Re: Code cleanup for detoast a expanded datum.

Michel Pelletier writes: > On Mon, Nov 18, 2024 at 7:42 PM Andy Fan wrote: > > Andy Fan writes: > > > > > make check-world passed after applying this patch. > > v2 changes the places of Assert, which is missed in v1 by mistakes. > > I'm not an expert in this end of the code but it looks cor

Re: Code cleanup for detoast a expanded datum.

On Mon, Nov 18, 2024 at 7:42 PM Andy Fan wrote: > Andy Fan writes: > > > > > make check-world passed after applying this patch. > > v2 changes the places of Assert, which is missed in v1 by mistakes. > I'm not an expert in this end of the code but it looks correct to me, my only comment would b

Re: Code cleanup for detoast a expanded datum.

Andy Fan writes: > > make check-world passed after applying this patch. v2 changes the places of Assert, which is missed in v1 by mistakes. -- Best Regards Andy Fan >From 0bbd242e034b8461e40bb17aec4fa354f5c97815 Mon Sep 17 00:00:00 2001 From: Andy Fan Date: Tue, 29 Oct 2024 14:05:05 +0800 Su

Code cleanup for detoast a expanded datum.

Hi, When detoasting an expanded datum in detoast_attr, we already know it is an expanded datum, but we use deotast_external_attr function which checks VARATT_IS_EXTERNAL_ONDISK & VARATT_IS_EXTERNAL_INDIRECT again which I think it is impossible to true in this case. so I think we should use a more