mirror of
https://git.code.sf.net/p/quake/quakeforge
synced 2024-11-29 23:52:22 +00:00
ef9b04ba83
It seems that i686 code generation is all over the place reguarding sse2 vs fp, with the resulting differences in carried precision. I'm not sure I'm happy with the situation, but at least it's being tested to a certain extent. Not sure if this broke basic (no sse) i686 tests.
473 lines
12 KiB
C
473 lines
12 KiB
C
#ifdef HAVE_CONFIG_H
|
|
# include "config.h"
|
|
#endif
|
|
|
|
#include "QF/mathlib.h"
|
|
|
|
static struct {
|
|
quat_t q1;
|
|
quat_t q2;
|
|
quat_t expect;
|
|
} quat_mult_tests[] = {
|
|
{{1, 2, 3, 4}, {5, 6, 7, 8}, {24, 48, 48, -6}},
|
|
};
|
|
#define num_quat_mult_tests (sizeof (quat_mult_tests) / sizeof (quat_mult_tests[0]))
|
|
|
|
//PITCH YAW ROLL
|
|
static vec3_t test_angles[] = {
|
|
{0, 0, 0},
|
|
{45, 0, 0},
|
|
{0, 45, 0},
|
|
{0, 0, 45},
|
|
{45, 45, 0},
|
|
{0, 45, 45},
|
|
{45, 0, 45},
|
|
{45, 45, 45},
|
|
{0, 180, 180},
|
|
{180, 0, 180},
|
|
{180, 180, 0},
|
|
};
|
|
#define num_angle_tests (sizeof (test_angles) / sizeof (test_angles[0]))
|
|
|
|
static struct {
|
|
vec3_t a;
|
|
vec3_t b;
|
|
quat_t expect;
|
|
} quat_vector_rotation_tests[] = {
|
|
{{1, 0, 0}, {1, 0, 0}, {0, 0, 0, 1}},
|
|
{{0, 1, 0}, {0, 1, 0}, {0, 0, 0, 1}},
|
|
{{0, 0, 1}, {0, 0, 1}, {0, 0, 0, 1}},
|
|
{{1, 0, 0}, {8, 0, 0}, {0, 0, 0, 1}},
|
|
{{0, 8, 0}, {0, 1, 0}, {0, 0, 0, 1}},
|
|
{{0, 0, 8}, {0, 0, 8}, {0, 0, 0, 1}},
|
|
{{1, 0, 0}, {-1, 0, 0}, {0, 0, 0, 0}}, // x, y, z = NaN
|
|
{{0, -1, 0}, {0, 1, 0}, {0, 0, 0, 0}}, // x, y, z = NaN
|
|
{{0, 0, 1}, {0, 0, -1}, {0, 0, 0, 0}}, // x, y, z = NaN
|
|
{{-1, 0, 0}, {8, 0, 0}, {0, 0, 0, 0}}, // x, y, z = NaN
|
|
{{0, 8, 0}, {0, -1, 0}, {0, 0, 0, 0}}, // x, y, z = NaN
|
|
{{0, 0, -8}, {0, 0, 8}, {0, 0, 0, 0}}, // x, y, z = NaN
|
|
// excessive for float, but if vec_t becomes double...
|
|
// 1/50 second orbiting JNSQ Kerbin at 120km altitude. While this has
|
|
// nothing to do with quakeforge (yet?), it came up when testing camera
|
|
// rotation in KSP and Unity failed miserably. I don't remember the exact
|
|
// details, but I could see significant snapping in the rotation (I thought
|
|
// it was a few times per second, but these numbers indicate it must have
|
|
// been every few seconds).
|
|
// The quaternion is unit to 9 sigfigs (a little larger)
|
|
{{1720000, 0, 0}, {1719999.9983028059, 76.409082595828366, 0}, {0, 0, 2.22119434e-5, 1}},
|
|
// 1/20 second, same situation
|
|
{{1720000, 0, 0}, {1719999.9893925365, 191.02270615971355, 0}, {0, 0, 5.55298575e-5, 1}},
|
|
// 1/5 second, same situation
|
|
{{1720000, 0, 0}, {1719999.8302805868, 764.09080107761736, 0}, {0, 0, 2.2211943e-4, 1}},
|
|
// 1/4 second, same situation
|
|
{{1720000, 0, 0}, {1719999.7348134194, 955.11348367609469, 0}, {0, 0, 2.77649262e-4, 1}},
|
|
// 1/3 second, same situation. This is (float) 1 ulp in w: about 0.0424
|
|
// degrees.
|
|
// The quaternion is unit to 9 sigfigs (a little larger)
|
|
{{1720000, 0, 0}, {1719999.5285571995, 1273.4845939975546, 0}, {0, 0, 3.70199094e-4, 0.99999994}},
|
|
};
|
|
#define num_quat_vector_rotation_tests (sizeof (quat_vector_rotation_tests) / sizeof (quat_vector_rotation_tests[0]))
|
|
|
|
// return true if a and b are close enough (yay, floats)
|
|
static int
|
|
compare (vec_t a, vec_t b)
|
|
{
|
|
vec_t diff = a - b;
|
|
return diff * diff < 0.001;
|
|
}
|
|
|
|
static int
|
|
test_quat_mult(const quat_t q1, const quat_t q2, const quat_t expect)
|
|
{
|
|
int i;
|
|
quat_t r;
|
|
|
|
QuatMult (q1, q2, r);
|
|
|
|
for (i = 0; i < 4; i++)
|
|
if (!compare (r[i], expect[i]))
|
|
goto fail;
|
|
return 1;
|
|
fail:
|
|
printf ("%11.9g %11.9g %11.9g %11.9g\n", QuatExpand (q1));
|
|
printf ("%11.9g %11.9g %11.9g %11.9g\n", QuatExpand (q2));
|
|
printf ("%11.9g %11.9g %11.9g %11.9g\n", QuatExpand (r));
|
|
printf ("%11.9g %11.9g %11.9g %11.9g\n", QuatExpand (expect));
|
|
return 0;
|
|
}
|
|
|
|
|
|
static void
|
|
rotate_vec (const quat_t r, const vec3_t v, vec3_t out)
|
|
{
|
|
quat_t qv = {0, 0, 0, 0};
|
|
quat_t t;
|
|
|
|
VectorCopy (v, qv);
|
|
|
|
QuatConj (r, t);
|
|
QuatMult (qv, t, t);
|
|
QuatMult (r, t, t);
|
|
VectorCopy (t, out);
|
|
}
|
|
|
|
static int
|
|
test_rotation (const vec3_t angles)
|
|
{
|
|
int i;
|
|
vec3_t forward, right, up;
|
|
|
|
quat_t quat, conj, f, r, u, t;
|
|
quat_t qf = {1, 0, 0, 0};
|
|
quat_t qr = {0, -1, 0, 0};
|
|
quat_t qu = {0, 0, 1, 0};
|
|
|
|
AngleVectors (angles, forward, right, up);
|
|
|
|
AngleQuat (angles, quat);
|
|
QuatConj (quat, conj);
|
|
// rotate forward vector
|
|
QuatMult (qf, conj, t);
|
|
QuatMult (quat, t, f);
|
|
// rotate right vector
|
|
QuatMult (qr, conj, t);
|
|
QuatMult (quat, t, r);
|
|
// rotate up vector
|
|
QuatMult (qu, conj, t);
|
|
QuatMult (quat, t, u);
|
|
|
|
if (!compare (f[3], 0))
|
|
goto fail;
|
|
for (i = 0; i < 3; i++)
|
|
if (!compare (forward[i], f[i]))
|
|
goto fail;
|
|
|
|
if (!compare (r[3], 0))
|
|
goto fail;
|
|
for (i = 0; i < 3; i++)
|
|
if (!compare (right[i], r[i]))
|
|
goto fail;
|
|
|
|
if (!compare (u[3], 0))
|
|
goto fail;
|
|
for (i = 0; i < 3; i++)
|
|
if (!compare (up[i], u[i]))
|
|
goto fail;
|
|
return 1;
|
|
fail:
|
|
printf ("\ntest_rotation\n");
|
|
printf ("%11.9g %11.9g %11.9g\n", VectorExpand (angles));
|
|
printf ("%11.9g %11.9g %11.9g %11.9g\n", QuatExpand (quat));
|
|
printf ("%11.9g %11.9g %11.9g %11.9g\n\n", QuatExpand (conj));
|
|
printf ("%11.9g %11.9g %11.9g\n", VectorExpand (forward));
|
|
printf ("%11.9g %11.9g %11.9g\n", VectorExpand (right));
|
|
printf ("%11.9g %11.9g %11.9g\n\n", VectorExpand (up));
|
|
|
|
printf ("%11.9g %11.9g %11.9g %11.9g\n", QuatExpand (f));
|
|
printf ("%11.9g %11.9g %11.9g %11.9g\n", QuatExpand (r));
|
|
printf ("%11.9g %11.9g %11.9g %11.9g\n", QuatExpand (u));
|
|
return 0;
|
|
}
|
|
|
|
static int
|
|
test_rotation2 (const vec3_t angles)
|
|
{
|
|
int i;
|
|
vec3_t forward, right, up;
|
|
|
|
quat_t quat;
|
|
vec3_t f, r, u;
|
|
vec3_t vf = {1, 0, 0};
|
|
vec3_t vr = {0, -1, 0};
|
|
vec3_t vu = {0, 0, 1};
|
|
|
|
AngleVectors (angles, forward, right, up);
|
|
|
|
AngleQuat (angles, quat);
|
|
// rotate forward vector
|
|
QuatMultVec (quat, vf, f);
|
|
// rotate right vector
|
|
QuatMultVec (quat, vr, r);
|
|
// rotate up vector
|
|
QuatMultVec (quat, vu, u);
|
|
|
|
for (i = 0; i < 3; i++)
|
|
if (!compare (forward[i], f[i]))
|
|
goto fail;
|
|
|
|
for (i = 0; i < 3; i++)
|
|
if (!compare (right[i], r[i]))
|
|
goto fail;
|
|
|
|
for (i = 0; i < 3; i++)
|
|
if (!compare (up[i], u[i]))
|
|
goto fail;
|
|
return 1;
|
|
fail:
|
|
printf ("\ntest_rotation2\n");
|
|
printf ("\n\n%11.9g %11.9g %11.9g\n\n", angles[0], angles[1], angles[2]);
|
|
printf ("%11.9g %11.9g %11.9g\n", forward[0], forward[1], forward[2]);
|
|
printf ("%11.9g %11.9g %11.9g\n", right[0], right[1], right[2]);
|
|
printf ("%11.9g %11.9g %11.9g\n\n", up[0], up[1], up[2]);
|
|
|
|
printf ("%11.9g %11.9g %11.9g\n", f[0], f[1], f[2]);
|
|
printf ("%11.9g %11.9g %11.9g\n", r[0], r[1], r[2]);
|
|
printf ("%11.9g %11.9g %11.9g\n", u[0], u[1], u[2]);
|
|
return 0;
|
|
}
|
|
|
|
static int
|
|
test_rotation3 (const vec3_t angles)
|
|
{
|
|
int i;
|
|
quat_t quat;
|
|
vec3_t v = {1, 1, 1};
|
|
vec3_t a, b;
|
|
|
|
AngleQuat (angles, quat);
|
|
QuatMultVec (quat, v, a);
|
|
rotate_vec (quat, v, b);
|
|
|
|
for (i = 0; i < 3; i++)
|
|
if (!compare (a[i], b[i]))
|
|
goto fail;
|
|
return 1;
|
|
fail:
|
|
printf ("\ntest_rotation3\n");
|
|
printf ("%11.9g %11.9g %11.9g\n", VectorExpand(a));
|
|
printf ("%11.9g %11.9g %11.9g\n", VectorExpand(b));
|
|
return 0;
|
|
}
|
|
|
|
// XXX FIXME see usage in test_rotation4. need to investigate whether
|
|
// -ffast-math is any real benefit
|
|
#define ISNAN(x) (((x) & 0x7f800000) == 0x7f800000 && ((x) & 0x7fffff))
|
|
|
|
// FIXME differences in precision between archs
|
|
static int
|
|
cmp (float a, float b)
|
|
{
|
|
typedef union {
|
|
float f;
|
|
int i;
|
|
} fi;
|
|
fi ax;
|
|
fi bx;
|
|
int x;
|
|
|
|
ax.f = a;
|
|
bx.f = b;
|
|
x = ax.i - bx.i;
|
|
if (x < 0) {
|
|
x = -x;
|
|
}
|
|
return (x & 0x7ffffffc) == 0;
|
|
}
|
|
|
|
static int
|
|
test_rotation4 (const vec3_t a, const vec3_t b, const quat_t expect)
|
|
{
|
|
int i;
|
|
union { int x[4]; vec_t q[4]; } q;
|
|
vec_t *quat = q.q;
|
|
vec3_t t;
|
|
vec_t d = 0;
|
|
|
|
VectorZero (t);
|
|
|
|
// find the rotation vector between a and b
|
|
QuatRotation (a, b, quat);
|
|
|
|
if (quat[3] == 0) {
|
|
if (expect[3] != 0) {
|
|
goto fail;
|
|
}
|
|
// expect NaN for the vector components because the vectors are
|
|
// anti-parallel and thus the rotation axis is undefined
|
|
//XXX FIXME(?) still using -ffast-math which implies
|
|
// -ffinite-math-only which in turn disables nan/inf checks, so have
|
|
// to do it by hand
|
|
// if (!(isnan(quat[0]) && isnan(quat[1]) && isnan(quat[2]))) {
|
|
if (!(ISNAN(q.x[0]) && ISNAN(q.x[1]) && ISNAN(q.x[2]))) {
|
|
goto fail;
|
|
}
|
|
} else {
|
|
// the vectors are not anti-parallel and thus the rotation axis is
|
|
// defined, so NaN is invalid
|
|
// XXX FIXME see above
|
|
//if (isnan(quat[0]) || isnan(quat[1]) || isnan(quat[2])) {
|
|
if (ISNAN(q.x[0]) || ISNAN(q.x[1]) || ISNAN(q.x[2])) {
|
|
goto fail;
|
|
}
|
|
for (i = 0; i < 4; i++) {
|
|
// yes, float precision will make it difficult to set up expect
|
|
// but it is at least consistent (ie, the "errors" are not at all
|
|
// random and thus will be the same from run to run)
|
|
if (!cmp (quat[i], expect[i])) {
|
|
goto fail;
|
|
}
|
|
}
|
|
QuatMultVec(quat, a, t);
|
|
|
|
d = DotProduct (t, b) / (VectorLength (t) * VectorLength (b)) - 1;
|
|
if (d * d > 1e-8) {
|
|
goto fail;
|
|
}
|
|
}
|
|
return 1;
|
|
fail:
|
|
printf ("\ntest_rotation4\n");
|
|
printf ("a: %11.9g %11.9g %11.9g\n", VectorExpand(a));
|
|
printf ("b: %11.9g %11.9g %11.9g\n", VectorExpand(b));
|
|
printf ("q: %11.9g %11.9g %11.9g %11.9g\n", QuatExpand(quat));
|
|
printf ("e: %11.9g %11.9g %11.9g %11.9g\n", QuatExpand(expect));
|
|
printf ("t: %11.9g %11.9g %11.9g\n", VectorExpand(t));
|
|
printf ("d: %11.9g\n", d);
|
|
return 0;
|
|
}
|
|
|
|
#define s05 0.70710678118654757
|
|
|
|
static struct {
|
|
quat_t q;
|
|
vec_t expect[9];
|
|
} quat_mat_tests[] = {
|
|
{{0, 0, 0, 1},
|
|
{1, 0, 0,
|
|
0, 1, 0,
|
|
0, 0, 1}},
|
|
{{1, 0, 0, 0},
|
|
{1, 0, 0,
|
|
0, -1, 0,
|
|
0, 0, -1}},
|
|
{{0, 1, 0, 0},
|
|
{-1, 0, 0,
|
|
0, 1, 0,
|
|
0, 0, -1}},
|
|
{{0, 0, 1, 0},
|
|
{-1, 0, 0,
|
|
0, -1, 0,
|
|
0, 0, 1}},
|
|
{{0.5, 0.5, 0.5, 0.5},
|
|
{0, 0, 1,
|
|
1, 0, 0,
|
|
0, 1, 0}},
|
|
#if defined(__i686__) && defined(__OPTIMIZE__)
|
|
// the fp unit carries more precision than a 32-bit float, so
|
|
// the close-to-zero errors are different
|
|
{{s05, 0.0, 0.0, s05},
|
|
{1, 0, 0,
|
|
0, 3.42285418e-08, -0.99999994,
|
|
0, 0.99999994, 3.42285418e-08}},
|
|
#else
|
|
{{s05, 0.0, 0.0, s05},
|
|
{1, 0, 0,
|
|
0, 5.96046448e-8, -0.99999994,
|
|
0, 0.99999994, 5.96046448e-8}},
|
|
#endif
|
|
};
|
|
#define num_quat_mat_tests (sizeof (quat_mat_tests) / sizeof (quat_mat_tests[0]))
|
|
|
|
static int
|
|
test_quat_mat(const quat_t q, const quat_t expect)
|
|
{
|
|
int i;
|
|
vec_t m[9];
|
|
|
|
QuatToMatrix (q, m, 0, 0);
|
|
|
|
for (i = 0; i < 9; i++)
|
|
if (m[i] != expect[i]) // exact tests here
|
|
goto fail;
|
|
return 1;
|
|
fail:
|
|
printf ("\ntest_quat_mat\n");
|
|
printf ("%11.9g %11.9g %11.9g %11.9g\n", QuatExpand (q));
|
|
printf ("%11.9g %11.9g %11.9g %11.9g %11.9g %11.9g\n",
|
|
VectorExpand (m + 0), VectorExpand (expect + 0));
|
|
printf ("%11.9g %11.9g %11.9g %11.9g %11.9g %11.9g\n",
|
|
VectorExpand (m + 3), VectorExpand (expect + 3));
|
|
printf ("%11.9g %11.9g %11.9g %11.9g %11.9g %11.9g\n",
|
|
VectorExpand (m + 6), VectorExpand (expect + 6));
|
|
return 0;
|
|
}
|
|
|
|
static int
|
|
test_quat_mat2(const quat_t q, const quat_t expect)
|
|
{
|
|
int i;
|
|
vec_t m[9];
|
|
|
|
QuatToMatrix (q, m, 0, 1);
|
|
Mat3Transpose (m, m);
|
|
|
|
for (i = 0; i < 9; i++)
|
|
if (m[i] != expect[i]) // exact tests here
|
|
goto fail;
|
|
return 1;
|
|
fail:
|
|
printf ("\ntest_quat_mat\n");
|
|
printf ("%11.9g %11.9g %11.9g %11.9g\n", QuatExpand (q));
|
|
printf ("%11.9g %11.9g %11.9g %11.9g %11.9g %11.9g\n",
|
|
VectorExpand (m + 0), VectorExpand (expect + 0));
|
|
printf ("%11.9g %11.9g %11.9g %11.9g %11.9g %11.9g\n",
|
|
VectorExpand (m + 3), VectorExpand (expect + 3));
|
|
printf ("%11.9g %11.9g %11.9g %11.9g %11.9g %11.9g\n",
|
|
VectorExpand (m + 6), VectorExpand (expect + 6));
|
|
return 0;
|
|
}
|
|
|
|
int
|
|
main (int argc, const char **argv)
|
|
{
|
|
int res = 0;
|
|
size_t i;
|
|
|
|
for (i = 0; i < num_quat_mult_tests; i ++) {
|
|
vec_t *q1 = quat_mult_tests[i].q1;
|
|
vec_t *q2 = quat_mult_tests[i].q2;
|
|
vec_t *expect = quat_mult_tests[i].expect;
|
|
if (!test_quat_mult (q1, q2, expect))
|
|
res = 1;
|
|
}
|
|
|
|
for (i = 0; i < num_angle_tests; i ++) {
|
|
if (!test_rotation (test_angles[i]))
|
|
res = 1;
|
|
}
|
|
|
|
for (i = 0; i < num_angle_tests; i ++) {
|
|
if (!test_rotation2 (test_angles[i]))
|
|
res = 1;
|
|
}
|
|
|
|
for (i = 0; i < num_angle_tests; i ++) {
|
|
if (!test_rotation3 (test_angles[i]))
|
|
res = 1;
|
|
}
|
|
|
|
for (i = 0; i < num_quat_vector_rotation_tests; i++) {
|
|
vec_t *a = quat_vector_rotation_tests[i].a;
|
|
vec_t *b = quat_vector_rotation_tests[i].b;
|
|
vec_t *expect = quat_vector_rotation_tests[i].expect;
|
|
if (!test_rotation4 (a, b, expect)) {
|
|
res = 1;
|
|
}
|
|
}
|
|
|
|
for (i = 0; i < num_quat_mat_tests; i ++) {
|
|
vec_t *q = quat_mat_tests[i].q;
|
|
vec_t *expect = quat_mat_tests[i].expect;
|
|
if (!test_quat_mat (q, expect))
|
|
res = 1;
|
|
}
|
|
|
|
for (i = 0; i < num_quat_mat_tests; i ++) {
|
|
vec_t *q = quat_mat_tests[i].q;
|
|
vec_t *expect = quat_mat_tests[i].expect;
|
|
if (!test_quat_mat2 (q, expect))
|
|
res = 1;
|
|
}
|
|
|
|
return res;
|
|
}
|