On Feb 22, 2013, at 9:08 AM, Jordan Rose <[email protected]> wrote:
> My personal opinion on this is that we should actually remove the Invalid > state from CFGStmt. It seems dangerous for exactly the reasons Ted has > outlined. Agreed. If one has a CFGStmt it should be usable. Period. If there is an optimization trick we want to do to make Optional<CFGStmt> more efficient by stealing a bit from CFGStmt that's a different story, but the question of whether a CFGStmt is "valid" should not be something that can be queried by a client.
_______________________________________________ cfe-commits mailing list [email protected] http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits
