restore FUSE reconnection after suspend and resume#283
Merged
Conversation
![github-advanced-security[bot]](https://avatars.githubusercontent.com/in/57789?s=60&v=4){kind=small}
🤖 Augment PR SummarySummary: This PR fixes regressions in FUSE suspend/resume reconnects introduced by recent AFP reconnect hardening. Changes:
Technical Notes: Connection generation updates/reads are now atomic helpers, and attention threads release server refs safely after stale-generation checks. 🤖 Was this summary useful? React with 👍 or 👎 |
![augmentcode[bot]](https://avatars.githubusercontent.com/in/1027498?s=60&v=4){kind=small}
The FUSE suspend/resume operations got broken when we hardened AFP session reconnection recently. Keep a per-mount FUSE reconnect credential after the initial authenticated login, since the shared AFP server password cache is scrubbed for stateless session safety and resume commands do not carry a fresh password. Make DSI attention handling generation-aware by copying packets into the worker context and ignoring stale attentions from older connections before they can disconnect a newly resumed socket. Also register FPZzzzz as a blank-reply AFP command and initialize FUSE reconnect error reporting so suspend and resume diagnostics reflect real failures.
Member
Author
|
augment review |
rdmark
changed the title
Install an explicit SIGPIPE handler in afpfsd before any FUSE mount thread can run so pthread_kill(SIGPIPE) cannot terminate the daemon if libfuse has not installed its own handler yet. Keep using SIGPIPE only as the documented libfuse wakeup after fuse_exit(), allowing the FUSE thread's fuse_main() path to own unmount and session cleanup.
Add helper accessors for the AFP server connection generation counter and use them anywhere DSI snapshots or compares the value. This removes the data race between reconnect generation increments and detached attention handling, keeping stale-attention and stale-reply filtering reliable across reconnects.
Member
Author
|
@andylemin with this changeset, it should be easier and more consistent to run the FUSE tests locally prepare the environment as described in test/TESTS.md and invoke it like this I have attempted several times to create a CI job for the FUSE tests, but I'm pretty sure this is where we'd have to create our own self hosted runner to make it happen |
before, any file system operation on a suspended mount would reactivate it, but now only an explicit 'resume' command will cause it to wake up
on macOS we have to call umount directly, and cannot rely on afp_client unmount
|
andylemin
approved these changes
Jul 7, 2026
Contributor
|
Lots of great UX improvements |
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
3 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.
The FUSE suspend/resume operations got broken when we hardened AFP session reconnection recently.
Keep a per-mount FUSE reconnect credential after the initial authenticated login, since the shared AFP server password cache is scrubbed for stateless session safety and resume commands do not carry a fresh password.
Make DSI attention handling generation-aware by copying packets into the worker context and ignoring stale attentions from older connections before they can disconnect a newly resumed socket.
Also register FPZzzzz as a blank-reply AFP command and initialize FUSE reconnect error reporting so suspend and resume diagnostics reflect real failures.
introduce explicit suspended flag for FUSE mounts: before, any file system operation on a suspended mount would reactivate it, but now only an explicit 'resume' command will cause it to wake up
Add FUSE client tests for suspend and resume, while improving the resilience, configurability, and documentation of the FUSE test runner.
Make a number of adjacent improvements and bug fixes: