 No.22638
Firewave Senior Tester
Dec 20, 2024, 14:34
|
Ignoring MAME exception: LSI53C810: reg_w: Unknown reg F0, 00 terminate called after throwing an instance of 'emu_fatalerror' what(): LSI53C810: reg_w: Unknown reg F0, 00 Looks like there is a missing `catch` somewhere. Sets are flagged MNW so this report should be invalid. |
|---|---|
 No.22651
hap Developer
Dec 20, 2024, 17:18
|
If it's fatalerroring (instead of crash), then yeah, MT report can be closed. Unless the game was half-working before and it's a regression. |
 No.22652
cuavas Administrator
Dec 21, 2024, 00:52
|
> Looks like there is a missing `catch` somewhere. Recompiled code has no unwind tables, so exceptions can’t propagate across it. |
 No.22668
JimCarlTay Tester
Dec 22, 2024, 10:19
edited on: Dec 22, 2024, 10:44
|
Confirmed. It's likely a complete freeze at first, and then MAME closes itself afterward. Likely a crash. |
|
No.23329
JimCarlTay Tester
9 days ago
|
As of build f1f9ec3: Attempting to load the game (from those affected sets) causes the emulator to crash without freezing and gives a fatal error. The error message appears after a brief black screen. The message (from vs29915): "Fatal error: LSI53C810: reg_w: Unknown reg F0, 00" |
 No.23333
Robbbert Moderator
9 days ago
|
There's a long-term bug when run in Windows that prevents certain fatal errors from showing, including this one. When run normally only the dump appears. When run under debug, there's nothing at all - it just exits to the command line. |
 No.23335
Kale Developer
9 days ago
edited on: 9 days ago
|
vs29815 - massive memory trashing and page faults As for the crash, game trashes $f9000000-$f900ffff area for some reason, which happens to touch SCSI registers and the chip is not even supposed to have registers beyond $5f. It may be setting up PCI BARs in a noticeably different way than other Model 3 games causing the crash. |