@ (c) Copyright 2007, Grazvydas "notaz" Ignotas\r
@ All Rights Reserved\r
\r
+.include "port_config.s"\r
\r
.extern Pico\r
.extern PicoOpt\r
@ int cells; // 0x14\r
@ };\r
\r
-@ int DrawLayer(int plane, int *hcache, int maxcells, int sh)\r
+@ void DrawLayer(int plane_sh, int *hcache, int cellskip, int maxcells);\r
\r
-.global DrawLayer @ int plane, int *hcache, int maxcells, int sh\r
+.global DrawLayer\r
\r
DrawLayer:\r
stmfd sp!, {r4-r11,lr}\r
ldr r11, =(Pico+0x22228) @ Pico.video\r
mov r8, #1\r
\r
- ldrb r7, [r11, #16] @ ??hh??ww\r
+ ldrb r7, [r11, #16] @ ??vv??hh\r
\r
mov r6, r1 @ hcache\r
- orr r9, r2, r3, lsl #31 @ r9=maxcells|(sh<<31)\r
+ orr r9, r3, r0, lsl #30\r
+ orr r9, r9, r2, lsl #8 @ r9=sh[31]|cellskip[15:8]|maxcells[7:0] (tmp)\r
\r
mov r1, r7, lsl #4\r
orr r1, r1, #0x00ff\r
sub r5, r5, #1 @ r5=xmask\r
\r
@ Find name table:\r
- tst r0, r0\r
+ ands r0, r0, #1\r
ldreqb r12, [r11, #2]\r
ldrneb r12, [r11, #4]\r
\r
orrne r10,r10, #1<<23 @ r10=(cells<<24|sh<<23|hi_not_empty<<22|had_output<<21|ty)\r
movne r3, #0x40 @ default to shadowed pal on sh mode\r
\r
- mvn r9, #0 @ r9=prevcode=-1\r
-\r
cmp r7, #8\r
addne r10,r10, #0x01000000 @ we will loop cells+1 times if there is scroll\r
\r
+ and r9, r9, #0xff00\r
+ add r8, r8, r9, lsr #8 @ tilex+=cellskip\r
+ add r7, r7, r9, lsr #5 @ dx+=cellskip<<3;\r
+ sub r10,r10,r9, lsl #16 @ cells-=cellskip\r
+\r
@ cache some stuff to avoid mem access\r
+.if OVERRIDE_HIGHCOL\r
+ ldr r11,=HighCol\r
+ mov r0, #0xf\r
+ ldr r11,[r11]\r
+.else\r
ldr r11,=HighCol\r
mov r0, #0xf\r
+.endif\r
+\r
+ mvn r9, #0 @ r9=prevcode=-1\r
add r1, r11, r7 @ r1=pdest\r
\r
\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[15:0])\r
+ orr r8, r8, r5, lsl #25 @ r8=(xmask[31:25]|had_output[24]|tilex[23:0])\r
\r
ldr r4, =Scanline\r
orr r5, r1, r10, lsl #24\r
add r7, r1, #1 @ r7=dx=((ts->hscroll-1)&7)+1\r
\r
mov r10,r9, lsl #16\r
- tst r0, r0\r
+ tst r0, #1\r
orrne r10,r10, #0x8000\r
tst r9, #1<<31\r
mov r3, #0\r
orr r10,r10, #0xff000000 @ will be adjusted on entering loop\r
- orrne r10,r10, #1<<23 @ r10=(cells[31:24]|sh[23]|hi_not_empty[22]|cells_max[21:16]|plane[15]|ty[14:0])\r
+ orrne r10,r10, #1<<23 @ r10=(cell[31:24]|sh[23]|hi_not_empty[22]|cells_max[21:16]|plane[15]|ty[14:0])\r
movne r3, #0x40 @ default to shadowed pal on sh mode\r
\r
- mvn r9, #0 @ r9=prevcode=-1\r
+ cmp r7, #8\r
+ subne r10,r10, #0x01000000 @ have hscroll, start with negative cell\r
+\r
+ and r9, r9, #0xff00\r
+ add r8, r8, r9, lsr #8 @ tilex+=cellskip\r
+ add r7, r7, r9, lsr #5 @ dx+=cellskip<<3;\r
+ add r10,r10,r9, lsl #16 @ cell+=cellskip\r
\r
@ cache some stuff to avoid mem access\r
+.if OVERRIDE_HIGHCOL\r
ldr r11,=HighCol\r
mov r0, #0xf\r
- add r1, r11, r7 @ r1=pdest\r
-\r
- cmp r7, #8\r
- subne r10,r10, #0x01000000 @ have hscroll, start with negative cell\r
+ ldr r11,[r11]\r
+.else\r
+ ldr r11,=HighCol\r
+ mov r0, #0xf\r
+.endif\r
\r
+ mvn r9, #0 @ r9=prevcode=-1\r
+ add r1, r11, r7 @ r1=pdest\r
\r
@ r4 & r7 are scratch in this loop\r
.dsloop_vs_subr1:\r
BackFill:\r
stmfd sp!, {r4-r9,lr}\r
\r
+.if OVERRIDE_HIGHCOL\r
+ ldr lr, =HighCol\r
+ mov r0, r0, lsl #26\r
+ ldr lr, [lr]\r
+ mov r0, r0, lsr #26\r
+ add lr, lr, #8\r
+.else\r
ldr lr, =(HighCol+8)\r
-\r
mov r0, r0, lsl #26\r
mov r0, r0, lsr #26\r
+.endif\r
+\r
orr r0, r0, r1, lsl #6\r
orr r0, r0, r0, lsl #8\r
orr r0, r0, r0, lsl #16\r
stmfd sp!, {r4-r8,r11,lr}\r
\r
@ cache some stuff to avoid mem access\r
+.if OVERRIDE_HIGHCOL\r
+ ldr r11,=HighCol\r
+ mov r12,#0xf\r
+ ldr r11,[r11]\r
+.else\r
ldr r11,=HighCol\r
- ldr lr, =(Pico+0x10000) @ lr=Pico.vram\r
mov r12,#0xf\r
+.endif\r
+ ldr lr, =(Pico+0x10000) @ lr=Pico.vram\r
\r
mvn r5, #0 @ r5=prevcode=-1\r
ands r8, r1, #1\r
stmfd sp!, {r4-r11,lr}\r
\r
@ cache some stuff to avoid mem access\r
+.if OVERRIDE_HIGHCOL\r
ldr r11,=HighCol\r
+ mov r12,#0xf\r
+ ldr r11,[r11]\r
+.else\r
+ ldr r11,=HighCol\r
+ mov r12,#0xf\r
+.endif\r
ldr lr, =(Pico+0x10000) @ lr=Pico.vram\r
mov r6, r1, lsl #31\r
orr r6, r6, #1<<30\r
- mov r12,#0xf\r
\r
mov r10, r0\r
\r
bne .dspr_cache @ if(code&0x8000) // high priority - cache it\r
\r
@ cache some stuff to avoid mem access\r
+.if OVERRIDE_HIGHCOL\r
+ ldr r11,=HighCol\r
+ mov r12,#0xf\r
+ ldr r11,[r11]\r
+.else\r
ldr r11,=HighCol\r
- ldr lr, =(Pico+0x10000) @ lr=Pico.vram\r
mov r12,#0xf\r
+.endif\r
+ ldr lr, =(Pico+0x10000) @ lr=Pico.vram\r
\r
mov r5, r5, lsl #4 @ delta<<=4; // Delta of address\r
and r4, r9, #0x6000\r
orr r6, r6, r3, lsl #8 @ shadow mode\r
\r
sub r8, r1, r0\r
- mov r8, r8, lsl #1 @ cells\r
-\r
- mvn r9, #0 @ r9=prevcode=-1\r
\r
@ cache some stuff to avoid mem access\r
+.if OVERRIDE_HIGHCOL\r
+ ldr r11,=HighCol\r
+ mov r8, r8, lsl #1 @ cells\r
+ ldr r11,[r11]\r
+ mvn r9, #0 @ r9=prevcode=-1\r
+ add r11,r11,#8\r
+.else\r
ldr r11,=(HighCol+8)\r
+ mov r8, r8, lsl #1 @ cells\r
+ mvn r9, #0 @ r9=prevcode=-1\r
+.endif\r
add r1, r11, r0, lsl #4 @ r1=pdest\r
mov r0, #0xf\r
b .dwloop_enter\r
\r
\r
.fl_noshBGR444:\r
+.if OVERRIDE_HIGHCOL\r
+ ldr r1, =HighCol\r
+ mov lr, #0xff\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
mov lr, lr, lsl #1\r
+.endif\r
\r
.fl_loopBGR444:\r
\r
orr \reg, \reg, r3 @ add blue back\r
.endm\r
\r
+.global vidConvCpyRGB565\r
+\r
vidConvCpyRGB565: @ void *to, void *from, int pixels\r
stmfd sp!, {r4-r9,lr}\r
\r
- mov r12, r2, lsr #3 @ repeats\r
+ mov r12, r2, lsr #3 @ repeats\r
mov lr, #0x001c0000\r
orr lr, lr, #0x01c @ lr == pattern 0x001c001c\r
mov r8, #0x00030000\r
- orr r8, r8, #0x003 @ lr == pattern 0x001c001c\r
+ orr r8, r8, #0x003\r
\r
.loopRGB565:\r
ldmia r1!, {r4-r7}\r
sub r3, r3, #0x40*2\r
\r
.fl_noshRGB555:\r
+.if OVERRIDE_HIGHCOL\r
+ ldr r1, =HighCol\r
+ ldr r0, =DrawLineDest\r
+ ldr r1, [r1]\r
+ ldr r0, [r0]\r
+ add r1, r1, #8\r
+.else\r
ldr r0, =DrawLineDest\r
ldr r1, =(HighCol+8)\r
ldr r0, [r0]\r
+.endif\r
\r
ldrb r12, [r8, #12]\r
mov lr, #0xff\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