Open
Conversation
Signed-off-by: Carl Pearson <cwpears@sandia.gov>
Signed-off-by: Carl Pearson <cwpears@sandia.gov>
In ref/ cg_solve.hpp, we have waxpby(one, r, beta, p, p) In kokkos/ we had waxpby(beta, p, one, r, p) These are algebraicly equivalent: waxpby(alpha, x, beta, y, w) computes w = alpha*x + beta*y, so ref/: p = 1*r + beta * p kokkos/: p = beta*p + 1*r However, p is length A.num_cols and includes ghost entries after make_local_matrix, while r has only local rows. For mpi runs, when p is the x argument it trips y.local_size < x.local_size || w.local_size < x.local_size This PR swaps the Kokkos arguments back to the ref ones, so when p is extra long the check doesn't trip. Signed-off-by: Carl Pearson <cwpears@sandia.gov>
cwpearson
force-pushed
the
ci/compare-ref-kokkos
branch
from
901ffd3 to
6778f48
Compare
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
This adds a CI job that does a serial+openmpi run and uses MINIFE_DEBUG to dump the raw values from inside the solver to compare the two implementations.
Along the way, two issues fixed:
kokkos/dump thing was expectingstd::vector, notKokkos::vectorkokkos/waxpy had a bug in the decomposed MPI caseWith these changes, the internal vector/matrix states match before and after the solves.