version 4.1

README.md

@@ -12,9 +12,18 @@ Prog8 - Structured Programming Language for 8-bit 6502/65c02 microprocessors
 
 This is a structured programming language for the 8-bit 6502/6510/65c02 microprocessor from the late 1970's and 1980's
 as used in many home computers from that era. It is a medium to low level programming language,
-which aims to provide many conveniences over raw assembly code (even when using a macro assembler):
+which aims to provide many conveniences over raw assembly code (even when using a macro assembler).
 
-- reduction of source code length
+Documentation
+-------------
+Full documentation (syntax reference, how to use the language and the compiler, etc.) can be found at:
+https://prog8.readthedocs.io/
+
+
+What use Prog8 provide?
+-----------------------
+
+- reduction of source code length over raw assembly
 - modularity, symbol scoping, subroutines
 - various data types other than just bytes (16-bit words, floats, strings)
 - automatic variable allocations, automatic string and array variables and string sharing
@@ -29,7 +38,7 @@ which aims to provide many conveniences over raw assembly code (even when using
 - inline assembly allows you to have full control when every cycle or byte matters
 - many built-in functions such as ``sin``, ``cos``, ``rnd``, ``abs``, ``min``, ``max``, ``sqrt``, ``msb``, ``rol``, ``ror``, ``swap``, ``memset``, ``memcopy``, ``sort`` and ``reverse``
 
-Rapid edit-compile-run-debug cycle:
+*Rapid edit-compile-run-debug cycle:*
 
 - use a modern PC to do the work on
 - very quick compilation times
@@ -37,17 +46,16 @@ Rapid edit-compile-run-debug cycle:
 - breakpoints, that let the Vice emulator drop into the monitor if execution hits them
 - source code labels automatically loaded in Vice emulator so it can show them in disassembly
 
-Prog8 is mainly targeted at the Commodore-64 machine.
-Preliminary support for the [CommanderX16](https://www.commanderx16.com) is available as a second compilation target.
-Contributions to improve these or to add support for other machines are welcome!
+*Two supported compiler targets* (contributions to improve these or to add support for other machines are welcome!):
+
+- "c64": Commodore-64  (6510 CPU = almost a 6502)  premium support.
+- "cx16": [CommanderX16](https://www.commanderx16.com)  (65c02 CPU)  experimental support.
 
 
-Documentation/manual
---------------------
-https://prog8.readthedocs.io/
 
-Required tools
---------------
+
+Additional required tools
+-------------------------
 
 [64tass](https://sourceforge.net/projects/tass64/) - cross assembler. Install this on your shell path.
 A recent .exe version of this tool for Windows can be obtained from my [clone](https://github.com/irmen/64tass/releases) of this project.

compiler/res/prog8lib/c64floats.asm

@@ -8,7 +8,8 @@ ub2float	.proc
 		sta  P8ZP_SCRATCH_W2
 		sty  P8ZP_SCRATCH_W2+1
 		ldy  P8ZP_SCRATCH_B1
-		jsr  FREADUY
+		lda  #0
+		jsr  GIVAYF
 _fac_to_mem	ldx  P8ZP_SCRATCH_W2
 		ldy  P8ZP_SCRATCH_W2+1
 		jsr  MOVMF
@@ -74,7 +75,8 @@ stack_ub2float	.proc
 		lda  P8ESTACK_LO,x
 		stx  P8ZP_SCRATCH_REG_X
 		tay
-		jsr  FREADUY
+		lda  #0
+		jsr  GIVAYF
 		jmp  push_fac1_as_result
 		.pend
 
@@ -663,8 +665,8 @@ _largest_pos_float	.byte  255,127,255,255,255		; largest positive float
 		.pend
 
 func_sum_f	.proc
-		lda  #<FL_ZERO
-		ldy  #>FL_ZERO
+		lda  #<ZERO
+		ldy  #>ZERO
 		jsr  MOVFM
 		jsr  prog8_lib.pop_array_and_lengthmin1Y
 		stx  P8ZP_SCRATCH_REG_X

compiler/res/prog8lib/c64flt.p8

@@ -6,12 +6,12 @@
 
 %option enable_floats
 
-
 c64flt {
 	; ---- this block contains C-64 floating point related functions ----
 
         const  float  PI        = 3.141592653589793
         const  float  TWOPI	    = 6.283185307179586
+        const  float  ZERO      = 0.0
 
 
 ; ---- C64 basic and kernal ROM float constants and functions ----
@@ -35,13 +35,11 @@ c64flt {
 		&float  FL_TWOPI	= $e2e5  ; 2 * PI
 		&float  FL_FR4		= $e2ea  ; .25
 		; oddly enough, 0.0 isn't available in the kernel.
-        float   FL_ZERO     = 0.0    ; oddly enough 0.0 isn't available in the kernel
 
 
 ; note: fac1/2 might get clobbered even if not mentioned in the function's name.
 ; note: for subtraction and division, the left operand is in fac2, the right operand in fac1.
 
-; checked functions below:
 romsub $bba2 = MOVFM(uword mflpt @ AY) clobbers(A,Y)        ; load mflpt value from memory  in A/Y into fac1
 romsub $bba6 = FREADMEM() clobbers(A,Y)                     ; load mflpt value from memory  in $22/$23 into fac1
 romsub $ba8c = CONUPK(uword mflpt @ AY) clobbers(A,Y)       ; load mflpt value from memory  in A/Y into fac2
@@ -91,6 +89,7 @@ romsub $bb12 = FDIVT() clobbers(A,X,Y)                      ; fac1 = fac2/fac1
 romsub $bb0f = FDIV(uword mflpt @ AY) clobbers(A,X,Y)       ; fac1 = mflpt in A/Y / fac1  (remainder in fac2)
 romsub $bf7b = FPWRT() clobbers(A,X,Y)                      ; fac1 = fac2 ** fac1
 romsub $bf78 = FPWR(uword mflpt @ AY) clobbers(A,X,Y)       ; fac1 = fac2 ** mflpt from A/Y
+romsub $bd7e = FINLOG(byte value @A) clobbers (A, X, Y)     ; fac1 += signed byte in A
 
 romsub $aed4 = NOTOP() clobbers(A,X,Y)                      ; fac1 = NOT(fac1)
 romsub $bccc = INT() clobbers(A,X,Y)                        ; INT() truncates, use FADDH first to round instead of trunc
@@ -193,33 +192,26 @@ asmsub  GETADRAY  () clobbers(X) -> uword @ AY  {
 }
 
 sub  print_f  (float value) {
-	; ---- prints the floating point value (without a newline) using basic rom routines.
+	; ---- prints the floating point value (without a newline).
 	%asm {{
 		stx  P8ZP_SCRATCH_REG_X
 		lda  #<value
 		ldy  #>value
 		jsr  MOVFM		; load float into fac1
 		jsr  FOUT		; fac1 to string in A/Y
-		jsr  c64.STROUT			; print string in A/Y
+		sta  P8ZP_SCRATCH_B1
+		sty  P8ZP_SCRATCH_REG
+		ldy  #0
+-		lda  (P8ZP_SCRATCH_B1),y
+		beq  +
+		jsr  c64.CHROUT
+		iny
+		bne  -
 		ldx  P8ZP_SCRATCH_REG_X
-		rts
++		rts
 	}}
 }
 
-sub  print_fln  (float value) {
-	; ---- prints the floating point value (with a newline at the end) using basic rom routines
-	%asm {{
-		stx  P8ZP_SCRATCH_REG_X
-		lda  #<value
-		ldy  #>value
-		jsr  MOVFM		; load float into fac1
-		jsr  FPRINTLN		; print fac1 with newline
-		ldx  P8ZP_SCRATCH_REG_X
-		rts
-	}}
-
-}
-
 %asminclude "library:c64floats.asm", ""
 
-}  ; ------ end of block c64flt
+}

compiler/res/prog8lib/c64graphics.p8

@@ -0,0 +1,243 @@
+%import c64textio
+
+; bitmap pixel graphics module for the C64
+; only black/white monchrome for now
+; assumes bitmap screen memory is $2000-$3fff
+
+graphics {
+    const uword bitmap_address = $2000
+
+    sub enable_bitmap_mode() {
+        ; enable bitmap screen, erase it and set colors to black/white.
+        c64.SCROLY |= %00100000
+        c64.VMCSB = (c64.VMCSB & %11110000) | %00001000   ; $2000-$3fff
+        clear_screen(1, 0)
+    }
+
+    sub clear_screen(ubyte pixelcolor, ubyte bgcolor) {
+        memset(bitmap_address, 320*200/8, 0)
+        txt.fill_screen(pixelcolor << 4 | bgcolor, 0)
+    }
+
+    sub line(uword @zp x1, ubyte @zp y1, uword @zp x2, ubyte @zp y2) {
+        ; Bresenham algorithm.
+        ; This code special cases various quadrant loops to allow simple ++ and -- operations.
+        ; TODO rewrite this in optimized assembly
+        if y1>y2 {
+            ; make sure dy is always positive to avoid 8 instead of just 4 special cases
+            swap(x1, x2)
+            swap(y1, y2)
+        }
+        word @zp d = 0
+        ubyte positive_ix = true
+        word @zp dx = x2 - x1 as word
+        word @zp dy = y2 as word - y1 as word
+        if dx < 0 {
+            dx = -dx
+            positive_ix = false
+        }
+        dx *= 2
+        dy *= 2
+        internal_plotx = x1
+
+        if dx >= dy {
+            if positive_ix {
+                repeat {
+                    internal_plot(y1)
+                    if internal_plotx==x2
+                        return
+                    internal_plotx++
+                    d += dy
+                    if d > dx {
+                        y1++
+                        d -= dx
+                    }
+                }
+            } else {
+                repeat {
+                    internal_plot(y1)
+                    if internal_plotx==x2
+                        return
+                    internal_plotx--
+                    d += dy
+                    if d > dx {
+                        y1++
+                        d -= dx
+                    }
+                }
+            }
+        }
+        else {
+            if positive_ix {
+                repeat {
+                    internal_plot(y1)
+                    if y1 == y2
+                        return
+                    y1++
+                    d += dx
+                    if d > dy {
+                        internal_plotx++
+                        d -= dy
+                    }
+                }
+            } else {
+                repeat {
+                    internal_plot(y1)
+                    if y1 == y2
+                        return
+                    y1++
+                    d += dx
+                    if d > dy {
+                        internal_plotx--
+                        d -= dy
+                    }
+                }
+            }
+        }
+    }
+
+    sub circle(uword xcenter, ubyte ycenter, ubyte radius) {
+        ; Midpoint algorithm
+        ubyte @zp ploty
+        ubyte @zp xx = radius
+        ubyte @zp yy = 0
+        byte @zp decisionOver2 = 1-xx as byte
+
+        while xx>=yy {
+            internal_plotx = xcenter + xx
+            ploty = ycenter + yy
+            internal_plot(ploty)
+            internal_plotx = xcenter - xx
+            internal_plot(ploty)
+            internal_plotx = xcenter + xx
+            ploty = ycenter - yy
+            internal_plot(ploty)
+            internal_plotx = xcenter - xx
+            internal_plot(ploty)
+            internal_plotx = xcenter + yy
+            ploty = ycenter + xx
+            internal_plot(ploty)
+            internal_plotx = xcenter - yy
+            internal_plot(ploty)
+            internal_plotx = xcenter + yy
+            ploty = ycenter - xx
+            internal_plot(ploty)
+            internal_plotx = xcenter - yy
+            internal_plot(ploty)
+            yy++
+            if decisionOver2<=0
+                decisionOver2 += 2*yy+1
+            else {
+                xx--
+                decisionOver2 += 2*(yy-xx)+1
+            }
+        }
+    }
+
+    sub disc(uword cx, ubyte cy, ubyte radius) {
+        ; Midpoint algorithm, filled
+        ubyte xx = radius
+        ubyte yy = 0
+        byte decisionOver2 = 1-xx as byte
+
+        while xx>=yy {
+            ubyte cy_plus_yy = cy + yy
+            ubyte cy_min_yy = cy - yy
+            ubyte cy_plus_xx = cy + xx
+            ubyte cy_min_xx = cy - xx
+
+            for internal_plotx in cx to cx+xx {
+                internal_plot(cy_plus_yy)
+                internal_plot(cy_min_yy)
+            }
+            for internal_plotx in cx-xx to cx-1 {
+                internal_plot(cy_plus_yy)
+                internal_plot(cy_min_yy)
+            }
+            for internal_plotx in cx to cx+yy {
+                internal_plot(cy_plus_xx)
+                internal_plot(cy_min_xx)
+            }
+            for internal_plotx in cx-yy to cx {
+                internal_plot(cy_plus_xx)
+                internal_plot(cy_min_xx)
+            }
+            yy++
+            if decisionOver2<=0
+                decisionOver2 += 2*yy+1
+            else {
+                xx--
+                decisionOver2 += 2*(yy-xx)+1
+            }
+        }
+    }
+
+
+; here is the non-asm code for the plot routine below:
+;    sub plot_nonasm(uword px, ubyte py) {
+;        ubyte[] ormask = [128, 64, 32, 16, 8, 4, 2, 1]
+;        uword addr = bitmap_address + 320*(py>>3) + (py & 7) + (px & %0000000111111000)
+;        @(addr) |= ormask[lsb(px) & 7]
+;    }
+
+    ; TODO fix the use of X (or XY) as parameter so we can actually use this plot() routine
+    ;   calling it with a byte results in a compiler crash, calling it with word results in clobbering X register I think
+    asmsub  plotXXX(uword plotx @XY, ubyte ploty @A) {
+        %asm {{
+            stx  internal_plotx
+            sty  internal_plotx+1
+            jmp  internal_plot
+        }}
+    }
+
+    uword internal_plotx     ; 0..319        ; separate 'parameter' for internal_plot()
+
+    asmsub  internal_plot(ubyte ploty @A) {      ; internal_plotx is 16 bits 0 to 319... doesn't fit in a register
+        %asm {{
+        tay
+        stx  P8ZP_SCRATCH_REG_X
+        lda  internal_plotx+1
+        sta  P8ZP_SCRATCH_W2+1
+        lsr  a            ; 0
+        sta  P8ZP_SCRATCH_W2
+        lda  internal_plotx
+        pha
+        and  #7
+        tax
+
+        lda  _y_lookup_lo,y
+        clc
+        adc  P8ZP_SCRATCH_W2
+        sta  P8ZP_SCRATCH_W2
+        lda  _y_lookup_hi,y
+        adc  P8ZP_SCRATCH_W2+1
+        sta  P8ZP_SCRATCH_W2+1
+
+        pla     ; internal_plotx
+        and  #%11111000
+        tay
+        lda  (P8ZP_SCRATCH_W2),y
+        ora  _ormask,x
+        sta  (P8ZP_SCRATCH_W2),y
+
+        ldx  P8ZP_SCRATCH_REG_X
+        rts
+
+_ormask     .byte 128, 64, 32, 16, 8, 4, 2, 1
+
+; note: this can be even faster if we also have a 256 byte x-lookup table, but hey.
+; see http://codebase64.org/doku.php?id=base:various_techniques_to_calculate_adresses_fast_common_screen_formats_for_pixel_graphics
+; the y lookup tables encodes this formula:  bitmap_address + 320*(py>>3) + (py & 7)    (y from 0..199)
+; We use the 64tass syntax for range expressions to calculate this table on assembly time.
+
+_plot_y_values := $2000 + 320*(range(200)>>3) + (range(200) & 7)
+
+_y_lookup_lo    .byte  <_plot_y_values
+_y_lookup_hi    .byte  >_plot_y_values
+
+        }}
+    }
+
+}
+
+

compiler/res/prog8lib/c64textio.p8

@@ -11,8 +11,16 @@
 
 txt {
 
-asmsub  clear_screen (ubyte char @ A, ubyte color @ Y) clobbers(A)  {
-	; ---- clear the character screen with the given fill character and character color.
+asmsub  clear_screen() {
+    %asm {{
+        lda  #' '
+        jmp  clear_screenchars
+    }}
+}
+
+
+asmsub  fill_screen (ubyte char @ A, ubyte charcolor @ Y) clobbers(A)  {
+	; ---- fill the character screen with the given fill character and character color.
 	;      (assumes screen and color matrix are at their default addresses)
 
 	%asm {{
@@ -41,7 +49,7 @@ _loop		sta  c64.Screen,y
         }}
 }
 
-asmsub  clear_screencolors (ubyte color @ A) clobbers(Y)  {
+asmsub  clear_screencolors (ubyte scrcolor @ A) clobbers(Y)  {
 	; ---- clear the character screen colors with the given color (leaves characters).
 	;      (assumes color matrix is at the default address)
 	%asm {{
@@ -56,6 +64,10 @@ _loop		sta  c64.Colors,y
         }}
 }
 
+sub color (ubyte txtcol) {
+    c64.COLOR = txtcol
+}
+
 asmsub  scroll_left_full  (ubyte alsocolors @ Pc) clobbers(A, Y)  {
 	; ---- scroll the whole screen 1 character to the left
 	;      contents of the rightmost column are unchanged, you should clear/refill this yourself
@@ -504,7 +516,7 @@ _mod		lda  $ffff		; modified
 	}}
 }
 
-sub  setcc  (ubyte column, ubyte row, ubyte char, ubyte color)  {
+sub  setcc  (ubyte column, ubyte row, ubyte char, ubyte charcolor)  {
 	; ---- set char+color at the given position on the screen
 	%asm {{
 		lda  row
@@ -524,7 +536,7 @@ sub  setcc  (ubyte column, ubyte row, ubyte char, ubyte color)  {
 		inc  _colormod+2
 +		lda  char
 _charmod	sta  $ffff		; modified
-		lda  color
+		lda  charcolor
 _colormod	sta  $ffff		; modified
 		rts
 	}}

compiler/res/prog8lib/cx16flt.p8

@@ -0,0 +1,153 @@
+; Prog8 definitions for floating point handling on the CommanderX16
+;
+; Written by Irmen de Jong (irmen@razorvine.net) - license: GNU GPL 3.0
+;
+; indent format: TABS, size=8
+
+%option enable_floats
+
+c64flt {
+	; ---- this block contains C-64 floating point related functions ----
+
+        const  float  PI        = 3.141592653589793
+        const  float  TWOPI	    = 6.283185307179586
+        const  float  ZERO      = 0.0
+
+
+; ---- ROM float functions ----
+
+		; note: the fac1 and fac2 are working registers and take 6 bytes each,
+		; floats in memory  (and rom) are stored in 5-byte MFLPT packed format.
+
+; note: fac1/2 might get clobbered even if not mentioned in the function's name.
+; note: for subtraction and division, the left operand is in fac2, the right operand in fac1.
+
+romsub $fe00 = AYINT() clobbers(A,X,Y)          ; fac1-> signed word in 100-101 ($64-$65) MSB FIRST. (might throw ILLEGAL QUANTITY)
+
+; GIVAYF: signed word in Y/A (note different lsb/msb order) -> float in fac1
+; there is also c64flt.GIVUAYFAY - unsigned word in A/Y (lo/hi) to fac1
+; (tip: use GIVAYFAY to use A/Y input; lo/hi switched to normal order)
+romsub $fe03 = GIVAYF(ubyte lo @ Y, ubyte hi @ A) clobbers(A,X,Y)
+
+; fac1 -> unsigned word in Y/A (might throw ILLEGAL QUANTITY) (result also in $14/15)
+; (tip: use GETADRAY to get A/Y output; lo/hi switched to normal little endian order)
+romsub $fe06 = GETADR() clobbers(X) -> ubyte @ Y, ubyte @ A
+
+romsub $fe09 = FADDH() clobbers(A,X,Y)                      ; fac1 += 0.5, for rounding- call this before INT
+romsub $fe0c = FSUB(uword mflpt @ AY) clobbers(A,X,Y)       ; fac1 = mflpt from A/Y - fac1
+romsub $fe0f = FSUBT() clobbers(A,X,Y)                      ; fac1 = fac2-fac1   mind the order of the operands
+romsub $fe12 = FADD(uword mflpt @ AY) clobbers(A,X,Y)       ; fac1 += mflpt value from A/Y
+romsub $fe15 = FADDT() clobbers(A,X,Y)                      ; fac1 += fac2
+romsub $fe1b = ZEROFC() clobbers(A,X,Y)                     ; fac1 = 0
+romsub $fe1e = NORMAL() clobbers(A,X,Y)                     ; normalize fac1 (?)
+romsub $fe24 = LOG() clobbers(A,X,Y)                        ; fac1 = LN(fac1)  (natural log)
+romsub $fe27 = FMULT(uword mflpt @ AY) clobbers(A,X,Y)      ; fac1 *= mflpt value from A/Y
+romsub $fe2a = FMULTT() clobbers(A,X,Y)                     ; fac1 *= fac2
+romsub $fe33 = CONUPK(uword mflpt @ AY) clobbers(A,Y)       ; load mflpt value from memory  in A/Y into fac2
+romsub $fe36 = MUL10() clobbers(A,X,Y)                      ; fac1 *= 10
+romsub $fe3c = DIV10() clobbers(A,X,Y)                      ; fac1 /= 10 , CAUTION: result is always positive!
+romsub $fe3f = FDIV(uword mflpt @ AY) clobbers(A,X,Y)       ; fac1 = mflpt in A/Y / fac1  (remainder in fac2)
+romsub $fe42 = FDIVT() clobbers(A,X,Y)                      ; fac1 = fac2/fac1  (remainder in fac2)  mind the order of the operands
+
+romsub $fe48 = MOVFM(uword mflpt @ AY) clobbers(A,Y)        ; load mflpt value from memory  in A/Y into fac1
+romsub $fe4b = MOVMF(uword mflpt @ XY) clobbers(A,Y)        ; store fac1 to memory  X/Y as 5-byte mflpt
+romsub $fe4e = MOVFA() clobbers(A,X)                        ; copy fac2 to fac1
+romsub $fe51 = MOVAF() clobbers(A,X)                        ; copy fac1 to fac2  (rounded)
+romsub $fe54 = MOVEF() clobbers(A,X)                        ; copy fac1 to fac2
+romsub $fe5a = SIGN() -> ubyte @ A                          ; SIGN(fac1) to A, $ff, $0, $1 for negative, zero, positive
+romsub $fe5d = SGN() clobbers(A,X,Y)                        ; fac1 = SGN(fac1), result of SIGN (-1, 0 or 1)
+romsub $fe60 = FREADSA(byte value @ A) clobbers(A,X,Y)      ; 8 bit signed A -> float in fac1
+romsub $fe6c = ABS()                                        ; fac1 = ABS(fac1)
+romsub $fe6f = FCOMP(uword mflpt @ AY) clobbers(X,Y) -> ubyte @ A   ; A = compare fac1 to mflpt in A/Y, 0=equal 1=fac1 is greater, 255=fac1 is less than
+romsub $fe78 = INT() clobbers(A,X,Y)                        ; INT() truncates, use FADDH first to round instead of trunc
+romsub $fe7e = FINLOG(byte value @A) clobbers (A, X, Y)           ; fac1 += signed byte in A
+romsub $fe81 = FOUT() clobbers(X) -> uword @ AY             ; fac1 -> string, address returned in AY
+romsub $fe8a = SQR() clobbers(A,X,Y)                        ; fac1 = SQRT(fac1)
+romsub $fe8d = FPWRT() clobbers(A,X,Y)                      ; fac1 = fac2 ** fac1
+romsub $fe93 = NEGOP() clobbers(A)                          ; switch the sign of fac1
+romsub $fe96 = EXP() clobbers(A,X,Y)                        ; fac1 = EXP(fac1)  (e ** fac1)
+romsub $fe9f = RND2(byte value @A) clobbers(A,X,Y)                ; fac1 = RND(A) float random number generator
+romsub $fea2 = RND() clobbers(A,X,Y)                        ; fac1 = RND(fac1) float random number generator
+romsub $fea5 = COS() clobbers(A,X,Y)                        ; fac1 = COS(fac1)
+romsub $fea8 = SIN() clobbers(A,X,Y)                        ; fac1 = SIN(fac1)
+romsub $feab = TAN() clobbers(A,X,Y)                        ; fac1 = TAN(fac1)
+romsub $feae = ATN() clobbers(A,X,Y)                        ; fac1 = ATN(fac1)
+
+
+asmsub  GIVUAYFAY  (uword value @ AY) clobbers(A,X,Y)  {
+	; ---- unsigned 16 bit word in A/Y (lo/hi) to fac1
+	%asm {{
+        phx
+        sta  P8ZP_SCRATCH_REG
+        sty  P8ZP_SCRATCH_B1
+        tya
+        ldy  P8ZP_SCRATCH_REG
+        jsr  GIVAYF                 ; load it as signed... correct afterwards
+        lda  P8ZP_SCRATCH_B1
+	    bpl  +
+	    lda  #<_flt65536
+	    ldy  #>_flt65536
+	    jsr  FADD
++	    plx
+        rts
+_flt65536    .byte 145,0,0,0,0       ; 65536.0
+	}}
+}
+
+
+asmsub  GIVAYFAY  (uword value @ AY) clobbers(A,X,Y)  {
+	; ---- signed 16 bit word in A/Y (lo/hi) to float in fac1
+	%asm {{
+		sta  P8ZP_SCRATCH_REG
+		tya
+		ldy  P8ZP_SCRATCH_REG
+		jmp  GIVAYF		; this uses the inverse order, Y/A
+	}}
+}
+
+asmsub  FTOSWRDAY  () clobbers(X) -> uword @ AY  {
+	; ---- fac1 to signed word in A/Y
+	%asm {{
+		jsr  FTOSWORDYA	; note the inverse Y/A order
+		sta  P8ZP_SCRATCH_REG
+		tya
+		ldy  P8ZP_SCRATCH_REG
+		rts
+	}}
+}
+
+asmsub  GETADRAY  () clobbers(X) -> uword @ AY  {
+	; ---- fac1 to unsigned word in A/Y
+	%asm {{
+		jsr  GETADR		; this uses the inverse order, Y/A
+		sta  P8ZP_SCRATCH_B1
+		tya
+		ldy  P8ZP_SCRATCH_B1
+		rts
+	}}
+}
+
+sub  print_f  (float value) {
+	; ---- prints the floating point value (without a newline).
+	%asm {{
+		stx  P8ZP_SCRATCH_REG_X
+		lda  #<value
+		ldy  #>value
+		jsr  MOVFM		; load float into fac1
+		jsr  FOUT		; fac1 to string in A/Y
+		sta  P8ZP_SCRATCH_B1
+		sty  P8ZP_SCRATCH_REG
+		ldy  #0
+-		lda  (P8ZP_SCRATCH_B1),y
+		beq  +
+		jsr  c64.CHROUT
+		iny
+		bne  -
+		ldx  P8ZP_SCRATCH_REG_X
++		rts
+	}}
+}
+
+%asminclude "library:c64floats.asm", ""
+
+}

compiler/res/prog8lib/cx16lib.p8

@@ -62,63 +62,72 @@ cx16 {
 ; spelling of the names is taken from the Commander X-16 rom sources
 
 ; the sixteen virtual 16-bit registers
-&ubyte r0  = $02
-&ubyte r0L = $02
-&ubyte r0H = $03
-&ubyte r1  = $04
-&ubyte r1L = $04
-&ubyte r1H = $05
-&ubyte r2  = $06
-&ubyte r2L = $06
-&ubyte r2H = $07
-&ubyte r3  = $08
-&ubyte r3L = $08
-&ubyte r3H = $09
-&ubyte r4  = $0a
-&ubyte r4L = $0a
-&ubyte r4H = $0b
-&ubyte r5  = $0c
-&ubyte r5L = $0c
-&ubyte r5H = $0d
-&ubyte r6  = $0e
-&ubyte r6L = $0e
-&ubyte r6H = $0f
-&ubyte r7  = $10
-&ubyte r7L = $10
-&ubyte r7H = $11
-&ubyte r8  = $12
-&ubyte r8L = $12
-&ubyte r8H = $13
-&ubyte r9  = $14
-&ubyte r9L = $14
-&ubyte r9H = $15
-&ubyte r10 = $16
-&ubyte r10L    = $16
-&ubyte r10H    = $17
-&ubyte r11     = $18
-&ubyte r11L    = $18
-&ubyte r11H    = $19
-&ubyte r12     = $1a
-&ubyte r12L    = $1a
-&ubyte r12H    = $1b
-&ubyte r13     = $1c
-&ubyte r13L    = $1c
-&ubyte r13H    = $1d
-&ubyte r14     = $1e
-&ubyte r14L    = $1e
-&ubyte r14H    = $1f
-&ubyte r15     = $20
-&ubyte r15L    = $20
-&ubyte r15H    = $21
+    &uword r0  = $02
+    &uword r1  = $04
+    &uword r2  = $06
+    &uword r3  = $08
+    &uword r4  = $0a
+    &uword r5  = $0c
+    &uword r6  = $0e
+    &uword r7  = $10
+    &uword r8  = $12
+    &uword r9  = $14
+    &uword r10 = $16
+    &uword r11 = $18
+    &uword r12 = $1a
+    &uword r13 = $1c
+    &uword r14 = $1e
+    &uword r15 = $20
 
+; VERA registers
+
+    const uword VERA_BASE = $9F20
+    &uword VERA_ADDR_L   	 = VERA_BASE + $00
+    &uword VERA_ADDR_M   	 = VERA_BASE + $01
+    &uword VERA_ADDR_H   	 = VERA_BASE + $02
+    &uword VERA_DATA0        = VERA_BASE + $03
+    &uword VERA_DATA1        = VERA_BASE + $04
+    &uword VERA_CTRL         = VERA_BASE + $05
+    &uword VERA_IEN          = VERA_BASE + $06
+    &uword VERA_ISR          = VERA_BASE + $07
+    &uword VERA_IRQ_LINE_L   = VERA_BASE + $08
+    &uword VERA_DC_VIDEO     = VERA_BASE + $09
+    &uword VERA_DC_HSCALE    = VERA_BASE + $0A
+    &uword VERA_DC_VSCALE    = VERA_BASE + $0B
+    &uword VERA_DC_BORDER    = VERA_BASE + $0C
+    &uword VERA_DC_HSTART    = VERA_BASE + $09
+    &uword VERA_DC_HSTOP     = VERA_BASE + $0A
+    &uword VERA_DC_VSTART    = VERA_BASE + $0B
+    &uword VERA_DC_VSTOP     = VERA_BASE + $0C
+    &uword VERA_L0_CONFIG    = VERA_BASE + $0D
+    &uword VERA_L0_MAPBASE   = VERA_BASE + $0E
+    &uword VERA_L0_TILEBASE  = VERA_BASE + $0F
+    &uword VERA_L0_HSCROLL_L = VERA_BASE + $10
+    &uword VERA_L0_HSCROLL_H = VERA_BASE + $11
+    &uword VERA_L0_VSCROLL_L = VERA_BASE + $12
+    &uword VERA_L0_VSCROLL_H = VERA_BASE + $13
+    &uword VERA_L1_CONFIG    = VERA_BASE + $14
+    &uword VERA_L1_MAPBASE   = VERA_BASE + $15
+    &uword VERA_L1_TILEBASE  = VERA_BASE + $16
+    &uword VERA_L1_HSCROLL_L = VERA_BASE + $17
+    &uword VERA_L1_HSCROLL_H = VERA_BASE + $18
+    &uword VERA_L1_VSCROLL_L = VERA_BASE + $19
+    &uword VERA_L1_VSCROLL_H = VERA_BASE + $1A
+    &uword VERA_AUDIO_CTRL   = VERA_BASE + $1B
+    &uword VERA_AUDIO_RATE   = VERA_BASE + $1C
+    &uword VERA_AUDIO_DATA   = VERA_BASE + $1D
+    &uword VERA_SPI_DATA     = VERA_BASE + $1E
+    &uword VERA_SPI_CTRL     = VERA_BASE + $1F
+; VERA_PSG_BASE     = $1F9C0
+; VERA_PALETTE_BASE = $1FA00
+; VERA_SPRITES_BASE = $1FC00
 
-; TODO subroutine args + soubroutine returnvalues + clobber registers
 
 ; supported C128 additions
 romsub $ff4a = close_all()
 romsub $ff59 = lkupla()
 romsub $ff5c = lkupsa()
-romsub $ff5f = screen_set_mode()
+romsub $ff5f = screen_set_mode(ubyte mode @A) clobbers(A, X, Y) -> ubyte @Pc
 romsub $ff62 = screen_set_charset(ubyte charset @A, uword charsetptr @XY) clobbers(A,X,Y)      ; incompatible with C128  dlchr()
 romsub $ff65 = pfkey()
 romsub $ff6e = jsrfar()
@@ -129,44 +138,59 @@ romsub $ff7d = primm()
 
 ; X16 additions
 romsub $ff44 = macptr()
-romsub $ff47 = enter_basic()
-romsub $ff68 = mouse_config()
-romsub $ff6b = mouse_get()
-romsub $ff71 = mouse_scan()
-romsub $ff53 = joystick_scan()
-romsub $ff56 = joystick_get()
-romsub $ff4d = clock_set_date_time()
-romsub $ff50 = clock_get_date_time()
+romsub $ff47 = enter_basic(ubyte cold_or_warm @Pc)
+romsub $ff68 = mouse_config(ubyte shape @A, ubyte scale @X) clobbers (A, X, Y)
+romsub $ff6b = mouse_get(ubyte zpdataptr @X) clobbers(A)
+romsub $ff71 = mouse_scan() clobbers(A, X, Y)
+romsub $ff53 = joystick_scan() clobbers(A, X, Y)
+romsub $ff56 = joystick_get(ubyte joynr @A) -> ubyte @A, ubyte @X, ubyte @Y
+romsub $ff4d = clock_set_date_time() clobbers(A, X, Y)      ; args: r0, r1, r2, r3L
+romsub $ff50 = clock_get_date_time() clobbers(A)            ; outout args: r0, r1, r2, r3L
 
 ; high level graphics & fonts
-romsub $ff20 = GRAPH_init()
+romsub $ff20 = GRAPH_init()             ; uses vectors=r0
 romsub $ff23 = GRAPH_clear()
-romsub $ff26 = GRAPH_set_window()
-romsub $ff29 = GRAPH_set_colors()
-romsub $ff2c = GRAPH_draw_line()
-romsub $ff2f = GRAPH_draw_rect()
-romsub $ff32 = GRAPH_move_rect()
-romsub $ff35 = GRAPH_draw_oval()
-romsub $ff38 = GRAPH_draw_image()
-romsub $ff3b = GRAPH_set_font()
-romsub $ff3e = GRAPH_get_char_size()
-romsub $ff41 = GRAPH_put_char()
+romsub $ff26 = GRAPH_set_window()       ; uses x=r0, y=r1, width=r2, height=r3
+romsub $ff29 = GRAPH_set_colors(ubyte stroke @A, ubyte fill @X, ubyte background @Y)
+romsub $ff2c = GRAPH_draw_line()        ; uses x1=r0, y1=r1, x2=r2, y2=r3
+romsub $ff2f = GRAPH_draw_rect(ubyte fill @Pc)        ; uses x=r0, y=r1, width=r2, height=r3, cornerradius=r4
+romsub $ff32 = GRAPH_move_rect()        ; uses sx=r0, sy=r1, tx=r2, ty=r3, width=r4, height=r5
+romsub $ff35 = GRAPH_draw_oval(ubyte fill @Pc)        ; uses x=r0, y=r1, width=r2, height=r3
+romsub $ff38 = GRAPH_draw_image()       ; uses x=r0, y=r1, ptr=r2, width=r3, height=r4
+romsub $ff3b = GRAPH_set_font()         ; uses ptr=r0
+romsub $ff3e = GRAPH_get_char_size(ubyte baseline @A, ubyte width @X, ubyte height_or_style @Y, ubyte is_control @Pc)
+romsub $ff41 = GRAPH_put_char(ubyte char @A)   ; uses x=r0, y=r1
 
-; TODO framebuffer API not yet included, include it
+; framebuffer
+romsub $fef6 = FB_init()
+romsub $fef9 = FB_get_info() -> byte @A    ; also outputs width=r0, height=r1
+romsub $fefc = FB_set_palette(ubyte index @A, ubyte bytecount @X)      ; also uses pointer=r0
+romsub $feff = FB_cursor_position()    ; uses x=r0, y=r1
+romsub $ff02 = FB_cursor_next_line()   ; uses x=r0
+romsub $ff05 = FB_get_pixel() -> ubyte @A
+romsub $ff08 = FB_get_pixels()         ; uses ptr=r0, count=r1
+romsub $ff0b = FB_set_pixel(ubyte color @A)
+romsub $ff0e = FB_set_pixels()         ; uses ptr=r0, count=r1
+romsub $ff11 = FB_set_8_pixels(ubyte pattern @A, ubyte color @X)
+romsub $ff14 = FB_set_8_pixels_opaque(ubyte pattern @A, ubyte color1 @X, ubyte color2 @Y)  ; also uses mask=r0L
+romsub $ff17 = FB_fill_pixels(ubyte color @A)   ; also uses count=r0, step=r1
+romsub $ff1a = FB_filter_pixels()      ; uses ptr=r0, count=r1
+romsub $ff1d = FB_move_pixels()        ; uses sx=r0, sy=r1, tx=r2, ty=r3, count=r4
 
-romsub $fef0 = sprite_set_image()
-romsub $fef3 = sprite_set_position()
-romsub $fee4 = memory_fill()
-romsub $fee7 = memory_copy()
-romsub $feea = memory_crc()
-romsub $feed = memory_decompress()
-romsub $fedb = console_init()
-romsub $fede = console_put_char()
-romsub $fee1 = console_get_char()
-romsub $fed8 = console_put_image()
-romsub $fed5 = console_set_paging_message()
-romsub $fed2 = kbdbuf_put()
-romsub $fecf = entropy_get()
+; misc
+romsub $fef0 = sprite_set_image(ubyte number @A, ubyte width @X, ubyte height @Y, ubyte apply_mask @Pc) -> ubyte @Pc  ; also uses pixels=r0, mask=r1, bpp=r2L
+romsub $fef3 = sprite_set_position(ubyte number @A)  ; also uses x=r0 and y=r1
+romsub $fee4 = memory_fill(ubyte value @A)      ; uses address=r0, num_bytes=r1
+romsub $fee7 = memory_copy()                    ; uses source=r0, target=r1, num_bytes=r2
+romsub $feea = memory_crc()                     ; uses address=r0, num_bytes=r1     result->r2
+romsub $feed = memory_decompress()              ; uses input=r0, output=r1     result->r1
+romsub $fedb = console_init()           ; uses x=r0, y=r1, width=r2, height=r3
+romsub $fede = console_put_char(ubyte char @A, ubyte wrapping @Pc)
+romsub $fee1 = console_get_char() -> ubyte @A
+romsub $fed8 = console_put_image()      ; uses ptr=r0, width=r1, height=r2
+romsub $fed5 = console_set_paging_message()     ; uses messageptr=r0
+romsub $fed2 = kbdbuf_put(ubyte key @A)
+romsub $fecf = entropy_get() -> ubyte @A, ubyte @X, ubyte @Y
 romsub $fecc = monitor()
 
 
@@ -179,9 +203,8 @@ asmsub init_system()  {
     %asm {{
         sei
         cld
-        lda  #0
-        sta  $00
-        sta  $01
+        stz  $00
+        stz  $01
         jsr  c64.IOINIT
         jsr  c64.RESTOR
         jsr  c64.CINT
@@ -191,9 +214,6 @@ asmsub init_system()  {
         clc
         clv
         cli
-        lda  #66
-        clc
-        jsr  console_put_char
         rts
     }}
 }

compiler/res/prog8lib/cx16textio.p8

@@ -11,16 +11,61 @@
 
 txt {
 
-asmsub  clear_screen (ubyte char @ A, ubyte color @ Y) clobbers(A)  {
-	; ---- clear the character screen with the given fill character and character color.
-	;      (assumes screen and color matrix are at their default addresses)
+sub  clear_screen() {
+    c64.CHROUT(147)         ; clear screen (spaces)
+}
 
+
+asmsub  fill_screen (ubyte char @ A, ubyte txtcolor @ Y) clobbers(A)  {
+	; ---- fill the character screen with the given fill character and character color.
 	%asm {{
-	    brk     ; TODO
-        }}
+        sta  P8ZP_SCRATCH_W1        ; fillchar
+        sty  P8ZP_SCRATCH_W1+1      ; textcolor
+        phx
+        jsr  c64.SCREEN             ; get dimensions in X/Y
+        dex
+        dey
+        txa
+        asl  a
+        adc  #1
+        sta  P8ZP_SCRATCH_B1
+-       ldx  P8ZP_SCRATCH_B1
+-       stz  cx16.VERA_ADDR_H
+        stx  cx16.VERA_ADDR_L
+        sty  cx16.VERA_ADDR_M
+        lda  cx16.VERA_DATA0
+        and  #$f0
+        ora  P8ZP_SCRATCH_W1+1
+        sta  cx16.VERA_DATA0
+        dex
+        stz  cx16.VERA_ADDR_H
+        stx  cx16.VERA_ADDR_L
+        sty  cx16.VERA_ADDR_M
+        lda  P8ZP_SCRATCH_W1
+        sta  cx16.VERA_DATA0
+        dex
+        cpx  #255
+        bne  -
+        dey
+        bpl  --
+        plx
+        rts
+    }}
 
 }
 
+ubyte[16] color_to_charcode = [$90,$05,$1c,$9f,$9c,$1e,$1f,$9e,$81,$95,$96,$97,$98,$99,$9a,$9b]
+
+sub color (ubyte txtcol) {
+    c64.CHROUT(color_to_charcode[txtcol & 15])
+}
+
+sub color2 (ubyte txtcol, ubyte bgcol) {
+    c64.CHROUT(color_to_charcode[bgcol & 15])
+    c64.CHROUT(1)       ; switch fg and bg colors
+    c64.CHROUT(color_to_charcode[txtcol & 15])
+}
+
 asmsub  print (str text @ AY) clobbers(A,Y)  {
 	; ---- print null terminated string from A/Y
 	; note: the compiler contains an optimization that will replace
@@ -42,7 +87,7 @@ asmsub  print (str text @ AY) clobbers(A,Y)  {
 asmsub  print_ub0  (ubyte value @ A) clobbers(A,Y)  {
 	; ---- print the ubyte in A in decimal form, with left padding 0s (3 positions total)
 	%asm {{
-		stx  P8ZP_SCRATCH_REG_X
+		phx
 		jsr  conv.ubyte2decimal
 		pha
 		tya
@@ -51,7 +96,7 @@ asmsub  print_ub0  (ubyte value @ A) clobbers(A,Y)  {
 		jsr  c64.CHROUT
 		txa
 		jsr  c64.CHROUT
-		ldx  P8ZP_SCRATCH_REG_X
+		plx
 		rts
 	}}
 }
@@ -59,7 +104,7 @@ asmsub  print_ub0  (ubyte value @ A) clobbers(A,Y)  {
 asmsub  print_ub  (ubyte value @ A) clobbers(A,Y)  {
 	; ---- print the ubyte in A in decimal form, without left padding 0s
 	%asm {{
-		stx  P8ZP_SCRATCH_REG_X
+		phx
 		jsr  conv.ubyte2decimal
 _print_byte_digits
 		pha
@@ -76,7 +121,7 @@ _print_byte_digits
         jsr  c64.CHROUT
 _ones   txa
 		jsr  c64.CHROUT
-		ldx  P8ZP_SCRATCH_REG_X
+		plx
 		rts
 	}}
 }
@@ -84,7 +129,7 @@ _ones   txa
 asmsub  print_b  (byte value @ A) clobbers(A,Y)  {
 	; ---- print the byte in A in decimal form, without left padding 0s
 	%asm {{
-		stx  P8ZP_SCRATCH_REG_X
+		phx
 		pha
 		cmp  #0
 		bpl  +
@@ -92,16 +137,14 @@ asmsub  print_b  (byte value @ A) clobbers(A,Y)  {
 		jsr  c64.CHROUT
 +		pla
 		jsr  conv.byte2decimal
-		jsr  print_ub._print_byte_digits
-		ldx  P8ZP_SCRATCH_REG_X
-		rts
+		jmp  print_ub._print_byte_digits
 	}}
 }
 
 asmsub  print_ubhex  (ubyte value @ A, ubyte prefix @ Pc) clobbers(A,Y)  {
 	; ---- print the ubyte in A in hex form (if Carry is set, a radix prefix '$' is printed as well)
 	%asm {{
-		stx  P8ZP_SCRATCH_REG_X
+		phx
 		bcc  +
 		pha
 		lda  #'$'
@@ -111,7 +154,7 @@ asmsub  print_ubhex  (ubyte value @ A, ubyte prefix @ Pc) clobbers(A,Y)  {
 		jsr  c64.CHROUT
 		tya
 		jsr  c64.CHROUT
-		ldx  P8ZP_SCRATCH_REG_X
+		plx
 		rts
 	}}
 }
@@ -119,7 +162,7 @@ asmsub  print_ubhex  (ubyte value @ A, ubyte prefix @ Pc) clobbers(A,Y)  {
 asmsub  print_ubbin  (ubyte value @ A, ubyte prefix @ Pc) clobbers(A,Y)  {
 	; ---- print the ubyte in A in binary form (if Carry is set, a radix prefix '%' is printed as well)
 	%asm {{
-		stx  P8ZP_SCRATCH_REG_X
+		phx
 		sta  P8ZP_SCRATCH_B1
 		bcc  +
 		lda  #'%'
@@ -132,7 +175,7 @@ asmsub  print_ubbin  (ubyte value @ A, ubyte prefix @ Pc) clobbers(A,Y)  {
 +		jsr  c64.CHROUT
 		dey
 		bne  -
-		ldx  P8ZP_SCRATCH_REG_X
+		plx
 		rts
 	}}
 }
@@ -165,7 +208,7 @@ asmsub  print_uwhex  (uword value @ AY, ubyte prefix @ Pc) clobbers(A,Y)  {
 asmsub  print_uw0  (uword value @ AY) clobbers(A,Y)  {
 	; ---- print the uword in A/Y in decimal form, with left padding 0s (5 positions total)
 	%asm {{
-	    stx  P8ZP_SCRATCH_REG_X
+	    phx
 		jsr  conv.uword2decimal
 		ldy  #0
 -		lda  conv.uword2decimal.decTenThousands,y
@@ -173,7 +216,7 @@ asmsub  print_uw0  (uword value @ AY) clobbers(A,Y)  {
 		jsr  c64.CHROUT
 		iny
 		bne  -
-+		ldx  P8ZP_SCRATCH_REG_X
++		plx
 		rts
 	}}
 }
@@ -181,9 +224,9 @@ asmsub  print_uw0  (uword value @ AY) clobbers(A,Y)  {
 asmsub  print_uw  (uword value @ AY) clobbers(A,Y)  {
 	; ---- print the uword in A/Y in decimal form, without left padding 0s
 	%asm {{
-	    stx  P8ZP_SCRATCH_REG_X
+	    phx
 		jsr  conv.uword2decimal
-		ldx  P8ZP_SCRATCH_REG_X
+		plx
 		ldy  #0
 -		lda  conv.uword2decimal.decTenThousands,y
 		beq  _allzero
@@ -225,14 +268,45 @@ asmsub  print_w  (word value @ AY) clobbers(A,Y)  {
 	}}
 }
 
+; TODO implement the "missing" txtio subroutines
+
+sub  setcc  (ubyte column, ubyte row, ubyte char, ubyte charcolor)  {
+	; ---- set char+color at the given position on the screen
+	%asm {{
+	    phx
+		lda  column
+		asl  a
+		tax
+		ldy  row
+		lda  charcolor
+		and  #$0f
+	    sta  P8ZP_SCRATCH_B1
+        stz  cx16.VERA_ADDR_H
+        stx  cx16.VERA_ADDR_L
+        sty  cx16.VERA_ADDR_M
+        lda  char
+        sta  cx16.VERA_DATA0
+        inx
+        stz  cx16.VERA_ADDR_H
+        stx  cx16.VERA_ADDR_L
+        sty  cx16.VERA_ADDR_M
+        lda  cx16.VERA_DATA0
+        and  #$f0
+        ora  P8ZP_SCRATCH_B1
+        sta  cx16.VERA_DATA0
+		plx
+		rts
+    }}
+}
+
 asmsub  plot  (ubyte col @ Y, ubyte row @ A) clobbers(A) {
 	; ---- safe wrapper around PLOT kernel routine, to save the X register.
 	%asm  {{
-		stx  P8ZP_SCRATCH_REG_X
+		phx
 		tax
 		clc
 		jsr  c64.PLOT
-		ldx  P8ZP_SCRATCH_REG_X
+		plx
 		rts
 	}}
 }

compiler/res/prog8lib/math.asm

@@ -59,7 +59,7 @@ multiply_words	.proc
 		sty  P8ZP_SCRATCH_W2+1
 		stx  P8ZP_SCRATCH_REG_X
 
-mult16		lda  #$00
+mult16		lda  #0
 		sta  result+2	; clear upper bits of product
 		sta  result+3
 		ldx  #16			; for all 16 bits...
@@ -86,8 +86,40 @@ result		.byte  0,0,0,0
 		.pend
 
 
+divmod_b_asm	.proc
+	; signed byte division: make everything positive and fix sign afterwards
+		sta  P8ZP_SCRATCH_B1
+		tya
+		eor  P8ZP_SCRATCH_B1
+		php			; save sign
+		lda  P8ZP_SCRATCH_B1
+		bpl  +
+		eor  #$ff
+		sec
+		adc  #0			; make it positive
++		pha
+		tya
+		bpl  +
+		eor  #$ff
+		sec
+		adc  #0			; make it positive
+		tay
++		pla
+		jsr  divmod_ub_asm
+		sta  _remainder
+		plp
+		bpl  +
+		tya
+		eor  #$ff
+		sec
+		adc  #0			; negate result
+		tay
++		rts
+_remainder	.byte  0
+		.pend
+
+
 divmod_ub_asm	.proc
-	; TODO divmod_ub_asm doesn't work correctly.  (remainder = ok, quotient = FAULTY)
 	; -- divide A by Y, result quotient in Y, remainder in A   (unsigned)
 	;    division by zero will result in quotient = 255 and remainder = original number
 		sty  P8ZP_SCRATCH_REG
@@ -109,6 +141,49 @@ divmod_ub_asm	.proc
 		rts
 		.pend
 
+divmod_w_asm	.proc
+	; signed word division: make everything positive and fix sign afterwards
+		sta  P8ZP_SCRATCH_W2
+		sty  P8ZP_SCRATCH_W2+1
+		lda  P8ZP_SCRATCH_W1+1
+		eor  P8ZP_SCRATCH_W2+1
+		php			; save sign
+		lda  P8ZP_SCRATCH_W1+1
+		bpl  +
+		lda  #0
+		sec
+		sbc  P8ZP_SCRATCH_W1
+		sta  P8ZP_SCRATCH_W1
+		lda  #0
+		sbc  P8ZP_SCRATCH_W1+1
+		sta  P8ZP_SCRATCH_W1+1
++		lda  P8ZP_SCRATCH_W2+1
+		bpl  +
+		lda  #0
+		sec
+		sbc  P8ZP_SCRATCH_W2
+		sta  P8ZP_SCRATCH_W2
+		lda  #0
+		sbc  P8ZP_SCRATCH_W2+1
+		sta  P8ZP_SCRATCH_W2+1
++		tay
+		lda  P8ZP_SCRATCH_W2
+		jsr  divmod_uw_asm
+		plp			; restore sign
+		bpl  +
+		sta  P8ZP_SCRATCH_W2
+		sty  P8ZP_SCRATCH_W2+1
+		lda  #0
+		sec
+		sbc  P8ZP_SCRATCH_W2
+		pha
+		lda  #0
+		sbc  P8ZP_SCRATCH_W2+1
+		tay
+		pla
++		rts
+		.pend
+
 divmod_uw_asm	.proc
 	; -- divide two unsigned words (16 bit each) into 16 bit results
 	;    input:  P8ZP_SCRATCH_W1 in ZP: 16 bit number, A/Y: 16 bit divisor

compiler/res/prog8lib/prog8lib.asm

@@ -960,7 +960,7 @@ _result_maxuw	.word  0
 		.pend
 
 func_max_w	.proc
-		lda  #$00
+		lda  #0
 		sta  _result_maxw
 		lda  #$80
 		sta  _result_maxw+1

compiler/res/version.txt

@@ -1 +1 @@
-4.0
+4.1

compiler/src/prog8/CompilerMain.kt

@@ -36,7 +36,7 @@ fun pathFrom(stringPath: String, vararg rest: String): Path  = FileSystems.getDe
 
 private fun compileMain(args: Array<String>) {
     val cli = CommandLineInterface("prog8compiler")
-    val startEmulator by cli.flagArgument("-emu", "auto-start the Vice C-64 emulator after successful compilation")
+    val startEmulator by cli.flagArgument("-emu", "auto-start emulator after successful compilation")
     val outputDir by cli.flagValueArgument("-out", "directory", "directory for output files instead of current directory", ".")
     val dontWriteAssembly by cli.flagArgument("-noasm", "don't create assembly code")
     val dontOptimize by cli.flagArgument("-noopt", "don't perform any optimizations")

compiler/src/prog8/ast/expressions/AstExpressions.kt

@@ -642,7 +642,14 @@ class RangeExpr(var from: Expression,
             fromDt istype DataType.STR && toDt istype DataType.STR -> InferredTypes.knownFor(DataType.STR)
             fromDt istype DataType.WORD || toDt istype DataType.WORD -> InferredTypes.knownFor(DataType.ARRAY_W)
             fromDt istype DataType.BYTE || toDt istype DataType.BYTE -> InferredTypes.knownFor(DataType.ARRAY_B)
-            else -> InferredTypes.knownFor(DataType.ARRAY_UB)
+            else -> {
+                val fdt = fromDt.typeOrElse(DataType.STRUCT)
+                val tdt = toDt.typeOrElse(DataType.STRUCT)
+                if(fdt largerThan tdt)
+                    InferredTypes.knownFor(ElementArrayTypes.getValue(fdt))
+                else
+                    InferredTypes.knownFor(ElementArrayTypes.getValue(tdt))
+            }
         }
     }
     override fun toString(): String {

compiler/src/prog8/ast/processing/AstChecker.kt

@@ -118,6 +118,11 @@ internal class AstChecker(private val program: Program,
             if(loopvar==null || loopvar.type== VarDeclType.CONST) {
                 errors.err("for loop requires a variable to loop with", forLoop.position)
             } else {
+
+                fun checkLoopRangeValues() {
+
+                }
+
                 when (loopvar.datatype) {
                     DataType.UBYTE -> {
                         if(iterableDt!= DataType.UBYTE && iterableDt!= DataType.ARRAY_UB && iterableDt != DataType.STR)
@@ -142,6 +147,22 @@ internal class AstChecker(private val program: Program,
                     }
                     else -> errors.err("loop variable must be numeric type", forLoop.position)
                 }
+                if(errors.isEmpty()) {
+                    // check loop range values
+                    val range = forLoop.iterable as? RangeExpr
+                    if(range!=null) {
+                        val from = range.from as? NumericLiteralValue
+                        val to = range.to as? NumericLiteralValue
+                        if(from != null)
+                            checkValueTypeAndRange(loopvar.datatype, from)
+                        else if(!range.from.inferType(program).istype(loopvar.datatype))
+                            errors.err("range start value is incompatible with loop variable type", range.position)
+                        if(to != null)
+                            checkValueTypeAndRange(loopvar.datatype, to)
+                        else if(!range.to.inferType(program).istype(loopvar.datatype))
+                            errors.err("range end value is incompatible with loop variable type", range.position)
+                    }
+                }
             }
         }
 
@@ -915,9 +936,6 @@ internal class AstChecker(private val program: Program,
                     val argIDt = arg.first.value.inferType(program)
                     if (!argIDt.isKnown)
                         return
-                    if (target.asmParameterRegisters[arg.first.index].registerOrPair in setOf(RegisterOrPair.AX, RegisterOrPair.XY, RegisterOrPair.X)
-                            && arg.first.value !is NumericLiteralValue && arg.first.value !is IdentifierReference)
-                        errors.warn("calling a subroutine that expects X as a parameter is problematic. If you see a compiler error/crash about this later, try to change this call", position)
                 }
             }
         }

compiler/src/prog8/ast/statements/AstStatements.kt

@@ -657,6 +657,7 @@ class Subroutine(override val name: String,
     }
 
     fun regXasResult() = asmReturnvaluesRegisters.any { it.registerOrPair in setOf(RegisterOrPair.X, RegisterOrPair.AX, RegisterOrPair.XY) }
+    fun regXasParam() = asmParameterRegisters.any { it.registerOrPair in setOf(RegisterOrPair.X, RegisterOrPair.AX, RegisterOrPair.XY) }
 
     fun amountOfRtsInAsm(): Int = statements
             .asSequence()

compiler/src/prog8/compiler/target/IMachineDefinition.kt

@@ -11,6 +11,11 @@ internal interface IMachineFloat {
     fun makeFloatFillAsm(): String
 }
 
+internal enum class CpuType {
+    CPU6502,
+    CPU65c02
+}
+
 internal interface IMachineDefinition {
     val FLOAT_MAX_NEGATIVE: Double
     val FLOAT_MAX_POSITIVE: Double
@@ -24,7 +29,7 @@ internal interface IMachineDefinition {
     val opcodeNames: Set<String>
     var zeropage: Zeropage
     val initSystemProcname: String
-    val cpu: String
+    val cpu: CpuType
 
     fun initializeZeropage(compilerOptions: CompilationOptions)
     fun getFloat(num: Number): IMachineFloat

compiler/src/prog8/compiler/target/c64/C64MachineDefinition.kt

@@ -2,6 +2,7 @@ package prog8.compiler.target.c64
 
 import prog8.ast.Program
 import prog8.compiler.*
+import prog8.compiler.target.CpuType
 import prog8.compiler.target.IMachineDefinition
 import prog8.compiler.target.IMachineFloat
 import prog8.parser.ModuleImporter
@@ -12,7 +13,7 @@ import kotlin.math.pow
 
 internal object C64MachineDefinition: IMachineDefinition {
 
-    override val cpu = "6502"
+    override val cpu = CpuType.CPU6502
 
     // 5-byte cbm MFLPT format limitations:
     override val FLOAT_MAX_POSITIVE = 1.7014118345e+38         // bytes: 255,127,255,255,255
@@ -37,7 +38,7 @@ internal object C64MachineDefinition: IMachineDefinition {
         val mflpt5 = Mflpt5.fromNumber(number)
         val floatbytes = shortArrayOf(mflpt5.b0, mflpt5.b1, mflpt5.b2, mflpt5.b3, mflpt5.b4)
         when {
-            floatbytes.contentEquals(shortArrayOf(0x00, 0x00, 0x00, 0x00, 0x00)) -> return "c64flt.FL_ZERO"
+            floatbytes.contentEquals(shortArrayOf(0x00, 0x00, 0x00, 0x00, 0x00)) -> return "c64flt.ZERO"        // not a ROM const
             floatbytes.contentEquals(shortArrayOf(0x82, 0x49, 0x0f, 0xda, 0xa1)) -> return "c64flt.FL_PIVAL"
             floatbytes.contentEquals(shortArrayOf(0x90, 0x80, 0x00, 0x00, 0x00)) -> return "c64flt.FL_N32768"
             floatbytes.contentEquals(shortArrayOf(0x81, 0x00, 0x00, 0x00, 0x00)) -> return "c64flt.FL_FONE"

compiler/src/prog8/compiler/target/c64/codegen/AsmGen.kt

@@ -9,6 +9,7 @@ import prog8.ast.expressions.*
 import prog8.ast.statements.*
 import prog8.compiler.*
 import prog8.compiler.target.CompilationTarget
+import prog8.compiler.target.CpuType
 import prog8.compiler.target.IAssemblyGenerator
 import prog8.compiler.target.IAssemblyProgram
 import prog8.compiler.target.c64.AssemblyProgram
@@ -78,7 +79,11 @@ internal class AsmGen(private val program: Program,
 
     private fun header() {
         val ourName = this.javaClass.name
-        val cpu = CompilationTarget.machine.cpu
+        val cpu = when(CompilationTarget.machine.cpu) {
+            CpuType.CPU6502 -> "6502"
+            CpuType.CPU65c02 -> "65c02"
+            else -> "unsupported"
+        }
 
         out("; $cpu assembly code for '${program.name}'")
         out("; generated by $ourName on ${LocalDateTime.now().withNano(0)}")
@@ -546,7 +551,12 @@ internal class AsmGen(private val program: Program,
     internal fun saveRegister(register: CpuRegister) {
         when(register) {
             CpuRegister.A -> out("  pha")
-            CpuRegister.X -> out("  txa | pha")
+            CpuRegister.X -> {
+                if(CompilationTarget.machine.cpu == CpuType.CPU65c02)
+                    out("  phx")
+                else
+                    out("  stx  P8ZP_SCRATCH_REG_X")
+            }
             CpuRegister.Y -> out("  tya | pha")
         }
     }
@@ -554,7 +564,12 @@ internal class AsmGen(private val program: Program,
     internal fun restoreRegister(register: CpuRegister) {
         when(register) {
             CpuRegister.A -> out("  pla")
-            CpuRegister.X -> out("  pla | tax")
+            CpuRegister.X -> {
+                if(CompilationTarget.machine.cpu == CpuType.CPU65c02)
+                    out("  plx")
+                else
+                    out("  ldx  P8ZP_SCRATCH_REG_X")
+            }
             CpuRegister.Y -> out("  pla | tay")
         }
     }

compiler/src/prog8/compiler/target/c64/codegen/AsmOptimizer.kt

@@ -151,7 +151,8 @@ private fun optimizeSameAssignments(linesByFourteen: List<List<IndexedValue<Stri
         }
 
         if(first.startsWith("lda") && second.startsWith("ldy") && third.startsWith("sta") && fourth.startsWith("sty") &&
-                fifth.startsWith("lda") && sixth.startsWith("ldy") && seventh.startsWith("jsr  c64flt.copy_float")) {
+                fifth.startsWith("lda") && sixth.startsWith("ldy") &&
+                (seventh.startsWith("jsr  c64flt.copy_float") || seventh.startsWith("jsr  cx16flt.copy_float"))) {
 
             val nineth = pair[8].value.trimStart()
             val tenth = pair[9].value.trimStart()
@@ -161,7 +162,8 @@ private fun optimizeSameAssignments(linesByFourteen: List<List<IndexedValue<Stri
             val fourteenth = pair[13].value.trimStart()
 
             if(eighth.startsWith("lda") && nineth.startsWith("ldy") && tenth.startsWith("sta") && eleventh.startsWith("sty") &&
-                    twelveth.startsWith("lda") && thirteenth.startsWith("ldy") && fourteenth.startsWith("jsr  c64flt.copy_float")) {
+                    twelveth.startsWith("lda") && thirteenth.startsWith("ldy") &&
+                    (fourteenth.startsWith("jsr  c64flt.copy_float") || fourteenth.startsWith("jsr  cx16flt.copy_float"))) {
 
                 if(first.substring(4) == eighth.substring(4) && second.substring(4)==nineth.substring(4)) {
                     // identical float init

compiler/src/prog8/compiler/target/c64/codegen/ExpressionsAsmGen.kt

@@ -4,6 +4,8 @@ import prog8.ast.Program
 import prog8.ast.base.*
 import prog8.ast.expressions.*
 import prog8.compiler.AssemblyError
+import prog8.compiler.target.CompilationTarget
+import prog8.compiler.target.CpuType
 import prog8.compiler.toHex
 import prog8.functions.BuiltinFunctions
 import kotlin.math.absoluteValue
@@ -45,15 +47,28 @@ internal class ExpressionsAsmGen(private val program: Program, private val asmge
                             RegisterOrPair.AY -> asmgen.out("  sta  P8ESTACK_LO,x |  tya |  sta  P8ESTACK_HI,x |  dex")
                             RegisterOrPair.X -> {
                                 // return value in X register has been discarded, just push a zero
-                                asmgen.out("  lda  #0 |  sta  P8ESTACK_LO,x |  dex")
+                                if(CompilationTarget.machine.cpu==CpuType.CPU65c02)
+                                    asmgen.out("  stz  P8ESTACK_LO,x")
+                                else
+                                    asmgen.out("  lda  #0 |  sta  P8ESTACK_LO,x")
+                                asmgen.out("  dex")
                             }
                             RegisterOrPair.AX -> {
                                 // return value in X register has been discarded, just push a zero in this place
-                                asmgen.out("  sta  P8ESTACK_LO,x |  lda  #0 |  sta  P8ESTACK_HI,x |  dex")
+                                asmgen.out("  sta  P8ESTACK_LO,x")
+                                if(CompilationTarget.machine.cpu==CpuType.CPU65c02)
+                                    asmgen.out("  stz  P8ESTACK_HI,x")
+                                else
+                                    asmgen.out("  lda  #0 |  sta  P8ESTACK_HI,x")
+                                asmgen.out("  dex")
                             }
                             RegisterOrPair.XY -> {
                                 // return value in X register has been discarded, just push a zero in this place
-                                asmgen.out("  lda  #0 |  sta  P8ESTACK_LO,x |  tya |  sta  P8ESTACK_HI,x |  dex")
+                                if(CompilationTarget.machine.cpu==CpuType.CPU65c02)
+                                    asmgen.out("  stz  P8ESTACK_LO,x")
+                                else
+                                    asmgen.out("  lda  #0 |  sta  P8ESTACK_LO,x")
+                                asmgen.out("  tya |  sta  P8ESTACK_HI,x |  dex")
                             }
                         }
                     }
@@ -69,7 +84,12 @@ internal class ExpressionsAsmGen(private val program: Program, private val asmge
             DataType.UBYTE -> {
                 when(expr.type) {
                     DataType.UBYTE, DataType.BYTE -> {}
-                    DataType.UWORD, DataType.WORD -> asmgen.out("  lda  #0  |  sta  P8ESTACK_HI+1,x")
+                    DataType.UWORD, DataType.WORD -> {
+                        if(CompilationTarget.machine.cpu==CpuType.CPU65c02)
+                            asmgen.out("  stz  P8ESTACK_HI+1,x")
+                        else
+                            asmgen.out("  lda  #0  |  sta  P8ESTACK_HI+1,x")
+                    }
                     DataType.FLOAT -> asmgen.out(" jsr  c64flt.stack_ub2float")
                     in PassByReferenceDatatypes -> throw AssemblyError("cannot cast to a pass-by-reference datatype")
                     else -> throw AssemblyError("weird type")
@@ -83,9 +103,14 @@ internal class ExpressionsAsmGen(private val program: Program, private val asmge
                         asmgen.out(""" 
                             lda  P8ESTACK_LO+1,x
                             ora  #$7f
-                            bmi  +
-                            lda  #0
-+                           sta  P8ESTACK_HI+1,x""")
+                            bmi  +""")
+                        if(CompilationTarget.machine.cpu==CpuType.CPU65c02)
+                            asmgen.out("""
++                               stz  P8ESTACK_HI+1,x""")
+                        else
+                            asmgen.out("""
+                                lda  #0
++                               sta  P8ESTACK_HI+1,x""")
                     }
                     DataType.FLOAT -> asmgen.out(" jsr  c64flt.stack_b2float")
                     in PassByReferenceDatatypes -> throw AssemblyError("cannot cast to a pass-by-reference datatype")

compiler/src/prog8/compiler/target/c64/codegen/ForLoopsAsmGen.kt

@@ -46,7 +46,7 @@ internal class ForLoopsAsmGen(private val program: Program, private val asmgen:
             DataType.ARRAY_B, DataType.ARRAY_UB -> {
                 if (stepsize==1 || stepsize==-1) {
 
-                    // bytes, step 1 or -1
+                    // bytes array, step 1 or -1
 
                     val incdec = if(stepsize==1) "inc" else "dec"
                     // loop over byte range via loopvar
@@ -258,13 +258,6 @@ $endLabel       inx""")
         asmgen.loopEndLabels.pop()
     }
 
-    private fun assignLoopvar(stmt: ForLoop, range: RangeExpr) {
-        val target = AsmAssignTarget(TargetStorageKind.VARIABLE, program, asmgen, stmt.loopVarDt(program).typeOrElse(DataType.STRUCT), variable=stmt.loopVar)
-        val src = AsmAssignSource.fromAstSource(range.from, program).adjustDataTypeToTarget(target)
-        val assign = AsmAssignment(src, target, false, range.position)
-        asmgen.translateNormalAssignment(assign)
-    }
-
     private fun translateForOverIterableVar(stmt: ForLoop, iterableDt: DataType, ident: IdentifierReference) {
         val loopLabel = asmgen.makeLabel("for_loop")
         val endLabel = asmgen.makeLabel("for_end")
@@ -615,4 +608,11 @@ $loopLabel""")
 $endLabel""")
         asmgen.loopEndLabels.pop()
     }
+
+    private fun assignLoopvar(stmt: ForLoop, range: RangeExpr) {
+        val target = AsmAssignTarget(TargetStorageKind.VARIABLE, program, asmgen, stmt.loopVarDt(program).typeOrElse(DataType.STRUCT), variable=stmt.loopVar)
+        val src = AsmAssignSource.fromAstSource(range.from, program).adjustDataTypeToTarget(target)
+        val assign = AsmAssignment(src, target, false, range.position)
+        asmgen.translateNormalAssignment(assign)
+    }
 }

compiler/src/prog8/compiler/target/c64/codegen/FunctionCallAsmGen.kt

@@ -16,9 +16,9 @@ internal class FunctionCallAsmGen(private val program: Program, private val asmg
         // output the code to setup the parameters and perform the actual call
         // does NOT output the code to deal with the result values!
         val sub = stmt.target.targetSubroutine(program.namespace) ?: throw AssemblyError("undefined subroutine ${stmt.target}")
-        val saveX = CpuRegister.X in sub.asmClobbers || sub.regXasResult()
+        val saveX = CpuRegister.X in sub.asmClobbers || sub.regXasResult() || sub.regXasParam()
         if(saveX)
-            asmgen.out("  stx  P8ZP_SCRATCH_REG_X")        // we only save X for now (required! is the eval stack pointer), screw A and Y...
+            asmgen.saveRegister(CpuRegister.X)  // we only save X for now (required! is the eval stack pointer), screw A and Y...
 
         val subName = asmgen.asmSymbolName(stmt.target)
         if(stmt.args.isNotEmpty()) {
@@ -57,7 +57,7 @@ internal class FunctionCallAsmGen(private val program: Program, private val asmg
         asmgen.out("  jsr  $subName")
 
         if(saveX)
-            asmgen.out("  ldx  P8ZP_SCRATCH_REG_X")        // restore X again
+            asmgen.restoreRegister(CpuRegister.X)
     }
 
     private fun registerArgsViaStackEvaluation(stmt: IFunctionCall, sub: Subroutine) {

compiler/src/prog8/compiler/target/c64/codegen/PostIncrDecrAsmGen.kt

@@ -97,7 +97,8 @@ internal class PostIncrDecrAsmGen(private val program: Program, private val asmg
                     }
                     else -> {
                         asmgen.loadScaledArrayIndexIntoRegister(targetArrayIdx, elementDt, CpuRegister.A)
-                        asmgen.out("  stx  P8ZP_SCRATCH_REG_X |  tax")
+                        asmgen.saveRegister(CpuRegister.X)
+                        asmgen.out("  tax")
                         when(elementDt) {
                             in ByteDatatypes -> {
                                 asmgen.out(if(incr) "  inc  $asmArrayvarname,x" else "  dec  $asmArrayvarname,x")
@@ -124,7 +125,7 @@ internal class PostIncrDecrAsmGen(private val program: Program, private val asmg
                             }
                             else -> throw AssemblyError("weird array elt dt")
                         }
-                        asmgen.out("  ldx  P8ZP_SCRATCH_REG_X")
+                        asmgen.restoreRegister(CpuRegister.X)
                     }
                 }
             }

compiler/src/prog8/compiler/target/c64/codegen/assignment/AssignmentAsmGen.kt

@@ -5,6 +5,8 @@ import prog8.ast.base.*
 import prog8.ast.expressions.*
 import prog8.ast.statements.*
 import prog8.compiler.AssemblyError
+import prog8.compiler.target.CompilationTarget
+import prog8.compiler.target.CpuType
 import prog8.compiler.target.c64.codegen.AsmGen
 import prog8.compiler.toHex
 
@@ -706,7 +708,16 @@ internal class AssignmentAsmGen(private val program: Program, private val asmgen
             // optimized case for float zero
             when(target.kind) {
                 TargetStorageKind.VARIABLE -> {
-                    asmgen.out("""
+                    if(CompilationTarget.machine.cpu == CpuType.CPU65c02)
+                        asmgen.out("""
+                            stz  ${target.asmVarname}
+                            stz  ${target.asmVarname}+1
+                            stz  ${target.asmVarname}+2
+                            stz  ${target.asmVarname}+3
+                            stz  ${target.asmVarname}+4
+                        """)
+                    else
+                        asmgen.out("""
                             lda  #0
                             sta  ${target.asmVarname}
                             sta  ${target.asmVarname}+1

compiler/src/prog8/compiler/target/c64/codegen/assignment/AugmentableAssignmentAsmGen.kt

@@ -5,6 +5,7 @@ import prog8.ast.base.*
 import prog8.ast.expressions.*
 import prog8.compiler.AssemblyError
 import prog8.compiler.target.CompilationTarget
+import prog8.compiler.target.CpuType
 import prog8.compiler.target.c64.codegen.AsmGen
 import prog8.compiler.toHex
 
@@ -125,7 +126,7 @@ internal class AugmentableAssignmentAsmGen(private val program: Program,
                             ident != null -> inplaceModification_word_variable_to_variable(target.asmVarname, target.datatype, operator, ident)
                             // TODO more specialized code for types such as memory read etc.
 //                            value is DirectMemoryRead -> {
-//                                println("warning: slow stack evaluation used (8):  $name $operator= ${value::class.simpleName} at ${value.position}") // TODO
+//                                println("warning: slow stack evaluation used (8):  $name $operator= ${value::class.simpleName} at ${value.position}")
 //                                // assignmentAsmGen.translateOtherAssignment(origAssign)
 //                                asmgen.translateExpression(value.addressExpression)
 //                                asmgen.out("""
@@ -134,7 +135,6 @@ internal class AugmentableAssignmentAsmGen(private val program: Program,
 //                                    inx
 //                                    """)
 //                                inplaceModification_word_value_to_variable(name, operator, )
-//                                // TODO
 //                            }
                             value is TypecastExpression -> {
                                 if (tryRemoveRedundantCast(value, target, operator)) return
@@ -250,16 +250,16 @@ internal class AugmentableAssignmentAsmGen(private val program: Program,
                 else
                     asmgen.out("  sta  (P8ZP_SCRATCH_W1),y")
             }
-            "*" -> TODO("mul")// asmgen.out("  jsr  prog8_lib.mul_byte")  //  the optimized routines should have been checked earlier
-            "/" -> TODO("div")// asmgen.out(if(types==DataType.UBYTE) "  jsr  prog8_lib.idiv_ub" else "  jsr  prog8_lib.idiv_b")
+            "*" -> TODO("mul mem byte")// asmgen.out("  jsr  prog8_lib.mul_byte")  //  the optimized routines should have been checked earlier
+            "/" -> TODO("div mem byte")// asmgen.out(if(types==DataType.UBYTE) "  jsr  prog8_lib.idiv_ub" else "  jsr  prog8_lib.idiv_b")
             "%" -> {
-                TODO("byte remainder")
+                TODO("mem byte remainder")
 //                if(types==DataType.BYTE)
 //                    throw AssemblyError("remainder of signed integers is not properly defined/implemented, use unsigned instead")
 //                asmgen.out("  jsr prog8_lib.remainder_ub")
             }
-            "<<" -> TODO("ubyte asl")
-            ">>" -> TODO("ubyte lsr")
+            "<<" -> TODO("mem ubyte asl")
+            ">>" -> TODO("mem ubyte lsr")
             "&" -> {
                 val (ptrOnZp, sourceName) = asmgen.loadByteFromPointerIntoA(pointervar)
                 asmgen.out(" and  P8ESTACK_LO+1,x")
@@ -309,16 +309,16 @@ internal class AugmentableAssignmentAsmGen(private val program: Program,
                 else
                     asmgen.out("  sta  (P8ZP_SCRATCH_W1),y")
             }
-            "*" -> TODO("mul")// asmgen.out("  jsr  prog8_lib.mul_byte")  //  the optimized routines should have been checked earlier
-            "/" -> TODO("div")// asmgen.out(if(types==DataType.UBYTE) "  jsr  prog8_lib.idiv_ub" else "  jsr  prog8_lib.idiv_b")
+            "*" -> TODO("mem mul")// asmgen.out("  jsr  prog8_lib.mul_byte")  //  the optimized routines should have been checked earlier
+            "/" -> TODO("mem div")// asmgen.out(if(types==DataType.UBYTE) "  jsr  prog8_lib.idiv_ub" else "  jsr  prog8_lib.idiv_b")
             "%" -> {
-                TODO("byte remainder")
+                TODO("mem byte remainder")
 //                if(types==DataType.BYTE)
 //                    throw AssemblyError("remainder of signed integers is not properly defined/implemented, use unsigned instead")
 //                asmgen.out("  jsr prog8_lib.remainder_ub")
             }
-            "<<" -> TODO("ubyte asl")
-            ">>" -> TODO("ubyte lsr")
+            "<<" -> TODO("mem ubyte asl")
+            ">>" -> TODO("mem ubyte lsr")
             "&" -> {
                 val (ptrOnZp, sourceName) = asmgen.loadByteFromPointerIntoA(pointervar)
                 asmgen.out(" and  $otherName")
@@ -367,19 +367,19 @@ internal class AugmentableAssignmentAsmGen(private val program: Program,
                     asmgen.out("  sta  (P8ZP_SCRATCH_W1),y")
             }
             "*" -> {
-                TODO("mul byte litval")
+                TODO("mem mul byte litval")
                 // asmgen.out("  jsr  prog8_lib.mul_byte")  //  the optimized routines should have been checked earlier
             }
             "/" -> {
                 if(value==0)
                     throw AssemblyError("division by zero")
-                TODO("div byte litval")
+                TODO("mem div byte litval")
                 // asmgen.out(if(types==DataType.UBYTE) "  jsr  prog8_lib.idiv_ub" else "  jsr  prog8_lib.idiv_b")
             }
             "%" -> {
                 if(value==0)
                     throw AssemblyError("division by zero")
-                TODO("byte remainder litval")
+                TODO("mem byte remainder litval")
 //                if(types==DataType.BYTE)
 //                    throw AssemblyError("remainder of signed integers is not properly defined/implemented, use unsigned instead")
 //                asmgen.out("  jsr prog8_lib.remainder_ub")
@@ -442,14 +442,14 @@ internal class AugmentableAssignmentAsmGen(private val program: Program,
             "+" -> asmgen.out(" lda  $name |  clc |  adc  P8ESTACK_LO+1,x |  sta  $name")
             "-" -> asmgen.out(" lda  $name |  sec |  sbc  P8ESTACK_LO+1,x |  sta  $name")
             "*" -> {
-                TODO("mul byte expr")
+                TODO("var mul byte expr")
                 // asmgen.out("  jsr  prog8_lib.mul_byte")  //  the optimized routines should have been checked earlier
             }
             "/" -> {
-                TODO("div byte expr")// asmgen.out(if(types==DataType.UBYTE) "  jsr  prog8_lib.idiv_ub" else "  jsr  prog8_lib.idiv_b")
+                TODO("var div byte expr")// asmgen.out(if(types==DataType.UBYTE) "  jsr  prog8_lib.idiv_ub" else "  jsr  prog8_lib.idiv_b")
             }
             "%" -> {
-                TODO("byte remainder expr")
+                TODO("var byte remainder expr")
 //                if(types==DataType.BYTE)
 //                    throw AssemblyError("remainder of signed integers is not properly defined/implemented, use unsigned instead")
 //                asmgen.out("  jsr prog8_lib.remainder_ub")
@@ -506,7 +506,7 @@ internal class AugmentableAssignmentAsmGen(private val program: Program,
             "*" -> asmgen.out(" lda  $name |  ldy  $otherName  |  jsr  math.multiply_bytes |  sta  $name")
             "/" -> {
                 if(dt==DataType.BYTE) {
-                    TODO("signed byte divide see prog8lib.idiv_b")
+                    asmgen.out(" lda  $name |  ldy  $otherName  |  jsr  math.divmod_b_asm |  sty  $name")
                 }
                 else {
                     asmgen.out(" lda  $name |  ldy  $otherName  |  jsr  math.divmod_ub_asm |  sty  $name")
@@ -573,7 +573,7 @@ internal class AugmentableAssignmentAsmGen(private val program: Program,
             "-" -> asmgen.out(" lda  $name |  sec |  sbc  #$value |  sta  $name")
             "*" -> {
                 // TODO what about the optimized mul_5 etc routines?
-                TODO("byte mul litval")
+                TODO("var byte mul litval")
                 // asmgen.out("  jsr  prog8_lib.mul_byte")  //  the optimized routines should have been checked earlier
             }
             "/" -> {
@@ -585,7 +585,7 @@ internal class AugmentableAssignmentAsmGen(private val program: Program,
                         sty  $name                                              
                     """)
                 } else {
-                    TODO("BYTE div  litval")
+                    TODO("var BYTE div  litval")
                 }
             }
             "%" -> {
@@ -687,7 +687,17 @@ internal class AugmentableAssignmentAsmGen(private val program: Program,
                 if(value==0)
                     throw AssemblyError("division by zero")
                 if(dt==DataType.WORD) {
-                    TODO("signed word divide see prog8lib.idiv_w")
+                    asmgen.out("""
+                        lda  $name
+                        ldy  $name+1
+                        sta  P8ZP_SCRATCH_W1
+                        sty  P8ZP_SCRATCH_W1+1
+                        lda  #<$value
+                        ldy  #>$value
+                        jsr  math.divmod_w_asm
+                        sta  $name
+                        sty  $name+1
+                    """)
                 }
                 else {
                     asmgen.out("""
@@ -736,9 +746,26 @@ internal class AugmentableAssignmentAsmGen(private val program: Program,
             }
             "&" -> {
                 when {
-                    value == 0 -> asmgen.out(" lda  #0 |  sta  $name |  sta  $name+1")
-                    value and 255 == 0 -> asmgen.out(" lda  #0 |  sta  $name |  lda  $name+1 |  and  #>$value |  sta  $name+1")
-                    value < 0x0100 -> asmgen.out(" lda  $name |  and  #$value |  sta  $name |  lda  #0 |  sta  $name+1")
+                    value == 0 -> {
+                        if(CompilationTarget.machine.cpu == CpuType.CPU65c02)
+                            asmgen.out(" stz  $name |  stz  $name+1")
+                        else
+                            asmgen.out(" lda  #0 |  sta  $name |  sta  $name+1")
+                    }
+                    value and 255 == 0 -> {
+                        if(CompilationTarget.machine.cpu == CpuType.CPU65c02)
+                            asmgen.out(" stz  $name")
+                        else
+                            asmgen.out(" lda  #0 |  sta  $name")
+                        asmgen.out(" lda  $name+1 |  and  #>$value |  sta  $name+1")
+                    }
+                    value < 0x0100 -> {
+                        asmgen.out(" lda  $name |  and  #$value |  sta  $name")
+                        if(CompilationTarget.machine.cpu == CpuType.CPU65c02)
+                            asmgen.out("  stz  $name+1")
+                        else
+                            asmgen.out("  lda  #0 |  sta  $name+1")
+                    }
                     else -> asmgen.out(" lda  $name |  and  #<$value |  sta  $name |  lda  $name+1 |  and  #>$value |  sta  $name+1")
                 }
             }
@@ -845,7 +872,17 @@ internal class AugmentableAssignmentAsmGen(private val program: Program,
                     }
                     "/" -> {
                         if(dt==DataType.WORD) {
-                            TODO("signed word divide see prog8lib.idiv_w")
+                            asmgen.out("""
+                                lda  $name
+                                ldy  $name+1
+                                sta  P8ZP_SCRATCH_W1
+                                sty  P8ZP_SCRATCH_W1+1
+                                lda  $otherName
+                                ldy  $otherName+1
+                                jsr  math.divmod_w_asm
+                                sta  $name
+                                sty  $name+1
+                            """)
                         }
                         else {
                             asmgen.out("""
@@ -989,7 +1026,17 @@ internal class AugmentableAssignmentAsmGen(private val program: Program,
                     }
                     "/" -> {
                         if (dt == DataType.WORD) {
-                            TODO("signed word divide see prog8lib.idiv_w")
+                            asmgen.out("""
+                                lda  $name
+                                ldy  $name+1
+                                sta  P8ZP_SCRATCH_W1
+                                sty  P8ZP_SCRATCH_W1+1
+                                lda  P8ESTACK_LO+1,x
+                                ldy  P8ESTACK_HI+1,x
+                                jsr  math.divmod_w_asm
+                                sta  $name
+                                sty  $name+1
+                            """)
                         } else {
                             asmgen.out("""
                                 lda  $name
@@ -1041,11 +1088,11 @@ internal class AugmentableAssignmentAsmGen(private val program: Program,
         // because the value is evaluated onto the eval stack (=slow).
         println("warning: slow stack evaluation used (2):  $name $operator= ${value::class.simpleName} at ${value.position}") // TODO
         asmgen.translateExpression(value)
+        asmgen.saveRegister(CpuRegister.X)
         when (operator) {
             "**" -> {
                 asmgen.out("""
                     jsr  c64flt.pop_float_fac1
-                    stx  P8ZP_SCRATCH_REG_X
                     lda  #<$name
                     ldy  #>$name
                     jsr  c64flt.CONUPK
@@ -1055,7 +1102,6 @@ internal class AugmentableAssignmentAsmGen(private val program: Program,
             "+" -> {
                 asmgen.out("""
                     jsr  c64flt.pop_float_fac1
-                    stx  P8ZP_SCRATCH_REG_X
                     lda  #<$name
                     ldy  #>$name
                     jsr  c64flt.FADD
@@ -1064,7 +1110,6 @@ internal class AugmentableAssignmentAsmGen(private val program: Program,
             "-" -> {
                 asmgen.out("""
                     jsr  c64flt.pop_float_fac1
-                    stx  P8ZP_SCRATCH_REG_X
                     lda  #<$name
                     ldy  #>$name
                     jsr  c64flt.FSUB
@@ -1073,7 +1118,6 @@ internal class AugmentableAssignmentAsmGen(private val program: Program,
             "*" -> {
                 asmgen.out("""
                     jsr  c64flt.pop_float_fac1
-                    stx  P8ZP_SCRATCH_REG_X
                     lda  #<$name
                     ldy  #>$name
                     jsr  c64flt.FMULT
@@ -1082,7 +1126,6 @@ internal class AugmentableAssignmentAsmGen(private val program: Program,
             "/" -> {
                 asmgen.out("""
                     jsr  c64flt.pop_float_fac1
-                    stx  P8ZP_SCRATCH_REG_X
                     lda  #<$name
                     ldy  #>$name
                     jsr  c64flt.FDIV
@@ -1095,8 +1138,8 @@ internal class AugmentableAssignmentAsmGen(private val program: Program,
             ldx  #<$name
             ldy  #>$name
             jsr  c64flt.MOVMF
-            ldx  P8ZP_SCRATCH_REG_X
         """)
+        asmgen.restoreRegister(CpuRegister.X)
     }
 
     private fun inplaceModification_float_variable_to_variable(name: String, operator: String, ident: IdentifierReference) {
@@ -1105,10 +1148,10 @@ internal class AugmentableAssignmentAsmGen(private val program: Program,
             throw AssemblyError("float variable expected")
 
         val otherName = asmgen.asmVariableName(ident)
+        asmgen.saveRegister(CpuRegister.X)
         when (operator) {
             "**" -> {
                 asmgen.out("""
-                    stx  P8ZP_SCRATCH_REG_X
                     lda  #<$name
                     ldy  #>$name
                     jsr  c64flt.CONUPK
@@ -1119,7 +1162,6 @@ internal class AugmentableAssignmentAsmGen(private val program: Program,
             }
             "+" -> {
                 asmgen.out("""
-                    stx  P8ZP_SCRATCH_REG_X
                     lda  #<$name
                     ldy  #>$name
                     jsr  c64flt.MOVFM
@@ -1130,7 +1172,6 @@ internal class AugmentableAssignmentAsmGen(private val program: Program,
             }
             "-" -> {
                 asmgen.out("""
-                    stx  P8ZP_SCRATCH_REG_X
                     lda  #<$otherName
                     ldy  #>$otherName
                     jsr  c64flt.MOVFM
@@ -1141,7 +1182,6 @@ internal class AugmentableAssignmentAsmGen(private val program: Program,
             }
             "*" -> {
                 asmgen.out("""
-                    stx  P8ZP_SCRATCH_REG_X
                     lda  #<$name
                     ldy  #>$name
                     jsr  c64flt.MOVFM
@@ -1152,7 +1192,6 @@ internal class AugmentableAssignmentAsmGen(private val program: Program,
             }
             "/" -> {
                 asmgen.out("""
-                    stx  P8ZP_SCRATCH_REG_X
                     lda  #<$otherName
                     ldy  #>$otherName
                     jsr  c64flt.MOVFM
@@ -1168,16 +1207,16 @@ internal class AugmentableAssignmentAsmGen(private val program: Program,
             ldx  #<$name
             ldy  #>$name
             jsr  c64flt.MOVMF
-            ldx  P8ZP_SCRATCH_REG_X
         """)
+        asmgen.restoreRegister(CpuRegister.X)
     }
 
     private fun inplaceModification_float_litval_to_variable(name: String, operator: String, value: Double) {
         val constValueName = asmgen.getFloatAsmConst(value)
+        asmgen.saveRegister(CpuRegister.X)
         when (operator) {
             "**" -> {
                 asmgen.out("""
-                    stx  P8ZP_SCRATCH_REG_X
                     lda  #<$name
                     ldy  #>$name
                     jsr  c64flt.CONUPK
@@ -1190,7 +1229,6 @@ internal class AugmentableAssignmentAsmGen(private val program: Program,
                 if (value == 0.0)
                     return
                 asmgen.out("""
-                    stx  P8ZP_SCRATCH_REG_X
                     lda  #<$name
                     ldy  #>$name
                     jsr  c64flt.MOVFM
@@ -1203,7 +1241,6 @@ internal class AugmentableAssignmentAsmGen(private val program: Program,
                 if (value == 0.0)
                     return
                 asmgen.out("""
-                    stx  P8ZP_SCRATCH_REG_X
                     lda  #<$constValueName
                     ldy  #>$constValueName
                     jsr  c64flt.MOVFM
@@ -1214,7 +1251,6 @@ internal class AugmentableAssignmentAsmGen(private val program: Program,
             }
             "*" -> {
                 asmgen.out("""
-                    stx  P8ZP_SCRATCH_REG_X
                     lda  #<$name
                     ldy  #>$name
                     jsr  c64flt.MOVFM
@@ -1227,7 +1263,6 @@ internal class AugmentableAssignmentAsmGen(private val program: Program,
                 if (value == 0.0)
                     throw AssemblyError("division by zero")
                 asmgen.out("""
-                    stx  P8ZP_SCRATCH_REG_X
                     lda  #<$constValueName
                     ldy  #>$constValueName
                     jsr  c64flt.MOVFM
@@ -1243,8 +1278,8 @@ internal class AugmentableAssignmentAsmGen(private val program: Program,
             ldx  #<$name
             ldy  #>$name
             jsr  c64flt.MOVMF
-            ldx  P8ZP_SCRATCH_REG_X
         """)
+        asmgen.restoreRegister(CpuRegister.X)
     }
 
     private fun inplaceCast(target: AsmAssignTarget, cast: TypecastExpression, position: Position) {
@@ -1259,12 +1294,25 @@ internal class AugmentableAssignmentAsmGen(private val program: Program,
                     when (outerCastDt) {
                         DataType.UBYTE, DataType.BYTE -> {
                             when(target.kind) {
-                                TargetStorageKind.VARIABLE -> asmgen.out(" lda  #0 |  sta  ${target.asmVarname}+1")
+                                TargetStorageKind.VARIABLE -> {
+                                    if(CompilationTarget.machine.cpu == CpuType.CPU65c02)
+                                        asmgen.out(" stz  ${target.asmVarname}+1")
+                                    else
+                                        asmgen.out(" lda  #0 |  sta  ${target.asmVarname}+1")
+                                }
                                 TargetStorageKind.ARRAY -> {
                                     asmgen.loadScaledArrayIndexIntoRegister(target.array!!, target.datatype, CpuRegister.Y, true)
-                                    asmgen.out("  lda  #0 |  sta  ${target.asmVarname},y")
+                                    if(CompilationTarget.machine.cpu == CpuType.CPU65c02)
+                                        asmgen.out("  stz  ${target.asmVarname},y")
+                                    else
+                                        asmgen.out("  lda  #0 |  sta  ${target.asmVarname},y")
+                                }
+                                TargetStorageKind.STACK -> {
+                                    if(CompilationTarget.machine.cpu == CpuType.CPU65c02)
+                                        asmgen.out(" stz  P8ESTACK_HI+1,x")
+                                    else
+                                        asmgen.out(" lda  #0 |  sta  P8ESTACK_HI+1,x")
                                 }
-                                TargetStorageKind.STACK -> asmgen.out(" lda  #0 |  sta  P8ESTACK_HI+1,x")
                                 else -> throw AssemblyError("weird target")
                             }
                         }
@@ -1468,8 +1516,8 @@ internal class AugmentableAssignmentAsmGen(private val program: Program,
             DataType.FLOAT -> {
                 when(target.kind) {
                     TargetStorageKind.VARIABLE -> {
+                        asmgen.saveRegister(CpuRegister.X)
                         asmgen.out("""
-                            stx  P8ZP_SCRATCH_REG_X
                             lda  #<${target.asmVarname}
                             ldy  #>${target.asmVarname}
                             jsr  c64flt.MOVFM
@@ -1477,8 +1525,8 @@ internal class AugmentableAssignmentAsmGen(private val program: Program,
                             ldx  #<${target.asmVarname}
                             ldy  #>${target.asmVarname}
                             jsr  c64flt.MOVMF
-                            ldx  P8ZP_SCRATCH_REG_X
                         """)
+                        asmgen.restoreRegister(CpuRegister.X)
                     }
                     TargetStorageKind.ARRAY -> TODO("in-place negate float array")
                     TargetStorageKind.STACK -> TODO("stack float negate")

compiler/src/prog8/compiler/target/cx16/CX16MachineDefinition.kt

@@ -2,6 +2,7 @@ package prog8.compiler.target.cx16
 
 import prog8.ast.Program
 import prog8.compiler.*
+import prog8.compiler.target.CpuType
 import prog8.compiler.target.IMachineDefinition
 import prog8.compiler.target.c64.C64MachineDefinition
 import prog8.parser.ModuleImporter
@@ -9,7 +10,7 @@ import java.io.IOException
 
 internal object CX16MachineDefinition: IMachineDefinition {
 
-    override val cpu = "65c02"
+    override val cpu = CpuType.CPU65c02
 
     // 5-byte cbm MFLPT format limitations:
     override val FLOAT_MAX_POSITIVE = 1.7014118345e+38         // bytes: 255,127,255,255,255
@@ -29,7 +30,7 @@ internal object CX16MachineDefinition: IMachineDefinition {
 
     override fun getFloat(num: Number) = C64MachineDefinition.Mflpt5.fromNumber(num)
 
-    override fun getFloatRomConst(number: Double): String? = null       // TODO Does Cx16 have ROM float locations?
+    override fun getFloatRomConst(number: Double): String? = null       // Cx16 has no pulblic ROM float locations
     override fun importLibs(compilerOptions: CompilationOptions, importer: ModuleImporter, program: Program) {
         if (compilerOptions.launcher == LauncherType.BASIC || compilerOptions.output == OutputType.PRG)
             importer.importLibraryModule(program, "cx16lib")
@@ -57,16 +58,17 @@ internal object CX16MachineDefinition: IMachineDefinition {
     }
 
     // 6502 opcodes (including aliases and illegal opcodes), these cannot be used as variable or label names
-    // TODO add 65C02 opcodes
-    override val opcodeNames = setOf("adc", "ahx", "alr", "anc", "and", "ane", "arr", "asl", "asr", "axs", "bcc", "bcs",
+    override val opcodeNames = setOf("adc", "and", "asl", "bcc", "bcs",
             "beq", "bge", "bit", "blt", "bmi", "bne", "bpl", "brk", "bvc", "bvs", "clc",
-            "cld", "cli", "clv", "cmp", "cpx", "cpy", "dcm", "dcp", "dec", "dex", "dey",
+            "cld", "cli", "clv", "cmp", "cpx", "cpy", "dec", "dex", "dey",
             "eor", "gcc", "gcs", "geq", "gge", "glt", "gmi", "gne", "gpl", "gvc", "gvs",
-            "inc", "ins", "inx", "iny", "isb", "isc", "jam", "jmp", "jsr", "lae", "las",
-            "lax", "lda", "lds", "ldx", "ldy", "lsr", "lxa", "nop", "ora", "pha", "php",
-            "pla", "plp", "rla", "rol", "ror", "rra", "rti", "rts", "sax", "sbc", "sbx",
-            "sec", "sed", "sei", "sha", "shl", "shr", "shs", "shx", "shy", "slo", "sre",
-            "sta", "stx", "sty", "tas", "tax", "tay", "tsx", "txa", "txs", "tya", "xaa")
+            "inc", "inx", "iny", "jmp", "jsr",
+            "lda", "ldx", "ldy", "lsr", "nop", "ora", "pha", "php",
+            "pla", "plp", "rol", "ror", "rti", "rts", "sbc",
+            "sec", "sed", "sei",
+            "sta", "stx", "sty", "tax", "tay", "tsx", "txa", "txs", "tya",
+            "bra", "phx", "phy", "plx", "ply", "stz", "trb", "tsb", "bbr", "bbs",
+            "rmb", "smb", "stp", "wai")
 
 
     internal class CX16Zeropage(options: CompilationOptions) : Zeropage(options) {

compiler/src/prog8/optimizer/ConstantFoldingOptimizer.kt

@@ -318,11 +318,11 @@ internal class ConstantFoldingOptimizer(private val program: Program) : AstWalke
     }
 
     override fun after(forLoop: ForLoop, parent: Node): Iterable<IAstModification> {
-        fun adjustRangeDt(rangeFrom: NumericLiteralValue, targetDt: DataType, rangeTo: NumericLiteralValue, stepLiteral: NumericLiteralValue?, range: RangeExpr): RangeExpr {
+        fun adjustRangeDt(rangeFrom: NumericLiteralValue, targetDt: DataType, rangeTo: NumericLiteralValue, stepLiteral: NumericLiteralValue?, range: RangeExpr): RangeExpr? {
             val fromCast = rangeFrom.cast(targetDt)
             val toCast = rangeTo.cast(targetDt)
             if(!fromCast.isValid || !toCast.isValid)
-                return range
+                return null
 
             val newStep =
                 if(stepLiteral!=null) {
@@ -351,28 +351,32 @@ internal class ConstantFoldingOptimizer(private val program: Program) : AstWalke
                 if(rangeFrom.type!= DataType.UBYTE) {
                     // attempt to translate the iterable into ubyte values
                     val newIter = adjustRangeDt(rangeFrom, loopvar.datatype, rangeTo, stepLiteral, iterableRange)
-                    return listOf(IAstModification.ReplaceNode(forLoop.iterable, newIter, forLoop))
+                    if(newIter!=null)
+                        return listOf(IAstModification.ReplaceNode(forLoop.iterable, newIter, forLoop))
                 }
             }
             DataType.BYTE -> {
                 if(rangeFrom.type!= DataType.BYTE) {
                     // attempt to translate the iterable into byte values
                     val newIter = adjustRangeDt(rangeFrom, loopvar.datatype, rangeTo, stepLiteral, iterableRange)
-                    return listOf(IAstModification.ReplaceNode(forLoop.iterable, newIter, forLoop))
+                    if(newIter!=null)
+                        return listOf(IAstModification.ReplaceNode(forLoop.iterable, newIter, forLoop))
                 }
             }
             DataType.UWORD -> {
                 if(rangeFrom.type!= DataType.UWORD) {
                     // attempt to translate the iterable into uword values
                     val newIter = adjustRangeDt(rangeFrom, loopvar.datatype, rangeTo, stepLiteral, iterableRange)
-                    return listOf(IAstModification.ReplaceNode(forLoop.iterable, newIter, forLoop))
+                    if(newIter!=null)
+                        return listOf(IAstModification.ReplaceNode(forLoop.iterable, newIter, forLoop))
                 }
             }
             DataType.WORD -> {
                 if(rangeFrom.type!= DataType.WORD) {
                     // attempt to translate the iterable into word values
                     val newIter = adjustRangeDt(rangeFrom, loopvar.datatype, rangeTo, stepLiteral, iterableRange)
-                    return listOf(IAstModification.ReplaceNode(forLoop.iterable, newIter, forLoop))
+                    if(newIter!=null)
+                        return listOf(IAstModification.ReplaceNode(forLoop.iterable, newIter, forLoop))
                 }
             }
             else -> throw FatalAstException("invalid loopvar datatype $loopvar")

docs/source/building.rst

@@ -83,7 +83,7 @@ For normal use the compiler is invoked with the command:
 By default, assembly code is generated and written to ``sourcefile.asm``.
 It is then (automatically) fed to the `64tass <https://sourceforge.net/projects/tass64/>`_ cross assembler tool
 that assembles it into the final program.
-If you use the option to let the compiler auto-start a C-64 emulator, it will do so after
+If you use the option to let the compiler auto-start an emulator, it will do so after
 a successful compilation. This will load your program and the symbol and breakpoint lists
 (for the machine code monitor) into the emulator.
 

docs/source/index.rst

@@ -12,6 +12,7 @@ What is Prog8?
 
 This is an experimental compiled programming language targeting the 8-bit
 `6502 <https://en.wikipedia.org/wiki/MOS_Technology_6502>`_ /
+`65c02 <https://en.wikipedia.org/wiki/MOS_Technology_65C02>`_ /
 `6510 <https://en.wikipedia.org/wiki/MOS_Technology_6510>`_ microprocessor.
 This CPU is from the late 1970's and early 1980's and was used in many home computers from that era,
 such as the `Commodore-64 <https://en.wikipedia.org/wiki/Commodore_64>`_.
@@ -95,7 +96,6 @@ when compiled an ran on a C-64 you get this:
     :alt: result when run on C-64
 
 
-// TODO fix code example
 The following programs shows a use of the high level ``struct`` type::
 
 
@@ -179,6 +179,7 @@ Fnd for Windows it's possible to get that as well. Check out `AdoptOpenJDK <http
 
 Finally: a **C-64 emulator** (or a real C-64 ofcourse) can be nice to test and run your programs on.
 The compiler assumes the presence of the `Vice emulator <http://vice-emu.sourceforge.net/>`_.
+If you're targeting the CommanderX16, there's the `x16emu <https://github.com/commanderx16/x16-emulator>`_.
 
 .. important::
     **Building the compiler itself:** (*Only needed if you have not downloaded a pre-built 'fat-jar'*)

docs/source/targetsystem.rst

@@ -9,10 +9,12 @@ Prog8 targets the following hardware:
 - optional use of memory mapped I/O registers
 - optional use of system ROM routines
 
-The main target machine is the well-known Commodore-64, which is an example of this.
-Another (preliminary) supported target machine is the `CommanderX16 <https://www.commanderx16.com/>`_ .
+Currently there are two machines that are supported as compiler target (via the ``-target`` compiler argument):
 
-This chapter explains the relevant system details of such machines.
+- 'c64': the well-known Commodore-64, premium support
+- 'cx16': the `CommanderX16 <https://www.commanderx16.com/>`_ a project from the 8-Bit Guy. Support for this is still experimental.
+
+This chapter explains the relevant system details of these machines.
 
 
 Memory Model

examples/arithmetic/div.p8

@@ -2,8 +2,6 @@
 %import c64textio
 %zeropage basicsafe
 
-; TODO implement signed byte/word DIV asm generation, fix unsigned DIV asm generation (for in-place)
-
 main {
 
     sub start() {
@@ -11,17 +9,17 @@ main {
         div_ubyte(100, 6, 16)
         div_ubyte(255, 2, 127)
 
-        ;div_byte(0, 1, 0)      ; TODO implement
-        ;div_byte(100, -6, -16) ; TODO implement
-        ;div_byte(127, -2, -63) ; TODO implement
+        div_byte(0, 1, 0)
+        div_byte(100, -6, -16)
+        div_byte(127, -2, -63)
 
         div_uword(0,1,0)
         div_uword(40000,500,80)
         div_uword(43211,2,21605)
 
-        ;div_word(0,1,0)            ; TODO implement
-        ;div_word(-20000,500,-40)   ; TODO implement
-        ;div_word(-2222,2,-1111)    ; TODO implement
+        div_word(0,1,0)
+        div_word(-20000,500,-40)
+        div_word(-2222,2,-1111)
 
         div_float(0,1,0)
         div_float(999.9,111.0,9.008108108108107)

examples/bdmusic.p8

@@ -47,10 +47,10 @@ sub delay() {
 sub print_notes(ubyte n1, ubyte n2) {
     c64.CHROUT('\n')
     txt.plot(n1/2, 24)
-    c64.COLOR=7
+    txt.color(7)
     c64.CHROUT('Q')
     txt.plot(n2/2, 24)
-    c64.COLOR=4
+    txt.color(4)
     c64.CHROUT('Q')
 }
 

examples/c64graphics.p8

@@ -1,256 +0,0 @@
-%import c64textio
-
-; bitmap pixel graphics module for the C64
-; only black/white monchrome for now
-
-; you could put this code at $4000 which is after the bitmap screen in memory ($2000-$3fff),
-; this leaves more space for user program code.
-
-graphics {
-    const uword bitmap_address = $2000
-
-    sub enable_bitmap_mode() {
-        ; enable bitmap screen, erase it and set colors to black/white.
-        c64.SCROLY |= %00100000
-        c64.VMCSB = (c64.VMCSB & %11110000) | %00001000   ; $2000-$3fff
-        clear_screen()
-    }
-
-    sub clear_screen() {
-        memset(bitmap_address, 320*200/8, 0)
-        txt.clear_screen($10, 0)         ; pixel color $1 (white) backround $0 (black)
-    }
-
-    sub line(uword x1, ubyte y1, uword x2, ubyte y2) {
-        ; Bresenham algorithm.
-        ; This code special cases various quadrant loops to allow simple ++ and -- operations.
-        if y1>y2 {
-            ; make sure dy is always positive to avoid 8 instead of just 4 special cases
-            swap(x1, x2)
-            swap(y1, y2)
-        }
-        word d = 0
-        ubyte positive_ix = true
-        word dx = x2 - x1 as word
-        word dy = y2 as word - y1 as word
-        if dx < 0 {
-            dx = -dx
-            positive_ix = false
-        }
-        dx *= 2
-        dy *= 2
-        plotx = x1
-
-        if dx >= dy {
-            if positive_ix {
-                repeat {
-                    plot(y1)
-                    if plotx==x2
-                        return
-                    plotx++
-                    d += dy
-                    if d > dx {
-                        y1++
-                        d -= dx
-                    }
-                }
-            } else {
-                repeat {
-                    plot(y1)
-                    if plotx==x2
-                        return
-                    plotx--
-                    d += dy
-                    if d > dx {
-                        y1++
-                        d -= dx
-                    }
-                }
-            }
-        }
-        else {
-            if positive_ix {
-                repeat {
-                    plot(y1)
-                    if y1 == y2
-                        return
-                    y1++
-                    d += dx
-                    if d > dy {
-                        plotx++
-                        d -= dy
-                    }
-                }
-            } else {
-                repeat {
-                    plot(y1)
-                    if y1 == y2
-                        return
-                    y1++
-                    d += dx
-                    if d > dy {
-                        plotx--
-                        d -= dy
-                    }
-                }
-            }
-        }
-    }
-
-    sub circle(uword xcenter, ubyte ycenter, ubyte radius) {
-        ; Midpoint algorithm
-        ubyte ploty
-        ubyte xx = radius
-        ubyte yy = 0
-        byte decisionOver2 = 1-xx as byte
-
-        while xx>=yy {
-            plotx = xcenter + xx
-            ploty = ycenter + yy
-            plot(ploty)
-            plotx = xcenter - xx
-            plot(ploty)
-            plotx = xcenter + xx
-            ploty = ycenter - yy
-            plot(ploty)
-            plotx = xcenter - xx
-            plot(ploty)
-            plotx = xcenter + yy
-            ploty = ycenter + xx
-            plot(ploty)
-            plotx = xcenter - yy
-            plot(ploty)
-            plotx = xcenter + yy
-            ploty = ycenter - xx
-            plot(ploty)
-            plotx = xcenter - yy
-            plot(ploty)
-            yy++
-            if decisionOver2<=0
-                decisionOver2 += 2*yy+1
-            else {
-                xx--
-                decisionOver2 += 2*(yy-xx)+1
-            }
-        }
-    }
-
-    sub disc(uword cx, ubyte cy, ubyte radius) {
-        ; Midpoint algorithm, filled
-        ubyte xx = radius
-        ubyte yy = 0
-        byte decisionOver2 = 1-xx as byte
-
-        while xx>=yy {
-            ubyte cy_plus_yy = cy + yy
-            ubyte cy_min_yy = cy - yy
-            ubyte cy_plus_xx = cy + xx
-            ubyte cy_min_xx = cy - xx
-
-            for plotx in cx to cx+xx {
-                plot(cy_plus_yy)
-                plot(cy_min_yy)
-            }
-            for plotx in cx-xx to cx-1 {
-                plot(cy_plus_yy)
-                plot(cy_min_yy)
-            }
-            for plotx in cx to cx+yy {
-                plot(cy_plus_xx)
-                plot(cy_min_xx)
-            }
-            for plotx in cx-yy to cx {
-                plot(cy_plus_xx)
-                plot(cy_min_xx)
-            }
-            yy++
-            if decisionOver2<=0
-                decisionOver2 += 2*yy+1
-            else {
-                xx--
-                decisionOver2 += 2*(yy-xx)+1
-            }
-        }
-    }
-
-
-; here is the non-asm code for the plot routine below:
-;    sub plot_nonasm(uword px, ubyte py) {
-;        ubyte[] ormask = [128, 64, 32, 16, 8, 4, 2, 1]
-;        uword addr = bitmap_address + 320*(py>>3) + (py & 7) + (px & %0000000111111000)
-;        @(addr) |= ormask[lsb(px) & 7]
-;    }
-
-    uword plotx     ; 0..319        ; separate 'parameter' for plot()
-
-    asmsub plot(ubyte ploty @A) {           ; plotx is 16 bits 0 to 319... doesn't fit in a register
-        %asm {{
-        tay
-        stx  P8ZP_SCRATCH_REG_X
-        lda  plotx+1
-        sta  P8ZP_SCRATCH_W2+1
-        lsr  a            ; 0
-        sta  P8ZP_SCRATCH_W2
-        lda  plotx
-        pha
-        and  #7
-        tax
-
-        lda  _y_lookup_lo,y
-        clc
-        adc  P8ZP_SCRATCH_W2
-        sta  P8ZP_SCRATCH_W2
-        lda  _y_lookup_hi,y
-        adc  P8ZP_SCRATCH_W2+1
-        sta  P8ZP_SCRATCH_W2+1
-
-        pla     ; plotx
-        and  #%11111000
-        tay
-        lda  (P8ZP_SCRATCH_W2),y
-        ora  _ormask,x
-        sta  (P8ZP_SCRATCH_W2),y
-
-        ldx  P8ZP_SCRATCH_REG_X
-        rts
-
-_ormask     .byte 128, 64, 32, 16, 8, 4, 2, 1
-
-; note: this can be even faster if we also have a 256 byte x-lookup table, but hey.
-; see http://codebase64.org/doku.php?id=base:various_techniques_to_calculate_adresses_fast_common_screen_formats_for_pixel_graphics
-; the y lookup tables encodes this formula:  bitmap_address + 320*(py>>3) + (py & 7)    (y from 0..199)
-_y_lookup_hi
-            .byte  $20, $20, $20, $20, $20, $20, $20, $20, $21, $21, $21, $21, $21, $21, $21, $21
-            .byte  $22, $22, $22, $22, $22, $22, $22, $22, $23, $23, $23, $23, $23, $23, $23, $23
-            .byte  $25, $25, $25, $25, $25, $25, $25, $25, $26, $26, $26, $26, $26, $26, $26, $26
-            .byte  $27, $27, $27, $27, $27, $27, $27, $27, $28, $28, $28, $28, $28, $28, $28, $28
-            .byte  $2a, $2a, $2a, $2a, $2a, $2a, $2a, $2a, $2b, $2b, $2b, $2b, $2b, $2b, $2b, $2b
-            .byte  $2c, $2c, $2c, $2c, $2c, $2c, $2c, $2c, $2d, $2d, $2d, $2d, $2d, $2d, $2d, $2d
-            .byte  $2f, $2f, $2f, $2f, $2f, $2f, $2f, $2f, $30, $30, $30, $30, $30, $30, $30, $30
-            .byte  $31, $31, $31, $31, $31, $31, $31, $31, $32, $32, $32, $32, $32, $32, $32, $32
-            .byte  $34, $34, $34, $34, $34, $34, $34, $34, $35, $35, $35, $35, $35, $35, $35, $35
-            .byte  $36, $36, $36, $36, $36, $36, $36, $36, $37, $37, $37, $37, $37, $37, $37, $37
-            .byte  $39, $39, $39, $39, $39, $39, $39, $39, $3a, $3a, $3a, $3a, $3a, $3a, $3a, $3a
-            .byte  $3b, $3b, $3b, $3b, $3b, $3b, $3b, $3b, $3c, $3c, $3c, $3c, $3c, $3c, $3c, $3c
-            .byte  $3e, $3e, $3e, $3e, $3e, $3e, $3e, $3e
-
-_y_lookup_lo
-            .byte $00, $01, $02, $03, $04, $05, $06, $07, $40, $41, $42, $43, $44, $45, $46, $47
-            .byte $80, $81, $82, $83, $84, $85, $86, $87, $c0, $c1, $c2, $c3, $c4, $c5, $c6, $c7
-            .byte $00, $01, $02, $03, $04, $05, $06, $07, $40, $41, $42, $43, $44, $45, $46, $47
-            .byte $80, $81, $82, $83, $84, $85, $86, $87, $c0, $c1, $c2, $c3, $c4, $c5, $c6, $c7
-            .byte $00, $01, $02, $03, $04, $05, $06, $07, $40, $41, $42, $43, $44, $45, $46, $47
-            .byte $80, $81, $82, $83, $84, $85, $86, $87, $c0, $c1, $c2, $c3, $c4, $c5, $c6, $c7
-            .byte $00, $01, $02, $03, $04, $05, $06, $07, $40, $41, $42, $43, $44, $45, $46, $47
-            .byte $80, $81, $82, $83, $84, $85, $86, $87, $c0, $c1, $c2, $c3, $c4, $c5, $c6, $c7
-            .byte $00, $01, $02, $03, $04, $05, $06, $07, $40, $41, $42, $43, $44, $45, $46, $47
-            .byte $80, $81, $82, $83, $84, $85, $86, $87, $c0, $c1, $c2, $c3, $c4, $c5, $c6, $c7
-            .byte $00, $01, $02, $03, $04, $05, $06, $07, $40, $41, $42, $43, $44, $45, $46, $47
-            .byte $80, $81, $82, $83, $84, $85, $86, $87, $c0, $c1, $c2, $c3, $c4, $c5, $c6, $c7
-            .byte $00, $01, $02, $03, $04, $05, $06, $07
-        }}
-    }
-
-}
-
-

examples/cube3d-gfx.p8

@@ -32,7 +32,7 @@ main {
 
         repeat {
             rotate_vertices(msb(anglex), msb(angley), msb(anglez))
-            graphics.clear_screen()
+            graphics.clear_screen(1, 0)
             draw_lines()
             anglex-=500
             angley+=217

examples/cx16/floats.p8

@@ -0,0 +1,23 @@
+; CommanderX16 floating point example!
+; make sure to compile with the cx16 compiler target.
+
+%import cx16textio
+%import cx16flt
+%zeropage basicsafe
+
+main {
+
+    sub start() {
+        float f1 = 5.55
+        float f2 = 33.3
+        float f3 = f1 * f2
+
+        c64flt.print_f(f1)
+        c64.CHROUT('*')
+        c64flt.print_f(f2)
+        c64.CHROUT('=')
+        c64flt.print_f(f3)
+        c64.CHROUT('\n')
+    }
+}
+

examples/cx16/graphics.p8

@@ -0,0 +1,39 @@
+; CommanderX16 text clock example!
+; make sure to compile with the cx16 compiler target.
+
+%zeropage basicsafe
+
+main {
+
+    sub start() {
+
+        void cx16.screen_set_mode($80)
+        cx16.r0=0
+
+        cx16.FB_init()
+        cx16.r0 = 0
+        cx16.r1 = 0
+        cx16.FB_cursor_position()
+        uword xx
+        ubyte yy
+        for yy in 199 downto 0 {
+            for xx in 319 downto 0 {
+                cx16.FB_set_pixel( yy+lsb(xx))
+            }
+        }
+
+        ; cx16.GRAPH_init()
+        for xx in 0 to 319 step 32 {
+            ubyte q
+            for q in 0 to 31 {
+                cx16.GRAPH_set_colors(q, 2, 0)
+                cx16.r0 = xx+q
+                cx16.r1=0
+                cx16.r2=rnd()
+                cx16.r3=199
+                cx16.GRAPH_draw_line()
+            }
+        }
+    }
+}
+

examples/cx16/mandelbrot-gfx.bas

@@ -0,0 +1,31 @@
+100 REM MANDELBROT GFX - BASIC VERSION FOR COMMANDERX16
+110 REM MANDELBROT COORDINATES
+120 XL=-2.000:XU=0.500
+130 YL=-1.100:YU=1.100
+140 MI%=16 :REM MAXIMUM ITERATIONS
+150 WI%=256:HI%=200 :REM SCREEN DIMENSIONS
+160 DX=(XU-XL)/WI%
+170 DY=(YU-YL)/HI%
+200 SCREEN $80:TI$="000000"
+500 REM MAIN LOOP
+520 X0=0:Y0=0:GOSUB2000
+530 FOR YY=0 TO HI%-1
+540 Y0=YL+DY*YY
+610 FOR XX=0 TO WI%-1
+620 X0=XL+DX*XX
+700 REM CALCULATE MANDELBROT
+710 X=0:Y=0:IT%=0
+720 X2=0:Y2=0
+730 XY=X*Y
+740 X=X2-Y2+X0
+750 Y=2*XY+Y0
+760 IT%=IT%+1
+770 X2=X*X:Y2=Y*Y
+780 IF (X2+Y2 <= 4) AND (IT% < MI%) THEN GOTO 730
+790 PSET XX,YY,MI%-IT%
+800 NEXT XX
+810 GOSUB2000
+820 NEXT YY
+1000 GOTO 1000
+2000 REM OUTPUT TIME SPENT
+2010 PRINT CHR$(19)SPC(33)TI$:RETURN

examples/cx16/mandelbrot-gfx.p8

@@ -0,0 +1,94 @@
+%import cx16textio
+%import cx16flt
+%zeropage basicsafe
+
+main {
+    const uword width = 256
+    const uword height = 200
+    const ubyte max_iter = 16       ; 32 looks pretty nice
+
+    sub start()  {
+        initialize()
+        mandel()
+        repeat {
+            ; do nothing
+        }
+    }
+
+    sub mandel() {
+        const float XL=-2.000
+        const float XU=0.500
+        const float YL=-1.100
+        const float YU=1.100
+        float dx = (XU-XL)/width
+        float dy = (YU-YL)/height
+        ubyte pixelx
+        ubyte pixely
+
+        for pixely in 0 to height-1 {
+            float yy = YL+dy*(pixely as float)
+
+            cx16.r0 = 0
+            cx16.r1 = pixely
+            cx16.FB_cursor_position()
+
+            for pixelx in 0 to width-1 {
+                float xx = XL+dx*(pixelx as float)
+
+                float xsquared = 0.0
+                float ysquared = 0.0
+                float x = 0.0
+                float y = 0.0
+                ubyte iter = 0
+
+                while iter<max_iter and xsquared+ysquared<4.0 {
+                    y = x*y*2.0 + yy
+                    x = xsquared - ysquared + xx
+                    xsquared = x*x
+                    ysquared = y*y
+                    iter++
+                }
+                cx16.FB_set_pixel(max_iter-iter)
+            }
+
+            print_time()
+        }
+    }
+
+    sub initialize() {
+        void cx16.screen_set_mode($80)
+
+        txt.plot(32, 5)
+        txt.print("256*200")
+        txt.plot(32, 6)
+        txt.print("mandel-")
+        txt.plot(33, 7)
+        txt.print("brot")
+        txt.plot(32, 9)
+        txt.print("floats")
+        txt.plot(32, 10)
+        txt.print_b(max_iter)
+        txt.print(" iter")
+
+        cx16.r0 = 0
+        cx16.r1 = 0
+        cx16.r2 = 0
+        cx16.r3 = 0
+        cx16.clock_set_date_time()
+
+        cx16.r0=0
+        cx16.FB_init()
+    }
+
+    sub print_time() {
+        cx16.clock_get_date_time()
+        txt.plot(33, 12)
+        if lsb(cx16.r2) < 10
+            c64.CHROUT('0')
+        txt.print_ub(lsb(cx16.r2))
+        c64.CHROUT(':')
+        if msb(cx16.r2) < 10
+            c64.CHROUT('0')
+        txt.print_ub(msb(cx16.r2))
+    }
+}

examples/cx16/mandelbrot.p8

@@ -0,0 +1,42 @@
+%import cx16textio
+%import cx16flt
+%zeropage basicsafe
+
+main {
+    const uword width = 60
+    const uword height = 50
+    const ubyte max_iter = 16
+
+    sub start()  {
+        txt.print("calculating mandelbrot fractal...\n\n")
+
+        ubyte pixelx
+        ubyte pixely
+
+        for pixely in 0 to height-1 {
+            float yy = (pixely as float)/0.40/height - 1.3
+
+            for pixelx in 0 to width-1 {
+                float xx = (pixelx as float)/0.32/width - 2.2
+
+                float xsquared = 0.0
+                float ysquared = 0.0
+                float x = 0.0
+                float y = 0.0
+                ubyte iter = 0
+
+                while iter<max_iter and xsquared+ysquared<4.0 {
+                    y = x*y*2.0 + yy
+                    x = xsquared - ysquared + xx
+                    xsquared = x*x
+                    ysquared = y*y
+                    iter++
+                }
+                ; txt.setchr(pixelx, pixely, '*')
+                txt.color2(1, max_iter-iter)
+                c64.CHROUT(' ')
+            }
+            c64.CHROUT('\n')
+        }
+    }
+}

examples/cx16/swirl-float.p8

@@ -0,0 +1,27 @@
+%import cx16textio
+%import cx16flt
+%zeropage basicsafe
+
+main {
+
+    const uword width = 80
+    const uword height = 60
+
+    struct Ball {
+        float t
+        ubyte color
+    }
+
+    sub start()  {
+
+        Ball ball
+
+        repeat {
+            ubyte xx=(sin(ball.t) * width/2.1) + width/2.0 as ubyte
+            ubyte yy=(cos(ball.t*1.1356) * height/2.1) + height/2.0 as ubyte
+            txt.setcc(xx, yy, 81, ball.color)
+            ball.t  += 0.05
+            ball.color++
+        }
+    }
+}

examples/cx16/swirl.p8

@@ -0,0 +1,30 @@
+%import cx16lib
+%import cx16textio
+%zeropage basicsafe
+
+main {
+
+    const uword width = 79
+    const uword height = 59
+
+    struct Ball {
+        uword anglex
+        uword angley
+        ubyte color
+    }
+
+    sub start()  {
+
+        Ball ball
+
+        repeat {
+            ubyte x = msb(sin8u(msb(ball.anglex)) as uword * width)
+            ubyte y = msb(cos8u(msb(ball.angley)) as uword * height)
+            txt.setcc(x, y, 81, ball.color)
+
+            ball.anglex+=266
+            ball.angley+=215
+            ball.color++
+        }
+    }
+}

examples/cx16/textclock.p8

@@ -8,18 +8,34 @@ main {
 
     sub start() {
 
-        cx16.r0L = 2020 - 1900
-        cx16.r0H = 8
-        cx16.r1L = 27
-        cx16.r1H = 19
-        cx16.r2L = 16
-        cx16.r2H = 0
-        cx16.r3L = 0
+;        %asm {{
+;            lda  #$80
+;            jsr  cx16.screen_set_mode
+;        }}
+;        cx16.r0=0
+;        cx16.GRAPH_init()
+;        %asm {{
+;            lda  #4
+;            ldy  #0
+;            ldx  #1
+;            jsr  cx16.GRAPH_set_colors
+;        }}
+;        cx16.GRAPH_clear()
+;        cx16.r0=10
+;        cx16.r1=10
+;        cx16.r2=100
+;        cx16.r3=150
+;        cx16.GRAPH_draw_line()
+;
+;        repeat {
+;        }
+
+        cx16.r0 = mkword(8, 2020 - 1900)
+        cx16.r1 = mkword(19, 27)
+        cx16.r2 = mkword(0, 16)
+        cx16.r3 = 0
         cx16.clock_set_date_time()
-
-
-        cx16.screen_set_charset(3, 0)
-        ; c64.CHROUT(14)      ; lowercase charset
+        cx16.screen_set_charset(3, 0)  ; lowercase charset
 
         repeat {
             c64.CHROUT(19)      ; HOME
@@ -33,33 +49,33 @@ main {
     }
 
     sub print_date() {
-        txt.print_uw(1900 + cx16.r0L)
+        txt.print_uw(1900 + lsb(cx16.r0))
         c64.CHROUT('-')
-        if cx16.r0H < 10
+        if msb(cx16.r0) < 10
             c64.CHROUT('0')
-        txt.print_ub(cx16.r0H)
+        txt.print_ub(msb(cx16.r0))
         c64.CHROUT('-')
-        if cx16.r1L < 10
+        if lsb(cx16.r1) < 10
             c64.CHROUT('0')
-        txt.print_ub(cx16.r1L)
+        txt.print_ub(lsb(cx16.r1))
     }
 
     sub print_time() {
-        if cx16.r1H < 10
+        if msb(cx16.r1) < 10
            c64.CHROUT('0')
-        txt.print_ub(cx16.r1H)
+        txt.print_ub(msb(cx16.r1))
         c64.CHROUT(':')
-        if cx16.r2L < 10
+        if lsb(cx16.r2) < 10
             c64.CHROUT('0')
-        txt.print_ub(cx16.r2L)
+        txt.print_ub(lsb(cx16.r2))
         c64.CHROUT(':')
-        if cx16.r2H < 10
+        if msb(cx16.r2) < 10
             c64.CHROUT('0')
-        txt.print_ub(cx16.r2H)
+        txt.print_ub(msb(cx16.r2))
         c64.CHROUT('.')
-        if cx16.r3L < 10
+        if lsb(cx16.r3) < 10
             c64.CHROUT('0')
-        txt.print_ub(cx16.r3L)
+        txt.print_ub(lsb(cx16.r3))
     }
 }
 

examples/hello.p8

@@ -9,7 +9,7 @@ main {
     sub start() {
 
         ; set text color and activate lowercase charset
-        c64.COLOR = 13
+        txt.color(13)
         c64.VMCSB |= 2
 
         ; use optimized routine to write text

examples/line-circle-txt.p8

@@ -16,13 +16,13 @@ main {
         txt.print("enter for disc:")
         void c64.CHRIN()
         c64.CHROUT('\n')
-        txt.clear_screen(' ', 1)
+        txt.clear_screen()
         disc(20, 12, 12)
 
         txt.print("enter for lines:")
         void c64.CHRIN()
         c64.CHROUT('\n')
-        txt.clear_screen(' ', 1)
+        txt.clear_screen()
 
         line(1, 10, 38, 24)
         line(1, 20, 38, 2)
@@ -32,7 +32,7 @@ main {
         txt.print("enter for rectangles:")
         void c64.CHRIN()
         c64.CHROUT('\n')
-        txt.clear_screen(' ', 1)
+        txt.clear_screen()
 
         rect(4, 8, 37, 23, false)
         rect(20, 12, 30, 20, true)

examples/mandelbrot-gfx.p8

@@ -37,10 +37,10 @@ main {
                     iter++
                 }
 
-                if iter & 1 {
-                    graphics.plotx = pixelx
-                    graphics.plot(pixely)
-                }
+                if iter & 1
+                    ; TODO fix plot() so we don't have to use separate internal variable
+                    graphics.internal_plotx = pixelx
+                    graphics.internal_plot(pixely)
             }
         }
 

examples/plasma.p8

@@ -21,7 +21,7 @@ main {
     const ubyte PAGE2 = ((SCREEN2 >> 6) & $F0) | ((CHARSET >> 10) & $0E)
 
     sub start() {
-        c64.COLOR = 1
+        txt.color(1)
         txt.print("creating charset...\n")
         makechar()
 

examples/romfloats.p8

@@ -4,8 +4,47 @@
 main {
 
     sub start() {
+        print_rom_floats_values()
+       	c64.CHROUT('\n')
+        print_rom_floats()
+       	c64.CHROUT('\n')
+    }
 
+    sub print_rom_floats() {
+        c64flt.FL_PIVAL=9.9999
+    	c64flt.print_f(c64flt.FL_PIVAL)
+    	c64.CHROUT('\n')
+    	c64flt.print_f(c64flt.FL_N32768)
+    	c64.CHROUT('\n')
+    	c64flt.print_f(c64flt.FL_FONE)
+    	c64.CHROUT('\n')
+    	c64flt.print_f(c64flt.FL_SQRHLF)
+    	c64.CHROUT('\n')
+    	c64flt.print_f(c64flt.FL_SQRTWO)
+    	c64.CHROUT('\n')
+    	c64flt.print_f(c64flt.FL_NEGHLF)
+    	c64.CHROUT('\n')
+    	c64flt.print_f(c64flt.FL_LOG2)
+    	c64.CHROUT('\n')
+    	c64flt.print_f(c64flt.FL_TENC)
+    	c64.CHROUT('\n')
+    	c64flt.print_f(c64flt.FL_NZMIL)
+    	c64.CHROUT('\n')
+    	c64flt.print_f(c64flt.FL_FHALF)
+    	c64.CHROUT('\n')
+    	c64flt.print_f(c64flt.FL_LOGEB2)
+    	c64.CHROUT('\n')
+    	c64flt.print_f(c64flt.FL_PIHALF)
+    	c64.CHROUT('\n')
+    	c64flt.print_f(c64flt.FL_TWOPI)
+    	c64.CHROUT('\n')
+    	c64flt.print_f(c64flt.FL_FR4)
+    	c64.CHROUT('\n')
+    }
+
+    sub print_rom_floats_values() {
         ; these are all floating point constants defined in the ROM so no allocation required
+        ; the compiler recognises these and will substitute the ROM values automatically
 
         c64flt.print_f(3.141592653589793)
         c64.CHROUT('\n')

examples/swirl-float.p8

@@ -6,17 +6,21 @@ main {
     const uword width = 40
     const uword height = 25
 
-    sub start()  {
-
+    struct Ball {
         float t
         ubyte color
+    }
+
+    sub start()  {
+
+        Ball ball
 
         repeat {
-            ubyte xx=(sin(t) * width/2.2) + width/2.0 as ubyte
-            ubyte yy=(cos(t*1.1356) * height/2.2) + height/2.0 as ubyte
-            txt.setcc(xx, yy, 81, color)
-            t  += 0.08
-            color++
+            ubyte xx=(sin(ball.t) * width/2.1) + width/2.0 as ubyte
+            ubyte yy=(cos(ball.t*1.1356) * height/2.1) + height/2.0 as ubyte
+            txt.setcc(xx, yy, 81, ball.color)
+            ball.t  += 0.08
+            ball.color++
         }
     }
 }

examples/tehtriz.p8

@@ -280,12 +280,12 @@ waitkey:
 
     sub drawBoard() {
         c64.CLEARSCR()
-        c64.COLOR = 7
+        txt.color(7)
         txt.plot(1,1)
         txt.print("irmen's")
         txt.plot(2,2)
         txt.print("teh▁triz")
-        c64.COLOR = 5
+        txt.color(5)
         txt.plot(6,4)
         txt.print("hold:")
         txt.plot(2,22)
@@ -296,10 +296,10 @@ waitkey:
         txt.print("lines:")
         txt.plot(28,14)
         txt.print("score:")
-        c64.COLOR = 12
+        txt.color(12)
         txt.plot(27,18)
         txt.print("controls:")
-        c64.COLOR = 11
+        txt.color(11)
         txt.plot(28,19)
         txt.print(",/  move")
         txt.plot(28,20)
@@ -342,7 +342,7 @@ waitkey:
     }
 
     sub drawScore() {
-        c64.COLOR=1
+        txt.color(1)
         txt.plot(30,11)
         txt.print_uw(lines)
         txt.plot(30,15)

examples/test.p8

@@ -3,16 +3,13 @@
 
 main {
     sub start() {
-        ubyte b1 = 2
-        ubyte b2 = 13
-        ubyte b3 = 100
 
-        uword w1 = 2222
-        uword w2 = 11
-        uword w3 = 33
+;asmsub  clear_screen (ubyte char @ A, ubyte color @ Y) clobbers(A)  { ...}
+; TODO dont cause name conflict if we define sub or sub with param 'color' or even a var 'color' later.
+
+;   sub color(...) {}
+;   sub other(ubyte color) {}    ; TODO don't cause name conflict
 
-        w1 %= (w2+w3)
-        txt.print_uw(w1)
         c64.CHROUT('\n')
     }
 }