ipiv: Rename ipiv_init and ipiv_reducek functions as INSERT_TASK  and make them match the INSERT_TASK API

Mapv: Add some missing fixes, and make the map kernel non runtime dependent.

See merge request !445

parsec/map: Fix the map codelet to always use a tile interface and do not enforce the user to check for the runtime used in the map functions.

header/coreblas: add the tile version of the functions into the main header to ease the access in map functions

Modify the map function to take into parameters an variadic number of data

See merge request !201

compute/map: Improve the map function to take into account multiple descriptor and multiple implementations for the kernel to apply.

correct zgetrf filters

See merge request !442

doc: fix typo

See merge request !443

Added _starpu_mpi_redux_wrapup_data detection for MPI reductions

See merge request !441

Added _starpu_mpi_redux_wrapup_data detection so that automatic MPI reduction can be used for partial pivoting (see also RUNTIME_ipiv_reducek).

small fix (filter in insert_task getrf)

See merge request !439

Prevent using PARSEC with MPI cmake+doc

See merge request !433

Remove the use of CHAMELEON_PRECISIONS that is an upper duplicate of...

Closes #124

See merge request !432

Abel Calluaud authored 1 year ago and Mathieu Faverge committed 1 year ago

Remove the use of CHAMELEON_PRECISIONS that is an upper duplicate of CHAMELEON_PRECISION provided by cmake_morse (Fix #124)

Fix issue #124: RP_CHAMELEON_PRECISION is the set of supported precisions,...

Closes #124

See merge request !428

Fix issue 124: RP_CHAMELEON_PRECISION is the set of supported precisions, while CHAMELEON_PRECISION is the set of enabled precisions

Expose __half type only if the cuda_fp16.h header is included (see...

See merge request solverstack/chameleon!427

Expose __half type only if the cuda_fp16.h header is included (see gpucublas.h) where __half is actually defined.

Disable StarPU/lapcy optimization for now and Fix a major issue with descriptor global allocation in distributed

See merge request !425

descriptor: Fix a major issue with global allocation in distributed and update comments in the structure for better understanding

starpu/lacpy: Fix a huge memory leak whan using the starpu_data_cpy family functions instead of lacpy