Is the SPC700 big or little endian?

[Toasty]

Title says it all. Thanks in advance!

 

[bcrew1375]

You might be able to find some info on it here: http://www.zophar.net/tech/snes.html

 

[BGNG]

I believe it's big-endian, but don't hold me to that.

 

[Toasty]

I believe it's big-endian, but don't hold me to that.

That's what I thought too, but the docs on the instruction set have me a little confused - this one inparticular:

MOV    A, (dp+X)  E7    2[bytes]     6[cycles]     A <- ((dp+X+1)(dp+X))     N......Z

(This is just describing the E7 version of the MOV instruction.) What confuses me is that dp+X+1 is written before dp+X, implying that the MSB is the second byte in memory, not the first (which of course would mean little-endian). Anybody know what's what here?

 

[BGNG]

Well, the ((dp+X+1)(dp+X)) is transferring to the Accumulator, which would imply that whatever byte ordering the data is stored as is the inverse of the byte ordering the processor uses. So if the SPC data is big-endian, according to this instruction, then the processor uses little-endian.

 

[Toasty]

Soooo, it's little-endian?

 

[BGNG]

Depends on the processor. I've not done research on the SPC chip. But I would say the SPC file data is stored in big-endian... but again, I'm not sure.

Try it out. One will work and one won't. That's an easy way to tell.

 

[Toasty]

Try it out. One will work and one won't. That's an easy way to tell.

Good point. (After all I have an inline function that does all the 16-bit memory reads. If it turns out wrong I only have to change a couple lines of code. ) Thanks for the suggestion.

 

[Doomulation]

Hmm yes, endianess is very easy to detect in c/c++. Just a few lines of code. Good luck on your project.