Alvaro Herrera <alvhe...@2ndquadrant.com> writes: >> I can get behind that ... it makes the patch a lot smaller.
> Here it is. LGTM. > I noticed that ECPG's copy was setting errno. I had forgot to do that > in my previous patch, but on second look, malloc failure already sets > it, so doing it again is pointless. Right. regards, tom lane