There are (at least) two ways in which the reinterpret_cast is misused:

1. Used instead of static_cast:
   https://github.com/rapidsai/cudf/blob/f7fbc1160b17b969db2708e0cf6033d3db9fc1cf/cpp/src/io/avro/avro_gpu.cu#L95
   static cast should be used to cast from void*.
2. The use causes undefined behavior:
   https://github.com/rapidsai/cudf/blob/f7fbc1160b17b969db2708e0cf6033d3db9fc1cf/cpp/src/io

The version of device_copy_n which takes a CUDA stream argument
https://github.com/clara-parabricks/GenomeWorks/blob/74e6424c156a7ee15b4137d4788a4257ee6482c4/common/base/include/claraparabricks/genomeworks/utils/cudautils.hpp#L138
should be renamed to device_copy_n_async to make the non-blocking behavior more obvious.

The blocking three-argument version of device_copy_n should remain

When running the regression, the resulting logs seem to end up in third_party/tests/$TEST/.... This of course is not unnoticed by git, so a git status shows a ton of non-added new files added.

To reproduce:

make
make regression
git status   # observe all the files

Build or test artifacts should never clutter the rest of the code-base (we should regard them as read-only in

/ɛ, ɔ/ appear in a few-dozen phonemic transcriptions, but they either are allophones of, or incorrect transcriptions, of, /e, o/ in the modern language.