port from perforce

evoke-64k/trunk/ev10/intrin.cpp

@@ -0,0 +1,219 @@
+unsigned int myrand()
+{
+	static int i= 0;
+	i*= 78547803;
+	i+= 1747627;
+	return i;
+}
+
+#ifdef SUPERSMALL
+extern "C" int _fltused;
+int _fltused;
+
+extern "C" 
+{ 
+	__declspec(naked) void _ftol2() 
+	{
+		__asm
+		{
+			fistp    DWORD PTR [esp-12]
+			mov		 eax, DWORD PTR [esp-12]
+			mov		 ecx, DWORD PTR [esp-8]
+
+			ret
+		}
+	}
+	__declspec(naked) void _ftol2_sse() 
+	{
+		__asm
+		{
+			fistp    DWORD PTR [esp-12]
+			mov		 eax, DWORD PTR [esp-12]
+			mov		 ecx, DWORD PTR [esp-8]
+
+			ret
+		}
+	}
+}
+
+#endif
+
+extern "C" int ____security_cookie;
+int ____security_cookie;
+
+#ifdef SUPERSMALL
+/*extern "C" 
+{ 
+	__declspec(naked) void _ftol2() 
+	{
+		__asm
+		{
+			fistp    DWORD PTR [esp-12]
+			mov		 eax, DWORD PTR [esp-12]
+			mov		 ecx, DWORD PTR [esp-8]
+
+			ret
+		}
+	}
+	__declspec(naked) void _ftol2_sse() 
+	{
+		__asm
+		{
+			fistp    DWORD PTR [esp-12]
+			mov		 eax, DWORD PTR [esp-12]
+			mov		 ecx, DWORD PTR [esp-8]
+
+			ret
+		}
+	}
+}*/
+
+float mypow( float a,  float b)
+{
+  // faster pow based on code by agner fog
+  __asm
+  {
+    fld   b;
+	fld   a;
+
+//    ftst;
+//    fstsw ax;
+//    sahf;
+ //   jz    zero;
+
+    fyl2x;
+    fist  dword ptr [a];
+    sub   esp, 12;
+    mov   dword ptr [esp],0;
+    mov   dword ptr [esp+4],0x80000000;
+    fisub dword ptr [a];
+    mov   eax, dword ptr [a];
+    add   eax, 0x3fff;
+    mov   [esp+8], eax;
+//  jle   underflow;
+//  cmp   eax, 0x8000;
+//    jge   overflow;
+    f2xm1;
+    fld1;
+    fadd;
+    fld   tbyte ptr [esp];
+    add   esp, 12;
+    fmul;
+    //jmp   end;
+
+// underflow:
+//    fstp  st;
+//    fldz;
+//    add   esp, 12;
+//    jmp   end;
+
+//overflow:
+//    push  0x7f800000;
+//    fstp  st;
+//    fld   dword ptr [esp];
+//    add   esp, 16;
+//    jmp   end;
+
+//zero:
+//    fstp  st(1);
+
+//end:
+  }
+  //return st;
+}
+#endif
+
+void mymemcpy(void* av_Dest_, const void* av_Src_, size_t ai_Size)
+{
+	__asm mov esi, av_Src_
+	__asm mov edi, av_Dest_
+	__asm mov ecx, ai_Size
+	__asm rep movsb
+}
+
+void mymemset(void* dest, const char value, size_t ai_size)
+{
+	__asm mov al, value
+	__asm mov edi, dest
+	__asm mov ecx, ai_size
+	__asm rep stosb
+}
+
+void mymemzero(void* dest, size_t ai_size)
+{
+	__asm xor al, al
+	__asm mov edi, dest
+	__asm mov ecx, ai_size
+	__asm rep stosb
+}
+
+extern "C"
+{
+	double __cdecl sin(double);
+	double __cdecl cos(double);
+	double __cdecl sqrt(double);
+	double __cdecl fabs(double);
+
+	void* __cdecl memcpy( void* dest, const void* src, size_t count );
+}
+
+#pragma intrinsic ( sin, cos, sqrt, fabs ) // true intrinsic
+#pragma intrinsic ( memcpy ) // true intrinsic
+
+/*
+__forceinline double sin(double v)
+{
+ __asm fld v;
+ __asm fsin;
+ __asm fstp v;
+ return v;
+}
+
+
+float cos(float v)
+{
+ __asm fld v;
+ __asm fcos;
+ __asm fstp v;
+ return v;
+}
+*/
+/*float sqrt(float v)
+{
+ __asm finit;
+ __asm fld v;
+ __asm fsqrt;
+ __asm fstp v;
+ return v;
+}*/
+
+/*float abs(float X)
+{
+	if(X < 0.0f)
+		return -X;
+
+	return X;
+}*/
+
+float maximum(float X, float Y)
+{
+	if(X < Y)
+		return Y;
+
+	return X;
+}
+					
+float minimum(float X, float Y)
+{
+	if(X > Y)
+		return Y;
+
+	return X;
+}
+
+float signpow( float a, float b )
+{
+	float fSign= ( a < 0.0f ? -1.0f : 1.0f );
+	return fSign * (float)mypow( (float)fabs( (float)a ), b );
+}
+