1
0
mirror of https://github.com/TomHarte/CLK.git synced 2024-11-21 21:33:54 +00:00

Modifies interface, starts on scheduler.

Probably corrects the pixel clock, which I think was scaled up by a factor of 4.
This commit is contained in:
Thomas Harte 2021-07-27 16:41:18 -04:00
parent d8f814f1c4
commit 3544746934
3 changed files with 91 additions and 30 deletions

View File

@ -79,7 +79,7 @@ class ConcreteMachine:
cia_b_.run_for(e_clocks);
}
const auto changes = chipset_.run_for(cycle.length, false);
const auto changes = chipset_.run_for(cycle.length);
cia_a_.advance_tod(changes.vsyncs);
cia_b_.advance_tod(changes.hsyncs);
mc68000_.set_interrupt_level(changes.interrupt_level);

View File

@ -42,20 +42,75 @@ Chipset::Chipset(uint16_t *ram, size_t size) :
crt_(908, 4, Outputs::Display::Type::PAL50, Outputs::Display::InputDataType::Red4Green4Blue4) {
}
Chipset::Changes Chipset::run_for(HalfCycles length, bool) {
Chipset::Changes Chipset::run_for(HalfCycles length) {
return run<false>(length);
}
Chipset::Changes Chipset::run_until_cpu_slot() {
return run<true>();
}
template <int cycle, bool stop_if_cpu> bool Chipset::perform_cycle() {
// TODO: actual CPU scheduling.
if constexpr (stop_if_cpu) {
return true;
}
return false;
}
template <bool stop_on_cpu> Chipset::Changes Chipset::run(HalfCycles length) {
Changes changes;
// This code uses 'pixels' as a measure, which is equivalent to one pixel clock time,
// or half a cycle.
auto pixels_remaining = length.as<int>();
// TODO: all DMA scheduling and actions, possibly stopping upon discovery of a CPU slot.
// Video memory collected here is spirited away for unspooling in the next loop.
// Update raster position, spooling out graphics.
while(pixels_remaining) {
// Determine number of pixels left on this line.
int line_pixels = std::min(pixels_remaining, line_length_ - x_);
int line_pixels = std::min(pixels_remaining, line_length_ - line_cycle_);
pixels_remaining -= line_pixels;
//
// Run DMA scheduler.
//
int final_slot = (line_cycle_ + line_pixels) >> 2;
if((line_cycle_ >> 2) == final_slot) {
// Not enough pixels left to fill any whole slots, just stop.
break;
}
#define C(x) \
case x: \
if constexpr(stop_on_cpu) { if(perform_cycle<x, stop_on_cpu>()) break; } else { perform_cycle<x, stop_on_cpu>(); } \
line_cycle_ += 4; \
if((line_cycle_ >> 2) == final_slot) break;
#define C10(x) C(x); C(x+1); C(x+2); C(x+3); C(x+4); C(x+5); C(x+6); C(x+7); C(x+8); C(x+9);
switch(line_cycle_ >> 2) {
C10(0); C10(10); C10(20); C10(30); C10(40);
C10(50); C10(60); C10(70); C10(80); C10(90);
C10(100); C10(110); C10(120); C10(130); C10(140);
C10(150); C10(160); C10(170); C10(180); C10(190);
C10(200); C10(210);
C(220); C(221); C(222); C(223); C(224);
C(225); C(226); C(227); C(228);
default: assert(false);
}
#undef C
// Update per the possibility that the above ended early.
final_slot = line_cycle_ >> 2;
int slot = line_cycle_ >> 2;
//
// Output video signal as implied by whatever happened above.
//
// Hardware stop is at 0x18;
// 12/64 * 227 = 42.5625
//
@ -79,11 +134,10 @@ Chipset::Changes Chipset::run_for(HalfCycles length, bool) {
static_assert(blank3 == 43);
#define LINK(location, action, length) \
if(x_ < (location) && final_x >= (location)) { \
if(slot < (location) && final_slot >= (location)) { \
crt_.action((length) * 4); \
}
const int final_x = x_ + line_pixels;
if(y_ < vertical_blank_height_) {
// Put three lines of sync at the centre of the vertical blank period.
// TODO: offset by half a line if interlaced and on an odd frame.
@ -124,34 +178,23 @@ Chipset::Changes Chipset::run_for(HalfCycles length, bool) {
LINK(burst, output_default_colour_burst, burst - blank2); // TODO: only if colour enabled.
LINK(blank3, output_blank, blank3 - burst);
// if(final_x > blank3) {
// const int start_x = std::max(blank3, x_);
//
// uint16_t *const pixels = reinterpret_cast<uint16_t *>(crt_.begin_data(2));
// if(pixels) {
// *pixels = palette_[0];
// }
// crt_.output_data((final_x - start_x) * 4, 1);
// }
// Output colour 0 to fill the rest of the line; Kickstart uses this
// colour to post the error code. TODO: actual pixels, etc.
if(final_x == line_length_) {
if(line_cycle_ >= line_length_) {
uint16_t *const pixels = reinterpret_cast<uint16_t *>(crt_.begin_data(1));
if(pixels) {
*pixels = palette_[0];
}
crt_.output_data((final_x - blank3) * 4, 1);
crt_.output_data(line_length_ - blank3 * 4, 1);
}
}
// Advance intraline counter and possibly ripple upwards into
// lines and fields.
x_ += line_pixels;
if(x_ == line_length_) {
if(line_cycle_ >= line_length_) {
++changes.hsyncs;
x_ = 0;
line_cycle_ = 0;
++y_;
if(y_ == frame_height_) {
@ -160,6 +203,9 @@ Chipset::Changes Chipset::run_for(HalfCycles length, bool) {
update_interrupts();
y_ = 0;
// TODO: the manual is vague on when this happens. Try to find out.
copper_address_ = copper_addresses_[0];
}
}
}
@ -221,7 +267,7 @@ void Chipset::perform(const CPU::MC68000::Microcycle &cycle) {
cycle.set_value16(position);
} break;
case Read(0x006): {
const uint16_t position = uint16_t((x_ << 8) | (y_ & 0xff));
const uint16_t position = uint16_t(((line_cycle_ << 6) & 0xff00) | (y_ & 0x00ff));
LOG("Read position low " << PADHEX(4) << position);
cycle.set_value16(position);
} break;
@ -395,21 +441,27 @@ void Chipset::perform(const CPU::MC68000::Microcycle &cycle) {
break;
case Write(0x080):
LOG("TODO: coprocessor first location register high " << PADHEX(4) << cycle.value16());
copper_addresses_[0] = (copper_addresses_[0] & 0x0000'ffff) | uint32_t(cycle.value16() << 16);
break;
case Write(0x082):
LOG("TODO: coprocessor first location register low " << PADHEX(4) << cycle.value16());
copper_addresses_[0] = (copper_addresses_[0] & 0xffff'0000) | uint32_t(cycle.value16() << 0);
break;
case Write(0x084):
LOG("TODO: coprocessor second location register high " << PADHEX(4) << cycle.value16());
copper_addresses_[1] = (copper_addresses_[1] & 0x0000'ffff) | uint32_t(cycle.value16() << 16);
break;
case Write(0x086):
LOG("TODO: coprocessor second location register low " << PADHEX(4) << cycle.value16());
copper_addresses_[1] = (copper_addresses_[1] & 0xffff'0000) | uint32_t(cycle.value16() << 0);
break;
case Write(0x088): case Read(0x088):
LOG("TODO: coprocessor restart at first location");
copper_address_ = copper_addresses_[0];
break;
case Write(0x08a): case Read(0x08a):
LOG("TODO: coprocessor restart at second location");
copper_address_ = copper_addresses_[1];
break;
case Write(0x08c):
LOG("TODO: coprocessor instruction fetch identity " << PADHEX(4) << cycle.value16());

View File

@ -34,8 +34,11 @@ class Chipset {
HalfCycles duration;
};
/// Advances the stated amount of time, possibly stopping if a CPU slot is found.
Changes run_for(HalfCycles, bool stop_on_cpu_slot);
/// Advances the stated amount of time.
Changes run_for(HalfCycles);
/// Advances to the next available CPU slot.
Changes run_until_cpu_slot();
/// Performs the provided microcycle, which the caller guarantees to be a memory access.
void perform(const CPU::MC68000::Microcycle &);
@ -61,7 +64,12 @@ class Chipset {
void update_interrupts();
// MARK: - DMA Control and Blitter.
// MARK: - Scheduler.
template <bool stop_on_cpu> Changes run(HalfCycles duration = HalfCycles());
template <int cycle, bool stop_if_cpu> bool perform_cycle();
// MARK: - DMA Control, Scheduler and Blitter.
uint16_t dma_control_ = 0;
Blitter blitter_;
@ -77,8 +85,8 @@ class Chipset {
// MARK: - Raster.
int x_ = 0, y_ = 0;
int line_length_ = 227;
int line_cycle_ = 0, y_ = 0;
int line_length_ = 227 * 4;
int frame_height_ = 312;
int vertical_blank_height_ = 29;
@ -88,7 +96,8 @@ class Chipset {
// MARK: - Copper.
uint16_t copper_address_ = 0;
uint32_t copper_address_ = 0;
uint32_t copper_addresses_[2]{};
// MARK: - Pixel output.