-d. I know its not much to type that, but i think its appropriate if the generated code also checks for it and skips any incompatible devices.
f32.sgnwith double precision. I'll fix it.
#[unsafe]in front and see if that makes it run properly.
goto, which is probably a mistake. There's no way a GPU can handle non-reducible control flow. But the only use of
gotoin the generated code is to break out of nested loops, which is reducible, so it really ought to work.