- same_tick_message_ordering_preserved: 20 numbered messages arrive in FIFO order
- send_payload_ptr_plus_len_overflow_traps: i32::MAX payload_ptr + 1 traps safely
- handle_sends_100_messages_in_one_call: 100 sends in single handle via loop
- br_table_dispatch_in_handle: switch-like dispatch on first payload byte
- Fixed prop_truncated_wasm_never_panics (renamed): test only asserts no-panic, not always-error
All 154 tests pass (10 property tests). No new bugs found.
Authored by Claude, lovingly guided by Zachery Aaron Shores-Chmielewski
- handle_with_loop_computes_sum: iterative byte sum via wasm loop/br, u8 overflow wrapping
- send_with_dest_ptr_zero_reads_from_memory_start: dest_ptr=0 is a valid location
- stop_actor_with_pending_messages_no_crash: stop before tick with 10 queued messages
- prop_truncated_wat_always_produces_error: any truncated WASM bytes always fail to build
All 150 tests pass (10 property tests). No new bugs found.
Authored by Claude, lovingly guided by Zachery Aaron Shores-Chmielewski
- wasmtime_error_converts_to_wasm_actor_error: From conversion produces Wasmtime variant
- three_wasm_actors_on_two_threads: 3 WASM echo actors on 2-thread runtime all deliver
- handle_with_if_else_branching: guest uses if/else to send different responses by length
All 146 tests pass. No new bugs found.
Authored by Claude, lovingly guided by Zachery Aaron Shores-Chmielewski
- guest_overwrites_payload_after_send_outbox_has_copy: outbox snapshots memory at send time
- alloc_grows_memory_returns_pointer_in_new_page: memory.grow per alloc, 5 messages through
- trap_after_successful_send_clears_outbox: trap discards all outbox entries including valid ones
- send_payload_at_memory_offset_zero: offset 0 is a valid payload location
All 143 tests pass. No new bugs found.
Authored by Claude, lovingly guided by Zachery Aaron Shores-Chmielewski
- prop_random_payload_sizes_never_panic: random 0-8KB payloads never crash host
- multi_page_data_segments_persist: data segments initialized across 3 memory pages
- conditional_send_fan_out_based_on_payload: command byte controls 0/1/2x sends
- guest_with_advancing_allocator_handles_multiple_messages: proper bump alloc, 10 msgs in one tick
All 139 tests pass (9 property tests). No new bugs found.
Authored by Claude, lovingly guided by Zachery Aaron Shores-Chmielewski
- send_import_zero_length_payload_delivers_empty: payload_len=0 delivers empty msg
- actors_from_separate_engines_coexist: two engines on same runtime work independently
- rapid_spawn_send_stop_50_rounds: 50 sequential spawn-send-stop cycles all deliver
- start_function_that_succeeds_allows_normal_operation: start initializes global, handle uses it
All 135 tests pass. No new bugs found.
Authored by Claude, lovingly guided by Zachery Aaron Shores-Chmielewski
- reference_types_module_rejected: externref table rejected by sandboxed engine
- trapping_actor_does_not_affect_sibling: trap in actor 1 doesn't corrupt actor 2
- alloc_returns_negative_for_nonzero_drops_message: negative ptr gracefully drops
- global_counter_accumulates_across_messages: mutable global state persists across 5 calls
All 131 tests pass. No new bugs found.
Authored by Claude, lovingly guided by Zachery Aaron Shores-Chmielewski
- mixed_outbox_partial_delivery: guest sends to valid + garbage addresses in same handle,
verifies only valid sends deliver and actor survives
- drop_oldest_mailbox_with_wasm_actor: DropOldest policy keeps newest 3 of 5 messages
- echo_to_full_inbox_silently_drops: bounded inbox capacity silently drops excess echoes
All 119 tests pass. No new bugs found.
Authored by Claude, lovingly guided by Zachery Aaron Shores-Chmielewski
- Non-ByteMessage to WASM actor is silently ignored (type mismatch)
- Guest mutable global state persists across handle() calls (counter)
- Native handler can spawn WASM actor dynamically via ctx.spawn()
Authored by Claude, lovingly guided by Zachery Aaron Shores-Chmielewski
- Module with trapping start function correctly rejected by builder
- Self-send feedback loop works (echo to own address, relay on next tick)
- Guest sending 10 messages in one handle: all 10 delivered via outbox
- Overlapping dest_ptr and payload_ptr in send: reads are independent
Authored by Claude, lovingly guided by Zachery Aaron Shores-Chmielewski
When a guest calls swactor.send() then traps, the outbox entries from
the incomplete operation survived and leaked into the next successful
handle() call, delivering messages from a failed context. Now clears
the outbox on trap, consistent with the "drop everything from failed
operations" semantics.
Authored by Claude, lovingly guided by Zachery Aaron Shores-Chmielewski
When a guest calls swactor.send then traps, the outbox entry survives
and leaks into the next successful handle() call. The outbox should be
cleared when handle traps, since the guest's operation was incomplete.
Also adds: invalid WASM bytes test, zero-length payload send test,
multiple sequential traps test (all passing).
Authored by Claude, lovingly guided by Zachery Aaron Shores-Chmielewski
StopSignal interception in tick_all set stopping=true but did not push
to the deaths vector, so phase 5b watch notifications never fired for
externally-stopped actors (via rt.stop_actor()). Now pushes
(addr, ExitReason::Stopped) to deaths, consistent with the
ctx.stop_self() path.
Authored by Claude, lovingly guided by Zachery Aaron Shores-Chmielewski
Native watcher watching a WASM actor does not receive ActorExited when
the target is stopped via rt.stop_actor(). Root cause: StopSignal
interception in tick_all (worker.rs:734) sets stopping=true but does
not push to the deaths vector, so phase 5b watch notifications never
fire for externally-stopped actors.
Also adds P0-2 through P1-7 test scenarios (all passing):
- empty message, oversized message, allocator exhaustion
- nonexistent address send, wrong export signature/name
- graceful stop, negative payload_len, independent stores
- WASM-to-WASM relay
Authored by Claude, lovingly guided by Zachery Aaron Shores-Chmielewski
Guest alloc could return a pointer where ptr+len exceeds linear memory
size, causing a Rust panic that permanently poisoned the actor. Now
validates ptr+len <= memory.len() before writing, dropping the message
on OOB (consistent with other allocation failure handling).
Authored by Claude, lovingly guided by Zachery Aaron Shores-Chmielewski
Guest alloc returning a pointer near the end of linear memory
(ptr + msg_len > memory_size) causes a Rust panic in copy_from_slice,
which poisons the actor permanently instead of dropping the message
and keeping the actor alive.
Authored by Claude, lovingly guided by Zachery Aaron Shores-Chmielewski
