notaz.gp2x.de / picodrive.git / blame
? search:
summary | shortlog | log | commit | commitdiff | tree
blob | blame (incremental) | history | HEAD
some support for vdp debug reg
[picodrive.git] / pico / draw_arm.S
CommitLineData
cff531af 1/*\r
2 * assembly optimized versions of most funtions from draw.c\r
ea38612f 3 * (C) notaz, 2006-2010,2017\r
cff531af 4 *\r
5 * This work is licensed under the terms of MAME license.\r
6 * See COPYING file in the top-level directory.\r
7 *\r
8 * this is highly specialized, be careful if changing related C code!\r
9 */\r
cc68a136 10\r
ea38612f 11#include "pico_int_o32.h"\r
12\r
cc68a136 13.extern DrawStripInterlace\r
5a681086 14\r
283fec1b 15.equ PDRAW_SPRITES_MOVED, (1<<0)\r
16.equ PDRAW_WND_DIFF_PRIO, (1<<1)\r
17.equ PDRAW_ACC_SPRITES, (1<<2)\r
18.equ PDRAW_DIRTY_SPRITES, (1<<4)\r
19.equ PDRAW_PLANE_HI_PRIO, (1<<6)\r
20.equ PDRAW_SHHI_DONE, (1<<7)\r
cc68a136 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
bfa12428 68 andeq r4, r4, #0xbf\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
bfa12428 159 ldrgeb r7, [r1,#\ofs]\r
160 orrlt r7, r3, r4 @ normal\r
07abbab1 161\r
bfa12428 162 bicge r7, r7, #0xc0\r
163 orrge r7, r7, r4, lsl #6\r
164 strb r7, [r1,#\ofs]\r
07abbab1 1650:\r
cc68a136 166.endm\r
167\r
168@ TileFlipSh (r1=pdest, r2=pixels8, r3=pal) r4,r7: scratch, r0=sx, r12: register with helper pattern 0xf\r
169.macro TileFlipSh\r
170 TileDoShGenPixel 16, 0 @ #0x000f0000\r
171 TileDoShGenPixel 20, 1 @ #0x00f00000\r
172 TileDoShGenPixel 24, 2 @ #0x0f000000\r
173 TileDoShGenPixel 28, 3 @ #0xf0000000\r
174 TileDoShGenPixel 0, 4 @ #0x0000000f\r
175 TileDoShGenPixel 4, 5 @ #0x000000f0\r
176 TileDoShGenPixel 8, 6 @ #0x00000f00\r
177 TileDoShGenPixel 12, 7 @ #0x0000f000\r
178.endm\r
179\r
180@ TileNormSh (r1=pdest, r2=pixels8, r3=pal) r4,r7: scratch, r0=sx, r12: register with helper pattern 0xf\r
181.macro TileNormSh\r
182 TileDoShGenPixel 12, 0 @ #0x0000f000\r
183 TileDoShGenPixel 8, 1 @ #0x00000f00\r
184 TileDoShGenPixel 4, 2 @ #0x000000f0\r
185 TileDoShGenPixel 0, 3 @ #0x0000000f\r
186 TileDoShGenPixel 28, 4 @ #0xf0000000\r
187 TileDoShGenPixel 24, 5 @ #0x0f000000\r
188 TileDoShGenPixel 20, 6 @ #0x00f00000\r
189 TileDoShGenPixel 16, 7 @ #0x000f0000\r
190.endm\r
191\r
bfa12428 192.macro TileDoShGenPixel_markop shift ofs\r
07abbab1 193.if \shift\r
bfa12428 194 ands r4, r12, r2, lsr #\shift\r
07abbab1 195.else\r
bfa12428 196 ands r4, r12, r2\r
07abbab1 197.endif\r
bfa12428 198 beq 0f\r
199 cmp r4, #0xe\r
200 ldrgeb r4, [r1,#\ofs]\r
201 orrlt r4, r3, r4\r
e54507e8 202 orrge r4, r4, #0x80\r
bfa12428 203 strb r4, [r1,#\ofs]\r
2040:\r
07abbab1 205.endm\r
206\r
e54507e8 207.macro TileFlipSh_markop\r
bfa12428 208 TileDoShGenPixel_markop 16, 0 @ #0x000f0000\r
209 TileDoShGenPixel_markop 20, 1 @ #0x00f00000\r
210 TileDoShGenPixel_markop 24, 2 @ #0x0f000000\r
211 TileDoShGenPixel_markop 28, 3 @ #0xf0000000\r
212 TileDoShGenPixel_markop 0, 4 @ #0x0000000f\r
213 TileDoShGenPixel_markop 4, 5 @ #0x000000f0\r
214 TileDoShGenPixel_markop 8, 6 @ #0x00000f00\r
215 TileDoShGenPixel_markop 12, 7 @ #0x0000f000\r
07abbab1 216.endm\r
217\r
e54507e8 218.macro TileNormSh_markop\r
bfa12428 219 TileDoShGenPixel_markop 12, 0 @ #0x0000f000\r
220 TileDoShGenPixel_markop 8, 1 @ #0x00000f00\r
221 TileDoShGenPixel_markop 4, 2 @ #0x000000f0\r
222 TileDoShGenPixel_markop 0, 3 @ #0x0000000f\r
223 TileDoShGenPixel_markop 28, 4 @ #0xf0000000\r
224 TileDoShGenPixel_markop 24, 5 @ #0x0f000000\r
225 TileDoShGenPixel_markop 20, 6 @ #0x00f00000\r
226 TileDoShGenPixel_markop 16, 7 @ #0x000f0000\r
07abbab1 227.endm\r
228\r
229.macro TileDoShGenPixel_onlyop_lp shift ofs\r
230.if \shift\r
231 ands r7, r12, r2, lsr #\shift\r
232.else\r
233 ands r7, r12, r2\r
234.endif\r
235 ldrneb r4, [r1,#\ofs]\r
07abbab1 236 cmp r7, #0xe\r
bfa12428 237 blt 0f\r
238\r
239 tst r4, #0xc0\r
240 bicne r4, r4, #0xc0\r
241 orrne r4, r4, r7, lsl #6\r
242 strneb r4, [r1,#\ofs]\r
07abbab1 2430:\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
e0bcb7a9 281@ void DrawLayer(int lflags, int *hcache, int cellskip, int maxcells,\r
ea38612f 282@ struct PicoEState *est)\r
cc68a136 283\r
83c093a4 284.global DrawLayer\r
cc68a136 285\r
286DrawLayer:\r
ea38612f 287 ldr r12, [sp] @ est\r
cc68a136 288 stmfd sp!, {r4-r11,lr}\r
289\r
ea38612f 290 ldr r11, [r12, #OFS_Pico_video]\r
cc68a136 291 mov r8, #1\r
292\r
83c093a4 293 ldrb r7, [r11, #16] @ ??vv??hh\r
cc68a136 294\r
295 mov r6, r1 @ hcache\r
e0bcb7a9 296 orr r9, r3, r0, lsl #29 @ r9=force[31]|sh[30]|plane[29]\r
297 orr r9, r9, r2, lsl #8 @ |cellskip[15:8]|maxcells[7:0] (tmp)\r
cc68a136 298\r
299 mov r1, r7, lsl #4\r
300 orr r1, r1, #0x00ff\r
301\r
302 and r10, r7, #3\r
303 cmp r10, #1\r
304 biclt r1, r1, #0xfc00\r
305 biceq r1, r1, #0xfe00\r
eced0190 306 cmp r10, #2\r
307 moveq r1, #0x0007\r
308 movgt r1, #0x00ff @ r1=ymask=(height<<8)|0xff; ...; // Y Mask in pixels\r
cc68a136 309\r
310 add r10, r10, #5\r
311 cmp r10, #7\r
312 subge r10, r10, #1 @ r10=shift[width] (5,6,6,7)\r
313\r
ea38612f 314 ldr r2, [r12, #OFS_DrawScanline]\r
315 ldr lr, [r12, #OFS_Pico_vram]\r
cc68a136 316\r
317 @ Find name table:\r
83c093a4 318 ands r0, r0, #1\r
cc68a136 319 ldreqb r12, [r11, #2]\r
320 ldrneb r12, [r11, #4]\r
321\r
ea38612f 322 @ calculate xmask:\r
323 mov r5, r8, lsl r10\r
324 sub r5, r5, #1 @ r5=xmask\r
cc68a136 325\r
326 moveq r12, r12, lsl #10\r
327 movne r12, r12, lsl #13\r
328 and r12, r12, #(7<<13) @ r12=(ts->nametab<<1) (halfword compliant)\r
329\r
330 ldrh r8, [r11, #12]\r
331 ldrb r7, [r11, #11]\r
cc68a136 332\r
333 mov r4, r8, lsr #8 @ pvid->reg[13]\r
334 mov r4, r4, lsl #10 @ htab=pvid->reg[13]<<9; (halfwords)\r
335 tst r7, #2\r
b6d7ac70 336 addne r4, r4, r2, lsl #2 @ htab+=DrawScanline<<1; // Offset by line\r
cc68a136 337 tst r7, #1\r
338 biceq r4, r4, #0x1f @ htab&=~0xf; // Offset by tile\r
339 add r4, r4, r0, lsl #1 @ htab+=plane\r
340 bic r4, r4, #0x00ff0000 @ just in case\r
341 ldrh r3, [lr, r4] @ r3=hscroll\r
342\r
343 tst r7, #4\r
344 bne .DrawStrip_vsscroll\r
345\r
346 @ Get vertical scroll value:\r
347 add r7, lr, #0x012000\r
348 add r7, r7, #0x000180 @ r7=Pico.vsram (Pico+0x22180)\r
349 ldr r7, [r7]\r
350\r
351 tst r8, #2\r
352 tstne r8, #4\r
353 bne .DrawStrip_interlace\r
354\r
355 tst r0, r0\r
356 movne r7, r7, lsr #16\r
357\r
358 @ Find the line in the name table\r
359 add r2, r2, r7\r
360 and r2, r2, r1\r
361 mov r4, r2, lsr #3\r
362 add r10, r10, #1 @ shift[width]++\r
363 add r12, r12, r4, lsl r10 @ nametab+=(ts.line>>3)<<shift[width];\r
364\r
365 @ ldmia r0, {r1,r2,r3,r5,r6,r9} @ r2=line, r3=ts->hscroll, r5=ts->xmask, r6=ts->hc, r9=ts->cells\r
cc68a136 366\r
367 and r10,r2, #7\r
368 mov r10,r10, lsl #1 @ r10=ty=(ts->line&7)<<1;\r
369 orr r10,r10, r9, lsl #24\r
370\r
371 rsb r8, r3, #0\r
372 mov r8, r8, lsr #3 @ r8=tilex=(-ts->hscroll)>>3\r
373\r
374 sub r1, r3, #1\r
375 and r1, r1, #7\r
376 add r7, r1, #1 @ r7=dx=((ts->hscroll-1)&7)+1\r
377\r
e0bcb7a9 378 movs r3, r9, lsl #1 @ (force[31]|sh[30]) << 1\r
cc68a136 379 mov r3, #0\r
e0bcb7a9 380 orrmi r10,r10, #1<<23 @ r10=cells[31:24]|sh[23]|hi_not_empty[22]\r
381 orrcs r10,r10, #1<<20 @ |had_output[21]|force[20]|ty[15:0]\r
382 movmi r3, #0x40 @ default to shadowed pal on sh mode\r
cc68a136 383\r
cc68a136 384 cmp r7, #8\r
385 addne r10,r10, #0x01000000 @ we will loop cells+1 times if there is scroll\r
386\r
83c093a4 387 and r9, r9, #0xff00\r
388 add r8, r8, r9, lsr #8 @ tilex+=cellskip\r
389 add r7, r7, r9, lsr #5 @ dx+=cellskip<<3;\r
390 sub r10,r10,r9, lsl #16 @ cells-=cellskip\r
83c093a4 391\r
cc68a136 392 @ cache some stuff to avoid mem access\r
99bdfd31 393 ldr r11,[sp, #9*4] @ est\r
cc68a136 394 mov r0, #0xf\r
99bdfd31 395 ldr r11,[r11, #OFS_HighCol]\r
ea8c405f 396\r
397 mvn r9, #0 @ r9=prevcode=-1\r
99bdfd31 398 add r1, r11, r7 @ r1=pdest\r
cc68a136 399\r
400\r
401 @ r4 & r7 are scratch in this loop\r
402.dsloop_subr1:\r
403 sub r1, r1, #8\r
404.dsloop: @ 40-41 times\r
405 subs r10,r10, #0x01000000\r
406 bmi .dsloop_exit\r
407\r
408.dsloop_enter:\r
409 and r7, r5, r8\r
410 add r7, lr, r7, lsl #1 @ Pico.vram+((tilex&ts->xmask) as halfwords)\r
411 ldrh r7, [r7, r12] @ r7=code (int, but from unsigned, no sign extend)\r
412\r
413 add r1, r1, #8\r
414 add r8, r8, #1\r
415\r
416 tst r7, #0x8000\r
e0bcb7a9 417 tsteq r10, #1<<20 @ force?\r
cc68a136 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
e0bcb7a9 485 tst r10, #0x00d00000 @ sh[23]|hi_not_empty[22]|force[20]\r
740da8c6 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
ea38612f 498 ldr r1, [sp, #9*4] @ est\r
cc68a136 499 mov r0, #0\r
ea38612f 500 ldreq r2, [r1, #OFS_rendstatus]\r
cc68a136 501 str r0, [r6] @ terminate the cache list\r
283fec1b 502 orreq r2, r2, #PDRAW_PLANE_HI_PRIO @ had a layer with all hi-prio tiles\r
ea38612f 503 streq r2, [r1, #OFS_rendstatus]\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
ea38612f 516 ldr r11, [sp, #9*4] @ est\r
6d7acf9e 517 orr r5, r1, r10, lsl #24\r
ea38612f 518 ldr r4, [r11, #OFS_DrawScanline]\r
6d7acf9e 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
99bdfd31 542 ldr r11,[sp, #9*4] @ est\r
ea8c405f 543 mov r0, #0xf\r
99bdfd31 544 ldr r11,[r11, #OFS_HighCol]\r
6d7acf9e 545\r
ea8c405f 546 mvn r9, #0 @ r9=prevcode=-1\r
547 add r1, r11, r7 @ r1=pdest\r
6d7acf9e 548\r
549 @ r4 & r7 are scratch in this loop\r
550.dsloop_vs_subr1:\r
551 sub r1, r1, #8\r
552.dsloop_vs: @ 40-41 times\r
553 add r10,r10, #0x01000000\r
554 and r4, r10, #0x003f0000\r
555 cmp r4, r10, asr #8\r
740da8c6 556 ble .dsloop_vs_exit\r
6d7acf9e 557\r
558 @ calc offset and read tileline code to r7, also calc ty\r
559 add r7, lr, #0x012000\r
560 add r7, r7, #0x000180 @ r7=Pico.vsram (Pico+0x22180)\r
561 add r7, r7, r10,asr #23 @ vsram + ((cell&~1)<<1)\r
562 bic r7, r7, #3\r
563 tst r10,#0x8000 @ plane1?\r
564 addne r7, r7, #2\r
565 ldrh r7, [r7] @ r7=vscroll\r
566\r
567 bic r10,r10,#0xff @ clear old ty\r
7b802576 568 and r4, r5, #0xff0000 @ scanline\r
569 add r4, r4, r7, lsl #16 @ ... += vscroll\r
570 and r4, r4, r5, lsl #16 @ ... &= ymask\r
6d7acf9e 571 and r7, r4, #0x70000\r
572 orr r10,r10,r7, lsr #15 @ new ty\r
573\r
574 mov r4, r4, lsr #19\r
575 mov r7, r5, lsr #24\r
576 mov r4, r4, lsl r7 @ nametabadd\r
577\r
578 and r7, r8, r8, lsr #25\r
579 add r7, lr, r7, lsl #1 @ Pico.vram+((tilex&ts->xmask) as halfwords)\r
580 add r7, r7, r4, lsl #1\r
581 ldrh r7, [r7, r12] @ r7=code (int, but from unsigned, no sign extend)\r
582\r
583 add r1, r1, #8\r
584 add r8, r8, #1\r
585\r
586 tst r7, #0x8000\r
587 bne .DrawStrip_vs_hiprio\r
588\r
589 cmp r7, r9\r
590 beq .DrawStrip_vs_samecode @ we know stuff about this tile already\r
591\r
592 mov r9, r7 @ remember code\r
7b802576 593 orr r8, r8, #(1<<24)@ seen non hi-prio tile\r
6d7acf9e 594\r
595 movs r2, r9, lsl #20 @ if (code&0x1000)\r
596 mov r2, r2, lsl #1\r
597 add r2, r2, r10, lsl #17\r
598 mov r2, r2, lsr #17\r
599 eorcs r2, r2, #0x0e @ if (code&0x1000) addr^=0xe;\r
600\r
601 ldr r2, [lr, r2, lsl #1] @ pack=*(unsigned int *)(Pico.vram+addr); // Get 8 pixels\r
602\r
603 bic r7, r3, #0x3f\r
604 and r3, r9, #0x6000\r
605 add r3, r7, r3, lsr #9 @ r3=pal=((code&0x6000)>>9);\r
606\r
607.DrawStrip_vs_samecode:\r
608 tst r2, r2\r
609 beq .dsloop_vs @ tileline blank\r
610\r
611 cmp r2, r2, ror #4\r
612 beq .DrawStrip_vs_SingleColor @ tileline singlecolor \r
613\r
614 tst r9, #0x0800\r
7a7c6476 615 bne .DrawStrip_vs_TileFlip\r
6d7acf9e 616\r
617 @ (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r0: helper pattern\r
6d7acf9e 618.DrawStrip_vs_TileNorm:\r
619 TileNorm r0\r
620 b .dsloop_vs\r
621\r
7a7c6476 622.DrawStrip_vs_TileFlip:\r
623 TileFlip r0\r
624 b .dsloop_vs\r
625\r
6d7acf9e 626.DrawStrip_vs_SingleColor:\r
627 and r4, r2, #0xf\r
628 orr r4, r3, r4\r
629 orr r4, r4, r4, lsl #8\r
630 tst r1, #1 @ not aligned?\r
631 strneb r4, [r1], #1\r
632 streqh r4, [r1], #2\r
633 strh r4, [r1], #2\r
634 strh r4, [r1], #2\r
635 strh r4, [r1], #2\r
636 strneb r4, [r1], #1 @ have a remaining unaligned pixel?\r
637 b .dsloop_vs_subr1\r
638\r
639.DrawStrip_vs_hiprio:\r
640 tst r10, #0x00c00000\r
641 beq .DrawStrip_vs_hiprio_maybempt\r
642 sub r0, r1, r11\r
643 orr r7, r7, r0, lsl #16\r
644 orr r7, r7, r10, lsl #25 @ (ty<<25)\r
645 tst r7, #0x1000\r
646 eorne r7, r7, #7<<26 @ if(code&0x1000) cval^=7<<26;\r
647 str r7, [r6], #4 @ cache hi priority tile\r
648 mov r0, #0xf\r
649 b .dsloop_vs\r
650\r
651.DrawStrip_vs_hiprio_maybempt:\r
652 cmp r7, r9\r
653 beq .dsloop_vs @ must've been empty, otherwise we wouldn't get here\r
654 movs r2, r7, lsl #20 @ if (code&0x1000)\r
655 mov r2, r2, lsl #1\r
656 add r2, r2, r10, lsl #17\r
657 mov r2, r2, lsr #17\r
658 eorcs r2, r2, #0x0e @ if (code&0x1000) addr^=0xe;\r
659 ldr r2, [lr, r2, lsl #1] @ pack=*(unsigned int *)(Pico.vram+addr); // Get 8 pixels\r
660 mov r9, r7 @ remember code\r
661 tst r2, r2\r
662 orrne r10, r10, #1<<22\r
663 bne .DrawStrip_vs_hiprio\r
664 b .dsloop_vs\r
665\r
740da8c6 666.dsloop_vs_exit:\r
7b802576 667 tst r8, #(1<<24) @ seen non hi-prio tile\r
ea38612f 668 ldr r1, [sp, #9*4] @ est\r
740da8c6 669 mov r0, #0\r
ea38612f 670 ldreq r2, [r1, #OFS_rendstatus]\r
740da8c6 671 str r0, [r6] @ terminate the cache list\r
283fec1b 672 orreq r2, r2, #PDRAW_PLANE_HI_PRIO @ had a layer with all hi-prio tiles\r
ea38612f 673 streq r2, [r1, #OFS_rendstatus]\r
740da8c6 674\r
675 ldmfd sp!, {r4-r11,lr}\r
676 bx lr\r
677\r
6d7acf9e 678\r
679@ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r
cc68a136 680\r
681@ interlace mode 2? Sonic 2?\r
682.DrawStrip_interlace:\r
683 tst r0, r0\r
684 moveq r7, r7, lsl #21\r
685 movne r7, r7, lsl #5\r
686\r
687 @ Find the line in the name table\r
b6d7ac70 688 add r2, r7, r2, lsl #22 @ r2=(vscroll+(DrawScanline<<1))<<21 (11 bits);\r
cc68a136 689 orr r1, r1, #0x80000000\r
690 and r2, r2, r1, ror #10 @ &((ymask<<1)|1)<<21;\r
691 mov r2, r2, lsr #21\r
692 mov r4, r2, lsr #4\r
693 mov r12, r12, lsr #1 @ halfwords\r
694 add r0, r12, r4, lsl r10 @ nametab+=(ts.line>>4)<<shift[width];\r
695 and r9, r9, #0xff\r
696\r
697 sub sp, sp, #6*4\r
698 stmia sp, {r0,r2,r3,r5,r6,r9}\r
699\r
700 mov r0, sp\r
701 bl DrawStripInterlace @ struct TileStrip *ts\r
702\r
703 add sp, sp, #6*4\r
704 ldmfd sp!, {r4-r11,lr}\r
705 bx lr\r
706\r
707.pool\r
708\r
709@ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r
710\r
99bdfd31 711@ void BackFill(int reg7, int sh, struct PicoEState *est)\r
cc68a136 712\r
99bdfd31 713.global BackFill\r
cc68a136 714\r
715BackFill:\r
716 stmfd sp!, {r4-r9,lr}\r
717\r
ea8c405f 718 mov r0, r0, lsl #26\r
99bdfd31 719 ldr lr, [r2, #OFS_HighCol]\r
ea8c405f 720 mov r0, r0, lsr #26\r
721 add lr, lr, #8\r
ea8c405f 722\r
cc68a136 723 orr r0, r0, r1, lsl #6\r
724 orr r0, r0, r0, lsl #8\r
725 orr r0, r0, r0, lsl #16\r
726\r
727 mov r1, r0\r
728 mov r2, r0\r
729 mov r3, r0\r
730 mov r4, r0\r
731 mov r5, r0\r
732 mov r6, r0\r
733 mov r7, r0\r
734\r
735 @ go go go!\r
736 stmia lr!, {r0-r7} @ 10*8*4\r
737 stmia lr!, {r0-r7}\r
738 stmia lr!, {r0-r7}\r
739 stmia lr!, {r0-r7}\r
740 stmia lr!, {r0-r7}\r
741 stmia lr!, {r0-r7}\r
742 stmia lr!, {r0-r7}\r
743 stmia lr!, {r0-r7}\r
744 stmia lr!, {r0-r7}\r
745 stmia lr!, {r0-r7}\r
746\r
99bdfd31 747 ldmfd sp!, {r4-r9,lr}\r
748 bx lr\r
cc68a136 749\r
750\r
751@ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r
752\r
ea38612f 753@ void DrawTilesFromCache(int *hc, int sh, int rlim, struct PicoEState *est)\r
cc68a136 754\r
ea38612f 755.global DrawTilesFromCache\r
cc68a136 756\r
757DrawTilesFromCache:\r
ea38612f 758 stmfd sp!, {r4-r9,r11,lr}\r
cc68a136 759\r
cc68a136 760 @ cache some stuff to avoid mem access\r
99bdfd31 761 ldr r11,[r3, #OFS_HighCol]\r
cc68a136 762 mov r12,#0xf\r
ea38612f 763 ldr lr, [r3, #OFS_Pico_vram]\r
764 mov r9, r3 @ est\r
cc68a136 765\r
740da8c6 766 mvn r5, #0 @ r5=prevcode=-1\r
7a7c6476 767 ands r8, r1, #1\r
768 orr r8, r8, r2, lsl #1\r
740da8c6 769 bne .dtfc_check_rendflags\r
770\r
cc68a136 771 @ scratch: r4, r7\r
772.dtfc_loop:\r
773 ldr r6, [r0], #4 @ read code\r
774 movs r1, r6, lsr #16 @ r1=dx;\r
ea38612f 775 ldmeqfd sp!, {r4-r9,r11,pc} @ dx is never zero, this must be a terminator, return\r
7a7c6476 776 bic r4, r1, #0xfe00\r
777 add r1, r11, r4 @ r1=pdest\r
cc68a136 778\r
e0bcb7a9 779 movs r7, r6, lsl #16\r
780 bpl .dtfc_loop @ !(code & 0x8000)\r
cc68a136 781 cmp r5, r7, lsr #16\r
782 beq .dtfc_samecode @ if (code==prevcode)\r
783\r
784 mov r5, r7, lsr #16\r
785\r
786 mov r2, r5, lsl #21\r
787 mov r2, r2, lsr #17 @ r2=addr=(code&0x7ff)<<4;\r
788 add r2, r2, r6, lsr #25 @ addr+=ty\r
789\r
790 and r3, r5, #0x6000\r
791 mov r3, r3, lsr #9 @ r3=pal=((code&0x6000)>>9);\r
792\r
793 ldr r2, [lr, r2, lsl #1] @ pack=*(unsigned int *)(Pico.vram+addr); // Get 8 pixels\r
794\r
795.dtfc_samecode:\r
7a7c6476 796 rsbs r4, r4, r8, lsr #1\r
797 bmi .dtfc_cut_tile\r
798\r
799 tst r8, #1\r
cc68a136 800 bne .dtfc_shadow\r
801\r
802 tst r2, r2\r
803 beq .dtfc_loop\r
804\r
805 cmp r2, r2, ror #4\r
806 beq .dtfc_SingleColor @ tileline singlecolor \r
807\r
808 tst r5, #0x0800\r
7a7c6476 809 bne .dtfc_TileFlip\r
cc68a136 810\r
811 @ (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r12: helper pattern\r
cc68a136 812.dtfc_TileNorm:\r
813 TileNorm r12\r
814 b .dtfc_loop\r
815\r
7a7c6476 816.dtfc_TileFlip:\r
817 TileFlip r12\r
818 b .dtfc_loop\r
819\r
cc68a136 820.dtfc_SingleColor:\r
821 and r4, r2, #0xf\r
822 orr r4, r3, r4\r
823 orr r4, r4, r4, lsl #8\r
824 tst r1, #1 @ not aligned?\r
825 strneb r4, [r1], #1\r
826 streqh r4, [r1], #2\r
827 strh r4, [r1], #2\r
828 strh r4, [r1], #2\r
829 strh r4, [r1], #2\r
830 strneb r4, [r1], #1 @ have a remaining unaligned pixel?\r
831 b .dtfc_loop\r
832\r
833.dtfc_shadow:\r
834 tst r2, r2\r
835 beq .dtfc_shadow_blank\r
836\r
837 cmp r2, r2, ror #4\r
838 beq .dtfc_SingleColor @ tileline singlecolor \r
839\r
840 tst r5, #0x0800\r
7a7c6476 841 bne .dtfc_TileFlipShHP\r
cc68a136 842\r
843 @ (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r12: helper pattern\r
cc68a136 844.dtfc_TileNormShHP:\r
845 TileNormShHP\r
846 b .dtfc_loop\r
847\r
7a7c6476 848.dtfc_TileFlipShHP:\r
849 TileFlipShHP\r
850 b .dtfc_loop\r
851\r
cc68a136 852.dtfc_shadow_blank:\r
07abbab1 853 tst r1, #1\r
854 ldrneb r4, [r1]\r
bfa12428 855 mov r6, #0xbf\r
856 and r4, r4, #0xbf\r
07abbab1 857 strneb r4, [r1], #1\r
858 ldrh r4, [r1]\r
859 orr r6, r6, r6, lsl #8\r
860 and r4, r4, r6\r
861 strh r4, [r1], #2\r
862 ldrh r4, [r1]\r
863 and r4, r4, r6\r
864 strh r4, [r1], #2\r
865 ldrh r4, [r1]\r
866 and r4, r4, r6\r
867 strh r4, [r1], #2\r
868 ldrh r4, [r1]\r
869 and r4, r4, r6\r
870 streqh r4, [r1]\r
871 strneb r4, [r1]\r
cc68a136 872 b .dtfc_loop\r
873\r
7a7c6476 874.dtfc_cut_tile:\r
875 add r4, r4, #7 @ 0-6\r
876 mov r4, r4, lsl #2\r
877 mov r12,#0xf<<28\r
878 mov r12,r12,asr r4\r
879 mov r2, r2, ror #16\r
880 tst r5, #0x0800 @ flipped?\r
881 mvnne r12,r12\r
882 and r2, r2, r12\r
883 mov r2, r2, ror #16\r
884 mov r12,#0xf\r
885 tst r8, #1\r
886 bne .dtfc_shadow\r
887 tst r2, r2\r
888 beq .dtfc_loop\r
889 tst r5, #0x0800\r
890 beq .dtfc_TileNorm\r
891 b .dtfc_TileFlip\r
892\r
740da8c6 893@ check if we have detected layer covered with hi-prio tiles:\r
894.dtfc_check_rendflags:\r
ea38612f 895 ldr r2, [r9, #OFS_rendstatus]\r
283fec1b 896 tst r2, #(PDRAW_PLANE_HI_PRIO|PDRAW_SHHI_DONE)\r
740da8c6 897 beq .dtfc_loop\r
7a7c6476 898 bic r8, r8, #1 @ sh/hi mode off\r
283fec1b 899 tst r2, #PDRAW_SHHI_DONE\r
740da8c6 900 bne .dtfc_loop @ already processed\r
283fec1b 901 orr r2, r2, #PDRAW_SHHI_DONE\r
ea38612f 902 str r2, [r9, #OFS_rendstatus]\r
740da8c6 903\r
904 add r1, r11,#8\r
07abbab1 905 mov r3, #320/4/4\r
bfa12428 906 mov r6, #0xbf\r
740da8c6 907 orr r6, r6, r6, lsl #8\r
908 orr r6, r6, r6, lsl #16\r
909.dtfc_loop_shprep:\r
07abbab1 910 ldmia r1, {r2,r4,r5,r7}\r
740da8c6 911 subs r3, r3, #1\r
07abbab1 912 and r2, r2, r6\r
913 and r4, r4, r6\r
914 and r5, r5, r6\r
915 and r7, r7, r6\r
916 stmia r1!,{r2,r4,r5,r7}\r
917 bne .dtfc_loop_shprep\r
918\r
919 mvn r5, #0 @ r5=prevcode=-1\r
920 b .dtfc_loop\r
740da8c6 921\r
cc68a136 922.pool\r
923\r
924@ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r
925\r
926\r
ea38612f 927@ void DrawSpritesSHi(unsigned char *sprited, struct PicoEState *est)\r
928\r
929.global DrawSpritesSHi\r
e352c3af 930\r
931DrawSpritesSHi:\r
fbc65db7 932 ldr r3, [r0]\r
933 mov r12,#0xff\r
e352c3af 934 ands r3, r3, #0x7f\r
935 bxeq lr\r
cc68a136 936\r
ea38612f 937 stmfd sp!, {r1,r4-r11,lr} @ +est\r
fbc65db7 938 strb r12,[r0,#2] @ set end marker\r
939 add r10,r0, #3 @ r10=HighLnSpr end\r
e352c3af 940 add r10,r10,r3 @ r10=HighLnSpr end\r
941\r
99bdfd31 942 ldr r11,[r1, #OFS_HighCol]\r
ea8c405f 943 mov r12,#0xf\r
ea38612f 944 ldr lr, [r1, #OFS_Pico_vram]\r
cc68a136 945\r
cc68a136 946\r
e352c3af 947DrawSpriteSHi:\r
948 @ draw next sprite\r
949 ldrb r0, [r10,#-1]!\r
ea38612f 950 ldr r7, [sp] @ est\r
99bdfd31 951 ldr r1, [r7, #OFS_HighPreSpr]\r
e352c3af 952 cmp r0, #0xff\r
ea38612f 953 ldmeqfd sp!, {r1,r4-r11,pc} @ end of list\r
e352c3af 954 and r0, r0, #0x7f\r
955 add r0, r1, r0, lsl #3\r
956\r
957 ldr r9, [r0, #4] @ sprite[1]\r
958 mov r2, r9, asr #16 @ r2=sx\r
cc68a136 959\r
e352c3af 960 mov r9, r9, lsl #16\r
961 mov r3, r9, lsr #31 @ priority\r
962 mov r9, r9, lsr #16\r
fbc65db7 963@ orr r9, r9, r8, lsl #31 @ r9=code|sh[31] @@ sh is always on here now\r
e352c3af 964 and r4, r9, #0x6000\r
965 orr r9, r9, r4, lsl #16\r
fbc65db7 966 orr r9, r9, #0x90000000 @ r9=scc1 ???? ... <code> (s=shadow/hilight, cc=pal)\r
e352c3af 967 cmp r12,r9, lsr #28 @ sh/hi with pal3?\r
e54507e8 968 cmpne r3, #1 @ if not, is it hi prio?\r
e352c3af 969 bne DrawSpriteSHi @ non-operator low sprite, already drawn\r
cc68a136 970\r
e352c3af 971 ldr r3, [r0] @ sprite[0]\r
e352c3af 972 mov r6, r3, lsr #28\r
973 sub r6, r6, #1 @ r6=width-1 (inc later)\r
974 mov r5, r3, lsr #24\r
975 and r5, r5, #7 @ r5=height\r
cc68a136 976\r
ea38612f 977 ldr r7, [r7, #OFS_DrawScanline]\r
e352c3af 978 mov r0, r3, lsl #16 @ r4=sy<<16 (tmp)\r
cc68a136 979\r
e352c3af 980 sub r7, r7, r0, asr #16 @ r7=row=DrawScanline-sy\r
cc68a136 981\r
e352c3af 982 tst r9, #0x1000\r
983 movne r0, r5, lsl #3\r
984 subne r0, r0, #1\r
985 subne r7, r0, r7 @ if (code&0x1000) row=(height<<3)-1-row; // Flip Y\r
cc68a136 986\r
e352c3af 987 add r8, r9, r7, lsr #3 @ tile+=row>>3; // Tile number increases going down\r
988 tst r9, #0x0800\r
989 mlane r8, r5, r6, r8 @ if (code&0x0800) { tile+=delta*(width-1);\r
990 rsbne r5, r5, #0 @ delta=-delta; } // r5=delta now\r
cc68a136 991\r
e352c3af 992 mov r8, r8, lsl #21\r
993 mov r8, r8, lsr #17\r
994 and r7, r7, #7\r
995 add r8, r8, r7, lsl #1 @ tile+=(row&7)<<1; // Tile address\r
996\r
997 mov r5, r5, lsl #4 @ delta<<=4; // Delta of address\r
e352c3af 998 mov r3, r4, lsr #9 @ r3=pal=((code>>9)&0x30);\r
999\r
1000 add r6, r6, #1 @ inc now\r
1001 adds r0, r2, #0 @ mov sx to r0 and set ZV flags\r
1002 b .dsprShi_loop_enter\r
1003\r
1004.dsprShi_loop:\r
1005 subs r6, r6, #1 @ width--\r
1006 beq DrawSpriteSHi\r
1007 adds r0, r0, #8 @ sx+=8\r
1008 add r8, r8, r5 @ tile+=delta\r
1009\r
1010.dsprShi_loop_enter:\r
1011 ble .dsprShi_loop @ sx <= 0\r
cc68a136 1012 cmp r0, #328\r
e352c3af 1013 bge DrawSpriteSHi\r
cc68a136 1014\r
1015 mov r8, r8, lsl #17\r
e352c3af 1016 mov r8, r8, lsr #17 @ tile&=0x7fff; // Clip tile address\r
cc68a136 1017\r
e352c3af 1018 ldr r2, [lr, r8, lsl #1] @ pack=*(unsigned int *)(Pico.vram+addr); // Get 8 pixels\r
1019 add r1, r11, r0 @ r1=pdest\r
cc68a136 1020 tst r2, r2\r
e352c3af 1021 beq .dsprShi_loop\r
cc68a136 1022\r
e352c3af 1023 cmp r12, r9, lsr #28\r
1024 beq .dsprShi_shadow\r
cc68a136 1025\r
1026 cmp r2, r2, ror #4\r
e352c3af 1027 beq .dsprShi_SingleColor @ tileline singlecolor \r
cc68a136 1028\r
e352c3af 1029 tst r9, #0x0800\r
1030 bne .dsprShi_TileFlip\r
cc68a136 1031\r
e352c3af 1032 @ (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r12: helper pattern\r
1033@ scratch: r4, r7\r
1034.dsprShi_TileNorm:\r
cc68a136 1035 TileNorm r12\r
e352c3af 1036 b .dsprShi_loop\r
cc68a136 1037\r
e352c3af 1038.dsprShi_TileFlip:\r
7a7c6476 1039 TileFlip r12\r
e352c3af 1040 b .dsprShi_loop\r
7a7c6476 1041\r
e352c3af 1042.dsprShi_SingleColor:\r
cc68a136 1043 and r4, r2, #0xf\r
1044 orr r4, r3, r4\r
1045 orr r4, r4, r4, lsl #8\r
e352c3af 1046 tst r0, #1 @ not aligned?\r
cc68a136 1047 strneb r4, [r1], #1\r
1048 streqh r4, [r1], #2\r
1049 strh r4, [r1], #2\r
1050 strh r4, [r1], #2\r
1051 strh r4, [r1], #2\r
1052 strneb r4, [r1], #1\r
e352c3af 1053 b .dsprShi_loop\r
cc68a136 1054\r
e352c3af 1055.dsprShi_shadow:\r
1056 tst r9, #0x8000\r
1057 beq .dsprShi_shadow_lowpri\r
07abbab1 1058\r
cc68a136 1059 cmp r2, r2, ror #4\r
e352c3af 1060 beq .dsprShi_singlec_sh\r
cc68a136 1061\r
e352c3af 1062 tst r9, #0x0800\r
1063 bne .dsprShi_TileFlip_sh\r
cc68a136 1064\r
e54507e8 1065 @ (r1=pdest, r2=pixels8, r3=pal) r4, r7: scratch, r12: helper pattern\r
e352c3af 1066.dsprShi_TileNorm_sh:\r
cc68a136 1067 TileNormSh\r
e352c3af 1068 b .dsprShi_loop\r
cc68a136 1069\r
e352c3af 1070.dsprShi_TileFlip_sh:\r
7a7c6476 1071 TileFlipSh\r
e352c3af 1072 b .dsprShi_loop\r
7a7c6476 1073\r
e352c3af 1074.dsprShi_singlec_sh:\r
cc68a136 1075 cmp r2, #0xe0000000\r
e352c3af 1076 bcc .dsprShi_SingleColor @ normal singlecolor tileline (carry inverted in ARM)\r
cc68a136 1077 tst r2, #0x10000000\r
e352c3af 1078 bne .dsprShi_sh_sh\r
cc68a136 1079 TileSingleHi\r
e352c3af 1080 b .dsprShi_loop\r
cc68a136 1081\r
e352c3af 1082.dsprShi_sh_sh:\r
cc68a136 1083 TileSingleSh\r
e352c3af 1084 b .dsprShi_loop\r
cc68a136 1085\r
e352c3af 1086.dsprShi_shadow_lowpri:\r
1087 tst r9, #0x800\r
1088 bne .dsprShi_TileFlip_sh_lp\r
07abbab1 1089\r
e352c3af 1090.dsprShi_TileNorm_sh_lp:\r
07abbab1 1091 TileNormSh_onlyop_lp\r
e352c3af 1092 b .dsprShi_loop\r
07abbab1 1093\r
e352c3af 1094.dsprShi_TileFlip_sh_lp:\r
07abbab1 1095 TileFlipSh_onlyop_lp\r
e352c3af 1096 b .dsprShi_loop\r
07abbab1 1097\r
cc68a136 1098.pool\r
1099\r
1100@ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r
1101\r
ea38612f 1102@ void DrawAllSprites(unsigned char *sprited, int prio, int sh,\r
1103@ struct PicoEState *est)\r
1104\r
1105.global DrawAllSprites\r
283fec1b 1106\r
1107DrawAllSprites:\r
fbc65db7 1108 orr r1, r2, r1, lsl #1\r
ea38612f 1109 ldr r2, [r0]\r
1110 ands r2, r2, #0x7f\r
283fec1b 1111 bxeq lr\r
1112\r
1113 @ time to do some real work\r
ea38612f 1114 stmfd sp!, {r1,r3-r11,lr} @ +sh|prio<<1 +est\r
283fec1b 1115 mov r12,#0xff\r
fbc65db7 1116 strb r12,[r0,#2] @ set end marker\r
1117 add r10,r0, #3\r
ea38612f 1118 add r10,r10,r2 @ r10=HighLnSpr end\r
283fec1b 1119\r
99bdfd31 1120 ldr r11,[r3, #OFS_HighCol]\r
283fec1b 1121 mov r12,#0xf\r
ea38612f 1122 ldr lr, [r3, #OFS_Pico_vram]\r
283fec1b 1123\r
1124@ + 0 : hhhhvvvv ----hhvv yyyyyyyy yyyyyyyy // v, h: horiz. size\r
cc68a136 1125@ + 4 : xxxxxxxx xxxxxxxx pccvhnnn nnnnnnnn // x: x coord + 8\r
1126\r
ea38612f 1127DrawSprite:\r
283fec1b 1128 @ draw next sprite\r
1129 ldrb r0, [r10,#-1]!\r
ea38612f 1130 ldr r8, [sp] @ sh|prio<<1\r
1131 ldr r7, [sp, #4] @ est\r
e352c3af 1132 mov r2, r0, lsr #7\r
1133 cmp r0, #0xff\r
ea38612f 1134 ldmeqfd sp!, {r1,r3-r11,pc} @ end of list\r
97a7f774 1135 cmp r2, r8, lsr #1\r
e352c3af 1136 bne DrawSprite @ wrong priority\r
99bdfd31 1137 ldr r1, [r7, #OFS_HighPreSpr]\r
283fec1b 1138 and r0, r0, #0x7f\r
1139 add r0, r1, r0, lsl #3\r
cc68a136 1140\r
cc68a136 1141 ldr r3, [r0] @ sprite[0]\r
ea38612f 1142 ldr r7, [r7, #OFS_DrawScanline]\r
cc68a136 1143 mov r6, r3, lsr #28\r
1144 sub r6, r6, #1 @ r6=width-1 (inc later)\r
1145 mov r5, r3, lsr #24\r
1146 and r5, r5, #7 @ r5=height\r
1147\r
1148 mov r4, r3, lsl #16 @ r4=sy<<16 (tmp)\r
1149\r
cc68a136 1150 ldr r9, [r0, #4]\r
b6d7ac70 1151 sub r7, r7, r4, asr #16 @ r7=row=DrawScanline-sy\r
cc68a136 1152\r
cc68a136 1153 mov r2, r9, asr #16 @ r2=sx\r
e5fa9817 1154 mov r9, r9, lsl #16\r
1155 mov r9, r9, lsr #16\r
97a7f774 1156 orr r9, r9, r8, lsl #31 @ r9=code|sh[31]\r
cc68a136 1157\r
1158 tst r9, #0x1000\r
1159 movne r4, r5, lsl #3\r
1160 subne r4, r4, #1\r
1161 subne r7, r4, r7 @ if (code&0x1000) row=(height<<3)-1-row; // Flip Y\r
1162\r
07abbab1 1163 add r8, r9, r7, lsr #3 @ tile+=row>>3; // Tile number increases going down\r
cc68a136 1164 tst r9, #0x0800\r
1165 mlane r8, r5, r6, r8 @ if (code&0x0800) { tile+=delta*(width-1);\r
1166 rsbne r5, r5, #0 @ delta=-delta; } // r5=delta now\r
1167\r
07abbab1 1168 mov r8, r8, lsl #21\r
1169 mov r8, r8, lsr #17\r
cc68a136 1170 and r7, r7, #7\r
1171 add r8, r8, r7, lsl #1 @ tile+=(row&7)<<1; // Tile address\r
1172\r
e5fa9817 1173.dspr_continue:\r
cc68a136 1174 @ cache some stuff to avoid mem access\r
cc68a136 1175 mov r5, r5, lsl #4 @ delta<<=4; // Delta of address\r
1176 and r4, r9, #0x6000\r
1177 orr r9, r9, r4, lsl #16\r
97a7f774 1178 orrs r9, r9, #0x10000000 @ r9=scc1 ???? ... <code> (s=shadow/hilight, cc=pal)\r
cc68a136 1179\r
cc68a136 1180 mov r3, r4, lsr #9 @ r3=pal=((code>>9)&0x30);\r
97a7f774 1181 orrmi r3, r3, #0x40 @ for sh/hi\r
cc68a136 1182\r
1183 add r6, r6, #1 @ inc now\r
1184 adds r0, r2, #0 @ mov sx to r0 and set ZV flags\r
1185 b .dspr_loop_enter\r
1186\r
1187.dspr_loop:\r
1188 subs r6, r6, #1 @ width--\r
283fec1b 1189 beq DrawSprite\r
cc68a136 1190 adds r0, r0, #8 @ sx+=8\r
1191 add r8, r8, r5 @ tile+=delta\r
1192\r
1193.dspr_loop_enter:\r
1194 ble .dspr_loop @ sx <= 0\r
1195 cmp r0, #328\r
283fec1b 1196 bge DrawSprite\r
cc68a136 1197\r
1198 mov r8, r8, lsl #17\r
1199 mov r8, r8, lsr #17 @ tile&=0x7fff; // Clip tile address\r
1200\r
1201 ldr r2, [lr, r8, lsl #1] @ pack=*(unsigned int *)(Pico.vram+addr); // Get 8 pixels\r
1202 add r1, r11, r0 @ r1=pdest\r
1203 tst r2, r2\r
1204 beq .dspr_loop\r
1205\r
1206 cmp r12, r9, lsr #28\r
1207 beq .dspr_shadow\r
1208\r
1209 cmp r2, r2, ror #4\r
1210 beq .dspr_SingleColor @ tileline singlecolor \r
1211\r
1212 tst r9, #0x0800\r
7a7c6476 1213 bne .dspr_TileFlip\r
cc68a136 1214\r
1215 @ (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r12: helper pattern\r
cc68a136 1216@ scratch: r4, r7\r
1217.dspr_TileNorm:\r
1218 TileNorm r12\r
1219 b .dspr_loop\r
1220\r
7a7c6476 1221.dspr_TileFlip:\r
1222 TileFlip r12\r
1223 b .dspr_loop\r
1224\r
07abbab1 1225.dspr_singlec_sh:\r
1226 cmp r2, #0xe0000000\r
e54507e8 1227 bcs .dspr_TileNorm_sh @ op. tileline, markop. XXX: maybe add a spec. handler?\r
07abbab1 1228\r
cc68a136 1229.dspr_SingleColor:\r
1230 and r4, r2, #0xf\r
1231 orr r4, r3, r4\r
1232 orr r4, r4, r4, lsl #8\r
1233 tst r0, #1 @ not aligned?\r
1234 strneb r4, [r1], #1\r
1235 streqh r4, [r1], #2\r
1236 strh r4, [r1], #2\r
1237 strh r4, [r1], #2\r
1238 strh r4, [r1], #2\r
1239 strneb r4, [r1], #1\r
1240 b .dspr_loop\r
1241\r
1242.dspr_shadow:\r
1243 cmp r2, r2, ror #4\r
1244 beq .dspr_singlec_sh\r
1245\r
1246 tst r9, #0x0800\r
7a7c6476 1247 bne .dspr_TileFlip_sh\r
cc68a136 1248\r
1249 @ (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r12: helper pattern\r
cc68a136 1250.dspr_TileNorm_sh:\r
e54507e8 1251 TileNormSh_markop\r
cc68a136 1252 b .dspr_loop\r
1253\r
7a7c6476 1254.dspr_TileFlip_sh:\r
e54507e8 1255 TileFlipSh_markop\r
cc68a136 1256 b .dspr_loop\r
1257\r
1258\r
cc68a136 1259@ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r
1260\r
ea38612f 1261@ void DrawWindow(int tstart, int tend, int prio, int sh\r
1262@ struct PicoEState *est)\r
1263\r
1264.global DrawWindow\r
cc68a136 1265\r
1266DrawWindow:\r
ea38612f 1267 ldr r12, [sp] @ est\r
cc68a136 1268 stmfd sp!, {r4-r11,lr}\r
1269\r
ea38612f 1270 ldr r6, [r12, #OFS_Pico_video]\r
1271 ldr r10, [r12, #OFS_DrawScanline]\r
1272 mov r11, r12 @ est\r
1273 ldrb r12, [r6, #3] @ pvid->reg[3]\r
cc68a136 1274\r
ea38612f 1275 ldr r4, [r6, #12]\r
cc68a136 1276 mov r5, r10, lsr #3\r
1277 and r10, r10, #7\r
1278 mov r10, r10, lsl #1 @ r10=ty\r
1279\r
1280 mov r12, r12, lsl #10\r
1281\r
1282 tst r4, #1 @ 40 cell mode?\r
1283 andne r12, r12, #0xf000 @ 0x3c<<10\r
1284 andeq r12, r12, #0xf800\r
1285 addne r12, r12, r5, lsl #7\r
1286 addeq r12, r12, r5, lsl #6 @ nametab\r
1287 add r12, r12, r0, lsl #2 @ +starttile\r
1288\r
ea38612f 1289 ldr lr, [r11, #OFS_Pico_vram]\r
1290 ldr r6, [r11, #OFS_rendstatus]\r
cc68a136 1291\r
1292 @ fetch the first code now\r
1293 ldrh r7, [lr, r12]\r
1294\r
283fec1b 1295 ands r6, r6, #PDRAW_WND_DIFF_PRIO\r
cc68a136 1296 orr r6, r6, r2\r
cc68a136 1297\r
7292c709 1298 eoreq r8, r2, r7, lsr #15 @ do prio bits differ?\r
1299 cmpeq r8, #1\r
1300 ldmeqfd sp!, {r4-r11,pc} @ yes, assume that whole window uses same priority\r
cc68a136 1301\r
cc68a136 1302 orr r6, r6, r3, lsl #8 @ shadow mode\r
1303\r
1304 sub r8, r1, r0\r
cc68a136 1305\r
1306 @ cache some stuff to avoid mem access\r
99bdfd31 1307 ldr r11, [r11, #OFS_HighCol]\r
ea8c405f 1308 mov r8, r8, lsl #1 @ cells\r
ea8c405f 1309 add r11,r11,#8\r
ea8c405f 1310 mvn r9, #0 @ r9=prevcode=-1\r
07abbab1 1311 add r1, r11, r0, lsl #4 @ r1=pdest\r
cc68a136 1312 mov r0, #0xf\r
1313 b .dwloop_enter\r
1314\r
07abbab1 1315 @ r4,r5 are scratch in this loop\r
cc68a136 1316.dwloop:\r
1317 add r1, r1, #8\r
1318.dwloop_nor1:\r
1319 add r12, r12, #2 @ halfwords\r
1320 ldrh r7, [lr, r12] @ r7=code (int, but from unsigned, no sign extend)\r
1321 subs r8, r8, #1\r
1322 beq .dwloop_end @ done\r
1323\r
1324 eor r5, r6, r7, lsr #15\r
1325 tst r5, #1\r
1326 orrne r6, r6, #2 @ wrong pri\r
1327 bne .dwloop\r
1328\r
1329 cmp r7, r9\r
1330 beq .dw_samecode @ we know stuff about this tile already\r
1331\r
1332.dwloop_enter:\r
1333 mov r9, r7 @ remember code\r
1334\r
1335 movs r2, r9, lsl #20 @ if (code&0x1000)\r
1336 mov r2, r2, lsl #1\r
1337 add r2, r10, r2, lsr #17 @ r2=addr=(code&0x7ff)<<4; addr+=ty\r
1338 eorcs r2, r2, #0x0e @ if (code&0x1000) addr^=0xe;\r
1339\r
1340 and r3, r9, #0x6000\r
1341 mov r3, r3, lsr #9 @ r3=pal=((code&0x6000)>>9);\r
1342\r
1343 ldr r2, [lr, r2, lsl #1] @ pack=*(unsigned int *)(Pico.vram+addr); // Get 8 pixels\r
1344\r
1345.dw_samecode:\r
1346 tst r6, #0x100\r
1347 bne .dw_shadow\r
1348.dw_shadow_done:\r
1349 tst r2, r2\r
1350 beq .dwloop @ tileline blank\r
1351\r
1352 cmp r2, r2, ror #4\r
1353 beq .dw_SingleColor @ tileline singlecolor \r
1354\r
1355 tst r9, #0x0800\r
7a7c6476 1356 bne .dw_TileFlip\r
cc68a136 1357\r
1358 @ (r1=pdest, r2=pixels8, r3=pal) r4: scratch, r0: helper pattern\r
cc68a136 1359.dw_TileNorm:\r
1360 TileNorm r0\r
1361 b .dwloop\r
1362\r
7a7c6476 1363.dw_TileFlip:\r
1364 TileFlip r0\r
1365 b .dwloop\r
1366\r
cc68a136 1367.dw_SingleColor:\r
1368 and r4, r0, r2 @ #0x0000000f\r
1369 orr r4, r3, r4\r
1370 orr r4, r4, r4, lsl #8\r
1371 orr r4, r4, r4, lsl #16\r
1372 mov r5, r4\r
1373 stmia r1!, {r4,r5}\r
1374 b .dwloop_nor1 @ we incremeted r1 ourselves\r
1375\r
1376.dw_shadow:\r
1377 tst r6, #1 @ hi pri?\r
1378 orreq r3, r3, #0x40\r
1379 beq .dw_shadow_done\r
1380 ldr r4, [r1]\r
07abbab1 1381 mov r5, #0x3f\r
1382 orr r5, r5, r5, lsl #8\r
1383 orr r5, r5, r5, lsl #16\r
1384 and r4, r4, r5\r
cc68a136 1385 str r4, [r1]\r
1386 ldr r4, [r1,#4]\r
07abbab1 1387 and r4, r4, r5\r
cc68a136 1388 str r4, [r1,#4]\r
1389 b .dw_shadow_done\r
1390\r
1391.dwloop_end:\r
ea38612f 1392 and r2, r6, #PDRAW_WND_DIFF_PRIO\r
1393 ldmfd sp!, {r4-r11,lr}\r
1394 ldr r0, [sp]\r
1395 ldr r1, [r0, #OFS_rendstatus]\r
1396 orr r1, r1, r2\r
1397 str r1, [r0, #OFS_rendstatus]\r
cc68a136 1398\r
ea38612f 1399 bx lr\r
cc68a136 1400\r
1401\r
1402@ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r
1403\r
1404\r
1405@ hilights 2 pixels in RGB444/BGR444 format\r
1406.macro TileDoShHi2Pixels444 reg\r
1407 mov \reg, \reg, ror #12\r
1408 adds \reg, \reg, #0x40000000\r
1409 orrcs \reg, \reg, #0xf0000000\r
1410 mov \reg, \reg, ror #28\r
1411 adds \reg, \reg, #0x40000000\r
1412 orrcs \reg, \reg, #0xf0000000\r
1413 mov \reg, \reg, ror #28\r
1414 adds \reg, \reg, #0x40000000\r
1415 orrcs \reg, \reg, #0xf0000000\r
1416 mov \reg, \reg, ror #24\r
1417 adds \reg, \reg, #0x40000000\r
1418 orrcs \reg, \reg, #0xf0000000\r
1419 mov \reg, \reg, ror #28\r
1420 adds \reg, \reg, #0x40000000\r
1421 orrcs \reg, \reg, #0xf0000000\r
1422 mov \reg, \reg, ror #28\r
1423 adds \reg, \reg, #0x40000000\r
1424 orrcs \reg, \reg, #0xf0000000\r
1425 mov \reg, \reg, ror #12\r
1426.endm\r
1427\r
1428\r
cc68a136 1429@ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r
1430\r
1431\r
cc68a136 1432@ Convert 0000bbb0 ggg0rrr0\r
1433@ to rrrrrggg gggbbbbb\r
1434\r
a39d8ba5 1435@ r2,r3 - scratch, lr = 0x001c001c, r8 = 0x08610861\r
cc68a136 1436.macro convRGB565 reg\r
a39d8ba5 1437 and r2, lr, \reg,lsr #7 @ b\r
1438 and r3, lr, \reg,lsr #3 @ g\r
1439 and \reg, lr, \reg,lsl #1 @ r\r
1440 orr r2, r2, r3, lsl #6\r
1441 orr \reg, r2, \reg,lsl #11\r
1442\r
1443 and r2, r8, \reg,lsr #4\r
1444 orr \reg, \reg, r2\r
cc68a136 1445.endm\r
1446\r
a39d8ba5 1447@ trashes: r2-r8,r12,lr; r8 = 0x08610861; r0,r1 are advanced\r
b2305d08 1448.macro vidConvCpyRGB565_local\r
2ec14aec 1449 mov r12, r2, lsr #3 @ repeats\r
cc68a136 1450 mov lr, #0x001c0000\r
1451 orr lr, lr, #0x01c @ lr == pattern 0x001c001c\r
cc68a136 1452\r
b2305d08 14530:\r
cc68a136 1454 ldmia r1!, {r4-r7}\r
1455 subs r12, r12, #1\r
1456 convRGB565 r4\r
1457 str r4, [r0], #4\r
1458 convRGB565 r5\r
1459 str r5, [r0], #4\r
1460 convRGB565 r6\r
1461 str r6, [r0], #4\r
1462 convRGB565 r7\r
1463 str r7, [r0], #4\r
1464\r
b2305d08 1465 bgt 0b\r
1466.endm\r
1467\r
1468\r
1469.global vidConvCpyRGB565\r
cc68a136 1470\r
b2305d08 1471vidConvCpyRGB565: @ void *to, void *from, int pixels\r
1472 stmfd sp!, {r4-r9,lr}\r
a39d8ba5 1473 mov r8, #0x0061\r
1474 orr r8, r8, #0x0800\r
f4750ee0 1475 orr r8, r8, r8, lsl #16\r
b2305d08 1476 vidConvCpyRGB565_local\r
cc68a136 1477 ldmfd sp!, {r4-r9,lr}\r
1478 bx lr\r
1479\r
1480\r
ea38612f 1481@ void PicoDoHighPal555(int sh, int line, struct PicoEState *est)\r
1482\r
1483.global PicoDoHighPal555\r
cc68a136 1484\r
b2305d08 1485PicoDoHighPal555:\r
ea38612f 1486 stmfd sp!, {r4-r10,lr}\r
1487 mov r10,r2 @ est\r
b2305d08 1488 mov r1, #0\r
ea38612f 1489 ldr r8, [r10, #OFS_Pico_video]\r
cc68a136 1490\r
b2305d08 1491PicoDoHighPal555_nopush:\r
a39d8ba5 1492 orr r9, r1, r0, lsl #31 @ 0:called from FinalizeLine555, 31: s/h\r
b2305d08 1493\r
98a27142 1494 add r0, r10, #OFS_HighPal\r
b2305d08 1495\r
cc68a136 1496 mov r1, #0\r
b2305d08 1497 strb r1, [r8, #-0x1a] @ 0x2220e ~ dirtyPal\r
1498\r
3d48f143 1499 sub r1, r8, #0x128 @ r1=Pico.cram\r
cc68a136 1500 mov r2, #0x40\r
a39d8ba5 1501 mov r8, #0x0061\r
1502 orr r8, r8, #0x0800\r
f4750ee0 1503 orr r8, r8, r8, lsl #16\r
a39d8ba5 1504\r
b2305d08 1505 vidConvCpyRGB565_local\r
cc68a136 1506\r
a39d8ba5 1507 tst r9, #(1<<31)\r
b2305d08 1508 beq PicoDoHighPal555_end\r
1509\r
98a27142 1510 add r3, r10, #OFS_HighPal\r
cc68a136 1511\r
1512 @ shadowed pixels:\r
1513 mov r12, #0x008e\r
cc68a136 1514 add r4, r3, #0x40*2\r
b2305d08 1515 orr r12,r12,#0x7300\r
cc68a136 1516 add r5, r3, #0xc0*2\r
b2305d08 1517 orr r12,r12,r12,lsl #16\r
cc68a136 1518 mov lr, #0x40/4\r
1519.fl_loopcpRGB555_sh:\r
1520 ldmia r3!, {r1,r6}\r
1521 subs lr, lr, #1\r
1522 and r1, r12, r1, lsr #1\r
1523 and r6, r12, r6, lsr #1\r
1524 stmia r4!, {r1,r6}\r
1525 stmia r5!, {r1,r6}\r
1526 bne .fl_loopcpRGB555_sh\r
1527\r
1528 @ hilighted pixels:\r
a39d8ba5 1529 @ t = ((dpal[i] >> 1) & 0x738e738e) + 0x738e738e;\r
1530 @ t |= (t >> 4) & 0x08610861;\r
1531 @ r8=0x08610861\r
cc68a136 1532 sub r3, r3, #0x40*2\r
a39d8ba5 1533 mov lr, #0x40/4\r
cc68a136 1534.fl_loopcpRGB555_hi:\r
a39d8ba5 1535 ldmia r3!, {r1,r6}\r
1536 and r1, r12, r1, lsr #1\r
1537 and r6, r12, r6, lsr #1\r
1538 add r1, r12, r1\r
1539 add r6, r12, r6\r
1540 and r5, r8, r1, lsr #4\r
1541 and r7, r8, r6, lsr #4\r
1542 orr r1, r1, r5\r
1543 orr r6, r6, r7\r
1544 stmia r4!, {r1,r6}\r
cc68a136 1545 subs lr, lr, #1\r
1546 bne .fl_loopcpRGB555_hi\r
b2305d08 1547 mov r0, #1\r
cc68a136 1548\r
b2305d08 1549PicoDoHighPal555_end:\r
a39d8ba5 1550 tst r9, #1\r
ea38612f 1551 ldmeqfd sp!, {r4-r10,pc}\r
b2305d08 1552\r
ea38612f 1553 ldr r8, [r10, #OFS_Pico_video]\r
b2305d08 1554 b FinalizeLineRGB555_pal_done\r
1555\r
1556\r
ea38612f 1557@ void FinalizeLine555(int sh, int line, struct PicoEState *est)\r
1558\r
1559.global FinalizeLine555\r
b2305d08 1560\r
5a681086 1561FinalizeLine555:\r
ea38612f 1562 stmfd sp!, {r4-r10,lr}\r
1563 mov r10,r2 @ est\r
1564 ldr r8, [r10, #OFS_Pico_video]\r
b2305d08 1565\r
1566 ldrb r2, [r8, #-0x1a] @ 0x2220e ~ dirtyPal\r
1567 mov r1, #1\r
1568 tst r2, r2\r
1569 bne PicoDoHighPal555_nopush\r
1570\r
1571FinalizeLineRGB555_pal_done:\r
98a27142 1572 add r3, r10, #OFS_HighPal\r
cc68a136 1573\r
ea38612f 1574 ldr r12, [r10, #OFS_rendstatus]\r
b2305d08 1575 eors r0, r0, #1 @ sh is 0\r
e5fa9817 1576 mov lr, #0xff\r
283fec1b 1577 tstne r12,#PDRAW_ACC_SPRITES\r
e5fa9817 1578 movne lr, #0x3f\r
1579\r
99bdfd31 1580 ldr r1, [r10, #OFS_HighCol]\r
1581 ldr r0, [r10, #OFS_DrawLineDest]\r
ea8c405f 1582 add r1, r1, #8\r
3d48f143 1583\r
1584 ldrb r12, [r8, #12]\r
cc68a136 1585 mov lr, lr, lsl #1\r
1586\r
3d48f143 1587 tst r12, #1\r
1588 movne r2, #320/8 @ len\r
1589 bne .fl_no32colRGB555\r
99bdfd31 1590 ldr r4, [r10, #OFS_PicoOpt]\r
3d48f143 1591 mov r2, #256/8\r
1592 ldr r4, [r4]\r
1593 tst r4, #0x4000\r
1594 bne .fl_32scale_RGB555\r
1595 tst r4, #0x0100\r
1596 addeq r0, r0, #32*2\r
1597\r
1598.fl_no32colRGB555:\r
cc68a136 1599\r
f62850ba 1600#ifdef UNALIGNED_DRAWLINEDEST\r
499a0be3 1601 @ this is basically for Gizmondo, which has unaligned odd lines in the framebuffer\r
1602 tst r0, #2\r
1603 bne .fl_RGB555u\r
f62850ba 1604#endif\r
499a0be3 1605\r
1606.fl_loopRGB555:\r
cc68a136 1607 ldr r12, [r1], #4\r
1608 ldr r7, [r1], #4\r
1609\r
1610 and r4, lr, r12, lsl #1\r
1611 ldrh r4, [r3, r4]\r
1612 and r5, lr, r12, lsr #7\r
1613 ldrh r5, [r3, r5]\r
1614 and r6, lr, r12, lsr #15\r
1615 ldrh r6, [r3, r6]\r
1616 orr r4, r4, r5, lsl #16\r
1617\r
1618 and r5, lr, r12, lsr #23\r
1619 ldrh r5, [r3, r5]\r
1620 and r8, lr, r7, lsl #1\r
1621 ldrh r8, [r3, r8]\r
1622 orr r5, r6, r5, lsl #16\r
1623\r
1624 and r6, lr, r7, lsr #7\r
1625 ldrh r6, [r3, r6]\r
1626 and r12,lr, r7, lsr #15\r
1627 ldrh r12,[r3, r12]\r
499a0be3 1628 and r7, lr, r7, lsr #23\r
1629 ldrh r7, [r3, r7]\r
cc68a136 1630 orr r8, r8, r6, lsl #16\r
1631\r
cc68a136 1632 subs r2, r2, #1\r
499a0be3 1633 orr r12,r12, r7, lsl #16\r
cc68a136 1634\r
1635 stmia r0!, {r4,r5,r8,r12}\r
1636 bne .fl_loopRGB555\r
1637\r
ea38612f 1638 ldmfd sp!, {r4-r10,lr}\r
3d48f143 1639 bx lr\r
1640\r
1641\r
1642.fl_32scale_RGB555:\r
3d48f143 1643 mov r9, #0x3900 @ f800 07e0 001f | e000 0780 001c | 3800 01e0 0007\r
1644 orr r9, r9, #0x00e7\r
1645\r
f62850ba 1646#ifdef UNALIGNED_DRAWLINEDEST\r
499a0be3 1647 tst r0, #2\r
1648 bne .fl_32scale_RGB555u\r
f62850ba 1649#endif\r
499a0be3 1650\r
3d48f143 1651.fl_loop32scale_RGB555:\r
1652 ldr r12, [r1], #4\r
1653 ldr r7, [r1], #4\r
cc68a136 1654\r
3d48f143 1655 and r4, lr, r12,lsl #1\r
1656 ldrh r4, [r3, r4]\r
1657 and r5, lr, r12,lsr #7\r
1658 ldrh r5, [r3, r5]\r
1659 and r4, r4, r9, lsl #2\r
1660 orr r4, r4, r4, lsl #14 @ r4[31:16] = 1/4 pix_s 0\r
1661 and r5, r5, r9, lsl #2\r
1662 sub r6, r5, r5, lsr #2 @ r6 = 3/4 pix_s 1\r
1663 add r4, r4, r6, lsl #16 @ pix_d 0, 1\r
1664 and r6, lr, r12,lsr #15\r
1665 ldrh r6, [r3, r6]\r
1666 and r12,lr, r12,lsr #23\r
1667 ldrh r12,[r3, r12]\r
1668 and r6, r6, r9, lsl #2\r
1669 add r5, r5, r6\r
1670 mov r5, r5, lsr #1\r
1671 sub r6, r6, r6, lsr #2 @ r6 = 3/4 pix_s 2\r
1672 orr r5, r5, r6, lsl #16\r
1673\r
1674 and r6, lr, r7, lsl #1\r
1675 ldrh r6, [r3, r6]\r
1676 and r12,r12,r9, lsl #2\r
1677 add r5, r5, r12,lsl #14 @ pix_d 2, 3\r
1678 and r6, r6, r9, lsl #2\r
1679 orr r6, r12,r6, lsl #16 @ pix_d 4, 5\r
1680\r
1681 and r12,lr, r7, lsr #7\r
1682 ldrh r12,[r3, r12]\r
1683 and r10,lr, r7, lsr #15\r
1684 ldrh r10,[r3, r10]\r
1685 and r12,r12,r9, lsl #2\r
1686 sub r8, r12,r12,lsr #2 @ r8 = 3/4 pix_s 1\r
1687 add r8, r8, r6, lsr #18\r
1688 and r7, lr, r7, lsr #23\r
1689 ldrh r7, [r3, r7]\r
1690 and r10,r10,r9, lsl #2\r
1691 orr r8, r8, r10,lsl #15\r
1692 add r8, r8, r12,lsl #15 @ pix_d 6, 7\r
1693 sub r10,r10,r10,lsr #2 @ r10= 3/4 pix_s 2\r
1694 and r7, r7, r9, lsl #2\r
1695 add r10,r10,r7, lsr #2 @ += 1/4 pix_s 3\r
1696 orr r10,r10,r7, lsl #16 @ pix_d 8, 9\r
1697\r
1698 subs r2, r2, #1\r
1699\r
1700 stmia r0!, {r4,r5,r6,r8,r10}\r
1701 bne .fl_loop32scale_RGB555\r
1702\r
ea38612f 1703 ldmfd sp!, {r4-r10,lr}\r
3d48f143 1704 bx lr\r
1705\r
f62850ba 1706#ifdef UNALIGNED_DRAWLINEDEST\r
499a0be3 1707 @ unaligned versions of loops\r
9839d126 1708 @ warning: starts drawing 2bytes before dst\r
499a0be3 1709\r
1710.fl_RGB555u:\r
9839d126 1711 sub r0, r0, #2 @ initial adjustment\r
1712 mov r8, #0\r
499a0be3 1713\r
1714.fl_loopRGB555u:\r
1715 ldr r12, [r1], #4\r
1716 ldr r7, [r1], #4\r
1717\r
9839d126 1718 and r6, lr, r12,lsl #1\r
1719 ldrh r6, [r3, r6]\r
1720 and r5, lr, r12,lsr #7\r
499a0be3 1721 ldrh r5, [r3, r5]\r
9839d126 1722 orr r4, r8, r6, lsl #16\r
499a0be3 1723\r
9839d126 1724 and r6, lr, r12,lsr #15\r
499a0be3 1725 ldrh r6, [r3, r6]\r
9839d126 1726 and r8, lr, r12,lsr #23\r
499a0be3 1727 ldrh r8, [r3, r8]\r
9839d126 1728 orr r5, r5, r6, lsl #16\r
499a0be3 1729\r
9839d126 1730 and r6, lr, r7, lsl #1\r
499a0be3 1731 ldrh r6, [r3, r6]\r
9839d126 1732 and r12,lr, r7, lsr #7\r
499a0be3 1733 ldrh r12,[r3, r12]\r
9839d126 1734 orr r6, r8, r6, lsl #16\r
1735\r
1736 and r8, lr, r7, lsr #15\r
499a0be3 1737 ldrh r8, [r3, r8]\r
9839d126 1738 and r7, lr, r7, lsr #23\r
499a0be3 1739\r
1740 subs r2, r2, #1\r
9839d126 1741 orr r12,r12,r8, lsl #16\r
1742 ldrh r8, [r3, r7]\r
1743\r
1744 stmia r0!, {r4,r5,r6,r12}\r
499a0be3 1745 bne .fl_loopRGB555u\r
1746\r
1747 strh r8, [r0], #2\r
1748\r
ea38612f 1749 ldmfd sp!, {r4-r10,lr}\r
499a0be3 1750 bx lr\r
1751\r
1752\r
1753.fl_32scale_RGB555u:\r
9839d126 1754 sub r0, r0, #2 @ initial adjustment\r
1755 mov r4, #0\r
499a0be3 1756\r
1757 @ r9 f800 07e0 001f | e000 0780 001c | 3800 01e0 0007\r
1758.fl_loop32scale_RGB555u:\r
1759 ldr r12, [r1], #4\r
1760 ldr r7, [r1], #4\r
1761\r
9839d126 1762 and r6, lr, r12,lsl #1\r
1763 ldrh r6, [r3, r6]\r
499a0be3 1764 and r5, lr, r12,lsr #7\r
1765 ldrh r5, [r3, r5]\r
9839d126 1766 and r6, r6, r9, lsl #2\r
1767 orr r4, r4, r6, lsl #16 @ r4 = pix_d -1, 0\r
499a0be3 1768\r
1769 and r5, r5, r9, lsl #2\r
9839d126 1770 sub r8, r5, r5, lsr #2 @ r8 = 3/4 pix_s 1\r
1771 add r6, r8, r6, lsr #2 @ r6 = (1/4 pix_s 0) + (3/4 pix_s 1)\r
1772 orr r5, r6, r5, lsl #15\r
499a0be3 1773\r
1774 and r6, lr, r12,lsr #15\r
1775 ldrh r6, [r3, r6]\r
1776 and r12,lr, r12,lsr #23\r
1777 ldrh r12,[r3, r12]\r
1778 and r6, r6, r9, lsl #2\r
9839d126 1779 add r5, r5, r6, lsl #15 @ r5 = pix_d 1, 2\r
499a0be3 1780\r
9839d126 1781 and r8, lr, r7, lsl #1\r
1782 ldrh r8, [r3, r8]\r
1783 and r10,lr, r7, lsr #7\r
1784 ldrh r10,[r3, r10]\r
499a0be3 1785 and r12,r12,r9, lsl #2\r
9839d126 1786 sub r6, r6, r6, lsr #2 @ r6 = 3/4 pix_s 2\r
1787 add r6, r6, r12,lsr #2\r
1788 orr r6, r6, r12,lsl #16 @ r6 = pix_d 3, 4\r
499a0be3 1789\r
9839d126 1790 and r8, r8, r9, lsl #2\r
1791 and r10,r10,r9, lsl #2\r
1792 sub r12,r10,r10,lsr #2 @ r12 = 3/4 pix_s 5\r
1793 orr r8, r8, r8, lsl #14\r
1794 add r8, r8, r12,lsl #16 @ r8 = pix_d 5, 6\r
1795 and r12,lr, r7, lsr #15\r
499a0be3 1796 ldrh r12,[r3, r12]\r
499a0be3 1797 and r7, lr, r7, lsr #23\r
1798 ldrh r7, [r3, r7]\r
9839d126 1799 and r12,r12,r9, lsl #2\r
1800 add r10,r10,r12\r
1801 mov r10,r10, lsr #1\r
1802 sub r12,r12,r12,lsr #2 @ r12 = 3/4 pix_s 6\r
1803 orr r10,r10,r12,lsl #16\r
499a0be3 1804 and r7, r7, r9, lsl #2\r
9839d126 1805 add r10,r10,r7, lsl #14 @ r10 = pix_d 7, 8\r
499a0be3 1806\r
1807 subs r2, r2, #1\r
1808\r
1809 stmia r0!, {r4,r5,r6,r8,r10}\r
9839d126 1810 mov r4, r7\r
499a0be3 1811 bne .fl_loop32scale_RGB555u\r
1812\r
9839d126 1813 strh r4, [r0], #2\r
1814\r
ea38612f 1815 ldmfd sp!, {r4-r10,lr}\r
499a0be3 1816 bx lr\r
1817\r
f62850ba 1818#endif /* UNALIGNED_DRAWLINEDEST */\r
499a0be3 1819\r
cc68a136 1820\r
1821@ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@\r
1822\r
1823@ utility\r
1824.global blockcpy @ void *dst, void *src, size_t n\r
1825\r
1826blockcpy:\r
1827 stmfd sp!, {r4,r5}\r
1828 mov r2, r2, lsr #4\r
1829blockcpy_loop:\r
1830 ldmia r1!, {r3-r5,r12}\r
1831 subs r2, r2, #1\r
1832 stmia r0!, {r3-r5,r12}\r
1833 bne blockcpy_loop\r
1834 ldmfd sp!, {r4,r5}\r
1835 bx lr\r
1836\r
1837\r
1838.global blockcpy_or @ void *dst, void *src, size_t n, int pat\r
1839\r
1840blockcpy_or:\r
1841 stmfd sp!, {r4-r6}\r
1842 orr r3, r3, r3, lsl #8\r
1843 orr r3, r3, r3, lsl #16\r
1844 mov r2, r2, lsr #4\r
1845blockcpy_loop_or:\r
1846 ldmia r1!, {r4-r6,r12}\r
1847 subs r2, r2, #1\r
1848 orr r4, r4, r3\r
1849 orr r5, r5, r3\r
1850 orr r6, r6, r3\r
1851 orr r12,r12,r3\r
1852 stmia r0!, {r4-r6,r12}\r
1853 bne blockcpy_loop_or\r
1854 ldmfd sp!, {r4-r6}\r
1855 bx lr\r
1856\r
cff531af 1857@ vim:filetype=armasm\r
fast MegaDrive/MegaCD/32X emulator