mirrors/raylib
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

optimized matrixrotate function by removing identity matrix

Browse Source
This commit is contained in:
user 2017-12-14 11:48:33 +01:00
parent b5395b22cc
commit d51f382ef8
1 changed files with 18 additions and 28 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

46
src/raymath.h
View File

@ -759,8 +759,6 @@ RMDEF Matrix MatrixRotate(Vector3 axis, float angle)
{
Matrix result;
Matrix mat = MatrixIdentity();
float x = axis.x, y = axis.y, z = axis.z;
float length = sqrtf(x*x + y*y + z*z);
@ -777,33 +775,25 @@ RMDEF Matrix MatrixRotate(Vector3 axis, float angle)
float cosres = cosf(angle);
float t = 1.0f - cosres;
// Cache some matrix values (speed optimization)
float a00 = mat.m0, a01 = mat.m1, a02 = mat.m2, a03 = mat.m3;
float a10 = mat.m4, a11 = mat.m5, a12 = mat.m6, a13 = mat.m7;
float a20 = mat.m8, a21 = mat.m9, a22 = mat.m10, a23 = mat.m11;
result.m0 = x*x*t + cosres;
result.m1 = y*x*t + z*sinres;
result.m2 = z*x*t - y*sinres;
result.m3 = 0.0f;
// Construct the elements of the rotation matrix
float b00 = x*x*t + cosres, b01 = y*x*t + z*sinres, b02 = z*x*t - y*sinres;
float b10 = x*y*t - z*sinres, b11 = y*y*t + cosres, b12 = z*y*t + x*sinres;
float b20 = x*z*t + y*sinres, b21 = y*z*t - x*sinres, b22 = z*z*t + cosres;
// Perform rotation-specific matrix multiplication
result.m0 = a00*b00 + a10*b01 + a20*b02;
result.m1 = a01*b00 + a11*b01 + a21*b02;
result.m2 = a02*b00 + a12*b01 + a22*b02;
result.m3 = a03*b00 + a13*b01 + a23*b02;
result.m4 = a00*b10 + a10*b11 + a20*b12;
result.m5 = a01*b10 + a11*b11 + a21*b12;
result.m6 = a02*b10 + a12*b11 + a22*b12;
result.m7 = a03*b10 + a13*b11 + a23*b12;
result.m8 = a00*b20 + a10*b21 + a20*b22;
result.m9 = a01*b20 + a11*b21 + a21*b22;
result.m10 = a02*b20 + a12*b21 + a22*b22;
result.m11 = a03*b20 + a13*b21 + a23*b22;
result.m12 = mat.m12;
result.m13 = mat.m13;
result.m14 = mat.m14;
result.m15 = mat.m15;
result.m4 = x*y*t - z*sinres;
result.m5 = y*y*t + cosres;
result.m6 = z*y*t + x*sinres;
result.m7 = 0.0f;
result.m8 = x*z*t + y*sinres;
result.m9 = y*z*t - x*sinres;
result.m10 = z*z*t + cosres;
result.m11 = 0.0f;
result.m12 = 0.0f;
result.m13 = 0.0f;
result.m14 = 0.0f;
result.m15 = 1.0f;
return result;
}