)]}'
{
  "commit": "ff086354c9fc93e1da1d2f4d255456624fbcd099",
  "tree": "5249eeb3b9d81ef2749ef3e3ed559c31aefe4d57",
  "parents": [
    "f82485388963286763f8c8a9261ac227627f6a18",
    "6a07088b1b8659d28bec14695388f323a3fa323c"
  ],
  "author": {
    "name": "bors",
    "email": "bors@rust-lang.org",
    "time": "Fri Mar 06 10:21:32 2026 +0000"
  },
  "committer": {
    "name": "bors",
    "email": "bors@rust-lang.org",
    "time": "Fri Mar 06 10:21:32 2026 +0000"
  },
  "message": "Auto merge of #153316 - nnethercote:rm-with_related_context, r\u003doli-obk\n\nRemove `tls::with_related_context`.\n\nThis function gets the current `ImplicitCtxt` and checks that its `tcx` matches the passed-in `tcx`. It\u0027s an extra bit of sanity checking: when you already have a `tcx`, and you need access to the non-`tcx` parts of `ImplicitCtxt`, check that your `tcx` matches the one in `ImplicitCtxt`.\r\n\r\nHowever, it\u0027s only used in two places: `start_query` and `current_query_job`. The non-checked alternatives (`with_context`, `with_context_opt`) are used in more places, including some where a `tcx` is available. And things would have to go catastrophically wrong for the check to fail -- e.g. if we somehow end up with multiple `TyCtxt`s. In my opinion it\u0027s just an extra case to understand in `tls.rs` that adds little value.\r\n\r\nThis commit removes it. This avoids the need for `tcx` parameters in a couple of places. The commit also adjusts how `start_query` sets up its `ImplicitCtxt` to more closely match how similar functions do it, i.e. with `..icx.clone()` for the unchanged fields.\r\n\r\nr? @oli-obk",
  "tree_diff": []
}