Viewing Issue Advanced Details
ID Category [?] Severity [?] Reproducibility Date Submitted Last Update
02668 DIP/Input Minor Always Nov 23, 2008, 05:30 Nov 24, 2008, 02:04
Tester Tafoid View Status Public Platform
Assigned To robiza Resolution Fixed OS
Status [?] Resolved Driver
Version 0.128u3 Fixed in Version 0.128u4 Build
Fixed in Git Commit Github Pull Request #
Summary 02668: vendetta + clones: NVRAM being corrupted
Description Starting at 0.127u1, the NVRAM is being corrupted. The settings are not being saved properly and running those sets give a B16 .. BAD message which the typical F2/F3 Konami EEPROM creation trick will not fix. The game will run ok so long as there is no .NV file and no changes are made in service.
Steps To Reproduce
Additional Information Mentioned here:
http://www.mameworld.info/ubbthreads/showthreaded.php?Cat=&Number=172181
Github Commit
Flags
Regression Version 0.127u1
Affected Sets / Systems vendetta + clones
Attached Files
 
Relationships
There are no relationship linked to this issue.
Notes
7
User avatar
No.03085
Mamesick
Senior Tester
Nov 23, 2008, 07:38
edited on: Nov 23, 2008, 07:41
Someone in 0.127u1 changed a bit in the eeprom data read function. I think this is the cause but at the moment I cannot compile and verify by myself.

From:
	res |= input_port_read(machine, "EEPROM") & 0x0c;	/* test switch */
To:
	res |= input_port_read(machine, "EEPROM") & 0x0d;	/* test switch */
User avatar
No.03088
robiza
Developer
Nov 23, 2008, 10:45
you have right
i'll committ the change
User avatar
No.03089
robiza
Developer
Nov 23, 2008, 10:47
fix by mamesick
User avatar
No.03091
Fujix
Administrator
Nov 23, 2008, 15:39
Updated Mantis configuration to allow assigning Senior Tester level users to bugs.
I changed mamesick to Senior Tester and updated the report accordingly.
User avatar
No.03092
etabeta
Developer
Nov 23, 2008, 16:03
This fix is wrong. This way you're not reading anymore the eeprom bit (previously the reading was done setting res = eeprom_read_bit(); at the beginning, now the reading is done through a PORT_CUSTOM)

the correct fix is probably to set IP_ACTIVE_HIGH the 0x01 bits in both PORT_START("EEPROM")
User avatar
No.03093
Mamesick
Senior Tester
Nov 23, 2008, 17:29
Etabeta is right. The result is the same but correct fix is set to IP_ACTIVE_HIGH bit 0x01 in the eeprom input port.
User avatar
No.03098
robiza
Developer
Nov 24, 2008, 02:04
fixed with etabeta suggestions