07862: pitfight3: Pit-Fighter Rev. 3 - Game crashes around match 6 onward

ID	Category [?]	Severity [?]	Reproducibility	Date Submitted	Last Update
07862	Crash/Freeze	Critical (emulation)	Always	Jan 28, 2021, 20:08	Jan 29, 2021, 19:49

Tester	rcoltrane	View Status	Public	Platform	MAME (Official Binary)
Assigned To		Resolution	Unable to reproduce	OS	Windows 10 (64-bit)
Status [?]	Closed			Driver	atari/atarig1.cpp
Version	0.227	Fixed in Version		Build	64-bit
		Fixed in Git Commit		Github Pull Request #

Summary	07862: pitfight3: Pit-Fighter Rev. 3 - Game crashes around match 6 onward
Description	It looks like a slapstic bug because the behavior is quite similar of when the slapstic was not fully emulated (or simulated idk). After playing the game for quite come time, the game crashes. Sometimes, it starts glitching up and crashes, sometimes it freezes and crashes, sometimes the screen goes black during gameplay, etc. In general, it seems to start on later mathces (match 6 onwards). It's a regression since this game was working fine before. I think it started in version 0.227.
Steps To Reproduce	You just have to play the game to the end. At some point it will start glitching up or will crash all of a sudden. It seems to happen more from match 6 onward.
Additional Information	Seems to be a regression that started on version 0.227. It seems that a dev. called Cuavas commited a source cleanup in both slapstic.cpp and atarig1 files, which may be the cause of this regression. See screenshots here.
Github Commit

Flags
Regression Version	0.227
Affected Sets / Systems	pitfight3

Attached Files

No.18419 Robbbert Senior Tester Jan 29, 2021, 00:25	Cuavas did a srcclean. Srcclean only fixes incorrect code spacing and tabs, etc - it has no effect on how anything runs. You'll have to find someone else to blame.
No.18420 rcoltrane Tester Jan 29, 2021, 01:19 edited on: Jan 29, 2021, 01:24	I'm not here to blame anybody. I'm here to report bugs, and this bug started to happen on version 0.227. You may be right and so this is just a coincidence. Fact is thay this bug started to happen after these cleanups, unless someone else touched this driver in the past few months, which I highly doubt since Aaron wrote this driver all alone. The only other movement that I think could have a relation with this would be the Rampart changes, and since it also uses a slapstic, things could have gone wrong there, idk. Just saying.
No.18421 cuavas Administrator Jan 29, 2021, 04:41	You can refer to revisions by hash, posting mobile screenshots of your browser is useless. Also, Robbbert is right, that change definitely didn’t change behaviour of this driver.
No.18423 hap Developer Jan 29, 2021, 10:55	You're a fan of Pit Fighter (and as a result pretty good at it), right? Can you make an INP for 0.228 that triggers the bug? Then we can poke O.Galibert to have a look at it.
No.18424 rcoltrane Tester Jan 29, 2021, 13:25	Ok Hap, thank you for the suggestion, I will do that. :)
No.18425 Tafoid Administrator Jan 29, 2021, 13:47	Just a reminder if you intend to make an .inp file, be sure you have no NVRAM/eeprom file present (save a copy elsewhere if needed) and let the game use the default created one.
No.18426 rcoltrane Tester Jan 29, 2021, 16:54 edited on: Jan 29, 2021, 16:56	After updating MAME today with the latest 0.228 version (I was using 0.227 when I reported this bug) these bugs are gone! I've played the game several times, reaching later matches with one coin, and those bugs were not there anymore. I suppose that I may had old and/or corrupted files from previous releases that were overwritten and it happened to "fix" these random bugs I reported. My apologies to Cuavas if I sounded offensive by pointing the source cleanup as a possible origin of the bug. It was not my intention ;) Moderator please, close this report. I will keep testing it though and recording new .inp files following Tafoid's directions here. If it happens again then I will open a new report.