Sprinter/mame
2
0
Fork 0
mirror of https://github.com/holub/mame synced 2025-04-18 22:49:58 +03:00
Code Issues Actions Packages Projects Releases Wiki Activity

namcos10: Reevaluate assumption about how later blocks are handled (#11071)

Browse Source
This commit is contained in:
987123879113 2023-04-03 11:34:38 +09:00 committed by GitHub
parent 6dc6efdbb5
commit 9ef759905c
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 17 additions and 12 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

29
src/mame/namco/namcos10.cpp
View File

@ -417,8 +417,12 @@ try all permutations of 0-15 bit swaps which will take much longer or manually w
Known issues:
- Opening the operator menu sometimes can crash Mr. Driller 2
- nflclsfb: confirmed working but boots to a black screen. internal error says "namcoS10Sio0Init error :no EXIO!!!", making the check pass lets the game boot like normal
- mrdrilr2, mrdrilr2j: Opening the operator menu sometimes can crash Mr. Driller 2
- nflclsfb: confirmed working but boots to a black screen. internal error says "namcoS10Sio0Init error :no EXIO!!!", making the check pass lets the game boot like normal (patch the instruction at 8001428c to 10000006)
- knpuzzle: bad dumps? where are the encrypted program blocks?
- gjspace: bad dump? first program block is at 0x3fe even though the redirect table says 0x3ff and works if decrypted, but I can't find what should've been at 0x3fe
- mrdrilrg: bad dump? program code block isn't where expected. mrdrilrga looks good though
User data note:
- the games store settings / rankings / bookkeeping data in the first NAND ROM - at 0x4200 for MEM(N) and 0x40000 for MEM(M), the ROMs used should be defaulted where possible
@ -726,7 +730,9 @@ TIMER_DEVICE_CALLBACK_MEMBER(namcos10_state::io_update_interrupt_callback)
void namcos10_state::namcos10_map_inner(address_map &map)
{
// ram? stores block index to offset lookup table
// ram?
// Contains the NAND block table with redirected blocks generated
// based on the invalid block table found at the beginning of a game's first NAND
map(0xf500000, 0xf5fffff).ram().share("share3");
map(0xfb60000, 0xfb60003).noprw(); // ?
@ -1103,16 +1109,15 @@ uint16_t namcos10_memn_state::nand_data_r()
if (m_decrypter.found() && m_decrypter->is_active())
return m_decrypter->decrypt(data);
// There should never be scrambled data beyond 0x3ec
// Only the first device's beginning blocks are treated specially and can't contain scrambled data
// because it's guaranteed to always have important NAND layout information + EEP data
if ((BIT(m_nand_address, 8, 16) >> 5) < 0x3ec
&& (m_nand_device_idx != 0 || (m_nand_device_idx == 0 && (BIT(m_nand_address, 8, 16) >> 5) >= 0x0a)))
{
data = m_unscrambler(data ^ 0xaaaa);
}
// Block 0 is guaranteed to always be good according to the NAND datasheet, and it seems to always contain the bad block redirect table
// Block 1 appears to always be used by games as the EEPROM data block
// Dumped ROMs always have block 1 as plaintext so data doesn't appear scrambled when writing to block 1
// So treat only block 0 and block 1 as special blocks for now
// TODO: Is there a more reliable way to detect when data should be unscrambled? (m_ctrl_reg?)
if (m_nand_device_idx == 0 && (BIT(m_nand_address, 8, 16) >> 5) < 2)
return data;
return data;
return m_unscrambler(data ^ 0xaaaa);
}
void namcos10_memn_state::nand_data_w(offs_t offset, uint16_t data, uint16_t mem_mask)