lock SendData to fix random connection failures #1623
Merged
Conversation
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
mus65
requested review from
Rob-Hague,
WojciechNagorski and
drieseng
as code owners
|
Nice! It makes sense: the SFTP layer has its own length-prefixed packets so when they are being (wrongly) split up in the channel layer, the server reads an SFTP packet, tries to read the length of the next SFTP packet but ends up reading some random bytes in the middle of a different packet. Good find |
Rob-Hague
approved these changes
Mar 29, 2025
This was referenced Mar 29, 2025
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Outdated suggestions cannot be applied.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.
Fixes #1253 and probably #1493 .
This adds a lock to SendData to avoid the loop to interleave between multiple messages, causing a "bad message" error in the server (see #1253 (comment)).
This may totally be the wrong fix since I'm not really familiar with the protocol and I couldn't find anything specifically about this in the RFCs. But it seems sensible to me that the server doesn't expect this to happen.
By adding some logging I confirmed that in the failure cases of the test, the loop did indeed interleave between different messages. In the success cases, there simply were never more than 1 iteration, so it couldn't interleave.
Without this fix, I had the test fail about every third run. With this fix, I have run the test for about 500 times without a failure.