On Tue, 13 Nov 2018 16:53:07 +0100 Stefano Brivio <sbri...@redhat.com> wrote:
> But to make that effective, you would need to protect the read too, and > that makes your optimisation not really overzealous I think. > > I'd rather go with an additional comment, if that doesn't become > unreadable. Oh, and of course, this whole thing makes sense only if for some reason you end up keeping this function. Maybe ignore my comments on this patch at least for the moment being ;) -- Stefano
