mirror/86Box
1
0
Fork 0
mirror of https://github.com/86Box/86Box.git synced 2026-02-22 01:25:33 -07:00
Code Issues Packages Projects Releases Activity

Merge pull request #6419 from pixel-jupiter/tandy-shake

Browse Source 
Tandy: Implement CRTC-based display shifting (R2 & R7 registers)
This commit is contained in:
Miran Grča
2025-10-29 18:11:59 +01:00
committed by GitHub
parent 74f9314d27 258c89151c
commit 8b02672a2d
2 changed files with 242 additions and 93 deletions
Download Patch File Download Diff File Expand all files Collapse all files

8
src/include/86box/m_tandy.h
View File

@@ -33,6 +33,13 @@ typedef struct t1kvid_t {
uint8_t planar_ctrl;
uint8_t lp_strobe;
uint8_t baseline_hsyncpos;
uint8_t baseline_vsyncpos;
bool baseline_ready;
int hsync_offset;
int vsync_offset;
int vsync_offset_pending;
int linepos;
int displine;
int scanline;
@@ -51,6 +58,7 @@ typedef struct t1kvid_t {
uint64_t dispontime;
uint64_t dispofftime;
pc_timer_t timer;
pc_timer_t calib_timer;
int firstline;
int lastline;

327
src/video/vid_tandy.c
View File

@@ -16,6 +16,7 @@
* Copyright 2016-2019 Miran Grca.
* Copyright 2025 starfrost
*/
#include <stdbool.h>
#include <stdarg.h>
#include <stdint.h>
#include <stdio.h>
@@ -162,6 +163,48 @@ vid_update_latch(t1kvid_t *vid)
vid->crtc[0x11] = lp_latch & 0xff;
}
static void
baseline_calib_start(t1kvid_t *vid)
{
vid->baseline_ready = false;
vid->hsync_offset = 0;
vid->vsync_offset = 0;
timer_on_auto(&vid->calib_timer, 100000.0);
}
static void
baseline_calib_finish(void *priv)
{
tandy_t *dev = (tandy_t *) priv;
t1kvid_t *vid = dev->vid;
vid->baseline_hsyncpos = vid->crtc[2];
vid->baseline_vsyncpos = vid->crtc[7];
vid->baseline_ready = true;
}
static void
vid_update_display_offset(t1kvid_t *vid, uint8_t reg)
{
int hsync_scale = (vid->mode & 1) ? 8 : 16;
int vsync_scale = vid->crtc[9] + 1;
if (!vid->baseline_ready) {
vid->hsync_offset = 0;
vid->vsync_offset = 0;
return;
} else {
switch (reg) {
case 2:
vid->hsync_offset = ((int)vid->baseline_hsyncpos - (int)vid->crtc[2]) * hsync_scale;
break;
case 7:
vid->vsync_offset_pending = ((int)vid->baseline_vsyncpos - (int)vid->crtc[7]) * vsync_scale;
break;
}
}
}
void
tandy_vid_out(uint16_t addr, uint8_t val, void *priv)
{
@@ -188,12 +231,16 @@ tandy_vid_out(uint16_t addr, uint8_t val, void *priv)
vid->fullchange = changeframecount;
recalc_timings(dev);
}
if (vid->crtcreg == 0x02 || vid->crtcreg == 0x07)
vid_update_display_offset(vid, vid->crtcreg);
}
break;
case 0x03d8:
old = vid->mode;
vid->mode = val;
if (old != val)
baseline_calib_start(vid);
if ((old ^ val) & 0x01)
recalc_timings(dev);
if (!dev->is_sl2)
@@ -337,7 +384,7 @@ vid_read(uint32_t addr, void *priv)
}
static void
vid_render(tandy_t *dev, int line)
vid_render(tandy_t *dev, int line, int hos_offs)
{
t1kvid_t *vid = dev->vid;
uint16_t cursoraddr = (vid->crtc[15] | (vid->crtc[14] << 8)) & 0x3fff;
@@ -349,53 +396,91 @@ vid_render(tandy_t *dev, int line)
uint16_t dat;
int col;
int cols[4];
uint32_t border_val;
int out_x;
int c_start;
cols[0] = (vid->array[2] & 0xf) + 16;
for (c = 0; c < 8; c++) {
if (vid->array[3] & 4) {
buffer32->line[line][c] = cols[0];
if (vid->mode & 1)
buffer32->line[line][c + (vid->crtc[1] << 3) + 8] = cols[0];
else
buffer32->line[line][c + (vid->crtc[1] << 4) + 8] = cols[0];
} else if ((vid->mode & 0x12) == 0x12) {
buffer32->line[line][c] = 0;
if (vid->mode & 1)
buffer32->line[line][c + (vid->crtc[1] << 3) + 8] = 0;
else
buffer32->line[line][c + (vid->crtc[1] << 4) + 8] = 0;
if (line < 0) {
if (dev->is_sl2 && (vid->array[5] & 1)) {
vid->memaddr += vid->crtc[1] * 2;
} else {
buffer32->line[line][c] = buffer32->line[(line) + 1][c] = (vid->col & 15) + 16;
if (vid->mode & 1)
buffer32->line[line][c + (vid->crtc[1] << 3) + 8] = (vid->col & 15) + 16;
else
buffer32->line[line][c + (vid->crtc[1] << 4) + 8] = (vid->col & 15) + 16;
vid->memaddr += vid->crtc[1];
}
return;
}
if (vid->array[3] & 4) {
border_val = cols[0];
} else if ((vid->mode & 0x12) == 0x12) {
border_val = 0;
} else {
border_val = (vid->col & 15) + 16;
}
if (vid->array[3] & 4) {
for (c = 0; c < hos_offs; c++) {
buffer32->line[line][c] = border_val;
}
} else if ((vid->mode & 0x12) == 0x12) {
for (c = 0; c < hos_offs; c++) {
buffer32->line[line][c] = border_val;
}
} else {
for (c = 0; c < hos_offs; c++) {
buffer32->line[line][c] = buffer32->line[(line) + 1][c] = border_val;
}
}
if (vid->mode & 1) {
out_x = (vid->crtc[1] << 3) + hos_offs;
} else {
out_x = (vid->crtc[1] << 4) + hos_offs;
}
c_start = out_x < 0 ? -out_x : 0;
for (c = c_start; c < 8 - vid->hsync_offset; c++) {
buffer32->line[line][out_x + c] = border_val;
}
if (dev->is_sl2 && (vid->array[5] & 1)) { /*640x200x16*/
for (x = 0; x < vid->crtc[1] * 2; x++) {
dat = (vid->vram[(vid->memaddr << 1) & 0xffff] << 8) | vid->vram[((vid->memaddr << 1) + 1) & 0xffff];
vid->memaddr++;
buffer32->line[line][(x << 2) + 8] = vid->array[((dat >> 12) & 0xf) + 16] + 16;
buffer32->line[line][(x << 2) + 9] = vid->array[((dat >> 8) & 0xf) + 16] + 16;
buffer32->line[line][(x << 2) + 10] = vid->array[((dat >> 4) & 0xf) + 16] + 16;
buffer32->line[line][(x << 2) + 11] = vid->array[(dat & 0xf) + 16] + 16;
out_x = (x << 2) + hos_offs;
if (out_x >= 0) {
buffer32->line[line][out_x] = vid->array[((dat >> 12) & 0xf) + 16] + 16;
buffer32->line[line][out_x + 1] = vid->array[((dat >> 8) & 0xf) + 16] + 16;
buffer32->line[line][out_x + 2] = vid->array[((dat >> 4) & 0xf) + 16] + 16;
buffer32->line[line][out_x + 3] = vid->array[(dat & 0xf) + 16] + 16;
} else if (out_x > -4) {
for (c = -out_x; c < 4; c++) {
buffer32->line[line][out_x + c] =
vid->array[((dat >> (12 - c * 4)) & 0xf) + 16] + 16;
}
}
}
} else if ((vid->array[3] & 0x10) && (vid->mode & 1)) { /*320x200x16*/
for (x = 0; x < vid->crtc[1]; x++) {
dat = (vid->vram[((vid->memaddr << 1) & 0x1fff) + ((vid->scanline & 3) * 0x2000)] << 8) |
vid->vram[((vid->memaddr << 1) & 0x1fff) + ((vid->scanline & 3) * 0x2000) + 1];
vid->memaddr++;
buffer32->line[line][(x << 3) + 8] = buffer32->line[line][(x << 3) + 9] =
vid->array[((dat >> 12) & vid->array[1] & 0x0f) + 16] + 16;
buffer32->line[line][(x << 3) + 10] = buffer32->line[line][(x << 3) + 11] =
vid->array[((dat >> 8) & vid->array[1] & 0x0f) + 16] + 16;
buffer32->line[line][(x << 3) + 12] = buffer32->line[line][(x << 3) + 13] =
vid->array[((dat >> 4) & vid->array[1] & 0x0f) + 16] + 16;
buffer32->line[line][(x << 3) + 14] = buffer32->line[line][(x << 3) + 15] =
vid->array[(dat & vid->array[1] & 0x0f) + 16] + 16;
out_x = (x << 3) + hos_offs;
if (out_x >= 0) {
buffer32->line[line][out_x] = buffer32->line[line][out_x + 1] =
vid->array[((dat >> 12) & vid->array[1] & 0x0f) + 16] + 16;
buffer32->line[line][out_x + 2] = buffer32->line[line][out_x + 3] =
vid->array[((dat >> 8) & vid->array[1] & 0x0f) + 16] + 16;
buffer32->line[line][out_x + 4] = buffer32->line[line][out_x + 5] =
vid->array[((dat >> 4) & vid->array[1] & 0x0f) + 16] + 16;
buffer32->line[line][out_x + 6] = buffer32->line[line][out_x + 7] =
vid->array[(dat & vid->array[1] & 0x0f) + 16] + 16;
} else if (out_x > -8) {
for (c = -out_x; c < 8; c++) {
buffer32->line[line][out_x + c] =
vid->array[((dat >> 12 - (c >> 1) * 4) & vid->array[1] & 0x0f) + 16] + 16;
}
}
}
} else if (vid->array[3] & 0x10) { /*160x200x16*/
for (x = 0; x < vid->crtc[1]; x++) {
@@ -406,30 +491,43 @@ vid_render(tandy_t *dev, int line)
dat = (vid->vram[((vid->memaddr << 1) & 0x1fff) + ((vid->scanline & 3) * 0x2000)] << 8) |
vid->vram[((vid->memaddr << 1) & 0x1fff) + ((vid->scanline & 3) * 0x2000) + 1];
vid->memaddr++;
buffer32->line[line][(x << 4) + 8] = buffer32->line[line][(x << 4) + 9] =
buffer32->line[line][(x << 4) + 10] = buffer32->line[line][(x << 4) + 11] =
vid->array[((dat >> 12) & vid->array[1] & 0x0f) + 16] + 16;
buffer32->line[line][(x << 4) + 12] = buffer32->line[line][(x << 4) + 13] =
buffer32->line[line][(x << 4) + 14] = buffer32->line[line][(x << 4) + 15] =
vid->array[((dat >> 8) & vid->array[1] & 0x0f) + 16] + 16;
buffer32->line[line][(x << 4) + 16] = buffer32->line[line][(x << 4) + 17] =
buffer32->line[line][(x << 4) + 18] = buffer32->line[line][(x << 4) + 19] =
vid->array[((dat >> 4) & vid->array[1] & 0x0f) + 16] + 16;
buffer32->line[line][(x << 4) + 20] = buffer32->line[line][(x << 4) + 21] =
buffer32->line[line][(x << 4) + 22] = buffer32->line[line][(x << 4) + 23] =
vid->array[(dat & vid->array[1] & 0x0f) + 16] + 16;
out_x = (x << 4) + hos_offs;
if (out_x >= 0) {
buffer32->line[line][out_x] = buffer32->line[line][out_x + 1] =
buffer32->line[line][out_x + 2] = buffer32->line[line][out_x + 3] =
vid->array[((dat >> 12) & vid->array[1] & 0x0f) + 16] + 16;
buffer32->line[line][out_x + 4] = buffer32->line[line][out_x + 5] =
buffer32->line[line][out_x + 6] = buffer32->line[line][out_x + 7] =
vid->array[((dat >> 8) & vid->array[1] & 0x0f) + 16] + 16;
buffer32->line[line][out_x + 8] = buffer32->line[line][out_x + 9] =
buffer32->line[line][out_x + 10] = buffer32->line[line][out_x + 11] =
vid->array[((dat >> 4) & vid->array[1] & 0x0f) + 16] + 16;
buffer32->line[line][out_x + 12] = buffer32->line[line][out_x + 13] =
buffer32->line[line][out_x + 14] = buffer32->line[line][out_x + 15] =
vid->array[(dat & vid->array[1] & 0x0f) + 16] + 16;
} else if (out_x > -16) {
for (c = -out_x; c < 16; c++) {
buffer32->line[line][out_x + c] =
vid->array[((dat >> 12 - (c >> 2) * 4) & vid->array[1] & 0x0f) + 16] + 16;
}
}
}
} else if (vid->array[3] & 0x08) { /*640x200x4 - this implementation is a complete guess!*/
for (x = 0; x < vid->crtc[1]; x++) {
dat = (vid->vram[((vid->memaddr << 1) & 0x1fff) + ((vid->scanline & 3) * 0x2000)] << 8) |
vid->vram[((vid->memaddr << 1) & 0x1fff) + ((vid->scanline & 3) * 0x2000) + 1];
vid->memaddr++;
for (c = 0; c < 8; c++) {
chr = (dat >> 6) & 2;
chr |= ((dat >> 15) & 1);
buffer32->line[line][(x << 3) + 8 + c] =
vid->array[(chr & vid->array[1]) + 16] + 16;
dat <<= 1;
out_x = (x << 3) + hos_offs;
if (out_x > -8) {
c_start = out_x < 0 ? -out_x : 0;
dat <<= c_start;
for (c = c_start; c < 8; c++) {
chr = (dat >> 6) & 2;
chr |= ((dat >> 15) & 1);
buffer32->line[line][out_x + c] =
vid->array[(chr & vid->array[1]) + 16] + 16;
dat <<= 1;
}
}
}
} else if (vid->mode & 1) {
@@ -446,17 +544,22 @@ vid_render(tandy_t *dev, int line)
cols[1] = vid->array[((attr & 15) & vid->array[1]) + 16] + 16;
cols[0] = vid->array[((attr >> 4) & vid->array[1]) + 16] + 16;
}
if (vid->scanline & 8) for (c = 0; c < 8; c++)
buffer32->line[line][(x << 3) + c + 8] =
((chr >= 0xb3) && (chr <= 0xdf)) ? cols[(fontdat[chr][7] & (1 << (c ^ 7))) ? 1 : 0] : cols[0];
else for (c = 0; c < 8; c++) {
if (vid->scanline == 8)
buffer32->line[line][(x << 3) + c + 8] = cols[(fontdat[chr][7] & (1 << (c ^ 7))) ? 1 : 0];
else
buffer32->line[line][(x << 3) + c + 8] = cols[(fontdat[chr][vid->scanline & 7] & (1 << (c ^ 7))) ? 1 : 0];
out_x = (x << 3) + hos_offs;
if (out_x > -8) {
c_start = out_x < 0 ? -out_x : 0;
if (vid->scanline & 8) for (c = c_start; c < 8; c++) {
buffer32->line[line][out_x + c] =
((chr >= 0xb3) && (chr <= 0xdf)) ? cols[(fontdat[chr][7] & (1 << (c ^ 7))) ? 1 : 0] : cols[0];
} else for (c = c_start; c < 8; c++) {
if (vid->scanline == 8)
buffer32->line[line][out_x + c] = cols[(fontdat[chr][7] & (1 << (c ^ 7))) ? 1 : 0];
else
buffer32->line[line][out_x + c] = cols[(fontdat[chr][vid->scanline & 7] & (1 << (c ^ 7))) ? 1 : 0];
}
if (drawcursor) for (c = c_start; c < 8; c++) {
buffer32->line[line][out_x + c] ^= 15;
}
}
if (drawcursor) for (c = 0; c < 8; c++)
buffer32->line[line][(x << 3) + c + 8] ^= 15;
vid->memaddr++;
}
} else if (!(vid->mode & 2)) {
@@ -474,22 +577,31 @@ vid_render(tandy_t *dev, int line)
cols[0] = vid->array[((attr >> 4) & vid->array[1]) + 16] + 16;
}
vid->memaddr++;
if (vid->scanline & 8) for (c = 0; c < 8; c++)
buffer32->line[line][(x << 4) + (c << 1) + 8] =
buffer32->line[line][(x << 4) + (c << 1) + 1 + 8] =
((chr >= 0xb3) && (chr <= 0xdf)) ? cols[(fontdat[chr][7] & (1 << (c ^ 7))) ? 1 : 0] : cols[0];
else for (c = 0; c < 8; c++) {
if (vid->scanline == 8)
buffer32->line[line][(x << 4) + (c << 1) + 8] =
buffer32->line[line][(x << 4) + (c << 1) + 1 + 8] =
cols[(fontdat[chr][7] & (1 << (c ^ 7))) ? 1 : 0];
else
buffer32->line[line][(x << 4) + (c << 1) + 8] =
buffer32->line[line][(x << 4) + (c << 1) + 1 + 8] =
cols[(fontdat[chr][vid->scanline & 7] & (1 << (c ^ 7))) ? 1 : 0];
out_x = (x << 4) + hos_offs;
if (out_x > -16) {
c_start = out_x < 0 ? -out_x >> 1 : 0;
if (vid->scanline & 8) for (c = c_start; c < 8; c++) {
buffer32->line[line][out_x + (c << 1)] =
buffer32->line[line][out_x + 1 + (c << 1)] =
((chr >= 0xb3) && (chr <= 0xdf)) ? cols[(fontdat[chr][7] & (1 << (c ^ 7))) ? 1 : 0] : cols[0];
} else for (c = c_start; c < 8; c++) {
if (vid->scanline == 8) {
buffer32->line[line][out_x + (c << 1)] =
buffer32->line[line][out_x + 1 + (c << 1)] =
cols[(fontdat[chr][7] & (1 << (c ^ 7))) ? 1 : 0];
} else {
buffer32->line[line][out_x + (c << 1)] =
buffer32->line[line][out_x + 1 + (c << 1)] =
cols[(fontdat[chr][vid->scanline & 7] & (1 << (c ^ 7))) ? 1 : 0];
}
}
if (drawcursor) {
c_start = out_x < 0 ? -out_x : 0;
for (c = c_start; c < 16; c++) {
buffer32->line[line][out_x + c] ^= 15;
}
}
}
if (drawcursor) for (c = 0; c < 16; c++)
buffer32->line[line][(x << 4) + c + 8] ^= 15;
}
} else if (!(vid->mode & 16)) {
cols[0] = (vid->col & 15);
@@ -515,10 +627,15 @@ vid_render(tandy_t *dev, int line)
dat = (vid->vram[((vid->memaddr << 1) & 0x1fff) + ((vid->scanline & 1) * 0x2000)] << 8) |
vid->vram[((vid->memaddr << 1) & 0x1fff) + ((vid->scanline & 1) * 0x2000) + 1];
vid->memaddr++;
for (c = 0; c < 8; c++) {
buffer32->line[line][(x << 4) + (c << 1) + 8] =
buffer32->line[line][(x << 4) + (c << 1) + 1 + 8] = cols[dat >> 14];
dat <<= 2;
out_x = (x << 4) + hos_offs;
if (out_x > -16) {
c_start = out_x < 0 ? -out_x >> 1 : 0;
dat <<= (c_start > 0 ? c_start << 1 : 0);
for (c = c_start; c < 8; c++) {
buffer32->line[line][out_x + (c << 1)] =
buffer32->line[line][out_x + 1 + (c << 1)] = cols[dat >> 14];
dat <<= 2;
}
}
}
} else {
@@ -528,9 +645,14 @@ vid_render(tandy_t *dev, int line)
dat = (vid->vram[((vid->memaddr << 1) & 0x1fff) + ((vid->scanline & 1) * 0x2000)] << 8) |
vid->vram[((vid->memaddr << 1) & 0x1fff) + ((vid->scanline & 1) * 0x2000) + 1];
vid->memaddr++;
for (c = 0; c < 16; c++) {
buffer32->line[line][(x << 4) + c + 8] = buffer32->line[(line) + 1][(x << 4) + c + 8] = cols[dat >> 15];
dat <<= 1;
out_x = (x << 4) + hos_offs;
if (out_x > -16) {
c_start = out_x < 0 ? -out_x : 0;
dat <<= (c_start > 0 ? c_start : 0);
for (c = c_start; c < 16; c++) {
buffer32->line[line][out_x + c] = buffer32->line[(line) + 1][out_x + c] = cols[dat >> 15];
dat <<= 1;
}
}
}
}
@@ -541,6 +663,9 @@ vid_render_blank(tandy_t *dev, int line)
{
t1kvid_t *vid = dev->vid;
if (line < 0)
return;
if (vid->array[3] & 4) {
if (vid->mode & 1)
hline(buffer32, 0, line, (vid->crtc[1] << 3) + 16, (vid->array[2] & 0xf) + 16);
@@ -562,6 +687,9 @@ vid_render_process(tandy_t *dev, int line)
t1kvid_t *vid = dev->vid;
int x;
if (line < 0)
return;
if (vid->mode & 1)
x = (vid->crtc[1] << 3) + 16;
else
@@ -584,6 +712,9 @@ vid_poll(void *priv)
int oldvc;
int scanline_old;
int old_ma;
int hos_offs = 8 + vid->hsync_offset;
int displine_offs = vid->displine + vid->vsync_offset;
int displine_offs_double = (vid->displine << 1) + (vid->vsync_offset << 1);
if (!vid->linepos) {
timer_advance_u64(&vid->timer, vid->dispofftime);
@@ -600,39 +731,39 @@ vid_poll(void *priv)
vid->lastline = vid->displine;
switch (vid->double_type) {
default:
vid_render(dev, vid->displine << 1);
vid_render_blank(dev, (vid->displine << 1) + 1);
vid_render(dev, displine_offs_double, hos_offs);
vid_render_blank(dev, displine_offs_double + 1);
break;
case DOUBLE_NONE:
vid_render(dev, vid->displine);
vid_render(dev, displine_offs, hos_offs);
break;
case DOUBLE_SIMPLE:
old_ma = vid->memaddr;
vid_render(dev, vid->displine << 1);
vid_render(dev, displine_offs_double, hos_offs);
vid->memaddr = old_ma;
vid_render(dev, (vid->displine << 1) + 1);
vid_render(dev, displine_offs_double + 1, hos_offs);
break;
}
} else switch (vid->double_type) {
default:
vid_render_blank(dev, vid->displine << 1);
vid_render_blank(dev, displine_offs_double);
break;
case DOUBLE_NONE:
vid_render_blank(dev, vid->displine);
vid_render_blank(dev, displine_offs);
break;
case DOUBLE_SIMPLE:
vid_render_blank(dev, vid->displine << 1);
vid_render_blank(dev, (vid->displine << 1) + 1);
vid_render_blank(dev, displine_offs_double);
vid_render_blank(dev, displine_offs_double + 1);
break;
}
switch (vid->double_type) {
default:
vid_render_process(dev, vid->displine << 1);
vid_render_process(dev, (vid->displine << 1) + 1);
vid_render_process(dev, displine_offs_double);
vid_render_process(dev, displine_offs_double + 1);
break;
case DOUBLE_NONE:
vid_render_process(dev, vid->displine);
vid_render_process(dev, displine_offs);
break;
}
@@ -696,6 +827,7 @@ vid_poll(void *priv)
vid->cursoron = vid->blink & 16;
}
if (vid->vc == vid->crtc[7]) {
vid->vsync_offset = vid->vsync_offset_pending;
vid->dispon = 0;
vid->displine = 0;
vid->vsynctime = 16;
@@ -784,6 +916,8 @@ tandy_vid_close(void *priv)
{
tandy_t *dev = (tandy_t *) priv;
timer_on_auto(&dev->vid->calib_timer, 0.0);
free(dev->vid);
dev->vid = NULL;
}
@@ -795,6 +929,12 @@ tandy_vid_init(tandy_t *dev)
t1kvid_t *vid;
vid = calloc(1, sizeof(t1kvid_t));
vid->baseline_hsyncpos = 0;
vid->baseline_vsyncpos = 0;
vid->baseline_ready = false;
vid->hsync_offset = 0;
vid->vsync_offset = 0;
vid->vsync_offset_pending = 0;
vid->memctrl = -1;
video_inform(VIDEO_FLAG_TYPE_CGA, &timing_dram);
@@ -814,6 +954,7 @@ tandy_vid_init(tandy_t *dev)
vid->b8000_mask = 0x3fff;
timer_add(&vid->timer, vid_poll, dev, 1);
timer_add(&vid->calib_timer, baseline_calib_finish, dev, 0);
mem_mapping_add(&vid->mapping, 0xb8000, 0x08000,
vid_read, NULL, NULL, vid_write, NULL, NULL, NULL, 0, dev);
io_sethandler(0x03d0, 16,