Edit: I was asking where witx-bindgen was living, and obviously I thought of doing a github-wide search only just after searching, so... Just found it's in bytecodealliance/wasi
Actually, reusing this topic for an actual question... Is it known that witx-bindgen is broken? I've got a function that takes in (array u8), and witx-bindgen generates a function that takes &'a [u8]
(instead of &mut [u8]
) and does not declare the 'a
lifetime
I just upgraded witx-bindgen to the latest version of witx, which fixes the CI: https://github.com/bytecodealliance/wasi/pull/56
@Pat Hickey or @Alex Crichton, can you take a look?
@Léo Gaspard, for the lifetime issue you bring up, is it the same as https://github.com/bytecodealliance/wasi/issues/47?
@Alex Crichton, that was fast :smile:
thanks, idk why that broke CI, do you mind pointing me to the failure?
id like to make sure when i do a witx release in the future i dont break it, or i fix it myself :)
found https://github.com/bytecodealliance/wasi/pull/55/checks but it doesnt make sense to me :/ version 0.8.7 should have been available on crates.io still...
hm... perhaps it was that the Git submodule was looking at an old version of witx?
it's a local path dependency, not a crates.io dependency
hmm, yeah, so the cargo.toml didnt agree with the path dependency prior to that CI run
im not sure how that got introduced without CI catching it
could have been a lot of things, the submodule hadnt been updated since feb 2020. hopefully this doesnt happen again. thanks!
@Andrew Brown I don't think so, I've submitted issue 57 with a simple reproducer :)
Last updated: Dec 23 2024 at 12:05 UTC