bsnes/sfc/cpu/mmio/mmio.cpp

520 lines
12 KiB
C++
Raw Normal View History

#ifdef CPU_CPP
uint8 CPU::pio() { return status.pio; }
bool CPU::joylatch() { return status.joypad_strobe_latch; }
//WMDATA
uint8 CPU::mmio_r2180() {
Updated to v067r23 release. byuu says: Added missing $4200 IRQ lock, which fixes Chou Aniki on the fast CPU core, so slower PCs can get their brotherly love on. Added range-based controller IOBit latching to the fast CPU core, which enables Super Scope and Justifier support. Uses the priority queue as well, so there is zero speed-hit. Given the way range-testing works, the trigger point may vary by 1-2 pixels when firing at the same spot. Not really a big deal when it avoids a massive speed penalty. Fixed PAL and interlace-mode HVIRQs at V=0,H<2 on the fast CPU core. Added the dot-renderer's sprite list update-on-OAM-write functionality to the scanline-based PPU renderer. Unfortunately it looks like all the speed gain was already taken from the global dirty flag I was using before, but this certainly won't hurt speed any, so whatever. Added #ifdef to stop CoInitialize(0) on non-Windows ports. Added #ifdefs to stop gradient fade on Windows port. Not going to fuck over the Linux port aesthetic because of Qt bug #47,326,927. If there's a way to tell what Qt theme is being used, I can leave it enabled for XP/Vista themes. Moved HDMA trigger from 1104 to 1112, and reduced channel overhead from 24 to 16, to better simulate one-cycle DMA->CPU sync. Code clarity: I've re-added my varint.hpp classes, and am actively using them in the accuracy cores. So far, I haven't done anything that would detriment speed, but it is certainly cool. The APU ports exposed by the CPU and SMP now take uint2 address arguments, the CPU WRAM address register is a uint17, and the IRQ H/VTIME values are uint10. This basically allows the source to clearly convey the data sizes, and eliminates the need to manually mask values when writing to registers or reading from memory. I'm going to be doing this everywhere, and it will have a speed impact eventually, because the automation means we can't skip masks when we know the data is already masked off. Source: archive contains the launcher code, so that I can look into why it's crashing on XP tomorrow. It doesn't look like Circuit USA's flags are going to work too well with this new CPU core. Still not sure what the hell Robocop vs The Terminator is doing, I'll read through the mega SNES thread for clues tomorrow. Speedy Gonzales is definitely broken, as modifying the MDR was breaking things with my current core. Probably because the new CPU core doesn't wait for a cycle edge to trigger. I was thinking that perhaps we could keep some form of cheat codes list to work as game-specific hacks for the performance core. Keeps the hacks out of the emulator, but could allow the remaining bugs to be worked around for people who have no choice but to use the performance core.
2010-08-16 09:42:20 +00:00
return bus.read(0x7e0000 | status.wram_addr++);
}
//WMDATA
void CPU::mmio_w2180(uint8 data) {
Updated to v067r23 release. byuu says: Added missing $4200 IRQ lock, which fixes Chou Aniki on the fast CPU core, so slower PCs can get their brotherly love on. Added range-based controller IOBit latching to the fast CPU core, which enables Super Scope and Justifier support. Uses the priority queue as well, so there is zero speed-hit. Given the way range-testing works, the trigger point may vary by 1-2 pixels when firing at the same spot. Not really a big deal when it avoids a massive speed penalty. Fixed PAL and interlace-mode HVIRQs at V=0,H<2 on the fast CPU core. Added the dot-renderer's sprite list update-on-OAM-write functionality to the scanline-based PPU renderer. Unfortunately it looks like all the speed gain was already taken from the global dirty flag I was using before, but this certainly won't hurt speed any, so whatever. Added #ifdef to stop CoInitialize(0) on non-Windows ports. Added #ifdefs to stop gradient fade on Windows port. Not going to fuck over the Linux port aesthetic because of Qt bug #47,326,927. If there's a way to tell what Qt theme is being used, I can leave it enabled for XP/Vista themes. Moved HDMA trigger from 1104 to 1112, and reduced channel overhead from 24 to 16, to better simulate one-cycle DMA->CPU sync. Code clarity: I've re-added my varint.hpp classes, and am actively using them in the accuracy cores. So far, I haven't done anything that would detriment speed, but it is certainly cool. The APU ports exposed by the CPU and SMP now take uint2 address arguments, the CPU WRAM address register is a uint17, and the IRQ H/VTIME values are uint10. This basically allows the source to clearly convey the data sizes, and eliminates the need to manually mask values when writing to registers or reading from memory. I'm going to be doing this everywhere, and it will have a speed impact eventually, because the automation means we can't skip masks when we know the data is already masked off. Source: archive contains the launcher code, so that I can look into why it's crashing on XP tomorrow. It doesn't look like Circuit USA's flags are going to work too well with this new CPU core. Still not sure what the hell Robocop vs The Terminator is doing, I'll read through the mega SNES thread for clues tomorrow. Speedy Gonzales is definitely broken, as modifying the MDR was breaking things with my current core. Probably because the new CPU core doesn't wait for a cycle edge to trigger. I was thinking that perhaps we could keep some form of cheat codes list to work as game-specific hacks for the performance core. Keeps the hacks out of the emulator, but could allow the remaining bugs to be worked around for people who have no choice but to use the performance core.
2010-08-16 09:42:20 +00:00
bus.write(0x7e0000 | status.wram_addr++, data);
}
//WMADDL
void CPU::mmio_w2181(uint8 data) {
Updated to v067r23 release. byuu says: Added missing $4200 IRQ lock, which fixes Chou Aniki on the fast CPU core, so slower PCs can get their brotherly love on. Added range-based controller IOBit latching to the fast CPU core, which enables Super Scope and Justifier support. Uses the priority queue as well, so there is zero speed-hit. Given the way range-testing works, the trigger point may vary by 1-2 pixels when firing at the same spot. Not really a big deal when it avoids a massive speed penalty. Fixed PAL and interlace-mode HVIRQs at V=0,H<2 on the fast CPU core. Added the dot-renderer's sprite list update-on-OAM-write functionality to the scanline-based PPU renderer. Unfortunately it looks like all the speed gain was already taken from the global dirty flag I was using before, but this certainly won't hurt speed any, so whatever. Added #ifdef to stop CoInitialize(0) on non-Windows ports. Added #ifdefs to stop gradient fade on Windows port. Not going to fuck over the Linux port aesthetic because of Qt bug #47,326,927. If there's a way to tell what Qt theme is being used, I can leave it enabled for XP/Vista themes. Moved HDMA trigger from 1104 to 1112, and reduced channel overhead from 24 to 16, to better simulate one-cycle DMA->CPU sync. Code clarity: I've re-added my varint.hpp classes, and am actively using them in the accuracy cores. So far, I haven't done anything that would detriment speed, but it is certainly cool. The APU ports exposed by the CPU and SMP now take uint2 address arguments, the CPU WRAM address register is a uint17, and the IRQ H/VTIME values are uint10. This basically allows the source to clearly convey the data sizes, and eliminates the need to manually mask values when writing to registers or reading from memory. I'm going to be doing this everywhere, and it will have a speed impact eventually, because the automation means we can't skip masks when we know the data is already masked off. Source: archive contains the launcher code, so that I can look into why it's crashing on XP tomorrow. It doesn't look like Circuit USA's flags are going to work too well with this new CPU core. Still not sure what the hell Robocop vs The Terminator is doing, I'll read through the mega SNES thread for clues tomorrow. Speedy Gonzales is definitely broken, as modifying the MDR was breaking things with my current core. Probably because the new CPU core doesn't wait for a cycle edge to trigger. I was thinking that perhaps we could keep some form of cheat codes list to work as game-specific hacks for the performance core. Keeps the hacks out of the emulator, but could allow the remaining bugs to be worked around for people who have no choice but to use the performance core.
2010-08-16 09:42:20 +00:00
status.wram_addr = (status.wram_addr & 0x01ff00) | (data << 0);
}
//WMADDM
void CPU::mmio_w2182(uint8 data) {
Updated to v067r23 release. byuu says: Added missing $4200 IRQ lock, which fixes Chou Aniki on the fast CPU core, so slower PCs can get their brotherly love on. Added range-based controller IOBit latching to the fast CPU core, which enables Super Scope and Justifier support. Uses the priority queue as well, so there is zero speed-hit. Given the way range-testing works, the trigger point may vary by 1-2 pixels when firing at the same spot. Not really a big deal when it avoids a massive speed penalty. Fixed PAL and interlace-mode HVIRQs at V=0,H<2 on the fast CPU core. Added the dot-renderer's sprite list update-on-OAM-write functionality to the scanline-based PPU renderer. Unfortunately it looks like all the speed gain was already taken from the global dirty flag I was using before, but this certainly won't hurt speed any, so whatever. Added #ifdef to stop CoInitialize(0) on non-Windows ports. Added #ifdefs to stop gradient fade on Windows port. Not going to fuck over the Linux port aesthetic because of Qt bug #47,326,927. If there's a way to tell what Qt theme is being used, I can leave it enabled for XP/Vista themes. Moved HDMA trigger from 1104 to 1112, and reduced channel overhead from 24 to 16, to better simulate one-cycle DMA->CPU sync. Code clarity: I've re-added my varint.hpp classes, and am actively using them in the accuracy cores. So far, I haven't done anything that would detriment speed, but it is certainly cool. The APU ports exposed by the CPU and SMP now take uint2 address arguments, the CPU WRAM address register is a uint17, and the IRQ H/VTIME values are uint10. This basically allows the source to clearly convey the data sizes, and eliminates the need to manually mask values when writing to registers or reading from memory. I'm going to be doing this everywhere, and it will have a speed impact eventually, because the automation means we can't skip masks when we know the data is already masked off. Source: archive contains the launcher code, so that I can look into why it's crashing on XP tomorrow. It doesn't look like Circuit USA's flags are going to work too well with this new CPU core. Still not sure what the hell Robocop vs The Terminator is doing, I'll read through the mega SNES thread for clues tomorrow. Speedy Gonzales is definitely broken, as modifying the MDR was breaking things with my current core. Probably because the new CPU core doesn't wait for a cycle edge to trigger. I was thinking that perhaps we could keep some form of cheat codes list to work as game-specific hacks for the performance core. Keeps the hacks out of the emulator, but could allow the remaining bugs to be worked around for people who have no choice but to use the performance core.
2010-08-16 09:42:20 +00:00
status.wram_addr = (status.wram_addr & 0x0100ff) | (data << 8);
}
//WMADDH
void CPU::mmio_w2183(uint8 data) {
Updated to v067r23 release. byuu says: Added missing $4200 IRQ lock, which fixes Chou Aniki on the fast CPU core, so slower PCs can get their brotherly love on. Added range-based controller IOBit latching to the fast CPU core, which enables Super Scope and Justifier support. Uses the priority queue as well, so there is zero speed-hit. Given the way range-testing works, the trigger point may vary by 1-2 pixels when firing at the same spot. Not really a big deal when it avoids a massive speed penalty. Fixed PAL and interlace-mode HVIRQs at V=0,H<2 on the fast CPU core. Added the dot-renderer's sprite list update-on-OAM-write functionality to the scanline-based PPU renderer. Unfortunately it looks like all the speed gain was already taken from the global dirty flag I was using before, but this certainly won't hurt speed any, so whatever. Added #ifdef to stop CoInitialize(0) on non-Windows ports. Added #ifdefs to stop gradient fade on Windows port. Not going to fuck over the Linux port aesthetic because of Qt bug #47,326,927. If there's a way to tell what Qt theme is being used, I can leave it enabled for XP/Vista themes. Moved HDMA trigger from 1104 to 1112, and reduced channel overhead from 24 to 16, to better simulate one-cycle DMA->CPU sync. Code clarity: I've re-added my varint.hpp classes, and am actively using them in the accuracy cores. So far, I haven't done anything that would detriment speed, but it is certainly cool. The APU ports exposed by the CPU and SMP now take uint2 address arguments, the CPU WRAM address register is a uint17, and the IRQ H/VTIME values are uint10. This basically allows the source to clearly convey the data sizes, and eliminates the need to manually mask values when writing to registers or reading from memory. I'm going to be doing this everywhere, and it will have a speed impact eventually, because the automation means we can't skip masks when we know the data is already masked off. Source: archive contains the launcher code, so that I can look into why it's crashing on XP tomorrow. It doesn't look like Circuit USA's flags are going to work too well with this new CPU core. Still not sure what the hell Robocop vs The Terminator is doing, I'll read through the mega SNES thread for clues tomorrow. Speedy Gonzales is definitely broken, as modifying the MDR was breaking things with my current core. Probably because the new CPU core doesn't wait for a cycle edge to trigger. I was thinking that perhaps we could keep some form of cheat codes list to work as game-specific hacks for the performance core. Keeps the hacks out of the emulator, but could allow the remaining bugs to be worked around for people who have no choice but to use the performance core.
2010-08-16 09:42:20 +00:00
status.wram_addr = (status.wram_addr & 0x00ffff) | (data << 16);
}
//JOYSER0
//bit 0 is shared between JOYSER0 and JOYSER1, therefore
//strobing $4016.d0 affects both controller port latches.
//$4017 bit 0 writes are ignored.
void CPU::mmio_w4016(uint8 data) {
device.controllerPort1->latch(data & 1);
device.controllerPort2->latch(data & 1);
}
//JOYSER0
//7-2 = MDR
//1-0 = Joypad serial data
uint8 CPU::mmio_r4016() {
uint8 r = regs.mdr & 0xfc;
r |= device.controllerPort1->data();
return r;
}
//JOYSER1
//7-5 = MDR
//4-2 = Always 1 (pins are connected to GND)
//1-0 = Joypad serial data
uint8 CPU::mmio_r4017() {
uint8 r = (regs.mdr & 0xe0) | 0x1c;
r |= device.controllerPort2->data();
return r;
}
//NMITIMEN
void CPU::mmio_w4200(uint8 data) {
Updated to v067r23 release. byuu says: Added missing $4200 IRQ lock, which fixes Chou Aniki on the fast CPU core, so slower PCs can get their brotherly love on. Added range-based controller IOBit latching to the fast CPU core, which enables Super Scope and Justifier support. Uses the priority queue as well, so there is zero speed-hit. Given the way range-testing works, the trigger point may vary by 1-2 pixels when firing at the same spot. Not really a big deal when it avoids a massive speed penalty. Fixed PAL and interlace-mode HVIRQs at V=0,H<2 on the fast CPU core. Added the dot-renderer's sprite list update-on-OAM-write functionality to the scanline-based PPU renderer. Unfortunately it looks like all the speed gain was already taken from the global dirty flag I was using before, but this certainly won't hurt speed any, so whatever. Added #ifdef to stop CoInitialize(0) on non-Windows ports. Added #ifdefs to stop gradient fade on Windows port. Not going to fuck over the Linux port aesthetic because of Qt bug #47,326,927. If there's a way to tell what Qt theme is being used, I can leave it enabled for XP/Vista themes. Moved HDMA trigger from 1104 to 1112, and reduced channel overhead from 24 to 16, to better simulate one-cycle DMA->CPU sync. Code clarity: I've re-added my varint.hpp classes, and am actively using them in the accuracy cores. So far, I haven't done anything that would detriment speed, but it is certainly cool. The APU ports exposed by the CPU and SMP now take uint2 address arguments, the CPU WRAM address register is a uint17, and the IRQ H/VTIME values are uint10. This basically allows the source to clearly convey the data sizes, and eliminates the need to manually mask values when writing to registers or reading from memory. I'm going to be doing this everywhere, and it will have a speed impact eventually, because the automation means we can't skip masks when we know the data is already masked off. Source: archive contains the launcher code, so that I can look into why it's crashing on XP tomorrow. It doesn't look like Circuit USA's flags are going to work too well with this new CPU core. Still not sure what the hell Robocop vs The Terminator is doing, I'll read through the mega SNES thread for clues tomorrow. Speedy Gonzales is definitely broken, as modifying the MDR was breaking things with my current core. Probably because the new CPU core doesn't wait for a cycle edge to trigger. I was thinking that perhaps we could keep some form of cheat codes list to work as game-specific hacks for the performance core. Keeps the hacks out of the emulator, but could allow the remaining bugs to be worked around for people who have no choice but to use the performance core.
2010-08-16 09:42:20 +00:00
status.auto_joypad_poll = data & 1;
nmitimen_update(data);
}
//WRIO
void CPU::mmio_w4201(uint8 data) {
Updated to v067r23 release. byuu says: Added missing $4200 IRQ lock, which fixes Chou Aniki on the fast CPU core, so slower PCs can get their brotherly love on. Added range-based controller IOBit latching to the fast CPU core, which enables Super Scope and Justifier support. Uses the priority queue as well, so there is zero speed-hit. Given the way range-testing works, the trigger point may vary by 1-2 pixels when firing at the same spot. Not really a big deal when it avoids a massive speed penalty. Fixed PAL and interlace-mode HVIRQs at V=0,H<2 on the fast CPU core. Added the dot-renderer's sprite list update-on-OAM-write functionality to the scanline-based PPU renderer. Unfortunately it looks like all the speed gain was already taken from the global dirty flag I was using before, but this certainly won't hurt speed any, so whatever. Added #ifdef to stop CoInitialize(0) on non-Windows ports. Added #ifdefs to stop gradient fade on Windows port. Not going to fuck over the Linux port aesthetic because of Qt bug #47,326,927. If there's a way to tell what Qt theme is being used, I can leave it enabled for XP/Vista themes. Moved HDMA trigger from 1104 to 1112, and reduced channel overhead from 24 to 16, to better simulate one-cycle DMA->CPU sync. Code clarity: I've re-added my varint.hpp classes, and am actively using them in the accuracy cores. So far, I haven't done anything that would detriment speed, but it is certainly cool. The APU ports exposed by the CPU and SMP now take uint2 address arguments, the CPU WRAM address register is a uint17, and the IRQ H/VTIME values are uint10. This basically allows the source to clearly convey the data sizes, and eliminates the need to manually mask values when writing to registers or reading from memory. I'm going to be doing this everywhere, and it will have a speed impact eventually, because the automation means we can't skip masks when we know the data is already masked off. Source: archive contains the launcher code, so that I can look into why it's crashing on XP tomorrow. It doesn't look like Circuit USA's flags are going to work too well with this new CPU core. Still not sure what the hell Robocop vs The Terminator is doing, I'll read through the mega SNES thread for clues tomorrow. Speedy Gonzales is definitely broken, as modifying the MDR was breaking things with my current core. Probably because the new CPU core doesn't wait for a cycle edge to trigger. I was thinking that perhaps we could keep some form of cheat codes list to work as game-specific hacks for the performance core. Keeps the hacks out of the emulator, but could allow the remaining bugs to be worked around for people who have no choice but to use the performance core.
2010-08-16 09:42:20 +00:00
if((status.pio & 0x80) && !(data & 0x80)) ppu.latch_counters();
status.pio = data;
}
//WRMPYA
void CPU::mmio_w4202(uint8 data) {
status.wrmpya = data;
}
//WRMPYB
void CPU::mmio_w4203(uint8 data) {
status.rdmpy = 0;
if(alu.mpyctr || alu.divctr) return;
status.wrmpyb = data;
status.rddiv = (status.wrmpyb << 8) | status.wrmpya;
alu.mpyctr = 8; //perform multiplication over the next eight cycles
alu.shift = status.wrmpyb;
}
//WRDIVL
void CPU::mmio_w4204(uint8 data) {
Updated to v067r23 release. byuu says: Added missing $4200 IRQ lock, which fixes Chou Aniki on the fast CPU core, so slower PCs can get their brotherly love on. Added range-based controller IOBit latching to the fast CPU core, which enables Super Scope and Justifier support. Uses the priority queue as well, so there is zero speed-hit. Given the way range-testing works, the trigger point may vary by 1-2 pixels when firing at the same spot. Not really a big deal when it avoids a massive speed penalty. Fixed PAL and interlace-mode HVIRQs at V=0,H<2 on the fast CPU core. Added the dot-renderer's sprite list update-on-OAM-write functionality to the scanline-based PPU renderer. Unfortunately it looks like all the speed gain was already taken from the global dirty flag I was using before, but this certainly won't hurt speed any, so whatever. Added #ifdef to stop CoInitialize(0) on non-Windows ports. Added #ifdefs to stop gradient fade on Windows port. Not going to fuck over the Linux port aesthetic because of Qt bug #47,326,927. If there's a way to tell what Qt theme is being used, I can leave it enabled for XP/Vista themes. Moved HDMA trigger from 1104 to 1112, and reduced channel overhead from 24 to 16, to better simulate one-cycle DMA->CPU sync. Code clarity: I've re-added my varint.hpp classes, and am actively using them in the accuracy cores. So far, I haven't done anything that would detriment speed, but it is certainly cool. The APU ports exposed by the CPU and SMP now take uint2 address arguments, the CPU WRAM address register is a uint17, and the IRQ H/VTIME values are uint10. This basically allows the source to clearly convey the data sizes, and eliminates the need to manually mask values when writing to registers or reading from memory. I'm going to be doing this everywhere, and it will have a speed impact eventually, because the automation means we can't skip masks when we know the data is already masked off. Source: archive contains the launcher code, so that I can look into why it's crashing on XP tomorrow. It doesn't look like Circuit USA's flags are going to work too well with this new CPU core. Still not sure what the hell Robocop vs The Terminator is doing, I'll read through the mega SNES thread for clues tomorrow. Speedy Gonzales is definitely broken, as modifying the MDR was breaking things with my current core. Probably because the new CPU core doesn't wait for a cycle edge to trigger. I was thinking that perhaps we could keep some form of cheat codes list to work as game-specific hacks for the performance core. Keeps the hacks out of the emulator, but could allow the remaining bugs to be worked around for people who have no choice but to use the performance core.
2010-08-16 09:42:20 +00:00
status.wrdiva = (status.wrdiva & 0xff00) | (data << 0);
}
//WRDIVH
void CPU::mmio_w4205(uint8 data) {
status.wrdiva = (status.wrdiva & 0x00ff) | (data << 8);
}
//WRDIVB
void CPU::mmio_w4206(uint8 data) {
status.rdmpy = status.wrdiva;
if(alu.mpyctr || alu.divctr) return;
status.wrdivb = data;
alu.divctr = 16; //perform division over the next sixteen cycles
alu.shift = status.wrdivb << 16;
}
//HTIMEL
void CPU::mmio_w4207(uint8 data) {
Updated to v067r23 release. byuu says: Added missing $4200 IRQ lock, which fixes Chou Aniki on the fast CPU core, so slower PCs can get their brotherly love on. Added range-based controller IOBit latching to the fast CPU core, which enables Super Scope and Justifier support. Uses the priority queue as well, so there is zero speed-hit. Given the way range-testing works, the trigger point may vary by 1-2 pixels when firing at the same spot. Not really a big deal when it avoids a massive speed penalty. Fixed PAL and interlace-mode HVIRQs at V=0,H<2 on the fast CPU core. Added the dot-renderer's sprite list update-on-OAM-write functionality to the scanline-based PPU renderer. Unfortunately it looks like all the speed gain was already taken from the global dirty flag I was using before, but this certainly won't hurt speed any, so whatever. Added #ifdef to stop CoInitialize(0) on non-Windows ports. Added #ifdefs to stop gradient fade on Windows port. Not going to fuck over the Linux port aesthetic because of Qt bug #47,326,927. If there's a way to tell what Qt theme is being used, I can leave it enabled for XP/Vista themes. Moved HDMA trigger from 1104 to 1112, and reduced channel overhead from 24 to 16, to better simulate one-cycle DMA->CPU sync. Code clarity: I've re-added my varint.hpp classes, and am actively using them in the accuracy cores. So far, I haven't done anything that would detriment speed, but it is certainly cool. The APU ports exposed by the CPU and SMP now take uint2 address arguments, the CPU WRAM address register is a uint17, and the IRQ H/VTIME values are uint10. This basically allows the source to clearly convey the data sizes, and eliminates the need to manually mask values when writing to registers or reading from memory. I'm going to be doing this everywhere, and it will have a speed impact eventually, because the automation means we can't skip masks when we know the data is already masked off. Source: archive contains the launcher code, so that I can look into why it's crashing on XP tomorrow. It doesn't look like Circuit USA's flags are going to work too well with this new CPU core. Still not sure what the hell Robocop vs The Terminator is doing, I'll read through the mega SNES thread for clues tomorrow. Speedy Gonzales is definitely broken, as modifying the MDR was breaking things with my current core. Probably because the new CPU core doesn't wait for a cycle edge to trigger. I was thinking that perhaps we could keep some form of cheat codes list to work as game-specific hacks for the performance core. Keeps the hacks out of the emulator, but could allow the remaining bugs to be worked around for people who have no choice but to use the performance core.
2010-08-16 09:42:20 +00:00
status.hirq_pos = (status.hirq_pos & 0x0100) | (data << 0);
}
//HTIMEH
void CPU::mmio_w4208(uint8 data) {
Updated to v067r23 release. byuu says: Added missing $4200 IRQ lock, which fixes Chou Aniki on the fast CPU core, so slower PCs can get their brotherly love on. Added range-based controller IOBit latching to the fast CPU core, which enables Super Scope and Justifier support. Uses the priority queue as well, so there is zero speed-hit. Given the way range-testing works, the trigger point may vary by 1-2 pixels when firing at the same spot. Not really a big deal when it avoids a massive speed penalty. Fixed PAL and interlace-mode HVIRQs at V=0,H<2 on the fast CPU core. Added the dot-renderer's sprite list update-on-OAM-write functionality to the scanline-based PPU renderer. Unfortunately it looks like all the speed gain was already taken from the global dirty flag I was using before, but this certainly won't hurt speed any, so whatever. Added #ifdef to stop CoInitialize(0) on non-Windows ports. Added #ifdefs to stop gradient fade on Windows port. Not going to fuck over the Linux port aesthetic because of Qt bug #47,326,927. If there's a way to tell what Qt theme is being used, I can leave it enabled for XP/Vista themes. Moved HDMA trigger from 1104 to 1112, and reduced channel overhead from 24 to 16, to better simulate one-cycle DMA->CPU sync. Code clarity: I've re-added my varint.hpp classes, and am actively using them in the accuracy cores. So far, I haven't done anything that would detriment speed, but it is certainly cool. The APU ports exposed by the CPU and SMP now take uint2 address arguments, the CPU WRAM address register is a uint17, and the IRQ H/VTIME values are uint10. This basically allows the source to clearly convey the data sizes, and eliminates the need to manually mask values when writing to registers or reading from memory. I'm going to be doing this everywhere, and it will have a speed impact eventually, because the automation means we can't skip masks when we know the data is already masked off. Source: archive contains the launcher code, so that I can look into why it's crashing on XP tomorrow. It doesn't look like Circuit USA's flags are going to work too well with this new CPU core. Still not sure what the hell Robocop vs The Terminator is doing, I'll read through the mega SNES thread for clues tomorrow. Speedy Gonzales is definitely broken, as modifying the MDR was breaking things with my current core. Probably because the new CPU core doesn't wait for a cycle edge to trigger. I was thinking that perhaps we could keep some form of cheat codes list to work as game-specific hacks for the performance core. Keeps the hacks out of the emulator, but could allow the remaining bugs to be worked around for people who have no choice but to use the performance core.
2010-08-16 09:42:20 +00:00
status.hirq_pos = (status.hirq_pos & 0x00ff) | (data << 8);
}
//VTIMEL
void CPU::mmio_w4209(uint8 data) {
Updated to v067r23 release. byuu says: Added missing $4200 IRQ lock, which fixes Chou Aniki on the fast CPU core, so slower PCs can get their brotherly love on. Added range-based controller IOBit latching to the fast CPU core, which enables Super Scope and Justifier support. Uses the priority queue as well, so there is zero speed-hit. Given the way range-testing works, the trigger point may vary by 1-2 pixels when firing at the same spot. Not really a big deal when it avoids a massive speed penalty. Fixed PAL and interlace-mode HVIRQs at V=0,H<2 on the fast CPU core. Added the dot-renderer's sprite list update-on-OAM-write functionality to the scanline-based PPU renderer. Unfortunately it looks like all the speed gain was already taken from the global dirty flag I was using before, but this certainly won't hurt speed any, so whatever. Added #ifdef to stop CoInitialize(0) on non-Windows ports. Added #ifdefs to stop gradient fade on Windows port. Not going to fuck over the Linux port aesthetic because of Qt bug #47,326,927. If there's a way to tell what Qt theme is being used, I can leave it enabled for XP/Vista themes. Moved HDMA trigger from 1104 to 1112, and reduced channel overhead from 24 to 16, to better simulate one-cycle DMA->CPU sync. Code clarity: I've re-added my varint.hpp classes, and am actively using them in the accuracy cores. So far, I haven't done anything that would detriment speed, but it is certainly cool. The APU ports exposed by the CPU and SMP now take uint2 address arguments, the CPU WRAM address register is a uint17, and the IRQ H/VTIME values are uint10. This basically allows the source to clearly convey the data sizes, and eliminates the need to manually mask values when writing to registers or reading from memory. I'm going to be doing this everywhere, and it will have a speed impact eventually, because the automation means we can't skip masks when we know the data is already masked off. Source: archive contains the launcher code, so that I can look into why it's crashing on XP tomorrow. It doesn't look like Circuit USA's flags are going to work too well with this new CPU core. Still not sure what the hell Robocop vs The Terminator is doing, I'll read through the mega SNES thread for clues tomorrow. Speedy Gonzales is definitely broken, as modifying the MDR was breaking things with my current core. Probably because the new CPU core doesn't wait for a cycle edge to trigger. I was thinking that perhaps we could keep some form of cheat codes list to work as game-specific hacks for the performance core. Keeps the hacks out of the emulator, but could allow the remaining bugs to be worked around for people who have no choice but to use the performance core.
2010-08-16 09:42:20 +00:00
status.virq_pos = (status.virq_pos & 0x0100) | (data << 0);
}
//VTIMEH
void CPU::mmio_w420a(uint8 data) {
Updated to v067r23 release. byuu says: Added missing $4200 IRQ lock, which fixes Chou Aniki on the fast CPU core, so slower PCs can get their brotherly love on. Added range-based controller IOBit latching to the fast CPU core, which enables Super Scope and Justifier support. Uses the priority queue as well, so there is zero speed-hit. Given the way range-testing works, the trigger point may vary by 1-2 pixels when firing at the same spot. Not really a big deal when it avoids a massive speed penalty. Fixed PAL and interlace-mode HVIRQs at V=0,H<2 on the fast CPU core. Added the dot-renderer's sprite list update-on-OAM-write functionality to the scanline-based PPU renderer. Unfortunately it looks like all the speed gain was already taken from the global dirty flag I was using before, but this certainly won't hurt speed any, so whatever. Added #ifdef to stop CoInitialize(0) on non-Windows ports. Added #ifdefs to stop gradient fade on Windows port. Not going to fuck over the Linux port aesthetic because of Qt bug #47,326,927. If there's a way to tell what Qt theme is being used, I can leave it enabled for XP/Vista themes. Moved HDMA trigger from 1104 to 1112, and reduced channel overhead from 24 to 16, to better simulate one-cycle DMA->CPU sync. Code clarity: I've re-added my varint.hpp classes, and am actively using them in the accuracy cores. So far, I haven't done anything that would detriment speed, but it is certainly cool. The APU ports exposed by the CPU and SMP now take uint2 address arguments, the CPU WRAM address register is a uint17, and the IRQ H/VTIME values are uint10. This basically allows the source to clearly convey the data sizes, and eliminates the need to manually mask values when writing to registers or reading from memory. I'm going to be doing this everywhere, and it will have a speed impact eventually, because the automation means we can't skip masks when we know the data is already masked off. Source: archive contains the launcher code, so that I can look into why it's crashing on XP tomorrow. It doesn't look like Circuit USA's flags are going to work too well with this new CPU core. Still not sure what the hell Robocop vs The Terminator is doing, I'll read through the mega SNES thread for clues tomorrow. Speedy Gonzales is definitely broken, as modifying the MDR was breaking things with my current core. Probably because the new CPU core doesn't wait for a cycle edge to trigger. I was thinking that perhaps we could keep some form of cheat codes list to work as game-specific hacks for the performance core. Keeps the hacks out of the emulator, but could allow the remaining bugs to be worked around for people who have no choice but to use the performance core.
2010-08-16 09:42:20 +00:00
status.virq_pos = (status.virq_pos & 0x00ff) | (data << 8);
}
//DMAEN
void CPU::mmio_w420b(uint8 data) {
for(unsigned i = 0; i < 8; i++) {
channel[i].dma_enabled = data & (1 << i);
}
if(data) status.dma_pending = true;
}
//HDMAEN
void CPU::mmio_w420c(uint8 data) {
for(unsigned i = 0; i < 8; i++) {
channel[i].hdma_enabled = data & (1 << i);
}
}
//MEMSEL
void CPU::mmio_w420d(uint8 data) {
status.rom_speed = (data & 1 ? 6 : 8);
}
//RDNMI
//7 = NMI acknowledge
//6-4 = MDR
//3-0 = CPU (5a22) version
uint8 CPU::mmio_r4210() {
uint8 r = (regs.mdr & 0x70);
r |= (uint8)(rdnmi()) << 7;
r |= (cpu_version & 0x0f);
return r;
}
//TIMEUP
//7 = IRQ acknowledge
//6-0 = MDR
uint8 CPU::mmio_r4211() {
uint8 r = (regs.mdr & 0x7f);
r |= (uint8)(timeup()) << 7;
return r;
}
//HVBJOY
//7 = VBLANK acknowledge
//6 = HBLANK acknowledge
//5-1 = MDR
//0 = JOYPAD acknowledge
uint8 CPU::mmio_r4212() {
uint8 r = (regs.mdr & 0x3e);
Update to v073r01 release. byuu says: While perhaps not perfect, pretty good is better than nothing ... I've added emulation of auto-joypad poll timing. Going off ikari_01's confirmation of what we suspected, that the strobe happens every 256 clocks, I've set up emulation as follows: Upon reset, our clock counter is reset to zero. At the start of each frame, our poll counter is reset to zero. Every 256 clocks, we call the step_auto_joypad_poll() function. If we are at V=225/240+ (based on overscan setting), we check the poll counter. At zero, we poll the actual controller and set the joypad polling flag in $4212.d0 to 1. From zero through fifteen, we read in one bit for each controller and shift it into the register. At sixteen, we turn off the joypad polling flag. The 256-clock divider allows the start point of polling for each frame to fluctuate wildly like real hardware. I count regardless of auto joypad enable, as per $4212.d0's behavior; but only poll when it's actually enabled. I do not consume any actual time from this polling. I honestly don't know if I even should, or if it manages to do it in the background. If it should consume time, then this most likely happens between opcode edges and we'll have to adjust the code a good bit. All commercial games should continue to work fine, but this will likely break some hacks/translations not tested on hardware. Without the timing emulation, reading $4218-421f before V=~228 would basically give you the valid input controller values of the previous frame. Now, like hardware, it should give you a state that is part previous frame, part current frame shifted into it. Button positions won't be reliable and will shift every 256 clocks. I've also removed the Qt GUI, and renamed ui-phoenix to just ui. This removes 400kb of source code (phoenix is a lean 130kb), and drops the archive size from 564KB to 475KB. Combined with the DSP HLE, and we've knocked off ~570KB of source cruft from the entire project. I am looking forward to not having to specify which GUI is included anymore.
2010-12-27 07:29:57 +00:00
if(status.auto_joypad_active) r |= 0x01;
Updated to v067r23 release. byuu says: Added missing $4200 IRQ lock, which fixes Chou Aniki on the fast CPU core, so slower PCs can get their brotherly love on. Added range-based controller IOBit latching to the fast CPU core, which enables Super Scope and Justifier support. Uses the priority queue as well, so there is zero speed-hit. Given the way range-testing works, the trigger point may vary by 1-2 pixels when firing at the same spot. Not really a big deal when it avoids a massive speed penalty. Fixed PAL and interlace-mode HVIRQs at V=0,H<2 on the fast CPU core. Added the dot-renderer's sprite list update-on-OAM-write functionality to the scanline-based PPU renderer. Unfortunately it looks like all the speed gain was already taken from the global dirty flag I was using before, but this certainly won't hurt speed any, so whatever. Added #ifdef to stop CoInitialize(0) on non-Windows ports. Added #ifdefs to stop gradient fade on Windows port. Not going to fuck over the Linux port aesthetic because of Qt bug #47,326,927. If there's a way to tell what Qt theme is being used, I can leave it enabled for XP/Vista themes. Moved HDMA trigger from 1104 to 1112, and reduced channel overhead from 24 to 16, to better simulate one-cycle DMA->CPU sync. Code clarity: I've re-added my varint.hpp classes, and am actively using them in the accuracy cores. So far, I haven't done anything that would detriment speed, but it is certainly cool. The APU ports exposed by the CPU and SMP now take uint2 address arguments, the CPU WRAM address register is a uint17, and the IRQ H/VTIME values are uint10. This basically allows the source to clearly convey the data sizes, and eliminates the need to manually mask values when writing to registers or reading from memory. I'm going to be doing this everywhere, and it will have a speed impact eventually, because the automation means we can't skip masks when we know the data is already masked off. Source: archive contains the launcher code, so that I can look into why it's crashing on XP tomorrow. It doesn't look like Circuit USA's flags are going to work too well with this new CPU core. Still not sure what the hell Robocop vs The Terminator is doing, I'll read through the mega SNES thread for clues tomorrow. Speedy Gonzales is definitely broken, as modifying the MDR was breaking things with my current core. Probably because the new CPU core doesn't wait for a cycle edge to trigger. I was thinking that perhaps we could keep some form of cheat codes list to work as game-specific hacks for the performance core. Keeps the hacks out of the emulator, but could allow the remaining bugs to be worked around for people who have no choice but to use the performance core.
2010-08-16 09:42:20 +00:00
if(hcounter() <= 2 || hcounter() >= 1096) r |= 0x40; //hblank
Update to v073r01 release. byuu says: While perhaps not perfect, pretty good is better than nothing ... I've added emulation of auto-joypad poll timing. Going off ikari_01's confirmation of what we suspected, that the strobe happens every 256 clocks, I've set up emulation as follows: Upon reset, our clock counter is reset to zero. At the start of each frame, our poll counter is reset to zero. Every 256 clocks, we call the step_auto_joypad_poll() function. If we are at V=225/240+ (based on overscan setting), we check the poll counter. At zero, we poll the actual controller and set the joypad polling flag in $4212.d0 to 1. From zero through fifteen, we read in one bit for each controller and shift it into the register. At sixteen, we turn off the joypad polling flag. The 256-clock divider allows the start point of polling for each frame to fluctuate wildly like real hardware. I count regardless of auto joypad enable, as per $4212.d0's behavior; but only poll when it's actually enabled. I do not consume any actual time from this polling. I honestly don't know if I even should, or if it manages to do it in the background. If it should consume time, then this most likely happens between opcode edges and we'll have to adjust the code a good bit. All commercial games should continue to work fine, but this will likely break some hacks/translations not tested on hardware. Without the timing emulation, reading $4218-421f before V=~228 would basically give you the valid input controller values of the previous frame. Now, like hardware, it should give you a state that is part previous frame, part current frame shifted into it. Button positions won't be reliable and will shift every 256 clocks. I've also removed the Qt GUI, and renamed ui-phoenix to just ui. This removes 400kb of source code (phoenix is a lean 130kb), and drops the archive size from 564KB to 475KB. Combined with the DSP HLE, and we've knocked off ~570KB of source cruft from the entire project. I am looking forward to not having to specify which GUI is included anymore.
2010-12-27 07:29:57 +00:00
if(vcounter() >= (ppu.overscan() == false ? 225 : 240)) r |= 0x80; //vblank
return r;
}
//RDIO
uint8 CPU::mmio_r4213() {
return status.pio;
}
//RDDIVL
uint8 CPU::mmio_r4214() {
Updated to v067r23 release. byuu says: Added missing $4200 IRQ lock, which fixes Chou Aniki on the fast CPU core, so slower PCs can get their brotherly love on. Added range-based controller IOBit latching to the fast CPU core, which enables Super Scope and Justifier support. Uses the priority queue as well, so there is zero speed-hit. Given the way range-testing works, the trigger point may vary by 1-2 pixels when firing at the same spot. Not really a big deal when it avoids a massive speed penalty. Fixed PAL and interlace-mode HVIRQs at V=0,H<2 on the fast CPU core. Added the dot-renderer's sprite list update-on-OAM-write functionality to the scanline-based PPU renderer. Unfortunately it looks like all the speed gain was already taken from the global dirty flag I was using before, but this certainly won't hurt speed any, so whatever. Added #ifdef to stop CoInitialize(0) on non-Windows ports. Added #ifdefs to stop gradient fade on Windows port. Not going to fuck over the Linux port aesthetic because of Qt bug #47,326,927. If there's a way to tell what Qt theme is being used, I can leave it enabled for XP/Vista themes. Moved HDMA trigger from 1104 to 1112, and reduced channel overhead from 24 to 16, to better simulate one-cycle DMA->CPU sync. Code clarity: I've re-added my varint.hpp classes, and am actively using them in the accuracy cores. So far, I haven't done anything that would detriment speed, but it is certainly cool. The APU ports exposed by the CPU and SMP now take uint2 address arguments, the CPU WRAM address register is a uint17, and the IRQ H/VTIME values are uint10. This basically allows the source to clearly convey the data sizes, and eliminates the need to manually mask values when writing to registers or reading from memory. I'm going to be doing this everywhere, and it will have a speed impact eventually, because the automation means we can't skip masks when we know the data is already masked off. Source: archive contains the launcher code, so that I can look into why it's crashing on XP tomorrow. It doesn't look like Circuit USA's flags are going to work too well with this new CPU core. Still not sure what the hell Robocop vs The Terminator is doing, I'll read through the mega SNES thread for clues tomorrow. Speedy Gonzales is definitely broken, as modifying the MDR was breaking things with my current core. Probably because the new CPU core doesn't wait for a cycle edge to trigger. I was thinking that perhaps we could keep some form of cheat codes list to work as game-specific hacks for the performance core. Keeps the hacks out of the emulator, but could allow the remaining bugs to be worked around for people who have no choice but to use the performance core.
2010-08-16 09:42:20 +00:00
return status.rddiv >> 0;
}
//RDDIVH
uint8 CPU::mmio_r4215() {
return status.rddiv >> 8;
}
//RDMPYL
uint8 CPU::mmio_r4216() {
Updated to v067r23 release. byuu says: Added missing $4200 IRQ lock, which fixes Chou Aniki on the fast CPU core, so slower PCs can get their brotherly love on. Added range-based controller IOBit latching to the fast CPU core, which enables Super Scope and Justifier support. Uses the priority queue as well, so there is zero speed-hit. Given the way range-testing works, the trigger point may vary by 1-2 pixels when firing at the same spot. Not really a big deal when it avoids a massive speed penalty. Fixed PAL and interlace-mode HVIRQs at V=0,H<2 on the fast CPU core. Added the dot-renderer's sprite list update-on-OAM-write functionality to the scanline-based PPU renderer. Unfortunately it looks like all the speed gain was already taken from the global dirty flag I was using before, but this certainly won't hurt speed any, so whatever. Added #ifdef to stop CoInitialize(0) on non-Windows ports. Added #ifdefs to stop gradient fade on Windows port. Not going to fuck over the Linux port aesthetic because of Qt bug #47,326,927. If there's a way to tell what Qt theme is being used, I can leave it enabled for XP/Vista themes. Moved HDMA trigger from 1104 to 1112, and reduced channel overhead from 24 to 16, to better simulate one-cycle DMA->CPU sync. Code clarity: I've re-added my varint.hpp classes, and am actively using them in the accuracy cores. So far, I haven't done anything that would detriment speed, but it is certainly cool. The APU ports exposed by the CPU and SMP now take uint2 address arguments, the CPU WRAM address register is a uint17, and the IRQ H/VTIME values are uint10. This basically allows the source to clearly convey the data sizes, and eliminates the need to manually mask values when writing to registers or reading from memory. I'm going to be doing this everywhere, and it will have a speed impact eventually, because the automation means we can't skip masks when we know the data is already masked off. Source: archive contains the launcher code, so that I can look into why it's crashing on XP tomorrow. It doesn't look like Circuit USA's flags are going to work too well with this new CPU core. Still not sure what the hell Robocop vs The Terminator is doing, I'll read through the mega SNES thread for clues tomorrow. Speedy Gonzales is definitely broken, as modifying the MDR was breaking things with my current core. Probably because the new CPU core doesn't wait for a cycle edge to trigger. I was thinking that perhaps we could keep some form of cheat codes list to work as game-specific hacks for the performance core. Keeps the hacks out of the emulator, but could allow the remaining bugs to be worked around for people who have no choice but to use the performance core.
2010-08-16 09:42:20 +00:00
return status.rdmpy >> 0;
}
//RDMPYH
uint8 CPU::mmio_r4217() {
return status.rdmpy >> 8;
}
Update to v073r01 release. byuu says: While perhaps not perfect, pretty good is better than nothing ... I've added emulation of auto-joypad poll timing. Going off ikari_01's confirmation of what we suspected, that the strobe happens every 256 clocks, I've set up emulation as follows: Upon reset, our clock counter is reset to zero. At the start of each frame, our poll counter is reset to zero. Every 256 clocks, we call the step_auto_joypad_poll() function. If we are at V=225/240+ (based on overscan setting), we check the poll counter. At zero, we poll the actual controller and set the joypad polling flag in $4212.d0 to 1. From zero through fifteen, we read in one bit for each controller and shift it into the register. At sixteen, we turn off the joypad polling flag. The 256-clock divider allows the start point of polling for each frame to fluctuate wildly like real hardware. I count regardless of auto joypad enable, as per $4212.d0's behavior; but only poll when it's actually enabled. I do not consume any actual time from this polling. I honestly don't know if I even should, or if it manages to do it in the background. If it should consume time, then this most likely happens between opcode edges and we'll have to adjust the code a good bit. All commercial games should continue to work fine, but this will likely break some hacks/translations not tested on hardware. Without the timing emulation, reading $4218-421f before V=~228 would basically give you the valid input controller values of the previous frame. Now, like hardware, it should give you a state that is part previous frame, part current frame shifted into it. Button positions won't be reliable and will shift every 256 clocks. I've also removed the Qt GUI, and renamed ui-phoenix to just ui. This removes 400kb of source code (phoenix is a lean 130kb), and drops the archive size from 564KB to 475KB. Combined with the DSP HLE, and we've knocked off ~570KB of source cruft from the entire project. I am looking forward to not having to specify which GUI is included anymore.
2010-12-27 07:29:57 +00:00
uint8 CPU::mmio_r4218() { return status.joy1 >> 0; } //JOY1L
uint8 CPU::mmio_r4219() { return status.joy1 >> 8; } //JOY1H
uint8 CPU::mmio_r421a() { return status.joy2 >> 0; } //JOY2L
uint8 CPU::mmio_r421b() { return status.joy2 >> 8; } //JOY2H
uint8 CPU::mmio_r421c() { return status.joy3 >> 0; } //JOY3L
uint8 CPU::mmio_r421d() { return status.joy3 >> 8; } //JOY3H
uint8 CPU::mmio_r421e() { return status.joy4 >> 0; } //JOY4L
uint8 CPU::mmio_r421f() { return status.joy4 >> 8; } //JOY4H
//DMAPx
uint8 CPU::mmio_r43x0(uint8 i) {
return (channel[i].direction << 7)
| (channel[i].indirect << 6)
| (channel[i].unused << 5)
| (channel[i].reverse_transfer << 4)
| (channel[i].fixed_transfer << 3)
| (channel[i].transfer_mode << 0);
}
//BBADx
uint8 CPU::mmio_r43x1(uint8 i) {
return channel[i].dest_addr;
}
//A1TxL
uint8 CPU::mmio_r43x2(uint8 i) {
Updated to v067r23 release. byuu says: Added missing $4200 IRQ lock, which fixes Chou Aniki on the fast CPU core, so slower PCs can get their brotherly love on. Added range-based controller IOBit latching to the fast CPU core, which enables Super Scope and Justifier support. Uses the priority queue as well, so there is zero speed-hit. Given the way range-testing works, the trigger point may vary by 1-2 pixels when firing at the same spot. Not really a big deal when it avoids a massive speed penalty. Fixed PAL and interlace-mode HVIRQs at V=0,H<2 on the fast CPU core. Added the dot-renderer's sprite list update-on-OAM-write functionality to the scanline-based PPU renderer. Unfortunately it looks like all the speed gain was already taken from the global dirty flag I was using before, but this certainly won't hurt speed any, so whatever. Added #ifdef to stop CoInitialize(0) on non-Windows ports. Added #ifdefs to stop gradient fade on Windows port. Not going to fuck over the Linux port aesthetic because of Qt bug #47,326,927. If there's a way to tell what Qt theme is being used, I can leave it enabled for XP/Vista themes. Moved HDMA trigger from 1104 to 1112, and reduced channel overhead from 24 to 16, to better simulate one-cycle DMA->CPU sync. Code clarity: I've re-added my varint.hpp classes, and am actively using them in the accuracy cores. So far, I haven't done anything that would detriment speed, but it is certainly cool. The APU ports exposed by the CPU and SMP now take uint2 address arguments, the CPU WRAM address register is a uint17, and the IRQ H/VTIME values are uint10. This basically allows the source to clearly convey the data sizes, and eliminates the need to manually mask values when writing to registers or reading from memory. I'm going to be doing this everywhere, and it will have a speed impact eventually, because the automation means we can't skip masks when we know the data is already masked off. Source: archive contains the launcher code, so that I can look into why it's crashing on XP tomorrow. It doesn't look like Circuit USA's flags are going to work too well with this new CPU core. Still not sure what the hell Robocop vs The Terminator is doing, I'll read through the mega SNES thread for clues tomorrow. Speedy Gonzales is definitely broken, as modifying the MDR was breaking things with my current core. Probably because the new CPU core doesn't wait for a cycle edge to trigger. I was thinking that perhaps we could keep some form of cheat codes list to work as game-specific hacks for the performance core. Keeps the hacks out of the emulator, but could allow the remaining bugs to be worked around for people who have no choice but to use the performance core.
2010-08-16 09:42:20 +00:00
return channel[i].source_addr >> 0;
}
//A1TxH
uint8 CPU::mmio_r43x3(uint8 i) {
return channel[i].source_addr >> 8;
}
//A1Bx
uint8 CPU::mmio_r43x4(uint8 i) {
return channel[i].source_bank;
}
//DASxL
//union { uint16 transfer_size; uint16 indirect_addr; };
uint8 CPU::mmio_r43x5(uint8 i) {
Updated to v067r23 release. byuu says: Added missing $4200 IRQ lock, which fixes Chou Aniki on the fast CPU core, so slower PCs can get their brotherly love on. Added range-based controller IOBit latching to the fast CPU core, which enables Super Scope and Justifier support. Uses the priority queue as well, so there is zero speed-hit. Given the way range-testing works, the trigger point may vary by 1-2 pixels when firing at the same spot. Not really a big deal when it avoids a massive speed penalty. Fixed PAL and interlace-mode HVIRQs at V=0,H<2 on the fast CPU core. Added the dot-renderer's sprite list update-on-OAM-write functionality to the scanline-based PPU renderer. Unfortunately it looks like all the speed gain was already taken from the global dirty flag I was using before, but this certainly won't hurt speed any, so whatever. Added #ifdef to stop CoInitialize(0) on non-Windows ports. Added #ifdefs to stop gradient fade on Windows port. Not going to fuck over the Linux port aesthetic because of Qt bug #47,326,927. If there's a way to tell what Qt theme is being used, I can leave it enabled for XP/Vista themes. Moved HDMA trigger from 1104 to 1112, and reduced channel overhead from 24 to 16, to better simulate one-cycle DMA->CPU sync. Code clarity: I've re-added my varint.hpp classes, and am actively using them in the accuracy cores. So far, I haven't done anything that would detriment speed, but it is certainly cool. The APU ports exposed by the CPU and SMP now take uint2 address arguments, the CPU WRAM address register is a uint17, and the IRQ H/VTIME values are uint10. This basically allows the source to clearly convey the data sizes, and eliminates the need to manually mask values when writing to registers or reading from memory. I'm going to be doing this everywhere, and it will have a speed impact eventually, because the automation means we can't skip masks when we know the data is already masked off. Source: archive contains the launcher code, so that I can look into why it's crashing on XP tomorrow. It doesn't look like Circuit USA's flags are going to work too well with this new CPU core. Still not sure what the hell Robocop vs The Terminator is doing, I'll read through the mega SNES thread for clues tomorrow. Speedy Gonzales is definitely broken, as modifying the MDR was breaking things with my current core. Probably because the new CPU core doesn't wait for a cycle edge to trigger. I was thinking that perhaps we could keep some form of cheat codes list to work as game-specific hacks for the performance core. Keeps the hacks out of the emulator, but could allow the remaining bugs to be worked around for people who have no choice but to use the performance core.
2010-08-16 09:42:20 +00:00
return channel[i].transfer_size >> 0;
}
//DASxH
//union { uint16 transfer_size; uint16 indirect_addr; };
uint8 CPU::mmio_r43x6(uint8 i) {
return channel[i].transfer_size >> 8;
}
//DASBx
uint8 CPU::mmio_r43x7(uint8 i) {
return channel[i].indirect_bank;
}
//A2AxL
uint8 CPU::mmio_r43x8(uint8 i) {
Updated to v067r23 release. byuu says: Added missing $4200 IRQ lock, which fixes Chou Aniki on the fast CPU core, so slower PCs can get their brotherly love on. Added range-based controller IOBit latching to the fast CPU core, which enables Super Scope and Justifier support. Uses the priority queue as well, so there is zero speed-hit. Given the way range-testing works, the trigger point may vary by 1-2 pixels when firing at the same spot. Not really a big deal when it avoids a massive speed penalty. Fixed PAL and interlace-mode HVIRQs at V=0,H<2 on the fast CPU core. Added the dot-renderer's sprite list update-on-OAM-write functionality to the scanline-based PPU renderer. Unfortunately it looks like all the speed gain was already taken from the global dirty flag I was using before, but this certainly won't hurt speed any, so whatever. Added #ifdef to stop CoInitialize(0) on non-Windows ports. Added #ifdefs to stop gradient fade on Windows port. Not going to fuck over the Linux port aesthetic because of Qt bug #47,326,927. If there's a way to tell what Qt theme is being used, I can leave it enabled for XP/Vista themes. Moved HDMA trigger from 1104 to 1112, and reduced channel overhead from 24 to 16, to better simulate one-cycle DMA->CPU sync. Code clarity: I've re-added my varint.hpp classes, and am actively using them in the accuracy cores. So far, I haven't done anything that would detriment speed, but it is certainly cool. The APU ports exposed by the CPU and SMP now take uint2 address arguments, the CPU WRAM address register is a uint17, and the IRQ H/VTIME values are uint10. This basically allows the source to clearly convey the data sizes, and eliminates the need to manually mask values when writing to registers or reading from memory. I'm going to be doing this everywhere, and it will have a speed impact eventually, because the automation means we can't skip masks when we know the data is already masked off. Source: archive contains the launcher code, so that I can look into why it's crashing on XP tomorrow. It doesn't look like Circuit USA's flags are going to work too well with this new CPU core. Still not sure what the hell Robocop vs The Terminator is doing, I'll read through the mega SNES thread for clues tomorrow. Speedy Gonzales is definitely broken, as modifying the MDR was breaking things with my current core. Probably because the new CPU core doesn't wait for a cycle edge to trigger. I was thinking that perhaps we could keep some form of cheat codes list to work as game-specific hacks for the performance core. Keeps the hacks out of the emulator, but could allow the remaining bugs to be worked around for people who have no choice but to use the performance core.
2010-08-16 09:42:20 +00:00
return channel[i].hdma_addr >> 0;
}
//A2AxH
uint8 CPU::mmio_r43x9(uint8 i) {
return channel[i].hdma_addr >> 8;
}
//NTRLx
uint8 CPU::mmio_r43xa(uint8 i) {
return channel[i].line_counter;
}
//???
uint8 CPU::mmio_r43xb(uint8 i) {
return channel[i].unknown;
}
//DMAPx
void CPU::mmio_w43x0(uint8 i, uint8 data) {
channel[i].direction = data & 0x80;
channel[i].indirect = data & 0x40;
channel[i].unused = data & 0x20;
channel[i].reverse_transfer = data & 0x10;
channel[i].fixed_transfer = data & 0x08;
channel[i].transfer_mode = data & 0x07;
}
//DDBADx
void CPU::mmio_w43x1(uint8 i, uint8 data) {
channel[i].dest_addr = data;
}
//A1TxL
void CPU::mmio_w43x2(uint8 i, uint8 data) {
Updated to v067r23 release. byuu says: Added missing $4200 IRQ lock, which fixes Chou Aniki on the fast CPU core, so slower PCs can get their brotherly love on. Added range-based controller IOBit latching to the fast CPU core, which enables Super Scope and Justifier support. Uses the priority queue as well, so there is zero speed-hit. Given the way range-testing works, the trigger point may vary by 1-2 pixels when firing at the same spot. Not really a big deal when it avoids a massive speed penalty. Fixed PAL and interlace-mode HVIRQs at V=0,H<2 on the fast CPU core. Added the dot-renderer's sprite list update-on-OAM-write functionality to the scanline-based PPU renderer. Unfortunately it looks like all the speed gain was already taken from the global dirty flag I was using before, but this certainly won't hurt speed any, so whatever. Added #ifdef to stop CoInitialize(0) on non-Windows ports. Added #ifdefs to stop gradient fade on Windows port. Not going to fuck over the Linux port aesthetic because of Qt bug #47,326,927. If there's a way to tell what Qt theme is being used, I can leave it enabled for XP/Vista themes. Moved HDMA trigger from 1104 to 1112, and reduced channel overhead from 24 to 16, to better simulate one-cycle DMA->CPU sync. Code clarity: I've re-added my varint.hpp classes, and am actively using them in the accuracy cores. So far, I haven't done anything that would detriment speed, but it is certainly cool. The APU ports exposed by the CPU and SMP now take uint2 address arguments, the CPU WRAM address register is a uint17, and the IRQ H/VTIME values are uint10. This basically allows the source to clearly convey the data sizes, and eliminates the need to manually mask values when writing to registers or reading from memory. I'm going to be doing this everywhere, and it will have a speed impact eventually, because the automation means we can't skip masks when we know the data is already masked off. Source: archive contains the launcher code, so that I can look into why it's crashing on XP tomorrow. It doesn't look like Circuit USA's flags are going to work too well with this new CPU core. Still not sure what the hell Robocop vs The Terminator is doing, I'll read through the mega SNES thread for clues tomorrow. Speedy Gonzales is definitely broken, as modifying the MDR was breaking things with my current core. Probably because the new CPU core doesn't wait for a cycle edge to trigger. I was thinking that perhaps we could keep some form of cheat codes list to work as game-specific hacks for the performance core. Keeps the hacks out of the emulator, but could allow the remaining bugs to be worked around for people who have no choice but to use the performance core.
2010-08-16 09:42:20 +00:00
channel[i].source_addr = (channel[i].source_addr & 0xff00) | (data << 0);
}
//A1TxH
void CPU::mmio_w43x3(uint8 i, uint8 data) {
channel[i].source_addr = (channel[i].source_addr & 0x00ff) | (data << 8);
}
//A1Bx
void CPU::mmio_w43x4(uint8 i, uint8 data) {
channel[i].source_bank = data;
}
//DASxL
//union { uint16 transfer_size; uint16 indirect_addr; };
void CPU::mmio_w43x5(uint8 i, uint8 data) {
Updated to v067r23 release. byuu says: Added missing $4200 IRQ lock, which fixes Chou Aniki on the fast CPU core, so slower PCs can get their brotherly love on. Added range-based controller IOBit latching to the fast CPU core, which enables Super Scope and Justifier support. Uses the priority queue as well, so there is zero speed-hit. Given the way range-testing works, the trigger point may vary by 1-2 pixels when firing at the same spot. Not really a big deal when it avoids a massive speed penalty. Fixed PAL and interlace-mode HVIRQs at V=0,H<2 on the fast CPU core. Added the dot-renderer's sprite list update-on-OAM-write functionality to the scanline-based PPU renderer. Unfortunately it looks like all the speed gain was already taken from the global dirty flag I was using before, but this certainly won't hurt speed any, so whatever. Added #ifdef to stop CoInitialize(0) on non-Windows ports. Added #ifdefs to stop gradient fade on Windows port. Not going to fuck over the Linux port aesthetic because of Qt bug #47,326,927. If there's a way to tell what Qt theme is being used, I can leave it enabled for XP/Vista themes. Moved HDMA trigger from 1104 to 1112, and reduced channel overhead from 24 to 16, to better simulate one-cycle DMA->CPU sync. Code clarity: I've re-added my varint.hpp classes, and am actively using them in the accuracy cores. So far, I haven't done anything that would detriment speed, but it is certainly cool. The APU ports exposed by the CPU and SMP now take uint2 address arguments, the CPU WRAM address register is a uint17, and the IRQ H/VTIME values are uint10. This basically allows the source to clearly convey the data sizes, and eliminates the need to manually mask values when writing to registers or reading from memory. I'm going to be doing this everywhere, and it will have a speed impact eventually, because the automation means we can't skip masks when we know the data is already masked off. Source: archive contains the launcher code, so that I can look into why it's crashing on XP tomorrow. It doesn't look like Circuit USA's flags are going to work too well with this new CPU core. Still not sure what the hell Robocop vs The Terminator is doing, I'll read through the mega SNES thread for clues tomorrow. Speedy Gonzales is definitely broken, as modifying the MDR was breaking things with my current core. Probably because the new CPU core doesn't wait for a cycle edge to trigger. I was thinking that perhaps we could keep some form of cheat codes list to work as game-specific hacks for the performance core. Keeps the hacks out of the emulator, but could allow the remaining bugs to be worked around for people who have no choice but to use the performance core.
2010-08-16 09:42:20 +00:00
channel[i].transfer_size = (channel[i].transfer_size & 0xff00) | (data << 0);
}
//DASxH
//union { uint16 transfer_size; uint16 indirect_addr; };
void CPU::mmio_w43x6(uint8 i, uint8 data) {
channel[i].transfer_size = (channel[i].transfer_size & 0x00ff) | (data << 8);
}
//DASBx
void CPU::mmio_w43x7(uint8 i, uint8 data) {
channel[i].indirect_bank = data;
}
//A2AxL
void CPU::mmio_w43x8(uint8 i, uint8 data) {
Updated to v067r23 release. byuu says: Added missing $4200 IRQ lock, which fixes Chou Aniki on the fast CPU core, so slower PCs can get their brotherly love on. Added range-based controller IOBit latching to the fast CPU core, which enables Super Scope and Justifier support. Uses the priority queue as well, so there is zero speed-hit. Given the way range-testing works, the trigger point may vary by 1-2 pixels when firing at the same spot. Not really a big deal when it avoids a massive speed penalty. Fixed PAL and interlace-mode HVIRQs at V=0,H<2 on the fast CPU core. Added the dot-renderer's sprite list update-on-OAM-write functionality to the scanline-based PPU renderer. Unfortunately it looks like all the speed gain was already taken from the global dirty flag I was using before, but this certainly won't hurt speed any, so whatever. Added #ifdef to stop CoInitialize(0) on non-Windows ports. Added #ifdefs to stop gradient fade on Windows port. Not going to fuck over the Linux port aesthetic because of Qt bug #47,326,927. If there's a way to tell what Qt theme is being used, I can leave it enabled for XP/Vista themes. Moved HDMA trigger from 1104 to 1112, and reduced channel overhead from 24 to 16, to better simulate one-cycle DMA->CPU sync. Code clarity: I've re-added my varint.hpp classes, and am actively using them in the accuracy cores. So far, I haven't done anything that would detriment speed, but it is certainly cool. The APU ports exposed by the CPU and SMP now take uint2 address arguments, the CPU WRAM address register is a uint17, and the IRQ H/VTIME values are uint10. This basically allows the source to clearly convey the data sizes, and eliminates the need to manually mask values when writing to registers or reading from memory. I'm going to be doing this everywhere, and it will have a speed impact eventually, because the automation means we can't skip masks when we know the data is already masked off. Source: archive contains the launcher code, so that I can look into why it's crashing on XP tomorrow. It doesn't look like Circuit USA's flags are going to work too well with this new CPU core. Still not sure what the hell Robocop vs The Terminator is doing, I'll read through the mega SNES thread for clues tomorrow. Speedy Gonzales is definitely broken, as modifying the MDR was breaking things with my current core. Probably because the new CPU core doesn't wait for a cycle edge to trigger. I was thinking that perhaps we could keep some form of cheat codes list to work as game-specific hacks for the performance core. Keeps the hacks out of the emulator, but could allow the remaining bugs to be worked around for people who have no choice but to use the performance core.
2010-08-16 09:42:20 +00:00
channel[i].hdma_addr = (channel[i].hdma_addr & 0xff00) | (data << 0);
}
//A2AxH
void CPU::mmio_w43x9(uint8 i, uint8 data) {
channel[i].hdma_addr = (channel[i].hdma_addr & 0x00ff) | (data << 8);
}
//NTRLx
void CPU::mmio_w43xa(uint8 i, uint8 data) {
channel[i].line_counter = data;
}
//???
void CPU::mmio_w43xb(uint8 i, uint8 data) {
channel[i].unknown = data;
}
void CPU::mmio_power() {
}
void CPU::mmio_reset() {
Updated to v067r23 release. byuu says: Added missing $4200 IRQ lock, which fixes Chou Aniki on the fast CPU core, so slower PCs can get their brotherly love on. Added range-based controller IOBit latching to the fast CPU core, which enables Super Scope and Justifier support. Uses the priority queue as well, so there is zero speed-hit. Given the way range-testing works, the trigger point may vary by 1-2 pixels when firing at the same spot. Not really a big deal when it avoids a massive speed penalty. Fixed PAL and interlace-mode HVIRQs at V=0,H<2 on the fast CPU core. Added the dot-renderer's sprite list update-on-OAM-write functionality to the scanline-based PPU renderer. Unfortunately it looks like all the speed gain was already taken from the global dirty flag I was using before, but this certainly won't hurt speed any, so whatever. Added #ifdef to stop CoInitialize(0) on non-Windows ports. Added #ifdefs to stop gradient fade on Windows port. Not going to fuck over the Linux port aesthetic because of Qt bug #47,326,927. If there's a way to tell what Qt theme is being used, I can leave it enabled for XP/Vista themes. Moved HDMA trigger from 1104 to 1112, and reduced channel overhead from 24 to 16, to better simulate one-cycle DMA->CPU sync. Code clarity: I've re-added my varint.hpp classes, and am actively using them in the accuracy cores. So far, I haven't done anything that would detriment speed, but it is certainly cool. The APU ports exposed by the CPU and SMP now take uint2 address arguments, the CPU WRAM address register is a uint17, and the IRQ H/VTIME values are uint10. This basically allows the source to clearly convey the data sizes, and eliminates the need to manually mask values when writing to registers or reading from memory. I'm going to be doing this everywhere, and it will have a speed impact eventually, because the automation means we can't skip masks when we know the data is already masked off. Source: archive contains the launcher code, so that I can look into why it's crashing on XP tomorrow. It doesn't look like Circuit USA's flags are going to work too well with this new CPU core. Still not sure what the hell Robocop vs The Terminator is doing, I'll read through the mega SNES thread for clues tomorrow. Speedy Gonzales is definitely broken, as modifying the MDR was breaking things with my current core. Probably because the new CPU core doesn't wait for a cycle edge to trigger. I was thinking that perhaps we could keep some form of cheat codes list to work as game-specific hacks for the performance core. Keeps the hacks out of the emulator, but could allow the remaining bugs to be worked around for people who have no choice but to use the performance core.
2010-08-16 09:42:20 +00:00
//$2140-217f
for(auto& port : status.port) port = 0x00;
Updated to v067r23 release. byuu says: Added missing $4200 IRQ lock, which fixes Chou Aniki on the fast CPU core, so slower PCs can get their brotherly love on. Added range-based controller IOBit latching to the fast CPU core, which enables Super Scope and Justifier support. Uses the priority queue as well, so there is zero speed-hit. Given the way range-testing works, the trigger point may vary by 1-2 pixels when firing at the same spot. Not really a big deal when it avoids a massive speed penalty. Fixed PAL and interlace-mode HVIRQs at V=0,H<2 on the fast CPU core. Added the dot-renderer's sprite list update-on-OAM-write functionality to the scanline-based PPU renderer. Unfortunately it looks like all the speed gain was already taken from the global dirty flag I was using before, but this certainly won't hurt speed any, so whatever. Added #ifdef to stop CoInitialize(0) on non-Windows ports. Added #ifdefs to stop gradient fade on Windows port. Not going to fuck over the Linux port aesthetic because of Qt bug #47,326,927. If there's a way to tell what Qt theme is being used, I can leave it enabled for XP/Vista themes. Moved HDMA trigger from 1104 to 1112, and reduced channel overhead from 24 to 16, to better simulate one-cycle DMA->CPU sync. Code clarity: I've re-added my varint.hpp classes, and am actively using them in the accuracy cores. So far, I haven't done anything that would detriment speed, but it is certainly cool. The APU ports exposed by the CPU and SMP now take uint2 address arguments, the CPU WRAM address register is a uint17, and the IRQ H/VTIME values are uint10. This basically allows the source to clearly convey the data sizes, and eliminates the need to manually mask values when writing to registers or reading from memory. I'm going to be doing this everywhere, and it will have a speed impact eventually, because the automation means we can't skip masks when we know the data is already masked off. Source: archive contains the launcher code, so that I can look into why it's crashing on XP tomorrow. It doesn't look like Circuit USA's flags are going to work too well with this new CPU core. Still not sure what the hell Robocop vs The Terminator is doing, I'll read through the mega SNES thread for clues tomorrow. Speedy Gonzales is definitely broken, as modifying the MDR was breaking things with my current core. Probably because the new CPU core doesn't wait for a cycle edge to trigger. I was thinking that perhaps we could keep some form of cheat codes list to work as game-specific hacks for the performance core. Keeps the hacks out of the emulator, but could allow the remaining bugs to be worked around for people who have no choice but to use the performance core.
2010-08-16 09:42:20 +00:00
//$2181-$2183
status.wram_addr = 0x000000;
//$4016-$4017
status.joypad_strobe_latch = 0;
status.joypad1_bits = ~0;
status.joypad2_bits = ~0;
//$4200
Updated to v067r23 release. byuu says: Added missing $4200 IRQ lock, which fixes Chou Aniki on the fast CPU core, so slower PCs can get their brotherly love on. Added range-based controller IOBit latching to the fast CPU core, which enables Super Scope and Justifier support. Uses the priority queue as well, so there is zero speed-hit. Given the way range-testing works, the trigger point may vary by 1-2 pixels when firing at the same spot. Not really a big deal when it avoids a massive speed penalty. Fixed PAL and interlace-mode HVIRQs at V=0,H<2 on the fast CPU core. Added the dot-renderer's sprite list update-on-OAM-write functionality to the scanline-based PPU renderer. Unfortunately it looks like all the speed gain was already taken from the global dirty flag I was using before, but this certainly won't hurt speed any, so whatever. Added #ifdef to stop CoInitialize(0) on non-Windows ports. Added #ifdefs to stop gradient fade on Windows port. Not going to fuck over the Linux port aesthetic because of Qt bug #47,326,927. If there's a way to tell what Qt theme is being used, I can leave it enabled for XP/Vista themes. Moved HDMA trigger from 1104 to 1112, and reduced channel overhead from 24 to 16, to better simulate one-cycle DMA->CPU sync. Code clarity: I've re-added my varint.hpp classes, and am actively using them in the accuracy cores. So far, I haven't done anything that would detriment speed, but it is certainly cool. The APU ports exposed by the CPU and SMP now take uint2 address arguments, the CPU WRAM address register is a uint17, and the IRQ H/VTIME values are uint10. This basically allows the source to clearly convey the data sizes, and eliminates the need to manually mask values when writing to registers or reading from memory. I'm going to be doing this everywhere, and it will have a speed impact eventually, because the automation means we can't skip masks when we know the data is already masked off. Source: archive contains the launcher code, so that I can look into why it's crashing on XP tomorrow. It doesn't look like Circuit USA's flags are going to work too well with this new CPU core. Still not sure what the hell Robocop vs The Terminator is doing, I'll read through the mega SNES thread for clues tomorrow. Speedy Gonzales is definitely broken, as modifying the MDR was breaking things with my current core. Probably because the new CPU core doesn't wait for a cycle edge to trigger. I was thinking that perhaps we could keep some form of cheat codes list to work as game-specific hacks for the performance core. Keeps the hacks out of the emulator, but could allow the remaining bugs to be worked around for people who have no choice but to use the performance core.
2010-08-16 09:42:20 +00:00
status.nmi_enabled = false;
status.hirq_enabled = false;
status.virq_enabled = false;
status.auto_joypad_poll = false;
//$4201
status.pio = 0xff;
//$4202-$4203
status.wrmpya = 0xff;
status.wrmpyb = 0xff;
//$4204-$4206
status.wrdiva = 0xffff;
status.wrdivb = 0xff;
//$4207-$420a
status.hirq_pos = 0x01ff;
status.virq_pos = 0x01ff;
//$420d
status.rom_speed = 8;
//$4214-$4217
status.rddiv = 0x0000;
status.rdmpy = 0x0000;
//$4218-$421f
Update to v073r01 release. byuu says: While perhaps not perfect, pretty good is better than nothing ... I've added emulation of auto-joypad poll timing. Going off ikari_01's confirmation of what we suspected, that the strobe happens every 256 clocks, I've set up emulation as follows: Upon reset, our clock counter is reset to zero. At the start of each frame, our poll counter is reset to zero. Every 256 clocks, we call the step_auto_joypad_poll() function. If we are at V=225/240+ (based on overscan setting), we check the poll counter. At zero, we poll the actual controller and set the joypad polling flag in $4212.d0 to 1. From zero through fifteen, we read in one bit for each controller and shift it into the register. At sixteen, we turn off the joypad polling flag. The 256-clock divider allows the start point of polling for each frame to fluctuate wildly like real hardware. I count regardless of auto joypad enable, as per $4212.d0's behavior; but only poll when it's actually enabled. I do not consume any actual time from this polling. I honestly don't know if I even should, or if it manages to do it in the background. If it should consume time, then this most likely happens between opcode edges and we'll have to adjust the code a good bit. All commercial games should continue to work fine, but this will likely break some hacks/translations not tested on hardware. Without the timing emulation, reading $4218-421f before V=~228 would basically give you the valid input controller values of the previous frame. Now, like hardware, it should give you a state that is part previous frame, part current frame shifted into it. Button positions won't be reliable and will shift every 256 clocks. I've also removed the Qt GUI, and renamed ui-phoenix to just ui. This removes 400kb of source code (phoenix is a lean 130kb), and drops the archive size from 564KB to 475KB. Combined with the DSP HLE, and we've knocked off ~570KB of source cruft from the entire project. I am looking forward to not having to specify which GUI is included anymore.
2010-12-27 07:29:57 +00:00
status.joy1 = 0x0000;
status.joy2 = 0x0000;
status.joy3 = 0x0000;
status.joy4 = 0x0000;
//ALU
alu.mpyctr = 0;
alu.divctr = 0;
Updated to v067r23 release. byuu says: Added missing $4200 IRQ lock, which fixes Chou Aniki on the fast CPU core, so slower PCs can get their brotherly love on. Added range-based controller IOBit latching to the fast CPU core, which enables Super Scope and Justifier support. Uses the priority queue as well, so there is zero speed-hit. Given the way range-testing works, the trigger point may vary by 1-2 pixels when firing at the same spot. Not really a big deal when it avoids a massive speed penalty. Fixed PAL and interlace-mode HVIRQs at V=0,H<2 on the fast CPU core. Added the dot-renderer's sprite list update-on-OAM-write functionality to the scanline-based PPU renderer. Unfortunately it looks like all the speed gain was already taken from the global dirty flag I was using before, but this certainly won't hurt speed any, so whatever. Added #ifdef to stop CoInitialize(0) on non-Windows ports. Added #ifdefs to stop gradient fade on Windows port. Not going to fuck over the Linux port aesthetic because of Qt bug #47,326,927. If there's a way to tell what Qt theme is being used, I can leave it enabled for XP/Vista themes. Moved HDMA trigger from 1104 to 1112, and reduced channel overhead from 24 to 16, to better simulate one-cycle DMA->CPU sync. Code clarity: I've re-added my varint.hpp classes, and am actively using them in the accuracy cores. So far, I haven't done anything that would detriment speed, but it is certainly cool. The APU ports exposed by the CPU and SMP now take uint2 address arguments, the CPU WRAM address register is a uint17, and the IRQ H/VTIME values are uint10. This basically allows the source to clearly convey the data sizes, and eliminates the need to manually mask values when writing to registers or reading from memory. I'm going to be doing this everywhere, and it will have a speed impact eventually, because the automation means we can't skip masks when we know the data is already masked off. Source: archive contains the launcher code, so that I can look into why it's crashing on XP tomorrow. It doesn't look like Circuit USA's flags are going to work too well with this new CPU core. Still not sure what the hell Robocop vs The Terminator is doing, I'll read through the mega SNES thread for clues tomorrow. Speedy Gonzales is definitely broken, as modifying the MDR was breaking things with my current core. Probably because the new CPU core doesn't wait for a cycle edge to trigger. I was thinking that perhaps we could keep some form of cheat codes list to work as game-specific hacks for the performance core. Keeps the hacks out of the emulator, but could allow the remaining bugs to be worked around for people who have no choice but to use the performance core.
2010-08-16 09:42:20 +00:00
alu.shift = 0;
}
uint8 CPU::mmio_read(unsigned addr) {
addr &= 0xffff;
//APU
if((addr & 0xffc0) == 0x2140) { //$2140-$217f
synchronize_smp();
return smp.portRead(addr);
}
//DMA
if((addr & 0xff80) == 0x4300) { //$4300-$437f
unsigned i = (addr >> 4) & 7;
switch(addr & 0xf) {
case 0x0: return mmio_r43x0(i);
case 0x1: return mmio_r43x1(i);
case 0x2: return mmio_r43x2(i);
case 0x3: return mmio_r43x3(i);
case 0x4: return mmio_r43x4(i);
case 0x5: return mmio_r43x5(i);
case 0x6: return mmio_r43x6(i);
case 0x7: return mmio_r43x7(i);
case 0x8: return mmio_r43x8(i);
case 0x9: return mmio_r43x9(i);
case 0xa: return mmio_r43xa(i);
case 0xb: return mmio_r43xb(i);
case 0xc: return regs.mdr; //unmapped
case 0xd: return regs.mdr; //unmapped
case 0xe: return regs.mdr; //unmapped
case 0xf: return mmio_r43xb(i); //mirror of $43xb
}
}
switch(addr) {
case 0x2180: return mmio_r2180();
case 0x4016: return mmio_r4016();
case 0x4017: return mmio_r4017();
case 0x4210: return mmio_r4210();
case 0x4211: return mmio_r4211();
case 0x4212: return mmio_r4212();
case 0x4213: return mmio_r4213();
case 0x4214: return mmio_r4214();
case 0x4215: return mmio_r4215();
case 0x4216: return mmio_r4216();
case 0x4217: return mmio_r4217();
case 0x4218: return mmio_r4218();
case 0x4219: return mmio_r4219();
case 0x421a: return mmio_r421a();
case 0x421b: return mmio_r421b();
case 0x421c: return mmio_r421c();
case 0x421d: return mmio_r421d();
case 0x421e: return mmio_r421e();
case 0x421f: return mmio_r421f();
}
return regs.mdr;
}
void CPU::mmio_write(unsigned addr, uint8 data) {
addr &= 0xffff;
//APU
if((addr & 0xffc0) == 0x2140) { //$2140-$217f
synchronize_smp();
Updated to v067r23 release. byuu says: Added missing $4200 IRQ lock, which fixes Chou Aniki on the fast CPU core, so slower PCs can get their brotherly love on. Added range-based controller IOBit latching to the fast CPU core, which enables Super Scope and Justifier support. Uses the priority queue as well, so there is zero speed-hit. Given the way range-testing works, the trigger point may vary by 1-2 pixels when firing at the same spot. Not really a big deal when it avoids a massive speed penalty. Fixed PAL and interlace-mode HVIRQs at V=0,H<2 on the fast CPU core. Added the dot-renderer's sprite list update-on-OAM-write functionality to the scanline-based PPU renderer. Unfortunately it looks like all the speed gain was already taken from the global dirty flag I was using before, but this certainly won't hurt speed any, so whatever. Added #ifdef to stop CoInitialize(0) on non-Windows ports. Added #ifdefs to stop gradient fade on Windows port. Not going to fuck over the Linux port aesthetic because of Qt bug #47,326,927. If there's a way to tell what Qt theme is being used, I can leave it enabled for XP/Vista themes. Moved HDMA trigger from 1104 to 1112, and reduced channel overhead from 24 to 16, to better simulate one-cycle DMA->CPU sync. Code clarity: I've re-added my varint.hpp classes, and am actively using them in the accuracy cores. So far, I haven't done anything that would detriment speed, but it is certainly cool. The APU ports exposed by the CPU and SMP now take uint2 address arguments, the CPU WRAM address register is a uint17, and the IRQ H/VTIME values are uint10. This basically allows the source to clearly convey the data sizes, and eliminates the need to manually mask values when writing to registers or reading from memory. I'm going to be doing this everywhere, and it will have a speed impact eventually, because the automation means we can't skip masks when we know the data is already masked off. Source: archive contains the launcher code, so that I can look into why it's crashing on XP tomorrow. It doesn't look like Circuit USA's flags are going to work too well with this new CPU core. Still not sure what the hell Robocop vs The Terminator is doing, I'll read through the mega SNES thread for clues tomorrow. Speedy Gonzales is definitely broken, as modifying the MDR was breaking things with my current core. Probably because the new CPU core doesn't wait for a cycle edge to trigger. I was thinking that perhaps we could keep some form of cheat codes list to work as game-specific hacks for the performance core. Keeps the hacks out of the emulator, but could allow the remaining bugs to be worked around for people who have no choice but to use the performance core.
2010-08-16 09:42:20 +00:00
port_write(addr, data);
return;
}
//DMA
if((addr & 0xff80) == 0x4300) { //$4300-$437f
unsigned i = (addr >> 4) & 7;
switch(addr & 0xf) {
case 0x0: mmio_w43x0(i, data); return;
case 0x1: mmio_w43x1(i, data); return;
case 0x2: mmio_w43x2(i, data); return;
case 0x3: mmio_w43x3(i, data); return;
case 0x4: mmio_w43x4(i, data); return;
case 0x5: mmio_w43x5(i, data); return;
case 0x6: mmio_w43x6(i, data); return;
case 0x7: mmio_w43x7(i, data); return;
case 0x8: mmio_w43x8(i, data); return;
case 0x9: mmio_w43x9(i, data); return;
case 0xa: mmio_w43xa(i, data); return;
case 0xb: mmio_w43xb(i, data); return;
case 0xc: return; //unmapped
case 0xd: return; //unmapped
case 0xe: return; //unmapped
case 0xf: mmio_w43xb(i, data); return; //mirror of $43xb
}
}
switch(addr) {
case 0x2180: mmio_w2180(data); return;
case 0x2181: mmio_w2181(data); return;
case 0x2182: mmio_w2182(data); return;
case 0x2183: mmio_w2183(data); return;
case 0x4016: mmio_w4016(data); return;
case 0x4017: return; //unmapped
case 0x4200: mmio_w4200(data); return;
case 0x4201: mmio_w4201(data); return;
case 0x4202: mmio_w4202(data); return;
case 0x4203: mmio_w4203(data); return;
case 0x4204: mmio_w4204(data); return;
case 0x4205: mmio_w4205(data); return;
case 0x4206: mmio_w4206(data); return;
case 0x4207: mmio_w4207(data); return;
case 0x4208: mmio_w4208(data); return;
case 0x4209: mmio_w4209(data); return;
case 0x420a: mmio_w420a(data); return;
case 0x420b: mmio_w420b(data); return;
case 0x420c: mmio_w420c(data); return;
case 0x420d: mmio_w420d(data); return;
}
}
#endif