)]}'
{
  "commit": "ef2c71c3cc74108f6aae47fb77912ded4a8aebc2",
  "tree": "f7b65db8dfd4c1bed1ad0aea35263db51870186f",
  "parents": [
    "297cdc8137480637f97c3529ae2caa9a0788381b",
    "023f38fe73f5e392ea94cc15e53611bd2074302e"
  ],
  "author": {
    "name": "Jonathan Brouwer",
    "email": "jonathantbrouwer@gmail.com",
    "time": "Tue Dec 16 20:21:08 2025 +0100"
  },
  "committer": {
    "name": "GitHub",
    "email": "noreply@github.com",
    "time": "Tue Dec 16 20:21:08 2025 +0100"
  },
  "message": "Rollup merge of #149967 - folkertdev:va-list-hexagon, r\u003dworkingjubilee\n\ncustom `VaList` layout for Hexagon\n\nI noticed while browsing LLVM source that we use an incorrect `VaList` definition for the musl hexagon target.\n\nrelevant links\n\n- https://github.com/llvm/llvm-project/blob/0cdc1b6dd4a870fc41d4b15ad97e0001882aba58/clang/include/clang/Basic/TargetInfo.h#L333\n- https://github.com/llvm/llvm-project/blob/0cdc1b6dd4a870fc41d4b15ad97e0001882aba58/clang/lib/CodeGen/Targets/Hexagon.cpp#L407-L417\n\ncc target maintainer `@androm3da` can you confirm that this looks OK? In particular the `#[rustc_pass_indirectly_in_non_rustic_abis]` attribute is used to simulate pointer decay (like if the struct were wrapped in a 1-element array in C). The clang comment suggests that the Tag is wrapped in such a single-element array, but I haven\u0027t actually been able to confirm it.\n\nFor stabilizing `c_variadic` (on the hexagon targets) we will also need a custom `va_arg` implementation to mirror the one in `clang` in [va_arg.rs](https://github.com/rust-lang/rust/blob/main/compiler/rustc_codegen_llvm/src/va_arg.rs). Would you be able to contribute one?\n\nr? `@workingjubilee`\n",
  "tree_diff": []
}