rust/tests/codegen/enum/enum-aggregate.rs, branch perf-tmphttps://github.com/rust-lang/rust
http://git.dreamy.place/mirrors/rust/atom?h=perf-tmp2025-07-22T12:28:48+00:00Rename `tests/codegen` into `tests/codegen-llvm`2025-07-22T12:28:48+00:00Guillaume Gomezguillaume1.gomez@gmail.com2025-07-21T12:34:12+00:00urn:sha1:a27f3e3fd1e4d16160f8885b6b06665b5319f56cLet `rvalue_creates_operand` return true for *all* `Rvalue::Aggregate`s2025-07-08T06:08:10+00:00Scott McMurrayscottmcm@users.noreply.github.com2025-07-05T06:16:41+00:00urn:sha1:8cf2c71243f8c093e7728370e6653aa940dbaf22
Inspired by <https://github.com/rust-lang/rust/pull/138759#discussion_r2156375342> where I noticed that we were nearly at this point, plus the comments I was writing in 143410 that reminded me a type-dependent `true` is fine.
This PR splits the `OperandRef::builder` logic out to a separate type, with the updates needed to handle SIMD as well. In doing so, that makes the existing `Aggregate` path in `codegen_rvalue_operand` capable of handing SIMD values just fine.
As a result, we no longer need to do layout calculations for aggregate result types when running the analysis to determine which things can be SSA in codegen.
Allow all MIR `Aggregate`s to take the operand path (if layout permits)2025-07-04T05:53:19+00:00Scott McMurrayscottmcm@users.noreply.github.com2025-06-19T06:32:15+00:00urn:sha1:caeacba78a790031ebf262ef952f79795fdc9cc9