.DrawStrip_vsscroll:\r
rsb r8, r3, #0\r
mov r8, r8, lsr #3 @ r8=tilex=(-ts->hscroll)>>3\r
- bic r8, r8, #0xff000000\r
- orr r8, r8, r5, lsl #25 @ r8=(xmask[31:25]|had_output[24]|tilex[23:0])\r
+ bic r8, r8, #0x3fc00000\r
+ orr r8, r8, r5, lsl #25 @ r8=(xmask[31:25]|had_output[24]|tilex[21:0])\r
\r
ldr r4, =Scanline\r
orr r5, r1, r10, lsl #24\r
ldrh r7, [r7] @ r7=vscroll\r
\r
bic r10,r10,#0xff @ clear old ty\r
- and r4, r5, #0xff0000\r
- add r4, r4, r7, lsl #16\r
- and r4, r4, r5, lsl #16 @ r4=line<<16\r
+ and r4, r5, #0xff0000 @ scanline\r
+ add r4, r4, r7, lsl #16 @ ... += vscroll\r
+ and r4, r4, r5, lsl #16 @ ... &= ymask\r
and r7, r4, #0x70000\r
orr r10,r10,r7, lsr #15 @ new ty\r
\r
beq .DrawStrip_vs_samecode @ we know stuff about this tile already\r
\r
mov r9, r7 @ remember code\r
- orr r8, r8, #1<<24 @ seen non hi-prio tile\r
+ orr r8, r8, #(1<<24)@ seen non hi-prio tile\r
\r
movs r2, r9, lsl #20 @ if (code&0x1000)\r
mov r2, r2, lsl #1\r
b .dsloop_vs\r
\r
.dsloop_vs_exit:\r
- tst r8, #1<<24 @ seen non hi-prio tile\r
+ tst r8, #(1<<24) @ seen non hi-prio tile\r
ldreq r1, =rendstatus\r
mov r0, #0\r
ldreq r2, [r1]\r
@ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r
\r
\r
-.global DrawSpritesFromCache @ int *hc, int sh\r
+.global DrawSpritesFromCache @ int *hc, int maxwidth, int prio, int sh\r
\r
DrawSpritesFromCache:\r
stmfd sp!, {r4-r11,lr}\r
mov r12,#0xf\r
.endif\r
ldr lr, =(Pico+0x10000) @ lr=Pico.vram\r
- mov r6, r1, lsl #31\r
+ mov r6, r3, lsl #31\r
orr r6, r6, #1<<30\r
\r
mov r10, r0\r
@ + 0 : hhhhvvvv ab--hhvv yyyyyyyy yyyyyyyy // a: offscreen h, b: offs. v, h: horiz. size\r
@ + 4 : xxxxxxxx xxxxxxxx pccvhnnn nnnnnnnn // x: x coord + 8\r
\r
-.global DrawSprite @ unsigned int *sprite, int **hc, int sh\r
+.global DrawSprite @ unsigned int *sprite, int **hc, int sh, int acc_sprites\r
\r
DrawSprite:\r
stmfd sp!, {r4-r9,r11,lr}\r
\r
+ orr r8, r3, r2, lsl #4\r
ldr r3, [r0] @ sprite[0]\r
ldr r7, =Scanline\r
mov r6, r3, lsr #28\r
ldr r9, [r0, #4]\r
sub r7, r7, r4, asr #16 @ r7=row=Scanline-sy\r
\r
- tst r2, r2\r
mov r2, r9, asr #16 @ r2=sx\r
- bic r9, r9, #0xfe000000\r
- orrne r9, r9, #1<<31 @ r9=code|(sh<<31)\r
+ mov r9, r9, lsl #16\r
+ mov r9, r9, lsr #16\r
+ orr r9, r9, r8, lsl #27 @ r9=code|sh[31]|as[27]\r
\r
tst r9, #0x1000\r
movne r4, r5, lsl #3\r
tst r9, #0x8000\r
bne .dspr_cache @ if(code&0x8000) // high priority - cache it\r
\r
+.dspr_continue:\r
@ cache some stuff to avoid mem access\r
.if OVERRIDE_HIGHCOL\r
ldr r11,=HighCol\r
mov r5, r5, lsl #4 @ delta<<=4; // Delta of address\r
and r4, r9, #0x6000\r
orr r9, r9, r4, lsl #16\r
- orr r9, r9, #0x10000000 @ r9=scc1 ???? ... <code> (s=shadow/hilight, cc=pal)\r
+ orrs r9, r9, #0x10000000 @ r9=scc1 a??? ... <code> (s=shadow/hilight, cc=pal, a=acc_spr)\r
\r
- tst r9, #1<<31\r
mov r3, r4, lsr #9 @ r3=pal=((code>>9)&0x30);\r
- orrne r3, r3, #0x40 @ shadow by default\r
+ orrmi r3, r3, #0x40 @ shadow by default\r
\r
add r6, r6, #1 @ inc now\r
adds r0, r2, #0 @ mov sx to r0 and set ZV flags\r
mov r4, r8, lsl #16 @ tile\r
tst r9, #0x0800\r
orrne r4, r4, #0x10000 @ code&0x0800\r
- mov r2, r2, lsl #22\r
- orr r4, r4, r2, lsr #16 @ (sx<<6)&0x0000ffc0\r
- and r2, r9, #0x6000\r
- orr r4, r4, r2, lsr #9 @ (code>>9)&0x30\r
+ mov r0, r2, lsl #22\r
+ orr r4, r4, r0, lsr #16 @ (sx<<6)&0x0000ffc0\r
+ and r0, r9, #0x6000\r
+ orr r4, r4, r0, lsr #9 @ (code>>9)&0x30\r
mov r3, r3, lsl #12\r
- ldr r2, [r1]\r
+ ldr r0, [r1]\r
orr r4, r4, r3, lsr #28 @ (sprite[0]>>24)&0xf\r
\r
- str r4, [r2], #4\r
- str r2, [r1]\r
+ str r4, [r0], #4\r
+ str r0, [r1]\r
\r
- ldmfd sp!, {r4-r9,r11,lr}\r
- bx lr\r
+ tst r9, #(1<<27)\r
+ ldmeqfd sp!, {r4-r9,r11,lr}\r
+ bne .dspr_continue @ draw anyway if accurate sprites enabled\r
+ bxeq lr\r
\r
@ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r
\r
bne .fl_loopcpBGR444_hi\r
\r
sub r3, r4, #0x40*3*2\r
+ mov r6, #1\r
\r
\r
.fl_noshBGR444:\r
+ ldr r12,=rendstatus\r
+ eors r6, r6, #1 @ sh is 0\r
+ ldr r12,[r12]\r
+ mov lr, #0xff\r
+ tstne r12,#(1<<2) @ and PDRAW_ACC_SPRITES\r
+\r
.if OVERRIDE_HIGHCOL\r
ldr r1, =HighCol\r
- mov lr, #0xff\r
+ movne lr, #0x3f\r
ldr r1, [r1]\r
mov lr, lr, lsl #1\r
add r1, r1, #8\r
.else\r
ldr r1, =(HighCol+8)\r
- mov lr, #0xff\r
+ movne lr, #0x3f\r
mov lr, lr, lsl #1\r
.endif\r
\r
.fl_loopBGR444:\r
-\r
ldr r12, [r1], #4\r
subs r2, r2, #1\r
\r
ldrh r5, [r3, r5]\r
and r6, lr, r12, lsr #15\r
ldrh r6, [r3, r6]\r
+ and r12,lr, r12, lsr #23\r
+ ldrh r12,[r3, r12] @ 1c.i.\r
orr r4, r4, r5, lsl #16\r
-\r
- and r5, lr, r12, lsr #23\r
- ldrh r5, [r3, r5] @ 2c.i.\r
- orr r5, r6, r5, lsl #16\r
+ orr r5, r6, r12,lsl #16\r
\r
stmia r0!, {r4,r5}\r
bne .fl_loopBGR444\r
bne .fl_loopcpRGB555_hi\r
\r
sub r3, r3, #0x40*2\r
+ mov r6, #1\r
\r
.fl_noshRGB555:\r
+ ldr r12,=rendstatus\r
+ eors r6, r6, #1 @ sh is 0\r
+ ldr r12,[r12]\r
+ mov lr, #0xff\r
+ tstne r12,#(1<<2) @ and PDRAW_ACC_SPRITES\r
+ movne lr, #0x3f\r
+\r
.if OVERRIDE_HIGHCOL\r
ldr r1, =HighCol\r
ldr r0, =DrawLineDest\r
.endif\r
\r
ldrb r12, [r8, #12]\r
- mov lr, #0xff\r
mov lr, lr, lsl #1\r
\r
tst r12, #1\r
addeq r0, r0, #32*2\r
\r
.fl_no32colRGB555:\r
-.fl_loopRGB555:\r
\r
+.if UNALIGNED_DRAWLINEDEST\r
+ @ this is basically for Gizmondo, which has unaligned odd lines in the framebuffer\r
+ tst r0, #2\r
+ bne .fl_RGB555u\r
+.endif\r
+\r
+.fl_loopRGB555:\r
ldr r12, [r1], #4\r
ldr r7, [r1], #4\r
\r
ldrh r6, [r3, r6]\r
and r12,lr, r7, lsr #15\r
ldrh r12,[r3, r12]\r
+ and r7, lr, r7, lsr #23\r
+ ldrh r7, [r3, r7]\r
orr r8, r8, r6, lsl #16\r
\r
- and r6, lr, r7, lsr #23\r
- ldrh r6, [r3, r6] @ 1 cycle interlock here (r6)\r
subs r2, r2, #1\r
- orr r12,r12, r6, lsl #16\r
+ orr r12,r12, r7, lsl #16\r
\r
stmia r0!, {r4,r5,r8,r12}\r
bne .fl_loopRGB555\r
mov r9, #0x3900 @ f800 07e0 001f | e000 0780 001c | 3800 01e0 0007\r
orr r9, r9, #0x00e7\r
\r
+.if UNALIGNED_DRAWLINEDEST\r
+ tst r0, #2\r
+ bne .fl_32scale_RGB555u\r
+.endif\r
+\r
.fl_loop32scale_RGB555:\r
ldr r12, [r1], #4\r
ldr r7, [r1], #4\r
ldmfd sp!, {r4-r8,lr}\r
bx lr\r
\r
+.if UNALIGNED_DRAWLINEDEST\r
+ @ unaligned versions of loops\r
+ @ warning: starts drawing 2bytes before dst\r
+\r
+.fl_RGB555u:\r
+ sub r0, r0, #2 @ initial adjustment\r
+ mov r8, #0\r
+\r
+.fl_loopRGB555u:\r
+ ldr r12, [r1], #4\r
+ ldr r7, [r1], #4\r
+\r
+ and r6, lr, r12,lsl #1\r
+ ldrh r6, [r3, r6]\r
+ and r5, lr, r12,lsr #7\r
+ ldrh r5, [r3, r5]\r
+ orr r4, r8, r6, lsl #16\r
+\r
+ and r6, lr, r12,lsr #15\r
+ ldrh r6, [r3, r6]\r
+ and r8, lr, r12,lsr #23\r
+ ldrh r8, [r3, r8]\r
+ orr r5, r5, r6, lsl #16\r
+\r
+ and r6, lr, r7, lsl #1\r
+ ldrh r6, [r3, r6]\r
+ and r12,lr, r7, lsr #7\r
+ ldrh r12,[r3, r12]\r
+ orr r6, r8, r6, lsl #16\r
+\r
+ and r8, lr, r7, lsr #15\r
+ ldrh r8, [r3, r8]\r
+ and r7, lr, r7, lsr #23\r
+\r
+ subs r2, r2, #1\r
+ orr r12,r12,r8, lsl #16\r
+ ldrh r8, [r3, r7]\r
+\r
+ stmia r0!, {r4,r5,r6,r12}\r
+ bne .fl_loopRGB555u\r
+\r
+ strh r8, [r0], #2\r
+\r
+ ldmfd sp!, {r4-r8,lr}\r
+ bx lr\r
+\r
+\r
+.fl_32scale_RGB555u:\r
+ sub r0, r0, #2 @ initial adjustment\r
+ mov r4, #0\r
+\r
+ @ r9 f800 07e0 001f | e000 0780 001c | 3800 01e0 0007\r
+.fl_loop32scale_RGB555u:\r
+ ldr r12, [r1], #4\r
+ ldr r7, [r1], #4\r
+\r
+ and r6, lr, r12,lsl #1\r
+ ldrh r6, [r3, r6]\r
+ and r5, lr, r12,lsr #7\r
+ ldrh r5, [r3, r5]\r
+ and r6, r6, r9, lsl #2\r
+ orr r4, r4, r6, lsl #16 @ r4 = pix_d -1, 0\r
+\r
+ and r5, r5, r9, lsl #2\r
+ sub r8, r5, r5, lsr #2 @ r8 = 3/4 pix_s 1\r
+ add r6, r8, r6, lsr #2 @ r6 = (1/4 pix_s 0) + (3/4 pix_s 1)\r
+ orr r5, r6, r5, lsl #15\r
+\r
+ and r6, lr, r12,lsr #15\r
+ ldrh r6, [r3, r6]\r
+ and r12,lr, r12,lsr #23\r
+ ldrh r12,[r3, r12]\r
+ and r6, r6, r9, lsl #2\r
+ add r5, r5, r6, lsl #15 @ r5 = pix_d 1, 2\r
+\r
+ and r8, lr, r7, lsl #1\r
+ ldrh r8, [r3, r8]\r
+ and r10,lr, r7, lsr #7\r
+ ldrh r10,[r3, r10]\r
+ and r12,r12,r9, lsl #2\r
+ sub r6, r6, r6, lsr #2 @ r6 = 3/4 pix_s 2\r
+ add r6, r6, r12,lsr #2\r
+ orr r6, r6, r12,lsl #16 @ r6 = pix_d 3, 4\r
+\r
+ and r8, r8, r9, lsl #2\r
+ and r10,r10,r9, lsl #2\r
+ sub r12,r10,r10,lsr #2 @ r12 = 3/4 pix_s 5\r
+ orr r8, r8, r8, lsl #14\r
+ add r8, r8, r12,lsl #16 @ r8 = pix_d 5, 6\r
+ and r12,lr, r7, lsr #15\r
+ ldrh r12,[r3, r12]\r
+ and r7, lr, r7, lsr #23\r
+ ldrh r7, [r3, r7]\r
+ and r12,r12,r9, lsl #2\r
+ add r10,r10,r12\r
+ mov r10,r10, lsr #1\r
+ sub r12,r12,r12,lsr #2 @ r12 = 3/4 pix_s 6\r
+ orr r10,r10,r12,lsl #16\r
+ and r7, r7, r9, lsl #2\r
+ add r10,r10,r7, lsl #14 @ r10 = pix_d 7, 8\r
+\r
+ subs r2, r2, #1\r
+\r
+ stmia r0!, {r4,r5,r6,r8,r10}\r
+ mov r4, r7\r
+ bne .fl_loop32scale_RGB555u\r
+\r
+ strh r4, [r0], #2\r
+\r
+ ldmfd sp!, {r9,r10}\r
+ ldmfd sp!, {r4-r8,lr}\r
+ bx lr\r
+\r
+.endif @ UNALIGNED_DRAWLINEDEST\r
+\r
\r
@ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r
\r
notaz.gp2x.de / picodrive.git / blobdiff