Tracking Issue for "vectorcall" ABI

[beetrees]

This is a tracking issue for the "vectorcall" and "vectorcall-unwind" ABIs.
The feature gate for the issue is #![feature(abi_vectorcall)].

(Note that this feature was added in 2015, but didn't have a tracking issue until now.)

About tracking issues

Tracking issues are used to record the overall progress of implementation.
They are also used as hubs connecting to other relevant issues, e.g., bugs or open design questions.
A tracking issue is however not meant for large scale discussion, questions, or bug reports about a feature.
Instead, open a dedicated issue for the specific matter and add the relevant feature gate label.
Discussion comments will get marked as off-topic or deleted.
Repeated discussions on the tracking issue may lead to the tracking issue getting locked.

Steps

• Implement (llvm: Add support for vectorcall (X86_VectorCall) convention #30567)
• Adjust documentation (see instructions on rustc-dev-guide)
• Formatting for new syntax has been added to the Style Guide (nightly-style-procedure)
• Stabilization PR (see instructions on rustc-dev-guide)

Unresolved Questions

Currently an LLVM error is emitted if there is an attempt to use the ABI on a target without the required target features (e.g. i586), but only if the specific usage of the ABI would require use of the disabled registers (e.g. on i586 targets extern "vectorcall" fn f() -> i64 { ... } will compile fine but extern "vectorcall" fn f() -> f64 { ... } will cause an LLVM error). Should the compiler always error if the ABI is used and SSE2 is disabled? According to https://learn.microsoft.com/en-us/cpp/cpp/vectorcall the ABI is only supported when SSE2 is enabled. Closely related to #116558.

Implementation history

"vectorcall" added in #30567.
"vectorcall-unwind" added in #93561.

[beetrees]

@rustbot label +A-abi +O-windows +T-lang