Re: [PATCH 3/5] 64-bit obstack support, part 1

[Roland McGrath]

It would be easier to review if you separated the very trivial changes from
the others.  The size_t change is clearly fine and can be approved quickly.
The refactoring requires more attention.