Skip to content

⚡️ Speed up method Matrix4f.invertLocal by 45% #41

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
codeflash-ai wants to merge 1 commit into
base: master
Choose a base branch
from
+3,302 −3,299
Open

⚡️ Speed up method Matrix4f.invertLocal by 45% #41

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
97 changes: 50 additions & 47 deletions jme3-core/src/main/java/com/jme3/math/Matrix4f.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1546,61 +1546,64 @@ public Matrix4f invert(Matrix4f store) {
* @return the (inverted) current instance (for chaining)
*/
public Matrix4f invertLocal() {

float fA0 = m00 * m11 - m01 * m10;
float fA1 = m00 * m12 - m02 * m10;
float fA2 = m00 * m13 - m03 * m10;
float fA3 = m01 * m12 - m02 * m11;
float fA4 = m01 * m13 - m03 * m11;
float fA5 = m02 * m13 - m03 * m12;
float fB0 = m20 * m31 - m21 * m30;
float fB1 = m20 * m32 - m22 * m30;
float fB2 = m20 * m33 - m23 * m30;
float fB3 = m21 * m32 - m22 * m31;
float fB4 = m21 * m33 - m23 * m31;
float fB5 = m22 * m33 - m23 * m32;
// Store matrix elements in local variables to reduce field access overhead
float l_m00 = m00, l_m01 = m01, l_m02 = m02, l_m03 = m03;
float l_m10 = m10, l_m11 = m11, l_m12 = m12, l_m13 = m13;
float l_m20 = m20, l_m21 = m21, l_m22 = m22, l_m23 = m23;
float l_m30 = m30, l_m31 = m31, l_m32 = m32, l_m33 = m33;

float fA0 = l_m00 * l_m11 - l_m01 * l_m10;
float fA1 = l_m00 * l_m12 - l_m02 * l_m10;
float fA2 = l_m00 * l_m13 - l_m03 * l_m10;
float fA3 = l_m01 * l_m12 - l_m02 * l_m11;
float fA4 = l_m01 * l_m13 - l_m03 * l_m11;
float fA5 = l_m02 * l_m13 - l_m03 * l_m12;
float fB0 = l_m20 * l_m31 - l_m21 * l_m30;
float fB1 = l_m20 * l_m32 - l_m22 * l_m30;
float fB2 = l_m20 * l_m33 - l_m23 * l_m30;
float fB3 = l_m21 * l_m32 - l_m22 * l_m31;
float fB4 = l_m21 * l_m33 - l_m23 * l_m31;
float fB5 = l_m22 * l_m33 - l_m23 * l_m32;
float fDet = fA0 * fB5 - fA1 * fB4 + fA2 * fB3 + fA3 * fB2 - fA4 * fB1 + fA5 * fB0;

if (FastMath.abs(fDet) <= 0f) {
return zero();
}

float f00 = +m11 * fB5 - m12 * fB4 + m13 * fB3;
float f10 = -m10 * fB5 + m12 * fB2 - m13 * fB1;
float f20 = +m10 * fB4 - m11 * fB2 + m13 * fB0;
float f30 = -m10 * fB3 + m11 * fB1 - m12 * fB0;
float f01 = -m01 * fB5 + m02 * fB4 - m03 * fB3;
float f11 = +m00 * fB5 - m02 * fB2 + m03 * fB1;
float f21 = -m00 * fB4 + m01 * fB2 - m03 * fB0;
float f31 = +m00 * fB3 - m01 * fB1 + m02 * fB0;
float f02 = +m31 * fA5 - m32 * fA4 + m33 * fA3;
float f12 = -m30 * fA5 + m32 * fA2 - m33 * fA1;
float f22 = +m30 * fA4 - m31 * fA2 + m33 * fA0;
float f32 = -m30 * fA3 + m31 * fA1 - m32 * fA0;
float f03 = -m21 * fA5 + m22 * fA4 - m23 * fA3;
float f13 = +m20 * fA5 - m22 * fA2 + m23 * fA1;
float f23 = -m20 * fA4 + m21 * fA2 - m23 * fA0;
float f33 = +m20 * fA3 - m21 * fA1 + m22 * fA0;

m00 = f00;
m01 = f01;
m02 = f02;
m03 = f03;
m10 = f10;
m11 = f11;
m12 = f12;
m13 = f13;
m20 = f20;
m21 = f21;
m22 = f22;
m23 = f23;
m30 = f30;
m31 = f31;
m32 = f32;
m33 = f33;
float f00 = +l_m11 * fB5 - l_m12 * fB4 + l_m13 * fB3;
float f10 = -l_m10 * fB5 + l_m12 * fB2 - l_m13 * fB1;
float f20 = +l_m10 * fB4 - l_m11 * fB2 + l_m13 * fB0;
float f30 = -l_m10 * fB3 + l_m11 * fB1 - l_m12 * fB0;
float f01 = -l_m01 * fB5 + l_m02 * fB4 - l_m03 * fB3;
float f11 = +l_m00 * fB5 - l_m02 * fB2 + l_m03 * fB1;
float f21 = -l_m00 * fB4 + l_m01 * fB2 - l_m03 * fB0;
float f31 = +l_m00 * fB3 - l_m01 * fB1 + l_m02 * fB0;
float f02 = +l_m31 * fA5 - l_m32 * fA4 + l_m33 * fA3;
float f12 = -l_m30 * fA5 + l_m32 * fA2 - l_m33 * fA1;
float f22 = +l_m30 * fA4 - l_m31 * fA2 + l_m33 * fA0;
float f32 = -l_m30 * fA3 + l_m31 * fA1 - l_m32 * fA0;
float f03 = -l_m21 * fA5 + l_m22 * fA4 - l_m23 * fA3;
float f13 = +l_m20 * fA5 - l_m22 * fA2 + l_m23 * fA1;
float f23 = -l_m20 * fA4 + l_m21 * fA2 - l_m23 * fA0;
float f33 = +l_m20 * fA3 - l_m21 * fA1 + l_m22 * fA0;

float fInvDet = 1.0f / fDet;
multLocal(fInvDet);
m00 = f00 * fInvDet;
m01 = f01 * fInvDet;
m02 = f02 * fInvDet;
m03 = f03 * fInvDet;
m10 = f10 * fInvDet;
m11 = f11 * fInvDet;
m12 = f12 * fInvDet;
m13 = f13 * fInvDet;
m20 = f20 * fInvDet;
m21 = f21 * fInvDet;
m22 = f22 * fInvDet;
m23 = f23 * fInvDet;
m30 = f30 * fInvDet;
m31 = f31 * fInvDet;
m32 = f32 * fInvDet;
m33 = f33 * fInvDet;

return this;
}
Expand Down
Loading