stop the preliminary zelda ucode from overreading from its own internal buffers.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@3468 8ced0084-cf51-0410-be5f-012b33b47a6e
This commit is contained in:
parent
4ed15aedf5
commit
81f6d431dc
|
@ -143,6 +143,7 @@ void CUCode_Zelda::HandleMail(u32 _uMail)
|
|||
|
||||
if (m_step >= m_numSteps)
|
||||
{
|
||||
DEBUG_LOG(DSPHLE, "Executing %i-step list.", m_numSteps);
|
||||
ExecuteList();
|
||||
m_bListInProgress = false;
|
||||
}
|
||||
|
@ -159,6 +160,11 @@ void CUCode_Zelda::HandleMail(u32 _uMail)
|
|||
m_bListInProgress = true;
|
||||
m_numSteps = _uMail;
|
||||
m_step = 0;
|
||||
|
||||
// make sure we never read outside the buffer by mistake.
|
||||
// Before deleting extra reads in ExecuteList, we were getting these
|
||||
// values.
|
||||
memset(m_Buffer, 0xcc, sizeof(m_Buffer));
|
||||
}
|
||||
else
|
||||
{
|
||||
|
@ -176,9 +182,11 @@ void CUCode_Zelda::ExecuteList()
|
|||
// begin with the list
|
||||
m_readOffset = 0;
|
||||
|
||||
// First figure out what command we're dealing with.
|
||||
u32 CmdMail = Read32();
|
||||
u32 Command = (CmdMail >> 24) & 0x7f;
|
||||
u32 Sync = CmdMail >> 16;
|
||||
u16 ExtraData = CmdMail & 0xFFFF; // not yet used
|
||||
|
||||
DEBUG_LOG(DSPHLE, "==============================================================================");
|
||||
DEBUG_LOG(DSPHLE, "Zelda UCode - execute dlist (cmd: 0x%04x : sync: 0x%04x)", Command, Sync);
|
||||
|
@ -202,7 +210,11 @@ void CUCode_Zelda::ExecuteList()
|
|||
// This points to some strange data table.
|
||||
DEBUG_LOG(DSPHLE, "DSPRES_FILTER (size: 0x40): 0x%08x", tmp[1]);
|
||||
|
||||
// See 0x02, same thing as DSPADPCM_FILTER there.
|
||||
// Zelda WW: This points to a 64-byte array of coefficients, which are EXACTLY the same
|
||||
// as the AFC ADPCM coef array in decode.c of the in_cube winamp plugin,
|
||||
// which can play Zelda audio.
|
||||
// There's also a lot more table-looking data immediately after - maybe alternative
|
||||
// tables? I wonder where the parameter blocks are?
|
||||
DEBUG_LOG(DSPHLE, "DSPADPCM_FILTER (size: 0x500): 0x%08x", tmp[2]);
|
||||
DEBUG_LOG(DSPHLE, "Some other mixing buffer: 0x%08x", tmp[3]);
|
||||
}
|
||||
|
@ -214,7 +226,6 @@ void CUCode_Zelda::ExecuteList()
|
|||
u32 tmp[3];
|
||||
tmp[0] = Read32();
|
||||
tmp[1] = Read32();
|
||||
tmp[2] = Read32();
|
||||
|
||||
// We're ready to mix
|
||||
// soundStream->GetMixer()->SetHLEReady(true);
|
||||
|
@ -232,13 +243,6 @@ void CUCode_Zelda::ExecuteList()
|
|||
// but not at, the ADMA read addresses.
|
||||
DEBUG_LOG(DSPHLE, "Left mixing buffer? 0x%08x", tmp[0]);
|
||||
DEBUG_LOG(DSPHLE, "Right mixing buffer? 0x%08x", tmp[1]);
|
||||
|
||||
// Zelda WW: This points to a 64-byte array of coefficients, which are EXACTLY the same
|
||||
// as the AFC ADPCM coef array in decode.c of the in_cube winamp plugin,
|
||||
// which can play Zelda audio.
|
||||
// There's also a lot more table-looking data immediately after - maybe alternative
|
||||
// tables? I wonder where the parameter blocks are?
|
||||
DEBUG_LOG(DSPHLE, "DSPADPCM_FILTER (size: 0x40): 0x%08x", tmp[2]);
|
||||
}
|
||||
return;
|
||||
break;
|
||||
|
@ -259,11 +263,9 @@ void CUCode_Zelda::ExecuteList()
|
|||
{
|
||||
u32 tmp[2];
|
||||
tmp[0] = Read32();
|
||||
tmp[1] = Read32();
|
||||
|
||||
DEBUG_LOG(DSPHLE, "DSetDolbyDelay");
|
||||
DEBUG_LOG(DSPHLE, "DOLBY2_DELAY_BUF (size 0x960): 0x%08x", tmp[0]);
|
||||
DEBUG_LOG(DSPHLE, "DSPRES_FILTER (size 0x500): 0x%08x", tmp[1]);
|
||||
}
|
||||
break;
|
||||
|
||||
|
|
Loading…
Reference in New Issue