Fix several issues during cross compiling

[topjohnwu]

• When cross compiling LLVM on an arm64 macOS machine to x86_64, CMake will produce universal binaries by default, causing link errors. Explicitly set CMAKE_OSX_ARCHITECTURES to the one single target architecture so that the executables and libraries will be single architecture.
• When cross compiling rustc with llvm.clang = true, CLANG_TABLEGEN has to be set to the host clang-tblgen executable to build clang.

[rust-highfive]

Thanks for the pull request, and welcome! The Rust team is excited to review your changes, and you should hear from @Mark-Simulacrum (or someone else) soon.

Please see the contribution instructions for more information.

[topjohnwu]

@jyn514 since you've reviewed my other PR, would you want to check this out too?

[jyn514]

When cross compiling LLVM on an arm64 macOS machine to x86_64, CMake will produce universal binaries by default, causing link errors.

Why does this cause link errors?

[topjohnwu]

I don't actually know why, but the linker will complain file too small for architecture. I think it's quite important that we do NOT build universal binaries for LLVM, as the rest of rustc does not.

[jyn514]

Hmm, ok, seems good to have LLVM's build be consistent with rustc's.

[jyn514]

@bors r+

[bors]

📌 Commit 91c7dd0 has been approved by jyn514

[JohnTitor]

Failed in rollup: #98442 (comment)
@bors r-

[topjohnwu]

@jyn514 turns out that LLVM_NM might still be required for cross compiling on MSVC. I added it back, hopefully this time all ci tests will pass 🙏

@JohnTitor I have pushed a new update to the PR

[jyn514]

@bors r+

[bors]

📌 Commit 7180afb has been approved by jyn514

[topjohnwu]

@jyn514 hello, I have added a few changes to my PR

• Only check the existence of clang-tblgen if not during a dry run
• Add a fix to the libLLVM.dylib symlink workaround. More details in the commit message

r? @jyn514

[jyn514]

@bors r-

@topjohnwu can we please merge the existing changes before you add more? You can make a new PR if you want to get a review before this one is merged.

[topjohnwu]

@jyn514 sure, I'll split the 3rd commit into its own PR

[jyn514]

@bors r+

[bors]

📌 Commit 600026a has been approved by jyn514

[bors]

⌛ Testing commit 600026a with merge 73443a0...

[alex]

@jyn514 fwiw the source of those errors is #55235

[bors]

☀️ Test successful - checks-actions
Approved by: jyn514
Pushing 73443a0 to master...

[rust-timer]

Finished benchmarking commit (73443a0): comparison url.

Instruction count

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

Max RSS (memory usage)

Results

• Primary benchmarks: no relevant changes found
• Secondary benchmarks: 🎉 relevant improvements found

	mean1	max	count2
Regressions 😿 (primary)	N/A	N/A	0
Regressions 😿 (secondary)	N/A	N/A	0
Improvements 🎉 (primary)	N/A	N/A	0
Improvements 🎉 (secondary)	-4.6%	-8.7%	4
All 😿🎉 (primary)	N/A	N/A	0

Cycles

Results

• Primary benchmarks: mixed results
• Secondary benchmarks: 🎉 relevant improvement found

	mean1	max	count2
Regressions 😿 (primary)	3.5%	3.7%	2
Regressions 😿 (secondary)	N/A	N/A	0
Improvements 🎉 (primary)	-7.1%	-7.1%	1
Improvements 🎉 (secondary)	-2.2%	-2.2%	1
All 😿🎉 (primary)	-0.1%	-7.1%	3

If you disagree with this performance assessment, please file an issue in rust-lang/rustc-perf.

@rustbot label: -perf-regression

Footnotes

1. the arithmetic mean of the percent change ↩ ↩²

2. number of relevant changes ↩ ↩²