fix(#111): stage4 I4.1 — check the java child's /dev/input fds, not bwrap#130
Merged
Conversation
The old I4.1 counted /dev/input fds on the bwrap SUPERVISOR and expected
exactly 2 (an event+js pair) -> 4x false 'expected=2 actual=0' fails on
healthy slots in HW-1. Two stale assumptions:
1. the device fds live in the JAVA child (.slots[slot].pid), not the
bwrap supervisor (.slots[slot].bwrap_pid);
2. under the MCSS_RAW_BINDING=1 default only the jsN node is dev-bound
(event+js only under the legacy virtual-mapper), so the count is
mode-dependent, not a fixed 2.
Rewrite: read the ground-truth set of input nodes bwrap actually bound for
the slot from its argv (mode-agnostic), then assert the java process's open
/dev/input fds are non-empty and all within that bound set (no cross-slot
leakage). Every path logs java pid + bound nodes + observed fds so a Deck run
is self-diagnosing. I4.2/I4.3/I4.4 untouched.
Validated: tests/hardware/run_all.sh stage4 on the Deck (HW-2).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Hm7DmagTa2EttH8aHxGji2
This was referenced Jul 22, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Part of the pre-HW-2 test-harness batch. Hardware-gated — validated by
tests/hardware/run_all.sh stage4on the Deck (HW-2), not in CI (tests/hardware/is Deck-only).Problem
HW-1 (2026-07-19) recorded I4.1 failing ×4 (
expected=2 actual=0) on every healthy slot. Two stale assumptions:/dev/inputfds on the bwrap supervisor — but the device fds live in the java child (.slots[slot].pid, not.slots[slot].bwrap_pid);MCSS_RAW_BINDING=1default only thejsNnode is dev-bound, so the count is mode-dependent.Fix
Read the ground-truth set of input nodes bwrap actually bound for the slot from its argv (mode-agnostic — jsN under raw, event+js under legacy, virtual pair under the future #38 proxy), then assert the java process's open
/dev/inputfds are non-empty and all within that bound set (no cross-slot leakage). Every path logs the java pid, bound nodes, and observed fds so a Deck run is self-diagnosing. I4.2/I4.3/I4.4 untouched.Closes #111 (on the green stage4 run).
🤖 Generated with Claude Code
https://claude.ai/code/session_01Hm7DmagTa2EttH8aHxGji2