3d48f143 |
1 | @ vim:filetype=armasm\r |
2 | \r |
cc68a136 |
3 | @ assembly "optimized" version of some funtions from draw.c\r |
4 | @ this is highly specialized, be careful if changing related C code!\r |
5 | \r |
6cadc2da |
6 | @ (c) Copyright 2007, Grazvydas "notaz" Ignotas\r |
cc68a136 |
7 | @ All Rights Reserved\r |
8 | \r |
9 | \r |
10 | .extern Pico\r |
11 | .extern PicoOpt\r |
12 | .extern HighCol\r |
13 | .extern Scanline\r |
14 | .extern HighSprZ\r |
15 | .extern rendstatus\r |
16 | .extern DrawLineDest\r |
cc68a136 |
17 | .extern DrawStripInterlace\r |
18 | \r |
19 | \r |
20 | @ helper\r |
21 | .macro TilePixel pat lsrr offs\r |
22 | .if !\lsrr\r |
23 | ands r4, \pat, r2\r |
24 | .else\r |
25 | ands r4, \pat, r2, lsr #\lsrr\r |
26 | .endif\r |
27 | orrne r4, r3, r4\r |
28 | strneb r4, [r1,#\offs]\r |
29 | .endm\r |
30 | \r |
31 | @ TileNorm (r1=pdest, r2=pixels8, r3=pal) r4: scratch, pat: register with helper pattern 0xf\r |
32 | .macro TileNorm pat\r |
33 | TilePixel \pat, 12, 0 @ #0x0000f000\r |
34 | TilePixel \pat, 8, 1 @ #0x00000f00\r |
35 | TilePixel \pat, 4, 2 @ #0x000000f0\r |
36 | TilePixel \pat, 0, 3 @ #0x0000000f\r |
37 | TilePixel \pat, 28, 4 @ #0xf0000000\r |
38 | TilePixel \pat, 24, 5 @ #0x0f000000\r |
39 | TilePixel \pat, 20, 6 @ #0x00f00000\r |
40 | TilePixel \pat, 16, 7 @ #0x000f0000\r |
41 | .endm\r |
42 | \r |
43 | @ TileFlip (r1=pdest, r2=pixels8, r3=pal) r4: scratch, pat: register with helper pattern 0xf\r |
44 | .macro TileFlip pat\r |
45 | TilePixel \pat, 16, 0 @ #0x000f0000\r |
46 | TilePixel \pat, 20, 1 @ #0x00f00000\r |
47 | TilePixel \pat, 24, 2 @ #0x0f000000\r |
48 | TilePixel \pat, 28, 3 @ #0xf0000000\r |
49 | TilePixel \pat, 0, 4 @ #0x0000000f\r |
50 | TilePixel \pat, 4, 5 @ #0x000000f0\r |
51 | TilePixel \pat, 8, 6 @ #0x00000f00\r |
52 | TilePixel \pat, 12, 7 @ #0x0000f000\r |
53 | .endm\r |
54 | \r |
55 | @ shadow/hilight mode\r |
56 | \r |
57 | @ this one is for hi priority layer\r |
58 | .macro TilePixelShHP lsrr offs\r |
59 | .if !\lsrr\r |
60 | ands r4, r12, r2\r |
61 | .else\r |
62 | ands r4, r12, r2, lsr #\lsrr\r |
63 | .endif\r |
64 | ldreqb r4, [r1,#\offs]\r |
65 | orrne r4, r3, r4\r |
66 | strneb r4, [r1,#\offs]\r |
67 | tsteq r4, #0x80\r |
68 | andeq r4, r4, #0x3f\r |
69 | streqb r4, [r1,#\offs]\r |
70 | .endm\r |
71 | \r |
7a7c6476 |
72 | @ TileNormShHP (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r12: register with helper pattern 0xf, touches r3 high bits\r |
cc68a136 |
73 | .macro TileNormShHP\r |
74 | TilePixelShHP 12, 0 @ #0x0000f000\r |
75 | TilePixelShHP 8, 1 @ #0x00000f00\r |
76 | TilePixelShHP 4, 2 @ #0x000000f0\r |
77 | TilePixelShHP 0, 3 @ #0x0000000f\r |
78 | TilePixelShHP 28, 4 @ #0xf0000000\r |
79 | TilePixelShHP 24, 5 @ #0x0f000000\r |
80 | TilePixelShHP 20, 6 @ #0x00f00000\r |
81 | TilePixelShHP 16, 7 @ #0x000f0000\r |
82 | .endm\r |
83 | \r |
7a7c6476 |
84 | @ TileFlipShHP (r1=pdest, r2=pixels8, r3=pal) r4: scratch, pat: register with helper pattern 0xf\r |
cc68a136 |
85 | .macro TileFlipShHP\r |
86 | TilePixelShHP 16, 0 @ #0x000f0000\r |
87 | TilePixelShHP 20, 1 @ #0x00f00000\r |
88 | TilePixelShHP 24, 2 @ #0x0f000000\r |
89 | TilePixelShHP 28, 3 @ #0xf0000000\r |
90 | TilePixelShHP 0, 4 @ #0x0000000f\r |
91 | TilePixelShHP 4, 5 @ #0x000000f0\r |
92 | TilePixelShHP 8, 6 @ #0x00000f00\r |
93 | TilePixelShHP 12, 7 @ #0x0000f000\r |
94 | .endm\r |
95 | \r |
96 | \r |
97 | @ TileSingleSh (r1=pdest, r2=pixels8, r3=pal) r4,r7: scratch, r0=sx; r12: helper pattern 0xf\r |
98 | .macro TileSingleSh\r |
99 | tst r0, #1 @ not aligned?\r |
100 | mov r7, #0x00c000\r |
101 | orr r7, r7, #0xc0\r |
102 | ldrneb r4, [r1]\r |
103 | ldreqh r4, [r1]\r |
104 | orr r4, r4, r7\r |
105 | strneb r4, [r1], #1\r |
106 | streqh r4, [r1], #2\r |
107 | ldrh r4, [r1]\r |
108 | orr r4, r4, r7\r |
109 | strh r4, [r1], #2\r |
110 | ldrh r4, [r1]\r |
111 | orr r4, r4, r7\r |
112 | strh r4, [r1], #2\r |
113 | ldrh r4, [r1]\r |
114 | orr r4, r4, r7\r |
115 | strh r4, [r1], #2\r |
116 | ldrneb r4, [r1]\r |
117 | orr r4, r4, r7\r |
118 | strneb r4, [r1], #1\r |
119 | .endm\r |
120 | \r |
121 | @ TileSingleHi (r1=pdest, r2=pixels8, r3=pal) r4,r7: scratch, r0=sx, r12: register with helper pattern 0xf\r |
122 | .macro TileSingleHi\r |
123 | tst r1, #1 @ not aligned?\r |
124 | mov r7, #0x008000\r |
125 | orr r7, r7, #0x80\r |
126 | ldrneb r4, [r1], #1\r |
127 | ldreqh r4, [r1], #2 @ 1ci\r |
128 | ldrh r12, [r1], #2\r |
129 | bic r4, r4, r7, lsr #1\r |
130 | orr r4, r4, r7\r |
131 | strneb r4, [r1, #-3]\r |
132 | streqh r4, [r1, #-4]\r |
133 | ldrh r4, [r1], #2\r |
134 | bic r12, r12, r7, lsr #1\r |
135 | orr r12, r12, r7\r |
136 | strh r12, [r1, #-4]\r |
137 | ldrh r12, [r1], #2\r |
138 | bic r4, r4, r7, lsr #1\r |
139 | orr r4, r4, r7\r |
140 | strh r4, [r1, #-4]\r |
141 | ldrneb r4, [r1]\r |
142 | bic r12, r12, r7, lsr #1\r |
143 | orr r12, r12, r7\r |
144 | strh r12, [r1, #-2]\r |
145 | bicne r4, r4, r7, lsr #1\r |
146 | orrne r4, r4, r7\r |
147 | strneb r4, [r1], #1\r |
148 | mov r12, #0xf\r |
149 | .endm\r |
150 | \r |
151 | .macro TileDoShGenPixel shift ofs\r |
152 | .if \shift\r |
153 | ands r4, r12, r2, lsr #\shift\r |
154 | .else\r |
155 | ands r4, r12, r2\r |
156 | .endif\r |
157 | beq 3f\r |
158 | cmp r4, #0xe\r |
159 | beq 2f\r |
160 | bgt 1f\r |
161 | orr r4, r3, r4\r |
162 | strb r4, [r1,#\ofs]\r |
163 | b 3f\r |
164 | 1:\r |
165 | ldrb r4, [r1,#\ofs] @ 2ci\r |
166 | orr r4, r4, #0xc0\r |
167 | strb r4, [r1,#\ofs]\r |
168 | b 3f\r |
169 | 2:\r |
170 | ldrb r4, [r1,#\ofs] @ 2ci\r |
171 | bic r4, r4, #0xc0\r |
172 | orr r4, r4, #0x80\r |
173 | strb r4, [r1,#\ofs]\r |
174 | 3:\r |
175 | .endm\r |
176 | \r |
177 | @ TileFlipSh (r1=pdest, r2=pixels8, r3=pal) r4,r7: scratch, r0=sx, r12: register with helper pattern 0xf\r |
178 | .macro TileFlipSh\r |
179 | TileDoShGenPixel 16, 0 @ #0x000f0000\r |
180 | TileDoShGenPixel 20, 1 @ #0x00f00000\r |
181 | TileDoShGenPixel 24, 2 @ #0x0f000000\r |
182 | TileDoShGenPixel 28, 3 @ #0xf0000000\r |
183 | TileDoShGenPixel 0, 4 @ #0x0000000f\r |
184 | TileDoShGenPixel 4, 5 @ #0x000000f0\r |
185 | TileDoShGenPixel 8, 6 @ #0x00000f00\r |
186 | TileDoShGenPixel 12, 7 @ #0x0000f000\r |
187 | .endm\r |
188 | \r |
189 | @ TileNormSh (r1=pdest, r2=pixels8, r3=pal) r4,r7: scratch, r0=sx, r12: register with helper pattern 0xf\r |
190 | .macro TileNormSh\r |
191 | TileDoShGenPixel 12, 0 @ #0x0000f000\r |
192 | TileDoShGenPixel 8, 1 @ #0x00000f00\r |
193 | TileDoShGenPixel 4, 2 @ #0x000000f0\r |
194 | TileDoShGenPixel 0, 3 @ #0x0000000f\r |
195 | TileDoShGenPixel 28, 4 @ #0xf0000000\r |
196 | TileDoShGenPixel 24, 5 @ #0x0f000000\r |
197 | TileDoShGenPixel 20, 6 @ #0x00f00000\r |
198 | TileDoShGenPixel 16, 7 @ #0x000f0000\r |
199 | .endm\r |
200 | \r |
201 | \r |
202 | @ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r |
203 | \r |
204 | @ struct TileStrip\r |
205 | @ {\r |
206 | @ int nametab; // 0x00\r |
207 | @ int line; // 0x04\r |
208 | @ int hscroll; // 0x08\r |
209 | @ int xmask; // 0x0C\r |
210 | @ int *hc; // 0x10 (pointer to cache buffer)\r |
211 | @ int cells; // 0x14\r |
212 | @ };\r |
213 | \r |
83c093a4 |
214 | @ void DrawLayer(int plane_sh, int *hcache, int cellskip, int maxcells);\r |
cc68a136 |
215 | \r |
83c093a4 |
216 | .global DrawLayer\r |
cc68a136 |
217 | \r |
218 | DrawLayer:\r |
219 | stmfd sp!, {r4-r11,lr}\r |
220 | \r |
221 | ldr r11, =(Pico+0x22228) @ Pico.video\r |
222 | mov r8, #1\r |
223 | \r |
83c093a4 |
224 | ldrb r7, [r11, #16] @ ??vv??hh\r |
cc68a136 |
225 | \r |
226 | mov r6, r1 @ hcache\r |
83c093a4 |
227 | orr r9, r3, r0, lsl #30\r |
228 | orr r9, r9, r2, lsl #8 @ r9=sh[31]|cellskip[15:8]|maxcells[7:0] (tmp)\r |
cc68a136 |
229 | \r |
230 | mov r1, r7, lsl #4\r |
231 | orr r1, r1, #0x00ff\r |
232 | \r |
233 | and r10, r7, #3\r |
234 | cmp r10, #1\r |
235 | biclt r1, r1, #0xfc00\r |
236 | biceq r1, r1, #0xfe00\r |
237 | bicgt r1, r1, #0xff00 @ r1=ymask=(height<<8)|0xff; ...; // Y Mask in pixels\r |
238 | \r |
239 | add r10, r10, #5\r |
240 | cmp r10, #7\r |
241 | subge r10, r10, #1 @ r10=shift[width] (5,6,6,7)\r |
242 | \r |
243 | @ calculate xmask:\r |
244 | mov r5, r8, lsl r10\r |
245 | sub r5, r5, #1 @ r5=xmask\r |
246 | \r |
247 | @ Find name table:\r |
83c093a4 |
248 | ands r0, r0, #1\r |
cc68a136 |
249 | ldreqb r12, [r11, #2]\r |
250 | ldrneb r12, [r11, #4]\r |
251 | \r |
252 | ldr r2, =Scanline @ trying to make good use of pipeline here\r |
253 | ldr lr, =(Pico+0x10000) @ lr=Pico.vram\r |
254 | \r |
255 | moveq r12, r12, lsl #10\r |
256 | movne r12, r12, lsl #13\r |
257 | and r12, r12, #(7<<13) @ r12=(ts->nametab<<1) (halfword compliant)\r |
258 | \r |
259 | ldrh r8, [r11, #12]\r |
260 | ldrb r7, [r11, #11]\r |
261 | ldr r2, [r2]\r |
262 | \r |
263 | mov r4, r8, lsr #8 @ pvid->reg[13]\r |
264 | mov r4, r4, lsl #10 @ htab=pvid->reg[13]<<9; (halfwords)\r |
265 | tst r7, #2\r |
266 | addne r4, r4, r2, lsl #2 @ htab+=Scanline<<1; // Offset by line\r |
267 | tst r7, #1\r |
268 | biceq r4, r4, #0x1f @ htab&=~0xf; // Offset by tile\r |
269 | add r4, r4, r0, lsl #1 @ htab+=plane\r |
270 | bic r4, r4, #0x00ff0000 @ just in case\r |
271 | ldrh r3, [lr, r4] @ r3=hscroll\r |
272 | \r |
273 | tst r7, #4\r |
274 | bne .DrawStrip_vsscroll\r |
275 | \r |
276 | @ Get vertical scroll value:\r |
277 | add r7, lr, #0x012000\r |
278 | add r7, r7, #0x000180 @ r7=Pico.vsram (Pico+0x22180)\r |
279 | ldr r7, [r7]\r |
280 | \r |
281 | tst r8, #2\r |
282 | tstne r8, #4\r |
283 | bne .DrawStrip_interlace\r |
284 | \r |
285 | tst r0, r0\r |
286 | movne r7, r7, lsr #16\r |
287 | \r |
288 | @ Find the line in the name table\r |
289 | add r2, r2, r7\r |
290 | and r2, r2, r1\r |
291 | mov r4, r2, lsr #3\r |
292 | add r10, r10, #1 @ shift[width]++\r |
293 | add r12, r12, r4, lsl r10 @ nametab+=(ts.line>>3)<<shift[width];\r |
294 | \r |
295 | @ ldmia r0, {r1,r2,r3,r5,r6,r9} @ r2=line, r3=ts->hscroll, r5=ts->xmask, r6=ts->hc, r9=ts->cells\r |
cc68a136 |
296 | \r |
297 | and r10,r2, #7\r |
298 | mov r10,r10, lsl #1 @ r10=ty=(ts->line&7)<<1;\r |
299 | orr r10,r10, r9, lsl #24\r |
300 | \r |
301 | rsb r8, r3, #0\r |
302 | mov r8, r8, lsr #3 @ r8=tilex=(-ts->hscroll)>>3\r |
303 | \r |
304 | sub r1, r3, #1\r |
305 | and r1, r1, #7\r |
306 | add r7, r1, #1 @ r7=dx=((ts->hscroll-1)&7)+1\r |
307 | \r |
308 | tst r9, #1<<31\r |
309 | mov r3, #0\r |
740da8c6 |
310 | orrne r10,r10, #1<<23 @ r10=(cells<<24|sh<<23|hi_not_empty<<22|had_output<<21|ty)\r |
cc68a136 |
311 | movne r3, #0x40 @ default to shadowed pal on sh mode\r |
312 | \r |
cc68a136 |
313 | cmp r7, #8\r |
314 | addne r10,r10, #0x01000000 @ we will loop cells+1 times if there is scroll\r |
315 | \r |
83c093a4 |
316 | and r9, r9, #0xff00\r |
317 | add r8, r8, r9, lsr #8 @ tilex+=cellskip\r |
318 | add r7, r7, r9, lsr #5 @ dx+=cellskip<<3;\r |
319 | sub r10,r10,r9, lsl #16 @ cells-=cellskip\r |
320 | mvn r9, #0 @ r9=prevcode=-1\r |
321 | \r |
cc68a136 |
322 | @ cache some stuff to avoid mem access\r |
323 | ldr r11,=HighCol\r |
324 | mov r0, #0xf\r |
325 | add r1, r11, r7 @ r1=pdest\r |
326 | \r |
327 | \r |
328 | @ r4 & r7 are scratch in this loop\r |
329 | .dsloop_subr1:\r |
330 | sub r1, r1, #8\r |
331 | .dsloop: @ 40-41 times\r |
332 | subs r10,r10, #0x01000000\r |
333 | bmi .dsloop_exit\r |
334 | \r |
335 | .dsloop_enter:\r |
336 | and r7, r5, r8\r |
337 | add r7, lr, r7, lsl #1 @ Pico.vram+((tilex&ts->xmask) as halfwords)\r |
338 | ldrh r7, [r7, r12] @ r7=code (int, but from unsigned, no sign extend)\r |
339 | \r |
340 | add r1, r1, #8\r |
341 | add r8, r8, #1\r |
342 | \r |
343 | tst r7, #0x8000\r |
344 | bne .DrawStrip_hiprio\r |
345 | \r |
346 | cmp r7, r9\r |
347 | beq .DrawStrip_samecode @ we know stuff about this tile already\r |
348 | \r |
349 | mov r9, r7 @ remember code\r |
740da8c6 |
350 | orr r10, r10, #1<<21 @ seen non hi-prio tile\r |
cc68a136 |
351 | \r |
352 | movs r2, r9, lsl #20 @ if (code&0x1000)\r |
353 | mov r2, r2, lsl #1\r |
354 | add r2, r2, r10, lsl #17\r |
355 | mov r2, r2, lsr #17\r |
356 | eorcs r2, r2, #0x0e @ if (code&0x1000) addr^=0xe;\r |
357 | \r |
358 | ldr r2, [lr, r2, lsl #1] @ pack=*(unsigned int *)(Pico.vram+addr); // Get 8 pixels\r |
359 | \r |
360 | bic r7, r3, #0x3f\r |
361 | and r3, r9, #0x6000\r |
362 | add r3, r7, r3, lsr #9 @ r3=pal=((code&0x6000)>>9);\r |
363 | \r |
364 | .DrawStrip_samecode:\r |
365 | tst r2, r2\r |
366 | beq .dsloop @ tileline blank\r |
367 | \r |
368 | cmp r2, r2, ror #4\r |
369 | beq .DrawStrip_SingleColor @ tileline singlecolor \r |
370 | \r |
371 | tst r9, #0x0800\r |
7a7c6476 |
372 | bne .DrawStrip_TileFlip\r |
cc68a136 |
373 | \r |
374 | @ (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r0: helper pattern\r |
cc68a136 |
375 | .DrawStrip_TileNorm:\r |
376 | TileNorm r0\r |
377 | b .dsloop\r |
378 | \r |
7a7c6476 |
379 | .DrawStrip_TileFlip:\r |
380 | TileFlip r0\r |
381 | b .dsloop\r |
382 | \r |
cc68a136 |
383 | .DrawStrip_SingleColor:\r |
384 | and r4, r2, #0xf\r |
385 | orr r4, r3, r4\r |
386 | orr r4, r4, r4, lsl #8\r |
387 | tst r1, #1 @ not aligned?\r |
388 | strneb r4, [r1], #1\r |
389 | streqh r4, [r1], #2\r |
390 | strh r4, [r1], #2\r |
391 | strh r4, [r1], #2\r |
392 | strh r4, [r1], #2\r |
393 | strneb r4, [r1], #1 @ have a remaining unaligned pixel?\r |
394 | b .dsloop_subr1\r |
395 | \r |
cc68a136 |
396 | .DrawStrip_hiprio_maybempt:\r |
397 | cmp r7, r9\r |
398 | beq .dsloop @ must've been empty, otherwise we wouldn't get here\r |
399 | movs r2, r7, lsl #20 @ if (code&0x1000)\r |
400 | mov r2, r2, lsl #1\r |
401 | add r2, r2, r10, lsl #17\r |
402 | mov r2, r2, lsr #17\r |
403 | eorcs r2, r2, #0x0e @ if (code&0x1000) addr^=0xe;\r |
404 | ldr r2, [lr, r2, lsl #1] @ pack=*(unsigned int *)(Pico.vram+addr); // Get 8 pixels\r |
405 | mov r9, r7 @ remember code\r |
406 | tst r2, r2\r |
740da8c6 |
407 | beq .dsloop\r |
408 | orr r10, r10, #1<<22\r |
409 | \r |
410 | .DrawStrip_hiprio:\r |
411 | tst r10, #0x00c00000\r |
412 | beq .DrawStrip_hiprio_maybempt\r |
413 | sub r0, r1, r11\r |
414 | orr r7, r7, r0, lsl #16\r |
415 | orr r7, r7, r10, lsl #25 @ (ty<<25)\r |
416 | tst r7, #0x1000\r |
417 | eorne r7, r7, #7<<26 @ if(code&0x1000) cval^=7<<26;\r |
418 | str r7, [r6], #4 @ cache hi priority tile\r |
419 | mov r0, #0xf\r |
cc68a136 |
420 | b .dsloop\r |
421 | \r |
422 | .dsloop_exit:\r |
740da8c6 |
423 | tst r10, #1<<21 @ seen non hi-prio tile\r |
424 | ldreq r1, =rendstatus\r |
cc68a136 |
425 | mov r0, #0\r |
740da8c6 |
426 | ldreq r2, [r1]\r |
cc68a136 |
427 | str r0, [r6] @ terminate the cache list\r |
740da8c6 |
428 | orreq r2, r2, #0x40 @ had a layer with all hi-prio tiles\r |
429 | streq r2, [r1]\r |
cc68a136 |
430 | \r |
431 | ldmfd sp!, {r4-r11,lr}\r |
432 | bx lr\r |
433 | \r |
6d7acf9e |
434 | @ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r |
cc68a136 |
435 | \r |
436 | .DrawStrip_vsscroll:\r |
6d7acf9e |
437 | rsb r8, r3, #0\r |
438 | mov r8, r8, lsr #3 @ r8=tilex=(-ts->hscroll)>>3\r |
439 | bic r8, r8, #0xff000000\r |
83c093a4 |
440 | orr r8, r8, r5, lsl #25 @ r8=(xmask[31:25]|had_output[24]|tilex[23:0])\r |
cc68a136 |
441 | \r |
6d7acf9e |
442 | ldr r4, =Scanline\r |
443 | orr r5, r1, r10, lsl #24\r |
444 | ldr r4, [r4]\r |
445 | sub r1, r3, #1\r |
446 | orr r5, r5, r4, lsl #16 @ r5=(shift_width[31:24]|scanline[23:16]|ymask[15:0])\r |
447 | and r1, r1, #7\r |
448 | add r7, r1, #1 @ r7=dx=((ts->hscroll-1)&7)+1\r |
cc68a136 |
449 | \r |
6d7acf9e |
450 | mov r10,r9, lsl #16\r |
83c093a4 |
451 | tst r0, #1\r |
6d7acf9e |
452 | orrne r10,r10, #0x8000\r |
453 | tst r9, #1<<31\r |
454 | mov r3, #0\r |
455 | orr r10,r10, #0xff000000 @ will be adjusted on entering loop\r |
83c093a4 |
456 | 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 |
457 | movne r3, #0x40 @ default to shadowed pal on sh mode\r |
cc68a136 |
458 | \r |
83c093a4 |
459 | cmp r7, #8\r |
460 | subne r10,r10, #0x01000000 @ have hscroll, start with negative cell\r |
461 | \r |
462 | and r9, r9, #0xff00\r |
463 | add r8, r8, r9, lsr #8 @ tilex+=cellskip\r |
464 | add r7, r7, r9, lsr #5 @ dx+=cellskip<<3;\r |
465 | add r10,r10,r9, lsl #16 @ cell+=cellskip\r |
466 | mvn r9, #0 @ r9=prevcode=-1\r |
cc68a136 |
467 | \r |
6d7acf9e |
468 | @ cache some stuff to avoid mem access\r |
469 | ldr r11,=HighCol\r |
470 | mov r0, #0xf\r |
471 | add r1, r11, r7 @ r1=pdest\r |
472 | \r |
6d7acf9e |
473 | \r |
474 | @ r4 & r7 are scratch in this loop\r |
475 | .dsloop_vs_subr1:\r |
476 | sub r1, r1, #8\r |
477 | .dsloop_vs: @ 40-41 times\r |
478 | add r10,r10, #0x01000000\r |
479 | and r4, r10, #0x003f0000\r |
480 | cmp r4, r10, asr #8\r |
740da8c6 |
481 | ble .dsloop_vs_exit\r |
6d7acf9e |
482 | \r |
483 | @ calc offset and read tileline code to r7, also calc ty\r |
484 | add r7, lr, #0x012000\r |
485 | add r7, r7, #0x000180 @ r7=Pico.vsram (Pico+0x22180)\r |
486 | add r7, r7, r10,asr #23 @ vsram + ((cell&~1)<<1)\r |
487 | bic r7, r7, #3\r |
488 | tst r10,#0x8000 @ plane1?\r |
489 | addne r7, r7, #2\r |
490 | ldrh r7, [r7] @ r7=vscroll\r |
491 | \r |
492 | bic r10,r10,#0xff @ clear old ty\r |
493 | and r4, r5, #0xff0000\r |
494 | add r4, r4, r7, lsl #16\r |
495 | and r4, r4, r5, lsl #16 @ r4=line<<16\r |
496 | and r7, r4, #0x70000\r |
497 | orr r10,r10,r7, lsr #15 @ new ty\r |
498 | \r |
499 | mov r4, r4, lsr #19\r |
500 | mov r7, r5, lsr #24\r |
501 | mov r4, r4, lsl r7 @ nametabadd\r |
502 | \r |
503 | and r7, r8, r8, lsr #25\r |
504 | add r7, lr, r7, lsl #1 @ Pico.vram+((tilex&ts->xmask) as halfwords)\r |
505 | add r7, r7, r4, lsl #1\r |
506 | ldrh r7, [r7, r12] @ r7=code (int, but from unsigned, no sign extend)\r |
507 | \r |
508 | add r1, r1, #8\r |
509 | add r8, r8, #1\r |
510 | \r |
511 | tst r7, #0x8000\r |
512 | bne .DrawStrip_vs_hiprio\r |
513 | \r |
514 | cmp r7, r9\r |
515 | beq .DrawStrip_vs_samecode @ we know stuff about this tile already\r |
516 | \r |
517 | mov r9, r7 @ remember code\r |
740da8c6 |
518 | orr r8, r8, #1<<24 @ seen non hi-prio tile\r |
6d7acf9e |
519 | \r |
520 | movs r2, r9, lsl #20 @ if (code&0x1000)\r |
521 | mov r2, r2, lsl #1\r |
522 | add r2, r2, r10, lsl #17\r |
523 | mov r2, r2, lsr #17\r |
524 | eorcs r2, r2, #0x0e @ if (code&0x1000) addr^=0xe;\r |
525 | \r |
526 | ldr r2, [lr, r2, lsl #1] @ pack=*(unsigned int *)(Pico.vram+addr); // Get 8 pixels\r |
527 | \r |
528 | bic r7, r3, #0x3f\r |
529 | and r3, r9, #0x6000\r |
530 | add r3, r7, r3, lsr #9 @ r3=pal=((code&0x6000)>>9);\r |
531 | \r |
532 | .DrawStrip_vs_samecode:\r |
533 | tst r2, r2\r |
534 | beq .dsloop_vs @ tileline blank\r |
535 | \r |
536 | cmp r2, r2, ror #4\r |
537 | beq .DrawStrip_vs_SingleColor @ tileline singlecolor \r |
538 | \r |
539 | tst r9, #0x0800\r |
7a7c6476 |
540 | bne .DrawStrip_vs_TileFlip\r |
6d7acf9e |
541 | \r |
542 | @ (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r0: helper pattern\r |
6d7acf9e |
543 | .DrawStrip_vs_TileNorm:\r |
544 | TileNorm r0\r |
545 | b .dsloop_vs\r |
546 | \r |
7a7c6476 |
547 | .DrawStrip_vs_TileFlip:\r |
548 | TileFlip r0\r |
549 | b .dsloop_vs\r |
550 | \r |
6d7acf9e |
551 | .DrawStrip_vs_SingleColor:\r |
552 | and r4, r2, #0xf\r |
553 | orr r4, r3, r4\r |
554 | orr r4, r4, r4, lsl #8\r |
555 | tst r1, #1 @ not aligned?\r |
556 | strneb r4, [r1], #1\r |
557 | streqh r4, [r1], #2\r |
558 | strh r4, [r1], #2\r |
559 | strh r4, [r1], #2\r |
560 | strh r4, [r1], #2\r |
561 | strneb r4, [r1], #1 @ have a remaining unaligned pixel?\r |
562 | b .dsloop_vs_subr1\r |
563 | \r |
564 | .DrawStrip_vs_hiprio:\r |
565 | tst r10, #0x00c00000\r |
566 | beq .DrawStrip_vs_hiprio_maybempt\r |
567 | sub r0, r1, r11\r |
568 | orr r7, r7, r0, lsl #16\r |
569 | orr r7, r7, r10, lsl #25 @ (ty<<25)\r |
570 | tst r7, #0x1000\r |
571 | eorne r7, r7, #7<<26 @ if(code&0x1000) cval^=7<<26;\r |
572 | str r7, [r6], #4 @ cache hi priority tile\r |
573 | mov r0, #0xf\r |
574 | b .dsloop_vs\r |
575 | \r |
576 | .DrawStrip_vs_hiprio_maybempt:\r |
577 | cmp r7, r9\r |
578 | beq .dsloop_vs @ must've been empty, otherwise we wouldn't get here\r |
579 | movs r2, r7, lsl #20 @ if (code&0x1000)\r |
580 | mov r2, r2, lsl #1\r |
581 | add r2, r2, r10, lsl #17\r |
582 | mov r2, r2, lsr #17\r |
583 | eorcs r2, r2, #0x0e @ if (code&0x1000) addr^=0xe;\r |
584 | ldr r2, [lr, r2, lsl #1] @ pack=*(unsigned int *)(Pico.vram+addr); // Get 8 pixels\r |
585 | mov r9, r7 @ remember code\r |
586 | tst r2, r2\r |
587 | orrne r10, r10, #1<<22\r |
588 | bne .DrawStrip_vs_hiprio\r |
589 | b .dsloop_vs\r |
590 | \r |
740da8c6 |
591 | .dsloop_vs_exit:\r |
592 | tst r8, #1<<24 @ seen non hi-prio tile\r |
593 | ldreq r1, =rendstatus\r |
594 | mov r0, #0\r |
595 | ldreq r2, [r1]\r |
596 | str r0, [r6] @ terminate the cache list\r |
597 | orreq r2, r2, #0x40 @ had a layer with all hi-prio tiles\r |
598 | streq r2, [r1]\r |
599 | \r |
600 | ldmfd sp!, {r4-r11,lr}\r |
601 | bx lr\r |
602 | \r |
6d7acf9e |
603 | \r |
604 | @ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r |
cc68a136 |
605 | \r |
606 | @ interlace mode 2? Sonic 2?\r |
607 | .DrawStrip_interlace:\r |
608 | tst r0, r0\r |
609 | moveq r7, r7, lsl #21\r |
610 | movne r7, r7, lsl #5\r |
611 | \r |
612 | @ Find the line in the name table\r |
613 | add r2, r7, r2, lsl #22 @ r2=(vscroll+(Scanline<<1))<<21 (11 bits);\r |
614 | orr r1, r1, #0x80000000\r |
615 | and r2, r2, r1, ror #10 @ &((ymask<<1)|1)<<21;\r |
616 | mov r2, r2, lsr #21\r |
617 | mov r4, r2, lsr #4\r |
618 | mov r12, r12, lsr #1 @ halfwords\r |
619 | add r0, r12, r4, lsl r10 @ nametab+=(ts.line>>4)<<shift[width];\r |
620 | and r9, r9, #0xff\r |
621 | \r |
622 | sub sp, sp, #6*4\r |
623 | stmia sp, {r0,r2,r3,r5,r6,r9}\r |
624 | \r |
625 | mov r0, sp\r |
626 | bl DrawStripInterlace @ struct TileStrip *ts\r |
627 | \r |
628 | add sp, sp, #6*4\r |
629 | ldmfd sp!, {r4-r11,lr}\r |
630 | bx lr\r |
631 | \r |
632 | .pool\r |
633 | \r |
634 | @ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r |
635 | \r |
636 | \r |
637 | .global BackFill @ int reg7, int sh\r |
638 | \r |
639 | BackFill:\r |
640 | stmfd sp!, {r4-r9,lr}\r |
641 | \r |
642 | ldr lr, =(HighCol+8)\r |
643 | \r |
644 | mov r0, r0, lsl #26\r |
645 | mov r0, r0, lsr #26\r |
646 | orr r0, r0, r1, lsl #6\r |
647 | orr r0, r0, r0, lsl #8\r |
648 | orr r0, r0, r0, lsl #16\r |
649 | \r |
650 | mov r1, r0\r |
651 | mov r2, r0\r |
652 | mov r3, r0\r |
653 | mov r4, r0\r |
654 | mov r5, r0\r |
655 | mov r6, r0\r |
656 | mov r7, r0\r |
657 | \r |
658 | @ go go go!\r |
659 | stmia lr!, {r0-r7} @ 10*8*4\r |
660 | stmia lr!, {r0-r7}\r |
661 | stmia lr!, {r0-r7}\r |
662 | stmia lr!, {r0-r7}\r |
663 | stmia lr!, {r0-r7}\r |
664 | stmia lr!, {r0-r7}\r |
665 | stmia lr!, {r0-r7}\r |
666 | stmia lr!, {r0-r7}\r |
667 | stmia lr!, {r0-r7}\r |
668 | stmia lr!, {r0-r7}\r |
669 | \r |
670 | ldmfd sp!, {r4-r9,r12}\r |
671 | bx r12\r |
672 | \r |
673 | \r |
674 | @ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r |
675 | \r |
676 | \r |
7a7c6476 |
677 | .global DrawTilesFromCache @ int *hc, int sh, int rlim\r |
cc68a136 |
678 | \r |
679 | DrawTilesFromCache:\r |
680 | stmfd sp!, {r4-r8,r11,lr}\r |
681 | \r |
cc68a136 |
682 | @ cache some stuff to avoid mem access\r |
683 | ldr r11,=HighCol\r |
684 | ldr lr, =(Pico+0x10000) @ lr=Pico.vram\r |
685 | mov r12,#0xf\r |
686 | \r |
740da8c6 |
687 | mvn r5, #0 @ r5=prevcode=-1\r |
7a7c6476 |
688 | ands r8, r1, #1\r |
689 | orr r8, r8, r2, lsl #1\r |
740da8c6 |
690 | bne .dtfc_check_rendflags\r |
691 | \r |
cc68a136 |
692 | @ scratch: r4, r7\r |
693 | .dtfc_loop:\r |
694 | ldr r6, [r0], #4 @ read code\r |
695 | movs r1, r6, lsr #16 @ r1=dx;\r |
696 | ldmeqfd sp!, {r4-r8,r11,pc} @ dx is never zero, this must be a terminator, return\r |
7a7c6476 |
697 | bic r4, r1, #0xfe00\r |
698 | add r1, r11, r4 @ r1=pdest\r |
cc68a136 |
699 | \r |
700 | mov r7, r6, lsl #16\r |
701 | cmp r5, r7, lsr #16\r |
702 | beq .dtfc_samecode @ if (code==prevcode)\r |
703 | \r |
704 | mov r5, r7, lsr #16\r |
705 | \r |
706 | mov r2, r5, lsl #21\r |
707 | mov r2, r2, lsr #17 @ r2=addr=(code&0x7ff)<<4;\r |
708 | add r2, r2, r6, lsr #25 @ addr+=ty\r |
709 | \r |
710 | and r3, r5, #0x6000\r |
711 | mov r3, r3, lsr #9 @ r3=pal=((code&0x6000)>>9);\r |
712 | \r |
713 | ldr r2, [lr, r2, lsl #1] @ pack=*(unsigned int *)(Pico.vram+addr); // Get 8 pixels\r |
714 | \r |
715 | .dtfc_samecode:\r |
7a7c6476 |
716 | rsbs r4, r4, r8, lsr #1\r |
717 | bmi .dtfc_cut_tile\r |
718 | \r |
719 | tst r8, #1\r |
cc68a136 |
720 | bne .dtfc_shadow\r |
721 | \r |
722 | tst r2, r2\r |
723 | beq .dtfc_loop\r |
724 | \r |
725 | cmp r2, r2, ror #4\r |
726 | beq .dtfc_SingleColor @ tileline singlecolor \r |
727 | \r |
728 | tst r5, #0x0800\r |
7a7c6476 |
729 | bne .dtfc_TileFlip\r |
cc68a136 |
730 | \r |
731 | @ (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r12: helper pattern\r |
cc68a136 |
732 | .dtfc_TileNorm:\r |
733 | TileNorm r12\r |
734 | b .dtfc_loop\r |
735 | \r |
7a7c6476 |
736 | .dtfc_TileFlip:\r |
737 | TileFlip r12\r |
738 | b .dtfc_loop\r |
739 | \r |
cc68a136 |
740 | .dtfc_SingleColor:\r |
741 | and r4, r2, #0xf\r |
742 | orr r4, r3, r4\r |
743 | orr r4, r4, r4, lsl #8\r |
744 | tst r1, #1 @ not aligned?\r |
745 | strneb r4, [r1], #1\r |
746 | streqh r4, [r1], #2\r |
747 | strh r4, [r1], #2\r |
748 | strh r4, [r1], #2\r |
749 | strh r4, [r1], #2\r |
750 | strneb r4, [r1], #1 @ have a remaining unaligned pixel?\r |
751 | b .dtfc_loop\r |
752 | \r |
753 | .dtfc_shadow:\r |
754 | tst r2, r2\r |
755 | beq .dtfc_shadow_blank\r |
756 | \r |
757 | cmp r2, r2, ror #4\r |
758 | beq .dtfc_SingleColor @ tileline singlecolor \r |
759 | \r |
760 | tst r5, #0x0800\r |
7a7c6476 |
761 | bne .dtfc_TileFlipShHP\r |
cc68a136 |
762 | \r |
763 | @ (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r12: helper pattern\r |
cc68a136 |
764 | .dtfc_TileNormShHP:\r |
765 | TileNormShHP\r |
766 | b .dtfc_loop\r |
767 | \r |
7a7c6476 |
768 | .dtfc_TileFlipShHP:\r |
769 | TileFlipShHP\r |
770 | b .dtfc_loop\r |
771 | \r |
cc68a136 |
772 | .dtfc_shadow_blank:\r |
773 | ldrb r4, [r1] @ 1ci\r |
774 | ldrb r12,[r1,#1]\r |
775 | tst r4, #0x80\r |
776 | andeq r4, r4,#0x3f\r |
777 | streqb r4, [r1]\r |
778 | tst r12,#0x80\r |
779 | ldrb r4, [r1,#2]\r |
780 | andeq r12,r12,#0x3f\r |
781 | streqb r12,[r1,#1]\r |
782 | tst r4, #0x80\r |
783 | ldrb r12,[r1,#3]\r |
784 | andeq r4, r4,#0x3f\r |
785 | streqb r4, [r1,#2]\r |
786 | tst r12,#0x80\r |
787 | ldrb r4, [r1,#4]\r |
788 | andeq r12,r12,#0x3f\r |
789 | streqb r12,[r1,#3]\r |
790 | tst r4, #0x80\r |
791 | ldrb r12,[r1,#5]\r |
792 | andeq r4, r4,#0x3f\r |
793 | streqb r4, [r1,#4]\r |
794 | tst r12,#0x80\r |
795 | ldrb r4, [r1,#6]\r |
796 | andeq r12,r12,#0x3f\r |
797 | streqb r12,[r1,#5]\r |
798 | tst r4, #0x80\r |
799 | ldrb r12,[r1,#7]\r |
800 | andeq r4, r4,#0x3f\r |
801 | streqb r4, [r1,#6]\r |
802 | tst r12,#0x80\r |
803 | andeq r12,r12,#0x3f\r |
804 | streqb r12,[r1,#7]\r |
805 | mov r12, #0xf\r |
806 | b .dtfc_loop\r |
807 | \r |
7a7c6476 |
808 | .dtfc_cut_tile:\r |
809 | add r4, r4, #7 @ 0-6\r |
810 | mov r4, r4, lsl #2\r |
811 | mov r12,#0xf<<28\r |
812 | mov r12,r12,asr r4\r |
813 | mov r2, r2, ror #16\r |
814 | tst r5, #0x0800 @ flipped?\r |
815 | mvnne r12,r12\r |
816 | and r2, r2, r12\r |
817 | mov r2, r2, ror #16\r |
818 | mov r12,#0xf\r |
819 | tst r8, #1\r |
820 | bne .dtfc_shadow\r |
821 | tst r2, r2\r |
822 | beq .dtfc_loop\r |
823 | tst r5, #0x0800\r |
824 | beq .dtfc_TileNorm\r |
825 | b .dtfc_TileFlip\r |
826 | \r |
740da8c6 |
827 | @ check if we have detected layer covered with hi-prio tiles:\r |
828 | .dtfc_check_rendflags:\r |
829 | ldr r1, =rendstatus\r |
830 | ldr r2, [r1]\r |
831 | tst r2, #0xc0\r |
832 | beq .dtfc_loop\r |
7a7c6476 |
833 | bic r8, r8, #1 @ sh/hi mode off\r |
740da8c6 |
834 | tst r2, #0x80\r |
835 | bne .dtfc_loop @ already processed\r |
836 | orr r2, r2, #0x80\r |
837 | str r2, [r1]\r |
838 | \r |
839 | add r1, r11,#8\r |
840 | mov r3, #320/4\r |
841 | mov r7, #0x80\r |
842 | orr r7, r7, r7, lsl #8\r |
843 | orr r7, r7, r7, lsl #16\r |
844 | mov r6, #0x3f\r |
845 | orr r6, r6, r6, lsl #8\r |
846 | orr r6, r6, r6, lsl #16\r |
847 | .dtfc_loop_shprep:\r |
848 | subs r3, r3, #1\r |
849 | bmi .dtfc_loop @ done\r |
850 | ldr r2, [r1]\r |
851 | tst r2, r7\r |
852 | andeq r2, r2, r6\r |
853 | streq r2, [r1], #4\r |
854 | beq .dtfc_loop_shprep\r |
855 | tst r2, #0x80000000\r |
856 | biceq r2, r2, #0xc0000000\r |
857 | tst r2, #0x00800000\r |
858 | biceq r2, r2, #0x00c00000\r |
859 | tst r2, #0x00008000\r |
860 | biceq r2, r2, #0x0000c000\r |
861 | tst r2, #0x00000080\r |
862 | biceq r2, r2, #0x000000c0\r |
863 | str r2, [r1], #4\r |
864 | b .dtfc_loop_shprep\r |
865 | \r |
cc68a136 |
866 | .pool\r |
867 | \r |
868 | @ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r |
869 | \r |
870 | \r |
871 | .global DrawSpritesFromCache @ int *hc, int sh\r |
872 | \r |
873 | DrawSpritesFromCache:\r |
874 | stmfd sp!, {r4-r11,lr}\r |
875 | \r |
876 | @ cache some stuff to avoid mem access\r |
877 | ldr r11,=HighCol\r |
878 | ldr lr, =(Pico+0x10000) @ lr=Pico.vram\r |
879 | mov r6, r1, lsl #31\r |
880 | orr r6, r6, #1<<30\r |
881 | mov r12,#0xf\r |
882 | \r |
883 | mov r10, r0\r |
884 | \r |
885 | .dsfc_loop:\r |
886 | ldr r9, [r10], #4 @ read code\r |
887 | bic r6, r6, #7 @ using pipeline\r |
888 | tst r9, r9\r |
889 | ldmeqfd sp!, {r4-r11,pc}\r |
890 | \r |
891 | mov r4, r9, lsl #28\r |
892 | orr r6, r6, r4, lsr #30\r |
893 | add r6, r6, #1 @ r6=s1cc???? ... ?????www (s=shadow/hilight, cc=pal, w=width)\r |
894 | \r |
895 | and r5, r9, #3\r |
896 | add r5, r5, #1 @ r5=delta\r |
897 | tst r9, #0x10000\r |
898 | rsbne r5, r5, #0 @ Flip X\r |
899 | mov r5, r5, lsl #4\r |
900 | \r |
901 | mov r2, r9, lsr #17\r |
902 | mov r8, r2, lsl #1 @ tile=((unsigned int)code>>17)<<1;\r |
903 | \r |
904 | and r3, r9, #0x30 @ r3=pal=(code&0x30);\r |
905 | \r |
906 | bic r6, r6, #3<<28\r |
907 | orr r6, r6, r3, lsl #24\r |
908 | \r |
909 | mov r0, r9, lsl #16\r |
910 | mov r0, r0, asr #22 @ sx=(code<<16)>>22\r |
911 | adds r0, r0, #0 @ set ZV\r |
912 | b .dsfc_inloop_enter\r |
913 | \r |
914 | @ scratch: r4, r7\r |
915 | .dsfc_inloop:\r |
916 | sub r6, r6, #1\r |
917 | tst r6, #7\r |
918 | beq .dsfc_loop\r |
919 | adds r0, r0, #8\r |
920 | add r8, r8, r5\r |
921 | \r |
922 | .dsfc_inloop_enter:\r |
923 | ble .dsfc_inloop\r |
924 | cmp r0, #328\r |
925 | bge .dsfc_loop\r |
926 | \r |
927 | mov r8, r8, lsl #17\r |
928 | mov r8, r8, lsr #17 @ tile&=0x7fff; // Clip tile address\r |
929 | \r |
930 | ldr r2, [lr, r8, lsl #1] @ pack=*(unsigned int *)(Pico.vram+tile); // Get 8 pixels\r |
931 | add r1, r11, r0 @ r1=pdest\r |
932 | tst r2, r2\r |
933 | beq .dsfc_inloop\r |
934 | \r |
935 | cmp r12, r6, lsr #28\r |
936 | beq .dsfc_shadow\r |
937 | \r |
938 | cmp r2, r2, ror #4\r |
939 | beq .dsfc_SingleColor @ tileline singlecolor \r |
940 | \r |
941 | tst r9, #0x10000\r |
7a7c6476 |
942 | bne .dsfc_TileFlip\r |
cc68a136 |
943 | \r |
944 | @ TileFlip (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r12: helper pattern\r |
cc68a136 |
945 | .dsfc_TileNorm:\r |
946 | TileNorm r12\r |
947 | b .dsfc_inloop\r |
948 | \r |
7a7c6476 |
949 | .dsfc_TileFlip:\r |
950 | TileFlip r12\r |
951 | b .dsfc_inloop\r |
952 | \r |
cc68a136 |
953 | .dsfc_SingleColor:\r |
954 | tst r0, #1 @ not aligned?\r |
955 | and r4, r2, #0xf\r |
956 | orr r4, r3, r4\r |
957 | orr r4, r4, r4, lsl #8\r |
958 | strneb r4, [r1], #1\r |
959 | streqh r4, [r1], #2\r |
960 | strh r4, [r1], #2\r |
961 | strh r4, [r1], #2\r |
962 | strh r4, [r1], #2\r |
963 | strneb r4, [r1], #1\r |
964 | b .dsfc_inloop\r |
965 | \r |
966 | .dsfc_shadow:\r |
967 | cmp r2, r2, ror #4\r |
968 | beq .dsfc_singlec_sh\r |
969 | \r |
970 | tst r9, #0x10000\r |
7a7c6476 |
971 | bne .dsfc_TileFlip_sh\r |
cc68a136 |
972 | \r |
973 | @ (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r12: helper pattern\r |
cc68a136 |
974 | .dsfc_TileNorm_sh:\r |
975 | TileNormSh\r |
976 | b .dsfc_inloop\r |
977 | \r |
7a7c6476 |
978 | .dsfc_TileFlip_sh:\r |
979 | TileFlipSh\r |
980 | b .dsfc_inloop\r |
981 | \r |
cc68a136 |
982 | .dsfc_singlec_sh:\r |
983 | cmp r2, #0xe0000000\r |
984 | bcc .dsfc_SingleColor @ normal singlecolor tileline (carry inverted in ARM)\r |
985 | tst r2, #0x10000000\r |
986 | bne .dsfc_sh_sh\r |
987 | TileSingleHi\r |
988 | b .dsfc_inloop\r |
989 | \r |
990 | .dsfc_sh_sh:\r |
991 | TileSingleSh\r |
992 | b .dsfc_inloop\r |
993 | \r |
994 | .pool\r |
995 | \r |
996 | @ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r |
997 | \r |
998 | @ + 0 : hhhhvvvv ab--hhvv yyyyyyyy yyyyyyyy // a: offscreen h, b: offs. v, h: horiz. size\r |
999 | @ + 4 : xxxxxxxx xxxxxxxx pccvhnnn nnnnnnnn // x: x coord + 8\r |
1000 | \r |
1001 | .global DrawSprite @ unsigned int *sprite, int **hc, int sh\r |
1002 | \r |
1003 | DrawSprite:\r |
1004 | stmfd sp!, {r4-r9,r11,lr}\r |
1005 | \r |
1006 | ldr r3, [r0] @ sprite[0]\r |
1007 | ldr r7, =Scanline\r |
1008 | mov r6, r3, lsr #28\r |
1009 | sub r6, r6, #1 @ r6=width-1 (inc later)\r |
1010 | mov r5, r3, lsr #24\r |
1011 | and r5, r5, #7 @ r5=height\r |
1012 | \r |
1013 | mov r4, r3, lsl #16 @ r4=sy<<16 (tmp)\r |
1014 | \r |
1015 | ldr r7, [r7]\r |
1016 | ldr r9, [r0, #4]\r |
1017 | sub r7, r7, r4, asr #16 @ r7=row=Scanline-sy\r |
1018 | \r |
1019 | tst r2, r2\r |
1020 | mov r2, r9, asr #16 @ r2=sx\r |
1021 | bic r9, r9, #0xfe000000\r |
1022 | orrne r9, r9, #1<<31 @ r9=code|(sh<<31)\r |
1023 | \r |
1024 | tst r9, #0x1000\r |
1025 | movne r4, r5, lsl #3\r |
1026 | subne r4, r4, #1\r |
1027 | subne r7, r4, r7 @ if (code&0x1000) row=(height<<3)-1-row; // Flip Y\r |
1028 | \r |
1029 | mov r8, r9, lsl #21\r |
1030 | mov r8, r8, lsr #21\r |
1031 | add r8, r8, r7, lsr #3 @ tile+=row>>3; // Tile number increases going down\r |
1032 | \r |
1033 | tst r9, #0x0800\r |
1034 | mlane r8, r5, r6, r8 @ if (code&0x0800) { tile+=delta*(width-1);\r |
1035 | rsbne r5, r5, #0 @ delta=-delta; } // r5=delta now\r |
1036 | \r |
1037 | mov r8, r8, lsl #4\r |
1038 | and r7, r7, #7\r |
1039 | add r8, r8, r7, lsl #1 @ tile+=(row&7)<<1; // Tile address\r |
1040 | \r |
1041 | tst r9, #0x8000\r |
1042 | bne .dspr_cache @ if(code&0x8000) // high priority - cache it\r |
1043 | \r |
1044 | @ cache some stuff to avoid mem access\r |
1045 | ldr r11,=HighCol\r |
1046 | ldr lr, =(Pico+0x10000) @ lr=Pico.vram\r |
1047 | mov r12,#0xf\r |
1048 | \r |
1049 | mov r5, r5, lsl #4 @ delta<<=4; // Delta of address\r |
1050 | and r4, r9, #0x6000\r |
1051 | orr r9, r9, r4, lsl #16\r |
1052 | orr r9, r9, #0x10000000 @ r9=scc1 ???? ... <code> (s=shadow/hilight, cc=pal)\r |
1053 | \r |
1054 | tst r9, #1<<31\r |
1055 | mov r3, r4, lsr #9 @ r3=pal=((code>>9)&0x30);\r |
1056 | orrne r3, r3, #0x40 @ shadow by default\r |
1057 | \r |
1058 | add r6, r6, #1 @ inc now\r |
1059 | adds r0, r2, #0 @ mov sx to r0 and set ZV flags\r |
1060 | b .dspr_loop_enter\r |
1061 | \r |
1062 | .dspr_loop:\r |
1063 | subs r6, r6, #1 @ width--\r |
1064 | ldmeqfd sp!, {r4-r9,r11,pc}@ return\r |
1065 | adds r0, r0, #8 @ sx+=8\r |
1066 | add r8, r8, r5 @ tile+=delta\r |
1067 | \r |
1068 | .dspr_loop_enter:\r |
1069 | ble .dspr_loop @ sx <= 0\r |
1070 | cmp r0, #328\r |
1071 | ldmgefd sp!, {r4-r9,r11,pc}@ return\r |
1072 | \r |
1073 | mov r8, r8, lsl #17\r |
1074 | mov r8, r8, lsr #17 @ tile&=0x7fff; // Clip tile address\r |
1075 | \r |
1076 | ldr r2, [lr, r8, lsl #1] @ pack=*(unsigned int *)(Pico.vram+addr); // Get 8 pixels\r |
1077 | add r1, r11, r0 @ r1=pdest\r |
1078 | tst r2, r2\r |
1079 | beq .dspr_loop\r |
1080 | \r |
1081 | cmp r12, r9, lsr #28\r |
1082 | beq .dspr_shadow\r |
1083 | \r |
1084 | cmp r2, r2, ror #4\r |
1085 | beq .dspr_SingleColor @ tileline singlecolor \r |
1086 | \r |
1087 | tst r9, #0x0800\r |
7a7c6476 |
1088 | bne .dspr_TileFlip\r |
cc68a136 |
1089 | \r |
1090 | @ (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r12: helper pattern\r |
cc68a136 |
1091 | @ scratch: r4, r7\r |
1092 | .dspr_TileNorm:\r |
1093 | TileNorm r12\r |
1094 | b .dspr_loop\r |
1095 | \r |
7a7c6476 |
1096 | .dspr_TileFlip:\r |
1097 | TileFlip r12\r |
1098 | b .dspr_loop\r |
1099 | \r |
cc68a136 |
1100 | .dspr_SingleColor:\r |
1101 | and r4, r2, #0xf\r |
1102 | orr r4, r3, r4\r |
1103 | orr r4, r4, r4, lsl #8\r |
1104 | tst r0, #1 @ not aligned?\r |
1105 | strneb r4, [r1], #1\r |
1106 | streqh r4, [r1], #2\r |
1107 | strh r4, [r1], #2\r |
1108 | strh r4, [r1], #2\r |
1109 | strh r4, [r1], #2\r |
1110 | strneb r4, [r1], #1\r |
1111 | b .dspr_loop\r |
1112 | \r |
1113 | .dspr_shadow:\r |
1114 | cmp r2, r2, ror #4\r |
1115 | beq .dspr_singlec_sh\r |
1116 | \r |
1117 | tst r9, #0x0800\r |
7a7c6476 |
1118 | bne .dspr_TileFlip_sh\r |
cc68a136 |
1119 | \r |
1120 | @ (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r12: helper pattern\r |
cc68a136 |
1121 | .dspr_TileNorm_sh:\r |
1122 | TileNormSh\r |
1123 | b .dspr_loop\r |
1124 | \r |
7a7c6476 |
1125 | .dspr_TileFlip_sh:\r |
1126 | TileFlipSh\r |
1127 | b .dspr_loop\r |
1128 | \r |
cc68a136 |
1129 | .dspr_singlec_sh:\r |
1130 | cmp r2, #0xe0000000\r |
1131 | bcc .dspr_SingleColor @ normal tileline\r |
1132 | tst r2, #0x10000000\r |
1133 | bne .dspr_sh_sh\r |
1134 | TileSingleHi\r |
1135 | b .dspr_loop\r |
1136 | \r |
1137 | .dspr_sh_sh:\r |
1138 | TileSingleSh\r |
1139 | b .dspr_loop\r |
1140 | \r |
1141 | \r |
1142 | .dspr_cache:\r |
1143 | @ *(*hc)++ = (tile<<16)|((code&0x0800)<<5)|((sx<<6)&0x0000ffc0)|((code>>9)&0x30)|((sprite[0]>>24)&0xf);\r |
1144 | mov r4, r8, lsl #16 @ tile\r |
1145 | tst r9, #0x0800\r |
1146 | orrne r4, r4, #0x10000 @ code&0x0800\r |
1147 | mov r2, r2, lsl #22\r |
1148 | orr r4, r4, r2, lsr #16 @ (sx<<6)&0x0000ffc0\r |
1149 | and r2, r9, #0x6000\r |
1150 | orr r4, r4, r2, lsr #9 @ (code>>9)&0x30\r |
1151 | mov r3, r3, lsl #12\r |
1152 | ldr r2, [r1]\r |
1153 | orr r4, r4, r3, lsr #28 @ (sprite[0]>>24)&0xf\r |
1154 | \r |
1155 | str r4, [r2], #4\r |
1156 | str r2, [r1]\r |
1157 | \r |
1158 | ldmfd sp!, {r4-r9,r11,lr}\r |
1159 | bx lr\r |
1160 | \r |
1161 | @ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r |
1162 | \r |
1163 | .global DrawWindow @ int tstart, int tend, int prio, int sh // int *hcache\r |
1164 | \r |
1165 | DrawWindow:\r |
1166 | stmfd sp!, {r4-r11,lr}\r |
1167 | \r |
1168 | ldr r11, =(Pico+0x22228) @ Pico.video\r |
1169 | ldr r10, =Scanline\r |
1170 | ldrb r12, [r11, #3] @ pvid->reg[3]\r |
1171 | \r |
1172 | ldr r10, [r10]\r |
1173 | ldr r4, [r11, #12]\r |
1174 | mov r5, r10, lsr #3\r |
1175 | and r10, r10, #7\r |
1176 | mov r10, r10, lsl #1 @ r10=ty\r |
1177 | \r |
1178 | mov r12, r12, lsl #10\r |
1179 | \r |
1180 | tst r4, #1 @ 40 cell mode?\r |
1181 | andne r12, r12, #0xf000 @ 0x3c<<10\r |
1182 | andeq r12, r12, #0xf800\r |
1183 | addne r12, r12, r5, lsl #7\r |
1184 | addeq r12, r12, r5, lsl #6 @ nametab\r |
1185 | add r12, r12, r0, lsl #2 @ +starttile\r |
1186 | \r |
1187 | ldr r6, =rendstatus\r |
1188 | ldr lr, =(Pico+0x10000) @ lr=Pico.vram\r |
1189 | ldrb r6, [r6]\r |
1190 | \r |
1191 | @ fetch the first code now\r |
1192 | ldrh r7, [lr, r12]\r |
1193 | \r |
1194 | ands r6, r6, #2 @ we care about bit 1 only\r |
1195 | orr r6, r6, r2\r |
1196 | bne .dw_no_sameprio\r |
1197 | \r |
1198 | cmp r2, r7, lsr #15\r |
1199 | ldmnefd sp!, {r4-r11,pc} @ assume that whole window uses same priority\r |
1200 | \r |
1201 | .dw_no_sameprio:\r |
1202 | orr r6, r6, r3, lsl #8 @ shadow mode\r |
1203 | \r |
1204 | sub r8, r1, r0\r |
1205 | mov r8, r8, lsl #1 @ cells\r |
1206 | \r |
1207 | mvn r9, #0 @ r9=prevcode=-1\r |
1208 | \r |
1209 | @ cache some stuff to avoid mem access\r |
1210 | ldr r11,=(HighCol+8)\r |
1211 | add r1, r11, r0, lsl #4 @ r1=pdest\r |
1212 | mov r0, #0xf\r |
1213 | b .dwloop_enter\r |
1214 | \r |
1215 | @ r4,r5 & r7 are scratch in this loop\r |
1216 | .dwloop:\r |
1217 | add r1, r1, #8\r |
1218 | .dwloop_nor1:\r |
1219 | add r12, r12, #2 @ halfwords\r |
1220 | ldrh r7, [lr, r12] @ r7=code (int, but from unsigned, no sign extend)\r |
1221 | subs r8, r8, #1\r |
1222 | beq .dwloop_end @ done\r |
1223 | \r |
1224 | eor r5, r6, r7, lsr #15\r |
1225 | tst r5, #1\r |
1226 | orrne r6, r6, #2 @ wrong pri\r |
1227 | bne .dwloop\r |
1228 | \r |
1229 | cmp r7, r9\r |
1230 | beq .dw_samecode @ we know stuff about this tile already\r |
1231 | \r |
1232 | .dwloop_enter:\r |
1233 | mov r9, r7 @ remember code\r |
1234 | \r |
1235 | movs r2, r9, lsl #20 @ if (code&0x1000)\r |
1236 | mov r2, r2, lsl #1\r |
1237 | add r2, r10, r2, lsr #17 @ r2=addr=(code&0x7ff)<<4; addr+=ty\r |
1238 | eorcs r2, r2, #0x0e @ if (code&0x1000) addr^=0xe;\r |
1239 | \r |
1240 | and r3, r9, #0x6000\r |
1241 | mov r3, r3, lsr #9 @ r3=pal=((code&0x6000)>>9);\r |
1242 | \r |
1243 | ldr r2, [lr, r2, lsl #1] @ pack=*(unsigned int *)(Pico.vram+addr); // Get 8 pixels\r |
1244 | \r |
1245 | .dw_samecode:\r |
1246 | tst r6, #0x100\r |
1247 | bne .dw_shadow\r |
1248 | .dw_shadow_done:\r |
1249 | tst r2, r2\r |
1250 | beq .dwloop @ tileline blank\r |
1251 | \r |
1252 | cmp r2, r2, ror #4\r |
1253 | beq .dw_SingleColor @ tileline singlecolor \r |
1254 | \r |
1255 | tst r9, #0x0800\r |
7a7c6476 |
1256 | bne .dw_TileFlip\r |
cc68a136 |
1257 | \r |
1258 | @ (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r0: helper pattern\r |
cc68a136 |
1259 | .dw_TileNorm:\r |
1260 | TileNorm r0\r |
1261 | b .dwloop\r |
1262 | \r |
7a7c6476 |
1263 | .dw_TileFlip:\r |
1264 | TileFlip r0\r |
1265 | b .dwloop\r |
1266 | \r |
cc68a136 |
1267 | .dw_SingleColor:\r |
1268 | and r4, r0, r2 @ #0x0000000f\r |
1269 | orr r4, r3, r4\r |
1270 | orr r4, r4, r4, lsl #8\r |
1271 | orr r4, r4, r4, lsl #16\r |
1272 | mov r5, r4\r |
1273 | stmia r1!, {r4,r5}\r |
1274 | b .dwloop_nor1 @ we incremeted r1 ourselves\r |
1275 | \r |
1276 | .dw_shadow:\r |
1277 | tst r6, #1 @ hi pri?\r |
1278 | orreq r3, r3, #0x40\r |
1279 | beq .dw_shadow_done\r |
1280 | ldr r4, [r1]\r |
1281 | tst r4, #0x00000080\r |
1282 | biceq r4, r4, #0x000000c0\r |
1283 | tst r4, #0x00008000\r |
1284 | biceq r4, r4, #0x0000c000\r |
1285 | tst r4, #0x00800000\r |
1286 | biceq r4, r4, #0x00c00000\r |
1287 | tst r4, #0x80000000\r |
1288 | biceq r4, r4, #0xc0000000\r |
1289 | str r4, [r1]\r |
1290 | ldr r4, [r1,#4]\r |
1291 | tst r4, #0x00000080\r |
1292 | biceq r4, r4, #0x000000c0\r |
1293 | tst r4, #0x00008000\r |
1294 | biceq r4, r4, #0x0000c000\r |
1295 | tst r4, #0x00800000\r |
1296 | biceq r4, r4, #0x00c00000\r |
1297 | tst r4, #0x80000000\r |
1298 | biceq r4, r4, #0xc0000000\r |
1299 | str r4, [r1,#4]\r |
1300 | b .dw_shadow_done\r |
1301 | \r |
1302 | .dwloop_end:\r |
1303 | ldr r0, =rendstatus\r |
1304 | ldr r1, [r0]\r |
1305 | and r6, r6, #2\r |
1306 | orr r1, r1, r6\r |
1307 | str r1, [r0]\r |
1308 | \r |
1309 | ldmfd sp!, {r4-r11,r12}\r |
1310 | bx r12\r |
1311 | \r |
1312 | \r |
1313 | @ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r |
1314 | \r |
1315 | \r |
1316 | @ hilights 2 pixels in RGB444/BGR444 format\r |
1317 | .macro TileDoShHi2Pixels444 reg\r |
1318 | mov \reg, \reg, ror #12\r |
1319 | adds \reg, \reg, #0x40000000\r |
1320 | orrcs \reg, \reg, #0xf0000000\r |
1321 | mov \reg, \reg, ror #28\r |
1322 | adds \reg, \reg, #0x40000000\r |
1323 | orrcs \reg, \reg, #0xf0000000\r |
1324 | mov \reg, \reg, ror #28\r |
1325 | adds \reg, \reg, #0x40000000\r |
1326 | orrcs \reg, \reg, #0xf0000000\r |
1327 | mov \reg, \reg, ror #24\r |
1328 | adds \reg, \reg, #0x40000000\r |
1329 | orrcs \reg, \reg, #0xf0000000\r |
1330 | mov \reg, \reg, ror #28\r |
1331 | adds \reg, \reg, #0x40000000\r |
1332 | orrcs \reg, \reg, #0xf0000000\r |
1333 | mov \reg, \reg, ror #28\r |
1334 | adds \reg, \reg, #0x40000000\r |
1335 | orrcs \reg, \reg, #0xf0000000\r |
1336 | mov \reg, \reg, ror #12\r |
1337 | .endm\r |
1338 | \r |
1339 | \r |
1340 | .global FinalizeLineBGR444 @ int sh\r |
1341 | \r |
1342 | FinalizeLineBGR444:\r |
1343 | stmfd sp!, {r4-r6,lr}\r |
1344 | mov r6, r0\r |
1345 | ldr lr, =(Pico+0x22228) @ Pico.video\r |
1346 | ldr r0, =DrawLineDest\r |
1347 | ldrb r12, [lr, #12]\r |
1348 | ldr r0, [r0]\r |
1349 | sub r3, lr, #0x128 @ r3=Pico.cram\r |
1350 | \r |
1351 | tst r12, #1\r |
1352 | movne r2, #320/4 @ len\r |
1353 | bne .fl_no32colBGR444\r |
1354 | ldr r4, =PicoOpt\r |
1355 | mov r2, #256/4\r |
1356 | ldr r4, [r4]\r |
1357 | tst r4, #0x100\r |
1358 | addeq r0, r0, #32*2\r |
1359 | \r |
1360 | .fl_no32colBGR444:\r |
1361 | tst r6, r6\r |
1362 | beq .fl_noshBGR444\r |
1363 | \r |
1364 | ldr r4, =HighPal\r |
1365 | \r |
1366 | ldrb r12, [lr, #-0x1a] @ 0x2220e ~ dirtyPal\r |
1367 | tst r12, r12\r |
1368 | moveq r3, r4\r |
1369 | beq .fl_noshBGR444\r |
1370 | mov r12, #0\r |
1371 | strb r12, [lr, #-0x1a]\r |
1372 | \r |
1373 | mov lr, #0x40/8\r |
1374 | @ copy pal:\r |
1375 | .fl_loopcpBGR444:\r |
1376 | ldmia r3!, {r1,r5,r6,r12}\r |
1377 | subs lr, lr, #1\r |
1378 | stmia r4!, {r1,r5,r6,r12}\r |
1379 | bne .fl_loopcpBGR444\r |
1380 | \r |
1381 | @ shadowed pixels:\r |
1382 | mov r12, #0x0077\r |
1383 | orr r12,r12,#0x0700\r |
1384 | orr r12,r12,r12,lsl #16\r |
1385 | sub r3, r3, #0x40*2\r |
1386 | add r5, r4, #0x80*2\r |
1387 | mov lr, #0x40/4\r |
1388 | .fl_loopcpBGR444_sh:\r |
1389 | ldmia r3!, {r1,r6}\r |
1390 | subs lr, lr, #1\r |
1391 | and r1, r12, r1, lsr #1\r |
1392 | and r6, r12, r6, lsr #1\r |
1393 | stmia r4!, {r1,r6}\r |
1394 | stmia r5!, {r1,r6}\r |
1395 | bne .fl_loopcpBGR444_sh\r |
1396 | \r |
1397 | @ hilighted pixels:\r |
1398 | sub r3, r3, #0x40*2\r |
1399 | mov lr, #0x40/2\r |
1400 | .fl_loopcpBGR444_hi:\r |
1401 | ldr r1, [r3], #4\r |
1402 | TileDoShHi2Pixels444 r1\r |
1403 | str r1, [r4], #4\r |
1404 | subs lr, lr, #1\r |
1405 | bne .fl_loopcpBGR444_hi\r |
1406 | \r |
1407 | sub r3, r4, #0x40*3*2\r |
1408 | \r |
1409 | \r |
1410 | .fl_noshBGR444:\r |
1411 | ldr r1, =(HighCol+8)\r |
1412 | mov lr, #0xff\r |
1413 | mov lr, lr, lsl #1\r |
1414 | \r |
1415 | .fl_loopBGR444:\r |
1416 | \r |
1417 | ldr r12, [r1], #4\r |
1418 | subs r2, r2, #1\r |
1419 | \r |
1420 | and r4, lr, r12, lsl #1\r |
1421 | ldrh r4, [r3, r4]\r |
1422 | and r5, lr, r12, lsr #7\r |
1423 | ldrh r5, [r3, r5]\r |
1424 | and r6, lr, r12, lsr #15\r |
1425 | ldrh r6, [r3, r6]\r |
1426 | orr r4, r4, r5, lsl #16\r |
1427 | \r |
1428 | and r5, lr, r12, lsr #23\r |
1429 | ldrh r5, [r3, r5] @ 2c.i.\r |
1430 | orr r5, r6, r5, lsl #16\r |
1431 | \r |
1432 | stmia r0!, {r4,r5}\r |
1433 | bne .fl_loopBGR444\r |
1434 | \r |
1435 | \r |
1436 | ldmfd sp!, {r4-r6,lr}\r |
1437 | bx lr\r |
1438 | \r |
1439 | \r |
1440 | @ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r |
1441 | \r |
1442 | \r |
1443 | @ hilights 2 pixels in RGB555/BGR555 format\r |
1444 | .macro TileDoShHi2Pixels555 reg\r |
1445 | adds \reg, \reg, #0x40000000\r |
1446 | orrcs \reg, \reg, #0xf0000000\r |
1447 | mov \reg, \reg, ror #27\r |
1448 | adds \reg, \reg, #0x40000000\r |
1449 | orrcs \reg, \reg, #0xf0000000\r |
1450 | mov \reg, \reg, ror #26\r |
1451 | adds \reg, \reg, #0x40000000\r |
1452 | orrcs \reg, \reg, #0xf0000000\r |
1453 | mov \reg, \reg, ror #27\r |
1454 | adds \reg, \reg, #0x40000000\r |
1455 | orrcs \reg, \reg, #0xf0000000\r |
1456 | mov \reg, \reg, ror #27\r |
1457 | adds \reg, \reg, #0x40000000\r |
1458 | orrcs \reg, \reg, #0xf0000000\r |
1459 | mov \reg, \reg, ror #26\r |
1460 | adds \reg, \reg, #0x40000000\r |
1461 | orrcs \reg, \reg, #0xf0000000\r |
1462 | mov \reg, \reg, ror #27\r |
1463 | .endm\r |
1464 | \r |
1465 | \r |
1466 | @ Convert 0000bbb0 ggg0rrr0\r |
1467 | @ to rrrrrggg gggbbbbb\r |
1468 | \r |
1469 | @ r2,r3,r9 - scratch, lr = 0x001c001c, r8 = 0x00030003\r |
1470 | .macro convRGB565 reg\r |
1471 | and r2, lr, \reg,lsl #1\r |
1472 | and r9, r8, \reg,lsr #2\r |
1473 | orr r2, r2, r9 @ r2=red\r |
1474 | and r3, lr, \reg,lsr #7\r |
1475 | and r9, r8, \reg,lsr #10\r |
1476 | orr r3, r3, r9 @ r3=blue\r |
1477 | and \reg, \reg, lr, lsl #3\r |
1478 | orr \reg, \reg, \reg,lsl #3 @ green\r |
1479 | orr \reg, \reg, r2, lsl #11 @ add red back\r |
1480 | orr \reg, \reg, r3 @ add blue back\r |
1481 | .endm\r |
1482 | \r |
1483 | vidConvCpyRGB565: @ void *to, void *from, int pixels\r |
1484 | stmfd sp!, {r4-r9,lr}\r |
1485 | \r |
1486 | mov r12, r2, lsr #3 @ repeats\r |
1487 | mov lr, #0x001c0000\r |
1488 | orr lr, lr, #0x01c @ lr == pattern 0x001c001c\r |
1489 | mov r8, #0x00030000\r |
1490 | orr r8, r8, #0x003 @ lr == pattern 0x001c001c\r |
1491 | \r |
1492 | .loopRGB565:\r |
1493 | ldmia r1!, {r4-r7}\r |
1494 | subs r12, r12, #1\r |
1495 | convRGB565 r4\r |
1496 | str r4, [r0], #4\r |
1497 | convRGB565 r5\r |
1498 | str r5, [r0], #4\r |
1499 | convRGB565 r6\r |
1500 | str r6, [r0], #4\r |
1501 | convRGB565 r7\r |
1502 | str r7, [r0], #4\r |
1503 | \r |
1504 | bgt .loopRGB565\r |
1505 | \r |
1506 | ldmfd sp!, {r4-r9,lr}\r |
1507 | bx lr\r |
1508 | \r |
1509 | \r |
1510 | \r |
1511 | .global FinalizeLineRGB555 @ int sh\r |
1512 | \r |
1513 | FinalizeLineRGB555:\r |
1514 | stmfd sp!, {r4-r8,lr}\r |
3d48f143 |
1515 | ldr r8, =(Pico+0x22228) @ Pico.video\r |
cc68a136 |
1516 | ldr r4, =HighPal\r |
1517 | \r |
3d48f143 |
1518 | ldrb r7, [r8, #-0x1a] @ 0x2220e ~ dirtyPal\r |
cc68a136 |
1519 | mov r6, r0\r |
1520 | mov r1, #0\r |
1521 | tst r7, r7\r |
1522 | beq .fl_noconvRGB555\r |
3d48f143 |
1523 | strb r1, [r8, #-0x1a]\r |
1524 | sub r1, r8, #0x128 @ r1=Pico.cram\r |
cc68a136 |
1525 | mov r0, r4\r |
1526 | mov r2, #0x40\r |
1527 | bl vidConvCpyRGB565\r |
1528 | \r |
1529 | .fl_noconvRGB555:\r |
cc68a136 |
1530 | mov r3, r4\r |
1531 | tst r6, r6\r |
1532 | beq .fl_noshRGB555\r |
1533 | tst r7, r7\r |
1534 | beq .fl_noshRGB555\r |
1535 | \r |
1536 | @ shadowed pixels:\r |
1537 | mov r12, #0x008e\r |
1538 | orr r12,r12,#0x7300\r |
1539 | orr r12,r12,r12,lsl #16\r |
1540 | add r4, r3, #0x40*2\r |
1541 | add r5, r3, #0xc0*2\r |
1542 | mov lr, #0x40/4\r |
1543 | .fl_loopcpRGB555_sh:\r |
1544 | ldmia r3!, {r1,r6}\r |
1545 | subs lr, lr, #1\r |
1546 | and r1, r12, r1, lsr #1\r |
1547 | and r6, r12, r6, lsr #1\r |
1548 | stmia r4!, {r1,r6}\r |
1549 | stmia r5!, {r1,r6}\r |
1550 | bne .fl_loopcpRGB555_sh\r |
1551 | \r |
1552 | @ hilighted pixels:\r |
1553 | sub r3, r3, #0x40*2\r |
1554 | mov lr, #0x40/2\r |
1555 | .fl_loopcpRGB555_hi:\r |
1556 | ldr r1, [r3], #4\r |
1557 | TileDoShHi2Pixels555 r1\r |
1558 | str r1, [r4], #4\r |
1559 | subs lr, lr, #1\r |
1560 | bne .fl_loopcpRGB555_hi\r |
1561 | \r |
1562 | sub r3, r3, #0x40*2\r |
1563 | \r |
cc68a136 |
1564 | .fl_noshRGB555:\r |
3d48f143 |
1565 | ldr r0, =DrawLineDest\r |
cc68a136 |
1566 | ldr r1, =(HighCol+8)\r |
3d48f143 |
1567 | ldr r0, [r0]\r |
1568 | \r |
1569 | ldrb r12, [r8, #12]\r |
cc68a136 |
1570 | mov lr, #0xff\r |
1571 | mov lr, lr, lsl #1\r |
1572 | \r |
3d48f143 |
1573 | tst r12, #1\r |
1574 | movne r2, #320/8 @ len\r |
1575 | bne .fl_no32colRGB555\r |
1576 | ldr r4, =PicoOpt\r |
1577 | mov r2, #256/8\r |
1578 | ldr r4, [r4]\r |
1579 | tst r4, #0x4000\r |
1580 | bne .fl_32scale_RGB555\r |
1581 | tst r4, #0x0100\r |
1582 | addeq r0, r0, #32*2\r |
1583 | \r |
1584 | .fl_no32colRGB555:\r |
cc68a136 |
1585 | .fl_loopRGB555:\r |
1586 | \r |
1587 | ldr r12, [r1], #4\r |
1588 | ldr r7, [r1], #4\r |
1589 | \r |
1590 | and r4, lr, r12, lsl #1\r |
1591 | ldrh r4, [r3, r4]\r |
1592 | and r5, lr, r12, lsr #7\r |
1593 | ldrh r5, [r3, r5]\r |
1594 | and r6, lr, r12, lsr #15\r |
1595 | ldrh r6, [r3, r6]\r |
1596 | orr r4, r4, r5, lsl #16\r |
1597 | \r |
1598 | and r5, lr, r12, lsr #23\r |
1599 | ldrh r5, [r3, r5]\r |
1600 | and r8, lr, r7, lsl #1\r |
1601 | ldrh r8, [r3, r8]\r |
1602 | orr r5, r6, r5, lsl #16\r |
1603 | \r |
1604 | and r6, lr, r7, lsr #7\r |
1605 | ldrh r6, [r3, r6]\r |
1606 | and r12,lr, r7, lsr #15\r |
1607 | ldrh r12,[r3, r12]\r |
1608 | orr r8, r8, r6, lsl #16\r |
1609 | \r |
1610 | and r6, lr, r7, lsr #23\r |
1611 | ldrh r6, [r3, r6] @ 1 cycle interlock here (r6)\r |
1612 | subs r2, r2, #1\r |
1613 | orr r12,r12, r6, lsl #16\r |
1614 | \r |
1615 | stmia r0!, {r4,r5,r8,r12}\r |
1616 | bne .fl_loopRGB555\r |
1617 | \r |
3d48f143 |
1618 | ldmfd sp!, {r4-r8,lr}\r |
1619 | bx lr\r |
1620 | \r |
1621 | \r |
1622 | .fl_32scale_RGB555:\r |
1623 | stmfd sp!, {r9,r10}\r |
1624 | mov r9, #0x3900 @ f800 07e0 001f | e000 0780 001c | 3800 01e0 0007\r |
1625 | orr r9, r9, #0x00e7\r |
1626 | \r |
1627 | .fl_loop32scale_RGB555:\r |
1628 | ldr r12, [r1], #4\r |
1629 | ldr r7, [r1], #4\r |
cc68a136 |
1630 | \r |
3d48f143 |
1631 | and r4, lr, r12,lsl #1\r |
1632 | ldrh r4, [r3, r4]\r |
1633 | and r5, lr, r12,lsr #7\r |
1634 | ldrh r5, [r3, r5]\r |
1635 | and r4, r4, r9, lsl #2\r |
1636 | orr r4, r4, r4, lsl #14 @ r4[31:16] = 1/4 pix_s 0\r |
1637 | and r5, r5, r9, lsl #2\r |
1638 | sub r6, r5, r5, lsr #2 @ r6 = 3/4 pix_s 1\r |
1639 | add r4, r4, r6, lsl #16 @ pix_d 0, 1\r |
1640 | and r6, lr, r12,lsr #15\r |
1641 | ldrh r6, [r3, r6]\r |
1642 | and r12,lr, r12,lsr #23\r |
1643 | ldrh r12,[r3, r12]\r |
1644 | and r6, r6, r9, lsl #2\r |
1645 | add r5, r5, r6\r |
1646 | mov r5, r5, lsr #1\r |
1647 | sub r6, r6, r6, lsr #2 @ r6 = 3/4 pix_s 2\r |
1648 | orr r5, r5, r6, lsl #16\r |
1649 | \r |
1650 | and r6, lr, r7, lsl #1\r |
1651 | ldrh r6, [r3, r6]\r |
1652 | and r12,r12,r9, lsl #2\r |
1653 | add r5, r5, r12,lsl #14 @ pix_d 2, 3\r |
1654 | and r6, r6, r9, lsl #2\r |
1655 | orr r6, r12,r6, lsl #16 @ pix_d 4, 5\r |
1656 | \r |
1657 | and r12,lr, r7, lsr #7\r |
1658 | ldrh r12,[r3, r12]\r |
1659 | and r10,lr, r7, lsr #15\r |
1660 | ldrh r10,[r3, r10]\r |
1661 | and r12,r12,r9, lsl #2\r |
1662 | sub r8, r12,r12,lsr #2 @ r8 = 3/4 pix_s 1\r |
1663 | add r8, r8, r6, lsr #18\r |
1664 | and r7, lr, r7, lsr #23\r |
1665 | ldrh r7, [r3, r7]\r |
1666 | and r10,r10,r9, lsl #2\r |
1667 | orr r8, r8, r10,lsl #15\r |
1668 | add r8, r8, r12,lsl #15 @ pix_d 6, 7\r |
1669 | sub r10,r10,r10,lsr #2 @ r10= 3/4 pix_s 2\r |
1670 | and r7, r7, r9, lsl #2\r |
1671 | add r10,r10,r7, lsr #2 @ += 1/4 pix_s 3\r |
1672 | orr r10,r10,r7, lsl #16 @ pix_d 8, 9\r |
1673 | \r |
1674 | subs r2, r2, #1\r |
1675 | \r |
1676 | stmia r0!, {r4,r5,r6,r8,r10}\r |
1677 | bne .fl_loop32scale_RGB555\r |
1678 | \r |
1679 | ldmfd sp!, {r9,r10}\r |
cc68a136 |
1680 | ldmfd sp!, {r4-r8,lr}\r |
3d48f143 |
1681 | bx lr\r |
1682 | \r |
cc68a136 |
1683 | \r |
1684 | @ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r |
1685 | \r |
1686 | @ utility\r |
1687 | .global blockcpy @ void *dst, void *src, size_t n\r |
1688 | \r |
1689 | blockcpy:\r |
1690 | stmfd sp!, {r4,r5}\r |
1691 | mov r2, r2, lsr #4\r |
1692 | blockcpy_loop:\r |
1693 | ldmia r1!, {r3-r5,r12}\r |
1694 | subs r2, r2, #1\r |
1695 | stmia r0!, {r3-r5,r12}\r |
1696 | bne blockcpy_loop\r |
1697 | ldmfd sp!, {r4,r5}\r |
1698 | bx lr\r |
1699 | \r |
1700 | \r |
1701 | .global blockcpy_or @ void *dst, void *src, size_t n, int pat\r |
1702 | \r |
1703 | blockcpy_or:\r |
1704 | stmfd sp!, {r4-r6}\r |
1705 | orr r3, r3, r3, lsl #8\r |
1706 | orr r3, r3, r3, lsl #16\r |
1707 | mov r2, r2, lsr #4\r |
1708 | blockcpy_loop_or:\r |
1709 | ldmia r1!, {r4-r6,r12}\r |
1710 | subs r2, r2, #1\r |
1711 | orr r4, r4, r3\r |
1712 | orr r5, r5, r3\r |
1713 | orr r6, r6, r3\r |
1714 | orr r12,r12,r3\r |
1715 | stmia r0!, {r4-r6,r12}\r |
1716 | bne blockcpy_loop_or\r |
1717 | ldmfd sp!, {r4-r6}\r |
1718 | bx lr\r |
1719 | \r |