On Thu, Apr 14, 2016 at 3:45 PM, Jakub Jelinek <ja...@redhat.com> wrote: > On Thu, Apr 14, 2016 at 09:43:26AM -0400, Jason Merrill wrote: >> On 04/14/2016 09:25 AM, Nathan Sidwell wrote: >> >On 04/13/16 15:41, Jason Merrill wrote: >> > >> >>The fini_constexpr stuff is OK immediately. >> > >> >As those two objects are currently GTY((deletable)) I don't think that's >> >necessary. Have I missed something? >> >> True, I suppose it doesn't make much difference. > > Right now our debugging shows that the remaining issue is just an unrelated > ipa-devirt bug (where the aggressive pruning of BLOCK trees with -g0 results > in different devirt decisions from non-pruned one). > > So we hopefully can get away with deletable.
Seems not so - see PR70675. I'd make the tables non-deletable but flush them after parsing finished (I don't see them deleted anywhere right now). Richard. > Jakub
