tree e57a4fe518e5f18c45640019c32f20ee66a8a6e4
parent 9aa52c6ff74358d1b91542221e0545bb61bdedcc
author Dmitriy Korchemkin <dmitriy.korchemkin@gmail.com> 1667394408 +0300
committer Dmitriy Korchemkin <dmitriy.korchemkin@gmail.com> 1668268686 +0300

Parallel for with iteration costs and left product

Parallel for with user-supplied [cumulative] iteration costs allows to
get performance improvements on problems with significantly different
time requirements per parallel loop iteration.

One of those problems is left multiplication with block-sparse matrix.
Using number of non-zero values per column block, we partition column
blocks into contiguous sets with approximately equal number of
operations to be performed.

Change-Id: I4a862a10a586cdfbec22e8168a3423537039abc2