Apple: Refactor deployment target version parsing

[madsmtm]

Refactor deployment target parsing to make it easier to do #129342 (I wanted to make sure of all the places that std::env::var is called).

Specifically, my goal was to minimize the amount of target-specific configuration, so to that end I renamed the opts function that generates the TargetOptions to base, and made it return the LLVM target and target_arch too. In the future, I would like to move even more out of the target files and into spec::apple, as it makes it easier for me to maintain.

For example, this fixed a bug in aarch64-apple-watchos, which wasn't passing the deployment target as part of the LLVM triple. This (probably) fixes #123582 and fixes #107630.

We also now parse the patch version of deployment targets, allowing the user to specify e.g. MACOSX_DEPLOYMENT_TARGET=10.12.6.

Finally, this fixes the LLVM target name for visionOS, it should be *-apple-xros and not *-apple-visionos.

Since I have changed all the Apple targets here, I smoke-tested my changes by running the following:

# Build each target
./x build library --target="aarch64-apple-darwin,aarch64-apple-ios,aarch64-apple-ios-macabi,aarch64-apple-ios-sim,aarch64-apple-tvos,aarch64-apple-tvos-sim,aarch64-apple-visionos,aarch64-apple-visionos-sim,aarch64-apple-watchos,aarch64-apple-watchos-sim,arm64_32-apple-watchos,arm64e-apple-ios,armv7k-apple-watchos,armv7s-apple-ios,i386-apple-ios,x86_64-apple-darwin,x86_64-apple-ios,x86_64-apple-ios-macabi,x86_64-apple-tvos,x86_64-apple-watchos-sim,x86_64h-apple-darwin"

# Test that we can still at least link basic projects
cargo new foobar && cd foobar && cargo +stage1 build --target=aarch64-apple-darwin --target=aarch64-apple-ios --target=aarch64-apple-ios-macabi --target=aarch64-apple-ios-sim --target=aarch64-apple-tvos --target=aarch64-apple-tvos-sim --target=aarch64-apple-visionos --target=aarch64-apple-visionos-sim --target=aarch64-apple-watchos --target=aarch64-apple-watchos-sim --target=arm64_32-apple-watchos --target=armv7s-apple-ios --target=i386-apple-ios --target=x86_64-apple-darwin --target=x86_64-apple-ios --target=x86_64-apple-ios-macabi --target=x86_64-apple-tvos --target=x86_64-apple-watchos-sim --target=x86_64h-apple-darwin

I couldn't build for the arm64e-apple-darwin target, the armv7k-apple-watchos and arm64e-apple-ios targets failed to link, and I know that the i686-apple-darwin target requires a bit of setup, but all of this is as it was before this PR.

r? thomcc

CC @BlackHoleFox

I would recommend using rollup=never when merging this, in case we need to bisect this later.

[rustbot]

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

[madsmtm]

@bors rollup=never

There are quite a few changes here, and we might need to bisect this later.

EDIT: Oh well, I don't have permission to configure that, but I'd still recommend it when merging.

[bors]

@madsmtm: 🔑 Insufficient privileges: not in try users

[thomcc]

@bors rollup=never

[thomcc]

This looks fine to me. Was a hassle to verify all the targets match what they used to.

[thomcc]

I'm not a compiler-team or compiler-contributor so I'll roll to someone who is, but I'd have R+ed this if I were.
r? compiler

[madsmtm]

r? compiler

[madsmtm]

Thanks for the review!

@rustbot ready

[BlackHoleFox]

Since this has been looked at someone on the compiler team, here's my lookover from the Apple perspective since I was CCed.

[petrochenkov]

@bors r+

[bors]

📌 Commit bd56857 has been approved by petrochenkov

It is now in the queue for this repository.

[bors]

⌛ Testing commit bd56857 with merge 9afe713...

[bors]

☀️ Test successful - checks-actions
Approved by: petrochenkov
Pushing 9afe713 to master...

[rust-timer]

Finished benchmarking commit (9afe713): comparison URL.

Overall result: no relevant changes - no action needed

@rustbot label: -perf-regression

Instruction count

This benchmark run did not return any relevant results for this metric.

Max RSS (memory usage)

Results (primary 2.7%)

This is a less reliable metric that may be of interest but was not used to determine the overall result at the top of this comment.

	mean	range	count
Regressions ❌ (primary)	2.7%	[2.7%, 2.7%]	1
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	2.7%	[2.7%, 2.7%]	1

Cycles

This benchmark run did not return any relevant results for this metric.

Binary size

This benchmark run did not return any relevant results for this metric.

Bootstrap: 756.217s -> 756.798s (0.08%)
Artifact size: 341.14 MiB -> 341.03 MiB (-0.03%)