notaz.gp2x.de / picodrive.git / blame
? search:
summary | shortlog | log | commit | commitdiff | tree
blob | blame (incremental) | history | HEAD
sprite acching removed in asm too
[picodrive.git] / Pico / Draw.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
283fec1b 22.equ PDRAW_SPRITES_MOVED, (1<<0)\r
23.equ PDRAW_WND_DIFF_PRIO, (1<<1)\r
24.equ PDRAW_ACC_SPRITES, (1<<2)\r
25.equ PDRAW_DIRTY_SPRITES, (1<<4)\r
26.equ PDRAW_PLANE_HI_PRIO, (1<<6)\r
27.equ PDRAW_SHHI_DONE, (1<<7)\r
28.equ MAX_LINE_SPRITES, 30\r
cc68a136 29\r
30@ helper\r
31.macro TilePixel pat lsrr offs\r
32.if !\lsrr\r
33 ands r4, \pat, r2\r
34.else\r
35 ands r4, \pat, r2, lsr #\lsrr\r
36.endif\r
37 orrne r4, r3, r4\r
38 strneb r4, [r1,#\offs]\r
39.endm\r
40\r
41@ TileNorm (r1=pdest, r2=pixels8, r3=pal) r4: scratch, pat: register with helper pattern 0xf\r
42.macro TileNorm pat\r
43 TilePixel \pat, 12, 0 @ #0x0000f000\r
44 TilePixel \pat, 8, 1 @ #0x00000f00\r
45 TilePixel \pat, 4, 2 @ #0x000000f0\r
46 TilePixel \pat, 0, 3 @ #0x0000000f\r
47 TilePixel \pat, 28, 4 @ #0xf0000000\r
48 TilePixel \pat, 24, 5 @ #0x0f000000\r
49 TilePixel \pat, 20, 6 @ #0x00f00000\r
50 TilePixel \pat, 16, 7 @ #0x000f0000\r
51.endm\r
52\r
53@ TileFlip (r1=pdest, r2=pixels8, r3=pal) r4: scratch, pat: register with helper pattern 0xf\r
54.macro TileFlip pat\r
55 TilePixel \pat, 16, 0 @ #0x000f0000\r
56 TilePixel \pat, 20, 1 @ #0x00f00000\r
57 TilePixel \pat, 24, 2 @ #0x0f000000\r
58 TilePixel \pat, 28, 3 @ #0xf0000000\r
59 TilePixel \pat, 0, 4 @ #0x0000000f\r
60 TilePixel \pat, 4, 5 @ #0x000000f0\r
61 TilePixel \pat, 8, 6 @ #0x00000f00\r
62 TilePixel \pat, 12, 7 @ #0x0000f000\r
63.endm\r
64\r
65@ shadow/hilight mode\r
66\r
67@ this one is for hi priority layer\r
68.macro TilePixelShHP lsrr offs\r
69.if !\lsrr\r
70 ands r4, r12, r2\r
71.else\r
72 ands r4, r12, r2, lsr #\lsrr\r
73.endif\r
74 ldreqb r4, [r1,#\offs]\r
75 orrne r4, r3, r4\r
cc68a136 76 andeq r4, r4, #0x3f\r
07abbab1 77 strb r4, [r1,#\offs]\r
cc68a136 78.endm\r
79\r
7a7c6476 80@ TileNormShHP (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r12: register with helper pattern 0xf, touches r3 high bits\r
cc68a136 81.macro TileNormShHP\r
82 TilePixelShHP 12, 0 @ #0x0000f000\r
83 TilePixelShHP 8, 1 @ #0x00000f00\r
84 TilePixelShHP 4, 2 @ #0x000000f0\r
85 TilePixelShHP 0, 3 @ #0x0000000f\r
86 TilePixelShHP 28, 4 @ #0xf0000000\r
87 TilePixelShHP 24, 5 @ #0x0f000000\r
88 TilePixelShHP 20, 6 @ #0x00f00000\r
89 TilePixelShHP 16, 7 @ #0x000f0000\r
90.endm\r
91\r
7a7c6476 92@ TileFlipShHP (r1=pdest, r2=pixels8, r3=pal) r4: scratch, pat: register with helper pattern 0xf\r
cc68a136 93.macro TileFlipShHP\r
94 TilePixelShHP 16, 0 @ #0x000f0000\r
95 TilePixelShHP 20, 1 @ #0x00f00000\r
96 TilePixelShHP 24, 2 @ #0x0f000000\r
97 TilePixelShHP 28, 3 @ #0xf0000000\r
98 TilePixelShHP 0, 4 @ #0x0000000f\r
99 TilePixelShHP 4, 5 @ #0x000000f0\r
100 TilePixelShHP 8, 6 @ #0x00000f00\r
101 TilePixelShHP 12, 7 @ #0x0000f000\r
102.endm\r
103\r
104\r
105@ TileSingleSh (r1=pdest, r2=pixels8, r3=pal) r4,r7: scratch, r0=sx; r12: helper pattern 0xf\r
106.macro TileSingleSh\r
107 tst r0, #1 @ not aligned?\r
108 mov r7, #0x00c000\r
109 orr r7, r7, #0xc0\r
110 ldrneb r4, [r1]\r
111 ldreqh r4, [r1]\r
112 orr r4, r4, r7\r
113 strneb r4, [r1], #1\r
114 streqh r4, [r1], #2\r
115 ldrh r4, [r1]\r
116 orr r4, r4, r7\r
117 strh r4, [r1], #2\r
118 ldrh r4, [r1]\r
119 orr r4, r4, r7\r
120 strh r4, [r1], #2\r
121 ldrh r4, [r1]\r
122 orr r4, r4, r7\r
123 strh r4, [r1], #2\r
124 ldrneb r4, [r1]\r
125 orr r4, r4, r7\r
126 strneb r4, [r1], #1\r
127.endm\r
128\r
129@ TileSingleHi (r1=pdest, r2=pixels8, r3=pal) r4,r7: scratch, r0=sx, r12: register with helper pattern 0xf\r
130.macro TileSingleHi\r
131 tst r1, #1 @ not aligned?\r
132 mov r7, #0x008000\r
133 orr r7, r7, #0x80\r
134 ldrneb r4, [r1], #1\r
135 ldreqh r4, [r1], #2 @ 1ci\r
136 ldrh r12, [r1], #2\r
137 bic r4, r4, r7, lsr #1\r
138 orr r4, r4, r7\r
139 strneb r4, [r1, #-3]\r
140 streqh r4, [r1, #-4]\r
141 ldrh r4, [r1], #2\r
142 bic r12, r12, r7, lsr #1\r
143 orr r12, r12, r7\r
144 strh r12, [r1, #-4]\r
145 ldrh r12, [r1], #2\r
146 bic r4, r4, r7, lsr #1\r
147 orr r4, r4, r7\r
148 strh r4, [r1, #-4]\r
149 ldrneb r4, [r1]\r
150 bic r12, r12, r7, lsr #1\r
151 orr r12, r12, r7\r
152 strh r12, [r1, #-2]\r
153 bicne r4, r4, r7, lsr #1\r
154 orrne r4, r4, r7\r
155 strneb r4, [r1], #1\r
156 mov r12, #0xf\r
157.endm\r
158\r
159.macro TileDoShGenPixel shift ofs\r
160.if \shift\r
161 ands r4, r12, r2, lsr #\shift\r
162.else\r
163 ands r4, r12, r2\r
164.endif\r
07abbab1 165 beq 0f\r
cc68a136 166 cmp r4, #0xe\r
07abbab1 167 ldrgeb r4, [r1,#\ofs]\r
168 orrlt r4, r3, r4 @ normal\r
169\r
170 biceq r4, r4, #0xc0 @ hilight\r
171 orreq r4, r4, #0x80\r
172 orrgt r4, r4, #0xc0 @ shadow\r
173\r
cc68a136 174 strb r4, [r1,#\ofs]\r
07abbab1 1750:\r
cc68a136 176.endm\r
177\r
178@ TileFlipSh (r1=pdest, r2=pixels8, r3=pal) r4,r7: scratch, r0=sx, r12: register with helper pattern 0xf\r
179.macro TileFlipSh\r
180 TileDoShGenPixel 16, 0 @ #0x000f0000\r
181 TileDoShGenPixel 20, 1 @ #0x00f00000\r
182 TileDoShGenPixel 24, 2 @ #0x0f000000\r
183 TileDoShGenPixel 28, 3 @ #0xf0000000\r
184 TileDoShGenPixel 0, 4 @ #0x0000000f\r
185 TileDoShGenPixel 4, 5 @ #0x000000f0\r
186 TileDoShGenPixel 8, 6 @ #0x00000f00\r
187 TileDoShGenPixel 12, 7 @ #0x0000f000\r
188.endm\r
189\r
190@ TileNormSh (r1=pdest, r2=pixels8, r3=pal) r4,r7: scratch, r0=sx, r12: register with helper pattern 0xf\r
191.macro TileNormSh\r
192 TileDoShGenPixel 12, 0 @ #0x0000f000\r
193 TileDoShGenPixel 8, 1 @ #0x00000f00\r
194 TileDoShGenPixel 4, 2 @ #0x000000f0\r
195 TileDoShGenPixel 0, 3 @ #0x0000000f\r
196 TileDoShGenPixel 28, 4 @ #0xf0000000\r
197 TileDoShGenPixel 24, 5 @ #0x0f000000\r
198 TileDoShGenPixel 20, 6 @ #0x00f00000\r
199 TileDoShGenPixel 16, 7 @ #0x000f0000\r
200.endm\r
201\r
07abbab1 202.macro TileDoShGenPixel_noop shift ofs\r
203.if \shift\r
204 and r4, r12, r2, lsr #\shift\r
205.else\r
206 and r4, r12, r2\r
207.endif\r
208 sub r7, r4, #1\r
209 cmp r7, #0xd\r
210 orrcc r4, r3, r4 @ 0-0xc (was 1-0xd)\r
211 strccb r4, [r1,#\ofs]\r
212.endm\r
213\r
214.macro TileFlipSh_noop\r
215 TileDoShGenPixel_noop 16, 0 @ #0x000f0000\r
216 TileDoShGenPixel_noop 20, 1 @ #0x00f00000\r
217 TileDoShGenPixel_noop 24, 2 @ #0x0f000000\r
218 TileDoShGenPixel_noop 28, 3 @ #0xf0000000\r
219 TileDoShGenPixel_noop 0, 4 @ #0x0000000f\r
220 TileDoShGenPixel_noop 4, 5 @ #0x000000f0\r
221 TileDoShGenPixel_noop 8, 6 @ #0x00000f00\r
222 TileDoShGenPixel_noop 12, 7 @ #0x0000f000\r
223.endm\r
224\r
225.macro TileNormSh_noop\r
226 TileDoShGenPixel_noop 12, 0 @ #0x0000f000\r
227 TileDoShGenPixel_noop 8, 1 @ #0x00000f00\r
228 TileDoShGenPixel_noop 4, 2 @ #0x000000f0\r
229 TileDoShGenPixel_noop 0, 3 @ #0x0000000f\r
230 TileDoShGenPixel_noop 28, 4 @ #0xf0000000\r
231 TileDoShGenPixel_noop 24, 5 @ #0x0f000000\r
232 TileDoShGenPixel_noop 20, 6 @ #0x00f00000\r
233 TileDoShGenPixel_noop 16, 7 @ #0x000f0000\r
234.endm\r
235\r
236.macro TileDoShGenPixel_onlyop_lp shift ofs\r
237.if \shift\r
238 ands r7, r12, r2, lsr #\shift\r
239.else\r
240 ands r7, r12, r2\r
241.endif\r
242 ldrneb r4, [r1,#\ofs]\r
243 tstne r4, #0x40\r
244 beq 0f\r
245\r
246 cmp r7, #0xe\r
247 biceq r4, r4, #0xc0 @ hilight\r
248 orreq r4, r4, #0x80\r
249 orrgt r4, r4, #0xc0 @ shadow\r
250 strgeb r4, [r1,#\ofs]\r
2510:\r
252.endm\r
253\r
254.macro TileFlipSh_onlyop_lp\r
255 TileDoShGenPixel_onlyop_lp 16, 0 @ #0x000f0000\r
256 TileDoShGenPixel_onlyop_lp 20, 1 @ #0x00f00000\r
257 TileDoShGenPixel_onlyop_lp 24, 2 @ #0x0f000000\r
258 TileDoShGenPixel_onlyop_lp 28, 3 @ #0xf0000000\r
259 TileDoShGenPixel_onlyop_lp 0, 4 @ #0x0000000f\r
260 TileDoShGenPixel_onlyop_lp 4, 5 @ #0x000000f0\r
261 TileDoShGenPixel_onlyop_lp 8, 6 @ #0x00000f00\r
262 TileDoShGenPixel_onlyop_lp 12, 7 @ #0x0000f000\r
263.endm\r
264\r
265.macro TileNormSh_onlyop_lp\r
266 TileDoShGenPixel_onlyop_lp 12, 0 @ #0x0000f000\r
267 TileDoShGenPixel_onlyop_lp 8, 1 @ #0x00000f00\r
268 TileDoShGenPixel_onlyop_lp 4, 2 @ #0x000000f0\r
269 TileDoShGenPixel_onlyop_lp 0, 3 @ #0x0000000f\r
270 TileDoShGenPixel_onlyop_lp 28, 4 @ #0xf0000000\r
271 TileDoShGenPixel_onlyop_lp 24, 5 @ #0x0f000000\r
272 TileDoShGenPixel_onlyop_lp 20, 6 @ #0x00f00000\r
273 TileDoShGenPixel_onlyop_lp 16, 7 @ #0x000f0000\r
274.endm\r
275\r
cc68a136 276\r
277@ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r
278\r
279@ struct TileStrip\r
280@ {\r
281@ int nametab; // 0x00\r
282@ int line; // 0x04\r
283@ int hscroll; // 0x08\r
284@ int xmask; // 0x0C\r
285@ int *hc; // 0x10 (pointer to cache buffer)\r
286@ int cells; // 0x14\r
287@ };\r
288\r
83c093a4 289@ void DrawLayer(int plane_sh, int *hcache, int cellskip, int maxcells);\r
cc68a136 290\r
83c093a4 291.global DrawLayer\r
cc68a136 292\r
293DrawLayer:\r
294 stmfd sp!, {r4-r11,lr}\r
295\r
296 ldr r11, =(Pico+0x22228) @ Pico.video\r
297 mov r8, #1\r
298\r
83c093a4 299 ldrb r7, [r11, #16] @ ??vv??hh\r
cc68a136 300\r
301 mov r6, r1 @ hcache\r
83c093a4 302 orr r9, r3, r0, lsl #30\r
303 orr r9, r9, r2, lsl #8 @ r9=sh[31]|cellskip[15:8]|maxcells[7:0] (tmp)\r
cc68a136 304\r
305 mov r1, r7, lsl #4\r
306 orr r1, r1, #0x00ff\r
307\r
308 and r10, r7, #3\r
309 cmp r10, #1\r
310 biclt r1, r1, #0xfc00\r
311 biceq r1, r1, #0xfe00\r
312 bicgt r1, r1, #0xff00 @ r1=ymask=(height<<8)|0xff; ...; // Y Mask in pixels\r
313\r
314 add r10, r10, #5\r
315 cmp r10, #7\r
316 subge r10, r10, #1 @ r10=shift[width] (5,6,6,7)\r
317\r
318 @ calculate xmask:\r
319 mov r5, r8, lsl r10\r
320 sub r5, r5, #1 @ r5=xmask\r
321\r
322 @ Find name table:\r
83c093a4 323 ands r0, r0, #1\r
cc68a136 324 ldreqb r12, [r11, #2]\r
325 ldrneb r12, [r11, #4]\r
326\r
b6d7ac70 327 ldr r2, =DrawScanline @ trying to make good use of pipeline here\r
cc68a136 328 ldr lr, =(Pico+0x10000) @ lr=Pico.vram\r
329\r
330 moveq r12, r12, lsl #10\r
331 movne r12, r12, lsl #13\r
332 and r12, r12, #(7<<13) @ r12=(ts->nametab<<1) (halfword compliant)\r
333\r
334 ldrh r8, [r11, #12]\r
335 ldrb r7, [r11, #11]\r
336 ldr r2, [r2]\r
337\r
338 mov r4, r8, lsr #8 @ pvid->reg[13]\r
339 mov r4, r4, lsl #10 @ htab=pvid->reg[13]<<9; (halfwords)\r
340 tst r7, #2\r
b6d7ac70 341 addne r4, r4, r2, lsl #2 @ htab+=DrawScanline<<1; // Offset by line\r
cc68a136 342 tst r7, #1\r
343 biceq r4, r4, #0x1f @ htab&=~0xf; // Offset by tile\r
344 add r4, r4, r0, lsl #1 @ htab+=plane\r
345 bic r4, r4, #0x00ff0000 @ just in case\r
346 ldrh r3, [lr, r4] @ r3=hscroll\r
347\r
348 tst r7, #4\r
349 bne .DrawStrip_vsscroll\r
350\r
351 @ Get vertical scroll value:\r
352 add r7, lr, #0x012000\r
353 add r7, r7, #0x000180 @ r7=Pico.vsram (Pico+0x22180)\r
354 ldr r7, [r7]\r
355\r
356 tst r8, #2\r
357 tstne r8, #4\r
358 bne .DrawStrip_interlace\r
359\r
360 tst r0, r0\r
361 movne r7, r7, lsr #16\r
362\r
363 @ Find the line in the name table\r
364 add r2, r2, r7\r
365 and r2, r2, r1\r
366 mov r4, r2, lsr #3\r
367 add r10, r10, #1 @ shift[width]++\r
368 add r12, r12, r4, lsl r10 @ nametab+=(ts.line>>3)<<shift[width];\r
369\r
370 @ ldmia r0, {r1,r2,r3,r5,r6,r9} @ r2=line, r3=ts->hscroll, r5=ts->xmask, r6=ts->hc, r9=ts->cells\r
cc68a136 371\r
372 and r10,r2, #7\r
373 mov r10,r10, lsl #1 @ r10=ty=(ts->line&7)<<1;\r
374 orr r10,r10, r9, lsl #24\r
375\r
376 rsb r8, r3, #0\r
377 mov r8, r8, lsr #3 @ r8=tilex=(-ts->hscroll)>>3\r
378\r
379 sub r1, r3, #1\r
380 and r1, r1, #7\r
381 add r7, r1, #1 @ r7=dx=((ts->hscroll-1)&7)+1\r
382\r
383 tst r9, #1<<31\r
384 mov r3, #0\r
740da8c6 385 orrne r10,r10, #1<<23 @ r10=(cells<<24|sh<<23|hi_not_empty<<22|had_output<<21|ty)\r
cc68a136 386 movne r3, #0x40 @ default to shadowed pal on sh mode\r
387\r
cc68a136 388 cmp r7, #8\r
389 addne r10,r10, #0x01000000 @ we will loop cells+1 times if there is scroll\r
390\r
83c093a4 391 and r9, r9, #0xff00\r
392 add r8, r8, r9, lsr #8 @ tilex+=cellskip\r
393 add r7, r7, r9, lsr #5 @ dx+=cellskip<<3;\r
394 sub r10,r10,r9, lsl #16 @ cells-=cellskip\r
83c093a4 395\r
cc68a136 396 @ cache some stuff to avoid mem access\r
ea8c405f 397.if OVERRIDE_HIGHCOL\r
398 ldr r11,=HighCol\r
399 mov r0, #0xf\r
400 ldr r11,[r11]\r
401.else\r
cc68a136 402 ldr r11,=HighCol\r
403 mov r0, #0xf\r
ea8c405f 404.endif\r
405\r
406 mvn r9, #0 @ r9=prevcode=-1\r
cc68a136 407 add r1, r11, r7 @ r1=pdest\r
408\r
409\r
410 @ r4 & r7 are scratch in this loop\r
411.dsloop_subr1:\r
412 sub r1, r1, #8\r
413.dsloop: @ 40-41 times\r
414 subs r10,r10, #0x01000000\r
415 bmi .dsloop_exit\r
416\r
417.dsloop_enter:\r
418 and r7, r5, r8\r
419 add r7, lr, r7, lsl #1 @ Pico.vram+((tilex&ts->xmask) as halfwords)\r
420 ldrh r7, [r7, r12] @ r7=code (int, but from unsigned, no sign extend)\r
421\r
422 add r1, r1, #8\r
423 add r8, r8, #1\r
424\r
425 tst r7, #0x8000\r
426 bne .DrawStrip_hiprio\r
427\r
428 cmp r7, r9\r
429 beq .DrawStrip_samecode @ we know stuff about this tile already\r
430\r
431 mov r9, r7 @ remember code\r
740da8c6 432 orr r10, r10, #1<<21 @ seen non hi-prio tile\r
cc68a136 433\r
434 movs r2, r9, lsl #20 @ if (code&0x1000)\r
435 mov r2, r2, lsl #1\r
436 add r2, r2, r10, lsl #17\r
437 mov r2, r2, lsr #17\r
438 eorcs r2, r2, #0x0e @ if (code&0x1000) addr^=0xe;\r
439\r
440 ldr r2, [lr, r2, lsl #1] @ pack=*(unsigned int *)(Pico.vram+addr); // Get 8 pixels\r
441\r
442 bic r7, r3, #0x3f\r
443 and r3, r9, #0x6000\r
444 add r3, r7, r3, lsr #9 @ r3=pal=((code&0x6000)>>9);\r
445\r
446.DrawStrip_samecode:\r
447 tst r2, r2\r
448 beq .dsloop @ tileline blank\r
449\r
450 cmp r2, r2, ror #4\r
451 beq .DrawStrip_SingleColor @ tileline singlecolor \r
452\r
453 tst r9, #0x0800\r
7a7c6476 454 bne .DrawStrip_TileFlip\r
cc68a136 455\r
456 @ (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r0: helper pattern\r
cc68a136 457.DrawStrip_TileNorm:\r
458 TileNorm r0\r
459 b .dsloop\r
460\r
7a7c6476 461.DrawStrip_TileFlip:\r
462 TileFlip r0\r
463 b .dsloop\r
464\r
cc68a136 465.DrawStrip_SingleColor:\r
466 and r4, r2, #0xf\r
467 orr r4, r3, r4\r
468 orr r4, r4, r4, lsl #8\r
469 tst r1, #1 @ not aligned?\r
470 strneb r4, [r1], #1\r
471 streqh r4, [r1], #2\r
472 strh r4, [r1], #2\r
473 strh r4, [r1], #2\r
474 strh r4, [r1], #2\r
475 strneb r4, [r1], #1 @ have a remaining unaligned pixel?\r
476 b .dsloop_subr1\r
477\r
cc68a136 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 movs r2, r7, lsl #20 @ if (code&0x1000)\r
482 mov r2, r2, lsl #1\r
483 add r2, r2, r10, lsl #17\r
484 mov r2, r2, lsr #17\r
485 eorcs r2, r2, #0x0e @ if (code&0x1000) addr^=0xe;\r
486 ldr r2, [lr, r2, lsl #1] @ pack=*(unsigned int *)(Pico.vram+addr); // Get 8 pixels\r
487 mov r9, r7 @ remember code\r
488 tst r2, r2\r
740da8c6 489 beq .dsloop\r
490 orr r10, r10, #1<<22\r
491\r
492.DrawStrip_hiprio:\r
493 tst r10, #0x00c00000\r
494 beq .DrawStrip_hiprio_maybempt\r
495 sub r0, r1, r11\r
496 orr r7, r7, r0, lsl #16\r
497 orr r7, r7, r10, lsl #25 @ (ty<<25)\r
498 tst r7, #0x1000\r
499 eorne r7, r7, #7<<26 @ if(code&0x1000) cval^=7<<26;\r
500 str r7, [r6], #4 @ cache hi priority tile\r
501 mov r0, #0xf\r
cc68a136 502 b .dsloop\r
503\r
504.dsloop_exit:\r
740da8c6 505 tst r10, #1<<21 @ seen non hi-prio tile\r
506 ldreq r1, =rendstatus\r
cc68a136 507 mov r0, #0\r
740da8c6 508 ldreq r2, [r1]\r
cc68a136 509 str r0, [r6] @ terminate the cache list\r
283fec1b 510 orreq r2, r2, #PDRAW_PLANE_HI_PRIO @ had a layer with all hi-prio tiles\r
740da8c6 511 streq r2, [r1]\r
cc68a136 512\r
513 ldmfd sp!, {r4-r11,lr}\r
514 bx lr\r
515\r
6d7acf9e 516@ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r
cc68a136 517\r
518.DrawStrip_vsscroll:\r
6d7acf9e 519 rsb r8, r3, #0\r
520 mov r8, r8, lsr #3 @ r8=tilex=(-ts->hscroll)>>3\r
7b802576 521 bic r8, r8, #0x3fc00000\r
522 orr r8, r8, r5, lsl #25 @ r8=(xmask[31:25]|had_output[24]|tilex[21:0])\r
cc68a136 523\r
b6d7ac70 524 ldr r4, =DrawScanline\r
6d7acf9e 525 orr r5, r1, r10, lsl #24\r
526 ldr r4, [r4]\r
527 sub r1, r3, #1\r
528 orr r5, r5, r4, lsl #16 @ r5=(shift_width[31:24]|scanline[23:16]|ymask[15:0])\r
529 and r1, r1, #7\r
530 add r7, r1, #1 @ r7=dx=((ts->hscroll-1)&7)+1\r
cc68a136 531\r
6d7acf9e 532 mov r10,r9, lsl #16\r
83c093a4 533 tst r0, #1\r
6d7acf9e 534 orrne r10,r10, #0x8000\r
535 tst r9, #1<<31\r
536 mov r3, #0\r
537 orr r10,r10, #0xff000000 @ will be adjusted on entering loop\r
83c093a4 538 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 539 movne r3, #0x40 @ default to shadowed pal on sh mode\r
cc68a136 540\r
83c093a4 541 cmp r7, #8\r
542 subne r10,r10, #0x01000000 @ have hscroll, start with negative cell\r
543\r
544 and r9, r9, #0xff00\r
545 add r8, r8, r9, lsr #8 @ tilex+=cellskip\r
546 add r7, r7, r9, lsr #5 @ dx+=cellskip<<3;\r
547 add r10,r10,r9, lsl #16 @ cell+=cellskip\r
cc68a136 548\r
6d7acf9e 549 @ cache some stuff to avoid mem access\r
ea8c405f 550.if OVERRIDE_HIGHCOL\r
6d7acf9e 551 ldr r11,=HighCol\r
552 mov r0, #0xf\r
ea8c405f 553 ldr r11,[r11]\r
554.else\r
555 ldr r11,=HighCol\r
556 mov r0, #0xf\r
557.endif\r
6d7acf9e 558\r
ea8c405f 559 mvn r9, #0 @ r9=prevcode=-1\r
560 add r1, r11, r7 @ r1=pdest\r
6d7acf9e 561\r
562 @ r4 & r7 are scratch in this loop\r
563.dsloop_vs_subr1:\r
564 sub r1, r1, #8\r
565.dsloop_vs: @ 40-41 times\r
566 add r10,r10, #0x01000000\r
567 and r4, r10, #0x003f0000\r
568 cmp r4, r10, asr #8\r
740da8c6 569 ble .dsloop_vs_exit\r
6d7acf9e 570\r
571 @ calc offset and read tileline code to r7, also calc ty\r
572 add r7, lr, #0x012000\r
573 add r7, r7, #0x000180 @ r7=Pico.vsram (Pico+0x22180)\r
574 add r7, r7, r10,asr #23 @ vsram + ((cell&~1)<<1)\r
575 bic r7, r7, #3\r
576 tst r10,#0x8000 @ plane1?\r
577 addne r7, r7, #2\r
578 ldrh r7, [r7] @ r7=vscroll\r
579\r
580 bic r10,r10,#0xff @ clear old ty\r
7b802576 581 and r4, r5, #0xff0000 @ scanline\r
582 add r4, r4, r7, lsl #16 @ ... += vscroll\r
583 and r4, r4, r5, lsl #16 @ ... &= ymask\r
6d7acf9e 584 and r7, r4, #0x70000\r
585 orr r10,r10,r7, lsr #15 @ new ty\r
586\r
587 mov r4, r4, lsr #19\r
588 mov r7, r5, lsr #24\r
589 mov r4, r4, lsl r7 @ nametabadd\r
590\r
591 and r7, r8, r8, lsr #25\r
592 add r7, lr, r7, lsl #1 @ Pico.vram+((tilex&ts->xmask) as halfwords)\r
593 add r7, r7, r4, lsl #1\r
594 ldrh r7, [r7, r12] @ r7=code (int, but from unsigned, no sign extend)\r
595\r
596 add r1, r1, #8\r
597 add r8, r8, #1\r
598\r
599 tst r7, #0x8000\r
600 bne .DrawStrip_vs_hiprio\r
601\r
602 cmp r7, r9\r
603 beq .DrawStrip_vs_samecode @ we know stuff about this tile already\r
604\r
605 mov r9, r7 @ remember code\r
7b802576 606 orr r8, r8, #(1<<24)@ seen non hi-prio tile\r
6d7acf9e 607\r
608 movs r2, r9, lsl #20 @ if (code&0x1000)\r
609 mov r2, r2, lsl #1\r
610 add r2, r2, r10, lsl #17\r
611 mov r2, r2, lsr #17\r
612 eorcs r2, r2, #0x0e @ if (code&0x1000) addr^=0xe;\r
613\r
614 ldr r2, [lr, r2, lsl #1] @ pack=*(unsigned int *)(Pico.vram+addr); // Get 8 pixels\r
615\r
616 bic r7, r3, #0x3f\r
617 and r3, r9, #0x6000\r
618 add r3, r7, r3, lsr #9 @ r3=pal=((code&0x6000)>>9);\r
619\r
620.DrawStrip_vs_samecode:\r
621 tst r2, r2\r
622 beq .dsloop_vs @ tileline blank\r
623\r
624 cmp r2, r2, ror #4\r
625 beq .DrawStrip_vs_SingleColor @ tileline singlecolor \r
626\r
627 tst r9, #0x0800\r
7a7c6476 628 bne .DrawStrip_vs_TileFlip\r
6d7acf9e 629\r
630 @ (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r0: helper pattern\r
6d7acf9e 631.DrawStrip_vs_TileNorm:\r
632 TileNorm r0\r
633 b .dsloop_vs\r
634\r
7a7c6476 635.DrawStrip_vs_TileFlip:\r
636 TileFlip r0\r
637 b .dsloop_vs\r
638\r
6d7acf9e 639.DrawStrip_vs_SingleColor:\r
640 and r4, r2, #0xf\r
641 orr r4, r3, r4\r
642 orr r4, r4, r4, lsl #8\r
643 tst r1, #1 @ not aligned?\r
644 strneb r4, [r1], #1\r
645 streqh r4, [r1], #2\r
646 strh r4, [r1], #2\r
647 strh r4, [r1], #2\r
648 strh r4, [r1], #2\r
649 strneb r4, [r1], #1 @ have a remaining unaligned pixel?\r
650 b .dsloop_vs_subr1\r
651\r
652.DrawStrip_vs_hiprio:\r
653 tst r10, #0x00c00000\r
654 beq .DrawStrip_vs_hiprio_maybempt\r
655 sub r0, r1, r11\r
656 orr r7, r7, r0, lsl #16\r
657 orr r7, r7, r10, lsl #25 @ (ty<<25)\r
658 tst r7, #0x1000\r
659 eorne r7, r7, #7<<26 @ if(code&0x1000) cval^=7<<26;\r
660 str r7, [r6], #4 @ cache hi priority tile\r
661 mov r0, #0xf\r
662 b .dsloop_vs\r
663\r
664.DrawStrip_vs_hiprio_maybempt:\r
665 cmp r7, r9\r
666 beq .dsloop_vs @ must've been empty, otherwise we wouldn't get here\r
667 movs r2, r7, lsl #20 @ if (code&0x1000)\r
668 mov r2, r2, lsl #1\r
669 add r2, r2, r10, lsl #17\r
670 mov r2, r2, lsr #17\r
671 eorcs r2, r2, #0x0e @ if (code&0x1000) addr^=0xe;\r
672 ldr r2, [lr, r2, lsl #1] @ pack=*(unsigned int *)(Pico.vram+addr); // Get 8 pixels\r
673 mov r9, r7 @ remember code\r
674 tst r2, r2\r
675 orrne r10, r10, #1<<22\r
676 bne .DrawStrip_vs_hiprio\r
677 b .dsloop_vs\r
678\r
740da8c6 679.dsloop_vs_exit:\r
7b802576 680 tst r8, #(1<<24) @ seen non hi-prio tile\r
740da8c6 681 ldreq r1, =rendstatus\r
682 mov r0, #0\r
683 ldreq r2, [r1]\r
684 str r0, [r6] @ terminate the cache list\r
283fec1b 685 orreq r2, r2, #PDRAW_PLANE_HI_PRIO @ had a layer with all hi-prio tiles\r
740da8c6 686 streq r2, [r1]\r
687\r
688 ldmfd sp!, {r4-r11,lr}\r
689 bx lr\r
690\r
6d7acf9e 691\r
692@ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r
cc68a136 693\r
694@ interlace mode 2? Sonic 2?\r
695.DrawStrip_interlace:\r
696 tst r0, r0\r
697 moveq r7, r7, lsl #21\r
698 movne r7, r7, lsl #5\r
699\r
700 @ Find the line in the name table\r
b6d7ac70 701 add r2, r7, r2, lsl #22 @ r2=(vscroll+(DrawScanline<<1))<<21 (11 bits);\r
cc68a136 702 orr r1, r1, #0x80000000\r
703 and r2, r2, r1, ror #10 @ &((ymask<<1)|1)<<21;\r
704 mov r2, r2, lsr #21\r
705 mov r4, r2, lsr #4\r
706 mov r12, r12, lsr #1 @ halfwords\r
707 add r0, r12, r4, lsl r10 @ nametab+=(ts.line>>4)<<shift[width];\r
708 and r9, r9, #0xff\r
709\r
710 sub sp, sp, #6*4\r
711 stmia sp, {r0,r2,r3,r5,r6,r9}\r
712\r
713 mov r0, sp\r
714 bl DrawStripInterlace @ struct TileStrip *ts\r
715\r
716 add sp, sp, #6*4\r
717 ldmfd sp!, {r4-r11,lr}\r
718 bx lr\r
719\r
720.pool\r
721\r
722@ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r
723\r
724\r
725.global BackFill @ int reg7, int sh\r
726\r
727BackFill:\r
728 stmfd sp!, {r4-r9,lr}\r
729\r
ea8c405f 730.if OVERRIDE_HIGHCOL\r
731 ldr lr, =HighCol\r
732 mov r0, r0, lsl #26\r
733 ldr lr, [lr]\r
734 mov r0, r0, lsr #26\r
735 add lr, lr, #8\r
736.else\r
cc68a136 737 ldr lr, =(HighCol+8)\r
cc68a136 738 mov r0, r0, lsl #26\r
739 mov r0, r0, lsr #26\r
ea8c405f 740.endif\r
741\r
cc68a136 742 orr r0, r0, r1, lsl #6\r
743 orr r0, r0, r0, lsl #8\r
744 orr r0, r0, r0, lsl #16\r
745\r
746 mov r1, r0\r
747 mov r2, r0\r
748 mov r3, r0\r
749 mov r4, r0\r
750 mov r5, r0\r
751 mov r6, r0\r
752 mov r7, r0\r
753\r
754 @ go go go!\r
755 stmia lr!, {r0-r7} @ 10*8*4\r
756 stmia lr!, {r0-r7}\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\r
766 ldmfd sp!, {r4-r9,r12}\r
767 bx r12\r
768\r
769\r
770@ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r
771\r
772\r
7a7c6476 773.global DrawTilesFromCache @ int *hc, int sh, int rlim\r
cc68a136 774\r
775DrawTilesFromCache:\r
776 stmfd sp!, {r4-r8,r11,lr}\r
777\r
cc68a136 778 @ cache some stuff to avoid mem access\r
ea8c405f 779.if OVERRIDE_HIGHCOL\r
cc68a136 780 ldr r11,=HighCol\r
cc68a136 781 mov r12,#0xf\r
ea8c405f 782 ldr r11,[r11]\r
783.else\r
784 ldr r11,=HighCol\r
785 mov r12,#0xf\r
786.endif\r
787 ldr lr, =(Pico+0x10000) @ lr=Pico.vram\r
cc68a136 788\r
740da8c6 789 mvn r5, #0 @ r5=prevcode=-1\r
7a7c6476 790 ands r8, r1, #1\r
791 orr r8, r8, r2, lsl #1\r
740da8c6 792 bne .dtfc_check_rendflags\r
793\r
cc68a136 794 @ scratch: r4, r7\r
795.dtfc_loop:\r
796 ldr r6, [r0], #4 @ read code\r
797 movs r1, r6, lsr #16 @ r1=dx;\r
798 ldmeqfd sp!, {r4-r8,r11,pc} @ dx is never zero, this must be a terminator, return\r
7a7c6476 799 bic r4, r1, #0xfe00\r
800 add r1, r11, r4 @ r1=pdest\r
cc68a136 801\r
802 mov r7, r6, lsl #16\r
803 cmp r5, r7, lsr #16\r
804 beq .dtfc_samecode @ if (code==prevcode)\r
805\r
806 mov r5, r7, lsr #16\r
807\r
808 mov r2, r5, lsl #21\r
809 mov r2, r2, lsr #17 @ r2=addr=(code&0x7ff)<<4;\r
810 add r2, r2, r6, lsr #25 @ addr+=ty\r
811\r
812 and r3, r5, #0x6000\r
813 mov r3, r3, lsr #9 @ r3=pal=((code&0x6000)>>9);\r
814\r
815 ldr r2, [lr, r2, lsl #1] @ pack=*(unsigned int *)(Pico.vram+addr); // Get 8 pixels\r
816\r
817.dtfc_samecode:\r
7a7c6476 818 rsbs r4, r4, r8, lsr #1\r
819 bmi .dtfc_cut_tile\r
820\r
821 tst r8, #1\r
cc68a136 822 bne .dtfc_shadow\r
823\r
824 tst r2, r2\r
825 beq .dtfc_loop\r
826\r
827 cmp r2, r2, ror #4\r
828 beq .dtfc_SingleColor @ tileline singlecolor \r
829\r
830 tst r5, #0x0800\r
7a7c6476 831 bne .dtfc_TileFlip\r
cc68a136 832\r
833 @ (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r12: helper pattern\r
cc68a136 834.dtfc_TileNorm:\r
835 TileNorm r12\r
836 b .dtfc_loop\r
837\r
7a7c6476 838.dtfc_TileFlip:\r
839 TileFlip r12\r
840 b .dtfc_loop\r
841\r
cc68a136 842.dtfc_SingleColor:\r
843 and r4, r2, #0xf\r
844 orr r4, r3, r4\r
845 orr r4, r4, r4, lsl #8\r
846 tst r1, #1 @ not aligned?\r
847 strneb r4, [r1], #1\r
848 streqh r4, [r1], #2\r
849 strh r4, [r1], #2\r
850 strh r4, [r1], #2\r
851 strh r4, [r1], #2\r
852 strneb r4, [r1], #1 @ have a remaining unaligned pixel?\r
853 b .dtfc_loop\r
854\r
855.dtfc_shadow:\r
856 tst r2, r2\r
857 beq .dtfc_shadow_blank\r
858\r
859 cmp r2, r2, ror #4\r
860 beq .dtfc_SingleColor @ tileline singlecolor \r
861\r
862 tst r5, #0x0800\r
7a7c6476 863 bne .dtfc_TileFlipShHP\r
cc68a136 864\r
865 @ (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r12: helper pattern\r
cc68a136 866.dtfc_TileNormShHP:\r
867 TileNormShHP\r
868 b .dtfc_loop\r
869\r
7a7c6476 870.dtfc_TileFlipShHP:\r
871 TileFlipShHP\r
872 b .dtfc_loop\r
873\r
cc68a136 874.dtfc_shadow_blank:\r
07abbab1 875 tst r1, #1\r
876 ldrneb r4, [r1]\r
877 mov r6, #0x3f\r
878 and r4, r4, #0x3f\r
879 strneb r4, [r1], #1\r
880 ldrh r4, [r1]\r
881 orr r6, r6, r6, lsl #8\r
882 and r4, r4, r6\r
883 strh r4, [r1], #2\r
884 ldrh r4, [r1]\r
885 and r4, r4, r6\r
886 strh r4, [r1], #2\r
887 ldrh r4, [r1]\r
888 and r4, r4, r6\r
889 strh r4, [r1], #2\r
890 ldrh r4, [r1]\r
891 and r4, r4, r6\r
892 streqh r4, [r1]\r
893 strneb r4, [r1]\r
cc68a136 894 b .dtfc_loop\r
895\r
7a7c6476 896.dtfc_cut_tile:\r
897 add r4, r4, #7 @ 0-6\r
898 mov r4, r4, lsl #2\r
899 mov r12,#0xf<<28\r
900 mov r12,r12,asr r4\r
901 mov r2, r2, ror #16\r
902 tst r5, #0x0800 @ flipped?\r
903 mvnne r12,r12\r
904 and r2, r2, r12\r
905 mov r2, r2, ror #16\r
906 mov r12,#0xf\r
907 tst r8, #1\r
908 bne .dtfc_shadow\r
909 tst r2, r2\r
910 beq .dtfc_loop\r
911 tst r5, #0x0800\r
912 beq .dtfc_TileNorm\r
913 b .dtfc_TileFlip\r
914\r
740da8c6 915@ check if we have detected layer covered with hi-prio tiles:\r
916.dtfc_check_rendflags:\r
917 ldr r1, =rendstatus\r
918 ldr r2, [r1]\r
283fec1b 919 tst r2, #(PDRAW_PLANE_HI_PRIO|PDRAW_SHHI_DONE)\r
740da8c6 920 beq .dtfc_loop\r
7a7c6476 921 bic r8, r8, #1 @ sh/hi mode off\r
283fec1b 922 tst r2, #PDRAW_SHHI_DONE\r
740da8c6 923 bne .dtfc_loop @ already processed\r
283fec1b 924 orr r2, r2, #PDRAW_SHHI_DONE\r
740da8c6 925 str r2, [r1]\r
926\r
927 add r1, r11,#8\r
07abbab1 928 mov r3, #320/4/4\r
740da8c6 929 mov r6, #0x3f\r
930 orr r6, r6, r6, lsl #8\r
931 orr r6, r6, r6, lsl #16\r
932.dtfc_loop_shprep:\r
07abbab1 933 ldmia r1, {r2,r4,r5,r7}\r
740da8c6 934 subs r3, r3, #1\r
07abbab1 935 and r2, r2, r6\r
936 and r4, r4, r6\r
937 and r5, r5, r6\r
938 and r7, r7, r6\r
939 stmia r1!,{r2,r4,r5,r7}\r
940 bne .dtfc_loop_shprep\r
941\r
942 mvn r5, #0 @ r5=prevcode=-1\r
943 b .dtfc_loop\r
740da8c6 944\r
cc68a136 945.pool\r
946\r
947@ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r
948\r
949\r
e352c3af 950.global DrawSpritesSHi @ int prio_unused, int sh\r
951\r
952DrawSpritesSHi:\r
953 ldr r3, =DrawScanline\r
954 ldr r2, =HighLnSpr\r
955 ldr r12,[r3]\r
956 mov r3, #(MAX_LINE_SPRITES+2)\r
957 mla r2, r12, r3, r2\r
958 ldr r3, [r2]\r
959 ands r3, r3, #0x7f\r
960 bxeq lr\r
cc68a136 961\r
cc68a136 962 stmfd sp!, {r4-r11,lr}\r
e352c3af 963 mov r12,#0xff\r
964 strb r12,[r2,#1] @ set end marker\r
965 add r10,r2, #2\r
966 add r10,r10,r3 @ r10=HighLnSpr end\r
967\r
968 str r1, [sp, #-4] @ no calls after this point\r
cc68a136 969\r
ea8c405f 970.if OVERRIDE_HIGHCOL\r
cc68a136 971 ldr r11,=HighCol\r
ea8c405f 972 mov r12,#0xf\r
973 ldr r11,[r11]\r
974.else\r
975 ldr r11,=HighCol\r
976 mov r12,#0xf\r
977.endif\r
cc68a136 978 ldr lr, =(Pico+0x10000) @ lr=Pico.vram\r
cc68a136 979\r
cc68a136 980\r
e352c3af 981DrawSpriteSHi:\r
982 @ draw next sprite\r
983 ldrb r0, [r10,#-1]!\r
984 ldr r1, =HighPreSpr\r
985 ldr r8, [sp, #-4]\r
986 cmp r0, #0xff\r
987 ldmeqfd sp!, {r4-r11,pc} @ end of list\r
988 and r0, r0, #0x7f\r
989 add r0, r1, r0, lsl #3\r
990\r
991 ldr r9, [r0, #4] @ sprite[1]\r
992 mov r2, r9, asr #16 @ r2=sx\r
cc68a136 993\r
e352c3af 994 mov r9, r9, lsl #16\r
995 mov r3, r9, lsr #31 @ priority\r
996 mov r9, r9, lsr #16\r
997 orr r9, r9, r8, lsl #31 @ r9=code|sh[31]\r
998 and r4, r9, #0x6000\r
999 orr r9, r9, r4, lsl #16\r
1000 orr r9, r9, #0x10000000 @ r9=scc1 ???? ... <code> (s=shadow/hilight, cc=pal)\r
1001 cmp r12,r9, lsr #28 @ sh/hi with pal3?\r
1002 cmpne r3, #1 @ if not, is hi prio\r
1003 bne DrawSpriteSHi @ non-operator low sprite, already drawn\r
cc68a136 1004\r
e352c3af 1005 ldr r3, [r0] @ sprite[0]\r
1006 ldr r7, =DrawScanline\r
1007 mov r6, r3, lsr #28\r
1008 sub r6, r6, #1 @ r6=width-1 (inc later)\r
1009 mov r5, r3, lsr #24\r
1010 and r5, r5, #7 @ r5=height\r
cc68a136 1011\r
e352c3af 1012 mov r0, r3, lsl #16 @ r4=sy<<16 (tmp)\r
cc68a136 1013\r
e352c3af 1014 ldr r7, [r7]\r
1015 sub r7, r7, r0, asr #16 @ r7=row=DrawScanline-sy\r
cc68a136 1016\r
e352c3af 1017 tst r9, #0x1000\r
1018 movne r0, r5, lsl #3\r
1019 subne r0, r0, #1\r
1020 subne r7, r0, r7 @ if (code&0x1000) row=(height<<3)-1-row; // Flip Y\r
cc68a136 1021\r
e352c3af 1022 add r8, r9, r7, lsr #3 @ tile+=row>>3; // Tile number increases going down\r
1023 tst r9, #0x0800\r
1024 mlane r8, r5, r6, r8 @ if (code&0x0800) { tile+=delta*(width-1);\r
1025 rsbne r5, r5, #0 @ delta=-delta; } // r5=delta now\r
cc68a136 1026\r
e352c3af 1027 mov r8, r8, lsl #21\r
1028 mov r8, r8, lsr #17\r
1029 and r7, r7, #7\r
1030 add r8, r8, r7, lsl #1 @ tile+=(row&7)<<1; // Tile address\r
1031\r
1032 mov r5, r5, lsl #4 @ delta<<=4; // Delta of address\r
1033\r
1034 orrs r3, r9, r9, lsl #4\r
1035 mov r3, r4, lsr #9 @ r3=pal=((code>>9)&0x30);\r
1036\r
1037 add r6, r6, #1 @ inc now\r
1038 adds r0, r2, #0 @ mov sx to r0 and set ZV flags\r
1039 b .dsprShi_loop_enter\r
1040\r
1041.dsprShi_loop:\r
1042 subs r6, r6, #1 @ width--\r
1043 beq DrawSpriteSHi\r
1044 adds r0, r0, #8 @ sx+=8\r
1045 add r8, r8, r5 @ tile+=delta\r
1046\r
1047.dsprShi_loop_enter:\r
1048 ble .dsprShi_loop @ sx <= 0\r
cc68a136 1049 cmp r0, #328\r
e352c3af 1050 bge DrawSpriteSHi\r
cc68a136 1051\r
1052 mov r8, r8, lsl #17\r
e352c3af 1053 mov r8, r8, lsr #17 @ tile&=0x7fff; // Clip tile address\r
cc68a136 1054\r
e352c3af 1055 ldr r2, [lr, r8, lsl #1] @ pack=*(unsigned int *)(Pico.vram+addr); // Get 8 pixels\r
1056 add r1, r11, r0 @ r1=pdest\r
cc68a136 1057 tst r2, r2\r
e352c3af 1058 beq .dsprShi_loop\r
cc68a136 1059\r
e352c3af 1060 cmp r12, r9, lsr #28\r
1061 beq .dsprShi_shadow\r
cc68a136 1062\r
1063 cmp r2, r2, ror #4\r
e352c3af 1064 beq .dsprShi_SingleColor @ tileline singlecolor \r
cc68a136 1065\r
e352c3af 1066 tst r9, #0x0800\r
1067 bne .dsprShi_TileFlip\r
cc68a136 1068\r
e352c3af 1069 @ (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r12: helper pattern\r
1070@ scratch: r4, r7\r
1071.dsprShi_TileNorm:\r
cc68a136 1072 TileNorm r12\r
e352c3af 1073 b .dsprShi_loop\r
cc68a136 1074\r
e352c3af 1075.dsprShi_TileFlip:\r
7a7c6476 1076 TileFlip r12\r
e352c3af 1077 b .dsprShi_loop\r
7a7c6476 1078\r
e352c3af 1079.dsprShi_SingleColor:\r
cc68a136 1080 and r4, r2, #0xf\r
1081 orr r4, r3, r4\r
1082 orr r4, r4, r4, lsl #8\r
e352c3af 1083 tst r0, #1 @ not aligned?\r
cc68a136 1084 strneb r4, [r1], #1\r
1085 streqh r4, [r1], #2\r
1086 strh r4, [r1], #2\r
1087 strh r4, [r1], #2\r
1088 strh r4, [r1], #2\r
1089 strneb r4, [r1], #1\r
e352c3af 1090 b .dsprShi_loop\r
cc68a136 1091\r
e352c3af 1092.dsprShi_shadow:\r
1093 tst r9, #0x8000\r
1094 beq .dsprShi_shadow_lowpri\r
07abbab1 1095\r
cc68a136 1096 cmp r2, r2, ror #4\r
e352c3af 1097 beq .dsprShi_singlec_sh\r
cc68a136 1098\r
e352c3af 1099 tst r9, #0x0800\r
1100 bne .dsprShi_TileFlip_sh\r
cc68a136 1101\r
1102 @ (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r12: helper pattern\r
e352c3af 1103.dsprShi_TileNorm_sh:\r
cc68a136 1104 TileNormSh\r
e352c3af 1105 b .dsprShi_loop\r
cc68a136 1106\r
e352c3af 1107.dsprShi_TileFlip_sh:\r
7a7c6476 1108 TileFlipSh\r
e352c3af 1109 b .dsprShi_loop\r
7a7c6476 1110\r
e352c3af 1111.dsprShi_singlec_sh:\r
cc68a136 1112 cmp r2, #0xe0000000\r
e352c3af 1113 bcc .dsprShi_SingleColor @ normal singlecolor tileline (carry inverted in ARM)\r
cc68a136 1114 tst r2, #0x10000000\r
e352c3af 1115 bne .dsprShi_sh_sh\r
cc68a136 1116 TileSingleHi\r
e352c3af 1117 b .dsprShi_loop\r
cc68a136 1118\r
e352c3af 1119.dsprShi_sh_sh:\r
cc68a136 1120 TileSingleSh\r
e352c3af 1121 b .dsprShi_loop\r
cc68a136 1122\r
e352c3af 1123.dsprShi_shadow_lowpri:\r
1124 tst r9, #0x800\r
1125 bne .dsprShi_TileFlip_sh_lp\r
07abbab1 1126\r
e352c3af 1127.dsprShi_TileNorm_sh_lp:\r
07abbab1 1128 TileNormSh_onlyop_lp\r
e352c3af 1129 b .dsprShi_loop\r
07abbab1 1130\r
e352c3af 1131.dsprShi_TileFlip_sh_lp:\r
07abbab1 1132 TileFlipSh_onlyop_lp\r
e352c3af 1133 b .dsprShi_loop\r
07abbab1 1134\r
cc68a136 1135.pool\r
1136\r
1137@ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r
1138\r
e352c3af 1139.global DrawAllSprites @ int prio, int sh\r
283fec1b 1140\r
1141DrawAllSprites:\r
1142 ldr r3, =rendstatus\r
e352c3af 1143 orr r1, r1, r0, lsl #1\r
283fec1b 1144 ldr r12,[r3]\r
1145 tst r12,#(PDRAW_ACC_SPRITES|PDRAW_SPRITES_MOVED)\r
1146 beq das_no_prep\r
e352c3af 1147 stmfd sp!, {r1,lr}\r
283fec1b 1148 and r0, r12,#PDRAW_DIRTY_SPRITES\r
1149 bic r12,r12,#(PDRAW_ACC_SPRITES|PDRAW_SPRITES_MOVED)\r
1150 str r12,[r3]\r
1151 bl PrepareSprites\r
e352c3af 1152 ldmfd sp!, {r1,lr}\r
283fec1b 1153\r
1154das_no_prep:\r
283fec1b 1155 ldr r3, =DrawScanline\r
e352c3af 1156 ldr r2, =HighLnSpr\r
283fec1b 1157 ldr r12,[r3]\r
1158 mov r3, #(MAX_LINE_SPRITES+2)\r
e352c3af 1159 mla r2, r12, r3, r2\r
1160 ldr r3, [r2]\r
283fec1b 1161 ands r3, r3, #0x7f\r
283fec1b 1162 bxeq lr\r
1163\r
1164 @ time to do some real work\r
1165 stmfd sp!, {r4-r11,lr}\r
1166 ldr r4, =rendstatus\r
1167 mov r12,#0xff\r
1168 ldr r4, [r4]\r
e352c3af 1169 strb r12,[r2,#1] @ set end marker\r
1170 add r10,r2, #2\r
283fec1b 1171 add r10,r10,r3 @ r10=HighLnSpr end\r
1172\r
e352c3af 1173 mov r8, r1, lsl #4\r
283fec1b 1174 tst r4, #PDRAW_ACC_SPRITES\r
1175 orrne r8, r8, #1\r
1176 str r8, [sp, #-4] @ no calls after this point\r
1177\r
1178.if OVERRIDE_HIGHCOL\r
1179 ldr r11,=HighCol\r
1180 mov r12,#0xf\r
1181 ldr r11,[r11]\r
1182.else\r
1183 ldr r11,=HighCol\r
1184 mov r12,#0xf\r
1185.endif\r
1186 ldr lr, =(Pico+0x10000) @ lr=Pico.vram\r
1187\r
1188@ + 0 : hhhhvvvv ----hhvv yyyyyyyy yyyyyyyy // v, h: horiz. size\r
cc68a136 1189@ + 4 : xxxxxxxx xxxxxxxx pccvhnnn nnnnnnnn // x: x coord + 8\r
1190\r
283fec1b 1191DrawSprite: @ was: unsigned int *sprite, int sh, int acc_sprites\r
1192 @ draw next sprite\r
1193 ldrb r0, [r10,#-1]!\r
1194 ldr r1, =HighPreSpr\r
283fec1b 1195 ldr r8, [sp, #-4]\r
e352c3af 1196 mov r2, r0, lsr #7\r
1197 cmp r0, #0xff\r
1198 ldmeqfd sp!, {r4-r11,pc} @ end of list\r
1199 cmp r2, r8, lsr #5\r
1200 bne DrawSprite @ wrong priority\r
283fec1b 1201 and r0, r0, #0x7f\r
1202 add r0, r1, r0, lsl #3\r
cc68a136 1203\r
283fec1b 1204@ stmfd sp!, {r4-r9,r11,lr}\r
1205@ orr r8, r2, r1, lsl #4\r
cc68a136 1206\r
1207 ldr r3, [r0] @ sprite[0]\r
b6d7ac70 1208 ldr r7, =DrawScanline\r
cc68a136 1209 mov r6, r3, lsr #28\r
1210 sub r6, r6, #1 @ r6=width-1 (inc later)\r
1211 mov r5, r3, lsr #24\r
1212 and r5, r5, #7 @ r5=height\r
1213\r
1214 mov r4, r3, lsl #16 @ r4=sy<<16 (tmp)\r
1215\r
1216 ldr r7, [r7]\r
1217 ldr r9, [r0, #4]\r
b6d7ac70 1218 sub r7, r7, r4, asr #16 @ r7=row=DrawScanline-sy\r
cc68a136 1219\r
cc68a136 1220 mov r2, r9, asr #16 @ r2=sx\r
e5fa9817 1221 mov r9, r9, lsl #16\r
1222 mov r9, r9, lsr #16\r
1223 orr r9, r9, r8, lsl #27 @ r9=code|sh[31]|as[27]\r
cc68a136 1224\r
1225 tst r9, #0x1000\r
1226 movne r4, r5, lsl #3\r
1227 subne r4, r4, #1\r
1228 subne r7, r4, r7 @ if (code&0x1000) row=(height<<3)-1-row; // Flip Y\r
1229\r
07abbab1 1230 add r8, r9, r7, lsr #3 @ tile+=row>>3; // Tile number increases going down\r
cc68a136 1231 tst r9, #0x0800\r
1232 mlane r8, r5, r6, r8 @ if (code&0x0800) { tile+=delta*(width-1);\r
1233 rsbne r5, r5, #0 @ delta=-delta; } // r5=delta now\r
1234\r
07abbab1 1235 mov r8, r8, lsl #21\r
1236 mov r8, r8, lsr #17\r
cc68a136 1237 and r7, r7, #7\r
1238 add r8, r8, r7, lsl #1 @ tile+=(row&7)<<1; // Tile address\r
1239\r
e5fa9817 1240.dspr_continue:\r
cc68a136 1241 @ cache some stuff to avoid mem access\r
cc68a136 1242 mov r5, r5, lsl #4 @ delta<<=4; // Delta of address\r
1243 and r4, r9, #0x6000\r
1244 orr r9, r9, r4, lsl #16\r
e352c3af 1245 orr r9, r9, #0x10000000 @ r9=scc1 a??? ... <code> (s=shadow/hilight, cc=pal, a=acc_spr)\r
cc68a136 1246\r
e352c3af 1247 orrs r3, r9, r9, lsl #4\r
cc68a136 1248 mov r3, r4, lsr #9 @ r3=pal=((code>>9)&0x30);\r
e352c3af 1249 orrmi r3, r3, #0x40 @ for shadow|as\r
cc68a136 1250\r
1251 add r6, r6, #1 @ inc now\r
1252 adds r0, r2, #0 @ mov sx to r0 and set ZV flags\r
1253 b .dspr_loop_enter\r
1254\r
1255.dspr_loop:\r
1256 subs r6, r6, #1 @ width--\r
283fec1b 1257 beq DrawSprite\r
cc68a136 1258 adds r0, r0, #8 @ sx+=8\r
1259 add r8, r8, r5 @ tile+=delta\r
1260\r
1261.dspr_loop_enter:\r
1262 ble .dspr_loop @ sx <= 0\r
1263 cmp r0, #328\r
283fec1b 1264 bge DrawSprite\r
cc68a136 1265\r
1266 mov r8, r8, lsl #17\r
1267 mov r8, r8, lsr #17 @ tile&=0x7fff; // Clip tile address\r
1268\r
1269 ldr r2, [lr, r8, lsl #1] @ pack=*(unsigned int *)(Pico.vram+addr); // Get 8 pixels\r
1270 add r1, r11, r0 @ r1=pdest\r
1271 tst r2, r2\r
1272 beq .dspr_loop\r
1273\r
1274 cmp r12, r9, lsr #28\r
1275 beq .dspr_shadow\r
1276\r
1277 cmp r2, r2, ror #4\r
1278 beq .dspr_SingleColor @ tileline singlecolor \r
1279\r
1280 tst r9, #0x0800\r
7a7c6476 1281 bne .dspr_TileFlip\r
cc68a136 1282\r
1283 @ (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r12: helper pattern\r
cc68a136 1284@ scratch: r4, r7\r
1285.dspr_TileNorm:\r
1286 TileNorm r12\r
1287 b .dspr_loop\r
1288\r
7a7c6476 1289.dspr_TileFlip:\r
1290 TileFlip r12\r
1291 b .dspr_loop\r
1292\r
07abbab1 1293.dspr_singlec_sh:\r
1294 cmp r2, #0xe0000000\r
1295 bcs .dspr_loop @ operator tileline, ignore\r
1296\r
cc68a136 1297.dspr_SingleColor:\r
1298 and r4, r2, #0xf\r
1299 orr r4, r3, r4\r
1300 orr r4, r4, r4, lsl #8\r
1301 tst r0, #1 @ not aligned?\r
1302 strneb r4, [r1], #1\r
1303 streqh r4, [r1], #2\r
1304 strh r4, [r1], #2\r
1305 strh r4, [r1], #2\r
1306 strh r4, [r1], #2\r
1307 strneb r4, [r1], #1\r
1308 b .dspr_loop\r
1309\r
1310.dspr_shadow:\r
1311 cmp r2, r2, ror #4\r
1312 beq .dspr_singlec_sh\r
1313\r
1314 tst r9, #0x0800\r
7a7c6476 1315 bne .dspr_TileFlip_sh\r
cc68a136 1316\r
1317 @ (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r12: helper pattern\r
cc68a136 1318.dspr_TileNorm_sh:\r
07abbab1 1319 TileNormSh_noop\r
cc68a136 1320 b .dspr_loop\r
1321\r
7a7c6476 1322.dspr_TileFlip_sh:\r
07abbab1 1323 TileFlipSh_noop\r
cc68a136 1324 b .dspr_loop\r
1325\r
1326\r
cc68a136 1327@ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r
1328\r
1329.global DrawWindow @ int tstart, int tend, int prio, int sh // int *hcache\r
1330\r
1331DrawWindow:\r
1332 stmfd sp!, {r4-r11,lr}\r
1333\r
1334 ldr r11, =(Pico+0x22228) @ Pico.video\r
b6d7ac70 1335 ldr r10, =DrawScanline\r
cc68a136 1336 ldrb r12, [r11, #3] @ pvid->reg[3]\r
1337\r
1338 ldr r10, [r10]\r
1339 ldr r4, [r11, #12]\r
1340 mov r5, r10, lsr #3\r
1341 and r10, r10, #7\r
1342 mov r10, r10, lsl #1 @ r10=ty\r
1343\r
1344 mov r12, r12, lsl #10\r
1345\r
1346 tst r4, #1 @ 40 cell mode?\r
1347 andne r12, r12, #0xf000 @ 0x3c<<10\r
1348 andeq r12, r12, #0xf800\r
1349 addne r12, r12, r5, lsl #7\r
1350 addeq r12, r12, r5, lsl #6 @ nametab\r
1351 add r12, r12, r0, lsl #2 @ +starttile\r
1352\r
1353 ldr r6, =rendstatus\r
1354 ldr lr, =(Pico+0x10000) @ lr=Pico.vram\r
07abbab1 1355 ldr r6, [r6]\r
cc68a136 1356\r
1357 @ fetch the first code now\r
1358 ldrh r7, [lr, r12]\r
1359\r
283fec1b 1360 ands r6, r6, #PDRAW_WND_DIFF_PRIO\r
cc68a136 1361 orr r6, r6, r2\r
cc68a136 1362\r
7292c709 1363 eoreq r8, r2, r7, lsr #15 @ do prio bits differ?\r
1364 cmpeq r8, #1\r
1365 ldmeqfd sp!, {r4-r11,pc} @ yes, assume that whole window uses same priority\r
cc68a136 1366\r
cc68a136 1367 orr r6, r6, r3, lsl #8 @ shadow mode\r
1368\r
1369 sub r8, r1, r0\r
cc68a136 1370\r
1371 @ cache some stuff to avoid mem access\r
ea8c405f 1372.if OVERRIDE_HIGHCOL\r
1373 ldr r11,=HighCol\r
1374 mov r8, r8, lsl #1 @ cells\r
1375 ldr r11,[r11]\r
1376 mvn r9, #0 @ r9=prevcode=-1\r
1377 add r11,r11,#8\r
1378.else\r
cc68a136 1379 ldr r11,=(HighCol+8)\r
ea8c405f 1380 mov r8, r8, lsl #1 @ cells\r
1381 mvn r9, #0 @ r9=prevcode=-1\r
1382.endif\r
07abbab1 1383 add r1, r11, r0, lsl #4 @ r1=pdest\r
cc68a136 1384 mov r0, #0xf\r
1385 b .dwloop_enter\r
1386\r
07abbab1 1387 @ r4,r5 are scratch in this loop\r
cc68a136 1388.dwloop:\r
1389 add r1, r1, #8\r
1390.dwloop_nor1:\r
1391 add r12, r12, #2 @ halfwords\r
1392 ldrh r7, [lr, r12] @ r7=code (int, but from unsigned, no sign extend)\r
1393 subs r8, r8, #1\r
1394 beq .dwloop_end @ done\r
1395\r
1396 eor r5, r6, r7, lsr #15\r
1397 tst r5, #1\r
1398 orrne r6, r6, #2 @ wrong pri\r
1399 bne .dwloop\r
1400\r
1401 cmp r7, r9\r
1402 beq .dw_samecode @ we know stuff about this tile already\r
1403\r
1404.dwloop_enter:\r
1405 mov r9, r7 @ remember code\r
1406\r
1407 movs r2, r9, lsl #20 @ if (code&0x1000)\r
1408 mov r2, r2, lsl #1\r
1409 add r2, r10, r2, lsr #17 @ r2=addr=(code&0x7ff)<<4; addr+=ty\r
1410 eorcs r2, r2, #0x0e @ if (code&0x1000) addr^=0xe;\r
1411\r
1412 and r3, r9, #0x6000\r
1413 mov r3, r3, lsr #9 @ r3=pal=((code&0x6000)>>9);\r
1414\r
1415 ldr r2, [lr, r2, lsl #1] @ pack=*(unsigned int *)(Pico.vram+addr); // Get 8 pixels\r
1416\r
1417.dw_samecode:\r
1418 tst r6, #0x100\r
1419 bne .dw_shadow\r
1420.dw_shadow_done:\r
1421 tst r2, r2\r
1422 beq .dwloop @ tileline blank\r
1423\r
1424 cmp r2, r2, ror #4\r
1425 beq .dw_SingleColor @ tileline singlecolor \r
1426\r
1427 tst r9, #0x0800\r
7a7c6476 1428 bne .dw_TileFlip\r
cc68a136 1429\r
1430 @ (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r0: helper pattern\r
cc68a136 1431.dw_TileNorm:\r
1432 TileNorm r0\r
1433 b .dwloop\r
1434\r
7a7c6476 1435.dw_TileFlip:\r
1436 TileFlip r0\r
1437 b .dwloop\r
1438\r
cc68a136 1439.dw_SingleColor:\r
1440 and r4, r0, r2 @ #0x0000000f\r
1441 orr r4, r3, r4\r
1442 orr r4, r4, r4, lsl #8\r
1443 orr r4, r4, r4, lsl #16\r
1444 mov r5, r4\r
1445 stmia r1!, {r4,r5}\r
1446 b .dwloop_nor1 @ we incremeted r1 ourselves\r
1447\r
1448.dw_shadow:\r
1449 tst r6, #1 @ hi pri?\r
1450 orreq r3, r3, #0x40\r
1451 beq .dw_shadow_done\r
1452 ldr r4, [r1]\r
07abbab1 1453 mov r5, #0x3f\r
1454 orr r5, r5, r5, lsl #8\r
1455 orr r5, r5, r5, lsl #16\r
1456 and r4, r4, r5\r
cc68a136 1457 str r4, [r1]\r
1458 ldr r4, [r1,#4]\r
07abbab1 1459 and r4, r4, r5\r
cc68a136 1460 str r4, [r1,#4]\r
1461 b .dw_shadow_done\r
1462\r
1463.dwloop_end:\r
1464 ldr r0, =rendstatus\r
1465 ldr r1, [r0]\r
1466 and r6, r6, #2\r
1467 orr r1, r1, r6\r
1468 str r1, [r0]\r
1469\r
1470 ldmfd sp!, {r4-r11,r12}\r
1471 bx r12\r
1472\r
1473\r
1474@ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r
1475\r
1476\r
1477@ hilights 2 pixels in RGB444/BGR444 format\r
1478.macro TileDoShHi2Pixels444 reg\r
1479 mov \reg, \reg, ror #12\r
1480 adds \reg, \reg, #0x40000000\r
1481 orrcs \reg, \reg, #0xf0000000\r
1482 mov \reg, \reg, ror #28\r
1483 adds \reg, \reg, #0x40000000\r
1484 orrcs \reg, \reg, #0xf0000000\r
1485 mov \reg, \reg, ror #28\r
1486 adds \reg, \reg, #0x40000000\r
1487 orrcs \reg, \reg, #0xf0000000\r
1488 mov \reg, \reg, ror #24\r
1489 adds \reg, \reg, #0x40000000\r
1490 orrcs \reg, \reg, #0xf0000000\r
1491 mov \reg, \reg, ror #28\r
1492 adds \reg, \reg, #0x40000000\r
1493 orrcs \reg, \reg, #0xf0000000\r
1494 mov \reg, \reg, ror #28\r
1495 adds \reg, \reg, #0x40000000\r
1496 orrcs \reg, \reg, #0xf0000000\r
1497 mov \reg, \reg, ror #12\r
1498.endm\r
1499\r
1500\r
1501.global FinalizeLineBGR444 @ int sh\r
1502\r
1503FinalizeLineBGR444:\r
1504 stmfd sp!, {r4-r6,lr}\r
1505 mov r6, r0\r
1506 ldr lr, =(Pico+0x22228) @ Pico.video\r
1507 ldr r0, =DrawLineDest\r
1508 ldrb r12, [lr, #12]\r
1509 ldr r0, [r0]\r
1510 sub r3, lr, #0x128 @ r3=Pico.cram\r
1511\r
1512 tst r12, #1\r
1513 movne r2, #320/4 @ len\r
1514 bne .fl_no32colBGR444\r
1515 ldr r4, =PicoOpt\r
1516 mov r2, #256/4\r
1517 ldr r4, [r4]\r
1518 tst r4, #0x100\r
1519 addeq r0, r0, #32*2\r
1520\r
1521.fl_no32colBGR444:\r
1522 tst r6, r6\r
1523 beq .fl_noshBGR444\r
1524\r
1525 ldr r4, =HighPal\r
1526\r
1527 ldrb r12, [lr, #-0x1a] @ 0x2220e ~ dirtyPal\r
1528 tst r12, r12\r
1529 moveq r3, r4\r
1530 beq .fl_noshBGR444\r
1531 mov r12, #0\r
1532 strb r12, [lr, #-0x1a]\r
1533\r
1534 mov lr, #0x40/8\r
1535 @ copy pal:\r
1536.fl_loopcpBGR444:\r
1537 ldmia r3!, {r1,r5,r6,r12}\r
1538 subs lr, lr, #1\r
1539 stmia r4!, {r1,r5,r6,r12}\r
1540 bne .fl_loopcpBGR444\r
1541\r
1542 @ shadowed pixels:\r
1543 mov r12, #0x0077\r
1544 orr r12,r12,#0x0700\r
1545 orr r12,r12,r12,lsl #16\r
1546 sub r3, r3, #0x40*2\r
1547 add r5, r4, #0x80*2\r
1548 mov lr, #0x40/4\r
1549.fl_loopcpBGR444_sh:\r
1550 ldmia r3!, {r1,r6}\r
1551 subs lr, lr, #1\r
1552 and r1, r12, r1, lsr #1\r
1553 and r6, r12, r6, lsr #1\r
1554 stmia r4!, {r1,r6}\r
1555 stmia r5!, {r1,r6}\r
1556 bne .fl_loopcpBGR444_sh\r
1557\r
1558 @ hilighted pixels:\r
1559 sub r3, r3, #0x40*2\r
1560 mov lr, #0x40/2\r
1561.fl_loopcpBGR444_hi:\r
1562 ldr r1, [r3], #4\r
1563 TileDoShHi2Pixels444 r1\r
1564 str r1, [r4], #4\r
1565 subs lr, lr, #1\r
1566 bne .fl_loopcpBGR444_hi\r
1567\r
1568 sub r3, r4, #0x40*3*2\r
e5fa9817 1569 mov r6, #1\r
cc68a136 1570\r
1571\r
1572.fl_noshBGR444:\r
e5fa9817 1573 ldr r12,=rendstatus\r
1574 eors r6, r6, #1 @ sh is 0\r
1575 ldr r12,[r12]\r
1576 mov lr, #0xff\r
283fec1b 1577 tstne r12,#PDRAW_ACC_SPRITES\r
e5fa9817 1578\r
ea8c405f 1579.if OVERRIDE_HIGHCOL\r
1580 ldr r1, =HighCol\r
e5fa9817 1581 movne lr, #0x3f\r
ea8c405f 1582 ldr r1, [r1]\r
1583 mov lr, lr, lsl #1\r
1584 add r1, r1, #8\r
1585.else\r
cc68a136 1586 ldr r1, =(HighCol+8)\r
e5fa9817 1587 movne lr, #0x3f\r
cc68a136 1588 mov lr, lr, lsl #1\r
ea8c405f 1589.endif\r
cc68a136 1590\r
1591.fl_loopBGR444:\r
cc68a136 1592 ldr r12, [r1], #4\r
1593 subs r2, r2, #1\r
1594\r
1595 and r4, lr, r12, lsl #1\r
1596 ldrh r4, [r3, r4]\r
1597 and r5, lr, r12, lsr #7\r
1598 ldrh r5, [r3, r5]\r
1599 and r6, lr, r12, lsr #15\r
1600 ldrh r6, [r3, r6]\r
e5fa9817 1601 and r12,lr, r12, lsr #23\r
1602 ldrh r12,[r3, r12] @ 1c.i.\r
cc68a136 1603 orr r4, r4, r5, lsl #16\r
e5fa9817 1604 orr r5, r6, r12,lsl #16\r
cc68a136 1605\r
1606 stmia r0!, {r4,r5}\r
1607 bne .fl_loopBGR444\r
1608\r
1609\r
1610 ldmfd sp!, {r4-r6,lr}\r
1611 bx lr\r
1612\r
1613\r
1614@ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r
1615\r
1616\r
1617@ hilights 2 pixels in RGB555/BGR555 format\r
1618.macro TileDoShHi2Pixels555 reg\r
1619 adds \reg, \reg, #0x40000000\r
1620 orrcs \reg, \reg, #0xf0000000\r
1621 mov \reg, \reg, ror #27\r
1622 adds \reg, \reg, #0x40000000\r
1623 orrcs \reg, \reg, #0xf0000000\r
1624 mov \reg, \reg, ror #26\r
1625 adds \reg, \reg, #0x40000000\r
1626 orrcs \reg, \reg, #0xf0000000\r
1627 mov \reg, \reg, ror #27\r
1628 adds \reg, \reg, #0x40000000\r
1629 orrcs \reg, \reg, #0xf0000000\r
1630 mov \reg, \reg, ror #27\r
1631 adds \reg, \reg, #0x40000000\r
1632 orrcs \reg, \reg, #0xf0000000\r
1633 mov \reg, \reg, ror #26\r
1634 adds \reg, \reg, #0x40000000\r
1635 orrcs \reg, \reg, #0xf0000000\r
1636 mov \reg, \reg, ror #27\r
1637.endm\r
1638\r
1639\r
1640@ Convert 0000bbb0 ggg0rrr0\r
1641@ to rrrrrggg gggbbbbb\r
1642\r
1643@ r2,r3,r9 - scratch, lr = 0x001c001c, r8 = 0x00030003\r
1644.macro convRGB565 reg\r
1645 and r2, lr, \reg,lsl #1\r
1646 and r9, r8, \reg,lsr #2\r
1647 orr r2, r2, r9 @ r2=red\r
1648 and r3, lr, \reg,lsr #7\r
1649 and r9, r8, \reg,lsr #10\r
1650 orr r3, r3, r9 @ r3=blue\r
1651 and \reg, \reg, lr, lsl #3\r
1652 orr \reg, \reg, \reg,lsl #3 @ green\r
1653 orr \reg, \reg, r2, lsl #11 @ add red back\r
1654 orr \reg, \reg, r3 @ add blue back\r
1655.endm\r
1656\r
ea8c405f 1657.global vidConvCpyRGB565\r
1658\r
cc68a136 1659vidConvCpyRGB565: @ void *to, void *from, int pixels\r
1660 stmfd sp!, {r4-r9,lr}\r
1661\r
2ec14aec 1662 mov r12, r2, lsr #3 @ repeats\r
cc68a136 1663 mov lr, #0x001c0000\r
1664 orr lr, lr, #0x01c @ lr == pattern 0x001c001c\r
1665 mov r8, #0x00030000\r
2ec14aec 1666 orr r8, r8, #0x003\r
cc68a136 1667\r
1668.loopRGB565:\r
1669 ldmia r1!, {r4-r7}\r
1670 subs r12, r12, #1\r
1671 convRGB565 r4\r
1672 str r4, [r0], #4\r
1673 convRGB565 r5\r
1674 str r5, [r0], #4\r
1675 convRGB565 r6\r
1676 str r6, [r0], #4\r
1677 convRGB565 r7\r
1678 str r7, [r0], #4\r
1679\r
1680 bgt .loopRGB565\r
1681\r
1682 ldmfd sp!, {r4-r9,lr}\r
1683 bx lr\r
1684\r
1685\r
1686\r
1687.global FinalizeLineRGB555 @ int sh\r
1688\r
1689FinalizeLineRGB555:\r
1690 stmfd sp!, {r4-r8,lr}\r
3d48f143 1691 ldr r8, =(Pico+0x22228) @ Pico.video\r
cc68a136 1692 ldr r4, =HighPal\r
1693\r
3d48f143 1694 ldrb r7, [r8, #-0x1a] @ 0x2220e ~ dirtyPal\r
cc68a136 1695 mov r6, r0\r
1696 mov r1, #0\r
1697 tst r7, r7\r
1698 beq .fl_noconvRGB555\r
3d48f143 1699 strb r1, [r8, #-0x1a]\r
1700 sub r1, r8, #0x128 @ r1=Pico.cram\r
cc68a136 1701 mov r0, r4\r
1702 mov r2, #0x40\r
1703 bl vidConvCpyRGB565\r
1704\r
1705.fl_noconvRGB555:\r
cc68a136 1706 mov r3, r4\r
1707 tst r6, r6\r
1708 beq .fl_noshRGB555\r
1709 tst r7, r7\r
1710 beq .fl_noshRGB555\r
1711\r
1712 @ shadowed pixels:\r
1713 mov r12, #0x008e\r
1714 orr r12,r12,#0x7300\r
1715 orr r12,r12,r12,lsl #16\r
1716 add r4, r3, #0x40*2\r
1717 add r5, r3, #0xc0*2\r
1718 mov lr, #0x40/4\r
1719.fl_loopcpRGB555_sh:\r
1720 ldmia r3!, {r1,r6}\r
1721 subs lr, lr, #1\r
1722 and r1, r12, r1, lsr #1\r
1723 and r6, r12, r6, lsr #1\r
1724 stmia r4!, {r1,r6}\r
1725 stmia r5!, {r1,r6}\r
1726 bne .fl_loopcpRGB555_sh\r
1727\r
1728 @ hilighted pixels:\r
1729 sub r3, r3, #0x40*2\r
1730 mov lr, #0x40/2\r
1731.fl_loopcpRGB555_hi:\r
1732 ldr r1, [r3], #4\r
1733 TileDoShHi2Pixels555 r1\r
1734 str r1, [r4], #4\r
1735 subs lr, lr, #1\r
1736 bne .fl_loopcpRGB555_hi\r
1737\r
1738 sub r3, r3, #0x40*2\r
e5fa9817 1739 mov r6, #1\r
cc68a136 1740\r
cc68a136 1741.fl_noshRGB555:\r
e5fa9817 1742 ldr r12,=rendstatus\r
1743 eors r6, r6, #1 @ sh is 0\r
1744 ldr r12,[r12]\r
1745 mov lr, #0xff\r
283fec1b 1746 tstne r12,#PDRAW_ACC_SPRITES\r
e5fa9817 1747 movne lr, #0x3f\r
1748\r
ea8c405f 1749.if OVERRIDE_HIGHCOL\r
1750 ldr r1, =HighCol\r
1751 ldr r0, =DrawLineDest\r
1752 ldr r1, [r1]\r
1753 ldr r0, [r0]\r
1754 add r1, r1, #8\r
1755.else\r
3d48f143 1756 ldr r0, =DrawLineDest\r
cc68a136 1757 ldr r1, =(HighCol+8)\r
3d48f143 1758 ldr r0, [r0]\r
ea8c405f 1759.endif\r
3d48f143 1760\r
1761 ldrb r12, [r8, #12]\r
cc68a136 1762 mov lr, lr, lsl #1\r
1763\r
3d48f143 1764 tst r12, #1\r
1765 movne r2, #320/8 @ len\r
1766 bne .fl_no32colRGB555\r
1767 ldr r4, =PicoOpt\r
1768 mov r2, #256/8\r
1769 ldr r4, [r4]\r
1770 tst r4, #0x4000\r
1771 bne .fl_32scale_RGB555\r
1772 tst r4, #0x0100\r
1773 addeq r0, r0, #32*2\r
1774\r
1775.fl_no32colRGB555:\r
cc68a136 1776\r
499a0be3 1777.if UNALIGNED_DRAWLINEDEST\r
1778 @ this is basically for Gizmondo, which has unaligned odd lines in the framebuffer\r
1779 tst r0, #2\r
1780 bne .fl_RGB555u\r
1781.endif\r
1782\r
1783.fl_loopRGB555:\r
cc68a136 1784 ldr r12, [r1], #4\r
1785 ldr r7, [r1], #4\r
1786\r
1787 and r4, lr, r12, lsl #1\r
1788 ldrh r4, [r3, r4]\r
1789 and r5, lr, r12, lsr #7\r
1790 ldrh r5, [r3, r5]\r
1791 and r6, lr, r12, lsr #15\r
1792 ldrh r6, [r3, r6]\r
1793 orr r4, r4, r5, lsl #16\r
1794\r
1795 and r5, lr, r12, lsr #23\r
1796 ldrh r5, [r3, r5]\r
1797 and r8, lr, r7, lsl #1\r
1798 ldrh r8, [r3, r8]\r
1799 orr r5, r6, r5, lsl #16\r
1800\r
1801 and r6, lr, r7, lsr #7\r
1802 ldrh r6, [r3, r6]\r
1803 and r12,lr, r7, lsr #15\r
1804 ldrh r12,[r3, r12]\r
499a0be3 1805 and r7, lr, r7, lsr #23\r
1806 ldrh r7, [r3, r7]\r
cc68a136 1807 orr r8, r8, r6, lsl #16\r
1808\r
cc68a136 1809 subs r2, r2, #1\r
499a0be3 1810 orr r12,r12, r7, lsl #16\r
cc68a136 1811\r
1812 stmia r0!, {r4,r5,r8,r12}\r
1813 bne .fl_loopRGB555\r
1814\r
3d48f143 1815 ldmfd sp!, {r4-r8,lr}\r
1816 bx lr\r
1817\r
1818\r
1819.fl_32scale_RGB555:\r
1820 stmfd sp!, {r9,r10}\r
1821 mov r9, #0x3900 @ f800 07e0 001f | e000 0780 001c | 3800 01e0 0007\r
1822 orr r9, r9, #0x00e7\r
1823\r
499a0be3 1824.if UNALIGNED_DRAWLINEDEST\r
1825 tst r0, #2\r
1826 bne .fl_32scale_RGB555u\r
1827.endif\r
1828\r
3d48f143 1829.fl_loop32scale_RGB555:\r
1830 ldr r12, [r1], #4\r
1831 ldr r7, [r1], #4\r
cc68a136 1832\r
3d48f143 1833 and r4, lr, r12,lsl #1\r
1834 ldrh r4, [r3, r4]\r
1835 and r5, lr, r12,lsr #7\r
1836 ldrh r5, [r3, r5]\r
1837 and r4, r4, r9, lsl #2\r
1838 orr r4, r4, r4, lsl #14 @ r4[31:16] = 1/4 pix_s 0\r
1839 and r5, r5, r9, lsl #2\r
1840 sub r6, r5, r5, lsr #2 @ r6 = 3/4 pix_s 1\r
1841 add r4, r4, r6, lsl #16 @ pix_d 0, 1\r
1842 and r6, lr, r12,lsr #15\r
1843 ldrh r6, [r3, r6]\r
1844 and r12,lr, r12,lsr #23\r
1845 ldrh r12,[r3, r12]\r
1846 and r6, r6, r9, lsl #2\r
1847 add r5, r5, r6\r
1848 mov r5, r5, lsr #1\r
1849 sub r6, r6, r6, lsr #2 @ r6 = 3/4 pix_s 2\r
1850 orr r5, r5, r6, lsl #16\r
1851\r
1852 and r6, lr, r7, lsl #1\r
1853 ldrh r6, [r3, r6]\r
1854 and r12,r12,r9, lsl #2\r
1855 add r5, r5, r12,lsl #14 @ pix_d 2, 3\r
1856 and r6, r6, r9, lsl #2\r
1857 orr r6, r12,r6, lsl #16 @ pix_d 4, 5\r
1858\r
1859 and r12,lr, r7, lsr #7\r
1860 ldrh r12,[r3, r12]\r
1861 and r10,lr, r7, lsr #15\r
1862 ldrh r10,[r3, r10]\r
1863 and r12,r12,r9, lsl #2\r
1864 sub r8, r12,r12,lsr #2 @ r8 = 3/4 pix_s 1\r
1865 add r8, r8, r6, lsr #18\r
1866 and r7, lr, r7, lsr #23\r
1867 ldrh r7, [r3, r7]\r
1868 and r10,r10,r9, lsl #2\r
1869 orr r8, r8, r10,lsl #15\r
1870 add r8, r8, r12,lsl #15 @ pix_d 6, 7\r
1871 sub r10,r10,r10,lsr #2 @ r10= 3/4 pix_s 2\r
1872 and r7, r7, r9, lsl #2\r
1873 add r10,r10,r7, lsr #2 @ += 1/4 pix_s 3\r
1874 orr r10,r10,r7, lsl #16 @ pix_d 8, 9\r
1875\r
1876 subs r2, r2, #1\r
1877\r
1878 stmia r0!, {r4,r5,r6,r8,r10}\r
1879 bne .fl_loop32scale_RGB555\r
1880\r
1881 ldmfd sp!, {r9,r10}\r
cc68a136 1882 ldmfd sp!, {r4-r8,lr}\r
3d48f143 1883 bx lr\r
1884\r
499a0be3 1885.if UNALIGNED_DRAWLINEDEST\r
1886 @ unaligned versions of loops\r
9839d126 1887 @ warning: starts drawing 2bytes before dst\r
499a0be3 1888\r
1889.fl_RGB555u:\r
9839d126 1890 sub r0, r0, #2 @ initial adjustment\r
1891 mov r8, #0\r
499a0be3 1892\r
1893.fl_loopRGB555u:\r
1894 ldr r12, [r1], #4\r
1895 ldr r7, [r1], #4\r
1896\r
9839d126 1897 and r6, lr, r12,lsl #1\r
1898 ldrh r6, [r3, r6]\r
1899 and r5, lr, r12,lsr #7\r
499a0be3 1900 ldrh r5, [r3, r5]\r
9839d126 1901 orr r4, r8, r6, lsl #16\r
499a0be3 1902\r
9839d126 1903 and r6, lr, r12,lsr #15\r
499a0be3 1904 ldrh r6, [r3, r6]\r
9839d126 1905 and r8, lr, r12,lsr #23\r
499a0be3 1906 ldrh r8, [r3, r8]\r
9839d126 1907 orr r5, r5, r6, lsl #16\r
499a0be3 1908\r
9839d126 1909 and r6, lr, r7, lsl #1\r
499a0be3 1910 ldrh r6, [r3, r6]\r
9839d126 1911 and r12,lr, r7, lsr #7\r
499a0be3 1912 ldrh r12,[r3, r12]\r
9839d126 1913 orr r6, r8, r6, lsl #16\r
1914\r
1915 and r8, lr, r7, lsr #15\r
499a0be3 1916 ldrh r8, [r3, r8]\r
9839d126 1917 and r7, lr, r7, lsr #23\r
499a0be3 1918\r
1919 subs r2, r2, #1\r
9839d126 1920 orr r12,r12,r8, lsl #16\r
1921 ldrh r8, [r3, r7]\r
1922\r
1923 stmia r0!, {r4,r5,r6,r12}\r
499a0be3 1924 bne .fl_loopRGB555u\r
1925\r
1926 strh r8, [r0], #2\r
1927\r
1928 ldmfd sp!, {r4-r8,lr}\r
1929 bx lr\r
1930\r
1931\r
1932.fl_32scale_RGB555u:\r
9839d126 1933 sub r0, r0, #2 @ initial adjustment\r
1934 mov r4, #0\r
499a0be3 1935\r
1936 @ r9 f800 07e0 001f | e000 0780 001c | 3800 01e0 0007\r
1937.fl_loop32scale_RGB555u:\r
1938 ldr r12, [r1], #4\r
1939 ldr r7, [r1], #4\r
1940\r
9839d126 1941 and r6, lr, r12,lsl #1\r
1942 ldrh r6, [r3, r6]\r
499a0be3 1943 and r5, lr, r12,lsr #7\r
1944 ldrh r5, [r3, r5]\r
9839d126 1945 and r6, r6, r9, lsl #2\r
1946 orr r4, r4, r6, lsl #16 @ r4 = pix_d -1, 0\r
499a0be3 1947\r
1948 and r5, r5, r9, lsl #2\r
9839d126 1949 sub r8, r5, r5, lsr #2 @ r8 = 3/4 pix_s 1\r
1950 add r6, r8, r6, lsr #2 @ r6 = (1/4 pix_s 0) + (3/4 pix_s 1)\r
1951 orr r5, r6, r5, lsl #15\r
499a0be3 1952\r
1953 and r6, lr, r12,lsr #15\r
1954 ldrh r6, [r3, r6]\r
1955 and r12,lr, r12,lsr #23\r
1956 ldrh r12,[r3, r12]\r
1957 and r6, r6, r9, lsl #2\r
9839d126 1958 add r5, r5, r6, lsl #15 @ r5 = pix_d 1, 2\r
499a0be3 1959\r
9839d126 1960 and r8, lr, r7, lsl #1\r
1961 ldrh r8, [r3, r8]\r
1962 and r10,lr, r7, lsr #7\r
1963 ldrh r10,[r3, r10]\r
499a0be3 1964 and r12,r12,r9, lsl #2\r
9839d126 1965 sub r6, r6, r6, lsr #2 @ r6 = 3/4 pix_s 2\r
1966 add r6, r6, r12,lsr #2\r
1967 orr r6, r6, r12,lsl #16 @ r6 = pix_d 3, 4\r
499a0be3 1968\r
9839d126 1969 and r8, r8, r9, lsl #2\r
1970 and r10,r10,r9, lsl #2\r
1971 sub r12,r10,r10,lsr #2 @ r12 = 3/4 pix_s 5\r
1972 orr r8, r8, r8, lsl #14\r
1973 add r8, r8, r12,lsl #16 @ r8 = pix_d 5, 6\r
1974 and r12,lr, r7, lsr #15\r
499a0be3 1975 ldrh r12,[r3, r12]\r
499a0be3 1976 and r7, lr, r7, lsr #23\r
1977 ldrh r7, [r3, r7]\r
9839d126 1978 and r12,r12,r9, lsl #2\r
1979 add r10,r10,r12\r
1980 mov r10,r10, lsr #1\r
1981 sub r12,r12,r12,lsr #2 @ r12 = 3/4 pix_s 6\r
1982 orr r10,r10,r12,lsl #16\r
499a0be3 1983 and r7, r7, r9, lsl #2\r
9839d126 1984 add r10,r10,r7, lsl #14 @ r10 = pix_d 7, 8\r
499a0be3 1985\r
1986 subs r2, r2, #1\r
1987\r
1988 stmia r0!, {r4,r5,r6,r8,r10}\r
9839d126 1989 mov r4, r7\r
499a0be3 1990 bne .fl_loop32scale_RGB555u\r
1991\r
9839d126 1992 strh r4, [r0], #2\r
1993\r
499a0be3 1994 ldmfd sp!, {r9,r10}\r
1995 ldmfd sp!, {r4-r8,lr}\r
1996 bx lr\r
1997\r
1998.endif @ UNALIGNED_DRAWLINEDEST\r
1999\r
cc68a136 2000\r
2001@ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r
2002\r
2003@ utility\r
2004.global blockcpy @ void *dst, void *src, size_t n\r
2005\r
2006blockcpy:\r
2007 stmfd sp!, {r4,r5}\r
2008 mov r2, r2, lsr #4\r
2009blockcpy_loop:\r
2010 ldmia r1!, {r3-r5,r12}\r
2011 subs r2, r2, #1\r
2012 stmia r0!, {r3-r5,r12}\r
2013 bne blockcpy_loop\r
2014 ldmfd sp!, {r4,r5}\r
2015 bx lr\r
2016\r
2017\r
2018.global blockcpy_or @ void *dst, void *src, size_t n, int pat\r
2019\r
2020blockcpy_or:\r
2021 stmfd sp!, {r4-r6}\r
2022 orr r3, r3, r3, lsl #8\r
2023 orr r3, r3, r3, lsl #16\r
2024 mov r2, r2, lsr #4\r
2025blockcpy_loop_or:\r
2026 ldmia r1!, {r4-r6,r12}\r
2027 subs r2, r2, #1\r
2028 orr r4, r4, r3\r
2029 orr r5, r5, r3\r
2030 orr r6, r6, r3\r
2031 orr r12,r12,r3\r
2032 stmia r0!, {r4-r6,r12}\r
2033 bne blockcpy_loop_or\r
2034 ldmfd sp!, {r4-r6}\r
2035 bx lr\r
2036\r
fast MegaDrive/MegaCD/32X emulator