From 009d3fbcf7e1e3ea94656b5233e0bc96fa7b5a5b Mon Sep 17 00:00:00 2001
From: Made <made111@gmx.de>
Date: Fri, 1 May 2020 18:59:48 +0200
Subject: decompile FX.s, FX_atan.s and partially decompile FX_ves.s

---
 arm9/arm9.lcf           |   1 +
 arm9/asm/FX_vec.s       | 181 ++++++++++++++++++++++++------------------------
 arm9/lib/src/FX.c       |  20 ++++++
 arm9/lib/src/FX_atan.c  | 161 ++++++++++++++++++++++++++++++++++++++++++
 arm9/lib/src/FX_vec_c.c |  34 +++++++++
 include/fx.h            |  41 +++++++++++
 6 files changed, 348 insertions(+), 90 deletions(-)
 create mode 100644 arm9/lib/src/FX.c
 create mode 100644 arm9/lib/src/FX_atan.c
 create mode 100644 arm9/lib/src/FX_vec_c.c
 create mode 100644 include/fx.h

diff --git a/arm9/arm9.lcf b/arm9/arm9.lcf
index 42b0a8ee..046ce41d 100644
--- a/arm9/arm9.lcf
+++ b/arm9/arm9.lcf
@@ -127,6 +127,7 @@ SECTIONS {
 		FX_mtx44.o (.text)
 		FX_cp.o (.text)
 		FX_vec.o (.text)
+        FX_vec_c.o (.text)
 		FX_atan.o (.text)
 		FX.o (.text)
 		GX.o (.text)
diff --git a/arm9/asm/FX_vec.s b/arm9/asm/FX_vec.s
index e6270162..08068eaf 100644
--- a/arm9/asm/FX_vec.s
+++ b/arm9/asm/FX_vec.s
@@ -310,93 +310,94 @@ VEC_CrossProduct:
 	ldmia sp!, {r4-r11,lr}
 	bx lr
 
-	arm_func_start VEC_Fx16DotProduct
-VEC_Fx16DotProduct: ; 0x020C5538
-	stmdb sp!, {r4-r5,lr}
-	sub sp, sp, #0x4
-	ldrsh lr, [r0, #0x2]
-	ldrsh r12, [r1, #0x2]
-	ldrsh r3, [r0, #0x4]
-	ldrsh r2, [r1, #0x4]
-	ldrsh r5, [r0, #0x0]
-	ldrsh r4, [r1, #0x0]
-	smulbb r1, lr, r12
-	smulbb r0, r3, r2
-	smlabb r1, r5, r4, r1
-	add r0, r0, #0x800
-	adds r2, r1, r0
-	mov r1, r1, asr #0x1f
-	adc r1, r1, r0, asr #0x1f
-	mov r0, r2, lsr #0xc
-	orr r0, r0, r1, lsl #0x14
-	add sp, sp, #0x4
-	ldmia sp!, {r4-r5,lr}
-	bx lr
-
-	arm_func_start VEC_DotProduct
-VEC_DotProduct:
-	stmdb sp!, {r4,lr}
-	ldr r3, [r0, #0x4]
-	ldr r2, [r1, #0x4]
-	ldr r12, [r0, #0x0]
-	smull r4, lr, r3, r2
-	ldr r2, [r1, #0x0]
-	ldr r3, [r0, #0x8]
-	smlal r4, lr, r12, r2
-	ldr r1, [r1, #0x8]
-	mov r0, #0x800
-	smlal r4, lr, r3, r1
-	adds r0, r4, r0
-	adc r1, lr, #0x0
-	mov r0, r0, lsr #0xc
-	orr r0, r0, r1, lsl #0x14
-	ldmia sp!, {r4,lr}
-	bx lr
-
-	arm_func_start VEC_Fx16Add
-VEC_Fx16Add: ; 0x020C55CC
-	ldrsh r12, [r0, #0x0]
-	ldrsh r3, [r1, #0x0]
-	add r3, r12, r3
-	strh r3, [r2, #0x0]
-	ldrsh r12, [r0, #0x2]
-	ldrsh r3, [r1, #0x2]
-	add r3, r12, r3
-	strh r3, [r2, #0x2]
-	ldrsh r3, [r0, #0x4]
-	ldrsh r0, [r1, #0x4]
-	add r0, r3, r0
-	strh r0, [r2, #0x4]
-	bx lr
-
-	arm_func_start VEC_Subtract
-VEC_Subtract: ; 0x020C5600
-	ldr r12, [r0, #0x0]
-	ldr r3, [r1, #0x0]
-	sub r3, r12, r3
-	str r3, [r2, #0x0]
-	ldr r12, [r0, #0x4]
-	ldr r3, [r1, #0x4]
-	sub r3, r12, r3
-	str r3, [r2, #0x4]
-	ldr r3, [r0, #0x8]
-	ldr r0, [r1, #0x8]
-	sub r0, r3, r0
-	str r0, [r2, #0x8]
-	bx lr
-
-	arm_func_start VEC_Add
-VEC_Add: ; 0x020C5634
-	ldr r12, [r0, #0x0]
-	ldr r3, [r1, #0x0]
-	add r3, r12, r3
-	str r3, [r2, #0x0]
-	ldr r12, [r0, #0x4]
-	ldr r3, [r1, #0x4]
-	add r3, r12, r3
-	str r3, [r2, #0x4]
-	ldr r3, [r0, #0x8]
-	ldr r0, [r1, #0x8]
-	add r0, r3, r0
-	str r0, [r2, #0x8]
-	bx lr
+;	arm_func_start VEC_Fx16DotProduct
+;VEC_Fx16DotProduct: ; 0x020C5538
+;	stmdb sp!, {r4-r5,lr}
+;	sub sp, sp, #0x4
+;	ldrsh lr, [r0, #0x2]
+;	ldrsh r12, [r1, #0x2]
+;	ldrsh r3, [r0, #0x4]
+;	ldrsh r2, [r1, #0x4]
+;	ldrsh r5, [r0, #0x0]
+;	ldrsh r4, [r1, #0x0]
+;	smulbb r1, lr, r12
+;	smulbb r0, r3, r2
+;	smlabb r1, r5, r4, r1
+;	add r0, r0, #0x800
+;	adds r2, r1, r0
+;	mov r1, r1, asr #0x1f
+;	adc r1, r1, r0, asr #0x1f
+;	mov r0, r2, lsr #0xc
+;	orr r0, r0, r1, lsl #0x14
+;	add sp, sp, #0x4
+;	ldmia sp!, {r4-r5,lr}
+;	bx lr
+;
+;	arm_func_start VEC_DotProduct
+;VEC_DotProduct:
+;	stmdb sp!, {r4,lr}
+;	ldr r3, [r0, #0x4]
+;	ldr r2, [r1, #0x4]
+;	ldr r12, [r0, #0x0]
+;	smull r4, lr, r3, r2
+;	ldr r2, [r1, #0x0]
+;	ldr r3, [r0, #0x8]
+;	smlal r4, lr, r12, r2
+;	ldr r1, [r1, #0x8]
+;	mov r0, #0x800
+;	smlal r4, lr, r3, r1
+;	adds r0, r4, r0
+;	adc r1, lr, #0x0
+;	mov r0, r0, lsr #0xc
+;	orr r0, r0, r1, lsl #0x14
+;	ldmia sp!, {r4,lr}
+;	bx lr
+;
+;	arm_func_start VEC_Fx16Add
+;VEC_Fx16Add: ; 0x020C55CC
+;	ldrsh r12, [r0, #0x0]
+;	ldrsh r3, [r1, #0x0]
+;	add r3, r12, r3
+;	strh r3, [r2, #0x0]
+;	ldrsh r12, [r0, #0x2]
+;	ldrsh r3, [r1, #0x2]
+;	add r3, r12, r3
+;	strh r3, [r2, #0x2]
+;	ldrsh r3, [r0, #0x4]
+;	ldrsh r0, [r1, #0x4]
+;	add r0, r3, r0
+;	strh r0, [r2, #0x4]
+;	bx lr
+;
+;	arm_func_start VEC_Subtract
+;VEC_Subtract: ; 0x020C5600
+;	ldr r12, [r0, #0x0]
+;	ldr r3, [r1, #0x0]
+;	sub r3, r12, r3
+;	str r3, [r2, #0x0]
+;	ldr r12, [r0, #0x4]
+;	ldr r3, [r1, #0x4]
+;	sub r3, r12, r3
+;	str r3, [r2, #0x4]
+;	ldr r3, [r0, #0x8]
+;	ldr r0, [r1, #0x8]
+;	sub r0, r3, r0
+;	str r0, [r2, #0x8]
+;	bx lr
+;
+;	arm_func_start VEC_Add
+;VEC_Add: ; 0x020C5634
+;	ldr r12, [r0, #0x0]
+;	ldr r3, [r1, #0x0]
+;	add r3, r12, r3
+;	str r3, [r2, #0x0]
+;	ldr r12, [r0, #0x4]
+;	ldr r3, [r1, #0x4]
+;	add r3, r12, r3
+;	str r3, [r2, #0x4]
+;	ldr r3, [r0, #0x8]
+;	ldr r0, [r1, #0x8]
+;	add r0, r3, r0
+;	str r0, [r2, #0x8]
+;	bx lr
+;
diff --git a/arm9/lib/src/FX.c b/arm9/lib/src/FX.c
new file mode 100644
index 00000000..fa6fa6ba
--- /dev/null
+++ b/arm9/lib/src/FX.c
@@ -0,0 +1,20 @@
+#include "global.h"
+#include "main.h"
+#include "fx.h"
+
+void FX_Init(){
+    return;
+}
+
+s32 FX_Modf(s32 x, s32 *iptr){
+    if (x >= 0)
+    {
+        *iptr = x & 0x7FFFF000;
+        return x & 0xFFF;
+    }
+    else
+    {
+        *iptr = -(-x & 0x7FFFF000);
+        return -(-x & 0xFFF);
+    }
+}
diff --git a/arm9/lib/src/FX_atan.c b/arm9/lib/src/FX_atan.c
new file mode 100644
index 00000000..684db81e
--- /dev/null
+++ b/arm9/lib/src/FX_atan.c
@@ -0,0 +1,161 @@
+#include "global.h"
+#include "main.h"
+#include "fx.h"
+
+extern s16 FX_AtanTable_[];
+
+u16 FX_Atan(s32 x){
+    if (x >= 0)
+    {
+        if (x > 0x1000)
+        {
+            x = FX_Inv(x);
+            s16 y = FX_AtanTable_[x >> 5];
+            return 0x4000 - y;
+        }
+        else if (x < 0x1000)
+        {
+            return FX_AtanTable_[x >> 5];
+        }
+        else
+        {
+            return 0x2000;
+        }
+    }
+    else
+    {
+        if (x < -0x1000)
+        {
+            x = FX_Inv(-x);
+            s16 y = FX_AtanTable_[x >> 5];
+            return y - 0x4000;
+        }
+        else if (x > -0x1000)
+        {
+            return -FX_AtanTable_[-x >> 5];
+        }
+        else
+        {
+            return -0x2000;
+        }
+    }
+}
+
+u16 FX_Atan2(s32 x, s32 y){
+    s32 result;
+    u32 positive, bias, denominator, numerator;
+    if (x > 0)
+    {
+        if (y > 0)
+        {
+            if (y > x)
+            {
+                numerator = x;
+                denominator = y;
+                bias = 0;
+                positive = TRUE;
+            }
+            else if (y < x)
+            {
+                numerator = y;
+                denominator = x;
+                bias = 0x4000;
+                positive = FALSE;
+            }
+            else
+            {
+                return 0x2000;
+            }
+        }
+        else if (y < 0)
+        {
+            y = -y;
+            if (y < x)
+            {
+                numerator = y;
+                denominator = x;
+                bias = 0x4000;
+                positive = TRUE;
+            }
+            else if (y > x)
+            {
+                numerator = x;
+                denominator = y;
+                bias = 0x8000;
+                positive = FALSE;
+            }
+            else
+            {
+                return 0x6000;
+            }
+        }
+        else
+        {
+            return 0x4000;
+        }
+    }
+    else if (x < 0)
+    {
+        x = -x;
+        if (y < 0)
+        {
+            y = -y;
+            if (y > x)
+            {
+                numerator = x;
+                denominator = y;
+                bias = -0x8000;
+                positive = TRUE;
+            }
+            else if (y < x)
+            {
+                numerator = y;
+                denominator = x;
+                bias = -0x4000;
+                positive = FALSE;
+            }
+            else
+            {
+                return 0xA000;
+            }
+        }
+        else if (y > 0)
+        {
+            if (y < x)
+            {
+                numerator = y;
+                denominator = x;
+                bias = -0x4000;
+                positive = TRUE;
+            }
+            else if (y > x)
+            {
+                numerator = x;
+                denominator = y;
+                bias = 0x0;
+                positive = FALSE;
+            }
+            else
+            {
+                return 0xE000;
+            }
+        }
+        else
+        {
+            return 0xC000;
+        }
+    }
+    else
+    {
+        if (y >= 0)
+            return 0x0;
+        else
+            return 0x8000;
+    }
+    if (denominator == 0x0)
+        return 0x0;
+    if (positive)
+        return bias + FX_AtanTable_[FX_Div(numerator, denominator) >> 5];
+    else
+        return bias - FX_AtanTable_[FX_Div(numerator, denominator) >> 5];
+}
diff --git a/arm9/lib/src/FX_vec_c.c b/arm9/lib/src/FX_vec_c.c
new file mode 100644
index 00000000..a2e0ffe4
--- /dev/null
+++ b/arm9/lib/src/FX_vec_c.c
@@ -0,0 +1,34 @@
+#include "global.h"
+#include "main.h"
+#include "fx.h"
+
+void VEC_Add(struct Vecx32 *x, struct Vecx32 *y, struct Vecx32 *dst){
+    dst->x = x->x + y->x;
+    dst->y = x->y + y->y;
+    dst->z = x->z + y->z;
+}
+
+void VEC_Subtract(struct Vecx32 *x, struct Vecx32 *y, struct Vecx32 *dst){
+    dst->x = x->x - y->x;
+    dst->y = x->y - y->y;
+    dst->z = x->z - y->z;
+}
+
+void VEC_Fx16Add(struct Vecx16 *x, struct Vecx16 *y, struct Vecx16 *dst){
+    dst->x = x->x + y->x;
+    dst->y = x->y + y->y;
+    dst->z = x->z + y->z;
+}
+
+s32 VEC_DotProduct(struct Vecx32 *x, struct Vecx32 *y){
+    return ((s64)x->x * y->x + (s64)x->y * y->y + (s64)x->z * y->z + (1 << (FX64_INT_SHIFT - 1))) >> FX64_INT_SHIFT;
+}
+
+s32 VEC_Fx16DotProduct(struct Vecx16 *x, struct Vecx16 *y){
+    s32 temp1, temp2;
+    temp1 = (x->x * y->x) + (x->y * y->y);
+    temp2 = (x->z * y->z) + (1 << (FX64_INT_SHIFT - 1));
+    return (s32)(((s64)temp1 + temp2) >> FX64_INT_SHIFT);
+}
+
+void VEC_CrossProduct(struct Vecx32 *x, struct Vecx32 *y, struct Vecx32 *);
diff --git a/include/fx.h b/include/fx.h
new file mode 100644
index 00000000..82938219
--- /dev/null
+++ b/include/fx.h
@@ -0,0 +1,41 @@
+#ifndef GUARD_FX_H
+#define GUARD_FX_H
+
+#define FX32_INT_MASK     0xFFFFF000
+#define FX32_INT_ABS_MASK 0x7FFFF000
+#define FX32_FRAC_MASK    0x00000FFF
+
+#define FX32_INT_SHIFT    0xC
+#define FX64_INT_SHIFT    0xC
+
+#define FX32_INT(x)       (((x) & FX32_INT_MASK) >> FX32_INT_SHIFT)
+#define FX32_INT_ABS(x)   (((x) & FX32_INT_ABS_MASK) >> FX32_INT_SHIFT)
+#define FX32_FRAC(x)      ((x) & FX32_FRAC_MASK)
+
+struct Vecx32
+{
+    s32 x;
+    s32 y;
+    s32 z;
+};
+
+struct Vecx16
+{
+    s16 x;
+    s16 y;
+    s16 z;
+};
+
+void FX_Init();
+s32 FX_Modf(s32 x, s32 *iptr);
+s32 FX_Inv(s32 x);
+u16 FX_Atan(s32 x);
+u16 FX_Atan2(s32 x, s32 y);
+
+//Vec
+void VEC_Add(struct Vecx32 *x, struct Vecx32 *y, struct Vecx32 *dst);
+void VEC_Subtract(struct Vecx32 *x, struct Vecx32 *y, struct Vecx32 *dst);
+void VEC_Fx16Add(struct Vecx16 *x, struct Vecx16 *y, struct Vecx16 *dst);
+s32 VEC_DotProduct(struct Vecx32 *x, struct Vecx32 *y);
+
+#endif //GUARD_FX_H
-- 
cgit v1.2.3


From fefbe6b2dbc36e45fd521a75970fa58df5690615 Mon Sep 17 00:00:00 2001
From: Made <made111@gmx.de>
Date: Fri, 1 May 2020 19:06:20 +0200
Subject: better vector names

---
 arm9/lib/src/FX_vec_c.c | 36 ++++++++++++++++++------------------
 1 file changed, 18 insertions(+), 18 deletions(-)

diff --git a/arm9/lib/src/FX_vec_c.c b/arm9/lib/src/FX_vec_c.c
index a2e0ffe4..bebf7ca0 100644
--- a/arm9/lib/src/FX_vec_c.c
+++ b/arm9/lib/src/FX_vec_c.c
@@ -2,33 +2,33 @@
 #include "main.h"
 #include "fx.h"
 
-void VEC_Add(struct Vecx32 *x, struct Vecx32 *y, struct Vecx32 *dst){
-    dst->x = x->x + y->x;
-    dst->y = x->y + y->y;
-    dst->z = x->z + y->z;
+void VEC_Add(struct Vecx32 *a, struct Vecx32 *b, struct Vecx32 *dst){
+    dst->x = a->x + b->x;
+    dst->y = a->y + b->y;
+    dst->z = a->z + b->z;
 }
 
-void VEC_Subtract(struct Vecx32 *x, struct Vecx32 *y, struct Vecx32 *dst){
-    dst->x = x->x - y->x;
-    dst->y = x->y - y->y;
-    dst->z = x->z - y->z;
+void VEC_Subtract(struct Vecx32 *a, struct Vecx32 *b, struct Vecx32 *dst){
+    dst->x = a->x - b->x;
+    dst->y = a->y - b->y;
+    dst->z = a->z - b->z;
 }
 
-void VEC_Fx16Add(struct Vecx16 *x, struct Vecx16 *y, struct Vecx16 *dst){
-    dst->x = x->x + y->x;
-    dst->y = x->y + y->y;
-    dst->z = x->z + y->z;
+void VEC_Fx16Add(struct Vecx16 *a, struct Vecx16 *b, struct Vecx16 *dst){
+    dst->x = a->x + b->x;
+    dst->y = a->y + b->y;
+    dst->z = a->z + b->z;
 }
 
-s32 VEC_DotProduct(struct Vecx32 *x, struct Vecx32 *y){
-    return ((s64)x->x * y->x + (s64)x->y * y->y + (s64)x->z * y->z + (1 << (FX64_INT_SHIFT - 1))) >> FX64_INT_SHIFT;
+s32 VEC_DotProduct(struct Vecx32 *a, struct Vecx32 *b){
+    return ((s64)a->x * b->x + (s64)a->y * b->y + (s64)a->z * b->z + (1 << (FX64_INT_SHIFT - 1))) >> FX64_INT_SHIFT;
 }
 
-s32 VEC_Fx16DotProduct(struct Vecx16 *x, struct Vecx16 *y){
+s32 VEC_Fx16DotProduct(struct Vecx16 *a, struct Vecx16 *b){
     s32 temp1, temp2;
-    temp1 = (x->x * y->x) + (x->y * y->y);
-    temp2 = (x->z * y->z) + (1 << (FX64_INT_SHIFT - 1));
+    temp1 = (a->x * b->x) + (a->y * b->y);
+    temp2 = (a->z * b->z) + (1 << (FX64_INT_SHIFT - 1));
     return (s32)(((s64)temp1 + temp2) >> FX64_INT_SHIFT);
 }
 
-void VEC_CrossProduct(struct Vecx32 *x, struct Vecx32 *y, struct Vecx32 *);
+void VEC_CrossProduct(struct Vecx32 *a, struct Vecx32 *b, struct Vecx32 *dst);
-- 
cgit v1.2.3


From 3a51b862659071a4dab8296f1c9cfbe38d287887 Mon Sep 17 00:00:00 2001
From: Made <made111@gmx.de>
Date: Fri, 1 May 2020 21:57:41 +0200
Subject: finish decompiling FX_vec.s

---
 arm9/arm9.lcf           |   1 -
 arm9/asm/FX_vec.s       | 403 ------------------------------------------------
 arm9/lib/src/FX_vec.c   | 125 +++++++++++++++
 arm9/lib/src/FX_vec_c.c |  34 ----
 4 files changed, 125 insertions(+), 438 deletions(-)
 delete mode 100644 arm9/asm/FX_vec.s
 create mode 100644 arm9/lib/src/FX_vec.c
 delete mode 100644 arm9/lib/src/FX_vec_c.c

diff --git a/arm9/arm9.lcf b/arm9/arm9.lcf
index 046ce41d..42b0a8ee 100644
--- a/arm9/arm9.lcf
+++ b/arm9/arm9.lcf
@@ -127,7 +127,6 @@ SECTIONS {
 		FX_mtx44.o (.text)
 		FX_cp.o (.text)
 		FX_vec.o (.text)
-        FX_vec_c.o (.text)
 		FX_atan.o (.text)
 		FX.o (.text)
 		GX.o (.text)
diff --git a/arm9/asm/FX_vec.s b/arm9/asm/FX_vec.s
deleted file mode 100644
index 08068eaf..00000000
--- a/arm9/asm/FX_vec.s
+++ /dev/null
@@ -1,403 +0,0 @@
-	.include "asm/macros.inc"
-	.include "global.inc"
-
-    .text
-
-	arm_func_start VEC_MultAdd
-VEC_MultAdd: ; 0x020C50D4
-	stmdb sp!, {r4,lr}
-	ldr r4, [r1, #0x0]
-	ldr lr, [r2, #0x0]
-	smull r12, r4, r0, r4
-	mov r12, r12, lsr #0xc
-	orr r12, r12, r4, lsl #0x14
-	add r4, lr, r12
-	str r4, [r3, #0x0]
-	ldr r12, [r1, #0x4]
-	ldr r4, [r2, #0x4]
-	smull lr, r12, r0, r12
-	mov lr, lr, lsr #0xc
-	orr lr, lr, r12, lsl #0x14
-	add r4, r4, lr
-	str r4, [r3, #0x4]
-	ldr r1, [r1, #0x8]
-	ldr r12, [r2, #0x8]
-	smull r2, r1, r0, r1
-	mov r0, r2, lsr #0xc
-	orr r0, r0, r1, lsl #0x14
-	add r0, r12, r0
-	str r0, [r3, #0x8]
-	ldmia sp!, {r4,lr}
-	bx lr
-
-	arm_func_start VEC_Fx16Normalize
-VEC_Fx16Normalize: ; 0x020C5134
-	stmdb sp!, {r4-r8,lr}
-	ldrsh r4, [r0, #0x0]
-	ldrsh r3, [r0, #0x2]
-	ldrsh r2, [r0, #0x4]
-	mul r8, r4, r4
-	mul r7, r3, r3
-	mul r6, r2, r2
-	ldr r2, _020C526C ; =0x04000280
-	mov r3, #0x2
-	ldr r4, _020C5270 ; =0x04000290
-	strh r3, [r2, #0x0]
-	mov r2, #0x0
-	str r2, [r4, #0x0]
-	mov r5, #0x1000000
-	mov r2, r8, asr #0x1f
-	adds r3, r8, r7
-	adc r2, r2, r7, asr #0x1f
-	adds r7, r3, r6
-	adc r6, r2, r6, asr #0x1f
-	mov r3, r6, lsl #0x2
-	ldr r2, _020C5274 ; =0x04000298
-	str r5, [r4, #0x4]
-	str r7, [r2, #0x0]
-	ldr r5, _020C5278 ; =0x040002B0
-	str r6, [r2, #0x4]
-	mov r4, #0x1
-	ldr r2, _020C527C ; =0x040002B8
-	strh r4, [r5, #0x0]
-	mov r4, r7, lsl #0x2
-	str r4, [r2, #0x0]
-	orr r3, r3, r7, lsr #0x1e
-	str r3, [r2, #0x4]
-_020C51B4:
-	ldrh r2, [r5, #0x0]
-	ands r2, r2, #0x8000
-	bne _020C51B4
-	ldr r2, _020C5280 ; =0x040002B4
-	ldr r12, [r2, #0x0]
-	ldr r3, _020C526C ; =0x04000280
-_020C51CC:
-	ldrh r2, [r3, #0x0]
-	ands r2, r2, #0x8000
-	bne _020C51CC
-	ldr r7, _020C5284 ; =0x040002A0
-	ldrsh r5, [r0, #0x0]
-	ldr r6, [r7, #0x0]
-	mov r4, r12, asr #0x1f
-	umull r3, r2, r6, r12
-	mla r2, r6, r4, r2
-	ldr r8, [r7, #0x4]
-	mov r4, r5, asr #0x1f
-	umull r7, r6, r3, r5
-	mov lr, #0x0
-	mla r2, r8, r12, r2
-	mla r6, r3, r4, r6
-	mla r6, r2, r5, r6
-	adds r4, r7, lr
-	adc r4, r6, #0x1000
-	mov r4, r4, asr #0xd
-	strh r4, [r1, #0x0]
-	ldrsh r5, [r0, #0x2]
-	umull r6, r12, r3, r5
-	mov r4, r5, asr #0x1f
-	mla r12, r3, r4, r12
-	mla r12, r2, r5, r12
-	adds r4, r6, lr
-	adc r4, r12, #0x1000
-	mov r4, r4, asr #0xd
-	strh r4, [r1, #0x2]
-	ldrsh r12, [r0, #0x4]
-	umull r5, r4, r3, r12
-	mov r0, r12, asr #0x1f
-	adds r5, r5, lr
-	mla r4, r3, r0, r4
-	mla r4, r2, r12, r4
-	adc r0, r4, #0x1000
-	mov r0, r0, asr #0xd
-	strh r0, [r1, #0x4]
-	ldmia sp!, {r4-r8,lr}
-	bx lr
-	.balign 4
-_020C526C: .word 0x04000280
-_020C5270: .word 0x04000290
-_020C5274: .word 0x04000298
-_020C5278: .word 0x040002B0
-_020C527C: .word 0x040002B8
-_020C5280: .word 0x040002B4
-_020C5284: .word 0x040002A0
-
-	arm_func_start VEC_Normalize
-VEC_Normalize:
-	stmdb sp!, {r4-r8,lr}
-	ldr r3, [r0, #0x4]
-	ldr r4, [r0, #0x0]
-	smull r2, r7, r3, r3
-	smlal r2, r7, r4, r4
-	ldr r4, [r0, #0x8]
-	ldr r3, _020C53AC ; =0x04000280
-	smlal r2, r7, r4, r4
-	mov r6, #0x2
-	mov r4, r7, lsl #0x2
-	ldr r5, _020C53B0 ; =0x04000290
-	strh r6, [r3, #0x0]
-	mov r3, #0x0
-	str r3, [r5, #0x0]
-	mov r6, #0x1000000
-	ldr r3, _020C53B4 ; =0x04000298
-	str r6, [r5, #0x4]
-	str r2, [r3, #0x0]
-	ldr r6, _020C53B8 ; =0x040002B0
-	str r7, [r3, #0x4]
-	mov r5, #0x1
-	ldr r3, _020C53BC ; =0x040002B8
-	strh r5, [r6, #0x0]
-	mov r5, r2, lsl #0x2
-	str r5, [r3, #0x0]
-	orr r4, r4, r2, lsr #0x1e
-	str r4, [r3, #0x4]
-_020C52F4:
-	ldrh r2, [r6, #0x0]
-	ands r2, r2, #0x8000
-	bne _020C52F4
-	ldr r2, _020C53C0 ; =0x040002B4
-	ldr r12, [r2, #0x0]
-	ldr r3, _020C53AC ; =0x04000280
-_020C530C:
-	ldrh r2, [r3, #0x0]
-	ands r2, r2, #0x8000
-	bne _020C530C
-	ldr r7, _020C53C4 ; =0x040002A0
-	ldr r5, [r0, #0x0]
-	ldr r6, [r7, #0x0]
-	mov r4, r12, asr #0x1f
-	umull r3, r2, r6, r12
-	mla r2, r6, r4, r2
-	ldr r8, [r7, #0x4]
-	mov r4, r5, asr #0x1f
-	umull r7, r6, r3, r5
-	mov lr, #0x0
-	mla r2, r8, r12, r2
-	mla r6, r3, r4, r6
-	mla r6, r2, r5, r6
-	adds r4, r7, lr
-	adc r4, r6, #0x1000
-	mov r4, r4, asr #0xd
-	str r4, [r1, #0x0]
-	ldr r5, [r0, #0x4]
-	umull r6, r12, r3, r5
-	mov r4, r5, asr #0x1f
-	mla r12, r3, r4, r12
-	mla r12, r2, r5, r12
-	adds r4, r6, lr
-	adc r4, r12, #0x1000
-	mov r4, r4, asr #0xd
-	str r4, [r1, #0x4]
-	ldr r12, [r0, #0x8]
-	umull r5, r4, r3, r12
-	mov r0, r12, asr #0x1f
-	adds r5, r5, lr
-	mla r4, r3, r0, r4
-	mla r4, r2, r12, r4
-	adc r0, r4, #0x1000
-	mov r0, r0, asr #0xd
-	str r0, [r1, #0x8]
-	ldmia sp!, {r4-r8,lr}
-	bx lr
-	.balign 4
-_020C53AC: .word 0x04000280
-_020C53B0: .word 0x04000290
-_020C53B4: .word 0x04000298
-_020C53B8: .word 0x040002B0
-_020C53BC: .word 0x040002B8
-_020C53C0: .word 0x040002B4
-_020C53C4: .word 0x040002A0
-
-	arm_func_start VEC_Mag
-VEC_Mag:
-	ldr r2, [r0, #0x4]
-	ldr r3, [r0, #0x0]
-	smull r12, r1, r2, r2
-	smlal r12, r1, r3, r3
-	ldr r0, [r0, #0x8]
-	ldr r3, _020C5424 ; =0x040002B0
-	smlal r12, r1, r0, r0
-	mov r2, #0x1
-	mov r1, r1, lsl #0x2
-	ldr r0, _020C5428 ; =0x040002B8
-	strh r2, [r3, #0x0]
-	mov r2, r12, lsl #0x2
-	str r2, [r0, #0x0]
-	orr r1, r1, r12, lsr #0x1e
-	str r1, [r0, #0x4]
-_020C5404:
-	ldrh r0, [r3, #0x0]
-	ands r0, r0, #0x8000
-	bne _020C5404
-	ldr r0, _020C542C ; =0x040002B4
-	ldr r0, [r0, #0x0]
-	add r0, r0, #0x1
-	mov r0, r0, asr #0x1
-	bx lr
-	.balign 4
-_020C5424: .word 0x040002B0
-_020C5428: .word 0x040002B8
-_020C542C: .word 0x040002B4
-
-	arm_func_start VEC_Fx16CrossProduct
-VEC_Fx16CrossProduct: ; 0x020C5430
-	stmdb sp!, {r4-r6,lr}
-	ldrsh r4, [r1, #0x4]
-	ldrsh r12, [r0, #0x2]
-	ldrsh lr, [r0, #0x0]
-	ldrsh r3, [r1, #0x2]
-	ldrsh r6, [r0, #0x4]
-	ldrsh r1, [r1, #0x0]
-	mul r5, r12, r4
-	mul r0, r6, r3
-	sub r0, r5, r0
-	add r0, r0, #0x800
-	mov r0, r0, asr #0xc
-	mul r5, r6, r1
-	mul r4, lr, r4
-	sub r4, r5, r4
-	add r4, r4, #0x800
-	mul r3, lr, r3
-	mul r1, r12, r1
-	sub r1, r3, r1
-	add r1, r1, #0x800
-	strh r0, [r2, #0x0]
-	mov r0, r4, asr #0xc
-	strh r0, [r2, #0x2]
-	mov r0, r1, asr #0xc
-	strh r0, [r2, #0x4]
-	ldmia sp!, {r4-r6,lr}
-	bx lr
-
-	arm_func_start VEC_CrossProduct
-VEC_CrossProduct:
-	stmdb sp!, {r4-r11,lr}
-	sub sp, sp, #0x4
-	ldr r12, [r1, #0x8]
-	ldr r9, [r0, #0x4]
-	ldr r3, [r0, #0x0]
-	ldr r11, [r1, #0x4]
-	ldr r0, [r0, #0x8]
-	ldr r10, [r1, #0x0]
-	smull r4, r7, r9, r12
-	smull r1, r6, r0, r11
-	subs r8, r4, r1
-	smull r5, r4, r0, r10
-	smull lr, r12, r3, r12
-	smull r1, r0, r3, r11
-	smull r10, r3, r9, r10
-	sbc r6, r7, r6
-	mov r9, #0x800
-	adds r7, r8, r9
-	adc r6, r6, #0x0
-	subs r8, r5, lr
-	sbc r5, r4, r12
-	adds r8, r8, r9
-	mov r4, r7, lsr #0xc
-	adc r5, r5, #0x0
-	subs r1, r1, r10
-	orr r4, r4, r6, lsl #0x14
-	mov r6, r8, lsr #0xc
-	sbc r0, r0, r3
-	adds r1, r1, r9
-	str r4, [r2, #0x0]
-	orr r6, r6, r5, lsl #0x14
-	adc r0, r0, #0x0
-	mov r1, r1, lsr #0xc
-	str r6, [r2, #0x4]
-	orr r1, r1, r0, lsl #0x14
-	str r1, [r2, #0x8]
-	add sp, sp, #0x4
-	ldmia sp!, {r4-r11,lr}
-	bx lr
-
-;	arm_func_start VEC_Fx16DotProduct
-;VEC_Fx16DotProduct: ; 0x020C5538
-;	stmdb sp!, {r4-r5,lr}
-;	sub sp, sp, #0x4
-;	ldrsh lr, [r0, #0x2]
-;	ldrsh r12, [r1, #0x2]
-;	ldrsh r3, [r0, #0x4]
-;	ldrsh r2, [r1, #0x4]
-;	ldrsh r5, [r0, #0x0]
-;	ldrsh r4, [r1, #0x0]
-;	smulbb r1, lr, r12
-;	smulbb r0, r3, r2
-;	smlabb r1, r5, r4, r1
-;	add r0, r0, #0x800
-;	adds r2, r1, r0
-;	mov r1, r1, asr #0x1f
-;	adc r1, r1, r0, asr #0x1f
-;	mov r0, r2, lsr #0xc
-;	orr r0, r0, r1, lsl #0x14
-;	add sp, sp, #0x4
-;	ldmia sp!, {r4-r5,lr}
-;	bx lr
-;
-;	arm_func_start VEC_DotProduct
-;VEC_DotProduct:
-;	stmdb sp!, {r4,lr}
-;	ldr r3, [r0, #0x4]
-;	ldr r2, [r1, #0x4]
-;	ldr r12, [r0, #0x0]
-;	smull r4, lr, r3, r2
-;	ldr r2, [r1, #0x0]
-;	ldr r3, [r0, #0x8]
-;	smlal r4, lr, r12, r2
-;	ldr r1, [r1, #0x8]
-;	mov r0, #0x800
-;	smlal r4, lr, r3, r1
-;	adds r0, r4, r0
-;	adc r1, lr, #0x0
-;	mov r0, r0, lsr #0xc
-;	orr r0, r0, r1, lsl #0x14
-;	ldmia sp!, {r4,lr}
-;	bx lr
-;
-;	arm_func_start VEC_Fx16Add
-;VEC_Fx16Add: ; 0x020C55CC
-;	ldrsh r12, [r0, #0x0]
-;	ldrsh r3, [r1, #0x0]
-;	add r3, r12, r3
-;	strh r3, [r2, #0x0]
-;	ldrsh r12, [r0, #0x2]
-;	ldrsh r3, [r1, #0x2]
-;	add r3, r12, r3
-;	strh r3, [r2, #0x2]
-;	ldrsh r3, [r0, #0x4]
-;	ldrsh r0, [r1, #0x4]
-;	add r0, r3, r0
-;	strh r0, [r2, #0x4]
-;	bx lr
-;
-;	arm_func_start VEC_Subtract
-;VEC_Subtract: ; 0x020C5600
-;	ldr r12, [r0, #0x0]
-;	ldr r3, [r1, #0x0]
-;	sub r3, r12, r3
-;	str r3, [r2, #0x0]
-;	ldr r12, [r0, #0x4]
-;	ldr r3, [r1, #0x4]
-;	sub r3, r12, r3
-;	str r3, [r2, #0x4]
-;	ldr r3, [r0, #0x8]
-;	ldr r0, [r1, #0x8]
-;	sub r0, r3, r0
-;	str r0, [r2, #0x8]
-;	bx lr
-;
-;	arm_func_start VEC_Add
-;VEC_Add: ; 0x020C5634
-;	ldr r12, [r0, #0x0]
-;	ldr r3, [r1, #0x0]
-;	add r3, r12, r3
-;	str r3, [r2, #0x0]
-;	ldr r12, [r0, #0x4]
-;	ldr r3, [r1, #0x4]
-;	add r3, r12, r3
-;	str r3, [r2, #0x4]
-;	ldr r3, [r0, #0x8]
-;	ldr r0, [r1, #0x8]
-;	add r0, r3, r0
-;	str r0, [r2, #0x8]
-;	bx lr
-;
diff --git a/arm9/lib/src/FX_vec.c b/arm9/lib/src/FX_vec.c
new file mode 100644
index 00000000..dc28bedf
--- /dev/null
+++ b/arm9/lib/src/FX_vec.c
@@ -0,0 +1,125 @@
+#include "global.h"
+#include "main.h"
+#include "fx.h"
+
+void VEC_Add(struct Vecx32 *a, struct Vecx32 *b, struct Vecx32 *dst){
+    dst->x = a->x + b->x;
+    dst->y = a->y + b->y;
+    dst->z = a->z + b->z;
+}
+
+void VEC_Subtract(struct Vecx32 *a, struct Vecx32 *b, struct Vecx32 *dst){
+    dst->x = a->x - b->x;
+    dst->y = a->y - b->y;
+    dst->z = a->z - b->z;
+}
+
+void VEC_Fx16Add(struct Vecx16 *a, struct Vecx16 *b, struct Vecx16 *dst){
+    dst->x = a->x + b->x;
+    dst->y = a->y + b->y;
+    dst->z = a->z + b->z;
+}
+
+s32 VEC_DotProduct(struct Vecx32 *a, struct Vecx32 *b){
+    return ((s64)a->x * b->x + (s64)a->y * b->y + (s64)a->z * b->z + (1 << (FX64_INT_SHIFT - 1))) >> FX64_INT_SHIFT;
+}
+
+s32 VEC_Fx16DotProduct(struct Vecx16 *a, struct Vecx16 *b){
+    s32 temp1, temp2;
+    temp1 = (a->x * b->x) + (a->y * b->y);
+    temp2 = (a->z * b->z) + (1 << (FX64_INT_SHIFT - 1));
+    return (s32)(((s64)temp1 + temp2) >> FX64_INT_SHIFT);
+}
+
+void VEC_CrossProduct(struct Vecx32 *a, struct Vecx32 *b, struct Vecx32 *dst){
+    s32 x, y, z;
+    x = (s32)(((s64)a->y * b->z - (s64)a->z * b->y + (1 << (FX64_INT_SHIFT - 1))) >> FX64_INT_SHIFT);
+    y = (s32)(((s64)a->z * b->x - (s64)a->x * b->z + (1 << (FX64_INT_SHIFT - 1))) >> FX64_INT_SHIFT);
+    z = (s32)(((s64)a->x * b->y - (s64)a->y * b->x + (1 << (FX64_INT_SHIFT - 1))) >> FX64_INT_SHIFT);
+    dst->x = x;
+    dst->y = y;
+    dst->z = z;
+}
+
+void VEC_Fx16CrossProduct(struct Vecx16 *a, struct Vecx16 *b, struct Vecx16 *dst){
+    s32 x, y, z;
+    x = ((a->y * b->z - a->z * b->y + (1 << (FX64_INT_SHIFT - 1))) >> FX64_INT_SHIFT);
+    y = ((a->z * b->x - a->x * b->z + (1 << (FX64_INT_SHIFT - 1))) >> FX64_INT_SHIFT);
+    z = ((a->x * b->y - a->y * b->x + (1 << (FX64_INT_SHIFT - 1))) >> FX64_INT_SHIFT);
+    dst->x = x;
+    dst->y = y;
+    dst->z = z;
+}
+
+#define HW_REG_DIVCNT              0x04000280
+#define HW_REG_DIV_NUMER           0x04000290
+#define HW_REG_DIV_DENOM           0x04000298
+#define HW_REG_DIV_RESULT          0x040002A0
+#define HW_REG_DIVREM_RESULT       0x040002A8
+
+#define HW_REG_SQRTCNT             0x040002B0
+#define HW_REG_SQRT_RESULT         0x040002B4
+#define HW_REG_SQRT_PARAM          0x040002B8
+
+#define SETREG16(x, y)             ((*(vu16 *)x) = y)
+#define SETREG32(x, y)             ((*(vu32 *)x) = y)
+#define SETREG64(x, y)             ((*(vu64 *)x) = y)
+#define READREG16(x)               (*(vu16 *)x)
+#define READREG32(x)               (*(vu32 *)x)
+#define READREG64(x)               (*(vu64 *)x)
+
+s32 VEC_Mag(struct Vecx32 *a){
+    s64 l2 = (s64)a->x * a->x;
+    l2 += (s64)a->y * a->y;
+    l2 += (s64)a->z * a->z;
+    SETREG16(HW_REG_SQRTCNT, 0x1);
+    SETREG64(HW_REG_SQRT_PARAM, l2 * 4);
+    while (READREG16(HW_REG_SQRTCNT) & 0x8000); //wait for coprocessor to finish
+    return ((s32)READREG32(HW_REG_SQRT_RESULT) + 1) >> 1;
+}
+
+void VEC_Normalize(struct Vecx32 *a, struct Vecx32 *dst){
+    s64 l2 = (s64)a->x * a->x;
+    l2 += (s64)a->y * a->y;
+    l2 += (s64)a->z * a->z;
+    //1/sqrt(l) is computed by calculating sqrt(l)*(1/l)
+    SETREG16(HW_REG_DIVCNT, 0x2);
+    SETREG64(HW_REG_DIV_NUMER, 0x0100000000000000);
+    SETREG64(HW_REG_DIV_DENOM, l2);
+    SETREG16(HW_REG_SQRTCNT, 0x1);
+    SETREG64(HW_REG_SQRT_PARAM, l2 * 4);
+    while (READREG16(HW_REG_SQRTCNT) & 0x8000); //wait for sqrt to finish
+    s32 sqrtresult = READREG32(HW_REG_SQRT_RESULT);
+    while (READREG16(HW_REG_DIVCNT) & 0x8000); //wait for division to finish
+    l2 = READREG64(HW_REG_DIV_RESULT);
+    l2 = sqrtresult * l2;
+    dst->x = (l2 * a->x + (1LL << (0x2D - 1))) >> 0x2D;
+    dst->y = (l2 * a->y + (1LL << (0x2D - 1))) >> 0x2D;
+    dst->z = (l2 * a->z + (1LL << (0x2D - 1))) >> 0x2D;
+}
+
+void VEC_Fx16Normalize(struct Vecx16 *a, struct Vecx16 *dst){
+    s64 l2 = a->x * a->x;
+    l2 += a->y * a->y;
+    l2 += a->z * a->z;
+    //1/sqrt(l) is computed by calculating sqrt(l)*(1/l)
+    SETREG16(HW_REG_DIVCNT, 0x2);
+    SETREG64(HW_REG_DIV_NUMER, 0x0100000000000000);
+    SETREG64(HW_REG_DIV_DENOM, l2);
+    SETREG16(HW_REG_SQRTCNT, 0x1);
+    SETREG64(HW_REG_SQRT_PARAM, l2 * 4);
+    while (READREG16(HW_REG_SQRTCNT) & 0x8000); //wait for sqrt to finish
+    s32 sqrtresult = READREG32(HW_REG_SQRT_RESULT);
+    while (READREG16(HW_REG_DIVCNT) & 0x8000); //wait for division to finish
+    l2 = READREG64(HW_REG_DIV_RESULT);
+    l2 = sqrtresult * l2;
+    dst->x = (l2 * a->x + (1LL << (0x2D - 1))) >> 0x2D;
+    dst->y = (l2 * a->y + (1LL << (0x2D - 1))) >> 0x2D;
+    dst->z = (l2 * a->z + (1LL << (0x2D - 1))) >> 0x2D;
+}
+
+void VEC_MultAdd(s32 factor, struct Vecx32  *a, struct Vecx32 *b, struct Vecx32 *dst){
+    dst->x = (s32)(((s64)factor * a->x) >> FX32_INT_SHIFT) + b->x;
+    dst->y = (s32)(((s64)factor * a->y) >> FX32_INT_SHIFT) + b->y;
+    dst->z = (s32)(((s64)factor * a->z) >> FX32_INT_SHIFT) + b->z;
+}
diff --git a/arm9/lib/src/FX_vec_c.c b/arm9/lib/src/FX_vec_c.c
deleted file mode 100644
index bebf7ca0..00000000
--- a/arm9/lib/src/FX_vec_c.c
+++ /dev/null
@@ -1,34 +0,0 @@
-#include "global.h"
-#include "main.h"
-#include "fx.h"
-
-void VEC_Add(struct Vecx32 *a, struct Vecx32 *b, struct Vecx32 *dst){
-    dst->x = a->x + b->x;
-    dst->y = a->y + b->y;
-    dst->z = a->z + b->z;
-}
-
-void VEC_Subtract(struct Vecx32 *a, struct Vecx32 *b, struct Vecx32 *dst){
-    dst->x = a->x - b->x;
-    dst->y = a->y - b->y;
-    dst->z = a->z - b->z;
-}
-
-void VEC_Fx16Add(struct Vecx16 *a, struct Vecx16 *b, struct Vecx16 *dst){
-    dst->x = a->x + b->x;
-    dst->y = a->y + b->y;
-    dst->z = a->z + b->z;
-}
-
-s32 VEC_DotProduct(struct Vecx32 *a, struct Vecx32 *b){
-    return ((s64)a->x * b->x + (s64)a->y * b->y + (s64)a->z * b->z + (1 << (FX64_INT_SHIFT - 1))) >> FX64_INT_SHIFT;
-}
-
-s32 VEC_Fx16DotProduct(struct Vecx16 *a, struct Vecx16 *b){
-    s32 temp1, temp2;
-    temp1 = (a->x * b->x) + (a->y * b->y);
-    temp2 = (a->z * b->z) + (1 << (FX64_INT_SHIFT - 1));
-    return (s32)(((s64)temp1 + temp2) >> FX64_INT_SHIFT);
-}
-
-void VEC_CrossProduct(struct Vecx32 *a, struct Vecx32 *b, struct Vecx32 *dst);
-- 
cgit v1.2.3


From dba3fd7f6eb8d0895bec720db75a6b37fc7259a0 Mon Sep 17 00:00:00 2001
From: Made <made111@gmx.de>
Date: Sat, 2 May 2020 17:39:44 +0200
Subject: Decompile FX_cp.s

---
 arm9/asm/FX_cp.s      | 190 --------------------------------------------------
 arm9/lib/src/FX_cp.c  |  74 ++++++++++++++++++++
 arm9/lib/src/FX_vec.c |  17 -----
 include/fx.h          |  29 +++++++-
 4 files changed, 102 insertions(+), 208 deletions(-)
 delete mode 100644 arm9/asm/FX_cp.s
 create mode 100644 arm9/lib/src/FX_cp.c

diff --git a/arm9/asm/FX_cp.s b/arm9/asm/FX_cp.s
deleted file mode 100644
index 1c5f0350..00000000
--- a/arm9/asm/FX_cp.s
+++ /dev/null
@@ -1,190 +0,0 @@
-	.include "asm/macros.inc"
-	.include "global.inc"
-
-	.text
-
-	arm_func_start FX_ModS32
-FX_ModS32: ; 0x020C4E98
-	stmdb sp!, {lr}
-	sub sp, sp, #0x4
-	ldr ip, _020C4EE0 ; =0x04000280
-	mov lr, #0x0
-	ldr r3, _020C4EE4 ; =0x04000290
-	strh lr, [r12, #0x0]
-	ldr r2, _020C4EE8 ; =0x04000298
-	str r0, [r3, #0x0]
-	str r1, [r2, #0x0]
-	str lr, [r2, #0x4]
-_020C4EC0:
-	ldrh r0, [r12, #0x0]
-	ands r0, r0, #0x8000
-	bne _020C4EC0
-	ldr r0, _020C4EEC ; =0x040002A8
-	ldr r0, [r0, #0x0]
-	add sp, sp, #0x4
-	ldmia sp!, {lr}
-	bx lr
-	.balign 4
-_020C4EE0: .word 0x04000280
-_020C4EE4: .word 0x04000290
-_020C4EE8: .word 0x04000298
-_020C4EEC: .word 0x040002A8
-
-	arm_func_start FX_DivS32
-FX_DivS32: ; 0x020C4EF0
-	stmdb sp!, {lr}
-	sub sp, sp, #0x4
-	ldr ip, _020C4F38 ; =0x04000280
-	mov lr, #0x0
-	ldr r3, _020C4F3C ; =0x04000290
-	strh lr, [r12, #0x0]
-	ldr r2, _020C4F40 ; =0x04000298
-	str r0, [r3, #0x0]
-	str r1, [r2, #0x0]
-	str lr, [r2, #0x4]
-_020C4F18:
-	ldrh r0, [r12, #0x0]
-	ands r0, r0, #0x8000
-	bne _020C4F18
-	ldr r0, _020C4F44 ; =0x040002A0
-	ldr r0, [r0, #0x0]
-	add sp, sp, #0x4
-	ldmia sp!, {lr}
-	bx lr
-	.balign 4
-_020C4F38: .word 0x04000280
-_020C4F3C: .word 0x04000290
-_020C4F40: .word 0x04000298
-_020C4F44: .word 0x040002A0
-
-	arm_func_start FX_DivAsync
-FX_DivAsync: ; 0x020C4F48
-	ldr r2, _020C4F74 ; =0x04000280
-	mov r12, #0x1
-	ldr r3, _020C4F78 ; =0x04000290
-	strh r12, [r2, #0x0]
-	mov r12, #0x0
-	str r12, [r3, #0x0]
-	ldr r2, _020C4F7C ; =0x04000298
-	str r0, [r3, #0x4]
-	str r1, [r2, #0x0]
-	str r12, [r2, #0x4]
-	bx lr
-	.balign 4
-_020C4F74: .word 0x04000280
-_020C4F78: .word 0x04000290
-_020C4F7C: .word 0x04000298
-
-	arm_func_start FX_GetSqrtResult
-FX_GetSqrtResult: ; 0x020C4F80
-	ldr r1, _020C4FA4 ; =0x040002B0
-_020C4F84:
-	ldrh r0, [r1, #0x0]
-	ands r0, r0, #0x8000
-	bne _020C4F84
-	ldr r0, _020C4FA8 ; =0x040002B4
-	ldr r0, [r0, #0x0]
-	add r0, r0, #0x200
-	mov r0, r0, lsr #0xa
-	bx lr
-	.balign 4
-_020C4FA4: .word 0x040002B0
-_020C4FA8: .word 0x040002B4
-
-	arm_func_start FX_InvAsync
-FX_InvAsync:
-	ldr r1, _020C4FDC ; =0x04000280
-	mov r3, #0x1
-	ldr r2, _020C4FE0 ; =0x04000290
-	strh r3, [r1, #0x0]
-	mov r12, #0x0
-	str r12, [r2, #0x0]
-	mov r3, #0x1000
-	ldr r1, _020C4FE4 ; =0x04000298
-	str r3, [r2, #0x4]
-	str r0, [r1, #0x0]
-	str r12, [r1, #0x4]
-	bx lr
-	.balign 4
-_020C4FDC: .word 0x04000280
-_020C4FE0: .word 0x04000290
-_020C4FE4: .word 0x04000298
-
-	arm_func_start FX_GetDivResult
-FX_GetDivResult: ; 0x020C4FE8
-	ldr r1, _020C501C ; =0x04000280
-_020C4FEC:
-	ldrh r0, [r1, #0x0]
-	ands r0, r0, #0x8000
-	bne _020C4FEC
-	ldr r1, _020C5020 ; =0x040002A0
-	mov r0, #0x80000
-	ldr r2, [r1, #0x0]
-	ldr r1, [r1, #0x4]
-	adds r0, r2, r0
-	adc r1, r1, #0x0
-	mov r0, r0, lsr #0x14
-	orr r0, r0, r1, lsl #0xc
-	bx lr
-	.balign 4
-_020C501C: .word 0x04000280
-_020C5020: .word 0x040002A0
-
-	arm_func_start FX_GetDivResultFx64c
-FX_GetDivResultFx64c: ; 0x020C5024
-	ldr r1, _020C5044 ; =0x04000280
-_020C5028:
-	ldrh r0, [r1, #0x0]
-	ands r0, r0, #0x8000
-	bne _020C5028
-	ldr r1, _020C5048 ; =0x040002A0
-	ldr r0, [r1, #0x0]
-	ldr r1, [r1, #0x4]
-	bx lr
-	.balign 4
-_020C5044: .word 0x04000280
-_020C5048: .word 0x040002A0
-
-	arm_func_start FX_Sqrt
-FX_Sqrt: ; 0x020C504C
-	stmdb sp!, {lr}
-	sub sp, sp, #0x4
-	cmp r0, #0x0
-	addle sp, sp, #0x4
-	movle r0, #0x0
-	ldmleia sp!, {lr}
-	bxle lr
-	ldr r2, _020C5094 ; =0x040002B0
-	mov r3, #0x1
-	strh r3, [r2, #0x0]
-	ldr r1, _020C5098 ; =0x040002B8
-	mov r2, #0x0
-	str r2, [r1, #0x0]
-	str r0, [r1, #0x4]
-	bl FX_GetSqrtResult
-	add sp, sp, #0x4
-	ldmia sp!, {lr}
-	bx lr
-	.balign 4
-_020C5094: .word 0x040002B0
-_020C5098: .word 0x040002B8
-
-	arm_func_start FX_Inv
-FX_Inv: ; 0x020C509C
-	stmdb sp!, {lr}
-	sub sp, sp, #0x4
-	bl FX_InvAsync
-	bl FX_GetDivResult
-	add sp, sp, #0x4
-	ldmia sp!, {lr}
-	bx lr
-
-	arm_func_start FX_Div
-FX_Div: ; 0x020C50B8
-	stmdb sp!, {lr}
-	sub sp, sp, #0x4
-	bl FX_DivAsync
-	bl FX_GetDivResult
-	add sp, sp, #0x4
-	ldmia sp!, {lr}
-	bx lr
diff --git a/arm9/lib/src/FX_cp.c b/arm9/lib/src/FX_cp.c
new file mode 100644
index 00000000..a8da0a8f
--- /dev/null
+++ b/arm9/lib/src/FX_cp.c
@@ -0,0 +1,74 @@
+#include "global.h"
+#include "main.h"
+#include "fx.h"
+
+s32 FX_GetDivResult();
+s32 FX_GetSqrtResult();
+void FX_DivAsync(s32 numerator, s32 denominator);
+void FX_InvAsync(s32 x);
+
+s32 FX_Div(s32 numerator, s32 denominator){
+    FX_DivAsync(numerator, denominator);
+    return FX_GetDivResult();
+}
+
+s32 FX_Inv(s32 x){
+    FX_InvAsync(x);
+    return FX_GetDivResult();
+}
+
+s32 FX_Sqrt(s32 x){
+    if (x > 0)
+    {
+        SETREG16(HW_REG_SQRTCNT, 0x1);
+        SETREG64(HW_REG_SQRT_PARAM, (s64)x << 32);
+        return FX_GetSqrtResult();
+    }
+    else
+    {
+        return 0;
+    }
+}
+
+s64 FX_GetDivResultFx64c(){
+    while (READREG16(HW_REG_DIVCNT) & 0x8000);
+    return READREG64(HW_REG_DIV_RESULT);
+}
+
+s32 FX_GetDivResult(){
+    while (READREG16(HW_REG_DIVCNT) & 0x8000);
+    return (READREG64(HW_REG_DIV_RESULT) + (1 << (0x14 - 1))) >> 0x14;
+}
+
+void FX_InvAsync(s32 x){
+    SETREG16(HW_REG_DIVCNT, 0x1);
+    SETREG64(HW_REG_DIV_NUMER, (s64)0x00001000 << 32);
+    SETREG64(HW_REG_DIV_DENOM, (u32)x);
+}
+
+s32 FX_GetSqrtResult(){
+    while (READREG16(HW_REG_SQRTCNT) & 0x8000);
+    return (READREG32(HW_REG_SQRT_RESULT) + (1 << (0xA - 1))) >> 0xA;
+}
+
+void FX_DivAsync(s32 numerator, s32 denominator){
+    SETREG16(HW_REG_DIVCNT, 0x1);
+    SETREG64(HW_REG_DIV_NUMER, (s64)numerator << 32);
+    SETREG64(HW_REG_DIV_DENOM, (u32)denominator);
+}
+
+ s32 FX_DivS32(s32 numerator, s32 denominator){
+     SETREG16(HW_REG_DIVCNT, 0x0);
+     SETREG32(HW_REG_DIV_NUMER, (u32)numerator); //32bit write for some reason
+     SETREG64(HW_REG_DIV_DENOM, (u32)denominator);
+     while (READREG16(HW_REG_DIVCNT) & 0x8000);
+     return READREG32(HW_REG_DIV_RESULT);
+ }
+
+ s32 FX_ModS32(s32 num, s32 mod){
+     SETREG16(HW_REG_DIVCNT, 0x0);
+     SETREG32(HW_REG_DIV_NUMER, (u32)num); //32bit write for some reason
+     SETREG64(HW_REG_DIV_DENOM, (u32)mod);
+     while (READREG16(HW_REG_DIVCNT) & 0x8000);
+     return READREG32(HW_REG_DIVREM_RESULT);
+ }
diff --git a/arm9/lib/src/FX_vec.c b/arm9/lib/src/FX_vec.c
index dc28bedf..7b838829 100644
--- a/arm9/lib/src/FX_vec.c
+++ b/arm9/lib/src/FX_vec.c
@@ -51,23 +51,6 @@ void VEC_Fx16CrossProduct(struct Vecx16 *a, struct Vecx16 *b, struct Vecx16 *dst
     dst->z = z;
 }
 
-#define HW_REG_DIVCNT              0x04000280
-#define HW_REG_DIV_NUMER           0x04000290
-#define HW_REG_DIV_DENOM           0x04000298
-#define HW_REG_DIV_RESULT          0x040002A0
-#define HW_REG_DIVREM_RESULT       0x040002A8
-
-#define HW_REG_SQRTCNT             0x040002B0
-#define HW_REG_SQRT_RESULT         0x040002B4
-#define HW_REG_SQRT_PARAM          0x040002B8
-
-#define SETREG16(x, y)             ((*(vu16 *)x) = y)
-#define SETREG32(x, y)             ((*(vu32 *)x) = y)
-#define SETREG64(x, y)             ((*(vu64 *)x) = y)
-#define READREG16(x)               (*(vu16 *)x)
-#define READREG32(x)               (*(vu32 *)x)
-#define READREG64(x)               (*(vu64 *)x)
-
 s32 VEC_Mag(struct Vecx32 *a){
     s64 l2 = (s64)a->x * a->x;
     l2 += (s64)a->y * a->y;
diff --git a/include/fx.h b/include/fx.h
index 82938219..e753eead 100644
--- a/include/fx.h
+++ b/include/fx.h
@@ -4,14 +4,41 @@
 #define FX32_INT_MASK     0xFFFFF000
 #define FX32_INT_ABS_MASK 0x7FFFF000
 #define FX32_FRAC_MASK    0x00000FFF
-
 #define FX32_INT_SHIFT    0xC
+
+#define FX64_INT_MASK     0xFFFFFFFFFFFFF000
+#define FX64_INT_ABS_MASK 0x7FFFFFFFFFFFF000
+#define FX64_FRAC_MASK    0x0000000000000FFF
 #define FX64_INT_SHIFT    0xC
 
+
 #define FX32_INT(x)       (((x) & FX32_INT_MASK) >> FX32_INT_SHIFT)
 #define FX32_INT_ABS(x)   (((x) & FX32_INT_ABS_MASK) >> FX32_INT_SHIFT)
 #define FX32_FRAC(x)      ((x) & FX32_FRAC_MASK)
 
+#define FX64_INT(x)       (((x) & FX64_INT_MASK) >> FX64_INT_SHIFT)
+#define FX64_INT_ABS(x)   (((x) & FX64_INT_ABS_MASK) >> FX64_INT_SHIFT)
+#define FX64_FRAC(x)      ((x) & FX64_FRAC_MASK)
+
+
+
+#define HW_REG_DIVCNT              0x04000280
+#define HW_REG_DIV_NUMER           0x04000290
+#define HW_REG_DIV_DENOM           0x04000298
+#define HW_REG_DIV_RESULT          0x040002A0
+#define HW_REG_DIVREM_RESULT       0x040002A8
+
+#define HW_REG_SQRTCNT             0x040002B0
+#define HW_REG_SQRT_RESULT         0x040002B4
+#define HW_REG_SQRT_PARAM          0x040002B8
+
+#define SETREG16(x, y)             ((*(vu16 *)x) = y)
+#define SETREG32(x, y)             ((*(vu32 *)x) = y)
+#define SETREG64(x, y)             ((*(vu64 *)x) = y)
+#define READREG16(x)               (*(vu16 *)x)
+#define READREG32(x)               (*(vu32 *)x)
+#define READREG64(x)               (*(vu64 *)x)
+
 struct Vecx32
 {
     s32 x;
-- 
cgit v1.2.3


From dbd9afd038a082bc58c1d93c105db8095a24a35f Mon Sep 17 00:00:00 2001
From: Made <made111@gmx.de>
Date: Sat, 2 May 2020 18:14:17 +0200
Subject: header cleanup

---
 include/fx.h | 56 +++++++++++++++++++++++++++++++++++++++++---------------
 1 file changed, 41 insertions(+), 15 deletions(-)

diff --git a/include/fx.h b/include/fx.h
index e753eead..22ef3d3a 100644
--- a/include/fx.h
+++ b/include/fx.h
@@ -1,24 +1,36 @@
 #ifndef GUARD_FX_H
 #define GUARD_FX_H
 
-#define FX32_INT_MASK     0xFFFFF000
-#define FX32_INT_ABS_MASK 0x7FFFF000
-#define FX32_FRAC_MASK    0x00000FFF
-#define FX32_INT_SHIFT    0xC
+#define FX32_INT_MASK              0xFFFFF000
+#define FX32_INT_ABS_MASK          0x7FFFF000
+#define FX32_FRAC_MASK             0x00000FFF
+#define FX32_INT_SHIFT             0xC
 
-#define FX64_INT_MASK     0xFFFFFFFFFFFFF000
-#define FX64_INT_ABS_MASK 0x7FFFFFFFFFFFF000
-#define FX64_FRAC_MASK    0x0000000000000FFF
-#define FX64_INT_SHIFT    0xC
+#define FX64_INT_MASK              0xFFFFFFFFFFFFF000
+#define FX64_INT_ABS_MASK          0x7FFFFFFFFFFFF000
+#define FX64_FRAC_MASK             0x0000000000000FFF
+#define FX64_INT_SHIFT             0xC
 
+#define FX64C_INT_MASK             0xFFFFFFFF00000000
+#define FX64C_INT_ABS_MASK         0x7FFFFFFF00000000
+#define FX64C_FRAC_MASK            0x00000000FFFFFFFF
+#define FX64C_INT_SHIFT            0x20
 
-#define FX32_INT(x)       (((x) & FX32_INT_MASK) >> FX32_INT_SHIFT)
-#define FX32_INT_ABS(x)   (((x) & FX32_INT_ABS_MASK) >> FX32_INT_SHIFT)
-#define FX32_FRAC(x)      ((x) & FX32_FRAC_MASK)
+#define FX_INT(TYPE, x)            (((x) & TYPE ## _INT_MASK) >> TYPE ## _INT_SHIFT)
+#define FX_INT_ABS(TYPE, x)        (((x) & TYPE ## _INT_ABS_MASK) >> TYPE ## _INT_SHIFT)
+#define FX_FRAC(TYPE, x)           ((x) & TYPE ## _FRAC_MASK)
 
-#define FX64_INT(x)       (((x) & FX64_INT_MASK) >> FX64_INT_SHIFT)
-#define FX64_INT_ABS(x)   (((x) & FX64_INT_ABS_MASK) >> FX64_INT_SHIFT)
-#define FX64_FRAC(x)      ((x) & FX64_FRAC_MASK)
+#define FX32_INT(x)                FX_INT(FX32, x)
+#define FX32_INT_ABS(x)            FX_INT_ABS(FX32, x)
+#define FX32_FRAC(x)               FX_FRAC(FX32, x)
+
+#define FX64_INT(x)                FX_INT(FX64, x)
+#define FX64_INT_ABS(x)            FX_INT_ABS(FX64, x)
+#define FX64_FRAC(x)               FX_FRAC(FX64, x)
+
+#define FX64C_INT(x)               FX_INT(FX64C, x)
+#define FX64C_INT_ABS(x)           FX_INT_ABS(FX64C, x)
+#define FX64C_FRAC(x)              FX_FRAC(FX64C, x)
 
 
 
@@ -53,9 +65,11 @@ struct Vecx16
     s16 z;
 };
 
+//FX
 void FX_Init();
 s32 FX_Modf(s32 x, s32 *iptr);
-s32 FX_Inv(s32 x);
+
+//Atan
 u16 FX_Atan(s32 x);
 u16 FX_Atan2(s32 x, s32 y);
 
@@ -65,4 +79,16 @@ void VEC_Subtract(struct Vecx32 *x, struct Vecx32 *y, struct Vecx32 *dst);
 void VEC_Fx16Add(struct Vecx16 *x, struct Vecx16 *y, struct Vecx16 *dst);
 s32 VEC_DotProduct(struct Vecx32 *x, struct Vecx32 *y);
 
+//CP
+s32 FX_Div(s32 numerator, s32 denominator);
+s32 FX_Inv(s32 x);
+s32 FX_Sqrt(s32 x);
+s64 FX_GetDivResultFx64c();
+s32 FX_GetDivResult();
+void FX_InvAsync(s32 x);
+s32 FX_GetSqrtResult();
+void FX_DivAsync(s32 numerator, s32 denominator);
+s32 FX_DivS32(s32 numerator, s32 denominator);
+s32 FX_ModS32(s32 num, s32 mod);
+
 #endif //GUARD_FX_H
-- 
cgit v1.2.3


From a382cad64f6d03d03ae87a9eae33c9cfbf314fde Mon Sep 17 00:00:00 2001
From: Made <made111@gmx.de>
Date: Sun, 3 May 2020 14:45:11 +0200
Subject: remove unnecessary asm

---
 arm9/asm/FX.s      |  32 ---------
 arm9/asm/FX_atan.s | 201 -----------------------------------------------------
 2 files changed, 233 deletions(-)
 delete mode 100644 arm9/asm/FX.s
 delete mode 100644 arm9/asm/FX_atan.s

diff --git a/arm9/asm/FX.s b/arm9/asm/FX.s
deleted file mode 100644
index 873a3afd..00000000
--- a/arm9/asm/FX.s
+++ /dev/null
@@ -1,32 +0,0 @@
-	.include "asm/macros.inc"
-	.include "global.inc"
-
-    .text
-
-	arm_func_start FX_Modf
-FX_Modf: ; 0x020C5928
-	cmp r0, #0x0
-	blt _020C5948
-	ldr r2, _020C596C ; =0x7FFFF000
-	ldr r3, _020C5970 ; =0x00000FFF
-	and r2, r0, r2
-	str r2, [r1, #0x0]
-	and r0, r0, r3
-	bx lr
-_020C5948:
-	ldr r2, _020C596C ; =0x7FFFF000
-	rsb r3, r0, #0x0
-	ldr r0, _020C5970 ; =0x00000FFF
-	and r2, r3, r2
-	rsb r2, r2, #0x0
-	and r0, r3, r0
-	str r2, [r1, #0x0]
-	rsb r0, r0, #0x0
-	bx lr
-	.balign 4
-_020C596C: .word 0x7FFFF000
-_020C5970: .word 0x00000FFF
-
-	arm_func_start FX_Init
-FX_Init: ; 0x020C5974
-	bx lr
diff --git a/arm9/asm/FX_atan.s b/arm9/asm/FX_atan.s
deleted file mode 100644
index 232a7211..00000000
--- a/arm9/asm/FX_atan.s
+++ /dev/null
@@ -1,201 +0,0 @@
-    .include "asm/macros.inc"
-    .include "global.inc"
-
-    .text
-
-	arm_func_start FX_Atan2
-FX_Atan2: ; 0x020C5668
-	stmdb sp!, {r4,lr}
-	cmp r0, #0x0
-	ble _020C5704
-	cmp r1, #0x0
-	ble _020C56B4
-	cmp r1, r0
-	movgt r3, r0
-	movgt r4, #0x0
-	movgt r0, #0x1
-	bgt _020C57C8
-	cmp r1, r0
-	movlt r3, r1
-	movlt r1, r0
-	movlt r4, #0x4000
-	movlt r0, #0x0
-	blt _020C57C8
-	mov r0, #0x2000
-	ldmia sp!, {r4,lr}
-	bx lr
-_020C56B4:
-	cmp r1, #0x0
-	bge _020C56F8
-	rsb r1, r1, #0x0
-	cmp r1, r0
-	movlt r3, r1
-	movlt r1, r0
-	movlt r4, #0x4000
-	movlt r0, #0x1
-	blt _020C57C8
-	cmp r1, r0
-	movgt r3, r0
-	movgt r4, #0x8000
-	movgt r0, #0x0
-	bgt _020C57C8
-	mov r0, #0x6000
-	ldmia sp!, {r4,lr}
-	bx lr
-_020C56F8:
-	mov r0, #0x4000
-	ldmia sp!, {r4,lr}
-	bx lr
-_020C5704:
-	cmp r0, #0x0
-	bge _020C57B4
-	cmp r1, #0x0
-	rsb r2, r0, #0x0
-	bge _020C5760
-	rsb r1, r1, #0x0
-	cmp r1, r2
-	movgt r0, #0x8000
-	rsbgt r4, r0, #0x0
-	movgt r3, r2
-	movgt r0, #0x1
-	bgt _020C57C8
-	cmp r1, r2
-	bge _020C5754
-	mov r0, #0x4000
-	mov r3, r1
-	mov r1, r2
-	rsb r4, r0, #0x0
-	mov r0, #0x0
-	b _020C57C8
-_020C5754:
-	mov r0, #0xa000
-	ldmia sp!, {r4,lr}
-	bx lr
-_020C5760:
-	cmp r1, #0x0
-	ble _020C57A8
-	cmp r1, r2
-	bge _020C5788
-	mov r0, #0x4000
-	mov r3, r1
-	mov r1, r2
-	rsb r4, r0, #0x0
-	mov r0, #0x1
-	b _020C57C8
-_020C5788:
-	cmp r1, r2
-	movgt r4, #0x0
-	movgt r3, r2
-	movgt r0, r4
-	bgt _020C57C8
-	mov r0, #0xe000
-	ldmia sp!, {r4,lr}
-	bx lr
-_020C57A8:
-	mov r0, #0xc000
-	ldmia sp!, {r4,lr}
-	bx lr
-_020C57B4:
-	cmp r1, #0x0
-	movge r0, #0x0
-	movlt r0, #0x8000
-	ldmia sp!, {r4,lr}
-	bx lr
-_020C57C8:
-	cmp r1, #0x0
-	moveq r0, #0x0
-	ldmeqia sp!, {r4,lr}
-	bxeq lr
-	cmp r0, #0x0
-	beq _020C580C
-	mov r0, r3
-	bl FX_Div
-	mov r1, r0, asr #0x5
-	ldr r0, _020C5838 ; =FX_AtanTable_
-	mov r1, r1, lsl #0x1
-	ldrsh r0, [r0, r1]
-	add r0, r4, r0
-	mov r0, r0, lsl #0x10
-	mov r0, r0, lsr #0x10
-	ldmia sp!, {r4,lr}
-	bx lr
-_020C580C:
-	mov r0, r3
-	bl FX_Div
-	mov r1, r0, asr #0x5
-	ldr r0, _020C5838 ; =FX_AtanTable_
-	mov r1, r1, lsl #0x1
-	ldrsh r0, [r0, r1]
-	sub r0, r4, r0
-	mov r0, r0, lsl #0x10
-	mov r0, r0, lsr #0x10
-	ldmia sp!, {r4,lr}
-	bx lr
-	.balign 4
-_020C5838: .word FX_AtanTable_
-
-	arm_func_start FX_Atan
-FX_Atan: ; 0x020C583C
-	stmdb sp!, {lr}
-	sub sp, sp, #0x4
-	cmp r0, #0x0
-	blt _020C58A4
-	cmp r0, #0x1000
-	ble _020C5880
-	bl FX_Inv
-	mov r1, r0, asr #0x5
-	ldr r0, _020C5924 ; =FX_AtanTable_
-	mov r1, r1, lsl #0x1
-	ldrsh r0, [r0, r1]
-	add sp, sp, #0x4
-	rsb r0, r0, #0x4000
-	mov r0, r0, lsl #0x10
-	mov r0, r0, lsr #0x10
-	ldmia sp!, {lr}
-	bx lr
-_020C5880:
-	cmp r0, #0x1000
-	movlt r1, r0, asr #0x5
-	ldrlt r0, _020C5924 ; =FX_AtanTable_
-	movlt r1, r1, lsl #0x1
-	ldrlth r0, [r0, r1]
-	add sp, sp, #0x4
-	movge r0, #0x2000
-	ldmia sp!, {lr}
-	bx lr
-_020C58A4:
-	mov r1, #0x1000
-	rsb r1, r1, #0x0
-	cmp r0, r1
-	bge _020C58E4
-	rsb r0, r0, #0x0
-	bl FX_Inv
-	mov r1, r0, asr #0x5
-	ldr r0, _020C5924 ; =FX_AtanTable_
-	mov r1, r1, lsl #0x1
-	ldrsh r0, [r0, r1]
-	add sp, sp, #0x4
-	sub r0, r0, #0x4000
-	mov r0, r0, lsl #0x10
-	mov r0, r0, lsr #0x10
-	ldmia sp!, {lr}
-	bx lr
-_020C58E4:
-	cmp r0, r1
-	addle sp, sp, #0x4
-	movle r0, #0xe000
-	ldmleia sp!, {lr}
-	bxle lr
-	rsb r0, r0, #0x0
-	mov r1, r0, asr #0x5
-	ldr r0, _020C5924 ; =FX_AtanTable_
-	mov r1, r1, lsl #0x1
-	ldrsh r0, [r0, r1]
-	rsb r0, r0, #0x0
-	mov r0, r0, lsl #0x10
-	mov r0, r0, lsr #0x10
-	add sp, sp, #0x4
-	ldmia sp!, {lr}
-	bx lr
-	.balign 4
-_020C5924: .word FX_AtanTable_
-- 
cgit v1.2.3


From 34dbccce4c71ba80c52cae8f1254c25d950f50c4 Mon Sep 17 00:00:00 2001
From: Made <made111@gmx.de>
Date: Sun, 3 May 2020 14:49:24 +0200
Subject: decompile FX_mtx44 and update fx header

---
 arm9/asm/FX_mtx44.s     | 362 ------------------------------------------------
 arm9/lib/src/FX_cp.c    |  34 ++---
 arm9/lib/src/FX_mtx44.c | 172 +++++++++++++++++++++++
 include/fx.h            |  50 ++++++-
 4 files changed, 236 insertions(+), 382 deletions(-)
 delete mode 100644 arm9/asm/FX_mtx44.s
 create mode 100644 arm9/lib/src/FX_mtx44.c

diff --git a/arm9/asm/FX_mtx44.s b/arm9/asm/FX_mtx44.s
deleted file mode 100644
index 0b042dee..00000000
--- a/arm9/asm/FX_mtx44.s
+++ /dev/null
@@ -1,362 +0,0 @@
-	.include "asm/macros.inc"
-	.include "global.inc"
-
-    .text
-
-	arm_func_start MTX_Identity44_
-MTX_Identity44_: ; 0x020C49C4
-	mov r2, #0x1000
-	mov r3, #0x0
-	stmia r0!, {r2-r3}
-	mov r1, #0x0
-	stmia r0!, {r1,r3}
-	stmia r0!, {r1-r3}
-	stmia r0!, {r1,r3}
-	stmia r0!, {r1-r3}
-	stmia r0!, {r1,r3}
-	stmia r0!, {r1-r2}
-	bx lr
-
-	arm_func_start MTX_Copy44To43_
-MTX_Copy44To43_:
-	ldmia r0!, {r2-r3,r12}
-	add r0, r0, #0x4
-	stmia r1!, {r2-r3,r12}
-	ldmia r0!, {r2-r3,r12}
-	add r0, r0, #0x4
-	stmia r1!, {r2-r3,r12}
-	ldmia r0!, {r2-r3,r12}
-	add r0, r0, #0x4
-	stmia r1!, {r2-r3,r12}
-	ldmia r0!, {r2-r3,r12}
-	add r0, r0, #0x4
-	stmia r1!, {r2-r3,r12}
-	bx lr
-
-	thumb_func_start MTX_RotX44_
-MTX_RotX44_: ; 0x020C4A24
-	str r2, [r0, #0x14]
-	str r2, [r0, #0x28]
-	str r1, [r0, #0x18]
-	neg r1, r1
-	str r1, [r0, #0x24]
-	mov r1, #0x1
-	mov r2, #0x0
-	lsl r1, r1, #0xc
-	mov r3, #0x0
-	stmia r0!, {r1-r3}
-	stmia r0!, {r2-r3}
-	add r0, #0x8
-	stmia r0!, {r2-r3}
-	add r0, #0x8
-	stmia r0!, {r2-r3}
-	stmia r0!, {r2-r3}
-	str r1, [r0, #0x0]
-	bx lr
-
-	thumb_func_start MTX_RotY44_
-MTX_RotY44_: ; 0x020C4A48
-	str r2, [r0, #0x0]
-	str r2, [r0, #0x28]
-	str r1, [r0, #0x20]
-	neg r1, r1
-	str r1, [r0, #0x8]
-	mov r3, #0x1
-	mov r1, #0x0
-	lsl r3, r3, #0xc
-	mov r2, #0x0
-	str r2, [r0, #0x4]
-	add r0, #0xc
-	stmia r0!, {r1-r3}
-	stmia r0!, {r1-r2}
-	str r2, [r0, #0x4]
-	add r0, #0xc
-	stmia r0!, {r1-r2}
-	stmia r0!, {r1-r3}
-	bx lr
-
-	thumb_func_start MTX_RotZ44_
-MTX_RotZ44_: ; 0x020C4A6C
-	str r2, [r0, #0x0]
-	str r2, [r0, #0x14]
-	str r1, [r0, #0x4]
-	neg r1, r1
-	str r1, [r0, #0x10]
-	mov r3, #0x1
-	mov r1, #0x0
-	lsl r3, r3, #0xc
-	mov r2, #0x0
-	add r0, #0x8
-	stmia r0!, {r1-r2}
-	add r0, #0x8
-	stmia r0!, {r1-r2}
-	stmia r0!, {r1-r3}
-	stmia r0!, {r1-r2}
-	stmia r0!, {r1-r3}
-	bx lr
-	.balign 4
-
-	arm_func_start MTX_Concat44
-MTX_Concat44: ; 0x020C4A90
-	stmdb sp!, {r4-r11,lr}
-	sub sp, sp, #0x44
-	ldr r5, [r0, #0x4]
-	ldr r3, [r1, #0x10]
-	mov r12, r2
-	smull r6, r4, r5, r3
-	ldr r9, [r0, #0x0]
-	ldr r3, [r1, #0x0]
-	cmp r12, r1
-	smlal r6, r4, r9, r3
-	ldr r10, [r0, #0x8]
-	ldr r3, [r1, #0x20]
-	ldr r8, [r0, #0xc]
-	smlal r6, r4, r10, r3
-	ldr r3, [r1, #0x30]
-	addeq r2, sp, #0x0
-	smlal r6, r4, r8, r3
-	mov r3, r6, lsr #0xc
-	orr r3, r3, r4, lsl #0x14
-	str r3, [r2, #0x0]
-	ldr r3, [r1, #0x14]
-	ldr r4, [r1, #0x4]
-	smull r7, r6, r5, r3
-	smlal r7, r6, r9, r4
-	ldr r3, [r1, #0x24]
-	ldr r4, [r1, #0x34]
-	smlal r7, r6, r10, r3
-	smlal r7, r6, r8, r4
-	mov r3, r7, lsr #0xc
-	orr r3, r3, r6, lsl #0x14
-	str r3, [r2, #0x4]
-	ldr r3, [r1, #0x1c]
-	ldr r4, [r1, #0xc]
-	smull r7, r6, r5, r3
-	smlal r7, r6, r9, r4
-	ldr r3, [r1, #0x2c]
-	ldr r4, [r1, #0x3c]
-	smlal r7, r6, r10, r3
-	smlal r7, r6, r8, r4
-	mov r3, r7, lsr #0xc
-	orr r3, r3, r6, lsl #0x14
-	str r3, [r2, #0xc]
-	ldr r6, [r1, #0x18]
-	ldr r7, [r1, #0x8]
-	smull r4, r3, r5, r6
-	smlal r4, r3, r9, r7
-	ldr r9, [r1, #0x28]
-	ldr r5, [r1, #0x38]
-	smlal r4, r3, r10, r9
-	smlal r4, r3, r8, r5
-	mov r4, r4, lsr #0xc
-	orr r4, r4, r3, lsl #0x14
-	str r4, [r2, #0x8]
-	ldr r8, [r0, #0x14]
-	ldr r10, [r0, #0x10]
-	smull r4, r3, r8, r6
-	smlal r4, r3, r10, r7
-	ldr r7, [r0, #0x18]
-	ldr r6, [r0, #0x1c]
-	smlal r4, r3, r7, r9
-	smlal r4, r3, r6, r5
-	mov r4, r4, lsr #0xc
-	orr r4, r4, r3, lsl #0x14
-	str r4, [r2, #0x18]
-	ldr r3, [r1, #0x14]
-	ldr r4, [r1, #0x4]
-	smull r11, r3, r8, r3
-	smlal r11, r3, r10, r4
-	ldr r5, [r1, #0x24]
-	ldr r9, [r1, #0x34]
-	smlal r11, r3, r7, r5
-	smlal r11, r3, r6, r9
-	mov r4, r11, lsr #0xc
-	orr r4, r4, r3, lsl #0x14
-	str r4, [r2, #0x14]
-	ldr r3, [r1, #0x1c]
-	ldr r4, [r1, #0xc]
-	smull r11, r3, r8, r3
-	smlal r11, r3, r10, r4
-	ldr r5, [r1, #0x2c]
-	ldr r9, [r1, #0x3c]
-	smlal r11, r3, r7, r5
-	smlal r11, r3, r6, r9
-	mov r4, r11, lsr #0xc
-	orr r4, r4, r3, lsl #0x14
-	str r4, [r2, #0x1c]
-	ldr r9, [r1, #0x10]
-	ldr r11, [r1, #0x0]
-	smull r4, r3, r8, r9
-	smlal r4, r3, r10, r11
-	ldr r5, [r1, #0x20]
-	ldr r8, [r1, #0x30]
-	smlal r4, r3, r7, r5
-	smlal r4, r3, r6, r8
-	mov r4, r4, lsr #0xc
-	orr r4, r4, r3, lsl #0x14
-	str r4, [r2, #0x10]
-	ldr r6, [r0, #0x24]
-	ldr r7, [r0, #0x20]
-	smull r4, r3, r6, r9
-	smlal r4, r3, r7, r11
-	ldr r10, [r0, #0x28]
-	add lr, sp, #0x0
-	smlal r4, r3, r10, r5
-	ldr r5, [r0, #0x2c]
-	cmp r2, lr
-	smlal r4, r3, r5, r8
-	mov r4, r4, lsr #0xc
-	orr r4, r4, r3, lsl #0x14
-	str r4, [r2, #0x20]
-	ldr r3, [r1, #0x14]
-	ldr r4, [r1, #0x4]
-	smull r11, r3, r6, r3
-	smlal r11, r3, r7, r4
-	ldr r8, [r1, #0x24]
-	ldr r9, [r1, #0x34]
-	smlal r11, r3, r10, r8
-	smlal r11, r3, r5, r9
-	mov r4, r11, lsr #0xc
-	orr r4, r4, r3, lsl #0x14
-	str r4, [r2, #0x24]
-	ldr r3, [r1, #0x1c]
-	ldr r4, [r1, #0xc]
-	smull r11, r3, r6, r3
-	smlal r11, r3, r7, r4
-	ldr r8, [r1, #0x2c]
-	ldr r9, [r1, #0x3c]
-	smlal r11, r3, r10, r8
-	smlal r11, r3, r5, r9
-	mov r4, r11, lsr #0xc
-	orr r4, r4, r3, lsl #0x14
-	str r4, [r2, #0x2c]
-	ldr r9, [r1, #0x18]
-	ldr r8, [r1, #0x8]
-	smull r4, r3, r6, r9
-	smlal r4, r3, r7, r8
-	ldr r7, [r1, #0x28]
-	ldr r6, [r1, #0x38]
-	smlal r4, r3, r10, r7
-	smlal r4, r3, r5, r6
-	mov r4, r4, lsr #0xc
-	orr r4, r4, r3, lsl #0x14
-	str r4, [r2, #0x28]
-	ldr r5, [r0, #0x34]
-	ldr r4, [r0, #0x30]
-	smull r10, r9, r5, r9
-	ldr r3, [r0, #0x38]
-	smlal r10, r9, r4, r8
-	smlal r10, r9, r3, r7
-	ldr r0, [r0, #0x3c]
-	addne sp, sp, #0x44
-	smlal r10, r9, r0, r6
-	mov r6, r10, lsr #0xc
-	orr r6, r6, r9, lsl #0x14
-	str r6, [r2, #0x38]
-	ldr r6, [r1, #0x14]
-	ldr r7, [r1, #0x4]
-	smull r10, r6, r5, r6
-	smlal r10, r6, r4, r7
-	ldr r8, [r1, #0x24]
-	ldr r9, [r1, #0x34]
-	smlal r10, r6, r3, r8
-	smlal r10, r6, r0, r9
-	mov r7, r10, lsr #0xc
-	orr r7, r7, r6, lsl #0x14
-	str r7, [r2, #0x34]
-	ldr r6, [r1, #0x10]
-	ldr r7, [r1, #0x0]
-	smull r10, r6, r5, r6
-	smlal r10, r6, r4, r7
-	ldr r8, [r1, #0x20]
-	ldr r9, [r1, #0x30]
-	smlal r10, r6, r3, r8
-	smlal r10, r6, r0, r9
-	mov r7, r10, lsr #0xc
-	orr r7, r7, r6, lsl #0x14
-	str r7, [r2, #0x30]
-	ldr r8, [r1, #0x3c]
-	ldr r7, [r1, #0x2c]
-	ldr r6, [r1, #0xc]
-	ldr r1, [r1, #0x1c]
-	smull r9, r1, r5, r1
-	smlal r9, r1, r4, r6
-	smlal r9, r1, r3, r7
-	smlal r9, r1, r0, r8
-	mov r0, r9, lsr #0xc
-	orr r0, r0, r1, lsl #0x14
-	str r0, [r2, #0x3c]
-	ldmneia sp!, {r4-r11,lr}
-	bxne lr
-	ldmia lr!, {r0-r3}
-	stmia r12!, {r0-r3}
-	ldmia lr!, {r0-r3}
-	stmia r12!, {r0-r3}
-	ldmia lr!, {r0-r3}
-	stmia r12!, {r0-r3}
-	ldmia lr, {r0-r3}
-	stmia r12, {r0-r3}
-	add sp, sp, #0x44
-	ldmia sp!, {r4-r11,lr}
-	bx lr
-
-	arm_func_start MTX_TransApply44
-MTX_TransApply44: ; 0x020C4DB4
-	stmdb sp!, {r4-r7,lr}
-	sub sp, sp, #0x4
-	mov r7, r0
-	mov r6, r1
-	mov r5, r2
-	mov r4, r3
-	cmp r7, r6
-	beq _020C4DD8
-	bl MI_Copy48B
-_020C4DD8:
-	ldr r0, [r7, #0x10]
-	ldr r1, [r7, #0x0]
-	smull r3, r2, r4, r0
-	smlal r3, r2, r5, r1
-	ldr r0, [sp, #0x18]
-	ldr r1, [r7, #0x20]
-	ldr r12, [r7, #0x30]
-	smlal r3, r2, r0, r1
-	mov r1, r3, lsr #0xc
-	orr r1, r1, r2, lsl #0x14
-	add r1, r12, r1
-	str r1, [r6, #0x30]
-	ldr r1, [r7, #0x14]
-	ldr r2, [r7, #0x4]
-	smull r12, r3, r4, r1
-	smlal r12, r3, r5, r2
-	ldr r1, [r7, #0x24]
-	ldr r2, [r7, #0x34]
-	smlal r12, r3, r0, r1
-	mov r1, r12, lsr #0xc
-	orr r1, r1, r3, lsl #0x14
-	add r1, r2, r1
-	str r1, [r6, #0x34]
-	ldr r1, [r7, #0x18]
-	ldr r2, [r7, #0x8]
-	smull r12, r3, r4, r1
-	smlal r12, r3, r5, r2
-	ldr r1, [r7, #0x28]
-	ldr r2, [r7, #0x38]
-	smlal r12, r3, r0, r1
-	mov r1, r12, lsr #0xc
-	orr r1, r1, r3, lsl #0x14
-	add r1, r2, r1
-	str r1, [r6, #0x38]
-	ldr r1, [r7, #0x1c]
-	ldr r2, [r7, #0xc]
-	smull r12, r3, r4, r1
-	smlal r12, r3, r5, r2
-	ldr r1, [r7, #0x2c]
-	ldr r2, [r7, #0x3c]
-	smlal r12, r3, r0, r1
-	mov r0, r12, lsr #0xc
-	orr r0, r0, r3, lsl #0x14
-	add r0, r2, r0
-	str r0, [r6, #0x3c]
-	add sp, sp, #0x4
-	ldmia sp!, {r4-r7,lr}
-	bx lr
diff --git a/arm9/lib/src/FX_cp.c b/arm9/lib/src/FX_cp.c
index a8da0a8f..a2d8307b 100644
--- a/arm9/lib/src/FX_cp.c
+++ b/arm9/lib/src/FX_cp.c
@@ -1,11 +1,7 @@
 #include "global.h"
 #include "main.h"
 #include "fx.h"
-
-s32 FX_GetDivResult();
-s32 FX_GetSqrtResult();
-void FX_DivAsync(s32 numerator, s32 denominator);
-void FX_InvAsync(s32 x);
+ 
 
 s32 FX_Div(s32 numerator, s32 denominator){
     FX_DivAsync(numerator, denominator);
@@ -57,18 +53,18 @@ void FX_DivAsync(s32 numerator, s32 denominator){
     SETREG64(HW_REG_DIV_DENOM, (u32)denominator);
 }
 
- s32 FX_DivS32(s32 numerator, s32 denominator){
-     SETREG16(HW_REG_DIVCNT, 0x0);
-     SETREG32(HW_REG_DIV_NUMER, (u32)numerator); //32bit write for some reason
-     SETREG64(HW_REG_DIV_DENOM, (u32)denominator);
-     while (READREG16(HW_REG_DIVCNT) & 0x8000);
-     return READREG32(HW_REG_DIV_RESULT);
- }
+s32 FX_DivS32(s32 numerator, s32 denominator){
+    SETREG16(HW_REG_DIVCNT, 0x0);
+    SETREG32(HW_REG_DIV_NUMER, (u32)numerator); //32bit write for some reason
+    SETREG64(HW_REG_DIV_DENOM, (u32)denominator);
+    while (READREG16(HW_REG_DIVCNT) & 0x8000);
+    return READREG32(HW_REG_DIV_RESULT);
+}
 
- s32 FX_ModS32(s32 num, s32 mod){
-     SETREG16(HW_REG_DIVCNT, 0x0);
-     SETREG32(HW_REG_DIV_NUMER, (u32)num); //32bit write for some reason
-     SETREG64(HW_REG_DIV_DENOM, (u32)mod);
-     while (READREG16(HW_REG_DIVCNT) & 0x8000);
-     return READREG32(HW_REG_DIVREM_RESULT);
- }
+s32 FX_ModS32(s32 num, s32 mod){
+    SETREG16(HW_REG_DIVCNT, 0x0);
+    SETREG32(HW_REG_DIV_NUMER, (u32)num); //32bit write for some reason
+    SETREG64(HW_REG_DIV_DENOM, (u32)mod);
+    while (READREG16(HW_REG_DIVCNT) & 0x8000);
+    return READREG32(HW_REG_DIVREM_RESULT);
+}
diff --git a/arm9/lib/src/FX_mtx44.c b/arm9/lib/src/FX_mtx44.c
new file mode 100644
index 00000000..3c91b4df
--- /dev/null
+++ b/arm9/lib/src/FX_mtx44.c
@@ -0,0 +1,172 @@
+#include "global.h"
+#include "main.h"
+#include "fx.h"
+
+void MI_Copy48B(void *src, void *dst);
+
+
+void MTX_TransApply44(struct Mtx44 *mtx, struct Mtx44 *dst, s32 x, s32 y, s32 z){
+    if(mtx != dst)
+        MI_Copy48B(mtx, dst);
+    dst->_[12] = mtx->_[12] + (s32)(((s64)x * mtx->_[0] + (s64)y * mtx->_[4] + (s64)z * mtx->_[8] ) >> FX32_INT_SHIFT);
+    dst->_[13] = mtx->_[13] + (s32)(((s64)x * mtx->_[1] + (s64)y * mtx->_[5] + (s64)z * mtx->_[9] ) >> FX32_INT_SHIFT);
+    dst->_[14] = mtx->_[14] + (s32)(((s64)x * mtx->_[2] + (s64)y * mtx->_[6] + (s64)z * mtx->_[10]) >> FX32_INT_SHIFT);
+    dst->_[15] = mtx->_[15] + (s32)(((s64)x * mtx->_[3] + (s64)y * mtx->_[7] + (s64)z * mtx->_[11]) >> FX32_INT_SHIFT);
+}
+
+void MTX_Concat44(struct Mtx44 *a, struct Mtx44 *b, struct Mtx44 *c){
+    struct Mtx44 temp;
+    struct Mtx44 *dst;
+    s32 a0, a1, a2, a3;
+    s32 b0, b1, b2, b3;
+
+    if (c == b)
+        dst = &temp;
+    else
+        dst = c;
+
+    a0 = a->_[0];
+    a1 = a->_[1];
+    a2 = a->_[2];
+    a3 = a->_[3];
+    dst->_[0] =  (((s64)a0  * b->_[0] + (s64)a1  * b->_[4] + (s64)a2  * b->_[8]  + (s64)a3  * b->_[12]) >> FX32_INT_SHIFT);
+    dst->_[1] =  (((s64)a0  * b->_[1] + (s64)a1  * b->_[5] + (s64)a2  * b->_[9]  + (s64)a3  * b->_[13]) >> FX32_INT_SHIFT);
+    dst->_[3] =  (((s64)a0  * b->_[3] + (s64)a1  * b->_[7] + (s64)a2  * b->_[11] + (s64)a3  * b->_[15]) >> FX32_INT_SHIFT);
+    b0 = b->_[2];
+    b1 = b->_[6];
+    b2 = b->_[10];
+    b3 = b->_[14];
+    dst->_[2] =  (((s64)a0  * b0 + (s64)a1  * b1 + (s64)a2  * b2 + (s64)a3  * b3) >> FX32_INT_SHIFT);
+    a0 = a->_[4];
+    a1 = a->_[5];
+    a2 = a->_[6];
+    a3 = a->_[7];
+    dst->_[6] =  (((s64)a0  * b0 + (s64)a1  * b1 + (s64)a2  * b2 + (s64)a3  * b3) >> FX32_INT_SHIFT);
+    dst->_[5] =  (((s64)a0  * b->_[1] + (s64)a1  * b->_[5] + (s64)a2  * b->_[9]  + (s64)a3  * b->_[13]) >> FX32_INT_SHIFT);
+    dst->_[7] =  (((s64)a0  * b->_[3] + (s64)a1  * b->_[7] + (s64)a2  * b->_[11] + (s64)a3  * b->_[15]) >> FX32_INT_SHIFT);
+    b0 = b->_[0];
+    b1 = b->_[4];
+    b2 = b->_[8];
+    b3 = b->_[12];
+    dst->_[4] =  (((s64)a0  * b0 + (s64)a1  * b1 + (s64)a2  * b2  + (s64)a3  * b3) >> FX32_INT_SHIFT);
+    a0 = a->_[8];
+    a1 = a->_[9];
+    a2 = a->_[10];
+    a3 = a->_[11];
+    dst->_[8] =  (((s64)a0  * b0 + (s64)a1  * b1 + (s64)a2 * b2  + (s64)a3 * b3) >> FX32_INT_SHIFT);
+    dst->_[9] =  (((s64)a0  * b->_[1] + (s64)a1  * b->_[5] + (s64)a2 * b->_[9]  + (s64)a3 * b->_[13]) >> FX32_INT_SHIFT);
+    dst->_[11] = (((s64)a0  * b->_[3] + (s64)a1  * b->_[7] + (s64)a2 * b->_[11] + (s64)a3 * b->_[15]) >> FX32_INT_SHIFT);
+    b0 = b->_[2];
+    b1 = b->_[6];
+    b2 = b->_[10];
+    b3 = b->_[14];
+    dst->_[10] = (((s64)a0  * b0 + (s64)a1  * b1 + (s64)a2 * b2 + (s64)a3 * b3) >> FX32_INT_SHIFT);
+    a0 = a->_[12];
+    a1 = a->_[13];
+    a2 = a->_[14];
+    a3 = a->_[15];
+    dst->_[14] = (((s64)a0 * b0 + (s64)a1 * b1 + (s64)a2 * b2 + (s64)a3 * b3) >> FX32_INT_SHIFT);
+    dst->_[13] = (((s64)a0 * b->_[1] + (s64)a1 * b->_[5] + (s64)a2 * b->_[9]  + (s64)a3 * b->_[13]) >> FX32_INT_SHIFT);
+    dst->_[12] = (((s64)a0 * b->_[0] + (s64)a1 * b->_[4] + (s64)a2 * b->_[8]  + (s64)a3 * b->_[12]) >> FX32_INT_SHIFT);
+    dst->_[15] = (((s64)a0 * b->_[3] + (s64)a1 * b->_[7] + (s64)a2 * b->_[11] + (s64)a3 * b->_[15]) >> FX32_INT_SHIFT);
+    if (dst == &temp)
+        *c = temp;
+}
+
+asm void MTX_Identity44_(struct Mtx44 *dst){
+    mov r2, #0x1000
+    mov r3, #0x0
+    stmia r0!, {r2-r3}
+    mov r1, #0x0
+    stmia r0!, {r1,r3}
+    stmia r0!, {r1-r3}
+    stmia r0!, {r1,r3}
+    stmia r0!, {r1-r3}
+    stmia r0!, {r1,r3}
+    stmia r0!, {r1-r2}
+    bx lr
+}
+
+asm void MTX_Copy44To43_(struct Mtx44 *src, struct Mtx43 *dst){
+    ldmia r0!, {r2-r3,r12}
+    add r0, r0, #0x4
+    stmia r1!, {r2-r3,r12}
+    ldmia r0!, {r2-r3,r12}
+    add r0, r0, #0x4
+    stmia r1!, {r2-r3,r12}
+    ldmia r0!, {r2-r3,r12}
+    add r0, r0, #0x4
+    stmia r1!, {r2-r3,r12}
+    ldmia r0!, {r2-r3,r12}
+    add r0, r0, #0x4
+    stmia r1!, {r2-r3,r12}
+    bx lr
+}
+
+#pragma thumb on
+asm void MTX_RotX44_(struct Mtx44 *mtx, s32 sinphi, s32 cosphi){
+    str r2, [r0, #0x14]
+	str r2, [r0, #0x28]
+	str r1, [r0, #0x18]
+	neg r1, r1
+	str r1, [r0, #0x24]
+	mov r1, #0x1
+	mov r2, #0x0
+	lsl r1, r1, #0xc
+	mov r3, #0x0
+	stmia r0!, {r1-r3}
+	stmia r0!, {r2-r3}
+	add r0, #0x8
+	stmia r0!, {r2-r3}
+	add r0, #0x8
+	stmia r0!, {r2-r3}
+	stmia r0!, {r2-r3}
+	str r1, [r0, #0x0]
+	bx lr
+}
+#pragma thumb off
+
+#pragma thumb on
+asm void MTX_RotY44_(struct Mtx44 *mtx, s32 sinphi, s32 cosphi){
+    str r2, [r0, #0x0]
+	str r2, [r0, #0x28]
+	str r1, [r0, #0x20]
+	neg r1, r1
+	str r1, [r0, #0x8]
+	mov r3, #0x1
+	mov r1, #0x0
+	lsl r3, r3, #0xc
+	mov r2, #0x0
+	str r2, [r0, #0x4]
+	add r0, #0xc
+	stmia r0!, {r1-r3}
+	stmia r0!, {r1-r2}
+	str r2, [r0, #0x4]
+	add r0, #0xc
+	stmia r0!, {r1-r2}
+	stmia r0!, {r1-r3}
+	bx lr
+}
+#pragma thumb off
+
+#pragma thumb on
+asm void MTX_RotZ44_(struct Mtx44 *mtx, s32 sinphi, s32 cosphi){
+    str r2, [r0, #0x0]
+    str r2, [r0, #0x14]
+    str r1, [r0, #0x4]
+    neg r1, r1
+    str r1, [r0, #0x10]
+    mov r3, #0x1
+    mov r1, #0x0
+    lsl r3, r3, #0xc
+    mov r2, #0x0
+    add r0, #0x8
+    stmia r0!, {r1-r2}
+    add r0, #0x8
+    stmia r0!, {r1-r2}
+    stmia r0!, {r1-r3}
+    stmia r0!, {r1-r2}
+    stmia r0!, {r1-r3}
+    bx lr
+}
+#pragma thumb off
diff --git a/include/fx.h b/include/fx.h
index 22ef3d3a..1e9eb291 100644
--- a/include/fx.h
+++ b/include/fx.h
@@ -1,6 +1,11 @@
 #ifndef GUARD_FX_H
 #define GUARD_FX_H
 
+#define FX16_INT_MASK              0xF000
+#define FX16_INT_ABS_MASK          0x7000
+#define FX16_FRAC_MASK             0x0FFF
+#define FX16_INT_SHIFT             0xC
+
 #define FX32_INT_MASK              0xFFFFF000
 #define FX32_INT_ABS_MASK          0x7FFFF000
 #define FX32_FRAC_MASK             0x00000FFF
@@ -20,6 +25,10 @@
 #define FX_INT_ABS(TYPE, x)        (((x) & TYPE ## _INT_ABS_MASK) >> TYPE ## _INT_SHIFT)
 #define FX_FRAC(TYPE, x)           ((x) & TYPE ## _FRAC_MASK)
 
+#define FX16_INT(x)                FX_INT(FX16, x)
+#define FX16_INT_ABS(x)            FX_INT_ABS(FX16, x)
+#define FX16_FRAC(x)               FX_FRAC(FX16, x)
+
 #define FX32_INT(x)                FX_INT(FX32, x)
 #define FX32_INT_ABS(x)            FX_INT_ABS(FX32, x)
 #define FX32_FRAC(x)               FX_FRAC(FX32, x)
@@ -33,7 +42,6 @@
 #define FX64C_FRAC(x)              FX_FRAC(FX64C, x)
 
 
-
 #define HW_REG_DIVCNT              0x04000280
 #define HW_REG_DIV_NUMER           0x04000290
 #define HW_REG_DIV_DENOM           0x04000298
@@ -65,6 +73,26 @@ struct Vecx16
     s16 z;
 };
 
+struct Mtx44
+{
+    s32 _[16];
+};
+
+struct Mtx43
+{
+    s32 _[12];
+};
+
+struct Mtx33
+{
+    s32 _[9];
+};
+
+struct Mtx22
+{
+    s32 _[4];
+};
+
 //FX
 void FX_Init();
 s32 FX_Modf(s32 x, s32 *iptr);
@@ -78,6 +106,13 @@ void VEC_Add(struct Vecx32 *x, struct Vecx32 *y, struct Vecx32 *dst);
 void VEC_Subtract(struct Vecx32 *x, struct Vecx32 *y, struct Vecx32 *dst);
 void VEC_Fx16Add(struct Vecx16 *x, struct Vecx16 *y, struct Vecx16 *dst);
 s32 VEC_DotProduct(struct Vecx32 *x, struct Vecx32 *y);
+s32 VEC_Fx16DotProduct(struct Vecx16 *a, struct Vecx16 *b);
+void VEC_CrossProduct(struct Vecx32 *a, struct Vecx32 *b, struct Vecx32 *dst);
+void VEC_Fx16CrossProduct(struct Vecx16 *a, struct Vecx16 *b, struct Vecx16 *dst);
+s32 VEC_Mag(struct Vecx32 *a);
+void VEC_Normalize(struct Vecx32 *a, struct Vecx32 *dst);
+void VEC_Fx16Normalize(struct Vecx16 *a, struct Vecx16 *dst);
+void VEC_MultAdd(s32 factor, struct Vecx32  *a, struct Vecx32 *b, struct Vecx32 *dst);
 
 //CP
 s32 FX_Div(s32 numerator, s32 denominator);
@@ -91,4 +126,17 @@ void FX_DivAsync(s32 numerator, s32 denominator);
 s32 FX_DivS32(s32 numerator, s32 denominator);
 s32 FX_ModS32(s32 num, s32 mod);
 
+//Mtx
+//The functions ending in underscores seem to be in assembly originally
+//at least I see no way to recreate the stmia copies and some of them are actually in thumb
+//Mtx44
+void MTX_TransApply44(struct Mtx44 *mtx, struct Mtx44 *dst, s32 x, s32 y, s32 z);
+void MTX_Concat44(struct Mtx44 *a, struct Mtx44 *b, struct Mtx44 *c);
+void MTX_Identity44_(struct Mtx44 *dst);
+void MTX_Copy44To43_(struct Mtx44 *src, struct Mtx43 *dst);
+void MTX_RotX44_(struct Mtx44 *mtx, s32 sinphi, s32 cosphi);
+void MTX_RotY44_(struct Mtx44 *mtx, s32 sinphi, s32 cosphi);
+void MTX_RotZ44_(struct Mtx44 *mtx, s32 sinphi, s32 cosphi);
+
+
 #endif //GUARD_FX_H
-- 
cgit v1.2.3


From 8cd2ae6b66d4e2b2d5dbbf4e54f5a21cb21d7ebf Mon Sep 17 00:00:00 2001
From: Made <made111@gmx.de>
Date: Mon, 4 May 2020 15:57:14 +0200
Subject: Decompile FX_mtx43.s

---
 arm9/asm/FX_mtx43.s     | 523 ------------------------------------------------
 arm9/lib/src/FX_mtx43.c | 207 +++++++++++++++++++
 include/fx.h            |  12 ++
 3 files changed, 219 insertions(+), 523 deletions(-)
 delete mode 100644 arm9/asm/FX_mtx43.s
 create mode 100644 arm9/lib/src/FX_mtx43.c

diff --git a/arm9/asm/FX_mtx43.s b/arm9/asm/FX_mtx43.s
deleted file mode 100644
index 601bf236..00000000
--- a/arm9/asm/FX_mtx43.s
+++ /dev/null
@@ -1,523 +0,0 @@
-	.include "asm/macros.inc"
-	.include "global.inc"
-
-	.text
-
-	arm_func_start MTX_Identity43_
-MTX_Identity43_: ; 0x020C426C
-	mov r2, #0x1000
-	mov r3, #0x0
-	stmia r0!, {r2-r3}
-	mov r1, #0x0
-	stmia r0!, {r1,r3}
-	stmia r0!, {r2-r3}
-	stmia r0!, {r1,r3}
-	stmia r0!, {r2-r3}
-	stmia r0!, {r1,r3}
-	bx lr
-
-	arm_func_start MTX_Copy43To44_
-MTX_Copy43To44_: ; 0x020C4294
-	stmdb sp!, {r4}
-	mov r12, #0x0
-	ldmia r0!, {r2-r4}
-	stmia r1!, {r2-r4,r12}
-	ldmia r0!, {r2-r4}
-	stmia r1!, {r2-r4,r12}
-	ldmia r0!, {r2-r4}
-	stmia r1!, {r2-r4,r12}
-	mov r12, #0x1000
-	ldmia r0!, {r2-r4}
-	stmia r1!, {r2-r4,r12}
-	ldmia sp!, {r4}
-	bx lr
-
-	thumb_func_start MTX_Scale43_
-MTX_Scale43_: ; 0x020C42C8
-	stmia r0!, {r1}
-	mov r1, #0x0
-	str r3, [r0, #0x1c]
-	mov r3, #0x0
-	stmia r0!, {r1,r3}
-	stmia r0!, {r1-r3}
-	mov r2, #0x0
-	stmia r0!, {r1,r3}
-	add r0, #0x4
-	stmia r0!, {r1-r3}
-	bx lr
-	.balign 4
-
-	thumb_func_start MTX_RotX43_
-MTX_RotX43_: ; 0x020C42E0
-	str r1, [r0, #0x14]
-	neg r1, r1
-	str r1, [r0, #0x1c]
-	mov r1, #0x1
-	lsl r1, r1, #0xc
-	stmia r0!, {r1}
-	mov r3, #0x0
-	mov r1, #0x0
-	stmia r0!, {r1,r3}
-	stmia r0!, {r1-r2}
-	str r1, [r0, #0x4]
-	add r0, #0xc
-	stmia r0!, {r2-r3}
-	stmia r0!, {r1,r3}
-	bx lr
-	.balign 4
-
-	thumb_func_start MTX_RotY43_
-MTX_RotY43_: ; 0x020C4300
-	str r1, [r0, #0x18]
-	mov r3, #0x0
-	stmia r0!, {r2-r3}
-	neg r1, r1
-	stmia r0!, {r1,r3}
-	mov r1, #0x1
-	lsl r1, r1, #0xc
-	stmia r0!, {r1,r3}
-	add r0, #0x4
-	mov r1, #0x0
-	stmia r0!, {r1-r3}
-	stmia r0!, {r1,r3}
-	bx lr
-	.balign 4
-
-	arm_func_start MTX_MultVec43
-MTX_MultVec43: ; 0x020C431C
-	stmdb sp!, {r4-r6,lr}
-	ldr r12, [r0, #0x4]
-	ldr r4, [r1, #0xc]
-	ldr r3, [r0, #0x0]
-	smull r6, r5, r12, r4
-	ldr r4, [r1, #0x0]
-	ldr r0, [r0, #0x8]
-	smlal r6, r5, r3, r4
-	ldr r4, [r1, #0x18]
-	smlal r6, r5, r0, r4
-	mov r4, r6, lsr #0xc
-	orr r4, r4, r5, lsl #0x14
-	str r4, [r2, #0x0]
-	ldr r5, [r2, #0x0]
-	ldr r4, [r1, #0x24]
-	add r4, r5, r4
-	str r4, [r2, #0x0]
-	ldr r4, [r1, #0x10]
-	ldr r5, [r1, #0x4]
-	smull r6, lr, r12, r4
-	smlal r6, lr, r3, r5
-	ldr r4, [r1, #0x1c]
-	smlal r6, lr, r0, r4
-	mov r4, r6, lsr #0xc
-	orr r4, r4, lr, lsl #0x14
-	str r4, [r2, #0x4]
-	ldr r5, [r2, #0x4]
-	ldr r4, [r1, #0x28]
-	add r4, r5, r4
-	str r4, [r2, #0x4]
-	ldr lr, [r1, #0x14]
-	ldr r4, [r1, #0x8]
-	smull r5, lr, r12, lr
-	smlal r5, lr, r3, r4
-	ldr r3, [r1, #0x20]
-	smlal r5, lr, r0, r3
-	mov r0, r5, lsr #0xc
-	orr r0, r0, lr, lsl #0x14
-	str r0, [r2, #0x8]
-	ldr r3, [r2, #0x8]
-	ldr r0, [r1, #0x2c]
-	add r0, r3, r0
-	str r0, [r2, #0x8]
-	ldmia sp!, {r4-r6,lr}
-	bx lr
-
-	arm_func_start MTX_Concat43
-MTX_Concat43: ; 0x020C43D0
-	stmdb sp!, {r4-r10,lr}
-	sub sp, sp, #0x30
-	ldr r5, [r0, #0x4]
-	ldr r3, [r1, #0xc]
-	mov r12, r2
-	smull r8, r7, r5, r3
-	ldr r6, [r0, #0x0]
-	ldr r3, [r1, #0x0]
-	cmp r12, r1
-	smlal r8, r7, r6, r3
-	ldr r4, [r0, #0x8]
-	ldr r3, [r1, #0x18]
-	addeq r2, sp, #0x0
-	smlal r8, r7, r4, r3
-	mov r3, r8, lsr #0xc
-	orr r3, r3, r7, lsl #0x14
-	str r3, [r2, #0x0]
-	ldr r3, [r1, #0x10]
-	ldr r7, [r1, #0x4]
-	smull r9, r8, r5, r3
-	smlal r9, r8, r6, r7
-	ldr r3, [r1, #0x1c]
-	add lr, sp, #0x0
-	smlal r9, r8, r4, r3
-	mov r3, r9, lsr #0xc
-	orr r3, r3, r8, lsl #0x14
-	str r3, [r2, #0x4]
-	ldr r3, [r1, #0x14]
-	ldr r7, [r1, #0x8]
-	smull r9, r8, r5, r3
-	smlal r9, r8, r6, r7
-	ldr r6, [r1, #0x20]
-	smlal r9, r8, r4, r6
-	mov r4, r9, lsr #0xc
-	orr r4, r4, r8, lsl #0x14
-	str r4, [r2, #0x8]
-	ldr r4, [r0, #0x10]
-	ldr r5, [r0, #0xc]
-	smull r9, r8, r4, r3
-	smlal r9, r8, r5, r7
-	ldr r3, [r0, #0x14]
-	smlal r9, r8, r3, r6
-	mov r6, r9, lsr #0xc
-	orr r6, r6, r8, lsl #0x14
-	str r6, [r2, #0x14]
-	ldr r6, [r1, #0x10]
-	ldr r7, [r1, #0x4]
-	smull r9, r8, r4, r6
-	smlal r9, r8, r5, r7
-	ldr r6, [r1, #0x1c]
-	smlal r9, r8, r3, r6
-	mov r6, r9, lsr #0xc
-	orr r6, r6, r8, lsl #0x14
-	str r6, [r2, #0x10]
-	ldr r7, [r1, #0xc]
-	ldr r8, [r1, #0x0]
-	smull r10, r9, r4, r7
-	smlal r10, r9, r5, r8
-	ldr r6, [r1, #0x18]
-	smlal r10, r9, r3, r6
-	mov r3, r10, lsr #0xc
-	orr r3, r3, r9, lsl #0x14
-	str r3, [r2, #0xc]
-	ldr r4, [r0, #0x1c]
-	ldr r5, [r0, #0x18]
-	smull r9, r7, r4, r7
-	smlal r9, r7, r5, r8
-	ldr r3, [r0, #0x20]
-	smlal r9, r7, r3, r6
-	mov r6, r9, lsr #0xc
-	orr r6, r6, r7, lsl #0x14
-	str r6, [r2, #0x18]
-	ldr r6, [r1, #0x10]
-	ldr r7, [r1, #0x4]
-	smull r9, r6, r4, r6
-	smlal r9, r6, r5, r7
-	ldr r8, [r1, #0x1c]
-	smlal r9, r6, r3, r8
-	mov r7, r9, lsr #0xc
-	orr r7, r7, r6, lsl #0x14
-	str r7, [r2, #0x1c]
-	ldr r7, [r1, #0x14]
-	ldr r6, [r1, #0x8]
-	smull r9, r8, r4, r7
-	smlal r9, r8, r5, r6
-	ldr r5, [r1, #0x20]
-	smlal r9, r8, r3, r5
-	mov r3, r9, lsr #0xc
-	orr r3, r3, r8, lsl #0x14
-	str r3, [r2, #0x20]
-	ldr r4, [r0, #0x28]
-	ldr r3, [r0, #0x24]
-	smull r8, r7, r4, r7
-	smlal r8, r7, r3, r6
-	ldr r0, [r0, #0x2c]
-	ldr r6, [r1, #0x2c]
-	smlal r8, r7, r0, r5
-	mov r5, r8, lsr #0xc
-	orr r5, r5, r7, lsl #0x14
-	adds r5, r6, r5
-	str r5, [r2, #0x2c]
-	ldr r5, [r1, #0x10]
-	ldr r6, [r1, #0x4]
-	smull r8, r5, r4, r5
-	smlal r8, r5, r3, r6
-	ldr r7, [r1, #0x1c]
-	ldr r9, [r1, #0x28]
-	smlal r8, r5, r0, r7
-	mov r6, r8, lsr #0xc
-	orr r6, r6, r5, lsl #0x14
-	adds r5, r9, r6
-	str r5, [r2, #0x28]
-	ldr r8, [r1, #0x24]
-	ldr r6, [r1, #0x18]
-	ldr r5, [r1, #0x0]
-	ldr r1, [r1, #0xc]
-	smull r7, r1, r4, r1
-	smlal r7, r1, r3, r5
-	smlal r7, r1, r0, r6
-	mov r0, r7, lsr #0xc
-	orr r0, r0, r1, lsl #0x14
-	adds r0, r8, r0
-	cmp r2, lr
-	str r0, [r2, #0x24]
-	addne sp, sp, #0x30
-	ldmneia sp!, {r4-r10,lr}
-	bxne lr
-	ldmia lr!, {r0-r3}
-	stmia r12!, {r0-r3}
-	ldmia lr!, {r0-r3}
-	stmia r12!, {r0-r3}
-	ldmia lr, {r0-r3}
-	stmia r12, {r0-r3}
-	add sp, sp, #0x30
-	ldmia sp!, {r4-r10,lr}
-	bx lr
-
-	arm_func_start MTX_Inverse43
-MTX_Inverse43: ; 0x020C45F0
-	stmdb sp!, {r4-r11,lr}
-	sub sp, sp, #0x5c
-	mov r10, r0
-	ldr r11, [r10, #0x14]
-	cmp r10, r1
-	addeq r9, sp, #0x2c
-	ldr r3, [r10, #0x18]
-	ldr r2, [r10, #0xc]
-	ldr r8, [r10, #0x20]
-	str r1, [sp, #0x0]
-	movne r9, r1
-	smull r7, r5, r2, r8
-	smull r6, r1, r11, r3
-	ldr r4, [r10, #0x1c]
-	subs r12, r7, r6
-	ldr r0, [r10, #0x10]
-	smull r6, lr, r11, r4
-	sbc r11, r5, r1
-	smull r5, r4, r2, r4
-	mov r1, #0x800
-	smull r8, r7, r0, r8
-	smull r3, r2, r0, r3
-	adds r12, r12, r1
-	adc r0, r11, #0x0
-	subs r8, r8, r6
-	sbc r6, r7, lr
-	adds r7, r8, r1
-	adc r11, r6, #0x0
-	mov r7, r7, lsr #0xc
-	mov r8, r12, lsr #0xc
-	orr r7, r7, r11, lsl #0x14
-	orr r8, r8, r0, lsl #0x14
-	subs r0, r5, r3
-	sbc r12, r4, r2
-	ldr r6, [r10, #0x4]
-	ldr r11, [r10, #0x0]
-	smull r3, r2, r6, r8
-	adds r0, r0, r1
-	smull r5, r4, r11, r7
-	adc r11, r12, #0x0
-	mov r6, r0, lsr #0xc
-	subs r3, r5, r3
-	ldr r0, [r10, #0x8]
-	orr r6, r6, r11, lsl #0x14
-	sbc r2, r4, r2
-	smlal r3, r2, r0, r6
-	adds r0, r3, r1
-	mov r1, r8, asr #0x1f
-	str r1, [sp, #0x4]
-	mov r1, r7, asr #0x1f
-	str r1, [sp, #0x8]
-	mov r1, r6, asr #0x1f
-	adc r2, r2, #0x0
-	mov r0, r0, lsr #0xc
-	orrs r0, r0, r2, lsl #0x14
-	str r1, [sp, #0xc]
-	addeq sp, sp, #0x5c
-	mvneq r0, #0x0
-	ldmeqia sp!, {r4-r11,lr}
-	bxeq lr
-	bl FX_InvAsync
-	ldr r1, [r10, #0x8]
-	ldr r2, [r10, #0x1c]
-	ldr r11, [r10, #0x10]
-	smull r0, r4, r2, r1
-	smull r3, r2, r11, r1
-	ldr r5, [r10, #0x18]
-	str r2, [sp, #0x18]
-	str r3, [sp, #0x14]
-	smull r3, r2, r5, r1
-	str r2, [sp, #0x20]
-	ldr r2, [r10, #0x20]
-	ldr lr, [r10, #0x4]
-	str r2, [sp, #0x10]
-	ldr r5, [sp, #0x10]
-	str r3, [sp, #0x1c]
-	smull r12, r5, lr, r5
-	subs r0, r12, r0
-	ldr r2, [r10, #0xc]
-	sbc r4, r5, r4
-	smull r1, r5, r2, r1
-	str r5, [sp, #0x28]
-	mov r5, r0, lsr #0xc
-	ldr r11, [r10, #0x14]
-	orr r5, r5, r4, lsl #0x14
-	smull r4, r2, lr, r11
-	ldr r0, [sp, #0x14]
-	ldr r3, [r10, #0x0]
-	subs r4, r4, r0
-	ldr r0, [sp, #0x18]
-	mov r4, r4, lsr #0xc
-	sbc r0, r2, r0
-	orr r4, r4, r0, lsl #0x14
-	ldr r0, [sp, #0x10]
-	ldr r2, [sp, #0x1c]
-	smull r12, r0, r3, r0
-	subs r2, r12, r2
-	ldr r12, [sp, #0x20]
-	sbc r0, r0, r12
-	smull r12, r11, r3, r11
-	ldr r3, [sp, #0x28]
-	subs r1, r12, r1
-	sbc r3, r11, r3
-	mov r11, r2, lsr #0xc
-	orr r11, r11, r0, lsl #0x14
-	mov r0, r1, lsr #0xc
-	str r0, [sp, #0x24]
-	orr r0, r0, r3, lsl #0x14
-	str r0, [sp, #0x24]
-	bl FX_GetDivResult
-	smull r2, r1, r0, r5
-	mov r2, r2, lsr #0xc
-	orr r2, r2, r1, lsl #0x14
-	rsb lr, r2, #0x0
-	smull r2, r1, r0, r4
-	mov r4, r2, lsr #0xc
-	orr r4, r4, r1, lsl #0x14
-	smull r2, r1, r0, r11
-	mov r3, r2, lsr #0xc
-	orr r3, r3, r1, lsl #0x14
-	ldr r1, [sp, #0x24]
-	umull r11, r5, r0, r7
-	smull r2, r1, r0, r1
-	mov r2, r2, lsr #0xc
-	orr r2, r2, r1, lsl #0x14
-	ldr r1, [sp, #0x8]
-	mov r12, r0, asr #0x1f
-	mla r5, r0, r1, r5
-	mla r5, r12, r7, r5
-	mov r1, r11, lsr #0xc
-	orr r1, r1, r5, lsl #0x14
-	str r1, [r9, #0x0]
-	str lr, [r9, #0x4]
-	str r4, [r9, #0x8]
-	ldr r1, [sp, #0x4]
-	umull r5, r4, r0, r8
-	mla r4, r0, r1, r4
-	mla r4, r12, r8, r4
-	mov r1, r5, lsr #0xc
-	orr r1, r1, r4, lsl #0x14
-	rsb r1, r1, #0x0
-	str r1, [r9, #0xc]
-	ldr r1, [sp, #0xc]
-	rsb r2, r2, #0x0
-	str r3, [r9, #0x10]
-	str r2, [r9, #0x14]
-	umull r3, r2, r0, r6
-	mla r2, r0, r1, r2
-	mla r2, r12, r6, r2
-	mov r1, r3, lsr #0xc
-	orr r1, r1, r2, lsl #0x14
-	str r1, [r9, #0x18]
-	ldr r3, [r10, #0x0]
-	ldr r1, [r10, #0x1c]
-	ldr r2, [r10, #0x18]
-	smull r5, r4, r3, r1
-	ldr r1, [r10, #0x4]
-	smull r3, r1, r2, r1
-	subs r2, r5, r3
-	sbc r1, r4, r1
-	mov r2, r2, lsr #0xc
-	orr r2, r2, r1, lsl #0x14
-	smull r2, r1, r0, r2
-	mov r2, r2, lsr #0xc
-	orr r2, r2, r1, lsl #0x14
-	rsb r1, r2, #0x0
-	str r1, [r9, #0x1c]
-	ldr r3, [r10, #0x0]
-	ldr r1, [r10, #0x10]
-	ldr r2, [r10, #0xc]
-	smull r5, r4, r3, r1
-	ldr r1, [r10, #0x4]
-	smull r3, r1, r2, r1
-	subs r2, r5, r3
-	sbc r1, r4, r1
-	mov r2, r2, lsr #0xc
-	orr r2, r2, r1, lsl #0x14
-	smull r2, r1, r0, r2
-	mov r0, r2, lsr #0xc
-	orr r0, r0, r1, lsl #0x14
-	str r0, [r9, #0x20]
-	add r0, sp, #0x2c
-	ldr r2, [r9, #0xc]
-	ldr r1, [r10, #0x28]
-	ldr r4, [r9, #0x0]
-	smull r7, r1, r2, r1
-	ldr r3, [r10, #0x24]
-	ldr r6, [r9, #0x18]
-	smlal r7, r1, r4, r3
-	ldr r5, [r10, #0x2c]
-	cmp r9, r0
-	smlal r7, r1, r6, r5
-	mov r2, r7, lsr #0xc
-	orr r2, r2, r1, lsl #0x14
-	rsb r1, r2, #0x0
-	str r1, [r9, #0x24]
-	ldr r2, [r9, #0x10]
-	ldr r1, [r10, #0x28]
-	ldr r4, [r9, #0x4]
-	smull r7, r1, r2, r1
-	ldr r3, [r10, #0x24]
-	ldr r6, [r9, #0x1c]
-	smlal r7, r1, r4, r3
-	ldr r5, [r10, #0x2c]
-	smlal r7, r1, r6, r5
-	mov r2, r7, lsr #0xc
-	orr r2, r2, r1, lsl #0x14
-	rsb r1, r2, #0x0
-	str r1, [r9, #0x28]
-	ldr r1, [r10, #0x28]
-	ldr r2, [r9, #0x14]
-	ldr r4, [r9, #0x8]
-	smull r7, r1, r2, r1
-	ldr r3, [r10, #0x24]
-	ldr r6, [r9, #0x20]
-	smlal r7, r1, r4, r3
-	ldr r5, [r10, #0x2c]
-	smlal r7, r1, r6, r5
-	mov r2, r7, lsr #0xc
-	orr r2, r2, r1, lsl #0x14
-	rsb r1, r2, #0x0
-	str r1, [r9, #0x2c]
-	bne _020C4974
-	ldr r1, [sp, #0x0]
-	bl MI_Copy48B
-_020C4974:
-	mov r0, #0x0
-	add sp, sp, #0x5c
-	ldmia sp!, {r4-r11,lr}
-	bx lr
-
-	arm_func_start MTX_ScaleApply43
-MTX_ScaleApply43: ; 0x020C4984
-	stmdb sp!, {r4-r5,lr}
-	sub sp, sp, #0x4
-	ldr r12, [sp, #0x10]
-	mov r5, r0
-	str r12, [sp, #0x0]
-	mov r4, r1
-	bl MTX_ScaleApply33
-	ldr r0, [r5, #0x24]
-	str r0, [r4, #0x24]
-	ldr r0, [r5, #0x28]
-	str r0, [r4, #0x28]
-	ldr r0, [r5, #0x2c]
-	str r0, [r4, #0x2c]
-	add sp, sp, #0x4
-	ldmia sp!, {r4-r5,lr}
-	bx lr
diff --git a/arm9/lib/src/FX_mtx43.c b/arm9/lib/src/FX_mtx43.c
new file mode 100644
index 00000000..c5d227b8
--- /dev/null
+++ b/arm9/lib/src/FX_mtx43.c
@@ -0,0 +1,207 @@
+#include "global.h"
+#include "main.h"
+#include "fx.h"
+
+
+
+
+
+void MTX_ScaleApply43(struct Mtx43 *mtx, struct Mtx43 *dst, s32 x, s32 y, s32 z){
+    //this works because matrices are indexed columns first
+    MTX_ScaleApply33((struct Mtx33 *)mtx, (struct Mtx33 *)dst, x, y, z);
+    dst->_[9] = mtx->_[9];
+    dst->_[10] = mtx->_[10];
+    dst->_[11] = mtx->_[11];
+}
+
+s32 MTX_Inverse43(struct Mtx43 *mtx, struct Mtx43 *inv){
+    struct Mtx43 tempmat;
+    struct Mtx43 *dst;
+    s32 det0, det1, det2, det;
+    s32 var0, var1, var2, var3;
+    if (mtx == inv)
+        dst = &tempmat;
+    else
+        dst = inv;
+    //subdeterminants
+    det0 = ((s64)mtx->_[4] * mtx->_[8] - (s64)mtx->_[5] * mtx->_[7] + (s64)(1 << (FX32_INT_SHIFT - 1))) >> FX32_INT_SHIFT;
+    det1 = ((s64)mtx->_[3] * mtx->_[8] - (s64)mtx->_[5] * mtx->_[6] + (s64)(1 << (FX32_INT_SHIFT - 1))) >> FX32_INT_SHIFT;
+    det2 = ((s64)mtx->_[3] * mtx->_[7] - (s64)mtx->_[4] * mtx->_[6] + (s64)(1 << (FX32_INT_SHIFT - 1))) >> FX32_INT_SHIFT;
+    //matrix determinant
+    det = ((s64)mtx->_[0] * det0 - (s64)mtx->_[1] * det1 + (s64)mtx->_[2] * det2 + (s64)(1 << (FX32_INT_SHIFT - 1))) >> FX32_INT_SHIFT;
+
+    if (det == 0)
+        return -1; //not invertible
+
+    FX_InvAsync(det);
+
+    var0 = ((s64)mtx->_[1] * mtx->_[8] - (s64)mtx->_[7] * mtx->_[2]) >> FX32_INT_SHIFT;
+    var1 = ((s64)mtx->_[1] * mtx->_[5] - (s64)mtx->_[4] * mtx->_[2]) >> FX32_INT_SHIFT;
+    var2 = ((s64)mtx->_[0] * mtx->_[8] - (s64)mtx->_[6] * mtx->_[2]) >> FX32_INT_SHIFT;
+    var3 = ((s64)mtx->_[0] * mtx->_[5] - (s64)mtx->_[3] * mtx->_[2]) >> FX32_INT_SHIFT;
+
+    s32 ret = FX_GetDivResult();
+    dst->_[0] =  (s32)(((s64)ret * det0) >> FX32_INT_SHIFT);
+    dst->_[1] = -(s32)(((s64)ret * var0) >> FX32_INT_SHIFT);
+    dst->_[2] =  (s32)(((s64)ret * var1) >> FX32_INT_SHIFT);
+    dst->_[3] = -(s32)(((s64)ret * det1) >> FX32_INT_SHIFT);
+    dst->_[4] =  (s32)(((s64)ret * var2) >> FX32_INT_SHIFT);
+    dst->_[5] = -(s32)(((s64)ret * var3) >> FX32_INT_SHIFT);
+
+    dst->_[6] = (s32)(((s64)ret * det2) >> FX32_INT_SHIFT);
+    s32 temp = (s32)(((s64)mtx->_[0] * mtx->_[7] - (s64)mtx->_[6] * mtx->_[1]) >> FX32_INT_SHIFT);
+    dst->_[7] = -(s32)(((s64)ret * temp) >> FX32_INT_SHIFT);
+    s32 temp1 = (s32)(((s64)mtx->_[0] * mtx->_[4] - (s64)mtx->_[3] * mtx->_[1]) >> FX32_INT_SHIFT);
+    dst->_[8] = (s32)(((s64)ret * temp1) >> FX32_INT_SHIFT);
+    dst->_[9] =  -(s32)(((s64)dst->_[0] * mtx->_[9] + (s64)dst->_[3] * mtx->_[10] + (s64)dst->_[6] * mtx->_[11]) >> FX32_INT_SHIFT);
+    dst->_[10] = -(s32)(((s64)dst->_[1] * mtx->_[9] + (s64)dst->_[4] * mtx->_[10] + (s64)dst->_[7] * mtx->_[11]) >> FX32_INT_SHIFT);
+    dst->_[11] = -(s32)(((s64)dst->_[2] * mtx->_[9] + (s64)dst->_[5] * mtx->_[10] + (s64)dst->_[8] * mtx->_[11]) >> FX32_INT_SHIFT);
+
+    if (dst == &tempmat)
+        MI_Copy48B(&tempmat, inv);
+    return 0;
+}
+
+void MTX_Concat43(struct Mtx43 *a, struct Mtx43 *b, struct Mtx43 *c){
+    struct Mtx43 temp;
+    struct Mtx43 *dst;
+    s32 a0, a1, a2;
+    s32 b0, b1, b2;
+
+    if (c == b)
+        dst = &temp;
+    else
+        dst = c;
+
+    a0 = a->_[0];
+    a1 = a->_[1];
+    a2 = a->_[2];
+    dst->_[0] =  (((s64)a0  * b->_[0] + (s64)a1  * b->_[3] + (s64)a2  * b->_[6] ) >> FX32_INT_SHIFT);
+    dst->_[1] =  (((s64)a0  * b->_[1] + (s64)a1  * b->_[4] + (s64)a2  * b->_[7] ) >> FX32_INT_SHIFT);
+    b0 = b->_[2];
+    b1 = b->_[5];
+    b2 = b->_[8];
+    dst->_[2] =  (((s64)a0  * b0 + (s64)a1  * b1 + (s64)a2  * b2) >> FX32_INT_SHIFT);
+    a0 = a->_[3];
+    a1 = a->_[4];
+    a2 = a->_[5];
+    dst->_[5] =  (((s64)a0  * b0 + (s64)a1  * b1 + (s64)a2  * b2) >> FX32_INT_SHIFT);
+    dst->_[4] =  (((s64)a0  * b->_[1] + (s64)a1  * b->_[4] + (s64)a2  * b->_[7] ) >> FX32_INT_SHIFT);
+    b0 = b->_[0];
+    b1 = b->_[3];
+    b2 = b->_[6];
+    dst->_[3] =  (((s64)a0  * b0 + (s64)a1  * b1 + (s64)a2  * b2) >> FX32_INT_SHIFT);
+    a0 = a->_[6];
+    a1 = a->_[7];
+    a2 = a->_[8];
+    dst->_[6] =  (((s64)a0  * b0 + (s64)a1  * b1 + (s64)a2 * b2) >> FX32_INT_SHIFT);
+    dst->_[7] =  (((s64)a0  * b->_[1] + (s64)a1  * b->_[4] + (s64)a2 * b->_[7] ) >> FX32_INT_SHIFT);
+    b0 = b->_[2];
+    b1 = b->_[5];
+    b2 = b->_[8];
+    dst->_[8] = (((s64)a0  * b0 + (s64)a1  * b1 + (s64)a2 * b2) >> FX32_INT_SHIFT);
+    a0 = a->_[9];
+    a1 = a->_[10];
+    a2 = a->_[11];
+    dst->_[11] = ((((s64)a0 * b0 + (s64)a1 * b1 + (s64)a2 * b2) >> FX32_INT_SHIFT) + b->_[11]);
+    dst->_[10] = ((((s64)a0 * b->_[1] + (s64)a1 * b->_[4] + (s64)a2 * b->_[7]) >> FX32_INT_SHIFT) + b->_[10]);
+    dst->_[9] = ((((s64)a0 * b->_[0] + (s64)a1 * b->_[3] + (s64)a2 * b->_[6]) >> FX32_INT_SHIFT) + b->_[9]);
+    if (dst == &temp)
+        *c = temp;
+}
+
+void MTX_MultVec43(struct Vecx32 *vec, struct Mtx43 *mtx, struct Vecx32 *dst){
+    s32 x, y, z;
+    x = vec->x;
+    y = vec->y;
+    z = vec->z;
+    dst->x = ((s64)x * mtx->_[0] + (s64)y * mtx->_[3] + (s64)z * mtx->_[6]) >> FX32_INT_SHIFT;
+    dst->x += mtx->_[9];
+    dst->y = ((s64)x * mtx->_[1] + (s64)y * mtx->_[4] + (s64)z * mtx->_[7]) >> FX32_INT_SHIFT;
+    dst->y += mtx->_[10];
+    dst->z = ((s64)x * mtx->_[2] + (s64)y * mtx->_[5] + (s64)z * mtx->_[8]) >> FX32_INT_SHIFT;
+    dst->z += mtx->_[11];
+}
+
+asm void MTX_Identity43_(struct Mtx43 *mtx){
+    mov r2, #0x1000
+    mov r3, #0x0
+    stmia r0!, {r2-r3}
+    mov r1, #0x0
+    stmia r0!, {r1,r3}
+    stmia r0!, {r2-r3}
+    stmia r0!, {r1,r3}
+    stmia r0!, {r2-r3}
+    stmia r0!, {r1,r3}
+    bx lr
+}
+
+asm void MTX_Copy43To44_(struct Mtx43 *src, struct Mtx44 *dst){
+    stmdb sp!, {r4}
+    mov r12, #0x0
+    ldmia r0!, {r2-r4}
+    stmia r1!, {r2-r4,r12}
+    ldmia r0!, {r2-r4}
+    stmia r1!, {r2-r4,r12}
+    ldmia r0!, {r2-r4}
+    stmia r1!, {r2-r4,r12}
+    mov r12, #0x1000
+    ldmia r0!, {r2-r4}
+    stmia r1!, {r2-r4,r12}
+    ldmia sp!, {r4}
+    bx lr
+}
+
+#pragma thumb on
+asm void MTX_Scale43_(struct Mtx43 *dst, s32 x, s32 y, s32 z){
+    stmia r0!, {r1}
+    mov r1, #0x0
+    str r3, [r0, #0x1c]
+    mov r3, #0x0
+    stmia r0!, {r1,r3}
+    stmia r0!, {r1-r3}
+    mov r2, #0x0
+    stmia r0!, {r1,r3}
+    add r0, #0x4
+    stmia r0!, {r1-r3}
+    bx lr
+}
+#pragma thumb off
+
+#pragma thumb on
+asm void MTX_RotX43_(struct Mtx43 *mtx, s32 sinphi, s32 cosphi){
+    str r1, [r0, #0x14]
+	neg r1, r1
+	str r1, [r0, #0x1c]
+	mov r1, #0x1
+	lsl r1, r1, #0xc
+	stmia r0!, {r1}
+	mov r3, #0x0
+	mov r1, #0x0
+	stmia r0!, {r1,r3}
+	stmia r0!, {r1-r2}
+	str r1, [r0, #0x4]
+	add r0, #0xc
+	stmia r0!, {r2-r3}
+	stmia r0!, {r1,r3}
+	bx lr
+}
+#pragma thumb off
+
+#pragma thumb on
+asm void MTX_RotY43_(struct Mtx43 *mtx, s32 sinphi, s32 cosphi){
+    str r1, [r0, #0x18]
+	mov r3, #0x0
+	stmia r0!, {r2-r3}
+	neg r1, r1
+	stmia r0!, {r1,r3}
+	mov r1, #0x1
+	lsl r1, r1, #0xc
+	stmia r0!, {r1,r3}
+	add r0, #0x4
+	mov r1, #0x0
+	stmia r0!, {r1-r3}
+	stmia r0!, {r1,r3}
+	bx lr
+}
+#pragma thumb off
diff --git a/include/fx.h b/include/fx.h
index 1e9eb291..ee436883 100644
--- a/include/fx.h
+++ b/include/fx.h
@@ -73,6 +73,7 @@ struct Vecx16
     s16 z;
 };
 
+//Matrices are indexed as [column][row]
 struct Mtx44
 {
     s32 _[16];
@@ -138,5 +139,16 @@ void MTX_RotX44_(struct Mtx44 *mtx, s32 sinphi, s32 cosphi);
 void MTX_RotY44_(struct Mtx44 *mtx, s32 sinphi, s32 cosphi);
 void MTX_RotZ44_(struct Mtx44 *mtx, s32 sinphi, s32 cosphi);
 
+//Mtx43
+void MTX_ScaleApply33(struct Mtx33 *mtx, struct Mtx33 *dst, s32 x, s32 y, s32 z);
+s32 MTX_Inverse43(struct Mtx43 *mtx, struct Mtx43 *inv);
+void MTX_Concat43(struct Mtx43 *a, struct Mtx43 *b, struct Mtx43 *c);
+void MTX_MultVec43(struct Vecx32 *vec, struct Mtx43 *mtx, struct Vecx32 *dst);
+void MTX_Identity43_(struct Mtx43 *mtx);
+void MTX_Copy43To44_(struct Mtx43 *src, struct Mtx44 *dst);
+void MTX_Scale43_(struct Mtx43 *dst, s32 x, s32 y, s32 z);
+void MTX_RotX43_(struct Mtx43 *mtx, s32 sinphi, s32 cosphi);
+void MTX_RotY43_(struct Mtx43 *mtx, s32 sinphi, s32 cosphi);
+
 
 #endif //GUARD_FX_H
-- 
cgit v1.2.3


From 8edaffd4aaddac79b3d15148783a74fa508c3edc Mon Sep 17 00:00:00 2001
From: Made <made111@gmx.de>
Date: Mon, 4 May 2020 16:36:36 +0200
Subject: Decompile FX_mtx33.s

---
 arm9/asm/FX_mtx33.s     | 259 ------------------------------------------------
 arm9/lib/src/FX_mtx33.c | 135 +++++++++++++++++++++++++
 arm9/lib/src/FX_mtx43.c |   3 -
 include/fx.h            |  13 ++-
 4 files changed, 147 insertions(+), 263 deletions(-)
 delete mode 100644 arm9/asm/FX_mtx33.s
 create mode 100644 arm9/lib/src/FX_mtx33.c

diff --git a/arm9/asm/FX_mtx33.s b/arm9/asm/FX_mtx33.s
deleted file mode 100644
index 2b31fd6c..00000000
--- a/arm9/asm/FX_mtx33.s
+++ /dev/null
@@ -1,259 +0,0 @@
-	.include "asm/macros.inc"
-	.include "global.inc"
-
-	.text
-
-	arm_func_start MTX_Identity33_
-MTX_Identity33_: ; 0x020C3F14
-	mov r2, #0x1000
-	str r2, [r0, #0x20]
-	mov r3, #0x0
-	stmia r0!, {r2-r3}
-	mov r1, #0x0
-	stmia r0!, {r1,r3}
-	stmia r0!, {r2-r3}
-	stmia r0!, {r1,r3}
-	bx lr
-
-	thumb_func_start MTX_RotX33_
-MTX_RotX33_: ; 0x020C3F38
-	mov r3, #0x1
-	lsl r3, r3, #0xc
-	str r3, [r0, #0x0]
-	mov r3, #0x0
-	str r3, [r0, #0x4]
-	str r3, [r0, #0x8]
-	str r3, [r0, #0xc]
-	str r2, [r0, #0x10]
-	str r1, [r0, #0x14]
-	str r3, [r0, #0x18]
-	neg r1, r1
-	str r1, [r0, #0x1c]
-	str r2, [r0, #0x20]
-	bx lr
-
-	thumb_func_start MTX_RotY33_
-MTX_RotY33_: ; 0x020C3F54
-	str r2, [r0, #0x0]
-	str r2, [r0, #0x20]
-	mov r3, #0x0
-	str r3, [r0, #0x4]
-	str r3, [r0, #0xc]
-	str r3, [r0, #0x14]
-	str r3, [r0, #0x1c]
-	neg r2, r1
-	mov r3, #0x1
-	lsl r3, r3, #0xc
-	str r1, [r0, #0x18]
-	str r2, [r0, #0x8]
-	str r3, [r0, #0x10]
-	bx lr
-
-	thumb_func_start MTX_RotZ33_
-MTX_RotZ33_: ; 0x020C3F70
-	stmia r0!, {r2}
-	mov r3, #0x0
-	stmia r0!, {r1,r3}
-	neg r1, r1
-	stmia r0!, {r1-r2}
-	mov r1, #0x1
-	lsl r1, r1, #0xc
-	str r3, [r0, #0x0]
-	str r3, [r0, #0x4]
-	str r3, [r0, #0x8]
-	str r1, [r0, #0xc]
-	bx lr
-
-	arm_func_start MTX_MultVec33
-MTX_MultVec33: ; 0x020C3F88
-	stmdb sp!, {r4-r6,lr}
-	ldr r12, [r0, #0x4]
-	ldr r4, [r1, #0xc]
-	ldr r3, [r0, #0x0]
-	smull r6, r5, r12, r4
-	ldr r4, [r1, #0x0]
-	ldr r0, [r0, #0x8]
-	smlal r6, r5, r3, r4
-	ldr r4, [r1, #0x18]
-	smlal r6, r5, r0, r4
-	mov r4, r6, lsr #0xc
-	orr r4, r4, r5, lsl #0x14
-	str r4, [r2, #0x0]
-	ldr r4, [r1, #0x10]
-	ldr r5, [r1, #0x4]
-	smull r6, lr, r12, r4
-	smlal r6, lr, r3, r5
-	ldr r4, [r1, #0x1c]
-	smlal r6, lr, r0, r4
-	mov r4, r6, lsr #0xc
-	orr r4, r4, lr, lsl #0x14
-	str r4, [r2, #0x4]
-	ldr lr, [r1, #0x14]
-	ldr r4, [r1, #0x8]
-	smull r5, lr, r12, lr
-	smlal r5, lr, r3, r4
-	ldr r1, [r1, #0x20]
-	smlal r5, lr, r0, r1
-	mov r0, r5, lsr #0xc
-	orr r0, r0, lr, lsl #0x14
-	str r0, [r2, #0x8]
-	ldmia sp!, {r4-r6,lr}
-	bx lr
-
-	arm_func_start MTX_Concat33
-MTX_Concat33: ; 0x020C400C
-	stmdb sp!, {r4-r9,lr}
-	sub sp, sp, #0x24
-	ldr r5, [r0, #0x4]
-	ldr r3, [r1, #0xc]
-	mov r12, r2
-	smull r8, r7, r5, r3
-	ldr r6, [r0, #0x0]
-	ldr r3, [r1, #0x0]
-	cmp r12, r1
-	smlal r8, r7, r6, r3
-	ldr r4, [r0, #0x8]
-	ldr r3, [r1, #0x18]
-	addeq r2, sp, #0x0
-	smlal r8, r7, r4, r3
-	mov r3, r8, lsr #0xc
-	orr r3, r3, r7, lsl #0x14
-	str r3, [r2, #0x0]
-	ldr r3, [r1, #0x10]
-	ldr r7, [r1, #0x4]
-	smull r9, r8, r5, r3
-	smlal r9, r8, r6, r7
-	ldr r3, [r1, #0x1c]
-	add lr, sp, #0x0
-	smlal r9, r8, r4, r3
-	mov r3, r9, lsr #0xc
-	orr r3, r3, r8, lsl #0x14
-	str r3, [r2, #0x4]
-	ldr r3, [r1, #0x14]
-	ldr r7, [r1, #0x8]
-	smull r9, r8, r5, r3
-	smlal r9, r8, r6, r7
-	ldr r6, [r1, #0x20]
-	cmp r2, lr
-	smlal r9, r8, r4, r6
-	mov r4, r9, lsr #0xc
-	orr r4, r4, r8, lsl #0x14
-	str r4, [r2, #0x8]
-	ldr r4, [r0, #0x10]
-	ldr r5, [r0, #0xc]
-	smull r9, r8, r4, r3
-	smlal r9, r8, r5, r7
-	ldr r3, [r0, #0x14]
-	addne sp, sp, #0x24
-	smlal r9, r8, r3, r6
-	mov r6, r9, lsr #0xc
-	orr r6, r6, r8, lsl #0x14
-	str r6, [r2, #0x14]
-	ldr r6, [r1, #0x10]
-	ldr r7, [r1, #0x4]
-	smull r9, r8, r4, r6
-	smlal r9, r8, r5, r7
-	ldr r6, [r1, #0x1c]
-	smlal r9, r8, r3, r6
-	mov r6, r9, lsr #0xc
-	orr r6, r6, r8, lsl #0x14
-	str r6, [r2, #0x10]
-	ldr r7, [r1, #0xc]
-	ldr r6, [r1, #0x0]
-	smull r9, r8, r4, r7
-	smlal r9, r8, r5, r6
-	ldr r5, [r1, #0x18]
-	smlal r9, r8, r3, r5
-	mov r3, r9, lsr #0xc
-	orr r3, r3, r8, lsl #0x14
-	str r3, [r2, #0xc]
-	ldr r4, [r0, #0x1c]
-	ldr r3, [r0, #0x18]
-	smull r8, r7, r4, r7
-	smlal r8, r7, r3, r6
-	ldr r0, [r0, #0x20]
-	smlal r8, r7, r0, r5
-	mov r5, r8, lsr #0xc
-	orr r5, r5, r7, lsl #0x14
-	str r5, [r2, #0x18]
-	ldr r5, [r1, #0x10]
-	ldr r6, [r1, #0x4]
-	smull r8, r5, r4, r5
-	smlal r8, r5, r3, r6
-	ldr r7, [r1, #0x1c]
-	smlal r8, r5, r0, r7
-	mov r6, r8, lsr #0xc
-	orr r6, r6, r5, lsl #0x14
-	str r6, [r2, #0x1c]
-	ldr r6, [r1, #0x20]
-	ldr r5, [r1, #0x8]
-	ldr r1, [r1, #0x14]
-	smull r7, r1, r4, r1
-	smlal r7, r1, r3, r5
-	smlal r7, r1, r0, r6
-	mov r0, r7, lsr #0xc
-	orr r0, r0, r1, lsl #0x14
-	str r0, [r2, #0x20]
-	ldmneia sp!, {r4-r9,lr}
-	bxne lr
-	ldmia lr!, {r0-r3}
-	stmia r12!, {r0-r3}
-	ldmia lr!, {r0-r3}
-	stmia r12!, {r0-r3}
-	ldr r0, [lr, #0x0]
-	str r0, [r12, #0x0]
-	add sp, sp, #0x24
-	ldmia sp!, {r4-r9,lr}
-	bx lr
-
-	arm_func_start MTX_ScaleApply33
-MTX_ScaleApply33: ; 0x020C41A8
-	stmdb sp!, {r4,lr}
-	ldr r4, [r0, #0x0]
-	ldr r12, [sp, #0x8]
-	smull lr, r4, r2, r4
-	mov lr, lr, lsr #0xc
-	orr lr, lr, r4, lsl #0x14
-	str lr, [r1, #0x0]
-	ldr r4, [r0, #0x4]
-	smull lr, r4, r2, r4
-	mov lr, lr, lsr #0xc
-	orr lr, lr, r4, lsl #0x14
-	str lr, [r1, #0x4]
-	ldr lr, [r0, #0x8]
-	smull r4, lr, r2, lr
-	mov r2, r4, lsr #0xc
-	orr r2, r2, lr, lsl #0x14
-	str r2, [r1, #0x8]
-	ldr r2, [r0, #0xc]
-	smull lr, r2, r3, r2
-	mov lr, lr, lsr #0xc
-	orr lr, lr, r2, lsl #0x14
-	str lr, [r1, #0xc]
-	ldr r2, [r0, #0x10]
-	smull lr, r2, r3, r2
-	mov lr, lr, lsr #0xc
-	orr lr, lr, r2, lsl #0x14
-	str lr, [r1, #0x10]
-	ldr r2, [r0, #0x14]
-	smull lr, r2, r3, r2
-	mov r3, lr, lsr #0xc
-	orr r3, r3, r2, lsl #0x14
-	str r3, [r1, #0x14]
-	ldr r2, [r0, #0x18]
-	smull r3, r2, r12, r2
-	mov r3, r3, lsr #0xc
-	orr r3, r3, r2, lsl #0x14
-	str r3, [r1, #0x18]
-	ldr r2, [r0, #0x1c]
-	smull r3, r2, r12, r2
-	mov r3, r3, lsr #0xc
-	orr r3, r3, r2, lsl #0x14
-	str r3, [r1, #0x1c]
-	ldr r0, [r0, #0x20]
-	smull r2, r0, r12, r0
-	mov r2, r2, lsr #0xc
-	orr r2, r2, r0, lsl #0x14
-	str r2, [r1, #0x20]
-	ldmia sp!, {r4,lr}
-	bx lr
diff --git a/arm9/lib/src/FX_mtx33.c b/arm9/lib/src/FX_mtx33.c
new file mode 100644
index 00000000..32b1ad2e
--- /dev/null
+++ b/arm9/lib/src/FX_mtx33.c
@@ -0,0 +1,135 @@
+#include "global.h"
+#include "main.h"
+#include "fx.h"
+
+void MTX_ScaleApply33(struct Mtx33 *mtx, struct Mtx33 *dst, s32 x, s32 y, s32 z){
+    dst->_[0] = ((s64)x * mtx->_[0]) >> FX32_INT_SHIFT;
+    dst->_[1] = ((s64)x * mtx->_[1]) >> FX32_INT_SHIFT;
+    dst->_[2] = ((s64)x * mtx->_[2]) >> FX32_INT_SHIFT;
+    dst->_[3] = ((s64)y * mtx->_[3]) >> FX32_INT_SHIFT;
+    dst->_[4] = ((s64)y * mtx->_[4]) >> FX32_INT_SHIFT;
+    dst->_[5] = ((s64)y * mtx->_[5]) >> FX32_INT_SHIFT;
+    dst->_[6] = ((s64)z * mtx->_[6]) >> FX32_INT_SHIFT;
+    dst->_[7] = ((s64)z * mtx->_[7]) >> FX32_INT_SHIFT;
+    dst->_[8] = ((s64)z * mtx->_[8]) >> FX32_INT_SHIFT;
+}
+
+void MTX_Concat33(struct Mtx33 *a, struct Mtx33 *b, struct Mtx33 *c){
+    struct Mtx33 temp;
+    struct Mtx33 *dst;
+    s32 a0, a1, a2;
+    s32 b0, b1, b2;
+
+    if (c == b)
+        dst = &temp;
+    else
+        dst = c;
+
+    a0 = a->_[0];
+    a1 = a->_[1];
+    a2 = a->_[2];
+    dst->_[0] =  (((s64)a0  * b->_[0] + (s64)a1  * b->_[3] + (s64)a2  * b->_[6] ) >> FX32_INT_SHIFT);
+    dst->_[1] =  (((s64)a0  * b->_[1] + (s64)a1  * b->_[4] + (s64)a2  * b->_[7] ) >> FX32_INT_SHIFT);
+    b0 = b->_[2];
+    b1 = b->_[5];
+    b2 = b->_[8];
+    dst->_[2] =  (((s64)a0  * b0 + (s64)a1  * b1 + (s64)a2  * b2) >> FX32_INT_SHIFT);
+    a0 = a->_[3];
+    a1 = a->_[4];
+    a2 = a->_[5];
+    dst->_[5] =  (((s64)a0  * b0 + (s64)a1  * b1 + (s64)a2  * b2) >> FX32_INT_SHIFT);
+    dst->_[4] =  (((s64)a0  * b->_[1] + (s64)a1  * b->_[4] + (s64)a2  * b->_[7] ) >> FX32_INT_SHIFT);
+    b0 = b->_[0];
+    b1 = b->_[3];
+    b2 = b->_[6];
+    dst->_[3] =  (((s64)a0  * b0 + (s64)a1  * b1 + (s64)a2  * b2) >> FX32_INT_SHIFT);
+    a0 = a->_[6];
+    a1 = a->_[7];
+    a2 = a->_[8];
+    dst->_[6] =  (((s64)a0  * b0 + (s64)a1  * b1 + (s64)a2 * b2) >> FX32_INT_SHIFT);
+    dst->_[7] =  (((s64)a0  * b->_[1] + (s64)a1  * b->_[4] + (s64)a2 * b->_[7] ) >> FX32_INT_SHIFT);
+    b0 = b->_[2];
+    b1 = b->_[5];
+    b2 = b->_[8];
+    dst->_[8] = (((s64)a0  * b0 + (s64)a1  * b1 + (s64)a2 * b2) >> FX32_INT_SHIFT);
+
+    if (dst == &temp)
+        *c = temp;
+}
+
+void MTX_MultVec33(struct Vecx32 *vec, struct Mtx33 *mtx, struct Vecx32 *dst){
+    s32 x, y, z;
+    x = vec->x;
+    y = vec->y;
+    z = vec->z;
+    dst->x = ((s64)x * mtx->_[0] + (s64)y * mtx->_[3] + (s64)z * mtx->_[6]) >> FX32_INT_SHIFT;
+    dst->y = ((s64)x * mtx->_[1] + (s64)y * mtx->_[4] + (s64)z * mtx->_[7]) >> FX32_INT_SHIFT;
+    dst->z = ((s64)x * mtx->_[2] + (s64)y * mtx->_[5] + (s64)z * mtx->_[8]) >> FX32_INT_SHIFT;
+}
+
+asm void MTX_Identity33_(struct Mtx33 *mtx){
+    mov r2, #0x1000
+    str r2, [r0, #0x20]
+    mov r3, #0x0
+    stmia r0!, {r2-r3}
+    mov r1, #0x0
+    stmia r0!, {r1,r3}
+    stmia r0!, {r2-r3}
+    stmia r0!, {r1,r3}
+    bx lr
+}
+
+#pragma thumb on
+asm void MTX_RotX33_(struct Mtx33 *mtx, s32 sinphi, s32 cosphi){
+    mov r3, #0x1
+	lsl r3, r3, #0xc
+	str r3, [r0, #0x0]
+	mov r3, #0x0
+	str r3, [r0, #0x4]
+	str r3, [r0, #0x8]
+	str r3, [r0, #0xc]
+	str r2, [r0, #0x10]
+	str r1, [r0, #0x14]
+	str r3, [r0, #0x18]
+	neg r1, r1
+	str r1, [r0, #0x1c]
+	str r2, [r0, #0x20]
+	bx lr
+}
+#pragma thumb off
+
+#pragma thumb on
+asm void MTX_RotY33_(struct Mtx33 *mtx, s32 sinphi, s32 cosphi){
+    str r2, [r0, #0x0]
+	str r2, [r0, #0x20]
+	mov r3, #0x0
+	str r3, [r0, #0x4]
+	str r3, [r0, #0xc]
+	str r3, [r0, #0x14]
+	str r3, [r0, #0x1c]
+	neg r2, r1
+	mov r3, #0x1
+	lsl r3, r3, #0xc
+	str r1, [r0, #0x18]
+	str r2, [r0, #0x8]
+	str r3, [r0, #0x10]
+	bx lr
+}
+#pragma thumb off
+
+#pragma thumb on
+asm void MTX_RotZ33_(struct Mtx33 *mtx, s32 sinphi, s32 cosphi){
+    stmia r0!, {r2}
+	mov r3, #0x0
+	stmia r0!, {r1,r3}
+	neg r1, r1
+	stmia r0!, {r1-r2}
+	mov r1, #0x1
+	lsl r1, r1, #0xc
+	str r3, [r0, #0x0]
+	str r3, [r0, #0x4]
+	str r3, [r0, #0x8]
+	str r1, [r0, #0xc]
+	bx lr
+}
+#pragma thumb off
diff --git a/arm9/lib/src/FX_mtx43.c b/arm9/lib/src/FX_mtx43.c
index c5d227b8..9cb8aa6e 100644
--- a/arm9/lib/src/FX_mtx43.c
+++ b/arm9/lib/src/FX_mtx43.c
@@ -3,9 +3,6 @@
 #include "fx.h"
 
 
-
-
-
 void MTX_ScaleApply43(struct Mtx43 *mtx, struct Mtx43 *dst, s32 x, s32 y, s32 z){
     //this works because matrices are indexed columns first
     MTX_ScaleApply33((struct Mtx33 *)mtx, (struct Mtx33 *)dst, x, y, z);
diff --git a/include/fx.h b/include/fx.h
index ee436883..319c823d 100644
--- a/include/fx.h
+++ b/include/fx.h
@@ -59,6 +59,9 @@
 #define READREG32(x)               (*(vu32 *)x)
 #define READREG64(x)               (*(vu64 *)x)
 
+//TODO: add muladd aund mulsub macros
+//need the extra term on mulsub to round up?
+
 struct Vecx32
 {
     s32 x;
@@ -129,7 +132,6 @@ s32 FX_ModS32(s32 num, s32 mod);
 
 //Mtx
 //The functions ending in underscores seem to be in assembly originally
-//at least I see no way to recreate the stmia copies and some of them are actually in thumb
 //Mtx44
 void MTX_TransApply44(struct Mtx44 *mtx, struct Mtx44 *dst, s32 x, s32 y, s32 z);
 void MTX_Concat44(struct Mtx44 *a, struct Mtx44 *b, struct Mtx44 *c);
@@ -150,5 +152,14 @@ void MTX_Scale43_(struct Mtx43 *dst, s32 x, s32 y, s32 z);
 void MTX_RotX43_(struct Mtx43 *mtx, s32 sinphi, s32 cosphi);
 void MTX_RotY43_(struct Mtx43 *mtx, s32 sinphi, s32 cosphi);
 
+//Mtx33
+void MTX_ScaleApply33(struct Mtx33 *mtx, struct Mtx33 *dst, s32 x, s32 y, s32 z);
+void MTX_Concat33(struct Mtx33 *a, struct Mtx33 *b, struct Mtx33 *c);
+void MTX_MultVec33(struct Vecx32 *vec, struct Mtx33 *mtx, struct Vecx32 *dst);
+void MTX_Identity33_(struct Mtx33 *mtx);
+void MTX_RotX33_(struct Mtx33 *mtx, s32 sinphi, s32 cosphi);
+void MTX_RotY33_(struct Mtx33 *mtx, s32 sinphi, s32 cosphi);
+void MTX_RotZ33_(struct Mtx33 *mtx, s32 sinphi, s32 cosphi);
+
 
 #endif //GUARD_FX_H
-- 
cgit v1.2.3


From 246d8051ac7b7535c53d5d180eaf1d97451d31f2 Mon Sep 17 00:00:00 2001
From: Made <made111@gmx.de>
Date: Mon, 4 May 2020 17:01:35 +0200
Subject: Decompile FX_mtx22.s

---
 arm9/asm/FX_mtx22.s     | 50 -------------------------
 arm9/lib/src/FX_mtx22.c | 30 +++++++++++++++
 include/fx.h            | 97 +++++++++++++++++++++++++++----------------------
 3 files changed, 84 insertions(+), 93 deletions(-)
 delete mode 100644 arm9/asm/FX_mtx22.s
 create mode 100644 arm9/lib/src/FX_mtx22.c

diff --git a/arm9/asm/FX_mtx22.s b/arm9/asm/FX_mtx22.s
deleted file mode 100644
index 571180d2..00000000
--- a/arm9/asm/FX_mtx22.s
+++ /dev/null
@@ -1,50 +0,0 @@
-    .include "asm/macros.inc"
-    .include "global.inc"
-
-    .text
-
-	arm_func_start MTX_Identity22_
-MTX_Identity22_: ; 0x020C3E8C
-	mov r1, #0x0
-	mov r2, #0x1000
-	mov r3, #0x0
-	stmia r0!, {r2-r3}
-	stmia r0!, {r1-r2}
-	bx lr
-
-	thumb_func_start MTX_Rot22_
-MTX_Rot22_: ; 0x020C3EA4
-	str r2, [r0, #0x0]
-	str r1, [r0, #0x4]
-	neg r1, r1
-	str r1, [r0, #0x8]
-	str r2, [r0, #0xc]
-	bx lr
-
-	arm_func_start MTX_ScaleApply22
-MTX_ScaleApply22: ; 0x020C3EB0
-	stmdb sp!, {lr}
-	sub sp, sp, #0x4
-	ldr r12, [r0, #0x0]
-	smull lr, r12, r2, r12
-	mov lr, lr, lsr #0xc
-	orr lr, lr, r12, lsl #0x14
-	str lr, [r1, #0x0]
-	ldr r12, [r0, #0x4]
-	smull lr, r12, r2, r12
-	mov r2, lr, lsr #0xc
-	orr r2, r2, r12, lsl #0x14
-	str r2, [r1, #0x4]
-	ldr r2, [r0, #0x8]
-	smull r12, r2, r3, r2
-	mov r12, r12, lsr #0xc
-	orr r12, r12, r2, lsl #0x14
-	str r12, [r1, #0x8]
-	ldr r0, [r0, #0xc]
-	smull r2, r0, r3, r0
-	mov r2, r2, lsr #0xc
-	orr r2, r2, r0, lsl #0x14
-	str r2, [r1, #0xc]
-	add sp, sp, #0x4
-	ldmia sp!, {lr}
-	bx lr
diff --git a/arm9/lib/src/FX_mtx22.c b/arm9/lib/src/FX_mtx22.c
new file mode 100644
index 00000000..fcd634b1
--- /dev/null
+++ b/arm9/lib/src/FX_mtx22.c
@@ -0,0 +1,30 @@
+#include "global.h"
+#include "main.h"
+#include "fx.h"
+
+void MTX_ScaleApply22(struct Mtx22 *mtx, struct Mtx22 *dst, s32 x, s32 y){
+    dst->_[0] = ((s64)x * mtx->_[0]) >> FX32_INT_SHIFT;
+    dst->_[1] = ((s64)x * mtx->_[1]) >> FX32_INT_SHIFT;
+    dst->_[2] = ((s64)y * mtx->_[2]) >> FX32_INT_SHIFT;
+    dst->_[3] = ((s64)y * mtx->_[3]) >> FX32_INT_SHIFT;
+}
+
+asm void MTX_Identity22_(struct Mtx22 *mtx){
+    mov r1, #0x0
+    mov r2, #0x1000
+    mov r3, #0x0
+    stmia r0!, {r2-r3}
+    stmia r0!, {r1-r2}
+    bx lr
+}
+
+#pragma thumb on
+asm void MTX_Rot22_(struct Mtx22 *mtx, s32 sinphi, s32 cosphi){
+    str r2, [r0, #0x0]
+	str r1, [r0, #0x4]
+	neg r1, r1
+	str r1, [r0, #0x8]
+	str r2, [r0, #0xc]
+	bx lr
+}
+#pragma thumb off
diff --git a/include/fx.h b/include/fx.h
index 319c823d..bbbf621a 100644
--- a/include/fx.h
+++ b/include/fx.h
@@ -1,6 +1,11 @@
 #ifndef GUARD_FX_H
 #define GUARD_FX_H
 
+typedef s16 fx16;
+typedef s32 fx32;
+typedef s64 fx64;
+typedef s64 fx64c;
+
 #define FX16_INT_MASK              0xF000
 #define FX16_INT_ABS_MASK          0x7000
 #define FX16_FRAC_MASK             0x0FFF
@@ -59,107 +64,113 @@
 #define READREG32(x)               (*(vu32 *)x)
 #define READREG64(x)               (*(vu64 *)x)
 
-//TODO: add muladd aund mulsub macros
-//need the extra term on mulsub to round up?
+#define FX32_MUL(a, b)               ((fx32)(((fx64)a * b) >> FX32_INT_SHIFT))
+#define FX32_MUL_ADD_MUL(a, b, c, d) ((fx32)(((fx64)a * b + (fx64)c * d) >> FX32_INT_SHIFT))
+//the extra term here is for rounding
+#define FX32_MUL_SUB_MUL(a, b, c, d) ((fx32)(((fx64)a * b - (fx64)c * d + (1 << (FX32_INT_SHIFT - 1))) >> FX32_INT_SHIFT))
 
 struct Vecx32
 {
-    s32 x;
-    s32 y;
-    s32 z;
+    fx32 x;
+    fx32 y;
+    fx32 z;
 };
 
 struct Vecx16
 {
-    s16 x;
-    s16 y;
-    s16 z;
+    fx16 x;
+    fx16 y;
+    fx16 z;
 };
 
 //Matrices are indexed as [column][row]
 struct Mtx44
 {
-    s32 _[16];
+    fx32 _[16];
 };
 
 struct Mtx43
 {
-    s32 _[12];
+    fx32 _[12];
 };
 
 struct Mtx33
 {
-    s32 _[9];
+    fx32 _[9];
 };
 
 struct Mtx22
 {
-    s32 _[4];
+    fx32 _[4];
 };
 
 //FX
 void FX_Init();
-s32 FX_Modf(s32 x, s32 *iptr);
+fx32 FX_Modf(fx32 x, fx32 *iptr);
 
 //Atan
-u16 FX_Atan(s32 x);
-u16 FX_Atan2(s32 x, s32 y);
+u16 FX_Atan(fx32 x);
+u16 FX_Atan2(fx32 x, fx32 y);
 
 //Vec
 void VEC_Add(struct Vecx32 *x, struct Vecx32 *y, struct Vecx32 *dst);
 void VEC_Subtract(struct Vecx32 *x, struct Vecx32 *y, struct Vecx32 *dst);
 void VEC_Fx16Add(struct Vecx16 *x, struct Vecx16 *y, struct Vecx16 *dst);
-s32 VEC_DotProduct(struct Vecx32 *x, struct Vecx32 *y);
-s32 VEC_Fx16DotProduct(struct Vecx16 *a, struct Vecx16 *b);
+fx32 VEC_DotProduct(struct Vecx32 *x, struct Vecx32 *y);
+fx32 VEC_Fx16DotProduct(struct Vecx16 *a, struct Vecx16 *b);
 void VEC_CrossProduct(struct Vecx32 *a, struct Vecx32 *b, struct Vecx32 *dst);
 void VEC_Fx16CrossProduct(struct Vecx16 *a, struct Vecx16 *b, struct Vecx16 *dst);
-s32 VEC_Mag(struct Vecx32 *a);
+fx32 VEC_Mag(struct Vecx32 *a);
 void VEC_Normalize(struct Vecx32 *a, struct Vecx32 *dst);
 void VEC_Fx16Normalize(struct Vecx16 *a, struct Vecx16 *dst);
-void VEC_MultAdd(s32 factor, struct Vecx32  *a, struct Vecx32 *b, struct Vecx32 *dst);
+void VEC_MultAdd(fx32 factor, struct Vecx32  *a, struct Vecx32 *b, struct Vecx32 *dst);
 
 //CP
-s32 FX_Div(s32 numerator, s32 denominator);
-s32 FX_Inv(s32 x);
-s32 FX_Sqrt(s32 x);
-s64 FX_GetDivResultFx64c();
-s32 FX_GetDivResult();
-void FX_InvAsync(s32 x);
-s32 FX_GetSqrtResult();
-void FX_DivAsync(s32 numerator, s32 denominator);
-s32 FX_DivS32(s32 numerator, s32 denominator);
-s32 FX_ModS32(s32 num, s32 mod);
+fx32 FX_Div(fx32 numerator, fx32 denominator);
+fx32 FX_Inv(fx32 x);
+fx32 FX_Sqrt(fx32 x);
+fx64 FX_GetDivResultFx64c();
+fx32 FX_GetDivResult();
+void FX_InvAsync(fx32 x);
+fx32 FX_GetSqrtResult();
+void FX_DivAsync(fx32 numerator, fx32 denominator);
+fx32 FX_Divfx32(fx32 numerator, fx32 denominator);
+fx32 FX_Modfx32(fx32 num, fx32 mod);
 
 //Mtx
 //The functions ending in underscores seem to be in assembly originally
 //Mtx44
-void MTX_TransApply44(struct Mtx44 *mtx, struct Mtx44 *dst, s32 x, s32 y, s32 z);
+void MTX_TransApply44(struct Mtx44 *mtx, struct Mtx44 *dst, fx32 x, fx32 y, fx32 z);
 void MTX_Concat44(struct Mtx44 *a, struct Mtx44 *b, struct Mtx44 *c);
 void MTX_Identity44_(struct Mtx44 *dst);
 void MTX_Copy44To43_(struct Mtx44 *src, struct Mtx43 *dst);
-void MTX_RotX44_(struct Mtx44 *mtx, s32 sinphi, s32 cosphi);
-void MTX_RotY44_(struct Mtx44 *mtx, s32 sinphi, s32 cosphi);
-void MTX_RotZ44_(struct Mtx44 *mtx, s32 sinphi, s32 cosphi);
+void MTX_RotX44_(struct Mtx44 *mtx, fx32 sinphi, fx32 cosphi);
+void MTX_RotY44_(struct Mtx44 *mtx, fx32 sinphi, fx32 cosphi);
+void MTX_RotZ44_(struct Mtx44 *mtx, fx32 sinphi, fx32 cosphi);
 
 //Mtx43
-void MTX_ScaleApply33(struct Mtx33 *mtx, struct Mtx33 *dst, s32 x, s32 y, s32 z);
-s32 MTX_Inverse43(struct Mtx43 *mtx, struct Mtx43 *inv);
+void MTX_ScaleApply33(struct Mtx33 *mtx, struct Mtx33 *dst, fx32 x, fx32 y, fx32 z);
+fx32 MTX_Inverse43(struct Mtx43 *mtx, struct Mtx43 *inv);
 void MTX_Concat43(struct Mtx43 *a, struct Mtx43 *b, struct Mtx43 *c);
 void MTX_MultVec43(struct Vecx32 *vec, struct Mtx43 *mtx, struct Vecx32 *dst);
 void MTX_Identity43_(struct Mtx43 *mtx);
 void MTX_Copy43To44_(struct Mtx43 *src, struct Mtx44 *dst);
-void MTX_Scale43_(struct Mtx43 *dst, s32 x, s32 y, s32 z);
-void MTX_RotX43_(struct Mtx43 *mtx, s32 sinphi, s32 cosphi);
-void MTX_RotY43_(struct Mtx43 *mtx, s32 sinphi, s32 cosphi);
+void MTX_Scale43_(struct Mtx43 *dst, fx32 x, fx32 y, fx32 z);
+void MTX_RotX43_(struct Mtx43 *mtx, fx32 sinphi, fx32 cosphi);
+void MTX_RotY43_(struct Mtx43 *mtx, fx32 sinphi, fx32 cosphi);
 
 //Mtx33
-void MTX_ScaleApply33(struct Mtx33 *mtx, struct Mtx33 *dst, s32 x, s32 y, s32 z);
+void MTX_ScaleApply33(struct Mtx33 *mtx, struct Mtx33 *dst, fx32 x, fx32 y, fx32 z);
 void MTX_Concat33(struct Mtx33 *a, struct Mtx33 *b, struct Mtx33 *c);
 void MTX_MultVec33(struct Vecx32 *vec, struct Mtx33 *mtx, struct Vecx32 *dst);
 void MTX_Identity33_(struct Mtx33 *mtx);
-void MTX_RotX33_(struct Mtx33 *mtx, s32 sinphi, s32 cosphi);
-void MTX_RotY33_(struct Mtx33 *mtx, s32 sinphi, s32 cosphi);
-void MTX_RotZ33_(struct Mtx33 *mtx, s32 sinphi, s32 cosphi);
-
+void MTX_RotX33_(struct Mtx33 *mtx, fx32 sinphi, fx32 cosphi);
+void MTX_RotY33_(struct Mtx33 *mtx, fx32 sinphi, fx32 cosphi);
+void MTX_RotZ33_(struct Mtx33 *mtx, fx32 sinphi, fx32 cosphi);
+
+//Mtx22
+void MTX_ScaleApply22(struct Mtx22 *mtx, struct Mtx22 *dst, fx32 x, fx32 y);
+void MTX_Identity22_(struct Mtx22 *mtx);
+void MTX_Rot22_(struct Mtx22 *mtx, fx32 sinphi, fx32 cosphi);
 
 #endif //GUARD_FX_H
-- 
cgit v1.2.3


From ab28b0b37f2fd9fba6010e48e15bd31170d678d7 Mon Sep 17 00:00:00 2001
From: Made <made111@gmx.de>
Date: Mon, 4 May 2020 17:07:22 +0200
Subject: Add typedefs for fixed point types

---
 arm9/lib/src/FX.c       |   2 +-
 arm9/lib/src/FX_atan.c  |  12 +++---
 arm9/lib/src/FX_cp.c    |  28 +++++++-------
 arm9/lib/src/FX_mtx22.c |  12 +++---
 arm9/lib/src/FX_mtx33.c |  56 +++++++++++++--------------
 arm9/lib/src/FX_mtx43.c | 100 ++++++++++++++++++++++++------------------------
 arm9/lib/src/FX_mtx44.c |  52 ++++++++++++-------------
 arm9/lib/src/FX_vec.c   |  50 ++++++++++++------------
 8 files changed, 156 insertions(+), 156 deletions(-)

diff --git a/arm9/lib/src/FX.c b/arm9/lib/src/FX.c
index fa6fa6ba..f39b1995 100644
--- a/arm9/lib/src/FX.c
+++ b/arm9/lib/src/FX.c
@@ -6,7 +6,7 @@ void FX_Init(){
     return;
 }
 
-s32 FX_Modf(s32 x, s32 *iptr){
+fx32 FX_Modf(fx32 x, fx32 *iptr){
     if (x >= 0)
     {
         *iptr = x & 0x7FFFF000;
diff --git a/arm9/lib/src/FX_atan.c b/arm9/lib/src/FX_atan.c
index 684db81e..085763f8 100644
--- a/arm9/lib/src/FX_atan.c
+++ b/arm9/lib/src/FX_atan.c
@@ -2,15 +2,15 @@
 #include "main.h"
 #include "fx.h"
 
-extern s16 FX_AtanTable_[];
+extern fx16 FX_AtanTable_[];
 
-u16 FX_Atan(s32 x){
+u16 FX_Atan(fx32 x){
     if (x >= 0)
     {
         if (x > 0x1000)
         {
             x = FX_Inv(x);
-            s16 y = FX_AtanTable_[x >> 5];
+            fx16 y = FX_AtanTable_[x >> 5];
             return 0x4000 - y;
         }
         else if (x < 0x1000)
@@ -27,7 +27,7 @@ u16 FX_Atan(s32 x){
         if (x < -0x1000)
         {
             x = FX_Inv(-x);
-            s16 y = FX_AtanTable_[x >> 5];
+            fx16 y = FX_AtanTable_[x >> 5];
             return y - 0x4000;
         }
         else if (x > -0x1000)
@@ -41,8 +41,8 @@ u16 FX_Atan(s32 x){
     }
 }
 
-u16 FX_Atan2(s32 x, s32 y){
-    s32 result;
+u16 FX_Atan2(fx32 x, fx32 y){
+    fx32 result;
     u32 positive, bias, denominator, numerator;
     if (x > 0)
     {
diff --git a/arm9/lib/src/FX_cp.c b/arm9/lib/src/FX_cp.c
index a2d8307b..76dea4b4 100644
--- a/arm9/lib/src/FX_cp.c
+++ b/arm9/lib/src/FX_cp.c
@@ -1,23 +1,23 @@
 #include "global.h"
 #include "main.h"
 #include "fx.h"
- 
 
-s32 FX_Div(s32 numerator, s32 denominator){
+
+fx32 FX_Div(fx32 numerator, fx32 denominator){
     FX_DivAsync(numerator, denominator);
     return FX_GetDivResult();
 }
 
-s32 FX_Inv(s32 x){
+fx32 FX_Inv(fx32 x){
     FX_InvAsync(x);
     return FX_GetDivResult();
 }
 
-s32 FX_Sqrt(s32 x){
+fx32 FX_Sqrt(fx32 x){
     if (x > 0)
     {
         SETREG16(HW_REG_SQRTCNT, 0x1);
-        SETREG64(HW_REG_SQRT_PARAM, (s64)x << 32);
+        SETREG64(HW_REG_SQRT_PARAM, (fx64)x << 32);
         return FX_GetSqrtResult();
     }
     else
@@ -26,34 +26,34 @@ s32 FX_Sqrt(s32 x){
     }
 }
 
-s64 FX_GetDivResultFx64c(){
+fx64c FX_GetDivResultFx64c(){
     while (READREG16(HW_REG_DIVCNT) & 0x8000);
     return READREG64(HW_REG_DIV_RESULT);
 }
 
-s32 FX_GetDivResult(){
+fx32 FX_GetDivResult(){
     while (READREG16(HW_REG_DIVCNT) & 0x8000);
     return (READREG64(HW_REG_DIV_RESULT) + (1 << (0x14 - 1))) >> 0x14;
 }
 
-void FX_InvAsync(s32 x){
+void FX_InvAsync(fx32 x){
     SETREG16(HW_REG_DIVCNT, 0x1);
-    SETREG64(HW_REG_DIV_NUMER, (s64)0x00001000 << 32);
+    SETREG64(HW_REG_DIV_NUMER, (fx64)0x00001000 << 32);
     SETREG64(HW_REG_DIV_DENOM, (u32)x);
 }
 
-s32 FX_GetSqrtResult(){
+fx32 FX_GetSqrtResult(){
     while (READREG16(HW_REG_SQRTCNT) & 0x8000);
     return (READREG32(HW_REG_SQRT_RESULT) + (1 << (0xA - 1))) >> 0xA;
 }
 
-void FX_DivAsync(s32 numerator, s32 denominator){
+void FX_DivAsync(fx32 numerator, fx32 denominator){
     SETREG16(HW_REG_DIVCNT, 0x1);
-    SETREG64(HW_REG_DIV_NUMER, (s64)numerator << 32);
+    SETREG64(HW_REG_DIV_NUMER, (fx64)numerator << 32);
     SETREG64(HW_REG_DIV_DENOM, (u32)denominator);
 }
 
-s32 FX_DivS32(s32 numerator, s32 denominator){
+fx32 FX_Divfx32(fx32 numerator, fx32 denominator){
     SETREG16(HW_REG_DIVCNT, 0x0);
     SETREG32(HW_REG_DIV_NUMER, (u32)numerator); //32bit write for some reason
     SETREG64(HW_REG_DIV_DENOM, (u32)denominator);
@@ -61,7 +61,7 @@ s32 FX_DivS32(s32 numerator, s32 denominator){
     return READREG32(HW_REG_DIV_RESULT);
 }
 
-s32 FX_ModS32(s32 num, s32 mod){
+fx32 FX_Modfx32(fx32 num, fx32 mod){
     SETREG16(HW_REG_DIVCNT, 0x0);
     SETREG32(HW_REG_DIV_NUMER, (u32)num); //32bit write for some reason
     SETREG64(HW_REG_DIV_DENOM, (u32)mod);
diff --git a/arm9/lib/src/FX_mtx22.c b/arm9/lib/src/FX_mtx22.c
index fcd634b1..19504574 100644
--- a/arm9/lib/src/FX_mtx22.c
+++ b/arm9/lib/src/FX_mtx22.c
@@ -2,11 +2,11 @@
 #include "main.h"
 #include "fx.h"
 
-void MTX_ScaleApply22(struct Mtx22 *mtx, struct Mtx22 *dst, s32 x, s32 y){
-    dst->_[0] = ((s64)x * mtx->_[0]) >> FX32_INT_SHIFT;
-    dst->_[1] = ((s64)x * mtx->_[1]) >> FX32_INT_SHIFT;
-    dst->_[2] = ((s64)y * mtx->_[2]) >> FX32_INT_SHIFT;
-    dst->_[3] = ((s64)y * mtx->_[3]) >> FX32_INT_SHIFT;
+void MTX_ScaleApply22(struct Mtx22 *mtx, struct Mtx22 *dst, fx32 x, fx32 y){
+    dst->_[0] = ((fx64)x * mtx->_[0]) >> FX32_INT_SHIFT;
+    dst->_[1] = ((fx64)x * mtx->_[1]) >> FX32_INT_SHIFT;
+    dst->_[2] = ((fx64)y * mtx->_[2]) >> FX32_INT_SHIFT;
+    dst->_[3] = ((fx64)y * mtx->_[3]) >> FX32_INT_SHIFT;
 }
 
 asm void MTX_Identity22_(struct Mtx22 *mtx){
@@ -19,7 +19,7 @@ asm void MTX_Identity22_(struct Mtx22 *mtx){
 }
 
 #pragma thumb on
-asm void MTX_Rot22_(struct Mtx22 *mtx, s32 sinphi, s32 cosphi){
+asm void MTX_Rot22_(struct Mtx22 *mtx, fx32 sinphi, fx32 cosphi){
     str r2, [r0, #0x0]
 	str r1, [r0, #0x4]
 	neg r1, r1
diff --git a/arm9/lib/src/FX_mtx33.c b/arm9/lib/src/FX_mtx33.c
index 32b1ad2e..391a6104 100644
--- a/arm9/lib/src/FX_mtx33.c
+++ b/arm9/lib/src/FX_mtx33.c
@@ -2,23 +2,23 @@
 #include "main.h"
 #include "fx.h"
 
-void MTX_ScaleApply33(struct Mtx33 *mtx, struct Mtx33 *dst, s32 x, s32 y, s32 z){
-    dst->_[0] = ((s64)x * mtx->_[0]) >> FX32_INT_SHIFT;
-    dst->_[1] = ((s64)x * mtx->_[1]) >> FX32_INT_SHIFT;
-    dst->_[2] = ((s64)x * mtx->_[2]) >> FX32_INT_SHIFT;
-    dst->_[3] = ((s64)y * mtx->_[3]) >> FX32_INT_SHIFT;
-    dst->_[4] = ((s64)y * mtx->_[4]) >> FX32_INT_SHIFT;
-    dst->_[5] = ((s64)y * mtx->_[5]) >> FX32_INT_SHIFT;
-    dst->_[6] = ((s64)z * mtx->_[6]) >> FX32_INT_SHIFT;
-    dst->_[7] = ((s64)z * mtx->_[7]) >> FX32_INT_SHIFT;
-    dst->_[8] = ((s64)z * mtx->_[8]) >> FX32_INT_SHIFT;
+void MTX_ScaleApply33(struct Mtx33 *mtx, struct Mtx33 *dst, fx32 x, fx32 y, fx32 z){
+    dst->_[0] = ((fx64)x * mtx->_[0]) >> FX32_INT_SHIFT;
+    dst->_[1] = ((fx64)x * mtx->_[1]) >> FX32_INT_SHIFT;
+    dst->_[2] = ((fx64)x * mtx->_[2]) >> FX32_INT_SHIFT;
+    dst->_[3] = ((fx64)y * mtx->_[3]) >> FX32_INT_SHIFT;
+    dst->_[4] = ((fx64)y * mtx->_[4]) >> FX32_INT_SHIFT;
+    dst->_[5] = ((fx64)y * mtx->_[5]) >> FX32_INT_SHIFT;
+    dst->_[6] = ((fx64)z * mtx->_[6]) >> FX32_INT_SHIFT;
+    dst->_[7] = ((fx64)z * mtx->_[7]) >> FX32_INT_SHIFT;
+    dst->_[8] = ((fx64)z * mtx->_[8]) >> FX32_INT_SHIFT;
 }
 
 void MTX_Concat33(struct Mtx33 *a, struct Mtx33 *b, struct Mtx33 *c){
     struct Mtx33 temp;
     struct Mtx33 *dst;
-    s32 a0, a1, a2;
-    s32 b0, b1, b2;
+    fx32 a0, a1, a2;
+    fx32 b0, b1, b2;
 
     if (c == b)
         dst = &temp;
@@ -28,43 +28,43 @@ void MTX_Concat33(struct Mtx33 *a, struct Mtx33 *b, struct Mtx33 *c){
     a0 = a->_[0];
     a1 = a->_[1];
     a2 = a->_[2];
-    dst->_[0] =  (((s64)a0  * b->_[0] + (s64)a1  * b->_[3] + (s64)a2  * b->_[6] ) >> FX32_INT_SHIFT);
-    dst->_[1] =  (((s64)a0  * b->_[1] + (s64)a1  * b->_[4] + (s64)a2  * b->_[7] ) >> FX32_INT_SHIFT);
+    dst->_[0] =  (((fx64)a0  * b->_[0] + (fx64)a1  * b->_[3] + (fx64)a2  * b->_[6] ) >> FX32_INT_SHIFT);
+    dst->_[1] =  (((fx64)a0  * b->_[1] + (fx64)a1  * b->_[4] + (fx64)a2  * b->_[7] ) >> FX32_INT_SHIFT);
     b0 = b->_[2];
     b1 = b->_[5];
     b2 = b->_[8];
-    dst->_[2] =  (((s64)a0  * b0 + (s64)a1  * b1 + (s64)a2  * b2) >> FX32_INT_SHIFT);
+    dst->_[2] =  (((fx64)a0  * b0 + (fx64)a1  * b1 + (fx64)a2  * b2) >> FX32_INT_SHIFT);
     a0 = a->_[3];
     a1 = a->_[4];
     a2 = a->_[5];
-    dst->_[5] =  (((s64)a0  * b0 + (s64)a1  * b1 + (s64)a2  * b2) >> FX32_INT_SHIFT);
-    dst->_[4] =  (((s64)a0  * b->_[1] + (s64)a1  * b->_[4] + (s64)a2  * b->_[7] ) >> FX32_INT_SHIFT);
+    dst->_[5] =  (((fx64)a0  * b0 + (fx64)a1  * b1 + (fx64)a2  * b2) >> FX32_INT_SHIFT);
+    dst->_[4] =  (((fx64)a0  * b->_[1] + (fx64)a1  * b->_[4] + (fx64)a2  * b->_[7] ) >> FX32_INT_SHIFT);
     b0 = b->_[0];
     b1 = b->_[3];
     b2 = b->_[6];
-    dst->_[3] =  (((s64)a0  * b0 + (s64)a1  * b1 + (s64)a2  * b2) >> FX32_INT_SHIFT);
+    dst->_[3] =  (((fx64)a0  * b0 + (fx64)a1  * b1 + (fx64)a2  * b2) >> FX32_INT_SHIFT);
     a0 = a->_[6];
     a1 = a->_[7];
     a2 = a->_[8];
-    dst->_[6] =  (((s64)a0  * b0 + (s64)a1  * b1 + (s64)a2 * b2) >> FX32_INT_SHIFT);
-    dst->_[7] =  (((s64)a0  * b->_[1] + (s64)a1  * b->_[4] + (s64)a2 * b->_[7] ) >> FX32_INT_SHIFT);
+    dst->_[6] =  (((fx64)a0  * b0 + (fx64)a1  * b1 + (fx64)a2 * b2) >> FX32_INT_SHIFT);
+    dst->_[7] =  (((fx64)a0  * b->_[1] + (fx64)a1  * b->_[4] + (fx64)a2 * b->_[7] ) >> FX32_INT_SHIFT);
     b0 = b->_[2];
     b1 = b->_[5];
     b2 = b->_[8];
-    dst->_[8] = (((s64)a0  * b0 + (s64)a1  * b1 + (s64)a2 * b2) >> FX32_INT_SHIFT);
+    dst->_[8] = (((fx64)a0  * b0 + (fx64)a1  * b1 + (fx64)a2 * b2) >> FX32_INT_SHIFT);
 
     if (dst == &temp)
         *c = temp;
 }
 
 void MTX_MultVec33(struct Vecx32 *vec, struct Mtx33 *mtx, struct Vecx32 *dst){
-    s32 x, y, z;
+    fx32 x, y, z;
     x = vec->x;
     y = vec->y;
     z = vec->z;
-    dst->x = ((s64)x * mtx->_[0] + (s64)y * mtx->_[3] + (s64)z * mtx->_[6]) >> FX32_INT_SHIFT;
-    dst->y = ((s64)x * mtx->_[1] + (s64)y * mtx->_[4] + (s64)z * mtx->_[7]) >> FX32_INT_SHIFT;
-    dst->z = ((s64)x * mtx->_[2] + (s64)y * mtx->_[5] + (s64)z * mtx->_[8]) >> FX32_INT_SHIFT;
+    dst->x = ((fx64)x * mtx->_[0] + (fx64)y * mtx->_[3] + (fx64)z * mtx->_[6]) >> FX32_INT_SHIFT;
+    dst->y = ((fx64)x * mtx->_[1] + (fx64)y * mtx->_[4] + (fx64)z * mtx->_[7]) >> FX32_INT_SHIFT;
+    dst->z = ((fx64)x * mtx->_[2] + (fx64)y * mtx->_[5] + (fx64)z * mtx->_[8]) >> FX32_INT_SHIFT;
 }
 
 asm void MTX_Identity33_(struct Mtx33 *mtx){
@@ -80,7 +80,7 @@ asm void MTX_Identity33_(struct Mtx33 *mtx){
 }
 
 #pragma thumb on
-asm void MTX_RotX33_(struct Mtx33 *mtx, s32 sinphi, s32 cosphi){
+asm void MTX_RotX33_(struct Mtx33 *mtx, fx32 sinphi, fx32 cosphi){
     mov r3, #0x1
 	lsl r3, r3, #0xc
 	str r3, [r0, #0x0]
@@ -99,7 +99,7 @@ asm void MTX_RotX33_(struct Mtx33 *mtx, s32 sinphi, s32 cosphi){
 #pragma thumb off
 
 #pragma thumb on
-asm void MTX_RotY33_(struct Mtx33 *mtx, s32 sinphi, s32 cosphi){
+asm void MTX_RotY33_(struct Mtx33 *mtx, fx32 sinphi, fx32 cosphi){
     str r2, [r0, #0x0]
 	str r2, [r0, #0x20]
 	mov r3, #0x0
@@ -118,7 +118,7 @@ asm void MTX_RotY33_(struct Mtx33 *mtx, s32 sinphi, s32 cosphi){
 #pragma thumb off
 
 #pragma thumb on
-asm void MTX_RotZ33_(struct Mtx33 *mtx, s32 sinphi, s32 cosphi){
+asm void MTX_RotZ33_(struct Mtx33 *mtx, fx32 sinphi, fx32 cosphi){
     stmia r0!, {r2}
 	mov r3, #0x0
 	stmia r0!, {r1,r3}
diff --git a/arm9/lib/src/FX_mtx43.c b/arm9/lib/src/FX_mtx43.c
index 9cb8aa6e..862bd6f2 100644
--- a/arm9/lib/src/FX_mtx43.c
+++ b/arm9/lib/src/FX_mtx43.c
@@ -3,7 +3,7 @@
 #include "fx.h"
 
 
-void MTX_ScaleApply43(struct Mtx43 *mtx, struct Mtx43 *dst, s32 x, s32 y, s32 z){
+void MTX_ScaleApply43(struct Mtx43 *mtx, struct Mtx43 *dst, fx32 x, fx32 y, fx32 z){
     //this works because matrices are indexed columns first
     MTX_ScaleApply33((struct Mtx33 *)mtx, (struct Mtx33 *)dst, x, y, z);
     dst->_[9] = mtx->_[9];
@@ -11,48 +11,48 @@ void MTX_ScaleApply43(struct Mtx43 *mtx, struct Mtx43 *dst, s32 x, s32 y, s32 z)
     dst->_[11] = mtx->_[11];
 }
 
-s32 MTX_Inverse43(struct Mtx43 *mtx, struct Mtx43 *inv){
+fx32 MTX_Inverse43(struct Mtx43 *mtx, struct Mtx43 *inv){
     struct Mtx43 tempmat;
     struct Mtx43 *dst;
-    s32 det0, det1, det2, det;
-    s32 var0, var1, var2, var3;
+    fx32 det0, det1, det2, det;
+    fx32 var0, var1, var2, var3;
     if (mtx == inv)
         dst = &tempmat;
     else
         dst = inv;
     //subdeterminants
-    det0 = ((s64)mtx->_[4] * mtx->_[8] - (s64)mtx->_[5] * mtx->_[7] + (s64)(1 << (FX32_INT_SHIFT - 1))) >> FX32_INT_SHIFT;
-    det1 = ((s64)mtx->_[3] * mtx->_[8] - (s64)mtx->_[5] * mtx->_[6] + (s64)(1 << (FX32_INT_SHIFT - 1))) >> FX32_INT_SHIFT;
-    det2 = ((s64)mtx->_[3] * mtx->_[7] - (s64)mtx->_[4] * mtx->_[6] + (s64)(1 << (FX32_INT_SHIFT - 1))) >> FX32_INT_SHIFT;
+    det0 = ((fx64)mtx->_[4] * mtx->_[8] - (fx64)mtx->_[5] * mtx->_[7] + (fx64)(1 << (FX32_INT_SHIFT - 1))) >> FX32_INT_SHIFT;
+    det1 = ((fx64)mtx->_[3] * mtx->_[8] - (fx64)mtx->_[5] * mtx->_[6] + (fx64)(1 << (FX32_INT_SHIFT - 1))) >> FX32_INT_SHIFT;
+    det2 = ((fx64)mtx->_[3] * mtx->_[7] - (fx64)mtx->_[4] * mtx->_[6] + (fx64)(1 << (FX32_INT_SHIFT - 1))) >> FX32_INT_SHIFT;
     //matrix determinant
-    det = ((s64)mtx->_[0] * det0 - (s64)mtx->_[1] * det1 + (s64)mtx->_[2] * det2 + (s64)(1 << (FX32_INT_SHIFT - 1))) >> FX32_INT_SHIFT;
+    det = ((fx64)mtx->_[0] * det0 - (fx64)mtx->_[1] * det1 + (fx64)mtx->_[2] * det2 + (fx64)(1 << (FX32_INT_SHIFT - 1))) >> FX32_INT_SHIFT;
 
     if (det == 0)
         return -1; //not invertible
 
     FX_InvAsync(det);
 
-    var0 = ((s64)mtx->_[1] * mtx->_[8] - (s64)mtx->_[7] * mtx->_[2]) >> FX32_INT_SHIFT;
-    var1 = ((s64)mtx->_[1] * mtx->_[5] - (s64)mtx->_[4] * mtx->_[2]) >> FX32_INT_SHIFT;
-    var2 = ((s64)mtx->_[0] * mtx->_[8] - (s64)mtx->_[6] * mtx->_[2]) >> FX32_INT_SHIFT;
-    var3 = ((s64)mtx->_[0] * mtx->_[5] - (s64)mtx->_[3] * mtx->_[2]) >> FX32_INT_SHIFT;
-
-    s32 ret = FX_GetDivResult();
-    dst->_[0] =  (s32)(((s64)ret * det0) >> FX32_INT_SHIFT);
-    dst->_[1] = -(s32)(((s64)ret * var0) >> FX32_INT_SHIFT);
-    dst->_[2] =  (s32)(((s64)ret * var1) >> FX32_INT_SHIFT);
-    dst->_[3] = -(s32)(((s64)ret * det1) >> FX32_INT_SHIFT);
-    dst->_[4] =  (s32)(((s64)ret * var2) >> FX32_INT_SHIFT);
-    dst->_[5] = -(s32)(((s64)ret * var3) >> FX32_INT_SHIFT);
-
-    dst->_[6] = (s32)(((s64)ret * det2) >> FX32_INT_SHIFT);
-    s32 temp = (s32)(((s64)mtx->_[0] * mtx->_[7] - (s64)mtx->_[6] * mtx->_[1]) >> FX32_INT_SHIFT);
-    dst->_[7] = -(s32)(((s64)ret * temp) >> FX32_INT_SHIFT);
-    s32 temp1 = (s32)(((s64)mtx->_[0] * mtx->_[4] - (s64)mtx->_[3] * mtx->_[1]) >> FX32_INT_SHIFT);
-    dst->_[8] = (s32)(((s64)ret * temp1) >> FX32_INT_SHIFT);
-    dst->_[9] =  -(s32)(((s64)dst->_[0] * mtx->_[9] + (s64)dst->_[3] * mtx->_[10] + (s64)dst->_[6] * mtx->_[11]) >> FX32_INT_SHIFT);
-    dst->_[10] = -(s32)(((s64)dst->_[1] * mtx->_[9] + (s64)dst->_[4] * mtx->_[10] + (s64)dst->_[7] * mtx->_[11]) >> FX32_INT_SHIFT);
-    dst->_[11] = -(s32)(((s64)dst->_[2] * mtx->_[9] + (s64)dst->_[5] * mtx->_[10] + (s64)dst->_[8] * mtx->_[11]) >> FX32_INT_SHIFT);
+    var0 = ((fx64)mtx->_[1] * mtx->_[8] - (fx64)mtx->_[7] * mtx->_[2]) >> FX32_INT_SHIFT;
+    var1 = ((fx64)mtx->_[1] * mtx->_[5] - (fx64)mtx->_[4] * mtx->_[2]) >> FX32_INT_SHIFT;
+    var2 = ((fx64)mtx->_[0] * mtx->_[8] - (fx64)mtx->_[6] * mtx->_[2]) >> FX32_INT_SHIFT;
+    var3 = ((fx64)mtx->_[0] * mtx->_[5] - (fx64)mtx->_[3] * mtx->_[2]) >> FX32_INT_SHIFT;
+
+    fx32 ret = FX_GetDivResult();
+    dst->_[0] =  (fx32)(((fx64)ret * det0) >> FX32_INT_SHIFT);
+    dst->_[1] = -(fx32)(((fx64)ret * var0) >> FX32_INT_SHIFT);
+    dst->_[2] =  (fx32)(((fx64)ret * var1) >> FX32_INT_SHIFT);
+    dst->_[3] = -(fx32)(((fx64)ret * det1) >> FX32_INT_SHIFT);
+    dst->_[4] =  (fx32)(((fx64)ret * var2) >> FX32_INT_SHIFT);
+    dst->_[5] = -(fx32)(((fx64)ret * var3) >> FX32_INT_SHIFT);
+
+    dst->_[6] = (fx32)(((fx64)ret * det2) >> FX32_INT_SHIFT);
+    fx32 temp = (fx32)(((fx64)mtx->_[0] * mtx->_[7] - (fx64)mtx->_[6] * mtx->_[1]) >> FX32_INT_SHIFT);
+    dst->_[7] = -(fx32)(((fx64)ret * temp) >> FX32_INT_SHIFT);
+    fx32 temp1 = (fx32)(((fx64)mtx->_[0] * mtx->_[4] - (fx64)mtx->_[3] * mtx->_[1]) >> FX32_INT_SHIFT);
+    dst->_[8] = (fx32)(((fx64)ret * temp1) >> FX32_INT_SHIFT);
+    dst->_[9] =  -(fx32)(((fx64)dst->_[0] * mtx->_[9] + (fx64)dst->_[3] * mtx->_[10] + (fx64)dst->_[6] * mtx->_[11]) >> FX32_INT_SHIFT);
+    dst->_[10] = -(fx32)(((fx64)dst->_[1] * mtx->_[9] + (fx64)dst->_[4] * mtx->_[10] + (fx64)dst->_[7] * mtx->_[11]) >> FX32_INT_SHIFT);
+    dst->_[11] = -(fx32)(((fx64)dst->_[2] * mtx->_[9] + (fx64)dst->_[5] * mtx->_[10] + (fx64)dst->_[8] * mtx->_[11]) >> FX32_INT_SHIFT);
 
     if (dst == &tempmat)
         MI_Copy48B(&tempmat, inv);
@@ -62,8 +62,8 @@ s32 MTX_Inverse43(struct Mtx43 *mtx, struct Mtx43 *inv){
 void MTX_Concat43(struct Mtx43 *a, struct Mtx43 *b, struct Mtx43 *c){
     struct Mtx43 temp;
     struct Mtx43 *dst;
-    s32 a0, a1, a2;
-    s32 b0, b1, b2;
+    fx32 a0, a1, a2;
+    fx32 b0, b1, b2;
 
     if (c == b)
         dst = &temp;
@@ -73,50 +73,50 @@ void MTX_Concat43(struct Mtx43 *a, struct Mtx43 *b, struct Mtx43 *c){
     a0 = a->_[0];
     a1 = a->_[1];
     a2 = a->_[2];
-    dst->_[0] =  (((s64)a0  * b->_[0] + (s64)a1  * b->_[3] + (s64)a2  * b->_[6] ) >> FX32_INT_SHIFT);
-    dst->_[1] =  (((s64)a0  * b->_[1] + (s64)a1  * b->_[4] + (s64)a2  * b->_[7] ) >> FX32_INT_SHIFT);
+    dst->_[0] =  (((fx64)a0  * b->_[0] + (fx64)a1  * b->_[3] + (fx64)a2  * b->_[6] ) >> FX32_INT_SHIFT);
+    dst->_[1] =  (((fx64)a0  * b->_[1] + (fx64)a1  * b->_[4] + (fx64)a2  * b->_[7] ) >> FX32_INT_SHIFT);
     b0 = b->_[2];
     b1 = b->_[5];
     b2 = b->_[8];
-    dst->_[2] =  (((s64)a0  * b0 + (s64)a1  * b1 + (s64)a2  * b2) >> FX32_INT_SHIFT);
+    dst->_[2] =  (((fx64)a0  * b0 + (fx64)a1  * b1 + (fx64)a2  * b2) >> FX32_INT_SHIFT);
     a0 = a->_[3];
     a1 = a->_[4];
     a2 = a->_[5];
-    dst->_[5] =  (((s64)a0  * b0 + (s64)a1  * b1 + (s64)a2  * b2) >> FX32_INT_SHIFT);
-    dst->_[4] =  (((s64)a0  * b->_[1] + (s64)a1  * b->_[4] + (s64)a2  * b->_[7] ) >> FX32_INT_SHIFT);
+    dst->_[5] =  (((fx64)a0  * b0 + (fx64)a1  * b1 + (fx64)a2  * b2) >> FX32_INT_SHIFT);
+    dst->_[4] =  (((fx64)a0  * b->_[1] + (fx64)a1  * b->_[4] + (fx64)a2  * b->_[7] ) >> FX32_INT_SHIFT);
     b0 = b->_[0];
     b1 = b->_[3];
     b2 = b->_[6];
-    dst->_[3] =  (((s64)a0  * b0 + (s64)a1  * b1 + (s64)a2  * b2) >> FX32_INT_SHIFT);
+    dst->_[3] =  (((fx64)a0  * b0 + (fx64)a1  * b1 + (fx64)a2  * b2) >> FX32_INT_SHIFT);
     a0 = a->_[6];
     a1 = a->_[7];
     a2 = a->_[8];
-    dst->_[6] =  (((s64)a0  * b0 + (s64)a1  * b1 + (s64)a2 * b2) >> FX32_INT_SHIFT);
-    dst->_[7] =  (((s64)a0  * b->_[1] + (s64)a1  * b->_[4] + (s64)a2 * b->_[7] ) >> FX32_INT_SHIFT);
+    dst->_[6] =  (((fx64)a0  * b0 + (fx64)a1  * b1 + (fx64)a2 * b2) >> FX32_INT_SHIFT);
+    dst->_[7] =  (((fx64)a0  * b->_[1] + (fx64)a1  * b->_[4] + (fx64)a2 * b->_[7] ) >> FX32_INT_SHIFT);
     b0 = b->_[2];
     b1 = b->_[5];
     b2 = b->_[8];
-    dst->_[8] = (((s64)a0  * b0 + (s64)a1  * b1 + (s64)a2 * b2) >> FX32_INT_SHIFT);
+    dst->_[8] = (((fx64)a0  * b0 + (fx64)a1  * b1 + (fx64)a2 * b2) >> FX32_INT_SHIFT);
     a0 = a->_[9];
     a1 = a->_[10];
     a2 = a->_[11];
-    dst->_[11] = ((((s64)a0 * b0 + (s64)a1 * b1 + (s64)a2 * b2) >> FX32_INT_SHIFT) + b->_[11]);
-    dst->_[10] = ((((s64)a0 * b->_[1] + (s64)a1 * b->_[4] + (s64)a2 * b->_[7]) >> FX32_INT_SHIFT) + b->_[10]);
-    dst->_[9] = ((((s64)a0 * b->_[0] + (s64)a1 * b->_[3] + (s64)a2 * b->_[6]) >> FX32_INT_SHIFT) + b->_[9]);
+    dst->_[11] = ((((fx64)a0 * b0 + (fx64)a1 * b1 + (fx64)a2 * b2) >> FX32_INT_SHIFT) + b->_[11]);
+    dst->_[10] = ((((fx64)a0 * b->_[1] + (fx64)a1 * b->_[4] + (fx64)a2 * b->_[7]) >> FX32_INT_SHIFT) + b->_[10]);
+    dst->_[9] = ((((fx64)a0 * b->_[0] + (fx64)a1 * b->_[3] + (fx64)a2 * b->_[6]) >> FX32_INT_SHIFT) + b->_[9]);
     if (dst == &temp)
         *c = temp;
 }
 
 void MTX_MultVec43(struct Vecx32 *vec, struct Mtx43 *mtx, struct Vecx32 *dst){
-    s32 x, y, z;
+    fx32 x, y, z;
     x = vec->x;
     y = vec->y;
     z = vec->z;
-    dst->x = ((s64)x * mtx->_[0] + (s64)y * mtx->_[3] + (s64)z * mtx->_[6]) >> FX32_INT_SHIFT;
+    dst->x = ((fx64)x * mtx->_[0] + (fx64)y * mtx->_[3] + (fx64)z * mtx->_[6]) >> FX32_INT_SHIFT;
     dst->x += mtx->_[9];
-    dst->y = ((s64)x * mtx->_[1] + (s64)y * mtx->_[4] + (s64)z * mtx->_[7]) >> FX32_INT_SHIFT;
+    dst->y = ((fx64)x * mtx->_[1] + (fx64)y * mtx->_[4] + (fx64)z * mtx->_[7]) >> FX32_INT_SHIFT;
     dst->y += mtx->_[10];
-    dst->z = ((s64)x * mtx->_[2] + (s64)y * mtx->_[5] + (s64)z * mtx->_[8]) >> FX32_INT_SHIFT;
+    dst->z = ((fx64)x * mtx->_[2] + (fx64)y * mtx->_[5] + (fx64)z * mtx->_[8]) >> FX32_INT_SHIFT;
     dst->z += mtx->_[11];
 }
 
@@ -150,7 +150,7 @@ asm void MTX_Copy43To44_(struct Mtx43 *src, struct Mtx44 *dst){
 }
 
 #pragma thumb on
-asm void MTX_Scale43_(struct Mtx43 *dst, s32 x, s32 y, s32 z){
+asm void MTX_Scale43_(struct Mtx43 *dst, fx32 x, fx32 y, fx32 z){
     stmia r0!, {r1}
     mov r1, #0x0
     str r3, [r0, #0x1c]
@@ -166,7 +166,7 @@ asm void MTX_Scale43_(struct Mtx43 *dst, s32 x, s32 y, s32 z){
 #pragma thumb off
 
 #pragma thumb on
-asm void MTX_RotX43_(struct Mtx43 *mtx, s32 sinphi, s32 cosphi){
+asm void MTX_RotX43_(struct Mtx43 *mtx, fx32 sinphi, fx32 cosphi){
     str r1, [r0, #0x14]
 	neg r1, r1
 	str r1, [r0, #0x1c]
@@ -186,7 +186,7 @@ asm void MTX_RotX43_(struct Mtx43 *mtx, s32 sinphi, s32 cosphi){
 #pragma thumb off
 
 #pragma thumb on
-asm void MTX_RotY43_(struct Mtx43 *mtx, s32 sinphi, s32 cosphi){
+asm void MTX_RotY43_(struct Mtx43 *mtx, fx32 sinphi, fx32 cosphi){
     str r1, [r0, #0x18]
 	mov r3, #0x0
 	stmia r0!, {r2-r3}
diff --git a/arm9/lib/src/FX_mtx44.c b/arm9/lib/src/FX_mtx44.c
index 3c91b4df..8d158602 100644
--- a/arm9/lib/src/FX_mtx44.c
+++ b/arm9/lib/src/FX_mtx44.c
@@ -5,20 +5,20 @@
 void MI_Copy48B(void *src, void *dst);
 
 
-void MTX_TransApply44(struct Mtx44 *mtx, struct Mtx44 *dst, s32 x, s32 y, s32 z){
+void MTX_TransApply44(struct Mtx44 *mtx, struct Mtx44 *dst, fx32 x, fx32 y, fx32 z){
     if(mtx != dst)
         MI_Copy48B(mtx, dst);
-    dst->_[12] = mtx->_[12] + (s32)(((s64)x * mtx->_[0] + (s64)y * mtx->_[4] + (s64)z * mtx->_[8] ) >> FX32_INT_SHIFT);
-    dst->_[13] = mtx->_[13] + (s32)(((s64)x * mtx->_[1] + (s64)y * mtx->_[5] + (s64)z * mtx->_[9] ) >> FX32_INT_SHIFT);
-    dst->_[14] = mtx->_[14] + (s32)(((s64)x * mtx->_[2] + (s64)y * mtx->_[6] + (s64)z * mtx->_[10]) >> FX32_INT_SHIFT);
-    dst->_[15] = mtx->_[15] + (s32)(((s64)x * mtx->_[3] + (s64)y * mtx->_[7] + (s64)z * mtx->_[11]) >> FX32_INT_SHIFT);
+    dst->_[12] = mtx->_[12] + (fx32)(((fx64)x * mtx->_[0] + (fx64)y * mtx->_[4] + (fx64)z * mtx->_[8] ) >> FX32_INT_SHIFT);
+    dst->_[13] = mtx->_[13] + (fx32)(((fx64)x * mtx->_[1] + (fx64)y * mtx->_[5] + (fx64)z * mtx->_[9] ) >> FX32_INT_SHIFT);
+    dst->_[14] = mtx->_[14] + (fx32)(((fx64)x * mtx->_[2] + (fx64)y * mtx->_[6] + (fx64)z * mtx->_[10]) >> FX32_INT_SHIFT);
+    dst->_[15] = mtx->_[15] + (fx32)(((fx64)x * mtx->_[3] + (fx64)y * mtx->_[7] + (fx64)z * mtx->_[11]) >> FX32_INT_SHIFT);
 }
 
 void MTX_Concat44(struct Mtx44 *a, struct Mtx44 *b, struct Mtx44 *c){
     struct Mtx44 temp;
     struct Mtx44 *dst;
-    s32 a0, a1, a2, a3;
-    s32 b0, b1, b2, b3;
+    fx32 a0, a1, a2, a3;
+    fx32 b0, b1, b2, b3;
 
     if (c == b)
         dst = &temp;
@@ -29,46 +29,46 @@ void MTX_Concat44(struct Mtx44 *a, struct Mtx44 *b, struct Mtx44 *c){
     a1 = a->_[1];
     a2 = a->_[2];
     a3 = a->_[3];
-    dst->_[0] =  (((s64)a0  * b->_[0] + (s64)a1  * b->_[4] + (s64)a2  * b->_[8]  + (s64)a3  * b->_[12]) >> FX32_INT_SHIFT);
-    dst->_[1] =  (((s64)a0  * b->_[1] + (s64)a1  * b->_[5] + (s64)a2  * b->_[9]  + (s64)a3  * b->_[13]) >> FX32_INT_SHIFT);
-    dst->_[3] =  (((s64)a0  * b->_[3] + (s64)a1  * b->_[7] + (s64)a2  * b->_[11] + (s64)a3  * b->_[15]) >> FX32_INT_SHIFT);
+    dst->_[0] =  (((fx64)a0  * b->_[0] + (fx64)a1  * b->_[4] + (fx64)a2  * b->_[8]  + (fx64)a3  * b->_[12]) >> FX32_INT_SHIFT);
+    dst->_[1] =  (((fx64)a0  * b->_[1] + (fx64)a1  * b->_[5] + (fx64)a2  * b->_[9]  + (fx64)a3  * b->_[13]) >> FX32_INT_SHIFT);
+    dst->_[3] =  (((fx64)a0  * b->_[3] + (fx64)a1  * b->_[7] + (fx64)a2  * b->_[11] + (fx64)a3  * b->_[15]) >> FX32_INT_SHIFT);
     b0 = b->_[2];
     b1 = b->_[6];
     b2 = b->_[10];
     b3 = b->_[14];
-    dst->_[2] =  (((s64)a0  * b0 + (s64)a1  * b1 + (s64)a2  * b2 + (s64)a3  * b3) >> FX32_INT_SHIFT);
+    dst->_[2] =  (((fx64)a0  * b0 + (fx64)a1  * b1 + (fx64)a2  * b2 + (fx64)a3  * b3) >> FX32_INT_SHIFT);
     a0 = a->_[4];
     a1 = a->_[5];
     a2 = a->_[6];
     a3 = a->_[7];
-    dst->_[6] =  (((s64)a0  * b0 + (s64)a1  * b1 + (s64)a2  * b2 + (s64)a3  * b3) >> FX32_INT_SHIFT);
-    dst->_[5] =  (((s64)a0  * b->_[1] + (s64)a1  * b->_[5] + (s64)a2  * b->_[9]  + (s64)a3  * b->_[13]) >> FX32_INT_SHIFT);
-    dst->_[7] =  (((s64)a0  * b->_[3] + (s64)a1  * b->_[7] + (s64)a2  * b->_[11] + (s64)a3  * b->_[15]) >> FX32_INT_SHIFT);
+    dst->_[6] =  (((fx64)a0  * b0 + (fx64)a1  * b1 + (fx64)a2  * b2 + (fx64)a3  * b3) >> FX32_INT_SHIFT);
+    dst->_[5] =  (((fx64)a0  * b->_[1] + (fx64)a1  * b->_[5] + (fx64)a2  * b->_[9]  + (fx64)a3  * b->_[13]) >> FX32_INT_SHIFT);
+    dst->_[7] =  (((fx64)a0  * b->_[3] + (fx64)a1  * b->_[7] + (fx64)a2  * b->_[11] + (fx64)a3  * b->_[15]) >> FX32_INT_SHIFT);
     b0 = b->_[0];
     b1 = b->_[4];
     b2 = b->_[8];
     b3 = b->_[12];
-    dst->_[4] =  (((s64)a0  * b0 + (s64)a1  * b1 + (s64)a2  * b2  + (s64)a3  * b3) >> FX32_INT_SHIFT);
+    dst->_[4] =  (((fx64)a0  * b0 + (fx64)a1  * b1 + (fx64)a2  * b2  + (fx64)a3  * b3) >> FX32_INT_SHIFT);
     a0 = a->_[8];
     a1 = a->_[9];
     a2 = a->_[10];
     a3 = a->_[11];
-    dst->_[8] =  (((s64)a0  * b0 + (s64)a1  * b1 + (s64)a2 * b2  + (s64)a3 * b3) >> FX32_INT_SHIFT);
-    dst->_[9] =  (((s64)a0  * b->_[1] + (s64)a1  * b->_[5] + (s64)a2 * b->_[9]  + (s64)a3 * b->_[13]) >> FX32_INT_SHIFT);
-    dst->_[11] = (((s64)a0  * b->_[3] + (s64)a1  * b->_[7] + (s64)a2 * b->_[11] + (s64)a3 * b->_[15]) >> FX32_INT_SHIFT);
+    dst->_[8] =  (((fx64)a0  * b0 + (fx64)a1  * b1 + (fx64)a2 * b2  + (fx64)a3 * b3) >> FX32_INT_SHIFT);
+    dst->_[9] =  (((fx64)a0  * b->_[1] + (fx64)a1  * b->_[5] + (fx64)a2 * b->_[9]  + (fx64)a3 * b->_[13]) >> FX32_INT_SHIFT);
+    dst->_[11] = (((fx64)a0  * b->_[3] + (fx64)a1  * b->_[7] + (fx64)a2 * b->_[11] + (fx64)a3 * b->_[15]) >> FX32_INT_SHIFT);
     b0 = b->_[2];
     b1 = b->_[6];
     b2 = b->_[10];
     b3 = b->_[14];
-    dst->_[10] = (((s64)a0  * b0 + (s64)a1  * b1 + (s64)a2 * b2 + (s64)a3 * b3) >> FX32_INT_SHIFT);
+    dst->_[10] = (((fx64)a0  * b0 + (fx64)a1  * b1 + (fx64)a2 * b2 + (fx64)a3 * b3) >> FX32_INT_SHIFT);
     a0 = a->_[12];
     a1 = a->_[13];
     a2 = a->_[14];
     a3 = a->_[15];
-    dst->_[14] = (((s64)a0 * b0 + (s64)a1 * b1 + (s64)a2 * b2 + (s64)a3 * b3) >> FX32_INT_SHIFT);
-    dst->_[13] = (((s64)a0 * b->_[1] + (s64)a1 * b->_[5] + (s64)a2 * b->_[9]  + (s64)a3 * b->_[13]) >> FX32_INT_SHIFT);
-    dst->_[12] = (((s64)a0 * b->_[0] + (s64)a1 * b->_[4] + (s64)a2 * b->_[8]  + (s64)a3 * b->_[12]) >> FX32_INT_SHIFT);
-    dst->_[15] = (((s64)a0 * b->_[3] + (s64)a1 * b->_[7] + (s64)a2 * b->_[11] + (s64)a3 * b->_[15]) >> FX32_INT_SHIFT);
+    dst->_[14] = (((fx64)a0 * b0 + (fx64)a1 * b1 + (fx64)a2 * b2 + (fx64)a3 * b3) >> FX32_INT_SHIFT);
+    dst->_[13] = (((fx64)a0 * b->_[1] + (fx64)a1 * b->_[5] + (fx64)a2 * b->_[9]  + (fx64)a3 * b->_[13]) >> FX32_INT_SHIFT);
+    dst->_[12] = (((fx64)a0 * b->_[0] + (fx64)a1 * b->_[4] + (fx64)a2 * b->_[8]  + (fx64)a3 * b->_[12]) >> FX32_INT_SHIFT);
+    dst->_[15] = (((fx64)a0 * b->_[3] + (fx64)a1 * b->_[7] + (fx64)a2 * b->_[11] + (fx64)a3 * b->_[15]) >> FX32_INT_SHIFT);
     if (dst == &temp)
         *c = temp;
 }
@@ -104,7 +104,7 @@ asm void MTX_Copy44To43_(struct Mtx44 *src, struct Mtx43 *dst){
 }
 
 #pragma thumb on
-asm void MTX_RotX44_(struct Mtx44 *mtx, s32 sinphi, s32 cosphi){
+asm void MTX_RotX44_(struct Mtx44 *mtx, fx32 sinphi, fx32 cosphi){
     str r2, [r0, #0x14]
 	str r2, [r0, #0x28]
 	str r1, [r0, #0x18]
@@ -127,7 +127,7 @@ asm void MTX_RotX44_(struct Mtx44 *mtx, s32 sinphi, s32 cosphi){
 #pragma thumb off
 
 #pragma thumb on
-asm void MTX_RotY44_(struct Mtx44 *mtx, s32 sinphi, s32 cosphi){
+asm void MTX_RotY44_(struct Mtx44 *mtx, fx32 sinphi, fx32 cosphi){
     str r2, [r0, #0x0]
 	str r2, [r0, #0x28]
 	str r1, [r0, #0x20]
@@ -150,7 +150,7 @@ asm void MTX_RotY44_(struct Mtx44 *mtx, s32 sinphi, s32 cosphi){
 #pragma thumb off
 
 #pragma thumb on
-asm void MTX_RotZ44_(struct Mtx44 *mtx, s32 sinphi, s32 cosphi){
+asm void MTX_RotZ44_(struct Mtx44 *mtx, fx32 sinphi, fx32 cosphi){
     str r2, [r0, #0x0]
     str r2, [r0, #0x14]
     str r1, [r0, #0x4]
diff --git a/arm9/lib/src/FX_vec.c b/arm9/lib/src/FX_vec.c
index 7b838829..10a792be 100644
--- a/arm9/lib/src/FX_vec.c
+++ b/arm9/lib/src/FX_vec.c
@@ -20,29 +20,29 @@ void VEC_Fx16Add(struct Vecx16 *a, struct Vecx16 *b, struct Vecx16 *dst){
     dst->z = a->z + b->z;
 }
 
-s32 VEC_DotProduct(struct Vecx32 *a, struct Vecx32 *b){
-    return ((s64)a->x * b->x + (s64)a->y * b->y + (s64)a->z * b->z + (1 << (FX64_INT_SHIFT - 1))) >> FX64_INT_SHIFT;
+fx32 VEC_DotProduct(struct Vecx32 *a, struct Vecx32 *b){
+    return ((fx64)a->x * b->x + (fx64)a->y * b->y + (fx64)a->z * b->z + (1 << (FX64_INT_SHIFT - 1))) >> FX64_INT_SHIFT;
 }
 
-s32 VEC_Fx16DotProduct(struct Vecx16 *a, struct Vecx16 *b){
-    s32 temp1, temp2;
+fx32 VEC_Fx16DotProduct(struct Vecx16 *a, struct Vecx16 *b){
+    fx32 temp1, temp2;
     temp1 = (a->x * b->x) + (a->y * b->y);
     temp2 = (a->z * b->z) + (1 << (FX64_INT_SHIFT - 1));
-    return (s32)(((s64)temp1 + temp2) >> FX64_INT_SHIFT);
+    return (fx32)(((fx64)temp1 + temp2) >> FX64_INT_SHIFT);
 }
 
 void VEC_CrossProduct(struct Vecx32 *a, struct Vecx32 *b, struct Vecx32 *dst){
-    s32 x, y, z;
-    x = (s32)(((s64)a->y * b->z - (s64)a->z * b->y + (1 << (FX64_INT_SHIFT - 1))) >> FX64_INT_SHIFT);
-    y = (s32)(((s64)a->z * b->x - (s64)a->x * b->z + (1 << (FX64_INT_SHIFT - 1))) >> FX64_INT_SHIFT);
-    z = (s32)(((s64)a->x * b->y - (s64)a->y * b->x + (1 << (FX64_INT_SHIFT - 1))) >> FX64_INT_SHIFT);
+    fx32 x, y, z;
+    x = (fx32)(((fx64)a->y * b->z - (fx64)a->z * b->y + (1 << (FX64_INT_SHIFT - 1))) >> FX64_INT_SHIFT);
+    y = (fx32)(((fx64)a->z * b->x - (fx64)a->x * b->z + (1 << (FX64_INT_SHIFT - 1))) >> FX64_INT_SHIFT);
+    z = (fx32)(((fx64)a->x * b->y - (fx64)a->y * b->x + (1 << (FX64_INT_SHIFT - 1))) >> FX64_INT_SHIFT);
     dst->x = x;
     dst->y = y;
     dst->z = z;
 }
 
 void VEC_Fx16CrossProduct(struct Vecx16 *a, struct Vecx16 *b, struct Vecx16 *dst){
-    s32 x, y, z;
+    fx32 x, y, z;
     x = ((a->y * b->z - a->z * b->y + (1 << (FX64_INT_SHIFT - 1))) >> FX64_INT_SHIFT);
     y = ((a->z * b->x - a->x * b->z + (1 << (FX64_INT_SHIFT - 1))) >> FX64_INT_SHIFT);
     z = ((a->x * b->y - a->y * b->x + (1 << (FX64_INT_SHIFT - 1))) >> FX64_INT_SHIFT);
@@ -51,20 +51,20 @@ void VEC_Fx16CrossProduct(struct Vecx16 *a, struct Vecx16 *b, struct Vecx16 *dst
     dst->z = z;
 }
 
-s32 VEC_Mag(struct Vecx32 *a){
-    s64 l2 = (s64)a->x * a->x;
-    l2 += (s64)a->y * a->y;
-    l2 += (s64)a->z * a->z;
+fx32 VEC_Mag(struct Vecx32 *a){
+    fx64 l2 = (fx64)a->x * a->x;
+    l2 += (fx64)a->y * a->y;
+    l2 += (fx64)a->z * a->z;
     SETREG16(HW_REG_SQRTCNT, 0x1);
     SETREG64(HW_REG_SQRT_PARAM, l2 * 4);
     while (READREG16(HW_REG_SQRTCNT) & 0x8000); //wait for coprocessor to finish
-    return ((s32)READREG32(HW_REG_SQRT_RESULT) + 1) >> 1;
+    return ((fx32)READREG32(HW_REG_SQRT_RESULT) + 1) >> 1;
 }
 
 void VEC_Normalize(struct Vecx32 *a, struct Vecx32 *dst){
-    s64 l2 = (s64)a->x * a->x;
-    l2 += (s64)a->y * a->y;
-    l2 += (s64)a->z * a->z;
+    fx64 l2 = (fx64)a->x * a->x;
+    l2 += (fx64)a->y * a->y;
+    l2 += (fx64)a->z * a->z;
     //1/sqrt(l) is computed by calculating sqrt(l)*(1/l)
     SETREG16(HW_REG_DIVCNT, 0x2);
     SETREG64(HW_REG_DIV_NUMER, 0x0100000000000000);
@@ -72,7 +72,7 @@ void VEC_Normalize(struct Vecx32 *a, struct Vecx32 *dst){
     SETREG16(HW_REG_SQRTCNT, 0x1);
     SETREG64(HW_REG_SQRT_PARAM, l2 * 4);
     while (READREG16(HW_REG_SQRTCNT) & 0x8000); //wait for sqrt to finish
-    s32 sqrtresult = READREG32(HW_REG_SQRT_RESULT);
+    fx32 sqrtresult = READREG32(HW_REG_SQRT_RESULT);
     while (READREG16(HW_REG_DIVCNT) & 0x8000); //wait for division to finish
     l2 = READREG64(HW_REG_DIV_RESULT);
     l2 = sqrtresult * l2;
@@ -82,7 +82,7 @@ void VEC_Normalize(struct Vecx32 *a, struct Vecx32 *dst){
 }
 
 void VEC_Fx16Normalize(struct Vecx16 *a, struct Vecx16 *dst){
-    s64 l2 = a->x * a->x;
+    fx64 l2 = a->x * a->x;
     l2 += a->y * a->y;
     l2 += a->z * a->z;
     //1/sqrt(l) is computed by calculating sqrt(l)*(1/l)
@@ -92,7 +92,7 @@ void VEC_Fx16Normalize(struct Vecx16 *a, struct Vecx16 *dst){
     SETREG16(HW_REG_SQRTCNT, 0x1);
     SETREG64(HW_REG_SQRT_PARAM, l2 * 4);
     while (READREG16(HW_REG_SQRTCNT) & 0x8000); //wait for sqrt to finish
-    s32 sqrtresult = READREG32(HW_REG_SQRT_RESULT);
+    fx32 sqrtresult = READREG32(HW_REG_SQRT_RESULT);
     while (READREG16(HW_REG_DIVCNT) & 0x8000); //wait for division to finish
     l2 = READREG64(HW_REG_DIV_RESULT);
     l2 = sqrtresult * l2;
@@ -101,8 +101,8 @@ void VEC_Fx16Normalize(struct Vecx16 *a, struct Vecx16 *dst){
     dst->z = (l2 * a->z + (1LL << (0x2D - 1))) >> 0x2D;
 }
 
-void VEC_MultAdd(s32 factor, struct Vecx32  *a, struct Vecx32 *b, struct Vecx32 *dst){
-    dst->x = (s32)(((s64)factor * a->x) >> FX32_INT_SHIFT) + b->x;
-    dst->y = (s32)(((s64)factor * a->y) >> FX32_INT_SHIFT) + b->y;
-    dst->z = (s32)(((s64)factor * a->z) >> FX32_INT_SHIFT) + b->z;
+void VEC_MultAdd(fx32 factor, struct Vecx32  *a, struct Vecx32 *b, struct Vecx32 *dst){
+    dst->x = (fx32)(((fx64)factor * a->x) >> FX32_INT_SHIFT) + b->x;
+    dst->y = (fx32)(((fx64)factor * a->y) >> FX32_INT_SHIFT) + b->y;
+    dst->z = (fx32)(((fx64)factor * a->z) >> FX32_INT_SHIFT) + b->z;
 }
-- 
cgit v1.2.3


From 155f50c9890a93fbc191e078d304c161809e8fc6 Mon Sep 17 00:00:00 2001
From: Made <made111@gmx.de>
Date: Mon, 4 May 2020 17:26:31 +0200
Subject: Move header and fix accidental name changes

---
 arm9/lib/include/fx.h | 176 ++++++++++++++++++++++++++++++++++++++++++++++++++
 arm9/lib/src/FX_cp.c  |   4 +-
 include/fx.h          | 176 --------------------------------------------------
 3 files changed, 178 insertions(+), 178 deletions(-)
 create mode 100644 arm9/lib/include/fx.h
 delete mode 100644 include/fx.h

diff --git a/arm9/lib/include/fx.h b/arm9/lib/include/fx.h
new file mode 100644
index 00000000..7e74d079
--- /dev/null
+++ b/arm9/lib/include/fx.h
@@ -0,0 +1,176 @@
+#ifndef GUARD_FX_H
+#define GUARD_FX_H
+
+typedef s16 fx16;
+typedef s32 fx32;
+typedef s64 fx64;
+typedef s64 fx64c;
+
+#define FX16_INT_MASK              0xF000
+#define FX16_INT_ABS_MASK          0x7000
+#define FX16_FRAC_MASK             0x0FFF
+#define FX16_INT_SHIFT             0xC
+
+#define FX32_INT_MASK              0xFFFFF000
+#define FX32_INT_ABS_MASK          0x7FFFF000
+#define FX32_FRAC_MASK             0x00000FFF
+#define FX32_INT_SHIFT             0xC
+
+#define FX64_INT_MASK              0xFFFFFFFFFFFFF000
+#define FX64_INT_ABS_MASK          0x7FFFFFFFFFFFF000
+#define FX64_FRAC_MASK             0x0000000000000FFF
+#define FX64_INT_SHIFT             0xC
+
+#define FX64C_INT_MASK             0xFFFFFFFF00000000
+#define FX64C_INT_ABS_MASK         0x7FFFFFFF00000000
+#define FX64C_FRAC_MASK            0x00000000FFFFFFFF
+#define FX64C_INT_SHIFT            0x20
+
+#define FX_INT(TYPE, x)            (((x) & TYPE ## _INT_MASK) >> TYPE ## _INT_SHIFT)
+#define FX_INT_ABS(TYPE, x)        (((x) & TYPE ## _INT_ABS_MASK) >> TYPE ## _INT_SHIFT)
+#define FX_FRAC(TYPE, x)           ((x) & TYPE ## _FRAC_MASK)
+
+#define FX16_INT(x)                FX_INT(FX16, x)
+#define FX16_INT_ABS(x)            FX_INT_ABS(FX16, x)
+#define FX16_FRAC(x)               FX_FRAC(FX16, x)
+
+#define FX32_INT(x)                FX_INT(FX32, x)
+#define FX32_INT_ABS(x)            FX_INT_ABS(FX32, x)
+#define FX32_FRAC(x)               FX_FRAC(FX32, x)
+
+#define FX64_INT(x)                FX_INT(FX64, x)
+#define FX64_INT_ABS(x)            FX_INT_ABS(FX64, x)
+#define FX64_FRAC(x)               FX_FRAC(FX64, x)
+
+#define FX64C_INT(x)               FX_INT(FX64C, x)
+#define FX64C_INT_ABS(x)           FX_INT_ABS(FX64C, x)
+#define FX64C_FRAC(x)              FX_FRAC(FX64C, x)
+
+
+#define HW_REG_DIVCNT              0x04000280
+#define HW_REG_DIV_NUMER           0x04000290
+#define HW_REG_DIV_DENOM           0x04000298
+#define HW_REG_DIV_RESULT          0x040002A0
+#define HW_REG_DIVREM_RESULT       0x040002A8
+
+#define HW_REG_SQRTCNT             0x040002B0
+#define HW_REG_SQRT_RESULT         0x040002B4
+#define HW_REG_SQRT_PARAM          0x040002B8
+
+#define SETREG16(x, y)             ((*(vu16 *)x) = y)
+#define SETREG32(x, y)             ((*(vu32 *)x) = y)
+#define SETREG64(x, y)             ((*(vu64 *)x) = y)
+#define READREG16(x)               (*(vu16 *)x)
+#define READREG32(x)               (*(vu32 *)x)
+#define READREG64(x)               (*(vu64 *)x)
+
+#define FX32_MUL(a, b)               ((fx32)(((fx64)a * b) >> FX32_INT_SHIFT))
+#define FX32_MUL_ADD_MUL(a, b, c, d) ((fx32)(((fx64)a * b + (fx64)c * d) >> FX32_INT_SHIFT))
+//the extra term here is for rounding
+#define FX32_MUL_SUB_MUL(a, b, c, d) ((fx32)(((fx64)a * b - (fx64)c * d + (1 << (FX32_INT_SHIFT - 1))) >> FX32_INT_SHIFT))
+
+struct Vecx32
+{
+    fx32 x;
+    fx32 y;
+    fx32 z;
+};
+
+struct Vecx16
+{
+    fx16 x;
+    fx16 y;
+    fx16 z;
+};
+
+//Matrices are indexed as [column][row]
+struct Mtx44
+{
+    fx32 _[16];
+};
+
+struct Mtx43
+{
+    fx32 _[12];
+};
+
+struct Mtx33
+{
+    fx32 _[9];
+};
+
+struct Mtx22
+{
+    fx32 _[4];
+};
+
+//FX
+void FX_Init();
+fx32 FX_Modf(fx32 x, fx32 *iptr);
+
+//Atan
+u16 FX_Atan(fx32 x);
+u16 FX_Atan2(fx32 x, fx32 y);
+
+//Vec
+void VEC_Add(struct Vecx32 *x, struct Vecx32 *y, struct Vecx32 *dst);
+void VEC_Subtract(struct Vecx32 *x, struct Vecx32 *y, struct Vecx32 *dst);
+void VEC_Fx16Add(struct Vecx16 *x, struct Vecx16 *y, struct Vecx16 *dst);
+fx32 VEC_DotProduct(struct Vecx32 *x, struct Vecx32 *y);
+fx32 VEC_Fx16DotProduct(struct Vecx16 *a, struct Vecx16 *b);
+void VEC_CrossProduct(struct Vecx32 *a, struct Vecx32 *b, struct Vecx32 *dst);
+void VEC_Fx16CrossProduct(struct Vecx16 *a, struct Vecx16 *b, struct Vecx16 *dst);
+fx32 VEC_Mag(struct Vecx32 *a);
+void VEC_Normalize(struct Vecx32 *a, struct Vecx32 *dst);
+void VEC_Fx16Normalize(struct Vecx16 *a, struct Vecx16 *dst);
+void VEC_MultAdd(fx32 factor, struct Vecx32  *a, struct Vecx32 *b, struct Vecx32 *dst);
+
+//CP
+fx32 FX_Div(fx32 numerator, fx32 denominator);
+fx32 FX_Inv(fx32 x);
+fx32 FX_Sqrt(fx32 x);
+fx64 FX_GetDivResultFx64c();
+fx32 FX_GetDivResult();
+void FX_InvAsync(fx32 x);
+fx32 FX_GetSqrtResult();
+void FX_DivAsync(fx32 numerator, fx32 denominator);
+fx32 FX_DivS32(fx32 numerator, fx32 denominator);
+fx32 FX_ModS32(fx32 num, fx32 mod);
+
+//Mtx
+//The functions ending in underscores seem to be in assembly originally
+//Mtx44
+void MTX_TransApply44(struct Mtx44 *mtx, struct Mtx44 *dst, fx32 x, fx32 y, fx32 z);
+void MTX_Concat44(struct Mtx44 *a, struct Mtx44 *b, struct Mtx44 *c);
+void MTX_Identity44_(struct Mtx44 *dst);
+void MTX_Copy44To43_(struct Mtx44 *src, struct Mtx43 *dst);
+void MTX_RotX44_(struct Mtx44 *mtx, fx32 sinphi, fx32 cosphi);
+void MTX_RotY44_(struct Mtx44 *mtx, fx32 sinphi, fx32 cosphi);
+void MTX_RotZ44_(struct Mtx44 *mtx, fx32 sinphi, fx32 cosphi);
+
+//Mtx43
+void MTX_ScaleApply33(struct Mtx33 *mtx, struct Mtx33 *dst, fx32 x, fx32 y, fx32 z);
+fx32 MTX_Inverse43(struct Mtx43 *mtx, struct Mtx43 *inv);
+void MTX_Concat43(struct Mtx43 *a, struct Mtx43 *b, struct Mtx43 *c);
+void MTX_MultVec43(struct Vecx32 *vec, struct Mtx43 *mtx, struct Vecx32 *dst);
+void MTX_Identity43_(struct Mtx43 *mtx);
+void MTX_Copy43To44_(struct Mtx43 *src, struct Mtx44 *dst);
+void MTX_Scale43_(struct Mtx43 *dst, fx32 x, fx32 y, fx32 z);
+void MTX_RotX43_(struct Mtx43 *mtx, fx32 sinphi, fx32 cosphi);
+void MTX_RotY43_(struct Mtx43 *mtx, fx32 sinphi, fx32 cosphi);
+
+//Mtx33
+void MTX_ScaleApply33(struct Mtx33 *mtx, struct Mtx33 *dst, fx32 x, fx32 y, fx32 z);
+void MTX_Concat33(struct Mtx33 *a, struct Mtx33 *b, struct Mtx33 *c);
+void MTX_MultVec33(struct Vecx32 *vec, struct Mtx33 *mtx, struct Vecx32 *dst);
+void MTX_Identity33_(struct Mtx33 *mtx);
+void MTX_RotX33_(struct Mtx33 *mtx, fx32 sinphi, fx32 cosphi);
+void MTX_RotY33_(struct Mtx33 *mtx, fx32 sinphi, fx32 cosphi);
+void MTX_RotZ33_(struct Mtx33 *mtx, fx32 sinphi, fx32 cosphi);
+
+//Mtx22
+void MTX_ScaleApply22(struct Mtx22 *mtx, struct Mtx22 *dst, fx32 x, fx32 y);
+void MTX_Identity22_(struct Mtx22 *mtx);
+void MTX_Rot22_(struct Mtx22 *mtx, fx32 sinphi, fx32 cosphi);
+
+#endif //GUARD_FX_H
diff --git a/arm9/lib/src/FX_cp.c b/arm9/lib/src/FX_cp.c
index 76dea4b4..a5a8d9cb 100644
--- a/arm9/lib/src/FX_cp.c
+++ b/arm9/lib/src/FX_cp.c
@@ -53,7 +53,7 @@ void FX_DivAsync(fx32 numerator, fx32 denominator){
     SETREG64(HW_REG_DIV_DENOM, (u32)denominator);
 }
 
-fx32 FX_Divfx32(fx32 numerator, fx32 denominator){
+fx32 FX_DivS32(fx32 numerator, fx32 denominator){
     SETREG16(HW_REG_DIVCNT, 0x0);
     SETREG32(HW_REG_DIV_NUMER, (u32)numerator); //32bit write for some reason
     SETREG64(HW_REG_DIV_DENOM, (u32)denominator);
@@ -61,7 +61,7 @@ fx32 FX_Divfx32(fx32 numerator, fx32 denominator){
     return READREG32(HW_REG_DIV_RESULT);
 }
 
-fx32 FX_Modfx32(fx32 num, fx32 mod){
+fx32 FX_ModS32(fx32 num, fx32 mod){
     SETREG16(HW_REG_DIVCNT, 0x0);
     SETREG32(HW_REG_DIV_NUMER, (u32)num); //32bit write for some reason
     SETREG64(HW_REG_DIV_DENOM, (u32)mod);
diff --git a/include/fx.h b/include/fx.h
deleted file mode 100644
index bbbf621a..00000000
--- a/include/fx.h
+++ /dev/null
@@ -1,176 +0,0 @@
-#ifndef GUARD_FX_H
-#define GUARD_FX_H
-
-typedef s16 fx16;
-typedef s32 fx32;
-typedef s64 fx64;
-typedef s64 fx64c;
-
-#define FX16_INT_MASK              0xF000
-#define FX16_INT_ABS_MASK          0x7000
-#define FX16_FRAC_MASK             0x0FFF
-#define FX16_INT_SHIFT             0xC
-
-#define FX32_INT_MASK              0xFFFFF000
-#define FX32_INT_ABS_MASK          0x7FFFF000
-#define FX32_FRAC_MASK             0x00000FFF
-#define FX32_INT_SHIFT             0xC
-
-#define FX64_INT_MASK              0xFFFFFFFFFFFFF000
-#define FX64_INT_ABS_MASK          0x7FFFFFFFFFFFF000
-#define FX64_FRAC_MASK             0x0000000000000FFF
-#define FX64_INT_SHIFT             0xC
-
-#define FX64C_INT_MASK             0xFFFFFFFF00000000
-#define FX64C_INT_ABS_MASK         0x7FFFFFFF00000000
-#define FX64C_FRAC_MASK            0x00000000FFFFFFFF
-#define FX64C_INT_SHIFT            0x20
-
-#define FX_INT(TYPE, x)            (((x) & TYPE ## _INT_MASK) >> TYPE ## _INT_SHIFT)
-#define FX_INT_ABS(TYPE, x)        (((x) & TYPE ## _INT_ABS_MASK) >> TYPE ## _INT_SHIFT)
-#define FX_FRAC(TYPE, x)           ((x) & TYPE ## _FRAC_MASK)
-
-#define FX16_INT(x)                FX_INT(FX16, x)
-#define FX16_INT_ABS(x)            FX_INT_ABS(FX16, x)
-#define FX16_FRAC(x)               FX_FRAC(FX16, x)
-
-#define FX32_INT(x)                FX_INT(FX32, x)
-#define FX32_INT_ABS(x)            FX_INT_ABS(FX32, x)
-#define FX32_FRAC(x)               FX_FRAC(FX32, x)
-
-#define FX64_INT(x)                FX_INT(FX64, x)
-#define FX64_INT_ABS(x)            FX_INT_ABS(FX64, x)
-#define FX64_FRAC(x)               FX_FRAC(FX64, x)
-
-#define FX64C_INT(x)               FX_INT(FX64C, x)
-#define FX64C_INT_ABS(x)           FX_INT_ABS(FX64C, x)
-#define FX64C_FRAC(x)              FX_FRAC(FX64C, x)
-
-
-#define HW_REG_DIVCNT              0x04000280
-#define HW_REG_DIV_NUMER           0x04000290
-#define HW_REG_DIV_DENOM           0x04000298
-#define HW_REG_DIV_RESULT          0x040002A0
-#define HW_REG_DIVREM_RESULT       0x040002A8
-
-#define HW_REG_SQRTCNT             0x040002B0
-#define HW_REG_SQRT_RESULT         0x040002B4
-#define HW_REG_SQRT_PARAM          0x040002B8
-
-#define SETREG16(x, y)             ((*(vu16 *)x) = y)
-#define SETREG32(x, y)             ((*(vu32 *)x) = y)
-#define SETREG64(x, y)             ((*(vu64 *)x) = y)
-#define READREG16(x)               (*(vu16 *)x)
-#define READREG32(x)               (*(vu32 *)x)
-#define READREG64(x)               (*(vu64 *)x)
-
-#define FX32_MUL(a, b)               ((fx32)(((fx64)a * b) >> FX32_INT_SHIFT))
-#define FX32_MUL_ADD_MUL(a, b, c, d) ((fx32)(((fx64)a * b + (fx64)c * d) >> FX32_INT_SHIFT))
-//the extra term here is for rounding
-#define FX32_MUL_SUB_MUL(a, b, c, d) ((fx32)(((fx64)a * b - (fx64)c * d + (1 << (FX32_INT_SHIFT - 1))) >> FX32_INT_SHIFT))
-
-struct Vecx32
-{
-    fx32 x;
-    fx32 y;
-    fx32 z;
-};
-
-struct Vecx16
-{
-    fx16 x;
-    fx16 y;
-    fx16 z;
-};
-
-//Matrices are indexed as [column][row]
-struct Mtx44
-{
-    fx32 _[16];
-};
-
-struct Mtx43
-{
-    fx32 _[12];
-};
-
-struct Mtx33
-{
-    fx32 _[9];
-};
-
-struct Mtx22
-{
-    fx32 _[4];
-};
-
-//FX
-void FX_Init();
-fx32 FX_Modf(fx32 x, fx32 *iptr);
-
-//Atan
-u16 FX_Atan(fx32 x);
-u16 FX_Atan2(fx32 x, fx32 y);
-
-//Vec
-void VEC_Add(struct Vecx32 *x, struct Vecx32 *y, struct Vecx32 *dst);
-void VEC_Subtract(struct Vecx32 *x, struct Vecx32 *y, struct Vecx32 *dst);
-void VEC_Fx16Add(struct Vecx16 *x, struct Vecx16 *y, struct Vecx16 *dst);
-fx32 VEC_DotProduct(struct Vecx32 *x, struct Vecx32 *y);
-fx32 VEC_Fx16DotProduct(struct Vecx16 *a, struct Vecx16 *b);
-void VEC_CrossProduct(struct Vecx32 *a, struct Vecx32 *b, struct Vecx32 *dst);
-void VEC_Fx16CrossProduct(struct Vecx16 *a, struct Vecx16 *b, struct Vecx16 *dst);
-fx32 VEC_Mag(struct Vecx32 *a);
-void VEC_Normalize(struct Vecx32 *a, struct Vecx32 *dst);
-void VEC_Fx16Normalize(struct Vecx16 *a, struct Vecx16 *dst);
-void VEC_MultAdd(fx32 factor, struct Vecx32  *a, struct Vecx32 *b, struct Vecx32 *dst);
-
-//CP
-fx32 FX_Div(fx32 numerator, fx32 denominator);
-fx32 FX_Inv(fx32 x);
-fx32 FX_Sqrt(fx32 x);
-fx64 FX_GetDivResultFx64c();
-fx32 FX_GetDivResult();
-void FX_InvAsync(fx32 x);
-fx32 FX_GetSqrtResult();
-void FX_DivAsync(fx32 numerator, fx32 denominator);
-fx32 FX_Divfx32(fx32 numerator, fx32 denominator);
-fx32 FX_Modfx32(fx32 num, fx32 mod);
-
-//Mtx
-//The functions ending in underscores seem to be in assembly originally
-//Mtx44
-void MTX_TransApply44(struct Mtx44 *mtx, struct Mtx44 *dst, fx32 x, fx32 y, fx32 z);
-void MTX_Concat44(struct Mtx44 *a, struct Mtx44 *b, struct Mtx44 *c);
-void MTX_Identity44_(struct Mtx44 *dst);
-void MTX_Copy44To43_(struct Mtx44 *src, struct Mtx43 *dst);
-void MTX_RotX44_(struct Mtx44 *mtx, fx32 sinphi, fx32 cosphi);
-void MTX_RotY44_(struct Mtx44 *mtx, fx32 sinphi, fx32 cosphi);
-void MTX_RotZ44_(struct Mtx44 *mtx, fx32 sinphi, fx32 cosphi);
-
-//Mtx43
-void MTX_ScaleApply33(struct Mtx33 *mtx, struct Mtx33 *dst, fx32 x, fx32 y, fx32 z);
-fx32 MTX_Inverse43(struct Mtx43 *mtx, struct Mtx43 *inv);
-void MTX_Concat43(struct Mtx43 *a, struct Mtx43 *b, struct Mtx43 *c);
-void MTX_MultVec43(struct Vecx32 *vec, struct Mtx43 *mtx, struct Vecx32 *dst);
-void MTX_Identity43_(struct Mtx43 *mtx);
-void MTX_Copy43To44_(struct Mtx43 *src, struct Mtx44 *dst);
-void MTX_Scale43_(struct Mtx43 *dst, fx32 x, fx32 y, fx32 z);
-void MTX_RotX43_(struct Mtx43 *mtx, fx32 sinphi, fx32 cosphi);
-void MTX_RotY43_(struct Mtx43 *mtx, fx32 sinphi, fx32 cosphi);
-
-//Mtx33
-void MTX_ScaleApply33(struct Mtx33 *mtx, struct Mtx33 *dst, fx32 x, fx32 y, fx32 z);
-void MTX_Concat33(struct Mtx33 *a, struct Mtx33 *b, struct Mtx33 *c);
-void MTX_MultVec33(struct Vecx32 *vec, struct Mtx33 *mtx, struct Vecx32 *dst);
-void MTX_Identity33_(struct Mtx33 *mtx);
-void MTX_RotX33_(struct Mtx33 *mtx, fx32 sinphi, fx32 cosphi);
-void MTX_RotY33_(struct Mtx33 *mtx, fx32 sinphi, fx32 cosphi);
-void MTX_RotZ33_(struct Mtx33 *mtx, fx32 sinphi, fx32 cosphi);
-
-//Mtx22
-void MTX_ScaleApply22(struct Mtx22 *mtx, struct Mtx22 *dst, fx32 x, fx32 y);
-void MTX_Identity22_(struct Mtx22 *mtx);
-void MTX_Rot22_(struct Mtx22 *mtx, fx32 sinphi, fx32 cosphi);
-
-#endif //GUARD_FX_H
-- 
cgit v1.2.3


From f4c61411673207da9d6be013674f95eeb07408b5 Mon Sep 17 00:00:00 2001
From: Made <made111@gmx.de>
Date: Mon, 4 May 2020 17:48:38 +0200
Subject: Add function attributes and remove thumb pragmas

---
 arm9/lib/src/FX.c       |  4 ++--
 arm9/lib/src/FX_atan.c  |  4 ++--
 arm9/lib/src/FX_cp.c    | 20 ++++++++++----------
 arm9/lib/src/FX_mtx22.c |  8 +++-----
 arm9/lib/src/FX_mtx33.c | 20 +++++++-------------
 arm9/lib/src/FX_mtx43.c | 24 +++++++++---------------
 arm9/lib/src/FX_mtx44.c | 21 ++++++++-------------
 arm9/lib/src/FX_vec.c   | 22 +++++++++++-----------
 8 files changed, 52 insertions(+), 71 deletions(-)

diff --git a/arm9/lib/src/FX.c b/arm9/lib/src/FX.c
index f39b1995..4b9c213f 100644
--- a/arm9/lib/src/FX.c
+++ b/arm9/lib/src/FX.c
@@ -2,11 +2,11 @@
 #include "main.h"
 #include "fx.h"
 
-void FX_Init(){
+ARM_FUNC void FX_Init(){
     return;
 }
 
-fx32 FX_Modf(fx32 x, fx32 *iptr){
+ARM_FUNC fx32 FX_Modf(fx32 x, fx32 *iptr){
     if (x >= 0)
     {
         *iptr = x & 0x7FFFF000;
diff --git a/arm9/lib/src/FX_atan.c b/arm9/lib/src/FX_atan.c
index 085763f8..979895a8 100644
--- a/arm9/lib/src/FX_atan.c
+++ b/arm9/lib/src/FX_atan.c
@@ -4,7 +4,7 @@
 
 extern fx16 FX_AtanTable_[];
 
-u16 FX_Atan(fx32 x){
+ARM_FUNC u16 FX_Atan(fx32 x){
     if (x >= 0)
     {
         if (x > 0x1000)
@@ -41,7 +41,7 @@ u16 FX_Atan(fx32 x){
     }
 }
 
-u16 FX_Atan2(fx32 x, fx32 y){
+ARM_FUNC u16 FX_Atan2(fx32 x, fx32 y){
     fx32 result;
     u32 positive, bias, denominator, numerator;
     if (x > 0)
diff --git a/arm9/lib/src/FX_cp.c b/arm9/lib/src/FX_cp.c
index a5a8d9cb..2ca9d720 100644
--- a/arm9/lib/src/FX_cp.c
+++ b/arm9/lib/src/FX_cp.c
@@ -3,17 +3,17 @@
 #include "fx.h"
 
 
-fx32 FX_Div(fx32 numerator, fx32 denominator){
+ARM_FUNC fx32 FX_Div(fx32 numerator, fx32 denominator){
     FX_DivAsync(numerator, denominator);
     return FX_GetDivResult();
 }
 
-fx32 FX_Inv(fx32 x){
+ARM_FUNC fx32 FX_Inv(fx32 x){
     FX_InvAsync(x);
     return FX_GetDivResult();
 }
 
-fx32 FX_Sqrt(fx32 x){
+ARM_FUNC fx32 FX_Sqrt(fx32 x){
     if (x > 0)
     {
         SETREG16(HW_REG_SQRTCNT, 0x1);
@@ -26,34 +26,34 @@ fx32 FX_Sqrt(fx32 x){
     }
 }
 
-fx64c FX_GetDivResultFx64c(){
+ARM_FUNC fx64c FX_GetDivResultFx64c(){
     while (READREG16(HW_REG_DIVCNT) & 0x8000);
     return READREG64(HW_REG_DIV_RESULT);
 }
 
-fx32 FX_GetDivResult(){
+ARM_FUNC fx32 FX_GetDivResult(){
     while (READREG16(HW_REG_DIVCNT) & 0x8000);
     return (READREG64(HW_REG_DIV_RESULT) + (1 << (0x14 - 1))) >> 0x14;
 }
 
-void FX_InvAsync(fx32 x){
+ARM_FUNC void FX_InvAsync(fx32 x){
     SETREG16(HW_REG_DIVCNT, 0x1);
     SETREG64(HW_REG_DIV_NUMER, (fx64)0x00001000 << 32);
     SETREG64(HW_REG_DIV_DENOM, (u32)x);
 }
 
-fx32 FX_GetSqrtResult(){
+ARM_FUNC fx32 FX_GetSqrtResult(){
     while (READREG16(HW_REG_SQRTCNT) & 0x8000);
     return (READREG32(HW_REG_SQRT_RESULT) + (1 << (0xA - 1))) >> 0xA;
 }
 
-void FX_DivAsync(fx32 numerator, fx32 denominator){
+ARM_FUNC void FX_DivAsync(fx32 numerator, fx32 denominator){
     SETREG16(HW_REG_DIVCNT, 0x1);
     SETREG64(HW_REG_DIV_NUMER, (fx64)numerator << 32);
     SETREG64(HW_REG_DIV_DENOM, (u32)denominator);
 }
 
-fx32 FX_DivS32(fx32 numerator, fx32 denominator){
+ARM_FUNC fx32 FX_DivS32(fx32 numerator, fx32 denominator){
     SETREG16(HW_REG_DIVCNT, 0x0);
     SETREG32(HW_REG_DIV_NUMER, (u32)numerator); //32bit write for some reason
     SETREG64(HW_REG_DIV_DENOM, (u32)denominator);
@@ -61,7 +61,7 @@ fx32 FX_DivS32(fx32 numerator, fx32 denominator){
     return READREG32(HW_REG_DIV_RESULT);
 }
 
-fx32 FX_ModS32(fx32 num, fx32 mod){
+ARM_FUNC fx32 FX_ModS32(fx32 num, fx32 mod){
     SETREG16(HW_REG_DIVCNT, 0x0);
     SETREG32(HW_REG_DIV_NUMER, (u32)num); //32bit write for some reason
     SETREG64(HW_REG_DIV_DENOM, (u32)mod);
diff --git a/arm9/lib/src/FX_mtx22.c b/arm9/lib/src/FX_mtx22.c
index 19504574..38387c64 100644
--- a/arm9/lib/src/FX_mtx22.c
+++ b/arm9/lib/src/FX_mtx22.c
@@ -2,14 +2,14 @@
 #include "main.h"
 #include "fx.h"
 
-void MTX_ScaleApply22(struct Mtx22 *mtx, struct Mtx22 *dst, fx32 x, fx32 y){
+ARM_FUNC void MTX_ScaleApply22(struct Mtx22 *mtx, struct Mtx22 *dst, fx32 x, fx32 y){
     dst->_[0] = ((fx64)x * mtx->_[0]) >> FX32_INT_SHIFT;
     dst->_[1] = ((fx64)x * mtx->_[1]) >> FX32_INT_SHIFT;
     dst->_[2] = ((fx64)y * mtx->_[2]) >> FX32_INT_SHIFT;
     dst->_[3] = ((fx64)y * mtx->_[3]) >> FX32_INT_SHIFT;
 }
 
-asm void MTX_Identity22_(struct Mtx22 *mtx){
+ARM_FUNC asm void MTX_Identity22_(struct Mtx22 *mtx){
     mov r1, #0x0
     mov r2, #0x1000
     mov r3, #0x0
@@ -18,8 +18,7 @@ asm void MTX_Identity22_(struct Mtx22 *mtx){
     bx lr
 }
 
-#pragma thumb on
-asm void MTX_Rot22_(struct Mtx22 *mtx, fx32 sinphi, fx32 cosphi){
+THUMB_FUNC asm void MTX_Rot22_(struct Mtx22 *mtx, fx32 sinphi, fx32 cosphi){
     str r2, [r0, #0x0]
 	str r1, [r0, #0x4]
 	neg r1, r1
@@ -27,4 +26,3 @@ asm void MTX_Rot22_(struct Mtx22 *mtx, fx32 sinphi, fx32 cosphi){
 	str r2, [r0, #0xc]
 	bx lr
 }
-#pragma thumb off
diff --git a/arm9/lib/src/FX_mtx33.c b/arm9/lib/src/FX_mtx33.c
index 391a6104..90beddba 100644
--- a/arm9/lib/src/FX_mtx33.c
+++ b/arm9/lib/src/FX_mtx33.c
@@ -2,7 +2,7 @@
 #include "main.h"
 #include "fx.h"
 
-void MTX_ScaleApply33(struct Mtx33 *mtx, struct Mtx33 *dst, fx32 x, fx32 y, fx32 z){
+ARM_FUNC void MTX_ScaleApply33(struct Mtx33 *mtx, struct Mtx33 *dst, fx32 x, fx32 y, fx32 z){
     dst->_[0] = ((fx64)x * mtx->_[0]) >> FX32_INT_SHIFT;
     dst->_[1] = ((fx64)x * mtx->_[1]) >> FX32_INT_SHIFT;
     dst->_[2] = ((fx64)x * mtx->_[2]) >> FX32_INT_SHIFT;
@@ -14,7 +14,7 @@ void MTX_ScaleApply33(struct Mtx33 *mtx, struct Mtx33 *dst, fx32 x, fx32 y, fx32
     dst->_[8] = ((fx64)z * mtx->_[8]) >> FX32_INT_SHIFT;
 }
 
-void MTX_Concat33(struct Mtx33 *a, struct Mtx33 *b, struct Mtx33 *c){
+ARM_FUNC void MTX_Concat33(struct Mtx33 *a, struct Mtx33 *b, struct Mtx33 *c){
     struct Mtx33 temp;
     struct Mtx33 *dst;
     fx32 a0, a1, a2;
@@ -57,7 +57,7 @@ void MTX_Concat33(struct Mtx33 *a, struct Mtx33 *b, struct Mtx33 *c){
         *c = temp;
 }
 
-void MTX_MultVec33(struct Vecx32 *vec, struct Mtx33 *mtx, struct Vecx32 *dst){
+ARM_FUNC void MTX_MultVec33(struct Vecx32 *vec, struct Mtx33 *mtx, struct Vecx32 *dst){
     fx32 x, y, z;
     x = vec->x;
     y = vec->y;
@@ -67,7 +67,7 @@ void MTX_MultVec33(struct Vecx32 *vec, struct Mtx33 *mtx, struct Vecx32 *dst){
     dst->z = ((fx64)x * mtx->_[2] + (fx64)y * mtx->_[5] + (fx64)z * mtx->_[8]) >> FX32_INT_SHIFT;
 }
 
-asm void MTX_Identity33_(struct Mtx33 *mtx){
+ARM_FUNC asm void MTX_Identity33_(struct Mtx33 *mtx){
     mov r2, #0x1000
     str r2, [r0, #0x20]
     mov r3, #0x0
@@ -79,8 +79,7 @@ asm void MTX_Identity33_(struct Mtx33 *mtx){
     bx lr
 }
 
-#pragma thumb on
-asm void MTX_RotX33_(struct Mtx33 *mtx, fx32 sinphi, fx32 cosphi){
+THUMB_FUNC asm void MTX_RotX33_(struct Mtx33 *mtx, fx32 sinphi, fx32 cosphi){
     mov r3, #0x1
 	lsl r3, r3, #0xc
 	str r3, [r0, #0x0]
@@ -96,10 +95,8 @@ asm void MTX_RotX33_(struct Mtx33 *mtx, fx32 sinphi, fx32 cosphi){
 	str r2, [r0, #0x20]
 	bx lr
 }
-#pragma thumb off
 
-#pragma thumb on
-asm void MTX_RotY33_(struct Mtx33 *mtx, fx32 sinphi, fx32 cosphi){
+THUMB_FUNC asm void MTX_RotY33_(struct Mtx33 *mtx, fx32 sinphi, fx32 cosphi){
     str r2, [r0, #0x0]
 	str r2, [r0, #0x20]
 	mov r3, #0x0
@@ -115,10 +112,8 @@ asm void MTX_RotY33_(struct Mtx33 *mtx, fx32 sinphi, fx32 cosphi){
 	str r3, [r0, #0x10]
 	bx lr
 }
-#pragma thumb off
 
-#pragma thumb on
-asm void MTX_RotZ33_(struct Mtx33 *mtx, fx32 sinphi, fx32 cosphi){
+THUMB_FUNC asm void MTX_RotZ33_(struct Mtx33 *mtx, fx32 sinphi, fx32 cosphi){
     stmia r0!, {r2}
 	mov r3, #0x0
 	stmia r0!, {r1,r3}
@@ -132,4 +127,3 @@ asm void MTX_RotZ33_(struct Mtx33 *mtx, fx32 sinphi, fx32 cosphi){
 	str r1, [r0, #0xc]
 	bx lr
 }
-#pragma thumb off
diff --git a/arm9/lib/src/FX_mtx43.c b/arm9/lib/src/FX_mtx43.c
index 862bd6f2..96ab4062 100644
--- a/arm9/lib/src/FX_mtx43.c
+++ b/arm9/lib/src/FX_mtx43.c
@@ -3,7 +3,7 @@
 #include "fx.h"
 
 
-void MTX_ScaleApply43(struct Mtx43 *mtx, struct Mtx43 *dst, fx32 x, fx32 y, fx32 z){
+ARM_FUNC void MTX_ScaleApply43(struct Mtx43 *mtx, struct Mtx43 *dst, fx32 x, fx32 y, fx32 z){
     //this works because matrices are indexed columns first
     MTX_ScaleApply33((struct Mtx33 *)mtx, (struct Mtx33 *)dst, x, y, z);
     dst->_[9] = mtx->_[9];
@@ -11,7 +11,7 @@ void MTX_ScaleApply43(struct Mtx43 *mtx, struct Mtx43 *dst, fx32 x, fx32 y, fx32
     dst->_[11] = mtx->_[11];
 }
 
-fx32 MTX_Inverse43(struct Mtx43 *mtx, struct Mtx43 *inv){
+ARM_FUNC fx32 MTX_Inverse43(struct Mtx43 *mtx, struct Mtx43 *inv){
     struct Mtx43 tempmat;
     struct Mtx43 *dst;
     fx32 det0, det1, det2, det;
@@ -59,7 +59,7 @@ fx32 MTX_Inverse43(struct Mtx43 *mtx, struct Mtx43 *inv){
     return 0;
 }
 
-void MTX_Concat43(struct Mtx43 *a, struct Mtx43 *b, struct Mtx43 *c){
+ARM_FUNC void MTX_Concat43(struct Mtx43 *a, struct Mtx43 *b, struct Mtx43 *c){
     struct Mtx43 temp;
     struct Mtx43 *dst;
     fx32 a0, a1, a2;
@@ -107,7 +107,7 @@ void MTX_Concat43(struct Mtx43 *a, struct Mtx43 *b, struct Mtx43 *c){
         *c = temp;
 }
 
-void MTX_MultVec43(struct Vecx32 *vec, struct Mtx43 *mtx, struct Vecx32 *dst){
+ARM_FUNC void MTX_MultVec43(struct Vecx32 *vec, struct Mtx43 *mtx, struct Vecx32 *dst){
     fx32 x, y, z;
     x = vec->x;
     y = vec->y;
@@ -120,7 +120,7 @@ void MTX_MultVec43(struct Vecx32 *vec, struct Mtx43 *mtx, struct Vecx32 *dst){
     dst->z += mtx->_[11];
 }
 
-asm void MTX_Identity43_(struct Mtx43 *mtx){
+ARM_FUNC asm void MTX_Identity43_(struct Mtx43 *mtx){
     mov r2, #0x1000
     mov r3, #0x0
     stmia r0!, {r2-r3}
@@ -133,7 +133,7 @@ asm void MTX_Identity43_(struct Mtx43 *mtx){
     bx lr
 }
 
-asm void MTX_Copy43To44_(struct Mtx43 *src, struct Mtx44 *dst){
+ARM_FUNC asm void MTX_Copy43To44_(struct Mtx43 *src, struct Mtx44 *dst){
     stmdb sp!, {r4}
     mov r12, #0x0
     ldmia r0!, {r2-r4}
@@ -149,8 +149,7 @@ asm void MTX_Copy43To44_(struct Mtx43 *src, struct Mtx44 *dst){
     bx lr
 }
 
-#pragma thumb on
-asm void MTX_Scale43_(struct Mtx43 *dst, fx32 x, fx32 y, fx32 z){
+THUMB_FUNC asm void MTX_Scale43_(struct Mtx43 *dst, fx32 x, fx32 y, fx32 z){
     stmia r0!, {r1}
     mov r1, #0x0
     str r3, [r0, #0x1c]
@@ -163,10 +162,8 @@ asm void MTX_Scale43_(struct Mtx43 *dst, fx32 x, fx32 y, fx32 z){
     stmia r0!, {r1-r3}
     bx lr
 }
-#pragma thumb off
 
-#pragma thumb on
-asm void MTX_RotX43_(struct Mtx43 *mtx, fx32 sinphi, fx32 cosphi){
+THUMB_FUNC asm void MTX_RotX43_(struct Mtx43 *mtx, fx32 sinphi, fx32 cosphi){
     str r1, [r0, #0x14]
 	neg r1, r1
 	str r1, [r0, #0x1c]
@@ -183,10 +180,8 @@ asm void MTX_RotX43_(struct Mtx43 *mtx, fx32 sinphi, fx32 cosphi){
 	stmia r0!, {r1,r3}
 	bx lr
 }
-#pragma thumb off
 
-#pragma thumb on
-asm void MTX_RotY43_(struct Mtx43 *mtx, fx32 sinphi, fx32 cosphi){
+THUMB_FUNC asm void MTX_RotY43_(struct Mtx43 *mtx, fx32 sinphi, fx32 cosphi){
     str r1, [r0, #0x18]
 	mov r3, #0x0
 	stmia r0!, {r2-r3}
@@ -201,4 +196,3 @@ asm void MTX_RotY43_(struct Mtx43 *mtx, fx32 sinphi, fx32 cosphi){
 	stmia r0!, {r1,r3}
 	bx lr
 }
-#pragma thumb off
diff --git a/arm9/lib/src/FX_mtx44.c b/arm9/lib/src/FX_mtx44.c
index 8d158602..8152907d 100644
--- a/arm9/lib/src/FX_mtx44.c
+++ b/arm9/lib/src/FX_mtx44.c
@@ -5,7 +5,7 @@
 void MI_Copy48B(void *src, void *dst);
 
 
-void MTX_TransApply44(struct Mtx44 *mtx, struct Mtx44 *dst, fx32 x, fx32 y, fx32 z){
+ARM_FUNC void MTX_TransApply44(struct Mtx44 *mtx, struct Mtx44 *dst, fx32 x, fx32 y, fx32 z){
     if(mtx != dst)
         MI_Copy48B(mtx, dst);
     dst->_[12] = mtx->_[12] + (fx32)(((fx64)x * mtx->_[0] + (fx64)y * mtx->_[4] + (fx64)z * mtx->_[8] ) >> FX32_INT_SHIFT);
@@ -14,7 +14,7 @@ void MTX_TransApply44(struct Mtx44 *mtx, struct Mtx44 *dst, fx32 x, fx32 y, fx32
     dst->_[15] = mtx->_[15] + (fx32)(((fx64)x * mtx->_[3] + (fx64)y * mtx->_[7] + (fx64)z * mtx->_[11]) >> FX32_INT_SHIFT);
 }
 
-void MTX_Concat44(struct Mtx44 *a, struct Mtx44 *b, struct Mtx44 *c){
+ARM_FUNC void MTX_Concat44(struct Mtx44 *a, struct Mtx44 *b, struct Mtx44 *c){
     struct Mtx44 temp;
     struct Mtx44 *dst;
     fx32 a0, a1, a2, a3;
@@ -73,7 +73,7 @@ void MTX_Concat44(struct Mtx44 *a, struct Mtx44 *b, struct Mtx44 *c){
         *c = temp;
 }
 
-asm void MTX_Identity44_(struct Mtx44 *dst){
+ARM_FUNC asm void MTX_Identity44_(struct Mtx44 *dst){
     mov r2, #0x1000
     mov r3, #0x0
     stmia r0!, {r2-r3}
@@ -87,7 +87,7 @@ asm void MTX_Identity44_(struct Mtx44 *dst){
     bx lr
 }
 
-asm void MTX_Copy44To43_(struct Mtx44 *src, struct Mtx43 *dst){
+ARM_FUNC asm void MTX_Copy44To43_(struct Mtx44 *src, struct Mtx43 *dst){
     ldmia r0!, {r2-r3,r12}
     add r0, r0, #0x4
     stmia r1!, {r2-r3,r12}
@@ -103,8 +103,8 @@ asm void MTX_Copy44To43_(struct Mtx44 *src, struct Mtx43 *dst){
     bx lr
 }
 
-#pragma thumb on
-asm void MTX_RotX44_(struct Mtx44 *mtx, fx32 sinphi, fx32 cosphi){
+
+THUMB_FUNC asm void MTX_RotX44_(struct Mtx44 *mtx, fx32 sinphi, fx32 cosphi){
     str r2, [r0, #0x14]
 	str r2, [r0, #0x28]
 	str r1, [r0, #0x18]
@@ -124,10 +124,8 @@ asm void MTX_RotX44_(struct Mtx44 *mtx, fx32 sinphi, fx32 cosphi){
 	str r1, [r0, #0x0]
 	bx lr
 }
-#pragma thumb off
 
-#pragma thumb on
-asm void MTX_RotY44_(struct Mtx44 *mtx, fx32 sinphi, fx32 cosphi){
+THUMB_FUNC asm void MTX_RotY44_(struct Mtx44 *mtx, fx32 sinphi, fx32 cosphi){
     str r2, [r0, #0x0]
 	str r2, [r0, #0x28]
 	str r1, [r0, #0x20]
@@ -147,10 +145,8 @@ asm void MTX_RotY44_(struct Mtx44 *mtx, fx32 sinphi, fx32 cosphi){
 	stmia r0!, {r1-r3}
 	bx lr
 }
-#pragma thumb off
 
-#pragma thumb on
-asm void MTX_RotZ44_(struct Mtx44 *mtx, fx32 sinphi, fx32 cosphi){
+THUMB_FUNC asm void MTX_RotZ44_(struct Mtx44 *mtx, fx32 sinphi, fx32 cosphi){
     str r2, [r0, #0x0]
     str r2, [r0, #0x14]
     str r1, [r0, #0x4]
@@ -169,4 +165,3 @@ asm void MTX_RotZ44_(struct Mtx44 *mtx, fx32 sinphi, fx32 cosphi){
     stmia r0!, {r1-r3}
     bx lr
 }
-#pragma thumb off
diff --git a/arm9/lib/src/FX_vec.c b/arm9/lib/src/FX_vec.c
index 10a792be..af36fe89 100644
--- a/arm9/lib/src/FX_vec.c
+++ b/arm9/lib/src/FX_vec.c
@@ -2,36 +2,36 @@
 #include "main.h"
 #include "fx.h"
 
-void VEC_Add(struct Vecx32 *a, struct Vecx32 *b, struct Vecx32 *dst){
+ARM_FUNC void VEC_Add(struct Vecx32 *a, struct Vecx32 *b, struct Vecx32 *dst){
     dst->x = a->x + b->x;
     dst->y = a->y + b->y;
     dst->z = a->z + b->z;
 }
 
-void VEC_Subtract(struct Vecx32 *a, struct Vecx32 *b, struct Vecx32 *dst){
+ARM_FUNC void VEC_Subtract(struct Vecx32 *a, struct Vecx32 *b, struct Vecx32 *dst){
     dst->x = a->x - b->x;
     dst->y = a->y - b->y;
     dst->z = a->z - b->z;
 }
 
-void VEC_Fx16Add(struct Vecx16 *a, struct Vecx16 *b, struct Vecx16 *dst){
+ARM_FUNC void VEC_Fx16Add(struct Vecx16 *a, struct Vecx16 *b, struct Vecx16 *dst){
     dst->x = a->x + b->x;
     dst->y = a->y + b->y;
     dst->z = a->z + b->z;
 }
 
-fx32 VEC_DotProduct(struct Vecx32 *a, struct Vecx32 *b){
+ARM_FUNC fx32 VEC_DotProduct(struct Vecx32 *a, struct Vecx32 *b){
     return ((fx64)a->x * b->x + (fx64)a->y * b->y + (fx64)a->z * b->z + (1 << (FX64_INT_SHIFT - 1))) >> FX64_INT_SHIFT;
 }
 
-fx32 VEC_Fx16DotProduct(struct Vecx16 *a, struct Vecx16 *b){
+ARM_FUNC fx32 VEC_Fx16DotProduct(struct Vecx16 *a, struct Vecx16 *b){
     fx32 temp1, temp2;
     temp1 = (a->x * b->x) + (a->y * b->y);
     temp2 = (a->z * b->z) + (1 << (FX64_INT_SHIFT - 1));
     return (fx32)(((fx64)temp1 + temp2) >> FX64_INT_SHIFT);
 }
 
-void VEC_CrossProduct(struct Vecx32 *a, struct Vecx32 *b, struct Vecx32 *dst){
+ARM_FUNC void VEC_CrossProduct(struct Vecx32 *a, struct Vecx32 *b, struct Vecx32 *dst){
     fx32 x, y, z;
     x = (fx32)(((fx64)a->y * b->z - (fx64)a->z * b->y + (1 << (FX64_INT_SHIFT - 1))) >> FX64_INT_SHIFT);
     y = (fx32)(((fx64)a->z * b->x - (fx64)a->x * b->z + (1 << (FX64_INT_SHIFT - 1))) >> FX64_INT_SHIFT);
@@ -41,7 +41,7 @@ void VEC_CrossProduct(struct Vecx32 *a, struct Vecx32 *b, struct Vecx32 *dst){
     dst->z = z;
 }
 
-void VEC_Fx16CrossProduct(struct Vecx16 *a, struct Vecx16 *b, struct Vecx16 *dst){
+ARM_FUNC void VEC_Fx16CrossProduct(struct Vecx16 *a, struct Vecx16 *b, struct Vecx16 *dst){
     fx32 x, y, z;
     x = ((a->y * b->z - a->z * b->y + (1 << (FX64_INT_SHIFT - 1))) >> FX64_INT_SHIFT);
     y = ((a->z * b->x - a->x * b->z + (1 << (FX64_INT_SHIFT - 1))) >> FX64_INT_SHIFT);
@@ -51,7 +51,7 @@ void VEC_Fx16CrossProduct(struct Vecx16 *a, struct Vecx16 *b, struct Vecx16 *dst
     dst->z = z;
 }
 
-fx32 VEC_Mag(struct Vecx32 *a){
+ARM_FUNC fx32 VEC_Mag(struct Vecx32 *a){
     fx64 l2 = (fx64)a->x * a->x;
     l2 += (fx64)a->y * a->y;
     l2 += (fx64)a->z * a->z;
@@ -61,7 +61,7 @@ fx32 VEC_Mag(struct Vecx32 *a){
     return ((fx32)READREG32(HW_REG_SQRT_RESULT) + 1) >> 1;
 }
 
-void VEC_Normalize(struct Vecx32 *a, struct Vecx32 *dst){
+ARM_FUNC void VEC_Normalize(struct Vecx32 *a, struct Vecx32 *dst){
     fx64 l2 = (fx64)a->x * a->x;
     l2 += (fx64)a->y * a->y;
     l2 += (fx64)a->z * a->z;
@@ -81,7 +81,7 @@ void VEC_Normalize(struct Vecx32 *a, struct Vecx32 *dst){
     dst->z = (l2 * a->z + (1LL << (0x2D - 1))) >> 0x2D;
 }
 
-void VEC_Fx16Normalize(struct Vecx16 *a, struct Vecx16 *dst){
+ARM_FUNC void VEC_Fx16Normalize(struct Vecx16 *a, struct Vecx16 *dst){
     fx64 l2 = a->x * a->x;
     l2 += a->y * a->y;
     l2 += a->z * a->z;
@@ -101,7 +101,7 @@ void VEC_Fx16Normalize(struct Vecx16 *a, struct Vecx16 *dst){
     dst->z = (l2 * a->z + (1LL << (0x2D - 1))) >> 0x2D;
 }
 
-void VEC_MultAdd(fx32 factor, struct Vecx32  *a, struct Vecx32 *b, struct Vecx32 *dst){
+ARM_FUNC void VEC_MultAdd(fx32 factor, struct Vecx32  *a, struct Vecx32 *b, struct Vecx32 *dst){
     dst->x = (fx32)(((fx64)factor * a->x) >> FX32_INT_SHIFT) + b->x;
     dst->y = (fx32)(((fx64)factor * a->y) >> FX32_INT_SHIFT) + b->y;
     dst->z = (fx32)(((fx64)factor * a->z) >> FX32_INT_SHIFT) + b->z;
-- 
cgit v1.2.3