.text b initc2p # offset 0: initialization routines b doc2p_1X8 # offset 4: one pixel wide version, AGA b doc2p_1X6 # offset 8: one pixel wide version, EHB b doc2p_2X8 # offset 12: two pixel wide version, AGA b doc2p_2X6 # offset 16: two pixel wide version, EHB .word string # offset 20: address of descr string .byte "$VER: PowerPC 1.0 (16.07.98)",13,10,0 string: .byte "A chunky to planar routine by Peter McGavin. " .byte "Changes By Iain Barclay. " .byte "Ported to PowerPC by Frank Wille. " .byte "Supports 6/8 bitplane, single/double width pixels.",0 .align 4 .global initc2p initc2p: # create 2 tables for Gloom # r3=columns buffer to fill in (array of longs) for 1 wide pixs. # r4=columns buffer for 2 wide pixs # r5=how many columns (multiple of 32) # r6=palette remapping array (do 256) li r7,255 .init1: stbx r7,r6,r7 subic. r7,r7,1 bne .init1 # column offsets for 1 wide pixels # 0,2,4,6,8,10,12,14,1,3,5,7,9,11,13,15, # 16,18,20,... subi r3,r3,4 srwi r6,r5,4 li r0,8 li r8,0 .init2: li r7,2 .init3: mtctr r0 .init4: stwu r8,4(r3) addi r8,r8,2 bdnz .init4 subi r8,r8,15 subic. r7,r7,1 bne .init3 addi r7,r7,14 subic. r6,r6,1 bne .init2 # column offsets for 2 wide pixels # 0,4,8,12,1,5,9,13,2,6,10,14,3,7,11,15, # 16,20,24,28,17,... subi r4,r4,4 srwi r6,r5,4 li r0,4 li r8,0 .init5: li r7,4 .init6: mtctr r0 .init7: stwu r8,4(r4) addi r8,r8,4 bdnz .init7 subi r8,r8,15 subic. r7,r7,1 bne .init6 addi r7,r7,12 subic. r6,r6,1 bne .init5 blr .type initc2p,@function .size initc2p,$-initc2p .align 4 .global doc2p_1X8 doc2p_1X8: # inputs: # r3=src chunky buffer # r4=dest chipmem bitmap # r5=width (in pixels - multiple of 32) to convert # r6=height (in pixels - even) # r7=modulo from one bitplane to next (copmod-ish) # r8=modulo from start of one line to start of next (linemod) stwu r1,-48(r1) # build stack frame, save registers stmw r24,8(r1) slwi r31,r7,3 srwi r30,r5,4 # r30 = num of 16 pix per row subi r31,r31,2 # r31 = 8*bpmod-2 add r9,r30,r30 sub r5,r8,r9 # r5 = linemod-bytesperrow lis r27,0xf0f0 lis r28,0xcccc lis r29,0xaaaa ori r27,r27,0xf0f0 # r27 = 4 bit mask ori r28,r28,0xcccc # r28 = 2 bit mask ori r29,r29,0x5555 # r29 = 1 bit mask subi r3,r3,4 # prepare source for pre-increment sub r4,r4,r7 # prepare dest. for pre-increment .lineloop1X8: mtctr r30 # CTR = num 16 pix per row .pixloop1X8: lwzu r8,4(r3) # get next 4 pixels 00 lwzu r9,4(r3) # get next 4 pixels 01 mr r24,r8 and r8,r8,r27 lwzu r10,4(r3) # get next 4 pixels 02 xor r24,r24,r8 slwi r24,r24,4 and r25,r10,r27 xor r10,r10,r25 srwi r25,r25,4 lwzu r11,4(r3) # get next 4 pixels 03 or r10,r10,r24 or r8,r8,r25 # 00x02 -> 10 11 mr r24,r9 and r9,r9,r27 xor r24,r24,r9 and r25,r11,r27 xor r11,r11,r25 srwi r25,r25,4 slwi r24,r24,4 or r11,r11,r24 or r9,r9,r25 # 01x03 -> 12 13 mr r24,r10 and r10,r10,r28 xor r24,r24,r10 and r25,r11,r28 xor r11,r11,r25 slwi r24,r24,2 or r11,r11,r24 # 11x13b -> 23 mr r24,r11 and r11,r11,r29 xor r24,r24,r11 rlwimi r26,r24,15,1,15 rlwimi r26,r24,17,16,30 or r11,r11,r26 # 23x23 -> 33 sthux r11,r4,r7 # 33 -> plane 0 srwi r25,r25,2 or r10,10,r25 # 11x13a -> 22 mr r24,r10 and r10,r10,r29 xor r24,r24,r10 rlwimi r11,r11,16,16,31 sthux r11,r4,r7 # 33 -> plane 1 rlwimi r26,r24,15,1,15 rlwimi r26,r24,17,16,30 or r10,r10,r26 # 22x22 -> 32 mr r24,r8 and r8,r8,r28 xor r24,r24,r8 sthux r10,r4,r7 # 32 -> plane 2 and r25,r9,r28 xor r9,r9,r25 slwi r24,r24,2 or r9,r9,r24 # 10x12b -> 21 rlwimi r10,r10,16,16,31 sthux r10,r4,r7 # 32 -> plane 3 mr r24,r9 and r9,r9,r29 xor r24,r24,r9 rlwimi r26,r24,15,1,15 rlwimi r26,r24,17,16,30 or r9,r9,r26 # 21x21 -> 31 sthux r9,r4,r7 # 31 -> plane 4 srwi r25,r25,2 or r8,r8,r25 # 10x12a -> 20 mr r25,r8 and r8,r8,r29 xor r25,r25,r8 rlwimi r9,r9,16,16,31 sthux r9,r4,r7 # 31 -> plane 5 rlwimi r26,r25,15,1,15 rlwimi r26,r25,17,16,30 or r8,r8,r26 # 20x20 -> 30 mr r9,r8 sthux r8,r4,r7 # 30 -> plane 6 rlwimi r9,r9,16,16,31 sthux r9,r4,r7 # 30 -> plane 7 sub r4,r4,r31 # next 16 pix, destptr. to 1st plane bdnz .pixloop1X8 add r4,r4,r5 # + linemod subic. r6,r6,1 bne .lineloop1X8 lmw r24,8(r1) # restore regs, delete stack frame addi r1,r1,48 blr .type doc2p_1X8,@function .size doc2p_1X8,$-doc2p_1X8 .align 4 .global doc2p_1X6 doc2p_1X6: # inputs: # r3=src chunky buffer # r4=dest chipmem bitmap # r5=width (in pixels - multiple of 32) to convert # r6=height (in pixels - even) # r7=modulo from one bitplane to next (copmod-ish) # r8=modulo from start of one line to start of next (linemod) stwu r1,-48(r1) # build stack frame, save registers stmw r24,8(r1) mulli r31,r7,6 srwi r30,r5,4 # r30 = num of 16 pix per row subi r31,r31,2 # r31 = 6*bpmod-2 add r9,r30,r30 sub r5,r8,r9 # r5 = linemod-bytesperrow lis r27,0xf0f0 lis r28,0xcccc lis r29,0xaaaa ori r27,r27,0xf0f0 # r27 = 4 bit mask ori r28,r28,0xcccc # r28 = 2 bit mask ori r29,r29,0x5555 # r29 = 1 bit mask subi r3,r3,4 # prepare source for pre-increment sub r4,r4,r7 # prepare dest. for pre-increment .lineloop1X6: mtctr r30 # CTR = num 16 pix per row .pixloop1X6: lwzu r8,4(r3) # get next 4 pixels 00 lwzu r9,4(r3) # get next 4 pixels 01 mr r24,r8 and r8,r8,r27 lwzu r10,4(r3) # get next 4 pixels 02 xor r24,r24,r8 slwi r24,r24,4 and r25,r10,r27 xor r10,r10,r25 srwi r25,r25,4 lwzu r11,4(r3) # get next 4 pixels 03 or r10,r10,r24 or r8,r8,r25 # 00x02 -> 10 11 mr r24,r9 and r9,r9,r27 xor r24,r24,r9 and r25,r11,r27 xor r11,r11,r25 srwi r25,r25,4 slwi r24,r24,4 or r11,r11,r24 or r9,r9,r25 # 01x03 -> 12 13 mr r24,r10 and r10,r10,r28 xor r24,r24,r10 and r25,r11,r28 xor r11,r11,r25 slwi r24,r24,2 or r11,r11,r24 # 11x13b -> 23 mr r24,r11 and r11,r11,r29 xor r24,r24,r11 rlwimi r26,r24,15,1,15 rlwimi r26,r24,17,16,30 or r11,r11,r26 # 23x23 -> 33 sthux r11,r4,r7 # 33 -> plane 0 srwi r25,r25,2 or r10,10,r25 # 11x13a -> 22 mr r24,r10 and r10,r10,r29 xor r24,r24,r10 rlwimi r11,r11,16,16,31 sthux r11,r4,r7 # 33 -> plane 1 rlwimi r26,r24,15,1,15 rlwimi r26,r24,17,16,30 or r10,r10,r26 # 22x22 -> 32 slwi r8,r8,2 and r24,r9,r28 and r8,r8,r28 xor r9,r9,r24 sthux r10,r4,r7 # 32 -> plane 2 or r9,r9,r8 # 10x12b -> 21 mr r25,r9 and r9,r9,r29 xor r25,r25,r9 rlwimi r10,r10,16,16,31 sthux r10,r4,r7 # 32 -> plane 3 rlwimi r26,r25,15,1,15 rlwimi r26,r25,17,16,30 or r9,r9,r26 # 21x21 -> 31 mr r8,r9 sthux r8,r4,r7 # 31 -> plane 4 rlwimi r9,r9,16,16,31 sthux r9,r4,r7 # 31 -> plane 5 sub r4,r4,r31 # next 16 pix, destptr. to 1st plane bdnz .pixloop1X6 add r4,r4,r5 # + linemod subic. r6,r6,1 bne .lineloop1X6 lmw r24,8(r1) # restore regs, delete stack frame addi r1,r1,48 blr .type doc2p_1X6,@function .size doc2p_1X6,$-doc2p_1X6 .align 4 .global doc2p_2X8 doc2p_2X8: # inputs: # r3=src chunky buffer # r4=dest chipmem bitmap # r5=width (in pixels - multiple of 32) to convert # r6=height (in pixels - even) # r7=modulo from one bitplane to next (copmod-ish) # r8=modulo from start of one line to start of next (linemod) stwu r1,-48(r1) # build stack frame, save registers stmw r24,8(r1) slwi r31,r7,3 srwi r30,r5,4 # r30 = num of 16 pix per row subi r31,r31,4 # r31 = 8*bpmod-4 slwi r9,r30,2 sub r5,r8,r9 # r5 = linemod-bytesperrow lis r27,0xf0f0 lis r28,0xcccc lis r29,0xaaaa ori r27,r27,0xf0f0 # r27 = 4 bit mask ori r28,r28,0xcccc # r28 = 2 bit mask ori r29,r29,0xaaaa # r29 = 1 bit mask subi r3,r3,4 # prepare source for pre-increment sub r4,r4,r7 # prepare dest. for pre-increment .lineloop2X8: mtctr r30 # CTR = num 16 pix per row .pixloop2X8: lwzu r8,4(r3) # get next 4 pixels 00 lwzu r9,4(r3) # get next 4 pixels 01 mr r24,r8 and r8,r8,r27 lwzu r10,4(r3) # get next 4 pixels 02 xor r24,r24,r8 slwi r24,r24,4 and r25,r10,r27 xor r10,r10,r25 srwi r25,r25,4 lwzu r11,4(r3) # get next 4 pixels 03 or r10,r10,r24 or r8,r8,r25 # 00x02 -> 10 11 mr r24,r9 and r9,r9,r27 xor r24,r24,r9 and r25,r11,r27 xor r11,r11,r25 srwi r25,r25,4 slwi r24,r24,4 or r11,r11,r24 or r9,r9,r25 # 01x03 -> 12 13 mr r24,r8 and r8,r8,r28 xor r24,r24,r8 slwi r24,r24,2 and r25,r9,r28 xor r9,r9,r25 srwi r25,r25,2 or r9,r9,r24 or r8,r8,r25 # 10x12 -> 20 21 mr r24,r10 and r10,r10,r28 xor r24,r24,r10 slwi r24,r24,2 and r25,r11,r28 xor r11,r11,r25 srwi r25,r25,2 or r11,r11,r24 or r10,r10,r25 # 11x13 -> 22 23 mr r24,r11 and r11,r11,r29 xor r24,r24,r11 add r25,r24,r24 or r25,r25,r24 # 23x23b -> 37 stwux r25,r4,r7 # 37 -> plane 0 mr r24,r11 srwi r11,r11,1 or r24,r24,r11 # 23x23a -> 36 mr r25,r10 and r10,r10,r29 xor r25,r25,r10 stwux r24,r4,r7 # 36 -> plane 1 mr r11,r25 add r25,r25,r25 or r11,r11,r25 # 22x22b -> 35 mr r25,r10 srwi r10,r10,1 or r10,r10,r25 # 22x22a -> 34 stwux r11,r4,r7 # 35 -> plane 2 mr r24,r9 and r9,r9,r29 xor r24,r24,r9 mr r25,r8 and r8,r8,r29 xor r25,r25,r8 stwux r10,r4,r7 # 34 -> plane 3 mr r10,r24 add r24,r24,r24 or r10,r10,r24 # 21x21b -> 33 mr r26,r9 srwi r9,r9,1 or r26,r26,r9 # 21x21a -> 32 stwux r10,r4,r7 # 33 -> plane 4 add r25,r25,r25 or r25,r25,r11 # 20x20b -> 31 stwux r26,r4,r7 # 32 -> plane 5 mr r11,r8 srwi r8,r8,1 or r11,r11,r8 # 20x20a -> 30 stwux r25,r4,r7 # 31 -> plane 6 stwux r11,r4,r7 # 30 -> plane 7 sub r4,r4,r31 # next 16 pix, destptr. to 1st plane bdnz .pixloop2X8 add r4,r4,r5 # + linemod subic. r6,r6,1 bne .lineloop2X8 lmw r24,8(r1) # restore regs, delete stack frame addi r1,r1,48 blr .type doc2p_2X8,@function .size doc2p_2X8,$-doc2p_2X8 .align 4 .global doc2p_2X6 doc2p_2X6: # inputs: # r3=src chunky buffer # r4=dest chipmem bitmap # r5=width (in pixels - multiple of 32) to convert # r6=height (in pixels - even) # r7=modulo from one bitplane to next (copmod-ish) # r8=modulo from start of one line to start of next (linemod) stwu r1,-48(r1) # build stack frame, save registers stmw r24,8(r1) mulli r31,r7,6 srwi r30,r5,4 # r30 = num of 16 pix per row subi r31,r31,4 # r31 = 6*bpmod-4 slwi r9,r30,2 sub r5,r8,r9 # r5 = linemod-bytesperrow lis r27,0xf0f0 lis r28,0xcccc lis r29,0xaaaa ori r27,r27,0xf0f0 # r27 = 4 bit mask ori r28,r28,0xcccc # r28 = 2 bit mask ori r29,r29,0xaaaa # r29 = 1 bit mask subi r3,r3,4 # prepare source for pre-increment sub r4,r4,r7 # prepare dest. for pre-increment .lineloop2X6: mtctr r30 # CTR = num 16 pix per row .pixloop2X6: lwzu r8,4(r3) # get next 4 pixels 00 lwzu r9,4(r3) # get next 4 pixels 01 mr r24,r8 and r8,r8,r27 lwzu r10,4(r3) # get next 4 pixels 02 xor r24,r24,r8 slwi r24,r24,4 and r25,r10,r27 xor r10,r10,r25 srwi r25,r25,4 lwzu r11,4(r3) # get next 4 pixels 03 or r10,r10,r24 or r8,r8,r25 # 00x02 -> 10 11 mr r24,r9 and r9,r9,r27 xor r24,r24,r9 and r25,r11,r27 xor r11,r11,r25 srwi r25,r25,4 slwi r24,r24,4 or r11,r11,r24 or r9,r9,r25 # 01x03 -> 12 13 mr r24,r8 and r8,r8,r28 xor r24,r24,r8 slwi r24,r24,2 and r25,r9,r28 xor r9,r9,r25 srwi r25,r25,2 or r9,r9,r24 or r8,r8,r25 # 10x12 -> 20 21 mr r24,r10 and r10,r10,r28 xor r24,r24,r10 slwi r24,r24,2 and r25,r11,r28 xor r11,r11,r25 srwi r25,r25,2 or r11,r11,r24 or r10,r10,r25 # 11x13 -> 22 23 mr r24,r11 and r11,r11,r29 xor r24,r24,r11 add r25,r24,r24 or r25,r25,r24 # 23x23b -> 37 stwux r25,r4,r7 # 37 -> plane 0 mr r24,r11 srwi r11,r11,1 or r24,r24,r11 # 23x23a -> 36 mr r25,r10 and r10,r10,r29 xor r25,r25,r10 stwux r24,r4,r7 # 36 -> plane 1 mr r11,r25 add r25,r25,r25 or r11,r11,r25 # 22x22b -> 35 mr r25,r10 srwi r10,r10,1 or r10,r10,r25 # 22x22a -> 34 stwux r11,r4,r7 # 35 -> plane 2 mr r24,r9 and r9,r9,r29 xor r24,r24,r9 mr r25,r8 and r8,r8,r29 xor r25,r25,r8 stwux r10,r4,r7 # 34 -> plane 3 mr r10,r24 add r24,r24,r24 or r10,r10,r24 # 21x21b -> 33 stwux r10,r4,r7 # 33 -> plane 4 mr r26,r9 srwi r9,r9,1 or r26,r26,r9 # 21x21a -> 32 stwux r26,r4,r7 # 32 -> plane 5 sub r4,r4,r31 # next 16 pix, destptr. to 1st plane bdnz .pixloop2X6 add r4,r4,r5 # + linemod subic. r6,r6,1 bne .lineloop2X6 lmw r24,8(r1) # restore regs, delete stack frame addi r1,r1,48 blr .type doc2p_2X6,@function .size doc2p_2X6,$-doc2p_2X6