Wasmtime GitHub notifications bot (Feb 22 2026 at 11:27): Wasmtime GitHub notifications bot (Feb 22 2026 at 11:27):tschneidereit requested wasmtime-wasi-reviewers for a review on PR #12635.
Wasmtime GitHub notifications bot (Feb 22 2026 at 11:27):tschneidereit opened PR #12635 from tschneidereit:204-handling to bytecodealliance:main:
Per spec, an HTTP server isn't allowed to send a body for these status code. Nothing's stopping it from doing so nevertheless, though. This is admittedly very niche and might only actually happen in test suites. It does happen in those, leading to flaky results because whether we report an error here or not depends on whether the body is contained in the same packet as the headers or not.
I think it'd also be okay to instead make these cases always error, though for all I know it's possible that there are niche compatibility concerns with that.
I'm not sure how to add a test to this, because it requires an external server and is flaky, but I put together a repro in a gist. Note that that uses the wasip3 crate, which targets
0.3.0-rc-2026-01-06, so this won't work as-is with Wasmtime'smain.
Wasmtime GitHub notifications bot (Feb 23 2026 at 18:51):alexcrichton commented on PR #12635:
This is subtle enough that I think we'll probably want a test for this. Right now all wasi-http tests have access to a server running this in the background, so could that be updated with a few more paths, such as something upstream returning a 204 with a body, and that be used as part of the test?
Last updated: Feb 24 2026 at 04:36 UTC