[petsc-dev] Comparing binary output files in test harness

2021-02-18 Thread Blaise A Bourdin
Hi, I would like to write better tests for my exodus I/O functions and compare the binary files written to the drive instead of the output of the examples. For instance, would it be possible to do the following: ex26 -i -o output1.exo; mpirun -np 2 ex26 -i -o output2.exo; exodiff

Re: [petsc-dev] Understanding Vecscatter with Kokkos Vecs

2021-02-18 Thread Junchao Zhang
On Thu, Feb 18, 2021 at 4:04 PM Fande Kong wrote: > > > On Thu, Feb 18, 2021 at 1:55 PM Junchao Zhang > wrote: > >> VecScatter (i.e., SF, the two are the same thing) setup (building various >> index lists, rank lists) is done on the CPU. is1, is2 must be host data. >> > > Just out of

Re: [petsc-dev] Understanding Vecscatter with Kokkos Vecs

2021-02-18 Thread Fande Kong
On Thu, Feb 18, 2021 at 1:55 PM Junchao Zhang wrote: > VecScatter (i.e., SF, the two are the same thing) setup (building various > index lists, rank lists) is done on the CPU. is1, is2 must be host data. > Just out of curiosity, is1 and is2 can not be created on a GPU device in the first

Re: [petsc-dev] Understanding Vecscatter with Kokkos Vecs

2021-02-18 Thread Junchao Zhang
VecScatter (i.e., SF, the two are the same thing) setup (building various index lists, rank lists) is done on the CPU. is1, is2 must be host data. When the SF is used to communicate device data, indices are copied to the device.. --Junchao Zhang On Thu, Feb 18, 2021 at 11:50 AM Patrick Sanan

[petsc-dev] Understanding Vecscatter with Kokkos Vecs

2021-02-18 Thread Patrick Sanan
I'm trying to understand how VecScatters work with GPU-native Kokkos Vecs. Specifically, I'm interested in what will happen in code like in src/vec/vec/tests/ex22.c, ierr = VecScatterCreate(x,is1,y,is2,);CHKERRQ(ierr); (from

[petsc-dev] RSE and Postdoc openings at CU Boulder

2021-02-18 Thread Jed Brown
My research group has openings for a Research Software Engineer and a Postdoc. Details and application links below; feel free to email me with questions. ## Research Software Engineer CU Boulder’s PSAAP Multidisciplinary Simulation Center for Micromorphic Multiphysics Porous and Particulate

Re: [petsc-dev] TSSetConvergedReason(ts,TS_CONVERGED_USER);

2021-02-18 Thread Mark Adams
OK, makes sense. I didn't know monitors were called before the step. Thanks, On Thu, Feb 18, 2021 at 7:25 AM Matthew Knepley wrote: > Stefano is right. Changes in convergence should probably go in TSPostStep. > > Thanks, > > Matt > > On Thu, Feb 18, 2021 at 7:22 AM Stefano Zampini >

Re: [petsc-dev] TSSetConvergedReason(ts,TS_CONVERGED_USER);

2021-02-18 Thread Matthew Knepley
Stefano is right. Changes in convergence should probably go in TSPostStep. Thanks, Matt On Thu, Feb 18, 2021 at 7:22 AM Stefano Zampini wrote: > Mark > > monitors are not supposed to change the TS. You can think at monitors > being 'const' methods of the TS. Also, TSMonitor is called

Re: [petsc-dev] TSSetConvergedReason(ts,TS_CONVERGED_USER);

2021-02-18 Thread Stefano Zampini
Mark monitors are not supposed to change the TS. You can think at monitors being 'const' methods of the TS. Also, TSMonitor is called at the beginning of each step , see here https://gitlab.com/petsc/petsc/-/blob/master/src/ts/interface/ts.c#L4169 Il giorno gio 18 feb 2021 alle ore 15:16 Mark

[petsc-dev] TSSetConvergedReason(ts,TS_CONVERGED_USER);

2021-02-18 Thread Mark Adams
TSSetConvergedReason(ts,TS_CONVERGED_USER); does not seem to stop the iteration in a user monitor function. I have verified that it works from a post step method. Is this intentional? Mark