Re: [Emacs-diffs] master 5d4dd55: Fix lifetime error in previous patch

[Paul Eggert]

Stefan Monnier wrote:
> > Shouldn't we count the allocations that happened while GC was
> > inhibited and subtract them from consing_until_gc afterwards?
>
> I don't see why: they are counted already in consing_since_gc, no?

Yes and no. The code keeps subtracting from consing_since_gc while GC is 
inhibited (and consing_since_gc is enormous); but when allow_garbage_collection 
is called, all those subtractions are lost.

I installed the first part of Pip Cet's patch (rewritten slightly to I hope make 
it clearer) - thanks, Pip. The second part I didn't see the need for, as it made 
GC a tiny bit slower and arguably made GC harder to follow.