Ilya Enkovich comments

Results 27 comments of


                                            Ilya Enkovich

Remove unnecessary memory accesses from CPU codegen

I guess splitting functional and non-functional changes (more accurate IR metadata) would simplify the review process.

Shared execution context draft

> @ienkovich how is this branch looking in terms of moving out of draft? Also saw the asan failure in GeospatialTest, any ideas on that? The functionality of this patch...

Use tbb::parallel_sort instead of std::sort.

> @ienkovich can you rebase this? Done

Use tbb::parallel_sort instead of std::sort.

> The call to `ResultSet::topPermutation()` from https://github.com/intel-ai/omniscidb/blob/ienkovich/tbb-parallel_sort/QueryEngine/ResultSet.cpp#L613 is already done in a multi-threaded context using all available cpu threads. Will calling `tbb:parallel_sort()` help or hurt performance in this case? I...

Use tbb::parallel_sort instead of std::sort.

@mattpulver I added a flag to `ResultSet::topPermutation()` to allow multiple threads usage when we don't use it within a thread pool.

remove dead code

I just noticed that we have ra_exe_unit.join_quals.empty() as a part of removed condition and it's also checked earlier. So A || (A && B) == A.

Add data columns to join hash table payload

I understand the initial version is not good enough to become a default strategy. For now it can be a tool to find cases when payload really helps. Hopefully we...

Add data columns to join hash table payload

The patch was tested on not CUDA enabled machine. I'll fix the CUDA build.

Add data columns to join hash table payload

CUDA build should be fixed now but I found another issue, in some case a wrong payload is loaded into a hash table. I couldn't find the reason yet. Will...

Add data columns to join hash table payload

Fixed issue with wrong payload loaded.