Describe the enhancement requested

Currently, the Join result materialization is implemented by traversing the materialize_ list serially. I believe that using TaskGroup to achieve parallelism could improve the performance of hash join.

Component(s)

C++