Auto merge of #85385 - richkadel:simpler-simplify-with-coverage, r=wesleywiser

Reland - Report coverage `0` of dead blocks

Fixes: #84018

With `-Z instrument-coverage`, coverage reporting of dead blocks
(for example, blocks dropped because a conditional branch is dropped,
based on const evaluation) is now supported.

Note, this PR relands an earlier, reverted PR that failed when compiling
generators. The prior issues with generators has been resolved and a new
test was added to prevent future regressions.

Check out the resulting changes to test coverage of dead blocks in the
test coverage reports in this PR.

r? `@tmandry`
fyi: `@wesleywiser`

12 files changed, 100 insertions, 16 deletions

diff --git a/compiler/rustc_codegen_ssa/src/coverageinfo/map.rs b/compiler/rustc_codegen_ssa/src/coverageinfo/map.rs
index 08442c588f8..c1dfe1ef856 100644
--- a/compiler/rustc_codegen_ssa/src/coverageinfo/map.rs
+++ b/compiler/rustc_codegen_ssa/src/coverageinfo/map.rs
@@ -28,6 +28,7 @@ pub struct Expression {
 /// only whitespace or comments). According to LLVM Code Coverage Mapping documentation, "A count
 /// for a gap area is only used as the line execution count if there are no other regions on a
 /// line."
+#[derive(Debug)]
 pub struct FunctionCoverage<'tcx> {
     instance: Instance<'tcx>,
     source_hash: u64,
@@ -113,6 +114,14 @@ impl<'tcx> FunctionCoverage<'tcx> {
             expression_id, lhs, op, rhs, region
         );
         let expression_index = self.expression_index(u32::from(expression_id));
+        debug_assert!(
+            expression_index.as_usize() < self.expressions.len(),
+            "expression_index {} is out of range for expressions.len() = {}
+            for {:?}",
+            expression_index.as_usize(),
+            self.expressions.len(),
+            self,
+        );
         if let Some(previous_expression) = self.expressions[expression_index].replace(Expression {
             lhs,
             op,
diff --git a/compiler/rustc_mir/src/transform/const_goto.rs b/compiler/rustc_mir/src/transform/const_goto.rs
index b5c8b4bebc3..ba10b54c5ae 100644
--- a/compiler/rustc_mir/src/transform/const_goto.rs
+++ b/compiler/rustc_mir/src/transform/const_goto.rs
@@ -47,7 +47,7 @@ impl<'tcx> MirPass<'tcx> for ConstGoto {
         // if we applied optimizations, we potentially have some cfg to cleanup to
         // make it easier for further passes
         if should_simplify {
-            simplify_cfg(body);
+            simplify_cfg(tcx, body);
             simplify_locals(body, tcx);
         }
     }
diff --git a/compiler/rustc_mir/src/transform/deduplicate_blocks.rs b/compiler/rustc_mir/src/transform/deduplicate_blocks.rs
index c41e71e09a4..912505c6598 100644
--- a/compiler/rustc_mir/src/transform/deduplicate_blocks.rs
+++ b/compiler/rustc_mir/src/transform/deduplicate_blocks.rs
@@ -26,7 +26,7 @@ impl<'tcx> MirPass<'tcx> for DeduplicateBlocks {
         if has_opts_to_apply {
             let mut opt_applier = OptApplier { tcx, duplicates };
             opt_applier.visit_body(body);
-            simplify_cfg(body);
+            simplify_cfg(tcx, body);
         }
     }
 }
diff --git a/compiler/rustc_mir/src/transform/early_otherwise_branch.rs b/compiler/rustc_mir/src/transform/early_otherwise_branch.rs
index 7934d4ba849..07127042fa4 100644
--- a/compiler/rustc_mir/src/transform/early_otherwise_branch.rs
+++ b/compiler/rustc_mir/src/transform/early_otherwise_branch.rs
@@ -164,7 +164,7 @@ impl<'tcx> MirPass<'tcx> for EarlyOtherwiseBranch {
         // Since this optimization adds new basic blocks and invalidates others,
         // clean up the cfg to make it nicer for other passes
         if should_cleanup {
-            simplify_cfg(body);
+            simplify_cfg(tcx, body);
         }
     }
 }
diff --git a/compiler/rustc_mir/src/transform/generator.rs b/compiler/rustc_mir/src/transform/generator.rs
index 003003a8abb..3560b4b1e86 100644
--- a/compiler/rustc_mir/src/transform/generator.rs
+++ b/compiler/rustc_mir/src/transform/generator.rs
@@ -964,7 +964,7 @@ fn create_generator_drop_shim<'tcx>(
 
     // Make sure we remove dead blocks to remove
     // unrelated code from the resume part of the function
-    simplify::remove_dead_blocks(&mut body);
+    simplify::remove_dead_blocks(tcx, &mut body);
 
     dump_mir(tcx, None, "generator_drop", &0, &body, |_, _| Ok(()));
 
@@ -1137,7 +1137,7 @@ fn create_generator_resume_function<'tcx>(
 
     // Make sure we remove dead blocks to remove
     // unrelated code from the drop part of the function
-    simplify::remove_dead_blocks(body);
+    simplify::remove_dead_blocks(tcx, body);
 
     dump_mir(tcx, None, "generator_resume", &0, body, |_, _| Ok(()));
 }
diff --git a/compiler/rustc_mir/src/transform/inline.rs b/compiler/rustc_mir/src/transform/inline.rs
index b6f80763bc8..f1c95a84ade 100644
--- a/compiler/rustc_mir/src/transform/inline.rs
+++ b/compiler/rustc_mir/src/transform/inline.rs
@@ -57,7 +57,7 @@ impl<'tcx> MirPass<'tcx> for Inline {
         if inline(tcx, body) {
             debug!("running simplify cfg on {:?}", body.source);
             CfgSimplifier::new(body).simplify();
-            remove_dead_blocks(body);
+            remove_dead_blocks(tcx, body);
         }
     }
 }
diff --git a/compiler/rustc_mir/src/transform/match_branches.rs b/compiler/rustc_mir/src/transform/match_branches.rs
index f7a9835353e..21b208a08c2 100644
--- a/compiler/rustc_mir/src/transform/match_branches.rs
+++ b/compiler/rustc_mir/src/transform/match_branches.rs
@@ -167,7 +167,7 @@ impl<'tcx> MirPass<'tcx> for MatchBranchSimplification {
         }
 
         if should_cleanup {
-            simplify_cfg(body);
+            simplify_cfg(tcx, body);
         }
     }
 }
diff --git a/compiler/rustc_mir/src/transform/multiple_return_terminators.rs b/compiler/rustc_mir/src/transform/multiple_return_terminators.rs
index 4aaa0baa9f4..cd2db180552 100644
--- a/compiler/rustc_mir/src/transform/multiple_return_terminators.rs
+++ b/compiler/rustc_mir/src/transform/multiple_return_terminators.rs
@@ -38,6 +38,6 @@ impl<'tcx> MirPass<'tcx> for MultipleReturnTerminators {
             }
         }
 
-        simplify::remove_dead_blocks(body)
+        simplify::remove_dead_blocks(tcx, body)
     }
 }
diff --git a/compiler/rustc_mir/src/transform/remove_unneeded_drops.rs b/compiler/rustc_mir/src/transform/remove_unneeded_drops.rs
index 5144d48750d..02e45021a0a 100644
--- a/compiler/rustc_mir/src/transform/remove_unneeded_drops.rs
+++ b/compiler/rustc_mir/src/transform/remove_unneeded_drops.rs
@@ -36,7 +36,7 @@ impl<'tcx> MirPass<'tcx> for RemoveUnneededDrops {
         // if we applied optimizations, we potentially have some cfg to cleanup to
         // make it easier for further passes
         if should_simplify {
-            simplify_cfg(body);
+            simplify_cfg(tcx, body);
         }
     }
 }
diff --git a/compiler/rustc_mir/src/transform/simplify.rs b/compiler/rustc_mir/src/transform/simplify.rs
index 65e2d096b20..7aebca77e6f 100644
--- a/compiler/rustc_mir/src/transform/simplify.rs
+++ b/compiler/rustc_mir/src/transform/simplify.rs
@@ -29,6 +29,7 @@
 
 use crate::transform::MirPass;
 use rustc_index::vec::{Idx, IndexVec};
+use rustc_middle::mir::coverage::*;
 use rustc_middle::mir::visit::{MutVisitor, MutatingUseContext, PlaceContext, Visitor};
 use rustc_middle::mir::*;
 use rustc_middle::ty::TyCtxt;
@@ -46,9 +47,9 @@ impl SimplifyCfg {
     }
 }
 
-pub fn simplify_cfg(body: &mut Body<'_>) {
+pub fn simplify_cfg(tcx: TyCtxt<'tcx>, body: &mut Body<'_>) {
     CfgSimplifier::new(body).simplify();
-    remove_dead_blocks(body);
+    remove_dead_blocks(tcx, body);
 
     // FIXME: Should probably be moved into some kind of pass manager
     body.basic_blocks_mut().raw.shrink_to_fit();
@@ -59,9 +60,9 @@ impl<'tcx> MirPass<'tcx> for SimplifyCfg {
         Cow::Borrowed(&self.label)
     }
 
-    fn run_pass(&self, _tcx: TyCtxt<'tcx>, body: &mut Body<'tcx>) {
+    fn run_pass(&self, tcx: TyCtxt<'tcx>, body: &mut Body<'tcx>) {
         debug!("SimplifyCfg({:?}) - simplifying {:?}", self.label, body.source);
-        simplify_cfg(body);
+        simplify_cfg(tcx, body);
     }
 }
 
@@ -286,7 +287,7 @@ impl<'a, 'tcx> CfgSimplifier<'a, 'tcx> {
     }
 }
 
-pub fn remove_dead_blocks(body: &mut Body<'_>) {
+pub fn remove_dead_blocks(tcx: TyCtxt<'tcx>, body: &mut Body<'_>) {
     let reachable = traversal::reachable_as_bitset(body);
     let num_blocks = body.basic_blocks().len();
     if num_blocks == reachable.count() {
@@ -306,6 +307,11 @@ pub fn remove_dead_blocks(body: &mut Body<'_>) {
         }
         used_blocks += 1;
     }
+
+    if tcx.sess.instrument_coverage() {
+        save_unreachable_coverage(basic_blocks, used_blocks);
+    }
+
     basic_blocks.raw.truncate(used_blocks);
 
     for block in basic_blocks {
@@ -315,6 +321,75 @@ pub fn remove_dead_blocks(body: &mut Body<'_>) {
     }
 }
 
+/// Some MIR transforms can determine at compile time that a sequences of
+/// statements will never be executed, so they can be dropped from the MIR.
+/// For example, an `if` or `else` block that is guaranteed to never be executed
+/// because its condition can be evaluated at compile time, such as by const
+/// evaluation: `if false { ... }`.
+///
+/// Those statements are bypassed by redirecting paths in the CFG around the
+/// `dead blocks`; but with `-Z instrument-coverage`, the dead blocks usually
+/// include `Coverage` statements representing the Rust source code regions to
+/// be counted at runtime. Without these `Coverage` statements, the regions are
+/// lost, and the Rust source code will show no coverage information.
+///
+/// What we want to show in a coverage report is the dead code with coverage
+/// counts of `0`. To do this, we need to save the code regions, by injecting
+/// `Unreachable` coverage statements. These are non-executable statements whose
+/// code regions are still recorded in the coverage map, representing regions
+/// with `0` executions.
+fn save_unreachable_coverage(
+    basic_blocks: &mut IndexVec<BasicBlock, BasicBlockData<'_>>,
+    first_dead_block: usize,
+) {
+    let has_live_counters = basic_blocks.raw[0..first_dead_block].iter().any(|live_block| {
+        live_block.statements.iter().any(|statement| {
+            if let StatementKind::Coverage(coverage) = &statement.kind {
+                matches!(coverage.kind, CoverageKind::Counter { .. })
+            } else {
+                false
+            }
+        })
+    });
+    if !has_live_counters {
+        // If there are no live `Counter` `Coverage` statements anymore, don't
+        // move dead coverage to the `START_BLOCK`. Just allow the dead
+        // `Coverage` statements to be dropped with the dead blocks.
+        //
+        // The `generator::StateTransform` MIR pass can create atypical
+        // conditions, where all live `Counter`s are dropped from the MIR.
+        //
+        // At least one Counter per function is required by LLVM (and necessary,
+        // to add the `function_hash` to the counter's call to the LLVM
+        // intrinsic `instrprof.increment()`).
+        return;
+    }
+
+    // Retain coverage info for dead blocks, so coverage reports will still
+    // report `0` executions for the uncovered code regions.
+    let mut dropped_coverage = Vec::new();
+    for dead_block in basic_blocks.raw[first_dead_block..].iter() {
+        for statement in dead_block.statements.iter() {
+            if let StatementKind::Coverage(coverage) = &statement.kind {
+                if let Some(code_region) = &coverage.code_region {
+                    dropped_coverage.push((statement.source_info, code_region.clone()));
+                }
+            }
+        }
+    }
+
+    let start_block = &mut basic_blocks[START_BLOCK];
+    for (source_info, code_region) in dropped_coverage {
+        start_block.statements.push(Statement {
+            source_info,
+            kind: StatementKind::Coverage(box Coverage {
+                kind: CoverageKind::Unreachable,
+                code_region: Some(code_region),
+            }),
+        })
+    }
+}
+
 pub struct SimplifyLocals;
 
 impl<'tcx> MirPass<'tcx> for SimplifyLocals {
diff --git a/compiler/rustc_mir/src/transform/simplify_try.rs b/compiler/rustc_mir/src/transform/simplify_try.rs
index 89fddc95c98..dd2ec39c066 100644
--- a/compiler/rustc_mir/src/transform/simplify_try.rs
+++ b/compiler/rustc_mir/src/transform/simplify_try.rs
@@ -558,7 +558,7 @@ impl<'tcx> MirPass<'tcx> for SimplifyBranchSame {
 
         if did_remove_blocks {
             // We have dead blocks now, so remove those.
-            simplify::remove_dead_blocks(body);
+            simplify::remove_dead_blocks(tcx, body);
         }
     }
 }
diff --git a/compiler/rustc_mir/src/transform/unreachable_prop.rs b/compiler/rustc_mir/src/transform/unreachable_prop.rs
index 658c6b6e9db..e7fb6b4f6b4 100644
--- a/compiler/rustc_mir/src/transform/unreachable_prop.rs
+++ b/compiler/rustc_mir/src/transform/unreachable_prop.rs
@@ -60,7 +60,7 @@ impl MirPass<'_> for UnreachablePropagation {
         }
 
         if replaced {
-            simplify::remove_dead_blocks(body);
+            simplify::remove_dead_blocks(tcx, body);
         }
     }
 }