ast_lowering: Introduce lower_span for catching all spans entering HIR

[petrochenkov]

This PR cherry-picks the fn lower_span change from #84373.
I also introduced fn lower_ident for lowering spans in identifiers, and audited places where HIR structures with spans or identifiers are constructed and added a few missing lower_spans/lower_idents.

Having a hook for spans entering HIR can be useful for things other than #84373, e.g. #35148.
I also want to check whether this change causes perf regressions due to some accidental inlining issues.

r? @cjgillot

[petrochenkov]

@bors try @rust-timer queue

[rust-timer]

Awaiting bors try build completion.

@rustbot label: +S-waiting-on-perf

[bors]

⌛ Trying commit a611199277e1ee5f0af204d9d4132ee6c65ce811 with merge 6f99ad894b945a464742e58a6d3cbae7b7a382a6...

[bors]

☀️ Try build successful - checks-actions
Build commit: 6f99ad894b945a464742e58a6d3cbae7b7a382a6 (6f99ad894b945a464742e58a6d3cbae7b7a382a6)

[rust-timer]

Queued 6f99ad894b945a464742e58a6d3cbae7b7a382a6 with parent db002a0, future comparison URL.

[rust-timer]

Finished benchmarking try commit (6f99ad894b945a464742e58a6d3cbae7b7a382a6): comparison url.

Summary: This benchmark run did not return any relevant changes.

If you disagree with this performance assessment, please file an issue in rust-lang/rustc-perf.

Benchmarking this pull request likely means that it is perf-sensitive, so we're automatically marking it as not fit for rolling up.

@bors rollup=never
@rustbot label: +S-waiting-on-review -S-waiting-on-perf -perf-regression

[petrochenkov]

So this is not a source of regressions in #84373.
@bors r=maybe

[bors]

📌 Commit a611199277e1ee5f0af204d9d4132ee6c65ce811 has been approved by maybe

[petrochenkov]

@bors r-
@bors rollup=maybe

[Aaron1011]

It seems like it would be very easy to forget to call lower_span somewhere, since the input and output types are both Span. What will happen to relative span encoding if there's an 'unlowered' Span?

[petrochenkov]

@Aaron1011

What will happen to relative span encoding if there's an 'unlowered' Span?

Not setting the span's DefId parent means giving up on the specific incremental optimization introduced by #84373, but otherwise it doesn't affect correctness.
Spans decoded from other crates don't get parent DefIds at all, for example.

[cjgillot]

Normally, we would only have a spurious red DepNode and query recomputation.

[cjgillot]

I am not really the most neutral reviewer as I essentially wrote the same code. OTOH, this PR is harmless, up to making LLVM chew more code. @Aaron1011 do you have any concern about merging this?

[petrochenkov]

r? @Aaron1011

[Aaron1011]

r=me with the merge conflicts fixed

[petrochenkov]

@bors r=Aaron1011

[bors]

📌 Commit 59013cd has been approved by Aaron1011

[bors]

⌛ Testing commit 59013cd with merge ef52471...

[bors]

☀️ Test successful - checks-actions
Approved by: Aaron1011
Pushing ef52471 to master...