Skip to content
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

Jump to bottom

Dangerous matrix abstraction #4

Open
wants to merge 1 commit into
base: master
Choose a base branch
from
Open

Dangerous matrix abstraction #4

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
59 changes: 18 additions & 41 deletions swep1rcr.exe/vec.c
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -273,22 +273,7 @@ float *__cdecl sub_432030(float *a1, const float *a2) {
}

//----- (004323C0) --------------------------------------------------------
float* __cdecl sub_4323C0(float* a1, const float *a2) {
float v5; // edx
double v6; // st4
double v7; // rt2
double v8; // st5
double v11; // st7
double v12; // st7
double v13; // st2
double v14; // st7
float v15; // [esp+8h] [ebp+4h]
float v17; // [esp+8h] [ebp+4h]
float v18; // [esp+Ch] [ebp+8h]
float v20; // [esp+Ch] [ebp+8h]

//FIXME: This is a bit more complicated because a2 is still being read while a1 has been written.
// This means the order of operation suddenly becomes critical if regions in a1 and a2 overlap.
float* __cdecl sub_4323C0(float* restrict a1, const float* restrict a2) {

// Get squared length
float v19 = a2[0 * 4 + 0] * a2[0 * 4 + 0] +
Expand All @@ -301,43 +286,35 @@ float* __cdecl sub_4323C0(float* a1, const float *a2) {
a2[2 * 4 + 1] * a2[2 * 4 + 1] +
a2[2 * 4 + 2] * a2[2 * 4 + 2];

// Can't trust a2 after this (might have been written):
//FIXME: This is a bit more complicated because a2 is still being read after a1 has been written.
// This means the order of operation suddenly becomes critical if regions in a1 and a2 overlap.
// During this analysis I did break this order! So if the game depends on proper behaviour, it will break!
// Hence, the pointers have been marked as restricted.

a1[0 * 4 + 0] = a2[0 * 4 + 0] / v19;
a1[0 * 4 + 1] = a2[1 * 4 + 0] / v3;
a1[0 * 4 + 2] = a2[2 * 4 + 0] / v16;
a1[0 * 4 + 3] = 0.0f;

a1[1 * 4 + 0] = a2[0 * 4 + 1] / v19;
a1[1 * 4 + 1] = a2[1 * 4 + 1] / v3;
a1[1 * 4 + 2] = a2[2 * 4 + 1] / v16;
a1[1 * 4 + 3] = 0.0f;


a1[1 * 4 + 0] = a2[0 * 4 + 1] / v19;
a1[2 * 4 + 0] = a2[0 * 4 + 2] / v19;

a1[2 * 4 + 1] = a2[1 * 4 + 2] / v3;

a1[0 * 4 + 0] = a2[0 * 4 + 0] / v19;

a1[1 * 4 + 1] = a2[1 * 4 + 1] / v3;

v11 = a2[2 * 4 + 2] / v16;

a1[0 * 4 + 3] = 0.0f;
a1[1 * 4 + 3] = 0.0f;
a1[2 * 4 + 2] = a2[2 * 4 + 2] / v16;
a1[2 * 4 + 3] = 0.0f;
a1[3 * 4 + 3] = 1.0f;

a1[2 * 4 + 2] = v11;

v20 = a2[3 * 4 + 0];
v17 = a2[3 * 4 + 1];
v12 = a2[3 * 4 + 2];
float v20 = a2[3 * 4 + 0];
float v17 = a2[3 * 4 + 1];
double v12 = a2[3 * 4 + 2];

v13 = v12 * a1[2 * 4 + 1];

a1[3 * 4 + 0] = -(v17 * a1[1 * 4 + 0] + v12 * a1[2 * 4 + 0] + a1[0 * 4 + 0] * v20);
v14 = -(v17 * a1[1 * 4 + 2] + v20 * a1[0 * 4 + 2] + v12 * a1[2 * 4 + 2]);
a1[3 * 4 + 0] = -(v20 * a1[0 * 4 + 0] + v17 * a1[1 * 4 + 0] + v12 * a1[2 * 4 + 0]);
a1[3 * 4 + 1] = -(v20 * a1[0 * 4 + 1] + v17 * a1[1 * 4 + 1] + v12 * a1[2 * 4 + 1]);
a1[3 * 4 + 2] = -(v20 * a1[0 * 4 + 2] + v17 * a1[1 * 4 + 2] + v12 * a1[2 * 4 + 2]);
a1[3 * 4 + 3] = 1.0f;

a1[3 * 4 + 1] = -(v13 + v20 * a1[0 * 4 + 1] + v17 * a1[1 * 4 + 1];
a1[3 * 4 + 2] = v14;
return a1;
}

Expand Down