Skip to content
Snippets Groups Projects
Select Git revision
  • master default protected
1 result
You can move around the graph by using the arrow keys.
Created with Raphaël 2.2.028Nov26232221151413931Oct30262524232219181615121110854228Sep2721201917141312629Aug282724232221201615223Jul20191064319Jun151488May426Apr25242319181716131110929Mar28272623222120191615141398762128Feb272322191615141312987652131Jan30292625242322191817161512109854311Dec8764130Nov2928Fix another occurence of direct cache usageFix interoperability of classmethod generators with backend switchesAllow generators to be used as class methods![!286] Some changes for latest dune richardsCorrectly use base weight in the fastdg case[skip ci] Improve sumfact kernel interface[skip ci] Rename sumfact interface methodsFurther conditional restructuring[skip ci] Restructure conditionalsRestructure where permutation happens for sumfact vectorizationImplement operation counting for special loopy functions by delegatingIntroduce an option to skip sum factorization on boundary integrals[skip ci][wip] Move around things in realize_sumfact_kernel_function[skip ci][wip] Move input permutation to interface classesAdd permutation methods to Interface classesAdd direct_is_possible and the quadrature size to the parallel key[skip ci] Do not use matrix_sequence in VectorizedSumfactKernel[Bugfix] Fix cherry-pickMove quadrature_permutation to interface SumfactKernelInterfaceBase[Bugfix] Cherry pickUse cost permuted matrix sequence form SumfactKernel in realizationAdd cost permuted matrix sequence to SumfactKernelRename permuted_matrix_sequenceIf we have no SF kernels there is no costChoose vectorization strategy based on cost permuted matrix sequenceActivate gradvec vectorization for unstructured testsEdge consistent grids without detour over gmsh file + cleanup testsUpdate README.md after moving the project[!283] Renaming dune-perftool -> dune-codegenReadd submoduleFix pep8 after renamingSiwtch to cloning CI strategyRenaming part 1[!282] Autotune merge[!276] Feature/use custom geometry transformation for blockstructuredMerge branch 'master' into feature/autotuned-sumfact-kernel[!271] Add AVX512 single precision tranposes and a transpose testing suite[!263] Implement code generation for matrix inversion in C++[cmake] Skip all transpose tests on Clang - CI does not accept my partial exclusion[!277] Add a first implementation of hooks

Datenschutz und Impressum