notaz.gp2x.de / picodrive.git / blame
? search:
summary | shortlog | log | commit | commitdiff | tree
blob | blame (incremental) | history | HEAD
random minor fixes
[picodrive.git] / pico / draw_arm.s
CommitLineData
3d48f143 1@ vim:filetype=armasm\r
2\r
07abbab1 3@ ARM assembly versions of some funtions from draw.c\r
cc68a136 4@ this is highly specialized, be careful if changing related C code!\r
5\r
07abbab1 6@ (c) Copyright 2007-2008, Grazvydas "notaz" Ignotas\r
cc68a136 7@ All Rights Reserved\r
8\r
ea8c405f 9.include "port_config.s"\r
cc68a136 10\r
11.extern Pico\r
12.extern PicoOpt\r
13.extern HighCol\r
b6d7ac70 14.extern DrawScanline\r
cc68a136 15.extern HighSprZ\r
16.extern rendstatus\r
283fec1b 17.extern HighPreSpr\r
cc68a136 18.extern DrawLineDest\r
cc68a136 19.extern DrawStripInterlace\r
07abbab1 20.extern HighCacheS_ptr\r
cc68a136 21\r
5a681086 22.equiv OVERRIDE_HIGHCOL, 1\r
23\r
283fec1b 24.equ PDRAW_SPRITES_MOVED, (1<<0)\r
25.equ PDRAW_WND_DIFF_PRIO, (1<<1)\r
26.equ PDRAW_ACC_SPRITES, (1<<2)\r
27.equ PDRAW_DIRTY_SPRITES, (1<<4)\r
28.equ PDRAW_PLANE_HI_PRIO, (1<<6)\r
29.equ PDRAW_SHHI_DONE, (1<<7)\r
cc68a136 30\r
31@ helper\r
32.macro TilePixel pat lsrr offs\r
33.if !\lsrr\r
34 ands r4, \pat, r2\r
35.else\r
36 ands r4, \pat, r2, lsr #\lsrr\r
37.endif\r
38 orrne r4, r3, r4\r
39 strneb r4, [r1,#\offs]\r
40.endm\r
41\r
42@ TileNorm (r1=pdest, r2=pixels8, r3=pal) r4: scratch, pat: register with helper pattern 0xf\r
43.macro TileNorm pat\r
44 TilePixel \pat, 12, 0 @ #0x0000f000\r
45 TilePixel \pat, 8, 1 @ #0x00000f00\r
46 TilePixel \pat, 4, 2 @ #0x000000f0\r
47 TilePixel \pat, 0, 3 @ #0x0000000f\r
48 TilePixel \pat, 28, 4 @ #0xf0000000\r
49 TilePixel \pat, 24, 5 @ #0x0f000000\r
50 TilePixel \pat, 20, 6 @ #0x00f00000\r
51 TilePixel \pat, 16, 7 @ #0x000f0000\r
52.endm\r
53\r
54@ TileFlip (r1=pdest, r2=pixels8, r3=pal) r4: scratch, pat: register with helper pattern 0xf\r
55.macro TileFlip pat\r
56 TilePixel \pat, 16, 0 @ #0x000f0000\r
57 TilePixel \pat, 20, 1 @ #0x00f00000\r
58 TilePixel \pat, 24, 2 @ #0x0f000000\r
59 TilePixel \pat, 28, 3 @ #0xf0000000\r
60 TilePixel \pat, 0, 4 @ #0x0000000f\r
61 TilePixel \pat, 4, 5 @ #0x000000f0\r
62 TilePixel \pat, 8, 6 @ #0x00000f00\r
63 TilePixel \pat, 12, 7 @ #0x0000f000\r
64.endm\r
65\r
66@ shadow/hilight mode\r
67\r
68@ this one is for hi priority layer\r
69.macro TilePixelShHP lsrr offs\r
70.if !\lsrr\r
71 ands r4, r12, r2\r
72.else\r
73 ands r4, r12, r2, lsr #\lsrr\r
74.endif\r
75 ldreqb r4, [r1,#\offs]\r
76 orrne r4, r3, r4\r
bfa12428 77 andeq r4, r4, #0xbf\r
07abbab1 78 strb r4, [r1,#\offs]\r
cc68a136 79.endm\r
80\r
7a7c6476 81@ TileNormShHP (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r12: register with helper pattern 0xf, touches r3 high bits\r
cc68a136 82.macro TileNormShHP\r
83 TilePixelShHP 12, 0 @ #0x0000f000\r
84 TilePixelShHP 8, 1 @ #0x00000f00\r
85 TilePixelShHP 4, 2 @ #0x000000f0\r
86 TilePixelShHP 0, 3 @ #0x0000000f\r
87 TilePixelShHP 28, 4 @ #0xf0000000\r
88 TilePixelShHP 24, 5 @ #0x0f000000\r
89 TilePixelShHP 20, 6 @ #0x00f00000\r
90 TilePixelShHP 16, 7 @ #0x000f0000\r
91.endm\r
92\r
7a7c6476 93@ TileFlipShHP (r1=pdest, r2=pixels8, r3=pal) r4: scratch, pat: register with helper pattern 0xf\r
cc68a136 94.macro TileFlipShHP\r
95 TilePixelShHP 16, 0 @ #0x000f0000\r
96 TilePixelShHP 20, 1 @ #0x00f00000\r
97 TilePixelShHP 24, 2 @ #0x0f000000\r
98 TilePixelShHP 28, 3 @ #0xf0000000\r
99 TilePixelShHP 0, 4 @ #0x0000000f\r
100 TilePixelShHP 4, 5 @ #0x000000f0\r
101 TilePixelShHP 8, 6 @ #0x00000f00\r
102 TilePixelShHP 12, 7 @ #0x0000f000\r
103.endm\r
104\r
105\r
106@ TileSingleSh (r1=pdest, r2=pixels8, r3=pal) r4,r7: scratch, r0=sx; r12: helper pattern 0xf\r
107.macro TileSingleSh\r
108 tst r0, #1 @ not aligned?\r
109 mov r7, #0x00c000\r
110 orr r7, r7, #0xc0\r
111 ldrneb r4, [r1]\r
112 ldreqh r4, [r1]\r
113 orr r4, r4, r7\r
114 strneb r4, [r1], #1\r
115 streqh r4, [r1], #2\r
116 ldrh r4, [r1]\r
117 orr r4, r4, r7\r
118 strh r4, [r1], #2\r
119 ldrh r4, [r1]\r
120 orr r4, r4, r7\r
121 strh r4, [r1], #2\r
122 ldrh r4, [r1]\r
123 orr r4, r4, r7\r
124 strh r4, [r1], #2\r
125 ldrneb r4, [r1]\r
126 orr r4, r4, r7\r
127 strneb r4, [r1], #1\r
128.endm\r
129\r
130@ TileSingleHi (r1=pdest, r2=pixels8, r3=pal) r4,r7: scratch, r0=sx, r12: register with helper pattern 0xf\r
131.macro TileSingleHi\r
132 tst r1, #1 @ not aligned?\r
133 mov r7, #0x008000\r
134 orr r7, r7, #0x80\r
135 ldrneb r4, [r1], #1\r
136 ldreqh r4, [r1], #2 @ 1ci\r
137 ldrh r12, [r1], #2\r
138 bic r4, r4, r7, lsr #1\r
139 orr r4, r4, r7\r
140 strneb r4, [r1, #-3]\r
141 streqh r4, [r1, #-4]\r
142 ldrh r4, [r1], #2\r
143 bic r12, r12, r7, lsr #1\r
144 orr r12, r12, r7\r
145 strh r12, [r1, #-4]\r
146 ldrh r12, [r1], #2\r
147 bic r4, r4, r7, lsr #1\r
148 orr r4, r4, r7\r
149 strh r4, [r1, #-4]\r
150 ldrneb r4, [r1]\r
151 bic r12, r12, r7, lsr #1\r
152 orr r12, r12, r7\r
153 strh r12, [r1, #-2]\r
154 bicne r4, r4, r7, lsr #1\r
155 orrne r4, r4, r7\r
156 strneb r4, [r1], #1\r
157 mov r12, #0xf\r
158.endm\r
159\r
160.macro TileDoShGenPixel shift ofs\r
161.if \shift\r
162 ands r4, r12, r2, lsr #\shift\r
163.else\r
164 ands r4, r12, r2\r
165.endif\r
07abbab1 166 beq 0f\r
cc68a136 167 cmp r4, #0xe\r
bfa12428 168 ldrgeb r7, [r1,#\ofs]\r
169 orrlt r7, r3, r4 @ normal\r
07abbab1 170\r
bfa12428 171 bicge r7, r7, #0xc0\r
172 orrge r7, r7, r4, lsl #6\r
173 strb r7, [r1,#\ofs]\r
07abbab1 1740:\r
cc68a136 175.endm\r
176\r
177@ TileFlipSh (r1=pdest, r2=pixels8, r3=pal) r4,r7: scratch, r0=sx, r12: register with helper pattern 0xf\r
178.macro TileFlipSh\r
179 TileDoShGenPixel 16, 0 @ #0x000f0000\r
180 TileDoShGenPixel 20, 1 @ #0x00f00000\r
181 TileDoShGenPixel 24, 2 @ #0x0f000000\r
182 TileDoShGenPixel 28, 3 @ #0xf0000000\r
183 TileDoShGenPixel 0, 4 @ #0x0000000f\r
184 TileDoShGenPixel 4, 5 @ #0x000000f0\r
185 TileDoShGenPixel 8, 6 @ #0x00000f00\r
186 TileDoShGenPixel 12, 7 @ #0x0000f000\r
187.endm\r
188\r
189@ TileNormSh (r1=pdest, r2=pixels8, r3=pal) r4,r7: scratch, r0=sx, r12: register with helper pattern 0xf\r
190.macro TileNormSh\r
191 TileDoShGenPixel 12, 0 @ #0x0000f000\r
192 TileDoShGenPixel 8, 1 @ #0x00000f00\r
193 TileDoShGenPixel 4, 2 @ #0x000000f0\r
194 TileDoShGenPixel 0, 3 @ #0x0000000f\r
195 TileDoShGenPixel 28, 4 @ #0xf0000000\r
196 TileDoShGenPixel 24, 5 @ #0x0f000000\r
197 TileDoShGenPixel 20, 6 @ #0x00f00000\r
198 TileDoShGenPixel 16, 7 @ #0x000f0000\r
199.endm\r
200\r
bfa12428 201.macro TileDoShGenPixel_markop shift ofs\r
07abbab1 202.if \shift\r
bfa12428 203 ands r4, r12, r2, lsr #\shift\r
07abbab1 204.else\r
bfa12428 205 ands r4, r12, r2\r
07abbab1 206.endif\r
bfa12428 207 beq 0f\r
208 cmp r4, #0xe\r
209 ldrgeb r4, [r1,#\ofs]\r
210 orrlt r4, r3, r4\r
e54507e8 211 orrge r4, r4, #0x80\r
bfa12428 212 strb r4, [r1,#\ofs]\r
2130:\r
07abbab1 214.endm\r
215\r
e54507e8 216.macro TileFlipSh_markop\r
bfa12428 217 TileDoShGenPixel_markop 16, 0 @ #0x000f0000\r
218 TileDoShGenPixel_markop 20, 1 @ #0x00f00000\r
219 TileDoShGenPixel_markop 24, 2 @ #0x0f000000\r
220 TileDoShGenPixel_markop 28, 3 @ #0xf0000000\r
221 TileDoShGenPixel_markop 0, 4 @ #0x0000000f\r
222 TileDoShGenPixel_markop 4, 5 @ #0x000000f0\r
223 TileDoShGenPixel_markop 8, 6 @ #0x00000f00\r
224 TileDoShGenPixel_markop 12, 7 @ #0x0000f000\r
07abbab1 225.endm\r
226\r
e54507e8 227.macro TileNormSh_markop\r
bfa12428 228 TileDoShGenPixel_markop 12, 0 @ #0x0000f000\r
229 TileDoShGenPixel_markop 8, 1 @ #0x00000f00\r
230 TileDoShGenPixel_markop 4, 2 @ #0x000000f0\r
231 TileDoShGenPixel_markop 0, 3 @ #0x0000000f\r
232 TileDoShGenPixel_markop 28, 4 @ #0xf0000000\r
233 TileDoShGenPixel_markop 24, 5 @ #0x0f000000\r
234 TileDoShGenPixel_markop 20, 6 @ #0x00f00000\r
235 TileDoShGenPixel_markop 16, 7 @ #0x000f0000\r
07abbab1 236.endm\r
237\r
238.macro TileDoShGenPixel_onlyop_lp shift ofs\r
239.if \shift\r
240 ands r7, r12, r2, lsr #\shift\r
241.else\r
242 ands r7, r12, r2\r
243.endif\r
244 ldrneb r4, [r1,#\ofs]\r
07abbab1 245 cmp r7, #0xe\r
bfa12428 246 blt 0f\r
247\r
248 tst r4, #0xc0\r
249 bicne r4, r4, #0xc0\r
250 orrne r4, r4, r7, lsl #6\r
251 strneb r4, [r1,#\ofs]\r
07abbab1 2520:\r
253.endm\r
254\r
255.macro TileFlipSh_onlyop_lp\r
256 TileDoShGenPixel_onlyop_lp 16, 0 @ #0x000f0000\r
257 TileDoShGenPixel_onlyop_lp 20, 1 @ #0x00f00000\r
258 TileDoShGenPixel_onlyop_lp 24, 2 @ #0x0f000000\r
259 TileDoShGenPixel_onlyop_lp 28, 3 @ #0xf0000000\r
260 TileDoShGenPixel_onlyop_lp 0, 4 @ #0x0000000f\r
261 TileDoShGenPixel_onlyop_lp 4, 5 @ #0x000000f0\r
262 TileDoShGenPixel_onlyop_lp 8, 6 @ #0x00000f00\r
263 TileDoShGenPixel_onlyop_lp 12, 7 @ #0x0000f000\r
264.endm\r
265\r
266.macro TileNormSh_onlyop_lp\r
267 TileDoShGenPixel_onlyop_lp 12, 0 @ #0x0000f000\r
268 TileDoShGenPixel_onlyop_lp 8, 1 @ #0x00000f00\r
269 TileDoShGenPixel_onlyop_lp 4, 2 @ #0x000000f0\r
270 TileDoShGenPixel_onlyop_lp 0, 3 @ #0x0000000f\r
271 TileDoShGenPixel_onlyop_lp 28, 4 @ #0xf0000000\r
272 TileDoShGenPixel_onlyop_lp 24, 5 @ #0x0f000000\r
273 TileDoShGenPixel_onlyop_lp 20, 6 @ #0x00f00000\r
274 TileDoShGenPixel_onlyop_lp 16, 7 @ #0x000f0000\r
275.endm\r
276\r
cc68a136 277\r
278@ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r
279\r
280@ struct TileStrip\r
281@ {\r
282@ int nametab; // 0x00\r
283@ int line; // 0x04\r
284@ int hscroll; // 0x08\r
285@ int xmask; // 0x0C\r
286@ int *hc; // 0x10 (pointer to cache buffer)\r
287@ int cells; // 0x14\r
288@ };\r
289\r
83c093a4 290@ void DrawLayer(int plane_sh, int *hcache, int cellskip, int maxcells);\r
cc68a136 291\r
83c093a4 292.global DrawLayer\r
cc68a136 293\r
294DrawLayer:\r
295 stmfd sp!, {r4-r11,lr}\r
296\r
297 ldr r11, =(Pico+0x22228) @ Pico.video\r
298 mov r8, #1\r
299\r
83c093a4 300 ldrb r7, [r11, #16] @ ??vv??hh\r
cc68a136 301\r
302 mov r6, r1 @ hcache\r
83c093a4 303 orr r9, r3, r0, lsl #30\r
304 orr r9, r9, r2, lsl #8 @ r9=sh[31]|cellskip[15:8]|maxcells[7:0] (tmp)\r
cc68a136 305\r
306 mov r1, r7, lsl #4\r
307 orr r1, r1, #0x00ff\r
308\r
309 and r10, r7, #3\r
310 cmp r10, #1\r
311 biclt r1, r1, #0xfc00\r
312 biceq r1, r1, #0xfe00\r
313 bicgt r1, r1, #0xff00 @ r1=ymask=(height<<8)|0xff; ...; // Y Mask in pixels\r
314\r
315 add r10, r10, #5\r
316 cmp r10, #7\r
317 subge r10, r10, #1 @ r10=shift[width] (5,6,6,7)\r
318\r
319 @ calculate xmask:\r
320 mov r5, r8, lsl r10\r
321 sub r5, r5, #1 @ r5=xmask\r
322\r
323 @ Find name table:\r
83c093a4 324 ands r0, r0, #1\r
cc68a136 325 ldreqb r12, [r11, #2]\r
326 ldrneb r12, [r11, #4]\r
327\r
b6d7ac70 328 ldr r2, =DrawScanline @ trying to make good use of pipeline here\r
cc68a136 329 ldr lr, =(Pico+0x10000) @ lr=Pico.vram\r
330\r
331 moveq r12, r12, lsl #10\r
332 movne r12, r12, lsl #13\r
333 and r12, r12, #(7<<13) @ r12=(ts->nametab<<1) (halfword compliant)\r
334\r
335 ldrh r8, [r11, #12]\r
336 ldrb r7, [r11, #11]\r
337 ldr r2, [r2]\r
338\r
339 mov r4, r8, lsr #8 @ pvid->reg[13]\r
340 mov r4, r4, lsl #10 @ htab=pvid->reg[13]<<9; (halfwords)\r
341 tst r7, #2\r
b6d7ac70 342 addne r4, r4, r2, lsl #2 @ htab+=DrawScanline<<1; // Offset by line\r
cc68a136 343 tst r7, #1\r
344 biceq r4, r4, #0x1f @ htab&=~0xf; // Offset by tile\r
345 add r4, r4, r0, lsl #1 @ htab+=plane\r
346 bic r4, r4, #0x00ff0000 @ just in case\r
347 ldrh r3, [lr, r4] @ r3=hscroll\r
348\r
349 tst r7, #4\r
350 bne .DrawStrip_vsscroll\r
351\r
352 @ Get vertical scroll value:\r
353 add r7, lr, #0x012000\r
354 add r7, r7, #0x000180 @ r7=Pico.vsram (Pico+0x22180)\r
355 ldr r7, [r7]\r
356\r
357 tst r8, #2\r
358 tstne r8, #4\r
359 bne .DrawStrip_interlace\r
360\r
361 tst r0, r0\r
362 movne r7, r7, lsr #16\r
363\r
364 @ Find the line in the name table\r
365 add r2, r2, r7\r
366 and r2, r2, r1\r
367 mov r4, r2, lsr #3\r
368 add r10, r10, #1 @ shift[width]++\r
369 add r12, r12, r4, lsl r10 @ nametab+=(ts.line>>3)<<shift[width];\r
370\r
371 @ ldmia r0, {r1,r2,r3,r5,r6,r9} @ r2=line, r3=ts->hscroll, r5=ts->xmask, r6=ts->hc, r9=ts->cells\r
cc68a136 372\r
373 and r10,r2, #7\r
374 mov r10,r10, lsl #1 @ r10=ty=(ts->line&7)<<1;\r
375 orr r10,r10, r9, lsl #24\r
376\r
377 rsb r8, r3, #0\r
378 mov r8, r8, lsr #3 @ r8=tilex=(-ts->hscroll)>>3\r
379\r
380 sub r1, r3, #1\r
381 and r1, r1, #7\r
382 add r7, r1, #1 @ r7=dx=((ts->hscroll-1)&7)+1\r
383\r
384 tst r9, #1<<31\r
385 mov r3, #0\r
740da8c6 386 orrne r10,r10, #1<<23 @ r10=(cells<<24|sh<<23|hi_not_empty<<22|had_output<<21|ty)\r
cc68a136 387 movne r3, #0x40 @ default to shadowed pal on sh mode\r
388\r
cc68a136 389 cmp r7, #8\r
390 addne r10,r10, #0x01000000 @ we will loop cells+1 times if there is scroll\r
391\r
83c093a4 392 and r9, r9, #0xff00\r
393 add r8, r8, r9, lsr #8 @ tilex+=cellskip\r
394 add r7, r7, r9, lsr #5 @ dx+=cellskip<<3;\r
395 sub r10,r10,r9, lsl #16 @ cells-=cellskip\r
83c093a4 396\r
cc68a136 397 @ cache some stuff to avoid mem access\r
ea8c405f 398.if OVERRIDE_HIGHCOL\r
399 ldr r11,=HighCol\r
400 mov r0, #0xf\r
401 ldr r11,[r11]\r
402.else\r
cc68a136 403 ldr r11,=HighCol\r
404 mov r0, #0xf\r
ea8c405f 405.endif\r
406\r
407 mvn r9, #0 @ r9=prevcode=-1\r
cc68a136 408 add r1, r11, r7 @ r1=pdest\r
409\r
410\r
411 @ r4 & r7 are scratch in this loop\r
412.dsloop_subr1:\r
413 sub r1, r1, #8\r
414.dsloop: @ 40-41 times\r
415 subs r10,r10, #0x01000000\r
416 bmi .dsloop_exit\r
417\r
418.dsloop_enter:\r
419 and r7, r5, r8\r
420 add r7, lr, r7, lsl #1 @ Pico.vram+((tilex&ts->xmask) as halfwords)\r
421 ldrh r7, [r7, r12] @ r7=code (int, but from unsigned, no sign extend)\r
422\r
423 add r1, r1, #8\r
424 add r8, r8, #1\r
425\r
426 tst r7, #0x8000\r
427 bne .DrawStrip_hiprio\r
428\r
429 cmp r7, r9\r
430 beq .DrawStrip_samecode @ we know stuff about this tile already\r
431\r
432 mov r9, r7 @ remember code\r
740da8c6 433 orr r10, r10, #1<<21 @ seen non hi-prio tile\r
cc68a136 434\r
435 movs r2, r9, lsl #20 @ if (code&0x1000)\r
436 mov r2, r2, lsl #1\r
437 add r2, r2, r10, lsl #17\r
438 mov r2, r2, lsr #17\r
439 eorcs r2, r2, #0x0e @ if (code&0x1000) addr^=0xe;\r
440\r
441 ldr r2, [lr, r2, lsl #1] @ pack=*(unsigned int *)(Pico.vram+addr); // Get 8 pixels\r
442\r
443 bic r7, r3, #0x3f\r
444 and r3, r9, #0x6000\r
445 add r3, r7, r3, lsr #9 @ r3=pal=((code&0x6000)>>9);\r
446\r
447.DrawStrip_samecode:\r
448 tst r2, r2\r
449 beq .dsloop @ tileline blank\r
450\r
451 cmp r2, r2, ror #4\r
452 beq .DrawStrip_SingleColor @ tileline singlecolor \r
453\r
454 tst r9, #0x0800\r
7a7c6476 455 bne .DrawStrip_TileFlip\r
cc68a136 456\r
457 @ (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r0: helper pattern\r
cc68a136 458.DrawStrip_TileNorm:\r
459 TileNorm r0\r
460 b .dsloop\r
461\r
7a7c6476 462.DrawStrip_TileFlip:\r
463 TileFlip r0\r
464 b .dsloop\r
465\r
cc68a136 466.DrawStrip_SingleColor:\r
467 and r4, r2, #0xf\r
468 orr r4, r3, r4\r
469 orr r4, r4, r4, lsl #8\r
470 tst r1, #1 @ not aligned?\r
471 strneb r4, [r1], #1\r
472 streqh r4, [r1], #2\r
473 strh r4, [r1], #2\r
474 strh r4, [r1], #2\r
475 strh r4, [r1], #2\r
476 strneb r4, [r1], #1 @ have a remaining unaligned pixel?\r
477 b .dsloop_subr1\r
478\r
cc68a136 479.DrawStrip_hiprio_maybempt:\r
480 cmp r7, r9\r
481 beq .dsloop @ must've been empty, otherwise we wouldn't get here\r
482 movs r2, r7, lsl #20 @ if (code&0x1000)\r
483 mov r2, r2, lsl #1\r
484 add r2, r2, r10, lsl #17\r
485 mov r2, r2, lsr #17\r
486 eorcs r2, r2, #0x0e @ if (code&0x1000) addr^=0xe;\r
487 ldr r2, [lr, r2, lsl #1] @ pack=*(unsigned int *)(Pico.vram+addr); // Get 8 pixels\r
488 mov r9, r7 @ remember code\r
489 tst r2, r2\r
740da8c6 490 beq .dsloop\r
491 orr r10, r10, #1<<22\r
492\r
493.DrawStrip_hiprio:\r
494 tst r10, #0x00c00000\r
495 beq .DrawStrip_hiprio_maybempt\r
496 sub r0, r1, r11\r
497 orr r7, r7, r0, lsl #16\r
498 orr r7, r7, r10, lsl #25 @ (ty<<25)\r
499 tst r7, #0x1000\r
500 eorne r7, r7, #7<<26 @ if(code&0x1000) cval^=7<<26;\r
501 str r7, [r6], #4 @ cache hi priority tile\r
502 mov r0, #0xf\r
cc68a136 503 b .dsloop\r
504\r
505.dsloop_exit:\r
740da8c6 506 tst r10, #1<<21 @ seen non hi-prio tile\r
507 ldreq r1, =rendstatus\r
cc68a136 508 mov r0, #0\r
740da8c6 509 ldreq r2, [r1]\r
cc68a136 510 str r0, [r6] @ terminate the cache list\r
283fec1b 511 orreq r2, r2, #PDRAW_PLANE_HI_PRIO @ had a layer with all hi-prio tiles\r
740da8c6 512 streq r2, [r1]\r
cc68a136 513\r
514 ldmfd sp!, {r4-r11,lr}\r
515 bx lr\r
516\r
6d7acf9e 517@ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r
cc68a136 518\r
519.DrawStrip_vsscroll:\r
6d7acf9e 520 rsb r8, r3, #0\r
521 mov r8, r8, lsr #3 @ r8=tilex=(-ts->hscroll)>>3\r
7b802576 522 bic r8, r8, #0x3fc00000\r
523 orr r8, r8, r5, lsl #25 @ r8=(xmask[31:25]|had_output[24]|tilex[21:0])\r
cc68a136 524\r
b6d7ac70 525 ldr r4, =DrawScanline\r
6d7acf9e 526 orr r5, r1, r10, lsl #24\r
527 ldr r4, [r4]\r
528 sub r1, r3, #1\r
529 orr r5, r5, r4, lsl #16 @ r5=(shift_width[31:24]|scanline[23:16]|ymask[15:0])\r
530 and r1, r1, #7\r
531 add r7, r1, #1 @ r7=dx=((ts->hscroll-1)&7)+1\r
cc68a136 532\r
6d7acf9e 533 mov r10,r9, lsl #16\r
83c093a4 534 tst r0, #1\r
6d7acf9e 535 orrne r10,r10, #0x8000\r
536 tst r9, #1<<31\r
537 mov r3, #0\r
538 orr r10,r10, #0xff000000 @ will be adjusted on entering loop\r
83c093a4 539 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
6d7acf9e 540 movne r3, #0x40 @ default to shadowed pal on sh mode\r
cc68a136 541\r
83c093a4 542 cmp r7, #8\r
543 subne r10,r10, #0x01000000 @ have hscroll, start with negative cell\r
544\r
545 and r9, r9, #0xff00\r
546 add r8, r8, r9, lsr #8 @ tilex+=cellskip\r
547 add r7, r7, r9, lsr #5 @ dx+=cellskip<<3;\r
548 add r10,r10,r9, lsl #16 @ cell+=cellskip\r
cc68a136 549\r
6d7acf9e 550 @ cache some stuff to avoid mem access\r
ea8c405f 551.if OVERRIDE_HIGHCOL\r
6d7acf9e 552 ldr r11,=HighCol\r
553 mov r0, #0xf\r
ea8c405f 554 ldr r11,[r11]\r
555.else\r
556 ldr r11,=HighCol\r
557 mov r0, #0xf\r
558.endif\r
6d7acf9e 559\r
ea8c405f 560 mvn r9, #0 @ r9=prevcode=-1\r
561 add r1, r11, r7 @ r1=pdest\r
6d7acf9e 562\r
563 @ r4 & r7 are scratch in this loop\r
564.dsloop_vs_subr1:\r
565 sub r1, r1, #8\r
566.dsloop_vs: @ 40-41 times\r
567 add r10,r10, #0x01000000\r
568 and r4, r10, #0x003f0000\r
569 cmp r4, r10, asr #8\r
740da8c6 570 ble .dsloop_vs_exit\r
6d7acf9e 571\r
572 @ calc offset and read tileline code to r7, also calc ty\r
573 add r7, lr, #0x012000\r
574 add r7, r7, #0x000180 @ r7=Pico.vsram (Pico+0x22180)\r
575 add r7, r7, r10,asr #23 @ vsram + ((cell&~1)<<1)\r
576 bic r7, r7, #3\r
577 tst r10,#0x8000 @ plane1?\r
578 addne r7, r7, #2\r
579 ldrh r7, [r7] @ r7=vscroll\r
580\r
581 bic r10,r10,#0xff @ clear old ty\r
7b802576 582 and r4, r5, #0xff0000 @ scanline\r
583 add r4, r4, r7, lsl #16 @ ... += vscroll\r
584 and r4, r4, r5, lsl #16 @ ... &= ymask\r
6d7acf9e 585 and r7, r4, #0x70000\r
586 orr r10,r10,r7, lsr #15 @ new ty\r
587\r
588 mov r4, r4, lsr #19\r
589 mov r7, r5, lsr #24\r
590 mov r4, r4, lsl r7 @ nametabadd\r
591\r
592 and r7, r8, r8, lsr #25\r
593 add r7, lr, r7, lsl #1 @ Pico.vram+((tilex&ts->xmask) as halfwords)\r
594 add r7, r7, r4, lsl #1\r
595 ldrh r7, [r7, r12] @ r7=code (int, but from unsigned, no sign extend)\r
596\r
597 add r1, r1, #8\r
598 add r8, r8, #1\r
599\r
600 tst r7, #0x8000\r
601 bne .DrawStrip_vs_hiprio\r
602\r
603 cmp r7, r9\r
604 beq .DrawStrip_vs_samecode @ we know stuff about this tile already\r
605\r
606 mov r9, r7 @ remember code\r
7b802576 607 orr r8, r8, #(1<<24)@ seen non hi-prio tile\r
6d7acf9e 608\r
609 movs r2, r9, lsl #20 @ if (code&0x1000)\r
610 mov r2, r2, lsl #1\r
611 add r2, r2, r10, lsl #17\r
612 mov r2, r2, lsr #17\r
613 eorcs r2, r2, #0x0e @ if (code&0x1000) addr^=0xe;\r
614\r
615 ldr r2, [lr, r2, lsl #1] @ pack=*(unsigned int *)(Pico.vram+addr); // Get 8 pixels\r
616\r
617 bic r7, r3, #0x3f\r
618 and r3, r9, #0x6000\r
619 add r3, r7, r3, lsr #9 @ r3=pal=((code&0x6000)>>9);\r
620\r
621.DrawStrip_vs_samecode:\r
622 tst r2, r2\r
623 beq .dsloop_vs @ tileline blank\r
624\r
625 cmp r2, r2, ror #4\r
626 beq .DrawStrip_vs_SingleColor @ tileline singlecolor \r
627\r
628 tst r9, #0x0800\r
7a7c6476 629 bne .DrawStrip_vs_TileFlip\r
6d7acf9e 630\r
631 @ (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r0: helper pattern\r
6d7acf9e 632.DrawStrip_vs_TileNorm:\r
633 TileNorm r0\r
634 b .dsloop_vs\r
635\r
7a7c6476 636.DrawStrip_vs_TileFlip:\r
637 TileFlip r0\r
638 b .dsloop_vs\r
639\r
6d7acf9e 640.DrawStrip_vs_SingleColor:\r
641 and r4, r2, #0xf\r
642 orr r4, r3, r4\r
643 orr r4, r4, r4, lsl #8\r
644 tst r1, #1 @ not aligned?\r
645 strneb r4, [r1], #1\r
646 streqh r4, [r1], #2\r
647 strh r4, [r1], #2\r
648 strh r4, [r1], #2\r
649 strh r4, [r1], #2\r
650 strneb r4, [r1], #1 @ have a remaining unaligned pixel?\r
651 b .dsloop_vs_subr1\r
652\r
653.DrawStrip_vs_hiprio:\r
654 tst r10, #0x00c00000\r
655 beq .DrawStrip_vs_hiprio_maybempt\r
656 sub r0, r1, r11\r
657 orr r7, r7, r0, lsl #16\r
658 orr r7, r7, r10, lsl #25 @ (ty<<25)\r
659 tst r7, #0x1000\r
660 eorne r7, r7, #7<<26 @ if(code&0x1000) cval^=7<<26;\r
661 str r7, [r6], #4 @ cache hi priority tile\r
662 mov r0, #0xf\r
663 b .dsloop_vs\r
664\r
665.DrawStrip_vs_hiprio_maybempt:\r
666 cmp r7, r9\r
667 beq .dsloop_vs @ must've been empty, otherwise we wouldn't get here\r
668 movs r2, r7, lsl #20 @ if (code&0x1000)\r
669 mov r2, r2, lsl #1\r
670 add r2, r2, r10, lsl #17\r
671 mov r2, r2, lsr #17\r
672 eorcs r2, r2, #0x0e @ if (code&0x1000) addr^=0xe;\r
673 ldr r2, [lr, r2, lsl #1] @ pack=*(unsigned int *)(Pico.vram+addr); // Get 8 pixels\r
674 mov r9, r7 @ remember code\r
675 tst r2, r2\r
676 orrne r10, r10, #1<<22\r
677 bne .DrawStrip_vs_hiprio\r
678 b .dsloop_vs\r
679\r
740da8c6 680.dsloop_vs_exit:\r
7b802576 681 tst r8, #(1<<24) @ seen non hi-prio tile\r
740da8c6 682 ldreq r1, =rendstatus\r
683 mov r0, #0\r
684 ldreq r2, [r1]\r
685 str r0, [r6] @ terminate the cache list\r
283fec1b 686 orreq r2, r2, #PDRAW_PLANE_HI_PRIO @ had a layer with all hi-prio tiles\r
740da8c6 687 streq r2, [r1]\r
688\r
689 ldmfd sp!, {r4-r11,lr}\r
690 bx lr\r
691\r
6d7acf9e 692\r
693@ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r
cc68a136 694\r
695@ interlace mode 2? Sonic 2?\r
696.DrawStrip_interlace:\r
697 tst r0, r0\r
698 moveq r7, r7, lsl #21\r
699 movne r7, r7, lsl #5\r
700\r
701 @ Find the line in the name table\r
b6d7ac70 702 add r2, r7, r2, lsl #22 @ r2=(vscroll+(DrawScanline<<1))<<21 (11 bits);\r
cc68a136 703 orr r1, r1, #0x80000000\r
704 and r2, r2, r1, ror #10 @ &((ymask<<1)|1)<<21;\r
705 mov r2, r2, lsr #21\r
706 mov r4, r2, lsr #4\r
707 mov r12, r12, lsr #1 @ halfwords\r
708 add r0, r12, r4, lsl r10 @ nametab+=(ts.line>>4)<<shift[width];\r
709 and r9, r9, #0xff\r
710\r
711 sub sp, sp, #6*4\r
712 stmia sp, {r0,r2,r3,r5,r6,r9}\r
713\r
714 mov r0, sp\r
715 bl DrawStripInterlace @ struct TileStrip *ts\r
716\r
717 add sp, sp, #6*4\r
718 ldmfd sp!, {r4-r11,lr}\r
719 bx lr\r
720\r
721.pool\r
722\r
723@ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r
724\r
725\r
726.global BackFill @ int reg7, int sh\r
727\r
728BackFill:\r
729 stmfd sp!, {r4-r9,lr}\r
730\r
ea8c405f 731.if OVERRIDE_HIGHCOL\r
732 ldr lr, =HighCol\r
733 mov r0, r0, lsl #26\r
734 ldr lr, [lr]\r
735 mov r0, r0, lsr #26\r
736 add lr, lr, #8\r
737.else\r
cc68a136 738 ldr lr, =(HighCol+8)\r
cc68a136 739 mov r0, r0, lsl #26\r
740 mov r0, r0, lsr #26\r
ea8c405f 741.endif\r
742\r
cc68a136 743 orr r0, r0, r1, lsl #6\r
744 orr r0, r0, r0, lsl #8\r
745 orr r0, r0, r0, lsl #16\r
746\r
747 mov r1, r0\r
748 mov r2, r0\r
749 mov r3, r0\r
750 mov r4, r0\r
751 mov r5, r0\r
752 mov r6, r0\r
753 mov r7, r0\r
754\r
755 @ go go go!\r
756 stmia lr!, {r0-r7} @ 10*8*4\r
757 stmia lr!, {r0-r7}\r
758 stmia lr!, {r0-r7}\r
759 stmia lr!, {r0-r7}\r
760 stmia lr!, {r0-r7}\r
761 stmia lr!, {r0-r7}\r
762 stmia lr!, {r0-r7}\r
763 stmia lr!, {r0-r7}\r
764 stmia lr!, {r0-r7}\r
765 stmia lr!, {r0-r7}\r
766\r
767 ldmfd sp!, {r4-r9,r12}\r
768 bx r12\r
769\r
770\r
771@ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r
772\r
773\r
7a7c6476 774.global DrawTilesFromCache @ int *hc, int sh, int rlim\r
cc68a136 775\r
776DrawTilesFromCache:\r
777 stmfd sp!, {r4-r8,r11,lr}\r
778\r
cc68a136 779 @ cache some stuff to avoid mem access\r
ea8c405f 780.if OVERRIDE_HIGHCOL\r
cc68a136 781 ldr r11,=HighCol\r
cc68a136 782 mov r12,#0xf\r
ea8c405f 783 ldr r11,[r11]\r
784.else\r
785 ldr r11,=HighCol\r
786 mov r12,#0xf\r
787.endif\r
788 ldr lr, =(Pico+0x10000) @ lr=Pico.vram\r
cc68a136 789\r
740da8c6 790 mvn r5, #0 @ r5=prevcode=-1\r
7a7c6476 791 ands r8, r1, #1\r
792 orr r8, r8, r2, lsl #1\r
740da8c6 793 bne .dtfc_check_rendflags\r
794\r
cc68a136 795 @ scratch: r4, r7\r
796.dtfc_loop:\r
797 ldr r6, [r0], #4 @ read code\r
798 movs r1, r6, lsr #16 @ r1=dx;\r
799 ldmeqfd sp!, {r4-r8,r11,pc} @ dx is never zero, this must be a terminator, return\r
7a7c6476 800 bic r4, r1, #0xfe00\r
801 add r1, r11, r4 @ r1=pdest\r
cc68a136 802\r
803 mov r7, r6, lsl #16\r
804 cmp r5, r7, lsr #16\r
805 beq .dtfc_samecode @ if (code==prevcode)\r
806\r
807 mov r5, r7, lsr #16\r
808\r
809 mov r2, r5, lsl #21\r
810 mov r2, r2, lsr #17 @ r2=addr=(code&0x7ff)<<4;\r
811 add r2, r2, r6, lsr #25 @ addr+=ty\r
812\r
813 and r3, r5, #0x6000\r
814 mov r3, r3, lsr #9 @ r3=pal=((code&0x6000)>>9);\r
815\r
816 ldr r2, [lr, r2, lsl #1] @ pack=*(unsigned int *)(Pico.vram+addr); // Get 8 pixels\r
817\r
818.dtfc_samecode:\r
7a7c6476 819 rsbs r4, r4, r8, lsr #1\r
820 bmi .dtfc_cut_tile\r
821\r
822 tst r8, #1\r
cc68a136 823 bne .dtfc_shadow\r
824\r
825 tst r2, r2\r
826 beq .dtfc_loop\r
827\r
828 cmp r2, r2, ror #4\r
829 beq .dtfc_SingleColor @ tileline singlecolor \r
830\r
831 tst r5, #0x0800\r
7a7c6476 832 bne .dtfc_TileFlip\r
cc68a136 833\r
834 @ (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r12: helper pattern\r
cc68a136 835.dtfc_TileNorm:\r
836 TileNorm r12\r
837 b .dtfc_loop\r
838\r
7a7c6476 839.dtfc_TileFlip:\r
840 TileFlip r12\r
841 b .dtfc_loop\r
842\r
cc68a136 843.dtfc_SingleColor:\r
844 and r4, r2, #0xf\r
845 orr r4, r3, r4\r
846 orr r4, r4, r4, lsl #8\r
847 tst r1, #1 @ not aligned?\r
848 strneb r4, [r1], #1\r
849 streqh r4, [r1], #2\r
850 strh r4, [r1], #2\r
851 strh r4, [r1], #2\r
852 strh r4, [r1], #2\r
853 strneb r4, [r1], #1 @ have a remaining unaligned pixel?\r
854 b .dtfc_loop\r
855\r
856.dtfc_shadow:\r
857 tst r2, r2\r
858 beq .dtfc_shadow_blank\r
859\r
860 cmp r2, r2, ror #4\r
861 beq .dtfc_SingleColor @ tileline singlecolor \r
862\r
863 tst r5, #0x0800\r
7a7c6476 864 bne .dtfc_TileFlipShHP\r
cc68a136 865\r
866 @ (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r12: helper pattern\r
cc68a136 867.dtfc_TileNormShHP:\r
868 TileNormShHP\r
869 b .dtfc_loop\r
870\r
7a7c6476 871.dtfc_TileFlipShHP:\r
872 TileFlipShHP\r
873 b .dtfc_loop\r
874\r
cc68a136 875.dtfc_shadow_blank:\r
07abbab1 876 tst r1, #1\r
877 ldrneb r4, [r1]\r
bfa12428 878 mov r6, #0xbf\r
879 and r4, r4, #0xbf\r
07abbab1 880 strneb r4, [r1], #1\r
881 ldrh r4, [r1]\r
882 orr r6, r6, r6, lsl #8\r
883 and r4, r4, r6\r
884 strh r4, [r1], #2\r
885 ldrh r4, [r1]\r
886 and r4, r4, r6\r
887 strh r4, [r1], #2\r
888 ldrh r4, [r1]\r
889 and r4, r4, r6\r
890 strh r4, [r1], #2\r
891 ldrh r4, [r1]\r
892 and r4, r4, r6\r
893 streqh r4, [r1]\r
894 strneb r4, [r1]\r
cc68a136 895 b .dtfc_loop\r
896\r
7a7c6476 897.dtfc_cut_tile:\r
898 add r4, r4, #7 @ 0-6\r
899 mov r4, r4, lsl #2\r
900 mov r12,#0xf<<28\r
901 mov r12,r12,asr r4\r
902 mov r2, r2, ror #16\r
903 tst r5, #0x0800 @ flipped?\r
904 mvnne r12,r12\r
905 and r2, r2, r12\r
906 mov r2, r2, ror #16\r
907 mov r12,#0xf\r
908 tst r8, #1\r
909 bne .dtfc_shadow\r
910 tst r2, r2\r
911 beq .dtfc_loop\r
912 tst r5, #0x0800\r
913 beq .dtfc_TileNorm\r
914 b .dtfc_TileFlip\r
915\r
740da8c6 916@ check if we have detected layer covered with hi-prio tiles:\r
917.dtfc_check_rendflags:\r
918 ldr r1, =rendstatus\r
919 ldr r2, [r1]\r
283fec1b 920 tst r2, #(PDRAW_PLANE_HI_PRIO|PDRAW_SHHI_DONE)\r
740da8c6 921 beq .dtfc_loop\r
7a7c6476 922 bic r8, r8, #1 @ sh/hi mode off\r
283fec1b 923 tst r2, #PDRAW_SHHI_DONE\r
740da8c6 924 bne .dtfc_loop @ already processed\r
283fec1b 925 orr r2, r2, #PDRAW_SHHI_DONE\r
740da8c6 926 str r2, [r1]\r
927\r
928 add r1, r11,#8\r
07abbab1 929 mov r3, #320/4/4\r
bfa12428 930 mov r6, #0xbf\r
740da8c6 931 orr r6, r6, r6, lsl #8\r
932 orr r6, r6, r6, lsl #16\r
933.dtfc_loop_shprep:\r
07abbab1 934 ldmia r1, {r2,r4,r5,r7}\r
740da8c6 935 subs r3, r3, #1\r
07abbab1 936 and r2, r2, r6\r
937 and r4, r4, r6\r
938 and r5, r5, r6\r
939 and r7, r7, r6\r
940 stmia r1!,{r2,r4,r5,r7}\r
941 bne .dtfc_loop_shprep\r
942\r
943 mvn r5, #0 @ r5=prevcode=-1\r
944 b .dtfc_loop\r
740da8c6 945\r
cc68a136 946.pool\r
947\r
948@ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r
949\r
950\r
fbc65db7 951.global DrawSpritesSHi @ unsigned char *sprited\r
e352c3af 952\r
953DrawSpritesSHi:\r
fbc65db7 954 ldr r3, [r0]\r
955 mov r12,#0xff\r
e352c3af 956 ands r3, r3, #0x7f\r
957 bxeq lr\r
cc68a136 958\r
cc68a136 959 stmfd sp!, {r4-r11,lr}\r
fbc65db7 960 strb r12,[r0,#2] @ set end marker\r
961 add r10,r0, #3 @ r10=HighLnSpr end\r
e352c3af 962 add r10,r10,r3 @ r10=HighLnSpr end\r
963\r
ea8c405f 964.if OVERRIDE_HIGHCOL\r
cc68a136 965 ldr r11,=HighCol\r
ea8c405f 966 mov r12,#0xf\r
967 ldr r11,[r11]\r
968.else\r
969 ldr r11,=HighCol\r
970 mov r12,#0xf\r
971.endif\r
cc68a136 972 ldr lr, =(Pico+0x10000) @ lr=Pico.vram\r
cc68a136 973\r
cc68a136 974\r
e352c3af 975DrawSpriteSHi:\r
976 @ draw next sprite\r
977 ldrb r0, [r10,#-1]!\r
978 ldr r1, =HighPreSpr\r
fbc65db7 979@ ldr r8, [sp, #-4]\r
e352c3af 980 cmp r0, #0xff\r
981 ldmeqfd sp!, {r4-r11,pc} @ end of list\r
982 and r0, r0, #0x7f\r
983 add r0, r1, r0, lsl #3\r
984\r
985 ldr r9, [r0, #4] @ sprite[1]\r
986 mov r2, r9, asr #16 @ r2=sx\r
cc68a136 987\r
e352c3af 988 mov r9, r9, lsl #16\r
989 mov r3, r9, lsr #31 @ priority\r
990 mov r9, r9, lsr #16\r
fbc65db7 991@ orr r9, r9, r8, lsl #31 @ r9=code|sh[31] @@ sh is always on here now\r
e352c3af 992 and r4, r9, #0x6000\r
993 orr r9, r9, r4, lsl #16\r
fbc65db7 994 orr r9, r9, #0x90000000 @ r9=scc1 ???? ... <code> (s=shadow/hilight, cc=pal)\r
e352c3af 995 cmp r12,r9, lsr #28 @ sh/hi with pal3?\r
e54507e8 996 cmpne r3, #1 @ if not, is it hi prio?\r
e352c3af 997 bne DrawSpriteSHi @ non-operator low sprite, already drawn\r
cc68a136 998\r
e352c3af 999 ldr r3, [r0] @ sprite[0]\r
1000 ldr r7, =DrawScanline\r
1001 mov r6, r3, lsr #28\r
1002 sub r6, r6, #1 @ r6=width-1 (inc later)\r
1003 mov r5, r3, lsr #24\r
1004 and r5, r5, #7 @ r5=height\r
cc68a136 1005\r
e352c3af 1006 mov r0, r3, lsl #16 @ r4=sy<<16 (tmp)\r
cc68a136 1007\r
e352c3af 1008 ldr r7, [r7]\r
1009 sub r7, r7, r0, asr #16 @ r7=row=DrawScanline-sy\r
cc68a136 1010\r
e352c3af 1011 tst r9, #0x1000\r
1012 movne r0, r5, lsl #3\r
1013 subne r0, r0, #1\r
1014 subne r7, r0, r7 @ if (code&0x1000) row=(height<<3)-1-row; // Flip Y\r
cc68a136 1015\r
e352c3af 1016 add r8, r9, r7, lsr #3 @ tile+=row>>3; // Tile number increases going down\r
1017 tst r9, #0x0800\r
1018 mlane r8, r5, r6, r8 @ if (code&0x0800) { tile+=delta*(width-1);\r
1019 rsbne r5, r5, #0 @ delta=-delta; } // r5=delta now\r
cc68a136 1020\r
e352c3af 1021 mov r8, r8, lsl #21\r
1022 mov r8, r8, lsr #17\r
1023 and r7, r7, #7\r
1024 add r8, r8, r7, lsl #1 @ tile+=(row&7)<<1; // Tile address\r
1025\r
1026 mov r5, r5, lsl #4 @ delta<<=4; // Delta of address\r
e352c3af 1027 mov r3, r4, lsr #9 @ r3=pal=((code>>9)&0x30);\r
1028\r
1029 add r6, r6, #1 @ inc now\r
1030 adds r0, r2, #0 @ mov sx to r0 and set ZV flags\r
1031 b .dsprShi_loop_enter\r
1032\r
1033.dsprShi_loop:\r
1034 subs r6, r6, #1 @ width--\r
1035 beq DrawSpriteSHi\r
1036 adds r0, r0, #8 @ sx+=8\r
1037 add r8, r8, r5 @ tile+=delta\r
1038\r
1039.dsprShi_loop_enter:\r
1040 ble .dsprShi_loop @ sx <= 0\r
cc68a136 1041 cmp r0, #328\r
e352c3af 1042 bge DrawSpriteSHi\r
cc68a136 1043\r
1044 mov r8, r8, lsl #17\r
e352c3af 1045 mov r8, r8, lsr #17 @ tile&=0x7fff; // Clip tile address\r
cc68a136 1046\r
e352c3af 1047 ldr r2, [lr, r8, lsl #1] @ pack=*(unsigned int *)(Pico.vram+addr); // Get 8 pixels\r
1048 add r1, r11, r0 @ r1=pdest\r
cc68a136 1049 tst r2, r2\r
e352c3af 1050 beq .dsprShi_loop\r
cc68a136 1051\r
e352c3af 1052 cmp r12, r9, lsr #28\r
1053 beq .dsprShi_shadow\r
cc68a136 1054\r
1055 cmp r2, r2, ror #4\r
e352c3af 1056 beq .dsprShi_SingleColor @ tileline singlecolor \r
cc68a136 1057\r
e352c3af 1058 tst r9, #0x0800\r
1059 bne .dsprShi_TileFlip\r
cc68a136 1060\r
e352c3af 1061 @ (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r12: helper pattern\r
1062@ scratch: r4, r7\r
1063.dsprShi_TileNorm:\r
cc68a136 1064 TileNorm r12\r
e352c3af 1065 b .dsprShi_loop\r
cc68a136 1066\r
e352c3af 1067.dsprShi_TileFlip:\r
7a7c6476 1068 TileFlip r12\r
e352c3af 1069 b .dsprShi_loop\r
7a7c6476 1070\r
e352c3af 1071.dsprShi_SingleColor:\r
cc68a136 1072 and r4, r2, #0xf\r
1073 orr r4, r3, r4\r
1074 orr r4, r4, r4, lsl #8\r
e352c3af 1075 tst r0, #1 @ not aligned?\r
cc68a136 1076 strneb r4, [r1], #1\r
1077 streqh r4, [r1], #2\r
1078 strh r4, [r1], #2\r
1079 strh r4, [r1], #2\r
1080 strh r4, [r1], #2\r
1081 strneb r4, [r1], #1\r
e352c3af 1082 b .dsprShi_loop\r
cc68a136 1083\r
e352c3af 1084.dsprShi_shadow:\r
1085 tst r9, #0x8000\r
1086 beq .dsprShi_shadow_lowpri\r
07abbab1 1087\r
cc68a136 1088 cmp r2, r2, ror #4\r
e352c3af 1089 beq .dsprShi_singlec_sh\r
cc68a136 1090\r
e352c3af 1091 tst r9, #0x0800\r
1092 bne .dsprShi_TileFlip_sh\r
cc68a136 1093\r
e54507e8 1094 @ (r1=pdest, r2=pixels8, r3=pal) r4, r7: scratch, r12: helper pattern\r
e352c3af 1095.dsprShi_TileNorm_sh:\r
cc68a136 1096 TileNormSh\r
e352c3af 1097 b .dsprShi_loop\r
cc68a136 1098\r
e352c3af 1099.dsprShi_TileFlip_sh:\r
7a7c6476 1100 TileFlipSh\r
e352c3af 1101 b .dsprShi_loop\r
7a7c6476 1102\r
e352c3af 1103.dsprShi_singlec_sh:\r
cc68a136 1104 cmp r2, #0xe0000000\r
e352c3af 1105 bcc .dsprShi_SingleColor @ normal singlecolor tileline (carry inverted in ARM)\r
cc68a136 1106 tst r2, #0x10000000\r
e352c3af 1107 bne .dsprShi_sh_sh\r
cc68a136 1108 TileSingleHi\r
e352c3af 1109 b .dsprShi_loop\r
cc68a136 1110\r
e352c3af 1111.dsprShi_sh_sh:\r
cc68a136 1112 TileSingleSh\r
e352c3af 1113 b .dsprShi_loop\r
cc68a136 1114\r
e352c3af 1115.dsprShi_shadow_lowpri:\r
1116 tst r9, #0x800\r
1117 bne .dsprShi_TileFlip_sh_lp\r
07abbab1 1118\r
e352c3af 1119.dsprShi_TileNorm_sh_lp:\r
07abbab1 1120 TileNormSh_onlyop_lp\r
e352c3af 1121 b .dsprShi_loop\r
07abbab1 1122\r
e352c3af 1123.dsprShi_TileFlip_sh_lp:\r
07abbab1 1124 TileFlipSh_onlyop_lp\r
e352c3af 1125 b .dsprShi_loop\r
07abbab1 1126\r
cc68a136 1127.pool\r
1128\r
1129@ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r
1130\r
fbc65db7 1131.global DrawAllSprites @ unsigned char *sprited, int prio, int sh\r
283fec1b 1132\r
1133DrawAllSprites:\r
1134 ldr r3, =rendstatus\r
fbc65db7 1135 orr r1, r2, r1, lsl #1\r
283fec1b 1136 ldr r12,[r3]\r
0fc0e241 1137 tst r12,#(PDRAW_DIRTY_SPRITES|PDRAW_SPRITES_MOVED)\r
283fec1b 1138 beq das_no_prep\r
fbc65db7 1139 stmfd sp!, {r0,r1,lr}\r
283fec1b 1140 and r0, r12,#PDRAW_DIRTY_SPRITES\r
0fc0e241 1141 bic r12,r12,#(PDRAW_DIRTY_SPRITES|PDRAW_SPRITES_MOVED)\r
283fec1b 1142 str r12,[r3]\r
1143 bl PrepareSprites\r
fbc65db7 1144 ldmfd sp!, {r0,r1,lr}\r
283fec1b 1145\r
1146das_no_prep:\r
fbc65db7 1147 ldr r3, [r0]\r
283fec1b 1148 ands r3, r3, #0x7f\r
283fec1b 1149 bxeq lr\r
1150\r
1151 @ time to do some real work\r
1152 stmfd sp!, {r4-r11,lr}\r
283fec1b 1153 mov r12,#0xff\r
fbc65db7 1154 strb r12,[r0,#2] @ set end marker\r
1155 add r10,r0, #3\r
283fec1b 1156 add r10,r10,r3 @ r10=HighLnSpr end\r
1157\r
97a7f774 1158 str r1, [sp, #-4] @ no calls after this point\r
283fec1b 1159\r
1160.if OVERRIDE_HIGHCOL\r
1161 ldr r11,=HighCol\r
1162 mov r12,#0xf\r
1163 ldr r11,[r11]\r
1164.else\r
1165 ldr r11,=HighCol\r
1166 mov r12,#0xf\r
1167.endif\r
1168 ldr lr, =(Pico+0x10000) @ lr=Pico.vram\r
1169\r
1170@ + 0 : hhhhvvvv ----hhvv yyyyyyyy yyyyyyyy // v, h: horiz. size\r
cc68a136 1171@ + 4 : xxxxxxxx xxxxxxxx pccvhnnn nnnnnnnn // x: x coord + 8\r
1172\r
283fec1b 1173DrawSprite: @ was: unsigned int *sprite, int sh, int acc_sprites\r
1174 @ draw next sprite\r
1175 ldrb r0, [r10,#-1]!\r
1176 ldr r1, =HighPreSpr\r
283fec1b 1177 ldr r8, [sp, #-4]\r
e352c3af 1178 mov r2, r0, lsr #7\r
1179 cmp r0, #0xff\r
1180 ldmeqfd sp!, {r4-r11,pc} @ end of list\r
97a7f774 1181 cmp r2, r8, lsr #1\r
e352c3af 1182 bne DrawSprite @ wrong priority\r
283fec1b 1183 and r0, r0, #0x7f\r
1184 add r0, r1, r0, lsl #3\r
cc68a136 1185\r
283fec1b 1186@ stmfd sp!, {r4-r9,r11,lr}\r
1187@ orr r8, r2, r1, lsl #4\r
cc68a136 1188\r
1189 ldr r3, [r0] @ sprite[0]\r
b6d7ac70 1190 ldr r7, =DrawScanline\r
cc68a136 1191 mov r6, r3, lsr #28\r
1192 sub r6, r6, #1 @ r6=width-1 (inc later)\r
1193 mov r5, r3, lsr #24\r
1194 and r5, r5, #7 @ r5=height\r
1195\r
1196 mov r4, r3, lsl #16 @ r4=sy<<16 (tmp)\r
1197\r
1198 ldr r7, [r7]\r
1199 ldr r9, [r0, #4]\r
b6d7ac70 1200 sub r7, r7, r4, asr #16 @ r7=row=DrawScanline-sy\r
cc68a136 1201\r
cc68a136 1202 mov r2, r9, asr #16 @ r2=sx\r
e5fa9817 1203 mov r9, r9, lsl #16\r
1204 mov r9, r9, lsr #16\r
97a7f774 1205 orr r9, r9, r8, lsl #31 @ r9=code|sh[31]\r
cc68a136 1206\r
1207 tst r9, #0x1000\r
1208 movne r4, r5, lsl #3\r
1209 subne r4, r4, #1\r
1210 subne r7, r4, r7 @ if (code&0x1000) row=(height<<3)-1-row; // Flip Y\r
1211\r
07abbab1 1212 add r8, r9, r7, lsr #3 @ tile+=row>>3; // Tile number increases going down\r
cc68a136 1213 tst r9, #0x0800\r
1214 mlane r8, r5, r6, r8 @ if (code&0x0800) { tile+=delta*(width-1);\r
1215 rsbne r5, r5, #0 @ delta=-delta; } // r5=delta now\r
1216\r
07abbab1 1217 mov r8, r8, lsl #21\r
1218 mov r8, r8, lsr #17\r
cc68a136 1219 and r7, r7, #7\r
1220 add r8, r8, r7, lsl #1 @ tile+=(row&7)<<1; // Tile address\r
1221\r
e5fa9817 1222.dspr_continue:\r
cc68a136 1223 @ cache some stuff to avoid mem access\r
cc68a136 1224 mov r5, r5, lsl #4 @ delta<<=4; // Delta of address\r
1225 and r4, r9, #0x6000\r
1226 orr r9, r9, r4, lsl #16\r
97a7f774 1227 orrs r9, r9, #0x10000000 @ r9=scc1 ???? ... <code> (s=shadow/hilight, cc=pal)\r
cc68a136 1228\r
cc68a136 1229 mov r3, r4, lsr #9 @ r3=pal=((code>>9)&0x30);\r
97a7f774 1230 orrmi r3, r3, #0x40 @ for sh/hi\r
cc68a136 1231\r
1232 add r6, r6, #1 @ inc now\r
1233 adds r0, r2, #0 @ mov sx to r0 and set ZV flags\r
1234 b .dspr_loop_enter\r
1235\r
1236.dspr_loop:\r
1237 subs r6, r6, #1 @ width--\r
283fec1b 1238 beq DrawSprite\r
cc68a136 1239 adds r0, r0, #8 @ sx+=8\r
1240 add r8, r8, r5 @ tile+=delta\r
1241\r
1242.dspr_loop_enter:\r
1243 ble .dspr_loop @ sx <= 0\r
1244 cmp r0, #328\r
283fec1b 1245 bge DrawSprite\r
cc68a136 1246\r
1247 mov r8, r8, lsl #17\r
1248 mov r8, r8, lsr #17 @ tile&=0x7fff; // Clip tile address\r
1249\r
1250 ldr r2, [lr, r8, lsl #1] @ pack=*(unsigned int *)(Pico.vram+addr); // Get 8 pixels\r
1251 add r1, r11, r0 @ r1=pdest\r
1252 tst r2, r2\r
1253 beq .dspr_loop\r
1254\r
1255 cmp r12, r9, lsr #28\r
1256 beq .dspr_shadow\r
1257\r
1258 cmp r2, r2, ror #4\r
1259 beq .dspr_SingleColor @ tileline singlecolor \r
1260\r
1261 tst r9, #0x0800\r
7a7c6476 1262 bne .dspr_TileFlip\r
cc68a136 1263\r
1264 @ (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r12: helper pattern\r
cc68a136 1265@ scratch: r4, r7\r
1266.dspr_TileNorm:\r
1267 TileNorm r12\r
1268 b .dspr_loop\r
1269\r
7a7c6476 1270.dspr_TileFlip:\r
1271 TileFlip r12\r
1272 b .dspr_loop\r
1273\r
07abbab1 1274.dspr_singlec_sh:\r
1275 cmp r2, #0xe0000000\r
e54507e8 1276 bcs .dspr_TileNorm_sh @ op. tileline, markop. XXX: maybe add a spec. handler?\r
07abbab1 1277\r
cc68a136 1278.dspr_SingleColor:\r
1279 and r4, r2, #0xf\r
1280 orr r4, r3, r4\r
1281 orr r4, r4, r4, lsl #8\r
1282 tst r0, #1 @ not aligned?\r
1283 strneb r4, [r1], #1\r
1284 streqh r4, [r1], #2\r
1285 strh r4, [r1], #2\r
1286 strh r4, [r1], #2\r
1287 strh r4, [r1], #2\r
1288 strneb r4, [r1], #1\r
1289 b .dspr_loop\r
1290\r
1291.dspr_shadow:\r
1292 cmp r2, r2, ror #4\r
1293 beq .dspr_singlec_sh\r
1294\r
1295 tst r9, #0x0800\r
7a7c6476 1296 bne .dspr_TileFlip_sh\r
cc68a136 1297\r
1298 @ (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r12: helper pattern\r
cc68a136 1299.dspr_TileNorm_sh:\r
e54507e8 1300 TileNormSh_markop\r
cc68a136 1301 b .dspr_loop\r
1302\r
7a7c6476 1303.dspr_TileFlip_sh:\r
e54507e8 1304 TileFlipSh_markop\r
cc68a136 1305 b .dspr_loop\r
1306\r
1307\r
cc68a136 1308@ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r
1309\r
1310.global DrawWindow @ int tstart, int tend, int prio, int sh // int *hcache\r
1311\r
1312DrawWindow:\r
1313 stmfd sp!, {r4-r11,lr}\r
1314\r
1315 ldr r11, =(Pico+0x22228) @ Pico.video\r
b6d7ac70 1316 ldr r10, =DrawScanline\r
cc68a136 1317 ldrb r12, [r11, #3] @ pvid->reg[3]\r
1318\r
1319 ldr r10, [r10]\r
1320 ldr r4, [r11, #12]\r
1321 mov r5, r10, lsr #3\r
1322 and r10, r10, #7\r
1323 mov r10, r10, lsl #1 @ r10=ty\r
1324\r
1325 mov r12, r12, lsl #10\r
1326\r
1327 tst r4, #1 @ 40 cell mode?\r
1328 andne r12, r12, #0xf000 @ 0x3c<<10\r
1329 andeq r12, r12, #0xf800\r
1330 addne r12, r12, r5, lsl #7\r
1331 addeq r12, r12, r5, lsl #6 @ nametab\r
1332 add r12, r12, r0, lsl #2 @ +starttile\r
1333\r
1334 ldr r6, =rendstatus\r
0fc0e241 1335 ldr lr, =(Pico+0x10000) @ lr=Pico.vram\r
07abbab1 1336 ldr r6, [r6]\r
cc68a136 1337\r
1338 @ fetch the first code now\r
1339 ldrh r7, [lr, r12]\r
1340\r
283fec1b 1341 ands r6, r6, #PDRAW_WND_DIFF_PRIO\r
cc68a136 1342 orr r6, r6, r2\r
cc68a136 1343\r
7292c709 1344 eoreq r8, r2, r7, lsr #15 @ do prio bits differ?\r
1345 cmpeq r8, #1\r
1346 ldmeqfd sp!, {r4-r11,pc} @ yes, assume that whole window uses same priority\r
cc68a136 1347\r
cc68a136 1348 orr r6, r6, r3, lsl #8 @ shadow mode\r
1349\r
1350 sub r8, r1, r0\r
cc68a136 1351\r
1352 @ cache some stuff to avoid mem access\r
ea8c405f 1353.if OVERRIDE_HIGHCOL\r
1354 ldr r11,=HighCol\r
1355 mov r8, r8, lsl #1 @ cells\r
1356 ldr r11,[r11]\r
1357 mvn r9, #0 @ r9=prevcode=-1\r
1358 add r11,r11,#8\r
1359.else\r
cc68a136 1360 ldr r11,=(HighCol+8)\r
ea8c405f 1361 mov r8, r8, lsl #1 @ cells\r
1362 mvn r9, #0 @ r9=prevcode=-1\r
1363.endif\r
07abbab1 1364 add r1, r11, r0, lsl #4 @ r1=pdest\r
cc68a136 1365 mov r0, #0xf\r
1366 b .dwloop_enter\r
1367\r
07abbab1 1368 @ r4,r5 are scratch in this loop\r
cc68a136 1369.dwloop:\r
1370 add r1, r1, #8\r
1371.dwloop_nor1:\r
1372 add r12, r12, #2 @ halfwords\r
1373 ldrh r7, [lr, r12] @ r7=code (int, but from unsigned, no sign extend)\r
1374 subs r8, r8, #1\r
1375 beq .dwloop_end @ done\r
1376\r
1377 eor r5, r6, r7, lsr #15\r
1378 tst r5, #1\r
1379 orrne r6, r6, #2 @ wrong pri\r
1380 bne .dwloop\r
1381\r
1382 cmp r7, r9\r
1383 beq .dw_samecode @ we know stuff about this tile already\r
1384\r
1385.dwloop_enter:\r
1386 mov r9, r7 @ remember code\r
1387\r
1388 movs r2, r9, lsl #20 @ if (code&0x1000)\r
1389 mov r2, r2, lsl #1\r
1390 add r2, r10, r2, lsr #17 @ r2=addr=(code&0x7ff)<<4; addr+=ty\r
1391 eorcs r2, r2, #0x0e @ if (code&0x1000) addr^=0xe;\r
1392\r
1393 and r3, r9, #0x6000\r
1394 mov r3, r3, lsr #9 @ r3=pal=((code&0x6000)>>9);\r
1395\r
1396 ldr r2, [lr, r2, lsl #1] @ pack=*(unsigned int *)(Pico.vram+addr); // Get 8 pixels\r
1397\r
1398.dw_samecode:\r
1399 tst r6, #0x100\r
1400 bne .dw_shadow\r
1401.dw_shadow_done:\r
1402 tst r2, r2\r
1403 beq .dwloop @ tileline blank\r
1404\r
1405 cmp r2, r2, ror #4\r
1406 beq .dw_SingleColor @ tileline singlecolor \r
1407\r
1408 tst r9, #0x0800\r
7a7c6476 1409 bne .dw_TileFlip\r
cc68a136 1410\r
1411 @ (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r0: helper pattern\r
cc68a136 1412.dw_TileNorm:\r
1413 TileNorm r0\r
1414 b .dwloop\r
1415\r
7a7c6476 1416.dw_TileFlip:\r
1417 TileFlip r0\r
1418 b .dwloop\r
1419\r
cc68a136 1420.dw_SingleColor:\r
1421 and r4, r0, r2 @ #0x0000000f\r
1422 orr r4, r3, r4\r
1423 orr r4, r4, r4, lsl #8\r
1424 orr r4, r4, r4, lsl #16\r
1425 mov r5, r4\r
1426 stmia r1!, {r4,r5}\r
1427 b .dwloop_nor1 @ we incremeted r1 ourselves\r
1428\r
1429.dw_shadow:\r
1430 tst r6, #1 @ hi pri?\r
1431 orreq r3, r3, #0x40\r
1432 beq .dw_shadow_done\r
1433 ldr r4, [r1]\r
07abbab1 1434 mov r5, #0x3f\r
1435 orr r5, r5, r5, lsl #8\r
1436 orr r5, r5, r5, lsl #16\r
1437 and r4, r4, r5\r
cc68a136 1438 str r4, [r1]\r
1439 ldr r4, [r1,#4]\r
07abbab1 1440 and r4, r4, r5\r
cc68a136 1441 str r4, [r1,#4]\r
1442 b .dw_shadow_done\r
1443\r
1444.dwloop_end:\r
1445 ldr r0, =rendstatus\r
1446 ldr r1, [r0]\r
0fc0e241 1447 and r6, r6, #PDRAW_WND_DIFF_PRIO\r
cc68a136 1448 orr r1, r1, r6\r
1449 str r1, [r0]\r
1450\r
1451 ldmfd sp!, {r4-r11,r12}\r
1452 bx r12\r
1453\r
1454\r
1455@ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r
1456\r
1457\r
1458@ hilights 2 pixels in RGB444/BGR444 format\r
1459.macro TileDoShHi2Pixels444 reg\r
1460 mov \reg, \reg, ror #12\r
1461 adds \reg, \reg, #0x40000000\r
1462 orrcs \reg, \reg, #0xf0000000\r
1463 mov \reg, \reg, ror #28\r
1464 adds \reg, \reg, #0x40000000\r
1465 orrcs \reg, \reg, #0xf0000000\r
1466 mov \reg, \reg, ror #28\r
1467 adds \reg, \reg, #0x40000000\r
1468 orrcs \reg, \reg, #0xf0000000\r
1469 mov \reg, \reg, ror #24\r
1470 adds \reg, \reg, #0x40000000\r
1471 orrcs \reg, \reg, #0xf0000000\r
1472 mov \reg, \reg, ror #28\r
1473 adds \reg, \reg, #0x40000000\r
1474 orrcs \reg, \reg, #0xf0000000\r
1475 mov \reg, \reg, ror #28\r
1476 adds \reg, \reg, #0x40000000\r
1477 orrcs \reg, \reg, #0xf0000000\r
1478 mov \reg, \reg, ror #12\r
1479.endm\r
1480\r
1481\r
1482.global FinalizeLineBGR444 @ int sh\r
1483\r
1484FinalizeLineBGR444:\r
1485 stmfd sp!, {r4-r6,lr}\r
1486 mov r6, r0\r
1487 ldr lr, =(Pico+0x22228) @ Pico.video\r
1488 ldr r0, =DrawLineDest\r
1489 ldrb r12, [lr, #12]\r
1490 ldr r0, [r0]\r
1491 sub r3, lr, #0x128 @ r3=Pico.cram\r
1492\r
1493 tst r12, #1\r
1494 movne r2, #320/4 @ len\r
1495 bne .fl_no32colBGR444\r
1496 ldr r4, =PicoOpt\r
1497 mov r2, #256/4\r
1498 ldr r4, [r4]\r
1499 tst r4, #0x100\r
1500 addeq r0, r0, #32*2\r
1501\r
1502.fl_no32colBGR444:\r
1503 tst r6, r6\r
1504 beq .fl_noshBGR444\r
1505\r
1506 ldr r4, =HighPal\r
1507\r
1508 ldrb r12, [lr, #-0x1a] @ 0x2220e ~ dirtyPal\r
1509 tst r12, r12\r
1510 moveq r3, r4\r
1511 beq .fl_noshBGR444\r
1512 mov r12, #0\r
1513 strb r12, [lr, #-0x1a]\r
1514\r
1515 mov lr, #0x40/8\r
1516 @ copy pal:\r
1517.fl_loopcpBGR444:\r
1518 ldmia r3!, {r1,r5,r6,r12}\r
1519 subs lr, lr, #1\r
1520 stmia r4!, {r1,r5,r6,r12}\r
1521 bne .fl_loopcpBGR444\r
1522\r
1523 @ shadowed pixels:\r
1524 mov r12, #0x0077\r
1525 orr r12,r12,#0x0700\r
1526 orr r12,r12,r12,lsl #16\r
1527 sub r3, r3, #0x40*2\r
1528 add r5, r4, #0x80*2\r
1529 mov lr, #0x40/4\r
1530.fl_loopcpBGR444_sh:\r
1531 ldmia r3!, {r1,r6}\r
1532 subs lr, lr, #1\r
1533 and r1, r12, r1, lsr #1\r
1534 and r6, r12, r6, lsr #1\r
1535 stmia r4!, {r1,r6}\r
1536 stmia r5!, {r1,r6}\r
1537 bne .fl_loopcpBGR444_sh\r
1538\r
1539 @ hilighted pixels:\r
1540 sub r3, r3, #0x40*2\r
1541 mov lr, #0x40/2\r
1542.fl_loopcpBGR444_hi:\r
1543 ldr r1, [r3], #4\r
1544 TileDoShHi2Pixels444 r1\r
1545 str r1, [r4], #4\r
1546 subs lr, lr, #1\r
1547 bne .fl_loopcpBGR444_hi\r
1548\r
1549 sub r3, r4, #0x40*3*2\r
e5fa9817 1550 mov r6, #1\r
cc68a136 1551\r
1552\r
1553.fl_noshBGR444:\r
e5fa9817 1554 ldr r12,=rendstatus\r
1555 eors r6, r6, #1 @ sh is 0\r
1556 ldr r12,[r12]\r
1557 mov lr, #0xff\r
283fec1b 1558 tstne r12,#PDRAW_ACC_SPRITES\r
e5fa9817 1559\r
ea8c405f 1560.if OVERRIDE_HIGHCOL\r
1561 ldr r1, =HighCol\r
e5fa9817 1562 movne lr, #0x3f\r
ea8c405f 1563 ldr r1, [r1]\r
1564 mov lr, lr, lsl #1\r
1565 add r1, r1, #8\r
1566.else\r
cc68a136 1567 ldr r1, =(HighCol+8)\r
e5fa9817 1568 movne lr, #0x3f\r
cc68a136 1569 mov lr, lr, lsl #1\r
ea8c405f 1570.endif\r
cc68a136 1571\r
1572.fl_loopBGR444:\r
cc68a136 1573 ldr r12, [r1], #4\r
1574 subs r2, r2, #1\r
1575\r
1576 and r4, lr, r12, lsl #1\r
1577 ldrh r4, [r3, r4]\r
1578 and r5, lr, r12, lsr #7\r
1579 ldrh r5, [r3, r5]\r
1580 and r6, lr, r12, lsr #15\r
1581 ldrh r6, [r3, r6]\r
e5fa9817 1582 and r12,lr, r12, lsr #23\r
1583 ldrh r12,[r3, r12] @ 1c.i.\r
cc68a136 1584 orr r4, r4, r5, lsl #16\r
e5fa9817 1585 orr r5, r6, r12,lsl #16\r
cc68a136 1586\r
1587 stmia r0!, {r4,r5}\r
1588 bne .fl_loopBGR444\r
1589\r
1590\r
1591 ldmfd sp!, {r4-r6,lr}\r
1592 bx lr\r
1593\r
1594\r
1595@ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r
1596\r
1597\r
cc68a136 1598@ Convert 0000bbb0 ggg0rrr0\r
1599@ to rrrrrggg gggbbbbb\r
1600\r
a39d8ba5 1601@ r2,r3 - scratch, lr = 0x001c001c, r8 = 0x08610861\r
cc68a136 1602.macro convRGB565 reg\r
a39d8ba5 1603 and r2, lr, \reg,lsr #7 @ b\r
1604 and r3, lr, \reg,lsr #3 @ g\r
1605 and \reg, lr, \reg,lsl #1 @ r\r
1606 orr r2, r2, r3, lsl #6\r
1607 orr \reg, r2, \reg,lsl #11\r
1608\r
1609 and r2, r8, \reg,lsr #4\r
1610 orr \reg, \reg, r2\r
cc68a136 1611.endm\r
1612\r
a39d8ba5 1613@ trashes: r2-r8,r12,lr; r8 = 0x08610861; r0,r1 are advanced\r
b2305d08 1614.macro vidConvCpyRGB565_local\r
2ec14aec 1615 mov r12, r2, lsr #3 @ repeats\r
cc68a136 1616 mov lr, #0x001c0000\r
1617 orr lr, lr, #0x01c @ lr == pattern 0x001c001c\r
cc68a136 1618\r
b2305d08 16190:\r
cc68a136 1620 ldmia r1!, {r4-r7}\r
1621 subs r12, r12, #1\r
1622 convRGB565 r4\r
1623 str r4, [r0], #4\r
1624 convRGB565 r5\r
1625 str r5, [r0], #4\r
1626 convRGB565 r6\r
1627 str r6, [r0], #4\r
1628 convRGB565 r7\r
1629 str r7, [r0], #4\r
1630\r
b2305d08 1631 bgt 0b\r
1632.endm\r
1633\r
1634\r
1635.global vidConvCpyRGB565\r
cc68a136 1636\r
b2305d08 1637vidConvCpyRGB565: @ void *to, void *from, int pixels\r
1638 stmfd sp!, {r4-r9,lr}\r
a39d8ba5 1639 mov r8, #0x0061\r
1640 orr r8, r8, #0x0800\r
1641 orr r8, r8, lsl #16\r
b2305d08 1642 vidConvCpyRGB565_local\r
cc68a136 1643 ldmfd sp!, {r4-r9,lr}\r
1644 bx lr\r
1645\r
1646\r
b2305d08 1647.global PicoDoHighPal555 @ int sh\r
cc68a136 1648\r
b2305d08 1649PicoDoHighPal555:\r
1650 stmfd sp!, {r4-r9,lr}\r
1651 mov r1, #0\r
3d48f143 1652 ldr r8, =(Pico+0x22228) @ Pico.video\r
cc68a136 1653\r
b2305d08 1654PicoDoHighPal555_nopush:\r
a39d8ba5 1655 orr r9, r1, r0, lsl #31 @ 0:called from FinalizeLine555, 31: s/h\r
b2305d08 1656\r
b2305d08 1657 ldr r0, =HighPal\r
1658\r
cc68a136 1659 mov r1, #0\r
b2305d08 1660 strb r1, [r8, #-0x1a] @ 0x2220e ~ dirtyPal\r
1661\r
3d48f143 1662 sub r1, r8, #0x128 @ r1=Pico.cram\r
cc68a136 1663 mov r2, #0x40\r
a39d8ba5 1664 mov r8, #0x0061\r
1665 orr r8, r8, #0x0800\r
1666 orr r8, r8, lsl #16\r
1667\r
b2305d08 1668 vidConvCpyRGB565_local\r
cc68a136 1669\r
a39d8ba5 1670 tst r9, #(1<<31)\r
b2305d08 1671 beq PicoDoHighPal555_end\r
1672\r
1673 ldr r3, =HighPal\r
cc68a136 1674\r
1675 @ shadowed pixels:\r
1676 mov r12, #0x008e\r
cc68a136 1677 add r4, r3, #0x40*2\r
b2305d08 1678 orr r12,r12,#0x7300\r
cc68a136 1679 add r5, r3, #0xc0*2\r
b2305d08 1680 orr r12,r12,r12,lsl #16\r
cc68a136 1681 mov lr, #0x40/4\r
1682.fl_loopcpRGB555_sh:\r
1683 ldmia r3!, {r1,r6}\r
1684 subs lr, lr, #1\r
1685 and r1, r12, r1, lsr #1\r
1686 and r6, r12, r6, lsr #1\r
1687 stmia r4!, {r1,r6}\r
1688 stmia r5!, {r1,r6}\r
1689 bne .fl_loopcpRGB555_sh\r
1690\r
1691 @ hilighted pixels:\r
a39d8ba5 1692 @ t = ((dpal[i] >> 1) & 0x738e738e) + 0x738e738e;\r
1693 @ t |= (t >> 4) & 0x08610861;\r
1694 @ r8=0x08610861\r
cc68a136 1695 sub r3, r3, #0x40*2\r
a39d8ba5 1696 mov lr, #0x40/4\r
cc68a136 1697.fl_loopcpRGB555_hi:\r
a39d8ba5 1698 ldmia r3!, {r1,r6}\r
1699 and r1, r12, r1, lsr #1\r
1700 and r6, r12, r6, lsr #1\r
1701 add r1, r12, r1\r
1702 add r6, r12, r6\r
1703 and r5, r8, r1, lsr #4\r
1704 and r7, r8, r6, lsr #4\r
1705 orr r1, r1, r5\r
1706 orr r6, r6, r7\r
1707 stmia r4!, {r1,r6}\r
cc68a136 1708 subs lr, lr, #1\r
1709 bne .fl_loopcpRGB555_hi\r
b2305d08 1710 mov r0, #1\r
cc68a136 1711\r
b2305d08 1712PicoDoHighPal555_end:\r
a39d8ba5 1713 tst r9, #1\r
b2305d08 1714 ldmeqfd sp!, {r4-r9,pc}\r
1715\r
1716 ldr r8, =(Pico+0x22228) @ Pico.video\r
1717 b FinalizeLineRGB555_pal_done\r
1718\r
1719\r
5a681086 1720.global FinalizeLine555 @ int sh\r
b2305d08 1721\r
5a681086 1722FinalizeLine555:\r
b2305d08 1723 stmfd sp!, {r4-r9,lr}\r
1724 ldr r8, =(Pico+0x22228) @ Pico.video\r
1725\r
1726 ldrb r2, [r8, #-0x1a] @ 0x2220e ~ dirtyPal\r
1727 mov r1, #1\r
1728 tst r2, r2\r
1729 bne PicoDoHighPal555_nopush\r
1730\r
1731FinalizeLineRGB555_pal_done:\r
1732 ldr r3, =HighPal\r
cc68a136 1733\r
e5fa9817 1734 ldr r12,=rendstatus\r
b2305d08 1735 eors r0, r0, #1 @ sh is 0\r
e5fa9817 1736 ldr r12,[r12]\r
1737 mov lr, #0xff\r
283fec1b 1738 tstne r12,#PDRAW_ACC_SPRITES\r
e5fa9817 1739 movne lr, #0x3f\r
1740\r
ea8c405f 1741.if OVERRIDE_HIGHCOL\r
1742 ldr r1, =HighCol\r
1743 ldr r0, =DrawLineDest\r
1744 ldr r1, [r1]\r
1745 ldr r0, [r0]\r
1746 add r1, r1, #8\r
1747.else\r
3d48f143 1748 ldr r0, =DrawLineDest\r
cc68a136 1749 ldr r1, =(HighCol+8)\r
3d48f143 1750 ldr r0, [r0]\r
ea8c405f 1751.endif\r
3d48f143 1752\r
1753 ldrb r12, [r8, #12]\r
cc68a136 1754 mov lr, lr, lsl #1\r
1755\r
3d48f143 1756 tst r12, #1\r
1757 movne r2, #320/8 @ len\r
1758 bne .fl_no32colRGB555\r
1759 ldr r4, =PicoOpt\r
1760 mov r2, #256/8\r
1761 ldr r4, [r4]\r
1762 tst r4, #0x4000\r
1763 bne .fl_32scale_RGB555\r
1764 tst r4, #0x0100\r
1765 addeq r0, r0, #32*2\r
1766\r
1767.fl_no32colRGB555:\r
cc68a136 1768\r
499a0be3 1769.if UNALIGNED_DRAWLINEDEST\r
1770 @ this is basically for Gizmondo, which has unaligned odd lines in the framebuffer\r
1771 tst r0, #2\r
1772 bne .fl_RGB555u\r
1773.endif\r
1774\r
1775.fl_loopRGB555:\r
cc68a136 1776 ldr r12, [r1], #4\r
1777 ldr r7, [r1], #4\r
1778\r
1779 and r4, lr, r12, lsl #1\r
1780 ldrh r4, [r3, r4]\r
1781 and r5, lr, r12, lsr #7\r
1782 ldrh r5, [r3, r5]\r
1783 and r6, lr, r12, lsr #15\r
1784 ldrh r6, [r3, r6]\r
1785 orr r4, r4, r5, lsl #16\r
1786\r
1787 and r5, lr, r12, lsr #23\r
1788 ldrh r5, [r3, r5]\r
1789 and r8, lr, r7, lsl #1\r
1790 ldrh r8, [r3, r8]\r
1791 orr r5, r6, r5, lsl #16\r
1792\r
1793 and r6, lr, r7, lsr #7\r
1794 ldrh r6, [r3, r6]\r
1795 and r12,lr, r7, lsr #15\r
1796 ldrh r12,[r3, r12]\r
499a0be3 1797 and r7, lr, r7, lsr #23\r
1798 ldrh r7, [r3, r7]\r
cc68a136 1799 orr r8, r8, r6, lsl #16\r
1800\r
cc68a136 1801 subs r2, r2, #1\r
499a0be3 1802 orr r12,r12, r7, lsl #16\r
cc68a136 1803\r
1804 stmia r0!, {r4,r5,r8,r12}\r
1805 bne .fl_loopRGB555\r
1806\r
b2305d08 1807 ldmfd sp!, {r4-r9,lr}\r
3d48f143 1808 bx lr\r
1809\r
1810\r
1811.fl_32scale_RGB555:\r
b2305d08 1812 stmfd sp!, {r10}\r
3d48f143 1813 mov r9, #0x3900 @ f800 07e0 001f | e000 0780 001c | 3800 01e0 0007\r
1814 orr r9, r9, #0x00e7\r
1815\r
499a0be3 1816.if UNALIGNED_DRAWLINEDEST\r
1817 tst r0, #2\r
1818 bne .fl_32scale_RGB555u\r
1819.endif\r
1820\r
3d48f143 1821.fl_loop32scale_RGB555:\r
1822 ldr r12, [r1], #4\r
1823 ldr r7, [r1], #4\r
cc68a136 1824\r
3d48f143 1825 and r4, lr, r12,lsl #1\r
1826 ldrh r4, [r3, r4]\r
1827 and r5, lr, r12,lsr #7\r
1828 ldrh r5, [r3, r5]\r
1829 and r4, r4, r9, lsl #2\r
1830 orr r4, r4, r4, lsl #14 @ r4[31:16] = 1/4 pix_s 0\r
1831 and r5, r5, r9, lsl #2\r
1832 sub r6, r5, r5, lsr #2 @ r6 = 3/4 pix_s 1\r
1833 add r4, r4, r6, lsl #16 @ pix_d 0, 1\r
1834 and r6, lr, r12,lsr #15\r
1835 ldrh r6, [r3, r6]\r
1836 and r12,lr, r12,lsr #23\r
1837 ldrh r12,[r3, r12]\r
1838 and r6, r6, r9, lsl #2\r
1839 add r5, r5, r6\r
1840 mov r5, r5, lsr #1\r
1841 sub r6, r6, r6, lsr #2 @ r6 = 3/4 pix_s 2\r
1842 orr r5, r5, r6, lsl #16\r
1843\r
1844 and r6, lr, r7, lsl #1\r
1845 ldrh r6, [r3, r6]\r
1846 and r12,r12,r9, lsl #2\r
1847 add r5, r5, r12,lsl #14 @ pix_d 2, 3\r
1848 and r6, r6, r9, lsl #2\r
1849 orr r6, r12,r6, lsl #16 @ pix_d 4, 5\r
1850\r
1851 and r12,lr, r7, lsr #7\r
1852 ldrh r12,[r3, r12]\r
1853 and r10,lr, r7, lsr #15\r
1854 ldrh r10,[r3, r10]\r
1855 and r12,r12,r9, lsl #2\r
1856 sub r8, r12,r12,lsr #2 @ r8 = 3/4 pix_s 1\r
1857 add r8, r8, r6, lsr #18\r
1858 and r7, lr, r7, lsr #23\r
1859 ldrh r7, [r3, r7]\r
1860 and r10,r10,r9, lsl #2\r
1861 orr r8, r8, r10,lsl #15\r
1862 add r8, r8, r12,lsl #15 @ pix_d 6, 7\r
1863 sub r10,r10,r10,lsr #2 @ r10= 3/4 pix_s 2\r
1864 and r7, r7, r9, lsl #2\r
1865 add r10,r10,r7, lsr #2 @ += 1/4 pix_s 3\r
1866 orr r10,r10,r7, lsl #16 @ pix_d 8, 9\r
1867\r
1868 subs r2, r2, #1\r
1869\r
1870 stmia r0!, {r4,r5,r6,r8,r10}\r
1871 bne .fl_loop32scale_RGB555\r
1872\r
b2305d08 1873 ldmfd sp!, {r10}\r
1874 ldmfd sp!, {r4-r9,lr}\r
3d48f143 1875 bx lr\r
1876\r
499a0be3 1877.if UNALIGNED_DRAWLINEDEST\r
1878 @ unaligned versions of loops\r
9839d126 1879 @ warning: starts drawing 2bytes before dst\r
499a0be3 1880\r
1881.fl_RGB555u:\r
9839d126 1882 sub r0, r0, #2 @ initial adjustment\r
1883 mov r8, #0\r
499a0be3 1884\r
1885.fl_loopRGB555u:\r
1886 ldr r12, [r1], #4\r
1887 ldr r7, [r1], #4\r
1888\r
9839d126 1889 and r6, lr, r12,lsl #1\r
1890 ldrh r6, [r3, r6]\r
1891 and r5, lr, r12,lsr #7\r
499a0be3 1892 ldrh r5, [r3, r5]\r
9839d126 1893 orr r4, r8, r6, lsl #16\r
499a0be3 1894\r
9839d126 1895 and r6, lr, r12,lsr #15\r
499a0be3 1896 ldrh r6, [r3, r6]\r
9839d126 1897 and r8, lr, r12,lsr #23\r
499a0be3 1898 ldrh r8, [r3, r8]\r
9839d126 1899 orr r5, r5, r6, lsl #16\r
499a0be3 1900\r
9839d126 1901 and r6, lr, r7, lsl #1\r
499a0be3 1902 ldrh r6, [r3, r6]\r
9839d126 1903 and r12,lr, r7, lsr #7\r
499a0be3 1904 ldrh r12,[r3, r12]\r
9839d126 1905 orr r6, r8, r6, lsl #16\r
1906\r
1907 and r8, lr, r7, lsr #15\r
499a0be3 1908 ldrh r8, [r3, r8]\r
9839d126 1909 and r7, lr, r7, lsr #23\r
499a0be3 1910\r
1911 subs r2, r2, #1\r
9839d126 1912 orr r12,r12,r8, lsl #16\r
1913 ldrh r8, [r3, r7]\r
1914\r
1915 stmia r0!, {r4,r5,r6,r12}\r
499a0be3 1916 bne .fl_loopRGB555u\r
1917\r
1918 strh r8, [r0], #2\r
1919\r
b2305d08 1920 ldmfd sp!, {r4-r9,lr}\r
499a0be3 1921 bx lr\r
1922\r
1923\r
1924.fl_32scale_RGB555u:\r
9839d126 1925 sub r0, r0, #2 @ initial adjustment\r
1926 mov r4, #0\r
499a0be3 1927\r
1928 @ r9 f800 07e0 001f | e000 0780 001c | 3800 01e0 0007\r
1929.fl_loop32scale_RGB555u:\r
1930 ldr r12, [r1], #4\r
1931 ldr r7, [r1], #4\r
1932\r
9839d126 1933 and r6, lr, r12,lsl #1\r
1934 ldrh r6, [r3, r6]\r
499a0be3 1935 and r5, lr, r12,lsr #7\r
1936 ldrh r5, [r3, r5]\r
9839d126 1937 and r6, r6, r9, lsl #2\r
1938 orr r4, r4, r6, lsl #16 @ r4 = pix_d -1, 0\r
499a0be3 1939\r
1940 and r5, r5, r9, lsl #2\r
9839d126 1941 sub r8, r5, r5, lsr #2 @ r8 = 3/4 pix_s 1\r
1942 add r6, r8, r6, lsr #2 @ r6 = (1/4 pix_s 0) + (3/4 pix_s 1)\r
1943 orr r5, r6, r5, lsl #15\r
499a0be3 1944\r
1945 and r6, lr, r12,lsr #15\r
1946 ldrh r6, [r3, r6]\r
1947 and r12,lr, r12,lsr #23\r
1948 ldrh r12,[r3, r12]\r
1949 and r6, r6, r9, lsl #2\r
9839d126 1950 add r5, r5, r6, lsl #15 @ r5 = pix_d 1, 2\r
499a0be3 1951\r
9839d126 1952 and r8, lr, r7, lsl #1\r
1953 ldrh r8, [r3, r8]\r
1954 and r10,lr, r7, lsr #7\r
1955 ldrh r10,[r3, r10]\r
499a0be3 1956 and r12,r12,r9, lsl #2\r
9839d126 1957 sub r6, r6, r6, lsr #2 @ r6 = 3/4 pix_s 2\r
1958 add r6, r6, r12,lsr #2\r
1959 orr r6, r6, r12,lsl #16 @ r6 = pix_d 3, 4\r
499a0be3 1960\r
9839d126 1961 and r8, r8, r9, lsl #2\r
1962 and r10,r10,r9, lsl #2\r
1963 sub r12,r10,r10,lsr #2 @ r12 = 3/4 pix_s 5\r
1964 orr r8, r8, r8, lsl #14\r
1965 add r8, r8, r12,lsl #16 @ r8 = pix_d 5, 6\r
1966 and r12,lr, r7, lsr #15\r
499a0be3 1967 ldrh r12,[r3, r12]\r
499a0be3 1968 and r7, lr, r7, lsr #23\r
1969 ldrh r7, [r3, r7]\r
9839d126 1970 and r12,r12,r9, lsl #2\r
1971 add r10,r10,r12\r
1972 mov r10,r10, lsr #1\r
1973 sub r12,r12,r12,lsr #2 @ r12 = 3/4 pix_s 6\r
1974 orr r10,r10,r12,lsl #16\r
499a0be3 1975 and r7, r7, r9, lsl #2\r
9839d126 1976 add r10,r10,r7, lsl #14 @ r10 = pix_d 7, 8\r
499a0be3 1977\r
1978 subs r2, r2, #1\r
1979\r
1980 stmia r0!, {r4,r5,r6,r8,r10}\r
9839d126 1981 mov r4, r7\r
499a0be3 1982 bne .fl_loop32scale_RGB555u\r
1983\r
9839d126 1984 strh r4, [r0], #2\r
1985\r
b2305d08 1986 ldmfd sp!, {r10}\r
1987 ldmfd sp!, {r4-r9,lr}\r
499a0be3 1988 bx lr\r
1989\r
1990.endif @ UNALIGNED_DRAWLINEDEST\r
1991\r
cc68a136 1992\r
1993@ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r
1994\r
1995@ utility\r
1996.global blockcpy @ void *dst, void *src, size_t n\r
1997\r
1998blockcpy:\r
1999 stmfd sp!, {r4,r5}\r
2000 mov r2, r2, lsr #4\r
2001blockcpy_loop:\r
2002 ldmia r1!, {r3-r5,r12}\r
2003 subs r2, r2, #1\r
2004 stmia r0!, {r3-r5,r12}\r
2005 bne blockcpy_loop\r
2006 ldmfd sp!, {r4,r5}\r
2007 bx lr\r
2008\r
2009\r
2010.global blockcpy_or @ void *dst, void *src, size_t n, int pat\r
2011\r
2012blockcpy_or:\r
2013 stmfd sp!, {r4-r6}\r
2014 orr r3, r3, r3, lsl #8\r
2015 orr r3, r3, r3, lsl #16\r
2016 mov r2, r2, lsr #4\r
2017blockcpy_loop_or:\r
2018 ldmia r1!, {r4-r6,r12}\r
2019 subs r2, r2, #1\r
2020 orr r4, r4, r3\r
2021 orr r5, r5, r3\r
2022 orr r6, r6, r3\r
2023 orr r12,r12,r3\r
2024 stmia r0!, {r4-r6,r12}\r
2025 bne blockcpy_loop_or\r
2026 ldmfd sp!, {r4-r6}\r
2027 bx lr\r
2028\r
fast MegaDrive/MegaCD/32X emulator