fix: don't deduplicate binary math of unsigned types (#6848)

Co-authored-by: Maxim Vezenov <mvezenov@gmail.com>

compiler/noirc_evaluator/src/ssa/ir/instruction.rs

@@ -548,10 +548,25 @@ impl Instruction {
     /// If true the instruction will depend on `enable_side_effects` context during acir-gen.
     pub(crate) fn requires_acir_gen_predicate(&self, dfg: &DataFlowGraph) -> bool {
         match self {
-            Instruction::Binary(binary)
-                if matches!(binary.operator, BinaryOp::Div | BinaryOp::Mod) =>
-            {
-                true
+            Instruction::Binary(binary) => {
+                match binary.operator {
+                    BinaryOp::Add
+                    | BinaryOp::Sub
+                    | BinaryOp::Mul
+                    | BinaryOp::Div
+                    | BinaryOp::Mod => {
+                        // Some binary math can overflow or underflow, but this is only the case
+                        // for unsigned types (here we assume the type of binary.lhs is the same)
+                        dfg.type_of_value(binary.rhs).is_unsigned()
+                    }
+                    BinaryOp::Eq
+                    | BinaryOp::Lt
+                    | BinaryOp::And
+                    | BinaryOp::Or
+                    | BinaryOp::Xor
+                    | BinaryOp::Shl
+                    | BinaryOp::Shr => false,
+                }
             }
 
             Instruction::ArrayGet { array, index } => {
@@ -569,7 +584,6 @@ impl Instruction {
                 _ => false,
             },
             Instruction::Cast(_, _)
-            | Instruction::Binary(_)
             | Instruction::Not(_)
             | Instruction::Truncate { .. }
             | Instruction::Constrain(_, _, _)

compiler/noirc_evaluator/src/ssa/opt/loop_invariant.rs

@@ -263,17 +263,17 @@ mod test {
     fn simple_loop_invariant_code_motion() {
         let src = "
         brillig(inline) fn main f0 {
-          b0(v0: u32, v1: u32):
-              jmp b1(u32 0)
-          b1(v2: u32):
-              v5 = lt v2, u32 4
+          b0(v0: i32, v1: i32):
+              jmp b1(i32 0)
+          b1(v2: i32):
+              v5 = lt v2, i32 4
               jmpif v5 then: b3, else: b2
           b2():
               return
           b3():
               v6 = mul v0, v1
-              constrain v6 == u32 6
-              v8 = add v2, u32 1
+              constrain v6 == i32 6
+              v8 = add v2, i32 1
               jmp b1(v8)
         }
         ";
@@ -287,17 +287,17 @@ mod test {
         // `v6 = mul v0, v1` in b3 should now be `v3 = mul v0, v1` in b0
         let expected = "
         brillig(inline) fn main f0 {
-          b0(v0: u32, v1: u32):
+          b0(v0: i32, v1: i32):
             v3 = mul v0, v1
-            jmp b1(u32 0)
-          b1(v2: u32):
-            v6 = lt v2, u32 4
+            jmp b1(i32 0)
+          b1(v2: i32):
+            v6 = lt v2, i32 4
             jmpif v6 then: b3, else: b2
           b2():
             return
           b3():
-            constrain v3 == u32 6
-            v9 = add v2, u32 1
+            constrain v3 == i32 6
+            v9 = add v2, i32 1
             jmp b1(v9)
         }
         ";
@@ -312,25 +312,25 @@ mod test {
         // is hoisted to the parent loop's pre-header block.
         let src = "
         brillig(inline) fn main f0 {
-          b0(v0: u32, v1: u32):
-            jmp b1(u32 0)
-          b1(v2: u32):
-            v6 = lt v2, u32 4
+          b0(v0: i32, v1: i32):
+            jmp b1(i32 0)
+          b1(v2: i32):
+            v6 = lt v2, i32 4
             jmpif v6 then: b3, else: b2
           b2():
             return
           b3():
-            jmp b4(u32 0)
-          b4(v3: u32):
-            v7 = lt v3, u32 4
+            jmp b4(i32 0)
+          b4(v3: i32):
+            v7 = lt v3, i32 4
             jmpif v7 then: b6, else: b5
           b5():
-            v9 = add v2, u32 1
+            v9 = add v2, i32 1
             jmp b1(v9)
           b6():
             v10 = mul v0, v1
-            constrain v10 == u32 6
-            v12 = add v3, u32 1
+            constrain v10 == i32 6
+            v12 = add v3, i32 1
             jmp b4(v12)
         }
         ";
@@ -344,25 +344,25 @@ mod test {
         // `v10 = mul v0, v1` in b6 should now be `v4 = mul v0, v1` in b0
         let expected = "
         brillig(inline) fn main f0 {
-          b0(v0: u32, v1: u32):
+          b0(v0: i32, v1: i32):
             v4 = mul v0, v1
-            jmp b1(u32 0)
-          b1(v2: u32):
-            v7 = lt v2, u32 4
+            jmp b1(i32 0)
+          b1(v2: i32):
+            v7 = lt v2, i32 4
             jmpif v7 then: b3, else: b2
           b2():
             return
           b3():
-            jmp b4(u32 0)
-          b4(v3: u32):
-            v8 = lt v3, u32 4
+            jmp b4(i32 0)
+          b4(v3: i32):
+            v8 = lt v3, i32 4
             jmpif v8 then: b6, else: b5
           b5():
-            v10 = add v2, u32 1
+            v10 = add v2, i32 1
             jmp b1(v10)
           b6():
-            constrain v4 == u32 6
-            v12 = add v3, u32 1
+            constrain v4 == i32 6
+            v12 = add v3, i32 1
             jmp b4(v12)
         }
         ";
@@ -386,19 +386,19 @@ mod test {
         // hoist `v7 = mul v6, v0`.
         let src = "
         brillig(inline) fn main f0 {
-          b0(v0: u32, v1: u32):
-            jmp b1(u32 0)
-          b1(v2: u32):
-            v5 = lt v2, u32 4
+          b0(v0: i32, v1: i32):
+            jmp b1(i32 0)
+          b1(v2: i32):
+            v5 = lt v2, i32 4
             jmpif v5 then: b3, else: b2
           b2():
             return
           b3():
             v6 = mul v0, v1
             v7 = mul v6, v0
-            v8 = eq v7, u32 12
-            constrain v7 == u32 12
-            v9 = add v2, u32 1
+            v8 = eq v7, i32 12
+            constrain v7 == i32 12
+            v9 = add v2, i32 1
             jmp b1(v9)
         }
         ";
@@ -411,19 +411,19 @@ mod test {
 
         let expected = "
         brillig(inline) fn main f0 {
-          b0(v0: u32, v1: u32):
+          b0(v0: i32, v1: i32):
             v3 = mul v0, v1
             v4 = mul v3, v0
-            v6 = eq v4, u32 12
-            jmp b1(u32 0)
-          b1(v2: u32):
-            v9 = lt v2, u32 4
+            v6 = eq v4, i32 12
+            jmp b1(i32 0)
+          b1(v2: i32):
+            v9 = lt v2, i32 4
             jmpif v9 then: b3, else: b2
           b2():
             return
           b3():
-            constrain v4 == u32 12
-            v11 = add v2, u32 1
+            constrain v4 == i32 12
+            v11 = add v2, i32 1
             jmp b1(v11)
         }
         ";

test_programs/execution_success/regression_6834/Nargo.toml

@@ -0,0 +1,7 @@
+[package]
+name = "regression_6834"
+version = "0.1.0"
+type = "bin"
+authors = [""]
+
+[dependencies]

test_programs/execution_success/regression_6834/Prover.toml

@@ -0,0 +1,2 @@
+year = 1
+min_age = 1

test_programs/execution_success/regression_6834/src/main.nr

@@ -0,0 +1,9 @@
+fn main(year: u32, min_age: u8) -> pub u32 {
+    if min_age == 0 {
+        (year - 2) / 4
+    } else if year > 1 {
+        (year - 2) / 4
+    } else {
+        0
+    }
+}