Re: Backpatch b61d161c14 (Introduce vacuum errcontext ...) - Mailing list pgsql-hackers

From Justin Pryzby
Subject Re: Backpatch b61d161c14 (Introduce vacuum errcontext ...)
Date
Msg-id 20200623135354.GD4107@telsasoft.com
Whole thread Raw
In response to Re: Backpatch b61d161c14 (Introduce vacuum errcontext ...)  (Tom Lane <tgl@sss.pgh.pa.us>)
List pgsql-hackers
On Tue, Jun 23, 2020 at 12:14:40AM -0400, Tom Lane wrote:
> Andres Freund <andres@anarazel.de> writes:
> > I am only suggesting that where you save the old location, as currently
> > done with LVRelStats olderrinfo, you instead use a more specific
> > type. Not that you should pass that anywhere (except for
> > update_vacuum_error_info).
> 
> As things currently stand, I don't think we need another struct
> type at all.  ISTM we should hard-wire the handling of indname
> as I suggested above.  Then there are only two fields to be dealt
> with, and we could just as well save them in simple local variables.
> 
> If there's a clear future path to needing to save/restore more
> fields, then maybe another struct type would be useful ... but
> right now the struct type declaration itself would take more
> lines of code than it would save.

Updated patches for consideration.  I left the "struct" patch there to show
what it'd look like.

-- 
Justin

Attachment

pgsql-hackers by date:

Previous
From: Masahiko Sawada
Date:
Subject: Re: Internal key management system
Next
From: Tom Lane
Date:
Subject: Re: Building postgresql with higher major version of separate libpq package