Re: clang-built NetBSD and rust

[Greg Troxel <gdt%lexort.com@localhost>]

Havard Eidnes <he%NetBSD.org@localhost> writes:

> I've looked at
>
>   ftp://nyftp.netbsd.org/pub/NetBSD-daily/HEAD-llvm/latest/amd64/binary/sets/
>
> and specifically the base.tar.xz file, and it doesn't look like it has
> libgcc_s at all, but apparently MKGCC=yes will build it, but that
> doesn't appear to be the default (which is probably intentional).
>
> So ... what should I/we do about this?  Do we need a separate rust
> bootstrap kit built with clang and built with a clang-built "target
> root"?  It looks like there's no netbsd-9 nor netbsd-10 built with
> clang, only HEAD?

Basically binary packages are built for some environment, and you then
can't change the environment and expect them to work.

So yes, for NetBSD with LLVM and not GCC, I think we need different
bootstraps, and that's work to generate and awkward to implement.   I
don't think you should feel any duty to accomodate this, which I say
even if I'm a clang fan.

The big question to me is if a system with HAVE_LLVM can also build with
MKGCC.  The wiki says no:
  https://wiki.netbsd.org/tutorials/clang/
but that was likely adopted by me from list traffic and may be wrong.

I am unclear on whether NetBSD which 1) builds clang and 2) uses clang
to build the system can sanely also have GCC available, and whether we
should turn that on in public builds.  That's a bigger question than
rust, surely.

The real issue is that it's a bug that rust needs binary bootstraps, and
that there's no path from source with a base system.  This is
exacerbated by the rust.org implementation's practice of reuquiring the
previous rust version.  But other than the eventual gcc, and maybe
mrustc, I don't see that getting fixed, since they seem to view the
current situation as ok.