TriggerEngine::CheckSample returned early in every state but ARMED, so an
edge arriving while a capture was being collected or handed out was
dropped, and the automatic rearm then waited for a FRESH edge. The engine
was therefore blind from its own trigger point until the capture had been
harvested — a post-window — and for the holdoff on top of that.
On a sparse pulse train that rounds the capture spacing up to a whole
pulse period: at the default 1 s window the blind stretch is 1 s, so a
1 Hz train was caught at 0.5 Hz and a wider window lost whole multiples.
The comparator now keeps running through COLLECTING and TRIGGERED and
remembers the first edge at or past trigTime + max(postSec, holdoffSec).
The holdoff guards against re-triggering on the ringing of the same
event and is measured from the trigger point, so it overlaps the
post-window rather than adding to it. Rearm() fires on the remembered
edge; it also keeps the tracked level, so the first sample after it has
a real predecessor instead of being spent seeding one.
Arm() stays the operator's arm and discards the held edge — they asked
for the next event, not one already been and gone — and SetConfig() and
Disarm() drop it too, since it was never judged against the new window.
This is the same defect and the same remedy already validated in the Go
hub (wshub/trigger.go, trigger_sporadic_test.go); the C++ hub had been
left with the original semantics.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Reported as samples sporadically carrying a previous packet's timestamp:
holes on one side of the stream and collisions on the other, in both the
Go and the MARTe2 receiver. That it appeared in both is what located it
-- the shared cause is upstream of either client. Four independent
defects, all of which end in a packet's values being placed at a time
that is not theirs.
Reassembly slot exhaustion (the "Reassembly slots full; evicting oldest"
flood). Chunk size was learnt only from fragment 0, so an out-of-order
burst destroyed a packet whose bytes had all arrived and left the slot
occupied until the 2 s GC. Slots were keyed on the counter alone, but
DATA and CONFIG number independently, so equal counters merged the two
streams. The 32-byte received-mask covered 256 of the 512 fragments the
client accepts, so a duplicate above 255 was counted as new and the
packet was delivered with a hole of stale bytes in it. And one datagram
was read per Execute(), which cannot drain a fast producer. Fixed with a
pendingTail deferral, (counter, type) keying, a 64-byte mask, a
256-datagram drain, counter-age slot reclamation, and a 1 Hz aggregated
warning in place of the per-eviction flood.
UDPStreamer dropping whole Accumulate batches. EventSem::ResetWait is
Reset-then-Wait, so a Post() landing while the sender thread was inside
ServiceClients()/SendData() was destroyed by the next Reset. The batch
was then skipped with dataReady false, readyFill was never cleared, and
the following flush overwrote it: an entire run of RT cycles never
reached the wire. The record of pending work now lives in the buffers
rather than in the semaphore edge, which also removes up to
UDPS_DATA_WAIT_MS of latency; genuine backpressure overwrites are
counted and reported. Against the unfixed code the new test sees
2999/3000 batches never consumed.
Period inflation after loss. Accumulated scalars carry no SamplingRate,
so the receiver derives dt from the sender-clock gap -- but dividing it
by the previous packet's sample count is only right while nothing is
lost. One loss doubles the reported period, which spreads a batch a full
batch past its own end and into the range the next packet claims. That
is the hole and the collision, exactly. Inferring the cycle count from
the estimate's own period is not a way out: it has a stable fixed point
wherever gap/dt is an integer, so a real rate change locks it at the old
one for good (AccumDtGTest.FollowsSustainedRateChange).
The packet counter removes the ambiguity, so all three receivers now
order on it: a DATA packet that does not advance the counter is dropped
rather than delivered, because its values are older than data already
handed over. Ordering is on the signed difference so it survives the
uint32 wrap, and the sequence resets on reconnect, where the producer's
counter restarts independently of ours. The loss count that falls out of
the same delta feeds the period estimate as cycles = prevN * (1 + lost),
which reduces exactly to gap/prevN when nothing is lost and therefore
still tracks a genuine rate change. UDPSClient::AcceptDataCounter (C++),
udpsprotocol.SequenceGate (Go), decode_data (C).
The C client's existing gap counter was wrap-unsafe and let a stale
packet rewind last_counter, which made every subsequent gap wrong; it
uses the same code now. Docs/Protocol.md gains an Ordering DATA section
stating the requirement for any receiver, including ones outside this
repository.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Add a FileWriter-compatible per-source binary recorder: native-type encode
and TypeDescriptor mapping, FileWriter header serialization, subset/quantized/
ACCUMULATE row serialization, and push-thread double-buffer flush with size-cap
rotation, keep-N pruning, fdatasync cadence, disk-free and staging-overflow
guards. Covered by 12 GTests; full suite (83) green.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
