[FLASH-USERS] Error When Running LaserSlab On FLASH 4.6.1

[Jared Guzman]

Aaron,

I have fixed the space in my Makefile, but still the error continues.
Thanks for the tip though, and I have checked out that link but am not
entirely sure what it means. As I understand it, the file
"Burn_interface.o" and/or "for_main.o" do *not* have a main function, and
that is why the reference to main is undefined? I will look into building
with gfortran and formatting my Makefile as such. Thanks!

Klaus,

My version of GNU make is 3.82, but I will try to see why those flags dont
show up in the output. As well as trying out gmake

Also as a note to all, I am currently now trying to run the Sedov
simulation, and it produces the same identical error, to the letter. Not
sure if that helps but it's new information.

Thanks to you both,

Jared

On Thu, Jan 7, 2021 at 9:04 PM Klaus Weide <klaus at flash.uchicago.edu> wrote:

> On Thu, 7 Jan 2021, Jared Guzman wrote:
>
> > ./setup -auto LaserSlab -site=jag21791.vt.edu
>
> ...
>
> > I have the following modules loaded prior to running make:
> >
> > Currently Loaded Modules:
> >   1) intel/15.3   2) openmpi/3.0.0   3) phdf5/1.8.16
> >
> > My Makefile is also attached, which aligns with the modules I loaded in.
>
> Jared,
>
> I don't know what's going on there, but it appears that the output from
> the 'make' command you were showing does not fully correspond to the
> Makefile.h that you have attached.
>
> The first compilation command, from your output:
>
>   /opt/apps/intel15_3/openmpi/3.0.0/bin/mpif90  -I
> /opt/apps/intel15_3/openmpi3_0/phdf5/1.8.16/include -DH5_USE_16_API
> -DMAXBLOCKS=1000 -DNXB=8 -DNYB=8 -DNZB=1 -DN_DIM=2 Burn_interface.F90
>
> This is not correct, and cannot work properly. The command should include
> the flags that are set in your Makefile.h in the line
>
>   FFLAGS_OPT       = -c -g -r8 -i4 -O3 -real_size 64 -diag-disable 10120
>
> but for some reason they are missing.
>
> I wonder whether there is something unusual about your 'make' command
> (What version of GNU Make is it?) or the environment variables of your
> shell (perhaps set when you are loading modules?).
>
> Maybe you can try to run 'gmake' or 'make' in a new shell, without having
> run any commands to load any modules. Maybe the compilation command shown
> will be different? (The compilation would not be expected to succeed in
> this case, I am just suggesting this as a way to narrow down possible
> causes.)
>
>
> Btw., for correctly configuring LaserSlab you will need additional flags
> in your setup command (probably something including at least '+uhd3t
> +mtmmmt'), see the User's Guide.
>
> Klaus
>


-- 
*Jared Guzman*
*B.S. Physics 2019*
*M.S. Aerospace Engineering 2022*
*Virginia Tech*
*540-419-6027 | jag21791 at vt.edu <jag21791 at vt.edu>*
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://flash.rochester.edu/pipermail/flash-users/attachments/20210107/862d45a6/attachment-0001.htm>