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 |
14 | .extern Scanline\r |
15 | .extern HighSprZ\r |
16 | .extern rendstatus\r |
17 | .extern DrawLineDest\r |
cc68a136 |
18 | .extern DrawStripInterlace\r |
07abbab1 |
19 | .extern HighCacheS_ptr\r |
cc68a136 |
20 | \r |
21 | \r |
22 | @ helper\r |
23 | .macro TilePixel pat lsrr offs\r |
24 | .if !\lsrr\r |
25 | ands r4, \pat, r2\r |
26 | .else\r |
27 | ands r4, \pat, r2, lsr #\lsrr\r |
28 | .endif\r |
29 | orrne r4, r3, r4\r |
30 | strneb r4, [r1,#\offs]\r |
31 | .endm\r |
32 | \r |
33 | @ TileNorm (r1=pdest, r2=pixels8, r3=pal) r4: scratch, pat: register with helper pattern 0xf\r |
34 | .macro TileNorm pat\r |
35 | TilePixel \pat, 12, 0 @ #0x0000f000\r |
36 | TilePixel \pat, 8, 1 @ #0x00000f00\r |
37 | TilePixel \pat, 4, 2 @ #0x000000f0\r |
38 | TilePixel \pat, 0, 3 @ #0x0000000f\r |
39 | TilePixel \pat, 28, 4 @ #0xf0000000\r |
40 | TilePixel \pat, 24, 5 @ #0x0f000000\r |
41 | TilePixel \pat, 20, 6 @ #0x00f00000\r |
42 | TilePixel \pat, 16, 7 @ #0x000f0000\r |
43 | .endm\r |
44 | \r |
45 | @ TileFlip (r1=pdest, r2=pixels8, r3=pal) r4: scratch, pat: register with helper pattern 0xf\r |
46 | .macro TileFlip pat\r |
47 | TilePixel \pat, 16, 0 @ #0x000f0000\r |
48 | TilePixel \pat, 20, 1 @ #0x00f00000\r |
49 | TilePixel \pat, 24, 2 @ #0x0f000000\r |
50 | TilePixel \pat, 28, 3 @ #0xf0000000\r |
51 | TilePixel \pat, 0, 4 @ #0x0000000f\r |
52 | TilePixel \pat, 4, 5 @ #0x000000f0\r |
53 | TilePixel \pat, 8, 6 @ #0x00000f00\r |
54 | TilePixel \pat, 12, 7 @ #0x0000f000\r |
55 | .endm\r |
56 | \r |
57 | @ shadow/hilight mode\r |
58 | \r |
59 | @ this one is for hi priority layer\r |
60 | .macro TilePixelShHP lsrr offs\r |
61 | .if !\lsrr\r |
62 | ands r4, r12, r2\r |
63 | .else\r |
64 | ands r4, r12, r2, lsr #\lsrr\r |
65 | .endif\r |
66 | ldreqb r4, [r1,#\offs]\r |
67 | orrne r4, r3, r4\r |
cc68a136 |
68 | andeq r4, r4, #0x3f\r |
07abbab1 |
69 | strb r4, [r1,#\offs]\r |
cc68a136 |
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 |
07abbab1 |
157 | beq 0f\r |
cc68a136 |
158 | cmp r4, #0xe\r |
07abbab1 |
159 | ldrgeb r4, [r1,#\ofs]\r |
160 | orrlt r4, r3, r4 @ normal\r |
161 | \r |
162 | biceq r4, r4, #0xc0 @ hilight\r |
163 | orreq r4, r4, #0x80\r |
164 | orrgt r4, r4, #0xc0 @ shadow\r |
165 | \r |
cc68a136 |
166 | strb r4, [r1,#\ofs]\r |
07abbab1 |
167 | 0:\r |
cc68a136 |
168 | .endm\r |
169 | \r |
170 | @ TileFlipSh (r1=pdest, r2=pixels8, r3=pal) r4,r7: scratch, r0=sx, r12: register with helper pattern 0xf\r |
171 | .macro TileFlipSh\r |
172 | TileDoShGenPixel 16, 0 @ #0x000f0000\r |
173 | TileDoShGenPixel 20, 1 @ #0x00f00000\r |
174 | TileDoShGenPixel 24, 2 @ #0x0f000000\r |
175 | TileDoShGenPixel 28, 3 @ #0xf0000000\r |
176 | TileDoShGenPixel 0, 4 @ #0x0000000f\r |
177 | TileDoShGenPixel 4, 5 @ #0x000000f0\r |
178 | TileDoShGenPixel 8, 6 @ #0x00000f00\r |
179 | TileDoShGenPixel 12, 7 @ #0x0000f000\r |
180 | .endm\r |
181 | \r |
182 | @ TileNormSh (r1=pdest, r2=pixels8, r3=pal) r4,r7: scratch, r0=sx, r12: register with helper pattern 0xf\r |
183 | .macro TileNormSh\r |
184 | TileDoShGenPixel 12, 0 @ #0x0000f000\r |
185 | TileDoShGenPixel 8, 1 @ #0x00000f00\r |
186 | TileDoShGenPixel 4, 2 @ #0x000000f0\r |
187 | TileDoShGenPixel 0, 3 @ #0x0000000f\r |
188 | TileDoShGenPixel 28, 4 @ #0xf0000000\r |
189 | TileDoShGenPixel 24, 5 @ #0x0f000000\r |
190 | TileDoShGenPixel 20, 6 @ #0x00f00000\r |
191 | TileDoShGenPixel 16, 7 @ #0x000f0000\r |
192 | .endm\r |
193 | \r |
07abbab1 |
194 | .macro TileDoShGenPixel_noop shift ofs\r |
195 | .if \shift\r |
196 | and r4, r12, r2, lsr #\shift\r |
197 | .else\r |
198 | and r4, r12, r2\r |
199 | .endif\r |
200 | sub r7, r4, #1\r |
201 | cmp r7, #0xd\r |
202 | orrcc r4, r3, r4 @ 0-0xc (was 1-0xd)\r |
203 | strccb r4, [r1,#\ofs]\r |
204 | .endm\r |
205 | \r |
206 | .macro TileFlipSh_noop\r |
207 | TileDoShGenPixel_noop 16, 0 @ #0x000f0000\r |
208 | TileDoShGenPixel_noop 20, 1 @ #0x00f00000\r |
209 | TileDoShGenPixel_noop 24, 2 @ #0x0f000000\r |
210 | TileDoShGenPixel_noop 28, 3 @ #0xf0000000\r |
211 | TileDoShGenPixel_noop 0, 4 @ #0x0000000f\r |
212 | TileDoShGenPixel_noop 4, 5 @ #0x000000f0\r |
213 | TileDoShGenPixel_noop 8, 6 @ #0x00000f00\r |
214 | TileDoShGenPixel_noop 12, 7 @ #0x0000f000\r |
215 | .endm\r |
216 | \r |
217 | .macro TileNormSh_noop\r |
218 | TileDoShGenPixel_noop 12, 0 @ #0x0000f000\r |
219 | TileDoShGenPixel_noop 8, 1 @ #0x00000f00\r |
220 | TileDoShGenPixel_noop 4, 2 @ #0x000000f0\r |
221 | TileDoShGenPixel_noop 0, 3 @ #0x0000000f\r |
222 | TileDoShGenPixel_noop 28, 4 @ #0xf0000000\r |
223 | TileDoShGenPixel_noop 24, 5 @ #0x0f000000\r |
224 | TileDoShGenPixel_noop 20, 6 @ #0x00f00000\r |
225 | TileDoShGenPixel_noop 16, 7 @ #0x000f0000\r |
226 | .endm\r |
227 | \r |
228 | .macro TileDoShGenPixel_onlyop_lp shift ofs\r |
229 | .if \shift\r |
230 | ands r7, r12, r2, lsr #\shift\r |
231 | .else\r |
232 | ands r7, r12, r2\r |
233 | .endif\r |
234 | ldrneb r4, [r1,#\ofs]\r |
235 | tstne r4, #0x40\r |
236 | beq 0f\r |
237 | \r |
238 | cmp r7, #0xe\r |
239 | biceq r4, r4, #0xc0 @ hilight\r |
240 | orreq r4, r4, #0x80\r |
241 | orrgt r4, r4, #0xc0 @ shadow\r |
242 | strgeb r4, [r1,#\ofs]\r |
243 | 0:\r |
244 | .endm\r |
245 | \r |
246 | .macro TileFlipSh_onlyop_lp\r |
247 | TileDoShGenPixel_onlyop_lp 16, 0 @ #0x000f0000\r |
248 | TileDoShGenPixel_onlyop_lp 20, 1 @ #0x00f00000\r |
249 | TileDoShGenPixel_onlyop_lp 24, 2 @ #0x0f000000\r |
250 | TileDoShGenPixel_onlyop_lp 28, 3 @ #0xf0000000\r |
251 | TileDoShGenPixel_onlyop_lp 0, 4 @ #0x0000000f\r |
252 | TileDoShGenPixel_onlyop_lp 4, 5 @ #0x000000f0\r |
253 | TileDoShGenPixel_onlyop_lp 8, 6 @ #0x00000f00\r |
254 | TileDoShGenPixel_onlyop_lp 12, 7 @ #0x0000f000\r |
255 | .endm\r |
256 | \r |
257 | .macro TileNormSh_onlyop_lp\r |
258 | TileDoShGenPixel_onlyop_lp 12, 0 @ #0x0000f000\r |
259 | TileDoShGenPixel_onlyop_lp 8, 1 @ #0x00000f00\r |
260 | TileDoShGenPixel_onlyop_lp 4, 2 @ #0x000000f0\r |
261 | TileDoShGenPixel_onlyop_lp 0, 3 @ #0x0000000f\r |
262 | TileDoShGenPixel_onlyop_lp 28, 4 @ #0xf0000000\r |
263 | TileDoShGenPixel_onlyop_lp 24, 5 @ #0x0f000000\r |
264 | TileDoShGenPixel_onlyop_lp 20, 6 @ #0x00f00000\r |
265 | TileDoShGenPixel_onlyop_lp 16, 7 @ #0x000f0000\r |
266 | .endm\r |
267 | \r |
cc68a136 |
268 | \r |
269 | @ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r |
270 | \r |
271 | @ struct TileStrip\r |
272 | @ {\r |
273 | @ int nametab; // 0x00\r |
274 | @ int line; // 0x04\r |
275 | @ int hscroll; // 0x08\r |
276 | @ int xmask; // 0x0C\r |
277 | @ int *hc; // 0x10 (pointer to cache buffer)\r |
278 | @ int cells; // 0x14\r |
279 | @ };\r |
280 | \r |
83c093a4 |
281 | @ void DrawLayer(int plane_sh, int *hcache, int cellskip, int maxcells);\r |
cc68a136 |
282 | \r |
83c093a4 |
283 | .global DrawLayer\r |
cc68a136 |
284 | \r |
285 | DrawLayer:\r |
286 | stmfd sp!, {r4-r11,lr}\r |
287 | \r |
288 | ldr r11, =(Pico+0x22228) @ Pico.video\r |
289 | mov r8, #1\r |
290 | \r |
83c093a4 |
291 | ldrb r7, [r11, #16] @ ??vv??hh\r |
cc68a136 |
292 | \r |
293 | mov r6, r1 @ hcache\r |
83c093a4 |
294 | orr r9, r3, r0, lsl #30\r |
295 | orr r9, r9, r2, lsl #8 @ r9=sh[31]|cellskip[15:8]|maxcells[7:0] (tmp)\r |
cc68a136 |
296 | \r |
297 | mov r1, r7, lsl #4\r |
298 | orr r1, r1, #0x00ff\r |
299 | \r |
300 | and r10, r7, #3\r |
301 | cmp r10, #1\r |
302 | biclt r1, r1, #0xfc00\r |
303 | biceq r1, r1, #0xfe00\r |
304 | bicgt r1, r1, #0xff00 @ r1=ymask=(height<<8)|0xff; ...; // Y Mask in pixels\r |
305 | \r |
306 | add r10, r10, #5\r |
307 | cmp r10, #7\r |
308 | subge r10, r10, #1 @ r10=shift[width] (5,6,6,7)\r |
309 | \r |
310 | @ calculate xmask:\r |
311 | mov r5, r8, lsl r10\r |
312 | sub r5, r5, #1 @ r5=xmask\r |
313 | \r |
314 | @ Find name table:\r |
83c093a4 |
315 | ands r0, r0, #1\r |
cc68a136 |
316 | ldreqb r12, [r11, #2]\r |
317 | ldrneb r12, [r11, #4]\r |
318 | \r |
319 | ldr r2, =Scanline @ trying to make good use of pipeline here\r |
320 | ldr lr, =(Pico+0x10000) @ lr=Pico.vram\r |
321 | \r |
322 | moveq r12, r12, lsl #10\r |
323 | movne r12, r12, lsl #13\r |
324 | and r12, r12, #(7<<13) @ r12=(ts->nametab<<1) (halfword compliant)\r |
325 | \r |
326 | ldrh r8, [r11, #12]\r |
327 | ldrb r7, [r11, #11]\r |
328 | ldr r2, [r2]\r |
329 | \r |
330 | mov r4, r8, lsr #8 @ pvid->reg[13]\r |
331 | mov r4, r4, lsl #10 @ htab=pvid->reg[13]<<9; (halfwords)\r |
332 | tst r7, #2\r |
333 | addne r4, r4, r2, lsl #2 @ htab+=Scanline<<1; // Offset by line\r |
334 | tst r7, #1\r |
335 | biceq r4, r4, #0x1f @ htab&=~0xf; // Offset by tile\r |
336 | add r4, r4, r0, lsl #1 @ htab+=plane\r |
337 | bic r4, r4, #0x00ff0000 @ just in case\r |
338 | ldrh r3, [lr, r4] @ r3=hscroll\r |
339 | \r |
340 | tst r7, #4\r |
341 | bne .DrawStrip_vsscroll\r |
342 | \r |
343 | @ Get vertical scroll value:\r |
344 | add r7, lr, #0x012000\r |
345 | add r7, r7, #0x000180 @ r7=Pico.vsram (Pico+0x22180)\r |
346 | ldr r7, [r7]\r |
347 | \r |
348 | tst r8, #2\r |
349 | tstne r8, #4\r |
350 | bne .DrawStrip_interlace\r |
351 | \r |
352 | tst r0, r0\r |
353 | movne r7, r7, lsr #16\r |
354 | \r |
355 | @ Find the line in the name table\r |
356 | add r2, r2, r7\r |
357 | and r2, r2, r1\r |
358 | mov r4, r2, lsr #3\r |
359 | add r10, r10, #1 @ shift[width]++\r |
360 | add r12, r12, r4, lsl r10 @ nametab+=(ts.line>>3)<<shift[width];\r |
361 | \r |
362 | @ ldmia r0, {r1,r2,r3,r5,r6,r9} @ r2=line, r3=ts->hscroll, r5=ts->xmask, r6=ts->hc, r9=ts->cells\r |
cc68a136 |
363 | \r |
364 | and r10,r2, #7\r |
365 | mov r10,r10, lsl #1 @ r10=ty=(ts->line&7)<<1;\r |
366 | orr r10,r10, r9, lsl #24\r |
367 | \r |
368 | rsb r8, r3, #0\r |
369 | mov r8, r8, lsr #3 @ r8=tilex=(-ts->hscroll)>>3\r |
370 | \r |
371 | sub r1, r3, #1\r |
372 | and r1, r1, #7\r |
373 | add r7, r1, #1 @ r7=dx=((ts->hscroll-1)&7)+1\r |
374 | \r |
375 | tst r9, #1<<31\r |
376 | mov r3, #0\r |
740da8c6 |
377 | orrne r10,r10, #1<<23 @ r10=(cells<<24|sh<<23|hi_not_empty<<22|had_output<<21|ty)\r |
cc68a136 |
378 | movne r3, #0x40 @ default to shadowed pal on sh mode\r |
379 | \r |
cc68a136 |
380 | cmp r7, #8\r |
381 | addne r10,r10, #0x01000000 @ we will loop cells+1 times if there is scroll\r |
382 | \r |
83c093a4 |
383 | and r9, r9, #0xff00\r |
384 | add r8, r8, r9, lsr #8 @ tilex+=cellskip\r |
385 | add r7, r7, r9, lsr #5 @ dx+=cellskip<<3;\r |
386 | sub r10,r10,r9, lsl #16 @ cells-=cellskip\r |
83c093a4 |
387 | \r |
cc68a136 |
388 | @ cache some stuff to avoid mem access\r |
ea8c405f |
389 | .if OVERRIDE_HIGHCOL\r |
390 | ldr r11,=HighCol\r |
391 | mov r0, #0xf\r |
392 | ldr r11,[r11]\r |
393 | .else\r |
cc68a136 |
394 | ldr r11,=HighCol\r |
395 | mov r0, #0xf\r |
ea8c405f |
396 | .endif\r |
397 | \r |
398 | mvn r9, #0 @ r9=prevcode=-1\r |
cc68a136 |
399 | add r1, r11, r7 @ r1=pdest\r |
400 | \r |
401 | \r |
402 | @ r4 & r7 are scratch in this loop\r |
403 | .dsloop_subr1:\r |
404 | sub r1, r1, #8\r |
405 | .dsloop: @ 40-41 times\r |
406 | subs r10,r10, #0x01000000\r |
407 | bmi .dsloop_exit\r |
408 | \r |
409 | .dsloop_enter:\r |
410 | and r7, r5, r8\r |
411 | add r7, lr, r7, lsl #1 @ Pico.vram+((tilex&ts->xmask) as halfwords)\r |
412 | ldrh r7, [r7, r12] @ r7=code (int, but from unsigned, no sign extend)\r |
413 | \r |
414 | add r1, r1, #8\r |
415 | add r8, r8, #1\r |
416 | \r |
417 | tst r7, #0x8000\r |
418 | bne .DrawStrip_hiprio\r |
419 | \r |
420 | cmp r7, r9\r |
421 | beq .DrawStrip_samecode @ we know stuff about this tile already\r |
422 | \r |
423 | mov r9, r7 @ remember code\r |
740da8c6 |
424 | orr r10, r10, #1<<21 @ seen non hi-prio tile\r |
cc68a136 |
425 | \r |
426 | movs r2, r9, lsl #20 @ if (code&0x1000)\r |
427 | mov r2, r2, lsl #1\r |
428 | add r2, r2, r10, lsl #17\r |
429 | mov r2, r2, lsr #17\r |
430 | eorcs r2, r2, #0x0e @ if (code&0x1000) addr^=0xe;\r |
431 | \r |
432 | ldr r2, [lr, r2, lsl #1] @ pack=*(unsigned int *)(Pico.vram+addr); // Get 8 pixels\r |
433 | \r |
434 | bic r7, r3, #0x3f\r |
435 | and r3, r9, #0x6000\r |
436 | add r3, r7, r3, lsr #9 @ r3=pal=((code&0x6000)>>9);\r |
437 | \r |
438 | .DrawStrip_samecode:\r |
439 | tst r2, r2\r |
440 | beq .dsloop @ tileline blank\r |
441 | \r |
442 | cmp r2, r2, ror #4\r |
443 | beq .DrawStrip_SingleColor @ tileline singlecolor \r |
444 | \r |
445 | tst r9, #0x0800\r |
7a7c6476 |
446 | bne .DrawStrip_TileFlip\r |
cc68a136 |
447 | \r |
448 | @ (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r0: helper pattern\r |
cc68a136 |
449 | .DrawStrip_TileNorm:\r |
450 | TileNorm r0\r |
451 | b .dsloop\r |
452 | \r |
7a7c6476 |
453 | .DrawStrip_TileFlip:\r |
454 | TileFlip r0\r |
455 | b .dsloop\r |
456 | \r |
cc68a136 |
457 | .DrawStrip_SingleColor:\r |
458 | and r4, r2, #0xf\r |
459 | orr r4, r3, r4\r |
460 | orr r4, r4, r4, lsl #8\r |
461 | tst r1, #1 @ not aligned?\r |
462 | strneb r4, [r1], #1\r |
463 | streqh r4, [r1], #2\r |
464 | strh r4, [r1], #2\r |
465 | strh r4, [r1], #2\r |
466 | strh r4, [r1], #2\r |
467 | strneb r4, [r1], #1 @ have a remaining unaligned pixel?\r |
468 | b .dsloop_subr1\r |
469 | \r |
cc68a136 |
470 | .DrawStrip_hiprio_maybempt:\r |
471 | cmp r7, r9\r |
472 | beq .dsloop @ must've been empty, otherwise we wouldn't get here\r |
473 | movs r2, r7, lsl #20 @ if (code&0x1000)\r |
474 | mov r2, r2, lsl #1\r |
475 | add r2, r2, r10, lsl #17\r |
476 | mov r2, r2, lsr #17\r |
477 | eorcs r2, r2, #0x0e @ if (code&0x1000) addr^=0xe;\r |
478 | ldr r2, [lr, r2, lsl #1] @ pack=*(unsigned int *)(Pico.vram+addr); // Get 8 pixels\r |
479 | mov r9, r7 @ remember code\r |
480 | tst r2, r2\r |
740da8c6 |
481 | beq .dsloop\r |
482 | orr r10, r10, #1<<22\r |
483 | \r |
484 | .DrawStrip_hiprio:\r |
485 | tst r10, #0x00c00000\r |
486 | beq .DrawStrip_hiprio_maybempt\r |
487 | sub r0, r1, r11\r |
488 | orr r7, r7, r0, lsl #16\r |
489 | orr r7, r7, r10, lsl #25 @ (ty<<25)\r |
490 | tst r7, #0x1000\r |
491 | eorne r7, r7, #7<<26 @ if(code&0x1000) cval^=7<<26;\r |
492 | str r7, [r6], #4 @ cache hi priority tile\r |
493 | mov r0, #0xf\r |
cc68a136 |
494 | b .dsloop\r |
495 | \r |
496 | .dsloop_exit:\r |
740da8c6 |
497 | tst r10, #1<<21 @ seen non hi-prio tile\r |
498 | ldreq r1, =rendstatus\r |
cc68a136 |
499 | mov r0, #0\r |
740da8c6 |
500 | ldreq r2, [r1]\r |
cc68a136 |
501 | str r0, [r6] @ terminate the cache list\r |
740da8c6 |
502 | orreq r2, r2, #0x40 @ had a layer with all hi-prio tiles\r |
503 | streq r2, [r1]\r |
cc68a136 |
504 | \r |
505 | ldmfd sp!, {r4-r11,lr}\r |
506 | bx lr\r |
507 | \r |
6d7acf9e |
508 | @ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r |
cc68a136 |
509 | \r |
510 | .DrawStrip_vsscroll:\r |
6d7acf9e |
511 | rsb r8, r3, #0\r |
512 | mov r8, r8, lsr #3 @ r8=tilex=(-ts->hscroll)>>3\r |
7b802576 |
513 | bic r8, r8, #0x3fc00000\r |
514 | orr r8, r8, r5, lsl #25 @ r8=(xmask[31:25]|had_output[24]|tilex[21:0])\r |
cc68a136 |
515 | \r |
6d7acf9e |
516 | ldr r4, =Scanline\r |
517 | orr r5, r1, r10, lsl #24\r |
518 | ldr r4, [r4]\r |
519 | sub r1, r3, #1\r |
520 | orr r5, r5, r4, lsl #16 @ r5=(shift_width[31:24]|scanline[23:16]|ymask[15:0])\r |
521 | and r1, r1, #7\r |
522 | add r7, r1, #1 @ r7=dx=((ts->hscroll-1)&7)+1\r |
cc68a136 |
523 | \r |
6d7acf9e |
524 | mov r10,r9, lsl #16\r |
83c093a4 |
525 | tst r0, #1\r |
6d7acf9e |
526 | orrne r10,r10, #0x8000\r |
527 | tst r9, #1<<31\r |
528 | mov r3, #0\r |
529 | orr r10,r10, #0xff000000 @ will be adjusted on entering loop\r |
83c093a4 |
530 | 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 |
531 | movne r3, #0x40 @ default to shadowed pal on sh mode\r |
cc68a136 |
532 | \r |
83c093a4 |
533 | cmp r7, #8\r |
534 | subne r10,r10, #0x01000000 @ have hscroll, start with negative cell\r |
535 | \r |
536 | and r9, r9, #0xff00\r |
537 | add r8, r8, r9, lsr #8 @ tilex+=cellskip\r |
538 | add r7, r7, r9, lsr #5 @ dx+=cellskip<<3;\r |
539 | add r10,r10,r9, lsl #16 @ cell+=cellskip\r |
cc68a136 |
540 | \r |
6d7acf9e |
541 | @ cache some stuff to avoid mem access\r |
ea8c405f |
542 | .if OVERRIDE_HIGHCOL\r |
6d7acf9e |
543 | ldr r11,=HighCol\r |
544 | mov r0, #0xf\r |
ea8c405f |
545 | ldr r11,[r11]\r |
546 | .else\r |
547 | ldr r11,=HighCol\r |
548 | mov r0, #0xf\r |
549 | .endif\r |
6d7acf9e |
550 | \r |
ea8c405f |
551 | mvn r9, #0 @ r9=prevcode=-1\r |
552 | add r1, r11, r7 @ r1=pdest\r |
6d7acf9e |
553 | \r |
554 | @ r4 & r7 are scratch in this loop\r |
555 | .dsloop_vs_subr1:\r |
556 | sub r1, r1, #8\r |
557 | .dsloop_vs: @ 40-41 times\r |
558 | add r10,r10, #0x01000000\r |
559 | and r4, r10, #0x003f0000\r |
560 | cmp r4, r10, asr #8\r |
740da8c6 |
561 | ble .dsloop_vs_exit\r |
6d7acf9e |
562 | \r |
563 | @ calc offset and read tileline code to r7, also calc ty\r |
564 | add r7, lr, #0x012000\r |
565 | add r7, r7, #0x000180 @ r7=Pico.vsram (Pico+0x22180)\r |
566 | add r7, r7, r10,asr #23 @ vsram + ((cell&~1)<<1)\r |
567 | bic r7, r7, #3\r |
568 | tst r10,#0x8000 @ plane1?\r |
569 | addne r7, r7, #2\r |
570 | ldrh r7, [r7] @ r7=vscroll\r |
571 | \r |
572 | bic r10,r10,#0xff @ clear old ty\r |
7b802576 |
573 | and r4, r5, #0xff0000 @ scanline\r |
574 | add r4, r4, r7, lsl #16 @ ... += vscroll\r |
575 | and r4, r4, r5, lsl #16 @ ... &= ymask\r |
6d7acf9e |
576 | and r7, r4, #0x70000\r |
577 | orr r10,r10,r7, lsr #15 @ new ty\r |
578 | \r |
579 | mov r4, r4, lsr #19\r |
580 | mov r7, r5, lsr #24\r |
581 | mov r4, r4, lsl r7 @ nametabadd\r |
582 | \r |
583 | and r7, r8, r8, lsr #25\r |
584 | add r7, lr, r7, lsl #1 @ Pico.vram+((tilex&ts->xmask) as halfwords)\r |
585 | add r7, r7, r4, lsl #1\r |
586 | ldrh r7, [r7, r12] @ r7=code (int, but from unsigned, no sign extend)\r |
587 | \r |
588 | add r1, r1, #8\r |
589 | add r8, r8, #1\r |
590 | \r |
591 | tst r7, #0x8000\r |
592 | bne .DrawStrip_vs_hiprio\r |
593 | \r |
594 | cmp r7, r9\r |
595 | beq .DrawStrip_vs_samecode @ we know stuff about this tile already\r |
596 | \r |
597 | mov r9, r7 @ remember code\r |
7b802576 |
598 | orr r8, r8, #(1<<24)@ seen non hi-prio tile\r |
6d7acf9e |
599 | \r |
600 | movs r2, r9, lsl #20 @ if (code&0x1000)\r |
601 | mov r2, r2, lsl #1\r |
602 | add r2, r2, r10, lsl #17\r |
603 | mov r2, r2, lsr #17\r |
604 | eorcs r2, r2, #0x0e @ if (code&0x1000) addr^=0xe;\r |
605 | \r |
606 | ldr r2, [lr, r2, lsl #1] @ pack=*(unsigned int *)(Pico.vram+addr); // Get 8 pixels\r |
607 | \r |
608 | bic r7, r3, #0x3f\r |
609 | and r3, r9, #0x6000\r |
610 | add r3, r7, r3, lsr #9 @ r3=pal=((code&0x6000)>>9);\r |
611 | \r |
612 | .DrawStrip_vs_samecode:\r |
613 | tst r2, r2\r |
614 | beq .dsloop_vs @ tileline blank\r |
615 | \r |
616 | cmp r2, r2, ror #4\r |
617 | beq .DrawStrip_vs_SingleColor @ tileline singlecolor \r |
618 | \r |
619 | tst r9, #0x0800\r |
7a7c6476 |
620 | bne .DrawStrip_vs_TileFlip\r |
6d7acf9e |
621 | \r |
622 | @ (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r0: helper pattern\r |
6d7acf9e |
623 | .DrawStrip_vs_TileNorm:\r |
624 | TileNorm r0\r |
625 | b .dsloop_vs\r |
626 | \r |
7a7c6476 |
627 | .DrawStrip_vs_TileFlip:\r |
628 | TileFlip r0\r |
629 | b .dsloop_vs\r |
630 | \r |
6d7acf9e |
631 | .DrawStrip_vs_SingleColor:\r |
632 | and r4, r2, #0xf\r |
633 | orr r4, r3, r4\r |
634 | orr r4, r4, r4, lsl #8\r |
635 | tst r1, #1 @ not aligned?\r |
636 | strneb r4, [r1], #1\r |
637 | streqh r4, [r1], #2\r |
638 | strh r4, [r1], #2\r |
639 | strh r4, [r1], #2\r |
640 | strh r4, [r1], #2\r |
641 | strneb r4, [r1], #1 @ have a remaining unaligned pixel?\r |
642 | b .dsloop_vs_subr1\r |
643 | \r |
644 | .DrawStrip_vs_hiprio:\r |
645 | tst r10, #0x00c00000\r |
646 | beq .DrawStrip_vs_hiprio_maybempt\r |
647 | sub r0, r1, r11\r |
648 | orr r7, r7, r0, lsl #16\r |
649 | orr r7, r7, r10, lsl #25 @ (ty<<25)\r |
650 | tst r7, #0x1000\r |
651 | eorne r7, r7, #7<<26 @ if(code&0x1000) cval^=7<<26;\r |
652 | str r7, [r6], #4 @ cache hi priority tile\r |
653 | mov r0, #0xf\r |
654 | b .dsloop_vs\r |
655 | \r |
656 | .DrawStrip_vs_hiprio_maybempt:\r |
657 | cmp r7, r9\r |
658 | beq .dsloop_vs @ must've been empty, otherwise we wouldn't get here\r |
659 | movs r2, r7, lsl #20 @ if (code&0x1000)\r |
660 | mov r2, r2, lsl #1\r |
661 | add r2, r2, r10, lsl #17\r |
662 | mov r2, r2, lsr #17\r |
663 | eorcs r2, r2, #0x0e @ if (code&0x1000) addr^=0xe;\r |
664 | ldr r2, [lr, r2, lsl #1] @ pack=*(unsigned int *)(Pico.vram+addr); // Get 8 pixels\r |
665 | mov r9, r7 @ remember code\r |
666 | tst r2, r2\r |
667 | orrne r10, r10, #1<<22\r |
668 | bne .DrawStrip_vs_hiprio\r |
669 | b .dsloop_vs\r |
670 | \r |
740da8c6 |
671 | .dsloop_vs_exit:\r |
7b802576 |
672 | tst r8, #(1<<24) @ seen non hi-prio tile\r |
740da8c6 |
673 | ldreq r1, =rendstatus\r |
674 | mov r0, #0\r |
675 | ldreq r2, [r1]\r |
676 | str r0, [r6] @ terminate the cache list\r |
677 | orreq r2, r2, #0x40 @ had a layer with all hi-prio tiles\r |
678 | streq r2, [r1]\r |
679 | \r |
680 | ldmfd sp!, {r4-r11,lr}\r |
681 | bx lr\r |
682 | \r |
6d7acf9e |
683 | \r |
684 | @ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r |
cc68a136 |
685 | \r |
686 | @ interlace mode 2? Sonic 2?\r |
687 | .DrawStrip_interlace:\r |
688 | tst r0, r0\r |
689 | moveq r7, r7, lsl #21\r |
690 | movne r7, r7, lsl #5\r |
691 | \r |
692 | @ Find the line in the name table\r |
693 | add r2, r7, r2, lsl #22 @ r2=(vscroll+(Scanline<<1))<<21 (11 bits);\r |
694 | orr r1, r1, #0x80000000\r |
695 | and r2, r2, r1, ror #10 @ &((ymask<<1)|1)<<21;\r |
696 | mov r2, r2, lsr #21\r |
697 | mov r4, r2, lsr #4\r |
698 | mov r12, r12, lsr #1 @ halfwords\r |
699 | add r0, r12, r4, lsl r10 @ nametab+=(ts.line>>4)<<shift[width];\r |
700 | and r9, r9, #0xff\r |
701 | \r |
702 | sub sp, sp, #6*4\r |
703 | stmia sp, {r0,r2,r3,r5,r6,r9}\r |
704 | \r |
705 | mov r0, sp\r |
706 | bl DrawStripInterlace @ struct TileStrip *ts\r |
707 | \r |
708 | add sp, sp, #6*4\r |
709 | ldmfd sp!, {r4-r11,lr}\r |
710 | bx lr\r |
711 | \r |
712 | .pool\r |
713 | \r |
714 | @ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r |
715 | \r |
716 | \r |
717 | .global BackFill @ int reg7, int sh\r |
718 | \r |
719 | BackFill:\r |
720 | stmfd sp!, {r4-r9,lr}\r |
721 | \r |
ea8c405f |
722 | .if OVERRIDE_HIGHCOL\r |
723 | ldr lr, =HighCol\r |
724 | mov r0, r0, lsl #26\r |
725 | ldr lr, [lr]\r |
726 | mov r0, r0, lsr #26\r |
727 | add lr, lr, #8\r |
728 | .else\r |
cc68a136 |
729 | ldr lr, =(HighCol+8)\r |
cc68a136 |
730 | mov r0, r0, lsl #26\r |
731 | mov r0, r0, lsr #26\r |
ea8c405f |
732 | .endif\r |
733 | \r |
cc68a136 |
734 | orr r0, r0, r1, lsl #6\r |
735 | orr r0, r0, r0, lsl #8\r |
736 | orr r0, r0, r0, lsl #16\r |
737 | \r |
738 | mov r1, r0\r |
739 | mov r2, r0\r |
740 | mov r3, r0\r |
741 | mov r4, r0\r |
742 | mov r5, r0\r |
743 | mov r6, r0\r |
744 | mov r7, r0\r |
745 | \r |
746 | @ go go go!\r |
747 | stmia lr!, {r0-r7} @ 10*8*4\r |
748 | stmia lr!, {r0-r7}\r |
749 | stmia lr!, {r0-r7}\r |
750 | stmia lr!, {r0-r7}\r |
751 | stmia lr!, {r0-r7}\r |
752 | stmia lr!, {r0-r7}\r |
753 | stmia lr!, {r0-r7}\r |
754 | stmia lr!, {r0-r7}\r |
755 | stmia lr!, {r0-r7}\r |
756 | stmia lr!, {r0-r7}\r |
757 | \r |
758 | ldmfd sp!, {r4-r9,r12}\r |
759 | bx r12\r |
760 | \r |
761 | \r |
762 | @ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r |
763 | \r |
764 | \r |
7a7c6476 |
765 | .global DrawTilesFromCache @ int *hc, int sh, int rlim\r |
cc68a136 |
766 | \r |
767 | DrawTilesFromCache:\r |
768 | stmfd sp!, {r4-r8,r11,lr}\r |
769 | \r |
cc68a136 |
770 | @ cache some stuff to avoid mem access\r |
ea8c405f |
771 | .if OVERRIDE_HIGHCOL\r |
cc68a136 |
772 | ldr r11,=HighCol\r |
cc68a136 |
773 | mov r12,#0xf\r |
ea8c405f |
774 | ldr r11,[r11]\r |
775 | .else\r |
776 | ldr r11,=HighCol\r |
777 | mov r12,#0xf\r |
778 | .endif\r |
779 | ldr lr, =(Pico+0x10000) @ lr=Pico.vram\r |
cc68a136 |
780 | \r |
740da8c6 |
781 | mvn r5, #0 @ r5=prevcode=-1\r |
7a7c6476 |
782 | ands r8, r1, #1\r |
783 | orr r8, r8, r2, lsl #1\r |
740da8c6 |
784 | bne .dtfc_check_rendflags\r |
785 | \r |
cc68a136 |
786 | @ scratch: r4, r7\r |
787 | .dtfc_loop:\r |
788 | ldr r6, [r0], #4 @ read code\r |
789 | movs r1, r6, lsr #16 @ r1=dx;\r |
790 | ldmeqfd sp!, {r4-r8,r11,pc} @ dx is never zero, this must be a terminator, return\r |
7a7c6476 |
791 | bic r4, r1, #0xfe00\r |
792 | add r1, r11, r4 @ r1=pdest\r |
cc68a136 |
793 | \r |
794 | mov r7, r6, lsl #16\r |
795 | cmp r5, r7, lsr #16\r |
796 | beq .dtfc_samecode @ if (code==prevcode)\r |
797 | \r |
798 | mov r5, r7, lsr #16\r |
799 | \r |
800 | mov r2, r5, lsl #21\r |
801 | mov r2, r2, lsr #17 @ r2=addr=(code&0x7ff)<<4;\r |
802 | add r2, r2, r6, lsr #25 @ addr+=ty\r |
803 | \r |
804 | and r3, r5, #0x6000\r |
805 | mov r3, r3, lsr #9 @ r3=pal=((code&0x6000)>>9);\r |
806 | \r |
807 | ldr r2, [lr, r2, lsl #1] @ pack=*(unsigned int *)(Pico.vram+addr); // Get 8 pixels\r |
808 | \r |
809 | .dtfc_samecode:\r |
7a7c6476 |
810 | rsbs r4, r4, r8, lsr #1\r |
811 | bmi .dtfc_cut_tile\r |
812 | \r |
813 | tst r8, #1\r |
cc68a136 |
814 | bne .dtfc_shadow\r |
815 | \r |
816 | tst r2, r2\r |
817 | beq .dtfc_loop\r |
818 | \r |
819 | cmp r2, r2, ror #4\r |
820 | beq .dtfc_SingleColor @ tileline singlecolor \r |
821 | \r |
822 | tst r5, #0x0800\r |
7a7c6476 |
823 | bne .dtfc_TileFlip\r |
cc68a136 |
824 | \r |
825 | @ (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r12: helper pattern\r |
cc68a136 |
826 | .dtfc_TileNorm:\r |
827 | TileNorm r12\r |
828 | b .dtfc_loop\r |
829 | \r |
7a7c6476 |
830 | .dtfc_TileFlip:\r |
831 | TileFlip r12\r |
832 | b .dtfc_loop\r |
833 | \r |
cc68a136 |
834 | .dtfc_SingleColor:\r |
835 | and r4, r2, #0xf\r |
836 | orr r4, r3, r4\r |
837 | orr r4, r4, r4, lsl #8\r |
838 | tst r1, #1 @ not aligned?\r |
839 | strneb r4, [r1], #1\r |
840 | streqh r4, [r1], #2\r |
841 | strh r4, [r1], #2\r |
842 | strh r4, [r1], #2\r |
843 | strh r4, [r1], #2\r |
844 | strneb r4, [r1], #1 @ have a remaining unaligned pixel?\r |
845 | b .dtfc_loop\r |
846 | \r |
847 | .dtfc_shadow:\r |
848 | tst r2, r2\r |
849 | beq .dtfc_shadow_blank\r |
850 | \r |
851 | cmp r2, r2, ror #4\r |
852 | beq .dtfc_SingleColor @ tileline singlecolor \r |
853 | \r |
854 | tst r5, #0x0800\r |
7a7c6476 |
855 | bne .dtfc_TileFlipShHP\r |
cc68a136 |
856 | \r |
857 | @ (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r12: helper pattern\r |
cc68a136 |
858 | .dtfc_TileNormShHP:\r |
859 | TileNormShHP\r |
860 | b .dtfc_loop\r |
861 | \r |
7a7c6476 |
862 | .dtfc_TileFlipShHP:\r |
863 | TileFlipShHP\r |
864 | b .dtfc_loop\r |
865 | \r |
cc68a136 |
866 | .dtfc_shadow_blank:\r |
07abbab1 |
867 | tst r1, #1\r |
868 | ldrneb r4, [r1]\r |
869 | mov r6, #0x3f\r |
870 | and r4, r4, #0x3f\r |
871 | strneb r4, [r1], #1\r |
872 | ldrh r4, [r1]\r |
873 | orr r6, r6, r6, lsl #8\r |
874 | and r4, r4, r6\r |
875 | strh r4, [r1], #2\r |
876 | ldrh r4, [r1]\r |
877 | and r4, r4, r6\r |
878 | strh r4, [r1], #2\r |
879 | ldrh r4, [r1]\r |
880 | and r4, r4, r6\r |
881 | strh r4, [r1], #2\r |
882 | ldrh r4, [r1]\r |
883 | and r4, r4, r6\r |
884 | streqh r4, [r1]\r |
885 | strneb r4, [r1]\r |
cc68a136 |
886 | b .dtfc_loop\r |
887 | \r |
7a7c6476 |
888 | .dtfc_cut_tile:\r |
889 | add r4, r4, #7 @ 0-6\r |
890 | mov r4, r4, lsl #2\r |
891 | mov r12,#0xf<<28\r |
892 | mov r12,r12,asr r4\r |
893 | mov r2, r2, ror #16\r |
894 | tst r5, #0x0800 @ flipped?\r |
895 | mvnne r12,r12\r |
896 | and r2, r2, r12\r |
897 | mov r2, r2, ror #16\r |
898 | mov r12,#0xf\r |
899 | tst r8, #1\r |
900 | bne .dtfc_shadow\r |
901 | tst r2, r2\r |
902 | beq .dtfc_loop\r |
903 | tst r5, #0x0800\r |
904 | beq .dtfc_TileNorm\r |
905 | b .dtfc_TileFlip\r |
906 | \r |
740da8c6 |
907 | @ check if we have detected layer covered with hi-prio tiles:\r |
908 | .dtfc_check_rendflags:\r |
909 | ldr r1, =rendstatus\r |
910 | ldr r2, [r1]\r |
911 | tst r2, #0xc0\r |
912 | beq .dtfc_loop\r |
7a7c6476 |
913 | bic r8, r8, #1 @ sh/hi mode off\r |
740da8c6 |
914 | tst r2, #0x80\r |
915 | bne .dtfc_loop @ already processed\r |
916 | orr r2, r2, #0x80\r |
917 | str r2, [r1]\r |
918 | \r |
919 | add r1, r11,#8\r |
07abbab1 |
920 | mov r3, #320/4/4\r |
740da8c6 |
921 | mov r6, #0x3f\r |
922 | orr r6, r6, r6, lsl #8\r |
923 | orr r6, r6, r6, lsl #16\r |
924 | .dtfc_loop_shprep:\r |
07abbab1 |
925 | ldmia r1, {r2,r4,r5,r7}\r |
740da8c6 |
926 | subs r3, r3, #1\r |
07abbab1 |
927 | and r2, r2, r6\r |
928 | and r4, r4, r6\r |
929 | and r5, r5, r6\r |
930 | and r7, r7, r6\r |
931 | stmia r1!,{r2,r4,r5,r7}\r |
932 | bne .dtfc_loop_shprep\r |
933 | \r |
934 | mvn r5, #0 @ r5=prevcode=-1\r |
935 | b .dtfc_loop\r |
740da8c6 |
936 | \r |
cc68a136 |
937 | .pool\r |
938 | \r |
939 | @ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r |
940 | \r |
941 | \r |
e5fa9817 |
942 | .global DrawSpritesFromCache @ int *hc, int maxwidth, int prio, int sh\r |
cc68a136 |
943 | \r |
944 | DrawSpritesFromCache:\r |
945 | stmfd sp!, {r4-r11,lr}\r |
946 | \r |
947 | @ cache some stuff to avoid mem access\r |
ea8c405f |
948 | .if OVERRIDE_HIGHCOL\r |
cc68a136 |
949 | ldr r11,=HighCol\r |
ea8c405f |
950 | mov r12,#0xf\r |
951 | ldr r11,[r11]\r |
952 | .else\r |
953 | ldr r11,=HighCol\r |
954 | mov r12,#0xf\r |
955 | .endif\r |
cc68a136 |
956 | ldr lr, =(Pico+0x10000) @ lr=Pico.vram\r |
e5fa9817 |
957 | mov r6, r3, lsl #31\r |
cc68a136 |
958 | orr r6, r6, #1<<30\r |
cc68a136 |
959 | \r |
960 | mov r10, r0\r |
961 | \r |
962 | .dsfc_loop:\r |
963 | ldr r9, [r10], #4 @ read code\r |
964 | bic r6, r6, #7 @ using pipeline\r |
965 | tst r9, r9\r |
966 | ldmeqfd sp!, {r4-r11,pc}\r |
967 | \r |
968 | mov r4, r9, lsl #28\r |
969 | orr r6, r6, r4, lsr #30\r |
970 | add r6, r6, #1 @ r6=s1cc???? ... ?????www (s=shadow/hilight, cc=pal, w=width)\r |
971 | \r |
972 | and r5, r9, #3\r |
973 | add r5, r5, #1 @ r5=delta\r |
974 | tst r9, #0x10000\r |
975 | rsbne r5, r5, #0 @ Flip X\r |
976 | mov r5, r5, lsl #4\r |
977 | \r |
978 | mov r2, r9, lsr #17\r |
979 | mov r8, r2, lsl #1 @ tile=((unsigned int)code>>17)<<1;\r |
980 | \r |
981 | and r3, r9, #0x30 @ r3=pal=(code&0x30);\r |
982 | \r |
983 | bic r6, r6, #3<<28\r |
984 | orr r6, r6, r3, lsl #24\r |
985 | \r |
986 | mov r0, r9, lsl #16\r |
987 | mov r0, r0, asr #22 @ sx=(code<<16)>>22\r |
988 | adds r0, r0, #0 @ set ZV\r |
989 | b .dsfc_inloop_enter\r |
990 | \r |
991 | @ scratch: r4, r7\r |
992 | .dsfc_inloop:\r |
993 | sub r6, r6, #1\r |
994 | tst r6, #7\r |
995 | beq .dsfc_loop\r |
996 | adds r0, r0, #8\r |
997 | add r8, r8, r5\r |
998 | \r |
999 | .dsfc_inloop_enter:\r |
1000 | ble .dsfc_inloop\r |
1001 | cmp r0, #328\r |
1002 | bge .dsfc_loop\r |
1003 | \r |
1004 | mov r8, r8, lsl #17\r |
1005 | mov r8, r8, lsr #17 @ tile&=0x7fff; // Clip tile address\r |
1006 | \r |
1007 | ldr r2, [lr, r8, lsl #1] @ pack=*(unsigned int *)(Pico.vram+tile); // Get 8 pixels\r |
1008 | add r1, r11, r0 @ r1=pdest\r |
1009 | tst r2, r2\r |
1010 | beq .dsfc_inloop\r |
1011 | \r |
1012 | cmp r12, r6, lsr #28\r |
1013 | beq .dsfc_shadow\r |
1014 | \r |
1015 | cmp r2, r2, ror #4\r |
1016 | beq .dsfc_SingleColor @ tileline singlecolor \r |
1017 | \r |
1018 | tst r9, #0x10000\r |
7a7c6476 |
1019 | bne .dsfc_TileFlip\r |
cc68a136 |
1020 | \r |
1021 | @ TileFlip (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r12: helper pattern\r |
cc68a136 |
1022 | .dsfc_TileNorm:\r |
1023 | TileNorm r12\r |
1024 | b .dsfc_inloop\r |
1025 | \r |
7a7c6476 |
1026 | .dsfc_TileFlip:\r |
1027 | TileFlip r12\r |
1028 | b .dsfc_inloop\r |
1029 | \r |
cc68a136 |
1030 | .dsfc_SingleColor:\r |
1031 | tst r0, #1 @ not aligned?\r |
1032 | and r4, r2, #0xf\r |
1033 | orr r4, r3, r4\r |
1034 | orr r4, r4, r4, lsl #8\r |
1035 | strneb r4, [r1], #1\r |
1036 | streqh r4, [r1], #2\r |
1037 | strh r4, [r1], #2\r |
1038 | strh r4, [r1], #2\r |
1039 | strh r4, [r1], #2\r |
1040 | strneb r4, [r1], #1\r |
1041 | b .dsfc_inloop\r |
1042 | \r |
1043 | .dsfc_shadow:\r |
07abbab1 |
1044 | tst r9, #0x80000000\r |
1045 | beq .dsfc_shadow_lowpri\r |
1046 | \r |
cc68a136 |
1047 | cmp r2, r2, ror #4\r |
1048 | beq .dsfc_singlec_sh\r |
1049 | \r |
1050 | tst r9, #0x10000\r |
7a7c6476 |
1051 | bne .dsfc_TileFlip_sh\r |
cc68a136 |
1052 | \r |
1053 | @ (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r12: helper pattern\r |
cc68a136 |
1054 | .dsfc_TileNorm_sh:\r |
1055 | TileNormSh\r |
1056 | b .dsfc_inloop\r |
1057 | \r |
7a7c6476 |
1058 | .dsfc_TileFlip_sh:\r |
1059 | TileFlipSh\r |
1060 | b .dsfc_inloop\r |
1061 | \r |
cc68a136 |
1062 | .dsfc_singlec_sh:\r |
1063 | cmp r2, #0xe0000000\r |
1064 | bcc .dsfc_SingleColor @ normal singlecolor tileline (carry inverted in ARM)\r |
1065 | tst r2, #0x10000000\r |
1066 | bne .dsfc_sh_sh\r |
1067 | TileSingleHi\r |
1068 | b .dsfc_inloop\r |
1069 | \r |
1070 | .dsfc_sh_sh:\r |
1071 | TileSingleSh\r |
1072 | b .dsfc_inloop\r |
1073 | \r |
07abbab1 |
1074 | .dsfc_shadow_lowpri:\r |
1075 | tst r9, #0x10000\r |
1076 | bne .dsfc_TileFlip_sh_lp\r |
1077 | \r |
1078 | .dsfc_TileNorm_sh_lp:\r |
1079 | TileNormSh_onlyop_lp\r |
1080 | b .dsfc_inloop\r |
1081 | \r |
1082 | .dsfc_TileFlip_sh_lp:\r |
1083 | TileFlipSh_onlyop_lp\r |
1084 | b .dsfc_inloop\r |
1085 | \r |
cc68a136 |
1086 | .pool\r |
1087 | \r |
1088 | @ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r |
1089 | \r |
1090 | @ + 0 : hhhhvvvv ab--hhvv yyyyyyyy yyyyyyyy // a: offscreen h, b: offs. v, h: horiz. size\r |
1091 | @ + 4 : xxxxxxxx xxxxxxxx pccvhnnn nnnnnnnn // x: x coord + 8\r |
1092 | \r |
07abbab1 |
1093 | .global DrawSprite @ unsigned int *sprite, int sh, int acc_sprites\r |
cc68a136 |
1094 | \r |
1095 | DrawSprite:\r |
1096 | stmfd sp!, {r4-r9,r11,lr}\r |
1097 | \r |
07abbab1 |
1098 | orr r8, r2, r1, lsl #4\r |
cc68a136 |
1099 | ldr r3, [r0] @ sprite[0]\r |
1100 | ldr r7, =Scanline\r |
1101 | mov r6, r3, lsr #28\r |
1102 | sub r6, r6, #1 @ r6=width-1 (inc later)\r |
1103 | mov r5, r3, lsr #24\r |
1104 | and r5, r5, #7 @ r5=height\r |
1105 | \r |
1106 | mov r4, r3, lsl #16 @ r4=sy<<16 (tmp)\r |
1107 | \r |
1108 | ldr r7, [r7]\r |
1109 | ldr r9, [r0, #4]\r |
1110 | sub r7, r7, r4, asr #16 @ r7=row=Scanline-sy\r |
1111 | \r |
cc68a136 |
1112 | mov r2, r9, asr #16 @ r2=sx\r |
e5fa9817 |
1113 | mov r9, r9, lsl #16\r |
1114 | mov r9, r9, lsr #16\r |
1115 | orr r9, r9, r8, lsl #27 @ r9=code|sh[31]|as[27]\r |
cc68a136 |
1116 | \r |
1117 | tst r9, #0x1000\r |
1118 | movne r4, r5, lsl #3\r |
1119 | subne r4, r4, #1\r |
1120 | subne r7, r4, r7 @ if (code&0x1000) row=(height<<3)-1-row; // Flip Y\r |
1121 | \r |
07abbab1 |
1122 | add r8, r9, r7, lsr #3 @ tile+=row>>3; // Tile number increases going down\r |
cc68a136 |
1123 | tst r9, #0x0800\r |
1124 | mlane r8, r5, r6, r8 @ if (code&0x0800) { tile+=delta*(width-1);\r |
1125 | rsbne r5, r5, #0 @ delta=-delta; } // r5=delta now\r |
1126 | \r |
07abbab1 |
1127 | mov r8, r8, lsl #21\r |
1128 | mov r8, r8, lsr #17\r |
cc68a136 |
1129 | and r7, r7, #7\r |
1130 | add r8, r8, r7, lsl #1 @ tile+=(row&7)<<1; // Tile address\r |
1131 | \r |
1132 | tst r9, #0x8000\r |
07abbab1 |
1133 | tsteq r9, #(1<<27)\r |
1134 | bne .dspr_cache @ if(code&0x8000) || as\r |
1135 | tst r6, #0x4000\r |
1136 | tstne r6, #0x2000\r |
1137 | tstne r9, #(1<<31)\r |
1138 | bne .dspr_cache @ (sh && pal == 0x30)\r |
cc68a136 |
1139 | \r |
e5fa9817 |
1140 | .dspr_continue:\r |
cc68a136 |
1141 | @ cache some stuff to avoid mem access\r |
ea8c405f |
1142 | .if OVERRIDE_HIGHCOL\r |
1143 | ldr r11,=HighCol\r |
1144 | mov r12,#0xf\r |
1145 | ldr r11,[r11]\r |
1146 | .else\r |
cc68a136 |
1147 | ldr r11,=HighCol\r |
cc68a136 |
1148 | mov r12,#0xf\r |
ea8c405f |
1149 | .endif\r |
1150 | ldr lr, =(Pico+0x10000) @ lr=Pico.vram\r |
cc68a136 |
1151 | \r |
1152 | mov r5, r5, lsl #4 @ delta<<=4; // Delta of address\r |
1153 | and r4, r9, #0x6000\r |
1154 | orr r9, r9, r4, lsl #16\r |
e5fa9817 |
1155 | orrs r9, r9, #0x10000000 @ r9=scc1 a??? ... <code> (s=shadow/hilight, cc=pal, a=acc_spr)\r |
cc68a136 |
1156 | \r |
cc68a136 |
1157 | mov r3, r4, lsr #9 @ r3=pal=((code>>9)&0x30);\r |
e5fa9817 |
1158 | orrmi r3, r3, #0x40 @ shadow by default\r |
cc68a136 |
1159 | \r |
1160 | add r6, r6, #1 @ inc now\r |
1161 | adds r0, r2, #0 @ mov sx to r0 and set ZV flags\r |
1162 | b .dspr_loop_enter\r |
1163 | \r |
1164 | .dspr_loop:\r |
1165 | subs r6, r6, #1 @ width--\r |
1166 | ldmeqfd sp!, {r4-r9,r11,pc}@ return\r |
1167 | adds r0, r0, #8 @ sx+=8\r |
1168 | add r8, r8, r5 @ tile+=delta\r |
1169 | \r |
1170 | .dspr_loop_enter:\r |
1171 | ble .dspr_loop @ sx <= 0\r |
1172 | cmp r0, #328\r |
1173 | ldmgefd sp!, {r4-r9,r11,pc}@ return\r |
1174 | \r |
1175 | mov r8, r8, lsl #17\r |
1176 | mov r8, r8, lsr #17 @ tile&=0x7fff; // Clip tile address\r |
1177 | \r |
1178 | ldr r2, [lr, r8, lsl #1] @ pack=*(unsigned int *)(Pico.vram+addr); // Get 8 pixels\r |
1179 | add r1, r11, r0 @ r1=pdest\r |
1180 | tst r2, r2\r |
1181 | beq .dspr_loop\r |
1182 | \r |
1183 | cmp r12, r9, lsr #28\r |
1184 | beq .dspr_shadow\r |
1185 | \r |
1186 | cmp r2, r2, ror #4\r |
1187 | beq .dspr_SingleColor @ tileline singlecolor \r |
1188 | \r |
1189 | tst r9, #0x0800\r |
7a7c6476 |
1190 | bne .dspr_TileFlip\r |
cc68a136 |
1191 | \r |
1192 | @ (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r12: helper pattern\r |
cc68a136 |
1193 | @ scratch: r4, r7\r |
1194 | .dspr_TileNorm:\r |
1195 | TileNorm r12\r |
1196 | b .dspr_loop\r |
1197 | \r |
7a7c6476 |
1198 | .dspr_TileFlip:\r |
1199 | TileFlip r12\r |
1200 | b .dspr_loop\r |
1201 | \r |
07abbab1 |
1202 | .dspr_singlec_sh:\r |
1203 | cmp r2, #0xe0000000\r |
1204 | bcs .dspr_loop @ operator tileline, ignore\r |
1205 | \r |
cc68a136 |
1206 | .dspr_SingleColor:\r |
1207 | and r4, r2, #0xf\r |
1208 | orr r4, r3, r4\r |
1209 | orr r4, r4, r4, lsl #8\r |
1210 | tst r0, #1 @ not aligned?\r |
1211 | strneb r4, [r1], #1\r |
1212 | streqh r4, [r1], #2\r |
1213 | strh r4, [r1], #2\r |
1214 | strh r4, [r1], #2\r |
1215 | strh r4, [r1], #2\r |
1216 | strneb r4, [r1], #1\r |
1217 | b .dspr_loop\r |
1218 | \r |
1219 | .dspr_shadow:\r |
1220 | cmp r2, r2, ror #4\r |
1221 | beq .dspr_singlec_sh\r |
1222 | \r |
1223 | tst r9, #0x0800\r |
7a7c6476 |
1224 | bne .dspr_TileFlip_sh\r |
cc68a136 |
1225 | \r |
1226 | @ (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r12: helper pattern\r |
cc68a136 |
1227 | .dspr_TileNorm_sh:\r |
07abbab1 |
1228 | TileNormSh_noop\r |
cc68a136 |
1229 | b .dspr_loop\r |
1230 | \r |
7a7c6476 |
1231 | .dspr_TileFlip_sh:\r |
07abbab1 |
1232 | TileFlipSh_noop\r |
cc68a136 |
1233 | b .dspr_loop\r |
1234 | \r |
1235 | \r |
1236 | .dspr_cache:\r |
07abbab1 |
1237 | @ *HighCacheS_ptr++ = ((code&0x8000)<<16)|(tile<<16)|((code&0x0800)<<5)|((sx<<6)&0x0000ffc0)|pal|((sprite[0]>>16)&0xf);\r |
1238 | ldr r1, =HighCacheS_ptr\r |
cc68a136 |
1239 | mov r4, r8, lsl #16 @ tile\r |
1240 | tst r9, #0x0800\r |
1241 | orrne r4, r4, #0x10000 @ code&0x0800\r |
e5fa9817 |
1242 | mov r0, r2, lsl #22\r |
1243 | orr r4, r4, r0, lsr #16 @ (sx<<6)&0x0000ffc0\r |
1244 | and r0, r9, #0x6000\r |
1245 | orr r4, r4, r0, lsr #9 @ (code>>9)&0x30\r |
cc68a136 |
1246 | mov r3, r3, lsl #12\r |
cc68a136 |
1247 | orr r4, r4, r3, lsr #28 @ (sprite[0]>>24)&0xf\r |
1248 | \r |
07abbab1 |
1249 | ldr r0, [r1]\r |
1250 | tst r9, #0x8000\r |
1251 | orrne r4, r4, #0x80000000 @ prio\r |
1252 | \r |
e5fa9817 |
1253 | str r4, [r0], #4\r |
1254 | str r0, [r1]\r |
cc68a136 |
1255 | \r |
07abbab1 |
1256 | and r0, r9, #(1<<27) @ as\r |
1257 | teqne r0, #(1<<27) @ (code&0x8000) && !as\r |
1258 | ldmnefd sp!, {r4-r9,r11,pc}\r |
1259 | b .dspr_continue @ draw anyway if accurate sprites enabled\r |
cc68a136 |
1260 | \r |
1261 | @ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r |
1262 | \r |
1263 | .global DrawWindow @ int tstart, int tend, int prio, int sh // int *hcache\r |
1264 | \r |
1265 | DrawWindow:\r |
1266 | stmfd sp!, {r4-r11,lr}\r |
1267 | \r |
1268 | ldr r11, =(Pico+0x22228) @ Pico.video\r |
1269 | ldr r10, =Scanline\r |
1270 | ldrb r12, [r11, #3] @ pvid->reg[3]\r |
1271 | \r |
1272 | ldr r10, [r10]\r |
1273 | ldr r4, [r11, #12]\r |
1274 | mov r5, r10, lsr #3\r |
1275 | and r10, r10, #7\r |
1276 | mov r10, r10, lsl #1 @ r10=ty\r |
1277 | \r |
1278 | mov r12, r12, lsl #10\r |
1279 | \r |
1280 | tst r4, #1 @ 40 cell mode?\r |
1281 | andne r12, r12, #0xf000 @ 0x3c<<10\r |
1282 | andeq r12, r12, #0xf800\r |
1283 | addne r12, r12, r5, lsl #7\r |
1284 | addeq r12, r12, r5, lsl #6 @ nametab\r |
1285 | add r12, r12, r0, lsl #2 @ +starttile\r |
1286 | \r |
1287 | ldr r6, =rendstatus\r |
1288 | ldr lr, =(Pico+0x10000) @ lr=Pico.vram\r |
07abbab1 |
1289 | ldr r6, [r6]\r |
cc68a136 |
1290 | \r |
1291 | @ fetch the first code now\r |
1292 | ldrh r7, [lr, r12]\r |
1293 | \r |
1294 | ands r6, r6, #2 @ we care about bit 1 only\r |
1295 | orr r6, r6, r2\r |
cc68a136 |
1296 | \r |
07abbab1 |
1297 | teqne r2, r7, lsr #15 @ do prio bits differ?\r |
1298 | ldmnefd sp!, {r4-r11,pc} @ yes, assume that whole window uses same priority\r |
cc68a136 |
1299 | \r |
cc68a136 |
1300 | orr r6, r6, r3, lsl #8 @ shadow mode\r |
1301 | \r |
1302 | sub r8, r1, r0\r |
cc68a136 |
1303 | \r |
1304 | @ cache some stuff to avoid mem access\r |
ea8c405f |
1305 | .if OVERRIDE_HIGHCOL\r |
1306 | ldr r11,=HighCol\r |
1307 | mov r8, r8, lsl #1 @ cells\r |
1308 | ldr r11,[r11]\r |
1309 | mvn r9, #0 @ r9=prevcode=-1\r |
1310 | add r11,r11,#8\r |
1311 | .else\r |
cc68a136 |
1312 | ldr r11,=(HighCol+8)\r |
ea8c405f |
1313 | mov r8, r8, lsl #1 @ cells\r |
1314 | mvn r9, #0 @ r9=prevcode=-1\r |
1315 | .endif\r |
07abbab1 |
1316 | add r1, r11, r0, lsl #4 @ r1=pdest\r |
cc68a136 |
1317 | mov r0, #0xf\r |
1318 | b .dwloop_enter\r |
1319 | \r |
07abbab1 |
1320 | @ r4,r5 are scratch in this loop\r |
cc68a136 |
1321 | .dwloop:\r |
1322 | add r1, r1, #8\r |
1323 | .dwloop_nor1:\r |
1324 | add r12, r12, #2 @ halfwords\r |
1325 | ldrh r7, [lr, r12] @ r7=code (int, but from unsigned, no sign extend)\r |
1326 | subs r8, r8, #1\r |
1327 | beq .dwloop_end @ done\r |
1328 | \r |
1329 | eor r5, r6, r7, lsr #15\r |
1330 | tst r5, #1\r |
1331 | orrne r6, r6, #2 @ wrong pri\r |
1332 | bne .dwloop\r |
1333 | \r |
1334 | cmp r7, r9\r |
1335 | beq .dw_samecode @ we know stuff about this tile already\r |
1336 | \r |
1337 | .dwloop_enter:\r |
1338 | mov r9, r7 @ remember code\r |
1339 | \r |
1340 | movs r2, r9, lsl #20 @ if (code&0x1000)\r |
1341 | mov r2, r2, lsl #1\r |
1342 | add r2, r10, r2, lsr #17 @ r2=addr=(code&0x7ff)<<4; addr+=ty\r |
1343 | eorcs r2, r2, #0x0e @ if (code&0x1000) addr^=0xe;\r |
1344 | \r |
1345 | and r3, r9, #0x6000\r |
1346 | mov r3, r3, lsr #9 @ r3=pal=((code&0x6000)>>9);\r |
1347 | \r |
1348 | ldr r2, [lr, r2, lsl #1] @ pack=*(unsigned int *)(Pico.vram+addr); // Get 8 pixels\r |
1349 | \r |
1350 | .dw_samecode:\r |
1351 | tst r6, #0x100\r |
1352 | bne .dw_shadow\r |
1353 | .dw_shadow_done:\r |
1354 | tst r2, r2\r |
1355 | beq .dwloop @ tileline blank\r |
1356 | \r |
1357 | cmp r2, r2, ror #4\r |
1358 | beq .dw_SingleColor @ tileline singlecolor \r |
1359 | \r |
1360 | tst r9, #0x0800\r |
7a7c6476 |
1361 | bne .dw_TileFlip\r |
cc68a136 |
1362 | \r |
1363 | @ (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r0: helper pattern\r |
cc68a136 |
1364 | .dw_TileNorm:\r |
1365 | TileNorm r0\r |
1366 | b .dwloop\r |
1367 | \r |
7a7c6476 |
1368 | .dw_TileFlip:\r |
1369 | TileFlip r0\r |
1370 | b .dwloop\r |
1371 | \r |
cc68a136 |
1372 | .dw_SingleColor:\r |
1373 | and r4, r0, r2 @ #0x0000000f\r |
1374 | orr r4, r3, r4\r |
1375 | orr r4, r4, r4, lsl #8\r |
1376 | orr r4, r4, r4, lsl #16\r |
1377 | mov r5, r4\r |
1378 | stmia r1!, {r4,r5}\r |
1379 | b .dwloop_nor1 @ we incremeted r1 ourselves\r |
1380 | \r |
1381 | .dw_shadow:\r |
1382 | tst r6, #1 @ hi pri?\r |
1383 | orreq r3, r3, #0x40\r |
1384 | beq .dw_shadow_done\r |
1385 | ldr r4, [r1]\r |
07abbab1 |
1386 | mov r5, #0x3f\r |
1387 | orr r5, r5, r5, lsl #8\r |
1388 | orr r5, r5, r5, lsl #16\r |
1389 | and r4, r4, r5\r |
cc68a136 |
1390 | str r4, [r1]\r |
1391 | ldr r4, [r1,#4]\r |
07abbab1 |
1392 | and r4, r4, r5\r |
cc68a136 |
1393 | str r4, [r1,#4]\r |
1394 | b .dw_shadow_done\r |
1395 | \r |
1396 | .dwloop_end:\r |
1397 | ldr r0, =rendstatus\r |
1398 | ldr r1, [r0]\r |
1399 | and r6, r6, #2\r |
1400 | orr r1, r1, r6\r |
1401 | str r1, [r0]\r |
1402 | \r |
1403 | ldmfd sp!, {r4-r11,r12}\r |
1404 | bx r12\r |
1405 | \r |
1406 | \r |
1407 | @ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r |
1408 | \r |
1409 | \r |
1410 | @ hilights 2 pixels in RGB444/BGR444 format\r |
1411 | .macro TileDoShHi2Pixels444 reg\r |
1412 | mov \reg, \reg, ror #12\r |
1413 | adds \reg, \reg, #0x40000000\r |
1414 | orrcs \reg, \reg, #0xf0000000\r |
1415 | mov \reg, \reg, ror #28\r |
1416 | adds \reg, \reg, #0x40000000\r |
1417 | orrcs \reg, \reg, #0xf0000000\r |
1418 | mov \reg, \reg, ror #28\r |
1419 | adds \reg, \reg, #0x40000000\r |
1420 | orrcs \reg, \reg, #0xf0000000\r |
1421 | mov \reg, \reg, ror #24\r |
1422 | adds \reg, \reg, #0x40000000\r |
1423 | orrcs \reg, \reg, #0xf0000000\r |
1424 | mov \reg, \reg, ror #28\r |
1425 | adds \reg, \reg, #0x40000000\r |
1426 | orrcs \reg, \reg, #0xf0000000\r |
1427 | mov \reg, \reg, ror #28\r |
1428 | adds \reg, \reg, #0x40000000\r |
1429 | orrcs \reg, \reg, #0xf0000000\r |
1430 | mov \reg, \reg, ror #12\r |
1431 | .endm\r |
1432 | \r |
1433 | \r |
1434 | .global FinalizeLineBGR444 @ int sh\r |
1435 | \r |
1436 | FinalizeLineBGR444:\r |
1437 | stmfd sp!, {r4-r6,lr}\r |
1438 | mov r6, r0\r |
1439 | ldr lr, =(Pico+0x22228) @ Pico.video\r |
1440 | ldr r0, =DrawLineDest\r |
1441 | ldrb r12, [lr, #12]\r |
1442 | ldr r0, [r0]\r |
1443 | sub r3, lr, #0x128 @ r3=Pico.cram\r |
1444 | \r |
1445 | tst r12, #1\r |
1446 | movne r2, #320/4 @ len\r |
1447 | bne .fl_no32colBGR444\r |
1448 | ldr r4, =PicoOpt\r |
1449 | mov r2, #256/4\r |
1450 | ldr r4, [r4]\r |
1451 | tst r4, #0x100\r |
1452 | addeq r0, r0, #32*2\r |
1453 | \r |
1454 | .fl_no32colBGR444:\r |
1455 | tst r6, r6\r |
1456 | beq .fl_noshBGR444\r |
1457 | \r |
1458 | ldr r4, =HighPal\r |
1459 | \r |
1460 | ldrb r12, [lr, #-0x1a] @ 0x2220e ~ dirtyPal\r |
1461 | tst r12, r12\r |
1462 | moveq r3, r4\r |
1463 | beq .fl_noshBGR444\r |
1464 | mov r12, #0\r |
1465 | strb r12, [lr, #-0x1a]\r |
1466 | \r |
1467 | mov lr, #0x40/8\r |
1468 | @ copy pal:\r |
1469 | .fl_loopcpBGR444:\r |
1470 | ldmia r3!, {r1,r5,r6,r12}\r |
1471 | subs lr, lr, #1\r |
1472 | stmia r4!, {r1,r5,r6,r12}\r |
1473 | bne .fl_loopcpBGR444\r |
1474 | \r |
1475 | @ shadowed pixels:\r |
1476 | mov r12, #0x0077\r |
1477 | orr r12,r12,#0x0700\r |
1478 | orr r12,r12,r12,lsl #16\r |
1479 | sub r3, r3, #0x40*2\r |
1480 | add r5, r4, #0x80*2\r |
1481 | mov lr, #0x40/4\r |
1482 | .fl_loopcpBGR444_sh:\r |
1483 | ldmia r3!, {r1,r6}\r |
1484 | subs lr, lr, #1\r |
1485 | and r1, r12, r1, lsr #1\r |
1486 | and r6, r12, r6, lsr #1\r |
1487 | stmia r4!, {r1,r6}\r |
1488 | stmia r5!, {r1,r6}\r |
1489 | bne .fl_loopcpBGR444_sh\r |
1490 | \r |
1491 | @ hilighted pixels:\r |
1492 | sub r3, r3, #0x40*2\r |
1493 | mov lr, #0x40/2\r |
1494 | .fl_loopcpBGR444_hi:\r |
1495 | ldr r1, [r3], #4\r |
1496 | TileDoShHi2Pixels444 r1\r |
1497 | str r1, [r4], #4\r |
1498 | subs lr, lr, #1\r |
1499 | bne .fl_loopcpBGR444_hi\r |
1500 | \r |
1501 | sub r3, r4, #0x40*3*2\r |
e5fa9817 |
1502 | mov r6, #1\r |
cc68a136 |
1503 | \r |
1504 | \r |
1505 | .fl_noshBGR444:\r |
e5fa9817 |
1506 | ldr r12,=rendstatus\r |
1507 | eors r6, r6, #1 @ sh is 0\r |
1508 | ldr r12,[r12]\r |
1509 | mov lr, #0xff\r |
1510 | tstne r12,#(1<<2) @ and PDRAW_ACC_SPRITES\r |
1511 | \r |
ea8c405f |
1512 | .if OVERRIDE_HIGHCOL\r |
1513 | ldr r1, =HighCol\r |
e5fa9817 |
1514 | movne lr, #0x3f\r |
ea8c405f |
1515 | ldr r1, [r1]\r |
1516 | mov lr, lr, lsl #1\r |
1517 | add r1, r1, #8\r |
1518 | .else\r |
cc68a136 |
1519 | ldr r1, =(HighCol+8)\r |
e5fa9817 |
1520 | movne lr, #0x3f\r |
cc68a136 |
1521 | mov lr, lr, lsl #1\r |
ea8c405f |
1522 | .endif\r |
cc68a136 |
1523 | \r |
1524 | .fl_loopBGR444:\r |
cc68a136 |
1525 | ldr r12, [r1], #4\r |
1526 | subs r2, r2, #1\r |
1527 | \r |
1528 | and r4, lr, r12, lsl #1\r |
1529 | ldrh r4, [r3, r4]\r |
1530 | and r5, lr, r12, lsr #7\r |
1531 | ldrh r5, [r3, r5]\r |
1532 | and r6, lr, r12, lsr #15\r |
1533 | ldrh r6, [r3, r6]\r |
e5fa9817 |
1534 | and r12,lr, r12, lsr #23\r |
1535 | ldrh r12,[r3, r12] @ 1c.i.\r |
cc68a136 |
1536 | orr r4, r4, r5, lsl #16\r |
e5fa9817 |
1537 | orr r5, r6, r12,lsl #16\r |
cc68a136 |
1538 | \r |
1539 | stmia r0!, {r4,r5}\r |
1540 | bne .fl_loopBGR444\r |
1541 | \r |
1542 | \r |
1543 | ldmfd sp!, {r4-r6,lr}\r |
1544 | bx lr\r |
1545 | \r |
1546 | \r |
1547 | @ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r |
1548 | \r |
1549 | \r |
1550 | @ hilights 2 pixels in RGB555/BGR555 format\r |
1551 | .macro TileDoShHi2Pixels555 reg\r |
1552 | adds \reg, \reg, #0x40000000\r |
1553 | orrcs \reg, \reg, #0xf0000000\r |
1554 | mov \reg, \reg, ror #27\r |
1555 | adds \reg, \reg, #0x40000000\r |
1556 | orrcs \reg, \reg, #0xf0000000\r |
1557 | mov \reg, \reg, ror #26\r |
1558 | adds \reg, \reg, #0x40000000\r |
1559 | orrcs \reg, \reg, #0xf0000000\r |
1560 | mov \reg, \reg, ror #27\r |
1561 | adds \reg, \reg, #0x40000000\r |
1562 | orrcs \reg, \reg, #0xf0000000\r |
1563 | mov \reg, \reg, ror #27\r |
1564 | adds \reg, \reg, #0x40000000\r |
1565 | orrcs \reg, \reg, #0xf0000000\r |
1566 | mov \reg, \reg, ror #26\r |
1567 | adds \reg, \reg, #0x40000000\r |
1568 | orrcs \reg, \reg, #0xf0000000\r |
1569 | mov \reg, \reg, ror #27\r |
1570 | .endm\r |
1571 | \r |
1572 | \r |
1573 | @ Convert 0000bbb0 ggg0rrr0\r |
1574 | @ to rrrrrggg gggbbbbb\r |
1575 | \r |
1576 | @ r2,r3,r9 - scratch, lr = 0x001c001c, r8 = 0x00030003\r |
1577 | .macro convRGB565 reg\r |
1578 | and r2, lr, \reg,lsl #1\r |
1579 | and r9, r8, \reg,lsr #2\r |
1580 | orr r2, r2, r9 @ r2=red\r |
1581 | and r3, lr, \reg,lsr #7\r |
1582 | and r9, r8, \reg,lsr #10\r |
1583 | orr r3, r3, r9 @ r3=blue\r |
1584 | and \reg, \reg, lr, lsl #3\r |
1585 | orr \reg, \reg, \reg,lsl #3 @ green\r |
1586 | orr \reg, \reg, r2, lsl #11 @ add red back\r |
1587 | orr \reg, \reg, r3 @ add blue back\r |
1588 | .endm\r |
1589 | \r |
ea8c405f |
1590 | .global vidConvCpyRGB565\r |
1591 | \r |
cc68a136 |
1592 | vidConvCpyRGB565: @ void *to, void *from, int pixels\r |
1593 | stmfd sp!, {r4-r9,lr}\r |
1594 | \r |
2ec14aec |
1595 | mov r12, r2, lsr #3 @ repeats\r |
cc68a136 |
1596 | mov lr, #0x001c0000\r |
1597 | orr lr, lr, #0x01c @ lr == pattern 0x001c001c\r |
1598 | mov r8, #0x00030000\r |
2ec14aec |
1599 | orr r8, r8, #0x003\r |
cc68a136 |
1600 | \r |
1601 | .loopRGB565:\r |
1602 | ldmia r1!, {r4-r7}\r |
1603 | subs r12, r12, #1\r |
1604 | convRGB565 r4\r |
1605 | str r4, [r0], #4\r |
1606 | convRGB565 r5\r |
1607 | str r5, [r0], #4\r |
1608 | convRGB565 r6\r |
1609 | str r6, [r0], #4\r |
1610 | convRGB565 r7\r |
1611 | str r7, [r0], #4\r |
1612 | \r |
1613 | bgt .loopRGB565\r |
1614 | \r |
1615 | ldmfd sp!, {r4-r9,lr}\r |
1616 | bx lr\r |
1617 | \r |
1618 | \r |
1619 | \r |
1620 | .global FinalizeLineRGB555 @ int sh\r |
1621 | \r |
1622 | FinalizeLineRGB555:\r |
1623 | stmfd sp!, {r4-r8,lr}\r |
3d48f143 |
1624 | ldr r8, =(Pico+0x22228) @ Pico.video\r |
cc68a136 |
1625 | ldr r4, =HighPal\r |
1626 | \r |
3d48f143 |
1627 | ldrb r7, [r8, #-0x1a] @ 0x2220e ~ dirtyPal\r |
cc68a136 |
1628 | mov r6, r0\r |
1629 | mov r1, #0\r |
1630 | tst r7, r7\r |
1631 | beq .fl_noconvRGB555\r |
3d48f143 |
1632 | strb r1, [r8, #-0x1a]\r |
1633 | sub r1, r8, #0x128 @ r1=Pico.cram\r |
cc68a136 |
1634 | mov r0, r4\r |
1635 | mov r2, #0x40\r |
1636 | bl vidConvCpyRGB565\r |
1637 | \r |
1638 | .fl_noconvRGB555:\r |
cc68a136 |
1639 | mov r3, r4\r |
1640 | tst r6, r6\r |
1641 | beq .fl_noshRGB555\r |
1642 | tst r7, r7\r |
1643 | beq .fl_noshRGB555\r |
1644 | \r |
1645 | @ shadowed pixels:\r |
1646 | mov r12, #0x008e\r |
1647 | orr r12,r12,#0x7300\r |
1648 | orr r12,r12,r12,lsl #16\r |
1649 | add r4, r3, #0x40*2\r |
1650 | add r5, r3, #0xc0*2\r |
1651 | mov lr, #0x40/4\r |
1652 | .fl_loopcpRGB555_sh:\r |
1653 | ldmia r3!, {r1,r6}\r |
1654 | subs lr, lr, #1\r |
1655 | and r1, r12, r1, lsr #1\r |
1656 | and r6, r12, r6, lsr #1\r |
1657 | stmia r4!, {r1,r6}\r |
1658 | stmia r5!, {r1,r6}\r |
1659 | bne .fl_loopcpRGB555_sh\r |
1660 | \r |
1661 | @ hilighted pixels:\r |
1662 | sub r3, r3, #0x40*2\r |
1663 | mov lr, #0x40/2\r |
1664 | .fl_loopcpRGB555_hi:\r |
1665 | ldr r1, [r3], #4\r |
1666 | TileDoShHi2Pixels555 r1\r |
1667 | str r1, [r4], #4\r |
1668 | subs lr, lr, #1\r |
1669 | bne .fl_loopcpRGB555_hi\r |
1670 | \r |
1671 | sub r3, r3, #0x40*2\r |
e5fa9817 |
1672 | mov r6, #1\r |
cc68a136 |
1673 | \r |
cc68a136 |
1674 | .fl_noshRGB555:\r |
e5fa9817 |
1675 | ldr r12,=rendstatus\r |
1676 | eors r6, r6, #1 @ sh is 0\r |
1677 | ldr r12,[r12]\r |
1678 | mov lr, #0xff\r |
1679 | tstne r12,#(1<<2) @ and PDRAW_ACC_SPRITES\r |
1680 | movne lr, #0x3f\r |
1681 | \r |
ea8c405f |
1682 | .if OVERRIDE_HIGHCOL\r |
1683 | ldr r1, =HighCol\r |
1684 | ldr r0, =DrawLineDest\r |
1685 | ldr r1, [r1]\r |
1686 | ldr r0, [r0]\r |
1687 | add r1, r1, #8\r |
1688 | .else\r |
3d48f143 |
1689 | ldr r0, =DrawLineDest\r |
cc68a136 |
1690 | ldr r1, =(HighCol+8)\r |
3d48f143 |
1691 | ldr r0, [r0]\r |
ea8c405f |
1692 | .endif\r |
3d48f143 |
1693 | \r |
1694 | ldrb r12, [r8, #12]\r |
cc68a136 |
1695 | mov lr, lr, lsl #1\r |
1696 | \r |
3d48f143 |
1697 | tst r12, #1\r |
1698 | movne r2, #320/8 @ len\r |
1699 | bne .fl_no32colRGB555\r |
1700 | ldr r4, =PicoOpt\r |
1701 | mov r2, #256/8\r |
1702 | ldr r4, [r4]\r |
1703 | tst r4, #0x4000\r |
1704 | bne .fl_32scale_RGB555\r |
1705 | tst r4, #0x0100\r |
1706 | addeq r0, r0, #32*2\r |
1707 | \r |
1708 | .fl_no32colRGB555:\r |
cc68a136 |
1709 | \r |
499a0be3 |
1710 | .if UNALIGNED_DRAWLINEDEST\r |
1711 | @ this is basically for Gizmondo, which has unaligned odd lines in the framebuffer\r |
1712 | tst r0, #2\r |
1713 | bne .fl_RGB555u\r |
1714 | .endif\r |
1715 | \r |
1716 | .fl_loopRGB555:\r |
cc68a136 |
1717 | ldr r12, [r1], #4\r |
1718 | ldr r7, [r1], #4\r |
1719 | \r |
1720 | and r4, lr, r12, lsl #1\r |
1721 | ldrh r4, [r3, r4]\r |
1722 | and r5, lr, r12, lsr #7\r |
1723 | ldrh r5, [r3, r5]\r |
1724 | and r6, lr, r12, lsr #15\r |
1725 | ldrh r6, [r3, r6]\r |
1726 | orr r4, r4, r5, lsl #16\r |
1727 | \r |
1728 | and r5, lr, r12, lsr #23\r |
1729 | ldrh r5, [r3, r5]\r |
1730 | and r8, lr, r7, lsl #1\r |
1731 | ldrh r8, [r3, r8]\r |
1732 | orr r5, r6, r5, lsl #16\r |
1733 | \r |
1734 | and r6, lr, r7, lsr #7\r |
1735 | ldrh r6, [r3, r6]\r |
1736 | and r12,lr, r7, lsr #15\r |
1737 | ldrh r12,[r3, r12]\r |
499a0be3 |
1738 | and r7, lr, r7, lsr #23\r |
1739 | ldrh r7, [r3, r7]\r |
cc68a136 |
1740 | orr r8, r8, r6, lsl #16\r |
1741 | \r |
cc68a136 |
1742 | subs r2, r2, #1\r |
499a0be3 |
1743 | orr r12,r12, r7, lsl #16\r |
cc68a136 |
1744 | \r |
1745 | stmia r0!, {r4,r5,r8,r12}\r |
1746 | bne .fl_loopRGB555\r |
1747 | \r |
3d48f143 |
1748 | ldmfd sp!, {r4-r8,lr}\r |
1749 | bx lr\r |
1750 | \r |
1751 | \r |
1752 | .fl_32scale_RGB555:\r |
1753 | stmfd sp!, {r9,r10}\r |
1754 | mov r9, #0x3900 @ f800 07e0 001f | e000 0780 001c | 3800 01e0 0007\r |
1755 | orr r9, r9, #0x00e7\r |
1756 | \r |
499a0be3 |
1757 | .if UNALIGNED_DRAWLINEDEST\r |
1758 | tst r0, #2\r |
1759 | bne .fl_32scale_RGB555u\r |
1760 | .endif\r |
1761 | \r |
3d48f143 |
1762 | .fl_loop32scale_RGB555:\r |
1763 | ldr r12, [r1], #4\r |
1764 | ldr r7, [r1], #4\r |
cc68a136 |
1765 | \r |
3d48f143 |
1766 | and r4, lr, r12,lsl #1\r |
1767 | ldrh r4, [r3, r4]\r |
1768 | and r5, lr, r12,lsr #7\r |
1769 | ldrh r5, [r3, r5]\r |
1770 | and r4, r4, r9, lsl #2\r |
1771 | orr r4, r4, r4, lsl #14 @ r4[31:16] = 1/4 pix_s 0\r |
1772 | and r5, r5, r9, lsl #2\r |
1773 | sub r6, r5, r5, lsr #2 @ r6 = 3/4 pix_s 1\r |
1774 | add r4, r4, r6, lsl #16 @ pix_d 0, 1\r |
1775 | and r6, lr, r12,lsr #15\r |
1776 | ldrh r6, [r3, r6]\r |
1777 | and r12,lr, r12,lsr #23\r |
1778 | ldrh r12,[r3, r12]\r |
1779 | and r6, r6, r9, lsl #2\r |
1780 | add r5, r5, r6\r |
1781 | mov r5, r5, lsr #1\r |
1782 | sub r6, r6, r6, lsr #2 @ r6 = 3/4 pix_s 2\r |
1783 | orr r5, r5, r6, lsl #16\r |
1784 | \r |
1785 | and r6, lr, r7, lsl #1\r |
1786 | ldrh r6, [r3, r6]\r |
1787 | and r12,r12,r9, lsl #2\r |
1788 | add r5, r5, r12,lsl #14 @ pix_d 2, 3\r |
1789 | and r6, r6, r9, lsl #2\r |
1790 | orr r6, r12,r6, lsl #16 @ pix_d 4, 5\r |
1791 | \r |
1792 | and r12,lr, r7, lsr #7\r |
1793 | ldrh r12,[r3, r12]\r |
1794 | and r10,lr, r7, lsr #15\r |
1795 | ldrh r10,[r3, r10]\r |
1796 | and r12,r12,r9, lsl #2\r |
1797 | sub r8, r12,r12,lsr #2 @ r8 = 3/4 pix_s 1\r |
1798 | add r8, r8, r6, lsr #18\r |
1799 | and r7, lr, r7, lsr #23\r |
1800 | ldrh r7, [r3, r7]\r |
1801 | and r10,r10,r9, lsl #2\r |
1802 | orr r8, r8, r10,lsl #15\r |
1803 | add r8, r8, r12,lsl #15 @ pix_d 6, 7\r |
1804 | sub r10,r10,r10,lsr #2 @ r10= 3/4 pix_s 2\r |
1805 | and r7, r7, r9, lsl #2\r |
1806 | add r10,r10,r7, lsr #2 @ += 1/4 pix_s 3\r |
1807 | orr r10,r10,r7, lsl #16 @ pix_d 8, 9\r |
1808 | \r |
1809 | subs r2, r2, #1\r |
1810 | \r |
1811 | stmia r0!, {r4,r5,r6,r8,r10}\r |
1812 | bne .fl_loop32scale_RGB555\r |
1813 | \r |
1814 | ldmfd sp!, {r9,r10}\r |
cc68a136 |
1815 | ldmfd sp!, {r4-r8,lr}\r |
3d48f143 |
1816 | bx lr\r |
1817 | \r |
499a0be3 |
1818 | .if UNALIGNED_DRAWLINEDEST\r |
1819 | @ unaligned versions of loops\r |
9839d126 |
1820 | @ warning: starts drawing 2bytes before dst\r |
499a0be3 |
1821 | \r |
1822 | .fl_RGB555u:\r |
9839d126 |
1823 | sub r0, r0, #2 @ initial adjustment\r |
1824 | mov r8, #0\r |
499a0be3 |
1825 | \r |
1826 | .fl_loopRGB555u:\r |
1827 | ldr r12, [r1], #4\r |
1828 | ldr r7, [r1], #4\r |
1829 | \r |
9839d126 |
1830 | and r6, lr, r12,lsl #1\r |
1831 | ldrh r6, [r3, r6]\r |
1832 | and r5, lr, r12,lsr #7\r |
499a0be3 |
1833 | ldrh r5, [r3, r5]\r |
9839d126 |
1834 | orr r4, r8, r6, lsl #16\r |
499a0be3 |
1835 | \r |
9839d126 |
1836 | and r6, lr, r12,lsr #15\r |
499a0be3 |
1837 | ldrh r6, [r3, r6]\r |
9839d126 |
1838 | and r8, lr, r12,lsr #23\r |
499a0be3 |
1839 | ldrh r8, [r3, r8]\r |
9839d126 |
1840 | orr r5, r5, r6, lsl #16\r |
499a0be3 |
1841 | \r |
9839d126 |
1842 | and r6, lr, r7, lsl #1\r |
499a0be3 |
1843 | ldrh r6, [r3, r6]\r |
9839d126 |
1844 | and r12,lr, r7, lsr #7\r |
499a0be3 |
1845 | ldrh r12,[r3, r12]\r |
9839d126 |
1846 | orr r6, r8, r6, lsl #16\r |
1847 | \r |
1848 | and r8, lr, r7, lsr #15\r |
499a0be3 |
1849 | ldrh r8, [r3, r8]\r |
9839d126 |
1850 | and r7, lr, r7, lsr #23\r |
499a0be3 |
1851 | \r |
1852 | subs r2, r2, #1\r |
9839d126 |
1853 | orr r12,r12,r8, lsl #16\r |
1854 | ldrh r8, [r3, r7]\r |
1855 | \r |
1856 | stmia r0!, {r4,r5,r6,r12}\r |
499a0be3 |
1857 | bne .fl_loopRGB555u\r |
1858 | \r |
1859 | strh r8, [r0], #2\r |
1860 | \r |
1861 | ldmfd sp!, {r4-r8,lr}\r |
1862 | bx lr\r |
1863 | \r |
1864 | \r |
1865 | .fl_32scale_RGB555u:\r |
9839d126 |
1866 | sub r0, r0, #2 @ initial adjustment\r |
1867 | mov r4, #0\r |
499a0be3 |
1868 | \r |
1869 | @ r9 f800 07e0 001f | e000 0780 001c | 3800 01e0 0007\r |
1870 | .fl_loop32scale_RGB555u:\r |
1871 | ldr r12, [r1], #4\r |
1872 | ldr r7, [r1], #4\r |
1873 | \r |
9839d126 |
1874 | and r6, lr, r12,lsl #1\r |
1875 | ldrh r6, [r3, r6]\r |
499a0be3 |
1876 | and r5, lr, r12,lsr #7\r |
1877 | ldrh r5, [r3, r5]\r |
9839d126 |
1878 | and r6, r6, r9, lsl #2\r |
1879 | orr r4, r4, r6, lsl #16 @ r4 = pix_d -1, 0\r |
499a0be3 |
1880 | \r |
1881 | and r5, r5, r9, lsl #2\r |
9839d126 |
1882 | sub r8, r5, r5, lsr #2 @ r8 = 3/4 pix_s 1\r |
1883 | add r6, r8, r6, lsr #2 @ r6 = (1/4 pix_s 0) + (3/4 pix_s 1)\r |
1884 | orr r5, r6, r5, lsl #15\r |
499a0be3 |
1885 | \r |
1886 | and r6, lr, r12,lsr #15\r |
1887 | ldrh r6, [r3, r6]\r |
1888 | and r12,lr, r12,lsr #23\r |
1889 | ldrh r12,[r3, r12]\r |
1890 | and r6, r6, r9, lsl #2\r |
9839d126 |
1891 | add r5, r5, r6, lsl #15 @ r5 = pix_d 1, 2\r |
499a0be3 |
1892 | \r |
9839d126 |
1893 | and r8, lr, r7, lsl #1\r |
1894 | ldrh r8, [r3, r8]\r |
1895 | and r10,lr, r7, lsr #7\r |
1896 | ldrh r10,[r3, r10]\r |
499a0be3 |
1897 | and r12,r12,r9, lsl #2\r |
9839d126 |
1898 | sub r6, r6, r6, lsr #2 @ r6 = 3/4 pix_s 2\r |
1899 | add r6, r6, r12,lsr #2\r |
1900 | orr r6, r6, r12,lsl #16 @ r6 = pix_d 3, 4\r |
499a0be3 |
1901 | \r |
9839d126 |
1902 | and r8, r8, r9, lsl #2\r |
1903 | and r10,r10,r9, lsl #2\r |
1904 | sub r12,r10,r10,lsr #2 @ r12 = 3/4 pix_s 5\r |
1905 | orr r8, r8, r8, lsl #14\r |
1906 | add r8, r8, r12,lsl #16 @ r8 = pix_d 5, 6\r |
1907 | and r12,lr, r7, lsr #15\r |
499a0be3 |
1908 | ldrh r12,[r3, r12]\r |
499a0be3 |
1909 | and r7, lr, r7, lsr #23\r |
1910 | ldrh r7, [r3, r7]\r |
9839d126 |
1911 | and r12,r12,r9, lsl #2\r |
1912 | add r10,r10,r12\r |
1913 | mov r10,r10, lsr #1\r |
1914 | sub r12,r12,r12,lsr #2 @ r12 = 3/4 pix_s 6\r |
1915 | orr r10,r10,r12,lsl #16\r |
499a0be3 |
1916 | and r7, r7, r9, lsl #2\r |
9839d126 |
1917 | add r10,r10,r7, lsl #14 @ r10 = pix_d 7, 8\r |
499a0be3 |
1918 | \r |
1919 | subs r2, r2, #1\r |
1920 | \r |
1921 | stmia r0!, {r4,r5,r6,r8,r10}\r |
9839d126 |
1922 | mov r4, r7\r |
499a0be3 |
1923 | bne .fl_loop32scale_RGB555u\r |
1924 | \r |
9839d126 |
1925 | strh r4, [r0], #2\r |
1926 | \r |
499a0be3 |
1927 | ldmfd sp!, {r9,r10}\r |
1928 | ldmfd sp!, {r4-r8,lr}\r |
1929 | bx lr\r |
1930 | \r |
1931 | .endif @ UNALIGNED_DRAWLINEDEST\r |
1932 | \r |
cc68a136 |
1933 | \r |
1934 | @ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r |
1935 | \r |
1936 | @ utility\r |
1937 | .global blockcpy @ void *dst, void *src, size_t n\r |
1938 | \r |
1939 | blockcpy:\r |
1940 | stmfd sp!, {r4,r5}\r |
1941 | mov r2, r2, lsr #4\r |
1942 | blockcpy_loop:\r |
1943 | ldmia r1!, {r3-r5,r12}\r |
1944 | subs r2, r2, #1\r |
1945 | stmia r0!, {r3-r5,r12}\r |
1946 | bne blockcpy_loop\r |
1947 | ldmfd sp!, {r4,r5}\r |
1948 | bx lr\r |
1949 | \r |
1950 | \r |
1951 | .global blockcpy_or @ void *dst, void *src, size_t n, int pat\r |
1952 | \r |
1953 | blockcpy_or:\r |
1954 | stmfd sp!, {r4-r6}\r |
1955 | orr r3, r3, r3, lsl #8\r |
1956 | orr r3, r3, r3, lsl #16\r |
1957 | mov r2, r2, lsr #4\r |
1958 | blockcpy_loop_or:\r |
1959 | ldmia r1!, {r4-r6,r12}\r |
1960 | subs r2, r2, #1\r |
1961 | orr r4, r4, r3\r |
1962 | orr r5, r5, r3\r |
1963 | orr r6, r6, r3\r |
1964 | orr r12,r12,r3\r |
1965 | stmia r0!, {r4-r6,r12}\r |
1966 | bne blockcpy_loop_or\r |
1967 | ldmfd sp!, {r4-r6}\r |
1968 | bx lr\r |
1969 | \r |