diff --git a/include/c64/charwin.c b/include/c64/charwin.c
index 79a2c7f..f29939e 100644
--- a/include/c64/charwin.c
+++ b/include/c64/charwin.c
@@ -199,14 +199,12 @@ void cwin_write_string(CharWin * win, const char * buffer)
 	{
 		for(char x=0; x<win->wx; x++)
 		{
-			char c = *buffer;
-			if (c)
+			char ch = *buffer;
+			if (ch)
 			{
-				c &= 0xbf;
-				if (c & 0x80)
-					c ^= 0xc0;
+				ch = (ch & 0x3f) | ((ch & 0x80) >> 1);
 				
-				dp[x] = c;
+				dp[x] = ch;
 				buffer++;
 			}
 			else
@@ -254,9 +252,7 @@ void cwin_putat_char(CharWin * win, char x, char y, char ch, char color)
 {
 	int	offset = mul40[y] + x;
 
-	ch &= 0xbf;
-	if (ch & 0x80)
-		ch ^= 0xc0;
+	ch = (ch & 0x3f) | ((ch & 0x80) >> 1);
 
 	win->sp[offset] = ch;
 	win->cp[offset] = color;
@@ -275,9 +271,7 @@ void cwin_putat_chars(CharWin * win, char x, char y, const char * chars, char nu
 	{
 		char	ch = chars[i];
 
-		ch &= 0xbf;
-		if (ch & 0x80)
-			ch ^= 0xc0;
+		ch = (ch & 0x3f) | ((ch & 0x80) >> 1);
 
 		sp[i] = ch;
 		cp[i] = color;
@@ -296,9 +290,7 @@ char cwin_putat_string(CharWin * win, char x, char y, const char * str, char col
 	char	i = 0;
 	while (char	ch = str[i])
 	{
-		ch &= 0xbf;
-		if (ch & 0x80)
-			ch ^= 0xc0;
+		ch = (ch & 0x3f) | ((ch & 0x80) >> 1);
 
 		sp[i] = ch;
 		cp[i] = color;
diff --git a/include/crt.c b/include/crt.c
index 0a6a73f..79b127f 100644
--- a/include/crt.c
+++ b/include/crt.c
@@ -131,6 +131,7 @@ w2:
 // All native code
 		jsr	main
 pexec:
+tyexec:
 yexec:
 zexec:
 exec:
@@ -146,32 +147,27 @@ exec:
 		sta	ip + 1
 		
 pexec:
-		ldy	#$ff
+		ldy	#0
+		beq exec
+tyexec:
+		ldy	tmpy
 yexec:
 		iny
 exec:
-#if 0
-		tya		
-		clc		
-		adc	ip	
-		sta	ip	
-		bcc	W1
-		inc	ip + 1		
-W1:		ldy #0
-#endif
 		lda	(ip), y
 		sta	execjmp + 1
 		iny		
 execjmp:
 		jmp 	(0x0900)
 zexec:
-		tya		
+		tya	
+		ldy #0
 		clc		
 		adc	ip	
 		sta	ip	
-		bcc	pexec	
+		bcc	exec	
 		inc	ip + 1		
-		bne	pexec
+		bne	exec
 bcode:		
 		byt	BC_CALL_ABS * 2
 		byt	<main
@@ -902,8 +898,7 @@ __asm inp_load_abs_8
 L0:
 		lda	(addr), y
 		sta	$00, x
-		ldy	tmpy
-		jmp	startup.yexec
+		jmp	startup.tyexec
 inp_load_addr_8:
 		lda	(ip), y
 		tax
@@ -934,8 +929,7 @@ L0:
 		sta	$00, x
 		lda #0
 		sta $01, x
-		ldy	tmpy
-		jmp	startup.yexec
+		jmp	startup.tyexec
 inp_load_addr_u8:
 		lda	(ip), y
 		tax
@@ -967,8 +961,7 @@ L0:
 		iny
 		lda	(addr), y
 		sta	$01, x
-		ldy	tmpy
-		jmp	startup.yexec
+		jmp	startup.tyexec
 
 inp_load_addr_16:
 		lda	(ip), y
@@ -983,6 +976,28 @@ inp_load_addr_16:
 #pragma	bytecode(BC_LOAD_ABS_16, inp_load_abs_16)
 #pragma	bytecode(BC_LOAD_ADDR_16, inp_load_abs_16.inp_load_addr_16)
 
+__asm inp_load_abs_addr
+{
+		lda	(ip), y
+		sta	addr
+		iny
+		lda	(ip), y
+		sta	addr + 1
+		sty	tmpy
+
+		ldy #0
+		lda	(addr), y
+		tax
+		iny
+		lda	(addr), y
+		sta addr + 1
+		stx addr
+		jmp	startup.tyexec
+}
+
+#pragma	bytecode(BC_LOAD_ABS_ADDR, inp_load_abs_addr)
+
+
 __asm inp_load_abs_32
 {
 		lda	(ip), y
@@ -1008,8 +1023,7 @@ L0:
 		iny
 		lda	(addr), y
 		sta	$03, x
-		ldy	tmpy
-		jmp	startup.yexec
+		jmp	startup.tyexec
 
 inp_load_addr_32:
 		lda	(ip), y
@@ -1040,8 +1054,7 @@ L0:
 
 		lda	$00, x
 		sta	(addr), y
-		ldy	tmpy
-		jmp	startup.yexec
+		jmp	startup.tyexec
 
 inp_store_addr_8:
 		lda	(ip), y
@@ -1074,8 +1087,7 @@ L0:
 		iny
 		lda	$01, x
 		sta	(addr), y
-		ldy	tmpy
-		jmp	startup.yexec
+		jmp	startup.tyexec
 
 inp_store_addr_16:
 		lda	(ip), y
@@ -1114,8 +1126,7 @@ L0:
 		iny
 		lda	$03, x
 		sta	(addr), y
-		ldy	tmpy
-		jmp	startup.yexec
+		jmp	startup.tyexec
 
 inp_store_addr_32:
 		lda	(ip), y
@@ -1210,8 +1221,7 @@ __asm inp_load_local_16
 		iny
 		lda	(fp), y
 		sta	$01, x
-		ldy	tmpy
-		jmp	startup.yexec
+		jmp	startup.tyexec
 }
 
 #pragma	bytecode(BC_LOAD_LOCAL_16, inp_load_local_16)
@@ -1235,8 +1245,7 @@ __asm inp_load_local_32
 		iny
 		lda	(fp), y
 		sta	$03, x
-		ldy	tmpy
-		jmp	startup.yexec
+		jmp	startup.tyexec
 }
 
 #pragma	bytecode(BC_LOAD_LOCAL_32, inp_load_local_32)
@@ -1251,8 +1260,7 @@ __asm inp_load_local_8
 		tay
 		lda	(fp), y
 		sta	$00, x
-		ldy	tmpy
-		jmp	startup.yexec
+		jmp	startup.tyexec
 }
 		
 #pragma	bytecode(BC_LOAD_LOCAL_8, inp_load_local_8)
@@ -1269,8 +1277,7 @@ __asm inp_load_local_u8
 		sta	$00, x
 		lda #0
 		sta $01, x
-		ldy	tmpy
-		jmp	startup.yexec
+		jmp	startup.tyexec
 }
 		
 #pragma	bytecode(BC_LOAD_LOCAL_U8, inp_load_local_u8)
@@ -1285,8 +1292,7 @@ __asm inp_store_local_8
 		tay
 		lda	$00, x
 		sta	(fp), y
-		ldy	tmpy
-		jmp	startup.yexec
+		jmp	startup.tyexec
 }
 
 #pragma	bytecode(BC_STORE_LOCAL_8, inp_store_local_8)
@@ -1304,8 +1310,7 @@ __asm inp_store_local_16
 		iny
 		lda	$01, x
 		sta	(fp), y
-		ldy	tmpy
-		jmp	startup.yexec
+		jmp	startup.tyexec
 }
 
 #pragma	bytecode(BC_STORE_LOCAL_16, inp_store_local_16)
@@ -1329,8 +1334,7 @@ __asm inp_store_local_32
 		iny
 		lda	$03, x
 		sta	(fp), y
-		ldy	tmpy
-		jmp	startup.yexec
+		jmp	startup.tyexec
 }
 
 #pragma	bytecode(BC_STORE_LOCAL_32, inp_store_local_32)
@@ -1363,8 +1367,7 @@ __asm inp_store_frame_8
 		tay
 		lda	$00, x
 		sta	(sp), y
-		ldy	tmpy
-		jmp	startup.yexec
+		jmp	startup.tyexec
 }
 
 #pragma	bytecode(BC_STORE_FRAME_8, inp_store_frame_8)
@@ -1382,8 +1385,7 @@ __asm inp_store_frame_16
 		iny
 		lda	$01, x
 		sta	(sp), y
-		ldy	tmpy
-		jmp	startup.yexec
+		jmp	startup.tyexec
 }
 
 #pragma	bytecode(BC_STORE_FRAME_16, inp_store_frame_16)
@@ -1407,8 +1409,7 @@ __asm inp_store_frame_32
 		iny
 		lda	$03, x
 		sta	(sp), y
-		ldy	tmpy
-		jmp	startup.yexec
+		jmp	startup.tyexec
 }
 
 #pragma	bytecode(BC_STORE_FRAME_32, inp_store_frame_32)
diff --git a/include/crt.h b/include/crt.h
index ef6c3ef..f7766e7 100644
--- a/include/crt.h
+++ b/include/crt.h
@@ -26,6 +26,7 @@ enum ByteCode
 	BC_LOAD_ABS_U8,
 	BC_LOAD_ABS_16,
 	BC_LOAD_ABS_32,
+	BC_LOAD_ABS_ADDR,
 
 	BC_STORE_ABS_8,
 	BC_STORE_ABS_16,
diff --git a/oscar64/ByteCodeGenerator.cpp b/oscar64/ByteCodeGenerator.cpp
index 6d7f7a6..4cfc09d 100644
--- a/oscar64/ByteCodeGenerator.cpp
+++ b/oscar64/ByteCodeGenerator.cpp
@@ -26,6 +26,7 @@ static const char* ByteCodeNames[] = {
 	"LOAD_ABS_U8",
 	"LOAD_ABS_16",
 	"LOAD_ABS_32",
+	"LOAD_ABS_ADDR", 
 
 	"STORE_ABS_8",
 	"STORE_ABS_16",
@@ -147,9 +148,8 @@ static const char* ByteCodeNames[] = {
 	"SET_LT",
 	"SET_LE",
 
-	"JSR",		//113
+	"JSR",		//114
 
-	nullptr,
 	nullptr,
 	nullptr,
 
@@ -434,6 +434,9 @@ bool ByteCodeInstruction::CheckAccuSize(uint32 & used)
 		}
 		break;
 
+	case BC_LOAD_ABS_ADDR:
+		break;
+
 	case BC_LEA_ABS:
 	case BC_LEA_LOCAL:
 	case BC_LEA_FRAME:
@@ -954,6 +957,22 @@ void ByteCodeInstruction::Assemble(ByteCodeGenerator* generator, ByteCodeBasicBl
 		block->PutByte(mRegister);
 		break;
 
+	case BC_LOAD_ABS_ADDR:
+		block->PutCode(generator, mCode);
+		if (mRelocate)
+		{
+			LinkerReference	rl;
+			rl.mOffset = block->mCode.Size();
+			rl.mFlags = LREF_HIGHBYTE | LREF_LOWBYTE;
+			rl.mRefObject = mLinkerObject;
+			rl.mRefOffset = mValue;
+			block->mRelocations.Push(rl);
+			block->PutWord(0);
+		}
+		else
+			block->PutWord(mValue);
+		break;
+
 	case BC_LEA_ABS:
 	case BC_LEA_ABS_INDEX:
 	case BC_LEA_ABS_INDEX_U8:
@@ -4816,19 +4835,19 @@ bool ByteCodeBasicBlock::PeepHoleOptimizer(int phase)
 						mIns[i + 2].mCode = BC_NOP;
 						progress = true;
 					}
-					
+
 					else if (
 						mIns[i + 0].mCode == BC_LOAD_REG_8 &&
 						mIns[i + 1].mCode == BC_STORE_REG_16 &&
 						mIns[i + 2].mCode == BC_LEA_ABS_INDEX && mIns[i + 2].mRegister == mIns[i + 1].mRegister && !(mIns[i + 2].mLive & LIVE_ACCU) && mIns[i + 2].mRegisterFinal)
-					 {
-						 mIns[i + 2].mCode = BC_LEA_ABS_INDEX_U8;
-						 mIns[i + 2].mRegister = mIns[i + 0].mRegister;
-						 mIns[i + 2].mRegisterFinal = mIns[i + 0].mRegisterFinal;
-						 mIns[i + 0].mCode = BC_NOP;
-						 mIns[i + 1].mCode = BC_NOP;
-						 progress = true;
-					 }
+					{
+						mIns[i + 2].mCode = BC_LEA_ABS_INDEX_U8;
+						mIns[i + 2].mRegister = mIns[i + 0].mRegister;
+						mIns[i + 2].mRegisterFinal = mIns[i + 0].mRegisterFinal;
+						mIns[i + 0].mCode = BC_NOP;
+						mIns[i + 1].mCode = BC_NOP;
+						progress = true;
+					}
 
 					else if (
 						mIns[i + 0].mCode == BC_STORE_REG_16 &&
@@ -4872,7 +4891,7 @@ bool ByteCodeBasicBlock::PeepHoleOptimizer(int phase)
 						progress = true;
 					}
 					else if (
-						mIns[i + 0].mCode == BC_LOAD_REG_8 && 
+						mIns[i + 0].mCode == BC_LOAD_REG_8 &&
 						mIns[i + 1].mCode == BC_BINOP_SHRI_U16 &&
 						mIns[i + 2].mCode == BC_LOAD_REG_8 && mIns[i + 2].mRegister == BC_REG_ACCU)
 					{
@@ -4913,9 +4932,19 @@ bool ByteCodeBasicBlock::PeepHoleOptimizer(int phase)
 						mIns[i + 2].mCode = BC_BINOP_ADDA_16;
 						progress = true;
 					}
+					else if (
+						mIns[i + 0].mCode == BC_LOAD_REG_16 &&
+						mIns[i + 1].mCode == BC_BINOP_SHLI_16 && mIns[i + 1].mValue == 1 &&
+						mIns[i + 2].mCode == BC_STORE_REG_16 && mIns[i + 0].mRegister == mIns[i + 2].mRegister && !(mIns[i + 2].mLive & LIVE_ACCU))
+					{
+						mIns[i + 0].mRegisterFinal = false;
+						mIns[i + 1].mCode = BC_NOP;
+						mIns[i + 2].mCode = BC_BINOP_ADDA_16;
+						progress = true;
+					}
 					else if (
 						(mIns[i + 0].mCode == BC_LOAD_LOCAL_16 || mIns[i + 0].mCode == BC_LOAD_ABS_16 || mIns[i + 0].mCode == BC_LOAD_ADDR_16 ||
-						 mIns[i + 0].mCode == BC_LOAD_LOCAL_U8 || mIns[i + 0].mCode == BC_LOAD_ABS_U8 || mIns[i + 0].mCode == BC_LOAD_ADDR_U8) && 
+							mIns[i + 0].mCode == BC_LOAD_LOCAL_U8 || mIns[i + 0].mCode == BC_LOAD_ABS_U8 || mIns[i + 0].mCode == BC_LOAD_ADDR_U8) &&
 						mIns[i + 1].mCode == BC_BINOP_ADDR_16 && mIns[i + 0].mRegister == mIns[i + 1].mRegister && mIns[i + 1].mRegisterFinal &&
 						mIns[i + 2].mCode == BC_STORE_REG_16 && !(mIns[i + 2].mLive & LIVE_ACCU))
 					{
@@ -4936,6 +4965,18 @@ bool ByteCodeBasicBlock::PeepHoleOptimizer(int phase)
 						mIns[i + 1].mLive |= LIVE_ACCU;
 						progress = true;
 					}
+
+					else if (
+						mIns[i + 0].mCode == BC_LEA_ABS &&
+						mIns[i + 1].mCode == BC_BINOP_ADDA_16 && mIns[i + 0].mRegister == mIns[i + 1].mRegister &&
+						mIns[i + 2].mCode == BC_ADDR_REG && mIns[i + 0].mRegister == mIns[i + 2].mRegister && mIns[i + 2].mRegisterFinal)
+					{
+						mIns[i + 0].mCode = BC_LEA_ABS_INDEX;
+						mIns[i + 0].mRegister = BC_REG_ACCU;
+						mIns[i + 1].mCode = BC_NOP;
+						mIns[i + 2].mCode = BC_NOP;
+						progress = true;
+					}
 #if 1
 					else if (
 						i + 3 == mIns.Size() && mFalseJump &&
@@ -5283,7 +5324,26 @@ bool ByteCodeBasicBlock::PeepHoleOptimizer(int phase)
 						mIns[i + 0].mCode = BC_NOP;
 						progress = true;
 					}
+					else if (
+						mIns[i + 0].mCode == BC_BINOP_ADDA_16 &&
+						mIns[i + 1].mCode == BC_ADDR_REG && mIns[i + 0].mRegister == mIns[i + 1].mRegister && mIns[i + 1].mRegisterFinal)
+					{
+						mIns[i + 1].mCode = BC_LEA_ACCU_INDEX;
+						mIns[i + 0].mCode = BC_NOP;
+						progress = true;
+					}
 #endif
+#if 1
+					else if (
+						mIns[i + 0].mCode == BC_LOAD_ABS_16 && 
+						mIns[i + 1].mCode == BC_ADDR_REG && mIns[i + 0].mRegister == mIns[i + 1].mRegister && mIns[i + 1].mRegisterFinal)
+					{
+						mIns[i + 0].mCode = BC_LOAD_ABS_ADDR;
+						mIns[i + 1].mCode = BC_NOP;
+						progress = true;
+					}
+#endif
+
 #if 1
 					else if (
 						i + 2 == mIns.Size() && mFalseJump &&
diff --git a/oscar64/ByteCodeGenerator.h b/oscar64/ByteCodeGenerator.h
index 7e1cbf3..ddca152 100644
--- a/oscar64/ByteCodeGenerator.h
+++ b/oscar64/ByteCodeGenerator.h
@@ -27,6 +27,7 @@ enum ByteCode
 	BC_LOAD_ABS_U8,
 	BC_LOAD_ABS_16,
 	BC_LOAD_ABS_32,
+	BC_LOAD_ABS_ADDR,
 
 	BC_STORE_ABS_8,
 	BC_STORE_ABS_16,
diff --git a/oscar64/Disassembler.cpp b/oscar64/Disassembler.cpp
index d3fbd34..8daa508 100644
--- a/oscar64/Disassembler.cpp
+++ b/oscar64/Disassembler.cpp
@@ -155,6 +155,10 @@ void ByteCodeDisassembler::Disassemble(FILE* file, const uint8* memory, int star
 			fprintf(file, "MOVD\t%s, %s", TempName(memory[start + i + 2], tbuffer, proc), AddrName(uint16(memory[start + i + 0] + 256 * memory[start + i + 1]), abuffer, linker));
 			i += 3;
 			break;
+		case BC_LOAD_ABS_ADDR:
+			fprintf(file, "MOV\tADDR, %s", AddrName(uint16(memory[start + i + 0] + 256 * memory[start + i + 1]), abuffer, linker));
+			i += 2;
+			break;
 
 		case BC_LEA_ABS:
 			fprintf(file, "LEA\t%s, %s", TempName(memory[start + i + 0], tbuffer, proc), AddrName(uint16(memory[start + i + 1] + 256 * memory[start + i + 2]), abuffer, linker));