wasi-common publishing error? · wasmtime

Stream: wasmtime

Topic: wasi-common publishing error?

Peter Huene (Jul 09 2020 at 01:54):

It appears https://github.com/bytecodealliance/wasmtime/commit/f47133b229f420d82fa0a2abe06589d172049ec6 is tagged with v0.18.0, but the changes to wasi-common don't appear in the 0.18.0 crate on crates.io. @Alex Crichton any ideas?

Allow different Handles to act as stdio (#1600) · bytecodealliance/wasmtime@f47133b

* Allow any type which implements Handle to act as stdio There have been requests to allow more than just raw OS handles to act as stdio in `wasi-common`. This commit makes this possible by lo...

Peter Huene (Jul 09 2020 at 01:54):

The krustlet folks noticed it when trying to update to 0.18.0 to get the WASI proc_exit trap fix (see https://github.com/deislabs/krustlet/pull/319)

Fix kubelet exit on module failure by itowlson · Pull Request #319 · deislabs/krustlet

There is an issue where if a WASM module fails (for example, with Rust source, if main returns an Err), it takes down the entire krustlet-wasi process. This fixes that issue: now, if a module fail...

Alex Crichton (Jul 09 2020 at 14:18):

@Peter Huene hm IIRC the 0.18.0 tag I pushed manually a few weeks after the crates.io publish b/c I think it was accidentally forgotten, but maybe I tagged the wrong commit?

Alex Crichton (Jul 09 2020 at 14:19):

I tagged what I thought was the published commit but the disconnect may mean that the tag and crates.io differ

Peter Huene (Jul 09 2020 at 16:49):

Given the PR was merged on June 9th and the crate was published on June 11th, I'd expect the changes to be there unless the crate wasn't published from HEAD

Peter Huene (Jul 09 2020 at 16:55):

Erm correction. I was looking at wasmtime's publish date. wasi-common was the 9th, so that likely explains why it's missing for that crate and we published (and tagged) wasmtime's off a later commit?

Peter Huene (Jul 09 2020 at 16:56):

Weird that the publish dates were different

Last updated: Nov 22 2024 at 16:03 UTC