On 9/24/12 10:13 AM, Tom Lane wrote:
> FWIW, what I found out last time I touched this code is that on many
> systems setlocale doesn't bother to return a canonicalized spelling;
> it just gives back the string you gave it. It might be worth doing
> what Peter suggests, just to be consistent with what we are doing
> elsewhere, but I'm not sure how much it will help.
It might not have anything to do with the current problem, but if initdb
canonicalizes locale names, then pg_upgrade also has to. Otherwise,
whenever an operating system changes its locale canonicalization rules,
pg_upgrade will fail.