From 4b52a514bd4f6202cf5ba162ebcaee3a5d503b29 Mon Sep 17 00:00:00 2001 From: TC1995 Date: Fri, 5 Jan 2024 21:22:01 +0100 Subject: [PATCH] Huge fixes to 8514/A compatibles: 1. Properly implemented polygon filling in the BitBLT side of the ATI 8514/A compatibles (Mach8/32), this allows games like Mj8514 and demos like HDIDEMO from IBM to run under ATI's hdiload 1.1 properly. 2. Finally figured out the polygon filling command in the IBM one about read and write masks (Command 5 and Command 2 with polygon filling bits on, currently only for the read mask one), this allows the above samples to render properly with IBM's original hdiload 1.0 from 1987. --- src/include/86box/vid_8514a.h | 8 +- src/video/vid_8514a.c | 764 +++++++++++++++++----------------- src/video/vid_ati_mach8.c | 551 +++++++++++++----------- 3 files changed, 703 insertions(+), 620 deletions(-) diff --git a/src/include/86box/vid_8514a.h b/src/include/86box/vid_8514a.h index 49b4cb37d..4d7483f30 100644 --- a/src/include/86box/vid_8514a.h +++ b/src/include/86box/vid_8514a.h @@ -67,13 +67,14 @@ typedef struct ibm8514_t { uint16_t advfunc_cntl; uint8_t ext_advfunc_cntl; uint16_t cur_y; - uint16_t cur_y_bitres; uint16_t cur_x; - uint16_t cur_x_bitres; + int16_t destx; + int16_t desty; int16_t desty_axstp; int16_t destx_distp; int16_t err_term; int16_t maj_axis_pcnt; + int16_t maj_axis_pcnt_no_limit; uint16_t cmd; uint16_t cmd_back; uint16_t short_stroke; @@ -100,7 +101,9 @@ typedef struct ibm8514_t { int sys_cnt2; int temp_cnt; int16_t cx; + int16_t cx_back; int16_t cy; + int16_t oldcx; int16_t oldcy; int16_t sx; int16_t sy; @@ -133,6 +136,7 @@ typedef struct ibm8514_t { int fill_state; int xdir; int ydir; + int linedraw; uint32_t ge_offset; } accel; diff --git a/src/video/vid_8514a.c b/src/video/vid_8514a.c index e2a7da3ca..576562f23 100644 --- a/src/video/vid_8514a.c +++ b/src/video/vid_8514a.c @@ -447,14 +447,16 @@ ibm8514_accel_out_fifo(svga_t *svga, uint16_t port, uint32_t val, int len) { ibm8514_t *dev = (ibm8514_t *) svga->dev8514; + if (port != 0x9ae8 && port != 0xe2e8) + ibm8514_log("Port OUT FIFO=%04x, val=%04x, len=%d.\n", port, val, len); + switch (port) { case 0x82e8: case 0xc2e8: if (len == 1) { dev->accel.cur_y = (dev->accel.cur_y & 0x700) | val; - } else { + } else dev->accel.cur_y = val & 0x7ff; - } break; case 0x82e9: case 0xc2e9: @@ -465,11 +467,10 @@ ibm8514_accel_out_fifo(svga_t *svga, uint16_t port, uint32_t val, int len) case 0x86e8: case 0xc6e8: - if (len == 1) { + if (len == 1) dev->accel.cur_x = (dev->accel.cur_x & 0x700) | val; - } else { + else dev->accel.cur_x = val & 0x7ff; - } break; case 0x86e9: case 0xc6e9: @@ -483,6 +484,7 @@ ibm8514_accel_out_fifo(svga_t *svga, uint16_t port, uint32_t val, int len) if (len == 1) dev->accel.desty_axstp = (dev->accel.desty_axstp & 0x3f00) | val; else { + dev->accel.desty = val & 0x07ff; dev->accel.desty_axstp = val & 0x3fff; if (val & 0x2000) dev->accel.desty_axstp |= ~0x1fff; @@ -502,6 +504,7 @@ ibm8514_accel_out_fifo(svga_t *svga, uint16_t port, uint32_t val, int len) if (len == 1) dev->accel.destx_distp = (dev->accel.destx_distp & 0x3f00) | val; else { + dev->accel.destx = val & 0x07ff; dev->accel.destx_distp = val & 0x3fff; if (val & 0x2000) dev->accel.destx_distp |= ~0x1fff; @@ -544,6 +547,7 @@ ibm8514_accel_out_fifo(svga_t *svga, uint16_t port, uint32_t val, int len) dev->accel.maj_axis_pcnt = (dev->accel.maj_axis_pcnt & 0x700) | val; else { dev->accel.maj_axis_pcnt = val & 0x7ff; + dev->accel.maj_axis_pcnt_no_limit = val; } break; case 0x96e9: @@ -566,6 +570,7 @@ ibm8514_accel_out_fifo(svga_t *svga, uint16_t port, uint32_t val, int len) if (dev->accel.cmd & 0x100) dev->accel.cmd_back = 0; } + ibm8514_log("8514/A CMD=%04x.\n", dev->accel.cmd); ibm8514_accel_start(-1, 0, -1, 0, svga, len); } break; @@ -848,6 +853,7 @@ static void ibm8514_accel_out(uint16_t port, uint32_t val, svga_t *svga, int len) { ibm8514_t *dev = (ibm8514_t *) svga->dev8514; + uint8_t old = 0; if (port & 0x8000) { ibm8514_accel_out_fifo(svga, port, val, len); @@ -947,16 +953,40 @@ ibm8514_accel_out(uint16_t port, uint32_t val, svga_t *svga, int len) break; case 0x42e8: - if (len == 1) { - dev->subsys_stat &= ~val; - } else { - dev->subsys_stat &= ~(val & 0xff); + old = dev->subsys_stat; + if ((val & 0xff) & 1) + dev->subsys_stat &= ~1; + if ((val & 0xff) & 2) + dev->subsys_stat &= ~2; + if ((val & 0xff) & 4) + dev->subsys_stat &= ~4; + if ((val & 0xff) & 8) + dev->subsys_stat &= ~8; + if (len != 1) { + old = dev->subsys_cntl; dev->subsys_cntl = (val >> 8); + if ((old ^ dev->subsys_cntl) & 1) + dev->subsys_stat |= 1; + if ((old ^ dev->subsys_cntl) & 2) + dev->subsys_stat |= 2; + if ((old ^ dev->subsys_cntl) & 4) + dev->subsys_stat |= 4; + if ((old ^ dev->subsys_cntl) & 8) + dev->subsys_stat |= 8; } break; case 0x42e9: if (len == 1) { + old = dev->subsys_cntl; dev->subsys_cntl = val; + if ((old ^ val) & 1) + dev->subsys_stat |= 1; + if ((old ^ val) & 2) + dev->subsys_stat |= 2; + if ((old ^ val) & 4) + dev->subsys_stat |= 4; + if ((old ^ val) & 8) + dev->subsys_stat |= 8; } break; @@ -1038,6 +1068,7 @@ ibm8514_accel_in(uint16_t port, svga_t *svga, int len) break; case 0x42e8: + cmd = dev->accel.cmd >> 13; vpos = dev->vc & 0x7ff; if (vblankend > dev->v_total) { vblankend -= dev->v_total; @@ -1058,6 +1089,20 @@ ibm8514_accel_in(uint16_t port, svga_t *svga, int len) temp |= 0x80; break; + case 0x82e8: + case 0xc2e8: + if (len != 1) { + temp = dev->accel.cur_y; + } + break; + + case 0x86e8: + case 0xc6e8: + if (len != 1) { + temp = dev->accel.cur_x; + } + break; + case 0x92e8: if (len != 1) { temp = dev->test; @@ -1155,6 +1200,12 @@ ibm8514_short_stroke_start(int count, int cpu_input, uint32_t mix_dat, uint32_t ibm8514_accel_start(count, cpu_input, mix_dat, cpu_dat, svga, len); } +#define CLAMP(x) \ + do { \ + if ((x) & ~0xff) \ + x = ((x) < 0) ? 0 : 0xff; \ + } while (0) + void ibm8514_accel_start(int count, int cpu_input, uint32_t mix_dat, uint32_t cpu_dat, svga_t *svga, UNUSED(int len)) { @@ -1173,13 +1224,15 @@ ibm8514_accel_start(int count, int cpu_input, uint32_t mix_dat, uint32_t cpu_dat int compare_mode = dev->accel.multifunc[0x0a] & 0x38; int cmd = dev->accel.cmd >> 13; uint16_t wrt_mask = dev->accel.wrt_mask; + uint16_t wrt_mask_polygon = dev->accel.wrt_mask; uint16_t rd_mask = dev->accel.rd_mask; uint16_t rd_mask_polygon = dev->accel.rd_mask; uint16_t frgd_color = dev->accel.frgd_color; uint16_t bkgd_color = dev->accel.bkgd_color; uint32_t old_mix_dat; - int and3 = dev->accel.cur_x & 3; - uint16_t poly_src = 0; + int and3 = dev->accel.cur_x & 3; + uint16_t poly_src = 0; + uint16_t old_poly_src = 0; if (!dev->bpp) { compare &= 0xff; @@ -1298,10 +1351,6 @@ ibm8514_accel_start(int count, int cpu_input, uint32_t mix_dat, uint32_t cpu_dat /*Bit 4 of the Command register is the draw yes bit, which enables writing to memory/reading from memory when enabled. When this bit is disabled, no writing to memory/reading from memory is allowed. (This bit is almost meaningless on the NOP command)*/ - if (dev->accel.cmd == 0x43b3) { - ibm8514_log("CMD8514: CMD=%d, full=%04x, pixcntl=%x, count=%d, frcolor=%02x, bkcolor=%02x, polygon=%x, cpu=%08x, frgdmix=%02x, bkgdmix=%02x.\n", cmd, dev->accel.cmd, pixcntl, count, frgd_color, bkgd_color, dev->accel.multifunc[0x0a] & 6, cpu_dat, dev->accel.frgd_mix, dev->accel.bkgd_mix); - } - switch (cmd) { case 0: /*NOP (Short Stroke Vectors)*/ if (dev->accel.ssv_state == 0) @@ -1518,15 +1567,15 @@ ibm8514_accel_start(int count, int cpu_input, uint32_t mix_dat, uint32_t cpu_dat dev->accel.cx = dev->accel.cur_x; dev->accel.cy = dev->accel.cur_y; - if (dev->accel.cur_x >= 0x600) { + if (dev->accel.cur_x >= 0x600) dev->accel.cx |= ~0x5ff; - } - if (dev->accel.cur_y >= 0x600) { + + if (dev->accel.cur_y >= 0x600) dev->accel.cy |= ~0x5ff; - } dev->accel.sy = dev->accel.maj_axis_pcnt; + ibm8514_log("Line Draw 8514/A, frgdmix=%d, bkgdmix=%d, c(%d,%d), pixcntl=%d, sy=%d, polyfill=%x, selfrmix=%02x, selbkmix=%02x, bkgdcol=%02x, frgdcol=%02x, clipt=%d, clipb=%d.\n", frgd_mix, bkgd_mix, dev->accel.cx, dev->accel.cy, pixcntl, dev->accel.sy, dev->accel.multifunc[0x0a] & 6, dev->accel.frgd_mix & 0x1f, dev->accel.bkgd_mix & 0x1f, bkgd_color, frgd_color, dev->accel.clip_top, clip_b); if (ibm8514_cpu_src(svga)) { if (dev->accel.cmd & 2) { if (dev->accel.cmd & 8) { @@ -1586,7 +1635,7 @@ ibm8514_accel_start(int count, int cpu_input, uint32_t mix_dat, uint32_t cpu_dat READ((dev->accel.cy * dev->pitch) + dev->accel.cx, src_dat); if (pixcntl == 3) src_dat = ((src_dat & rd_mask) == rd_mask); - } else + } else { switch ((mix_dat & mix_mask) ? frgd_mix : bkgd_mix) { case 0: src_dat = bkgd_color; @@ -1604,6 +1653,7 @@ ibm8514_accel_start(int count, int cpu_input, uint32_t mix_dat, uint32_t cpu_dat default: break; } + } READ((dev->accel.cy * dev->pitch) + dev->accel.cx, dest_dat); @@ -1785,70 +1835,34 @@ ibm8514_accel_start(int count, int cpu_input, uint32_t mix_dat, uint32_t cpu_dat break; } - if (dev->accel.err_term >= dev->accel.maj_axis_pcnt) { - dev->accel.err_term += dev->accel.destx_distp; - /*Step minor axis*/ - switch (dev->accel.cmd & 0xe0) { - case 0x00: - dev->accel.cy--; - break; - case 0x20: - dev->accel.cy--; - break; - case 0x40: - dev->accel.cx--; - break; - case 0x60: - dev->accel.cx++; - break; - case 0x80: - dev->accel.cy++; - break; - case 0xa0: - dev->accel.cy++; - break; - case 0xc0: - dev->accel.cx--; - break; - case 0xe0: - dev->accel.cx++; - break; - - default: - break; - } - } else - dev->accel.err_term += dev->accel.desty_axstp; - - /*Step major axis*/ - switch (dev->accel.cmd & 0xe0) { - case 0x00: - dev->accel.cx--; - break; - case 0x20: - dev->accel.cx++; - break; - case 0x40: - dev->accel.cy--; - break; - case 0x60: - dev->accel.cy--; - break; - case 0x80: - dev->accel.cx--; - break; - case 0xa0: - dev->accel.cx++; - break; - case 0xc0: + if (dev->accel.cmd & 0x40) { + if (dev->accel.cmd & 0x80) dev->accel.cy++; - break; - case 0xe0: - dev->accel.cy++; - break; + else + dev->accel.cy--; - default: - break; + if (dev->accel.err_term >= 0) { + dev->accel.err_term += dev->accel.destx_distp; + if (dev->accel.cmd & 0x20) + dev->accel.cx++; + else + dev->accel.cx--; + } else + dev->accel.err_term += dev->accel.desty_axstp; + } else { + if (dev->accel.cmd & 0x20) + dev->accel.cx++; + else + dev->accel.cx--; + + if (dev->accel.err_term >= 0) { + dev->accel.err_term += dev->accel.destx_distp; + if (dev->accel.cmd & 0x80) + dev->accel.cy++; + else + dev->accel.cy--; + } else + dev->accel.err_term += dev->accel.desty_axstp; } dev->accel.sy--; @@ -1909,74 +1923,37 @@ ibm8514_accel_start(int count, int cpu_input, uint32_t mix_dat, uint32_t cpu_dat else cpu_dat >>= 8; - if (dev->accel.sy == 0) { + if (dev->accel.sy == 0) break; - } - if (dev->accel.err_term >= dev->accel.maj_axis_pcnt) { - dev->accel.err_term += dev->accel.destx_distp; - /*Step minor axis*/ - switch (dev->accel.cmd & 0xe0) { - case 0x00: - dev->accel.cy--; - break; - case 0x20: - dev->accel.cy--; - break; - case 0x40: - dev->accel.cx--; - break; - case 0x60: - dev->accel.cx++; - break; - case 0x80: - dev->accel.cy++; - break; - case 0xa0: - dev->accel.cy++; - break; - case 0xc0: - dev->accel.cx--; - break; - case 0xe0: - dev->accel.cx++; - break; - - default: - break; - } - } else - dev->accel.err_term += dev->accel.desty_axstp; - - /*Step major axis*/ - switch (dev->accel.cmd & 0xe0) { - case 0x00: - dev->accel.cx--; - break; - case 0x20: - dev->accel.cx++; - break; - case 0x40: - dev->accel.cy--; - break; - case 0x60: - dev->accel.cy--; - break; - case 0x80: - dev->accel.cx--; - break; - case 0xa0: - dev->accel.cx++; - break; - case 0xc0: + if (dev->accel.cmd & 0x40) { + if (dev->accel.cmd & 0x80) dev->accel.cy++; - break; - case 0xe0: - dev->accel.cy++; - break; + else + dev->accel.cy--; - default: - break; + if (dev->accel.err_term >= 0) { + dev->accel.err_term += dev->accel.destx_distp; + if (dev->accel.cmd & 0x20) + dev->accel.cx++; + else + dev->accel.cx--; + } else + dev->accel.err_term += dev->accel.desty_axstp; + } else { + if (dev->accel.cmd & 0x20) + dev->accel.cx++; + else + dev->accel.cx--; + + if (dev->accel.err_term >= 0) { + dev->accel.err_term += dev->accel.destx_distp; + if (dev->accel.cmd & 0x80) + dev->accel.cy++; + else + dev->accel.cy--; + } else + dev->accel.err_term += dev->accel.desty_axstp; } dev->accel.sy--; @@ -2010,13 +1987,11 @@ ibm8514_accel_start(int count, int cpu_input, uint32_t mix_dat, uint32_t cpu_dat if (dev->accel.cur_y >= 0x600) dev->accel.cy |= ~0x5ff; - if (((dev->local & 0xff) >= 0x02) && dev->accel.ge_offset && (dev->accel_bpp == 24)) + if (((dev->local & 0xff) >= 0x02) && dev->accel.ge_offset && ((dev->accel_bpp == 24) || (dev->accel_bpp == 8))) dev->accel.dest = (dev->accel.ge_offset << 2) + (dev->accel.cy * dev->pitch); else dev->accel.dest = dev->accel.cy * dev->pitch; - dev->accel.fill_state = 0; - if (cmd == 4) dev->accel.cmd |= 2; else if (cmd == 3) @@ -2062,7 +2037,7 @@ ibm8514_accel_start(int count, int cpu_input, uint32_t mix_dat, uint32_t cpu_dat if (!(dev->accel.cmd & 0x40) && (frgd_mix == 2) && (bkgd_mix == 2) && (pixcntl == 0) && (cmd == 2)) { if (!(dev->accel.sx & 1)) { dev->accel.output = 1; - if (((dev->local & 0xff) >= 0x02) && dev->accel.ge_offset && (dev->accel_bpp == 24)) + if (((dev->local & 0xff) >= 0x02) && dev->accel.ge_offset && ((dev->accel_bpp == 24) || (dev->accel_bpp == 8))) dev->accel.newdest_out = (dev->accel.ge_offset << 2) + ((dev->accel.cy + 1) * dev->pitch); else dev->accel.newdest_out = (dev->accel.cy + 1) * dev->pitch; @@ -2076,7 +2051,7 @@ ibm8514_accel_start(int count, int cpu_input, uint32_t mix_dat, uint32_t cpu_dat if (!(dev->accel.cmd & 2) && (frgd_mix == 2) && (pixcntl == 0) && (cmd == 2)) { if (!(dev->accel.sx & 1)) { dev->accel.input = 1; - if (((dev->local & 0xff) >= 0x02) && dev->accel.ge_offset && (dev->accel_bpp == 24)) + if (((dev->local & 0xff) >= 0x02) && dev->accel.ge_offset && ((dev->accel_bpp == 24) || (dev->accel_bpp == 8))) dev->accel.newdest_in = (dev->accel.ge_offset << 2) + ((dev->accel.cy + 1) * dev->pitch); else dev->accel.newdest_in = (dev->accel.cy + 1) * dev->pitch; @@ -2220,15 +2195,14 @@ rect_fill_pix: dev->accel.sx--; if (dev->accel.sx < 0) { dev->accel.sx = dev->accel.maj_axis_pcnt & 0x7ff; - if (and3 == 1) { + if (and3 == 1) dev->accel.sx += 4; - } else if (and3 == 2) { + else if (and3 == 2) dev->accel.sx += 5; - } else if (and3 == 3) { + else if (and3 == 3) dev->accel.sx += 6; - } else { + else dev->accel.sx += 3; - } if (dev->accel.cmd & 0x20) dev->accel.cx -= (dev->accel.sx + 1); @@ -2259,10 +2233,11 @@ rect_fill_pix: break; } - if (((dev->local & 0xff) >= 0x02) && dev->accel.ge_offset && (dev->accel_bpp == 24)) + if (((dev->local & 0xff) >= 0x02) && dev->accel.ge_offset && ((dev->accel_bpp == 24) || (dev->accel_bpp == 8))) dev->accel.dest = (dev->accel.ge_offset << 2) + (dev->accel.cy * dev->pitch); else dev->accel.dest = dev->accel.cy * dev->pitch; + dev->accel.sy--; return; } @@ -2344,7 +2319,7 @@ rect_fill_pix: else dev->accel.cy--; - if (((dev->local & 0xff) >= 0x02) && dev->accel.ge_offset && (dev->accel_bpp == 24)) + if (((dev->local & 0xff) >= 0x02) && dev->accel.ge_offset && ((dev->accel_bpp == 24) || (dev->accel_bpp == 8))) dev->accel.dest = (dev->accel.ge_offset << 2) + (dev->accel.cy * dev->pitch); else dev->accel.dest = dev->accel.cy * dev->pitch; @@ -2436,7 +2411,7 @@ rect_fill_pix: else dev->accel.cy--; - if (((dev->local & 0xff) >= 0x02) && dev->accel.ge_offset && (dev->accel_bpp == 24)) + if (((dev->local & 0xff) >= 0x02) && dev->accel.ge_offset && ((dev->accel_bpp == 24) || (dev->accel_bpp == 8))) dev->accel.dest = (dev->accel.ge_offset << 2) + (dev->accel.cy * dev->pitch); else dev->accel.dest = dev->accel.cy * dev->pitch; @@ -2495,7 +2470,7 @@ rect_fill_pix: else dev->accel.cy--; - if (((dev->local & 0xff) >= 0x02) && dev->accel.ge_offset && (dev->accel_bpp == 24)) { + if (((dev->local & 0xff) >= 0x02) && dev->accel.ge_offset && ((dev->accel_bpp == 24) || (dev->accel_bpp == 8))) { dev->accel.dest = (dev->accel.ge_offset << 2) + (dev->accel.cy * dev->pitch); dev->accel.newdest_in = (dev->accel.ge_offset << 2) + ((dev->accel.cy + 1) * dev->pitch); } else { @@ -2520,7 +2495,7 @@ rect_fill_pix: else dev->accel.cy--; - if (((dev->local & 0xff) >= 0x02) && dev->accel.ge_offset && (dev->accel_bpp == 24)) { + if (((dev->local & 0xff) >= 0x02) && dev->accel.ge_offset && ((dev->accel_bpp == 24) || (dev->accel_bpp == 8))) { dev->accel.dest = (dev->accel.ge_offset << 2) + (dev->accel.cy * dev->pitch); dev->accel.newdest_in = (dev->accel.ge_offset << 2) + ((dev->accel.cy + 1) * dev->pitch); } else { @@ -2576,7 +2551,7 @@ rect_fill_pix: else dev->accel.cy--; - if (((dev->local & 0xff) >= 0x02) && dev->accel.ge_offset && (dev->accel_bpp == 24)) { + if (((dev->local & 0xff) >= 0x02) && dev->accel.ge_offset && ((dev->accel_bpp == 24) || (dev->accel_bpp == 8))) { dev->accel.dest = (dev->accel.ge_offset << 2) + (dev->accel.cy * dev->pitch); dev->accel.newdest_out = (dev->accel.ge_offset << 2) + ((dev->accel.cy + 1) * dev->pitch); } else { @@ -2601,7 +2576,7 @@ rect_fill_pix: else dev->accel.cy--; - if (((dev->local & 0xff) >= 0x02) && dev->accel.ge_offset && (dev->accel_bpp == 24)) { + if (((dev->local & 0xff) >= 0x02) && dev->accel.ge_offset && ((dev->accel_bpp == 24) || (dev->accel_bpp == 8))) { dev->accel.dest = (dev->accel.ge_offset << 2) + (dev->accel.cy * dev->pitch); dev->accel.newdest_out = (dev->accel.ge_offset << 2) + ((dev->accel.cy + 1) * dev->pitch); } else { @@ -2691,7 +2666,7 @@ rect_fill_pix: else dev->accel.cy--; - if (((dev->local & 0xff) >= 0x02) && dev->accel.ge_offset && (dev->accel_bpp == 24)) + if (((dev->local & 0xff) >= 0x02) && dev->accel.ge_offset && ((dev->accel_bpp == 24) || (dev->accel_bpp == 8))) dev->accel.dest = (dev->accel.ge_offset << 2) + (dev->accel.cy * dev->pitch); else dev->accel.dest = dev->accel.cy * dev->pitch; @@ -2768,7 +2743,7 @@ rect_fill: else dev->accel.cy--; - if (((dev->local & 0xff) >= 0x02) && dev->accel.ge_offset && (dev->accel_bpp == 24)) + if (((dev->local & 0xff) >= 0x02) && dev->accel.ge_offset && ((dev->accel_bpp == 24) || (dev->accel_bpp == 8))) dev->accel.dest = (dev->accel.ge_offset << 2) + (dev->accel.cy * dev->pitch); else dev->accel.dest = dev->accel.cy * dev->pitch; @@ -2783,7 +2758,7 @@ rect_fill: } else { dev->accel.temp_cnt = 8; while (count-- && dev->accel.sy >= 0) { - if (dev->accel.temp_cnt == 0) { + if (!dev->accel.temp_cnt) { dev->accel.temp_cnt = 8; mix_dat = old_mix_dat; } @@ -2839,7 +2814,7 @@ rect_fill: else dev->accel.cy--; - if (((dev->local & 0xff) >= 0x02) && dev->accel.ge_offset && (dev->accel_bpp == 24)) + if (((dev->local & 0xff) >= 0x02) && dev->accel.ge_offset && ((dev->accel_bpp == 24) || (dev->accel_bpp == 8))) dev->accel.dest = (dev->accel.ge_offset << 2) + (dev->accel.cy * dev->pitch); else dev->accel.dest = dev->accel.cy * dev->pitch; @@ -2855,93 +2830,74 @@ rect_fill: } } } else { - if (dev->accel.multifunc[0x0a] & 6) { - while (count-- && dev->accel.sy >= 0) { + if ((dev->accel.multifunc[0x0a] & 6) == 4) { + while (count-- && (dev->accel.sy >= 0)) { if (dev->accel.cx >= dev->accel.clip_left && dev->accel.cx <= clip_r && dev->accel.cy >= dev->accel.clip_top && dev->accel.cy <= clip_b) { - switch ((mix_dat & mix_mask) ? frgd_mix : bkgd_mix) { - case 0: - src_dat = bkgd_color; - break; - case 1: - src_dat = frgd_color; - break; - case 2: - src_dat = 0; - break; - case 3: - src_dat = 0; - break; - - default: - break; - } - - READ(dev->accel.dest + dev->accel.cx, poly_src); - if (dev->accel.multifunc[0x0a] & 2) { - poly_src = ((poly_src & wrt_mask) == wrt_mask); - } else { - poly_src = ((poly_src & rd_mask_polygon) == rd_mask_polygon); - } - - if (poly_src) { - dev->accel.fill_state ^= 1; - } + READ(dev->accel.dest + dev->accel.cx, mix_dat); + if ((mix_dat & rd_mask_polygon) == rd_mask_polygon) + dev->accel.fill_state = !dev->accel.fill_state; + READ(dev->accel.dest + dev->accel.cx, dest_dat); + old_dest_dat = dest_dat; if (dev->accel.fill_state) { - READ(dev->accel.dest + dev->accel.cx, dest_dat); - - if ((compare_mode == 0) || ((compare_mode == 0x10) && (dest_dat >= compare)) || ((compare_mode == 0x18) && (dest_dat < compare)) || ((compare_mode == 0x20) && (dest_dat != compare)) || ((compare_mode == 0x28) && (dest_dat == compare)) || ((compare_mode == 0x30) && (dest_dat <= compare)) || ((compare_mode == 0x38) && (dest_dat > compare))) { - old_dest_dat = dest_dat; - MIX(mix_dat & mix_mask, dest_dat, src_dat); - dest_dat = (dest_dat & wrt_mask) | (old_dest_dat & ~wrt_mask); - WRITE(dev->accel.dest + dev->accel.cx, dest_dat); + if (!(rd_mask_polygon & 1) && (wrt_mask & 1)) { + MIX(mix_dat ^ rd_mask_polygon, dest_dat, mix_dat); + ibm8514_log("Filling c(%d,%d) without bit 0 of rdmask=%02x, wrtmask=%02x, mixdat=%02x, dest=%02x, old=%02x.\n", dev->accel.cx, dev->accel.cy, rd_mask_polygon, wrt_mask, mix_dat, dest_dat, old_dest_dat); + dest_dat &= ~rd_mask_polygon; + } else if ((rd_mask_polygon & 1) && (wrt_mask & 1)) { + ibm8514_log("Filling c(%d,%d) with bit 0 of rdmask=%02x, wrtmask=%02x.\n", dev->accel.cx, dev->accel.cy, rd_mask_polygon, wrt_mask); + dest_dat &= ~(rd_mask_polygon & wrt_mask); } + } else { + if (!(rd_mask_polygon & 1) && (wrt_mask & 1)) + dest_dat &= ~rd_mask_polygon; + else if ((rd_mask_polygon & 1) && (wrt_mask & 1)) + dest_dat &= ~(rd_mask_polygon & wrt_mask); + } + + dest_dat = (dest_dat & wrt_mask) | (old_dest_dat & ~wrt_mask); + + if ((compare_mode == 0) || ((compare_mode == 0x10) && (dest_dat >= compare)) || ((compare_mode == 0x18) && (dest_dat < compare)) || ((compare_mode == 0x20) && (dest_dat != compare)) || ((compare_mode == 0x28) && (dest_dat == compare)) || ((compare_mode == 0x30) && (dest_dat <= compare)) || ((compare_mode == 0x38) && (dest_dat > compare))) { + ibm8514_log("Results c(%d,%d):rdmask=%02x, wrtmask=%02x, mix=%02x, destdat=%02x, nowrite=%d.\n", dev->accel.cx, dev->accel.cy, rd_mask_polygon, wrt_mask, mix_dat, dest_dat, dev->accel.cx_back); + WRITE(dev->accel.dest + dev->accel.cx, dest_dat); } } - mix_dat <<= 1; - mix_dat |= 1; - - if (dev->accel.cmd & 0x20) { + if (dev->accel.cmd & 0x20) dev->accel.cx++; - } else { + else dev->accel.cx--; - } dev->accel.sx--; if (dev->accel.sx < 0) { - dev->accel.sx = dev->accel.maj_axis_pcnt & 0x7ff; dev->accel.fill_state = 0; + dev->accel.sx = dev->accel.maj_axis_pcnt & 0x7ff; - if (dev->accel.cmd & 0x20) { + if (dev->accel.cmd & 0x20) dev->accel.cx -= (dev->accel.sx) + 1; - } else { + else dev->accel.cx += (dev->accel.sx) + 1; - } if (dev->accel.cmd & 0x80) dev->accel.cy++; else dev->accel.cy--; - if (((dev->local & 0xff) >= 0x02) && dev->accel.ge_offset && (dev->accel_bpp == 24)) - dev->accel.dest = (dev->accel.ge_offset << 2) + (dev->accel.cy * dev->pitch); - else - dev->accel.dest = dev->accel.cy * dev->pitch; + dev->accel.dest = dev->accel.cy * dev->pitch; dev->accel.sy--; if (dev->accel.sy < 0) { - dev->accel.cur_x = dev->accel.cx; - dev->accel.cur_y = dev->accel.cy; + ibm8514_log(".\n"); return; } } } } else { + ibm8514_log("Rectangle Fill Normal CMD=%04x, CURRENT(%d,%d), sx=%d, FR(%02x), linedraw=%d.\n", dev->accel.cmd, dev->accel.cx, dev->accel.cy, dev->accel.sx, frgd_color, dev->accel.linedraw); while (count-- && dev->accel.sy >= 0) { if (dev->accel.cx >= dev->accel.clip_left && dev->accel.cx <= clip_r && dev->accel.cy >= dev->accel.clip_top && dev->accel.cy <= clip_b) { - switch (frgd_mix) { + switch ((mix_dat & mix_mask) ? frgd_mix : bkgd_mix) { case 0: src_dat = bkgd_color; if (!bkgd_mix && (dev->accel.cmd & 0x40) && ((dev->accel.frgd_mix & 0x1f) == 7) && ((dev->accel.bkgd_mix & 0x1f) == 3) && !dev->bpp && (bkgd_color == 0x00)) /*For some reason, the September 1992 Mach8/32 drivers for Win3.x don't set the background colors properly.*/ @@ -2961,13 +2917,17 @@ rect_fill: break; } + READ(dev->accel.dest + dev->accel.cx, dest_dat); if ((compare_mode == 0) || ((compare_mode == 0x10) && (dest_dat >= compare)) || ((compare_mode == 0x18) && (dest_dat < compare)) || ((compare_mode == 0x20) && (dest_dat != compare)) || ((compare_mode == 0x28) && (dest_dat == compare)) || ((compare_mode == 0x30) && (dest_dat <= compare)) || ((compare_mode == 0x38) && (dest_dat > compare))) { old_dest_dat = dest_dat; - MIX(1, dest_dat, src_dat); + MIX(mix_dat & mix_mask, dest_dat, src_dat); dest_dat = (dest_dat & wrt_mask) | (old_dest_dat & ~wrt_mask); - WRITE(dev->accel.dest + dev->accel.cx, dest_dat); + + if (dev->accel.cmd & 0x10) { + WRITE(dev->accel.dest + dev->accel.cx, dest_dat); + } } } @@ -2981,11 +2941,12 @@ rect_fill: dev->accel.sx--; if (dev->accel.sx < 0) { + dev->accel.fill_state = 0; dev->accel.sx = dev->accel.maj_axis_pcnt & 0x7ff; - if (dev->accel.cmd & 0x20) { + if (dev->accel.cmd & 0x20) dev->accel.cx -= (dev->accel.sx) + 1; - } else + else dev->accel.cx += (dev->accel.sx) + 1; if (dev->accel.cmd & 0x80) @@ -2993,7 +2954,7 @@ rect_fill: else dev->accel.cy--; - if (((dev->local & 0xff) >= 0x02) && dev->accel.ge_offset && (dev->accel_bpp == 24)) + if (((dev->local & 0xff) >= 0x02) && dev->accel.ge_offset && ((dev->accel_bpp == 24) || (dev->accel_bpp == 8))) dev->accel.dest = (dev->accel.ge_offset << 2) + (dev->accel.cy * dev->pitch); else dev->accel.dest = dev->accel.cy * dev->pitch; @@ -3013,17 +2974,25 @@ rect_fill: } break; - case 5: /*Draw Polygon Boundary Line*/ + case 5: /*Draw Polygon Boundary Line*/ { if (!cpu_input) { dev->accel.cx = dev->accel.cur_x; - dev->accel.cy = dev->accel.cur_y; if (dev->accel.cur_x >= 0x600) dev->accel.cx |= ~0x5ff; - + dev->accel.cy = dev->accel.cur_y; if (dev->accel.cur_y >= 0x600) dev->accel.cy |= ~0x5ff; - dev->accel.oldcy = dev->accel.cy; - dev->accel.sy = 0; + + dev->accel.sy = dev->accel.maj_axis_pcnt_no_limit; + + if (dev->accel.cmd & 0x80) + dev->accel.oldcy = dev->accel.cy + 1; + else + dev->accel.oldcy = dev->accel.cy - 1; + + dev->accel.oldcx = 0; + + ibm8514_log("Polygon Boundary activated=%04x, len=%d, cur(%d,%d), frgdmix=%02x, err=%d, clipping: l=%d, r=%d, t=%d, b=%d, pixcntl=%02x.\n", dev->accel.cmd, dev->accel.sy, dev->accel.cur_x_nolimit, dev->accel.cy, dev->accel.frgd_mix & 0x1f, dev->accel.err_term, dev->accel.clip_left, clip_r, dev->accel.clip_top, clip_b, compare_mode, dev->accel.multifunc[0x0a]); if (ibm8514_cpu_src(svga)) { dev->data_available = 0; @@ -3036,160 +3005,200 @@ rect_fill: } } - while (count-- && (dev->accel.sy >= 0)) { - if ((dev->accel.cx) >= dev->accel.clip_left && ((dev->accel.cx) <= clip_r) && (dev->accel.cy) >= dev->accel.clip_top && (dev->accel.cy) <= clip_b) { - switch ((mix_dat & mix_mask) ? frgd_mix : bkgd_mix) { - case 0: - src_dat = bkgd_color; + if (dev->accel.cmd & 8) { + while (count-- && (dev->accel.sy >= 0)) { + if (dev->accel.cx < 0) + dev->accel.cx = 0; + if (dev->accel.cy < 0) + dev->accel.cy = 0; + + if (dev->accel.cx >= dev->accel.clip_left && dev->accel.cx <= clip_r && dev->accel.cy >= dev->accel.clip_top && dev->accel.cy <= clip_b) { + switch ((mix_dat & mix_mask) ? frgd_mix : bkgd_mix) { + case 0: + src_dat = bkgd_color; + break; + case 1: + src_dat = frgd_color; + break; + case 2: + src_dat = cpu_dat; + break; + case 3: + src_dat = 0; + break; + + default: + break; + } + + + READ((dev->accel.cy * dev->pitch) + dev->accel.cx, dest_dat); + + if ((compare_mode == 0) || ((compare_mode == 0x10) && (dest_dat >= compare)) || ((compare_mode == 0x18) && (dest_dat < compare)) || ((compare_mode == 0x20) && (dest_dat != compare)) || ((compare_mode == 0x28) && (dest_dat == compare)) || ((compare_mode == 0x30) && (dest_dat <= compare)) || ((compare_mode == 0x38) && (dest_dat > compare))) { + old_dest_dat = dest_dat; + MIX(mix_dat & mix_mask, dest_dat, src_dat); + dest_dat = (dest_dat & wrt_mask) | (old_dest_dat & ~wrt_mask); + if (dev->accel.cmd & 0x10) { + if (dev->accel.sy && (dev->accel.cmd & 4)) { + if (dev->accel.oldcy != dev->accel.cy) { + WRITE((dev->accel.cy * dev->pitch) + (dev->accel.cx), dest_dat); + } + } else if (!(dev->accel.cmd & 4)) { + if (dev->accel.oldcy != dev->accel.cy) { + WRITE((dev->accel.cy * dev->pitch) + (dev->accel.cx), dest_dat); + } + } + } + } + } + + mix_dat <<= 1; + mix_dat |= 1; + if (dev->bpp) + cpu_dat >>= 16; + else + cpu_dat >>= 8; + + if (!dev->accel.sy) + break; + + switch (dev->accel.cmd & 0xe0) { + case 0x00: + dev->accel.cx++; break; - case 1: - src_dat = frgd_color; + case 0x20: + dev->accel.cx++; + dev->accel.oldcy = dev->accel.cy; + dev->accel.cy--; break; - case 2: - src_dat = cpu_dat; + case 0x40: + dev->accel.oldcy = dev->accel.cy; + dev->accel.cy--; break; - case 3: - src_dat = 0; + case 0x60: + dev->accel.cx--; + dev->accel.oldcy = dev->accel.cy; + dev->accel.cy--; + break; + case 0x80: + dev->accel.cx--; + break; + case 0xa0: + dev->accel.cx--; + dev->accel.oldcy = dev->accel.cy; + dev->accel.cy++; + break; + case 0xc0: + dev->accel.oldcy = dev->accel.cy; + dev->accel.cy++; + break; + case 0xe0: + dev->accel.cx++; + dev->accel.oldcy = dev->accel.cy; + dev->accel.cy++; break; default: break; } - READ((dev->accel.cy * dev->pitch) + dev->accel.cx, dest_dat); + dev->accel.sy--; + } + } else { + while (count-- && (dev->accel.sy >= 0)) { + if (dev->accel.cx < 0) + dev->accel.cx = 0; + if (dev->accel.cy < 0) + dev->accel.cy = 0; - if ((compare_mode == 0) || ((compare_mode == 0x10) && (dest_dat >= compare)) || ((compare_mode == 0x18) && (dest_dat < compare)) || ((compare_mode == 0x20) && (dest_dat != compare)) || ((compare_mode == 0x28) && (dest_dat == compare)) || ((compare_mode == 0x30) && (dest_dat <= compare)) || ((compare_mode == 0x38) && (dest_dat > compare))) { - old_dest_dat = dest_dat; - MIX(mix_dat & mix_mask, dest_dat, src_dat); - dest_dat = (dest_dat & wrt_mask) | (old_dest_dat & ~wrt_mask); - if (dev->accel.cmd & 4) { - if (dev->accel.sy < dev->accel.maj_axis_pcnt) { - if (dev->accel.cmd & 0x40) { - WRITE((dev->accel.cy * dev->pitch) + (dev->accel.cx), dest_dat); - } else { - if (dev->accel.cy == (dev->accel.oldcy + 1)) { - if (dev->accel.cmd & 0x20) { - if (dev->accel.err_term < (dev->accel.destx_distp + dev->accel.desty_axstp)) { - WRITE((dev->accel.cy * dev->pitch) + (dev->accel.cx), dest_dat); - } - } else { - if (dev->accel.err_term >= 0) { - WRITE((dev->accel.cy * dev->pitch) + (dev->accel.cx), dest_dat); - } - } - } - } - } - } else { - if (dev->accel.cmd & 0x40) { - WRITE((dev->accel.cy * dev->pitch) + (dev->accel.cx), dest_dat); - } else { - if (dev->accel.cy == (dev->accel.oldcy + 1)) { - if (dev->accel.cmd & 0x20) { - if (dev->accel.err_term < (dev->accel.destx_distp + dev->accel.desty_axstp)) { - WRITE((dev->accel.cy * dev->pitch) + (dev->accel.cx), dest_dat); - } + if (dev->accel.cx >= dev->accel.clip_left && dev->accel.cx <= clip_r && dev->accel.cy >= dev->accel.clip_top && dev->accel.cy <= clip_b) { + switch ((mix_dat & mix_mask) ? frgd_mix : bkgd_mix) { + case 0: + src_dat = bkgd_color; + break; + case 1: + src_dat = frgd_color; + break; + case 2: + src_dat = cpu_dat; + break; + case 3: + src_dat = 0; + break; + + default: + break; + } + + READ((dev->accel.cy * dev->pitch) + dev->accel.cx, dest_dat); + + if ((compare_mode == 0) || ((compare_mode == 0x10) && (dest_dat >= compare)) || ((compare_mode == 0x18) && (dest_dat < compare)) || ((compare_mode == 0x20) && (dest_dat != compare)) || ((compare_mode == 0x28) && (dest_dat == compare)) || ((compare_mode == 0x30) && (dest_dat <= compare)) || ((compare_mode == 0x38) && (dest_dat > compare))) { + old_dest_dat = dest_dat; + MIX(mix_dat & mix_mask, dest_dat, src_dat); + dest_dat = (dest_dat & wrt_mask) | (old_dest_dat & ~wrt_mask); + + if ((dev->accel.cmd & 0x14) == 0x14) { + if (dev->accel.sy) { + if (dev->accel.cmd & 0x40) { + WRITE((dev->accel.cy * dev->pitch) + dev->accel.cx, dest_dat); } else { - if (dev->accel.err_term >= 0) { - WRITE((dev->accel.cy * dev->pitch) + (dev->accel.cx), dest_dat); + if (dev->accel.oldcy != dev->accel.cy) { + WRITE((dev->accel.cy * dev->pitch) + dev->accel.cx, dest_dat); } } } } } } - } - mix_dat <<= 1; - mix_dat |= 1; - if (dev->bpp) - cpu_dat >>= 16; - else - cpu_dat >>= 8; + mix_dat <<= 1; + mix_dat |= 1; + if (dev->bpp) + cpu_dat >>= 16; + else + cpu_dat >>= 8; - if (dev->accel.sy == dev->accel.maj_axis_pcnt) { - break; - } - - /*Step major axis*/ - switch (dev->accel.cmd & 0xe0) { - case 0x00: - dev->accel.cx--; - break; - case 0x20: - dev->accel.cx++; - break; - case 0x40: - dev->accel.oldcy = dev->accel.cy; - dev->accel.cy--; - break; - case 0x60: - dev->accel.oldcy = dev->accel.cy; - dev->accel.cy--; - break; - case 0x80: - dev->accel.cx--; - break; - case 0xa0: - dev->accel.cx++; - break; - case 0xc0: - dev->accel.oldcy = dev->accel.cy; - dev->accel.cy++; - break; - case 0xe0: - dev->accel.oldcy = dev->accel.cy; - dev->accel.cy++; + if (!dev->accel.sy) break; - default: - break; - } - - if (dev->accel.err_term >= 0) { - dev->accel.err_term += dev->accel.destx_distp; - /*Step minor axis*/ - switch (dev->accel.cmd & 0xe0) { - case 0x00: - dev->accel.oldcy = dev->accel.cy; - dev->accel.cy--; - break; - case 0x20: - dev->accel.oldcy = dev->accel.cy; - dev->accel.cy--; - break; - case 0x40: - dev->accel.cx--; - break; - case 0x60: - dev->accel.cx++; - break; - case 0x80: - dev->accel.oldcy = dev->accel.cy; + if (dev->accel.cmd & 0x40) { + if (dev->accel.cmd & 0x80) dev->accel.cy++; - break; - case 0xa0: - dev->accel.oldcy = dev->accel.cy; - dev->accel.cy++; - break; - case 0xc0: - dev->accel.cx--; - break; - case 0xe0: - dev->accel.cx++; - break; + else + dev->accel.cy--; - default: - break; + if (dev->accel.err_term >= 0) { + dev->accel.err_term += dev->accel.destx_distp; + if (dev->accel.cmd & 0x20) + dev->accel.cx++; + else + dev->accel.cx--; + } else + dev->accel.err_term += dev->accel.desty_axstp; + } else { + if (dev->accel.cmd & 0x20) + dev->accel.cx++; + else + dev->accel.cx--; + + dev->accel.oldcy = dev->accel.cy; + if (dev->accel.err_term >= 0) { + dev->accel.err_term += dev->accel.destx_distp; + if (dev->accel.cmd & 0x80) + dev->accel.cy++; + else + dev->accel.cy--; + } else + dev->accel.err_term += dev->accel.desty_axstp; } - } else - dev->accel.err_term += dev->accel.desty_axstp; - dev->accel.sy++; + dev->accel.sy--; + } } - break; + } + break; - case 6: /*BitBlt*/ + case 6: /*BitBlt*/ if (!cpu_input) /*!cpu_input is trigger to start operation*/ { dev->accel.x_count = 0; @@ -3198,12 +3207,12 @@ rect_fill: dev->accel.sx = dev->accel.maj_axis_pcnt & 0x7ff; dev->accel.sy = dev->accel.multifunc[0] & 0x7ff; - dev->accel.dx = dev->accel.destx_distp; - dev->accel.dy = dev->accel.desty_axstp; + dev->accel.dx = dev->accel.destx; + dev->accel.dy = dev->accel.desty; - if (dev->accel.destx_distp >= 0x600) + if (dev->accel.destx >= 0x600) dev->accel.dx |= ~0x5ff; - if (dev->accel.desty_axstp >= 0x600) + if (dev->accel.desty >= 0x600) dev->accel.dy |= ~0x5ff; dev->accel.cx = dev->accel.cur_x; @@ -3216,6 +3225,7 @@ rect_fill: dev->accel.src = dev->accel.cy * dev->pitch; dev->accel.dest = dev->accel.dy * dev->pitch; + dev->accel.fill_state = 0; if (ibm8514_cpu_src(svga)) { if (dev->accel.cmd & 2) { @@ -3280,6 +3290,7 @@ bitblt_pix: if ((compare_mode == 0) || ((compare_mode == 0x10) && (dest_dat >= compare)) || ((compare_mode == 0x18) && (dest_dat < compare)) || ((compare_mode == 0x20) && (dest_dat != compare)) || ((compare_mode == 0x28) && (dest_dat == compare)) || ((compare_mode == 0x30) && (dest_dat <= compare)) || ((compare_mode == 0x38) && (dest_dat > compare))) { old_dest_dat = dest_dat; + MIX(mix_dat & mix_mask, dest_dat, src_dat); dest_dat = (dest_dat & wrt_mask) | (old_dest_dat & ~wrt_mask); WRITE(dev->accel.dest + dev->accel.dx, dest_dat); @@ -3413,8 +3424,8 @@ bitblt_pix: dev->accel.cx = dev->accel.cur_x; if (dev->accel.cur_x >= 0x600) dev->accel.cx |= ~0x5ff; - dev->accel.dx = dev->accel.destx_distp; - if (dev->accel.destx_distp >= 0x600) + dev->accel.dx = dev->accel.destx; + if (dev->accel.destx >= 0x600) dev->accel.dx |= ~0x5ff; } } @@ -3444,6 +3455,7 @@ bitblt_pix: } else { while (count-- && (dev->accel.sy >= 0)) { if (dev->accel.dx >= dev->accel.clip_left && dev->accel.dx <= clip_r && dev->accel.dy >= dev->accel.clip_top && dev->accel.dy <= clip_b) { + if (pixcntl == 3) { if (!(dev->accel.cmd & 0x10) && ((frgd_mix != 3) || (bkgd_mix != 3))) { READ(dev->accel.src + dev->accel.cx, mix_dat); @@ -3698,6 +3710,7 @@ bitblt: while (1) { if ((dx >= (((int64_t)dev->accel.clip_left) * 3)) && (dx <= (((uint64_t)clip_r) * 3)) && (dev->accel.dy >= (dev->accel.clip_top << 1)) && (dev->accel.dy <= (clip_b << 1))) { + READ(dev->accel.src + (dev->accel.ge_offset << 2) + cx, src_dat); READ(dev->accel.dest + (dev->accel.ge_offset << 2) + dx, dest_dat); @@ -3716,6 +3729,7 @@ bitblt: return; } + ibm8514_log("BitBLT 8514/A=%04x, selfrmix=%d, selbkmix=%d, d(%d,%d), c(%d,%d), pixcntl=%d, sy=%d, frgdmix=%02x, bkgdmix=%02x, rdmask=%02x, wrtmask=%02x, linedraw=%d.\n", dev->accel.cmd, frgd_mix, bkgd_mix, dev->accel.dx, dev->accel.dy, dev->accel.cx, dev->accel.cy, pixcntl, dev->accel.sy, dev->accel.frgd_mix & 0x1f, dev->accel.bkgd_mix & 0x1f, dev->accel.rd_mask, wrt_mask, dev->accel.linedraw); while (count-- && dev->accel.sy >= 0) { if ((dev->accel.dx >= dev->accel.clip_left) && (dev->accel.dx <= clip_r) && (dev->accel.dy >= dev->accel.clip_top) && (dev->accel.dy <= clip_b)) { @@ -3781,15 +3795,18 @@ bitblt: dev->accel.sx--; if (dev->accel.sx < 0) { + dev->accel.fill_state = 0; dev->accel.sx = dev->accel.maj_axis_pcnt & 0x7ff; - if (dev->accel.cmd & 0x20) { - dev->accel.dx -= (dev->accel.sx) + 1; - dev->accel.cx -= (dev->accel.sx) + 1; - } else { - dev->accel.dx += (dev->accel.sx) + 1; - dev->accel.cx += (dev->accel.sx) + 1; - } + dev->accel.dx = dev->accel.destx; + + if (dev->accel.destx >= 0x600) + dev->accel.dx |= ~0x5ff; + + dev->accel.cx = dev->accel.cur_x; + + if (dev->accel.cur_x >= 0x600) + dev->accel.cx |= ~0x5ff; if (dev->accel.cmd & 0x80) { dev->accel.dy++; @@ -3803,9 +3820,8 @@ bitblt: dev->accel.src = dev->accel.cy * dev->pitch; dev->accel.sy--; - if (dev->accel.sy < 0) { + if (dev->accel.sy < 0) return; - } } } } @@ -3818,6 +3834,8 @@ bitblt: } } +#undef CLAMP + void ibm8514_render_8bpp(svga_t *svga) { diff --git a/src/video/vid_ati_mach8.c b/src/video/vid_ati_mach8.c index 1fca83369..be034c6d5 100644 --- a/src/video/vid_ati_mach8.c +++ b/src/video/vid_ati_mach8.c @@ -149,6 +149,7 @@ typedef struct mach_t { int16_t cy_end; int16_t dx; int16_t dx_end; + int16_t dy; int16_t dy_end; int16_t dx_start; int16_t dy_start; @@ -165,6 +166,7 @@ typedef struct mach_t { int16_t width; int16_t src_width; int16_t height; + int16_t bleft, bright, btop, bbottom; int poly_src; int temp_cnt; int stepx; @@ -413,7 +415,7 @@ static void mach_accel_start(int cmd_type, int cpu_input, int count, uint32_t mix_dat, uint32_t cpu_dat, mach_t *mach, ibm8514_t *dev) { int compare_mode; - int poly_src = 0; + uint16_t poly_src = 0; uint16_t rd_mask = dev->accel.rd_mask; uint16_t wrt_mask = dev->accel.wrt_mask; uint16_t dest_cmp_clr = dev->accel.color_cmp; @@ -463,7 +465,7 @@ mach_accel_start(int cmd_type, int cpu_input, int count, uint32_t mix_dat, uint3 if (dev->accel_bpp == 24) mach_log("24BPP: CMDType=%d, cwh(%d,%d,%d,%d), dpconfig=%04x\n", cmd_type, clip_l, clip_r, clip_t, clip_b, mach->accel.dp_config); else - mach_log("BPP=%d, CMDType = %d, offs=%08x, DPCONFIG = %04x, cnt = %d, input = %d, mono_src = %d, frgdsel = %d, dstx = %d, dstxend = %d, pitch = %d, extcrt = %d, rw = %x, monpattern = %x.\n", dev->accel_bpp, cmd_type, mach->accel.ge_offset, mach->accel.dp_config, count, cpu_input, mono_src, frgd_sel, dev->accel.cur_x, mach->accel.dest_x_end, dev->ext_pitch, dev->ext_crt_pitch, mach->accel.dp_config & 1, mach->accel.mono_pattern_enable); + mach_log("RdMask=%04x, Clipping: l=%d, r=%d, t=%d, b=%d, LineDrawOpt=%04x, BPP=%d, CMDType = %d, offs=%08x, DPCONFIG = %04x, cnt = %d, input = %d, mono_src = %d, frgdsel = %d, d(%d,%d), dstxend = %d, pitch = %d, extcrt = %d, rw = %x, monpattern = %x.\n", rd_mask, clip_l, clip_r, clip_t, clip_b, mach->accel.linedraw_opt, dev->accel_bpp, cmd_type, mach->accel.ge_offset, mach->accel.dp_config, count, cpu_input, mono_src, frgd_sel, dev->accel.cur_x, dev->accel.cur_y, mach->accel.dest_x_end, dev->ext_pitch, dev->ext_crt_pitch, mach->accel.dp_config & 1, mach->accel.mono_pattern_enable); } switch (cmd_type) { @@ -568,6 +570,41 @@ mach_accel_start(int cmd_type, int cpu_input, int count, uint32_t mix_dat, uint3 } if (((dev->accel.dx) >= clip_l) && ((dev->accel.dx) <= clip_r) && ((dev->accel.dy) >= clip_t) && ((dev->accel.dy) <= clip_b)) { + switch (mix ? frgd_sel : bkgd_sel) { + case 0: + src_dat = dev->accel.bkgd_color; + break; + case 1: + src_dat = dev->accel.frgd_color; + break; + case 2: + src_dat = cpu_dat; + break; + case 3: + if (mach_pixel_read(mach)) + src_dat = cpu_dat; + else { + if (dev->bpp) { + READ((mach->accel.ge_offset << 1) + ((dev->accel.cy) * (dev->pitch)) + (dev->accel.cx), src_dat); + } else { + READ((mach->accel.ge_offset << 2) + ((dev->accel.cy) * (dev->pitch)) + (dev->accel.cx), src_dat); + } + if (mono_src == 3) { + src_dat = (src_dat & rd_mask) == rd_mask; + } + } + break; + case 5: + if (mix) { + src_dat = mach->accel.color_pattern[((dev->accel.dx) + ((dev->accel.dy) << 3)) & mach->accel.patt_len]; + } else + src_dat = 0; + break; + + default: + break; + } + if (mach->accel.linedraw_opt & 0x02) { if (dev->bpp) { READ((mach->accel.ge_offset << 1) + ((dev->accel.cy) * (dev->pitch)) + (dev->accel.cx), poly_src); @@ -579,99 +616,64 @@ mach_accel_start(int cmd_type, int cpu_input, int count, uint32_t mix_dat, uint3 mach->accel.poly_fill = !mach->accel.poly_fill; } - if (!mach->accel.poly_fill || !(mach->accel.linedraw_opt & 0x02)) { - switch (mix ? frgd_sel : bkgd_sel) { - case 0: - src_dat = dev->accel.bkgd_color; - break; + if (mach->accel.poly_fill || !(mach->accel.linedraw_opt & 0x02)) { + if (dev->bpp) { + READ((mach->accel.ge_offset << 1) + ((dev->accel.dy) * (dev->pitch)) + (dev->accel.dx), dest_dat); + } else { + READ((mach->accel.ge_offset << 2) + ((dev->accel.dy) * (dev->pitch)) + (dev->accel.dx), dest_dat); + } + + switch (compare_mode) { case 1: - src_dat = dev->accel.frgd_color; + compare = 1; break; case 2: - src_dat = cpu_dat; + compare = (dest_dat >= dest_cmp_clr) ? 0 : 1; break; case 3: - if (mach_pixel_read(mach)) - src_dat = cpu_dat; - else { - if (dev->bpp) { - READ((mach->accel.ge_offset << 1) + ((dev->accel.cy) * (dev->pitch)) + (dev->accel.cx), src_dat); - } else { - READ((mach->accel.ge_offset << 2) + ((dev->accel.cy) * (dev->pitch)) + (dev->accel.cx), src_dat); - } - if (mono_src == 3) { - src_dat = (src_dat & rd_mask) == rd_mask; - } - } + compare = (dest_dat < dest_cmp_clr) ? 0 : 1; + break; + case 4: + compare = (dest_dat != dest_cmp_clr) ? 0 : 1; break; case 5: - if (mix) { - src_dat = mach->accel.color_pattern[((dev->accel.dx) + ((dev->accel.dy) << 3)) & mach->accel.patt_len]; - } else - src_dat = 0; + compare = (dest_dat == dest_cmp_clr) ? 0 : 1; + break; + case 6: + compare = (dest_dat <= dest_cmp_clr) ? 0 : 1; + break; + case 7: + compare = (dest_dat > dest_cmp_clr) ? 0 : 1; break; default: break; } - if (dev->bpp) { - READ((mach->accel.ge_offset << 1) + ((dev->accel.dy) * (dev->pitch)) + (dev->accel.dx), dest_dat); - } else { - READ((mach->accel.ge_offset << 2) + ((dev->accel.dy) * (dev->pitch)) + (dev->accel.dx), dest_dat); + if (!compare) { + if (mach_pixel_write(mach)) { + old_dest_dat = dest_dat; + MIX(mix, dest_dat, src_dat); + dest_dat = (dest_dat & wrt_mask) | (old_dest_dat & ~wrt_mask); + } } - } - switch (compare_mode) { - case 1: - compare = 1; - break; - case 2: - compare = (dest_dat >= dest_cmp_clr) ? 0 : 1; - break; - case 3: - compare = (dest_dat < dest_cmp_clr) ? 0 : 1; - break; - case 4: - compare = (dest_dat != dest_cmp_clr) ? 0 : 1; - break; - case 5: - compare = (dest_dat == dest_cmp_clr) ? 0 : 1; - break; - case 6: - compare = (dest_dat <= dest_cmp_clr) ? 0 : 1; - break; - case 7: - compare = (dest_dat > dest_cmp_clr) ? 0 : 1; - break; - - default: - break; - } - - if (!compare) { - if (mach_pixel_write(mach)) { - old_dest_dat = dest_dat; - MIX(mix, dest_dat, src_dat); - dest_dat = (dest_dat & wrt_mask) | (old_dest_dat & ~wrt_mask); - } - } - - if (mach->accel.dp_config & 0x10) { - if (mach->accel.linedraw_opt & 0x04) { - if (dev->accel.sx < mach->accel.width) { + if (mach->accel.dp_config & 0x10) { + if (mach->accel.linedraw_opt & 0x04) { + if (((mono_src != 1) && (dev->accel.sx < mach->accel.width)) || ((mono_src == 1) && count)) { + if (dev->bpp) { + WRITE((mach->accel.ge_offset << 1) + ((dev->accel.dy) * (dev->pitch)) + (dev->accel.dx), dest_dat); + } else { + WRITE((mach->accel.ge_offset << 2) + ((dev->accel.dy) * (dev->pitch)) + (dev->accel.dx), dest_dat); + } + } + } else { if (dev->bpp) { WRITE((mach->accel.ge_offset << 1) + ((dev->accel.dy) * (dev->pitch)) + (dev->accel.dx), dest_dat); } else { WRITE((mach->accel.ge_offset << 2) + ((dev->accel.dy) * (dev->pitch)) + (dev->accel.dx), dest_dat); } } - } else { - if (dev->bpp) { - WRITE((mach->accel.ge_offset << 1) + ((dev->accel.dy) * (dev->pitch)) + (dev->accel.dx), dest_dat); - } else { - WRITE((mach->accel.ge_offset << 2) + ((dev->accel.dy) * (dev->pitch)) + (dev->accel.dx), dest_dat); - } } } } @@ -779,6 +781,41 @@ mach_accel_start(int cmd_type, int cpu_input, int count, uint32_t mix_dat, uint3 } if (((dev->accel.dx) >= clip_l) && ((dev->accel.dx) <= clip_r) && ((dev->accel.dy) >= clip_t) && ((dev->accel.dy) <= clip_b)) { + switch (mix ? frgd_sel : bkgd_sel) { + case 0: + src_dat = dev->accel.bkgd_color; + break; + case 1: + src_dat = dev->accel.frgd_color; + break; + case 2: + src_dat = cpu_dat; + break; + case 3: + if (mach_pixel_read(mach)) + src_dat = cpu_dat; + else { + if (dev->bpp) { + READ((mach->accel.ge_offset << 1) + ((dev->accel.cy) * (dev->pitch)) + (dev->accel.cx), src_dat); + } else { + READ((mach->accel.ge_offset << 2) + ((dev->accel.cy) * (dev->pitch)) + (dev->accel.cx), src_dat); + } + if (mono_src == 3) { + src_dat = (src_dat & rd_mask) == rd_mask; + } + } + break; + case 5: + if (mix) { + src_dat = mach->accel.color_pattern[((dev->accel.dx) + ((dev->accel.dy) << 3)) & mach->accel.patt_len]; + } else + src_dat = 0; + break; + + default: + break; + } + if (mach->accel.linedraw_opt & 0x02) { if (dev->bpp) { READ((mach->accel.ge_offset << 1) + ((dev->accel.cy) * (dev->pitch)) + (dev->accel.cx), poly_src); @@ -790,99 +827,64 @@ mach_accel_start(int cmd_type, int cpu_input, int count, uint32_t mix_dat, uint3 mach->accel.poly_fill = !mach->accel.poly_fill; } - if (!mach->accel.poly_fill || !(mach->accel.linedraw_opt & 0x02)) { - switch (mix ? frgd_sel : bkgd_sel) { - case 0: - src_dat = dev->accel.bkgd_color; - break; + if (mach->accel.poly_fill || !(mach->accel.linedraw_opt & 0x02)) { + if (dev->bpp) { + READ((mach->accel.ge_offset << 1) + ((dev->accel.dy) * (dev->pitch)) + (dev->accel.dx), dest_dat); + } else { + READ((mach->accel.ge_offset << 2) + ((dev->accel.dy) * (dev->pitch)) + (dev->accel.dx), dest_dat); + } + + switch (compare_mode) { case 1: - src_dat = dev->accel.frgd_color; + compare = 1; break; case 2: - src_dat = cpu_dat; + compare = (dest_dat >= dest_cmp_clr) ? 0 : 1; break; case 3: - if (mach_pixel_read(mach)) - src_dat = cpu_dat; - else { - if (dev->bpp) { - READ((mach->accel.ge_offset << 1) + ((dev->accel.cy) * (dev->pitch)) + (dev->accel.cx), src_dat); - } else { - READ((mach->accel.ge_offset << 2) + ((dev->accel.cy) * (dev->pitch)) + (dev->accel.cx), src_dat); - } - if (mono_src == 3) { - src_dat = (src_dat & rd_mask) == rd_mask; - } - } + compare = (dest_dat < dest_cmp_clr) ? 0 : 1; + break; + case 4: + compare = (dest_dat != dest_cmp_clr) ? 0 : 1; break; case 5: - if (mix) { - src_dat = mach->accel.color_pattern[((dev->accel.dx) + ((dev->accel.dy) << 3)) & mach->accel.patt_len]; - } else - src_dat = 0; + compare = (dest_dat == dest_cmp_clr) ? 0 : 1; + break; + case 6: + compare = (dest_dat <= dest_cmp_clr) ? 0 : 1; + break; + case 7: + compare = (dest_dat > dest_cmp_clr) ? 0 : 1; break; default: break; } - if (dev->bpp) { - READ((mach->accel.ge_offset << 1) + ((dev->accel.dy) * (dev->pitch)) + (dev->accel.dx), dest_dat); - } else { - READ((mach->accel.ge_offset << 2) + ((dev->accel.dy) * (dev->pitch)) + (dev->accel.dx), dest_dat); + if (!compare) { + if (mach_pixel_write(mach)) { + old_dest_dat = dest_dat; + MIX(mix, dest_dat, src_dat); + dest_dat = (dest_dat & wrt_mask) | (old_dest_dat & ~wrt_mask); + } } - } - switch (compare_mode) { - case 1: - compare = 1; - break; - case 2: - compare = (dest_dat >= dest_cmp_clr) ? 0 : 1; - break; - case 3: - compare = (dest_dat < dest_cmp_clr) ? 0 : 1; - break; - case 4: - compare = (dest_dat != dest_cmp_clr) ? 0 : 1; - break; - case 5: - compare = (dest_dat == dest_cmp_clr) ? 0 : 1; - break; - case 6: - compare = (dest_dat <= dest_cmp_clr) ? 0 : 1; - break; - case 7: - compare = (dest_dat > dest_cmp_clr) ? 0 : 1; - break; - - default: - break; - } - - if (!compare) { - if (mach_pixel_write(mach)) { - old_dest_dat = dest_dat; - MIX(mix, dest_dat, src_dat); - dest_dat = (dest_dat & wrt_mask) | (old_dest_dat & ~wrt_mask); - } - } - - if (mach->accel.dp_config & 0x10) { - if (mach->accel.linedraw_opt & 0x04) { - if (dev->accel.sx < mach->accel.width) { + if (mach->accel.dp_config & 0x10) { + if (mach->accel.linedraw_opt & 0x04) { + if (((mono_src != 1) && (dev->accel.sx < mach->accel.width)) || ((mono_src == 1) && count)) { + if (dev->bpp) { + WRITE((mach->accel.ge_offset << 1) + ((dev->accel.dy) * (dev->pitch)) + (dev->accel.dx), dest_dat); + } else { + WRITE((mach->accel.ge_offset << 2) + ((dev->accel.dy) * (dev->pitch)) + (dev->accel.dx), dest_dat); + } + } + } else { if (dev->bpp) { WRITE((mach->accel.ge_offset << 1) + ((dev->accel.dy) * (dev->pitch)) + (dev->accel.dx), dest_dat); } else { WRITE((mach->accel.ge_offset << 2) + ((dev->accel.dy) * (dev->pitch)) + (dev->accel.dx), dest_dat); } } - } else { - if (dev->bpp) { - WRITE((mach->accel.ge_offset << 1) + ((dev->accel.dy) * (dev->pitch)) + (dev->accel.dx), dest_dat); - } else { - WRITE((mach->accel.ge_offset << 2) + ((dev->accel.dy) * (dev->pitch)) + (dev->accel.dx), dest_dat); - } } } } @@ -928,6 +930,7 @@ mach_accel_start(int cmd_type, int cpu_input, int count, uint32_t mix_dat, uint3 dev->accel.sx++; } } + mach->accel.poly_fill = 0; dev->accel.cur_x = dev->accel.dx; dev->accel.cur_y = dev->accel.dy; break; @@ -1046,10 +1049,10 @@ mach_accel_start(int cmd_type, int cpu_input, int count, uint32_t mix_dat, uint3 else dev->accel.src = (mach->accel.ge_offset << 2) + (dev->accel.cy * (dev->pitch)); - if ((dev->accel_bpp == 24) && (frgd_sel == 5)) { + if ((dev->accel_bpp == 24) && (frgd_sel == 5)) mach_log("BitBLT=%04x, WH(%d,%d), SRCWidth=%d, c(%d,%d), s(%d,%d).\n", mach->accel.dp_config, mach->accel.width, mach->accel.height, mach->accel.src_width, dev->accel.dx, dev->accel.dy, dev->accel.cx, dev->accel.cy); - } else - mach_log("BitBLT=%04x, Pitch=%d, C(%d,%d), SRCWidth=%d, WH(%d,%d), geoffset=%08x.\n", mach->accel.dp_config, dev->ext_pitch, dev->accel.cx, dev->accel.cy, mach->accel.src_width, mach->accel.width, mach->accel.height, (mach->accel.ge_offset << 2)); + else if (mach->accel.dp_config & 0x02) + mach_log("BitBLT=%04x, Pitch=%d, C(%d,%d), D(%d,%d), SRCWidth=%d, SRCXStep=%d, WH(%d,%d), clipt=%d, clipb=%d, geoffset=%08x.\n", mach->accel.dp_config, dev->ext_pitch, mach->accel.src_x, mach->accel.src_y, dev->accel.cur_x, dev->accel.cur_y, mach->accel.src_width, mach->accel.src_stepx, mach->accel.width, mach->accel.height, clip_t, clip_b, (mach->accel.ge_offset << 2)); if (mono_src == 1) { if ((mach->accel.mono_pattern_enable) && !(mach->accel.patt_len_reg & 0x4000)) { @@ -1183,14 +1186,14 @@ mach_accel_start(int cmd_type, int cpu_input, int count, uint32_t mix_dat, uint3 } if (((dev->accel.dx) >= clip_l) && ((dev->accel.dx) <= clip_r) && ((dev->accel.dy) >= clip_t) && ((dev->accel.dy) <= clip_b)) { - if (mach->accel.dp_config & 0x02) { - READ(dev->accel.src + (dev->accel.cx), poly_src); + if ((mach->accel.dp_config & 0x02) || (mach->accel.linedraw_opt & 0x02)) { + READ(dev->accel.src + dev->accel.cx, poly_src); poly_src = ((poly_src & rd_mask) == rd_mask); if (poly_src) - mach->accel.poly_fill = !mach->accel.poly_fill; + mach->accel.poly_fill ^= 1; } - if (!mach->accel.poly_fill || !(mach->accel.dp_config & 0x02)) { + if (mach->accel.poly_fill || !(mach->accel.dp_config & 0x02) || !(mach->accel.linedraw_opt & 0x02)) { switch (mix ? frgd_sel : bkgd_sel) { case 0: src_dat = dev->accel.bkgd_color; @@ -1205,7 +1208,7 @@ mach_accel_start(int cmd_type, int cpu_input, int count, uint32_t mix_dat, uint3 if (mach_pixel_read(mach)) src_dat = cpu_dat; else { - READ(dev->accel.src + (dev->accel.cx), src_dat); + READ(dev->accel.src + dev->accel.cx, src_dat); if (mono_src == 3) src_dat = (src_dat & rd_mask) == rd_mask; } @@ -1223,62 +1226,62 @@ mach_accel_start(int cmd_type, int cpu_input, int count, uint32_t mix_dat, uint3 default: break; } - } - if ((dev->accel_bpp == 24) && (mono_src == 1) && (frgd_sel == 5) && (mach->accel.patt_len_reg & 0x4000)) { - if (dev->accel.sy & 1) { - READ(dev->accel.dest + dev->accel.dx - dev->ext_pitch, dest_dat); + if ((dev->accel_bpp == 24) && (mono_src == 1) && (frgd_sel == 5) && (mach->accel.patt_len_reg & 0x4000)) { + if (dev->accel.sy & 1) { + READ(dev->accel.dest + dev->accel.dx - dev->ext_pitch, dest_dat); + } else { + READ(dev->accel.dest + dev->accel.dx, dest_dat); + } } else { READ(dev->accel.dest + dev->accel.dx, dest_dat); } - } else { - READ(dev->accel.dest + dev->accel.dx, dest_dat); - } - switch (compare_mode) { - case 1: - compare = 1; - break; - case 2: - compare = (dest_dat >= dest_cmp_clr) ? 0 : 1; - break; - case 3: - compare = (dest_dat < dest_cmp_clr) ? 0 : 1; - break; - case 4: - compare = (dest_dat != dest_cmp_clr) ? 0 : 1; - break; - case 5: - compare = (dest_dat == dest_cmp_clr) ? 0 : 1; - break; - case 6: - compare = (dest_dat <= dest_cmp_clr) ? 0 : 1; - break; - case 7: - compare = (dest_dat > dest_cmp_clr) ? 0 : 1; - break; + switch (compare_mode) { + case 1: + compare = 1; + break; + case 2: + compare = (dest_dat >= dest_cmp_clr) ? 0 : 1; + break; + case 3: + compare = (dest_dat < dest_cmp_clr) ? 0 : 1; + break; + case 4: + compare = (dest_dat != dest_cmp_clr) ? 0 : 1; + break; + case 5: + compare = (dest_dat == dest_cmp_clr) ? 0 : 1; + break; + case 6: + compare = (dest_dat <= dest_cmp_clr) ? 0 : 1; + break; + case 7: + compare = (dest_dat > dest_cmp_clr) ? 0 : 1; + break; - default: - break; - } - - if (!compare) { - if (mach_pixel_write(mach)) { - old_dest_dat = dest_dat; - MIX(mix, dest_dat, src_dat); - dest_dat = (dest_dat & wrt_mask) | (old_dest_dat & ~wrt_mask); + default: + break; } - } - if (mach->accel.dp_config & 0x10) { - if ((dev->accel_bpp == 24) && (mono_src == 1) && (frgd_sel == 5) && (mach->accel.patt_len_reg & 0x4000)) { - if (dev->accel.sy & 1) { - WRITE(dev->accel.dest + dev->accel.dx - dev->ext_pitch, dest_dat); + if (!compare) { + if (mach_pixel_write(mach)) { + old_dest_dat = dest_dat; + MIX(mix, dest_dat, src_dat); + dest_dat = (dest_dat & wrt_mask) | (old_dest_dat & ~wrt_mask); + } + } + + if (mach->accel.dp_config & 0x10) { + if ((dev->accel_bpp == 24) && (mono_src == 1) && (frgd_sel == 5) && (mach->accel.patt_len_reg & 0x4000)) { + if (dev->accel.sy & 1) { + WRITE(dev->accel.dest + dev->accel.dx - dev->ext_pitch, dest_dat); + } else { + WRITE(dev->accel.dest + dev->accel.dx, dest_dat); + } } else { WRITE(dev->accel.dest + dev->accel.dx, dest_dat); } - } else { - WRITE(dev->accel.dest + dev->accel.dx, dest_dat); } } } @@ -1288,7 +1291,7 @@ mach_accel_start(int cmd_type, int cpu_input, int count, uint32_t mix_dat, uint3 else cpu_dat >>= 8; - if ((mono_src == 3) || (frgd_sel == 3) || (bkgd_sel == 3)) { + if ((mono_src == 3) || (frgd_sel == 3) || (bkgd_sel == 3) || (mach->accel.dp_config & 0x02)) { dev->accel.cx += mach->accel.src_stepx; mach->accel.sx++; if (mach->accel.sx >= mach->accel.src_width) { @@ -1334,7 +1337,6 @@ mach_accel_start(int cmd_type, int cpu_input, int count, uint32_t mix_dat, uint3 dev->accel.sx++; if (dev->accel.sx >= mach->accel.width) { - mach->accel.poly_fill = 0; dev->accel.sx = 0; if (mach->accel.stepx == -1) dev->accel.dx += mach->accel.width; @@ -1344,6 +1346,7 @@ mach_accel_start(int cmd_type, int cpu_input, int count, uint32_t mix_dat, uint3 dev->accel.dy += mach->accel.stepy; dev->accel.sy++; + mach->accel.poly_fill = 0; if (dev->bpp) dev->accel.dest = (mach->accel.ge_offset << 1) + (dev->accel.dy * (dev->pitch)); else @@ -1362,7 +1365,7 @@ mach_accel_start(int cmd_type, int cpu_input, int count, uint32_t mix_dat, uint3 return; } if (dev->accel.sy >= mach->accel.height) { - if ((mono_src == 2) || (mono_src == 3) || (frgd_sel == 3) || (bkgd_sel == 3)) + if ((mono_src == 2) || (mono_src == 3) || (frgd_sel == 3) || (bkgd_sel == 3) || (mach->accel.dp_config & 0x02) || (mach->accel.linedraw_opt & 0x02)) return; if ((mono_src == 1) && (frgd_sel == 5) && (dev->accel_bpp == 24) && (mach->accel.patt_len_reg & 0x4000)) return; @@ -1397,7 +1400,7 @@ mach_accel_start(int cmd_type, int cpu_input, int count, uint32_t mix_dat, uint3 dev->accel.sx = 0; - mach_log("Linedraw: c(%d,%d), d(%d,%d), cend(%d,%d).\n", dev->accel.cur_x, dev->accel.cur_y, dev->accel.dx, dev->accel.dy, mach->accel.cx_end_line, mach->accel.cy_end_line); + mach_log("Linedraw: c(%d,%d), d(%d,%d), cend(%d,%d), bounds: l=%d, r=%d, t=%d, b=%d.\n", dev->accel.cur_x, dev->accel.cur_y, dev->accel.dx, dev->accel.dy, mach->accel.cx_end_line, mach->accel.cy_end_line, mach->accel.bleft, mach->accel.bright, mach->accel.btop, mach->accel.bbottom); if ((mono_src == 2) || (bkgd_sel == 2) || (frgd_sel == 2) || mach_pixel_read(mach)) { if (mach_pixel_write(mach)) { @@ -1439,7 +1442,7 @@ mach_accel_start(int cmd_type, int cpu_input, int count, uint32_t mix_dat, uint3 mix_dat <<= 1; mix_dat |= 1; - if (((dev->accel.cx) >= clip_l) && ((dev->accel.cx) <= clip_r) && ((dev->accel.cy) >= clip_t) && ((dev->accel.cy) <= clip_b)) { + if ((dev->accel.cx >= clip_l) && (dev->accel.cx <= clip_r) && (dev->accel.cy >= clip_t) && (dev->accel.cy <= clip_b)) { mach->accel.clip_overrun = 0; switch (mix ? frgd_sel : bkgd_sel) { case 0: @@ -1454,14 +1457,13 @@ mach_accel_start(int cmd_type, int cpu_input, int count, uint32_t mix_dat, uint3 case 3: if (mach_pixel_read(mach)) src_dat = cpu_dat; - else { + else src_dat = 0; - } break; case 5: - if (mix) { + if (mix) src_dat = mach->accel.color_pattern[((dev->accel.cx) + ((dev->accel.cy) << 3)) & mach->accel.patt_len]; - } else + else src_dat = 0; break; @@ -1509,7 +1511,15 @@ mach_accel_start(int cmd_type, int cpu_input, int count, uint32_t mix_dat, uint3 dest_dat = (dest_dat & wrt_mask) | (old_dest_dat & ~wrt_mask); } } - if ((mach->accel.dp_config & 0x10) && (cmd_type == 3)) { + if (mach->accel.linedraw_opt & 0x04) { + if (count) { + if (dev->bpp) { + WRITE((mach->accel.ge_offset << 1) + ((dev->accel.cy) * (dev->pitch)) + (dev->accel.cx), dest_dat); + } else { + WRITE((mach->accel.ge_offset << 2) + ((dev->accel.cy) * (dev->pitch)) + (dev->accel.cx), dest_dat); + } + } + } else { if (dev->bpp) { WRITE((mach->accel.ge_offset << 1) + ((dev->accel.cy) * (dev->pitch)) + (dev->accel.cx), dest_dat); } else { @@ -1561,8 +1571,14 @@ mach_accel_start(int cmd_type, int cpu_input, int count, uint32_t mix_dat, uint3 break; } - if (((dev->accel.cx) >= clip_l) && ((dev->accel.cx) <= clip_r) && ((dev->accel.cy) >= clip_t) && ((dev->accel.cy) <= clip_b)) { + if ((dev->accel.cx >= clip_l) && (dev->accel.cx <= clip_r) && (dev->accel.cy >= clip_t) && (dev->accel.cy <= clip_b)) { mach->accel.clip_overrun = 0; + if (mach->accel.linedraw_opt & 0x02) { + READ((mach->accel.ge_offset << 2) + ((dev->accel.cy) * (dev->pitch)) + (dev->accel.cx), poly_src); + if (poly_src) + mach->accel.poly_fill = !mach->accel.poly_fill; + } + switch (mix ? frgd_sel : bkgd_sel) { case 0: src_dat = dev->accel.bkgd_color; @@ -1627,7 +1643,9 @@ mach_accel_start(int cmd_type, int cpu_input, int count, uint32_t mix_dat, uint3 if (!compare) { if (mach_pixel_write(mach)) { old_dest_dat = dest_dat; - MIX(mix, dest_dat, src_dat); + if (mach->accel.poly_fill || !(mach->accel.linedraw_opt & 0x02)) { + MIX(mix, dest_dat, src_dat); + } dest_dat = (dest_dat & wrt_mask) | (old_dest_dat & ~wrt_mask); } } @@ -1683,7 +1701,7 @@ mach_accel_start(int cmd_type, int cpu_input, int count, uint32_t mix_dat, uint3 mix_dat <<= 1; mix_dat |= 1; - if (((dev->accel.cx) >= clip_l) && ((dev->accel.cx) <= clip_r) && ((dev->accel.cy) >= clip_t) && ((dev->accel.cy) <= clip_b)) { + if ((dev->accel.cx >= clip_l) && (dev->accel.cx <= clip_r) && (dev->accel.cy >= clip_t) && (dev->accel.cy <= clip_b)) { mach->accel.clip_overrun = 0; switch (mix ? frgd_sel : bkgd_sel) { case 0: @@ -1754,10 +1772,20 @@ mach_accel_start(int cmd_type, int cpu_input, int count, uint32_t mix_dat, uint3 } if ((mach->accel.dp_config & 0x10) && (cmd_type == 3)) { - if (dev->bpp) { - WRITE((mach->accel.ge_offset << 1) + ((dev->accel.cy) * (dev->pitch)) + (dev->accel.cx), dest_dat); + if (mach->accel.linedraw_opt & 0x04) { + if (count) { + if (dev->bpp) { + WRITE((mach->accel.ge_offset << 1) + ((dev->accel.cy) * (dev->pitch)) + (dev->accel.cx), dest_dat); + } else { + WRITE((mach->accel.ge_offset << 2) + ((dev->accel.cy) * (dev->pitch)) + (dev->accel.cx), dest_dat); + } + } } else { - WRITE((mach->accel.ge_offset << 2) + ((dev->accel.cy) * (dev->pitch)) + (dev->accel.cx), dest_dat); + if (dev->bpp) { + WRITE((mach->accel.ge_offset << 1) + ((dev->accel.cy) * (dev->pitch)) + (dev->accel.cx), dest_dat); + } else { + WRITE((mach->accel.ge_offset << 2) + ((dev->accel.cy) * (dev->pitch)) + (dev->accel.cx), dest_dat); + } } } } else @@ -1805,8 +1833,9 @@ mach_accel_start(int cmd_type, int cpu_input, int count, uint32_t mix_dat, uint3 break; } - if (((dev->accel.cx) >= clip_l) && ((dev->accel.cx) <= clip_r) && ((dev->accel.cy) >= clip_t) && ((dev->accel.cy) <= clip_b)) { + if ((dev->accel.cx >= clip_l) && (dev->accel.cx <= clip_r) && (dev->accel.cy >= clip_t) && (dev->accel.cy <= clip_b)) { mach->accel.clip_overrun = 0; + switch (mix ? frgd_sel : bkgd_sel) { case 0: src_dat = dev->accel.bkgd_color; @@ -1915,6 +1944,7 @@ mach_accel_start(int cmd_type, int cpu_input, int count, uint32_t mix_dat, uint3 } } } + mach->accel.poly_fill = 0; mach->accel.line_array[(cmd_type == 4) ? 4 : 0] = dev->accel.cx; mach->accel.line_array[(cmd_type == 4) ? 5 : 1] = dev->accel.cy; dev->accel.cur_x = mach->accel.line_array[(cmd_type == 4) ? 4 : 0]; @@ -2828,9 +2858,8 @@ mach_accel_out_fifo(mach_t *mach, svga_t *svga, ibm8514_t *dev, uint16_t port, u case 0xf6ee: if (len == 1) { dev->accel.cur_y = (dev->accel.cur_y & 0x700) | val; - } else { + } else dev->accel.cur_y = val & 0x7ff; - } break; case 0x82e9: case 0xc2e9: @@ -2844,9 +2873,8 @@ mach_accel_out_fifo(mach_t *mach, svga_t *svga, ibm8514_t *dev, uint16_t port, u case 0xc6e8: if (len == 1) { dev->accel.cur_x = (dev->accel.cur_x & 0x700) | val; - } else { + } else dev->accel.cur_x = val & 0x7ff; - } break; case 0x86e9: case 0xc6e9: @@ -2861,6 +2889,7 @@ mach_accel_out_fifo(mach_t *mach, svga_t *svga, ibm8514_t *dev, uint16_t port, u dev->accel.desty_axstp = (dev->accel.desty_axstp & 0x3f00) | val; else { mach->accel.src_y = val; + dev->accel.desty = val & 0x07ff; dev->accel.desty_axstp = val & 0x3fff; if (val & 0x2000) dev->accel.desty_axstp |= ~0x1fff; @@ -2881,6 +2910,7 @@ mach_accel_out_fifo(mach_t *mach, svga_t *svga, ibm8514_t *dev, uint16_t port, u dev->accel.destx_distp = (dev->accel.destx_distp & 0x3f00) | val; else { mach->accel.src_x = val; + dev->accel.destx = val & 0x07ff; dev->accel.destx_distp = val & 0x3fff; if (val & 0x2000) dev->accel.destx_distp |= ~0x1fff; @@ -2901,7 +2931,6 @@ mach_accel_out_fifo(mach_t *mach, svga_t *svga, ibm8514_t *dev, uint16_t port, u fallthrough; case 0xd2e8: - mach_log("92E8 = %04x\n", val); if (len == 1) dev->accel.err_term = (dev->accel.err_term & 0x3f00) | val; else { @@ -2926,6 +2955,7 @@ mach_accel_out_fifo(mach_t *mach, svga_t *svga, ibm8514_t *dev, uint16_t port, u else { mach->accel.test = val & 0x1fff; dev->accel.maj_axis_pcnt = val & 0x07ff; + dev->accel.maj_axis_pcnt_no_limit = val; } break; case 0x96e9: @@ -2944,7 +2974,7 @@ mach_accel_out_fifo(mach_t *mach, svga_t *svga, ibm8514_t *dev, uint16_t port, u dev->data_available = 0; dev->data_available2 = 0; dev->accel.cmd = val; - mach_log("CMD8514 = %04x.\n", val); + mach_log("CMD8514=%04x, len=%d, pixcntl=%02x.\n", val, len, dev->accel.multifunc[0x0a]); mach->accel.cmd_type = -1; if (port == 0xdae8) { if (dev->accel.cmd & 0x100) @@ -3192,8 +3222,10 @@ mach_accel_out_fifo(mach_t *mach, svga_t *svga, ibm8514_t *dev, uint16_t port, u case 0xeae8: if (len == 1) dev->accel.wrt_mask = (dev->accel.wrt_mask & 0x00ff) | val; - else + else { dev->accel.wrt_mask = val; + mach_log("WrtMask=%04x.\n", val); + } break; case 0xaae9: case 0xeae9: @@ -3205,8 +3237,10 @@ mach_accel_out_fifo(mach_t *mach, svga_t *svga, ibm8514_t *dev, uint16_t port, u case 0xeee8: if (len == 1) dev->accel.rd_mask = (dev->accel.rd_mask & 0x00ff) | val; - else + else { dev->accel.rd_mask = val; + mach_log("ReadMask=%04x.\n", val); + } break; case 0xaee9: case 0xeee9: @@ -3344,7 +3378,7 @@ mach_accel_out_fifo(mach_t *mach, svga_t *svga, ibm8514_t *dev, uint16_t port, u mach->accel.bres_count = (mach->accel.bres_count & 0x700) | val; else { mach->accel.bres_count = val & 0x7ff; - mach_log("96EE line draw.\n"); + mach_log("BresenhamDraw = %04x.\n", mach->accel.dp_config); dev->data_available = 0; dev->data_available2 = 0; mach->accel.cmd_type = 1; @@ -3372,6 +3406,16 @@ mach_accel_out_fifo(mach_t *mach, svga_t *svga, ibm8514_t *dev, uint16_t port, u mach->accel.linedraw_opt = (mach->accel.linedraw_opt & 0xff00) | val; else { mach->accel.linedraw_opt = val; + mach->accel.bbottom = dev->accel.multifunc[3] & 0x7ff; + mach->accel.btop = dev->accel.clip_top & 0x7ff; + mach->accel.bleft = dev->accel.clip_left & 0x7ff; + mach->accel.bright = dev->accel.multifunc[4] & 0x7ff; + if (mach->accel.linedraw_opt & 0x100) { + mach->accel.bbottom = 2047; + mach->accel.btop = 0; + mach->accel.bleft = 0; + mach->accel.bright = 2047; + } } break; case 0xa2ef: @@ -3404,7 +3448,6 @@ mach_accel_out_fifo(mach_t *mach, svga_t *svga, ibm8514_t *dev, uint16_t port, u break; case 0xaeee: - mach_log("AEEE write val = %04x.\n", val); if (len == 1) mach->accel.dest_y_end = (mach->accel.dest_y_end & 0x700) | val; else { @@ -3471,7 +3514,6 @@ mach_accel_out_fifo(mach_t *mach, svga_t *svga, ibm8514_t *dev, uint16_t port, u break; case 0xcaee: - mach_log("CAEE write val = %04x.\n", val); if (len == 1) mach->accel.scan_to_x = (mach->accel.scan_to_x & 0x700) | val; else { @@ -3609,8 +3651,7 @@ mach_accel_out_fifo(mach_t *mach, svga_t *svga, ibm8514_t *dev, uint16_t port, u break; case 0xfeee: - if (mach->accel.dp_config == 0x2231 || mach->accel.dp_config == 0x2211) - mach_log("FEEE val = %d, lineidx = %d, DPCONFIG = %04x, CPUCX = %04x.\n", val, mach->accel.line_idx, mach->accel.dp_config, CX); + mach_log("LineDraw = %04x.\n", mach->accel.dp_config); if (len != 1) { mach->accel.line_array[mach->accel.line_idx] = val; dev->accel.cur_x = mach->accel.line_array[(mach->accel.line_idx == 4) ? 4 : 0]; @@ -3637,6 +3678,7 @@ mach_accel_out(uint16_t port, uint8_t val, mach_t *mach) { svga_t *svga = &mach->svga; ibm8514_t *dev = (ibm8514_t *) svga->dev8514; + uint8_t old = 0; mach_log("[%04X:%08X]: Port NORMAL OUT=%04x, val=%04x.\n", CS, cpu_state.pc, port, val); @@ -3709,10 +3751,27 @@ mach_accel_out(uint16_t port, uint8_t val, mach_t *mach) break; case 0x42e8: - dev->subsys_stat &= ~val; + old = dev->subsys_stat; + if (val & 1) + dev->subsys_stat &= ~1; + if (val & 2) + dev->subsys_stat &= ~2; + if (val & 4) + dev->subsys_stat &= ~4; + if (val & 8) + dev->subsys_stat &= ~8; break; case 0x42e9: + old = dev->subsys_cntl; dev->subsys_cntl = val; + if ((old ^ val) & 1) + dev->subsys_stat |= 1; + if ((old ^ val) & 2) + dev->subsys_stat |= 2; + if ((old ^ val) & 4) + dev->subsys_stat |= 4; + if ((old ^ val) & 8) + dev->subsys_stat |= 8; break; case 0x4ae8: @@ -3860,14 +3919,12 @@ mach_accel_out(uint16_t port, uint8_t val, mach_t *mach) case 0x52ef: mach_log("ATI 8514/A: (0x%04x) val = %04x.\n", port, val); WRITE8(port, mach->accel.scratch0, val); - mach->ext_on[port & 1] = 1; break; case 0x56ee: case 0x56ef: mach_log("ATI 8514/A: (0x%04x) val = %04x.\n", port, val); WRITE8(port, mach->accel.scratch1, val); - mach->ext_on[port & 1] = 1; break; case 0x5aee: @@ -4376,16 +4433,17 @@ mach_accel_in(uint16_t port, mach_t *mach) uint8_t temp = 0; uint16_t vpos = 0; uint16_t vblankend = svga->vblankstart + svga->crtc[0x16]; + int cmd; switch (port) { case 0x2e8: vpos = dev->vc & 0x7ff; if (vblankend > dev->v_total) { vblankend -= dev->v_total; - if (vpos >= svga->vblankstart || vpos <= vblankend) + if ((vpos >= svga->vblankstart) || (vpos <= vblankend)) temp |= 2; } else { - if (vpos >= svga->vblankstart && vpos <= vblankend) + if ((vpos >= svga->vblankstart) && (vpos <= vblankend)) temp |= 2; } break; @@ -4412,14 +4470,17 @@ mach_accel_in(uint16_t port, mach_t *mach) case 0x42e8: case 0x42e9: + cmd = dev->accel.cmd >> 13; vpos = dev->vc & 0x7ff; - if (vblankend > dev->v_total) { - vblankend -= dev->v_total; - if (vpos >= svga->vblankstart || vpos <= vblankend) - dev->subsys_stat |= 1; - } else { - if (vpos >= svga->vblankstart && vpos <= vblankend) - dev->subsys_stat |= 1; + if (!(dev->subsys_stat & 1)) { + if (vblankend > dev->v_total) { + vblankend -= dev->v_total; + if ((vpos >= svga->vblankstart) || (vpos <= vblankend)) + dev->subsys_stat |= 1; + } else { + if ((vpos >= svga->vblankstart) && (vpos <= vblankend)) + dev->subsys_stat |= 1; + } } if (port & 1) { @@ -4534,22 +4595,22 @@ mach_accel_in(uint16_t port, mach_t *mach) case 0x72ee: case 0x72ef: - READ8(port, dev->accel.clip_left); + READ8(port, (mach->accel.bleft)); break; case 0x76ee: case 0x76ef: - READ8(port, dev->accel.clip_top); + READ8(port, (mach->accel.btop)); break; case 0x7aee: case 0x7aef: - READ8(port, dev->accel.multifunc[4]); + READ8(port, (mach->accel.bright)); break; case 0x7eee: case 0x7eef: - READ8(port, dev->accel.multifunc[3]); + READ8(port, (mach->accel.bbottom)); break; default: