Andreas Rossberg
Andreas Rossberg
I am not aware of anything similar, unfortunately, but I may simply not know about it. For that reason alone I'd prefer if we didn't have to go there. Merely...
I agree that the embedder interface should have lookup functions for each sort of address. Should be straightforward to add, would anybody volunteer? :)
@penzn, under NaN-boxing, proper float NaNs must be canonical, otherwise they would be confused with pointers, e.g., by a GC. Hence, you want all float ops to produce canonical NaNs...
@KloudKoder: > the Release 2.0 draft of the WebAssembly Specification says in one place: > > "There is no undefined behavior, i.e., the execution rules cover all possible cases that...
@dbuenzli, yes, but that didn't quite fit my use case, which was already using Dynarray.to_array before. I'm okay with a copy, just don't wanna do two (or fall back to...
To be honest, I'm a bit confused what the use case for unsafe_to_iarray is. If I already know the number of elements, then surely I can directly use Iarray.init instead....
> guess a simple example is implementing: `sort` Well, for that, wouldn't a non-copying unsafe_array_to_iarray suffice? I don't see where the dynamic array size comes in.
Good point! Did you produce some that you could contribute?
Closing via #25 / upstream PR.
Thanks for the report. I was able to narrow the regression down to a bugfix in the substitution function. Previously, it cut off at deftypes, assuming they are closed, but...