Re: [AMBER-Developers] 3rd attempt to merge energy drift fix

From: David Cerutti <dscerutti.gmail.com>
Date: Fri, 8 Apr 2022 07:06:04 -0400

I sometimes see this. From my recollection it can be a number of things,
but there may be something with make -j8 if the various dependencies are
not engaged properly. The new cmake doesn't really let you choose how many
processes to engage. Maybe try make depends? Failing that, try a make
clean and remaking from scratch. Roll the dice again against what seems to
be a compilation Heisenbug.

Dave


On Thu, Apr 7, 2022 at 9:30 PM David A Case <david.case.rutgers.edu> wrote:

> On Thu, Apr 07, 2022, Scott Le Grand wrote:
>
> >Built yesterday, now seeing:
>
> So: what happened between yesterday and today?
>
> >Error: Symbol ‘ischeme’ at (1) has no IMPLICIT type
> >/home/slegrand/amber20_src/src/pmemd/src/prmtop_dat.F90:475:34:
> >
> > if (ischeme == 1 .and. therm_par .gt. 0.0) ndfmin = 0
>
> The routine here, init_prmtop_dat(), has this at line 230:
>
> use mdin_ctrl_dat_mod
>
> and the mdin_ctrl_dat.F90 file definitely declares therm_par and ischeme.
>
> You can go here:
>
> https://gitlab.ambermd.org/amber/amber/-/wikis/Singularity-Amber22
>
> to see the list of OS'es and compilers that have been tested, ranging from
> GNU 7.5 to GNU 11.2. Maybe something will ring a bell.
>
> ...dac
>
>
> _______________________________________________
> AMBER-Developers mailing list
> AMBER-Developers.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber-developers
>
_______________________________________________
AMBER-Developers mailing list
AMBER-Developers.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber-developers
Received on Fri Apr 08 2022 - 04:30:02 PDT
Custom Search