Fix race condition in IOUringEventLoop that prevents shutdown #227
Conversation
This file contains 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
Motivation: Short lived IOUringEventLoop's can fail to shutdown, being stuck in the read-completion phase. This is because under race conditions we can get a task that skips the submission of the eventFd and then exits. Modifications: - After setting the initial eventfd read event we flush it. This is a one time cost per-loop and avoids the case where a wakeup was requested but never received because the loop enqueued but didn't submit the interest in receiving those event. Result: Less stalled threads on shutdown.
|
For details about the debugging done to find this see #226 |
This was referenced Nov 15, 2023
bryce-anderson
commented
Nov 15, 2023
transport-classes-io_uring/src/main/java/io/netty/incubator/channel/uring/IOUringEventLoop.java
Outdated
Show resolved
Hide resolved
chrisvest
reviewed
Nov 15, 2023
...ort-native-io_uring/src/test/java/io/netty/incubator/channel/uring/IOUringEventLoopTest.java
Outdated
Show resolved
Hide resolved
...ort-native-io_uring/src/test/java/io/netty/incubator/channel/uring/IOUringEventLoopTest.java
Show resolved
Hide resolved
chrisvest
reviewed
Nov 15, 2023
...ort-native-io_uring/src/test/java/io/netty/incubator/channel/uring/IOUringEventLoopTest.java
Show resolved
Hide resolved
transport-classes-io_uring/src/main/java/io/netty/incubator/channel/uring/IOUringEventLoop.java
Outdated
Show resolved
Hide resolved
chrisvest
approved these changes
Nov 15, 2023
|
Discussed with @chrisvest offline, and presuming this is good to go, we should port the test to Netty5. |
transport-classes-io_uring/src/main/java/io/netty/incubator/channel/uring/IOUringEventLoop.java
Outdated
Show resolved
Hide resolved
normanmaurer
approved these changes
Nov 16, 2023
|
Thanks a lot for fixing it! |
|
@bryce-anderson can you also open a PR for netty 5 ? |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
Motivation:
Short lived IOUringEventLoop's can fail to shutdown, being stuck in the read-completion phase. This is because under race conditions we can get a task that skips the submission of the eventFd and then exits.
Modifications:
Result:
Less stalled threads on shutdown.