fix(core): deliver post-boot mountFs reconfiguration reliably to the native sidecar#1660
Merged
Conversation
Member
Author
|
Stack for rivet-dev/agentos
Get stack: |
This was referenced Jul 8, 2026
Merged
NathanFlurry
changed the base branch from
stack/fix-native-sidecar-resolve-js_bridge-mount-root-realpath-without-a-bridge-round-trip-wmqznzol
to
main
…ircuit - Builds on #1640: also short-circuit on the session record already marked closed (covers exits the kill RPC does not report), and skip the second wait when SIGKILL itself reports the process gone. - Mirror the short-circuit in the agentos-sidecar-core sync engine, with regression tests proving zero dead waits.
…bridge round-trip
- Resolve `realpath("/")` of a js_bridge mount locally (mirroring host_dir): host-side drivers often cannot canonicalize their own root and returned ENOENT, which aborted kernel permission-subject resolution before the readDir bridge call was ever issued.
- Regression test: readdir of a js_bridge mount root against a driver whose realpath always fails.
…native sidecar - `mountFs`/`unmountFs` now return a promise that settles once the sidecar has applied the mount; delivery failures reject instead of being swallowed. - Runtime mount reconfigures resend the boot `packages`/`packagesMountAt`/`toolShimCommands`: Rust `configure_vm` rebuilds the whole VM configuration from each payload, so omitting them stripped the `/opt/agentos` projections and tool shims as a side effect of a post-boot mount. - Runtime `linkSoftware` packages are retained too, so a later `mountFs` no longer unprojects them. - The Rust client `mount_fs` remains in-process-only; sidecar-visible runtime mounts there are a follow-up.
NathanFlurry
force-pushed
the
stack/fix-core-deliver-post-boot-mountfs-reconfiguration-reliably-to-the-native-sidecar-nyvvsyxp
branch
from
bda847c to
a46dab5
Compare
NathanFlurry
changed the base branch from
main
to
stack/fix-native-sidecar-resolve-js_bridge-mount-root-realpath-without-a-bridge-round-trip-wmqznzol
NathanFlurry
changed the base branch from
stack/fix-native-sidecar-resolve-js_bridge-mount-root-realpath-without-a-bridge-round-trip-wmqznzol
to
main
NathanFlurry
deleted the
stack/fix-core-deliver-post-boot-mountfs-reconfiguration-reliably-to-the-native-sidecar-nyvvsyxp
branch
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.
mountFs/unmountFsnow return a promise that settles once the sidecar has applied the mount; delivery failures reject instead of being swallowed.packages/packagesMountAt/toolShimCommands: Rustconfigure_vmrebuilds the whole VM configuration from each payload, so omitting them stripped the/opt/agentosprojections and tool shims as a side effect of a post-boot mount.linkSoftwarepackages are retained too, so a latermountFsno longer unprojects them.mount_fsremains in-process-only; sidecar-visible runtime mounts there are a follow-up.