Reuse active Sensor tracker wake deadlines

This commit is contained in:
mikecarper committed 2026-10-17 02:25:37 -07:00
1 parent 629557fcfe
commit 8a5f735bba
4 files changed
+305 -10

No files matched your search

+22 -5
View File
@@ -77,8 +77,8 @@ was added.
The test device was a RAK3401/nRF52840 with an external RAK13302 SX1262 on
Mercerwood. Tests used 908.5 MHz, BW250, SF9, CR5, power saving on, manual
RXPS 30000/10000 us, AGC 0, TX power 2, no forwarding/adverts and USB watchdog
off. A single private benchmark binary allowed changes without reflashing
between the decisive controls. Benchmark controls are not production features.
off. Runtime switches in private benchmark binaries allowed same-binary paired
controls without reflashing. Benchmark controls are not production features.
| Control | Capture | Time-weighted sleeping observations |
| --- | --- | --- |
@@ -89,7 +89,7 @@ between the decisive controls. Benchmark controls are not production features.
| No late SVC, non-yielding USB poll, without early clear | 30 s | 0% |
| Early clear, no late SVC, non-yielding USB poll | 90 s | 76.7% |
| Never-DTR client held open, old empty USB poll | 30 s | 0% |
| Same client held open, non-yielding USB poll | 30 s | 82.8% |
| Same client held open, non-yielding USB poll | 30 s | 82.9% |
The first production candidate restored sleeping observations with SoftDevice
off (6522/8331 samples, about 78% over 90 s), but not with SoftDevice on
@@ -120,8 +120,9 @@ The final frozen SDK-on diagnostics had 281272 completed late waits, only
than the capture; those counters are not its sleep-residency measurement.
All final samples had the same valid debugger/clock state described below.
The never-DTR client successfully issued `get powersaving` and held the same
file descriptor open across both captures. Early clearing and removal of the
In the controlled old/new USB-poll pair, the never-DTR client successfully
issued `get powersaving` and held the same file descriptor open across both
captures. Early clearing and removal of the
late state SVC were held constant in that pair, isolating the USB poll change.
A separate unsafe late-double-wait
control also slept, but was rejected because of the lost-work race above.
@@ -144,6 +145,22 @@ In particular the T096 FEM/LNA default differs from EasySkyMesh; this RAK
cannot measure that board-specific load. No RF delivery or long battery-life
claim follows from these captures.
## Sensor tracker wake timer
A separate tracker-mode issue posted `xTimerChangePeriod()` on every eligible
loop. Waking the higher-priority timer daemon recreates a scheduler event just
before the late wait. The helper now reuses an active timer whose remaining
time is nonzero and no later than the requested wake. Its active/expiry/tick
snapshot is protected by a short FreeRTOS critical section. Dormant, expired
or too-late timers are rearmed outside that section; allocation/queue failure
still keeps the node polling. No late event clearing was added.
Host tests execute the actual helper, including the original repeated-rearm
negative control, deadline changes, one-shot expiration, queue failures, tick
wrap and late IRQs. T1000e and RAK3401 Sensor builds pass. This timer follow-up
has not been physically qualified in GPS tracker mode; the RAK captures above
are repeater tests.
See [power-saving diagnostics](power_saving_diagnostics.md) for the permanent
on-demand CLI counters. Counters remain off by default. Wait-call counters
cover the late eligible wait, not the guaranteed-nonblocking early SDK