notaz.gp2x.de / picodrive.git / blame
? search:
summary | shortlog | log | commit | commitdiff | tree
blob | blame (incremental) | history | HEAD
sram handling refactored
[picodrive.git] / Pico / Draw.s
CommitLineData
3d48f143 1@ vim:filetype=armasm\r
2\r
cc68a136 3@ assembly "optimized" version of some funtions from draw.c\r
4@ this is highly specialized, be careful if changing related C code!\r
5\r
6cadc2da 6@ (c) Copyright 2007, Grazvydas "notaz" Ignotas\r
cc68a136 7@ All Rights Reserved\r
8\r
9\r
10.extern Pico\r
11.extern PicoOpt\r
12.extern HighCol\r
13.extern Scanline\r
14.extern HighSprZ\r
15.extern rendstatus\r
16.extern DrawLineDest\r
cc68a136 17.extern DrawStripInterlace\r
18\r
19\r
20@ helper\r
21.macro TilePixel pat lsrr offs\r
22.if !\lsrr\r
23 ands r4, \pat, r2\r
24.else\r
25 ands r4, \pat, r2, lsr #\lsrr\r
26.endif\r
27 orrne r4, r3, r4\r
28 strneb r4, [r1,#\offs]\r
29.endm\r
30\r
31@ TileNorm (r1=pdest, r2=pixels8, r3=pal) r4: scratch, pat: register with helper pattern 0xf\r
32.macro TileNorm pat\r
33 TilePixel \pat, 12, 0 @ #0x0000f000\r
34 TilePixel \pat, 8, 1 @ #0x00000f00\r
35 TilePixel \pat, 4, 2 @ #0x000000f0\r
36 TilePixel \pat, 0, 3 @ #0x0000000f\r
37 TilePixel \pat, 28, 4 @ #0xf0000000\r
38 TilePixel \pat, 24, 5 @ #0x0f000000\r
39 TilePixel \pat, 20, 6 @ #0x00f00000\r
40 TilePixel \pat, 16, 7 @ #0x000f0000\r
41.endm\r
42\r
43@ TileFlip (r1=pdest, r2=pixels8, r3=pal) r4: scratch, pat: register with helper pattern 0xf\r
44.macro TileFlip pat\r
45 TilePixel \pat, 16, 0 @ #0x000f0000\r
46 TilePixel \pat, 20, 1 @ #0x00f00000\r
47 TilePixel \pat, 24, 2 @ #0x0f000000\r
48 TilePixel \pat, 28, 3 @ #0xf0000000\r
49 TilePixel \pat, 0, 4 @ #0x0000000f\r
50 TilePixel \pat, 4, 5 @ #0x000000f0\r
51 TilePixel \pat, 8, 6 @ #0x00000f00\r
52 TilePixel \pat, 12, 7 @ #0x0000f000\r
53.endm\r
54\r
55@ shadow/hilight mode\r
56\r
57@ this one is for hi priority layer\r
58.macro TilePixelShHP lsrr offs\r
59.if !\lsrr\r
60 ands r4, r12, r2\r
61.else\r
62 ands r4, r12, r2, lsr #\lsrr\r
63.endif\r
64 ldreqb r4, [r1,#\offs]\r
65 orrne r4, r3, r4\r
66 strneb r4, [r1,#\offs]\r
67 tsteq r4, #0x80\r
68 andeq r4, r4, #0x3f\r
69 streqb r4, [r1,#\offs]\r
70.endm\r
71\r
72@ TileNorm (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r12: register with helper pattern 0xf, touches r3 high bits\r
73.macro TileNormShHP\r
74 TilePixelShHP 12, 0 @ #0x0000f000\r
75 TilePixelShHP 8, 1 @ #0x00000f00\r
76 TilePixelShHP 4, 2 @ #0x000000f0\r
77 TilePixelShHP 0, 3 @ #0x0000000f\r
78 TilePixelShHP 28, 4 @ #0xf0000000\r
79 TilePixelShHP 24, 5 @ #0x0f000000\r
80 TilePixelShHP 20, 6 @ #0x00f00000\r
81 TilePixelShHP 16, 7 @ #0x000f0000\r
82.endm\r
83\r
84@ TileFlip (r1=pdest, r2=pixels8, r3=pal) r4: scratch, pat: register with helper pattern 0xf\r
85.macro TileFlipShHP\r
86 TilePixelShHP 16, 0 @ #0x000f0000\r
87 TilePixelShHP 20, 1 @ #0x00f00000\r
88 TilePixelShHP 24, 2 @ #0x0f000000\r
89 TilePixelShHP 28, 3 @ #0xf0000000\r
90 TilePixelShHP 0, 4 @ #0x0000000f\r
91 TilePixelShHP 4, 5 @ #0x000000f0\r
92 TilePixelShHP 8, 6 @ #0x00000f00\r
93 TilePixelShHP 12, 7 @ #0x0000f000\r
94.endm\r
95\r
96\r
97@ TileSingleSh (r1=pdest, r2=pixels8, r3=pal) r4,r7: scratch, r0=sx; r12: helper pattern 0xf\r
98.macro TileSingleSh\r
99 tst r0, #1 @ not aligned?\r
100 mov r7, #0x00c000\r
101 orr r7, r7, #0xc0\r
102 ldrneb r4, [r1]\r
103 ldreqh r4, [r1]\r
104 orr r4, r4, r7\r
105 strneb r4, [r1], #1\r
106 streqh r4, [r1], #2\r
107 ldrh r4, [r1]\r
108 orr r4, r4, r7\r
109 strh r4, [r1], #2\r
110 ldrh r4, [r1]\r
111 orr r4, r4, r7\r
112 strh r4, [r1], #2\r
113 ldrh r4, [r1]\r
114 orr r4, r4, r7\r
115 strh r4, [r1], #2\r
116 ldrneb r4, [r1]\r
117 orr r4, r4, r7\r
118 strneb r4, [r1], #1\r
119.endm\r
120\r
121@ TileSingleHi (r1=pdest, r2=pixels8, r3=pal) r4,r7: scratch, r0=sx, r12: register with helper pattern 0xf\r
122.macro TileSingleHi\r
123 tst r1, #1 @ not aligned?\r
124 mov r7, #0x008000\r
125 orr r7, r7, #0x80\r
126 ldrneb r4, [r1], #1\r
127 ldreqh r4, [r1], #2 @ 1ci\r
128 ldrh r12, [r1], #2\r
129 bic r4, r4, r7, lsr #1\r
130 orr r4, r4, r7\r
131 strneb r4, [r1, #-3]\r
132 streqh r4, [r1, #-4]\r
133 ldrh r4, [r1], #2\r
134 bic r12, r12, r7, lsr #1\r
135 orr r12, r12, r7\r
136 strh r12, [r1, #-4]\r
137 ldrh r12, [r1], #2\r
138 bic r4, r4, r7, lsr #1\r
139 orr r4, r4, r7\r
140 strh r4, [r1, #-4]\r
141 ldrneb r4, [r1]\r
142 bic r12, r12, r7, lsr #1\r
143 orr r12, r12, r7\r
144 strh r12, [r1, #-2]\r
145 bicne r4, r4, r7, lsr #1\r
146 orrne r4, r4, r7\r
147 strneb r4, [r1], #1\r
148 mov r12, #0xf\r
149.endm\r
150\r
151.macro TileDoShGenPixel shift ofs\r
152.if \shift\r
153 ands r4, r12, r2, lsr #\shift\r
154.else\r
155 ands r4, r12, r2\r
156.endif\r
157 beq 3f\r
158 cmp r4, #0xe\r
159 beq 2f\r
160 bgt 1f\r
161 orr r4, r3, r4\r
162 strb r4, [r1,#\ofs]\r
163 b 3f\r
1641:\r
165 ldrb r4, [r1,#\ofs] @ 2ci\r
166 orr r4, r4, #0xc0\r
167 strb r4, [r1,#\ofs]\r
168 b 3f\r
1692:\r
170 ldrb r4, [r1,#\ofs] @ 2ci\r
171 bic r4, r4, #0xc0\r
172 orr r4, r4, #0x80\r
173 strb r4, [r1,#\ofs]\r
1743:\r
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
201\r
202@ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r
203\r
204@ struct TileStrip\r
205@ {\r
206@ int nametab; // 0x00\r
207@ int line; // 0x04\r
208@ int hscroll; // 0x08\r
209@ int xmask; // 0x0C\r
210@ int *hc; // 0x10 (pointer to cache buffer)\r
211@ int cells; // 0x14\r
212@ };\r
213\r
214@ int DrawLayer(int plane, int *hcache, int maxcells, int sh)\r
215\r
216.global DrawLayer @ int plane, int *hcache, int maxcells, int sh\r
217\r
218DrawLayer:\r
219 stmfd sp!, {r4-r11,lr}\r
220\r
221 ldr r11, =(Pico+0x22228) @ Pico.video\r
222 mov r8, #1\r
223\r
224 ldrb r7, [r11, #16] @ ??hh??ww\r
225\r
226 mov r6, r1 @ hcache\r
227 orr r9, r2, r3, lsl #31 @ r9=maxcells|(sh<<31)\r
228\r
229 mov r1, r7, lsl #4\r
230 orr r1, r1, #0x00ff\r
231\r
232 and r10, r7, #3\r
233 cmp r10, #1\r
234 biclt r1, r1, #0xfc00\r
235 biceq r1, r1, #0xfe00\r
236 bicgt r1, r1, #0xff00 @ r1=ymask=(height<<8)|0xff; ...; // Y Mask in pixels\r
237\r
238 add r10, r10, #5\r
239 cmp r10, #7\r
240 subge r10, r10, #1 @ r10=shift[width] (5,6,6,7)\r
241\r
242 @ calculate xmask:\r
243 mov r5, r8, lsl r10\r
244 sub r5, r5, #1 @ r5=xmask\r
245\r
246 @ Find name table:\r
247 tst r0, r0\r
248 ldreqb r12, [r11, #2]\r
249 ldrneb r12, [r11, #4]\r
250\r
251 ldr r2, =Scanline @ trying to make good use of pipeline here\r
252 ldr lr, =(Pico+0x10000) @ lr=Pico.vram\r
253\r
254 moveq r12, r12, lsl #10\r
255 movne r12, r12, lsl #13\r
256 and r12, r12, #(7<<13) @ r12=(ts->nametab<<1) (halfword compliant)\r
257\r
258 ldrh r8, [r11, #12]\r
259 ldrb r7, [r11, #11]\r
260 ldr r2, [r2]\r
261\r
262 mov r4, r8, lsr #8 @ pvid->reg[13]\r
263 mov r4, r4, lsl #10 @ htab=pvid->reg[13]<<9; (halfwords)\r
264 tst r7, #2\r
265 addne r4, r4, r2, lsl #2 @ htab+=Scanline<<1; // Offset by line\r
266 tst r7, #1\r
267 biceq r4, r4, #0x1f @ htab&=~0xf; // Offset by tile\r
268 add r4, r4, r0, lsl #1 @ htab+=plane\r
269 bic r4, r4, #0x00ff0000 @ just in case\r
270 ldrh r3, [lr, r4] @ r3=hscroll\r
271\r
272 tst r7, #4\r
273 bne .DrawStrip_vsscroll\r
274\r
275 @ Get vertical scroll value:\r
276 add r7, lr, #0x012000\r
277 add r7, r7, #0x000180 @ r7=Pico.vsram (Pico+0x22180)\r
278 ldr r7, [r7]\r
279\r
280 tst r8, #2\r
281 tstne r8, #4\r
282 bne .DrawStrip_interlace\r
283\r
284 tst r0, r0\r
285 movne r7, r7, lsr #16\r
286\r
287 @ Find the line in the name table\r
288 add r2, r2, r7\r
289 and r2, r2, r1\r
290 mov r4, r2, lsr #3\r
291 add r10, r10, #1 @ shift[width]++\r
292 add r12, r12, r4, lsl r10 @ nametab+=(ts.line>>3)<<shift[width];\r
293\r
294 @ ldmia r0, {r1,r2,r3,r5,r6,r9} @ r2=line, r3=ts->hscroll, r5=ts->xmask, r6=ts->hc, r9=ts->cells\r
cc68a136 295\r
296 and r10,r2, #7\r
297 mov r10,r10, lsl #1 @ r10=ty=(ts->line&7)<<1;\r
298 orr r10,r10, r9, lsl #24\r
299\r
300 rsb r8, r3, #0\r
301 mov r8, r8, lsr #3 @ r8=tilex=(-ts->hscroll)>>3\r
302\r
303 sub r1, r3, #1\r
304 and r1, r1, #7\r
305 add r7, r1, #1 @ r7=dx=((ts->hscroll-1)&7)+1\r
306\r
307 tst r9, #1<<31\r
308 mov r3, #0\r
740da8c6 309 orrne r10,r10, #1<<23 @ r10=(cells<<24|sh<<23|hi_not_empty<<22|had_output<<21|ty)\r
cc68a136 310 movne r3, #0x40 @ default to shadowed pal on sh mode\r
311\r
312 mvn r9, #0 @ r9=prevcode=-1\r
313\r
314 cmp r7, #8\r
315 addne r10,r10, #0x01000000 @ we will loop cells+1 times if there is scroll\r
316\r
317 @ cache some stuff to avoid mem access\r
318 ldr r11,=HighCol\r
319 mov r0, #0xf\r
320 add r1, r11, r7 @ r1=pdest\r
321\r
322\r
323 @ r4 & r7 are scratch in this loop\r
324.dsloop_subr1:\r
325 sub r1, r1, #8\r
326.dsloop: @ 40-41 times\r
327 subs r10,r10, #0x01000000\r
328 bmi .dsloop_exit\r
329\r
330.dsloop_enter:\r
331 and r7, r5, r8\r
332 add r7, lr, r7, lsl #1 @ Pico.vram+((tilex&ts->xmask) as halfwords)\r
333 ldrh r7, [r7, r12] @ r7=code (int, but from unsigned, no sign extend)\r
334\r
335 add r1, r1, #8\r
336 add r8, r8, #1\r
337\r
338 tst r7, #0x8000\r
339 bne .DrawStrip_hiprio\r
340\r
341 cmp r7, r9\r
342 beq .DrawStrip_samecode @ we know stuff about this tile already\r
343\r
344 mov r9, r7 @ remember code\r
740da8c6 345 orr r10, r10, #1<<21 @ seen non hi-prio tile\r
cc68a136 346\r
347 movs r2, r9, lsl #20 @ if (code&0x1000)\r
348 mov r2, r2, lsl #1\r
349 add r2, r2, r10, lsl #17\r
350 mov r2, r2, lsr #17\r
351 eorcs r2, r2, #0x0e @ if (code&0x1000) addr^=0xe;\r
352\r
353 ldr r2, [lr, r2, lsl #1] @ pack=*(unsigned int *)(Pico.vram+addr); // Get 8 pixels\r
354\r
355 bic r7, r3, #0x3f\r
356 and r3, r9, #0x6000\r
357 add r3, r7, r3, lsr #9 @ r3=pal=((code&0x6000)>>9);\r
358\r
359.DrawStrip_samecode:\r
360 tst r2, r2\r
361 beq .dsloop @ tileline blank\r
362\r
363 cmp r2, r2, ror #4\r
364 beq .DrawStrip_SingleColor @ tileline singlecolor \r
365\r
366 tst r9, #0x0800\r
367 beq .DrawStrip_TileNorm\r
368\r
369 @ (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r0: helper pattern\r
370 TileFlip r0\r
371 b .dsloop\r
372\r
373.DrawStrip_TileNorm:\r
374 TileNorm r0\r
375 b .dsloop\r
376\r
377.DrawStrip_SingleColor:\r
378 and r4, r2, #0xf\r
379 orr r4, r3, r4\r
380 orr r4, r4, r4, lsl #8\r
381 tst r1, #1 @ not aligned?\r
382 strneb r4, [r1], #1\r
383 streqh r4, [r1], #2\r
384 strh r4, [r1], #2\r
385 strh r4, [r1], #2\r
386 strh r4, [r1], #2\r
387 strneb r4, [r1], #1 @ have a remaining unaligned pixel?\r
388 b .dsloop_subr1\r
389\r
cc68a136 390.DrawStrip_hiprio_maybempt:\r
391 cmp r7, r9\r
392 beq .dsloop @ must've been empty, otherwise we wouldn't get here\r
393 movs r2, r7, lsl #20 @ if (code&0x1000)\r
394 mov r2, r2, lsl #1\r
395 add r2, r2, r10, lsl #17\r
396 mov r2, r2, lsr #17\r
397 eorcs r2, r2, #0x0e @ if (code&0x1000) addr^=0xe;\r
398 ldr r2, [lr, r2, lsl #1] @ pack=*(unsigned int *)(Pico.vram+addr); // Get 8 pixels\r
399 mov r9, r7 @ remember code\r
400 tst r2, r2\r
740da8c6 401 beq .dsloop\r
402 orr r10, r10, #1<<22\r
403\r
404.DrawStrip_hiprio:\r
405 tst r10, #0x00c00000\r
406 beq .DrawStrip_hiprio_maybempt\r
407 sub r0, r1, r11\r
408 orr r7, r7, r0, lsl #16\r
409 orr r7, r7, r10, lsl #25 @ (ty<<25)\r
410 tst r7, #0x1000\r
411 eorne r7, r7, #7<<26 @ if(code&0x1000) cval^=7<<26;\r
412 str r7, [r6], #4 @ cache hi priority tile\r
413 mov r0, #0xf\r
cc68a136 414 b .dsloop\r
415\r
416.dsloop_exit:\r
740da8c6 417 tst r10, #1<<21 @ seen non hi-prio tile\r
418 ldreq r1, =rendstatus\r
cc68a136 419 mov r0, #0\r
740da8c6 420 ldreq r2, [r1]\r
cc68a136 421 str r0, [r6] @ terminate the cache list\r
740da8c6 422 orreq r2, r2, #0x40 @ had a layer with all hi-prio tiles\r
423 streq r2, [r1]\r
cc68a136 424\r
425 ldmfd sp!, {r4-r11,lr}\r
426 bx lr\r
427\r
6d7acf9e 428@ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r
cc68a136 429\r
430.DrawStrip_vsscroll:\r
6d7acf9e 431 rsb r8, r3, #0\r
432 mov r8, r8, lsr #3 @ r8=tilex=(-ts->hscroll)>>3\r
433 bic r8, r8, #0xff000000\r
740da8c6 434 orr r8, r8, r5, lsl #25 @ r8=(xmask[31:25]|had_output[24]|tilex[15:0])\r
cc68a136 435\r
6d7acf9e 436 ldr r4, =Scanline\r
437 orr r5, r1, r10, lsl #24\r
438 ldr r4, [r4]\r
439 sub r1, r3, #1\r
440 orr r5, r5, r4, lsl #16 @ r5=(shift_width[31:24]|scanline[23:16]|ymask[15:0])\r
441 and r1, r1, #7\r
442 add r7, r1, #1 @ r7=dx=((ts->hscroll-1)&7)+1\r
cc68a136 443\r
6d7acf9e 444 mov r10,r9, lsl #16\r
445 tst r0, r0\r
446 orrne r10,r10, #0x8000\r
447 tst r9, #1<<31\r
448 mov r3, #0\r
449 orr r10,r10, #0xff000000 @ will be adjusted on entering loop\r
450 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
451 movne r3, #0x40 @ default to shadowed pal on sh mode\r
cc68a136 452\r
6d7acf9e 453 mvn r9, #0 @ r9=prevcode=-1\r
cc68a136 454\r
6d7acf9e 455 @ cache some stuff to avoid mem access\r
456 ldr r11,=HighCol\r
457 mov r0, #0xf\r
458 add r1, r11, r7 @ r1=pdest\r
459\r
460 cmp r7, #8\r
461 subne r10,r10, #0x01000000 @ have hscroll, start with negative cell\r
462\r
463\r
464 @ r4 & r7 are scratch in this loop\r
465.dsloop_vs_subr1:\r
466 sub r1, r1, #8\r
467.dsloop_vs: @ 40-41 times\r
468 add r10,r10, #0x01000000\r
469 and r4, r10, #0x003f0000\r
470 cmp r4, r10, asr #8\r
740da8c6 471 ble .dsloop_vs_exit\r
6d7acf9e 472\r
473 @ calc offset and read tileline code to r7, also calc ty\r
474 add r7, lr, #0x012000\r
475 add r7, r7, #0x000180 @ r7=Pico.vsram (Pico+0x22180)\r
476 add r7, r7, r10,asr #23 @ vsram + ((cell&~1)<<1)\r
477 bic r7, r7, #3\r
478 tst r10,#0x8000 @ plane1?\r
479 addne r7, r7, #2\r
480 ldrh r7, [r7] @ r7=vscroll\r
481\r
482 bic r10,r10,#0xff @ clear old ty\r
483 and r4, r5, #0xff0000\r
484 add r4, r4, r7, lsl #16\r
485 and r4, r4, r5, lsl #16 @ r4=line<<16\r
486 and r7, r4, #0x70000\r
487 orr r10,r10,r7, lsr #15 @ new ty\r
488\r
489 mov r4, r4, lsr #19\r
490 mov r7, r5, lsr #24\r
491 mov r4, r4, lsl r7 @ nametabadd\r
492\r
493 and r7, r8, r8, lsr #25\r
494 add r7, lr, r7, lsl #1 @ Pico.vram+((tilex&ts->xmask) as halfwords)\r
495 add r7, r7, r4, lsl #1\r
496 ldrh r7, [r7, r12] @ r7=code (int, but from unsigned, no sign extend)\r
497\r
498 add r1, r1, #8\r
499 add r8, r8, #1\r
500\r
501 tst r7, #0x8000\r
502 bne .DrawStrip_vs_hiprio\r
503\r
504 cmp r7, r9\r
505 beq .DrawStrip_vs_samecode @ we know stuff about this tile already\r
506\r
507 mov r9, r7 @ remember code\r
740da8c6 508 orr r8, r8, #1<<24 @ seen non hi-prio tile\r
6d7acf9e 509\r
510 movs r2, r9, lsl #20 @ if (code&0x1000)\r
511 mov r2, r2, lsl #1\r
512 add r2, r2, r10, lsl #17\r
513 mov r2, r2, lsr #17\r
514 eorcs r2, r2, #0x0e @ if (code&0x1000) addr^=0xe;\r
515\r
516 ldr r2, [lr, r2, lsl #1] @ pack=*(unsigned int *)(Pico.vram+addr); // Get 8 pixels\r
517\r
518 bic r7, r3, #0x3f\r
519 and r3, r9, #0x6000\r
520 add r3, r7, r3, lsr #9 @ r3=pal=((code&0x6000)>>9);\r
521\r
522.DrawStrip_vs_samecode:\r
523 tst r2, r2\r
524 beq .dsloop_vs @ tileline blank\r
525\r
526 cmp r2, r2, ror #4\r
527 beq .DrawStrip_vs_SingleColor @ tileline singlecolor \r
528\r
529 tst r9, #0x0800\r
530 beq .DrawStrip_vs_TileNorm\r
531\r
532 @ (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r0: helper pattern\r
533 TileFlip r0\r
534 b .dsloop_vs\r
535\r
536.DrawStrip_vs_TileNorm:\r
537 TileNorm r0\r
538 b .dsloop_vs\r
539\r
540.DrawStrip_vs_SingleColor:\r
541 and r4, r2, #0xf\r
542 orr r4, r3, r4\r
543 orr r4, r4, r4, lsl #8\r
544 tst r1, #1 @ not aligned?\r
545 strneb r4, [r1], #1\r
546 streqh r4, [r1], #2\r
547 strh r4, [r1], #2\r
548 strh r4, [r1], #2\r
549 strh r4, [r1], #2\r
550 strneb r4, [r1], #1 @ have a remaining unaligned pixel?\r
551 b .dsloop_vs_subr1\r
552\r
553.DrawStrip_vs_hiprio:\r
554 tst r10, #0x00c00000\r
555 beq .DrawStrip_vs_hiprio_maybempt\r
556 sub r0, r1, r11\r
557 orr r7, r7, r0, lsl #16\r
558 orr r7, r7, r10, lsl #25 @ (ty<<25)\r
559 tst r7, #0x1000\r
560 eorne r7, r7, #7<<26 @ if(code&0x1000) cval^=7<<26;\r
561 str r7, [r6], #4 @ cache hi priority tile\r
562 mov r0, #0xf\r
563 b .dsloop_vs\r
564\r
565.DrawStrip_vs_hiprio_maybempt:\r
566 cmp r7, r9\r
567 beq .dsloop_vs @ must've been empty, otherwise we wouldn't get here\r
568 movs r2, r7, lsl #20 @ if (code&0x1000)\r
569 mov r2, r2, lsl #1\r
570 add r2, r2, r10, lsl #17\r
571 mov r2, r2, lsr #17\r
572 eorcs r2, r2, #0x0e @ if (code&0x1000) addr^=0xe;\r
573 ldr r2, [lr, r2, lsl #1] @ pack=*(unsigned int *)(Pico.vram+addr); // Get 8 pixels\r
574 mov r9, r7 @ remember code\r
575 tst r2, r2\r
576 orrne r10, r10, #1<<22\r
577 bne .DrawStrip_vs_hiprio\r
578 b .dsloop_vs\r
579\r
740da8c6 580.dsloop_vs_exit:\r
581 tst r8, #1<<24 @ seen non hi-prio tile\r
582 ldreq r1, =rendstatus\r
583 mov r0, #0\r
584 ldreq r2, [r1]\r
585 str r0, [r6] @ terminate the cache list\r
586 orreq r2, r2, #0x40 @ had a layer with all hi-prio tiles\r
587 streq r2, [r1]\r
588\r
589 ldmfd sp!, {r4-r11,lr}\r
590 bx lr\r
591\r
6d7acf9e 592\r
593@ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r
cc68a136 594\r
595@ interlace mode 2? Sonic 2?\r
596.DrawStrip_interlace:\r
597 tst r0, r0\r
598 moveq r7, r7, lsl #21\r
599 movne r7, r7, lsl #5\r
600\r
601 @ Find the line in the name table\r
602 add r2, r7, r2, lsl #22 @ r2=(vscroll+(Scanline<<1))<<21 (11 bits);\r
603 orr r1, r1, #0x80000000\r
604 and r2, r2, r1, ror #10 @ &((ymask<<1)|1)<<21;\r
605 mov r2, r2, lsr #21\r
606 mov r4, r2, lsr #4\r
607 mov r12, r12, lsr #1 @ halfwords\r
608 add r0, r12, r4, lsl r10 @ nametab+=(ts.line>>4)<<shift[width];\r
609 and r9, r9, #0xff\r
610\r
611 sub sp, sp, #6*4\r
612 stmia sp, {r0,r2,r3,r5,r6,r9}\r
613\r
614 mov r0, sp\r
615 bl DrawStripInterlace @ struct TileStrip *ts\r
616\r
617 add sp, sp, #6*4\r
618 ldmfd sp!, {r4-r11,lr}\r
619 bx lr\r
620\r
621.pool\r
622\r
623@ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r
624\r
625\r
626.global BackFill @ int reg7, int sh\r
627\r
628BackFill:\r
629 stmfd sp!, {r4-r9,lr}\r
630\r
631 ldr lr, =(HighCol+8)\r
632\r
633 mov r0, r0, lsl #26\r
634 mov r0, r0, lsr #26\r
635 orr r0, r0, r1, lsl #6\r
636 orr r0, r0, r0, lsl #8\r
637 orr r0, r0, r0, lsl #16\r
638\r
639 mov r1, r0\r
640 mov r2, r0\r
641 mov r3, r0\r
642 mov r4, r0\r
643 mov r5, r0\r
644 mov r6, r0\r
645 mov r7, r0\r
646\r
647 @ go go go!\r
648 stmia lr!, {r0-r7} @ 10*8*4\r
649 stmia lr!, {r0-r7}\r
650 stmia lr!, {r0-r7}\r
651 stmia lr!, {r0-r7}\r
652 stmia lr!, {r0-r7}\r
653 stmia lr!, {r0-r7}\r
654 stmia lr!, {r0-r7}\r
655 stmia lr!, {r0-r7}\r
656 stmia lr!, {r0-r7}\r
657 stmia lr!, {r0-r7}\r
658\r
659 ldmfd sp!, {r4-r9,r12}\r
660 bx r12\r
661\r
662\r
663@ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r
664\r
665\r
666.global DrawTilesFromCache @ int *hc, int sh\r
667\r
668DrawTilesFromCache:\r
669 stmfd sp!, {r4-r8,r11,lr}\r
670\r
cc68a136 671 @ cache some stuff to avoid mem access\r
672 ldr r11,=HighCol\r
673 ldr lr, =(Pico+0x10000) @ lr=Pico.vram\r
674 mov r12,#0xf\r
675\r
740da8c6 676 mvn r5, #0 @ r5=prevcode=-1\r
677 movs r8, r1\r
678 bne .dtfc_check_rendflags\r
679\r
cc68a136 680 @ scratch: r4, r7\r
681.dtfc_loop:\r
682 ldr r6, [r0], #4 @ read code\r
683 movs r1, r6, lsr #16 @ r1=dx;\r
684 ldmeqfd sp!, {r4-r8,r11,pc} @ dx is never zero, this must be a terminator, return\r
685 bic r1, r1, #0xfe00\r
686 add r1, r11, r1 @ r1=pdest\r
687\r
688 mov r7, r6, lsl #16\r
689 cmp r5, r7, lsr #16\r
690 beq .dtfc_samecode @ if (code==prevcode)\r
691\r
692 mov r5, r7, lsr #16\r
693\r
694 mov r2, r5, lsl #21\r
695 mov r2, r2, lsr #17 @ r2=addr=(code&0x7ff)<<4;\r
696 add r2, r2, r6, lsr #25 @ addr+=ty\r
697\r
698 and r3, r5, #0x6000\r
699 mov r3, r3, lsr #9 @ r3=pal=((code&0x6000)>>9);\r
700\r
701 ldr r2, [lr, r2, lsl #1] @ pack=*(unsigned int *)(Pico.vram+addr); // Get 8 pixels\r
702\r
703.dtfc_samecode:\r
704 tst r8, r8\r
705 bne .dtfc_shadow\r
706\r
707 tst r2, r2\r
708 beq .dtfc_loop\r
709\r
710 cmp r2, r2, ror #4\r
711 beq .dtfc_SingleColor @ tileline singlecolor \r
712\r
713 tst r5, #0x0800\r
714 beq .dtfc_TileNorm\r
715\r
716 @ (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r12: helper pattern\r
717 TileFlip r12\r
718 b .dtfc_loop\r
719\r
720.dtfc_TileNorm:\r
721 TileNorm r12\r
722 b .dtfc_loop\r
723\r
724.dtfc_SingleColor:\r
725 and r4, r2, #0xf\r
726 orr r4, r3, r4\r
727 orr r4, r4, r4, lsl #8\r
728 tst r1, #1 @ not aligned?\r
729 strneb r4, [r1], #1\r
730 streqh r4, [r1], #2\r
731 strh r4, [r1], #2\r
732 strh r4, [r1], #2\r
733 strh r4, [r1], #2\r
734 strneb r4, [r1], #1 @ have a remaining unaligned pixel?\r
735 b .dtfc_loop\r
736\r
737.dtfc_shadow:\r
738 tst r2, r2\r
739 beq .dtfc_shadow_blank\r
740\r
741 cmp r2, r2, ror #4\r
742 beq .dtfc_SingleColor @ tileline singlecolor \r
743\r
744 tst r5, #0x0800\r
745 beq .dtfc_TileNormShHP\r
746\r
747 @ (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r12: helper pattern\r
748 TileFlipShHP\r
749 b .dtfc_loop\r
750\r
751.dtfc_TileNormShHP:\r
752 TileNormShHP\r
753 b .dtfc_loop\r
754\r
755.dtfc_shadow_blank:\r
756 ldrb r4, [r1] @ 1ci\r
757 ldrb r12,[r1,#1]\r
758 tst r4, #0x80\r
759 andeq r4, r4,#0x3f\r
760 streqb r4, [r1]\r
761 tst r12,#0x80\r
762 ldrb r4, [r1,#2]\r
763 andeq r12,r12,#0x3f\r
764 streqb r12,[r1,#1]\r
765 tst r4, #0x80\r
766 ldrb r12,[r1,#3]\r
767 andeq r4, r4,#0x3f\r
768 streqb r4, [r1,#2]\r
769 tst r12,#0x80\r
770 ldrb r4, [r1,#4]\r
771 andeq r12,r12,#0x3f\r
772 streqb r12,[r1,#3]\r
773 tst r4, #0x80\r
774 ldrb r12,[r1,#5]\r
775 andeq r4, r4,#0x3f\r
776 streqb r4, [r1,#4]\r
777 tst r12,#0x80\r
778 ldrb r4, [r1,#6]\r
779 andeq r12,r12,#0x3f\r
780 streqb r12,[r1,#5]\r
781 tst r4, #0x80\r
782 ldrb r12,[r1,#7]\r
783 andeq r4, r4,#0x3f\r
784 streqb r4, [r1,#6]\r
785 tst r12,#0x80\r
786 andeq r12,r12,#0x3f\r
787 streqb r12,[r1,#7]\r
788 mov r12, #0xf\r
789 b .dtfc_loop\r
790\r
740da8c6 791@ check if we have detected layer covered with hi-prio tiles:\r
792.dtfc_check_rendflags:\r
793 ldr r1, =rendstatus\r
794 ldr r2, [r1]\r
795 tst r2, #0xc0\r
796 beq .dtfc_loop\r
797 mov r8, #0 @ sh/hi mode off\r
798 tst r2, #0x80\r
799 bne .dtfc_loop @ already processed\r
800 orr r2, r2, #0x80\r
801 str r2, [r1]\r
802\r
803 add r1, r11,#8\r
804 mov r3, #320/4\r
805 mov r7, #0x80\r
806 orr r7, r7, r7, lsl #8\r
807 orr r7, r7, r7, lsl #16\r
808 mov r6, #0x3f\r
809 orr r6, r6, r6, lsl #8\r
810 orr r6, r6, r6, lsl #16\r
811.dtfc_loop_shprep:\r
812 subs r3, r3, #1\r
813 bmi .dtfc_loop @ done\r
814 ldr r2, [r1]\r
815 tst r2, r7\r
816 andeq r2, r2, r6\r
817 streq r2, [r1], #4\r
818 beq .dtfc_loop_shprep\r
819 tst r2, #0x80000000\r
820 biceq r2, r2, #0xc0000000\r
821 tst r2, #0x00800000\r
822 biceq r2, r2, #0x00c00000\r
823 tst r2, #0x00008000\r
824 biceq r2, r2, #0x0000c000\r
825 tst r2, #0x00000080\r
826 biceq r2, r2, #0x000000c0\r
827 str r2, [r1], #4\r
828 b .dtfc_loop_shprep\r
829\r
cc68a136 830.pool\r
831\r
832@ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r
833\r
834\r
835.global DrawSpritesFromCache @ int *hc, int sh\r
836\r
837DrawSpritesFromCache:\r
838 stmfd sp!, {r4-r11,lr}\r
839\r
840 @ cache some stuff to avoid mem access\r
841 ldr r11,=HighCol\r
842 ldr lr, =(Pico+0x10000) @ lr=Pico.vram\r
843 mov r6, r1, lsl #31\r
844 orr r6, r6, #1<<30\r
845 mov r12,#0xf\r
846\r
847 mov r10, r0\r
848\r
849.dsfc_loop:\r
850 ldr r9, [r10], #4 @ read code\r
851 bic r6, r6, #7 @ using pipeline\r
852 tst r9, r9\r
853 ldmeqfd sp!, {r4-r11,pc}\r
854\r
855 mov r4, r9, lsl #28\r
856 orr r6, r6, r4, lsr #30\r
857 add r6, r6, #1 @ r6=s1cc???? ... ?????www (s=shadow/hilight, cc=pal, w=width)\r
858\r
859 and r5, r9, #3\r
860 add r5, r5, #1 @ r5=delta\r
861 tst r9, #0x10000\r
862 rsbne r5, r5, #0 @ Flip X\r
863 mov r5, r5, lsl #4\r
864\r
865 mov r2, r9, lsr #17\r
866 mov r8, r2, lsl #1 @ tile=((unsigned int)code>>17)<<1;\r
867\r
868 and r3, r9, #0x30 @ r3=pal=(code&0x30);\r
869\r
870 bic r6, r6, #3<<28\r
871 orr r6, r6, r3, lsl #24\r
872\r
873 mov r0, r9, lsl #16\r
874 mov r0, r0, asr #22 @ sx=(code<<16)>>22\r
875 adds r0, r0, #0 @ set ZV\r
876 b .dsfc_inloop_enter\r
877\r
878@ scratch: r4, r7\r
879.dsfc_inloop:\r
880 sub r6, r6, #1\r
881 tst r6, #7\r
882 beq .dsfc_loop\r
883 adds r0, r0, #8\r
884 add r8, r8, r5\r
885\r
886.dsfc_inloop_enter:\r
887 ble .dsfc_inloop\r
888 cmp r0, #328\r
889 bge .dsfc_loop\r
890\r
891 mov r8, r8, lsl #17\r
892 mov r8, r8, lsr #17 @ tile&=0x7fff; // Clip tile address\r
893\r
894 ldr r2, [lr, r8, lsl #1] @ pack=*(unsigned int *)(Pico.vram+tile); // Get 8 pixels\r
895 add r1, r11, r0 @ r1=pdest\r
896 tst r2, r2\r
897 beq .dsfc_inloop\r
898\r
899 cmp r12, r6, lsr #28\r
900 beq .dsfc_shadow\r
901\r
902 cmp r2, r2, ror #4\r
903 beq .dsfc_SingleColor @ tileline singlecolor \r
904\r
905 tst r9, #0x10000\r
906 beq .dsfc_TileNorm\r
907\r
908 @ TileFlip (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r12: helper pattern\r
909 TileFlip r12\r
910 b .dsfc_inloop\r
911\r
912.dsfc_TileNorm:\r
913 TileNorm r12\r
914 b .dsfc_inloop\r
915\r
916.dsfc_SingleColor:\r
917 tst r0, #1 @ not aligned?\r
918 and r4, r2, #0xf\r
919 orr r4, r3, r4\r
920 orr r4, r4, r4, lsl #8\r
921 strneb r4, [r1], #1\r
922 streqh r4, [r1], #2\r
923 strh r4, [r1], #2\r
924 strh r4, [r1], #2\r
925 strh r4, [r1], #2\r
926 strneb r4, [r1], #1\r
927 b .dsfc_inloop\r
928\r
929.dsfc_shadow:\r
930 cmp r2, r2, ror #4\r
931 beq .dsfc_singlec_sh\r
932\r
933 tst r9, #0x10000\r
934 beq .dsfc_TileNorm_sh\r
935\r
936 @ (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r12: helper pattern\r
937 TileFlipSh\r
938 b .dsfc_inloop\r
939\r
940.dsfc_TileNorm_sh:\r
941 TileNormSh\r
942 b .dsfc_inloop\r
943\r
944.dsfc_singlec_sh:\r
945 cmp r2, #0xe0000000\r
946 bcc .dsfc_SingleColor @ normal singlecolor tileline (carry inverted in ARM)\r
947 tst r2, #0x10000000\r
948 bne .dsfc_sh_sh\r
949 TileSingleHi\r
950 b .dsfc_inloop\r
951\r
952.dsfc_sh_sh:\r
953 TileSingleSh\r
954 b .dsfc_inloop\r
955\r
956.pool\r
957\r
958@ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r
959\r
960@ + 0 : hhhhvvvv ab--hhvv yyyyyyyy yyyyyyyy // a: offscreen h, b: offs. v, h: horiz. size\r
961@ + 4 : xxxxxxxx xxxxxxxx pccvhnnn nnnnnnnn // x: x coord + 8\r
962\r
963.global DrawSprite @ unsigned int *sprite, int **hc, int sh\r
964\r
965DrawSprite:\r
966 stmfd sp!, {r4-r9,r11,lr}\r
967\r
968 ldr r3, [r0] @ sprite[0]\r
969 ldr r7, =Scanline\r
970 mov r6, r3, lsr #28\r
971 sub r6, r6, #1 @ r6=width-1 (inc later)\r
972 mov r5, r3, lsr #24\r
973 and r5, r5, #7 @ r5=height\r
974\r
975 mov r4, r3, lsl #16 @ r4=sy<<16 (tmp)\r
976\r
977 ldr r7, [r7]\r
978 ldr r9, [r0, #4]\r
979 sub r7, r7, r4, asr #16 @ r7=row=Scanline-sy\r
980\r
981 tst r2, r2\r
982 mov r2, r9, asr #16 @ r2=sx\r
983 bic r9, r9, #0xfe000000\r
984 orrne r9, r9, #1<<31 @ r9=code|(sh<<31)\r
985\r
986 tst r9, #0x1000\r
987 movne r4, r5, lsl #3\r
988 subne r4, r4, #1\r
989 subne r7, r4, r7 @ if (code&0x1000) row=(height<<3)-1-row; // Flip Y\r
990\r
991 mov r8, r9, lsl #21\r
992 mov r8, r8, lsr #21\r
993 add r8, r8, r7, lsr #3 @ tile+=row>>3; // Tile number increases going down\r
994 \r
995 tst r9, #0x0800\r
996 mlane r8, r5, r6, r8 @ if (code&0x0800) { tile+=delta*(width-1);\r
997 rsbne r5, r5, #0 @ delta=-delta; } // r5=delta now\r
998\r
999 mov r8, r8, lsl #4\r
1000 and r7, r7, #7\r
1001 add r8, r8, r7, lsl #1 @ tile+=(row&7)<<1; // Tile address\r
1002\r
1003 tst r9, #0x8000\r
1004 bne .dspr_cache @ if(code&0x8000) // high priority - cache it\r
1005\r
1006 @ cache some stuff to avoid mem access\r
1007 ldr r11,=HighCol\r
1008 ldr lr, =(Pico+0x10000) @ lr=Pico.vram\r
1009 mov r12,#0xf\r
1010\r
1011 mov r5, r5, lsl #4 @ delta<<=4; // Delta of address\r
1012 and r4, r9, #0x6000\r
1013 orr r9, r9, r4, lsl #16\r
1014 orr r9, r9, #0x10000000 @ r9=scc1 ???? ... <code> (s=shadow/hilight, cc=pal)\r
1015\r
1016 tst r9, #1<<31\r
1017 mov r3, r4, lsr #9 @ r3=pal=((code>>9)&0x30);\r
1018 orrne r3, r3, #0x40 @ shadow by default\r
1019\r
1020 add r6, r6, #1 @ inc now\r
1021 adds r0, r2, #0 @ mov sx to r0 and set ZV flags\r
1022 b .dspr_loop_enter\r
1023\r
1024.dspr_loop:\r
1025 subs r6, r6, #1 @ width--\r
1026 ldmeqfd sp!, {r4-r9,r11,pc}@ return\r
1027 adds r0, r0, #8 @ sx+=8\r
1028 add r8, r8, r5 @ tile+=delta\r
1029\r
1030.dspr_loop_enter:\r
1031 ble .dspr_loop @ sx <= 0\r
1032 cmp r0, #328\r
1033 ldmgefd sp!, {r4-r9,r11,pc}@ return\r
1034\r
1035 mov r8, r8, lsl #17\r
1036 mov r8, r8, lsr #17 @ tile&=0x7fff; // Clip tile address\r
1037\r
1038 ldr r2, [lr, r8, lsl #1] @ pack=*(unsigned int *)(Pico.vram+addr); // Get 8 pixels\r
1039 add r1, r11, r0 @ r1=pdest\r
1040 tst r2, r2\r
1041 beq .dspr_loop\r
1042\r
1043 cmp r12, r9, lsr #28\r
1044 beq .dspr_shadow\r
1045\r
1046 cmp r2, r2, ror #4\r
1047 beq .dspr_SingleColor @ tileline singlecolor \r
1048\r
1049 tst r9, #0x0800\r
1050 beq .dspr_TileNorm\r
1051\r
1052 @ (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r12: helper pattern\r
1053 TileFlip r12\r
1054 b .dspr_loop\r
1055\r
1056@ scratch: r4, r7\r
1057.dspr_TileNorm:\r
1058 TileNorm r12\r
1059 b .dspr_loop\r
1060\r
1061.dspr_SingleColor:\r
1062 and r4, r2, #0xf\r
1063 orr r4, r3, r4\r
1064 orr r4, r4, r4, lsl #8\r
1065 tst r0, #1 @ not aligned?\r
1066 strneb r4, [r1], #1\r
1067 streqh r4, [r1], #2\r
1068 strh r4, [r1], #2\r
1069 strh r4, [r1], #2\r
1070 strh r4, [r1], #2\r
1071 strneb r4, [r1], #1\r
1072 b .dspr_loop\r
1073\r
1074.dspr_shadow:\r
1075 cmp r2, r2, ror #4\r
1076 beq .dspr_singlec_sh\r
1077\r
1078 tst r9, #0x0800\r
1079 beq .dspr_TileNorm_sh\r
1080\r
1081 @ (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r12: helper pattern\r
1082 TileFlipSh\r
1083 b .dspr_loop\r
1084\r
1085.dspr_TileNorm_sh:\r
1086 TileNormSh\r
1087 b .dspr_loop\r
1088\r
1089.dspr_singlec_sh:\r
1090 cmp r2, #0xe0000000\r
1091 bcc .dspr_SingleColor @ normal tileline\r
1092 tst r2, #0x10000000\r
1093 bne .dspr_sh_sh\r
1094 TileSingleHi\r
1095 b .dspr_loop\r
1096\r
1097.dspr_sh_sh:\r
1098 TileSingleSh\r
1099 b .dspr_loop\r
1100\r
1101\r
1102.dspr_cache:\r
1103 @ *(*hc)++ = (tile<<16)|((code&0x0800)<<5)|((sx<<6)&0x0000ffc0)|((code>>9)&0x30)|((sprite[0]>>24)&0xf);\r
1104 mov r4, r8, lsl #16 @ tile\r
1105 tst r9, #0x0800\r
1106 orrne r4, r4, #0x10000 @ code&0x0800\r
1107 mov r2, r2, lsl #22\r
1108 orr r4, r4, r2, lsr #16 @ (sx<<6)&0x0000ffc0\r
1109 and r2, r9, #0x6000\r
1110 orr r4, r4, r2, lsr #9 @ (code>>9)&0x30\r
1111 mov r3, r3, lsl #12\r
1112 ldr r2, [r1]\r
1113 orr r4, r4, r3, lsr #28 @ (sprite[0]>>24)&0xf\r
1114\r
1115 str r4, [r2], #4\r
1116 str r2, [r1]\r
1117\r
1118 ldmfd sp!, {r4-r9,r11,lr}\r
1119 bx lr\r
1120\r
1121@ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r
1122\r
1123.global DrawWindow @ int tstart, int tend, int prio, int sh // int *hcache\r
1124\r
1125DrawWindow:\r
1126 stmfd sp!, {r4-r11,lr}\r
1127\r
1128 ldr r11, =(Pico+0x22228) @ Pico.video\r
1129 ldr r10, =Scanline\r
1130 ldrb r12, [r11, #3] @ pvid->reg[3]\r
1131\r
1132 ldr r10, [r10]\r
1133 ldr r4, [r11, #12]\r
1134 mov r5, r10, lsr #3\r
1135 and r10, r10, #7\r
1136 mov r10, r10, lsl #1 @ r10=ty\r
1137\r
1138 mov r12, r12, lsl #10\r
1139\r
1140 tst r4, #1 @ 40 cell mode?\r
1141 andne r12, r12, #0xf000 @ 0x3c<<10\r
1142 andeq r12, r12, #0xf800\r
1143 addne r12, r12, r5, lsl #7\r
1144 addeq r12, r12, r5, lsl #6 @ nametab\r
1145 add r12, r12, r0, lsl #2 @ +starttile\r
1146\r
1147 ldr r6, =rendstatus\r
1148 ldr lr, =(Pico+0x10000) @ lr=Pico.vram\r
1149 ldrb r6, [r6]\r
1150\r
1151 @ fetch the first code now\r
1152 ldrh r7, [lr, r12]\r
1153\r
1154 ands r6, r6, #2 @ we care about bit 1 only\r
1155 orr r6, r6, r2\r
1156 bne .dw_no_sameprio\r
1157\r
1158 cmp r2, r7, lsr #15\r
1159 ldmnefd sp!, {r4-r11,pc} @ assume that whole window uses same priority\r
1160\r
1161.dw_no_sameprio:\r
1162 orr r6, r6, r3, lsl #8 @ shadow mode\r
1163\r
1164 sub r8, r1, r0\r
1165 mov r8, r8, lsl #1 @ cells\r
1166\r
1167 mvn r9, #0 @ r9=prevcode=-1\r
1168\r
1169 @ cache some stuff to avoid mem access\r
1170 ldr r11,=(HighCol+8)\r
1171 add r1, r11, r0, lsl #4 @ r1=pdest\r
1172 mov r0, #0xf\r
1173 b .dwloop_enter\r
1174\r
1175 @ r4,r5 & r7 are scratch in this loop\r
1176.dwloop:\r
1177 add r1, r1, #8\r
1178.dwloop_nor1:\r
1179 add r12, r12, #2 @ halfwords\r
1180 ldrh r7, [lr, r12] @ r7=code (int, but from unsigned, no sign extend)\r
1181 subs r8, r8, #1\r
1182 beq .dwloop_end @ done\r
1183\r
1184 eor r5, r6, r7, lsr #15\r
1185 tst r5, #1\r
1186 orrne r6, r6, #2 @ wrong pri\r
1187 bne .dwloop\r
1188\r
1189 cmp r7, r9\r
1190 beq .dw_samecode @ we know stuff about this tile already\r
1191\r
1192.dwloop_enter:\r
1193 mov r9, r7 @ remember code\r
1194\r
1195 movs r2, r9, lsl #20 @ if (code&0x1000)\r
1196 mov r2, r2, lsl #1\r
1197 add r2, r10, r2, lsr #17 @ r2=addr=(code&0x7ff)<<4; addr+=ty\r
1198 eorcs r2, r2, #0x0e @ if (code&0x1000) addr^=0xe;\r
1199\r
1200 and r3, r9, #0x6000\r
1201 mov r3, r3, lsr #9 @ r3=pal=((code&0x6000)>>9);\r
1202\r
1203 ldr r2, [lr, r2, lsl #1] @ pack=*(unsigned int *)(Pico.vram+addr); // Get 8 pixels\r
1204\r
1205.dw_samecode:\r
1206 tst r6, #0x100\r
1207 bne .dw_shadow\r
1208.dw_shadow_done:\r
1209 tst r2, r2\r
1210 beq .dwloop @ tileline blank\r
1211\r
1212 cmp r2, r2, ror #4\r
1213 beq .dw_SingleColor @ tileline singlecolor \r
1214\r
1215 tst r9, #0x0800\r
1216 beq .dw_TileNorm\r
1217\r
1218 @ (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r0: helper pattern\r
1219 TileFlip r0\r
1220 b .dwloop\r
1221\r
1222.dw_TileNorm:\r
1223 TileNorm r0\r
1224 b .dwloop\r
1225\r
1226.dw_SingleColor:\r
1227 and r4, r0, r2 @ #0x0000000f\r
1228 orr r4, r3, r4\r
1229 orr r4, r4, r4, lsl #8\r
1230 orr r4, r4, r4, lsl #16\r
1231 mov r5, r4\r
1232 stmia r1!, {r4,r5}\r
1233 b .dwloop_nor1 @ we incremeted r1 ourselves\r
1234\r
1235.dw_shadow:\r
1236 tst r6, #1 @ hi pri?\r
1237 orreq r3, r3, #0x40\r
1238 beq .dw_shadow_done\r
1239 ldr r4, [r1]\r
1240 tst r4, #0x00000080\r
1241 biceq r4, r4, #0x000000c0\r
1242 tst r4, #0x00008000\r
1243 biceq r4, r4, #0x0000c000\r
1244 tst r4, #0x00800000\r
1245 biceq r4, r4, #0x00c00000\r
1246 tst r4, #0x80000000\r
1247 biceq r4, r4, #0xc0000000\r
1248 str r4, [r1]\r
1249 ldr r4, [r1,#4]\r
1250 tst r4, #0x00000080\r
1251 biceq r4, r4, #0x000000c0\r
1252 tst r4, #0x00008000\r
1253 biceq r4, r4, #0x0000c000\r
1254 tst r4, #0x00800000\r
1255 biceq r4, r4, #0x00c00000\r
1256 tst r4, #0x80000000\r
1257 biceq r4, r4, #0xc0000000\r
1258 str r4, [r1,#4]\r
1259 b .dw_shadow_done\r
1260\r
1261.dwloop_end:\r
1262 ldr r0, =rendstatus\r
1263 ldr r1, [r0]\r
1264 and r6, r6, #2\r
1265 orr r1, r1, r6\r
1266 str r1, [r0]\r
1267\r
1268 ldmfd sp!, {r4-r11,r12}\r
1269 bx r12\r
1270\r
1271\r
1272@ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r
1273\r
1274\r
1275@ hilights 2 pixels in RGB444/BGR444 format\r
1276.macro TileDoShHi2Pixels444 reg\r
1277 mov \reg, \reg, ror #12\r
1278 adds \reg, \reg, #0x40000000\r
1279 orrcs \reg, \reg, #0xf0000000\r
1280 mov \reg, \reg, ror #28\r
1281 adds \reg, \reg, #0x40000000\r
1282 orrcs \reg, \reg, #0xf0000000\r
1283 mov \reg, \reg, ror #28\r
1284 adds \reg, \reg, #0x40000000\r
1285 orrcs \reg, \reg, #0xf0000000\r
1286 mov \reg, \reg, ror #24\r
1287 adds \reg, \reg, #0x40000000\r
1288 orrcs \reg, \reg, #0xf0000000\r
1289 mov \reg, \reg, ror #28\r
1290 adds \reg, \reg, #0x40000000\r
1291 orrcs \reg, \reg, #0xf0000000\r
1292 mov \reg, \reg, ror #28\r
1293 adds \reg, \reg, #0x40000000\r
1294 orrcs \reg, \reg, #0xf0000000\r
1295 mov \reg, \reg, ror #12\r
1296.endm\r
1297\r
1298\r
1299.global FinalizeLineBGR444 @ int sh\r
1300\r
1301FinalizeLineBGR444:\r
1302 stmfd sp!, {r4-r6,lr}\r
1303 mov r6, r0\r
1304 ldr lr, =(Pico+0x22228) @ Pico.video\r
1305 ldr r0, =DrawLineDest\r
1306 ldrb r12, [lr, #12]\r
1307 ldr r0, [r0]\r
1308 sub r3, lr, #0x128 @ r3=Pico.cram\r
1309\r
1310 tst r12, #1\r
1311 movne r2, #320/4 @ len\r
1312 bne .fl_no32colBGR444\r
1313 ldr r4, =PicoOpt\r
1314 mov r2, #256/4\r
1315 ldr r4, [r4]\r
1316 tst r4, #0x100\r
1317 addeq r0, r0, #32*2\r
1318\r
1319.fl_no32colBGR444:\r
1320 tst r6, r6\r
1321 beq .fl_noshBGR444\r
1322\r
1323 ldr r4, =HighPal\r
1324\r
1325 ldrb r12, [lr, #-0x1a] @ 0x2220e ~ dirtyPal\r
1326 tst r12, r12\r
1327 moveq r3, r4\r
1328 beq .fl_noshBGR444\r
1329 mov r12, #0\r
1330 strb r12, [lr, #-0x1a]\r
1331\r
1332 mov lr, #0x40/8\r
1333 @ copy pal:\r
1334.fl_loopcpBGR444:\r
1335 ldmia r3!, {r1,r5,r6,r12}\r
1336 subs lr, lr, #1\r
1337 stmia r4!, {r1,r5,r6,r12}\r
1338 bne .fl_loopcpBGR444\r
1339\r
1340 @ shadowed pixels:\r
1341 mov r12, #0x0077\r
1342 orr r12,r12,#0x0700\r
1343 orr r12,r12,r12,lsl #16\r
1344 sub r3, r3, #0x40*2\r
1345 add r5, r4, #0x80*2\r
1346 mov lr, #0x40/4\r
1347.fl_loopcpBGR444_sh:\r
1348 ldmia r3!, {r1,r6}\r
1349 subs lr, lr, #1\r
1350 and r1, r12, r1, lsr #1\r
1351 and r6, r12, r6, lsr #1\r
1352 stmia r4!, {r1,r6}\r
1353 stmia r5!, {r1,r6}\r
1354 bne .fl_loopcpBGR444_sh\r
1355\r
1356 @ hilighted pixels:\r
1357 sub r3, r3, #0x40*2\r
1358 mov lr, #0x40/2\r
1359.fl_loopcpBGR444_hi:\r
1360 ldr r1, [r3], #4\r
1361 TileDoShHi2Pixels444 r1\r
1362 str r1, [r4], #4\r
1363 subs lr, lr, #1\r
1364 bne .fl_loopcpBGR444_hi\r
1365\r
1366 sub r3, r4, #0x40*3*2\r
1367\r
1368\r
1369.fl_noshBGR444:\r
1370 ldr r1, =(HighCol+8)\r
1371 mov lr, #0xff\r
1372 mov lr, lr, lsl #1\r
1373\r
1374.fl_loopBGR444:\r
1375\r
1376 ldr r12, [r1], #4\r
1377 subs r2, r2, #1\r
1378\r
1379 and r4, lr, r12, lsl #1\r
1380 ldrh r4, [r3, r4]\r
1381 and r5, lr, r12, lsr #7\r
1382 ldrh r5, [r3, r5]\r
1383 and r6, lr, r12, lsr #15\r
1384 ldrh r6, [r3, r6]\r
1385 orr r4, r4, r5, lsl #16\r
1386\r
1387 and r5, lr, r12, lsr #23\r
1388 ldrh r5, [r3, r5] @ 2c.i.\r
1389 orr r5, r6, r5, lsl #16\r
1390\r
1391 stmia r0!, {r4,r5}\r
1392 bne .fl_loopBGR444\r
1393\r
1394\r
1395 ldmfd sp!, {r4-r6,lr}\r
1396 bx lr\r
1397\r
1398\r
1399@ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r
1400\r
1401\r
1402@ hilights 2 pixels in RGB555/BGR555 format\r
1403.macro TileDoShHi2Pixels555 reg\r
1404 adds \reg, \reg, #0x40000000\r
1405 orrcs \reg, \reg, #0xf0000000\r
1406 mov \reg, \reg, ror #27\r
1407 adds \reg, \reg, #0x40000000\r
1408 orrcs \reg, \reg, #0xf0000000\r
1409 mov \reg, \reg, ror #26\r
1410 adds \reg, \reg, #0x40000000\r
1411 orrcs \reg, \reg, #0xf0000000\r
1412 mov \reg, \reg, ror #27\r
1413 adds \reg, \reg, #0x40000000\r
1414 orrcs \reg, \reg, #0xf0000000\r
1415 mov \reg, \reg, ror #27\r
1416 adds \reg, \reg, #0x40000000\r
1417 orrcs \reg, \reg, #0xf0000000\r
1418 mov \reg, \reg, ror #26\r
1419 adds \reg, \reg, #0x40000000\r
1420 orrcs \reg, \reg, #0xf0000000\r
1421 mov \reg, \reg, ror #27\r
1422.endm\r
1423\r
1424\r
1425@ Convert 0000bbb0 ggg0rrr0\r
1426@ to rrrrrggg gggbbbbb\r
1427\r
1428@ r2,r3,r9 - scratch, lr = 0x001c001c, r8 = 0x00030003\r
1429.macro convRGB565 reg\r
1430 and r2, lr, \reg,lsl #1\r
1431 and r9, r8, \reg,lsr #2\r
1432 orr r2, r2, r9 @ r2=red\r
1433 and r3, lr, \reg,lsr #7\r
1434 and r9, r8, \reg,lsr #10\r
1435 orr r3, r3, r9 @ r3=blue\r
1436 and \reg, \reg, lr, lsl #3\r
1437 orr \reg, \reg, \reg,lsl #3 @ green\r
1438 orr \reg, \reg, r2, lsl #11 @ add red back\r
1439 orr \reg, \reg, r3 @ add blue back\r
1440.endm\r
1441\r
1442vidConvCpyRGB565: @ void *to, void *from, int pixels\r
1443 stmfd sp!, {r4-r9,lr}\r
1444\r
1445 mov r12, r2, lsr #3 @ repeats\r
1446 mov lr, #0x001c0000\r
1447 orr lr, lr, #0x01c @ lr == pattern 0x001c001c\r
1448 mov r8, #0x00030000\r
1449 orr r8, r8, #0x003 @ lr == pattern 0x001c001c\r
1450\r
1451.loopRGB565:\r
1452 ldmia r1!, {r4-r7}\r
1453 subs r12, r12, #1\r
1454 convRGB565 r4\r
1455 str r4, [r0], #4\r
1456 convRGB565 r5\r
1457 str r5, [r0], #4\r
1458 convRGB565 r6\r
1459 str r6, [r0], #4\r
1460 convRGB565 r7\r
1461 str r7, [r0], #4\r
1462\r
1463 bgt .loopRGB565\r
1464\r
1465 ldmfd sp!, {r4-r9,lr}\r
1466 bx lr\r
1467\r
1468\r
1469\r
1470.global FinalizeLineRGB555 @ int sh\r
1471\r
1472FinalizeLineRGB555:\r
1473 stmfd sp!, {r4-r8,lr}\r
3d48f143 1474 ldr r8, =(Pico+0x22228) @ Pico.video\r
cc68a136 1475 ldr r4, =HighPal\r
1476\r
3d48f143 1477 ldrb r7, [r8, #-0x1a] @ 0x2220e ~ dirtyPal\r
cc68a136 1478 mov r6, r0\r
1479 mov r1, #0\r
1480 tst r7, r7\r
1481 beq .fl_noconvRGB555\r
3d48f143 1482 strb r1, [r8, #-0x1a]\r
1483 sub r1, r8, #0x128 @ r1=Pico.cram\r
cc68a136 1484 mov r0, r4\r
1485 mov r2, #0x40\r
1486 bl vidConvCpyRGB565\r
1487\r
1488.fl_noconvRGB555:\r
cc68a136 1489 mov r3, r4\r
1490 tst r6, r6\r
1491 beq .fl_noshRGB555\r
1492 tst r7, r7\r
1493 beq .fl_noshRGB555\r
1494\r
1495 @ shadowed pixels:\r
1496 mov r12, #0x008e\r
1497 orr r12,r12,#0x7300\r
1498 orr r12,r12,r12,lsl #16\r
1499 add r4, r3, #0x40*2\r
1500 add r5, r3, #0xc0*2\r
1501 mov lr, #0x40/4\r
1502.fl_loopcpRGB555_sh:\r
1503 ldmia r3!, {r1,r6}\r
1504 subs lr, lr, #1\r
1505 and r1, r12, r1, lsr #1\r
1506 and r6, r12, r6, lsr #1\r
1507 stmia r4!, {r1,r6}\r
1508 stmia r5!, {r1,r6}\r
1509 bne .fl_loopcpRGB555_sh\r
1510\r
1511 @ hilighted pixels:\r
1512 sub r3, r3, #0x40*2\r
1513 mov lr, #0x40/2\r
1514.fl_loopcpRGB555_hi:\r
1515 ldr r1, [r3], #4\r
1516 TileDoShHi2Pixels555 r1\r
1517 str r1, [r4], #4\r
1518 subs lr, lr, #1\r
1519 bne .fl_loopcpRGB555_hi\r
1520\r
1521 sub r3, r3, #0x40*2\r
1522\r
cc68a136 1523.fl_noshRGB555:\r
3d48f143 1524 ldr r0, =DrawLineDest\r
cc68a136 1525 ldr r1, =(HighCol+8)\r
3d48f143 1526 ldr r0, [r0]\r
1527\r
1528 ldrb r12, [r8, #12]\r
cc68a136 1529 mov lr, #0xff\r
1530 mov lr, lr, lsl #1\r
1531\r
3d48f143 1532 tst r12, #1\r
1533 movne r2, #320/8 @ len\r
1534 bne .fl_no32colRGB555\r
1535 ldr r4, =PicoOpt\r
1536 mov r2, #256/8\r
1537 ldr r4, [r4]\r
1538 tst r4, #0x4000\r
1539 bne .fl_32scale_RGB555\r
1540 tst r4, #0x0100\r
1541 addeq r0, r0, #32*2\r
1542\r
1543.fl_no32colRGB555:\r
cc68a136 1544.fl_loopRGB555:\r
1545\r
1546 ldr r12, [r1], #4\r
1547 ldr r7, [r1], #4\r
1548\r
1549 and r4, lr, r12, lsl #1\r
1550 ldrh r4, [r3, r4]\r
1551 and r5, lr, r12, lsr #7\r
1552 ldrh r5, [r3, r5]\r
1553 and r6, lr, r12, lsr #15\r
1554 ldrh r6, [r3, r6]\r
1555 orr r4, r4, r5, lsl #16\r
1556\r
1557 and r5, lr, r12, lsr #23\r
1558 ldrh r5, [r3, r5]\r
1559 and r8, lr, r7, lsl #1\r
1560 ldrh r8, [r3, r8]\r
1561 orr r5, r6, r5, lsl #16\r
1562\r
1563 and r6, lr, r7, lsr #7\r
1564 ldrh r6, [r3, r6]\r
1565 and r12,lr, r7, lsr #15\r
1566 ldrh r12,[r3, r12]\r
1567 orr r8, r8, r6, lsl #16\r
1568\r
1569 and r6, lr, r7, lsr #23\r
1570 ldrh r6, [r3, r6] @ 1 cycle interlock here (r6)\r
1571 subs r2, r2, #1\r
1572 orr r12,r12, r6, lsl #16\r
1573\r
1574 stmia r0!, {r4,r5,r8,r12}\r
1575 bne .fl_loopRGB555\r
1576\r
3d48f143 1577 ldmfd sp!, {r4-r8,lr}\r
1578 bx lr\r
1579\r
1580\r
1581.fl_32scale_RGB555:\r
1582 stmfd sp!, {r9,r10}\r
1583 mov r9, #0x3900 @ f800 07e0 001f | e000 0780 001c | 3800 01e0 0007\r
1584 orr r9, r9, #0x00e7\r
1585\r
1586.fl_loop32scale_RGB555:\r
1587 ldr r12, [r1], #4\r
1588 ldr r7, [r1], #4\r
cc68a136 1589\r
3d48f143 1590 and r4, lr, r12,lsl #1\r
1591 ldrh r4, [r3, r4]\r
1592 and r5, lr, r12,lsr #7\r
1593 ldrh r5, [r3, r5]\r
1594 and r4, r4, r9, lsl #2\r
1595 orr r4, r4, r4, lsl #14 @ r4[31:16] = 1/4 pix_s 0\r
1596 and r5, r5, r9, lsl #2\r
1597 sub r6, r5, r5, lsr #2 @ r6 = 3/4 pix_s 1\r
1598 add r4, r4, r6, lsl #16 @ pix_d 0, 1\r
1599 and r6, lr, r12,lsr #15\r
1600 ldrh r6, [r3, r6]\r
1601 and r12,lr, r12,lsr #23\r
1602 ldrh r12,[r3, r12]\r
1603 and r6, r6, r9, lsl #2\r
1604 add r5, r5, r6\r
1605 mov r5, r5, lsr #1\r
1606 sub r6, r6, r6, lsr #2 @ r6 = 3/4 pix_s 2\r
1607 orr r5, r5, r6, lsl #16\r
1608\r
1609 and r6, lr, r7, lsl #1\r
1610 ldrh r6, [r3, r6]\r
1611 and r12,r12,r9, lsl #2\r
1612 add r5, r5, r12,lsl #14 @ pix_d 2, 3\r
1613 and r6, r6, r9, lsl #2\r
1614 orr r6, r12,r6, lsl #16 @ pix_d 4, 5\r
1615\r
1616 and r12,lr, r7, lsr #7\r
1617 ldrh r12,[r3, r12]\r
1618 and r10,lr, r7, lsr #15\r
1619 ldrh r10,[r3, r10]\r
1620 and r12,r12,r9, lsl #2\r
1621 sub r8, r12,r12,lsr #2 @ r8 = 3/4 pix_s 1\r
1622 add r8, r8, r6, lsr #18\r
1623 and r7, lr, r7, lsr #23\r
1624 ldrh r7, [r3, r7]\r
1625 and r10,r10,r9, lsl #2\r
1626 orr r8, r8, r10,lsl #15\r
1627 add r8, r8, r12,lsl #15 @ pix_d 6, 7\r
1628 sub r10,r10,r10,lsr #2 @ r10= 3/4 pix_s 2\r
1629 and r7, r7, r9, lsl #2\r
1630 add r10,r10,r7, lsr #2 @ += 1/4 pix_s 3\r
1631 orr r10,r10,r7, lsl #16 @ pix_d 8, 9\r
1632\r
1633 subs r2, r2, #1\r
1634\r
1635 stmia r0!, {r4,r5,r6,r8,r10}\r
1636 bne .fl_loop32scale_RGB555\r
1637\r
1638 ldmfd sp!, {r9,r10}\r
cc68a136 1639 ldmfd sp!, {r4-r8,lr}\r
3d48f143 1640 bx lr\r
1641\r
cc68a136 1642\r
1643@ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r
1644\r
1645@ utility\r
1646.global blockcpy @ void *dst, void *src, size_t n\r
1647\r
1648blockcpy:\r
1649 stmfd sp!, {r4,r5}\r
1650 mov r2, r2, lsr #4\r
1651blockcpy_loop:\r
1652 ldmia r1!, {r3-r5,r12}\r
1653 subs r2, r2, #1\r
1654 stmia r0!, {r3-r5,r12}\r
1655 bne blockcpy_loop\r
1656 ldmfd sp!, {r4,r5}\r
1657 bx lr\r
1658\r
1659\r
1660.global blockcpy_or @ void *dst, void *src, size_t n, int pat\r
1661\r
1662blockcpy_or:\r
1663 stmfd sp!, {r4-r6}\r
1664 orr r3, r3, r3, lsl #8\r
1665 orr r3, r3, r3, lsl #16\r
1666 mov r2, r2, lsr #4\r
1667blockcpy_loop_or:\r
1668 ldmia r1!, {r4-r6,r12}\r
1669 subs r2, r2, #1\r
1670 orr r4, r4, r3\r
1671 orr r5, r5, r3\r
1672 orr r6, r6, r3\r
1673 orr r12,r12,r3\r
1674 stmia r0!, {r4-r6,r12}\r
1675 bne blockcpy_loop_or\r
1676 ldmfd sp!, {r4-r6}\r
1677 bx lr\r
1678\r
fast MegaDrive/MegaCD/32X emulator