claunia/86Box
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Ported over the AWE32 fixes from JosepMaJaz's PCem branch.

Browse Source
This commit is contained in:
OBattler
2017-06-22 23:51:57 +02:00
parent 08363e5ec2
commit 988fe85ea1
2 changed files with 58 additions and 25 deletions
Download Patch File Download Diff File Expand all files Collapse all files

79
src/SOUND/snd_emu8k.c
View File

@@ -46,14 +46,16 @@ static int32_t filt_w0[256];
static __inline int16_t EMU8K_READ(emu8k_t *emu8k, uint16_t addr)
{
addr &= 0xffffff;
if (addr < 0x80000)
addr &= EMU8K_MEM_ADDRESS_MASK;
/* TODO: I've read that the AWE64 Gold model had a 4MB Rom. It would be interesting
to find that rom and do some tests with it. */
if (addr < EMU8K_ROM_MEM_1MB_END)
return emu8k->rom[addr];
if (addr < 0x200000 || addr >= emu8k->ram_end_addr)
if (addr < EMU8K_RAM_MEM_START || addr >= emu8k->ram_end_addr)
return 0;
if (!emu8k->ram)
return 0;
return emu8k->ram[addr - 0x200000];
return emu8k->ram[addr - EMU8K_RAM_MEM_START];
}
static __inline int16_t EMU8K_READ_INTERP(emu8k_t *emu8k, uint16_t addr)
@@ -65,12 +67,20 @@ static __inline int16_t EMU8K_READ_INTERP(emu8k_t *emu8k, uint16_t addr)
static __inline void EMU8K_WRITE(emu8k_t *emu8k, uint16_t addr, uint16_t val)
{
addr &= 0xffffff;
if (emu8k->ram && addr >= 0x200000 && addr < emu8k->ram_end_addr)
emu8k->ram[addr - 0x200000] = val;
if ( !emu8k->ram || addr < EMU8K_RAM_MEM_START)
return;
/* It looks like if an application writes to a memory part outside of the available amount on the card,
it wraps, and opencubicplayer uses that to detect the amount of memory, as opposed to simply check
at the address that it has just tried to write. */
addr &= EMU8K_MEM_ADDRESS_MASK;
while (addr >= emu8k->ram_end_addr) {
addr -= emu8k->ram_end_addr - EMU8K_RAM_MEM_START;
}
emu8k->ram[addr - EMU8K_RAM_MEM_START] = val;
}
static int ff = 0;
static int voice_count = 0;
uint16_t emu8k_inw(uint16_t addr, void *p)
{
@@ -85,8 +95,11 @@ uint16_t emu8k_inw(uint16_t addr, void *p)
switch (emu8k->cur_reg)
{
case 0:
READ16(addr, emu8k->voice[emu8k->cur_voice].cpf);
return ret;
{
uint32_t var = (emu8k->voice[emu8k->cur_voice].cpf & 0xFFFF0000) | ((emu8k->voice[emu8k->cur_voice].addr >> 16) & 0xFFFF);
READ16(addr, var);
return ret;
}
case 1:
READ16(addr, emu8k->voice[emu8k->cur_voice].ptrx);
@@ -108,7 +121,7 @@ uint16_t emu8k_inw(uint16_t addr, void *p)
return ret;
case 7:
READ16(addr, emu8k->voice[emu8k->cur_voice].cpf);
READ16(addr, emu8k->voice[emu8k->cur_voice].csl);
return ret;
}
break;
@@ -118,6 +131,8 @@ uint16_t emu8k_inw(uint16_t addr, void *p)
{
case 0:
{
emu8k->voice[emu8k->cur_voice].ccca =
(emu8k->voice[emu8k->cur_voice].ccca & 0xFF000000) | ((emu8k->voice[emu8k->cur_voice].addr >> 32) & EMU8K_MEM_ADDRESS_MASK);
READ16(addr, emu8k->voice[emu8k->cur_voice].ccca);
return ret;
}
@@ -246,11 +261,12 @@ uint16_t emu8k_inw(uint16_t addr, void *p)
return 0x1c | ((emu8k->id & 0x0002) ? 0xff02 : 0);
}
break;
case 0xc02: /*Status - I think!*/
voice_count = (voice_count + 1) & 0x1f;
/* emu8k->c02_read ^= 0x1000;
pclog("Read status %04X\n", 0x803f | (voice_count << 8));*/
return 0x803f | (voice_count << 8);
case 0xc02:
/* LS five bits = channel number, next 3 bits = register number
and MS 8 bits = VLSI test register.
Impulse tracker tests the non variability of the LS byte and the variability
of the MS byte to determine that it really is an AWE32. */
return ((rand()&0xFF) << 8) | (emu8k->cur_reg << 5) | emu8k->cur_voice;
}
/* fatal("Bad EMU8K inw from %08X\n", addr);*/
return 0xffff;
@@ -258,6 +274,7 @@ uint16_t emu8k_inw(uint16_t addr, void *p)
void emu8k_outw(uint16_t addr, uint16_t val, void *p)
{
float q;
emu8k_t *emu8k = (emu8k_t *)p;
emu8k_update(emu8k);
@@ -270,6 +287,8 @@ void emu8k_outw(uint16_t addr, uint16_t val, void *p)
{
case 0:
WRITE16(addr, emu8k->voice[emu8k->cur_voice].cpf, val);
/* Ignoring any effect over writing to the "fractional address". The docs says that this value is constantly
updating, so it has no actual effect. */
return;
case 1:
@@ -286,7 +305,8 @@ void emu8k_outw(uint16_t addr, uint16_t val, void *p)
case 6:
WRITE16(addr, emu8k->voice[emu8k->cur_voice].psst, val);
emu8k->voice[emu8k->cur_voice].loop_start = (uint64_t)(emu8k->voice[emu8k->cur_voice].psst & 0xffffff) << 32;
/* TODO: Should we update only on MSB update, or this could be used as some sort of hack by applications? */
emu8k->voice[emu8k->cur_voice].loop_start = (uint64_t)(emu8k->voice[emu8k->cur_voice].psst & EMU8K_MEM_ADDRESS_MASK) << 32;
if (addr & 2)
{
emu8k->voice[emu8k->cur_voice].vol_l = val >> 8;
@@ -296,9 +316,10 @@ void emu8k_outw(uint16_t addr, uint16_t val, void *p)
return;
case 7:
WRITE16(addr, emu8k->voice[emu8k->cur_voice].cpf, val);
emu8k->voice[emu8k->cur_voice].loop_end = (uint64_t)(emu8k->voice[emu8k->cur_voice].cpf & 0xffffff) << 32;
/* pclog("emu8k_outl : write CPF %08X\n", emu8k->voice[emu8k->cur_voice].cpf);*/
WRITE16(addr, emu8k->voice[emu8k->cur_voice].csl, val);
/* TODO: Should we update only on MSB update, or this could be used as some sort of hack by applications? */
emu8k->voice[emu8k->cur_voice].loop_end = (uint64_t)(emu8k->voice[emu8k->cur_voice].csl & EMU8K_MEM_ADDRESS_MASK) << 32;
/* pclog("emu8k_outl : write CSL %08X\n", emu8k->voice[emu8k->cur_voice].csl);*/
return;
}
break;
@@ -308,7 +329,8 @@ void emu8k_outw(uint16_t addr, uint16_t val, void *p)
{
case 0:
WRITE16(addr, emu8k->voice[emu8k->cur_voice].ccca, val);
emu8k->voice[emu8k->cur_voice].addr = (uint64_t)(emu8k->voice[emu8k->cur_voice].ccca & 0xffffff) << 32;
/* TODO: Should we update only on MSB update, or this could be used as some sort of hack by applications? */
emu8k->voice[emu8k->cur_voice].addr = (uint64_t)(emu8k->voice[emu8k->cur_voice].ccca & EMU8K_MEM_ADDRESS_MASK) << 32;
/* pclog("emu8k_outl : write CCCA %08X\n", emu8k->voice[emu8k->cur_voice].ccca);*/
return;
@@ -385,10 +407,10 @@ void emu8k_outw(uint16_t addr, uint16_t val, void *p)
{
case 0:
{
float q;
WRITE16(addr, emu8k->voice[emu8k->cur_voice].ccca, val);
emu8k->voice[emu8k->cur_voice].addr = (uint64_t)(emu8k->voice[emu8k->cur_voice].ccca & 0xffffff) << 32;
emu8k->voice[emu8k->cur_voice].addr = (uint64_t)(emu8k->voice[emu8k->cur_voice].ccca & EMU8K_MEM_ADDRESS_MASK) << 32;
/* TODO: Since "fractional address" is a separate register (cpf), should we add its contents to .addr or we assume that it
is reset to zero? */
q = (float)(emu8k->voice[emu8k->cur_voice].ccca >> 28) / 15.0f;
q /= 10.0f; /*Horrible and wrong hack*/
@@ -680,8 +702,15 @@ void emu8k_init(emu8k_t *emu8k, int onboard_ram)
if (onboard_ram)
{
/* Clip to 28MB, since that's the max that we can address. */
if (onboard_ram > 0x7000) onboard_ram = 0x7000;
emu8k->ram = malloc(onboard_ram * 1024);
emu8k->ram_end_addr = 0x200000 + ((onboard_ram * 1024) / 2);
emu8k->ram_end_addr = EMU8K_RAM_MEM_START + ((onboard_ram * 1024) / 2);
}
else
{
emu8k->ram = 0;
emu8k->ram_end_addr = EMU8K_RAM_MEM_START;
}
emu8k->rom = malloc(1024 * 1024);

4
src/SOUND/snd_emu8k.h
View File

@@ -1,3 +1,7 @@
#define EMU8K_MEM_ADDRESS_MASK 0xFFFFFF
#define EMU8K_RAM_MEM_START 0x200000
#define EMU8K_ROM_MEM_1MB_END 0x80000
typedef struct emu8k_t
{
struct