Added AMD64 version of srE module

git-svn-id: https://svn.inf.ethz.ch/svn/lecturers/a2/trunk@7833 8c9fc860-2736-0410-a75d-ab315db34111

source/AMD64.srE.Mod

@@ -0,0 +1,90 @@
+MODULE srE;
+IMPORT SYSTEM, srBase;
+
+TYPE SREAL=srBase.SREAL;
+
+(* procedure for casting real array to integer array. From Patrik Reali ETHZ 2000*)
+
+PROCEDURE E*(VAR p: srBase.PT; VAR ijk: srBase.IPT);
+VAR
+	in: ARRAY 3 OF SREAL;
+	out: ARRAY 3 OF INTEGER;
+BEGIN
+	in[0]:=p.x;in[1]:=p.y; in[2]:=p.z;
+	Eprime(in,out);
+	ijk.i:=out[0]; ijk.j:=out[1]; ijk.k:=out[2];
+END E;
+
+PROCEDURE Eprime(VAR in:ARRAY OF SREAL; VAR out: ARRAY OF INTEGER);
+CODE {SYSTEM.AMD64, SYSTEM.FPU}
+   PUSH  RCX
+   MOV   RDI, [RBP+out+0]         ; dest   = ADR(out)
+   MOV   RCX, [RBP+out+8]       ; count  = LEN(out)
+   MOV   RSI, [RBP+in+0]        ; source = ADR(in)
+   CMP   RCX, [RBP+in+8]
+   JGE   Ok
+   PUSH  99                  ; LEN(in) > LEN(out)  then TRAP(99)
+   INT   3
+Ok:
+   SUB   RSP, 8              ; change FPU rounding to "chop"
+   FSTCW [RSP]
+   FWAIT
+   MOV   EBX, [RSP]
+   OR    EBX, 0400H          ; clear bit 10,11 (chop/truncate toward zero)
+   MOV   [RSP+4], EBX
+   FLDCW [RSP+4]
+   JMP   Check
+Loop:
+   DEC   RCX
+   FLD   DWORD [RSI+RCX*4]  ; in: SREAL
+   FISTP WORD [RDI+RCX*2]   ; out: INTEGER
+   FWAIT
+Check:
+   CMP   RCX, 0
+   JG    Loop
+   FLDCW [RSP]               ; restore original FPU configuration
+   ADD   RSP, 8
+   POP   RCX
+END Eprime;
+
+PROCEDURE E2*(in: srBase.PT; VAR out: srBase.IPT);
+(*
+BEGIN
+	ROUND(in[0], out[0]);
+	ROUND(in[1], out[1]);
+	ROUND(in[2], out[2]); *)
+END E2;
+
+(* PROCEDURE ROUND(x: SREAL; VAR y: LONGINT);
+CODE {SYSTEM.i386, SYSTEM.FPU}
+       FLD x[EBP]
+       MOV EAX, y[EBP]
+       FISTP DWORD 0[EAX]
+END ROUND;
+
+PROCEDURE ROUND(x: LONGSREAL; VAR y: LONGINT);
+CODE {SYSTEM.i386, SYSTEM.FPU}
+       FLD [EBP+x]
+       MOV EAX, [EBP+y]
+       FISTP DWORD [EAX]
+END ROUND; *)
+
+PROCEDURE -ROUND*(x: SREAL; VAR y: LONGINT);
+CODE {SYSTEM.AMD64, SYSTEM.FPU}
+       POP RAX
+       FLD DWORD [RBP]
+       ADD RSP, 8
+       FISTP DWORD [RAX]
+END ROUND;
+
+(* PROCEDURE -ROUND(x: LONGSREAL; VAR y: LONGINT);
+CODE {SYSTEM.i386, SYSTEM.FPU}
+       POP EAX
+       FLD QWORD [EBP]
+       ADD ESP, 8
+       FISTP DWORD [EAX]
+END ROUND;
+*)
+
+END srE.
+

source/Release.Tool

@@ -1479,7 +1479,10 @@ PACKAGE Contributions ARCHIVE "Contributions.zip" SOURCE "ContributionsSrc.zip"
 	SambaClient.Mod SambaServer.Mod
 
 	srRenderTools.Text srReadMe.Text	# Realtime Fractal-Voxel Raytracer (Soren Renner)
-	srBase.Mod srRayEngine.Mod srMath.Mod I386.srE.Mod srGL.Mod srHex.Mod srImage.Mod
+	srBase.Mod srRayEngine.Mod srMath.Mod
+	I386, WINORIG, WINGEN32, UNIX32, UNIXG32 { I386.srE.Mod }
+	AMD64, WINGEN64, UNIXG64 { AMD64.srE.Mod }
+	srGL.Mod srHex.Mod srImage.Mod
 	srVoxel.Mod srVoxel2.Mod srVoxel3.Mod srVolShader.Mod srVoxel4.Mod srVoxel5.Mod
 	srM2Space.Mod srM3Space.Mod srM5Space.Mod srM6Space.Mod srRastermovie.Mod
 	srTexVox.Mod srThermoCell.Mod srTree.Mod sr3DTexture.Mod srLifeVox.Mod