[llvm-dev] [Incubation] Request to incubate mlir-npcomp

Stella Laurenzo via llvm-dev llvm-dev at lists.llvm.org
Mon Jun 22 23:07:50 PDT 2020


Per the recent (seeming) consensus regarding incubating new projects under
the LLVM organization, I would like to trial the process by requesting to
incubate mlir-npcomp <https://github.com/google/mlir-npcomp>. The project
is still quite young and has been primarily developed part time by myself
and Sean Silva over the last ~2 months. We set it up following discussion
of a Numpy/Scipy op set
<https://llvm.discourse.group/t/numpy-scipy-op-set/768> and also in
conjunction with a proving ground for high level dialects/transforms for
lowering from "numpy-aligned" frontends (e.g. sometimes labeled TCF/TCP).

When I obtained permission from my employer (Google) to open source the
effort, it was with the understanding that it was being developed with an
eye towards eventual inclusion (in some fashion) under the LLVM umbrella.
As such, we set up licensing and repository organization in such a way as
to facilitate this down the road. We originally started it as a fork of the
LLVM repository, but transitioned to the MLIR standalone template
<https://github.com/llvm/llvm-project/tree/master/mlir/examples/standalone>,
and we found it more productive to iterate out of tree in this fashion,
bumping to the latest LLVM version every week or so as needed (note: the
ability to exist out of tree for MLIR dependent projects is actually quite
good, and the more of us who do it, the better it becomes).

As noted in the README, there are multiple directions that this project
could go in the future, possibly growing to be a full python/numpy compiler
frontend or just being a source of dialects, transforms and interface code:
we just need to keep building it to find out, and I am open to multiple
outcomes as things progress. In addition, there has been some inquiries
from the community regarding aligning this work with the LLVM foundation
explicitly, as it is an easier entity for contributors to align with.

So, as the first one walking through the door, what is the process we would
like to follow? I'm happy to provide more information/discussion, but I'd
also be happy if with just an LGTM and someone creating an "mlir-npcomp"
repository under the LLVM GitHub organization and working the rest out as
we go.

Thanks.
- Stella
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20200622/9ad3f189/attachment-0001.html>


More information about the llvm-dev mailing list