ShuriZma
/
BizHawk
mirror of https://github.com/TASEmulators/BizHawk.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

start clean up of APU and fix Saint Seiya

This commit is contained in:
alyosha-tas 2016-07-03 16:18:25 -04:00 committed by GitHub
parent c7e511bb2b
commit 99fde0c4ba
2 changed files with 20 additions and 24 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

31
BizHawk.Emulation.Cores/Consoles/Nintendo/NES/APU.cs
View File

@ -740,7 +740,8 @@ namespace BizHawk.Emulation.Cores.Nintendo.NES
//dmc delay per visual 2a03
int delay;
int timer;
// this timer never stops, ever, so it is convenient to use for even/odd timing used elsewhere
public int timer;
int user_address;
public uint user_length, sample_length;
int sample_address, sample_buffer;
@ -995,7 +996,6 @@ namespace BizHawk.Emulation.Cores.Nintendo.NES
ser.Sync("sequencer_irq_assert", ref sequencer_irq_assert);
ser.Sync("dmc_dma_countdown", ref dmc_dma_countdown);
ser.Sync("toggle", ref toggle);
ser.Sync("sample_length_delay", ref pending_length_change);
ser.Sync("dmc_called_from_write", ref call_from_write);
ser.Sync("sequencer_tick_delay", ref seq_tick);
@ -1048,8 +1048,6 @@ namespace BizHawk.Emulation.Cores.Nintendo.NES
}
//these figures are not valid for PAL. they must be recalculated with nintendulator's values above
//these values (the NTSC at least) are derived from nintendulator. they are all 2 higher than the specifications, due to some shortcoming in the emulation
//this is probably a hint that we're doing something a little wrong but making up for it with curcuitous chaos in other ways
static int[][] sequencer_lut = new int[][]{
new int[]{7457,14913,22371,29830},
new int[]{7457,14913,22371,29830,37282}
@ -1231,13 +1229,13 @@ namespace BizHawk.Emulation.Cores.Nintendo.NES
}
else if (addr == 0x4017)
{
if (toggle==0)
if (dmc.timer%2==0)
{
seq_tick = 4;
seq_tick = 3;
} else
{
seq_tick = 3;
seq_tick = 4;
}
seq_val = val;
@ -1291,7 +1289,6 @@ namespace BizHawk.Emulation.Cores.Nintendo.NES
public int DebugCallbackDivider;
public int DebugCallbackTimer;
int toggle = 0;
int pending_length_change;
@ -1340,12 +1337,12 @@ namespace BizHawk.Emulation.Cores.Nintendo.NES
//the current code simply matches known behaviour
if (pending_reg != -1)
{
if (pending_reg == 0x4015 || pending_reg == 0x4017 || pending_reg==0x4003)
if (pending_reg == 0x4015 || pending_reg == 0x4017 || pending_reg==0x4003 || pending_reg==0x4007)
{
_WriteReg(pending_reg, pending_val);
pending_reg = -1;
}
else if (toggle == 1)
else if (dmc.timer%2==0)
{
_WriteReg(pending_reg, pending_val);
pending_reg = -1;
@ -1356,15 +1353,6 @@ namespace BizHawk.Emulation.Cores.Nintendo.NES
sequencer_tick();
sequencer_write_tick(seq_val);
if (toggle==0)
{
toggle = 1;
} else
{
toggle = 0;
}
if (sequencer_irq_assert>0) {
sequencer_irq_assert--;
@ -1372,13 +1360,10 @@ namespace BizHawk.Emulation.Cores.Nintendo.NES
{
sequencer_irq = true;
}
}
}
SyncIRQ();
nes.irq_apu = irq_pending;
//since the units run concurrently, the APU frame sequencer is ran last because
//it can change the ouput values of the pulse/triangle channels

13
BizHawk.Emulation.Cores/Consoles/Nintendo/NES/PPU.regs.cs
View File

@ -501,7 +501,15 @@ namespace BizHawk.Emulation.Cores.Nintendo.NES
{
//does this take 4x longer? nestopia indicates so perhaps...
int addr = ppur.get_2007access() & 0x3FFF;
int addr = ppur.get_2007access();
if (ppuphase == PPUPHASE.BG)
{
if (reg_2001.show_bg)
{
addr = ppur.get_ntread();
}
}
if ((addr & 0x3F00) == 0x3F00)
{
//handle palette. this is being done nestopia style, because i found some documentation for it (appendix 1)
@ -519,7 +527,10 @@ namespace BizHawk.Emulation.Cores.Nintendo.NES
}
else
{
addr &= 0x3FFF;
ppubus_write(addr, value);
}
ppur.increment2007(ppur.status.rendering && reg_2001.PPUON, reg_2000.vram_incr32 != 0);