[llvm-dev] A libc in LLVM

[Siva Chandra via llvm-dev]

On Wed, Jun 26, 2019 at 9:02 AM Andrew Kelley via llvm-dev <
llvm-dev at lists.llvm.org> wrote:
> It seems reasonable to assume that Google is only
> interested in Linux. In this case I have to re-iterate my original
> question, what are the needs that are not being met by existing Linux
> libcs, such as musl?

First of all, let me make this clear: musl is great, and I have used it
personally to learn about how things work. We also evaluated the option of
adopting musl and modifying it to suit our needs (I listed our rough goals
here: http://lists.llvm.org/pipermail/llvm-dev/2019-June/133360.html)

However, considering the disruptive nature of some of the changes we want
to make (like ABI independence, sanitizer friendly, modularity, avoiding
complexity from dynamic linking where possible, etc.), it seemed comparable
to building from the ground up. That also offered an opportunity to
structure this as part of the overall LLVM project which has lots of
advantages on its own.

This does not mean we want to re-implement everything. If community members
already have parts of libc implementations ready and want to contribute
them to LLVM, that would be great. That could absolutely include parts of
musl if the authors are interested in contributing them to LLVM, but some
research indicated this wasn't likely (happy to be corrected if wrong
though). We would just need to figure out how to add pieces incrementally,
and support the goals outlined above.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20190626/f1906aab/attachment.html>