[PERF] stock: improve stock.move name_get memory usage

In the context of an export with a lot of records containing a 'stock.move' field (ex: stock.valuation.layer.stock_move_id):
 - Only the picking_id.origin, product_id.code and location(_dest)_id.name are needed.
 - Most of the data of picking_id, product_id and location(_dest)_id will be fetch and put in cache.
This can result in a memory error as multiple thousands of records values (at the very least) will be stored in the cache.

Customer cases:
STOCK.MOVE: 350_000
BEFORE:     980 MB
AFTER:      105 MB

OPW-3911127

closes odoo#165059

Signed-off-by: Arnold Moyaux (arm) <arm@odoo.com>

Author: DavidFesquet

addons/stock/models/stock_move.py

@@ -572,7 +572,7 @@ def default_get(self, fields_list):
 
     def name_get(self):
         res = []
-        for move in self:
+        for move in self.with_context(prefetch_fields=False):
             res.append((move.id, '%s%s%s>%s' % (
                 move.picking_id.origin and '%s/' % move.picking_id.origin or '',
                 move.product_id.code and '%s: ' % move.product_id.code or '',