summaryrefslogtreecommitdiff
path: root/src/qcommon/q_math.c
diff options
context:
space:
mode:
Diffstat (limited to 'src/qcommon/q_math.c')
-rw-r--r--src/qcommon/q_math.c1562
1 files changed, 1562 insertions, 0 deletions
diff --git a/src/qcommon/q_math.c b/src/qcommon/q_math.c
new file mode 100644
index 0000000..196d2f5
--- /dev/null
+++ b/src/qcommon/q_math.c
@@ -0,0 +1,1562 @@
+/*
+===========================================================================
+Copyright (C) 1999-2005 Id Software, Inc.
+Copyright (C) 2000-2006 Tim Angus
+
+This file is part of Tremulous.
+
+Tremulous is free software; you can redistribute it
+and/or modify it under the terms of the GNU General Public License as
+published by the Free Software Foundation; either version 2 of the License,
+or (at your option) any later version.
+
+Tremulous is distributed in the hope that it will be
+useful, but WITHOUT ANY WARRANTY; without even the implied warranty of
+MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+GNU General Public License for more details.
+
+You should have received a copy of the GNU General Public License
+along with Tremulous; if not, write to the Free Software
+Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
+===========================================================================
+*/
+//
+// q_math.c -- stateless support routines that are included in each code module
+
+// Some of the vector functions are static inline in q_shared.h. q3asm
+// doesn't understand static functions though, so we only want them in
+// one file. That's what this is about.
+#ifdef Q3_VM
+#define __Q3_VM_MATH
+#endif
+
+#include "q_shared.h"
+
+vec3_t vec3_origin = {0,0,0};
+vec3_t axisDefault[3] = { { 1, 0, 0 }, { 0, 1, 0 }, { 0, 0, 1 } };
+
+
+vec4_t colorBlack = {0, 0, 0, 1};
+vec4_t colorRed = {1, 0, 0, 1};
+vec4_t colorGreen = {0, 1, 0, 1};
+vec4_t colorBlue = {0, 0, 1, 1};
+vec4_t colorYellow = {1, 1, 0, 1};
+vec4_t colorMagenta= {1, 0, 1, 1};
+vec4_t colorCyan = {0, 1, 1, 1};
+vec4_t colorWhite = {1, 1, 1, 1};
+vec4_t colorLtGrey = {0.75, 0.75, 0.75, 1};
+vec4_t colorMdGrey = {0.5, 0.5, 0.5, 1};
+vec4_t colorDkGrey = {0.25, 0.25, 0.25, 1};
+
+vec4_t g_color_table[8] =
+ {
+ {0.0, 0.0, 0.0, 1.0},
+ {1.0, 0.0, 0.0, 1.0},
+ {0.0, 1.0, 0.0, 1.0},
+ {1.0, 1.0, 0.0, 1.0},
+ {0.0, 0.0, 1.0, 1.0},
+ {0.0, 1.0, 1.0, 1.0},
+ {1.0, 0.0, 1.0, 1.0},
+ {1.0, 1.0, 1.0, 1.0},
+ };
+
+
+vec3_t bytedirs[NUMVERTEXNORMALS] =
+{
+{-0.525731f, 0.000000f, 0.850651f}, {-0.442863f, 0.238856f, 0.864188f},
+{-0.295242f, 0.000000f, 0.955423f}, {-0.309017f, 0.500000f, 0.809017f},
+{-0.162460f, 0.262866f, 0.951056f}, {0.000000f, 0.000000f, 1.000000f},
+{0.000000f, 0.850651f, 0.525731f}, {-0.147621f, 0.716567f, 0.681718f},
+{0.147621f, 0.716567f, 0.681718f}, {0.000000f, 0.525731f, 0.850651f},
+{0.309017f, 0.500000f, 0.809017f}, {0.525731f, 0.000000f, 0.850651f},
+{0.295242f, 0.000000f, 0.955423f}, {0.442863f, 0.238856f, 0.864188f},
+{0.162460f, 0.262866f, 0.951056f}, {-0.681718f, 0.147621f, 0.716567f},
+{-0.809017f, 0.309017f, 0.500000f},{-0.587785f, 0.425325f, 0.688191f},
+{-0.850651f, 0.525731f, 0.000000f},{-0.864188f, 0.442863f, 0.238856f},
+{-0.716567f, 0.681718f, 0.147621f},{-0.688191f, 0.587785f, 0.425325f},
+{-0.500000f, 0.809017f, 0.309017f}, {-0.238856f, 0.864188f, 0.442863f},
+{-0.425325f, 0.688191f, 0.587785f}, {-0.716567f, 0.681718f, -0.147621f},
+{-0.500000f, 0.809017f, -0.309017f}, {-0.525731f, 0.850651f, 0.000000f},
+{0.000000f, 0.850651f, -0.525731f}, {-0.238856f, 0.864188f, -0.442863f},
+{0.000000f, 0.955423f, -0.295242f}, {-0.262866f, 0.951056f, -0.162460f},
+{0.000000f, 1.000000f, 0.000000f}, {0.000000f, 0.955423f, 0.295242f},
+{-0.262866f, 0.951056f, 0.162460f}, {0.238856f, 0.864188f, 0.442863f},
+{0.262866f, 0.951056f, 0.162460f}, {0.500000f, 0.809017f, 0.309017f},
+{0.238856f, 0.864188f, -0.442863f},{0.262866f, 0.951056f, -0.162460f},
+{0.500000f, 0.809017f, -0.309017f},{0.850651f, 0.525731f, 0.000000f},
+{0.716567f, 0.681718f, 0.147621f}, {0.716567f, 0.681718f, -0.147621f},
+{0.525731f, 0.850651f, 0.000000f}, {0.425325f, 0.688191f, 0.587785f},
+{0.864188f, 0.442863f, 0.238856f}, {0.688191f, 0.587785f, 0.425325f},
+{0.809017f, 0.309017f, 0.500000f}, {0.681718f, 0.147621f, 0.716567f},
+{0.587785f, 0.425325f, 0.688191f}, {0.955423f, 0.295242f, 0.000000f},
+{1.000000f, 0.000000f, 0.000000f}, {0.951056f, 0.162460f, 0.262866f},
+{0.850651f, -0.525731f, 0.000000f},{0.955423f, -0.295242f, 0.000000f},
+{0.864188f, -0.442863f, 0.238856f}, {0.951056f, -0.162460f, 0.262866f},
+{0.809017f, -0.309017f, 0.500000f}, {0.681718f, -0.147621f, 0.716567f},
+{0.850651f, 0.000000f, 0.525731f}, {0.864188f, 0.442863f, -0.238856f},
+{0.809017f, 0.309017f, -0.500000f}, {0.951056f, 0.162460f, -0.262866f},
+{0.525731f, 0.000000f, -0.850651f}, {0.681718f, 0.147621f, -0.716567f},
+{0.681718f, -0.147621f, -0.716567f},{0.850651f, 0.000000f, -0.525731f},
+{0.809017f, -0.309017f, -0.500000f}, {0.864188f, -0.442863f, -0.238856f},
+{0.951056f, -0.162460f, -0.262866f}, {0.147621f, 0.716567f, -0.681718f},
+{0.309017f, 0.500000f, -0.809017f}, {0.425325f, 0.688191f, -0.587785f},
+{0.442863f, 0.238856f, -0.864188f}, {0.587785f, 0.425325f, -0.688191f},
+{0.688191f, 0.587785f, -0.425325f}, {-0.147621f, 0.716567f, -0.681718f},
+{-0.309017f, 0.500000f, -0.809017f}, {0.000000f, 0.525731f, -0.850651f},
+{-0.525731f, 0.000000f, -0.850651f}, {-0.442863f, 0.238856f, -0.864188f},
+{-0.295242f, 0.000000f, -0.955423f}, {-0.162460f, 0.262866f, -0.951056f},
+{0.000000f, 0.000000f, -1.000000f}, {0.295242f, 0.000000f, -0.955423f},
+{0.162460f, 0.262866f, -0.951056f}, {-0.442863f, -0.238856f, -0.864188f},
+{-0.309017f, -0.500000f, -0.809017f}, {-0.162460f, -0.262866f, -0.951056f},
+{0.000000f, -0.850651f, -0.525731f}, {-0.147621f, -0.716567f, -0.681718f},
+{0.147621f, -0.716567f, -0.681718f}, {0.000000f, -0.525731f, -0.850651f},
+{0.309017f, -0.500000f, -0.809017f}, {0.442863f, -0.238856f, -0.864188f},
+{0.162460f, -0.262866f, -0.951056f}, {0.238856f, -0.864188f, -0.442863f},
+{0.500000f, -0.809017f, -0.309017f}, {0.425325f, -0.688191f, -0.587785f},
+{0.716567f, -0.681718f, -0.147621f}, {0.688191f, -0.587785f, -0.425325f},
+{0.587785f, -0.425325f, -0.688191f}, {0.000000f, -0.955423f, -0.295242f},
+{0.000000f, -1.000000f, 0.000000f}, {0.262866f, -0.951056f, -0.162460f},
+{0.000000f, -0.850651f, 0.525731f}, {0.000000f, -0.955423f, 0.295242f},
+{0.238856f, -0.864188f, 0.442863f}, {0.262866f, -0.951056f, 0.162460f},
+{0.500000f, -0.809017f, 0.309017f}, {0.716567f, -0.681718f, 0.147621f},
+{0.525731f, -0.850651f, 0.000000f}, {-0.238856f, -0.864188f, -0.442863f},
+{-0.500000f, -0.809017f, -0.309017f}, {-0.262866f, -0.951056f, -0.162460f},
+{-0.850651f, -0.525731f, 0.000000f}, {-0.716567f, -0.681718f, -0.147621f},
+{-0.716567f, -0.681718f, 0.147621f}, {-0.525731f, -0.850651f, 0.000000f},
+{-0.500000f, -0.809017f, 0.309017f}, {-0.238856f, -0.864188f, 0.442863f},
+{-0.262866f, -0.951056f, 0.162460f}, {-0.864188f, -0.442863f, 0.238856f},
+{-0.809017f, -0.309017f, 0.500000f}, {-0.688191f, -0.587785f, 0.425325f},
+{-0.681718f, -0.147621f, 0.716567f}, {-0.442863f, -0.238856f, 0.864188f},
+{-0.587785f, -0.425325f, 0.688191f}, {-0.309017f, -0.500000f, 0.809017f},
+{-0.147621f, -0.716567f, 0.681718f}, {-0.425325f, -0.688191f, 0.587785f},
+{-0.162460f, -0.262866f, 0.951056f}, {0.442863f, -0.238856f, 0.864188f},
+{0.162460f, -0.262866f, 0.951056f}, {0.309017f, -0.500000f, 0.809017f},
+{0.147621f, -0.716567f, 0.681718f}, {0.000000f, -0.525731f, 0.850651f},
+{0.425325f, -0.688191f, 0.587785f}, {0.587785f, -0.425325f, 0.688191f},
+{0.688191f, -0.587785f, 0.425325f}, {-0.955423f, 0.295242f, 0.000000f},
+{-0.951056f, 0.162460f, 0.262866f}, {-1.000000f, 0.000000f, 0.000000f},
+{-0.850651f, 0.000000f, 0.525731f}, {-0.955423f, -0.295242f, 0.000000f},
+{-0.951056f, -0.162460f, 0.262866f}, {-0.864188f, 0.442863f, -0.238856f},
+{-0.951056f, 0.162460f, -0.262866f}, {-0.809017f, 0.309017f, -0.500000f},
+{-0.864188f, -0.442863f, -0.238856f}, {-0.951056f, -0.162460f, -0.262866f},
+{-0.809017f, -0.309017f, -0.500000f}, {-0.681718f, 0.147621f, -0.716567f},
+{-0.681718f, -0.147621f, -0.716567f}, {-0.850651f, 0.000000f, -0.525731f},
+{-0.688191f, 0.587785f, -0.425325f}, {-0.587785f, 0.425325f, -0.688191f},
+{-0.425325f, 0.688191f, -0.587785f}, {-0.425325f, -0.688191f, -0.587785f},
+{-0.587785f, -0.425325f, -0.688191f}, {-0.688191f, -0.587785f, -0.425325f}
+};
+
+//==============================================================
+
+int Q_rand( int *seed ) {
+ *seed = (69069 * *seed + 1);
+ return *seed;
+}
+
+float Q_random( int *seed ) {
+ return ( Q_rand( seed ) & 0xffff ) / (float)0x10000;
+}
+
+float Q_crandom( int *seed ) {
+ return 2.0 * ( Q_random( seed ) - 0.5 );
+}
+
+//=======================================================
+
+signed char ClampChar( int i ) {
+ if ( i < -128 ) {
+ return -128;
+ }
+ if ( i > 127 ) {
+ return 127;
+ }
+ return i;
+}
+
+signed short ClampShort( int i ) {
+ if ( i < -32768 ) {
+ return -32768;
+ }
+ if ( i > 0x7fff ) {
+ return 0x7fff;
+ }
+ return i;
+}
+
+
+// this isn't a real cheap function to call!
+int DirToByte( vec3_t dir ) {
+ int i, best;
+ float d, bestd;
+
+ if ( !dir ) {
+ return 0;
+ }
+
+ bestd = 0;
+ best = 0;
+ for (i=0 ; i<NUMVERTEXNORMALS ; i++)
+ {
+ d = DotProduct (dir, bytedirs[i]);
+ if (d > bestd)
+ {
+ bestd = d;
+ best = i;
+ }
+ }
+
+ return best;
+}
+
+void ByteToDir( int b, vec3_t dir ) {
+ if ( b < 0 || b >= NUMVERTEXNORMALS ) {
+ VectorCopy( vec3_origin, dir );
+ return;
+ }
+ VectorCopy (bytedirs[b], dir);
+}
+
+
+unsigned ColorBytes3 (float r, float g, float b) {
+ unsigned i;
+
+ ( (byte *)&i )[0] = r * 255;
+ ( (byte *)&i )[1] = g * 255;
+ ( (byte *)&i )[2] = b * 255;
+
+ return i;
+}
+
+unsigned ColorBytes4 (float r, float g, float b, float a) {
+ unsigned i;
+
+ ( (byte *)&i )[0] = r * 255;
+ ( (byte *)&i )[1] = g * 255;
+ ( (byte *)&i )[2] = b * 255;
+ ( (byte *)&i )[3] = a * 255;
+
+ return i;
+}
+
+float NormalizeColor( const vec3_t in, vec3_t out ) {
+ float max;
+
+ max = in[0];
+ if ( in[1] > max ) {
+ max = in[1];
+ }
+ if ( in[2] > max ) {
+ max = in[2];
+ }
+
+ if ( !max ) {
+ VectorClear( out );
+ } else {
+ out[0] = in[0] / max;
+ out[1] = in[1] / max;
+ out[2] = in[2] / max;
+ }
+ return max;
+}
+
+
+/*
+=====================
+PlaneFromPoints
+
+Returns false if the triangle is degenrate.
+The normal will point out of the clock for clockwise ordered points
+=====================
+*/
+qboolean PlaneFromPoints( vec4_t plane, const vec3_t a, const vec3_t b, const vec3_t c ) {
+ vec3_t d1, d2;
+
+ VectorSubtract( b, a, d1 );
+ VectorSubtract( c, a, d2 );
+ CrossProduct( d2, d1, plane );
+ if ( VectorNormalize( plane ) == 0 ) {
+ return qfalse;
+ }
+
+ plane[3] = DotProduct( a, plane );
+ return qtrue;
+}
+
+/*
+===============
+RotatePointAroundVector
+
+This is not implemented very well...
+===============
+*/
+void RotatePointAroundVector( vec3_t dst, const vec3_t dir, const vec3_t point,
+ float degrees ) {
+ float sin_a;
+ float cos_a;
+ float cos_ia;
+ float i_i_ia;
+ float j_j_ia;
+ float k_k_ia;
+ float i_j_ia;
+ float i_k_ia;
+ float j_k_ia;
+ float a_sin;
+ float b_sin;
+ float c_sin;
+ float rot[3][3];
+
+ cos_ia = DEG2RAD(degrees);
+ sin_a = sin(cos_ia);
+ cos_a = cos(cos_ia);
+ cos_ia = 1.0F - cos_a;
+
+ i_i_ia = dir[0] * dir[0] * cos_ia;
+ j_j_ia = dir[1] * dir[1] * cos_ia;
+ k_k_ia = dir[2] * dir[2] * cos_ia;
+ i_j_ia = dir[0] * dir[1] * cos_ia;
+ i_k_ia = dir[0] * dir[2] * cos_ia;
+ j_k_ia = dir[1] * dir[2] * cos_ia;
+
+ a_sin = dir[0] * sin_a;
+ b_sin = dir[1] * sin_a;
+ c_sin = dir[2] * sin_a;
+
+ rot[0][0] = i_i_ia + cos_a;
+ rot[0][1] = i_j_ia - c_sin;
+ rot[0][2] = i_k_ia + b_sin;
+ rot[1][0] = i_j_ia + c_sin;
+ rot[1][1] = j_j_ia + cos_a;
+ rot[1][2] = j_k_ia - a_sin;
+ rot[2][0] = i_k_ia - b_sin;
+ rot[2][1] = j_k_ia + a_sin;
+ rot[2][2] = k_k_ia + cos_a;
+
+ dst[0] = point[0] * rot[0][0] + point[1] * rot[0][1] + point[2] * rot[0][2];
+ dst[1] = point[0] * rot[1][0] + point[1] * rot[1][1] + point[2] * rot[1][2];
+ dst[2] = point[0] * rot[2][0] + point[1] * rot[2][1] + point[2] * rot[2][2];
+}
+
+/*
+===============
+RotateAroundDirection
+===============
+*/
+void RotateAroundDirection( vec3_t axis[3], vec_t angle ) {
+ vec_t scale;
+
+ angle = DEG2RAD( angle );
+
+ // create an arbitrary axis[1]
+ PerpendicularVector( axis[ 1 ], axis[ 0 ] );
+
+ // cross to get axis[2]
+ CrossProduct( axis[ 0 ], axis[ 1 ], axis[ 2 ] );
+
+ // rotate
+ scale = cos( angle );
+ VectorScale( axis[ 1 ], scale, axis[ 1 ] );
+
+ scale = sin( angle );
+ VectorMA( axis[ 1 ], scale, axis[ 2 ], axis[ 1 ] );
+
+ // recalculate axis[2]
+ CrossProduct( axis[ 0 ], axis[ 1 ], axis[ 2 ] );
+}
+
+
+
+void vectoangles( const vec3_t value1, vec3_t angles ) {
+ float forward;
+ float yaw, pitch;
+
+ if ( value1[1] == 0 && value1[0] == 0 ) {
+ yaw = 0;
+ if ( value1[2] > 0 ) {
+ pitch = 90;
+ }
+ else {
+ pitch = 270;
+ }
+ }
+ else {
+ if ( value1[0] ) {
+ yaw = ( atan2 ( value1[1], value1[0] ) * 180 / M_PI );
+ }
+ else if ( value1[1] > 0 ) {
+ yaw = 90;
+ }
+ else {
+ yaw = 270;
+ }
+ if ( yaw < 0 ) {
+ yaw += 360;
+ }
+
+ forward = sqrt ( value1[0]*value1[0] + value1[1]*value1[1] );
+ pitch = ( atan2(value1[2], forward) * 180 / M_PI );
+ if ( pitch < 0 ) {
+ pitch += 360;
+ }
+ }
+
+ angles[PITCH] = -pitch;
+ angles[YAW] = yaw;
+ angles[ROLL] = 0;
+}
+
+
+/*
+=================
+AxisToAngles
+
+Takes an axis (forward + right + up)
+and returns angles -- including a roll
+=================
+*/
+void AxisToAngles( vec3_t axis[3], vec3_t angles ) {
+ float length1;
+ float yaw, pitch, roll = 0.0f;
+
+ if ( axis[0][1] == 0 && axis[0][0] == 0 ) {
+ yaw = 0;
+ if ( axis[0][2] > 0 ) {
+ pitch = 90;
+ }
+ else {
+ pitch = 270;
+ }
+ }
+ else {
+ if ( axis[0][0] ) {
+ yaw = ( atan2 ( axis[0][1], axis[0][0] ) * 180 / M_PI );
+ }
+ else if ( axis[0][1] > 0 ) {
+ yaw = 90;
+ }
+ else {
+ yaw = 270;
+ }
+ if ( yaw < 0 ) {
+ yaw += 360;
+ }
+
+ length1 = sqrt ( axis[0][0]*axis[0][0] + axis[0][1]*axis[0][1] );
+ pitch = ( atan2(axis[0][2], length1) * 180 / M_PI );
+ if ( pitch < 0 ) {
+ pitch += 360;
+ }
+
+ roll = ( atan2( axis[1][2], axis[2][2] ) * 180 / M_PI );
+ if ( roll < 0 ) {
+ roll += 360;
+ }
+ }
+
+ angles[PITCH] = -pitch;
+ angles[YAW] = yaw;
+ angles[ROLL] = roll;
+}
+
+/*
+=================
+AnglesToAxis
+=================
+*/
+void AnglesToAxis( const vec3_t angles, vec3_t axis[3] ) {
+ vec3_t right;
+
+ // angle vectors returns "right" instead of "y axis"
+ AngleVectors( angles, axis[0], right, axis[2] );
+ VectorSubtract( vec3_origin, right, axis[1] );
+}
+
+void AxisClear( vec3_t axis[3] ) {
+ axis[0][0] = 1;
+ axis[0][1] = 0;
+ axis[0][2] = 0;
+ axis[1][0] = 0;
+ axis[1][1] = 1;
+ axis[1][2] = 0;
+ axis[2][0] = 0;
+ axis[2][1] = 0;
+ axis[2][2] = 1;
+}
+
+void AxisCopy( vec3_t in[3], vec3_t out[3] ) {
+ VectorCopy( in[0], out[0] );
+ VectorCopy( in[1], out[1] );
+ VectorCopy( in[2], out[2] );
+}
+
+void ProjectPointOnPlane( vec3_t dst, const vec3_t p, const vec3_t normal )
+{
+ float d;
+ vec3_t n;
+ float inv_denom;
+
+ inv_denom = 1.0f / DotProduct( normal, normal );
+#ifndef Q3_VM
+ assert( Q_fabs(inv_denom) != 0.0f ); // bk010122 - zero vectors get here
+#endif
+ inv_denom = 1.0f / inv_denom;
+
+ d = DotProduct( normal, p ) * inv_denom;
+
+ n[0] = normal[0] * inv_denom;
+ n[1] = normal[1] * inv_denom;
+ n[2] = normal[2] * inv_denom;
+
+ dst[0] = p[0] - d * n[0];
+ dst[1] = p[1] - d * n[1];
+ dst[2] = p[2] - d * n[2];
+}
+
+/*
+================
+MakeNormalVectors
+
+Given a normalized forward vector, create two
+other perpendicular vectors
+================
+*/
+void MakeNormalVectors( const vec3_t forward, vec3_t right, vec3_t up) {
+ float d;
+
+ // this rotate and negate guarantees a vector
+ // not colinear with the original
+ right[1] = -forward[0];
+ right[2] = forward[1];
+ right[0] = forward[2];
+
+ d = DotProduct (right, forward);
+ VectorMA (right, -d, forward, right);
+ VectorNormalize (right);
+ CrossProduct (right, forward, up);
+}
+
+
+void VectorRotate( vec3_t in, vec3_t matrix[3], vec3_t out )
+{
+ out[0] = DotProduct( in, matrix[0] );
+ out[1] = DotProduct( in, matrix[1] );
+ out[2] = DotProduct( in, matrix[2] );
+}
+
+//============================================================================
+
+#if !idppc
+/*
+** float q_rsqrt( float number )
+*/
+float Q_rsqrt( float number )
+{
+ union {
+ float f;
+ int i;
+ } t;
+ float x2, y;
+ const float threehalfs = 1.5F;
+
+ x2 = number * 0.5F;
+ t.f = number;
+ t.i = 0x5f3759df - ( t.i >> 1 ); // what the fuck?
+ y = t.f;
+ y = y * ( threehalfs - ( x2 * y * y ) ); // 1st iteration
+// y = y * ( threehalfs - ( x2 * y * y ) ); // 2nd iteration, this can be removed
+
+ //assert( !isnan(y) ); // bk010122 - FPE?
+ return y;
+}
+
+float Q_fabs( float f ) {
+ int tmp = * ( int * ) &f;
+ tmp &= 0x7FFFFFFF;
+ return * ( float * ) &tmp;
+}
+#endif
+
+//============================================================
+
+/*
+===============
+LerpAngle
+
+===============
+*/
+float LerpAngle (float from, float to, float frac) {
+ float a;
+
+ if ( to - from > 180 ) {
+ to -= 360;
+ }
+ if ( to - from < -180 ) {
+ to += 360;
+ }
+ a = from + frac * (to - from);
+
+ return a;
+}
+
+
+/*
+=================
+AngleSubtract
+
+Always returns a value from -180 to 180
+=================
+*/
+float AngleSubtract( float a1, float a2 ) {
+ float a;
+
+ a = a1 - a2;
+ while ( a > 180 ) {
+ a -= 360;
+ }
+ while ( a < -180 ) {
+ a += 360;
+ }
+ return a;
+}
+
+
+void AnglesSubtract( vec3_t v1, vec3_t v2, vec3_t v3 ) {
+ v3[0] = AngleSubtract( v1[0], v2[0] );
+ v3[1] = AngleSubtract( v1[1], v2[1] );
+ v3[2] = AngleSubtract( v1[2], v2[2] );
+}
+
+
+float AngleMod(float a) {
+ a = (360.0/65536) * ((int)(a*(65536/360.0)) & 65535);
+ return a;
+}
+
+
+/*
+=================
+AngleNormalize360
+
+returns angle normalized to the range [0 <= angle < 360]
+=================
+*/
+float AngleNormalize360 ( float angle ) {
+ return (360.0 / 65536) * ((int)(angle * (65536 / 360.0)) & 65535);
+}
+
+
+/*
+=================
+AngleNormalize180
+
+returns angle normalized to the range [-180 < angle <= 180]
+=================
+*/
+float AngleNormalize180 ( float angle ) {
+ angle = AngleNormalize360( angle );
+ if ( angle > 180.0 ) {
+ angle -= 360.0;
+ }
+ return angle;
+}
+
+
+/*
+=================
+AngleDelta
+
+returns the normalized delta from angle1 to angle2
+=================
+*/
+float AngleDelta ( float angle1, float angle2 ) {
+ return AngleNormalize180( angle1 - angle2 );
+}
+
+
+//============================================================
+
+
+/*
+=================
+SetPlaneSignbits
+=================
+*/
+void SetPlaneSignbits (cplane_t *out) {
+ int bits, j;
+
+ // for fast box on planeside test
+ bits = 0;
+ for (j=0 ; j<3 ; j++) {
+ if (out->normal[j] < 0) {
+ bits |= 1<<j;
+ }
+ }
+ out->signbits = bits;
+}
+
+
+/*
+==================
+BoxOnPlaneSide
+
+Returns 1, 2, or 1 + 2
+
+// this is the slow, general version
+int BoxOnPlaneSide2 (vec3_t emins, vec3_t emaxs, struct cplane_s *p)
+{
+ int i;
+ float dist1, dist2;
+ int sides;
+ vec3_t corners[2];
+
+ for (i=0 ; i<3 ; i++)
+ {
+ if (p->normal[i] < 0)
+ {
+ corners[0][i] = emins[i];
+ corners[1][i] = emaxs[i];
+ }
+ else
+ {
+ corners[1][i] = emins[i];
+ corners[0][i] = emaxs[i];
+ }
+ }
+ dist1 = DotProduct (p->normal, corners[0]) - p->dist;
+ dist2 = DotProduct (p->normal, corners[1]) - p->dist;
+ sides = 0;
+ if (dist1 >= 0)
+ sides = 1;
+ if (dist2 < 0)
+ sides |= 2;
+
+ return sides;
+}
+
+==================
+*/
+
+#if !id386
+
+int BoxOnPlaneSide (vec3_t emins, vec3_t emaxs, struct cplane_s *p)
+{
+ float dist1, dist2;
+ int sides;
+
+// fast axial cases
+ if (p->type < 3)
+ {
+ if (p->dist <= emins[p->type])
+ return 1;
+ if (p->dist >= emaxs[p->type])
+ return 2;
+ return 3;
+ }
+
+// general case
+ switch (p->signbits)
+ {
+ case 0:
+ dist1 = p->normal[0]*emaxs[0] + p->normal[1]*emaxs[1] + p->normal[2]*emaxs[2];
+ dist2 = p->normal[0]*emins[0] + p->normal[1]*emins[1] + p->normal[2]*emins[2];
+ break;
+ case 1:
+ dist1 = p->normal[0]*emins[0] + p->normal[1]*emaxs[1] + p->normal[2]*emaxs[2];
+ dist2 = p->normal[0]*emaxs[0] + p->normal[1]*emins[1] + p->normal[2]*emins[2];
+ break;
+ case 2:
+ dist1 = p->normal[0]*emaxs[0] + p->normal[1]*emins[1] + p->normal[2]*emaxs[2];
+ dist2 = p->normal[0]*emins[0] + p->normal[1]*emaxs[1] + p->normal[2]*emins[2];
+ break;
+ case 3:
+ dist1 = p->normal[0]*emins[0] + p->normal[1]*emins[1] + p->normal[2]*emaxs[2];
+ dist2 = p->normal[0]*emaxs[0] + p->normal[1]*emaxs[1] + p->normal[2]*emins[2];
+ break;
+ case 4:
+ dist1 = p->normal[0]*emaxs[0] + p->normal[1]*emaxs[1] + p->normal[2]*emins[2];
+ dist2 = p->normal[0]*emins[0] + p->normal[1]*emins[1] + p->normal[2]*emaxs[2];
+ break;
+ case 5:
+ dist1 = p->normal[0]*emins[0] + p->normal[1]*emaxs[1] + p->normal[2]*emins[2];
+ dist2 = p->normal[0]*emaxs[0] + p->normal[1]*emins[1] + p->normal[2]*emaxs[2];
+ break;
+ case 6:
+ dist1 = p->normal[0]*emaxs[0] + p->normal[1]*emins[1] + p->normal[2]*emins[2];
+ dist2 = p->normal[0]*emins[0] + p->normal[1]*emaxs[1] + p->normal[2]*emaxs[2];
+ break;
+ case 7:
+ dist1 = p->normal[0]*emins[0] + p->normal[1]*emins[1] + p->normal[2]*emins[2];
+ dist2 = p->normal[0]*emaxs[0] + p->normal[1]*emaxs[1] + p->normal[2]*emaxs[2];
+ break;
+ default:
+ dist1 = dist2 = 0; // shut up compiler
+ break;
+ }
+
+ sides = 0;
+ if (dist1 >= p->dist)
+ sides = 1;
+ if (dist2 < p->dist)
+ sides |= 2;
+
+ return sides;
+}
+#elif __GNUC__
+// use matha.s
+#else
+#pragma warning( disable: 4035 )
+
+__declspec( naked ) int BoxOnPlaneSide (vec3_t emins, vec3_t emaxs, struct cplane_s *p)
+{
+ static int bops_initialized;
+ static int Ljmptab[8];
+
+ __asm {
+
+ push ebx
+
+ cmp bops_initialized, 1
+ je initialized
+ mov bops_initialized, 1
+
+ mov Ljmptab[0*4], offset Lcase0
+ mov Ljmptab[1*4], offset Lcase1
+ mov Ljmptab[2*4], offset Lcase2
+ mov Ljmptab[3*4], offset Lcase3
+ mov Ljmptab[4*4], offset Lcase4
+ mov Ljmptab[5*4], offset Lcase5
+ mov Ljmptab[6*4], offset Lcase6
+ mov Ljmptab[7*4], offset Lcase7
+
+initialized:
+
+ mov edx,dword ptr[4+12+esp]
+ mov ecx,dword ptr[4+4+esp]
+ xor eax,eax
+ mov ebx,dword ptr[4+8+esp]
+ mov al,byte ptr[17+edx]
+ cmp al,8
+ jge Lerror
+ fld dword ptr[0+edx]
+ fld st(0)
+ jmp dword ptr[Ljmptab+eax*4]
+Lcase0:
+ fmul dword ptr[ebx]
+ fld dword ptr[0+4+edx]
+ fxch st(2)
+ fmul dword ptr[ecx]
+ fxch st(2)
+ fld st(0)
+ fmul dword ptr[4+ebx]
+ fld dword ptr[0+8+edx]
+ fxch st(2)
+ fmul dword ptr[4+ecx]
+ fxch st(2)
+ fld st(0)
+ fmul dword ptr[8+ebx]
+ fxch st(5)
+ faddp st(3),st(0)
+ fmul dword ptr[8+ecx]
+ fxch st(1)
+ faddp st(3),st(0)
+ fxch st(3)
+ faddp st(2),st(0)
+ jmp LSetSides
+Lcase1:
+ fmul dword ptr[ecx]
+ fld dword ptr[0+4+edx]
+ fxch st(2)
+ fmul dword ptr[ebx]
+ fxch st(2)
+ fld st(0)
+ fmul dword ptr[4+ebx]
+ fld dword ptr[0+8+edx]
+ fxch st(2)
+ fmul dword ptr[4+ecx]
+ fxch st(2)
+ fld st(0)
+ fmul dword ptr[8+ebx]
+ fxch st(5)
+ faddp st(3),st(0)
+ fmul dword ptr[8+ecx]
+ fxch st(1)
+ faddp st(3),st(0)
+ fxch st(3)
+ faddp st(2),st(0)
+ jmp LSetSides
+Lcase2:
+ fmul dword ptr[ebx]
+ fld dword ptr[0+4+edx]
+ fxch st(2)
+ fmul dword ptr[ecx]
+ fxch st(2)
+ fld st(0)
+ fmul dword ptr[4+ecx]
+ fld dword ptr[0+8+edx]
+ fxch st(2)
+ fmul dword ptr[4+ebx]
+ fxch st(2)
+ fld st(0)
+ fmul dword ptr[8+ebx]
+ fxch st(5)
+ faddp st(3),st(0)
+ fmul dword ptr[8+ecx]
+ fxch st(1)
+ faddp st(3),st(0)
+ fxch st(3)
+ faddp st(2),st(0)
+ jmp LSetSides
+Lcase3:
+ fmul dword ptr[ecx]
+ fld dword ptr[0+4+edx]
+ fxch st(2)
+ fmul dword ptr[ebx]
+ fxch st(2)
+ fld st(0)
+ fmul dword ptr[4+ecx]
+ fld dword ptr[0+8+edx]
+ fxch st(2)
+ fmul dword ptr[4+ebx]
+ fxch st(2)
+ fld st(0)
+ fmul dword ptr[8+ebx]
+ fxch st(5)
+ faddp st(3),st(0)
+ fmul dword ptr[8+ecx]
+ fxch st(1)
+ faddp st(3),st(0)
+ fxch st(3)
+ faddp st(2),st(0)
+ jmp LSetSides
+Lcase4:
+ fmul dword ptr[ebx]
+ fld dword ptr[0+4+edx]
+ fxch st(2)
+ fmul dword ptr[ecx]
+ fxch st(2)
+ fld st(0)
+ fmul dword ptr[4+ebx]
+ fld dword ptr[0+8+edx]
+ fxch st(2)
+ fmul dword ptr[4+ecx]
+ fxch st(2)
+ fld st(0)
+ fmul dword ptr[8+ecx]
+ fxch st(5)
+ faddp st(3),st(0)
+ fmul dword ptr[8+ebx]
+ fxch st(1)
+ faddp st(3),st(0)
+ fxch st(3)
+ faddp st(2),st(0)
+ jmp LSetSides
+Lcase5:
+ fmul dword ptr[ecx]
+ fld dword ptr[0+4+edx]
+ fxch st(2)
+ fmul dword ptr[ebx]
+ fxch st(2)
+ fld st(0)
+ fmul dword ptr[4+ebx]
+ fld dword ptr[0+8+edx]
+ fxch st(2)
+ fmul dword ptr[4+ecx]
+ fxch st(2)
+ fld st(0)
+ fmul dword ptr[8+ecx]
+ fxch st(5)
+ faddp st(3),st(0)
+ fmul dword ptr[8+ebx]
+ fxch st(1)
+ faddp st(3),st(0)
+ fxch st(3)
+ faddp st(2),st(0)
+ jmp LSetSides
+Lcase6:
+ fmul dword ptr[ebx]
+ fld dword ptr[0+4+edx]
+ fxch st(2)
+ fmul dword ptr[ecx]
+ fxch st(2)
+ fld st(0)
+ fmul dword ptr[4+ecx]
+ fld dword ptr[0+8+edx]
+ fxch st(2)
+ fmul dword ptr[4+ebx]
+ fxch st(2)
+ fld st(0)
+ fmul dword ptr[8+ecx]
+ fxch st(5)
+ faddp st(3),st(0)
+ fmul dword ptr[8+ebx]
+ fxch st(1)
+ faddp st(3),st(0)
+ fxch st(3)
+ faddp st(2),st(0)
+ jmp LSetSides
+Lcase7:
+ fmul dword ptr[ecx]
+ fld dword ptr[0+4+edx]
+ fxch st(2)
+ fmul dword ptr[ebx]
+ fxch st(2)
+ fld st(0)
+ fmul dword ptr[4+ecx]
+ fld dword ptr[0+8+edx]
+ fxch st(2)
+ fmul dword ptr[4+ebx]
+ fxch st(2)
+ fld st(0)
+ fmul dword ptr[8+ecx]
+ fxch st(5)
+ faddp st(3),st(0)
+ fmul dword ptr[8+ebx]
+ fxch st(1)
+ faddp st(3),st(0)
+ fxch st(3)
+ faddp st(2),st(0)
+LSetSides:
+ faddp st(2),st(0)
+ fcomp dword ptr[12+edx]
+ xor ecx,ecx
+ fnstsw ax
+ fcomp dword ptr[12+edx]
+ and ah,1
+ xor ah,1
+ add cl,ah
+ fnstsw ax
+ and ah,1
+ add ah,ah
+ add cl,ah
+ pop ebx
+ mov eax,ecx
+ ret
+Lerror:
+ int 3
+ }
+}
+#pragma warning( default: 4035 )
+
+#endif
+
+/*
+=================
+RadiusFromBounds
+=================
+*/
+float RadiusFromBounds( const vec3_t mins, const vec3_t maxs ) {
+ int i;
+ vec3_t corner;
+ float a, b;
+
+ for (i=0 ; i<3 ; i++) {
+ a = fabs( mins[i] );
+ b = fabs( maxs[i] );
+ corner[i] = a > b ? a : b;
+ }
+
+ return VectorLength (corner);
+}
+
+
+void ClearBounds( vec3_t mins, vec3_t maxs ) {
+ mins[0] = mins[1] = mins[2] = 99999;
+ maxs[0] = maxs[1] = maxs[2] = -99999;
+}
+
+void AddPointToBounds( const vec3_t v, vec3_t mins, vec3_t maxs ) {
+ if ( v[0] < mins[0] ) {
+ mins[0] = v[0];
+ }
+ if ( v[0] > maxs[0]) {
+ maxs[0] = v[0];
+ }
+
+ if ( v[1] < mins[1] ) {
+ mins[1] = v[1];
+ }
+ if ( v[1] > maxs[1]) {
+ maxs[1] = v[1];
+ }
+
+ if ( v[2] < mins[2] ) {
+ mins[2] = v[2];
+ }
+ if ( v[2] > maxs[2]) {
+ maxs[2] = v[2];
+ }
+}
+
+
+vec_t VectorNormalize( vec3_t v ) {
+ // NOTE: TTimo - Apple G4 altivec source uses double?
+ float length, ilength;
+
+ length = v[0]*v[0] + v[1]*v[1] + v[2]*v[2];
+ length = sqrt (length);
+
+ if ( length ) {
+ ilength = 1/length;
+ v[0] *= ilength;
+ v[1] *= ilength;
+ v[2] *= ilength;
+ }
+
+ return length;
+}
+
+vec_t VectorNormalize2( const vec3_t v, vec3_t out) {
+ float length, ilength;
+
+ length = v[0]*v[0] + v[1]*v[1] + v[2]*v[2];
+ length = sqrt (length);
+
+ if (length)
+ {
+#ifndef Q3_VM // bk0101022 - FPE related
+// assert( ((Q_fabs(v[0])!=0.0f) || (Q_fabs(v[1])!=0.0f) || (Q_fabs(v[2])!=0.0f)) );
+#endif
+ ilength = 1/length;
+ out[0] = v[0]*ilength;
+ out[1] = v[1]*ilength;
+ out[2] = v[2]*ilength;
+ } else {
+#ifndef Q3_VM // bk0101022 - FPE related
+// assert( ((Q_fabs(v[0])==0.0f) && (Q_fabs(v[1])==0.0f) && (Q_fabs(v[2])==0.0f)) );
+#endif
+ VectorClear( out );
+ }
+
+ return length;
+
+}
+
+void _VectorMA( const vec3_t veca, float scale, const vec3_t vecb, vec3_t vecc) {
+ vecc[0] = veca[0] + scale*vecb[0];
+ vecc[1] = veca[1] + scale*vecb[1];
+ vecc[2] = veca[2] + scale*vecb[2];
+}
+
+
+vec_t _DotProduct( const vec3_t v1, const vec3_t v2 ) {
+ return v1[0]*v2[0] + v1[1]*v2[1] + v1[2]*v2[2];
+}
+
+void _VectorSubtract( const vec3_t veca, const vec3_t vecb, vec3_t out ) {
+ out[0] = veca[0]-vecb[0];
+ out[1] = veca[1]-vecb[1];
+ out[2] = veca[2]-vecb[2];
+}
+
+void _VectorAdd( const vec3_t veca, const vec3_t vecb, vec3_t out ) {
+ out[0] = veca[0]+vecb[0];
+ out[1] = veca[1]+vecb[1];
+ out[2] = veca[2]+vecb[2];
+}
+
+void _VectorCopy( const vec3_t in, vec3_t out ) {
+ out[0] = in[0];
+ out[1] = in[1];
+ out[2] = in[2];
+}
+
+void _VectorScale( const vec3_t in, vec_t scale, vec3_t out ) {
+ out[0] = in[0]*scale;
+ out[1] = in[1]*scale;
+ out[2] = in[2]*scale;
+}
+
+void Vector4Scale( const vec4_t in, vec_t scale, vec4_t out ) {
+ out[0] = in[0]*scale;
+ out[1] = in[1]*scale;
+ out[2] = in[2]*scale;
+ out[3] = in[3]*scale;
+}
+
+
+int Q_log2( int val ) {
+ int answer;
+
+ answer = 0;
+ while ( ( val>>=1 ) != 0 ) {
+ answer++;
+ }
+ return answer;
+}
+
+
+
+/*
+=================
+PlaneTypeForNormal
+=================
+*/
+/*
+int PlaneTypeForNormal (vec3_t normal) {
+ if ( normal[0] == 1.0 )
+ return PLANE_X;
+ if ( normal[1] == 1.0 )
+ return PLANE_Y;
+ if ( normal[2] == 1.0 )
+ return PLANE_Z;
+
+ return PLANE_NON_AXIAL;
+}
+*/
+
+
+/*
+================
+MatrixMultiply
+================
+*/
+void MatrixMultiply(float in1[3][3], float in2[3][3], float out[3][3]) {
+ out[0][0] = in1[0][0] * in2[0][0] + in1[0][1] * in2[1][0] +
+ in1[0][2] * in2[2][0];
+ out[0][1] = in1[0][0] * in2[0][1] + in1[0][1] * in2[1][1] +
+ in1[0][2] * in2[2][1];
+ out[0][2] = in1[0][0] * in2[0][2] + in1[0][1] * in2[1][2] +
+ in1[0][2] * in2[2][2];
+ out[1][0] = in1[1][0] * in2[0][0] + in1[1][1] * in2[1][0] +
+ in1[1][2] * in2[2][0];
+ out[1][1] = in1[1][0] * in2[0][1] + in1[1][1] * in2[1][1] +
+ in1[1][2] * in2[2][1];
+ out[1][2] = in1[1][0] * in2[0][2] + in1[1][1] * in2[1][2] +
+ in1[1][2] * in2[2][2];
+ out[2][0] = in1[2][0] * in2[0][0] + in1[2][1] * in2[1][0] +
+ in1[2][2] * in2[2][0];
+ out[2][1] = in1[2][0] * in2[0][1] + in1[2][1] * in2[1][1] +
+ in1[2][2] * in2[2][1];
+ out[2][2] = in1[2][0] * in2[0][2] + in1[2][1] * in2[1][2] +
+ in1[2][2] * in2[2][2];
+}
+
+/*
+================
+VectorMatrixMultiply
+================
+*/
+void VectorMatrixMultiply( const vec3_t p, vec3_t m[ 3 ], vec3_t out )
+{
+ out[ 0 ] = m[ 0 ][ 0 ] * p[ 0 ] + m[ 1 ][ 0 ] * p[ 1 ] + m[ 2 ][ 0 ] * p[ 2 ];
+ out[ 1 ] = m[ 0 ][ 1 ] * p[ 0 ] + m[ 1 ][ 1 ] * p[ 1 ] + m[ 2 ][ 1 ] * p[ 2 ];
+ out[ 2 ] = m[ 0 ][ 2 ] * p[ 0 ] + m[ 1 ][ 2 ] * p[ 1 ] + m[ 2 ][ 2 ] * p[ 2 ];
+}
+
+
+void AngleVectors( const vec3_t angles, vec3_t forward, vec3_t right, vec3_t up) {
+ float angle;
+ static float sr, sp, sy, cr, cp, cy;
+ // static to help MS compiler fp bugs
+
+ angle = angles[YAW] * (M_PI*2 / 360);
+ sy = sin(angle);
+ cy = cos(angle);
+ angle = angles[PITCH] * (M_PI*2 / 360);
+ sp = sin(angle);
+ cp = cos(angle);
+ angle = angles[ROLL] * (M_PI*2 / 360);
+ sr = sin(angle);
+ cr = cos(angle);
+
+ if (forward)
+ {
+ forward[0] = cp*cy;
+ forward[1] = cp*sy;
+ forward[2] = -sp;
+ }
+ if (right)
+ {
+ right[0] = (-1*sr*sp*cy+-1*cr*-sy);
+ right[1] = (-1*sr*sp*sy+-1*cr*cy);
+ right[2] = -1*sr*cp;
+ }
+ if (up)
+ {
+ up[0] = (cr*sp*cy+-sr*-sy);
+ up[1] = (cr*sp*sy+-sr*cy);
+ up[2] = cr*cp;
+ }
+}
+
+/*
+** assumes "src" is normalized
+*/
+void PerpendicularVector( vec3_t dst, const vec3_t src )
+{
+ int pos;
+ int i;
+ float minelem = 1.0F;
+ vec3_t tempvec;
+
+ /*
+ ** find the smallest magnitude axially aligned vector
+ */
+ for ( pos = 0, i = 0; i < 3; i++ )
+ {
+ if ( fabs( src[i] ) < minelem )
+ {
+ pos = i;
+ minelem = fabs( src[i] );
+ }
+ }
+ tempvec[0] = tempvec[1] = tempvec[2] = 0.0F;
+ tempvec[pos] = 1.0F;
+
+ /*
+ ** project the point onto the plane defined by src
+ */
+ ProjectPointOnPlane( dst, tempvec, src );
+
+ /*
+ ** normalize the result
+ */
+ VectorNormalize( dst );
+}
+
+/*
+=================
+pointToLineDistance
+
+Distance from a point to some line
+=================
+*/
+float pointToLineDistance( const vec3_t p0, const vec3_t p1, const vec3_t p2 )
+{
+ vec3_t v, w, y;
+ float c1, c2;
+
+ VectorSubtract( p2, p1, v );
+ VectorSubtract( p1, p0, w );
+
+ CrossProduct( w, v, y );
+ c1 = VectorLength( y );
+ c2 = VectorLength( v );
+
+ if( c2 == 0.0f )
+ return 0.0f;
+ else
+ return c1 / c2;
+}
+
+/*
+=================
+GetPerpendicularViewVector
+
+Used to find an "up" vector for drawing a sprite so that it always faces the view as best as possible
+=================
+*/
+void GetPerpendicularViewVector( const vec3_t point, const vec3_t p1, const vec3_t p2, vec3_t up )
+{
+ vec3_t v1, v2;
+
+ VectorSubtract( point, p1, v1 );
+ VectorNormalize( v1 );
+
+ VectorSubtract( point, p2, v2 );
+ VectorNormalize( v2 );
+
+ CrossProduct( v1, v2, up );
+ VectorNormalize( up );
+}
+
+/*
+================
+ProjectPointOntoVector
+================
+*/
+void ProjectPointOntoVector( vec3_t point, vec3_t vStart, vec3_t vEnd, vec3_t vProj )
+{
+ vec3_t pVec, vec;
+
+ VectorSubtract( point, vStart, pVec );
+ VectorSubtract( vEnd, vStart, vec );
+ VectorNormalize( vec );
+ // project onto the directional vector for this segment
+ VectorMA( vStart, DotProduct( pVec, vec ), vec, vProj );
+}
+
+/*
+================
+VectorMaxComponent
+
+Return the biggest component of some vector
+================
+*/
+float VectorMaxComponent( vec3_t v )
+{
+ float biggest = v[ 0 ];
+
+ if( v[ 1 ] > biggest )
+ biggest = v[ 1 ];
+
+ if( v[ 2 ] > biggest )
+ biggest = v[ 2 ];
+
+ return biggest;
+}
+
+/*
+================
+VectorMinComponent
+
+Return the smallest component of some vector
+================
+*/
+float VectorMinComponent( vec3_t v )
+{
+ float smallest = v[ 0 ];
+
+ if( v[ 1 ] < smallest )
+ smallest = v[ 1 ];
+
+ if( v[ 2 ] < smallest )
+ smallest = v[ 2 ];
+
+ return smallest;
+}
+
+
+#define LINE_DISTANCE_EPSILON 1e-05f
+
+/*
+================
+DistanceBetweenLineSegmentsSquared
+
+Return the smallest distance between two line segments, squared
+================
+*/
+vec_t DistanceBetweenLineSegmentsSquared(
+ const vec3_t sP0, const vec3_t sP1,
+ const vec3_t tP0, const vec3_t tP1,
+ float *s, float *t )
+{
+ vec3_t sMag, tMag, diff;
+ float a, b, c, d, e;
+ float D;
+ float sN, sD;
+ float tN, tD;
+ vec3_t separation;
+
+ VectorSubtract( sP1, sP0, sMag );
+ VectorSubtract( tP1, tP0, tMag );
+ VectorSubtract( sP0, tP0, diff );
+ a = DotProduct( sMag, sMag );
+ b = DotProduct( sMag, tMag );
+ c = DotProduct( tMag, tMag );
+ d = DotProduct( sMag, diff );
+ e = DotProduct( tMag, diff );
+ sD = tD = D = a * c - b * b;
+
+ if( D < LINE_DISTANCE_EPSILON )
+ {
+ // the lines are almost parallel
+ sN = 0.0; // force using point P0 on segment S1
+ sD = 1.0; // to prevent possible division by 0.0 later
+ tN = e;
+ tD = c;
+ }
+ else
+ {
+ // get the closest points on the infinite lines
+ sN = ( b * e - c * d );
+ tN = ( a * e - b * d );
+
+ if( sN < 0.0 )
+ {
+ // sN < 0 => the s=0 edge is visible
+ sN = 0.0;
+ tN = e;
+ tD = c;
+ }
+ else if( sN > sD )
+ {
+ // sN > sD => the s=1 edge is visible
+ sN = sD;
+ tN = e + b;
+ tD = c;
+ }
+ }
+
+ if( tN < 0.0 )
+ {
+ // tN < 0 => the t=0 edge is visible
+ tN = 0.0;
+
+ // recompute sN for this edge
+ if( -d < 0.0 )
+ sN = 0.0;
+ else if( -d > a )
+ sN = sD;
+ else
+ {
+ sN = -d;
+ sD = a;
+ }
+ }
+ else if( tN > tD )
+ {
+ // tN > tD => the t=1 edge is visible
+ tN = tD;
+
+ // recompute sN for this edge
+ if( ( -d + b ) < 0.0 )
+ sN = 0;
+ else if( ( -d + b ) > a )
+ sN = sD;
+ else
+ {
+ sN = ( -d + b );
+ sD = a;
+ }
+ }
+
+ // finally do the division to get *s and *t
+ *s = ( fabs( sN ) < LINE_DISTANCE_EPSILON ? 0.0 : sN / sD );
+ *t = ( fabs( tN ) < LINE_DISTANCE_EPSILON ? 0.0 : tN / tD );
+
+ // get the difference of the two closest points
+ VectorScale( sMag, *s, sMag );
+ VectorScale( tMag, *t, tMag );
+ VectorAdd( diff, sMag, separation );
+ VectorSubtract( separation, tMag, separation );
+
+ return VectorLengthSquared( separation );
+}
+
+/*
+================
+DistanceBetweenLineSegments
+
+Return the smallest distance between two line segments
+================
+*/
+vec_t DistanceBetweenLineSegments(
+ const vec3_t sP0, const vec3_t sP1,
+ const vec3_t tP0, const vec3_t tP1,
+ float *s, float *t )
+{
+ return (vec_t)sqrt( DistanceBetweenLineSegmentsSquared(
+ sP0, sP1, tP0, tP1, s, t ) );
+}
+
+/*
+=================
+Q_isnan
+
+Don't pass doubles to this
+================
+*/
+int Q_isnan( float x )
+{
+ union
+ {
+ float f;
+ unsigned int i;
+ } t;
+
+ t.f = x;
+ t.i &= 0x7FFFFFFF;
+ t.i = 0x7F800000 - t.i;
+
+ return (int)( (unsigned int)t.i >> 31 );
+}