GitLab

Implemented --build-dir option making it possible to have multiple external builds configured differently sharing the same source

Reorganized examples folder into tests/benchmarks/studies/examples and added configure mechanism adapted from Aquarius and NuLAB

Modified the build structure a bit to break it up more.

This commit contains a number of major changes and removals of unused source which aims at slimming down the source code in preparation for a release.

Switched to using CommData_t by value instead of by reference. MPI_Comms now freed via an std::set.

Got rid of the deprecated subdirectory.

Got rid of inner mapping functionalities which were not used due to being inefficient.

(somewhat questionable decision) got rid of old unit tests contained in unit_test subdirectory, including the code needed for the VERIFY functionality used in dist_tensor_op.cxx.
The old unit tests were rather limited and at this point somewhat redundant, but there may be reason to use VERIFY functionality in the future.

Hooked up add_to/from_subworld to fast redistribution kernel.  It should now be possible to use this function to add tensors
from an arbitrary sub MPI comm to/from a superset comm.

Also dist_tensor_internal.cxx is no longer included via cyclopstf.cxx though the organization here can be further improved.

Also, simplified interface.

Implemented a kernel that does a distributed reshuffle on the virtual block level when possible.

Cleaned up build system, now make should work immediately on linux, etc. Deprecated some unushed tests and code.

Deprecated C interface for cyclopstf.hpp. Added support for element-wise custom operations. Got rid of buffer and buffer_len parameters that were never used.  Added a test for weight
which uses a custom divide operaiton.

added matrix and vector types
added 3D FFT test

as Devin's interface.

2. Switched ctf.hpp to be templated and instantiated complex and double
3. Made diagonal rescaling work for SY and FFT test passes

TODO: Make diagonal rescaling require less dumb redistributions

Changed sizes from int64_t to int.  SY FFT test gets the wrong answer in parallel however, which is worrying

Added a new interface which is not fully complete nor tested