[FLASH-USERS] Use of Driver_getNumProcs.F90

[Lewis,Sean]

Thank you for the direction, Adam. Turns out my logfile also reports “Number of MPI tasks” as being “1” despite requesting >1 processors. I recall running into an issue like this when building FLASH with a serial version of hdf5 or with a non-parallel configured gcc library. This time though, I am confident that FLASH should be able to use all parallel functionality. Have you (or anyone else) seen a behavior like this before?

I also realized my line of questioning has shifted a bit, so please let me know if it’s preferred to start a new email chain for the sake of documentation.
Thanks,

Sean C. Lewis
Doctoral Candidate
Department of Physics
Drexel University

From: Adam Reyes <adam.reyes at rochester.edu>
Date: Monday, September 26, 2022 at 1:16 PM
To: "Lewis,Sean" <scl63 at drexel.edu>
Cc: "flash-users at flash.rochester.edu" <flash-users at flash.rochester.edu>
Subject: Re: [FLASH-USERS] Use of Driver_getNumProcs.F90


External.
Hi  Sean,

That is the correct way  to use that routine and should return the number of mpi ranks being used. It should match what gets written in the header of the logfile as "Number of MPI tasks:"

You might find it helpful to look at the Grid_mapParticlesToMesh<https://nam10.safelinks.protection.outlook.com/?url=https%3A%2F%2Fflash.rochester.edu%2Fsite%2Fflashcode%2Fuser_support%2Fflash4_ug_4p62%2Fnode62.html%23SECTION05192000000000000000&data=05%7C01%7Cscl63%40drexel.edu%7C806fecef4d484657dcc208da9fe2e9f0%7C3664e6fa47bd45a696708c4f080f8ca6%7C0%7C0%7C637998094188949227%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=4VdDzScPNFcrO4Sy1gr1QG6sU7sI91PsxEKvCyc5ebI%3D&reserved=0> routine for hints in how that one handles mapping particle attributes to field values at cells.

*************************************************************************
Adam Reyes
[Image removed by sender.]
Code Group Leader, Flash Center for Computational Science
Research Scientist, Dept. of Physics and Astronomy
University of Rochester
River Campus: Bausch and Lomb Hall, 369
500 Wilson Blvd. PO Box 270171, Rochester, NY 14627
Email adam.reyes at rochester.edu<mailto:adam.reyes at rochester.edu>
Web https://flash.rochester.edu<https://nam10.safelinks.protection.outlook.com/?url=https%3A%2F%2Fflash.rochester.edu%2F&data=05%7C01%7Cscl63%40drexel.edu%7C806fecef4d484657dcc208da9fe2e9f0%7C3664e6fa47bd45a696708c4f080f8ca6%7C0%7C0%7C637998094188949227%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=ecvaJJiiNC8Dq%2FKQQhc%2FObnWqfXSULYPJtbTS7FrEN0%3D&reserved=0>
 (he / him / his)
[Image removed by sender.]
*************************************************************************

On Mon, Sep 26, 2022 at 12:25 PM Lewis,Sean <scl63 at drexel.edu<mailto:scl63 at drexel.edu>> wrote:
Hi all,

In my application, I am attempting to gather all block cell data post-refinement on particle distribution so that I can then manually set cell field values. To accomplish this, I cycle over each processor and call Grid_getLocalNumBlks(). To know how many processors to cycle over, I first call Driver_getNumProcs(MESH_COMM, n), where MESH_ COMM is the Flash defined value and n is the output variable: total number of processors. In my application, Driver_getNumProcs() always returns “1” no matter how many processors I use. The same is true if I pass the GLOBAL_COMM variable. Is there a hidden functionality here that I am missing? Does Driver_getNumProcs() return 1 intentionally? Is there another way for FLASH to internally report the number of processors being used?

Thank you for any insight you can provide,
-Sean

Sean C. Lewis
Doctoral Candidate
Department of Physics
Drexel University
_______________________________________________
flash-users mailing list
flash-users at flash.rochester.edu<mailto:flash-users at flash.rochester.edu>

For list info, including unsubscribe:
https://flash.rochester.edu/mailman/listinfo/flash-users<https://nam10.safelinks.protection.outlook.com/?url=https%3A%2F%2Fflash.rochester.edu%2Fmailman%2Flistinfo%2Fflash-users&data=05%7C01%7Cscl63%40drexel.edu%7C806fecef4d484657dcc208da9fe2e9f0%7C3664e6fa47bd45a696708c4f080f8ca6%7C0%7C0%7C637998094189105463%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=8XWF12I5fkVlLmDfKKju7hfKQ2ThVCZx2O5ZnLZZBBI%3D&reserved=0>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://flash.rochester.edu/pipermail/flash-users/attachments/20220926/02b129cb/attachment.htm>