Avoid wrong unwrap suggestions for never option value

[chenyukang]

Fixes #162241

[rustbot]

r? @tiif

rustbot has assigned @tiif.
They will have a look at your PR within the next two weeks and either review your PR or reassign to another reviewer.

Use r? to explicitly pick a reviewer

Why was this reviewer chosen?

The reviewer was selected based on:

• Owners of files modified in this PR: compiler
• compiler expanded to 75 candidates
• Random selection from 20 candidates

[hanna-kruppe]

Mostly LGTM, but I'm not sure it makes sense to target only one of the two possible suggestions that this code can emit.

r? me @rustbot author

View changes since this review

[hanna-kruppe]

Does this suggestion make sense when inner_ty is ! or should we suppress the suggestion regardless of sugg_operator? If x: Option<!> then x? must be an unconditional early return, which seems dubious and certainly doesn't contribute to actually getting a value of the expected type.

(If we should suppress this suggestion to, then the control flow could also be a little simpler.)

[chenyukang]

Yes, I added a test and updated the code to remove the invalid suggestion.
This makes me thinking it's a more general issue because a lot of suggestion code are using may_coerce here https://github.com/rust-lang/rust/pull/162276/changes#diff-a188fa3dba59dd225d4a65e08961d2937f6418bc6126716e001303bf946d6ee3L2370, and it may not right for never type in suggestion scenario.