Stream: git-wasmtime

Topic: wasmtime / issue #2897 Implement RFC 11: Redesigning Wasm...

view this post on Zulip Wasmtime GitHub notifications bot (May 18 2021 at 15:04):

alexcrichton commented on issue #2897:

@peterhuene if you're up for it, mind reviewing the bits here ranging from the runtime internals, the wasmtime crate itself, and the C API? If that's too much though let me know! (I see you already assigned yourself as a reviewer but wanted to write something down too)

@pchickey would you be up for reviewing the wasi-common/wiggle/macro/etc bits?

view this post on Zulip Wasmtime GitHub notifications bot (May 18 2021 at 17:08):

pchickey commented on issue #2897:

Yes, I will review everything in that category, and I am also going to port a private repo to use this branch as well.

view this post on Zulip Wasmtime GitHub notifications bot (May 18 2021 at 21:00):

alexcrichton commented on issue #2897:

Ok all documentation should now be included, so I think that this is in theory good to land, pending review/RFC of course.

view this post on Zulip Wasmtime GitHub notifications bot (May 18 2021 at 23:27):

TerrorJack commented on issue #2897:

I reran the haskell bindgen and found the following symbols don't exist:

wasm_frame_copy
wasm_ref_as_trap
wasm_ref_as_trap_const
wasm_foreign_delete
wasm_foreign_copy
wasm_foreign_same
wasm_foreign_get_host_info
wasm_foreign_set_host_info
wasm_foreign_set_host_info_with_finalizer
wasm_foreign_as_ref
wasm_ref_as_foreign
wasm_foreign_as_ref_const
wasm_ref_as_foreign_const
wasm_foreign_new
wasm_ref_as_module
wasm_ref_as_module_const
wasm_ref_as_func
wasm_ref_as_func_const
wasm_ref_as_global
wasm_ref_as_global_const
wasm_ref_as_table
wasm_ref_as_table_const
wasm_ref_as_memory
wasm_ref_as_memory_const
wasm_ref_as_extern
wasm_ref_as_extern_const
wasm_ref_as_instance
wasm_ref_as_instance_const

Some of those were documented as "Unimplemented in Wasmtime, aborts the process if called.", which is incorrect. I recommend a consistent handling of unimplemented C API functions: either don't implement at all and be clear about that in the documentation, or always add a stub which panics.

view this post on Zulip Wasmtime GitHub notifications bot (May 19 2021 at 14:44):

alexcrichton commented on issue #2897:

@TerrorJack oh thanks for the reminder! I've gone through and implemented aborting stubs for those symbols.

view this post on Zulip Wasmtime GitHub notifications bot (May 21 2021 at 15:28):

TerrorJack commented on issue #2897:

We have distinct wasmtime_valkind_t and wasm_valkind_t, but there is no separate wasmtime_func_t as opposed to existing wasm_func_t. So it seems I can't call wasmtime_func_new to create a function which operates on the SIMD types.

view this post on Zulip Wasmtime GitHub notifications bot (May 21 2021 at 15:54):

alexcrichton commented on issue #2897:

@TerrorJack ah good point! I haven't done a ton of testing of the simd proposal in other bindings yet. I added in https://github.com/bytecodealliance/wasmtime/pull/2897/commits/7f4b340d69db55fbbb3bf91aca6669e14087a0d7 to recognize WASMTIME_V128 as a wasm_valkind_t

view this post on Zulip Wasmtime GitHub notifications bot (Jun 03 2021 at 14:09):

alexcrichton commented on issue #2897:

The RFC has been merged and this has been reviewed (thanks all!) so I'm going to merge.

Last updated: Jan 10 2026 at 02:36 UTC