Rollup merge of #132675 - Zalathar:empty-spans, r=jieyouxu

coverage: Restrict empty-span expansion to only cover `{` and `}`

Coverage instrumentation has some tricky code for converting a coverage-relevant `Span` into a set of start/end line/byte-column coordinates that will be embedded in the CGU's coverage metadata.

A big part of this complexity is special code for handling empty spans, which are expanded into non-empty spans (if possible) because LLVM's coverage reporter does not handle empty spans well.

This PR simplifies that code by restricting it to only apply in two specific situations: when the character after the empty span is `{`, or the character before the empty span is `}`.

(As an added benefit, this means that the expanded spans no longer extend awkwardly beyond the end of a physical line, which was common under the previous implementation.)

Along the way, this PR also removes some unhelpful code for dealing with function source code spread across multiple files. Functions currently can't have coverage spans in multiple files, and if that ever changes (e.g. to properly support expansion regions) then this code will need to be completely overhauled anyway.

Author: matthiaskrgr

compiler/rustc_codegen_llvm/src/coverageinfo/ffi.rs

@@ -1,5 +1,7 @@
 use rustc_middle::mir::coverage::{CounterId, CovTerm, ExpressionId, SourceRegion};
 
+use crate::coverageinfo::mapgen::LocalFileId;
+
 /// Must match the layout of `LLVMRustCounterKind`.
 #[derive(Copy, Clone, Debug)]
 #[repr(C)]
@@ -137,8 +139,12 @@ pub(crate) struct CoverageSpan {
 }
 
 impl CoverageSpan {
-    pub(crate) fn from_source_region(file_id: u32, code_region: &SourceRegion) -> Self {
-        let &SourceRegion { file_name: _, start_line, start_col, end_line, end_col } = code_region;
+    pub(crate) fn from_source_region(
+        local_file_id: LocalFileId,
+        code_region: &SourceRegion,
+    ) -> Self {
+        let file_id = local_file_id.as_u32();
+        let &SourceRegion { start_line, start_col, end_line, end_col } = code_region;
         // Internally, LLVM uses the high bit of `end_col` to distinguish between
         // code regions and gap regions, so it can't be used by the column number.
         assert!(end_col & (1u32 << 31) == 0, "high bit of `end_col` must be unset: {end_col:#X}");

compiler/rustc_codegen_llvm/src/coverageinfo/map_data.rs

@@ -6,7 +6,6 @@ use rustc_middle::mir::coverage::{
     SourceRegion,
 };
 use rustc_middle::ty::Instance;
-use rustc_span::Symbol;
 use tracing::{debug, instrument};
 
 use crate::coverageinfo::ffi::{Counter, CounterExpression, ExprKind};
@@ -180,7 +179,7 @@ impl<'tcx> FunctionCoverageCollector<'tcx> {
 }
 
 pub(crate) struct FunctionCoverage<'tcx> {
-    function_coverage_info: &'tcx FunctionCoverageInfo,
+    pub(crate) function_coverage_info: &'tcx FunctionCoverageInfo,
     is_used: bool,
 
     counters_seen: BitSet<CounterId>,
@@ -199,11 +198,6 @@ impl<'tcx> FunctionCoverage<'tcx> {
         if self.is_used { self.function_coverage_info.function_source_hash } else { 0 }
     }
 
-    /// Returns an iterator over all filenames used by this function's mappings.
-    pub(crate) fn all_file_names(&self) -> impl Iterator<Item = Symbol> + Captures<'_> {
-        self.function_coverage_info.mappings.iter().map(|mapping| mapping.source_region.file_name)
-    }
-
     /// Convert this function's coverage expression data into a form that can be
     /// passed through FFI to LLVM.
     pub(crate) fn counter_expressions(