github-actions[bot] commented on Issue #1996:
Subscribe to Label Action
cc @peterhuene
<details>
This issue or pull request has been labeled: "wasmtime:c-api"Thus the following users have been cc'd because of the following labels:
- peterhuene: wasmtime:c-api
To subscribe or unsubscribe from this label, edit the <code>.github/subscribe-to-label.json</code> configuration file.
Learn more.
</details>
fitzgen commented on Issue #1996:
I think I am going to go only one step towards the "raw"
wasm_val_t
andwasm_ref_t
representation you're describing: removing the doubleOption
bits, but not getting rid ofWasmRefInner
and its type tagging. This way,wasm_ref_copy
andwasm_ref_delete
can still be safely implemented, which I think is valuable, even if a bunch of the otherwasm_ref_*
stuff is not worthwhile.
alexcrichton commented on Issue #1996:
Hm ok that can work, but trying to think through this, presumably the reason is that
wasm_table_*
take/returnwasm_ref_t
, right? Forwasm_table_set
I think we should not take ownership anyway, so it's less of an issue there, but forwasm_table_get
you'd need some way to delete the reference that you acquire. Without tagging we don't know, in isolation, what to delete there.Given that it seems that to support that API we'd need to have internal tagging. I'll raise an issue on the wasm-c-api repo as well.
Last updated: Nov 22 2024 at 17:03 UTC