Fix inplace upgrade config updates (#9116)

IPUs currently sometimes break if new config fields have been added in
both the base schema and extension schemas. This is because the IPU
path calls get_existing_view_columns to make sure that columns are
ordered correctly, but get_existing_view_columns does not restrict the
query to only inspecting tables for the *current* version, and so gets
confused when both exist during the IPU.

Author: msullivan

edb/server/bootstrap.py

@@ -716,7 +716,8 @@ async def get_existing_view_columns(
     # Find all the config views (they are pg_classes where
     # there is also a table with the same name but "_dummy"
     # at the end) and collect all their columns in order.
-    return json.loads(await conn.sql_fetch_val('''\
+    schema = pg_common.versioned_schema("edgedbstd")
+    return json.loads(await conn.sql_fetch_val(f'''\
         select json_object_agg(v.relname, (
             select json_agg(a.attname order by a.attnum)
             from pg_catalog.pg_attribute as a
@@ -725,6 +726,10 @@ async def get_existing_view_columns(
         from pg_catalog.pg_class as v
         inner join pg_catalog.pg_tables as t
         on v.relname || '_dummy' = t.tablename
+        -- Filter for just our namespace!
+        inner join pg_catalog.pg_namespace as ns
+        on v.relnamespace = ns.oid
+        where ns.nspname = '{schema}' OR ns.nspname = 'edgedbpub'
 
     '''.encode('utf-8')))