Details | Last modification | View Log | RSS feed
| Rev | Author | Line No. | Line |
|---|---|---|---|
| 716 | lvd | 1 | #include "std.h" |
| 2 | |||
| 3 | #include "emul.h" |
||
| 4 | #include "vars.h" |
||
| 5 | #include "dxrend.h" |
||
| 6 | #include "dxrcopy.h" |
||
| 7 | #include "dxr_512.h" |
||
| 8 | #include "dxr_4bpp.h" |
||
| 9 | #include "dxr_prof.h" |
||
| 10 | #include "dxr_atm.h" |
||
| 11 | #include "draw.h" |
||
| 12 | #include "util.h" |
||
| 13 | |||
| 14 | void rend_small(unsigned char *dst, unsigned pitch) |
||
| 15 | { |
||
| 16 | if (temp.obpp == 8) { rend_copy8 (dst, pitch); return; } |
||
| 17 | if (temp.obpp == 16) { rend_copy16(dst, pitch); return; } |
||
| 18 | if (temp.obpp == 32) { rend_copy32(dst, pitch); return; } |
||
| 19 | } |
||
| 20 | |||
| 21 | void __fastcall render_small(unsigned char *dst, unsigned pitch) |
||
| 22 | { |
||
| 23 | if (conf.noflic) |
||
| 24 | { |
||
| 25 | if (temp.obpp == 8) { rend_copy8_nf (dst, pitch); } |
||
| 26 | if (temp.obpp == 16) { rend_copy16_nf(dst, pitch); } |
||
| 27 | if (temp.obpp == 32) { rend_copy32_nf(dst, pitch); } |
||
| 28 | memcpy(rbuf_s, rbuf, temp.scy*temp.scx/4); |
||
| 29 | return; |
||
| 30 | } |
||
| 31 | |||
| 32 | if (comp.pEFF7 & EFF7_4BPP) |
||
| 33 | { |
||
| 34 | rend_p4bpp_small(dst, pitch); |
||
| 35 | return; |
||
| 36 | } |
||
| 37 | |||
| 38 | if (conf.mem_model == MM_ATM450) |
||
| 39 | { |
||
| 40 | rend_atm_1_small(dst, pitch); |
||
| 41 | return; |
||
| 42 | } |
||
| 43 | |||
| 44 | if (conf.mem_model == MM_ATM710 || conf.mem_model == MM_ATM3) |
||
| 45 | { |
||
| 46 | rend_atm_2_small(dst, pitch); |
||
| 47 | return; |
||
| 48 | } |
||
| 49 | rend_small(dst, pitch); |
||
| 50 | } |
||
| 51 | |||
| 52 | void rend_dbl(unsigned char *dst, unsigned pitch) |
||
| 53 | { |
||
| 54 | if (temp.oy > temp.scy && conf.fast_sl) |
||
| 55 | pitch *= 2; |
||
| 56 | |||
| 57 | if (conf.noflic) |
||
| 58 | { |
||
| 59 | if (temp.obpp == 8) |
||
| 60 | { |
||
| 61 | if (conf.fast_sl) |
||
| 62 | rend_copy8d1_nf (dst, pitch); |
||
| 63 | else |
||
| 64 | rend_copy8d_nf (dst, pitch); |
||
| 65 | } |
||
| 66 | else if (temp.obpp == 16) |
||
| 67 | { |
||
| 68 | if (conf.fast_sl) |
||
| 69 | rend_copy16d1_nf(dst, pitch); |
||
| 70 | else |
||
| 71 | rend_copy16d_nf(dst, pitch); |
||
| 72 | } |
||
| 73 | else if (temp.obpp == 32) |
||
| 74 | { |
||
| 75 | if (conf.fast_sl) |
||
| 76 | rend_copy32d1_nf(dst, pitch); |
||
| 77 | else |
||
| 78 | rend_copy32d_nf(dst, pitch); |
||
| 79 | } |
||
| 80 | |||
| 81 | memcpy(rbuf_s, rbuf, temp.scy * temp.scx / 4); |
||
| 82 | } |
||
| 83 | else |
||
| 84 | { |
||
| 85 | if (temp.obpp == 8) |
||
| 86 | { |
||
| 87 | if (conf.fast_sl) |
||
| 88 | rend_copy8d1 (dst, pitch); |
||
| 89 | else |
||
| 90 | rend_copy8d (dst, pitch); |
||
| 91 | return; |
||
| 92 | } |
||
| 93 | if (temp.obpp == 16) |
||
| 94 | { |
||
| 95 | if (conf.fast_sl) |
||
| 96 | rend_copy16d1(dst, pitch); |
||
| 97 | else |
||
| 98 | rend_copy16d(dst, pitch); |
||
| 99 | return; |
||
| 100 | } |
||
| 101 | if (temp.obpp == 32) |
||
| 102 | { |
||
| 103 | if (conf.fast_sl) |
||
| 104 | rend_copy32d1(dst, pitch); |
||
| 105 | else |
||
| 106 | rend_copy32d(dst, pitch); |
||
| 107 | return; |
||
| 108 | } |
||
| 109 | } |
||
| 110 | } |
||
| 111 | |||
| 112 | void __fastcall render_dbl(unsigned char *dst, unsigned pitch) |
||
| 113 | { |
||
| 114 | #ifdef MOD_VID_VD |
||
| 115 | if ((comp.pVD & 8) && temp.obpp == 8) |
||
| 116 | { |
||
| 117 | rend_vd8dbl(dst, pitch); |
||
| 118 | return; |
||
| 119 | } |
||
| 120 | #endif |
||
| 121 | |||
| 122 | // todo: add ini option to show zx-screen with palette or with MC |
||
| 123 | if (comp.pEFF7 & EFF7_512) |
||
| 124 | { |
||
| 125 | rend_512(dst, pitch); |
||
| 126 | return; |
||
| 127 | } |
||
| 128 | if (comp.pEFF7 & EFF7_4BPP) |
||
| 129 | { |
||
| 130 | rend_p4bpp(dst, pitch); |
||
| 131 | return; |
||
| 132 | } |
||
| 133 | if ((comp.pDFFD & 0x80) && conf.mem_model == MM_PROFI) |
||
| 134 | { |
||
| 135 | rend_profi(dst, pitch); |
||
| 136 | return; |
||
| 137 | } |
||
| 138 | if (conf.mem_model == MM_ATM450) |
||
| 139 | { |
||
| 140 | rend_atm_1(dst, pitch); |
||
| 141 | return; |
||
| 142 | } |
||
| 143 | if (conf.mem_model == MM_ATM710 || conf.mem_model == MM_ATM3) |
||
| 144 | { |
||
| 145 | rend_atm_2(dst, pitch); |
||
| 146 | return; |
||
| 147 | } |
||
| 148 | |||
| 149 | rend_dbl(dst, pitch); |
||
| 150 | } |
||
| 151 | |||
| 152 | void __fastcall render_3x(unsigned char *dst, unsigned pitch) |
||
| 153 | { |
||
| 154 | if (conf.noflic) { |
||
| 155 | if (temp.obpp == 8) rend_copy8t_nf (dst, pitch); |
||
| 156 | if (temp.obpp == 16) rend_copy16t_nf(dst, pitch); |
||
| 157 | if (temp.obpp == 32) rend_copy32t_nf(dst, pitch); |
||
| 158 | memcpy(rbuf_s, rbuf, temp.scy*temp.scx/4); |
||
| 159 | } |
||
| 160 | else |
||
| 161 | { |
||
| 162 | if (temp.obpp == 8) { rend_copy8t (dst, pitch); return; } |
||
| 163 | if (temp.obpp == 16) { rend_copy16t(dst, pitch); return; } |
||
| 164 | if (temp.obpp == 32) { rend_copy32t(dst, pitch); return; } |
||
| 165 | } |
||
| 166 | } |
||
| 167 | |||
| 168 | void __fastcall render_quad(unsigned char *dst, unsigned pitch) |
||
| 169 | { |
||
| 170 | if (conf.noflic) { |
||
| 171 | if (temp.obpp == 8) rend_copy8q_nf (dst, pitch); |
||
| 172 | if (temp.obpp == 16) rend_copy16q_nf(dst, pitch); |
||
| 173 | if (temp.obpp == 32) rend_copy32q_nf(dst, pitch); |
||
| 174 | memcpy(rbuf_s, rbuf, temp.scy*temp.scx/4); |
||
| 175 | } else { |
||
| 176 | if (temp.obpp == 8) { rend_copy8q (dst, pitch); return; } |
||
| 177 | if (temp.obpp == 16) { rend_copy16q(dst, pitch); return; } |
||
| 178 | if (temp.obpp == 32) { rend_copy32q(dst, pitch); return; } |
||
| 179 | } |
||
| 180 | } |
||
| 181 | |||
| 182 | |||
| 183 | void __fastcall render_scale(unsigned char *dst, unsigned pitch) |
||
| 184 | { |
||
| 185 | unsigned char *src = rbuf; |
||
| 186 | unsigned dx = temp.scx / 4; |
||
| 187 | unsigned char buf[MAX_WIDTH*2]; |
||
| 188 | unsigned x; //Alone Coder 0.36.7 |
||
| 189 | for (unsigned y = 0; y < temp.scy-1; y++) |
||
| 190 | { |
||
| 191 | for (x = 0; x < dx; x += 2) |
||
| 192 | { |
||
| 193 | unsigned xx = (t.dbl[src[x]] << 16) + t.dbl[src[x+2]]; |
||
| 194 | unsigned yy = (t.dbl[src[x+dx]] << 16) + t.dbl[src[x+dx+2]]; |
||
| 195 | unsigned x1 = xx | (yy & ((xx>>1) | (xx<<1))); |
||
| 196 | unsigned *tab0 = t.sctab8[0] + (src[x+1] << 4); |
||
| 197 | *(unsigned*)(dst+x*8+ 0) = tab0[(x1>>28) & 0x0F]; |
||
| 198 | *(unsigned*)(dst+x*8+ 4) = tab0[(x1>>24) & 0x0F]; |
||
| 199 | *(unsigned*)(dst+x*8+ 8) = tab0[(x1>>20) & 0x0F]; |
||
| 200 | *(unsigned*)(dst+x*8+12) = tab0[(x1>>16) & 0x0F]; |
||
| 201 | unsigned *tab1 = t.sctab8[0] + src[x+3]; |
||
| 202 | *(unsigned*)(dst+x*8+16) = tab1[(x1>>12) & 0x0F]; |
||
| 203 | *(unsigned*)(dst+x*8+20) = tab1[(x1>> 8) & 0x0F]; |
||
| 204 | *(unsigned*)(dst+x*8+24) = tab1[(x1>> 4) & 0x0F]; |
||
| 205 | *(unsigned*)(dst+x*8+28) = tab1[(x1>> 0) & 0x0F]; |
||
| 206 | x1 = yy | (xx & ((yy>>1) | (yy<<1))); |
||
| 207 | *(unsigned*)(buf+x*8+ 0) = tab0[(x1>>28) & 0x0F]; |
||
| 208 | *(unsigned*)(buf+x*8+ 4) = tab0[(x1>>24) & 0x0F]; |
||
| 209 | *(unsigned*)(buf+x*8+ 8) = tab0[(x1>>20) & 0x0F]; |
||
| 210 | *(unsigned*)(buf+x*8+12) = tab0[(x1>>16) & 0x0F]; |
||
| 211 | *(unsigned*)(buf+x*8+16) = tab1[(x1>>12) & 0x0F]; |
||
| 212 | *(unsigned*)(buf+x*8+20) = tab1[(x1>> 8) & 0x0F]; |
||
| 213 | *(unsigned*)(buf+x*8+24) = tab1[(x1>> 4) & 0x0F]; |
||
| 214 | *(unsigned*)(buf+x*8+28) = tab1[(x1>> 0) & 0x0F]; |
||
| 215 | } |
||
| 216 | dst += pitch; |
||
| 217 | for (x = 0; x < temp.ox; x += 4) |
||
| 218 | *(unsigned*)(dst+x) = *(unsigned*)(buf+x); |
||
| 219 | src += dx; dst += pitch; |
||
| 220 | } |
||
| 221 | } |
||
| 222 | |||
| 223 | static u64 mask49 = 0x4949494949494949ULL; |
||
| 224 | static u64 mask92 = 0x9292929292929292ULL; |
||
| 225 | |||
| 226 | static void /*__declspec(naked)*/ __fastcall _bil_line1(unsigned char *dst, unsigned char *src) |
||
| 227 | { |
||
| 228 | for (unsigned i = 0; i < temp.scx; i += 2) |
||
| 229 | { |
||
| 230 | dst[i] = src[i]; |
||
| 231 | dst[i+1] = ((src[i] + src[i+1]) >> 1); |
||
| 232 | } |
||
| 233 | /* |
||
| 234 | __asm { |
||
| 235 | |||
| 236 | push ebx |
||
| 237 | push edi |
||
| 238 | push ebp |
||
| 239 | |||
| 240 | mov ebp, [temp.scx] |
||
| 241 | xor eax, eax |
||
| 242 | xor ebx, ebx // ebx - prev. pixel |
||
| 243 | shr ebp,1 |
||
| 244 | |||
| 245 | l1: |
||
| 246 | mov al, [edx] |
||
| 247 | xadd eax, ebx |
||
| 248 | shr eax, 1 |
||
| 249 | mov [ecx+1], bl |
||
| 250 | mov [ecx], al |
||
| 251 | mov al, [edx+1] |
||
| 252 | add ecx, 4 |
||
| 253 | xadd eax, ebx |
||
| 254 | add edx, 2 |
||
| 255 | shr eax, 1 |
||
| 256 | mov [ecx-1], bl |
||
| 257 | dec ebp |
||
| 258 | mov [ecx-2], al |
||
| 259 | jnz l1 |
||
| 260 | |||
| 261 | pop ebp |
||
| 262 | pop edi |
||
| 263 | pop ebx |
||
| 264 | retn |
||
| 265 | } |
||
| 266 | */ |
||
| 267 | } |
||
| 268 | |||
| 269 | static void /*__declspec(naked)*/ __fastcall _bil_line2(unsigned char *dst, unsigned char *s1) |
||
| 270 | { |
||
| 271 | u32 *s = (u32 *)s1; |
||
| 272 | u32 *d = (u32 *)dst; |
||
| 273 | |||
| 274 | for (unsigned j = 0; j < temp.ox/4; j++) |
||
| 275 | { |
||
| 276 | u32 a = s[j]; |
||
| 277 | u32 b = s[j+2*MAX_WIDTH/4]; |
||
| 278 | u32 x = a & b; |
||
| 279 | u32 y = (a ^ b) >> 1; |
||
| 280 | u32 z = a | b; |
||
| 281 | u32 n = x << 1; |
||
| 282 | u32 v1 = x ^ y; |
||
| 283 | v1 &= 0x49494949; |
||
| 284 | u32 v2 = z & n; |
||
| 285 | v2 |= x; |
||
| 286 | v2 &= 0x92929292; |
||
| 287 | |||
| 288 | d[j] = v1 | v2; |
||
| 289 | } |
||
| 290 | |||
| 291 | /* |
||
| 292 | __asm { |
||
| 293 | |||
| 294 | mov eax, [temp.ox] |
||
| 295 | movq mm2, [mask49] |
||
| 296 | movq mm3, [mask92] |
||
| 297 | shr eax, 3 |
||
| 298 | |||
| 299 | m2: movq mm0, [edx] |
||
| 300 | movq mm1, [edx+MAX_WIDTH*2] |
||
| 301 | movq mm4, mm0 |
||
| 302 | movq mm5, mm0 |
||
| 303 | pand mm4, mm1 // mm4 = a & b |
||
| 304 | pxor mm5, mm1 // mm5 = a ^ b |
||
| 305 | movq mm6, mm0 |
||
| 306 | psrlq mm5, 1 // mm5 = (a ^ b) >> 1 |
||
| 307 | por mm6, mm1 // mm6 = a | b |
||
| 308 | movq mm7, mm4 |
||
| 309 | pxor mm5, mm4 // mm5 = (a & b) ^ ((a ^ b) >> 1) |
||
| 310 | psllq mm7, 1 // mm7 = (a & b) << 1 |
||
| 311 | pand mm5, mm2 // mm5 = 0x49494949 & ((a & b) ^ ((a ^ b) >> 1)) |
||
| 312 | pand mm7, mm6 // mm7 = (a|b) & ((a & b) << 1) |
||
| 313 | por mm7, mm4 // mm7 = (a&b) | ((a|b)&((a&b)<<1)) |
||
| 314 | add ecx, 8 |
||
| 315 | pand mm7, mm3 // mm7 &= 0x92929292 |
||
| 316 | add edx, 8 |
||
| 317 | por mm7, mm5 |
||
| 318 | dec eax |
||
| 319 | movq [ecx-8], mm7 |
||
| 320 | jnz m2 |
||
| 321 | |||
| 322 | retn |
||
| 323 | } |
||
| 324 | */ |
||
| 325 | } |
||
| 326 | |||
| 327 | void __fastcall render_bil(unsigned char *dst, unsigned pitch) |
||
| 328 | { |
||
| 329 | render_small(snbuf, MAX_WIDTH); |
||
| 330 | |||
| 331 | unsigned char *src = snbuf; |
||
| 332 | unsigned char ATTR_ALIGN(16) l1[MAX_WIDTH*4]; |
||
| 333 | #define l2 (l1+MAX_WIDTH*2) |
||
| 334 | _bil_line1(l1, src); src += MAX_WIDTH; |
||
| 335 | memcpy(dst, l1, temp.ox); |
||
| 336 | dst += pitch; |
||
| 337 | |||
| 338 | for (unsigned i = temp.scy/2-1; i; i--) |
||
| 339 | { |
||
| 340 | _bil_line1(l2, src); src += MAX_WIDTH; |
||
| 341 | _bil_line2(dst, l1); dst += pitch; |
||
| 342 | memcpy(dst, l2, temp.ox); |
||
| 343 | dst += pitch; |
||
| 344 | |||
| 345 | _bil_line1(l1, src); src += MAX_WIDTH; |
||
| 346 | _bil_line2(dst, l1); dst += pitch; |
||
| 347 | memcpy(dst, l1, temp.ox); |
||
| 348 | dst += pitch; |
||
| 349 | } |
||
| 350 | _bil_line1(l2, src); src += MAX_WIDTH; |
||
| 351 | _bil_line2(dst, l1); dst += pitch; |
||
| 352 | memcpy(dst, l2, temp.ox); |
||
| 353 | dst += pitch; |
||
| 354 | memcpy(dst, l2, temp.ox); |
||
| 355 | #undef l2 |
||
| 356 | |||
| 357 | // _mm_empty(); |
||
| 358 | } |
||
| 359 | |||
| 360 | void __fastcall render_tv(unsigned char *dst, unsigned pitch) |
||
| 361 | { |
||
| 362 | // ripped from ccs and *highly* simplified and optimized |
||
| 363 | |||
| 364 | unsigned char midbuf[MAX_WIDTH*2]; |
||
| 365 | unsigned char line[MAX_WIDTH*2+4*2], line2[MAX_WIDTH*2]; |
||
| 366 | |||
| 367 | unsigned j; //Alone Coder 0.36.7 |
||
| 368 | for (/*unsigned*/ j = 0; j < MAX_WIDTH/2; j++) |
||
| 369 | *(unsigned*)(midbuf+j*4) = WORD4(0,0x80,0,0x80); |
||
| 370 | |||
| 371 | unsigned char *src = rbuf; unsigned delta = temp.scx/4; |
||
| 372 | |||
| 373 | for (unsigned i = temp.scy; i; i--) { |
||
| 374 | *(unsigned*)line = *(unsigned*)(line+4) = WORD4(0,0x80,0,0x80); |
||
| 375 | |||
| 376 | if (conf.noflic) line16_nf(line+8, src, t.sctab16_nf[0]); |
||
| 377 | else line16(line+8, src, t.sctab16[0]); |
||
| 378 | |||
| 379 | src += delta; |
||
| 380 | |||
| 381 | for (j = 0; j < temp.scx; j++) { |
||
| 382 | |||
| 383 | unsigned Y = line[j*2+8]*9+ |
||
| 384 | line[j*2-2+8]*4+ |
||
| 385 | line[j*2-4+8]*2+ |
||
| 386 | line[j*2-8+8]; |
||
| 387 | /* |
||
| 388 | unsigned U = line[j*2+8+1]*12 + |
||
| 389 | line[j*2-2+8+1]*2+ |
||
| 390 | line[j*2-4+8+1]+ |
||
| 391 | line[j*2-8+8+1]; |
||
| 392 | */ |
||
| 393 | line2[j*2] = Y>>4; |
||
| 394 | // line2[j*2+1] = U>>4; |
||
| 395 | line2[j*2+1] = line[j*2+9]; |
||
| 396 | } |
||
| 397 | // there must be only fixed length fader buffer |
||
| 398 | for (j = 0; j < temp.scx/2; j++) { |
||
| 399 | *(unsigned*)(midbuf+j*4) = *(unsigned*)(dst + j*4) = |
||
| 400 | ((*(unsigned*)(midbuf+j*4) & 0xFEFEFEFE)/2 + (*(unsigned*)(line2+j*4) & 0xFEFEFEFE)/2); |
||
| 401 | } |
||
| 402 | dst += pitch; |
||
| 403 | } |
||
| 404 | if (conf.noflic) memcpy(rbuf_s, rbuf, temp.scy*temp.scx/4); |
||
| 405 | } |
||
| 406 |