Code is up on github:

https://github.com/ct-clmsn/nim-hpx

A test program is here:

https://github.com/ct-clmsn/nim-hpx/blob/main/tests/test.nim

There are a couple issues with Nim's C++ integration support. By default,
Nim uses the copy constructor; also some of the Nim pragmas that are
documented to work around the copy constructor issue have some unfortunate
undocumented behavior...there are some language hacks in the wrapper that
work around that issue.

Traveler tools would be great - I'm not sure if the existing repository or
the Jetlag repository are preferred.

Chris

On Sat, May 14, 2022, 9:51 AM Hartmut Kaiser <hartmut.kai...@gmail.com>
wrote:

> Chris,
>
> > Not a problem, thank you for the reply and support.
> >
> > Quick update. The wrapper currently supports plain actions - serializing
> > arbitary nim types has been resolved. Local and remote asynchronous
> > function invocations work (futures work). Partitioned data types
> > (unordered_map and vector) work and a fair number of the parallel
> > algorithms have been wrapped. Components and component actions are in
> > progress. Dataflow does not seem far off.
>
> Wow, just wow! Would you be able to share your work?
>
> > At this point, providing nim implementations of existing hpx example
> > programs and some hpc benchmarks look feasible.
> >
> > Is the performance visualization tooling for C++ available (if so, which
> > repository)? I'd like to take a shot at integrating with the existing C++
> > functionality.
>
> Do you mean the Traveler tools or anything else?
>
> Thanks!
> Regards Hartmut
> ---------------
> https://stellar-group.org
> https://github.com/STEllAR-GROUP/hpx
>
>
> >
> > Thank you in advance,
> >
> > Chris
> >
> > On Wed, May 11, 2022, 7:45 AM Hartmut Kaiser
> > <mailto:hartmut.kai...@gmail.com> wrote:
> > Chris,
> >
> > Sorry for the late reply to your email.
> >
> > In general, I find the idea to combine HPX with Nim intriguing. Please
> let
> > us know if we can help in any way.
> >
> > Thanks!
> > Regards Hartmut
> > ---------------
> > https://stellar-group.org
> > https://github.com/STEllAR-GROUP/hpx
> >
> > From: mailto:hpx-users-boun...@stellar-group.org <mailto:hpx-users-
> > boun...@stellar-group.org> On Behalf Of ct clmsn
> > Sent: Thursday, April 21, 2022 9:21 PM
> > To: mailto:hpx-users@stellar-group.org
> > Subject: [hpx-users] nim wrapper for hpx
> >
> > To interested parties,
> >
> > I've put together a small nim wrapper for hpx. Nim's C/C++ interop is
> > nice. The language is statically typed, compiles to C relatively quickly,
> > and with hpx it provides a chapel/python experience. The process has been
> > straightforward (2 evenings of moderate work) and the wrapper works on a
> > single machine and has been tested on a slurm cluster. Link to the
> wrapper
> > is below:
> >
> > https://github.com/ct-clmsn/nim-hpx
> >
> > There's a bit of work to do with respect to getting nim typed variables
> > working with hpx serialization. Other than that, constructive feedback or
> > guidance regarding how to expand the feature set would be appreciated.
> >
> > v/r,
> > Chris
> > _______________________________________________
> > hpx-users mailing list
> > mailto:hpx-users@stellar-group.org
> > https://mail.cct.lsu.edu/mailman/listinfo/hpx-users
>
>
>
_______________________________________________
hpx-users mailing list
hpx-users@stellar-group.org
https://mail.cct.lsu.edu/mailman/listinfo/hpx-users

Reply via email to