The expansion of C.types[x] must be a structure type struct fieldtype*.
For the instructions which checking the types in C.types for function types, maybe have to modify to the similar description?
Take the call_ref x for example:
Let [t1*] -> [t2*] be the function type C.types[x].
Maybe should be modified to:
The expansion of C.types[x] must be a function type func [t1*] -> [t2*].
Indeed, thanks! And I found a few more inconsistencies around the typing of functions, since that changed from functype to a deftype that expands to a functype. Please see #458 for a fix.
For applying the GC proposal, the
C.types
in the validation context becomes the defined types.As the description in
struct.new x
for example:C.types[x]
must be a structure typestruct fieldtype*
.For the instructions which checking the types in
C.types
for function types, maybe have to modify to the similar description? Take thecall_ref x
for example:[t1*] -> [t2*]
be the function typeC.types[x]
.Maybe should be modified to:
C.types[x]
must be a function typefunc [t1*] -> [t2*]
.Thanks.