[petsc-dev] [issue1595] Issues of limited number of MPI communicators when having many instances of hypre boomerAMG with Moose

Rob Falgout added the comment: 1. Yes. AMG coarsens recursively based on matrix entries, so you can't know a priori which ranks will be active by the time you get to the coarsest grid. 2. If you use the other direct solver option, it will be a little less performant. Ulrike can probably com

Re: [petsc-dev] upcoming release and testing [v2]

On Wed, 4 Apr 2018, Satish Balay wrote: > I have the following branches in queue for master [via next-tmp - this build > will start now] > > origin/dalcinl/fix-scatter-destroy > origin/dalcinl/fix-vecscatter-initpkg > origin/knepley/fix-cylinder-mesh > origin/rmills/feature-aijmkl-add-ma

Re: [petsc-dev] [issue1595] Issues of limited number of MPI communicators when having many instances of hypre boomerAMG with Moose

1. Are you saying that the ranks involved in the coarse solve depends on the matrix entries? 2. Okay, but this might degrade performance, right? 4. I think you are right that *if* all sends and receives have been posted (may be hard to guarantee if the user is using threads) and MPI_ANY_SOURCE i

Re: [petsc-dev] Nullspace with right preconditioning

Matthew Knepley writes: > On Wed, Apr 4, 2018 at 3:23 PM, Smith, Barry F. wrote: > >> >> It may not be broken; it may just be misunderstood or incorrect approach. >> >> You'll need to explain why and where exactly you think it is broken. > > > I am solving Stokes. I will push the example for

Re: [petsc-dev] Nullspace with right preconditioning

On Wed, Apr 4, 2018 at 3:23 PM, Smith, Barry F. wrote: > > It may not be broken; it may just be misunderstood or incorrect approach. > > You'll need to explain why and where exactly you think it is broken. I am solving Stokes. I will push the example for you to run tonight. Basically, I put

Re: [petsc-dev] Nullspace with right preconditioning

It may not be broken; it may just be misunderstood or incorrect approach. You'll need to explain why and where exactly you think it is broken. Barry > On Apr 4, 2018, at 1:13 PM, Matthew Knepley wrote: > > This is broken. Did we know this was broken? > > Thanks, > > Matt >

[petsc-dev] Nullspace with right preconditioning

This is broken. Did we know this was broken? Thanks, Matt -- What most experimenters take for granted before they begin their experiments is infinitely more interesting than any results to which their experiments lead. -- Norbert Wiener https://www.cse.buffalo.edu/~knepley/

[petsc-dev] [issue1595] Issues of limited number of MPI communicators when having many instances of hypre boomerAMG with Moose

Rob Falgout added the comment: FYI, I meant to say "corresponding user recv requests" in 4 below. -Rob hypre Issue Tracker

Re: [petsc-dev] upcoming release and testing [v2]

FYI - I have the release related strings in https://bitbucket.org/petsc/petsc/branch/balay/release-3.9 This change should be similar to what was done for 3.8 release https://bitbucket.org/petsc/petsc/commits/0e50f9e530a7b78427514d3e384f6941d4a9cc62?at=v3.8 For now - I'm using tomorrow's date [b

[petsc-dev] [issue1595] Issues of limited number of MPI communicators when having many instances of hypre boomerAMG with Moose

Rob Falgout added the comment: Hi All, Some comments and questions: 1. The Comm_create is used to create a subcommunicator that involves only the currently active MPI tasks so that the Allgather() will happen only over that subset. I don't think we can create this once, attach it to a paren

[petsc-dev] Add PetscFree1() ?

Jed, just for the sake of consistency, don't we need PetscFree1()? Also, PetscFree() is a macro, and evaluates twice the first arg, I just had to fix a bug in my code that was doing while (ctx->size > 0) PetscFree(ctx->arrayOfPointers[--ctx->size]); PetscFree(ctx->arrayOfPointers) I guess w

Re: [petsc-dev] FAS indentation

On Wed, 4 Apr 2018, Dener, Alp wrote: > (I didn’t wanna push directly to next without approval). Just a note: we would never commit/push directly to next. All fixes should go to the corresponding feature branch - this way they will get to master when ready [and next is never directly merged

[petsc-dev] upcoming release and testing [v2]

[starting a new thread] As of now master builds are clean! http://ftp.mcs.anl.gov/pub/petsc/nightlylogs/archive/2018/04/03/master.html next is improved but still has issues that need fixing. http://ftp.mcs.anl.gov/pub/petsc/nightlylogs/archive/2018/04/04/next.html most-likely are are related t

Re: [petsc-dev] [issue1595] Issues of limited number of MPI communicators when having many instances of hypre boomerAMG with Moose

Yes, this is a real issue for MOOSE which sometimes has thousands of active single-field solvers. PETSc can limit the number of fine-level communicators by retaining the dup'd communicator so the same communicator can be passed to hypre for each solver, but cannot control the MPI_Comm_create fo

[petsc-dev] Fwd: In March, you had 482 users visit your website (Google Analytics)

Good month. Matt -- Forwarded message -- From: Google Analytics Date: Wed, Apr 4, 2018 at 5:21 AM Subject: In March, you had 482 users visit your website (Google Analytics) To: knep...@gmail.com Your Snapshot for March Want to get a monthly snapshot for more views? Edit your