Sparse streams #17
Open
Sparse streams #17
Conversation
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
docs: Remove unsupported Doxyfile.in tags
The new test
`TrilinosInstallTests_simpleBuildAgainstTrilinos_by_package_build_tree` merged
from PR #11863 fails because the subdirs ${CMAKE_CURRENT_BINARY_DIR}/common
and ${CMAKE_CURRENT_SOURCE_DIR}/common because this CMakeLists.txt file
already sits in the kokkos-kernels/common/ subdir.
I don't know why this error did not happen with PR testing for PR #11863 but
this is clearly the right thing to do.
Partially addresses issue kokkos#1749
Required for for TriBITS-compliance Discussion at trilinos/Trilinos#8866 (comment)
This reverts commit c0349db.
This reverts commit 6116419.
This reverts commit d55fb10.
Bsr compatibility
remove triplicate sanity checks in BsrMatrix
The solver was implemented along with some helper utilities: solver status and solver params. Four tests are added to check the interface, the correctness of single solves and the correctness in parallel solves that mimic its use in FE/FD code.
remove Intel 2017 code (no longer supported)
Newton solver
(HashmapAccumulator data races on Ada and Hopper architectures). To avoid checking for every architecture at every location, add a new macro KOKKOSKERNELS_CUDA_INDEPENDENT_THREADS which is defined if we're targeting an architecture with independent thread scheduling.
remove duplicate BSR SpMV tests
SpAdd perf test can call cusparse and MKL TPLs directly, but this can only be done if the default ordinal and offset type are both int. Check for this ahead of time and print a decent error message. Fixes build error when cusparse enabled, but offset int not enabled.
(insert a space between it and previous flag)
Fix SpAdd perf test when offset/ordinal is not int
Attempt to resolve "error: ‘pair’ was not declared in this scope" in cuda/11.2+gcc/8.3.0 builds
Test_ODE_Newton: Add template parameters for Kokkos::pair
Also on ger(): - better explanations on unit_tests, - its unit tests now have terminology similar to syr() unit tests, - replace 'one level' and 'two level' terminology on the impl file, following syr(). Also on syr(): - small improvement on the ROC tpl file
Remove printouts from the unit tests of ger() and syr()
including cusparse, rocsparse, onemkl wrappers. Fix some misc. bugs too: - Fallback check for cusparse was wrong - we were only calling the TPL for mode 'N', but TPL should be used for 'T' and 'H' also - In SYCL build, forgetting to #undef a macro caused warning in sparse tests - Can't call cuSPARSE or MKL spmv with mode 'H' and real scalars. In this case, just switch to mode 'T' since it's equivalent
(no longer needs to match AMatrix::execution_space)
Move them from KokkosSparse:: to KokkosSparse::Impl:: since they're not supposed to be part of the public interface. Keep an alias in KokkosSparse:: for backwards compatibility, but mark it [[deprecated]].
e10harvey
force-pushed
the
SparseStreams
branch
from
6212687 to
6cd36f9
Compare
e10harvey
pushed a commit
that referenced
this pull request
Sep 6, 2023
sparse/src: Add doxygen-style comment
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.
No description provided.