Closed fitzgen closed 4 months ago
Should we also have non-nullable variants of these tests?
There are a couple non-null types in there. Indeed, we avoid generating null-checks for GC barriers when we have non-null types. I think in an ideal world, yes, we would duplicate ~everything once for nullable and once more for non-nullable types. Don't really want to do that by hand though. Maybe if we were generating tests programmatically.
One nit: Can you match the folded style of the rest of this test?
Done!
(I don't have write access and can't merge this PR myself)
Thanks!
I noticed that Wasmtime was passing the spec tests despite having bits that were known to be unimplemented. This should help the tests exercise those corners of the spec.
Notably:
Setting
i31ref
globals.Initializing tables and globals with
(ref.i31 (global.get $g))
.Table operations on
i31ref
tables.Accessing
anyref
globals and tables that are actuallyi31ref
s. This is interesting to exercise outside of general subtyping because we have different paths in our inline GC barriers foranyref
s that are actuallyi31ref
s.@rossberg PTAL