mir: extract CSE into passes::cse and wire it in optimizer. Behavior preserved (count-only, no SSA rewrites). Build + PyVM Stage-2 smokes PASS.

src/mir/optimizer.rs

@@ -68,8 +68,11 @@ impl MirOptimizer {
             stats.dead_code_eliminated += eliminated;
         }
         
-        // Pass 2: Pure instruction CSE (Common Subexpression Elimination)
+        // Pass 2: Pure instruction CSE (modularized)
-        stats.merge(self.common_subexpression_elimination(module));
+        {
+            let eliminated = crate::mir::passes::cse::eliminate_common_subexpressions(module);
+            stats.cse_eliminated += eliminated;
+        }
         
         // Pass 3: Pure instruction reordering for better locality
         stats.merge(self.reorder_pure_instructions(module));

src/mir/passes/cse.rs

@@ -0,0 +1,56 @@
+//! Common Subexpression Elimination (CSE) for pure MIR instructions.
+//!
+//! Note: Current implementation mirrors the prior monolithic behavior and
+//! counts eliminations without rewriting uses (SSA update is TODO). This keeps
+//! behavior identical while modularizing the pass for future enhancement.
+
+use crate::mir::{MirInstruction, MirModule, MirFunction, ValueId};
+use std::collections::HashMap;
+
+/// Run CSE across the module. Returns the number of eliminated expressions.
+pub fn eliminate_common_subexpressions(module: &mut MirModule) -> usize {
+    let mut eliminated = 0usize;
+    for (_name, func) in module.functions.iter_mut() {
+        eliminated += cse_in_function(func);
+    }
+    eliminated
+}
+
+fn cse_in_function(function: &mut MirFunction) -> usize {
+    let mut expression_map: HashMap<String, ValueId> = HashMap::new();
+    let mut eliminated = 0usize;
+
+    for (_bid, block) in &mut function.blocks {
+        for inst in &mut block.instructions {
+            if inst.effects().is_pure() {
+                let key = instruction_key(inst);
+                if let Some(&existing) = expression_map.get(&key) {
+                    if let Some(dst) = inst.dst_value() {
+                        // Count as eliminated; rewriting uses is a future improvement.
+                        let _ = (existing, dst); // keep variables referenced
+                        eliminated += 1;
+                    }
+                } else if let Some(dst) = inst.dst_value() {
+                    expression_map.insert(key, dst);
+                }
+            }
+        }
+    }
+    eliminated
+}
+
+fn instruction_key(i: &MirInstruction) -> String {
+    match i {
+        MirInstruction::Const { value, .. } => format!("const_{:?}", value),
+        MirInstruction::BinOp { op, lhs, rhs, .. } =>
+            format!("binop_{:?}_{}_{}", op, lhs.as_u32(), rhs.as_u32()),
+        MirInstruction::Compare { op, lhs, rhs, .. } =>
+            format!("cmp_{:?}_{}_{}", op, lhs.as_u32(), rhs.as_u32()),
+        MirInstruction::Call { func, args, .. } => {
+            let args_str = args.iter().map(|v| v.as_u32().to_string()).collect::<Vec<_>>().join(",");
+            format!("call_{}_{}", func.as_u32(), args_str)
+        }
+        other => format!("other_{:?}", other),
+    }
+}
+

src/mir/passes/mod.rs

@@ -5,6 +5,7 @@ pub mod type_hints;
 pub mod escape;
 pub mod method_id_inject;
 pub mod dce;
+pub mod cse;
 
 /// Minimal pass trait for future expansion. Currently unused by the main
 /// optimizer pipeline but provided to guide modularization.