regression: duplicate matcher binding in macros

[Mark-Simulacrum]

root: nes - 1 detected crates which regressed due to this; cc @TrionProg

 * nes-0.1.1: [start](https://crater-reports.s3.amazonaws.com/beta-1.36-2/1.35.0/reg/nes-0.1.1/log.txt) v. [end](https://crater-reports.s3.amazonaws.com/beta-1.36-2/beta-2019-05-30/reg/nes-0.1.1/log.txt); cc @TrionProg

[petrochenkov]

The regression is expected and was introduced in #59858.
This was a deprecation warning previously.
cc @mark-i-m

[mark-i-m]

This regression was fixed in TrionProg/nes#2

The crate just hasn't been updated.

[mark-i-m]

And FWIW, it fixed a bug in that crate.

[mark-i-m]

@Mark-Simulacrum ^^^

[Mark-Simulacrum]

I'll leave this open for now for the compiler team to decide, but I expect that we'll simply close this as an expected regression (and perhaps the crate's author will publish a new version).

[Centril]

Closing per note in #61481 (comment). I also don't see the language team being OK with reintroducing this bug (I'm not).