* fix overflows during fixed-point multiply

* small fix to SwapBuffers

3 files changed, 113 insertions, 85 deletions

diff --git a/GPU3D.cpp b/GPU3D.cpp
index 01fdfb8..e77af69 100644
--- a/GPU3D.cpp
+++ b/GPU3D.cpp
@@ -270,25 +270,25 @@ void MatrixMult4x4(s32* m, s32* s)
     memcpy(tmp, m, 16*4);
 
     // m = s*m
-    m[0] = (s[0]*tmp[0] + s[1]*tmp[4] + s[2]*tmp[8] + s[3]*tmp[12]) >> 12;
-    m[1] = (s[0]*tmp[1] + s[1]*tmp[5] + s[2]*tmp[9] + s[3]*tmp[13]) >> 12;
-    m[2] = (s[0]*tmp[2] + s[1]*tmp[6] + s[2]*tmp[10] + s[3]*tmp[14]) >> 12;
-    m[3] = (s[0]*tmp[3] + s[1]*tmp[7] + s[2]*tmp[11] + s[3]*tmp[15]) >> 12;
-
-    m[4] = (s[4]*tmp[0] + s[5]*tmp[4] + s[6]*tmp[8] + s[7]*tmp[12]) >> 12;
-    m[5] = (s[4]*tmp[1] + s[5]*tmp[5] + s[6]*tmp[9] + s[7]*tmp[13]) >> 12;
-    m[6] = (s[4]*tmp[2] + s[5]*tmp[6] + s[6]*tmp[10] + s[7]*tmp[14]) >> 12;
-    m[7] = (s[4]*tmp[3] + s[5]*tmp[7] + s[6]*tmp[11] + s[7]*tmp[15]) >> 12;
-
-    m[8] = (s[8]*tmp[0] + s[9]*tmp[4] + s[10]*tmp[8] + s[11]*tmp[12]) >> 12;
-    m[9] = (s[8]*tmp[1] + s[9]*tmp[5] + s[10]*tmp[9] + s[11]*tmp[13]) >> 12;
-    m[10] = (s[8]*tmp[2] + s[9]*tmp[6] + s[10]*tmp[10] + s[11]*tmp[14]) >> 12;
-    m[11] = (s[8]*tmp[3] + s[9]*tmp[7] + s[10]*tmp[11] + s[11]*tmp[15]) >> 12;
-
-    m[12] = (s[12]*tmp[0] + s[13]*tmp[4] + s[14]*tmp[8] + s[15]*tmp[12]) >> 12;
-    m[13] = (s[12]*tmp[1] + s[13]*tmp[5] + s[14]*tmp[9] + s[15]*tmp[13]) >> 12;
-    m[14] = (s[12]*tmp[2] + s[13]*tmp[6] + s[14]*tmp[10] + s[15]*tmp[14]) >> 12;
-    m[15] = (s[12]*tmp[3] + s[13]*tmp[7] + s[14]*tmp[11] + s[15]*tmp[15]) >> 12;
+    m[0] = ((s64)s[0]*tmp[0] + (s64)s[1]*tmp[4] + (s64)s[2]*tmp[8] + (s64)s[3]*tmp[12]) >> 12;
+    m[1] = ((s64)s[0]*tmp[1] + (s64)s[1]*tmp[5] + (s64)s[2]*tmp[9] + (s64)s[3]*tmp[13]) >> 12;
+    m[2] = ((s64)s[0]*tmp[2] + (s64)s[1]*tmp[6] + (s64)s[2]*tmp[10] + (s64)s[3]*tmp[14]) >> 12;
+    m[3] = ((s64)s[0]*tmp[3] + (s64)s[1]*tmp[7] + (s64)s[2]*tmp[11] + (s64)s[3]*tmp[15]) >> 12;
+
+    m[4] = ((s64)s[4]*tmp[0] + (s64)s[5]*tmp[4] + (s64)s[6]*tmp[8] + (s64)s[7]*tmp[12]) >> 12;
+    m[5] = ((s64)s[4]*tmp[1] + (s64)s[5]*tmp[5] + (s64)s[6]*tmp[9] + (s64)s[7]*tmp[13]) >> 12;
+    m[6] = ((s64)s[4]*tmp[2] + (s64)s[5]*tmp[6] + (s64)s[6]*tmp[10] + (s64)s[7]*tmp[14]) >> 12;
+    m[7] = ((s64)s[4]*tmp[3] + (s64)s[5]*tmp[7] + (s64)s[6]*tmp[11] + (s64)s[7]*tmp[15]) >> 12;
+
+    m[8] = ((s64)s[8]*tmp[0] + (s64)s[9]*tmp[4] + (s64)s[10]*tmp[8] + (s64)s[11]*tmp[12]) >> 12;
+    m[9] = ((s64)s[8]*tmp[1] + (s64)s[9]*tmp[5] + (s64)s[10]*tmp[9] + (s64)s[11]*tmp[13]) >> 12;
+    m[10] = ((s64)s[8]*tmp[2] + (s64)s[9]*tmp[6] + (s64)s[10]*tmp[10] + (s64)s[11]*tmp[14]) >> 12;
+    m[11] = ((s64)s[8]*tmp[3] + (s64)s[9]*tmp[7] + (s64)s[10]*tmp[11] + (s64)s[11]*tmp[15]) >> 12;
+
+    m[12] = ((s64)s[12]*tmp[0] + (s64)s[13]*tmp[4] + (s64)s[14]*tmp[8] + (s64)s[15]*tmp[12]) >> 12;
+    m[13] = ((s64)s[12]*tmp[1] + (s64)s[13]*tmp[5] + (s64)s[14]*tmp[9] + (s64)s[15]*tmp[13]) >> 12;
+    m[14] = ((s64)s[12]*tmp[2] + (s64)s[13]*tmp[6] + (s64)s[14]*tmp[10] + (s64)s[15]*tmp[14]) >> 12;
+    m[15] = ((s64)s[12]*tmp[3] + (s64)s[13]*tmp[7] + (s64)s[14]*tmp[11] + (s64)s[15]*tmp[15]) >> 12;
 }
 
 void MatrixMult4x3(s32* m, s32* s)
@@ -296,26 +296,34 @@ void MatrixMult4x3(s32* m, s32* s)
     s32 tmp[16];
     memcpy(tmp, m, 16*4);
 
+    /*printf("4x3 matrix\n");
+    for (int j = 0; j < 12; j += 3)
+    {
+        for (int i = 0; i < 3; i++)
+            printf("%f ", s[i]/4096.0f);
+        printf("\n");
+    }*/
+
     // m = s*m
-    m[0] = (s[0]*tmp[0] + s[1]*tmp[4] + s[2]*tmp[8]) >> 12;
-    m[1] = (s[0]*tmp[1] + s[1]*tmp[5] + s[2]*tmp[9]) >> 12;
-    m[2] = (s[0]*tmp[2] + s[1]*tmp[6] + s[2]*tmp[10]) >> 12;
-    m[3] = (s[0]*tmp[3] + s[1]*tmp[7] + s[2]*tmp[11]) >> 12;
-
-    m[4] = (s[3]*tmp[0] + s[4]*tmp[4] + s[5]*tmp[8]) >> 12;
-    m[5] = (s[3]*tmp[1] + s[4]*tmp[5] + s[5]*tmp[9]) >> 12;
-    m[6] = (s[3]*tmp[2] + s[4]*tmp[6] + s[5]*tmp[10]) >> 12;
-    m[7] = (s[3]*tmp[3] + s[4]*tmp[7] + s[5]*tmp[11]) >> 12;
-
-    m[8] = (s[6]*tmp[0] + s[7]*tmp[4] + s[8]*tmp[8]) >> 12;
-    m[9] = (s[6]*tmp[1] + s[7]*tmp[5] + s[8]*tmp[9]) >> 12;
-    m[10] = (s[6]*tmp[2] + s[7]*tmp[6] + s[8]*tmp[10]) >> 12;
-    m[11] = (s[6]*tmp[3] + s[7]*tmp[7] + s[8]*tmp[11]) >> 12;
-
-    m[12] = (s[9]*tmp[0] + s[10]*tmp[4] + s[11]*tmp[8] + 0x1000*tmp[12]) >> 12;
-    m[13] = (s[9]*tmp[1] + s[10]*tmp[5] + s[11]*tmp[9] + 0x1000*tmp[13]) >> 12;
-    m[14] = (s[9]*tmp[2] + s[10]*tmp[6] + s[11]*tmp[10] + 0x1000*tmp[14]) >> 12;
-    m[15] = (s[9]*tmp[3] + s[10]*tmp[7] + s[11]*tmp[11] + 0x1000*tmp[15]) >> 12;
+    m[0] = ((s64)s[0]*tmp[0] + (s64)s[1]*tmp[4] + (s64)s[2]*tmp[8]) >> 12;
+    m[1] = ((s64)s[0]*tmp[1] + (s64)s[1]*tmp[5] + (s64)s[2]*tmp[9]) >> 12;
+    m[2] = ((s64)s[0]*tmp[2] + (s64)s[1]*tmp[6] + (s64)s[2]*tmp[10]) >> 12;
+    m[3] = ((s64)s[0]*tmp[3] + (s64)s[1]*tmp[7] + (s64)s[2]*tmp[11]) >> 12;
+
+    m[4] = ((s64)s[3]*tmp[0] + (s64)s[4]*tmp[4] + (s64)s[5]*tmp[8]) >> 12;
+    m[5] = ((s64)s[3]*tmp[1] + (s64)s[4]*tmp[5] + (s64)s[5]*tmp[9]) >> 12;
+    m[6] = ((s64)s[3]*tmp[2] + (s64)s[4]*tmp[6] + (s64)s[5]*tmp[10]) >> 12;
+    m[7] = ((s64)s[3]*tmp[3] + (s64)s[4]*tmp[7] + (s64)s[5]*tmp[11]) >> 12;
+
+    m[8] = ((s64)s[6]*tmp[0] + (s64)s[7]*tmp[4] + (s64)s[8]*tmp[8]) >> 12;
+    m[9] = ((s64)s[6]*tmp[1] + (s64)s[7]*tmp[5] + (s64)s[8]*tmp[9]) >> 12;
+    m[10] = ((s64)s[6]*tmp[2] + (s64)s[7]*tmp[6] + (s64)s[8]*tmp[10]) >> 12;
+    m[11] = ((s64)s[6]*tmp[3] + (s64)s[7]*tmp[7] + (s64)s[8]*tmp[11]) >> 12;
+
+    m[12] = ((s64)s[9]*tmp[0] + (s64)s[10]*tmp[4] + (s64)s[11]*tmp[8] + (s64)0x1000*tmp[12]) >> 12;
+    m[13] = ((s64)s[9]*tmp[1] + (s64)s[10]*tmp[5] + (s64)s[11]*tmp[9] + (s64)0x1000*tmp[13]) >> 12;
+    m[14] = ((s64)s[9]*tmp[2] + (s64)s[10]*tmp[6] + (s64)s[11]*tmp[10] + (s64)0x1000*tmp[14]) >> 12;
+    m[15] = ((s64)s[9]*tmp[3] + (s64)s[10]*tmp[7] + (s64)s[11]*tmp[11] + (s64)0x1000*tmp[15]) >> 12;
 }
 
 void MatrixMult3x3(s32* m, s32* s)
@@ -324,45 +332,45 @@ void MatrixMult3x3(s32* m, s32* s)
     memcpy(tmp, m, 12*4);
 
     // m = s*m
-    m[0] = (s[0]*tmp[0] + s[1]*tmp[4] + s[2]*tmp[8]) >> 12;
-    m[1] = (s[0]*tmp[1] + s[1]*tmp[5] + s[2]*tmp[9]) >> 12;
-    m[2] = (s[0]*tmp[2] + s[1]*tmp[6] + s[2]*tmp[10]) >> 12;
-    m[3] = (s[0]*tmp[3] + s[1]*tmp[7] + s[2]*tmp[11]) >> 12;
-
-    m[4] = (s[3]*tmp[0] + s[4]*tmp[4] + s[5]*tmp[8]) >> 12;
-    m[5] = (s[3]*tmp[1] + s[4]*tmp[5] + s[5]*tmp[9]) >> 12;
-    m[6] = (s[3]*tmp[2] + s[4]*tmp[6] + s[5]*tmp[10]) >> 12;
-    m[7] = (s[3]*tmp[3] + s[4]*tmp[7] + s[5]*tmp[11]) >> 12;
-
-    m[8] = (s[6]*tmp[0] + s[7]*tmp[4] + s[8]*tmp[8]) >> 12;
-    m[9] = (s[6]*tmp[1] + s[7]*tmp[5] + s[8]*tmp[9]) >> 12;
-    m[10] = (s[6]*tmp[2] + s[7]*tmp[6] + s[8]*tmp[10]) >> 12;
-    m[11] = (s[6]*tmp[3] + s[7]*tmp[7] + s[8]*tmp[11]) >> 12;
+    m[0] = ((s64)s[0]*tmp[0] + (s64)s[1]*tmp[4] + (s64)s[2]*tmp[8]) >> 12;
+    m[1] = ((s64)s[0]*tmp[1] + (s64)s[1]*tmp[5] + (s64)s[2]*tmp[9]) >> 12;
+    m[2] = ((s64)s[0]*tmp[2] + (s64)s[1]*tmp[6] + (s64)s[2]*tmp[10]) >> 12;
+    m[3] = ((s64)s[0]*tmp[3] + (s64)s[1]*tmp[7] + (s64)s[2]*tmp[11]) >> 12;
+
+    m[4] = ((s64)s[3]*tmp[0] + (s64)s[4]*tmp[4] + (s64)s[5]*tmp[8]) >> 12;
+    m[5] = ((s64)s[3]*tmp[1] + (s64)s[4]*tmp[5] + (s64)s[5]*tmp[9]) >> 12;
+    m[6] = ((s64)s[3]*tmp[2] + (s64)s[4]*tmp[6] + (s64)s[5]*tmp[10]) >> 12;
+    m[7] = ((s64)s[3]*tmp[3] + (s64)s[4]*tmp[7] + (s64)s[5]*tmp[11]) >> 12;
+
+    m[8] = ((s64)s[6]*tmp[0] + (s64)s[7]*tmp[4] + (s64)s[8]*tmp[8]) >> 12;
+    m[9] = ((s64)s[6]*tmp[1] + (s64)s[7]*tmp[5] + (s64)s[8]*tmp[9]) >> 12;
+    m[10] = ((s64)s[6]*tmp[2] + (s64)s[7]*tmp[6] + (s64)s[8]*tmp[10]) >> 12;
+    m[11] = ((s64)s[6]*tmp[3] + (s64)s[7]*tmp[7] + (s64)s[8]*tmp[11]) >> 12;
 }
 
 void MatrixScale(s32* m, s32* s)
 {
-    m[0] = (s[0]*m[0]) >> 12;
-    m[1] = (s[0]*m[1]) >> 12;
-    m[2] = (s[0]*m[2]) >> 12;
-    m[3] = (s[0]*m[3]) >> 12;
-
-    m[4] = (s[1]*m[4]) >> 12;
-    m[5] = (s[1]*m[5]) >> 12;
-    m[6] = (s[1]*m[6]) >> 12;
-    m[7] = (s[1]*m[7]) >> 12;
-
-    m[8] = (s[2]*m[8]) >> 12;
-    m[9] = (s[2]*m[9]) >> 12;
-    m[10] = (s[2]*m[10]) >> 12;
-    m[11] = (s[2]*m[11]) >> 12;
+    m[0] = ((s64)s[0]*m[0]) >> 12;
+    m[1] = ((s64)s[0]*m[1]) >> 12;
+    m[2] = ((s64)s[0]*m[2]) >> 12;
+    m[3] = ((s64)s[0]*m[3]) >> 12;
+
+    m[4] = ((s64)s[1]*m[4]) >> 12;
+    m[5] = ((s64)s[1]*m[5]) >> 12;
+    m[6] = ((s64)s[1]*m[6]) >> 12;
+    m[7] = ((s64)s[1]*m[7]) >> 12;
+
+    m[8] = ((s64)s[2]*m[8]) >> 12;
+    m[9] = ((s64)s[2]*m[9]) >> 12;
+    m[10] = ((s64)s[2]*m[10]) >> 12;
+    m[11] = ((s64)s[2]*m[11]) >> 12;
 }
 
 void MatrixTranslate(s32* m, s32* s)
 {
-    m[12] += (s[0]*m[0] + s[1]*m[4] + s[2]*m[8]) >> 12;
-    m[13] += (s[0]*m[1] + s[1]*m[5] + s[2]*m[9]) >> 12;
-    m[14] += (s[0]*m[2] + s[1]*m[6] + s[2]*m[10]) >> 12;
+    m[12] += ((s64)s[0]*m[0] + (s64)s[1]*m[4] + (s64)s[2]*m[8]) >> 12;
+    m[13] += ((s64)s[0]*m[1] + (s64)s[1]*m[5] + (s64)s[2]*m[9]) >> 12;
+    m[14] += ((s64)s[0]*m[2] + (s64)s[1]*m[6] + (s64)s[2]*m[10]) >> 12;
 }
 
 void UpdateClipMatrix()
@@ -379,7 +387,7 @@ void UpdateClipMatrix()
 template<int comp, s32 plane>
 void ClipSegment(Vertex* outbuf, int num, Vertex* vout, Vertex* vin)
 {
-    s32 factor = ((vin->Position[3] - (plane*vin->Position[comp])) << 12) /
+    s64 factor = ((vin->Position[3] - (plane*vin->Position[comp])) << 12) /
         ((vin->Position[3] - (plane*vin->Position[comp])) - (vout->Position[3] - (plane*vout->Position[comp])));
 
     Vertex mid;
@@ -412,6 +420,15 @@ void SubmitPolygon()
     int prev, next;
     int c;
 
+    /*if (NumPolygons == 91)
+    for (int i = 0; i < nverts; i++)
+    {
+        Vertex vtx = TempVertexBuffer[i];
+        printf("pre-clip v%d: %f %f %f %f\n", i,
+               vtx.Position[0]/4096.0f, vtx.Position[1]/4096.0f,
+               vtx.Position[2]/4096.0f, vtx.Position[3]/4096.0f);
+    }*/
+
     // X clipping
 
     prev = nverts-1; next = 1; c = 0;
@@ -603,7 +620,7 @@ void SubmitPolygon()
 
 void SubmitVertex()
 {
-    s32 vertex[4] = {(s32)CurVertex[0], (s32)CurVertex[1], (s32)CurVertex[2], 0x1000};
+    s64 vertex[4] = {(s64)CurVertex[0], (s64)CurVertex[1], (s64)CurVertex[2], 0x1000};
     //s32 vertextrans[4];
     Vertex* vertextrans = &TempVertexBuffer[VertexNumInPoly];
 
@@ -769,9 +786,12 @@ void ExecuteCommand()
     ExecParams[ExecParamCount] = entry.Param;
     ExecParamCount++;
 
+    //if ((entry.Command&0xF0)==0x10)
+    //    printf("MATRIX CMD %02X %08X\n", entry.Command, entry.Param);
+
     if (ExecParamCount >= CmdNumParams[entry.Command])
     {
-        CycleCount += CmdNumCycles[entry.Command];
+        //CycleCount += CmdNumCycles[entry.Command];
         ExecParamCount = 0;
 
         GXStat &= ~(1<<14);
@@ -1129,6 +1149,9 @@ void ExecuteCommand()
 
 void Run(s32 cycles)
 {
+    if (FlushRequest)
+        return;
+
     if (CycleCount <= 0)
     {
         while (CycleCount <= 0 && !CmdPIPE->IsEmpty())
@@ -1140,8 +1163,7 @@ void Run(s32 cycles)
     if (CycleCount <= 0 && CmdPIPE->IsEmpty())
     {
         CycleCount = 0;
-        if (!FlushRequest)
-            GXStat &= ~(1<<27);
+        GXStat &= ~(1<<27);
     }
 }
 
diff --git a/GPU3D_Soft.cpp b/GPU3D_Soft.cpp
index d47975d..6ac4e81 100644
--- a/GPU3D_Soft.cpp
+++ b/GPU3D_Soft.cpp
@@ -102,7 +102,7 @@ void RenderPolygon(Polygon* polygon)
             vbot = i;
         }
         //if (vtx->Color[0]==63 && vtx->Color[1]==0 && vtx->Color[2]==0)
-        //printf("v%d: %d,%d  W=%d\n", i, scrX, 191-scrY, vtx->Position[3]);
+        //printf("v%d: %d,%d  Z=%f  W=%f\n", i, scrX, 191-scrY, vtx->Position[2]/4096.0f, vtx->Position[3]/4096.0f);
     }
 
     // draw, line per line
@@ -176,11 +176,16 @@ void RenderPolygon(Polygon* polygon)
         s32 xl = scrcoords[lcur][0] + (((scrcoords[lnext][0] - scrcoords[lcur][0]) * lfactor) >> 12);
         s32 xr = scrcoords[rcur][0] + (((scrcoords[rnext][0] - scrcoords[rcur][0]) * rfactor) >> 12);
 
+        if (xl<0 || xr>255) continue; // hax
+
         //if (vlcur->Color[0]==0 && vlcur->Color[1]==63 && vlcur->Color[2]==0)
-        //    printf("y:%d  xleft:%d  xright:%d   %d,%d  %d,%d\n", y, xl, xr, lcur, rcur, vtop, vbot);
+            /*printf("y:%d  xleft:%d  xright:%d   %d,%d  %d,%d   |   left: %d to %d   right: %d to %d\n",
+                   y, xl, xr, lcur, rcur, vtop, vbot,
+                   scrcoords[lcur][0], scrcoords[lnext][0],
+                   scrcoords[rcur][0], scrcoords[rnext][0]);*/
 
-        //s32 zl = scrcoords[lcur][3] + (((scrcoords[lnext][3] - scrcoords[lcur][3]) * lfactor) >> 12);
-        //s32 zr = scrcoords[rcur][3] + (((scrcoords[rnext][3] - scrcoords[rcur][3]) * rfactor) >> 12);
+        //s32 zl = scrcoords[lcur][2] + (((scrcoords[lnext][2] - scrcoords[lcur][2]) * lfactor) >> 12);
+        //s32 zr = scrcoords[rcur][2] + (((scrcoords[rnext][2] - scrcoords[rcur][2]) * rfactor) >> 12);
 
         u8 rl = vlcur->Color[0] + (((vlnext->Color[0] - vlcur->Color[0]) * lfactor) >> 12);
         u8 gl = vlcur->Color[1] + (((vlnext->Color[1] - vlcur->Color[1]) * lfactor) >> 12);
@@ -200,12 +205,12 @@ void RenderPolygon(Polygon* polygon)
         {
             s32 xfactor = (x - xl) * xdiv;
 
-            //s32 z = (zl << 12) + ((zr - zl) * xfactor);
-            //z = zl + (((zr - zl) * xfactor) >> 12);
+            //s32 z = (((zr - zl) * xfactor) >> 12);
+            //if (zr!=zl) z = (z << 12) / (zr - zl);
 
             //s32 z_inv = ((z>>12)==0) ? 0x1000 : 0x1000000 / (z >> 12);
             //xfactor = (xfactor * z_inv) >> 12;
-            //xfactor = (xfactor << 12) / z;
+            //if (z) xfactor = (xfactor << 12) / z;
 
             // TODO: get rid of this shit
             if (x<0 || x>255 || y<0 || y>191)
@@ -248,6 +253,7 @@ void RenderFrame(Vertex* vertices, Polygon* polygons, int npolys)
                    polygons[i].Vertices[j]->Position[1]/4096.0f,
                    polygons[i].Vertices[j]->Position[2]/4096.0f);
 */
+        //printf("polygon %d\n", i);
         RenderPolygon(&polygons[i]);
     }
 }
diff --git a/melonDS.depend b/melonDS.depend
index 08e52b7..51e832c 100644
--- a/melonDS.depend
+++ b/melonDS.depend
@@ -1,5 +1,5 @@
 # depslib dependency file v1.0
-1486824787 source:c:\documents\sources\melonds\main.cpp
+1486993536 source:c:\documents\sources\melonds\main.cpp
 	<stdio.h>
 	<windows.h>
 	"NDS.h"
@@ -10,7 +10,7 @@
 
 1481161027 c:\documents\sources\melonds\types.h
 
-1486947856 source:c:\documents\sources\melonds\nds.cpp
+1486994139 source:c:\documents\sources\melonds\nds.cpp
 	<stdio.h>
 	<string.h>
 	"NDS.h"
@@ -148,14 +148,14 @@
 
 1486777933 c:\documents\sources\melonds\gpu3d.h
 
-1486947978 source:c:\documents\sources\melonds\gpu3d.cpp
+1486993935 source:c:\documents\sources\melonds\gpu3d.cpp
 	<stdio.h>
 	<string.h>
 	"NDS.h"
 	"GPU.h"
 	"FIFO.h"
 
-1486947027 source:c:\documents\sources\melonds\gpu3d_soft.cpp
+1486994049 source:c:\documents\sources\melonds\gpu3d_soft.cpp
 	<stdio.h>
 	<string.h>
 	"NDS.h"