Repository navigation
Report the nodes after their friends, on the hot-plug worker - #69
Merged
Merged
Conversation
A node was probed for its classes as soon as its path was known - before
its friends_group was broadcast, so a driver that answers the probe
through a friend could not: dmwm8994 answers DMDRVI_IOCTL_AUDIO_GET_INFO
over its I2C bus friend, failed, and the codec was never reported as
"audio" (no dmaudio@ service for it).
The nodes are now reported after the friends are broadcast. The nodes a
mount has when it becomes ready are reported by the hot-plug worker
(process_mount_ready()), not in dmfsi_dmdevfs_mounted(): dmod-boot mounts
/dev before the scheduler runs, and a probe that now talks to the hardware
waits - which there corrupted the context ("stack overflow" of
dmdevfs_hotplug). Their partitions are scanned there too, after the
report, as before.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
JohnAmadis
approved these changes
Oct 11, 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
2 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
A node was probed for its classes as soon as its path was known, before its
friends_groupwas broadcast. A driver that answers the probe through a friend could not answer. dmwm8994 answersDMDRVI_IOCTL_AUDIO_GET_INFOover its I2C bus friend, so the probe failed and the codec was never reported asaudio. As a result libsystemd started nodmaudio@wm8994for it.Changes:
report_node_ready()) after the friends are broadcast: at mount, for a configuration added to a mounted instance, and for hot-plugged nodes.process_mount_ready()), not indmfsi_dmdevfs_mounted(). dmod-boot mounts/devbefore the scheduler runs. A probe that now talks to the hardware waits for the I2C transfer, and waiting there corrupted the context: the board stopped with "Stack overflow detected in task: dmdevfs_hotplug". The partitions are scanned on the worker as well, after the report. A node a hot-plug event reported earlier is skipped.Testing
codec-with-touch.ini): the codec is reported asaudioanddmaudio@wm8994starts from its rule (choco-technologies/dmaudio, libdmaudio). The SD card is still automounted, and the display, touch, network and console work. The peak stack use ofdmdevfs_hotplugis unchanged (2468/4608).🤖 Generated with Claude Code