[Mitgcm-support] Re: How are we doing?

[mitgcm-support at dev.mitgcm.org]

Patrick Heimbach wrote:
> > If I just remove the #ifdef block from GAD_OPTIONS.h this will still do
> > what you say but it means adding #define DISABLE GAD_OPTIONS.h for the
> > experiments in question - which I don't have - which is why I did it
> > the other way to reduce impact.
> 
> Ehem, I didn't get that...

I was making excuses...

> 
> > Let's either add it back to CPP_OPTIONS.h and out of GAD_OPTIONS.h
> > keep the sense of DISABLE_ or keep it in GAD_OPTIONS.h and override it
> > like we do other packages (ie. carbon/code/GAD_OPTIONS.h)?
> 
> I think, keeping it as DISABLE... in GAD_OPTIONS.h is ok.

Thankyou.

> Then,
> - switch ifdef to ifndef in gad_advection, where they bracket
>   CADJ STORE with the "passkey" used
> CADJ STORE localTij(:,:)  = comlev1_bibj_pass, key=passkey,
> byte=isbyte
>   (NB: that's not all the STORE directives, some use "kkey";
>   those don't get ifdef-ed).
> - switch ifdef to ifndef in the_main_loop.F bracketing
> CADJ INIT comlev1_bibj_pass
> CADJ &    =
> COMMON,nchklev_1*nsx*nsy*nr*nthreads_chkpt*maxpass

Ehem, I didn't get that...

But seriously, I think that's what I already did (cos I'm clever?).

So I'll remove the #ifdef AUTO_ in GAD_OPTIONS.h in a moment...

A.