run_experiments snapshots the site registry once at entry. stage-x is
registered lazily (first causal_site! execution in the worker), so on a
1-core box the snapshot deterministically wins whenever a sibling test
has already paid the tsc_hz calibration — stage-x missed the sweep and
the summary assert tripped (silently, pre-propagation; the earlier
cascade attribution was incomplete). The worker now signals after its
first site entry and the test waits on it before starting the sweep.
Jarred separately: the lazy-registration trap is a lib UX hazard worth a
doc note or warm-up guidance.
A fresh/reused slot started with causal_delay=0 and causal_parked=true:
the first resume then 'forgave' the entire monotone global backlog, once
per spawn, booked as park forgiveness. Under close-mode conn churn (~95k
spawns/s) that is millions of phantom forgiven ms per 700ms window, even
in 0% cells — the books could not balance under actor churn while
ka-mode stayed plausible (few spawns). Impacts were unaffected: the
first resume always precedes the first check, so nobody ever spun.
reset_counters now installs Coz's new-thread rule: causal_delay starts
at the current global ledger and causal_parked starts false — a newborn
neither owes nor is forgiven the process's history, and delay injected
while it sits spawn-queued (runnable, not blocked) is owed and paid at
its first check, the semantics audit_zero_pct_window_absorbs_leftover_
debt pins. That test (born failing, unmasked by the run() propagation
fix) and the new actor_churn_between_experiments_forgives_nothing
regression test both go green.
The trampoline caught the root's panic, recorded it as Outcome::Panic on
the slot, and run() dropped the initial handle without reading it — every
assert inside run(), the standard test-suite pattern, was silently
vacuous (found live: a failing-first test passed). run() now reads the
root outcome before the handle drop and resume_unwinds the payload after
full teardown, so a caller's catch_unwind leaves the Runtime reusable;
Exit and Stopped return normally. The payload message is printed before
re-raising (the throw-site hook output was suppressed in-actor).
Correct the two tests this unmasked, both born failing and never run:
the select loser-arm test kept a closed arm in the set (the documented
closed-arm rule: a closed arm reports ready forever — observe the
disconnect and drop it); the send_after-to-dead test expected Ok(None)
from a closed+empty channel (documented: Err(RecvError), which proves
nothing-delivered even more strongly).
GPU sweep decomposed the ~23ms/700ms @50 injection deficit: every
ledger bucket is ~zero (drop park 0, discards 0, drop yield ~0.3ms),
books balance at absorbed+forgiven = 4x injected in all 48 cells, and
the 0%-cell contamination signature is absent. The probe pins the
residual: eff 0.933-0.943, a constant 22-27µs missing per site entry
= ~4.9 slice-expiry yields/entry x ~5.6µs runqueue wait. The
"deficit" is runnable off-CPU time inside the site — wall time the
probe's ground truth counts but on-CPU attribution correctly skips
(Coz model: speeding the site's code does not shrink queue-wait).
Measure-only reclassification, no behaviour change: a yield in the
target site stashes (tsc, experiment epoch) on the slot; the next
on_resume counts the gap into OFFCPU_IN_SITE_{CYCLES,N} (would-be
delta terms, MAX_SAMPLE_CYCLES-capped) iff the epoch still matches
and the word is live — a gap straddling end()/a same-word begin()
(live in the probe's 50,50 schedule) is dropped, never a leaked
cooldown. Parks excluded: blocked time is forgiveness territory.
New offcpu column in render_ledger_audit; LedgerCounters and
ExperimentResult grow the two fields. Fidelity footer now states
the on-CPU basis (deliberate wording change to the pinned summary;
the substring pin test still holds). +2 tests (counted gap; epoch
straddle) + render assert.
Measure-only counters for the deficit hunt (~23ms short per 700ms window
at 50% on the bottleneck site; superlinear vs 25%). Nothing here changes
injection or absorption; the sweep decides the fix.
Buckets, windowed per cell into new ExperimentResult fields (audit
snapshot taken at end() — spin/attribution freeze there, forgiveness
does not):
- spin_absorbed / park_forgiven: where owed delay actually went. Spin
during a 0% cell is the baseline-contamination signature — leftover
debt from a prior window being paid in a later one (checks gate on
the experiment word, so cooldowns pay nothing and debt carries over).
- drop_park / drop_yield (+counts): the deschedule path flushes no
sample tail — an in-target-site park or yield silently loses
[last sample -> now]; on_resume re-arms before the actor runs again.
New on_deschedule hook in all three intent arms (real park; explicit/
slice-expiry yield; requeued park counts as yield — it never blocked).
Slice-expiry yields sample at the descheduling checkpoint, so a fat
yield bucket points at explicit yield_now or requeued parks.
- discard_overmax (+count, in would-be delta terms so columns compare
against injected_cycles) / discard_unarmed: the attribute() clamps,
previously silent.
LedgerCounters + ledger_counters() expose cumulative totals (tests,
run-level prints); render_ledger_audit() is the per-cell companion to
render_summary, which stays byte-identical (pinned). ExperimentResult
now derives Default so literals survive future audit-field growth.
Tests: +7 (spin counted, forgiveness counted, in-site park drop, in-site
yield drop, overmax discard, 0%-window leftover absorption — synthesized
deterministically via inject_delay_cycles_for_test with no experiment
active — and the audit render). 22/22 causal.
One unconditional footer line whenever there are results:
"note: impacts are lower bounds — undershoot grows with speedup pct;
rankings unaffected" — surfacing the RFC 007 Validation fidelity statement
where users actually look, instead of only in the RFC. Wording pinned by
the summary test.
send_after_wall / send_after_named_wall (+ Timers::insert_send_wall) arm a
message-delivery timer that opts out of the RFC 007 virtual-time shift and
fires at its raw deadline regardless of injected delay — the Send-reason
sibling of sleep_wall, closing the jar item whose substrate efbc254 landed.
For deadlines that reflect the outside world (protocol timeouts, wall-clock
schedules) rather than workload pacing. cancel_timer is anchor-agnostic and
unchanged; without the feature the API exists and is identical to send_after.
The gen_server timer layer (send_after_to, RFC 015 §5) deliberately stays
virtual-only — an opt-out there means new options on the gen_server/statem
timeout API, out of scope for now.
Tests: wall send fires at raw deadline while a virtual sibling shifts;
cancel on a wall send with debt outstanding; featureless delivery/cancel
smokes through the public named API. 15/15 causal, 34/34 binaries both
feature configs, lib clippy clean both.
New timer anchor: insert_sleep_wall / scheduler::sleep_wall (exported) opt a
Sleep entry out of the RFC 007 virtual-time shift, so it fires at its raw
deadline regardless of injected delay. Featureless config is unchanged (the
API exists but is identical to sleep).
The causal controller's window/cooldown sleeps and the tsc_hz calibration
sleep use it on the actor path (the OS-thread path was already wall). This
fixes the controller's own sleeps dilating under its own injection —
experiment windows stretched ~2x at 50% speedup (337ms -> 646ms injected/
window). Deltas were rate-normalized so results were unbiased; this fixes
sweep cost, not bias. The general wall-anchored-timer-semantics jar item
(user-facing opt-out) remains open; this lands the substrate.
Test: wall_timer_ignores_injected_delay — wall entry fires at raw deadline
while a virtual sibling in the same heap shifts. 13/13 causal, 34/34
binaries both feature configs.
Injected delays dilate virtual time for the workload, but timer deadlines
stayed wall-anchored: a sleep or receive-timeout fired early in virtual
terms, so timeout/retry behaviour sped up relative to the dilated world
(v1 known gap #1).
Every heap Entry now carries delay_stamp — the global delay ledger at
(re-)queue time, cfg-gated on smarm-causal. pop_due converts any debt
accrued since the stamp to wall time (tsc_hz) and shifts the effective
deadline; a not-yet-due entry is re-queued at the shifted deadline with a
fresh stamp, so it keeps chasing delay injected while it waits. seq is
preserved across re-queues, keeping send_after cancellation identity
intact (cancelled entries are discarded before any shift). Zero debt is
byte-identical to the old path; peek_deadline may under-report, costing
one spurious scheduler wake per injected chunk (documented).
This also makes the park-gated resume credit *correct* rather than
forgiving for sleepers: a sleeping actor now physically pays its debt by
sleeping longer, so the on_resume fast-forward reflects real payment
(sleeping_actor_pays_injected_delay pins this end-to-end through the
runtime).
New test hooks: inject_delay_cycles_for_test (deterministic ledger
driver, eagerly TSC-calibrating so conversion never stalls a scheduler
loop) and cycles_to_duration. The ledger is process-global, so the
delta-sensitive causal tests now serialize on a shared test mutex — they
were racy under the parallel test harness before this, in principle.
Samples were taken only when maybe_preempt's cold block happened to fire
in-site, so the interval between the last check and SiteGuard drop was
discarded on every site entry. Measured live on the 24-core box:
22-29us lost per entry, a constant attribution efficiency of ~0.93-0.94,
which under-reported every impact (+83.5% where theory says +100%; the
observed shortfall fits 1/(1-pct*eff)-1 at both 25% and 50%).
SiteGuard enter/drop now call site_transition(): leaving the target site
flushes the pending interval into the ledger (sample-only, never spins,
so safe under no-preempt regions); entering the target site re-arms the
sample clock so pre-site time is never attributed (the symmetric
over-attribution). Winner attribution is factored into attribute(),
shared by the cold check and the flush, with the same interval clamps.
Adds examples/causal_attrib_probe.rs (ground-truth in-site time vs
ledger attribution, the probe that confirmed the leak) and the
site_boundaries_flush_tail regression test (a site entry that never
hits a cold check must still be attributed). Also gates causal_probe
on smarm-causal in Cargo.toml - it never was, so featureless builds
of the examples were broken.
causal_site! scoped site guards per actor slot, progress! throughput
points, and a Coz-style virtual-speedup engine hooked into
maybe_preempt's amortized cold block: target-site samples grow a global
delay ledger; bystanders spin-absorb their debt at the next causal
check, with timeslice extension so injected delay is not charged
against the slice.
Resume credit (Coz's blocked-thread rule) is gated on a causal_parked
slot bit set only by a real park: crediting on every resume made any
yield-cadence actor delay-immune and every experiment inert (found
live on a 24-core run — dead-flat deltas across all sites).
Report normalization uses a measured TSC frequency (~50ms calibration
on first use) instead of the crate-wide 3 GHz assumption, which
uniformly inflated impact numbers on a 3.7 GHz box. impact_pct() is
the machine-readable form of the summary for programmatic checks.
examples/causal_pipeline.rs burns fixed *work* (calibrated LCG loop),
not fixed wall time — a timed busy-wait absorbs injected delay into
its own budget and reads as a no-op. Self-checking: exits nonzero if
causal separation fails; skips the verdict below 4 cores. Validated
on a 24-core box: reserve (true bottleneck) +29.3%@25/+83.5%@50;
serialize and background-compaction ~0%.
Known v1 gaps (jar): timer-heap deadlines unshifted, no-check!/no-alloc
actors undelayable, multi-scheduler coherence best-effort Relaxed,
off-CPU blame punted, Instant::now() uncorrected.
Zero-cost with the feature off; clippy -D warnings clean both ways;
full suite green with and without smarm-causal.
Root cause of soak20 signature 2 (refcount_test.exs 'watcher crash',
110235x fast {:error, :server_down} probes over the full await window):
by_name mapped name -> slot *index*, so a name whose holder died (no stop
path unregisters; prune is lazy) and whose slot was then re-tenanted read
as live-held: register failed NameTaken{holder: <unrelated tenant>} (which
the bridge macro's generated start() swallows -> start_server/1 reports :ok
for a server that never came up), while name resolution reached the
tenant's mailbox, missed on the message TypeId and failed fast WITHOUT
pruning — the wedge self-sustained for the tenant's lifetime. Name-
addressed send additionally judged liveness on the slot's *current*
mailbox pid, so a same-typed tenant would have received the message
(misdelivery) and a differently typed one a misleading NoChannel.
Fix: by_name: HashMap<&'static str, Pid> — every reader judges the
*stored* holder with the generation-checked live(), so a recycled slot's
tenant no longer impersonates a dead holder, and every touch (register /
whereis / resolve / send) prunes and heals a stale name. prune(index)
becomes prune_holder(pid): names bound to the holder go; the mailbox goes
only while still the holder's own (a tenant's replacement mailbox is left
untouched). Introspection matches names to mailboxes by full pid, so a
stale name never annotates a slot's new tenant.
Deterministic regression test added first and shown to fail pre-fix
(tests/stale_name_slot_reuse.rs: tiny slab forces re-tenanting; old slot
(1,0) died, tenant (1,1) took the index; register -> NameTaken pre-fix).
Post-fix it asserts the healed contract: whereis -> None (pruned), call ->
ServerDown, re-register -> Ok. Suite 33 ok-binaries, clippy gate clean.
In the wild the window opened at every splice_test teardown:
Splice.terminate -> exit_server('subtree') left the name bound; width 20
raised the re-tenant probability. Downstream (smarm_beam): install_child's
unregister-before-register workaround becomes dead code (removed there);
the #[smarm_server] macro's swallowed register error becomes truthful
idempotency (a NameTaken now really is a live holder).
A queued Envelope::Call was stranded until the last Sender dropped, so a
caller parked in gen_server::call was never released with ServerDown when a
*named* server was request_stop'd — the registry's inbox Sender clone (lazy
prune) kept the channel Arc, and the queued reply_tx, alive indefinitely.
Receiver::Drop now drains the queue (items dropped after releasing the lock,
since a reply_tx drop reaches a different channel's lock + the scheduler),
restoring the documented ServerDown guarantee on every teardown path.
Adds tests/stop_with_queued_call.rs: deterministic pure-smarm reproducer.
A `=> postpone` row (cast/call/info) defers the current event untouched,
to be replayed after the next real transition. `handle` is now two-phase:
a borrow-only postpone pre-pass that hands the event back as
`Step::Postponed(ev)`, then the existing consuming `match (state, event)`.
The loop owns a FIFO queue, drained in the new state ahead of further
intake; a replayed event may postpone again. A postponed `call` keeps its
Reply, so a later state answers it.
`handle` returns `Step` (Postponed / Transitioned / Stayed) so the loop
can see both deferral and transition without reading the state cell.
`Resolution::Postpone` is removed: postpone is a pre-dispatch routing
decision, not a consuming-dispatch outcome.
Add the two timeout flavours, both surfacing as ordinary events matched
in on-state arms:
- cx.state_timeout(d): fires a state_timeout event after d in the current
state, auto-reset by the loop on every real transition.
- cx.timeout(name, d): fires a timeout(name) event after d, surviving state
changes, keyed by name, with cx.cancel_timeout(name).
Both ride the existing timer min-heap via send_after_to onto a new per-loop
system channel, selected above the inbox so a fire can't be starved by inbox
traffic. A local-id stamp on each fire lets a reset/cancel that loses the race
discard a stale fire. The macro grows an info: clause and folds three internal
Ev variants (Info, StateTimeout, Timeout) alongside cast/call, with new row
keywords info / state_timeout / timeout. Unmatched info silently drops (the
gen_server default); state/named timeouts have no default, so a state that can
see one must handle it or the match is non-exhaustive.
Rename the hand-written expansion-target example fused -> expanded, retire the
deprecated Switch demo machine (its round-trip / enter / panic-down coverage
moves onto the timer machine), and refresh the macro docs to the door machine.
cargo build --all-targets warning-free; cargo test green.
- rename StatemRef/StatemCallError/StatemSendError -> GenStatem*
- move the inline unit test out of src; consolidate the Switch coverage
onto a single macro-driven harness in tests/gen_statem.rs
- drop the redundant hand-written Switch test machine and the two
untracked rejected-direction probes (succ_enums, typed_edges)
- rename examples statem_{fused,macro}.rs -> gen_statem_{fused,macro}.rs
- strip RFC/chunk/spike provenance and fix the mislabeled "throwaway"
example header and dead cross-references
Runtime support layer for gen_statem, built against existing public API
(channel + scheduler::spawn), sibling to gen_server. No macro: per review,
build the primitives first and hand-write the Switch example to evaluate
whether a statem! macro earns its place before committing to one.
- src/statem.rs: Machine trait (on_start/handle), Resolution<S> with
From<S>, Cx (on_unhandled), Reply<T> move-only reply handle, StatemRef
(send/call), spawn + inbox loop. Real time only; Postpone + Cx timeout
arming are in the type surface but not yet acted on (chunks 2-3).
- examples/statem_switch.rs: the RFC Switch machine hand-written against the
primitives, tagged USER vs MACRO to mark what a macro would generate.
Asserts the RFC end-state (flips=1, enters=3).
- tests/statem.rs: call/cast round-trip, enter-on-start/transition-not-stay,
panicking-handler -> Down.
Reply<T> included (the call helper needs a handle type; keeps the example
true to the RFC surface) but isolated and trivially removable if we drop it.
A thin GenServer consumer of the Chunk-1 read primitive — the live
observer process (D12). ObserverRequest/ObserverReply are the wire
contract (D11); the version rides along on the snapshot/tree payloads,
which already carry SNAPSHOT_FORMAT_VERSION (D1). Behind the new
`observer` Cargo feature, off by default (D10): the primitive stays
always-on, only the transport is gated. Cast is Infallible, so the
server takes no async traffic and handle_cast is statically unreachable.
Gated integration test proves each verb relays exactly what the
corresponding primitive returns (snapshot/tree/actor_info), incl. a
forged-pid None and a live Parked classification.
Accumulate on-CPU cycles per actor as ActorInfo.budget_cycles, behind
the off-by-default budget-accounting feature (D6) — a reductions-style
work metric for relative comparison across runs.
- Approximate by design (per Mark): charge now - slice-start once at the
yield point, reusing the timestamp reset_timeslice already sets, so
one RDTSC per resume not two. Wake-slot resumes inherit the slice and
so slightly over-attribute the chain's time to the woken actor — noise
that averages out; we trade exactness for half the hot-path cost.
- Field/ActorInfo member are unconditional (keeps the snapshot shape
stable across the feature flag, D1); only the accumulation is gated,
so default builds are byte-identical and pay nothing. Reads return 0
when off. Single-writer Relaxed like the other counters; reset in
reset_counters (D7).
Matrix: default + feature-on + rq-mpmc + rq-striped + release + trace all
green; loom unaffected (feature off under --cfg loom).
Tally each dequeued message against the receiving actor, surfaced as
ActorInfo.messages_received (the 'is this actor draining slower than its
mailbox fills' signal).
- messages-received, not sent (D4): the receiver counts on its own
thread, so it's a single-writer Relaxed load+store on a hot Slot
AtomicU64, no atomic RMW (D5); reuses the stashed *const Slot from 2a.
- Incremented at all six channel dequeue-success sites (recv,
recv_timeout x2, recv_match, try_recv_match, try_recv) via
preempt::note_message_received; no-op outside an actor (null slot).
- Resets with overruns in reset_counters across the three lifecycle
sites (D7).
Matrix: debug default + rq-mpmc + rq-striped + release green; loom
slot_state/run_queue models pass (the 3 send_after_to loom failures are
pre-existing at fc014c4 — plain #[test]s under --cfg loom, not Chunk 2).
Tally overruns at the slice-expiry site in preempt.rs (the RFC 006
signal), surfaced as ActorInfo.overruns.
- Counter is a hot-region AtomicU64 on Slot, single-writer Relaxed
load+store (no atomic RMW), read Relaxed by the snapshot (D5).
- Reached from the rare expiry branch via a stashed *const Slot in a
preempt thread-local, set/cleared on the same resume/return boundary
as CURRENT_STOP — one TLS load, no runtime lookup. Shared infra for
the messages-received counter next.
- Reset in all three slot-lifecycle sites: Slot::vacant, reclaim_slot,
install_actor (standing invariant, D7) — per-incarnation counts.
tree() / tree_from() fold a Chunk-1 snapshot into a forest by grouping
each actor under its parent pid — one O(n) pass, no new reads. tree_from
is public so a held (or synthetic) snapshot can be folded without a
second scan.
- D8: actors parented at ROOT_PID are genuine roots; an actor whose
recorded parent is absent from the snapshot is re-rooted under the
sentinel and flagged orphaned, so the forest stays total. take()-on-
place doubles as a guard against re-entering a node.
- D9: the edge is parent/spawned-by, documented as not necessarily
supervision.
snapshot() / actor_info() return owned ActorInfo over the slab: pid,
names, fine-grained scheduling state, parent edge, trap flag, mailbox
depth, and monitor/link/joiner counts. Pure reads, no hot-path change.
- ActorState maps the packed slot word (no new storage); introspect.rs.
- D1: RuntimeSnapshot carries SNAPSHOT_FORMAT_VERSION from day one.
- D2: per-slot tearing (ps semantics); actor_info coherent per actor.
- D3: mailbox depth included. Registry channels now stored behind an
ErasedSender trait (downcast for clone_sender + type-erased
queued_len); depth summed over published channels under the registry
Leaf (Leaf -> Channel), kept out of the cold-lock pass so no two
Leaves are ever held at once. Depth covers published channels only.
- Done slots surface as root-less tombstones.
Add a cancellable message-delivery timer on the existing timer.rs min-heap,
the substrate the gen_server time idioms (idle/receive timeout, periodic tick,
debounce/backoff) need.
- Reason::Send { fire }: a type-erased delivery thunk. send_after (Pid<A>, via
send_to) and send_after_named (Name<M>, via send) capture dest+msg and
resolve the address on fire, not at arm time, so a dead target / restarted
name is observed when it fires; a failed resolve or closed inbox is dropped
(Erlang erlang:send_after semantics).
- Cancellation via an "armed" set keyed on the entry seq, exposed as an opaque
TimerId. Only Send timers use it; Sleep/WaitTimeout keep their inert-stale
behaviour untouched. pop_due fires a Send only while still armed and removes
it, so cancel returns true iff it landed before the fire (the race signal).
cancel is unscoped: any holder of the id can cancel (e.g. racing two servers
and cancelling the slow path).
- peek_deadline contract documented as "<= true next deadline" so a future
hierarchical timing wheel can back Timers without touching send_after or the
scheduler idle path. call_timeout left on recv_timeout (blast radius).
Tests: Timers-level (fire/cancel/race/clear/ordering) plus scheduler-level
delivery for both Name<M> and Pid<A>, cancel-prevents-delivery, and silent
drop on unresolved name / dead pid. Green on rq-mutex/rq-mpmc/rq-striped.
The explicit fallible fallback for when only an untyped Pid is in hand (off a
Down, out of a future members()), so the typed send_to / send-by-name stay the
default and callers don't grow workarounds around a missing primitive.
- send_dyn::<M>(Pid, msg): one line over the shared send_to_pid core. Same
identity-bound, no-redirect liveness as send_to (Dead once the incarnation is
gone); the difference is the explicit M, so NoChannel is a real runtime
outcome here rather than the debug_assert it collapses to on the typed paths.
- Takes Pid<Erased> specifically: a typed Pid<A> must .erase() to reach it, so
opting into the fallible downcast stays visible at the call site.
tests/registry.rs: send_dyn_delivers_and_reports_wrong_type (right type
delivers; live actor with no channel for the asked type -> NoChannel) and
send_dyn_to_dead_pid_is_dead. Full suite + order-checker green, warning-free.
This completes the RFC 014 send surface: Name<M> (re-resolving), Pid<A>
(typed direct), send_dyn (bare-pid). send_after (§6) can now target any of
them — its Dest question is answered.
Make a Pid<A> messageable: install + send_to, the direct addressing mode that
dies with the actor and never redirects (the counterpart to the re-resolving
Name).
- install::<A: Addressable>(tx: Sender<A::Msg>) -> Pid<A>: opt-in, lazy,
nameless publish (RFC 014 §5). Files the actor's inbox into the by_index
mailbox table under TypeId::of::<A::Msg>() and re-types self's identity as
Pid<A>. Infallible: no name to collide on, self is always live in run().
- register and install now share publish_channel (the insert-or-extend-mailbox
step factored out); register additionally binds the name. Byte-identical
storage, so name- and pid-addressing populate the same table.
- send_to::<A>(Pid<A>, A::Msg): resolve by raw pid, deliver with NO redirect.
The stored mailbox must be this exact incarnation (generation included) and
live, else SendError::Dead — even when the slot now holds a different live
actor, which is left untouched. Resolution clones the Sender under the Leaf
lock, releases, then sends (Leaf -> Channel), as name send / pg / finalize.
- send_to_pid is the shared raw-pid core; send_dyn (§4.6) lands on it next.
- SendError gains Dead(M), the pid-path counterpart to name-path Unresolved;
into_inner / Display / Debug updated. Name send never yields Dead, pid send
never yields Unresolved — disjoint by construction, documented on the enum.
tests/registry.rs: install_then_send_to_pid_delivers; and the load-bearing
send_to_does_not_redirect_after_takeover (slot reused by a new incarnation, the
stale Pid<A> send returns Dead and the new occupant gets only its own message).
Full suite + order-checker green, warning-free across all targets.
Rip out the old name<->pid bimap; rebuild registry.rs as the live mailbox
directory. A name resolves to a SINGLE actor (many-per-label is pg's job); that
actor owns a SET of typed channels (channels are typed, so no single mailbox),
keyed by message TypeId. Resolution: name -> pid (one actor) -> Mailbox ->
channel for type M. Same name + different type parameter selects different
channels of the one actor, so capability separation (§4.7) needs no new type.
- register<M>(Name<M>, Sender<M>): captures the current actor's channel under a
name (one step, per decision). Adds channels cumulatively; NameTaken only vs a
different LIVE holder; stale/dangling bindings pruned on contact.
- send<M>(Name<M>, msg): the payoff — resolve, clone the Sender UNDER the Leaf
lock, release, then send (Leaf -> Channel; a send can unpark). Returns the msg
on Unresolved / NoChannel / Closed.
- whereis -> single live pid; unregister frees only the name (mailbox/other
names survive). name_of (reverse lookup) dropped — deferred to introspection.
- Contained erasure: each channel is Box<dyn Any+Send> filed under TypeId::of M
and downcast to its own keying type, so the downcast can't fail on good data
(debug-asserted via the stored type_name). Phantom M on Name re-imposes type.
- One Leaf RawMutex (the fold), so a name send stays on a single Leaf — two
Leaves at once is a hard panic. runtime.rs field/init unchanged (same Registry
name + new()).
tests/registry.rs rewritten for the new semantics: send-by-name delivery,
many typed channels on one actor, same-name/different-type routing, NameTaken,
takeover across slot reuse, Unresolved/NoChannel. Also fold in a Phase 1 fixup:
the phantom test key was an enum whose variants tripped dead-code under the test
build (only caught now that I run -D warnings on --tests). Full suite +
order-checker green, warning-free across all targets.
Wire eviction to actor death via the monitor subsystem, using drain-on-
contact (no scheduler hot-path hook).
- Each membership now carries its own Monitor alongside the group entry;
join() installs monitor(pid) (registration races finalize_actor under the
cold lock, as every monitor does) and a redundant join tears its extra
monitor back down.
- reap_group: every group op drains the touched group's monitors with a
non-blocking try_recv (a delivered Down or closed channel = dead) and, on
the first death, sweeps that pid out of *every* group via remove_where —
so a death vanishes from all its groups on next contact.
- Lock discipline: monitor()/demonitor() (cold = Leaf) run outside the group
lock; try_recv (Channel) runs under it, permitted by the Leaf->Channel
order; evicted/rejected monitors are dropped after the lock releases so no
receiver-drop wakeup runs under it. Validated by the debug-build lock-order
checker passing under the integration tests.
- remove_where now returns the evicted monitors (for deferred drop) and
preserves intra-group insertion order.
- Tests: unit (synthetic monitors) for reap live/dead/closed + sweep;
integration (real actors) for death-vanishes-from-every-group, pick on an
all-dead group, peer survival, leave isolation, dead-at-join eviction.
Outstanding: a miri pass (cargo +nightly miri test --lib pg::) is untried —
first build exceeds the sandbox window and the futex RawMutex may need a miri
shim; the mechanical Leaf->Channel checker is the primary soundness guard.
request_stop sets the flag and issues a wildcard unpark, but unpark
no-ops on a Queued actor (the pending run "is" the wake). If the
actor's first action on resume is a blocking park — no allocation, no
check!() on the way — the wake-side-only check in park_current is
unreachable: the actor parks with the stop flag already set and no
wake is ever coming. The runtime then idles forever.
Fix: check_cancelled at the ENTRY of park_current, before the switch.
The remaining window (flag lands after the entry check, before the
park) is covered by the existing protocol: the stop's unpark finds
Running / the prep-to-park window, sets Notified, and the park-return
re-queues into the wake-side check. Unwinding from the entry is the
same unwind path as the wake side; leftover wait registrations are
stale-epoch / dead-generation and self-clean at their wakers' failed
CAS.
This is bug #1 of the urus chunk-2 session (bug #2, the terminal
wake, is the preceding commit). Repro: tests/cancel.rs
stop_flagged_while_queued_lands_at_first_park — deadlocks without the
fix (watchdog-bounded), passes in <50ms with it.
An idle scheduler thread blocks in poll_wake on a snapshot of
peek_deadline / io_outstanding. Only the io threads ever write the wake
pipe (on completion push); enqueue does not. Mid-flight that is masked —
the thread that caused an enqueue is awake and processes it — but at
termination the snapshot can go terminally stale, two ways:
- Stale io_outstanding (infinite hang): an actor parked in wait_readable
is request_stop'ped. Cancellation deregisters the waiter and produces
NO completion, so the wake pipe is never written. A sibling blocked on
io_outstanding > 0 with no timers pending sits in poll(-1) forever.
- Orphaned timer deadline (finite stall): an actor cancelled out of a
long sleep leaves its timer entry behind. Clearing the wheel at
AllDone doesn't help a sibling that already blocked on that deadline;
it sleeps it out in full.
In both cases the remaining actors finish on the other scheduler thread,
which reaches the live==0 && io_out==0 verdict and returns without
waking the blocked one. Runtime::run then stalls or hangs in its worker
join. Found via urus's graceful-shutdown tests (~10% flake, one test per
variant); confirmed by gdb dumps of hung processes showing run() in
JoinHandle::join over a sibling in poll(wake_fd, 59750ms) resp. poll(-1).
Fix: Io::wake() writes the wake pipe directly; the AllDone arm calls it
after the timer clear. One byte wakes every poller; each re-runs the
verdict, independently reaches AllDone, and re-wakes — idempotent.
tests/terminal_wake.rs reproduces both variants deterministically: the
root busy-spins (no timer entries, occupies one scheduler thread) so the
sibling settles into the stale idle wait before the stop is issued. Both
hang without the fix and pass in <0.5s with it.
Known residual gap, deliberately unfixed: the timers-pending /
no-io-subsystem idle branch blocks in thread::sleep and has no wake
mechanism at all — the same stall exists for runtimes that never
initialize io. Roadmap candidate alongside cross-thread unpark and
entry-side check_cancelled in park_current (the lossy-QUEUED-stop bug).
A thread-local Cell<Option<Pid>> per scheduler, checked before the shared
run queue. Runtime-selected via Config { wake_slot: bool }, default OFF
until the slot shootout accepts it (one binary benches both arms).
Push policy: slot-eligible iff the wake originates from actor context
(current_pid().is_some()) — the slot push replaces run_queue.push at the
tail of the unpark protocol's Parked → Queued CAS, so at-most-once-enqueued
holds verbatim as (slot ⊕ shared queue). Scheduler-context wakes (timer/IO
drain) and spawns always go shared (spawns never reach unpark_inner at all).
Displacement is Go semantics: newest wake takes the slot, occupant pushed
shared — moved, never copied.
Pop order: slot, then shared. Slot-popped actors skip reset_timeslice() and
inherit the waker's remaining slice; a handoff chain is bounded by one
slice, after which the preempt-yield re-enqueue goes shared (a yield is not
a wake) — the one-slice starvation bound, zero new counters. Idle and
AllDone are only reachable with an empty local slot by pop order; an
occupied slot elsewhere holds a Queued (live) actor, so the counter-first
termination argument is untouched.
Observability: per-thread slot_hits / slot_displacements (reset at run()
start so post-run stats() reads are per-run), SlotPush/SlotPop trace events.
Bench plan (roadmap v0.9 item 2): rq_runtime gains the slot on/off
dimension (SMARM_BENCH_SLOT, default "0 1") — ping-pong-pairs is the
target metric, yield-storm the regression guard, spawn-storm the
neutrality check. RQCSV grows a slot column; RQSLOT lines carry the
counters; bench_rq.sh aggregates both. Tests pin the push policy through
the counters (actor-context hits, spawn/join bypass, displacement,
default-off, per-run reset).
FdArm composes fd readiness with channel arms on one wait epoch.
Selectable grows fallible sel_register and an eager-cleanup hook;
losing/stop-unwound/timed-out fd arms are unregistered (waiters entry
+ kernel ONESHOT) so the fd is never poisoned. try_select /
try_select_timeout surface registration errors (EBADF, EMFILE,
AlreadyExists) instead of RFC 008's permanently-ready lean, which
would busy-loop a healthy-but-unregistrable fd; select/select_timeout
stay infallible for channel-only arms. Adds wait_readable_timeout /
wait_writable_timeout as one-arm selects.
Known benign race (pre-existing, slightly widened): a queued FdReady
racing the cleanup DEL can spuriously wake a fresh waiter on that fd;
absorbed by select's defensive re-loop. Fixable by epoch-stamping
completions.
A stopped actor unwinding out of wait_fd's park leaked its waiters entry
and the kernel-side EPOLLONESHOT registration; the stale entry then failed
every future wait_*() on that fd with AlreadyExists (the defensive bare DEL
in epoll_register sits behind the contains_key check, so it never ran).
Fix where the invariant breaks: a drop guard in wait_fd, armed after a
successful register and forgotten on the normal wake (where FdReady already
removed + DEL'd). On unwind it cleans up iff the entry is still this wait's
(pid, epoch) — an entry consumed by a racing FdReady means the fd may carry
another actor's fresh registration, which must be left alone. Closes the
v0.2 fd-hygiene TODO.
Monitors are created at runtime (you watch a worker you just spawned in a
handler), so down arms can't ride the static info list. init grows a
&ServerCtx parameter (breaking; no-op default) whose clonable Watcher hands
Monitors to the loop over a control arm; the loop selects the live down
arms ahead of everything else. Arm priority: downs → control → infos →
inbox. A delivered Down retires its arm (monitors are one-shot); a state
that never clones the Watcher closes the control arm after init and the
loop falls back to the plain-inbox park.
type Info + handle_info (no-op default) on GenServer; ServerBuilder
(with_info/under/start) so start variants don't multiply — start/start_under
stay as wrappers. The loop selects [infos.., inbox] in priority order when
info arms exist, and keeps the plain recv() park when none do. Closed info
arms are silently removed (closed-arm-is-ready-forever); a closed inbox
still means graceful shutdown.
The deadline is one more stamped waker on the select's wait epoch: a unit
TimerTarget whose on_timeout is a bare unpark_at. Nothing registers in any
arm for it, so there is nothing to cancel or leak — an arm winning leaves
the timer entry to die at its epoch CAS; the timer winning leaves the arms'
registrations to self-clean like any select loser's. Classification is pure
channel state (wakes are precise): some arm ready -> Some(first, priority
order); none -> the timer was the only remaining stamped waker -> None.
Message-first on a raced deadline, same as recv_timeout.
Registration pass factored into register_arms, which owns the retire_wait
obligation on the ready-now exit for both entry points.
select(&[&dyn Selectable]) -> usize registers (pid, epoch) in every arm
under one wait epoch, parks once, and returns the first ready index in
documented priority order (BEAM-style; no fairness promise). A closed arm
counts as ready — and stays ready forever, so callers drop it from the set
once observed. Losing arms need no cancellation pass: the winning wake
consumed the epoch, so their registrations die at their wakers' failed CAS
or are overwritten by the receiver's next wait on that channel (the
single-receiver debug_asserts relax to 'none or own pid' accordingly).
The one genuinely new protocol piece is the no-park exit: returning with an
arm ready at registration time leaves earlier arms holding LIVE-epoch
registrations, whose wakes could fault a later one-shot park as a pending
notification. scheduler::retire_wait closes it — bump the epoch (in-flight
wakes die at their CAS), eat a notification that already landed
(StateWord::clear_notify), then re-observe the stop flag, in that order.
Proved by the new loom theorem retire_eats_late_arm_notification; the
integration probes in tests/select.rs fire stale loser-arm wakes and assert
a subsequent sleep's one-shot park holds its full duration.
The slot epoch is THE wait identity, so the hand-rolled per-primitive copies
go away: channel loses cur_wait/next_wait_seq/timed_out, mutex loses Wait.seq
and next_seq, TimerTarget::on_timeout takes the epoch. Registrations become
(pid, epoch) — channel parked_receiver, mutex waiters, io fd waiters,
Blocking io completions, sleep timers, joiner lists — and their wakers move
to unpark_at. begin_wait is lock-free, so each primitive opens the wait
inside the same critical section that publishes the registration.
recv_timeout's wake-classification loop collapses: wakes are precise, so
queue → Ok, senders==0 → Disconnected, else Timeout — the 'Defensive'
re-register branch is now unreachable by protocol, not by audit. Same for
Mutex::lock_timeout's one-shot park.
End-state invariant, auditable in one sentence: the only wildcard wake is
request_stop, which is terminal.
Deviates deliberately from the roadmap's monitor-based sketch
(monitor + wait reply-or-Down-or-deadline + demonitor): server death is
already observable on the reply channel itself - the reply sender drops
as the server unwinds, closing the channel and waking the parked caller.
So a bounded call is exactly recv_timeout on the reply channel, mapping
Disconnected -> ServerDown and Timeout -> Timeout. No registration
exists, so a timed-out call leaks nothing by construction - the
guarantee the monitor design had to engineer.
Semantics on timeout match Erlang: the request stays in the inbox and
is still handled; the late reply's send fails harmlessly against the
dropped reply receiver. CallTimeoutError keeps ServerDown and Timeout
distinguishable; the infallible call() is unchanged.
Tests: reply within deadline, timeout against a slow (parking) handler
with elapsed bounds, server survival across an abandoned call (late
reply discarded, bounded and unbounded calls keep working), and
ServerDown-not-Timeout for both death paths (mid-call panic and
already-gone inbox).
The per-recv deadline the roadmap deferred, built exactly the way
Mutex::lock_timeout already works: register the wait with a per-wait
seq, arm a timer::Reason::WaitTimeout with the channel inner (now a
TimerTarget) as the target, park. On expiry on_timeout cancels the wait
only if that same seq is still parked; satisfied or abandoned waits
leave a stale heap entry that no-ops on seq mismatch, per the
no-cancellation convention in timer.rs.
Race resolution is message-first: a send that lands by the time the
woken receiver runs is delivered even if the deadline had also passed.
Closure is reported as RecvTimeoutError::Disconnected, keeping timeout
and server/sender death distinguishable for callers (gen_server call
timeout builds on exactly this distinction next).
The timer is armed outside the channel critical section (the timers
lock must never nest under a Channel-class lock); the unpark race this
opens is absorbed by the RunningNotified protocol.
Tests: immediate delivery, actual timeout (with elapsed check), prompt
wake on send, Disconnected on close, zero-duration poll, post-timeout
seq isolation (stale entry must not cancel later waits), and a
4-scheduler mixed-outcome run (12 fed / 12 timed out).
register/whereis/unregister plus the inverse lookup name_of. Erlang
semantics: one name per pid, one pid per name, registering over a live
binding errors; same-binding re-register is an idempotent Ok.
Bimap = two HashMaps held in exact inverse under one Leaf RawMutex in
RuntimeInner; the invariant is debug-asserted on every mutation. The
inverse direction costs one extra String per binding and buys O(1)
pid->name for supervisors/tracing/diagnostics.
Cleanup is lazy: no finalize_actor hook, no Slot field (which would buy
into the reset-in-three-places invariant). Liveness rides the
generation-checked slot word - pids are never reused, so a stale binding
is detectable, never misdirected - and bindings to dead actors behave as
absent, pruned on contact by whereis/name_of/register/unregister.
Liveness checks under the registry lock read only the atomic slot word,
so the leaf rule holds trivially.
send_by_name is deliberately absent: smarm has no per-pid mailbox, so
there is nothing generic to send *to* - the registry yields a Pid,
usable with monitor/link/request_stop and as routing for user channels.
Tests: roundtrip both directions, idempotence, NameTaken on live holder,
one-name-per-pid, NoProc, death-evaporates-binding (via lookup pruning
and via register's own eviction path), unregister, cross-actor lookup
wired into request_stop, and a 4-scheduler name-contention churn test.
Phase 2 surfaced the hole this closes: channel guards were held with
preemption enabled, so a timeslice switch inside a channel critical
section could release the pthread mutex from a different OS thread (UB).
RawMutex disables preemption for the guard span and is
cross-thread-release sound by construction; poison goes away with it.
The strict single-class leaf rule cannot survive the migration: finalize
clones the supervisor/trap senders and monitor() clones the Down sender,
all under a cold lock, and those senders live in the slot - the nesting
is structural. The leaf check is therefore generalized to two classes:
Leaf (cold locks, free list, stack pool) and Channel. Order is
Leaf -> Channel, at most one of each; both directions of violation are
debug-asserted at the acquisition site. Channel critical sections call
only the lock-free unpark protocol, so the order is acyclic.
Tests: class-ordering unit tests in raw_mutex (allowed nesting + all
three rejected shapes), plus a multi-scheduler integration test driving
channels through monitor churn and actor death so any ordering
regression trips the debug assert instead of deadlocking.
Claude (sandbox)