notaz.gp2x.de / picodrive.git / blame
? search:
summary | shortlog | log | commit | commitdiff | tree
blob | blame (incremental) | history | HEAD
trying to add movie support
[picodrive.git] / Pico / Draw_sh.s
CommitLineData
cc68a136 1@ assembly "optimized" version of some funtions from draw.c\r
2@ this is highly specialized, be careful if changing related C code!\r
3\r
4@ (c) Copyright 2006, notaz\r
5@ All Rights Reserved\r
6\r
7\r
8.extern Pico\r
9.extern PicoOpt\r
10.extern HighCol\r
11.extern Scanline\r
12.extern HighSprZ\r
13.extern rendstatus\r
14.extern DrawLineInt\r
15.extern DrawLineDest\r
16.extern DrawStripVSRam\r
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 pat lsrr offs\r
59 TilePixel \pat, \lsrr, \offs\r
60 ldreqb r4, [r1,#\offs]\r
61 tsteq r4, #0x80\r
62 andeq r4, r4, #0x3f\r
63 streqb r4, [r1,#\offs]\r
64.endm\r
65\r
66@ TileNorm (r1=pdest, r2=pixels8, r3=pal) r4: scratch, pat: register with helper pattern 0xf\r
67.macro TileNormShHP pat\r
68 TilePixelShHP \pat, 12, 0 @ #0x0000f000\r
69 TilePixelShHP \pat, 8, 1 @ #0x00000f00\r
70 TilePixelShHP \pat, 4, 2 @ #0x000000f0\r
71 TilePixelShHP \pat, 0, 3 @ #0x0000000f\r
72 TilePixelShHP \pat, 28, 4 @ #0xf0000000\r
73 TilePixelShHP \pat, 24, 5 @ #0x0f000000\r
74 TilePixelShHP \pat, 20, 6 @ #0x00f00000\r
75 TilePixelShHP \pat, 16, 7 @ #0x000f0000\r
76.endm\r
77\r
78@ TileFlip (r1=pdest, r2=pixels8, r3=pal) r4: scratch, pat: register with helper pattern 0xf\r
79.macro TileFlipShHP pat\r
80 TilePixelShHP \pat, 16, 0 @ #0x000f0000\r
81 TilePixelShHP \pat, 20, 1 @ #0x00f00000\r
82 TilePixelShHP \pat, 24, 2 @ #0x0f000000\r
83 TilePixelShHP \pat, 28, 3 @ #0xf0000000\r
84 TilePixelShHP \pat, 0, 4 @ #0x0000000f\r
85 TilePixelShHP \pat, 4, 5 @ #0x000000f0\r
86 TilePixelShHP \pat, 8, 6 @ #0x00000f00\r
87 TilePixelShHP \pat, 12, 7 @ #0x0000f000\r
88.endm\r
89\r
90\r
91@ TileSingleSh (r1=pdest, r2=pixels8, r3=pal) r4,r7: scratch, r0=sx; r12: helper pattern 0xf\r
92.macro TileSingleSh\r
93 tst r0, #1 @ not aligned?\r
94 mov r7, #0x00c000\r
95 orr r7, r7, #0xc0\r
96 ldrneb r4, [r1]\r
97 ldreqh r4, [r1]\r
98 orr r4, r4, r7\r
99 strneb r4, [r1], #1\r
100 streqh r4, [r1], #2\r
101 ldrh r4, [r1]\r
102 orr r4, r4, r7\r
103 strh r4, [r1], #2\r
104 ldrh r4, [r1]\r
105 orr r4, r4, r7\r
106 strh r4, [r1], #2\r
107 ldrh r4, [r1]\r
108 orr r4, r4, r7\r
109 strh r4, [r1], #2\r
110 ldrneb r4, [r1]\r
111 orr r4, r4, r7\r
112 strneb r4, [r1], #1\r
113.endm\r
114\r
115@ TileSingleHi (r1=pdest, r2=pixels8, r3=pal) r4,r7: scratch, r0=sx, r12: register with helper pattern 0xf\r
116.macro TileSingleHi\r
117 tst r1, #1 @ not aligned?\r
118 mov r7, #0x008000\r
119 orr r7, r7, #0x80\r
120 ldrneb r4, [r1]\r
121 ldreqh r4, [r1]\r
122 bic r4, r4, r7, lsr #1\r
123 orr r4, r4, r7\r
124 strneb r4, [r1], #1\r
125 streqh r4, [r1], #2\r
126 ldrh r4, [r1]\r
127 bic r4, r4, r7, lsr #1\r
128 orr r4, r4, r7\r
129 strh r4, [r1], #2\r
130 ldrh r4, [r1]\r
131 bic r4, r4, r7, lsr #1\r
132 orr r4, r4, r7\r
133 strh r4, [r1], #2\r
134 ldrh r4, [r1]\r
135 bic r4, r4, r7, lsr #1\r
136 orr r4, r4, r7\r
137 strh r4, [r1], #2\r
138 ldrneb r4, [r1]\r
139 bic r4, r4, r7, lsr #1\r
140 orr r4, r4, r7\r
141 strneb r4, [r1], #1\r
142.endm\r
143\r
144.macro TileDoShGenPixel shift ofs\r
145.if \shift\r
146 ands r4, r12, r2, lsr #\shift\r
147.else\r
148 ands r4, r12, r2\r
149.endif\r
150 beq 3f\r
151 cmp r4, #0xe\r
152 beq 2f\r
153 bgt 1f\r
154 orr r4, r3, r4\r
155 strb r4, [r1,#\ofs]\r
156 b 3f\r
1571:\r
158 ldrb r4, [r1,#\ofs]\r
159 orr r4, r4, #0xc0\r
160 strb r4, [r1,#\ofs]\r
161 b 3f\r
1622:\r
163 ldrb r4, [r1,#\ofs]\r
164 bic r4, r4, #0xc0\r
165 orr r4, r4, #0x80\r
166 strb r4, [r1,#\ofs]\r
1673:\r
168.endm\r
169\r
170@ TileFlipSh (r1=pdest, r2=pixels8, r3=pal) r4,r7: scratch, r0=sx, r12: register with helper pattern 0xf\r
171.macro TileFlipSh\r
172 TileDoShGenPixel 16, 0 @ #0x000f0000\r
173 TileDoShGenPixel 20, 1 @ #0x00f00000\r
174 TileDoShGenPixel 24, 2 @ #0x0f000000\r
175 TileDoShGenPixel 28, 3 @ #0xf0000000\r
176 TileDoShGenPixel 0, 4 @ #0x0000000f\r
177 TileDoShGenPixel 4, 5 @ #0x000000f0\r
178 TileDoShGenPixel 8, 6 @ #0x00000f00\r
179 TileDoShGenPixel 12, 7 @ #0x0000f000\r
180.endm\r
181\r
182@ TileNormSh (r1=pdest, r2=pixels8, r3=pal) r4,r7: scratch, r0=sx, r12: register with helper pattern 0xf\r
183.macro TileNormSh\r
184 TileDoShGenPixel 12, 0 @ #0x0000f000\r
185 TileDoShGenPixel 8, 1 @ #0x00000f00\r
186 TileDoShGenPixel 4, 2 @ #0x000000f0\r
187 TileDoShGenPixel 0, 3 @ #0x0000000f\r
188 TileDoShGenPixel 28, 4 @ #0xf0000000\r
189 TileDoShGenPixel 24, 5 @ #0x0f000000\r
190 TileDoShGenPixel 20, 6 @ #0x00f00000\r
191 TileDoShGenPixel 16, 7 @ #0x000f0000\r
192.endm\r
193\r
194\r
195@ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r
196\r
197@ struct TileStrip\r
198@ {\r
199@ int nametab; // 0x00\r
200@ int line; // 0x04\r
201@ int hscroll; // 0x08\r
202@ int xmask; // 0x0C\r
203@ int *hc; // 0x10 (pointer to cache buffer)\r
204@ int cells; // 0x14\r
205@ };\r
206\r
207@ int DrawLayer(int plane, int *hcache, int maxcells, int sh)\r
208\r
209.global DrawLayer @ int plane, int *hcache, int maxcells, int sh\r
210\r
211DrawLayer:\r
212 stmfd sp!, {r4-r11,lr}\r
213\r
214 ldr r11, =(Pico+0x22228) @ Pico.video\r
215\r
216 mov r6, r1 @ hcache\r
217 orr r9, r2, r3, lsl #31 @ r9=maxcells|(sh<<31)\r
218\r
219 ldrb r7, [r11, #16] @ ??hh??ww\r
220\r
221 mov r1, r7, lsl #4\r
222 orr r1, r1, #0x00ff\r
223\r
224 and r10, r7, #3\r
225 cmp r10, #1\r
226 biclt r1, r1, #0xfc00\r
227 biceq r1, r1, #0xfe00\r
228 bicgt r1, r1, #0xff00 @ r1=ymask=(height<<8)|0xff; ...; // Y Mask in pixels\r
229\r
230 add r10, r10, #5\r
231 cmp r10, #7\r
232 subge r10, r10, #1 @ r10=shift[width] (5,6,6,7)\r
233\r
234 @ calculate xmask:\r
235 mov r8, #1\r
236 mov r5, r8, lsl r10\r
237 sub r5, r5, #1 @ r5=xmask\r
238\r
239 @ Find name table:\r
240 tst r0, r0\r
241 ldreqb r12, [r11, #2]\r
242 moveq r12, r12, lsl #10\r
243 ldrneb r12, [r11, #4]\r
244 movne r12, r12, lsl #13\r
245 and r12, r12, #(7<<13) @ r12=(ts->nametab<<1) (halfword compliant)\r
246\r
247 ldr r2, =Scanline\r
248 ldr r2, [r2]\r
249 ldr lr, =(Pico+0x10000) @ lr=Pico.vram\r
250\r
251 ldrh r8, [r11, #12]\r
252 mov r4, r8, lsr #8 @ pvid->reg[13]\r
253 mov r4, r4, lsl #10 @ htab=pvid->reg[13]<<9; (halfwords)\r
254 ldrb r7, [r11, #11]\r
255 tst r7, #2\r
256 addne r4, r4, r2, lsl #2 @ htab+=Scanline<<1; // Offset by line\r
257 tst r7, #1\r
258 biceq r4, r4, #0x1f @ htab&=~0xf; // Offset by tile\r
259 add r4, r4, r0, lsl #1 @ htab+=plane\r
260 bic r4, r4, #0x00ff0000 @ just in case\r
261 ldrh r3, [lr, r4] @ r3=hscroll\r
262\r
263 tst r7, #4\r
264 bne .DrawStrip_vsscroll\r
265\r
266 @ Get vertical scroll value:\r
267 add r7, lr, #0x012000\r
268 add r7, r7, #0x000180 @ r7=Pico.vsram (Pico+0x22180)\r
269 ldr r7, [r7]\r
270\r
271 tst r8, #2\r
272 tstne r8, #4\r
273 bne .DrawStrip_interlace\r
274\r
275 tst r0, r0\r
276 movne r7, r7, lsr #16\r
277\r
278 @ Find the line in the name table\r
279 add r2, r2, r7\r
280 and r2, r2, r1\r
281 mov r4, r2, lsr #3\r
282 add r10, r10, #1 @ shift[width]++\r
283 add r12, r12, r4, lsl r10 @ nametab+=(ts.line>>3)<<shift[width];\r
284\r
285 @ ldmia r0, {r1,r2,r3,r5,r6,r9} @ r2=line, r3=ts->hscroll, r5=ts->xmask, r6=ts->hc, r9=ts->cells\r
286@ mov r12,r1, lsl #1 @ r12=(ts->nametab<<1) (halfword compliant)\r
287\r
288 and r10,r2, #7\r
289 mov r10,r10, lsl #1 @ r10=ty=(ts->line&7)<<1;\r
290 orr r10,r10, r9, lsl #24\r
291\r
292 rsb r8, r3, #0\r
293 mov r8, r8, lsr #3 @ r8=tilex=(-ts->hscroll)>>3\r
294\r
295 sub r1, r3, #1\r
296 and r1, r1, #7\r
297 add r4, r1, #1 @ r4=dx=((ts->hscroll-1)&7)+1\r
298\r
299 tst r9, #1<<31\r
300 mov r3, #0\r
301 orrne r10,r10, #1<<23 @ r10=(cells<<24|sh<<23|hi_not_empty<<22|ty)\r
302 movne r3, #0x40 @ default to shadowed pal on sh mode\r
303\r
304 mvn r9, #0 @ r9=prevcode=-1\r
305\r
306 @ cache some stuff to avoid mem access\r
307@ ldr r11,=HighCol\r
308 ldr r11,=DrawLineInt\r
309 ldr r11,[r11]\r
310 add r1, r11, r4 @ r1=pdest\r
311 mov r0, #0xf\r
312\r
313 cmp r4, #8\r
314 subeq r10,r10, #0x01000000 @ we will loop cells+1 times, so loop less when there is no scroll\r
315 beq .dsloop_enter\r
316\r
317 @ do first iteration with clipping\r
318 and r7, r5, r8\r
319 add r7, lr, r7, lsl #1 @ Pico.vram+((tilex&ts->xmask) as halfwords)\r
320 ldrh r7, [r7, r12] @ r7=code (int, but from unsigned, no sign extend)\r
321\r
322 tst r7, #0x8000\r
323 bne .DrawStrip_hiprio\r
324\r
325 mov r9, r7 @ remember code\r
326\r
327 movs r2, r9, lsl #20 @ if (code&0x1000)\r
328 mov r2, r2, lsl #1\r
329 add r2, r2, r10, lsl #17\r
330 mov r2, r2, lsr #17\r
331 eorcs r2, r2, #0x0e @ if (code&0x1000) addr^=0xe;\r
332\r
333 ldr r2, [lr, r2, lsl #1] @ pack=*(unsigned int *)(Pico.vram+addr); // Get 8 pixels\r
334\r
335 bic r7, r3, #0x3f\r
336 and r3, r9, #0x6000\r
337 add r3, r7, r3, lsr #9 @ r3=pal=((code&0x6000)>>9);\r
338\r
339 tst r2, r2\r
340 beq .dsloop @ tileline blank\r
341\r
342 tst r9, #0x0800\r
343 addne r4, r4, #8\r
344\r
345 ldr pc, [pc, r4, lsl #2]\r
346 nop\r
347 .word .ds_tn1_px1 @ should not happen\r
348 .word .ds_tn1_px1\r
349 .word .ds_tn1_px2\r
350 .word .ds_tn1_px3\r
351 .word .ds_tn1_px4\r
352 .word .ds_tn1_px5\r
353 .word .ds_tn1_px6\r
354 .word .ds_tn1_px7\r
355 .word .dsloop @ should not happen\r
356\r
357 .word .ds_tf1_px1 @ ...\r
358 .word .ds_tf1_px1\r
359 .word .ds_tf1_px2\r
360 .word .ds_tf1_px3\r
361 .word .ds_tf1_px4\r
362 .word .ds_tf1_px5\r
363 .word .ds_tf1_px6\r
364 .word .ds_tf1_px7\r
365 .word .dsloop @ ...\r
366\r
367 @ (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r0: helper pattern\r
368.ds_tn1_px1:\r
369 TilePixel r0, 8, 1 @ #0x00000f00\r
370.ds_tn1_px2:\r
371 TilePixel r0, 4, 2 @ #0x000000f0\r
372.ds_tn1_px3:\r
373 TilePixel r0, 0, 3 @ #0x0000000f\r
374.ds_tn1_px4:\r
375 TilePixel r0, 28, 4 @ #0xf0000000\r
376.ds_tn1_px5:\r
377 TilePixel r0, 24, 5 @ #0x0f000000\r
378.ds_tn1_px6:\r
379 TilePixel r0, 20, 6 @ #0x00f00000\r
380.ds_tn1_px7:\r
381 TilePixel r0, 16, 7 @ #0x000f0000\r
382 b .dsloop\r
383\r
384.ds_tf1_px1:\r
385 TilePixel r0, 20, 1 @ #0x00f00000\r
386.ds_tf1_px2:\r
387 TilePixel r0, 24, 2 @ #0x0f000000\r
388.ds_tf1_px3:\r
389 TilePixel r0, 28, 3 @ #0xf0000000\r
390.ds_tf1_px4:\r
391 TilePixel r0, 0, 4 @ #0x0000000f\r
392.ds_tf1_px5:\r
393 TilePixel r0, 4, 5 @ #0x000000f0\r
394.ds_tf1_px6:\r
395 TilePixel r0, 8, 6 @ #0x00000f00\r
396.ds_tf1_px7:\r
397 TilePixel r0, 12, 7 @ #0x0000f000\r
398\r
399\r
400 @ r4 & r7 are scratch in this loop\r
401.dsloop: @ 40-41 times\r
402 add r1, r1, #8\r
403.dsloop_nor1:\r
404 subs r10,r10, #0x01000000\r
405 add r8, r8, #1\r
406 bmi .dsloop_exit\r
407\r
408.dsloop_enter:\r
409 and r7, r5, r8\r
410 add r7, lr, r7, lsl #1 @ Pico.vram+((tilex&ts->xmask) as halfwords)\r
411 ldrh r7, [r7, r12] @ r7=code (int, but from unsigned, no sign extend)\r
412\r
413 tst r7, #0x8000\r
414 bne .DrawStrip_hiprio\r
415\r
416 cmp r7, r9\r
417 beq .DrawStrip_samecode @ we know stuff about this tile already\r
418\r
419 mov r9, r7 @ remember code\r
420\r
421 movs r2, r9, lsl #20 @ if (code&0x1000)\r
422 mov r2, r2, lsl #1\r
423@ bic r7, r10,#0xff000000\r
424@ add r2, r7, r2, lsr #17 @ r2=addr=(code&0x7ff)<<4; addr+=ty\r
425 add r2, r2, r10, lsl #17\r
426 mov r2, r2, lsr #17\r
427 eorcs r2, r2, #0x0e @ if (code&0x1000) addr^=0xe;\r
428\r
429 bic r7, r3, #0x3f\r
430 and r3, r9, #0x6000\r
431 add r3, r7, r3, lsr #9 @ r3=pal=((code&0x6000)>>9);\r
432\r
433 ldr r2, [lr, r2, lsl #1] @ pack=*(unsigned int *)(Pico.vram+addr); // Get 8 pixels\r
434\r
435.DrawStrip_samecode:\r
436 tst r2, r2\r
437 beq .dsloop @ tileline blank\r
438\r
439 cmp r2, r2, ror #4\r
440 beq .DrawStrip_SingleColor @ tileline singlecolor \r
441\r
442 tst r9, #0x0800\r
443 beq .DrawStrip_TileNorm\r
444\r
445 @ (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r0: helper pattern\r
446 TileFlip r0\r
447 b .dsloop\r
448\r
449.DrawStrip_TileNorm:\r
450 TileNorm r0\r
451 b .dsloop\r
452\r
453.DrawStrip_SingleColor:\r
454 and r4, r2, #0xf\r
455 orr r4, r3, r4\r
456 orr r4, r4, r4, lsl #8\r
457 tst r1, #1 @ not aligned?\r
458 strneb r4, [r1], #1\r
459 streqh r4, [r1], #2\r
460 strh r4, [r1], #2\r
461 strh r4, [r1], #2\r
462 strh r4, [r1], #2\r
463 strneb r4, [r1], #1 @ have a remaining unaligned pixel?\r
464 b .dsloop_nor1 @ we incremeted r1 ourselves\r
465\r
466.DrawStrip_hiprio:\r
467 tst r10, #0x00c00000\r
468 beq .DrawStrip_hiprio_maybempt\r
469 sub r0, r1, r11\r
470 orr r7, r7, r0, lsl #16\r
471 orr r7, r7, r10, lsl #25 @ (ty<<25)\r
472 tst r7, #0x1000\r
473 eorne r7, r7, #7<<26 @ if(code&0x1000) cval^=7<<26;\r
474 str r7, [r6], #4 @ cache hi priority tile\r
475 mov r0, #0xf\r
476 b .dsloop\r
477\r
478.DrawStrip_hiprio_maybempt:\r
479 cmp r7, r9\r
480 beq .dsloop @ must've been empty, otherwise we wouldn't get here\r
481 mov r9, r7 @ remember code\r
482 movs r2, r9, 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 tst r2, r2\r
489 orrne r10, r10, #1<<22\r
490 bne .DrawStrip_hiprio\r
491 b .dsloop\r
492\r
493.dsloop_exit:\r
494 mov r0, #0\r
495 str r0, [r6] @ terminate the cache list\r
496\r
497 ldmfd sp!, {r4-r11,lr}\r
498 bx lr\r
499\r
500\r
501.DrawStrip_vsscroll:\r
502 @ shit, we have 2-cell column based vscroll\r
503 @ let the c code handle this (for now)\r
504\r
505 @ int nametab; // 0x00\r
506 @ int line; // 0x04\r
507 @ int hscroll; // 0x08\r
508 @ int xmask; // 0x0C\r
509 @ int *hc; // 0x10 (pointer to cache buffer)\r
510 @ int cells; // 0x14\r
511\r
512 sub sp, sp, #6*4\r
513 orr r2, r1, r10, lsl #24 @ ts.line=ymask|(shift[width]<<24); // save some stuff instead of line\r
514 mov r1, r0 @ plane\r
515 mov r0, r12, lsr #1 @ halfwords\r
516 and r9, r9, #0xff\r
517 stmia sp, {r0,r2,r3,r5,r6,r9}\r
518\r
519 mov r0, sp\r
520 bl DrawStripVSRam @ struct TileStrip *ts, int plane\r
521\r
522 add sp, sp, #6*4\r
523 ldmfd sp!, {r4-r11,lr}\r
524 bx lr\r
525\r
526@ interlace mode 2? Sonic 2?\r
527.DrawStrip_interlace:\r
528 tst r0, r0\r
529 moveq r7, r7, lsl #21\r
530 movne r7, r7, lsl #5\r
531\r
532 @ Find the line in the name table\r
533 add r2, r7, r2, lsl #22 @ r2=(vscroll+(Scanline<<1))<<21 (11 bits);\r
534 orr r1, r1, #0x80000000\r
535 and r2, r2, r1, ror #10 @ &((ymask<<1)|1)<<21;\r
536 mov r2, r2, lsr #21\r
537 mov r4, r2, lsr #4\r
538 mov r12, r12, lsr #1 @ halfwords\r
539 add r0, r12, r4, lsl r10 @ nametab+=(ts.line>>4)<<shift[width];\r
540 and r9, r9, #0xff\r
541\r
542 sub sp, sp, #6*4\r
543 stmia sp, {r0,r2,r3,r5,r6,r9}\r
544\r
545 mov r0, sp\r
546 bl DrawStripInterlace @ struct TileStrip *ts\r
547\r
548 add sp, sp, #6*4\r
549 ldmfd sp!, {r4-r11,lr}\r
550 bx lr\r
551\r
552.pool\r
553\r
554@ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r
555\r
556\r
557.global BackFill @ int reg7, int sh\r
558\r
559BackFill:\r
560 stmfd sp!, {r4-r9,lr}\r
561\r
562@ ldr lr, =(HighCol+8)\r
563 ldr lr,=DrawLineInt\r
564 ldr lr,[lr]\r
565 add lr, lr, #8\r
566\r
567 mov r0, r0, lsl #26\r
568 mov r0, r0, lsr #26\r
569 orr r0, r0, r1, lsl #6\r
570 orr r0, r0, r0, lsl #8\r
571 orr r0, r0, r0, lsl #16\r
572\r
573 mov r1, r0\r
574 mov r2, r0\r
575 mov r3, r0\r
576 mov r4, r0\r
577 mov r5, r0\r
578 mov r6, r0\r
579 mov r7, r0\r
580\r
581 @ go go go!\r
582 stmia lr!, {r0-r7} @ 10*8*4\r
583 stmia lr!, {r0-r7}\r
584 stmia lr!, {r0-r7}\r
585 stmia lr!, {r0-r7}\r
586 stmia lr!, {r0-r7}\r
587 stmia lr!, {r0-r7}\r
588 stmia lr!, {r0-r7}\r
589 stmia lr!, {r0-r7}\r
590 stmia lr!, {r0-r7}\r
591 stmia lr!, {r0-r7}\r
592\r
593 ldmfd sp!, {r4-r9,r12}\r
594 bx r12\r
595\r
596\r
597@ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r
598\r
599\r
600.global DrawTilesFromCache @ int *hc, int sh\r
601\r
602DrawTilesFromCache:\r
603 stmfd sp!, {r4-r8,r11,lr}\r
604\r
605 mvn r5, #0 @ r5=prevcode=-1\r
606 mov r8, r1\r
607\r
608 @ cache some stuff to avoid mem access\r
609@ ldr r11,=HighCol\r
610 ldr r11,=DrawLineInt\r
611 ldr r11,[r11]\r
612 ldr lr, =(Pico+0x10000) @ lr=Pico.vram\r
613 mov r12,#0xf\r
614\r
615 @ scratch: r4, r7\r
616.dtfc_loop:\r
617 ldr r6, [r0], #4 @ read code\r
618 movs r1, r6, lsr #16 @ r1=dx;\r
619 ldmeqfd sp!, {r4-r8,r11,pc} @ dx is never zero, this must be a terminator, return\r
620 bic r1, r1, #0xfe00\r
621 add r1, r11, r1 @ r1=pdest\r
622\r
623@ tst r8, r8\r
624@ bne .dtfc_shadow @ this is a rare case, so we jump when it happens, not when it doesn't\r
625@.dtfc_shadow_done:\r
626\r
627 mov r7, r6, lsl #16\r
628 cmp r5, r7, lsr #16\r
629 beq .dtfc_samecode @ if (code==prevcode)\r
630\r
631 mov r5, r7, lsr #16\r
632\r
633 mov r2, r5, lsl #21\r
634 mov r2, r2, lsr #17 @ r2=addr=(code&0x7ff)<<4;\r
635 add r2, r2, r6, lsr #25 @ addr+=ty\r
636\r
637 and r3, r5, #0x6000\r
638 mov r3, r3, lsr #9 @ r3=pal=((code&0x6000)>>9);\r
639\r
640 ldr r2, [lr, r2, lsl #1] @ pack=*(unsigned int *)(Pico.vram+addr); // Get 8 pixels\r
641\r
642.dtfc_samecode:\r
643 tst r8, r8\r
644 bne .dtfc_shadow\r
645\r
646 tst r2, r2\r
647 beq .dtfc_loop\r
648\r
649 cmp r2, r2, ror #4\r
650 beq .dtfc_SingleColor @ tileline singlecolor \r
651\r
652 tst r5, #0x0800\r
653 beq .dtfc_TileNorm\r
654\r
655 @ (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r12: helper pattern\r
656 TileFlip r12\r
657 b .dtfc_loop\r
658\r
659.dtfc_TileNorm:\r
660 TileNorm r12\r
661 b .dtfc_loop\r
662\r
663.dtfc_SingleColor:\r
664 and r4, r2, #0xf\r
665 orr r4, r3, r4\r
666 orr r4, r4, r4, lsl #8\r
667 tst r1, #1 @ not aligned?\r
668 strneb r4, [r1], #1\r
669 streqh r4, [r1], #2\r
670 strh r4, [r1], #2\r
671 strh r4, [r1], #2\r
672 strh r4, [r1], #2\r
673 strneb r4, [r1], #1 @ have a remaining unaligned pixel?\r
674 b .dtfc_loop\r
675\r
676.dtfc_shadow:\r
677 tst r2, r2\r
678 beq .dtfc_shadow_blank\r
679\r
680 cmp r2, r2, ror #4\r
681 beq .dtfc_SingleColor @ tileline singlecolor \r
682\r
683 tst r5, #0x0800\r
684 beq .dtfc_TileNormShHP\r
685\r
686 @ (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r12: helper pattern\r
687 TileFlipShHP r12\r
688 b .dtfc_loop\r
689\r
690.dtfc_TileNormShHP:\r
691 TileNormShHP r12\r
692 b .dtfc_loop\r
693\r
694.dtfc_shadow_blank:\r
695 ldrb r4, [r1]\r
696 tst r4, #0x80\r
697 andeq r4, r4,#0x3f\r
698 streqb r4, [r1]\r
699 ldrb r4, [r1,#1]\r
700 tst r4, #0x80\r
701 andeq r4, r4,#0x3f\r
702 streqb r4, [r1,#1]\r
703 ldrb r4, [r1,#2]\r
704 tst r4, #0x80\r
705 andeq r4, r4,#0x3f\r
706 streqb r4, [r1,#2]\r
707 ldrb r4, [r1,#3]\r
708 tst r4, #0x80\r
709 andeq r4, r4,#0x3f\r
710 streqb r4, [r1,#3]\r
711 ldrb r4, [r1,#4]\r
712 tst r4, #0x80\r
713 andeq r4, r4,#0x3f\r
714 streqb r4, [r1,#4]\r
715 ldrb r4, [r1,#5]\r
716 tst r4, #0x80\r
717 andeq r4, r4,#0x3f\r
718 streqb r4, [r1,#5]\r
719 ldrb r4, [r1,#6]\r
720 tst r4, #0x80\r
721 andeq r4, r4,#0x3f\r
722 streqb r4, [r1,#6]\r
723 ldrb r4, [r1,#7]\r
724 tst r4, #0x80\r
725 andeq r4, r4,#0x3f\r
726 streqb r4, [r1,#7]\r
727 b .dtfc_loop\r
728\r
729.pool\r
730\r
731@ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r
732\r
733\r
734.global DrawSpritesFromCache @ int *hc, int sh\r
735\r
736DrawSpritesFromCache:\r
737 stmfd sp!, {r4-r11,lr}\r
738\r
739 @ cache some stuff to avoid mem access\r
740@ ldr r11,=HighCol\r
741 ldr r11,=DrawLineInt\r
742 ldr r11,[r11]\r
743 ldr lr, =(Pico+0x10000) @ lr=Pico.vram\r
744 mov r6, r1, lsl #31\r
745 orr r6, r6, #1<<30\r
746 mov r12,#0xf\r
747\r
748 mov r10, r0\r
749\r
750.dsfc_loop:\r
751 ldr r9, [r10], #4 @ read code\r
752 tst r9, r9\r
753 ldmeqfd sp!, {r4-r11,pc}\r
754\r
755 mov r4, r9, lsl #28\r
756 bic r6, r6, #7\r
757 orr r6, r6, r4, lsr #30\r
758 add r6, r6, #1 @ r6=s1cc???? ... ?????www (s=shadow/hilight, cc=pal, w=width)\r
759\r
760 and r5, r9, #3\r
761 add r5, r5, #1 @ r5=delta\r
762 tst r9, #0x10000\r
763 rsbne r5, r5, #0 @ Flip X\r
764 mov r5, r5, lsl #4\r
765\r
766 mov r2, r9, lsr #17\r
767 mov r8, r2, lsl #1 @ tile=((unsigned int)code>>17)<<1;\r
768\r
769 and r3, r9, #0x30 @ r3=pal=(code&0x30);\r
770\r
771 bic r6, r6, #3<<28\r
772 orr r6, r6, r3, lsl #24\r
773\r
774 mov r0, r9, lsl #16\r
775 mov r0, r0, asr #22 @ sx=(code<<16)>>22\r
776 adds r0, r0, #0 @ set ZV\r
777 b .dsfc_inloop_enter\r
778\r
779@ scratch: r4, r7\r
780.dsfc_inloop:\r
781 sub r6, r6, #1\r
782 tst r6, #7\r
783 beq .dsfc_loop\r
784 adds r0, r0, #8\r
785 add r8, r8, r5\r
786\r
787.dsfc_inloop_enter:\r
788 ble .dsfc_inloop\r
789 cmp r0, #328\r
790 bge .dsfc_loop\r
791\r
792 mov r8, r8, lsl #17\r
793 mov r8, r8, lsr #17 @ tile&=0x7fff; // Clip tile address\r
794\r
795 ldr r2, [lr, r8, lsl #1] @ pack=*(unsigned int *)(Pico.vram+tile); // Get 8 pixels\r
796 tst r2, r2\r
797 beq .dsfc_inloop\r
798\r
799 add r1, r11, r0 @ r1=pdest\r
800\r
801 cmp r12, r6, lsr #28\r
802 beq .dsfc_shadow\r
803\r
804 cmp r2, r2, ror #4\r
805 beq .dsfc_SingleColor @ tileline singlecolor \r
806\r
807 tst r9, #0x10000\r
808 beq .dsfc_TileNorm\r
809\r
810 @ TileFlip (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r12: helper pattern\r
811 TileFlip r12\r
812 b .dsfc_inloop\r
813\r
814.dsfc_TileNorm:\r
815 TileNorm r12\r
816 b .dsfc_inloop\r
817\r
818.dsfc_SingleColor:\r
819 tst r0, #1 @ not aligned?\r
820 and r4, r2, #0xf\r
821 orr r4, r3, r4\r
822 orr r4, r4, r4, lsl #8\r
823 strneb r4, [r1], #1\r
824 streqh r4, [r1], #2\r
825 strh r4, [r1], #2\r
826 strh r4, [r1], #2\r
827 strh r4, [r1], #2\r
828 strneb r4, [r1], #1\r
829 b .dsfc_inloop\r
830\r
831.dsfc_shadow:\r
832 cmp r2, r2, ror #4\r
833 beq .dsfc_singlec_sh\r
834\r
835 tst r9, #0x10000\r
836 beq .dsfc_TileNorm_sh\r
837\r
838 @ (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r12: helper pattern\r
839 TileFlipSh\r
840 b .dsfc_inloop\r
841\r
842.dsfc_TileNorm_sh:\r
843 TileNormSh\r
844 b .dsfc_inloop\r
845\r
846.dsfc_singlec_sh:\r
847 cmp r2, #0xe0000000\r
848 bcc .dsfc_SingleColor @ normal singlecolor tileline (carry inverted in ARM)\r
849 tst r2, #0x10000000\r
850 bne .dsfc_sh_sh\r
851 TileSingleHi\r
852 b .dsfc_inloop\r
853\r
854.dsfc_sh_sh:\r
855 TileSingleSh\r
856 b .dsfc_inloop\r
857\r
858.pool\r
859\r
860@ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r
861\r
862@ + 0 : hhhhvvvv ab--hhvv yyyyyyyy yyyyyyyy // a: offscreen h, b: offs. v, h: horiz. size\r
863@ + 4 : xxxxxxxx xxxxxxxx pccvhnnn nnnnnnnn // x: x coord + 8\r
864\r
865.global DrawSprite @ unsigned int *sprite, int **hc, int sh\r
866\r
867DrawSprite:\r
868 stmfd sp!, {r4-r9,r11,lr}\r
869\r
870 ldr r3, [r0] @ sprite[0]\r
871 mov r6, r3, lsr #28\r
872 sub r6, r6, #1 @ r6=width-1 (inc later)\r
873 mov r5, r3, lsr #24\r
874 and r5, r5, #7 @ r5=height\r
875\r
876 mov r4, r3, lsl #16 @ r4=sy<<16 (tmp)\r
877\r
878 ldr r7, =Scanline\r
879 ldr r7, [r7]\r
880 sub r7, r7, r4, asr #16 @ r7=row=Scanline-sy\r
881\r
882 tst r2, r2\r
883 ldr r9, [r0, #4]\r
884 mov r2, r9, asr #16 @ r2=sx\r
885 bic r9, r9, #0xfe000000\r
886 orrne r9, r9, #1<<31 @ r9=code|(sh<<31)\r
887\r
888 tst r9, #0x1000\r
889 movne r4, r5, lsl #3\r
890 subne r4, r4, #1\r
891 subne r7, r4, r7 @ if (code&0x1000) row=(height<<3)-1-row; // Flip Y\r
892\r
893 mov r8, r9, lsl #21\r
894 mov r8, r8, lsr #21\r
895 add r8, r8, r7, lsr #3 @ tile+=row>>3; // Tile number increases going down\r
896 \r
897 tst r9, #0x0800\r
898 mlane r8, r5, r6, r8 @ if (code&0x0800) { tile+=delta*(width-1);\r
899 rsbne r5, r5, #0 @ delta=-delta; } // r5=delta now\r
900\r
901 mov r8, r8, lsl #4\r
902 and r7, r7, #7\r
903 add r8, r8, r7, lsl #1 @ tile+=(row&7)<<1; // Tile address\r
904\r
905 tst r9, #0x8000\r
906 bne .dspr_cache @ if(code&0x8000) // high priority - cache it\r
907\r
908 @ cache some stuff to avoid mem access\r
909@ ldr r11,=HighCol\r
910 ldr r11,=DrawLineInt\r
911 ldr r11,[r11]\r
912 ldr lr, =(Pico+0x10000) @ lr=Pico.vram\r
913 mov r12,#0xf\r
914\r
915 mov r5, r5, lsl #4 @ delta<<=4; // Delta of address\r
916 and r4, r9, #0x6000\r
917 orr r9, r9, r4, lsl #16\r
918 orr r9, r9, #0x10000000 @ r9=scc1 ???? ... <code> (s=shadow/hilight, cc=pal)\r
919\r
920 tst r9, #1<<31\r
921 mov r3, r4, lsr #9 @ r3=pal=((code>>9)&0x30);\r
922 orrne r3, r3, #0x40 @ shadow by default\r
923\r
924 add r6, r6, #1 @ inc now\r
925 adds r0, r2, #0 @ mov sx to r0 and set ZV flags\r
926 b .dspr_loop_enter\r
927\r
928.dspr_loop:\r
929 subs r6, r6, #1 @ width--\r
930 ldmeqfd sp!, {r4-r9,r11,pc}@ return\r
931 adds r0, r0, #8 @ sx+=8\r
932 add r8, r8, r5 @ tile+=delta\r
933\r
934.dspr_loop_enter:\r
935 ble .dspr_loop @ sx <= 0\r
936 cmp r0, #328\r
937 ldmgefd sp!, {r4-r9,r11,pc}@ return\r
938\r
939 mov r8, r8, lsl #17\r
940 mov r8, r8, lsr #17 @ tile&=0x7fff; // Clip tile address\r
941\r
942 ldr r2, [lr, r8, lsl #1] @ pack=*(unsigned int *)(Pico.vram+addr); // Get 8 pixels\r
943 tst r2, r2\r
944 beq .dspr_loop\r
945\r
946 add r1, r11, r0 @ r1=pdest\r
947\r
948 cmp r12, r9, lsr #28\r
949 beq .dspr_shadow\r
950\r
951 cmp r2, r2, ror #4\r
952 beq .dspr_SingleColor @ tileline singlecolor \r
953\r
954 tst r9, #0x0800\r
955 beq .dspr_TileNorm\r
956\r
957 @ (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r12: helper pattern\r
958 TileFlip r12\r
959 b .dspr_loop\r
960\r
961@ scratch: r4, r7\r
962.dspr_TileNorm:\r
963 TileNorm r12\r
964 b .dspr_loop\r
965\r
966.dspr_SingleColor:\r
967 and r4, r2, #0xf\r
968 orr r4, r3, r4\r
969 orr r4, r4, r4, lsl #8\r
970 tst r0, #1 @ not aligned?\r
971 strneb r4, [r1], #1\r
972 streqh r4, [r1], #2\r
973 strh r4, [r1], #2\r
974 strh r4, [r1], #2\r
975 strh r4, [r1], #2\r
976 strneb r4, [r1], #1\r
977 b .dspr_loop\r
978\r
979.dspr_shadow:\r
980 cmp r2, r2, ror #4\r
981 beq .dspr_singlec_sh\r
982\r
983 tst r9, #0x0800\r
984 beq .dspr_TileNorm_sh\r
985\r
986 @ (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r12: helper pattern\r
987 TileFlipSh\r
988 b .dspr_loop\r
989\r
990.dspr_TileNorm_sh:\r
991 TileNormSh\r
992 b .dspr_loop\r
993\r
994.dspr_singlec_sh:\r
995 cmp r2, #0xe0000000\r
996 bcc .dspr_SingleColor @ normal tileline\r
997 tst r2, #0x10000000\r
998 bne .dspr_sh_sh\r
999 TileSingleHi\r
1000 b .dspr_loop\r
1001\r
1002.dspr_sh_sh:\r
1003 TileSingleSh\r
1004 b .dspr_loop\r
1005\r
1006\r
1007.dspr_cache:\r
1008 @ *(*hc)++ = (tile<<16)|((code&0x0800)<<5)|((sx<<6)&0x0000ffc0)|((code>>9)&0x30)|((sprite[0]>>24)&0xf);\r
1009 mov r4, r8, lsl #16 @ tile\r
1010 tst r9, #0x0800\r
1011 orrne r4, r4, #0x10000 @ code&0x0800\r
1012 mov r2, r2, lsl #22\r
1013 orr r4, r4, r2, lsr #16 @ (sx<<6)&0x0000ffc0\r
1014 and r2, r9, #0x6000\r
1015 orr r4, r4, r2, lsr #9 @ (code>>9)&0x30\r
1016 mov r2, r3, lsl #12\r
1017 orr r4, r4, r2, lsr #28 @ (sprite[0]>>24)&0xf\r
1018\r
1019 ldr r2, [r1]\r
1020 str r4, [r2], #4\r
1021 str r2, [r1]\r
1022\r
1023 ldmfd sp!, {r4-r9,r11,lr}\r
1024 bx lr\r
1025\r
1026@ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r
1027\r
1028.global DrawWindow @ int tstart, int tend, int prio, int sh // int *hcache\r
1029\r
1030DrawWindow:\r
1031 stmfd sp!, {r4-r11,lr}\r
1032\r
1033 ldr r11, =(Pico+0x22228) @ Pico.video\r
1034 ldrb r12, [r11, #3] @ pvid->reg[3]\r
1035 mov r12, r12, lsl #10\r
1036\r
1037 ldr r10, =Scanline\r
1038 ldr r10, [r10]\r
1039 mov r5, r10, lsr #3\r
1040 and r10, r10, #7\r
1041 mov r10, r10, lsl #1 @ r10=ty\r
1042\r
1043 ldr r4, [r11, #12]\r
1044 tst r4, #1 @ 40 cell mode?\r
1045 andne r12, r12, #0xf000 @ 0x3c<<10\r
1046 andeq r12, r12, #0xf800\r
1047 addne r12, r12, r5, lsl #7\r
1048 addeq r12, r12, r5, lsl #6 @ nametab\r
1049 add r12, r12, r0, lsl #2 @ +starttile\r
1050\r
1051 ldr lr, =(Pico+0x10000) @ lr=Pico.vram\r
1052\r
1053 @ fetch the first code now\r
1054 ldrh r7, [lr, r12]\r
1055\r
1056 ldr r6, =rendstatus\r
1057 ldrb r6, [r6]\r
1058 ands r6, r6, #2 @ we care about bit 1 only\r
1059 orr r6, r6, r2\r
1060 bne .dw_no_sameprio\r
1061\r
1062 cmp r2, r7, lsr #15\r
1063 ldmnefd sp!, {r4-r11,pc} @ assume that whole window uses same priority\r
1064\r
1065.dw_no_sameprio:\r
1066 orr r6, r6, r3, lsl #8 @ shadow mode\r
1067\r
1068 sub r8, r1, r0\r
1069 mov r8, r8, lsl #1 @ cells\r
1070\r
1071 mvn r9, #0 @ r9=prevcode=-1\r
1072\r
1073 @ cache some stuff to avoid mem access\r
1074@ ldr r11,=(HighCol+8)\r
1075 ldr r11,=DrawLineInt\r
1076 ldr r11,[r11]\r
1077 add r11,r11, #8\r
1078 add r1, r11, r0, lsl #4 @ r1=pdest\r
1079 mov r0, #0xf\r
1080 b .dwloop_enter\r
1081\r
1082 @ r4,r5 & r7 are scratch in this loop\r
1083.dwloop:\r
1084 add r1, r1, #8\r
1085.dwloop_nor1:\r
1086 subs r8, r8, #1\r
1087 add r12, r12, #2 @ halfwords\r
1088 beq .dwloop_end @ done\r
1089\r
1090 ldrh r7, [lr, r12] @ r7=code (int, but from unsigned, no sign extend)\r
1091\r
1092 eor r5, r6, r7, lsr #15\r
1093 tst r5, #1\r
1094 orrne r6, r6, #2 @ wrong pri\r
1095 bne .dwloop\r
1096\r
1097 cmp r7, r9\r
1098 beq .dw_samecode @ we know stuff about this tile already\r
1099\r
1100.dwloop_enter:\r
1101 mov r9, r7 @ remember code\r
1102\r
1103 movs r2, r9, lsl #20 @ if (code&0x1000)\r
1104 mov r2, r2, lsl #1\r
1105 add r2, r10, r2, lsr #17 @ r2=addr=(code&0x7ff)<<4; addr+=ty\r
1106 eorcs r2, r2, #0x0e @ if (code&0x1000) addr^=0xe;\r
1107\r
1108 and r3, r9, #0x6000\r
1109 mov r3, r3, lsr #9 @ r3=pal=((code&0x6000)>>9);\r
1110\r
1111 ldr r2, [lr, r2, lsl #1] @ pack=*(unsigned int *)(Pico.vram+addr); // Get 8 pixels\r
1112\r
1113.dw_samecode:\r
1114 tst r6, #0x100\r
1115 bne .dw_shadow\r
1116.dw_shadow_done:\r
1117 tst r2, r2\r
1118 beq .dwloop @ tileline blank\r
1119\r
1120 cmp r2, r2, ror #4\r
1121 beq .dw_SingleColor @ tileline singlecolor \r
1122\r
1123 tst r9, #0x0800\r
1124 beq .dw_TileNorm\r
1125\r
1126 @ (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r0: helper pattern\r
1127 TileFlip r0\r
1128 b .dwloop\r
1129\r
1130.dw_TileNorm:\r
1131 TileNorm r0\r
1132 b .dwloop\r
1133\r
1134.dw_SingleColor:\r
1135 and r4, r0, r2 @ #0x0000000f\r
1136 orr r4, r3, r4\r
1137 orr r4, r4, r4, lsl #8\r
1138 orr r4, r4, r4, lsl #16\r
1139 mov r5, r4\r
1140 stmia r1!, {r4,r5}\r
1141 b .dwloop_nor1 @ we incremeted r1 ourselves\r
1142\r
1143.dw_shadow:\r
1144 tst r6, #1 @ hi pri?\r
1145 orreq r3, r3, #0x40\r
1146 beq .dw_shadow_done\r
1147 ldr r4, [r1]\r
1148 tst r4, #0x00000080\r
1149 biceq r4, r4, #0x000000c0\r
1150 tst r4, #0x00008000\r
1151 biceq r4, r4, #0x0000c000\r
1152 tst r4, #0x00800000\r
1153 biceq r4, r4, #0x00c00000\r
1154 tst r4, #0x80000000\r
1155 biceq r4, r4, #0xc0000000\r
1156 str r4, [r1]\r
1157 ldr r4, [r1,#4]\r
1158 tst r4, #0x00000080\r
1159 biceq r4, r4, #0x000000c0\r
1160 tst r4, #0x00008000\r
1161 biceq r4, r4, #0x0000c000\r
1162 tst r4, #0x00800000\r
1163 biceq r4, r4, #0x00c00000\r
1164 tst r4, #0x80000000\r
1165 biceq r4, r4, #0xc0000000\r
1166 str r4, [r1,#4]\r
1167 b .dw_shadow_done\r
1168\r
1169.dwloop_end:\r
1170 ldr r0, =rendstatus\r
1171 ldr r1, [r0]\r
1172 and r6, r6, #2\r
1173 orr r1, r1, r6\r
1174 str r1, [r0]\r
1175\r
1176 ldmfd sp!, {r4-r11,r12}\r
1177 bx r12\r
1178\r
1179\r
1180@ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r
1181\r
1182\r
1183@ hilights 2 pixels in RGB444/BGR444 format\r
1184.macro TileDoShHi2Pixels444 reg\r
1185 mov \reg, \reg, ror #12\r
1186 adds \reg, \reg, #0x40000000\r
1187 orrcs \reg, \reg, #0xf0000000\r
1188 mov \reg, \reg, ror #28\r
1189 adds \reg, \reg, #0x40000000\r
1190 orrcs \reg, \reg, #0xf0000000\r
1191 mov \reg, \reg, ror #28\r
1192 adds \reg, \reg, #0x40000000\r
1193 orrcs \reg, \reg, #0xf0000000\r
1194 mov \reg, \reg, ror #24\r
1195 adds \reg, \reg, #0x40000000\r
1196 orrcs \reg, \reg, #0xf0000000\r
1197 mov \reg, \reg, ror #28\r
1198 adds \reg, \reg, #0x40000000\r
1199 orrcs \reg, \reg, #0xf0000000\r
1200 mov \reg, \reg, ror #28\r
1201 adds \reg, \reg, #0x40000000\r
1202 orrcs \reg, \reg, #0xf0000000\r
1203 mov \reg, \reg, ror #12\r
1204.endm\r
1205\r
1206\r
1207.global FinalizeLineBGR444 @ int sh\r
1208\r
1209FinalizeLineBGR444:\r
1210 stmfd sp!, {r4-r6,lr}\r
1211 mov r6, r0\r
1212 ldr r0, =DrawLineDest\r
1213 ldr r0, [r0]\r
1214 ldr lr, =(Pico+0x22228) @ Pico.video\r
1215 sub r3, lr, #0x128 @ r3=Pico.cram\r
1216\r
1217 ldrb r12, [lr, #12]\r
1218 tst r12, #1\r
1219 movne r2, #320/4 @ len\r
1220 moveq r2, #256/4\r
1221 addeq r0, r0, #32*2\r
1222 ldreq r4, =PicoOpt\r
1223 ldreq r4, [r4]\r
1224 tsteq r4, #0x100\r
1225 addeq r0, r0, #32*2\r
1226\r
1227 tst r6, r6\r
1228 beq .fl_noshBGR444\r
1229\r
1230 ldr r4, =HighPal\r
1231\r
1232 ldrb r12, [lr, #-0x1a] @ 0x2220e ~ dirtyPal\r
1233 tst r12, r12\r
1234 moveq r3, r4\r
1235 beq .fl_noshBGR444\r
1236 mov r12, #0\r
1237 strb r12, [lr, #-0x1a]\r
1238\r
1239 mov lr, #0x40/8\r
1240 @ copy pal:\r
1241.fl_loopcpBGR444:\r
1242 subs lr, lr, #1\r
1243 ldmia r3!, {r1,r5,r6,r12}\r
1244 stmia r4!, {r1,r5,r6,r12}\r
1245 bne .fl_loopcpBGR444\r
1246\r
1247 @ shadowed pixels:\r
1248 mov r12, #0x0077\r
1249 orr r12,r12,#0x0700\r
1250 orr r12,r12,r12,lsl #16\r
1251 sub r3, r3, #0x40*2\r
1252 add r5, r4, #0x80*2\r
1253 mov lr, #0x40/4\r
1254.fl_loopcpBGR444_sh:\r
1255 subs lr, lr, #1\r
1256 ldmia r3!, {r1,r6}\r
1257 and r1, r12, r1, lsr #1\r
1258 and r6, r12, r6, lsr #1\r
1259 stmia r4!, {r1,r6}\r
1260 stmia r5!, {r1,r6}\r
1261 bne .fl_loopcpBGR444_sh\r
1262\r
1263 @ hilighted pixels:\r
1264 sub r3, r3, #0x40*2\r
1265 mov lr, #0x40/2\r
1266.fl_loopcpBGR444_hi:\r
1267 ldr r1, [r3], #4\r
1268 TileDoShHi2Pixels444 r1\r
1269 str r1, [r4], #4\r
1270 subs lr, lr, #1\r
1271 bne .fl_loopcpBGR444_hi\r
1272\r
1273 sub r3, r4, #0x40*3*2\r
1274\r
1275\r
1276.fl_noshBGR444:\r
1277@ ldr r1, =(HighCol+8)\r
1278 ldr r1, =DrawLineInt\r
1279 ldr r1, [r1]\r
1280 add r1, r1, #8\r
1281 mov lr, #0xff\r
1282 mov lr, lr, lsl #1\r
1283\r
1284.fl_loopBGR444:\r
1285 subs r2, r2, #1\r
1286\r
1287 ldr r12, [r1], #4\r
1288\r
1289 and r4, lr, r12, lsl #1\r
1290 ldrh r4, [r3, r4]\r
1291 and r5, lr, r12, lsr #7\r
1292 ldrh r5, [r3, r5]\r
1293 orr r4, r4, r5, lsl #16\r
1294\r
1295 and r5, lr, r12, lsr #15\r
1296 ldrh r5, [r3, r5]\r
1297 and r6, lr, r12, lsr #23\r
1298 ldrh r6, [r3, r6]\r
1299 orr r5, r5, r6, lsl #16\r
1300\r
1301 stmia r0!, {r4,r5}\r
1302 bne .fl_loopBGR444\r
1303\r
1304\r
1305 ldmfd sp!, {r4-r6,lr}\r
1306 bx lr\r
1307\r
1308\r
1309@ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r
1310\r
1311\r
1312@ hilights 2 pixels in RGB555/BGR555 format\r
1313.macro TileDoShHi2Pixels555 reg\r
1314 adds \reg, \reg, #0x40000000\r
1315 orrcs \reg, \reg, #0xf0000000\r
1316 mov \reg, \reg, ror #27\r
1317 adds \reg, \reg, #0x40000000\r
1318 orrcs \reg, \reg, #0xf0000000\r
1319 mov \reg, \reg, ror #26\r
1320 adds \reg, \reg, #0x40000000\r
1321 orrcs \reg, \reg, #0xf0000000\r
1322 mov \reg, \reg, ror #27\r
1323 adds \reg, \reg, #0x40000000\r
1324 orrcs \reg, \reg, #0xf0000000\r
1325 mov \reg, \reg, ror #27\r
1326 adds \reg, \reg, #0x40000000\r
1327 orrcs \reg, \reg, #0xf0000000\r
1328 mov \reg, \reg, ror #26\r
1329 adds \reg, \reg, #0x40000000\r
1330 orrcs \reg, \reg, #0xf0000000\r
1331 mov \reg, \reg, ror #27\r
1332.endm\r
1333\r
1334\r
1335@ Convert 0000bbb0 ggg0rrr0\r
1336@ to rrrrrggg gggbbbbb\r
1337\r
1338@ r2,r3,r9 - scratch, lr = 0x001c001c, r8 = 0x00030003\r
1339.macro convRGB565 reg\r
1340 and r2, lr, \reg,lsl #1\r
1341 and r9, r8, \reg,lsr #2\r
1342 orr r2, r2, r9 @ r2=red\r
1343 and r3, lr, \reg,lsr #7\r
1344 and r9, r8, \reg,lsr #10\r
1345 orr r3, r3, r9 @ r3=blue\r
1346 and \reg, \reg, lr, lsl #3\r
1347 orr \reg, \reg, \reg,lsl #3 @ green\r
1348 orr \reg, \reg, r2, lsl #11 @ add red back\r
1349 orr \reg, \reg, r3 @ add blue back\r
1350.endm\r
1351\r
1352vidConvCpyRGB565: @ void *to, void *from, int pixels\r
1353 stmfd sp!, {r4-r9,lr}\r
1354\r
1355 mov r12, r2, lsr #3 @ repeats\r
1356 mov lr, #0x001c0000\r
1357 orr lr, lr, #0x01c @ lr == pattern 0x001c001c\r
1358 mov r8, #0x00030000\r
1359 orr r8, r8, #0x003 @ lr == pattern 0x001c001c\r
1360\r
1361.loopRGB565:\r
1362 subs r12, r12, #1\r
1363\r
1364 ldmia r1!, {r4-r7}\r
1365 convRGB565 r4\r
1366 str r4, [r0], #4\r
1367 convRGB565 r5\r
1368 str r5, [r0], #4\r
1369 convRGB565 r6\r
1370 str r6, [r0], #4\r
1371 convRGB565 r7\r
1372 str r7, [r0], #4\r
1373\r
1374 bgt .loopRGB565\r
1375\r
1376 ldmfd sp!, {r4-r9,lr}\r
1377 bx lr\r
1378\r
1379\r
1380\r
1381.global FinalizeLineRGB555 @ int sh\r
1382\r
1383FinalizeLineRGB555:\r
1384 stmfd sp!, {r4-r8,lr}\r
1385 ldr r5, =(Pico+0x22228) @ Pico.video\r
1386 ldr r4, =HighPal\r
1387 mov r6, r0\r
1388\r
1389 ldrb r7, [r5, #-0x1a] @ 0x2220e ~ dirtyPal\r
1390 tst r7, r7\r
1391 beq .fl_noconvRGB555\r
1392 mov r1, #0\r
1393 strb r1, [r5, #-0x1a]\r
1394 sub r1, r5, #0x128 @ r1=Pico.cram\r
1395 mov r0, r4\r
1396 mov r2, #0x40\r
1397 bl vidConvCpyRGB565\r
1398\r
1399.fl_noconvRGB555:\r
1400 ldr r0, =DrawLineDest\r
1401 ldr r0, [r0]\r
1402\r
1403 ldrb r12, [r5, #12]\r
1404 tst r12, #1\r
1405 movne r2, #320/8 @ len\r
1406 moveq r2, #256/8\r
1407 ldreq r3, =PicoOpt\r
1408 ldreq r3, [r3]\r
1409 tsteq r3, #0x100\r
1410 addeq r0, r0, #32*2\r
1411\r
1412 mov r3, r4\r
1413 tst r6, r6\r
1414 beq .fl_noshRGB555\r
1415 tst r7, r7\r
1416 beq .fl_noshRGB555\r
1417\r
1418 @ shadowed pixels:\r
1419 mov r12, #0x008e\r
1420 orr r12,r12,#0x7300\r
1421 orr r12,r12,r12,lsl #16\r
1422 add r4, r3, #0x40*2\r
1423 add r5, r3, #0xc0*2\r
1424 mov lr, #0x40/4\r
1425.fl_loopcpRGB555_sh:\r
1426 subs lr, lr, #1\r
1427 ldmia r3!, {r1,r6}\r
1428 and r1, r12, r1, lsr #1\r
1429 and r6, r12, r6, lsr #1\r
1430 stmia r4!, {r1,r6}\r
1431 stmia r5!, {r1,r6}\r
1432 bne .fl_loopcpRGB555_sh\r
1433\r
1434 @ hilighted pixels:\r
1435 sub r3, r3, #0x40*2\r
1436 mov lr, #0x40/2\r
1437.fl_loopcpRGB555_hi:\r
1438 ldr r1, [r3], #4\r
1439 TileDoShHi2Pixels555 r1\r
1440 str r1, [r4], #4\r
1441 subs lr, lr, #1\r
1442 bne .fl_loopcpRGB555_hi\r
1443\r
1444 sub r3, r3, #0x40*2\r
1445\r
1446\r
1447.fl_noshRGB555:\r
1448@ ldr r1, =(HighCol+8)\r
1449 ldr r1, =DrawLineInt\r
1450 ldr r1, [r1]\r
1451 add r1, r1, #8\r
1452 mov lr, #0xff\r
1453 mov lr, lr, lsl #1\r
1454\r
1455.fl_loopRGB555:\r
1456 subs r2, r2, #1\r
1457\r
1458 ldr r12, [r1], #4\r
1459 ldr r7, [r1], #4\r
1460\r
1461 and r4, lr, r12, lsl #1\r
1462 ldrh r4, [r3, r4]\r
1463 and r5, lr, r12, lsr #7\r
1464 ldrh r5, [r3, r5]\r
1465 orr r4, r4, r5, lsl #16\r
1466\r
1467 and r5, lr, r12, lsr #15\r
1468 ldrh r5, [r3, r5]\r
1469 and r6, lr, r12, lsr #23\r
1470 ldrh r6, [r3, r6]\r
1471 orr r5, r5, r6, lsl #16\r
1472\r
1473 and r8, lr, r7, lsl #1\r
1474 ldrh r8, [r3, r8]\r
1475 and r6, lr, r7, lsr #7\r
1476 ldrh r6, [r3, r6]\r
1477 orr r8, r8, r6, lsl #16\r
1478\r
1479 and r12,lr, r7, lsr #15\r
1480 ldrh r12,[r3, r12]\r
1481 and r6, lr, r7, lsr #23\r
1482 ldrh r6, [r3, r6]\r
1483 orr r12,r12, r6, lsl #16\r
1484\r
1485 stmia r0!, {r4,r5,r8,r12}\r
1486 bne .fl_loopRGB555\r
1487\r
1488\r
1489 ldmfd sp!, {r4-r8,lr}\r
1490 bx lr\r
1491\r
1492@ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r
1493\r
1494@ utility\r
1495.global blockcpy @ void *dst, void *src, size_t n\r
1496\r
1497blockcpy:\r
1498 stmfd sp!, {r4,r5}\r
1499 mov r2, r2, lsr #4\r
1500blockcpy_loop:\r
1501 subs r2, r2, #1\r
1502 ldmia r1!, {r3-r5,r12}\r
1503 stmia r0!, {r3-r5,r12}\r
1504 bne blockcpy_loop\r
1505 ldmfd sp!, {r4,r5}\r
1506 bx lr\r
1507\r
1508\r
1509.global blockcpy_or @ void *dst, void *src, size_t n, int pat\r
1510\r
1511blockcpy_or:\r
1512 stmfd sp!, {r4-r6}\r
1513 orr r3, r3, r3, lsl #8\r
1514 orr r3, r3, r3, lsl #16\r
1515 mov r2, r2, lsr #4\r
1516blockcpy_loop_or:\r
1517 subs r2, r2, #1\r
1518 ldmia r1!, {r4-r6,r12}\r
1519 orr r4, r4, r3\r
1520 orr r5, r5, r3\r
1521 orr r6, r6, r3\r
1522 orr r12,r12,r3\r
1523 stmia r0!, {r4-r6,r12}\r
1524 bne blockcpy_loop_or\r
1525 ldmfd sp!, {r4-r6}\r
1526 bx lr\r
1527\r
fast MegaDrive/MegaCD/32X emulator