Repository navigation
Leave the diagonal reorder to PETSc's own LU - #324
Merged
Merged
Conversation
singhharsh1708
referenced
this pull request
Oct 24, 2026
* Read the option names PETSc 3.25 renamed * Bump to 0.12.3
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
With PETSc 3.22,
-pc_factor_mat_solver_type mumps,umfpackorkluon a sparsejac_prototypecrashed Julia inMatReorderForNonzeroDiagonal_SeqAIJ. The package always adds-pc_factor_nonzeros_along_diagonal, andPCSetUp_LUin 3.22 (lu.c) reorders even when the factorization orders itself, so it reorders an ordering that was never made; 3.25 skips that case. The option is now left out when the user names a solver type other than petsc, which pivots on its own.All three solvers, and superlu_dist, now return Success on the 1-D Brusselator and on Robertson with a mass matrix and as a DAE, under PETSc 3.22 and 3.25.
AI Disclosure: Claude assisted with the code. Testing and review by @singhharsh1708.