JSON target specs: remove 'x86-softfloat' compatibility alias

[RalfJung]

This got introduced in #151154 as a compatibility alias. It's been a few months, so seems fine to remove the compatibility alias now. JSON targets are anyway unstable.

Cc @Darksonn @ojeda as this may affect Linux

[rustbot]

These commits modify compiler targets.
(See the Target Tier Policy.)

The Miri subtree was changed

cc @rust-lang/miri

[rustbot]

r? @adwinwhite

rustbot has assigned @adwinwhite.
They will have a look at your PR within the next two weeks and either review your PR or reassign to another reviewer.

Use r? to explicitly pick a reviewer

Why was this reviewer chosen?

The reviewer was selected based on:

• Owners of files modified in this PR: compiler
• compiler expanded to 73 candidates
• Random selection from 17 candidates

[Darksonn]

The relevant part of the RfL target is here.

if cfg.rustc_version_atleast(1, 86, 0) {
    ts.push("rustc-abi", "x86-softfloat");
}

It sounds like this should be updated to:

if cfg.rustc_version_atleast(1, 95, 0) {
    ts.push("rustc-abi", "softfloat");
} else if cfg.rustc_version_atleast(1, 86, 0) {
    ts.push("rustc-abi", "x86-softfloat");
}

[RalfJung]

Yes, that looks right. And I guess we need a patch somewhere in the rustc tree to satisfy the RfL CI job. I'd appreciate help with preparing that patch. :)

[ojeda]

Sounds good, I will prepare it.

[ojeda]

Please cherry-pick: ojeda@90f7e1c (not tested yet, though).

I used >= 1.98.0 since I would prefer to avoid changing how we build in existing compilers unless there is a reason not to (since we will need to backport it).

@rustbot label A-rust-for-linux

[RalfJung]

I used >= 1.98.0 since I would prefer to avoid changing how we build in existing compilers unless there is a reason not to (since we will need to backport it).

Ether way seems entirely fine for me. This is mapped to the same constant in the compiler already while parsing the JSON so there's absolutely no difference between the two values for compilers that accept both.

[ojeda]

try-job: x86_64-rust-for-linux
@bors try

[RalfJung]

That won't work. You need to put the try-job: into the PR description or
@bors try job=x86_64-rust-for-linux

[ojeda]

Ah, I always try from a PR of my own, so I didn't know the try-job wouldn't be read from a comment by the bot, thanks!

Ok, I tested it locally, it works, so the CI here should also pass :)

[rust-bors]

☀️ Try build successful (CI)
Build commit: 6110fc5 (6110fc500e7c14a83bc0581d04e16b8c3b2556e5, parent: 6eda7419e71fdbc1185ed5be7e6bff1a474ab5cd)