> On Oct 27, 2014, at 16:32, jahanian <[email protected]> wrote: > > >> On Oct 27, 2014, at 4:22 PM, Jordan Rose <[email protected]> wrote: >> >> What does maybeUndoReclaimObject do? Does that mean we previously treated >> this as a bridge at +1? > > That is not I’m my patch. It shows up in the diff because of rearrangement. > And there is comment on top of maybeUndoReclaimObject routine. > (You can ask John M. about it). > > - Fariborz
That seems very relevant to the content of this decision, including whether we've been leaking memory on previous OSs because people didn't know whether their CFArrays are +0 or +1. (The analyzer has assumed they're always +0.) Jordan _______________________________________________ cfe-commits mailing list [email protected] http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits
