Commit 4c9e2706 authored by Richard Gobeille's avatar Richard Gobeille
Browse files

Additional clipmove() reworks

This fixes a LOT of clipping errors but needs thorough testing.

From-SVN: r7829
parent 540e1bff
...@@ -1328,35 +1328,38 @@ int32_t clipmove(vec3_t * const pos, int16_t * const sectnum, int32_t xvect, int ...@@ -1328,35 +1328,38 @@ int32_t clipmove(vec3_t * const pos, int16_t * const sectnum, int32_t xvect, int
do do
{ {
for (native_t i=clipnum-1;i>=0;--i)
{
if (clipinsideboxline(pos->x, pos->y, clipit[i].x1, clipit[i].y1, clipit[i].x2, clipit[i].y2, walldist))
{
keepaway(&pos->x, &pos->y, i);
break;
}
}
vec2_t vec = goal; vec2_t vec = goal;
hitwall = raytrace(pos->x, pos->y, &vec.x, &vec.y); if ((hitwall = raytrace(pos->x, pos->y, &vec.x, &vec.y)) >= 0)
if (hitwall >= 0)
{ {
vec2_t const clipr = { clipit[hitwall].x2 - clipit[hitwall].x1, clipit[hitwall].y2 - clipit[hitwall].y1 }; vec2_t const clipr = { clipit[hitwall].x2 - clipit[hitwall].x1, clipit[hitwall].y2 - clipit[hitwall].y1 };
uint64_t const tempull = (int64_t)clipr.x * (int64_t)clipr.x + (int64_t)clipr.y * (int64_t)clipr.y; int64_t const templl = (int64_t)clipr.x * clipr.x + (int64_t)clipr.y * clipr.y;
if (tempull > 0 && tempull < INT32_MAX) if (templl > 0)
{ {
int32_t const tempint2 = (int32_t) tempull; int64_t const templl2 = (int64_t)(goal.x-vec.x)*clipr.x + (int64_t)(goal.y-vec.y)*clipr.y;
int32_t const tempint1 = (goal.x-vec.x)*clipr.x+(goal.y-vec.y)*clipr.y; int32_t const i = ((llabs(templl2)>>11) < templl) ? divscale64(templl2, templl, 20) : 0;
int32_t const i = ((klabs(tempint1)>>11) < tempint2) ? divscale20(tempint1, tempint2) : 0;
goal = { mulscale20(clipr.x, i)+vec.x, mulscale20(clipr.y, i)+vec.y }; goal = { mulscale20(clipr.x, i)+vec.x, mulscale20(clipr.y, i)+vec.y };
} }
int32_t const tempint1 = dmulscale6(clipr.x, move.x, clipr.y, move.y); int32_t const tempint = dmulscale6(clipr.x, move.x, clipr.y, move.y);
for (native_t i=cnt+1, j; i<=clipmoveboxtracenum; ++i) for (native_t i=cnt+1, j; i<=clipmoveboxtracenum; ++i)
{ {
j = hitwalls[i]; j = hitwalls[i];
int32_t const tempint2 = dmulscale6(clipit[j].x2 - clipit[j].x1, move.x, clipit[j].y2 - clipit[j].y1, move.y);
if ((tempint1^tempint2) < 0) if ((tempint ^ dmulscale6(clipit[j].x2-clipit[j].x1, move.x, clipit[j].y2-clipit[j].y1, move.y)) < 0)
{
clipupdatesector(*(vec2_t *)pos, sectnum, rad);
return clipReturn; return clipReturn;
}
} }
keepaway(&goal.x, &goal.y, hitwall); keepaway(&goal.x, &goal.y, hitwall);
...@@ -1368,17 +1371,11 @@ int32_t clipmove(vec3_t * const pos, int16_t * const sectnum, int32_t xvect, int ...@@ -1368,17 +1371,11 @@ int32_t clipmove(vec3_t * const pos, int16_t * const sectnum, int32_t xvect, int
hitwalls[cnt] = hitwall; hitwalls[cnt] = hitwall;
} }
int const osectnum = *sectnum;
clipupdatesector(vec, sectnum, rad); clipupdatesector(vec, sectnum, rad);
if (*sectnum != -1 && (*sectnum == osectnum || editstatus || !check_floor_curb(osectnum, *sectnum, flordist, pos->z, vec))) pos->x = vec.x;
{ pos->y = vec.y;
pos->x = vec.x; cnt--;
pos->y = vec.y;
cnt--;
}
else
*sectnum = osectnum;
} while ((xvect|yvect) != 0 && hitwall >= 0 && cnt > 0); } while ((xvect|yvect) != 0 && hitwall >= 0 && cnt > 0);
return clipReturn; return clipReturn;
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment