I meant the developer needs to add the locks, not the compiler. I think they're just advisory but everyone always rights well-behaved code, right? ;-)
On Mon, Oct 23, 2017 at 4:17 AM, Geoff Winkless <pgsqladmin@geoff.dj> wrote:
I wrote: > I assume that's because it can be certain at compile time that, since > both buff and buff2 are local, nothing else is going to modify the > source string (without some stack smashing, anyway).
Ugh. Of course, they're not local: they're malloc'd; however it's still reasonable to assume that (since the pointers aren't passed anywhere) the buffers are untouched elsewhere, unless other threads are misbehaving. Geoff