Merge pull request #3 from lscharen/expand-exception-handler-space

Expand exception handler space

demos/sprites/App.Main.s

@@ -36,10 +36,13 @@ DOWN_ARROW          equ        $0A
                     jsl        SetPalette
 
                     ldx        #5                        ; Mode 0 is full-screen, mode 5 is 256x160
+                    ldx        #320
+                    ldy        #144
                     jsl        SetScreenMode
 
 ; Set up our level data
                     jsr        BG0SetUp
+                    jsr        TileAnimInit
                     jsr        SetLimits
 
                     jsr        InitOverlay             ; Initialize the status bar
@@ -239,7 +242,11 @@ EvtLoop
                     ldy        PlayerY
                     jsl        UpdateSprite           ; Move the sprite to this local position
 
+; Update the timers
+                    jsl        DoTimers
+
 ; Let's see what it looks like!
+
                     lda        vsync
                     beq        :no_vsync
 :vsyncloop          jsl        GetVerticalCounter     ; 8-bit value
@@ -281,7 +288,7 @@ Exit
                     bcs        Fatal
 Fatal               brk        $00
 
-BG1DataFile         strl       '1/octane.c1'
+BG1DataFile         strl       '1/sunset.c1'
 
 ; Color palette
 ; MyPalette           dw         $068F,$0EDA,$0000,$0E51,$0BF1,$00A0,$0EEE,$0456,$0FA4,$0F59,$0E30,$01CE,$02E3,$0870,$0F93,$0FD7

demos/sprites/gen/App.TileMapBG0.s

@@ -208,7 +208,7 @@ App_TileMapBG0
         dw $1021,$1021,$1031,$1032,$1021,$1021,$100f,$0010,$0011,$1012,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$100f,$0010,$0011,$1012,$1021,$1021,$0007,$0008,$0007,$0008,$0007,$0008,$0007,$0008,$0007,$0008,$0007,$0008,$0007,$0008,$0007,$0008,$1021,$1021,$1021,$1021
         dw $1021,$1021,$1030,$0015,$0015,$0015,$0015,$1033,$1021,$1021,$1021,$1021,$1009,$100a,$1021,$1021,$1021,$1021,$1021,$1021,$001a,$001a,$001a,$001a,$0040,$0040,$001a,$001a,$001a,$001a,$1021,$1021
 
-        dw $1021,$1030,$0015,$0036,$0015,$0015,$0036,$0015,$1033,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1022,$1023,$1022,$1023,$1022,$1023,$1021,$1021,$1021,$1030,$0015,$0036,$1230,$1021,$1021,$1057,$1056,$1057,$1056,$1057,$1056,$1057,$1056,$1021,$1022,$1023,$1021,$1021,$1057,$1056,$1057,$1056,$1013,$0014,$0015,$1016,$1021,$1021,$1021,$1021
+        dw $1021,$1030,$0015,$0036,$0015,$0015,$0036,$0015,$1033,$1021,$1800,$1804,$0800,$0804,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1022,$1023,$1022,$1023,$1022,$1023,$1021,$1021,$1021,$1030,$0015,$0036,$1230,$1021,$1021,$1057,$1056,$1057,$1056,$1057,$1056,$1057,$1056,$1021,$1022,$1023,$1021,$1021,$1057,$1056,$1057,$1056,$1013,$0014,$0015,$1016,$1021,$1021,$1021,$1021
         dw $1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1013,$0014,$0015,$1016,$1021,$1021,$1021,$1021,$1022,$1023,$1022,$1023,$1021,$1021,$1021,$1021,$1013,$0014,$0015,$1016,$1021,$1030,$0015,$0036,$0015,$0015,$0036,$0015,$1033,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1013,$0014,$0015,$1016,$1021,$1021,$1022,$1023,$1022,$1023,$1022,$1023,$1021,$1021
         dw $1021,$1030,$0015,$0036,$1033,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1022,$1023,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1022,$1023,$1022,$1023,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021
         dw $1021,$1030,$0015,$0036,$0015,$0015,$0036,$0015,$1033,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1022,$1023,$1022,$1023,$1022,$1023,$1021,$1021,$1021,$1030,$0015,$0036,$1033,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1022,$1023,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021
@@ -216,7 +216,7 @@ App_TileMapBG0
         dw $1021,$1030,$0015,$0036,$1033,$1021,$1013,$0014,$0015,$1016,$1021,$1021,$1021,$1021,$0000,$1021,$1022,$1023,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1013,$0014,$0015,$1016,$0005,$0006,$0005,$0006,$0005,$0006,$0005,$0006,$0005,$0006,$0005,$0006,$0005,$0006,$0005,$0006,$0005,$0006,$1021,$1021,$1021,$1021
         dw $1021,$1030,$0015,$0036,$0015,$0015,$0036,$0015,$1033,$1021,$1021,$1021,$0005,$0006,$1021,$1021,$1021,$1021,$1021,$1021,$001a,$001a,$001a,$001a,$0040,$0040,$001a,$001a,$001a,$001a,$1021,$1021
 
-        dw $1030,$0015,$0015,$0015,$0015,$0015,$0015,$0015,$0015,$1033,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1024,$0025,$0025,$0025,$0025,$0025,$0025,$1026,$1021,$1030,$0015,$0015,$0015,$0015,$1230,$1021,$1077,$1076,$1077,$1076,$1077,$1076,$1077,$1076,$1024,$0025,$0025,$1026,$1021,$1077,$1076,$1077,$1076,$1013,$0014,$0015,$1016,$1021,$1021,$1021,$1021
+        dw $1030,$0015,$0015,$0015,$0015,$0015,$0015,$0015,$0015,$1033,$1808,$180C,$0808,$080C,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1024,$0025,$0025,$0025,$0025,$0025,$0025,$1026,$1021,$1030,$0015,$0015,$0015,$0015,$1230,$1021,$1077,$1076,$1077,$1076,$1077,$1076,$1077,$1076,$1024,$0025,$0025,$1026,$1021,$1077,$1076,$1077,$1076,$1013,$0014,$0015,$1016,$1021,$1021,$1021,$1021
         dw $1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1013,$0014,$0015,$1016,$1021,$1021,$1021,$1024,$0025,$0025,$0025,$0025,$1026,$1021,$1021,$1021,$1013,$0014,$0015,$1016,$1030,$0015,$0015,$0015,$0015,$0015,$0015,$0015,$0015,$1033,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1013,$0014,$0015,$1016,$1021,$1024,$0025,$0025,$0025,$0025,$0025,$0025,$1026,$1021
         dw $1030,$0015,$0015,$0015,$0015,$1033,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1024,$0025,$0025,$1026,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1024,$0025,$0025,$0025,$0025,$1026,$1021,$1021,$1021,$1021,$1021,$1021,$1021
         dw $1030,$0015,$0015,$0015,$0015,$0015,$0015,$0015,$0015,$1033,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1024,$0025,$0025,$0025,$0025,$0025,$0025,$1026,$1021,$1030,$0015,$0015,$0015,$0015,$1033,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1024,$0025,$0025,$1026,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021,$1021

macros/CORE.MACS.S

@@ -219,7 +219,7 @@ mixed           cmp   #$FFFF          ; All 1's in the mask is fully transparent
 ; This is the slowest path because there is a *lot* of work to do.  So much that it's
 ; worth it to change up the environment to optimize things a bit more.
 ;
-; Need to fill in the first 8 bytes of the JMP handler with the following code sequence
+; Need to fill in the first 10 bytes of the JMP handler with the following code sequence
 ;
 ;            lda  (00),y
 ;            and  #MASK
@@ -252,6 +252,9 @@ mixed           cmp   #$FFFF          ; All 1's in the mask is fully transparent
                 lda   ]1
                 sta:  $0006,y
 
+                lda   #$0D80          ; branch to the prologue (BRA *+15)
+                sta:  $0008,y
+
                 ldy   _Y_REG          ; restore original y-register value and move on
                 bra   next
 
@@ -321,6 +324,9 @@ mixed           cmp   #$FFFF          ; All 1's in the mask is fully transparent
                 ldal  ]1,x
                 sta:  $0006,y
 
+                lda   #$0D80          ; branch to the prologue (BRA *+15)
+                sta:  $0008,y
+
                 ldy   _Y_REG          ; restore original y-register value and move on
                 bra   next
 
@@ -333,103 +339,3 @@ transparent
                 sta:  ]3+1,y
 next
                 eom
-
-; Masked renderer for a dynamic tile. What's interesting about this renderer is that the mask
-; value is not used directly, but simply indicates if we can use a LDA 0,x / PHA sequence,
-; a LDA (00),y / PHA, or a JMP to a blended render
-;
-; If a dynamic tile is animated, there is the possibility to create a special mask that marks
-; words of the tile that a front / back / mixed across all frames.
-;
-; ]1 : tiledata offset
-; ]2 : tilemask offset
-; ]3 : code field offset
-CopyMaskedDWord MAC
-
-; Need to fill in the first 8 bytes of the JMP handler with the following code sequence
-;
-;            lda  (00),y
-;            and  $80,x
-;            ora  $00,x
-;            bcc  *+4
-
-                ldx   _X_REG          ; Get the addressing offset
-                ldal  JTableOffset,x  ; Get the address offset and add to the base address
-                adc   _BASE_ADDR      ; of the current code field line
-                adc   #{]1&$F000}     ; adjust for the current row offset
-                sta:  ]1+1,y
-
-                tax                   ; This becomes the new address that we use to patch in
-                lda   _X_REG          ; Get the offset and render a LDA (dp),y instruction
-
-                sep   #$20            ; Easier to do 8-bit operations
-                sta:  $0001,x         ; Set the LDA (00),y operand
-                lda   #$B1
-                sta:  $0000,x         ; Set the LDA (00),y opcode
-
-                lda   _T_PTR
-                sta:  $0005,x         ; Set ORA 00,x operand
-                ora   #$80
-                sta:  $0003,x         ; Set AND 00,x operand
-                lda   #$35
-                sta:  $0002,x         ; Set AND 00,x operand
-                lda   #$15
-                sta:  $0004,x         ; Set ORA 00,x operand
-                rep   #$30
-
-                lda   #$0290          ; BCC *+4
-                sta:  $0006,x
-                eom
-
-
-; Masked renderer for a dynamic tile with sprite data overlaid. What's interesting about this renderer is that the mask
-; value is not used directly, but simply indicates if we can use a LDA 0,x / PHA sequence,
-; a LDA (00),y / PHA, or a JMP to a blended render
-;
-; If a dynamic tile is animated, there is the possibility to create a special mask that marks
-; words of the tile that a front / back / mixed across all frames.
-;
-; ]1 : tiledata offset
-; ]2 : tilemask offset
-; ]3 : code field offset
-CopyMaskedDynSpriteWord MAC
-
-; Need to fill in the first 12(!!) bytes of the JMP handler with the following code sequence
-;
-;            lda  (00),y
-;            and  $80,x
-;            ora  $00,x
-;            and  #MASK
-;            ora  #DATA
-;
-; If MASK == 0, then we can do a PEA.  If MASK == $FFFF, then fall back to the simple Dynamic Masked
-; code.
-
-
-                ldx   _X_REG          ; Get the addressing offset
-                ldal  JTableOffset,x  ; Get the address offset and add to the base address
-                adc   _BASE_ADDR      ; of the current code field line
-                adc   #{]1&$F000}     ; adjust for the current row offset
-                sta:  ]1+1,y
-
-                tax                   ; This becomes the new address that we use to patch in
-                lda   _X_REG          ; Get the offset and render a LDA (dp),y instruction
-
-                sep   #$20            ; Easier to do 8-bit operations
-                sta:  $0001,x         ; Set the LDA (00),y operand
-                lda   #$B1
-                sta:  $0000,x         ; Set the LDA (00),y opcode
-
-                lda   _T_PTR
-                sta:  $0005,x         ; Set ORA 00,x operand
-                ora   #$80
-                sta:  $0003,x         ; Set AND 00,x operand
-                lda   #$35
-                sta:  $0002,x         ; Set AND 00,x operand
-                lda   #$15
-                sta:  $0004,x         ; Set ORA 00,x operand
-                rep   #$30
-
-                lda   #$0290          ; BCC *+4
-                sta:  $0006,x
-                eom

src/Core.s

@@ -8,7 +8,7 @@
                   use       .\Defs.s
 
 ; Feature flags
-NO_INTERRUPTS     equ       1                   ; turn off for crossrunner debugging
+NO_INTERRUPTS     equ       0                   ; turn off for crossrunner debugging
 NO_MUSIC          equ       1                   ; turn music + tool loading off
 
 ; External data provided by the main program segment
@@ -412,10 +412,15 @@ ReadControl       ENT
                   put       blitter/Tiles00011.s      ; dynamic masked tiles
 
                   put       blitter/Tiles10000.s      ; normal tiles + sprites
-;                  put       blitter/Tiles10001.s
+                  put       blitter/Tiles10001.s      ; dynamic tiles + sprites
                   put       blitter/Tiles10010.s      ; normal masked tiles + sprites
-;                  put       blitter/Tiles10011.s
+                  put       blitter/Tiles10011.s      ; dynamic masked tiles + sprites
-                  put       blitter/Tiles11000.s      ; normal high priority tile + sprites
+
+                  put       blitter/Tiles11000.s      ; normal high priority tiles + sprites
+                  put       blitter/Tiles11001.s      ; dynamic high priority tiles + sprites
+                  put       blitter/Tiles11010.s      ; normal high priority masked tiles + sprites
+                  put       blitter/Tiles11011.s      ; dynamic high priority masked tiles + sprites
+
                   put       blitter/TilesBG1.s
                   put       blitter/Vert.s
                   put       blitter/BG0.s

src/blitter/Template.s

@@ -37,8 +37,18 @@ PagePatches        da    {long_0-base+2}
                    da    {loop_back-base+2}
                    da    {loop_exit_3-base+2}
                    da    {even_exit-base+2}
+                   da    {jmp_rtn_1-base+2}
+                   da    {jmp_rtn_2-base+2}
+
+]index             equ   0
+                   lup   82                                 ; All the snippet addresses. The two JMP
+                   da    {snippets-base+{]index*32}+31}     ; instructino are at the end of each of
+                   da    {snippets-base+{]index*32}+28}     ; the 32-byte buffers
+]index             equ   ]index+1
+                   --^
 PagePatchNum       equ   *-PagePatches
 
+; Location that need a bank byte set for long addressing modes
 BankPatches        da    {long_0-base+3}
                    da    {long_1-base+3}
                    da    {long_2-base+3}
@@ -483,7 +493,9 @@ BuildBank
 ; 13 banks for a total of 208 lines, which is what is required to render 26 tiles
 ; to cover the full screen vertical scrolling.
 ;
-; The 'base' location is always assumed to be on a 4kb ($1000) boundary
+; The 'base' location is always assumed to be on a 4kb ($1000) boundary.  We make sure that
+; the code is assembled on a page boundary to help will alignment
+                   ds    \,$00                      ; pad to the next page boundary
 base
 entry_1            ldx   #0000                      ; Used for LDA 00,x addressing
 entry_2            ldy   #0000                      ; Used for LDA (00),y addressing
@@ -538,7 +550,7 @@ long_3             stal  *+5-base
 full_return        jml   blt_return                 ; Full exit
 
 ; Re-enable interrupts and continue -- the even_exit JMP from the previous line will jump here every
-; 8 or 16 lines in order to give the system some extra time to handle interrupts.
+; 8 or 16 lines in order to give the system time to handle interrupts.
 enable_int         ldal  stk_save+1                 ; restore the stack
                    tcs
                    sep   #$20                       ; 8-bit mode
@@ -554,13 +566,26 @@ enable_int         ldal  stk_save+1                 ; restore the stack
                    rep   #$20
                    bra   entry_1
 
+; The even/odd branch of this line's exception handler will return here.  This is mostly
+; a space-saving measure to allow for more code in the exeption handers themselved, but
+; also simplified the relocation process since we only have to update a single address
+; in each exception handler, rather than two.
+;
+; Oce working, this code should be able to be interleaved with the r_jmp_rtn code
+; above to eliminate a couple of branches
+jmp_rtn
+                   bvs   jmp_rtn_v                  ; overflow set means this is the right edge (entry)
+                   clc                              ; carry is set only for edge operations; force clear
+jmp_rtn_1          jmp   l_jmp_rtn-base
+jmp_rtn_v          rep   #$41                       ; clear V and C
+jmp_rtn_2          jmp   r_jmp_rtn-base
 
 ; This is the spot that needs to be page-aligned. In addition to simplifying the entry address
 ; and only needing to update a byte instad of a word, because the code breaks out of the
 ; code field with a BRA instruction, we keep everything within a page to avoid the 1-cycle
 ; page-crossing penalty of the branch.
 
-                   ds    166
+                   ds    \,$00                      ; pad to the next page boundary
 loop_exit_1        jmp   odd_exit-base              ; +0   Alternate exit point depending on whether the left edge is 
 loop_exit_2        jmp   even_exit-base             ; +3   odd-aligned
 
@@ -685,32 +710,25 @@ epilogue_1         tsc
 ;  r_edge    rep   #$41
 ;            brl  r_jmp_rtn     ; 3
 
-
+; Each snippet is provided 32 bytes of space.  The constant code is filled in from the end and
+; it is the responsibility of the code that fills in the hander to create valid program in the
+; first 23 bytes are available to be manipulated.
+;
+; Note that the code that's assembled in the first bytes of these snippets is just an example.  Every
+; routine that created an exception handler *MUST* write a full set of instructions since there is
+; no guarantee of what was written previously.
                    ds    \,$00                      ; pad to the next page boundary
 ]index             equ   0
 snippets           lup   82
-                   ds    2                          ; space for a 2-byte sequence; LDA (00),y  LDA 00,x  LDA 0,s
+                   ds    2                          ; space for all exception handlers
                    and   #$0000                     ; the mask operand will be set when the tile is drawn
                    ora   #$0000                     ; the data operand will be set when the tile is drawn
-                   bcs   *+6
+                   ds    15                         ; extra padding
-                   pha
+
-                   brl   loop+3+{3*]index}          ; use relative branch for convenience
+                   bcs   :byte                      ; if C = 0, just push the data and return
-                   bvs   *+6                        ; overflow set means this is the right edge (entry)
+                   pha                              ; 1 byte 
-                   clc                              ; carry is set only for edge operations; force clear
+                   jmp   loop+3+{3*]index}-base     ; 3 bytes : use relative branch for convenience
-                   brl   l_jmp_rtn
+:byte              jmp   jmp_rtn-base               ; 3 bytes
-                   rep   #$41                       ; clear V and C
-                   brl   r_jmp_rtn                  ; 25 bytes
-                   ds    7                          ; padding
 ]index             equ   ]index+1
                    --^
-top
+top
-
-
-
-
-
-
-
-
-
-

src/blitter/Tiles.s

@@ -48,6 +48,9 @@ _Y_REG           equ             tiletmp+2
 _T_PTR           equ             tiletmp+4                         ; Copy of the tile address pointer
 _BASE_ADDR       equ             tiletmp+6                         ; Copy of BTableLow for this tile
 _SPR_X_REG       equ             tiletmp+8                         ; Cache address of sprite plane source for a tile
+_JTBL_CACHE      equ             tiletmp+10                        ; Cache the offset to the exception handler for a column
+_OP_CACHE        equ             tiletmp+12                        ; Cache of a relevant operand / oeprator
+_TILE_ID         equ             tiletmp+14                        ; Copy of the tile descriptor
 
 ; Low-level function to take a tile descriptor and return the address in the tiledata
 ; bank.  This is not too useful in the fast-path because the fast-path does more
@@ -119,6 +122,7 @@ _RenderTile2
                  stx   _SPR_X_REG
 
 :nosprite
+                 sta   _TILE_ID                       ; Some tile blitters need to get the tile descriptor
                  and   #TILE_CTRL_MASK
                  xba
                  tax
@@ -179,12 +183,14 @@ TileProcs        dw     _TBSolidTile_00,_TBSolidTile_0H,_TBSolidTile_V0,_TBSolid
                  dw     _TBDynamicMaskTile_00,_TBDynamicMaskTile_00
 
 ; Here are all the sprite variants of the tiles
-                 dw     _TBSolidSpriteTile_00,_TBSolidSpriteTile_0H,
+                 dw     _TBSolidSpriteTile_00,_TBSolidSpriteTile_0H
                  dw     _TBSolidSpriteTile_V0,_TBSolidSpriteTile_VH                              ; 10000 : normal tiles w/sprite
-                 dw     _TBSolidTile_00,_TBSolidTile_0H,_TBSolidTile_V0,_TBSolidTile_VH          ; 10001 : dynamic tiles w/sprite
+                 dw     _TBDynamicSpriteTile_00,_TBDynamicSpriteTile_00
-                 dw     _TBMaskedSpriteTile_00,_TBMaskedSpriteTile_0H,
+                 dw     _TBDynamicSpriteTile_00,_TBDynamicSpriteTile_00                          ; 10001 : dynamic tiles w/sprite
+                 dw     _TBMaskedSpriteTile_00,_TBMaskedSpriteTile_0H
                  dw     _TBMaskedSpriteTile_V0,_TBMaskedSpriteTile_VH                            ; 10010 : masked normal tiles w/sprite
-                 dw     _TBSolidTile_00,_TBSolidTile_0H,_TBSolidTile_V0,_TBSolidTile_VH          ; 10011 : masked dynamic tiles w/sprite
+                 dw     _TBDynamicMaskedSpriteTile_00,_TBDynamicMaskedSpriteTile_00
+                 dw     _TBDynamicMaskedSpriteTile_00,_TBDynamicMaskedSpriteTile_00              ; 10011 : masked dynamic tiles w/sprite
 
                  dw     _TBSolidTile_00,_TBSolidTile_0H,_TBSolidTile_V0,_TBSolidTile_VH          ; 10100 : fringed normal tiles w/sprite
                  dw     _TBSolidTile_00,_TBSolidTile_0H,_TBSolidTile_V0,_TBSolidTile_VH          ; 10101 : fringed dynamic tiles w/sprite
@@ -193,9 +199,12 @@ TileProcs        dw     _TBSolidTile_00,_TBSolidTile_0H,_TBSolidTile_V0,_TBSolid
 
                  dw     _TBSolidPrioritySpriteTile_00,_TBSolidPrioritySpriteTile_0H,
                  dw     _TBSolidPrioritySpriteTile_V0,_TBSolidPrioritySpriteTile_VH              ; 11000 : high-priority normal tiles w/sprite
-                 dw     _TBSolidTile_00,_TBSolidTile_0H,_TBSolidTile_V0,_TBSolidTile_VH          ; 11001 : high-priority dynamic tiles w/sprite
+                 dw     _TBDynamicPrioritySpriteTile_00,_TBDynamicPrioritySpriteTile_00
-                 dw     _TBSolidTile_00,_TBSolidTile_0H,_TBSolidTile_V0,_TBSolidTile_VH          ; 11010 : high-priority masked normal tiles w/sprite
+                 dw     _TBDynamicPrioritySpriteTile_00,_TBDynamicPrioritySpriteTile_00          ; 11001 : high-priority dynamic tiles w/sprite
-                 dw     _TBSolidTile_00,_TBSolidTile_0H,_TBSolidTile_V0,_TBSolidTile_VH          ; 11011 : high-priority masked dynamic tiles w/sprite
+                 dw     _TBMaskedPrioritySpriteTile_00,_TBMaskedPrioritySpriteTile_0H
+                 dw     _TBMaskedPrioritySpriteTile_V0,_TBMaskedPrioritySpriteTile_VH            ; 11010 : high-priority masked normal tiles w/sprite
+                 dw     _TBDynamicMaskedPrioritySpriteTile_00,_TBDynamicMaskedPrioritySpriteTile_00
+                 dw     _TBDynamicMaskedPrioritySpriteTile_00,_TBDynamicMaskedPrioritySpriteTile_00 ; 11011 : high-priority masked dynamic tiles w/sprite
 
                  dw     _TBSolidTile_00,_TBSolidTile_0H,_TBSolidTile_V0,_TBSolidTile_VH          ; 11100 : high-priority fringed normal tiles w/sprite
                  dw     _TBSolidTile_00,_TBSolidTile_0H,_TBSolidTile_V0,_TBSolidTile_VH          ; 11101 : high-priority fringed dynamic tiles w/sprite

src/blitter/Tiles00001.s

@@ -17,11 +17,7 @@ _TBDynamicTile_00
 ; LDA 00,x / PHA where the operand is fixed when the tile is rendered
 ; $B5 $00 $48
 _TBDynamicData
-                 txa
+                 lda             _TILE_ID                          ; Get the original tile descriptor
-                 asl
-                 asl
-                 asl
-                 xba                                               ; Undo the x128 we just need x4
                  and             #$007F                            ; clamp to < (32 * 4)
                  ora             #$4800                            ; insert the PHA instruction
 

src/blitter/Tiles00011.s

@@ -8,9 +8,26 @@ _TBDynamicMaskTile_00
 
 ; A = dynamic tile id (must be <32)
 _TBDynamicDataAndMask
-                 and             #$007F                            ; clamp to < (32 * 4)
+                 sta             _X_REG               ; Cache some column values derived from _X_REG
-                 sta             _T_PTR
+                 tax
-                 stx             _X_REG
+                 ora             #$B100               ; Pre-calc the LDA (dp),y opcode + operand
+                 xba
+                 sta             _OP_CACHE
+
+                 clc
+                 ldal            JTableOffset,x       ; Get the address offset and add to the base address
+                 adc             _BASE_ADDR           ; of the current code field line
+                 sta             _JTBL_CACHE
+
+; We need to do an AND dp|$80,x / ORA dp,x.  The opcode values are $35 and $15, respectively.
+; We pre-calculate the AND opcode with the high bit of the operand set and then, in the macro
+; perform and EOR #$2080 to covert the opcode and operand in one instruction
+
+                 lda             _TILE_ID             ; Get the original tile descriptor
+                 and             #$007F               ; clamp to < (32 * 4)
+                 ora             #$3580               ; Pre-calc the AND $80,x opcode + operand
+                 xba
+                 sta             _T_PTR               ; This is an op to load the dynamic tile data
 
                  CopyMaskedDWord  $0003
                  CopyMaskedDWord  $1003
@@ -21,10 +38,18 @@ _TBDynamicDataAndMask
                  CopyMaskedDWord  $6003
                  CopyMaskedDWord  $7003
 
-                 inc             _T_PTR                            ; Move to the next column
+                 ldx             _X_REG
-                 inc             _T_PTR
+                 clc
-                 inc             _X_REG                            ; Move to the next column
+                 ldal            JTableOffset+2,x     ; Get the address offset and add to the base address
-                 inc             _X_REG
+                 adc             _BASE_ADDR           ; of the current code field line
+                 sta             _JTBL_CACHE
+
+                 lda             _OP_CACHE
+                 adc             #$0200
+                 sta             _OP_CACHE
+                 lda             _T_PTR
+                 adc             #$0200
+                 sta             _T_PTR
 
                  CopyMaskedDWord  $0000
                  CopyMaskedDWord  $1000
@@ -59,3 +84,39 @@ _TBFillJMPOpcode
                  sta             $7003,y
                  rep             #$20
                  rts
+
+
+; Masked renderer for a dynamic tile. What's interesting about this renderer is that the mask
+; value is not used directly, but simply indicates if we can use a LDA 0,x / PHA sequence,
+; a LDA (00),y / PHA, or a JMP to a blended render
+;
+; If a dynamic tile is animated, there is the possibility to create a special mask that marks
+; words of the tile that a front / back / mixed across all frames.
+;
+; ]1 : code field offset
+;
+; This macro does not set the opcode since they will all be JMP instructions, they can be 
+; filled more efficiently in a separate routine.
+CopyMaskedDWord MAC
+
+; Need to fill in the first 6 bytes of the JMP handler with the following code sequence
+;
+;            lda  (00),y
+;            and  $80,x
+;            ora  $00,x
+;            bra  *+17
+
+                lda   _JTBL_CACHE
+                ora   #{]1&$F000}     ; adjust for the current row offset
+                sta:  ]1+1,y
+
+                tax                   ; This becomes the new address that we use to patch in
+                lda   _OP_CACHE
+                sta:  $0000,x         ; LDA (00),y
+                lda   _T_PTR
+                sta:  $0002,x         ; AND $80,x
+                eor   #$8020          ; Switch the opcode to an ORA and remove the high bit of the operand
+                sta:  $0004,x         ; ORA $00,x
+                lda   #$0F80          ; branch to the prologue (BRA *+17)
+                sta:  $0006,x
+                eom

src/blitter/Tiles10001.s

@@ -0,0 +1,119 @@
+; _TBDynamicSpriteTile
+;
+; This tile type does not explicitly support horizontal or vertical flipping.  An appropriate tile
+; descriptor should be passed into CopyTileToDyn to put the horizontally or vertically flipped source
+; data into the dynamic tile buffer
+_TBDynamicSpriteTile_00
+                 sty             _Y_REG               ; This is restored in the macro
+
+                 sta             _X_REG               ; Cache some column values derived from _X_REG
+                 tax
+                 clc
+                 ldal            JTableOffset,x       ; Get the address offset and add to the base address
+                 adc             _BASE_ADDR           ; of the current code field line
+                 sta             _JTBL_CACHE
+
+                 lda             _TILE_ID             ; Get the original tile descriptor
+                 and             #$007F               ; clamp to < (32 * 4)
+                 ora             #$B500
+                 xba
+                 sta             _OP_CACHE            ; This is the 2-byte opcode for to load the data
+
+                 ldx             _SPR_X_REG
+
+                 CopyDynSpriteWord  {0*SPRITE_PLANE_SPAN};$0003
+                 CopyDynSpriteWord  {1*SPRITE_PLANE_SPAN};$1003
+                 CopyDynSpriteWord  {2*SPRITE_PLANE_SPAN};$2003
+                 CopyDynSpriteWord  {3*SPRITE_PLANE_SPAN};$3003
+                 CopyDynSpriteWord  {4*SPRITE_PLANE_SPAN};$4003
+                 CopyDynSpriteWord  {5*SPRITE_PLANE_SPAN};$5003
+                 CopyDynSpriteWord  {6*SPRITE_PLANE_SPAN};$6003
+                 CopyDynSpriteWord  {7*SPRITE_PLANE_SPAN};$7003
+
+                 ldx             _X_REG
+                 clc
+                 ldal            JTableOffset+2,x     ; Get the address offset and add to the base address
+                 adc             _BASE_ADDR           ; of the current code field line
+                 sta             _JTBL_CACHE
+
+                 lda             _OP_CACHE
+                 adc             #$0200
+                 sta             _OP_CACHE
+
+                 ldx             _SPR_X_REG
+
+                 CopyDynSpriteWord  {0*SPRITE_PLANE_SPAN}+2;$0000
+                 CopyDynSpriteWord  {1*SPRITE_PLANE_SPAN}+2;$1000
+                 CopyDynSpriteWord  {2*SPRITE_PLANE_SPAN}+2;$2000
+                 CopyDynSpriteWord  {3*SPRITE_PLANE_SPAN}+2;$3000
+                 CopyDynSpriteWord  {4*SPRITE_PLANE_SPAN}+2;$4000
+                 CopyDynSpriteWord  {5*SPRITE_PLANE_SPAN}+2;$5000
+                 CopyDynSpriteWord  {6*SPRITE_PLANE_SPAN}+2;$6000
+                 CopyDynSpriteWord  {7*SPRITE_PLANE_SPAN}+2;$7000
+
+                 rts
+
+
+; Masked renderer for a dynamic tile with sprite data overlaid.
+;
+; ]1 : sprite plane offset
+; ]2 : code field offset
+CopyDynSpriteWord MAC
+
+; Need to fill in the first 10 bytes of the JMP handler with the following code sequence where
+; the data and mask from from the sprite plane
+;
+;            lda  $00,x
+;            and  #MASK
+;            ora  #DATA
+;            bra  *+15
+;
+; If MASK == 0, then we can do a PEA.  If MASK == $FFFF, then fall back to the simple Dynamic Tile
+; code.
+                ldal  spritemask+]1,x            ; load the mask value
+                bne   mixed                      ; a non-zero value may be mixed
+
+; This is a solid word
+                lda   #$00F4          ; PEA instruction
+                sta:  ]2,y
+                ldal  spritedata+]1,x ; load the sprite data
+                sta:  ]2+1,y          ; PEA operand
+                bra   next
+
+mixed           cmp   #$FFFF          ; All 1's in the mask is a fully transparent sprite word
+                beq   transparent
+
+                lda   #$004C          ; JMP to handler
+                sta:  ]2,y
+                lda   _JTBL_CACHE     ; Get the offset to the exception handler for this column
+                ora   #{]2&$F000}     ; adjust for the current row offset
+                sta:  ]2+1,y
+                tay                   ; This becomes the new address that we use to patch in
+
+                lda   _OP_CACHE       ; Get the LDA dp,x instruction for this column
+                sta:  $0000,y
+
+                lda   #$0029          ; AND #SPRITE_MASK
+                sta:  $0002,y
+                ldal  spritemask+]1,x 
+                sta:  $0003,y
+
+                lda   #$0009          ; ORA #SPRITE_DATA
+                sta:  $0005,y
+                ldal  spritedata+]1,x
+                sta:  $0006,y
+
+                lda   #$0D80          ; branch to the prologue (BRA *+15)
+                sta:  $0008,y
+
+                ldy   _Y_REG          ; restore original y-register value and move on
+                bra   next
+
+; This is a transparent word, so just show the dynamic data
+transparent
+                lda   #$4800          ; Put the PHA in the third byte
+                sta:  ]2+1,y
+                lda   _OP_CACHE       ; Store the LDA dp,x instruction with operand
+                sta:  ]2,y
+next
+                eom

src/blitter/Tiles10010.s

@@ -4,7 +4,6 @@
 _TBMaskedSpriteTile_00
                  sta             _X_REG                       ; Immedately stash the parameters
                  sty             _Y_REG
-;                 stx             _T_PTR
 
                  jsr             _TBCopyTileDataToCBuff       ; Copy the tile data into the compositing buffer (using correct x-register)
                  jsr             _TBCopyTileMaskToCBuff       ; Copy the tile mask into the compositing buffer (using correct x-register)

src/blitter/Tiles10011.s

@@ -0,0 +1,134 @@
+; _TBDynamicMaskedSpriteTile
+;
+; This tile type does not explicitly support horizontal or vertical flipping.  An appropriate tile
+; descriptor should be passed into CopyTileToDyn to put the horizontally or vertically flipped source
+; data into the dynamic tile buffer
+;
+; When rendering, the background, via lda (dp),y, is shown behind the animate sprite
+_TBDynamicMaskedSpriteTile_00
+                 sty             _Y_REG               ; This is restored in the macro
+
+                 sta             _X_REG               ; Cache some column values derived from _X_REG
+                 tax
+                 ora             #$B100               ; Pre-calc the LDA (dp),y opcode + operand
+                 xba
+                 sta             _OP_CACHE
+
+                 clc
+                 ldal            JTableOffset,x       ; Get the address offset and add to the base address
+                 adc             _BASE_ADDR           ; of the current code field line
+                 sta             _JTBL_CACHE
+
+                 lda             _TILE_ID             ; Get the original tile descriptor
+                 and             #$007F               ; clamp to < (32 * 4)
+                 ora             #$3580               ; Pre-calc the AND $80,x opcode + operand
+                 xba
+                 sta             _T_PTR               ; This is an op to load the dynamic tile data
+
+                 ldx             _SPR_X_REG
+
+                 CopyDynMaskedSpriteWord  {0*SPRITE_PLANE_SPAN};$0003
+                 CopyDynMaskedSpriteWord  {1*SPRITE_PLANE_SPAN};$1003
+                 CopyDynMaskedSpriteWord  {2*SPRITE_PLANE_SPAN};$2003
+                 CopyDynMaskedSpriteWord  {3*SPRITE_PLANE_SPAN};$3003
+                 CopyDynMaskedSpriteWord  {4*SPRITE_PLANE_SPAN};$4003
+                 CopyDynMaskedSpriteWord  {5*SPRITE_PLANE_SPAN};$5003
+                 CopyDynMaskedSpriteWord  {6*SPRITE_PLANE_SPAN};$6003
+                 CopyDynMaskedSpriteWord  {7*SPRITE_PLANE_SPAN};$7003
+
+                 ldx             _X_REG
+                 clc
+                 ldal            JTableOffset+2,x     ; Get the address offset and add to the base address
+                 adc             _BASE_ADDR           ; of the current code field line
+                 sta             _JTBL_CACHE
+
+                 lda             _OP_CACHE
+                 adc             #$0200
+                 sta             _OP_CACHE
+                 lda             _T_PTR
+                 adc             #$0200
+                 sta             _T_PTR
+
+                 ldx             _SPR_X_REG
+
+                 CopyDynMaskedSpriteWord  {0*SPRITE_PLANE_SPAN}+2;$0000
+                 CopyDynMaskedSpriteWord  {1*SPRITE_PLANE_SPAN}+2;$1000
+                 CopyDynMaskedSpriteWord  {2*SPRITE_PLANE_SPAN}+2;$2000
+                 CopyDynMaskedSpriteWord  {3*SPRITE_PLANE_SPAN}+2;$3000
+                 CopyDynMaskedSpriteWord  {4*SPRITE_PLANE_SPAN}+2;$4000
+                 CopyDynMaskedSpriteWord  {5*SPRITE_PLANE_SPAN}+2;$5000
+                 CopyDynMaskedSpriteWord  {6*SPRITE_PLANE_SPAN}+2;$6000
+                 CopyDynMaskedSpriteWord  {7*SPRITE_PLANE_SPAN}+2;$7000
+
+                 rts
+
+
+; Masked renderer for a masked dynamic tile with sprite data overlaid.
+;
+; ]1 : sprite plane offset
+; ]2 : code field offset
+CopyDynMaskedSpriteWord MAC
+
+; Need to fill in the first 14 bytes of the JMP handler with the following code sequence where
+; the data and mask from from the sprite plane
+;
+;            lda  ($00),y
+;            and  $80,x
+;            ora  $00,x
+;            and  #MASK
+;            ora  #DATA
+;            bra  *+15
+;
+; If MASK == 0, then we can do a PEA.  If MASK == $FFFF, then fall back to the simple Dynamic Tile
+; code and eliminate the constanct AND/ORA instructions.
+
+                ldal  spritemask+]1,x            ; load the mask value
+                bne   mixed                      ; a non-zero value may be mixed
+
+; This is a solid word
+                lda   #$00F4          ; PEA instruction
+                sta:  ]2,y
+                ldal  spritedata+]1,x ; load the sprite data
+                sta:  ]2+1,y          ; PEA operand
+                bra   next
+
+; We will always do a JMP to the eception handler, so set that up, then check for sprite
+; transparency
+mixed
+                lda   #$004C          ; JMP to handler
+                sta:  ]2,y
+                lda   _JTBL_CACHE     ; Get the offset to the exception handler for this column
+                ora   #{]2&$F000}     ; adjust for the current row offset
+                sta:  ]2+1,y
+                tay                   ; This becomes the new address that we use to patch in
+
+                lda   _OP_CACHE
+                sta:  $0000,y         ; LDA (00),y
+                lda   _T_PTR
+                sta:  $0002,y         ; AND $80,x
+                eor   #$8020          ; Switch the opcode to an ORA and remove the high bit of the operand
+                sta:  $0004,y         ; ORA $00,x
+
+                lda   #$0029          ; AND #SPRITE_MASK
+                sta:  $0006,y
+                ldal  spritemask+]1,x 
+                cmp   #$FFFF          ; All 1's in the mask is a fully transparent sprite word
+                beq   transparent     ; so we can use the Tile00011 method
+                sta:  $0007,y
+
+                lda   #$0009          ; ORA #SPRITE_DATA
+                sta:  $0009,y
+                ldal  spritedata+]1,x
+                sta:  $000A,y
+
+                lda   #$0980          ; branch to the prologue (BRA *+11)
+                sta:  $000C,y
+                bra   next
+
+; This is a transparent word, so just show the dynamic data
+transparent
+                lda   #$0F80          ; branch to the epilogue (BRA *+17)
+                sta:  $0006,y
+next
+                ldy   _Y_REG          ; restore original y-register value and move on
+                eom

src/blitter/Tiles11001.s

@@ -0,0 +1,92 @@
+; _TBDynamicPrioritySpriteTile
+;
+; Variant of _TBDynamicSpriteTile (Tile10001), but draw the sprite data behind the dynamic tile 
+_TBDynamicPrioritySpriteTile_00
+                jsr      _TBDynamicPriorityDataAndMask
+                jmp      _TBFillJMPOpcode
+
+_TBDynamicPriorityDataAndMask
+                 sty             _Y_REG               ; This is restored in the macro
+
+                 sta             _X_REG               ; Cache some column values derived from _X_REG
+                 tax
+                 clc
+                 ldal            JTableOffset,x       ; Get the address offset and add to the base address
+                 adc             _BASE_ADDR           ; of the current code field line
+                 sta             _JTBL_CACHE
+
+                 lda             _TILE_ID             ; Get the original tile descriptor
+                 and             #$007F               ; clamp to < (32 * 4)
+                 ora             #$3580               ; Pre-calc the AND $80,x opcode + operand
+                 xba
+                 sta             _OP_CACHE            ; This is an op to load the dynamic tile data
+
+                 ldx             _SPR_X_REG
+
+                 CopyDynPriSpriteWord  {0*SPRITE_PLANE_SPAN};$0003
+                 CopyDynPriSpriteWord  {1*SPRITE_PLANE_SPAN};$1003
+                 CopyDynPriSpriteWord  {2*SPRITE_PLANE_SPAN};$2003
+                 CopyDynPriSpriteWord  {3*SPRITE_PLANE_SPAN};$3003
+                 CopyDynPriSpriteWord  {4*SPRITE_PLANE_SPAN};$4003
+                 CopyDynPriSpriteWord  {5*SPRITE_PLANE_SPAN};$5003
+                 CopyDynPriSpriteWord  {6*SPRITE_PLANE_SPAN};$6003
+                 CopyDynPriSpriteWord  {7*SPRITE_PLANE_SPAN};$7003
+
+                 ldx             _X_REG
+                 clc
+                 ldal            JTableOffset+2,x     ; Get the address offset and add to the base address
+                 adc             _BASE_ADDR           ; of the current code field line
+                 sta             _JTBL_CACHE
+
+                 lda             _OP_CACHE
+                 adc             #$0200
+                 sta             _OP_CACHE
+
+                 ldx             _SPR_X_REG
+
+                 CopyDynPriSpriteWord  {0*SPRITE_PLANE_SPAN}+2;$0000
+                 CopyDynPriSpriteWord  {1*SPRITE_PLANE_SPAN}+2;$1000
+                 CopyDynPriSpriteWord  {2*SPRITE_PLANE_SPAN}+2;$2000
+                 CopyDynPriSpriteWord  {3*SPRITE_PLANE_SPAN}+2;$3000
+                 CopyDynPriSpriteWord  {4*SPRITE_PLANE_SPAN}+2;$4000
+                 CopyDynPriSpriteWord  {5*SPRITE_PLANE_SPAN}+2;$5000
+                 CopyDynPriSpriteWord  {6*SPRITE_PLANE_SPAN}+2;$6000
+                 CopyDynPriSpriteWord  {7*SPRITE_PLANE_SPAN}+2;$7000
+
+                 rts
+
+
+; Masked renderer for a dynamic tile with sprite data overlaid.
+;
+; ]1 : sprite plane offset
+; ]2 : code field offset
+CopyDynPriSpriteWord MAC
+
+; Need to fill in the first 9 bytes of the JMP handler with the following code sequence where
+; the data and mask from from the sprite plane
+;
+;            lda  #DATA
+;            and  $80,x
+;            ora  $00,x
+;            bra  *+16
+
+                lda   _JTBL_CACHE     ; Get the offset to the exception handler for this column
+                ora   #{]2&$F000}     ; adjust for the current row offset
+                sta:  ]2+1,y
+                tay                   ; This becomes the new address that we use to patch in
+
+                lda   #$00A9          ; LDA #DATA
+                sta:  $0000,y
+                ldal  spritedata+]1,x
+                sta:  $0001,y
+
+                lda   _OP_CACHE
+                sta:  $0003,y         ; AND $80,x
+                eor   #$8020          ; Switch the opcode to an ORA and remove the high bit of the operand
+                sta:  $0005,y         ; ORA $00,x
+
+                lda   #$0E80          ; branch to the prologue (BRA *+16)
+                sta:  $0007,y
+
+                ldy   _Y_REG          ; restore original y-register value and move on
+                eom

src/blitter/Tiles11010.s

@@ -0,0 +1,61 @@
+; _TBMaskedPrioritySpriteTile
+;
+; Renders a composited tile with masking to the code field.  The sprite is underlaid
+_TBMaskedPrioritySpriteTile_00
+                 sta             _X_REG                       ; Immedately stash the parameters
+                 sty             _Y_REG
+
+                 jsr             _TBCopyTileDataToCBuff       ; Copy the tile data into the compositing buffer (using correct x-register)
+                 jsr             _TBCopyTileMaskToCBuff       ; Copy the tile mask into the compositing buffer (using correct x-register)
+                 jsr             _TBUnderlaySpriteDataAndMask ; Underlay the data and mask from the sprite plane into the compositing buffer
+                 jmp             _TBMaskedCBuff               ; Render the masked tile from the compositing buffer into the code field
+
+_TBMaskedPrioritySpriteTile_0H
+                 sta             _X_REG
+                 sty             _Y_REG
+                 jsr             _TBCopyTileDataToCBuffH
+                 jsr             _TBCopyTileMaskToCBuffH
+                 jsr             _TBUnderlaySpriteDataAndMask
+                 jmp             _TBMaskedCBuff
+
+_TBMaskedPrioritySpriteTile_V0
+                 sta             _X_REG
+                 sty             _Y_REG
+                 jsr             _TBCopyTileDataToCBuffV
+                 jsr             _TBCopyTileMaskToCBuffV
+                 jsr             _TBUnderlaySpriteDataAndMask
+                 jmp             _TBMaskedCBuff
+
+_TBMaskedPrioritySpriteTile_VH
+                 sta             _X_REG
+                 sty             _Y_REG
+                 jsr             _TBCopyTileDataToCBuffVH
+                 jsr             _TBCopyTileMaskToCBuffVH
+                 jsr             _TBUnderlaySpriteDataAndMask
+                 jmp             _TBMaskedCBuff
+
+_TBUnderlaySpriteDataAndMask
+                 ldx   _SPR_X_REG                               ; set to the unaligned tile block address in the sprite plane
+
+]line            equ   0
+                 lup   8
+                 ldal  spritedata+{]line*SPRITE_PLANE_SPAN},x
+                 and   blttmp+{]line*4}+32
+                 ora   blttmp+{]line*4}                         ; Maybe this can be a TSB???
+                 sta   blttmp+{]line*4}
+
+                 ldal  spritemask+{]line*SPRITE_PLANE_SPAN},x
+                 and   blttmp+{]line*4}+32
+                 sta   blttmp+{]line*4}+32
+
+                 ldal  spritedata+{]line*SPRITE_PLANE_SPAN}+2,x
+                 and   blttmp+{]line*4}+32+2
+                 ora   blttmp+{]line*4}+2
+                 sta   blttmp+{]line*4}+2
+
+                 ldal  spritemask+{]line*SPRITE_PLANE_SPAN}+2,x
+                 and   blttmp+{]line*4}+32+2
+                 sta   blttmp+{]line*4}+32+2
+]line            equ   ]line+1
+                 --^
+                 rts 

src/blitter/Tiles11011.s

@@ -0,0 +1,123 @@
+; _TBDynamicMaskedPrioritySpriteTile
+;
+; This tile type does not explicitly support horizontal or vertical flipping.  An appropriate tile
+; descriptor should be passed into CopyTileToDyn to put the horizontally or vertically flipped source
+; data into the dynamic tile buffer
+_TBDynamicMaskedPrioritySpriteTile_00
+                 sty             _Y_REG               ; This is restored in the macro
+
+                 sta             _X_REG               ; Cache some column values derived from _X_REG
+                 tax
+                 ora             #$B100               ; Pre-calc the LDA (dp),y opcode + operand
+                 xba
+                 sta             _OP_CACHE
+
+                 clc
+                 ldal            JTableOffset,x       ; Get the address offset and add to the base address
+                 adc             _BASE_ADDR           ; of the current code field line
+                 sta             _JTBL_CACHE
+
+                 lda             _TILE_ID             ; Get the original tile descriptor
+                 and             #$007F               ; clamp to < (32 * 4)
+                 ora             #$3580               ; Pre-calc the AND $80,x opcode + operand
+                 xba
+                 sta             _T_PTR               ; This is an op to load the dynamic tile data
+
+                 ldx             _SPR_X_REG
+
+                 CopyDynPrioMaskedSpriteWord  {0*SPRITE_PLANE_SPAN};$0003
+                 CopyDynPrioMaskedSpriteWord  {1*SPRITE_PLANE_SPAN};$1003
+                 CopyDynPrioMaskedSpriteWord  {2*SPRITE_PLANE_SPAN};$2003
+                 CopyDynPrioMaskedSpriteWord  {3*SPRITE_PLANE_SPAN};$3003
+                 CopyDynPrioMaskedSpriteWord  {4*SPRITE_PLANE_SPAN};$4003
+                 CopyDynPrioMaskedSpriteWord  {5*SPRITE_PLANE_SPAN};$5003
+                 CopyDynPrioMaskedSpriteWord  {6*SPRITE_PLANE_SPAN};$6003
+                 CopyDynPrioMaskedSpriteWord  {7*SPRITE_PLANE_SPAN};$7003
+
+                 ldx             _X_REG
+                 clc
+                 ldal            JTableOffset+2,x     ; Get the address offset and add to the base address
+                 adc             _BASE_ADDR           ; of the current code field line
+                 sta             _JTBL_CACHE
+
+                 lda             _OP_CACHE
+                 adc             #$0200
+                 sta             _OP_CACHE
+                 lda             _T_PTR
+                 adc             #$0200
+                 sta             _T_PTR
+
+                 ldx             _SPR_X_REG
+
+                 CopyDynPrioMaskedSpriteWord  {0*SPRITE_PLANE_SPAN}+2;$0000
+                 CopyDynPrioMaskedSpriteWord  {1*SPRITE_PLANE_SPAN}+2;$1000
+                 CopyDynPrioMaskedSpriteWord  {2*SPRITE_PLANE_SPAN}+2;$2000
+                 CopyDynPrioMaskedSpriteWord  {3*SPRITE_PLANE_SPAN}+2;$3000
+                 CopyDynPrioMaskedSpriteWord  {4*SPRITE_PLANE_SPAN}+2;$4000
+                 CopyDynPrioMaskedSpriteWord  {5*SPRITE_PLANE_SPAN}+2;$5000
+                 CopyDynPrioMaskedSpriteWord  {6*SPRITE_PLANE_SPAN}+2;$6000
+                 CopyDynPrioMaskedSpriteWord  {7*SPRITE_PLANE_SPAN}+2;$7000
+
+                 rts
+
+
+; Masked renderer for a masked dynamic tile with sprite data underlaid.
+;
+; ]1 : sprite plane offset
+; ]2 : code field offset
+CopyDynPrioMaskedSpriteWord MAC
+
+; Need to fill in the first 14 bytes of the JMP handler with the following code sequence where
+; the data and mask from from the sprite plane
+;
+;            lda  ($00),y
+;            and  #MASK
+;            ora  #DATA
+;            and  $80,x
+;            ora  $00,x
+;            bra  *+15
+
+                lda   #$004C          ; JMP to handler
+                sta:  ]2,y
+                lda   _JTBL_CACHE     ; Get the offset to the exception handler for this column
+                ora   #{]2&$F000}     ; adjust for the current row offset
+                sta:  ]2+1,y
+                tay                   ; This becomes the new address that we use to patch in
+
+                lda   _OP_CACHE
+                sta:  $0000,y         ; LDA (00),y
+
+                lda   #$0029          ; AND #SPRITE_MASK
+                sta:  $0002,y
+
+                ldal  spritemask+]1,x
+                cmp   #$FFFF          ; All 1's in the mask is a fully transparent sprite word
+                beq   transparent     ; so we can use the Tile00011 method
+                sta:  $0003,y
+
+                lda   #$0009          ; ORA #SPRITE_DATA
+                sta:  $0005,y
+                ldal  spritedata+]1,x
+                sta:  $0006,y
+
+                lda   _T_PTR
+                sta:  $0008,y         ; AND $80,x
+                eor   #$8020          ; Switch the opcode to an ORA and remove the high bit of the operand
+                sta:  $000A,y         ; ORA $00,x
+
+                lda   #$0980          ; branch to the prologue (BRA *+11)
+                sta:  $000C,y
+                bra   next
+
+; This is a transparent word, so just show the dynamic data
+transparent
+                lda   _T_PTR
+                sta:  $0002,y         ; AND $80,x
+                eor   #$8020          ; Switch the opcode to an ORA and remove the high bit of the operand
+                sta:  $0004,y         ; ORA $00,x
+
+                lda   #$0F80          ; branch to the epilogue (BRA *+17)
+                sta:  $0006,y
+next
+                ldy   _Y_REG          ; restore original y-register value and move on
+                eom

tools/png2iigs.js

@@ -308,11 +308,6 @@ function buildTile(options, buff, _mask, width, x, y) {
         if (mask.some(h => h != 0)) {
             tile.isSolid = false;
         }
-
-        if (x === 120 && y === 8) {
-            console.warn(`isSolid: ${tile.isSolid}` );
-            console.warn(data.map(d => d.toString(16)), mask);
-        }
     }
 
     for (dy = 0; dy < 8; dy += 1) {

tools/tiled2iigs.js

@@ -460,6 +460,7 @@ function convertTileID(tileId, tileset) {
     }
     const mask_bit = (!tileset[tileIndex - 1].isSolid || tileIndex === GLOBALS.emptyTile) && ((GLOBALS.tileLayers.length !== 1) || GLOBALS.forceMasked);
 
+    /*
     if (tileIndex === 48) {
         console.warn('isSolid: ', tileset[tileIndex - 1].isSolid);
         console.warn('GLOBALS.emptyTile: ', GLOBALS.emptyTile);
@@ -467,6 +468,7 @@ function convertTileID(tileId, tileset) {
         console.warn('GLOBALS.forceMasked: ', GLOBALS.forceMasked);
         console.warn('mask_bit: ', mask_bit);
     }
+    */
 
     // Build up a partial set of control bits
     let control_bits = (mask_bit ? GTE_MASK_BIT : 0) + (hflip ? GTE_HFLIP_BIT : 0) + (vflip ? GTE_VFLIP_BIT : 0);
@@ -477,6 +479,11 @@ function convertTileID(tileId, tileset) {
         const animation = tileset[tileIndex - 1].animation;
         tileId = animation.dynTileId;
         control_bits = GTE_DYN_BIT;
+
+        console.warn('Dyanmic animation tile found!');
+        console.warn('isSolid: ', tileset[tileIndex - 1].isSolid);
+        console.warn('dynTileId: ',  animation.dynTileId);
+        console.warn('mask_bit: ', mask_bit);
     }
 
     return (tileId & 0x1FFFFFFF) + control_bits;