[Resolved (ish)] Thumb Scratch Registers? Returning #0x0 from ASMC Routines?

[Crazycolorz5]

If you were reading the recent posts on my dump of ASM crap, you’d know that @AlfredKamon had a problem with one of my routines. So I went in and I couldn’t really find a problem. But then I changed a
push {r4-r5, lr} to a push {r0-r5, r14} and the equivalent pops. And now it worked. What’s going on!? Aren’t registers r0-r3 supposed to be scratch? Or is returning 0x0 in r0 a problem for ASMC routines?

[Icecube]

This?

@AlfredKamon

There’s only one problem: the Supply command. I think it’s hardcoded to 0x01 (Eirika) and 0xF (Ephraim) in their respective routes. Should I copy Fa/Shiori’s data to a new character slot to change their portraits, I’d lose the Supply command as well.

IMO, the second one(if r0 is the only changed register.).

[Crazycolorz5]

No, this problem. It’s with the give-all-items-to-convoy routine.

Crazycolorz5's Dump of ASM Crap

Yes, I am. Anyway I should prepare the ROM and the save (I’ve worked around the problem so I should replicate it now) and since I’m having an exam tomorrow, I won’t be able to pass you the stuff today. We’ll get back to this soon. P.S. I’m also using Cam’s lord shift hack in the same chapter, if that has any relevance. Also, does the specified unit need to be loaded when calling the ASM? EndingScene: FADI 0x5 _SETVAL 0x1 Eirika//set Fa as the lord ASMC 0xoffset//Cam's lord shift hack…

[CT075]

depending on where/what you wrote your hack into up it’s entirely possible that there were extraneous return values in r1(/r2/r3) that you inadvertently clobbered

[Crazycolorz5]

No, this was literally a standalone ASMC routine for events…

[CT075]

actual ASMC event routines typically don’t return a value (they wouldn’t do anything with it anyway); it shouldn’t actually matter but FE8 is probably weird???

[Crazycolorz5]

I’ll take “FE8 is probably weird” …