You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
This repository has been archived by the owner on Nov 3, 2021. It is now read-only.
We decided on set_memory/memset/mem.fill, but IMO this operation makes less sense when you look at the table equivalent, table.fill, since we currently have no good way to provide an immediate value.
Is it better to start with mem.clear and table.clear first, and add mem.fill later if desired?
The text was updated successfully, but these errors were encountered:
I kind of expect that we gonna have table.get/set in a similar timeframe, at which point table.fill would already work. So it may not be worth bothering with adding clear methods.
OK, sounds good. For now, I'll move forward with this proposal with mem.fill but without table.fill and will update later when we have something concrete for table.{get,set}.
With the reftypes proposal, table.clear == table.fill with ref.null even for table-of-funcref, and there is (as Andreas says) also table.get and table.set to complement that, for table-of-anyref. If it becomes hard to spec table.fill here let's move it into reftypes.
This topic came up a while ago, and again later.
We decided on
set_memory
/memset
/mem.fill
, but IMO this operation makes less sense when you look at the table equivalent,table.fill
, since we currently have no good way to provide an immediate value.Is it better to start with
mem.clear
andtable.clear
first, and addmem.fill
later if desired?The text was updated successfully, but these errors were encountered: