https://gcc.gnu.org/bugzilla/show_bug.cgi?id=65752
--- Comment #1 from joseph at codesourcery dot com <joseph at codesourcery dot com> --- On Mon, 13 Apr 2015, gcc at robbertkrebbers dot nl wrote: > NB 1: I do not think that DR #260 applies here Why not? It seems clear enough that optimizations based on pointer origins are intended to be allowed; that it's not allowed to produce a pointer based on another pointer out of thin air, because the optimizations are more useful in practice than code such as this.
