Name Last commit Last update
..
Makefile Cleaned up build system, now make should work immediately on linux, etc. Deprecated some unushed tests and code.
cyclopstf.cxx Optimized calc_cnt_displs to use bucket_offset, corrected some warnings
cyclopstf.hpp Optimized calc_cnt_displs to use bucket_offset, corrected some warnings
dist_tensor_fold.cxx Optimized calc_cnt_displs to use bucket_offset, corrected some warnings
dist_tensor_inner.cxx Implemented permuted slice. Working for multiworld reads/write. -1s not tested yet, and same world fast-redistribution not
dist_tensor_internal.cxx Optimized calc_cnt_displs to use bucket_offset, corrected some warnings
dist_tensor_internal.h Implemented permuted slice. Working for multiworld reads/write. -1s not tested yet, and same world fast-redistribution not
dist_tensor_map.cxx A small but important fix in a corner case of the mapping logic.
dist_tensor_op.cxx Optimized calc_cnt_displs to use bucket_offset, corrected some warnings
dt_aux_map.hxx Implemented permuted slice. Working for multiworld reads/write. -1s not tested yet, and same world fast-redistribution not
dt_aux_permute.hxx Optimized calc_cnt_displs to use bucket_offset, corrected some warnings
dt_aux_rw.hxx Optimized calc_cnt_displs to use bucket_offset, corrected some warnings
dt_aux_sort.hxx Optimized calc_cnt_displs to use bucket_offset, corrected some warnings
dt_aux_topo.hxx Standardized all allocs to CTF_free and CTF_alloc and implemented a memory management system that keeps track of memory usage.
mach.h Revealed profiling calls interface, activated with -DPROFILE
scala_backend.cxx Made some improvements into the cost model to take into account blockwise reshuffle and try to do it more often
sym_indices.hxx Fixed memory leaks. Added production mode which turns of memstack tracing (-DPRODUCTION).