08855: segfault during device traversal in device.h

ID	Category [?]	Severity [?]	Reproducibility	Date Submitted	Last Update
08855	Core	Critical (emulator)	Always	Apr 13, 2024, 14:48	Apr 18, 2024, 09:24

Tester	holub	View Status	Public	Platform	MAME (Self-compiled)
Assigned To		Resolution	Invalid report	OS	Linux (64-bit)
Status [?]	Closed			Driver
Version	0.264	Fixed in Version		Build	64-bit
		Fixed in Git Commit		Github Pull Request #

Summary	08855: segfault during device traversal in device.h
Description	MAME fails when doing traversal through removed device Presumably caused by replaced AY from the parent
Steps To Reproduce	Unreleased specnext: * Boot till introduction screen * press 'd' for disabling intro during next boot * System reboots.and crushing
Additional Information
Github Commit

Flags
Regression Version
Affected Sets / Systems

Attached Files	image.png (1,289,198 bytes) Apr 13, 2024, 14:48 Uploaded by holub
Attached Files	image2.png (363,384 bytes) Apr 13, 2024, 18:40 Uploaded by holub

No.22058 holub Tester Apr 13, 2024, 18:42	sound stream is under suspicion
No.22059 Robbbert Senior Tester Apr 13, 2024, 22:06	I'm not sure if we can accept bugs for a system that isn't in MAME.
No.22060 holub Tester Apr 13, 2024, 23:36	I think the problem is fundamental - WIP system it's just where it's accidentally exposed for some reason.
No.22061 holub Tester Apr 14, 2024, 20:46 edited on: Apr 14, 2024, 23:54	This particular issue was fixed with #12249. But it's really scary that random spi commands can lead to memory polition of entre machine.
No.22062 Robbbert Senior Tester Apr 15, 2024, 01:27	It hasn't been accepted yet. Let's wait for that process to complete.
No.22066 hap Developer Apr 18, 2024, 09:22	MT bug reports need to be reproduceable on MAME official release, not local WIP drivers. AFAIK this issue was resolved anyway and was due to some kind of array overflow? But if it isn't resolved, this report can be reopened once your driver is in a MAME release.