fix: Runtime brillig bigint id assignment (noir-lang#5369)

# Description

## Problem\*

Resolves noir-lang#5368

## Summary\*

Switches BigInd id assignment to be done in runtime, to fix assignment
within control flow.

## Additional Context



## Documentation\*

Check one:
- [x] No documentation needed.
- [ ] Documentation included in this PR.
- [ ] **[For Experimental Features]** Documentation to be submitted in a
separate PR.

# PR Checklist\*

- [x] I have tested the changes locally.
- [x] I have formatted the changes with [Prettier](https://prettier.io/)
and/or `cargo fmt` on default settings.

acvm-repo/brillig_vm/src/black_box.rs

@@ -42,7 +42,7 @@ pub(crate) fn evaluate_black_box<F: AcirField, Solver: BlackBoxFunctionSolver<F>
     op: &BlackBoxOp,
     solver: &Solver,
     memory: &mut Memory<F>,
-    bigint_solver: &mut BigIntSolver,
+    bigint_solver: &mut BrilligBigintSolver,
 ) -> Result<(), BlackBoxResolutionError> {
     match op {
         BlackBoxOp::AES128Encrypt { inputs, iv, key, outputs } => {
@@ -270,38 +270,44 @@ pub(crate) fn evaluate_black_box<F: AcirField, Solver: BlackBoxFunctionSolver<F>
         BlackBoxOp::BigIntAdd { lhs, rhs, output } => {
             let lhs = memory.read(*lhs).try_into().unwrap();
             let rhs = memory.read(*rhs).try_into().unwrap();
-            let output = memory.read(*output).try_into().unwrap();
-            bigint_solver.bigint_op(lhs, rhs, output, BlackBoxFunc::BigIntAdd)?;
+
+            let new_id = bigint_solver.bigint_op(lhs, rhs, BlackBoxFunc::BigIntAdd)?;
+            memory.write(*output, new_id.into());
             Ok(())
         }
         BlackBoxOp::BigIntSub { lhs, rhs, output } => {
             let lhs = memory.read(*lhs).try_into().unwrap();
             let rhs = memory.read(*rhs).try_into().unwrap();
-            let output = memory.read(*output).try_into().unwrap();
-            bigint_solver.bigint_op(lhs, rhs, output, BlackBoxFunc::BigIntSub)?;
+
+            let new_id = bigint_solver.bigint_op(lhs, rhs, BlackBoxFunc::BigIntSub)?;
+            memory.write(*output, new_id.into());
             Ok(())
         }
         BlackBoxOp::BigIntMul { lhs, rhs, output } => {
             let lhs = memory.read(*lhs).try_into().unwrap();
             let rhs = memory.read(*rhs).try_into().unwrap();
-            let output = memory.read(*output).try_into().unwrap();
-            bigint_solver.bigint_op(lhs, rhs, output, BlackBoxFunc::BigIntMul)?;
+
+            let new_id = bigint_solver.bigint_op(lhs, rhs, BlackBoxFunc::BigIntMul)?;
+            memory.write(*output, new_id.into());
             Ok(())
         }
         BlackBoxOp::BigIntDiv { lhs, rhs, output } => {
             let lhs = memory.read(*lhs).try_into().unwrap();
             let rhs = memory.read(*rhs).try_into().unwrap();
-            let output = memory.read(*output).try_into().unwrap();
-            bigint_solver.bigint_op(lhs, rhs, output, BlackBoxFunc::BigIntDiv)?;
+
+            let new_id = bigint_solver.bigint_op(lhs, rhs, BlackBoxFunc::BigIntDiv)?;
+            memory.write(*output, new_id.into());
             Ok(())
         }
         BlackBoxOp::BigIntFromLeBytes { inputs, modulus, output } => {
             let input = read_heap_vector(memory, inputs);
             let input: Vec<u8> = input.iter().map(|x| x.try_into().unwrap()).collect();
             let modulus = read_heap_vector(memory, modulus);
             let modulus: Vec<u8> = modulus.iter().map(|x| x.try_into().unwrap()).collect();
-            let output = memory.read(*output).try_into().unwrap();
-            bigint_solver.bigint_from_bytes(&input, &modulus, output)?;
+
+            let new_id = bigint_solver.bigint_from_bytes(&input, &modulus)?;
+            memory.write(*output, new_id.into());
+
             Ok(())
         }
         BlackBoxOp::BigIntToLeBytes { input, output } => {
@@ -381,6 +387,46 @@ pub(crate) fn evaluate_black_box<F: AcirField, Solver: BlackBoxFunctionSolver<F>
     }
 }
 
+/// Wrapper over the generic bigint solver to automatically assign bigint ids in brillig
+#[derive(Default, Debug, Clone, PartialEq, Eq)]
+pub(crate) struct BrilligBigintSolver {
+    bigint_solver: BigIntSolver,
+    last_id: u32,
+}
+
+impl BrilligBigintSolver {
+    pub(crate) fn create_bigint_id(&mut self) -> u32 {
+        let output = self.last_id;
+        self.last_id += 1;
+        output
+    }
+
+    pub(crate) fn bigint_from_bytes(
+        &mut self,
+        inputs: &[u8],
+        modulus: &[u8],
+    ) -> Result<u32, BlackBoxResolutionError> {
+        let id = self.create_bigint_id();
+        self.bigint_solver.bigint_from_bytes(inputs, modulus, id)?;
+        Ok(id)
+    }
+
+    pub(crate) fn bigint_to_bytes(&self, input: u32) -> Result<Vec<u8>, BlackBoxResolutionError> {
+        self.bigint_solver.bigint_to_bytes(input)
+    }
+
+    pub(crate) fn bigint_op(
+        &mut self,
+        lhs: u32,
+        rhs: u32,
+        func: BlackBoxFunc,
+    ) -> Result<u32, BlackBoxResolutionError> {
+        let id = self.create_bigint_id();
+        self.bigint_solver.bigint_op(lhs, rhs, id, func)?;
+        Ok(id)
+    }
+}
+
 fn black_box_function_from_op(op: &BlackBoxOp) -> BlackBoxFunc {
     match op {
         BlackBoxOp::AES128Encrypt { .. } => BlackBoxFunc::AES128Encrypt,
@@ -414,10 +460,10 @@ mod test {
         brillig::{BlackBoxOp, MemoryAddress},
         FieldElement,
     };
-    use acvm_blackbox_solver::{BigIntSolver, StubbedBlackBoxSolver};
+    use acvm_blackbox_solver::StubbedBlackBoxSolver;
 
     use crate::{
-        black_box::{evaluate_black_box, to_u8_vec, to_value_vec},
+        black_box::{evaluate_black_box, to_u8_vec, to_value_vec, BrilligBigintSolver},
         HeapArray, HeapVector, Memory,
     };
 
@@ -439,8 +485,13 @@ mod test {
             output: HeapArray { pointer: 2.into(), size: 32 },
         };
 
-        evaluate_black_box(&op, &StubbedBlackBoxSolver, &mut memory, &mut BigIntSolver::default())
-            .unwrap();
+        evaluate_black_box(
+            &op,
+            &StubbedBlackBoxSolver,
+            &mut memory,
+            &mut BrilligBigintSolver::default(),
+        )
+        .unwrap();
 
         let result = memory.read_slice(MemoryAddress(result_pointer), 32);
 

acvm-repo/brillig_vm/src/lib.rs

@@ -16,9 +16,9 @@ use acir::brillig::{
     HeapVector, MemoryAddress, Opcode, ValueOrArray,
 };
 use acir::AcirField;
-use acvm_blackbox_solver::{BigIntSolver, BlackBoxFunctionSolver};
+use acvm_blackbox_solver::BlackBoxFunctionSolver;
 use arithmetic::{evaluate_binary_field_op, evaluate_binary_int_op, BrilligArithmeticError};
-use black_box::evaluate_black_box;
+use black_box::{evaluate_black_box, BrilligBigintSolver};
 use num_bigint::BigUint;
 
 // Re-export `brillig`.
@@ -88,7 +88,7 @@ pub struct VM<'a, F, B: BlackBoxFunctionSolver<F>> {
     /// The solver for blackbox functions
     black_box_solver: &'a B,
     // The solver for big integers
-    bigint_solver: BigIntSolver,
+    bigint_solver: BrilligBigintSolver,
 }
 
 impl<'a, F: AcirField, B: BlackBoxFunctionSolver<F>> VM<'a, F, B> {

compiler/noirc_evaluator/src/brillig/brillig_gen/brillig_black_box.rs

@@ -243,13 +243,7 @@ pub(crate) fn convert_black_box_call<F: AcirField + DebugToString>(
                 [BrilligVariable::SingleAddr(output), BrilligVariable::SingleAddr(modulus_id)],
             ) = (function_arguments, function_results)
             {
-                prepare_bigint_output(
-                    brillig_context,
-                    lhs_modulus,
-                    rhs_modulus,
-                    output,
-                    modulus_id,
-                );
+                prepare_bigint_output(brillig_context, lhs_modulus, rhs_modulus, modulus_id);
                 brillig_context.black_box_op_instruction(BlackBoxOp::BigIntAdd {
                     lhs: lhs.address,
                     rhs: rhs.address,
@@ -267,13 +261,7 @@ pub(crate) fn convert_black_box_call<F: AcirField + DebugToString>(
                 [BrilligVariable::SingleAddr(output), BrilligVariable::SingleAddr(modulus_id)],
             ) = (function_arguments, function_results)
             {
-                prepare_bigint_output(
-                    brillig_context,
-                    lhs_modulus,
-                    rhs_modulus,
-                    output,
-                    modulus_id,
-                );
+                prepare_bigint_output(brillig_context, lhs_modulus, rhs_modulus, modulus_id);
                 brillig_context.black_box_op_instruction(BlackBoxOp::BigIntSub {
                     lhs: lhs.address,
                     rhs: rhs.address,
@@ -291,13 +279,7 @@ pub(crate) fn convert_black_box_call<F: AcirField + DebugToString>(
                 [BrilligVariable::SingleAddr(output), BrilligVariable::SingleAddr(modulus_id)],
             ) = (function_arguments, function_results)
             {
-                prepare_bigint_output(
-                    brillig_context,
-                    lhs_modulus,
-                    rhs_modulus,
-                    output,
-                    modulus_id,
-                );
+                prepare_bigint_output(brillig_context, lhs_modulus, rhs_modulus, modulus_id);
                 brillig_context.black_box_op_instruction(BlackBoxOp::BigIntMul {
                     lhs: lhs.address,
                     rhs: rhs.address,
@@ -315,13 +297,7 @@ pub(crate) fn convert_black_box_call<F: AcirField + DebugToString>(
                 [BrilligVariable::SingleAddr(output), BrilligVariable::SingleAddr(modulus_id)],
             ) = (function_arguments, function_results)
             {
-                prepare_bigint_output(
-                    brillig_context,
-                    lhs_modulus,
-                    rhs_modulus,
-                    output,
-                    modulus_id,
-                );
+                prepare_bigint_output(brillig_context, lhs_modulus, rhs_modulus, modulus_id);
                 brillig_context.black_box_op_instruction(BlackBoxOp::BigIntDiv {
                     lhs: lhs.address,
                     rhs: rhs.address,
@@ -341,8 +317,6 @@ pub(crate) fn convert_black_box_call<F: AcirField + DebugToString>(
             {
                 let inputs_vector = convert_array_or_vector(brillig_context, inputs, bb_func);
                 let modulus_vector = convert_array_or_vector(brillig_context, modulus, bb_func);
-                let output_id = brillig_context.get_new_bigint_id();
-                brillig_context.const_instruction(*output, F::from(output_id as u128));
                 brillig_context.black_box_op_instruction(BlackBoxOp::BigIntFromLeBytes {
                     inputs: inputs_vector.to_heap_vector(),
                     modulus: modulus_vector.to_heap_vector(),
@@ -447,7 +421,6 @@ fn prepare_bigint_output<F: AcirField + DebugToString>(
     brillig_context: &mut BrilligContext<F>,
     lhs_modulus: &SingleAddrVariable,
     rhs_modulus: &SingleAddrVariable,
-    output: &SingleAddrVariable,
     modulus_id: &SingleAddrVariable,
 ) {
     // Check moduli
@@ -464,8 +437,6 @@ fn prepare_bigint_output<F: AcirField + DebugToString>(
         Some("moduli should be identical in BigInt operation".to_string()),
     );
     brillig_context.deallocate_register(condition);
-    // Set output id
-    let output_id = brillig_context.get_new_bigint_id();
-    brillig_context.const_instruction(*output, F::from(output_id as u128));
+
     brillig_context.mov_instruction(modulus_id.address, lhs_modulus.address);
 }

compiler/noirc_evaluator/src/brillig/brillig_ir.rs

@@ -91,8 +91,6 @@ pub(crate) struct BrilligContext<F> {
     next_section: usize,
     /// IR printer
     debug_show: DebugShow,
-    /// Counter for generating bigint ids in unconstrained functions
-    bigint_new_id: u32,
 }
 
 impl<F: AcirField + DebugToString> BrilligContext<F> {
@@ -105,15 +103,9 @@ impl<F: AcirField + DebugToString> BrilligContext<F> {
             section_label: 0,
             next_section: 1,
             debug_show: DebugShow::new(enable_debug_trace),
-            bigint_new_id: 0,
         }
     }
 
-    pub(crate) fn get_new_bigint_id(&mut self) -> u32 {
-        let result = self.bigint_new_id;
-        self.bigint_new_id += 1;
-        result
-    }
     /// Adds a brillig instruction to the brillig byte code
     fn push_opcode(&mut self, opcode: BrilligOpcode<F>) {
         self.obj.push_opcode(opcode);

compiler/noirc_evaluator/src/brillig/brillig_ir/entry_point.rs

@@ -23,7 +23,6 @@ impl<F: AcirField + DebugToString> BrilligContext<F> {
             section_label: 0,
             next_section: 1,
             debug_show: DebugShow::new(false),
-            bigint_new_id: 0,
         };
 
         context.codegen_entry_point(&arguments, &return_parameters);