Writing a Gecko Register to an Address

[Almas]

I'm going to look like a fool for asking this, I fear.

I have a code that stores one of two values into a gecko register pending an If codetype. However, I am having issues writing this value to another place. The CT4/CST2 (Gecko Register Functions / Save Register To) codes write to [XXXXXXXX], but I can find no codes which do the equivalent to XXXXXXXX, which would make life a lot easier. I suppose it is possible for me to use a second register to permanently store the place where I'd like to write to, but I was wondering if there was a more efficient means to write to this point.

Thanks for your help.


EDIT: There was a mistake elsewhere in my code, it seems, and I managed to patch it up. It turns out the codetypes DO write to XXXXXXXX. Does this mean that the codetype document is incorrect, or did I make an error reading it? Sorry for the fuss.

[hetoan2]

you could always set the value to 0 and have your gecko Register be added to the value, which is basically moving it :|

[Romaap]

what are you trying to do?

because there is an option to store XXXXXXXX to gR

[hetoan2]

no he want's gR to XXXXXXXX

[Romaap]

gR to [XXXXXXXX] or to XXXXXXXX? because you can't store anything to a value xwink

use this:
84UYZZZN XXXXXXXX:
8400 : writes the 8bits in grN ZZZ times+1 at [XXXXXXXX]
8401 : writes the 8bits in grN ZZZ times+1 at [XXXXXXXX+ba]
9401 : writes the 8bits in grN ZZZ times+1 at [XXXXXXXX+po]
8410 : writes the 16bits in grN ZZZ times+1 at [XXXXXXXX]
8411 : writes the 16bits in grN ZZZ times+1 at [XXXXXXXX+ba]
9411 : writes the 16bits in grN ZZZ times+1 at [XXXXXXXX+po]
8420 : writes the 32bits in grN ZZZ times+1 at [XXXXXXXX]
8421 : writes the 32bits in grN ZZZ times+1 at [XXXXXXXX+ba]
9421 : writes the 32bits in grN ZZZ times+1 at [XXXXXXXX+po]

I think you should use 8420

Doesnt anybody read the codetypes? >_>

[Romaap]

eehh... yeah, sorry. but why put an 04 in front?

[Romaap]

but still, people shoud look at the codetypes before asking.

[Almas]

My initial concern was that there was no simple way to write grN to XXXXXXXX - thanks paprika for resolving this for me. However, I later discovered that the gecko register codetypes which SAY they write to [XXXXXXXX] ACTUALLY write to XXXXXXXX. Which is different, I thought.

[Romaap]

[XXXXXXXX] is a address and XXXXXXXX is a value

[Almas]

So [XXXXXXXX] means "the data stored at XXXXXXXX".

Does this not imply that:

8420 : writes the 32bits in grN ZZZ times+1 at [XXXXXXXX]

means:

Write grN ZZZ times+1 at the address stored at XXXXXXXX

?

[Romaap]

So [XXXXXXXX] means "the data stored at XXXXXXXX".

Does this not imply that:

8420 : writes the 32bits in grN ZZZ times+1 at [XXXXXXXX]

means:

Write grN ZZZ times+1 at the address stored at XXXXXXXX

?

no

[Almas]

Well then, my point is that the codetype doesn't do that. It writes the Gecko Register to the address XXXXXXXX, rather than the value stored at the address XXXXXXXX, as the codetype document would suggest (and you have just agreed with me upon).

[Romaap]

yeah, I'm sorry. it does write to the address, I didn't read it properly. my bad

[Romaap]

84200000 80123456

that code will write the value at gR0 to 80123456
so if gR0 is 11111111, 80123456 will become 11111111

[Almas]

No, it doesn't. Which is what my editted point was - the description of it in the codetype document is incorrect, I believe.