GitLab

Add option for more compact buffer allocation

See merge request exastencils/exastencils!90

# Conflicts:
#	Compiler/src/exastencils/app/ir/IR_LayerHandler.scala
#	Compiler/src/exastencils/config/Knowledge.scala

Parallel I/O interfaces

See merge request exastencils/exastencils!45

This reverts commit efcda7b5

Use fragment index from fielddata in CUDA transfer generation instead of...

See merge request exastencils/exastencils!86

devel/fix_cuda_initDomainFromFile

See merge request exastencils/exastencils!78

Use NUMA-aware zero-initialization for matrix-fields

See merge request exastencils/exastencils!75

devel/fix_poly_matrix_compounds

See merge request exastencils/exastencils!77

IR_PreItMOps: Fix missing duplicate when splitting up compound assignments and only apply trafo for HODTs

CI: Check if par I/O is compatible with optimizations and parallelization (poly, vect, OMP, CSE, unrolling)

# Conflicts:
#	Compiler/src/exastencils/app/ir/IR_LayerHandler.scala
#	Compiler/src/exastencils/config/Knowledge.scala

Use fragment index from fielddata in CUDA transfer generation instead of always assuming the fragment loop default iterator.

devel/fix_cuda_nodePos_comm

See merge request exastencils/exastencils!72

Devel/matrix init fix

See merge request exastencils/exastencils!54

Fix for CUDA communication: Consider neighbor fragment accesses with "dity data" for memory transfers.

# Conflicts:
#   .gitlab-ci.yml

This reverts commit 613066bf