Make warnings of renamed and removed lints themselves lints

[brson]

This adds the renamed_and_removed_lints warning, defaulting
to the warning level.

Fixes #31141

[brson]

Nominating for beta because this fixes the unsightly raw_pointer_derive warnings that show up in a lot of cargo build output.

[alexcrichton]

r=me, but the travis failure looks legit:

---- [run-pass] run-pass/ifmt.rs stdout ----

error: compilation failed!
status: exit code: 101
command: x86_64-unknown-linux-gnu/stage2/bin/rustc /home/travis/build/rust-lang/rust/src/test/run-pass/ifmt.rs -L x86_64-unknown-linux-gnu/test/run-pass/ --target=x86_64-unknown-linux-gnu -L x86_64-unknown-linux-gnu/test/run-pass/ifmt.stage2-x86_64-unknown-linux-gnu.run-pass.libaux -C prefer-dynamic -o x86_64-unknown-linux-gnu/test/run-pass/ifmt.stage2-x86_64-unknown-linux-gnu --cfg rtopt -C rpath -O -L x86_64-unknown-linux-gnu/rt
stdout:
------------------------------------------

------------------------------------------
stderr:
------------------------------------------
/home/travis/build/rust-lang/rust/src/test/run-pass/ifmt.rs:15:10: 15:26 error: lint unknown_features has been renamed to unused_features
/home/travis/build/rust-lang/rust/src/test/run-pass/ifmt.rs:15 #![allow(unknown_features)]
                                                                        ^~~~~~~~~~~~~~~~
/home/travis/build/rust-lang/rust/src/test/run-pass/ifmt.rs:13:9: 13:17 note: lint level defined here
/home/travis/build/rust-lang/rust/src/test/run-pass/ifmt.rs:13 #![deny(warnings)]
                                                                       ^~~~~~~~
error: aborting due to previous error

------------------------------------------

thread '[run-pass] run-pass/ifmt.rs' panicked at 'explicit panic', /home/travis/build/rust-lang/rust/src/compiletest/runtest.rs:1651
note: Run with `RUST_BACKTRACE=1` for a backtrace.


failures:
    [run-pass] run-pass/ifmt.rs

[nikomatsakis]

Discussed in @rust-lang/compiler meeting. We are not sure if this is a good candidate for beta uplift or not. It's not fixing a regression per se -- but I guess it will depend on how many such warnings show up. Also, the code is mildly non-trivial (as evidenced by travis failures).

[alexcrichton]

FWIW I still see this warning a few times per day due to crates bringing in older versions of winapi or older versions of libc

[brson]

Looking into the failure.

[brson]

@bors r=alexcrichton

[bors]

📌 Commit bcc291f has been approved by alexcrichton

[bors]

☔ The latest upstream changes (presumably #32309) made this pull request unmergeable. Please resolve the merge conflicts.

[bors]

🔒 Merge conflict

[brson]

@bors r=alexcrichton

[bors]

📌 Commit addde1f has been approved by alexcrichton

[bors]

⌛ Testing commit addde1f with merge dc1f683...

[alexcrichton]

This passed on all but android but apparently that bot disappeared so merging manually.

[pnkfelix]

compiler team discussion concludes that since actively developed (and warning-free) crates should not themselves cause problems downstream, and thus the only problem here is with older version or abandoned crates, combined with fact that such crates will eventually be painless to pull in (once this PR hits stable in <=12 weeks), means that its not crucial to backport this to beta.