Virtual machine for field arithmetic operations. More...

#include <field.fuzzer.hpp>

Classes
struct	ParsedInstruction
	Structure to hold parsed instruction data. More...

Public Member Functions
	FieldVM (bool with_debug=false, size_t max_steps=SIZE_MAX)
	Constructor for FieldVM.

size_t	execute_instruction (const unsigned char *data_ptr)
	Execute a single VM instruction.

std::pair< std::vector< ParsedInstruction >, size_t >	parse_instructions (const unsigned char *Data, size_t Size, size_t max_steps)
	Parse instructions from data buffer into a vector.

bool	execute_parsed_instruction (const ParsedInstruction &parsed)
	Execute a parsed instruction.

size_t	run (const unsigned char *Data, size_t Size, bool reset_steps=true)
	Run the VM on input data.

bool	check_internal_state () const
	Check internal state consistency between field and uint256_t representations.

std::vector< numeric::uint256_t >	export_uint_state () const
	Export the final uint state as a vector of uint256_t values.

size_t	get_step_count () const
	Get the number of steps executed.

bool	was_stopped_by_max_steps () const
	Check if the VM was stopped due to reaching max steps.

void	set_max_steps (size_t new_max_steps)
	Set a new step limit for the VM.

void	reset_step_count ()
	Reset the step counter to 0.

size_t	get_max_steps () const
	Get the current step limit.

bool	has_remaining_steps () const
	Check if there are remaining steps available.

bool	verify_initial_state (const std::vector< numeric::uint256_t > &state) const
	Verify that the initial state is correctly loaded.

Static Public Member Functions
static numeric::uint256_t	reduce_to_modulus (const numeric::uint256_t &value)
	Reduce a uint256_t value to the field's modulus.

Public Attributes
std::array< Field, INTERNAL_STATE_SIZE >	field_internal_state
	Internal state array of field elements.

std::array< numeric::uint256_t, INTERNAL_STATE_SIZE >	uint_internal_state
	Internal state array of uint256_t values.

bool	with_debug
	Flag to enable debug output.

VMSettings	settings
	VM settings controlling which operations are enabled.

size_t	max_steps
	Maximum number of steps the VM can execute.

size_t	step_count {}
	Number of steps executed so far.

Static Public Attributes
static constexpr bool	LARGE_MODULUS = (Field::modulus.data[3] >= 0x4000000000000000ULL)
	Flag indicating if the field has a large modulus requiring uint512_t for arithmetic.

static constexpr bool	SUPPORTS_SQRT
	Flag indicating if the field supports square root operations.

Detailed Description

template<typename Field>
struct bb::FieldVM< Field >

Virtual machine for field arithmetic operations.

Template Parameters

Field The field type to operate on

This template class implements a virtual machine that can execute field arithmetic operations. It maintains both field elements and uint256_t representations for verification purposes. The VM supports various field operations and includes comprehensive error checking and debugging capabilities.

Definition at line 146 of file field.fuzzer.hpp.

Constructor & Destructor Documentation

◆ FieldVM()

template<typename Field >

bb::FieldVM< Field >::FieldVM	(	bool	with_debug = `false`,
		size_t	max_steps = `SIZE_MAX`
	)

inline

Constructor for FieldVM.

Parameters

with_debug	Whether to enable debug output
max_steps	Maximum number of steps to execute

Initializes the VM with default settings and zero-initialized state

Definition at line 211 of file field.fuzzer.hpp.

Member Function Documentation

◆ check_internal_state()

template<typename Field >

bool bb::FieldVM< Field >::check_internal_state ( ) const

inline

Check internal state consistency between field and uint256_t representations.

Returns: bool True if all state elements are consistent, false otherwise

This method verifies that the field_internal_state and uint_internal_state arrays are consistent with each other. Any discrepancy indicates a potential bug in the field arithmetic implementation.

Definition at line 1240 of file field.fuzzer.hpp.

◆ execute_instruction()

template<typename Field >

size_t bb::FieldVM< Field >::execute_instruction ( const unsigned char * data_ptr )

inline

Execute a single VM instruction.

Parameters

data_ptr Pointer to the instruction data

Returns: size_t Number of bytes consumed by the instruction

This method parses and executes a single VM instruction. It handles all supported field operations and maintains consistency between field and uint256_t representations. The method returns the number of bytes consumed.

Helper function to get an index from data with wraparound

Parameters

data_ptr_index	Pointer to the data containing the index
offset	Offset into the data

Returns: size_t The index value wrapped around INTERNAL_STATE_SIZE

< Total number of instructions

Helper function to get a uint256_t value from data

Parameters

data_ptr_value	Pointer to the data containing the value
offset	Offset into the data

Returns: numeric::uint256_t The uint256_t value read from data

Helper function to get a uint64_t value from data

Parameters

data_ptr_value	Pointer to the data containing the value
offset	Offset into the data

Returns: uint64_t The uint64_t value read from data

Definition at line 258 of file field.fuzzer.hpp.

◆ execute_parsed_instruction()

template<typename Field >

bool bb::FieldVM< Field >::execute_parsed_instruction ( const ParsedInstruction & parsed )

inline

Execute a parsed instruction.

Parameters

parsed The parsed instruction to execute

Returns: bool True if execution should continue, false if should stop

This method executes a previously parsed instruction. It provides debug output if enabled and returns whether execution should continue.

Definition at line 1159 of file field.fuzzer.hpp.

◆ export_uint_state()

template<typename Field >

std::vector< numeric::uint256_t > bb::FieldVM< Field >::export_uint_state ( ) const

inline

Export the final uint state as a vector of uint256_t values.

Returns: std::vector<numeric::uint256_t> The final uint state

This method creates a copy of the uint_internal_state array for external use, typically for state transfer between VM phases.

Definition at line 1262 of file field.fuzzer.hpp.

◆ get_max_steps()

template<typename Field >

size_t bb::FieldVM< Field >::get_max_steps ( ) const

inline

Get the current step limit.

Returns: size_t The current maximum number of steps

Definition at line 1303 of file field.fuzzer.hpp.

◆ get_step_count()

template<typename Field >

size_t bb::FieldVM< Field >::get_step_count ( ) const

inline

Get the number of steps executed.

Returns: size_t The number of steps executed

Definition at line 1277 of file field.fuzzer.hpp.

◆ has_remaining_steps()

template<typename Field >

bool bb::FieldVM< Field >::has_remaining_steps ( ) const

inline

Check if there are remaining steps available.

Returns: bool True if more steps can be executed

Definition at line 1310 of file field.fuzzer.hpp.

◆ parse_instructions()

template<typename Field >

std::pair< std::vector< ParsedInstruction >, size_t > bb::FieldVM< Field >::parse_instructions	(	const unsigned char *	Data,
		size_t	Size,
		size_t	max_steps
	)

inline

Parse instructions from data buffer into a vector.

Parameters

Data	Pointer to the data buffer
Size	Size of the data buffer
max_steps	Maximum number of instructions to parse

Returns: std::pair<std::vector<ParsedInstruction>, size_t> Vector of parsed instructions and bytes consumed

This method parses the input data into a sequence of instructions. It handles settings parsing and instruction size calculation. The method returns both the parsed instructions and the number of bytes consumed during parsing.

< Total number of instructions

Definition at line 1016 of file field.fuzzer.hpp.

◆ reduce_to_modulus()

template<typename Field >

static numeric::uint256_t bb::FieldVM< Field >::reduce_to_modulus ( const numeric::uint256_t & value )

inlinestatic

Reduce a uint256_t value to the field's modulus.

Parameters

value The value to reduce

Returns: numeric::uint256_t The reduced value

This method handles modulus reduction for both large and small modulus fields. For large modulus fields, it uses uint512_t to prevent overflow.

Definition at line 1321 of file field.fuzzer.hpp.

◆ reset_step_count()

template<typename Field >

void bb::FieldVM< Field >::reset_step_count ( )

inline

Reset the step counter to 0.

Definition at line 1296 of file field.fuzzer.hpp.

◆ run()

template<typename Field >

size_t bb::FieldVM< Field >::run	(	const unsigned char *	Data,
		size_t	Size,
		bool	reset_steps = `true`
	)

inline

Run the VM on input data.

Parameters

Data	The data to run the VM on
Size	The size of the data
reset_steps	Whether to reset the step counter (default: true)

Returns: size_t The number of bytes consumed, or 0 if not enough data for settings

This method is the main entry point for VM execution. It parses all instructions from the input data and executes them sequentially. The method handles step counting, debug output, and returns the number of bytes consumed.

Definition at line 1197 of file field.fuzzer.hpp.

◆ set_max_steps()

template<typename Field >

void bb::FieldVM< Field >::set_max_steps ( size_t new_max_steps )

inline

Set a new step limit for the VM.

Parameters

new_max_steps The new maximum number of steps

Definition at line 1291 of file field.fuzzer.hpp.

◆ verify_initial_state()

template<typename Field >

bool bb::FieldVM< Field >::verify_initial_state ( const std::vector< numeric::uint256_t > & state ) const

inline

Verify that the initial state is correctly loaded.

Parameters

state The state vector to verify against

Returns: bool True if the state is correctly loaded

This method verifies that the imported state matches the current internal state after proper modulus reduction and field conversion.

Definition at line 1339 of file field.fuzzer.hpp.

◆ was_stopped_by_max_steps()

template<typename Field >

bool bb::FieldVM< Field >::was_stopped_by_max_steps ( ) const

inline

Check if the VM was stopped due to reaching max steps.

Returns: bool True if the VM was stopped due to max steps

Definition at line 1284 of file field.fuzzer.hpp.

Member Data Documentation

◆ field_internal_state

template<typename Field >

std::array<Field, INTERNAL_STATE_SIZE> bb::FieldVM< Field >::field_internal_state

Internal state array of field elements.

Used for actual field arithmetic operations

Definition at line 174 of file field.fuzzer.hpp.

◆ LARGE_MODULUS

template<typename Field >

constexpr bool bb::FieldVM< Field >::LARGE_MODULUS = (Field::modulus.data[3] >= 0x4000000000000000ULL)

staticconstexpr

Flag indicating if the field has a large modulus requiring uint512_t for arithmetic.

Fields with moduli >= 2^254 require uint512_t for safe addition/subtraction

Definition at line 152 of file field.fuzzer.hpp.

◆ max_steps

template<typename Field >

size_t bb::FieldVM< Field >::max_steps

Maximum number of steps the VM can execute.

Definition at line 196 of file field.fuzzer.hpp.

◆ settings

template<typename Field >

VMSettings bb::FieldVM< Field >::settings

VM settings controlling which operations are enabled.

Definition at line 191 of file field.fuzzer.hpp.

◆ step_count

template<typename Field >

size_t bb::FieldVM< Field >::step_count {}

Number of steps executed so far.

Definition at line 201 of file field.fuzzer.hpp.

◆ SUPPORTS_SQRT

template<typename Field >

constexpr bool bb::FieldVM< Field >::SUPPORTS_SQRT

staticconstexpr

Initial value:

= []() {
        if constexpr (requires { Field::primitive_root_log_size(); }) {
            
            return Field::primitive_root_log_size() >= 6;
        } else {
            
            return true;
        }
    }()

Flag indicating if the field supports square root operations.

Fields with very small 2-adicity (like secp256r1) have issues with Tonelli-Shanks

Definition at line 159 of file field.fuzzer.hpp.

◆ uint_internal_state

template<typename Field >

std::array<numeric::uint256_t, INTERNAL_STATE_SIZE> bb::FieldVM< Field >::uint_internal_state

Internal state array of uint256_t values.

Used as oracles for checking the correctness of field operations

Definition at line 181 of file field.fuzzer.hpp.

◆ with_debug

template<typename Field >

bool bb::FieldVM< Field >::with_debug

Flag to enable debug output.

Definition at line 186 of file field.fuzzer.hpp.

The documentation for this struct was generated from the following file:

src/barretenberg/ecc/fields/field.fuzzer.hpp

Classes

Public Member Functions

Static Public Member Functions

Public Attributes

Static Public Attributes

Detailed Description

Constructor & Destructor Documentation

◆ FieldVM()

Member Function Documentation

◆ check_internal_state()

◆ execute_instruction()

◆ execute_parsed_instruction()

◆ export_uint_state()

◆ get_max_steps()

◆ get_step_count()

◆ has_remaining_steps()

◆ parse_instructions()

◆ reduce_to_modulus()

◆ reset_step_count()

◆ run()

◆ set_max_steps()

◆ verify_initial_state()

◆ was_stopped_by_max_steps()

Member Data Documentation

◆ field_internal_state

◆ LARGE_MODULUS

◆ max_steps

◆ settings

◆ step_count

◆ SUPPORTS_SQRT

◆ uint_internal_state

◆ with_debug