07731: All sets in zr107.cpp: Mask rom tests as bad

ID	Category [?]	Severity [?]	Reproducibility	Date Submitted	Last Update
07731	Misc.	Trivial	Always	Aug 19, 2020, 23:21	Sep 4, 2020, 20:42

Tester	Hydreigon	View Status	Public	Platform	MAME (Official Binary)
Assigned To		Resolution	Open	OS	Windows 10 (64-bit)
Status [?]	Acknowledged			Driver	konami/zr107.cpp
Version	0.223	Fixed in Version		Build	64-bit
		Fixed in Git Commit		Github Pull Request #

Summary	07731: All sets in zr107.cpp: Mask rom tests as bad
Description	When doing a mask rom test in test mode, six (midnrun and windheat) or four (jetwave) will mark as bad. These roms being: midnrun/windheat: a-11 to a-16; a-11 and a-12 being the two k056832 tilemap roms and a-13 to a-16 are the texture data roms. jetwave: 4d, 9d, 13d and 18d; texture data roms
Steps To Reproduce	Press F2 to enter service mode, select "MASK ROM CHECK" and wait.
Additional Information	All sets in gticlub.cpp except for hangplt (freezes during its mask rom check different reasons) have their mask rom check pass. jetwave uses the same video board as gticlub would use yet jetwave fails its mask rom check while gticlub passes its mask rom check. Possible bad rom dumps or something else?
Github Commit

Flags
Regression Version
Affected Sets / Systems	All sets in zr107.cpp

Attached Files	windheat.png (4,711 bytes) Aug 19, 2020, 23:23 Uploaded by Hydreigon
	jetwave.png (2,586 bytes) Aug 19, 2020, 23:24 Uploaded by Hydreigon
	gticlub.png (6,695 bytes) Aug 19, 2020, 23:24 Uploaded by Hydreigon Passes mask rom checks
	finaldiff.png (263,426 bytes) Aug 26, 2020, 19:24 Uploaded by Osso

No.17911 Osso Moderator Aug 26, 2020, 19:24 edited on: Aug 26, 2020, 19:32	I can get jetwave to pass the mask ROM test by changing the ROM loading to match the one used for gticlub but then textures get worse (see attached screenshot - left current, center passing mask ROM test). EDIT: Actually nevermind, found the solution. So jetwave and clones are fixed, the other z107 games still fail.
No.17912 Hydreigon Tester Aug 26, 2020, 21:24	The solution was under my nose this entire time :S...mostly. I wonder how zr107's rom loading was incorrect all this time (but remained correct in gticlub)? I also ask what the right most screenshot was? Maybe the 056832 tilemap roms should be loaded as "ROM_LOAD32_WORD" after noticing how the tilemap roms were loaded in asterix.cpp and moo.cpp.
No.17913 Osso Moderator Aug 27, 2020, 04:51	56832 is difficult to deal with, it has per driver / hardware configs. The right screenshot is the difference between the left and central one. The uploaded file is what pngcmp outputs.
No.17936 Hydreigon Tester Sep 4, 2020, 20:42	No kidding. I attempted my said suggestion of loading those tilemap roms as "ROM_LOAD32_WORD" and made the tilemap graphics worse. The rom test still showed the tilemap roms as bad showing the same address. Something about Konami arcade hardware makes me feel uneasy.