brianhansonxyz/clearpilot
1
0
Fork 0
Code Issues Pull Requests Actions 54 Packages Projects Releases Wiki Activity

controlsd+calibrationd: suppress commIssue from valid=False cascade

Browse Source 
Two fixes that reinstate the pre-revert defenses against the "TAKE CONTROL
IMMEDIATELY / Communication Issue" banner that fires when self-driving
on the baseline modelrevert stack:

calibrationd: publish valid based on calStatus == calibrated, not
sm.all_checks(). Original gate cascaded upstream freq glitches into
liveCalibration.valid=False, which kept locationd.filterInitialized
False, which fed garbage into paramsd, which corrupted steerRatio
(erratic steering). "valid" here is a question about convergence, not
input freshness.

controlsd: narrow the commIssue trigger to genuine comm failures —
not_alive OR can_rcv_timeout. The `not sm.all_checks()` branch also
picked up valid=False, but paramsd / torqued / plannerd / frogpilot_planner
/ dmonitoringd all propagate their sm.all_checks() into msg.valid via
a polling-pattern artifact (freq_ok inside poll='...' subscribers
tracks gaps between drain bursts rather than the publish rate), so
the whole stack flaps valid and trips the banner during normal
driving. Content and rate are fine; just the flag.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
This commit is contained in:
Brian Hanson
2026-04-23 11:00:45 -05:00
parent f896dfe25a
commit dc7e0a2db7
2 changed files with 16 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files
selfdrive/controls/controlsd.py
+8 -5
View File
@@ -485,13 +485,16 @@ class Controls:
# generic catch-all. ideally, a more specific event should be added above instead
has_disable_events = self.events.contains(ET.NO_ENTRY) and (self.events.contains(ET.SOFT_DISABLE) or self.events.contains(ET.IMMEDIATE_DISABLE))
no_system_errors = (not has_disable_events) or (len(self.events) == num_events)
if (not self.sm.all_checks() or self.card.can_rcv_timeout) and no_system_errors and not model_suppress:
# CLEARPILOT: fire commIssue ONLY when messages actually aren't flowing (not_alive)
# or CAN RX is timing out. Don't fire on self-declared valid=False — that's the
# polling-pattern / all_checks cascade that paramsd/torqued/plannerd/frogpilot
# propagate even while their publish rate and content are fine.
comms_really_broken = (not self.sm.all_alive()) or self.card.can_rcv_timeout
if comms_really_broken and no_system_errors and not model_suppress:
if not self.sm.all_alive():
self.events.add(EventName.commIssue)
elif not self.sm.all_freq_ok():
self.events.add(EventName.commIssueAvgFreq)
else: # invalid or can_rcv_timeout.
self.events.add(EventName.commIssue)
else:
self.events.add(EventName.commIssue) # can_rcv_timeout path
logs = {
'invalid': [s for s, valid in self.sm.valid.items() if not valid],
selfdrive/locationd/calibrationd.py
+8 -1
View File
@@ -284,7 +284,14 @@ def main() -> NoReturn:
# 4Hz driven by cameraOdometry
if sm.frame % 5 == 0:
calibrator.send_data(pm, sm.all_checks())
# CLEARPILOT: publish valid based on calibration status, not upstream sm.all_checks().
# The original gate cascaded upstream freq glitches into liveCalibration.valid=False,
# which kept locationd.filterInitialized False, which fed garbage into paramsd, which
# corrupted steerRatio and caused erratic steering (and controlsd commIssue banners).
# "valid" here semantically means "the calibration data is trustworthy" — a question
# about convergence, not input freshness.
cal_valid = calibrator.cal_status == log.LiveCalibrationData.Status.calibrated
calibrator.send_data(pm, cal_valid)
if __name__ == "__main__":