[Oberon] RISC5: memory instructions and flags

[Hellwig Geisse]

Hi Paul,

thanks for your answer.

On Sa, 2021-02-06 at 13:53 +0000, Paul Reed wrote:
> 
> Sorry that this has caused you trouble.

No trouble at all - only an interesting morning... :-)
I'm happy that the error messages coming from my memory test
program didn't have anything to do with a dysfunctional SDRAM
controller, but were caused by details of setting the flags.

> There's another gotcha here, 
> that although the "co-optimised" version of the Verilog/Lola and the 
> compiler allow the shorter code sequence when testing the result of a 
> load, it doesn't work in all branch cases, because the C and V flags are 
> *not* set.
> 

It seems questionable anyway how carry and overflow flags
should be set on a load, as there is no arithmetic involved.

Hellwig