z80dma.cpp: Relaxed fatal error which were used for unknown commands catching. #14627
Conversation
|
I believe this fatal stayed enough to catch any commands we suspected MAME is missing. Time to be less destructive for the user. |
| // 0x8e: newtype on Sharp X1, unknown purpose | ||
| logerror("Z80DMA '%s' Unknown base register %02x\n", tag(), data); |
There was a problem hiding this comment.
logerror is a device_t member function and always includes the tag. You don’t need to duplicate it in the message.
There was a problem hiding this comment.
Will change in one of the following commits
There was a problem hiding this comment.
What are the other SWs that triggers this btw?
There was a problem hiding this comment.
I’ve already got it staged locally. I realise it was just an oversight converting from throwing emu_fatalerror to logerror.
There was a problem hiding this comment.
What are the other SWs that triggers this btw?
The Next's DMA extends this one. Recently we found an example where multiple software write 0xCA and this command is not documented anywhere (not only this one, but few other are crashing same way):
https://github.com/mikedailly/mod_player/blob/0c46930c3c0a2aa4679718452935853fe89c1774/mod_player/utils.asm#L627
db $ca ; unknown
DMA in Spectrum is cursed. Many soft is based on examples from Magazines which either had typo or used rare clone of z80dma chip.
4 participants
No description provided.