Update rust-analyzer

[lnicola]

to https://github.com/rust-analyzer/rust-analyzer/releases/tag/2021-01-11

[rust-highfive]

r? @Mark-Simulacrum

(rust-highfive has picked a reviewer for you, use r? to override)

[rust-highfive]

⚠️ Warning ⚠️

• These commits modify submodules.

[jonas-schievink]

@bors r+

[bors]

📌 Commit 61a6468 has been approved by jonas-schievink

[m-ou-se]

@bors r-

This failed in #81102 (comment)

error: could not compile `salsa`

That seems to be a rust-analyzer dependency.

Full logs here: https://github.com/rust-lang-ci/rust/runs/1715493685

[lnicola]

salsa compiles fine with the latest nightly. Could another PR have introduced those errors?

[lnicola]

Probably #80732.

[m-ou-se]

Could another PR have introduced those errors?

Maybe. I just searched for salsa because of that error message, which popped up a lot in rust-analyzer. Note that it failed on --target arm-unknown-linux-gnueabihf, which might be relevant.

If you think this PR is not the source of the problem, we can just try it again and see what happens.

[lnicola]

The target seems less relevant, that looks like a type error. I tried cargo +nightly build --target arm-unknown-linux-gnueabihf on salsa 0.16 and it works.

I thought #80732 might be related to this because the test files say "ambigunous", but the error messages are not identical.

[lnicola]

Confirmed in #80732 (comment).

[lnicola]

This needs another bors r+, but there'll be another rust-analyzer release tagged tonight, so I suppose we can wait until that.

[m-ou-se]

salsa 0.15.2 (the one used before this PR) compiles fine with #80732. So if you're not in a hurry, I'd prefer waiting a bit until we figured out if salsa 0.16 or #80732 is to blame, as only the combination of the two fails.