Tracking issue for release notes of #136698: Replace i686-unknown-redox target with i586-unknown-redox

[rustbot]

This issue tracks the release notes text for #136698.

Steps

• Proposed text is drafted by PR author (or team) making the noteworthy change.
• Issue is nominated for release team review of clarity for wider audience.
• Release team includes text in release notes/blog posts.

Release notes text

The responsible team for the underlying change should edit this section to replace the automatically generated link with a succinct description of what changed, drawing upon text proposed by the author (either in discussion or through direct editing).

# Category (e.g. Language, Compiler, Libraries, Compatibility notes, ...)
- [Replace i686-unknown-redox target with i586-unknown-redox](https://github.com/rust-lang/rust/pull/136698)

Tip

Use the previous releases categories to help choose which one(s) to use.
The category will be de-duplicated with all the other ones by the release team.

More than one section can be included if needed.

Release blog section

If the change is notable enough for inclusion in the blog post, the responsible team should add content to this section.
Otherwise leave it empty.

cc @jackpot51, @oli-obk -- origin issue/PR authors and assignees for starting to draft text

[briansmith]

If we're going to regress on the presence of CMOV then that is definitely worth putting in the release notes. I think we should reconsider removing CMOV from the baseline though.

[jackpot51]

The base CPU has been changed back to pentiumpro, so there is no change from the original i686 target other than naming.

[briansmith]

The base CPU has been changed back to pentiumpro, so there is no change from the original i686 target other than naming.

Great! Do you forsee an i686 target any time soon? My understanding is that it will be difficult for any <i686 target to become Tier 1 or Tier 2 any time soon.

[jackpot51]

The focus of Redox OS for the 32-bit x86 target is to support 32-bit only hardware, not to compile 32-bit programs that are intended to run on x86_64 capable hardware. This may be a fundamental difference with other 32-bit x86 targets, I am not sure.

A 32-bit target requiring SSE2, which was introduced with the Pentium 4 in 2000, would cover a vanishingly small set of 32-bit only CPUs as support for the x86_64 instruction set was common soon after. AMD's first processors to support SSE2 were the Athlon 64 and Opteron, introduced in 2003, which were also the first to support 64-bit long mode. For Intel, they lagged a bit in adopting the 64-bit instruction set, but there were still only a handful of generations that supported SSE2 but not x86_64.

I understand the concerns about the x87 yielding different results than standard 64-bit floating point math, but I cannot justify a Redox OS target for 32-bit x86 that would only support the very last of the 32-bit only processors.