[code-reflection] RFR: Computes the VarOp's result type from its operand type, if not overridden

Paul Sandoz psandoz at openjdk.org
Tue Jun 18 16:42:24 UTC 2024


On Tue, 18 Jun 2024 10:57:19 GMT, Maurizio Cimadamore <mcimadamore at openjdk.org> wrote:

> 
> I wonder if, like in the source code, we need a notion of `Var` without an explicit type (e.g. make "overridability of type" a property in the model).

The `CoreOp.var` factory factories are set up like that, and we could encode the "fixed result type" explicitly in the model if the current approach is too fragile. Maybe we should even think about exposing the "result type derived from operands" or "result type explicit" as some more generic operation property.

-------------

PR Comment: https://git.openjdk.org/babylon/pull/148#issuecomment-2176534289


More information about the babylon-dev mailing list