Working on new stencil blitter for IIgs.

.gitignore

@@ -11,3 +11,5 @@ notes.txt
 assets/JoeyLib Logo - 320x200.png
 assets/JoeyLib Logo.jpg
 assets/JoeyLib Logo.png
+*.stn
+*.img

imgconvert/main.c

@@ -97,7 +97,7 @@ void convertStencil(char *filename, char *basename) {
 	for (int y=0; y<image->h; y++) {
 		for (int x=0; x<image->w; x++) {
 			data <<= 1;
-			data |= (getPixel(image, x, y) == 0);
+			data |= (getPixel(image, x, y) != 0);
 			bits++;
 			if (bits > 7) {
 				bytes++;

joeylib/build.sh

@@ -99,8 +99,9 @@ function doIIgsBuild() {
 		rm "${IMPORT}" 2> /dev/null || true
 
 		cp "${OUT}/test" "${OUT}/Test#B30000"
-		cp "${JOEY}/joeylib/joeylib/src/kanga.sta" "${OUT}/kanga.sta#060000"
+		cp "${JOEY}/joeylib/joeylib/src/kanga.img" "${OUT}/kanga.img#060000"
-		cp "${JOEY}/joeylib/joeylib/src/font.sta" "${OUT}/font.sta#060000"
+		cp "${JOEY}/joeylib/joeylib/src/font.img" "${OUT}/font.img#060000"
+		cp "${JOEY}/joeylib/joeylib/src/font.stn" "${OUT}/font.stn#060000"
 		cp "${JOEY}/joeylib/joeylib/src/music" "${OUT}/music#D50000"
 		cp "${JOEY}/joeylib/joeylib/src/music.w" "${OUT}/music.w#060000"
 
@@ -108,8 +109,9 @@ function doIIgsBuild() {
 		"${CADIUS}" createfolder "${IMPORT}" ${VOL}/data > /dev/null
 		"${CADIUS}" addfile "${IMPORT}" ${VOL} "${JOEY}/dist/IIgs/Tool221#ba0000" > /dev/null
 		"${CADIUS}" addfile "${IMPORT}" ${VOL} "${OUT}/Test#b30000" > /dev/null
-		"${CADIUS}" addfile "${IMPORT}" ${VOL}/data "${OUT}/kanga.sta#060000" > /dev/null
+		"${CADIUS}" addfile "${IMPORT}" ${VOL}/data "${OUT}/kanga.img#060000" > /dev/null
-		"${CADIUS}" addfile "${IMPORT}" ${VOL}/data "${OUT}/font.sta#060000" > /dev/null
+		"${CADIUS}" addfile "${IMPORT}" ${VOL}/data "${OUT}/font.img#060000" > /dev/null
+		"${CADIUS}" addfile "${IMPORT}" ${VOL}/data "${OUT}/font.stn#060000" > /dev/null
 		"${CADIUS}" addfile "${IMPORT}" ${VOL}/data "${OUT}/music#D50000" > /dev/null
 		"${CADIUS}" addfile "${IMPORT}" ${VOL}/data "${OUT}/music.w#060000" > /dev/null
 

joeylib/src/jIIgs.asm

@@ -27,8 +27,8 @@
 
 
 GlobalData	data
-ScanTable	entry
+ScTable	ds	400	; 400 bytes for 200 scanline offsets
-	ds	400	; 400 bytes for scanline offsets
+StTable	ds	50	; 50 bytes for 25 stencil table offsets
 VblRate	ds	2	; Either 5 or 6 depending on PAL or NTSC.
 VblTime	ds	2	; Integer Counter
 	end
@@ -61,7 +61,7 @@ yc	equ	9	; Y Counter
 	asl	a	; y screen location is now in the accumulator
 	tay
 	clc
-	lda	ScanTable,y	; Load byte offset of y position from table
+	lda	ScTable,y	; Load byte offset of y position from table
 	adc	t	; Add t to scanline offset
 	sta	source	; Offset to start of source pixels
 
@@ -75,7 +75,7 @@ yc	equ	9	; Y Counter
 	lda	ty	; Load ty to get index into scanline table
 	asl	a	; Multiply by two to get offset into scanline table (two bytes per entry in table)
 	tay
-	lda	ScanTable,y	; Load byte offset of y position from table
+	lda	ScTable,y	; Load byte offset of y position from table
 	adc	tx	; Add x position to scanline offset
 	sta	target	; Offset to start of target pixels
 
@@ -128,7 +128,7 @@ blitTop	ldy	source	; Load Y register with source pixel offset
 ;----------------------------------------
 ; Blit an 8x8 block with alpha to (almost) anywhere on a surface.
 ;----------------------------------------
-asmB88a	start
+asmB88ax	start
 
 source	equ	1	; Source Pixels Offset
 target	equ	3	; Target Pixels Offset
@@ -153,7 +153,7 @@ yc	equ	11	; Y Counter
 	asl	a	; y screen location is now in the accumulator
 	tay
 	clc
-	lda	ScanTable,y	; Load byte offset of y position from table
+	lda	ScTable,y	; Load byte offset of y position from table
 	adc	t	; Add t to scanline offset
 	sta	source	; Offset to start of source pixels
 
@@ -175,7 +175,7 @@ yc	equ	11	; Y Counter
 	lda	ty	; Load ty to get index into scanline table
 	asl	a	; Multiply by two to get offset into scanline table (two bytes per entry in table)
 	tay
-	lda	ScanTable,y	; Load byte offset of y position from table
+	lda	ScTable,y	; Load byte offset of y position from table
 	adc	tx	; Add x position to scanline offset
 	sta	target	; Offset to start of target pixels
 
@@ -237,6 +237,150 @@ blitTop	ldy	target	; Load Y register with target pixel offset
 	end
 
 
+;----------------------------------------
+; Blit an 8x8 block with stencil to (almost) anywhere on a surface.
+;----------------------------------------
+asmB88a	start
+
+source	equ	1	; Source Pixels Offset
+target	equ	3	; Target Pixels Offset
+t	equ	5	; Temp
+xc	equ	7	; X Counter
+yc	equ	9	; Y Counter
+stenO	equ	11	; Stencil Buffer Offset
+stenB	equ	12	; Stencil Buffer Data - bits - two bytes, but we only use one
+maskW	equ	14	; Actual mask bytes for pixels
+
+	jsubroutine (4:surface,4:tiles,4:stencil,2:sx,2:sy,2:tx,2:ty),16
+	using GlobalData
+
+; Find offset into tile memory
+	lda	sx	; Multiply sx by 4 to get offset (two pixels per byte)
+	asl	a
+	asl	a
+	sta	t
+	clc
+	lda	sy	; Multiply sy by 16 to get index into scanline table
+	asl	a
+	asl	a
+	asl	a
+	asl	a	; y screen location is now in the accumulator
+	tay
+	clc
+	lda	ScTable,y	; Load byte offset of y position from table
+	adc	t	; Add t to scanline offset
+	sta	source	; Offset to start of source pixels
+
+; Divide target horizontal position by two since there are two pixels per byte
+	lda	tx	; Load X target position into accumulator
+	lsr	a	; Shift Right - divide by 2
+	clc
+	sta	tx	; Store it back into tx
+
+; Find offset into target surface memory
+	lda	ty	; Load ty to get index into scanline table
+	asl	a	; Multiply by two to get offset into scanline table (two bytes per entry in table)
+	tay
+	lda	ScTable,y	; Load byte offset of y position from table
+	adc	tx	; Add x position to scanline offset
+	sta	target	; Offset to start of target pixels
+
+; Find starting stencil byte
+	clc
+	ldy	ty	; Vertical cell position
+	lda	StTable,y	; Load offset location of start of this row in the stencil buffer
+	adc	tx	; Add x position to address offset
+	sta	stenO	; Offset to start of mask data
+
+
+	lda	#0	; Load 0 into X and Y counters
+	sta	xc
+	sta	yc
+
+blitTop	lda	[stenO]	; Load byte at stencil offset
+	xba		; Shift it into the high byte
+	sta	stenB	; Put it in to the stencil byte
+
+; Convert stencil byte to mask word (4 pixels)
+	ldx	#4	; How many pixels to process?
+	lda	#0	; Initially, the mask word is all zero bits
+	sta	maskW
+nextNib	lda	stenB	; Load stencil byte into accumulator
+	asl	a	; Shift first bit out of stencil byte into C
+	bcs	pixOn	; If C is set, we draw the pixel
+	lda	maskW	; Load mask word into accumulator
+	asl	a	; Shift mask word over a pixel (4 bits)
+	asl	a
+	asl	a
+	asl	a
+	and	#$fff0	; Do not draw pixel - mask it off
+	sta	maskW	; Store updated mask word
+	jmp 	nextPix	; Start next pixel
+pixOn	lda	maskW	; Load mask word into accumulator
+	asl	a	; Shift mask word over a pixel (4 bits)
+	asl	a
+	asl	a
+	asl	a
+	and	#$ffff	; Draw pixel
+	sta	maskW	; Store updated mask word
+nextPix	dex		; Decrement pixel counter in X
+	bne	nextNib	; Not zero?  Do next nibble/pixel
+
+; Draw four pixels
+	ldy	target	; Load Y register with target pixel offset
+	lda	[surface],y	; Load target pixels
+	and	maskW	; AND with mask word
+	ldy	source	; Load Y register with source pixel offset
+	ora	[tiles],y	; OR with source pixels
+	ldy	target	; Load Y register with target pixel offset
+	sta	[surface],y	; Store 4 pixels into screen
+
+	clc		; Increment to next pixel target quad
+	lda	#2
+	adc	target
+	sta	target
+	clc		; Increment to next pixel source quad
+	lda	#2
+	adc	source
+	sta	source
+	clc		; Increment to next stencil byte
+	lda	#1
+	adc	stenO
+	sta	stenO
+
+	clc
+	lda	xc	; Increment X counter
+	adc	#1
+	sta	xc
+	cmp	#2	; End of X pixels?
+	bcc	blitTop	; Nope!
+
+	lda	#0	; Reset X counter
+	sta	xc
+	clc		; Increment target offset
+	lda	#156
+	adc	target
+	sta	target
+	clc		; Increment source offset
+	lda	#156
+	adc	source
+	sta	source
+	clc		; Increment stencil mask offset
+	lda	#39
+	adc	stenO
+	sta	stenO
+	clc
+	lda	yc	; Increment Y counter
+	adc	#1
+	sta	yc
+	cmp	#8	; End of Y pixels?
+	bcs	blitDone	; Yep!
+	brl	blitTop	; Back for more
+
+blitDone	jreturn
+	end
+
+
 ;----------------------------------------
 ; Set Border Color
 ;----------------------------------------
@@ -879,7 +1023,7 @@ temp1	equ	1
 	lda	Yp	; Load accumulator with Y location
 	asl	a	; Shift accumulator left (multiply by 2) for word offset into table
 	tay		; Transfer accumulator to y register
-	lda	ScanTable,y	; Look up scan line address offset
+	lda	ScTable,y	; Look up scan line address offset
 	sta	temp1	; Store accumulator (row offset into SHR memory) into temp1
 	clc		; Clear carry flag
 	lda	Xp	; Place X position in accumulator
@@ -1100,7 +1244,7 @@ PBoth	sta	temp3	; Store accumulator in temp3
 	lda	Yp	; Load accumulator with Y location
 	asl	a	; Shift accumulator left (multiply by 2)
 	tay		; Transfer accumulator to y register
-	lda	ScanTable,y	; Look up scan line address offset
+	lda	ScTable,y	; Look up scan line address offset
 	adc	temp1	; Add X position to address with carry
 	tay		; Transfer accumulator to y
 
@@ -1348,13 +1492,24 @@ asmStart	start
 	ldy	#0	; Load 0 into y register
 	lda	#0	; Load 0 into accumulator
 	clc		; Clear carry flag
-buildTable	anop
+tScan	anop
-	sta	ScanTable,y	; Store accumulator in ScanTable+x
+	sta	ScTable,y	; Store accumulator in ScTable+x
-	adc	#160	; Add 160 to accumulator
+	adc	#160	; Add 160 to accumulator (320 pixels / 2 pixels per byte)
 	iny		; Increment y
 	iny		; Increment y
 	cpy	#400	; Compare y to 400 (200 scanlines)
-	bcc	buildTable	; Repeat Scan Table Loop until we do all 200 lines
+	bcc	tScan	; Repeat Scan Table Loop until we do all 200 lines
+
+	ldy	#0	; Load 0 into y register
+	lda	#0	; Load 0 into accumulator
+	clc		; Clear carry flag
+tSten	anop
+	sta	StTable,y	; Store accumulator in StencilTable+x
+	adc	#320	; Add 320 to accumulator (8 lines * 40 bytes)
+	iny		; Increment y
+	iny		; Increment y
+	cpy	#50	; Compare y to 50 (25 stencil cell offsets)
+	bcc	tSten	; Repeat Stencil Table Loop until we do all 25 cells
 
 	lda	hertz	; Store refresh rate / 10 in VblCount and VblRate
 	sta	>VblRate

joeylib/src/jIIgs.c

@@ -156,6 +156,29 @@ bool jlGameGetButton(byte which) {
 }
 
 
+bool _jlImgCreate(jlImgT **img) {
+	jlImgT *t = (jlImgT *)jlMalloc(sizeof(jlImgT));
+	if (t != NULL) {
+		memset(t, 0, sizeof(jlImgT));
+		t->id[0] = 'I';
+		t->id[1] = 'M';
+		t->id[2] = 'G';
+		t->version = 0;
+		memcpy(t->palette, SHRCOLORS, sizeof(t->palette));
+		memcpy(t->pixels, SHRPIXELS, sizeof(t->pixels));
+		*img = t;
+		return true;
+	}
+	return false;
+}
+
+
+void jlImgDisplay(jlImgT *img) {
+	memcpy(SHRCOLORS, img->palette, sizeof(img->palette));
+	memcpy(SHRPIXELS, img->pixels, sizeof(img->pixels));
+}
+
+
 char _jlKeyCheck(char key) {
 	if ((key < 8) || (key > 127)) {
 		key = 0;
@@ -206,8 +229,8 @@ void jlPaletteSet(byte index, byte r, byte g, byte b) {
 }
 
 
-void jlPaletteSetFromSta(jlStaT *sta) {
+void jlPaletteSetFromImg(jlImgT *img) {
-	memcpy(SHRCOLORS, sta->palette, sizeof(sta->palette));
+	memcpy(SHRCOLORS, img->palette, sizeof(img->palette));
 }
 
 
@@ -266,29 +289,6 @@ void jlSoundPlay(jlSoundT *sound) {
 }
 
 
-bool _jlStaCreate(jlStaT **sta) {
-	jlStaT *t = (jlStaT *)jlMalloc(sizeof(jlStaT));
-	if (t != NULL) {
-		memset(t, 0, sizeof(jlStaT));
-		t->id[0] = 'S';
-		t->id[1] = 'T';
-		t->id[2] = 'A';
-		t->version = 0;
-		memcpy(t->palette, SHRCOLORS, sizeof(t->palette));
-		memcpy(t->pixels, SHRPIXELS, sizeof(t->pixels));
-		*sta = t;
-		return true;
-	}
-	return false;
-}
-
-
-void jlStaDisplay(jlStaT *sta) {
-	memcpy(SHRCOLORS, sta->palette, sizeof(sta->palette));
-	memcpy(SHRPIXELS, sta->pixels, sizeof(sta->pixels));
-}
-
-
 void jlUtilIdle(void) {
 	// No need to pump the message loop on the IIgs
 }

joeylib/src/jSDL2.c

@@ -177,42 +177,6 @@ void jlDrawBlit8x8(jlSurfaceT source, jint16 sx, jint16 sy, jint16 tx, jint16 ty
 }
 
 
-/*
-void jlDrawBlit8x8a(jlSurfaceT source, jint16 sx, jint16 sy, jint16 tx, jint16 ty, byte offset) {
-	int           mo;  // Mask offset
-	int           so;  // Source Pixel Offset
-	int           to;  // Target Pixel Offset
-	int            x;
-	int            y;
-	jlPixelPairT   m;  // Mask
-	jlPixelPairT   s;  // Source Pixel
-	jlPixelPairT   t;  // Target Pixel
-	jlPixelPairT  *pixels = (jlPixelPairT *)source;
-	jlPixelPairT  *target = (jlPixelPairT *)_jlDrawTargetActual;
-
-	// We mask off unused bits in the source tile location so they can be used to hold other data.
-	mo = ((sy & 0x1f) * 8 * 160) + (((sx & 0x3f) + offset) * 4);  // This is in tiles
-	so = ((sy & 0x1f) * 8 * 160) + ((sx & 0x3f) * 4);  // This is in tiles
-	if (jlUtilIsOdd(tx)) tx--; // tx must be even because there are two pixels per byte
-	to = ty * 160 + (int)(tx * 0.5);  // This is in pixels...
-
-	for (y=0; y<8; y++) {
-		for (x=0; x<4; x++) {
-			m = pixels[mo++];
-			s = pixels[so++];
-			t = target[to];
-			t.r = (t.r & m.r) | s.r;
-			t.l = (t.l & m.l) | s.l;
-			target[to++] = t;
-		}
-		mo += 156;
-		so += 156;
-		to += 156;
-	}
-}
-*/
-
-
 void jlDrawBlit8x8a(jlSurfaceT source, jlStnT *stencil, jint16 sx, jint16 sy, jint16 tx, jint16 ty) {
 	int           mo;  // Mask offset
 	int           so;  // Source Pixel Offset
@@ -238,11 +202,11 @@ void jlDrawBlit8x8a(jlSurfaceT source, jlStnT *stencil, jint16 sx, jint16 sy, ji
 			s = pixels[so++];
 
 			//***FIX*** Another endian order issue.  Left & Right are swapped.
-			if ((stencil->pixels[mo] & (1 << b--)) == 0) {
+			if ((stencil->pixels[mo] & (1 << b--)) != 0) {
 				t.r = s.r;
 			}
 
-			if ((stencil->pixels[mo] & (1 << b--)) == 0) {
+			if ((stencil->pixels[mo] & (1 << b--)) != 0) {
 				t.l = s.l;
 			}
 

joeylib/src/joey.h

@@ -324,7 +324,7 @@ bool         _jlVecLoad(jlVecT **vec, char *filename);
 #ifdef JOEY_IIGS
 // Inlined functions - asm code
 extern void     asmB88(jlSurfaceT target, jlSurfaceT source, jint16 cx1, jint16 cy1, jint16 cx2, jint16 cy2);
-extern void     asmB88a(jlSurfaceT target, jlSurfaceT source, jint16 cx1, jint16 cy1, jint16 cx2, jint16 cy2, byte alphaOffset);
+extern void     asmB88a(jlSurfaceT target, jlSurfaceT source, byte *stencil, jint16 cx1, jint16 cy1, jint16 cx2, jint16 cy2);
 extern void     asmDrawBM(jlSurfaceT target, jint16 startX, jint16 startY, jint16 width, jint16 height, byte *mapData, juint16 stride, jlSurfaceT source);
 //extern void     asmDrawLine(jlSurfaceT target, jint16 color, jint16 x1, jint16 y1, jint16 x2, jint16 y2);
 extern jint16   asmGetPoint(jlSurfaceT target, jint16 x, jint16 y);
@@ -334,7 +334,7 @@ extern void     asmPoint(jlSurfaceT target, jint16 color, jint16 x, jint16 y);
 
 // Inlined functions
 #define jlDrawBlit8x8(source, cx1, cy1, cx2, cy2)                               asmB88(_jlDrawTargetActual, source, cx1, cy1, cx2, cy2)
-#define jlDrawBlit8x8a(source, cx1, cy1, cx2, cy2, offset)                      asmB88a(_jlDrawTargetActual, source, cx1, cy1, cx2, cy2, offset)
+#define jlDrawBlit8x8a(source, stencil, cx1, cy1, cx2, cy2)                     asmB88a(_jlDrawTargetActual, source, (byte *)stencil, cx1, cy1, cx2, cy2)
 #define jlDrawBlitMap(startX, startY, width, height, mapData, stride, source)   asmDrawBM(_jlDrawTargetActual, startX, startY, width, height, (byte *)mapData, stride, source)
 #define jlDrawPixelGet(x, y)                                                    asmGetPoint(_jlDrawTargetActual, x, y)
 #define jlDrawPixelSet(x, y)                                                    asmPoint(_jlDrawTargetActual, _jlDrawColorNibbles, x, y)

joeylib/src/test.c

@@ -41,8 +41,8 @@ extern void    asmDrawLine(jlSurfaceT target, jint16 color, jint16 x1, jint16 y1
 
 
 // Font hacking!
-__attribute__((__format__ (__printf__, 4, 0)))
+__attribute__((__format__ (__printf__, 5, 0)))
-void printAt(jlImgT *font, jint16 cx, jint16 cy, const char *what, ...) {
+void printAt(jlImgT *font, jlStnT *stencil, jint16 cx, jint16 cy, const char *what, ...) {
 	jint16  x;
 	jint16  y;
 	jint16  tx;
@@ -63,7 +63,11 @@ void printAt(jlImgT *font, jint16 cx, jint16 cy, const char *what, ...) {
 		//y = (msg[counter] - ' ') / 40;
 		x = msg[counter] % 40;
 		y = msg[counter] / 40;
+		if (stencil) {
+			jlDrawBlit8x8a(jlImgSurfaceGet(font), stencil, x, y, tx, ty);
+		} else {
 			jlDrawBlit8x8(jlImgSurfaceGet(font), x, y, tx, ty);
+		}
 		tx += 8;
 	}
 }
@@ -99,7 +103,7 @@ void lineTest(void) {
 	while (!jlKeyPressed()) {
 		y = 17;
 		//printAt(font, 1, y++, "R = %d   ", asmTest((jlSurfaceT)0x012000));
-		printAt(font, 1, y++, "Drawing %s    ", what);
+		printAt(font, NULL, 1, y++, "Drawing %s    ", what);
 
 		jlDrawColorSet((byte)color);
 		if (phase < 2) {
@@ -191,6 +195,12 @@ void blitTest(void) {
 	jlDrawClear();
 	jlPaletteSet(15, 15, 15, 15);
 
+	printAt(font, NULL, 1, 2, "%s", "Blitting without stencil buffer.");
+	printAt(font, stencil, 1, 4, "%s", "Blitting with stencil buffer.");
+	jlDisplayPresent();
+	jlKeyWaitForAny();
+
+	/*
 	y = 91;
 	while (!jlKeyPressed()) {
 		for (x=0; x<319-8; x++) {
@@ -202,7 +212,7 @@ void blitTest(void) {
 			}
 
 			jlDrawBlit8x8(jlImgSurfaceGet(font), 1, 0, x, y);
-			printAt(font, 1, 1, "Drawing at %d x %d   ", x, y);
+			printAt(font, NULL, 1, 1, "Drawing at %d x %d   ", x, y);
 			jlDisplayPresent();
 			jlDrawBlit8x8(jlImgSurfaceGet(font), 0, 0, x, y);
 			jlUtilSleep(1);
@@ -212,6 +222,7 @@ void blitTest(void) {
 		}
 	}
 	jlKeyRead();
+	*/
 
 	jlStnFree(stencil);
 	jlImgFree(font);