--- Comment #10 from Thomas Preud'homme <thopre01 at gcc dot> ---
(In reply to Jakub Jelinek from comment #9)
> Created attachment 43384 [details]
> gcc8-pr84272.patch
> So like this?  A few additional changes, Florian Weimer suggested using
> preincrement instead of postincrement with the C++ iterators, and I hope
> there are no pointers in between different forest, so it should be fine to
> free already when processing of the forest is done, rather than waiting
> until all forests are processed.
> Could somebody test it on aarch64-linux please?

Yes I think it should be fine. Sorry, got to head back home again but Kyrill
said he would be testing it.

Reply via email to