Mark Harris

[FEA] Separate types for begin and end input iterators

3

**Is your feature request related to a problem? Please describe.** We may want to consider making the begin/end iterator capable of having different types in order to improve future compatibility...

Automate include grouping order in .clang-format

3

Replaces #14993 ## Description This uses the `IncludeCategories` settings in .clang-format to attempt to enforce our documented `#include` order in libcudf. See https://docs.rapids.ai/api/libcudf/stable/developer_guide I realize that there was a [previous...

Currently the C++ include files in the RAPIDS codebase are not consistently ordered and grouped. Inconsistent ordering makes it harder to maintain the codebase, to understand the dependencies, and to...

Refactor RMM resources to implement `cuda::memory_resource` concepts and deprecate and remove `device_memory_resource` and `host_memory_resource` base classes.

1

Part of #1443 ```[tasklist] ### Tasks that should not require prior work - [ ] https://github.com/rapidsai/rmm/issues/1446 - [ ] https://github.com/rapidsai/rmm/issues/1448 - [ ] https://github.com/rapidsai/rmm/issues/1453 - [x] Add `rmm::device_check_resource_adaptor::get_upstream_resource` - [...

[FEA] Change semantics of RMM memory resource equality comparison

5

Part of #1443 We plan to adopt `cuda::mr::memory_resource` concepts as the basis for RMM memory resources. This refactoring will bring flexibility both in building and composing memory resources (e.g. machinery...

[FEA] Python bindings for `rmm::mr::pinned_host_memory_resource`

18

**Is your feature request related to a problem? Please describe.** Add Python bindings for `pinned_host_memory_resource` added in #1392 .

Part of #1443

The `` header provides a standard C++ interface for heterogeneous, stream-ordered memory allocation tailored to the needs of CUDA C++ developers. This design builds off of the success ofRMM and...

**Is your feature request related to a problem? Please describe.** Extend RMM `stream_pool` with the same features as were recently added to libcudf in https://github.com/rapidsai/cudf/pull/13922 * add a base stream_pool...

As we move to replace `device_memory_resource` and `host_memory_resource` as dichotomous base classes in favor of the `cuda::memory_resource` concept, I think where possible we should deprecate features of the former that...