2010-08-09 13:28:56 +00:00
|
|
|
#ifdef SA1_CPP
|
|
|
|
|
|
2011-01-16 13:22:51 +00:00
|
|
|
uint8 SA1::bus_read(unsigned addr) {
|
|
|
|
|
if((addr & 0x40fe00) == 0x002200) { //$00-3f|80-bf:2200-23ff
|
|
|
|
|
return mmio_read(addr);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if((addr & 0x408000) == 0x008000) { //$00-3f|80-bf:8000-ffff
|
Update to v089r17 release.
byuu says:
This implements the spec from the XML part 3 thread:
http://board.byuu.org/viewtopic.php?f=16&t=2998
It's also propagated the changes to nall and purify, so you can test
this one.
This is basically it, after years of effort I feel I finally have
a fully consistent and logical XML board format.
The only things left to change will be: modifications if emulation turns
out to be incorrect (eg we missed some MMIO mirrors, or mirrored too
much), and new additions.
And of course, I'm giving it a bit of time for good arguments against
the format.
Other than that, this release removes linear_vector and pointer_vector,
as vector is better than linear_vector and I've never used
pointer_vector.
vector also gets move(), which is a way to use move-semantics across
types. It lets you steal the underlying memory pool, effectively
destroying the vector without a copy.
This works really nicely with the move for read() functions to return
vector<uint8> instead of taking (uint8_t*&, unsigned&) parameters.
2012-07-15 13:02:27 +00:00
|
|
|
return mmcrom_read(addr);
|
2011-01-16 13:22:51 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if((addr & 0xc00000) == 0xc00000) { //$c0-ff:0000-ffff
|
Update to v089r17 release.
byuu says:
This implements the spec from the XML part 3 thread:
http://board.byuu.org/viewtopic.php?f=16&t=2998
It's also propagated the changes to nall and purify, so you can test
this one.
This is basically it, after years of effort I feel I finally have
a fully consistent and logical XML board format.
The only things left to change will be: modifications if emulation turns
out to be incorrect (eg we missed some MMIO mirrors, or mirrored too
much), and new additions.
And of course, I'm giving it a bit of time for good arguments against
the format.
Other than that, this release removes linear_vector and pointer_vector,
as vector is better than linear_vector and I've never used
pointer_vector.
vector also gets move(), which is a way to use move-semantics across
types. It lets you steal the underlying memory pool, effectively
destroying the vector without a copy.
This works really nicely with the move for read() functions to return
vector<uint8> instead of taking (uint8_t*&, unsigned&) parameters.
2012-07-15 13:02:27 +00:00
|
|
|
return mmcrom_read(addr);
|
2011-01-16 13:22:51 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if((addr & 0x40e000) == 0x006000) { //$00-3f|80-bf:6000-7fff
|
|
|
|
|
return mmc_sa1_read(addr);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if((addr & 0x40f800) == 0x000000) { //$00-3f|80-bf:0000-07ff
|
Update to v074r11 release.
byuu says:
Changelog:
- debugger compiles on all three profiles
- libsnes compiles on all three platforms (no API changes to libsnes)
- memory.cpp : namespace memory removed (wram -> cpu, apuram -> smp,
vram, oam, cgram -> ppu)
- sa1.cpp : namespace memory removed (SA-1 specific functions merged
inline to SA1::bus_read,write)
- GameBoy: added serial link support with interrupts and proper 8192hz
timing, but obviously it acts as if no other GB is connected to it
- GameBoy: added STAT OAM interrupt, and better STAT d1,d0 mode values
- UI: since Qt is dead, I've renamed the config files back to bsnes.cfg
and bsnes-geometry.cfg
- SA1: IRAM was not syncing to CPU on SA-1 side
- PPU/Accuracy and PPU/Performance needed Sprite oam renamed to Sprite
sprite; so that I could add uint8 oam[544]
- makes more sense anyway, OAM = object attribute memory, obj or
sprite are better names for Sprite rendering class
- more cleanup
2011-01-24 09:03:17 +00:00
|
|
|
synchronize_cpu();
|
|
|
|
|
return iram.read(addr & 2047);
|
2011-01-16 13:22:51 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if((addr & 0x40f800) == 0x003000) { //$00-3f|80-bf:3000-37ff
|
Update to v074r11 release.
byuu says:
Changelog:
- debugger compiles on all three profiles
- libsnes compiles on all three platforms (no API changes to libsnes)
- memory.cpp : namespace memory removed (wram -> cpu, apuram -> smp,
vram, oam, cgram -> ppu)
- sa1.cpp : namespace memory removed (SA-1 specific functions merged
inline to SA1::bus_read,write)
- GameBoy: added serial link support with interrupts and proper 8192hz
timing, but obviously it acts as if no other GB is connected to it
- GameBoy: added STAT OAM interrupt, and better STAT d1,d0 mode values
- UI: since Qt is dead, I've renamed the config files back to bsnes.cfg
and bsnes-geometry.cfg
- SA1: IRAM was not syncing to CPU on SA-1 side
- PPU/Accuracy and PPU/Performance needed Sprite oam renamed to Sprite
sprite; so that I could add uint8 oam[544]
- makes more sense anyway, OAM = object attribute memory, obj or
sprite are better names for Sprite rendering class
- more cleanup
2011-01-24 09:03:17 +00:00
|
|
|
synchronize_cpu();
|
|
|
|
|
return iram.read(addr & 2047);
|
2011-01-16 13:22:51 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if((addr & 0xf00000) == 0x400000) { //$40-4f:0000-ffff
|
Update to v074r11 release.
byuu says:
Changelog:
- debugger compiles on all three profiles
- libsnes compiles on all three platforms (no API changes to libsnes)
- memory.cpp : namespace memory removed (wram -> cpu, apuram -> smp,
vram, oam, cgram -> ppu)
- sa1.cpp : namespace memory removed (SA-1 specific functions merged
inline to SA1::bus_read,write)
- GameBoy: added serial link support with interrupts and proper 8192hz
timing, but obviously it acts as if no other GB is connected to it
- GameBoy: added STAT OAM interrupt, and better STAT d1,d0 mode values
- UI: since Qt is dead, I've renamed the config files back to bsnes.cfg
and bsnes-geometry.cfg
- SA1: IRAM was not syncing to CPU on SA-1 side
- PPU/Accuracy and PPU/Performance needed Sprite oam renamed to Sprite
sprite; so that I could add uint8 oam[544]
- makes more sense anyway, OAM = object attribute memory, obj or
sprite are better names for Sprite rendering class
- more cleanup
2011-01-24 09:03:17 +00:00
|
|
|
synchronize_cpu();
|
2012-07-08 02:57:34 +00:00
|
|
|
return bwram.read(addr & (bwram.size() - 1));
|
2011-01-16 13:22:51 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if((addr & 0xf00000) == 0x600000) { //$60-6f:0000-ffff
|
Update to v074r11 release.
byuu says:
Changelog:
- debugger compiles on all three profiles
- libsnes compiles on all three platforms (no API changes to libsnes)
- memory.cpp : namespace memory removed (wram -> cpu, apuram -> smp,
vram, oam, cgram -> ppu)
- sa1.cpp : namespace memory removed (SA-1 specific functions merged
inline to SA1::bus_read,write)
- GameBoy: added serial link support with interrupts and proper 8192hz
timing, but obviously it acts as if no other GB is connected to it
- GameBoy: added STAT OAM interrupt, and better STAT d1,d0 mode values
- UI: since Qt is dead, I've renamed the config files back to bsnes.cfg
and bsnes-geometry.cfg
- SA1: IRAM was not syncing to CPU on SA-1 side
- PPU/Accuracy and PPU/Performance needed Sprite oam renamed to Sprite
sprite; so that I could add uint8 oam[544]
- makes more sense anyway, OAM = object attribute memory, obj or
sprite are better names for Sprite rendering class
- more cleanup
2011-01-24 09:03:17 +00:00
|
|
|
synchronize_cpu();
|
|
|
|
|
return bitmap_read(addr & 0x0fffff);
|
2011-01-16 13:22:51 +00:00
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
void SA1::bus_write(unsigned addr, uint8 data) {
|
|
|
|
|
if((addr & 0x40fe00) == 0x002200) { //$00-3f|80-bf:2200-23ff
|
|
|
|
|
return mmio_write(addr, data);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if((addr & 0x40e000) == 0x006000) { //$00-3f|80-bf:6000-7fff
|
|
|
|
|
return mmc_sa1_write(addr, data);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if((addr & 0x40f800) == 0x000000) { //$00-3f|80-bf:0000-07ff
|
Update to v074r11 release.
byuu says:
Changelog:
- debugger compiles on all three profiles
- libsnes compiles on all three platforms (no API changes to libsnes)
- memory.cpp : namespace memory removed (wram -> cpu, apuram -> smp,
vram, oam, cgram -> ppu)
- sa1.cpp : namespace memory removed (SA-1 specific functions merged
inline to SA1::bus_read,write)
- GameBoy: added serial link support with interrupts and proper 8192hz
timing, but obviously it acts as if no other GB is connected to it
- GameBoy: added STAT OAM interrupt, and better STAT d1,d0 mode values
- UI: since Qt is dead, I've renamed the config files back to bsnes.cfg
and bsnes-geometry.cfg
- SA1: IRAM was not syncing to CPU on SA-1 side
- PPU/Accuracy and PPU/Performance needed Sprite oam renamed to Sprite
sprite; so that I could add uint8 oam[544]
- makes more sense anyway, OAM = object attribute memory, obj or
sprite are better names for Sprite rendering class
- more cleanup
2011-01-24 09:03:17 +00:00
|
|
|
synchronize_cpu();
|
|
|
|
|
return iram.write(addr & 2047, data);
|
2011-01-16 13:22:51 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if((addr & 0x40f800) == 0x003000) { //$00-3f|80-bf:3000-37ff
|
Update to v074r11 release.
byuu says:
Changelog:
- debugger compiles on all three profiles
- libsnes compiles on all three platforms (no API changes to libsnes)
- memory.cpp : namespace memory removed (wram -> cpu, apuram -> smp,
vram, oam, cgram -> ppu)
- sa1.cpp : namespace memory removed (SA-1 specific functions merged
inline to SA1::bus_read,write)
- GameBoy: added serial link support with interrupts and proper 8192hz
timing, but obviously it acts as if no other GB is connected to it
- GameBoy: added STAT OAM interrupt, and better STAT d1,d0 mode values
- UI: since Qt is dead, I've renamed the config files back to bsnes.cfg
and bsnes-geometry.cfg
- SA1: IRAM was not syncing to CPU on SA-1 side
- PPU/Accuracy and PPU/Performance needed Sprite oam renamed to Sprite
sprite; so that I could add uint8 oam[544]
- makes more sense anyway, OAM = object attribute memory, obj or
sprite are better names for Sprite rendering class
- more cleanup
2011-01-24 09:03:17 +00:00
|
|
|
synchronize_cpu();
|
|
|
|
|
return iram.write(addr & 2047, data);
|
2011-01-16 13:22:51 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if((addr & 0xf00000) == 0x400000) { //$40-4f:0000-ffff
|
Update to v074r11 release.
byuu says:
Changelog:
- debugger compiles on all three profiles
- libsnes compiles on all three platforms (no API changes to libsnes)
- memory.cpp : namespace memory removed (wram -> cpu, apuram -> smp,
vram, oam, cgram -> ppu)
- sa1.cpp : namespace memory removed (SA-1 specific functions merged
inline to SA1::bus_read,write)
- GameBoy: added serial link support with interrupts and proper 8192hz
timing, but obviously it acts as if no other GB is connected to it
- GameBoy: added STAT OAM interrupt, and better STAT d1,d0 mode values
- UI: since Qt is dead, I've renamed the config files back to bsnes.cfg
and bsnes-geometry.cfg
- SA1: IRAM was not syncing to CPU on SA-1 side
- PPU/Accuracy and PPU/Performance needed Sprite oam renamed to Sprite
sprite; so that I could add uint8 oam[544]
- makes more sense anyway, OAM = object attribute memory, obj or
sprite are better names for Sprite rendering class
- more cleanup
2011-01-24 09:03:17 +00:00
|
|
|
synchronize_cpu();
|
2012-07-08 02:57:34 +00:00
|
|
|
return bwram.write(addr & (bwram.size() - 1), data);
|
2011-01-16 13:22:51 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if((addr & 0xf00000) == 0x600000) { //$60-6f:0000-ffff
|
Update to v074r11 release.
byuu says:
Changelog:
- debugger compiles on all three profiles
- libsnes compiles on all three platforms (no API changes to libsnes)
- memory.cpp : namespace memory removed (wram -> cpu, apuram -> smp,
vram, oam, cgram -> ppu)
- sa1.cpp : namespace memory removed (SA-1 specific functions merged
inline to SA1::bus_read,write)
- GameBoy: added serial link support with interrupts and proper 8192hz
timing, but obviously it acts as if no other GB is connected to it
- GameBoy: added STAT OAM interrupt, and better STAT d1,d0 mode values
- UI: since Qt is dead, I've renamed the config files back to bsnes.cfg
and bsnes-geometry.cfg
- SA1: IRAM was not syncing to CPU on SA-1 side
- PPU/Accuracy and PPU/Performance needed Sprite oam renamed to Sprite
sprite; so that I could add uint8 oam[544]
- makes more sense anyway, OAM = object attribute memory, obj or
sprite are better names for Sprite rendering class
- more cleanup
2011-01-24 09:03:17 +00:00
|
|
|
synchronize_cpu();
|
|
|
|
|
return bitmap_write(addr & 0x0fffff, data);
|
2011-01-16 13:22:51 +00:00
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
//$230c (VDPL), $230d (VDPH) use this bus to read variable-length data.
|
|
|
|
|
//this is used both to keep VBR-reads from accessing MMIO registers, and
|
|
|
|
|
//to avoid syncing the S-CPU and SA-1*; as both chips are able to access
|
|
|
|
|
//these ports.
|
|
|
|
|
uint8 SA1::vbr_read(unsigned addr) {
|
|
|
|
|
if((addr & 0x408000) == 0x008000) { //$00-3f|80-bf:8000-ffff
|
Update to v089r17 release.
byuu says:
This implements the spec from the XML part 3 thread:
http://board.byuu.org/viewtopic.php?f=16&t=2998
It's also propagated the changes to nall and purify, so you can test
this one.
This is basically it, after years of effort I feel I finally have
a fully consistent and logical XML board format.
The only things left to change will be: modifications if emulation turns
out to be incorrect (eg we missed some MMIO mirrors, or mirrored too
much), and new additions.
And of course, I'm giving it a bit of time for good arguments against
the format.
Other than that, this release removes linear_vector and pointer_vector,
as vector is better than linear_vector and I've never used
pointer_vector.
vector also gets move(), which is a way to use move-semantics across
types. It lets you steal the underlying memory pool, effectively
destroying the vector without a copy.
This works really nicely with the move for read() functions to return
vector<uint8> instead of taking (uint8_t*&, unsigned&) parameters.
2012-07-15 13:02:27 +00:00
|
|
|
return mmcrom_read(addr);
|
2011-01-16 13:22:51 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if((addr & 0xc00000) == 0xc00000) { //$c0-ff:0000-ffff
|
Update to v089r17 release.
byuu says:
This implements the spec from the XML part 3 thread:
http://board.byuu.org/viewtopic.php?f=16&t=2998
It's also propagated the changes to nall and purify, so you can test
this one.
This is basically it, after years of effort I feel I finally have
a fully consistent and logical XML board format.
The only things left to change will be: modifications if emulation turns
out to be incorrect (eg we missed some MMIO mirrors, or mirrored too
much), and new additions.
And of course, I'm giving it a bit of time for good arguments against
the format.
Other than that, this release removes linear_vector and pointer_vector,
as vector is better than linear_vector and I've never used
pointer_vector.
vector also gets move(), which is a way to use move-semantics across
types. It lets you steal the underlying memory pool, effectively
destroying the vector without a copy.
This works really nicely with the move for read() functions to return
vector<uint8> instead of taking (uint8_t*&, unsigned&) parameters.
2012-07-15 13:02:27 +00:00
|
|
|
return mmcrom_read(addr);
|
2011-01-16 13:22:51 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if((addr & 0x40e000) == 0x006000) { //$00-3f|80-bf:6000-7fff
|
2012-07-08 02:57:34 +00:00
|
|
|
return bwram.read(addr & (bwram.size() - 1));
|
2011-01-16 13:22:51 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if((addr & 0xf00000) == 0x400000) { //$40-4f:0000-ffff
|
2012-07-08 02:57:34 +00:00
|
|
|
return bwram.read(addr & (bwram.size() - 1));
|
2011-01-16 13:22:51 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if((addr & 0x40f800) == 0x000000) { //$00-3f|80-bf:0000-07ff
|
Update to v074r11 release.
byuu says:
Changelog:
- debugger compiles on all three profiles
- libsnes compiles on all three platforms (no API changes to libsnes)
- memory.cpp : namespace memory removed (wram -> cpu, apuram -> smp,
vram, oam, cgram -> ppu)
- sa1.cpp : namespace memory removed (SA-1 specific functions merged
inline to SA1::bus_read,write)
- GameBoy: added serial link support with interrupts and proper 8192hz
timing, but obviously it acts as if no other GB is connected to it
- GameBoy: added STAT OAM interrupt, and better STAT d1,d0 mode values
- UI: since Qt is dead, I've renamed the config files back to bsnes.cfg
and bsnes-geometry.cfg
- SA1: IRAM was not syncing to CPU on SA-1 side
- PPU/Accuracy and PPU/Performance needed Sprite oam renamed to Sprite
sprite; so that I could add uint8 oam[544]
- makes more sense anyway, OAM = object attribute memory, obj or
sprite are better names for Sprite rendering class
- more cleanup
2011-01-24 09:03:17 +00:00
|
|
|
return iram.read(addr & 2047);
|
2011-01-16 13:22:51 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if((addr & 0x40f800) == 0x003000) { //$00-3f|80-bf:3000-37ff
|
Update to v074r11 release.
byuu says:
Changelog:
- debugger compiles on all three profiles
- libsnes compiles on all three platforms (no API changes to libsnes)
- memory.cpp : namespace memory removed (wram -> cpu, apuram -> smp,
vram, oam, cgram -> ppu)
- sa1.cpp : namespace memory removed (SA-1 specific functions merged
inline to SA1::bus_read,write)
- GameBoy: added serial link support with interrupts and proper 8192hz
timing, but obviously it acts as if no other GB is connected to it
- GameBoy: added STAT OAM interrupt, and better STAT d1,d0 mode values
- UI: since Qt is dead, I've renamed the config files back to bsnes.cfg
and bsnes-geometry.cfg
- SA1: IRAM was not syncing to CPU on SA-1 side
- PPU/Accuracy and PPU/Performance needed Sprite oam renamed to Sprite
sprite; so that I could add uint8 oam[544]
- makes more sense anyway, OAM = object attribute memory, obj or
sprite are better names for Sprite rendering class
- more cleanup
2011-01-24 09:03:17 +00:00
|
|
|
return iram.read(addr & 0x2047);
|
2011-01-16 13:22:51 +00:00
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
2010-08-09 13:28:56 +00:00
|
|
|
//ROM, I-RAM and MMIO registers are accessed at ~10.74MHz (2 clock ticks)
|
|
|
|
|
//BW-RAM is accessed at ~5.37MHz (4 clock ticks)
|
|
|
|
|
//tick() == 2 clock ticks
|
|
|
|
|
//note: bus conflict delays are not emulated at this time
|
|
|
|
|
|
|
|
|
|
void SA1::op_io() {
|
|
|
|
|
tick();
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
uint8 SA1::op_read(unsigned addr) {
|
|
|
|
|
tick();
|
|
|
|
|
if(((addr & 0x40e000) == 0x006000) || ((addr & 0xd00000) == 0x400000)) tick();
|
2011-01-16 13:22:51 +00:00
|
|
|
return bus_read(addr);
|
2010-08-09 13:28:56 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
void SA1::op_write(unsigned addr, uint8 data) {
|
|
|
|
|
tick();
|
|
|
|
|
if(((addr & 0x40e000) == 0x006000) || ((addr & 0xd00000) == 0x400000)) tick();
|
2011-01-16 13:22:51 +00:00
|
|
|
bus_write(addr, data);
|
2010-08-09 13:28:56 +00:00
|
|
|
}
|
|
|
|
|
|
Update to v089r17 release.
byuu says:
This implements the spec from the XML part 3 thread:
http://board.byuu.org/viewtopic.php?f=16&t=2998
It's also propagated the changes to nall and purify, so you can test
this one.
This is basically it, after years of effort I feel I finally have
a fully consistent and logical XML board format.
The only things left to change will be: modifications if emulation turns
out to be incorrect (eg we missed some MMIO mirrors, or mirrored too
much), and new additions.
And of course, I'm giving it a bit of time for good arguments against
the format.
Other than that, this release removes linear_vector and pointer_vector,
as vector is better than linear_vector and I've never used
pointer_vector.
vector also gets move(), which is a way to use move-semantics across
types. It lets you steal the underlying memory pool, effectively
destroying the vector without a copy.
This works really nicely with the move for read() functions to return
vector<uint8> instead of taking (uint8_t*&, unsigned&) parameters.
2012-07-15 13:02:27 +00:00
|
|
|
uint8 SA1::mmcrom_read(unsigned addr) {
|
Update to v074r03 release.
byuu says:
You guys are going to hate the hell out of this one. It's twenty hours
of non-stop work, no exaggeration at all. Started at 4AM, just wrapped
up now at 8PM.
I rewrote the entire memory subsystem.
Old system:
65536 pages that map 256 bytes each
Mapping a new page overwrites old page
Granularity capped at 256 bytes minimum, requiring ST-001x to map
60:0000-00ff instead of 60:0000,0001
Classes inherit from MMIO and Memory, forcing only one mappable function
per class, and fixed names
MMIO sub-mapper inside memory: 00-3f:2000-5fff for one-byte granularity
Can dynamically change the map at run-time, MMC register settings
perform dynamic remapping
New system:
XML mapping is still based around banklo-bankhi:addrlo-addrhi, as that
shapes almost everything on the SNES very well
Internally, 2048 pages that map 8192 bytes each
Pages are vectors, scans O(n) from last to first (O(log n) would not
help, n is never > 3)
Can multi-cast writes, but not reads [for the obvious reason of: which
read do you return?]
Can map reads and writes separately
Granularity of one for entire 24-bit address range, no need for MMIO
- whatever is in XML is exactly what you get
Read/Write tables bind function callbacks, so I can have any number of
functions with any names from any classes with no inheritance (no
more uPD7725DR, uPD7725SR helpers, etc)
Less memory usage overall due to less tables [ I tried 16 million tables
and it used 2GB of RAM >_o ]
Cannot dynamically change the map at run-time, MMC read/write functions
perform address translation [worse average case speed, better worst
case speed]
Now the hate me part, functors can't beat virtual functions for speed.
There are speed penalties involved:
-4.5% on average games
-11% on SuperFX games (SFX has its own bus)
-15% on SA-1 games (SA-1 has two buses)
Of course the two that need the speed the most get the biggest hits.
I'm afraid there's really not a lot of wiggle room to boost speed back
up.
I suppose one bright spot is that we can much more easily try out
entirely new mapping systems now, since the dynamic portions have been
eliminated.
2011-01-15 04:30:29 +00:00
|
|
|
if((addr & 0xffffe0) == 0x00ffe0) {
|
|
|
|
|
if(addr == 0xffea && sa1.mmio.cpu_nvsw) return sa1.mmio.snv >> 0;
|
|
|
|
|
if(addr == 0xffeb && sa1.mmio.cpu_nvsw) return sa1.mmio.snv >> 8;
|
|
|
|
|
if(addr == 0xffee && sa1.mmio.cpu_ivsw) return sa1.mmio.siv >> 0;
|
|
|
|
|
if(addr == 0xffef && sa1.mmio.cpu_ivsw) return sa1.mmio.siv >> 8;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static auto read = [](unsigned addr) {
|
2012-07-08 02:57:34 +00:00
|
|
|
return sa1.rom.read(bus.mirror(addr, sa1.rom.size()));
|
Update to v074r03 release.
byuu says:
You guys are going to hate the hell out of this one. It's twenty hours
of non-stop work, no exaggeration at all. Started at 4AM, just wrapped
up now at 8PM.
I rewrote the entire memory subsystem.
Old system:
65536 pages that map 256 bytes each
Mapping a new page overwrites old page
Granularity capped at 256 bytes minimum, requiring ST-001x to map
60:0000-00ff instead of 60:0000,0001
Classes inherit from MMIO and Memory, forcing only one mappable function
per class, and fixed names
MMIO sub-mapper inside memory: 00-3f:2000-5fff for one-byte granularity
Can dynamically change the map at run-time, MMC register settings
perform dynamic remapping
New system:
XML mapping is still based around banklo-bankhi:addrlo-addrhi, as that
shapes almost everything on the SNES very well
Internally, 2048 pages that map 8192 bytes each
Pages are vectors, scans O(n) from last to first (O(log n) would not
help, n is never > 3)
Can multi-cast writes, but not reads [for the obvious reason of: which
read do you return?]
Can map reads and writes separately
Granularity of one for entire 24-bit address range, no need for MMIO
- whatever is in XML is exactly what you get
Read/Write tables bind function callbacks, so I can have any number of
functions with any names from any classes with no inheritance (no
more uPD7725DR, uPD7725SR helpers, etc)
Less memory usage overall due to less tables [ I tried 16 million tables
and it used 2GB of RAM >_o ]
Cannot dynamically change the map at run-time, MMC read/write functions
perform address translation [worse average case speed, better worst
case speed]
Now the hate me part, functors can't beat virtual functions for speed.
There are speed penalties involved:
-4.5% on average games
-11% on SuperFX games (SFX has its own bus)
-15% on SA-1 games (SA-1 has two buses)
Of course the two that need the speed the most get the biggest hits.
I'm afraid there's really not a lot of wiggle room to boost speed back
up.
I suppose one bright spot is that we can much more easily try out
entirely new mapping systems now, since the dynamic portions have been
eliminated.
2011-01-15 04:30:29 +00:00
|
|
|
};
|
|
|
|
|
|
|
|
|
|
if((addr & 0xe08000) == 0x008000) { //$00-1f:8000-ffff
|
|
|
|
|
addr = ((addr & 0x1f0000) >> 1) | (addr & 0x007fff);
|
|
|
|
|
if(mmio.cbmode == 0) return read(0x000000 | addr);
|
|
|
|
|
return read((mmio.cb << 20) | addr);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if((addr & 0xe08000) == 0x208000) { //$20-3f:8000-ffff
|
|
|
|
|
addr = ((addr & 0x1f0000) >> 1) | (addr & 0x007fff);
|
|
|
|
|
if(mmio.dbmode == 0) return read(0x100000 | addr);
|
|
|
|
|
return read((mmio.db << 20) | addr);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if((addr & 0xe08000) == 0x808000) { //$80-9f:8000-ffff
|
|
|
|
|
addr = ((addr & 0x1f0000) >> 1) | (addr & 0x007fff);
|
|
|
|
|
if(mmio.ebmode == 0) return read(0x200000 | addr);
|
|
|
|
|
return read((mmio.eb << 20) | addr);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if((addr & 0xe08000) == 0xa08000) { //$a0-bf:8000-ffff
|
|
|
|
|
addr = ((addr & 0x1f0000) >> 1) | (addr & 0x007fff);
|
|
|
|
|
if(mmio.fbmode == 0) return read(0x300000 | addr);
|
|
|
|
|
return read((mmio.fb << 20) | addr);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if((addr & 0xf00000) == 0xc00000) { //$c0-cf:0000-ffff
|
|
|
|
|
return read((mmio.cb << 20) | (addr & 0x0fffff));
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if((addr & 0xf00000) == 0xd00000) { //$d0-df:0000-ffff
|
|
|
|
|
return read((mmio.db << 20) | (addr & 0x0fffff));
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if((addr & 0xf00000) == 0xe00000) { //$e0-ef:0000-ffff
|
|
|
|
|
return read((mmio.eb << 20) | (addr & 0x0fffff));
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if((addr & 0xf00000) == 0xf00000) { //$f0-ff:0000-ffff
|
|
|
|
|
return read((mmio.fb << 20) | (addr & 0x0fffff));
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
return 0x00;
|
|
|
|
|
}
|
|
|
|
|
|
Update to v089r17 release.
byuu says:
This implements the spec from the XML part 3 thread:
http://board.byuu.org/viewtopic.php?f=16&t=2998
It's also propagated the changes to nall and purify, so you can test
this one.
This is basically it, after years of effort I feel I finally have
a fully consistent and logical XML board format.
The only things left to change will be: modifications if emulation turns
out to be incorrect (eg we missed some MMIO mirrors, or mirrored too
much), and new additions.
And of course, I'm giving it a bit of time for good arguments against
the format.
Other than that, this release removes linear_vector and pointer_vector,
as vector is better than linear_vector and I've never used
pointer_vector.
vector also gets move(), which is a way to use move-semantics across
types. It lets you steal the underlying memory pool, effectively
destroying the vector without a copy.
This works really nicely with the move for read() functions to return
vector<uint8> instead of taking (uint8_t*&, unsigned&) parameters.
2012-07-15 13:02:27 +00:00
|
|
|
void SA1::mmcrom_write(unsigned addr, uint8 data) {
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
uint8 SA1::mmcbwram_read(unsigned addr) {
|
|
|
|
|
if((addr & 0x40e000) == 0x006000) { //$00-3f|80-bf:6000-7fff
|
|
|
|
|
cpu.synchronize_coprocessors();
|
|
|
|
|
addr = bus.mirror(mmio.sbm * 0x2000 + (addr & 0x1fff), cpubwram.size());
|
|
|
|
|
return cpubwram.read(addr);
|
2012-07-08 02:57:34 +00:00
|
|
|
}
|
Update to v074r03 release.
byuu says:
You guys are going to hate the hell out of this one. It's twenty hours
of non-stop work, no exaggeration at all. Started at 4AM, just wrapped
up now at 8PM.
I rewrote the entire memory subsystem.
Old system:
65536 pages that map 256 bytes each
Mapping a new page overwrites old page
Granularity capped at 256 bytes minimum, requiring ST-001x to map
60:0000-00ff instead of 60:0000,0001
Classes inherit from MMIO and Memory, forcing only one mappable function
per class, and fixed names
MMIO sub-mapper inside memory: 00-3f:2000-5fff for one-byte granularity
Can dynamically change the map at run-time, MMC register settings
perform dynamic remapping
New system:
XML mapping is still based around banklo-bankhi:addrlo-addrhi, as that
shapes almost everything on the SNES very well
Internally, 2048 pages that map 8192 bytes each
Pages are vectors, scans O(n) from last to first (O(log n) would not
help, n is never > 3)
Can multi-cast writes, but not reads [for the obvious reason of: which
read do you return?]
Can map reads and writes separately
Granularity of one for entire 24-bit address range, no need for MMIO
- whatever is in XML is exactly what you get
Read/Write tables bind function callbacks, so I can have any number of
functions with any names from any classes with no inheritance (no
more uPD7725DR, uPD7725SR helpers, etc)
Less memory usage overall due to less tables [ I tried 16 million tables
and it used 2GB of RAM >_o ]
Cannot dynamically change the map at run-time, MMC read/write functions
perform address translation [worse average case speed, better worst
case speed]
Now the hate me part, functors can't beat virtual functions for speed.
There are speed penalties involved:
-4.5% on average games
-11% on SuperFX games (SFX has its own bus)
-15% on SA-1 games (SA-1 has two buses)
Of course the two that need the speed the most get the biggest hits.
I'm afraid there's really not a lot of wiggle room to boost speed back
up.
I suppose one bright spot is that we can much more easily try out
entirely new mapping systems now, since the dynamic portions have been
eliminated.
2011-01-15 04:30:29 +00:00
|
|
|
|
Update to v089r17 release.
byuu says:
This implements the spec from the XML part 3 thread:
http://board.byuu.org/viewtopic.php?f=16&t=2998
It's also propagated the changes to nall and purify, so you can test
this one.
This is basically it, after years of effort I feel I finally have
a fully consistent and logical XML board format.
The only things left to change will be: modifications if emulation turns
out to be incorrect (eg we missed some MMIO mirrors, or mirrored too
much), and new additions.
And of course, I'm giving it a bit of time for good arguments against
the format.
Other than that, this release removes linear_vector and pointer_vector,
as vector is better than linear_vector and I've never used
pointer_vector.
vector also gets move(), which is a way to use move-semantics across
types. It lets you steal the underlying memory pool, effectively
destroying the vector without a copy.
This works really nicely with the move for read() functions to return
vector<uint8> instead of taking (uint8_t*&, unsigned&) parameters.
2012-07-15 13:02:27 +00:00
|
|
|
if((addr & 0xf00000) == 0x400000) { //$40-4f:0000-ffff
|
|
|
|
|
return cpubwram.read(addr & 0x0fffff);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
return cpu.regs.mdr;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
void SA1::mmcbwram_write(unsigned addr, uint8 data) {
|
2012-07-08 02:57:34 +00:00
|
|
|
if((addr & 0x40e000) == 0x006000) { //$00-3f|80-bf:6000-7fff
|
|
|
|
|
cpu.synchronize_coprocessors();
|
|
|
|
|
addr = bus.mirror(mmio.sbm * 0x2000 + (addr & 0x1fff), cpubwram.size());
|
|
|
|
|
return cpubwram.write(addr, data);
|
|
|
|
|
}
|
Update to v074r03 release.
byuu says:
You guys are going to hate the hell out of this one. It's twenty hours
of non-stop work, no exaggeration at all. Started at 4AM, just wrapped
up now at 8PM.
I rewrote the entire memory subsystem.
Old system:
65536 pages that map 256 bytes each
Mapping a new page overwrites old page
Granularity capped at 256 bytes minimum, requiring ST-001x to map
60:0000-00ff instead of 60:0000,0001
Classes inherit from MMIO and Memory, forcing only one mappable function
per class, and fixed names
MMIO sub-mapper inside memory: 00-3f:2000-5fff for one-byte granularity
Can dynamically change the map at run-time, MMC register settings
perform dynamic remapping
New system:
XML mapping is still based around banklo-bankhi:addrlo-addrhi, as that
shapes almost everything on the SNES very well
Internally, 2048 pages that map 8192 bytes each
Pages are vectors, scans O(n) from last to first (O(log n) would not
help, n is never > 3)
Can multi-cast writes, but not reads [for the obvious reason of: which
read do you return?]
Can map reads and writes separately
Granularity of one for entire 24-bit address range, no need for MMIO
- whatever is in XML is exactly what you get
Read/Write tables bind function callbacks, so I can have any number of
functions with any names from any classes with no inheritance (no
more uPD7725DR, uPD7725SR helpers, etc)
Less memory usage overall due to less tables [ I tried 16 million tables
and it used 2GB of RAM >_o ]
Cannot dynamically change the map at run-time, MMC read/write functions
perform address translation [worse average case speed, better worst
case speed]
Now the hate me part, functors can't beat virtual functions for speed.
There are speed penalties involved:
-4.5% on average games
-11% on SuperFX games (SFX has its own bus)
-15% on SA-1 games (SA-1 has two buses)
Of course the two that need the speed the most get the biggest hits.
I'm afraid there's really not a lot of wiggle room to boost speed back
up.
I suppose one bright spot is that we can much more easily try out
entirely new mapping systems now, since the dynamic portions have been
eliminated.
2011-01-15 04:30:29 +00:00
|
|
|
|
2012-07-08 02:57:34 +00:00
|
|
|
if((addr & 0xf00000) == 0x400000) { //$40-4f:0000-ffff
|
|
|
|
|
return cpubwram.write(addr & 0x0fffff, data);
|
|
|
|
|
}
|
Update to v074r03 release.
byuu says:
You guys are going to hate the hell out of this one. It's twenty hours
of non-stop work, no exaggeration at all. Started at 4AM, just wrapped
up now at 8PM.
I rewrote the entire memory subsystem.
Old system:
65536 pages that map 256 bytes each
Mapping a new page overwrites old page
Granularity capped at 256 bytes minimum, requiring ST-001x to map
60:0000-00ff instead of 60:0000,0001
Classes inherit from MMIO and Memory, forcing only one mappable function
per class, and fixed names
MMIO sub-mapper inside memory: 00-3f:2000-5fff for one-byte granularity
Can dynamically change the map at run-time, MMC register settings
perform dynamic remapping
New system:
XML mapping is still based around banklo-bankhi:addrlo-addrhi, as that
shapes almost everything on the SNES very well
Internally, 2048 pages that map 8192 bytes each
Pages are vectors, scans O(n) from last to first (O(log n) would not
help, n is never > 3)
Can multi-cast writes, but not reads [for the obvious reason of: which
read do you return?]
Can map reads and writes separately
Granularity of one for entire 24-bit address range, no need for MMIO
- whatever is in XML is exactly what you get
Read/Write tables bind function callbacks, so I can have any number of
functions with any names from any classes with no inheritance (no
more uPD7725DR, uPD7725SR helpers, etc)
Less memory usage overall due to less tables [ I tried 16 million tables
and it used 2GB of RAM >_o ]
Cannot dynamically change the map at run-time, MMC read/write functions
perform address translation [worse average case speed, better worst
case speed]
Now the hate me part, functors can't beat virtual functions for speed.
There are speed penalties involved:
-4.5% on average games
-11% on SuperFX games (SFX has its own bus)
-15% on SA-1 games (SA-1 has two buses)
Of course the two that need the speed the most get the biggest hits.
I'm afraid there's really not a lot of wiggle room to boost speed back
up.
I suppose one bright spot is that we can much more easily try out
entirely new mapping systems now, since the dynamic portions have been
eliminated.
2011-01-15 04:30:29 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
uint8 SA1::mmc_sa1_read(unsigned addr) {
|
|
|
|
|
synchronize_cpu();
|
|
|
|
|
if(mmio.sw46 == 0) {
|
|
|
|
|
//$40-43:0000-ffff x 32 projection
|
2012-07-08 02:57:34 +00:00
|
|
|
addr = bus.mirror((mmio.cbm & 0x1f) * 0x2000 + (addr & 0x1fff), bwram.size());
|
|
|
|
|
return bwram.read(addr);
|
Update to v074r03 release.
byuu says:
You guys are going to hate the hell out of this one. It's twenty hours
of non-stop work, no exaggeration at all. Started at 4AM, just wrapped
up now at 8PM.
I rewrote the entire memory subsystem.
Old system:
65536 pages that map 256 bytes each
Mapping a new page overwrites old page
Granularity capped at 256 bytes minimum, requiring ST-001x to map
60:0000-00ff instead of 60:0000,0001
Classes inherit from MMIO and Memory, forcing only one mappable function
per class, and fixed names
MMIO sub-mapper inside memory: 00-3f:2000-5fff for one-byte granularity
Can dynamically change the map at run-time, MMC register settings
perform dynamic remapping
New system:
XML mapping is still based around banklo-bankhi:addrlo-addrhi, as that
shapes almost everything on the SNES very well
Internally, 2048 pages that map 8192 bytes each
Pages are vectors, scans O(n) from last to first (O(log n) would not
help, n is never > 3)
Can multi-cast writes, but not reads [for the obvious reason of: which
read do you return?]
Can map reads and writes separately
Granularity of one for entire 24-bit address range, no need for MMIO
- whatever is in XML is exactly what you get
Read/Write tables bind function callbacks, so I can have any number of
functions with any names from any classes with no inheritance (no
more uPD7725DR, uPD7725SR helpers, etc)
Less memory usage overall due to less tables [ I tried 16 million tables
and it used 2GB of RAM >_o ]
Cannot dynamically change the map at run-time, MMC read/write functions
perform address translation [worse average case speed, better worst
case speed]
Now the hate me part, functors can't beat virtual functions for speed.
There are speed penalties involved:
-4.5% on average games
-11% on SuperFX games (SFX has its own bus)
-15% on SA-1 games (SA-1 has two buses)
Of course the two that need the speed the most get the biggest hits.
I'm afraid there's really not a lot of wiggle room to boost speed back
up.
I suppose one bright spot is that we can much more easily try out
entirely new mapping systems now, since the dynamic portions have been
eliminated.
2011-01-15 04:30:29 +00:00
|
|
|
} else {
|
|
|
|
|
//$60-6f:0000-ffff x 128 projection
|
Update to v074r11 release.
byuu says:
Changelog:
- debugger compiles on all three profiles
- libsnes compiles on all three platforms (no API changes to libsnes)
- memory.cpp : namespace memory removed (wram -> cpu, apuram -> smp,
vram, oam, cgram -> ppu)
- sa1.cpp : namespace memory removed (SA-1 specific functions merged
inline to SA1::bus_read,write)
- GameBoy: added serial link support with interrupts and proper 8192hz
timing, but obviously it acts as if no other GB is connected to it
- GameBoy: added STAT OAM interrupt, and better STAT d1,d0 mode values
- UI: since Qt is dead, I've renamed the config files back to bsnes.cfg
and bsnes-geometry.cfg
- SA1: IRAM was not syncing to CPU on SA-1 side
- PPU/Accuracy and PPU/Performance needed Sprite oam renamed to Sprite
sprite; so that I could add uint8 oam[544]
- makes more sense anyway, OAM = object attribute memory, obj or
sprite are better names for Sprite rendering class
- more cleanup
2011-01-24 09:03:17 +00:00
|
|
|
addr = bus.mirror(mmio.cbm * 0x2000 + (addr & 0x1fff), 0x100000);
|
|
|
|
|
return bitmap_read(addr);
|
Update to v074r03 release.
byuu says:
You guys are going to hate the hell out of this one. It's twenty hours
of non-stop work, no exaggeration at all. Started at 4AM, just wrapped
up now at 8PM.
I rewrote the entire memory subsystem.
Old system:
65536 pages that map 256 bytes each
Mapping a new page overwrites old page
Granularity capped at 256 bytes minimum, requiring ST-001x to map
60:0000-00ff instead of 60:0000,0001
Classes inherit from MMIO and Memory, forcing only one mappable function
per class, and fixed names
MMIO sub-mapper inside memory: 00-3f:2000-5fff for one-byte granularity
Can dynamically change the map at run-time, MMC register settings
perform dynamic remapping
New system:
XML mapping is still based around banklo-bankhi:addrlo-addrhi, as that
shapes almost everything on the SNES very well
Internally, 2048 pages that map 8192 bytes each
Pages are vectors, scans O(n) from last to first (O(log n) would not
help, n is never > 3)
Can multi-cast writes, but not reads [for the obvious reason of: which
read do you return?]
Can map reads and writes separately
Granularity of one for entire 24-bit address range, no need for MMIO
- whatever is in XML is exactly what you get
Read/Write tables bind function callbacks, so I can have any number of
functions with any names from any classes with no inheritance (no
more uPD7725DR, uPD7725SR helpers, etc)
Less memory usage overall due to less tables [ I tried 16 million tables
and it used 2GB of RAM >_o ]
Cannot dynamically change the map at run-time, MMC read/write functions
perform address translation [worse average case speed, better worst
case speed]
Now the hate me part, functors can't beat virtual functions for speed.
There are speed penalties involved:
-4.5% on average games
-11% on SuperFX games (SFX has its own bus)
-15% on SA-1 games (SA-1 has two buses)
Of course the two that need the speed the most get the biggest hits.
I'm afraid there's really not a lot of wiggle room to boost speed back
up.
I suppose one bright spot is that we can much more easily try out
entirely new mapping systems now, since the dynamic portions have been
eliminated.
2011-01-15 04:30:29 +00:00
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
void SA1::mmc_sa1_write(unsigned addr, uint8 data) {
|
|
|
|
|
synchronize_cpu();
|
|
|
|
|
if(mmio.sw46 == 0) {
|
|
|
|
|
//$40-43:0000-ffff x 32 projection
|
2012-07-08 02:57:34 +00:00
|
|
|
addr = bus.mirror((mmio.cbm & 0x1f) * 0x2000 + (addr & 0x1fff), bwram.size());
|
|
|
|
|
bwram.write(addr, data);
|
Update to v074r03 release.
byuu says:
You guys are going to hate the hell out of this one. It's twenty hours
of non-stop work, no exaggeration at all. Started at 4AM, just wrapped
up now at 8PM.
I rewrote the entire memory subsystem.
Old system:
65536 pages that map 256 bytes each
Mapping a new page overwrites old page
Granularity capped at 256 bytes minimum, requiring ST-001x to map
60:0000-00ff instead of 60:0000,0001
Classes inherit from MMIO and Memory, forcing only one mappable function
per class, and fixed names
MMIO sub-mapper inside memory: 00-3f:2000-5fff for one-byte granularity
Can dynamically change the map at run-time, MMC register settings
perform dynamic remapping
New system:
XML mapping is still based around banklo-bankhi:addrlo-addrhi, as that
shapes almost everything on the SNES very well
Internally, 2048 pages that map 8192 bytes each
Pages are vectors, scans O(n) from last to first (O(log n) would not
help, n is never > 3)
Can multi-cast writes, but not reads [for the obvious reason of: which
read do you return?]
Can map reads and writes separately
Granularity of one for entire 24-bit address range, no need for MMIO
- whatever is in XML is exactly what you get
Read/Write tables bind function callbacks, so I can have any number of
functions with any names from any classes with no inheritance (no
more uPD7725DR, uPD7725SR helpers, etc)
Less memory usage overall due to less tables [ I tried 16 million tables
and it used 2GB of RAM >_o ]
Cannot dynamically change the map at run-time, MMC read/write functions
perform address translation [worse average case speed, better worst
case speed]
Now the hate me part, functors can't beat virtual functions for speed.
There are speed penalties involved:
-4.5% on average games
-11% on SuperFX games (SFX has its own bus)
-15% on SA-1 games (SA-1 has two buses)
Of course the two that need the speed the most get the biggest hits.
I'm afraid there's really not a lot of wiggle room to boost speed back
up.
I suppose one bright spot is that we can much more easily try out
entirely new mapping systems now, since the dynamic portions have been
eliminated.
2011-01-15 04:30:29 +00:00
|
|
|
} else {
|
|
|
|
|
//$60-6f:0000-ffff x 128 projection
|
Update to v074r11 release.
byuu says:
Changelog:
- debugger compiles on all three profiles
- libsnes compiles on all three platforms (no API changes to libsnes)
- memory.cpp : namespace memory removed (wram -> cpu, apuram -> smp,
vram, oam, cgram -> ppu)
- sa1.cpp : namespace memory removed (SA-1 specific functions merged
inline to SA1::bus_read,write)
- GameBoy: added serial link support with interrupts and proper 8192hz
timing, but obviously it acts as if no other GB is connected to it
- GameBoy: added STAT OAM interrupt, and better STAT d1,d0 mode values
- UI: since Qt is dead, I've renamed the config files back to bsnes.cfg
and bsnes-geometry.cfg
- SA1: IRAM was not syncing to CPU on SA-1 side
- PPU/Accuracy and PPU/Performance needed Sprite oam renamed to Sprite
sprite; so that I could add uint8 oam[544]
- makes more sense anyway, OAM = object attribute memory, obj or
sprite are better names for Sprite rendering class
- more cleanup
2011-01-24 09:03:17 +00:00
|
|
|
addr = bus.mirror(mmio.cbm * 0x2000 + (addr & 0x1fff), 0x100000);
|
|
|
|
|
bitmap_write(addr, data);
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
uint8 SA1::bitmap_read(unsigned addr) {
|
|
|
|
|
if(mmio.bbf == 0) {
|
|
|
|
|
//4bpp
|
|
|
|
|
unsigned shift = addr & 1;
|
2012-07-08 02:57:34 +00:00
|
|
|
addr = (addr >> 1) & (bwram.size() - 1);
|
Update to v074r11 release.
byuu says:
Changelog:
- debugger compiles on all three profiles
- libsnes compiles on all three platforms (no API changes to libsnes)
- memory.cpp : namespace memory removed (wram -> cpu, apuram -> smp,
vram, oam, cgram -> ppu)
- sa1.cpp : namespace memory removed (SA-1 specific functions merged
inline to SA1::bus_read,write)
- GameBoy: added serial link support with interrupts and proper 8192hz
timing, but obviously it acts as if no other GB is connected to it
- GameBoy: added STAT OAM interrupt, and better STAT d1,d0 mode values
- UI: since Qt is dead, I've renamed the config files back to bsnes.cfg
and bsnes-geometry.cfg
- SA1: IRAM was not syncing to CPU on SA-1 side
- PPU/Accuracy and PPU/Performance needed Sprite oam renamed to Sprite
sprite; so that I could add uint8 oam[544]
- makes more sense anyway, OAM = object attribute memory, obj or
sprite are better names for Sprite rendering class
- more cleanup
2011-01-24 09:03:17 +00:00
|
|
|
switch(shift) { default:
|
2012-07-08 02:57:34 +00:00
|
|
|
case 0: return (bwram.read(addr) >> 0) & 15;
|
|
|
|
|
case 1: return (bwram.read(addr) >> 4) & 15;
|
Update to v074r11 release.
byuu says:
Changelog:
- debugger compiles on all three profiles
- libsnes compiles on all three platforms (no API changes to libsnes)
- memory.cpp : namespace memory removed (wram -> cpu, apuram -> smp,
vram, oam, cgram -> ppu)
- sa1.cpp : namespace memory removed (SA-1 specific functions merged
inline to SA1::bus_read,write)
- GameBoy: added serial link support with interrupts and proper 8192hz
timing, but obviously it acts as if no other GB is connected to it
- GameBoy: added STAT OAM interrupt, and better STAT d1,d0 mode values
- UI: since Qt is dead, I've renamed the config files back to bsnes.cfg
and bsnes-geometry.cfg
- SA1: IRAM was not syncing to CPU on SA-1 side
- PPU/Accuracy and PPU/Performance needed Sprite oam renamed to Sprite
sprite; so that I could add uint8 oam[544]
- makes more sense anyway, OAM = object attribute memory, obj or
sprite are better names for Sprite rendering class
- more cleanup
2011-01-24 09:03:17 +00:00
|
|
|
}
|
|
|
|
|
} else {
|
|
|
|
|
//2bpp
|
|
|
|
|
unsigned shift = addr & 3;
|
2012-07-08 02:57:34 +00:00
|
|
|
addr = (addr >> 2) & (bwram.size() - 1);
|
Update to v074r11 release.
byuu says:
Changelog:
- debugger compiles on all three profiles
- libsnes compiles on all three platforms (no API changes to libsnes)
- memory.cpp : namespace memory removed (wram -> cpu, apuram -> smp,
vram, oam, cgram -> ppu)
- sa1.cpp : namespace memory removed (SA-1 specific functions merged
inline to SA1::bus_read,write)
- GameBoy: added serial link support with interrupts and proper 8192hz
timing, but obviously it acts as if no other GB is connected to it
- GameBoy: added STAT OAM interrupt, and better STAT d1,d0 mode values
- UI: since Qt is dead, I've renamed the config files back to bsnes.cfg
and bsnes-geometry.cfg
- SA1: IRAM was not syncing to CPU on SA-1 side
- PPU/Accuracy and PPU/Performance needed Sprite oam renamed to Sprite
sprite; so that I could add uint8 oam[544]
- makes more sense anyway, OAM = object attribute memory, obj or
sprite are better names for Sprite rendering class
- more cleanup
2011-01-24 09:03:17 +00:00
|
|
|
switch(shift) { default:
|
2012-07-08 02:57:34 +00:00
|
|
|
case 0: return (bwram.read(addr) >> 0) & 3;
|
|
|
|
|
case 1: return (bwram.read(addr) >> 2) & 3;
|
|
|
|
|
case 2: return (bwram.read(addr) >> 4) & 3;
|
|
|
|
|
case 3: return (bwram.read(addr) >> 6) & 3;
|
Update to v074r11 release.
byuu says:
Changelog:
- debugger compiles on all three profiles
- libsnes compiles on all three platforms (no API changes to libsnes)
- memory.cpp : namespace memory removed (wram -> cpu, apuram -> smp,
vram, oam, cgram -> ppu)
- sa1.cpp : namespace memory removed (SA-1 specific functions merged
inline to SA1::bus_read,write)
- GameBoy: added serial link support with interrupts and proper 8192hz
timing, but obviously it acts as if no other GB is connected to it
- GameBoy: added STAT OAM interrupt, and better STAT d1,d0 mode values
- UI: since Qt is dead, I've renamed the config files back to bsnes.cfg
and bsnes-geometry.cfg
- SA1: IRAM was not syncing to CPU on SA-1 side
- PPU/Accuracy and PPU/Performance needed Sprite oam renamed to Sprite
sprite; so that I could add uint8 oam[544]
- makes more sense anyway, OAM = object attribute memory, obj or
sprite are better names for Sprite rendering class
- more cleanup
2011-01-24 09:03:17 +00:00
|
|
|
}
|
Update to v074r03 release.
byuu says:
You guys are going to hate the hell out of this one. It's twenty hours
of non-stop work, no exaggeration at all. Started at 4AM, just wrapped
up now at 8PM.
I rewrote the entire memory subsystem.
Old system:
65536 pages that map 256 bytes each
Mapping a new page overwrites old page
Granularity capped at 256 bytes minimum, requiring ST-001x to map
60:0000-00ff instead of 60:0000,0001
Classes inherit from MMIO and Memory, forcing only one mappable function
per class, and fixed names
MMIO sub-mapper inside memory: 00-3f:2000-5fff for one-byte granularity
Can dynamically change the map at run-time, MMC register settings
perform dynamic remapping
New system:
XML mapping is still based around banklo-bankhi:addrlo-addrhi, as that
shapes almost everything on the SNES very well
Internally, 2048 pages that map 8192 bytes each
Pages are vectors, scans O(n) from last to first (O(log n) would not
help, n is never > 3)
Can multi-cast writes, but not reads [for the obvious reason of: which
read do you return?]
Can map reads and writes separately
Granularity of one for entire 24-bit address range, no need for MMIO
- whatever is in XML is exactly what you get
Read/Write tables bind function callbacks, so I can have any number of
functions with any names from any classes with no inheritance (no
more uPD7725DR, uPD7725SR helpers, etc)
Less memory usage overall due to less tables [ I tried 16 million tables
and it used 2GB of RAM >_o ]
Cannot dynamically change the map at run-time, MMC read/write functions
perform address translation [worse average case speed, better worst
case speed]
Now the hate me part, functors can't beat virtual functions for speed.
There are speed penalties involved:
-4.5% on average games
-11% on SuperFX games (SFX has its own bus)
-15% on SA-1 games (SA-1 has two buses)
Of course the two that need the speed the most get the biggest hits.
I'm afraid there's really not a lot of wiggle room to boost speed back
up.
I suppose one bright spot is that we can much more easily try out
entirely new mapping systems now, since the dynamic portions have been
eliminated.
2011-01-15 04:30:29 +00:00
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
Update to v074r11 release.
byuu says:
Changelog:
- debugger compiles on all three profiles
- libsnes compiles on all three platforms (no API changes to libsnes)
- memory.cpp : namespace memory removed (wram -> cpu, apuram -> smp,
vram, oam, cgram -> ppu)
- sa1.cpp : namespace memory removed (SA-1 specific functions merged
inline to SA1::bus_read,write)
- GameBoy: added serial link support with interrupts and proper 8192hz
timing, but obviously it acts as if no other GB is connected to it
- GameBoy: added STAT OAM interrupt, and better STAT d1,d0 mode values
- UI: since Qt is dead, I've renamed the config files back to bsnes.cfg
and bsnes-geometry.cfg
- SA1: IRAM was not syncing to CPU on SA-1 side
- PPU/Accuracy and PPU/Performance needed Sprite oam renamed to Sprite
sprite; so that I could add uint8 oam[544]
- makes more sense anyway, OAM = object attribute memory, obj or
sprite are better names for Sprite rendering class
- more cleanup
2011-01-24 09:03:17 +00:00
|
|
|
void SA1::bitmap_write(unsigned addr, uint8 data) {
|
|
|
|
|
if(mmio.bbf == 0) {
|
|
|
|
|
//4bpp
|
|
|
|
|
unsigned shift = addr & 1;
|
2012-07-08 02:57:34 +00:00
|
|
|
addr = (addr >> 1) & (bwram.size() - 1);
|
Update to v074r11 release.
byuu says:
Changelog:
- debugger compiles on all three profiles
- libsnes compiles on all three platforms (no API changes to libsnes)
- memory.cpp : namespace memory removed (wram -> cpu, apuram -> smp,
vram, oam, cgram -> ppu)
- sa1.cpp : namespace memory removed (SA-1 specific functions merged
inline to SA1::bus_read,write)
- GameBoy: added serial link support with interrupts and proper 8192hz
timing, but obviously it acts as if no other GB is connected to it
- GameBoy: added STAT OAM interrupt, and better STAT d1,d0 mode values
- UI: since Qt is dead, I've renamed the config files back to bsnes.cfg
and bsnes-geometry.cfg
- SA1: IRAM was not syncing to CPU on SA-1 side
- PPU/Accuracy and PPU/Performance needed Sprite oam renamed to Sprite
sprite; so that I could add uint8 oam[544]
- makes more sense anyway, OAM = object attribute memory, obj or
sprite are better names for Sprite rendering class
- more cleanup
2011-01-24 09:03:17 +00:00
|
|
|
switch(shift) { default:
|
2012-07-08 02:57:34 +00:00
|
|
|
case 0: data = (bwram.read(addr) & 0xf0) | ((data & 15) << 0); break;
|
|
|
|
|
case 1: data = (bwram.read(addr) & 0x0f) | ((data & 15) << 4); break;
|
Update to v074r11 release.
byuu says:
Changelog:
- debugger compiles on all three profiles
- libsnes compiles on all three platforms (no API changes to libsnes)
- memory.cpp : namespace memory removed (wram -> cpu, apuram -> smp,
vram, oam, cgram -> ppu)
- sa1.cpp : namespace memory removed (SA-1 specific functions merged
inline to SA1::bus_read,write)
- GameBoy: added serial link support with interrupts and proper 8192hz
timing, but obviously it acts as if no other GB is connected to it
- GameBoy: added STAT OAM interrupt, and better STAT d1,d0 mode values
- UI: since Qt is dead, I've renamed the config files back to bsnes.cfg
and bsnes-geometry.cfg
- SA1: IRAM was not syncing to CPU on SA-1 side
- PPU/Accuracy and PPU/Performance needed Sprite oam renamed to Sprite
sprite; so that I could add uint8 oam[544]
- makes more sense anyway, OAM = object attribute memory, obj or
sprite are better names for Sprite rendering class
- more cleanup
2011-01-24 09:03:17 +00:00
|
|
|
}
|
|
|
|
|
} else {
|
|
|
|
|
//2bpp
|
|
|
|
|
unsigned shift = addr & 3;
|
2012-07-08 02:57:34 +00:00
|
|
|
addr = (addr >> 2) & (bwram.size() - 1);
|
Update to v074r11 release.
byuu says:
Changelog:
- debugger compiles on all three profiles
- libsnes compiles on all three platforms (no API changes to libsnes)
- memory.cpp : namespace memory removed (wram -> cpu, apuram -> smp,
vram, oam, cgram -> ppu)
- sa1.cpp : namespace memory removed (SA-1 specific functions merged
inline to SA1::bus_read,write)
- GameBoy: added serial link support with interrupts and proper 8192hz
timing, but obviously it acts as if no other GB is connected to it
- GameBoy: added STAT OAM interrupt, and better STAT d1,d0 mode values
- UI: since Qt is dead, I've renamed the config files back to bsnes.cfg
and bsnes-geometry.cfg
- SA1: IRAM was not syncing to CPU on SA-1 side
- PPU/Accuracy and PPU/Performance needed Sprite oam renamed to Sprite
sprite; so that I could add uint8 oam[544]
- makes more sense anyway, OAM = object attribute memory, obj or
sprite are better names for Sprite rendering class
- more cleanup
2011-01-24 09:03:17 +00:00
|
|
|
switch(shift) { default:
|
2012-07-08 02:57:34 +00:00
|
|
|
case 0: data = (bwram.read(addr) & 0xfc) | ((data & 3) << 0); break;
|
|
|
|
|
case 1: data = (bwram.read(addr) & 0xf3) | ((data & 3) << 2); break;
|
|
|
|
|
case 2: data = (bwram.read(addr) & 0xcf) | ((data & 3) << 4); break;
|
|
|
|
|
case 3: data = (bwram.read(addr) & 0x3f) | ((data & 3) << 6); break;
|
Update to v074r11 release.
byuu says:
Changelog:
- debugger compiles on all three profiles
- libsnes compiles on all three platforms (no API changes to libsnes)
- memory.cpp : namespace memory removed (wram -> cpu, apuram -> smp,
vram, oam, cgram -> ppu)
- sa1.cpp : namespace memory removed (SA-1 specific functions merged
inline to SA1::bus_read,write)
- GameBoy: added serial link support with interrupts and proper 8192hz
timing, but obviously it acts as if no other GB is connected to it
- GameBoy: added STAT OAM interrupt, and better STAT d1,d0 mode values
- UI: since Qt is dead, I've renamed the config files back to bsnes.cfg
and bsnes-geometry.cfg
- SA1: IRAM was not syncing to CPU on SA-1 side
- PPU/Accuracy and PPU/Performance needed Sprite oam renamed to Sprite
sprite; so that I could add uint8 oam[544]
- makes more sense anyway, OAM = object attribute memory, obj or
sprite are better names for Sprite rendering class
- more cleanup
2011-01-24 09:03:17 +00:00
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
2012-07-08 02:57:34 +00:00
|
|
|
bwram.write(addr, data);
|
Update to v074r11 release.
byuu says:
Changelog:
- debugger compiles on all three profiles
- libsnes compiles on all three platforms (no API changes to libsnes)
- memory.cpp : namespace memory removed (wram -> cpu, apuram -> smp,
vram, oam, cgram -> ppu)
- sa1.cpp : namespace memory removed (SA-1 specific functions merged
inline to SA1::bus_read,write)
- GameBoy: added serial link support with interrupts and proper 8192hz
timing, but obviously it acts as if no other GB is connected to it
- GameBoy: added STAT OAM interrupt, and better STAT d1,d0 mode values
- UI: since Qt is dead, I've renamed the config files back to bsnes.cfg
and bsnes-geometry.cfg
- SA1: IRAM was not syncing to CPU on SA-1 side
- PPU/Accuracy and PPU/Performance needed Sprite oam renamed to Sprite
sprite; so that I could add uint8 oam[544]
- makes more sense anyway, OAM = object attribute memory, obj or
sprite are better names for Sprite rendering class
- more cleanup
2011-01-24 09:03:17 +00:00
|
|
|
}
|
|
|
|
|
|
2010-08-09 13:28:56 +00:00
|
|
|
#endif
|