SDL_Rect *r)
xend = 2 * (region->x + region->w); xend = 2 * (region->x + region->w);
yend = 2 * (region->y + region->h); yend = 2 * (region->y + region->h);
dsrc = src->w - region->w; dsrc = src->w - region->w;
ddst = dst->w - 2 * region->w; ddst = 2 * dst->w - 2 * region->w;
// precompute the blending masks and shifts // precompute the blending masks and shifts
Scale_CalcChannelMasks (fmt, 16, &blend16); Scale_CalcChannelMasks (fmt, 16, &blend16);
@@ -2512,7 +2512,7 @@ void Scale_BilinearFilter (SDL_Surface *src, SDL_Surface *dst, SDL_Rect *r)
{ {
case 2: case 2:
{ {
Uint16 *src_p = (Uint16 *) src->pixels, *src_p2; Uint16 *src_p = (Uint16 *) src->pixels;
Uint16 *dst_p = (Uint16 *) dst->pixels; Uint16 *dst_p = (Uint16 *) dst->pixels;
Uint32 p[4]; Uint32 p[4];
@@ -2520,11 +2520,8 @@ void Scale_BilinearFilter (SDL_Surface *src, SDL_Surface *dst, SDL_Rect *r)
src_p += src->w * region->y + region->x; src_p += src->w * region->y + region->x;
dst_p += (dst->w * region->y + region->x) * 2; dst_p += (dst->w * region->y + region->x) * 2;
for (y = dry; y < yend; ++y, dst_p += ddst) for (y = dry; y < yend; y += 2, dst_p += ddst, src_p += dsrc)
{ {
src_p2 = src_p;
j = i;
p[0] = src_p[0]; p[0] = src_p[0];
if (y < h - 2) if (y < h - 2)
p[2] = src_p[src->w]; p[2] = src_p[src->w];
@@ -2533,7 +2530,7 @@ void Scale_BilinearFilter (SDL_Surface *src, SDL_Surface *dst, SDL_Rect *r)
for (x = drx; x < xend; ++x, ++dst_p) for (x = drx; x < xend; ++x, ++dst_p)
{ {
if (j == i) if (j == 0)
{ {
if (y < h - 2) if (y < h - 2)
{ {
@@ -2562,32 +2559,27 @@ void Scale_BilinearFilter (SDL_Surface *src, SDL_Surface *dst, SDL_Rect *r)
blend16, p, bilinear_table[j] blend16, p, bilinear_table[j]
); );
if (j & 2) *(dst_p + w) = Scale_BlendPixels_bilinear (
blend16, p, bilinear_table[j + 1]
);
if (j == 2)
{ {
j = i; j = 0;
src_p++; src_p++;
p[0] = p[1]; p[0] = p[1];
p[2] = p[3]; p[2] = p[3];
} }
else else
{ j = 2;
j = i | 2;
}
} }
if (!i)
src_p = src_p2;
else
src_p += dsrc;
i = !i;
} }
break; break;
} }
case 3: case 3:
{ {
PIXEL_24BIT *src_p = (PIXEL_24BIT *) src->pixels, *src_p2; PIXEL_24BIT *src_p = (PIXEL_24BIT *) src->pixels;
PIXEL_24BIT *dst_p = (PIXEL_24BIT *) dst->pixels; PIXEL_24BIT *dst_p = (PIXEL_24BIT *) dst->pixels;
Uint32 p[4]; Uint32 p[4];
@@ -2595,11 +2587,8 @@ void Scale_BilinearFilter (SDL_Surface *src, SDL_Surface *dst, SDL_Rect *r)
src_p += src->w * region->y + region->x; src_p += src->w * region->y + region->x;
dst_p += (dst->w * region->y + region->x) * 2; dst_p += (dst->w * region->y + region->x) * 2;
for (y = dry; y < yend; ++y, dst_p += ddst) for (y = dry; y < yend; y += 2, dst_p += ddst, src_p += dsrc)
{ {
src_p2 = src_p;
j = i;
p[0] = GET_PIX_24BIT (src_p); p[0] = GET_PIX_24BIT (src_p);
if (y < h - 2) if (y < h - 2)
p[2] = GET_PIX_24BIT (src_p + src->w); p[2] = GET_PIX_24BIT (src_p + src->w);
@@ -2608,7 +2597,7 @@ void Scale_BilinearFilter (SDL_Surface *src, SDL_Surface *dst, SDL_Rect *r)
for (x = drx; x < xend; ++x, ++dst_p) for (x = drx; x < xend; ++x, ++dst_p)
{ {
if (j == i) if (j == 0)
{ {
if (y < h - 2) if (y < h - 2)
{ {
@@ -2637,32 +2626,27 @@ void Scale_BilinearFilter (SDL_Surface *src, SDL_Surface *dst, SDL_Rect *r)
blend16, p, bilinear_table[j] blend16, p, bilinear_table[j]
)); ));
if (j & 2) SET_PIX_24BIT (dst_p + w, Scale_BlendPixels_bilinear (
blend16, p, bilinear_table[j + 1]
));
if (j == 2)
{ {
j = i; j = 0;
src_p++; src_p++;
p[0] = p[1]; p[0] = p[1];
p[2] = p[3]; p[2] = p[3];
} }
else else
{ j = 2;
j = i | 2;
}
} }
if (!i)
src_p = src_p2;
else
src_p += dsrc;
i = !i;
} }
break; break;
} }
case 4: case 4:
{ {
Uint32 *src_p = (Uint32 *) src->pixels, *src_p2; Uint32 *src_p = (Uint32 *) src->pixels;
Uint32 *dst_p = (Uint32 *) dst->pixels; Uint32 *dst_p = (Uint32 *) dst->pixels;
Uint32 p[4]; Uint32 p[4];
@@ -2670,11 +2654,8 @@ void Scale_BilinearFilter (SDL_Surface *src, SDL_Surface *dst, SDL_Rect *r)
src_p += src->w * region->y + region->x; src_p += src->w * region->y + region->x;
dst_p += (dst->w * region->y + region->x) * 2; dst_p += (dst->w * region->y + region->x) * 2;
for (y = dry; y < yend; ++y, dst_p += ddst) for (y = dry; y < yend; y += 2, dst_p += ddst, src_p += dsrc)
{ {
src_p2 = src_p;
j = i;
p[0] = src_p[0]; p[0] = src_p[0];
if (y < h - 2) if (y < h - 2)
p[2] = src_p[src->w]; p[2] = src_p[src->w];
@@ -2683,7 +2664,7 @@ void Scale_BilinearFilter (SDL_Surface *src, SDL_Surface *dst, SDL_Rect *r)
for (x = drx; x < xend; ++x, ++dst_p) for (x = drx; x < xend; ++x, ++dst_p)
{ {
if (j == i) if (j == 0)
{ {
if (y < h - 2) if (y < h - 2)
{ {
@@ -2712,25 +2693,20 @@ void Scale_BilinearFilter (SDL_Surface *src, SDL_Surface *dst, SDL_Rect *r)
blend16, p, bilinear_table[j] blend16, p, bilinear_table[j]
); );
if (j & 2) *(dst_p + w) = Scale_BlendPixels_bilinear (
blend16, p, bilinear_table[j + 1]
);
if (j == 2)
{ {
j = i; j = 0;
src_p++; src_p++;
p[0] = p[1]; p[0] = p[1];
p[2] = p[3]; p[2] = p[3];
} }
else else
{ j = 2;
j = i | 2;
}
} }
if (!i)
src_p = src_p2;
else
src_p += dsrc;
i = !i;
} }
break; break;