dbcsr
dbcsr copied to clipboard
DBCSR: Distributed Block Compressed Sparse Row matrix library
This is a placeholder for future code modifications to OO: - [ ] use member functions instead of separation between types and functions for proper data encapsulation - [ ]...
Currently `dbcsr_mp_new` creates `mp_env_obj` and does not initialize `prow_group` and `pcol_group` https://github.com/cp2k/dbcsr/blob/d97f28787a62e99c02aec38295bc8d88ea6a40e1/src/mpi/dbcsr_mp_methods.F#L80 This is done by the function `dbcsr_mp_grid_setup` https://github.com/cp2k/dbcsr/blob/d97f28787a62e99c02aec38295bc8d88ea6a40e1/src/mpi/dbcsr_mp_methods.F#L127 It is not clear, that one needs to call both...
Drop all ``` LOGICAL, PARAMETER :: debug_mod = .FALSE. LOGICAL, PARAMETER :: careful_mod = .FALSE. ``` .... ``` IF (debug_mod) THEN ... ENDIF IF (careful_mod) THEN ... ENDIF ```
See https://github.com/cp2k/cp2k/issues/212 Libsmm has to be exaplained (and possibly the code has to be improved). We have to explain when (instead of libxsmm, for instance) and how to use the...
1) Macros for compiling the library 2) External library dependencies
* Citation: "Setting cudaLimitPrintfFifoSize must not be performed after launching any kernel that uses the printf() device system call - in such case cudaErrorInvalidValue will be returned." * Since DeviceSetLimit...
updates: - [github.com/astral-sh/ruff-pre-commit: v0.3.4 → v0.3.5](https://github.com/astral-sh/ruff-pre-commit/compare/v0.3.4...v0.3.5) - [github.com/pre-commit/pre-commit-hooks: v4.5.0 → v4.6.0](https://github.com/pre-commit/pre-commit-hooks/compare/v4.5.0...v4.6.0)
* Allow to omit to preprocess kernel when dumping code (ACC_OPENCL_CPP). * Dump source even if CPP is not available. * Minor improvements (acc_opencl.sh).
This PR seems to cause: > CUDA RUNTIME API error: DeviceSetLimit failed with error cudaErrorInvalidValue. ( tested on H100 device ) _Originally posted by @hfp in https://github.com/cp2k/dbcsr/issues/767#issuecomment-2034752764_