Fix uncatchable IO exceptions. - #1137
Closed
mefyl wants to merge 6 commits into
Closed
Conversation
One test would change the global client's caching policy, impacting
following tests.
RFC 7230 §3.3.1 forbids Transfer-Encoding in 1xx, 204, and 304 responses. The previous cohttp 6.x change that defaulted Unknown encoding to Chunked did not account for these response types, causing e.g. WebSocket 101 Switching Protocols responses to include a spurious Transfer-Encoding: chunked header and a trailing chunked body terminator (0\r\n\r\n), breaking the WebSocket handshake for conforming clients.
…ncoding-v2 fix: don't add Transfer-Encoding to responses that cannot have a body
Contributor
Author
|
Superseded by #1140 |
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
There is currently a bug where IO write exception are "uncatchable": even if the underlying
IO_erroris handled, it would still find its way out and kill the main loop as demonstrated in this test.Bamboozled by the backtrace, I initially reported this in lwt and lived with it for some time. I finally had enough and jumped into the rabbit hole, copiously hitting my head every meter during the fall, but finally hunted it down to this Lwt.dont_wait. Cruel irony, I already addressed this very issue but was too shy. This is the harsher version, which I think is the correct one:
resshould be reported twice.Connection.close, but I don't think any exception there should kill the Lwt loop either.