1 |
On Mon, 2021-02-08 at 18:56 +0100, Alessandro Barbieri wrote: |
2 |
> Il Lun 8 Feb 2021, 12:19 Michał Górny <mgorny@g.o> ha scritto: |
3 |
> |
4 |
> > Hi, |
5 |
> > |
6 |
> > FYI the developers of dev-python/cryptography decided that Rust is going |
7 |
> > to be mandatory for 1.5+ versions. It's unlikely that they're going to |
8 |
> > provide LTS support or security fixes for the old versions. |
9 |
> > |
10 |
> > Since cryptography is a very important package in the Python ecosystem, |
11 |
> > and it is an indirect dependency of Portage, this means that we will |
12 |
> > probably have to entirely drop support for architectures that are not |
13 |
> > supported by Rust. |
14 |
> > |
15 |
> > According to upstream platform support information [1], this probably |
16 |
> > means (eventually) entirely removing the following architectures: |
17 |
> > - alpha (stable) |
18 |
> > - hppa (stable) |
19 |
> > - ia64 (stable) |
20 |
> > - m68k (exp) |
21 |
> > - s390 (except for s390x, exp) |
22 |
> > |
23 |
> > Furthermore, the Gentoo Rust packages are missing support |
24 |
> > for the following platforms, apparently supported upstream: |
25 |
> > - mips (exp) |
26 |
> > - ppc (32) (stable) |
27 |
> > - sparc (stable) |
28 |
> > - s390x (exp) |
29 |
> > - riscv (stable) |
30 |
> > |
31 |
> > Apparently it's non-trivial to bootstrap Rust on these platforms, |
32 |
> > so it's unclear when Gentoo is going to start providing Rust on them. |
33 |
> > |
34 |
> > I've raised a protest on the cryptography bug tracker [2] but apparently |
35 |
> > upstream considers Rust's 'memory safety' more important than ability to |
36 |
> > actually use the package. |
37 |
> > |
38 |
> > Honestly, I don't think it likely that Rust will gain support for these |
39 |
> > platforms. This involves a lot of work, starting with writing a new |
40 |
> > LLVM backend and getting it accepted (getting new code into LLVM is very |
41 |
> > hard unless you're doing that on behalf one of the big companies). You |
42 |
> > can imagine how much effort that involves compared to rewriting the new |
43 |
> > code from Cryptography into C. |
44 |
> > |
45 |
> > If we can't convince upstream, I'm afraid we'll either have to drop |
46 |
> > these architectures entirely or fork Cryptography. |
47 |
> > |
48 |
> > |
49 |
> > [1] https://doc.rust-lang.org/nightly/rustc/platform-support.html |
50 |
> > [2] https://github.com/pyca/cryptography/issues/5771 |
51 |
> > |
52 |
> > -- |
53 |
> > Best regards, |
54 |
> > Michał Górny |
55 |
> > |
56 |
> |
57 |
> Should we shed tears for those legacy architectures or move forward? Does |
58 |
> anyone really use them in production? |
59 |
|
60 |
I'm pretty sure people do use them in production. Whether we want to |
61 |
really continue supporting them is a separate topic. But the last thing |
62 |
I'd like to do is drop support for arches because of a single upstream |
63 |
maintainer being a Rust lover. |
64 |
|
65 |
-- |
66 |
Best regards, |
67 |
Michał Górny |