Skip to main content Crate crucible_test_context Copy item path Source pub use schema::register_account_schemas ;pub use schema::AccountSchema ;pub use coverage::build_cached_analysis ;pub use coverage::generate_bytecode_lcov ;pub use coverage::generate_coverage_html ;pub use coverage::generate_coverage_html_cached ;pub use coverage::generate_source_lcov ;pub use coverage::build_dwarf_source_map ;pub use coverage::DwarfSourceMap ;pub use coverage::SourceLocation ;pub use coverage::CachedFunctionInfo ;pub use coverage::CachedProgramAnalysis ;pub use coverage::CoverageStats ;pub use coverage::CoverageWriteStats ;pub use coverage::FunctionInfo ;pub use coverage::ReachableAnalysis ;pub use litesvm ;pub use serde_json ;coverage Coverage analysis and visualization for BPF programs. fuzz_types schema Account schema registry for semantic field-level diffs. snapshot Dirty account tracking and snapshot/restore for fast fuzzing iterations. fuzz_assert Assert a condition is true fuzz_assert_approx_eq Assert two values are approximately equal within a delta (absolute difference) fuzz_assert_eq Assert two values are equal fuzz_assert_ge Assert a >= b fuzz_assert_gt Assert a > b fuzz_assert_le Assert a <= b fuzz_assert_lt Assert a < b fuzz_assert_ne Assert two values are not equal ActionRecord Record of a single action execution for crash metadata CrashMetadata Complete crash metadata for .meta.json files EmptyInvocationCallback Empty callback that does nothing - used during setup to avoid DefaultRegisterTracingCallback
trying to find .so files on disk for built-in programs. FieldDelta A single field-level semantic diff (used by schema registry for rich crash output). FxBuildHasher An implementation of BuildHasherFxHasher GenericAccountBuilder InstructionBuilder MintAccountBuilder MockPythOracleBuilder Builder for creating mock Pyth price feed accounts PriceFeedMessage Pyth price feed message containing price data PriceUpdateV2 Pyth PriceUpdateV2 account structure (Pyth Solana Receiver format) ProgramBuilder ProgramData Stores program data for reloading into debuggable SVMs TestContext TokenAccountBuilder TransactionBuilder TxError Error type for TxOutcome::into_result() TxOutcome Parsed transaction outcome from litesvm execution VerificationLevel Pyth verification level for price updates DEFAULT_PYTH_RECEIVER_ID Default Pyth Solana Receiver program ID
rec5EKMGg6MxZYaMdyBps2bnnCNHi6KCYuQedA7GsAuW PYTH_DISCRIMINATOR PriceUpdateV2 discriminator: sha256(“account:PriceUpdateV2”)[0..8] TOTAL_ACTIONS_DISPATCHED Global counter of total actions dispatched across all iterations.
Used with TOTAL_EXECUTIONS to compute average actions per execution. TOTAL_ACTIONS_SUCCEEDED Global counter of total actions that succeeded across all iterations.
Used by monitor to display success rate (ok/total). TOTAL_ACTION_VARIANTS Total number of action variants available (set once at startup).
Used by monitor to display “discovered: N/M actions”. AccountBuilderBase IntoActionSuccess Trait to normalize action return values to success/failure.
Allows actions to return either () (always success) or Result<(), E> (success/failure). InvocationInspectCallback backfill_action_params Backfill the params for a specific action record in the history. build_crash_metadata Build crash metadata from current state clear_action_history Clear the action history (called at start of each iteration) clear_iteration_state Clear all per-iteration state: action history + violation tracking.
Convenience wrapper — call at the start of each fuzzing iteration. clear_violation_tracking Clear all violation tracking state (called at start of each iteration) compute_svm_debug_hash Hash SVM state for tracked accounts. Returns (hash, clock_slot).
Accounts sorted by (lamports, data_len, data_hash) for cross-run stability. format_action_sequence Print the action sequence to stderr (for debugging crashes)
Format the current action sequence from TLS history into a string.
Must be called on the same thread that executed the actions (reads TLS). format_all_actions_oneline Format ALL actions in the current history as one-line descriptions, newline-separated.
Used by stateful mode to store the full chain description in a single action_desc field. format_json_value Format a JSON value for display (compact format) format_last_action_oneline Format the most recent action from TLS history as a one-line summary.
e.g. “action_deposit(user=0, amount=500) -> OK”
Returns empty string if no action in history. get_action_history Get a copy of the action history get_current_instruction Get the current Anchor instruction name get_current_iteration Get the current iteration number get_current_test_name Get the current test name get_first_action_success Check if the first action in history succeeded.
Avoids cloning the entire history Vec — O(1) instead of O(n). get_iteration_dispatch_count Get the per-iteration dispatch count (call after fn_name).
Returns at least 1 to avoid zero-counting iterations. get_registered_discriminators Get all registered discriminators (for debugging) get_send_batch_timers Get send_batch sub-phase timers: (pre_ns, svm_ns, post_ns). get_send_tx_breakdown Get send_transaction sub-breakdown: (blockhash_ns, sign_ns, exec_ns). get_violation_action_index Get the action index that triggered the violation (if any) has_variant_succeeded Check if a given action variant index has ever succeeded. has_violation Check if a violation has been recorded (without consuming it) increment_action_count Increment the global action counter by one.
Also increments the thread-local per-iteration counter for accurate
multicore exec/sec reporting (global atomic has cross-thread noise). increment_action_success_count Increment the global succeeded action counter by one. is_corpus_loading Check if corpus loading is in progress. is_debug_replay Check FUZZ_DEBUG_REPLAY env var (cached per-thread). is_stateful_chain_mode Check if stateful chain mode is active. lookup_instruction_by_discriminator Look up instruction name from discriminator bytes at the start of instruction data.
Automatically uses the correct discriminator length (4 or 8 bytes).
Returns None if discriminator is not registered or if the data is too short.
Lock-free after initialization. mark_variant_succeeded Mark an action variant index as having succeeded at least once. parse_action_desc Parse an action description string like “withdraw(authority=null, lamports=5) -> OK”
back into an ActionRecord. Used by stateful mode to reconstruct full crash metadata
from pool descriptions. parse_error_code Parse litesvm TransactionError to extract error code.
Extracts Custom(N) error codes from InstructionError variants. parse_instruction_index Parse litesvm TransactionError to extract instruction index. print_action_sequence push_action_record Push an action record to the history and update cumulative stats. push_action_record_lite Lite action record: only records name and success, defers JSON params.
Avoids serde_json::Value allocation on every action of every iteration.
The params field is set to null and should be backfilled via
backfill_action_params only when needed (crash/violation). record_violation Record an invariant violation (used by fuzz_assert! macros) register_instruction_discriminators Register instruction discriminators for per-instruction coverage tracking.
Call this once at harness initialization with discriminators from the program’s IDL.
Supports variable-length discriminators (4 bytes for bincode, 8 bytes for Anchor).
Subsequent calls are ignored (OnceLock can only be set once). reset_iteration_dispatch_count Reset the per-iteration dispatch counter (call before fn_name). reset_send_batch_timers Reset send_batch sub-phase accumulators (call before fn_name). set_corpus_loading Set corpus loading flag (suppresses monitor output during loading). set_current_iteration Set the current iteration number set_current_test_name Set the current test name (called at test start) set_last_error_code Set the last error code from a transaction result (called by tx_result_to_outcome) set_stateful_chain_mode Set stateful chain mode flag. When true, invariant test loops break on
first action failure (failed actions produce dead-end states in stateful mode). set_total_actions Set the total number of actions in the current sequence (for early exit tracking) set_violation_action_index Record which action index triggered a violation (only records first violation) succeeded_variant_count Get the number of action variants that have ever succeeded. take_violation Take the current violation (clearing it). Returns Some if violated. tx_result_to_outcome Convert litesvm transaction result to TxOutcome write_crash_metadata Write crash metadata to a .meta.json file and save input bytes for replay write_crash_metadata_for_id Write crash metadata for a known crash ID (used by tmin to update metadata in place) write_crash_metadata_with_actions Write crash metadata with an explicit full action chain (for stateful mode).
If full_actions is Some, it overrides the TLS action history. FastHashMap Type alias for fast HashMap (uses FxHash) FastHashSet Type alias for fast HashSet (uses FxHash) FxHashMap Type alias for a hash map that uses the Fx hashing algorithm. FxHashSet Type alias for a hash set that uses the Fx hashing algorithm.