Horton

Re: Fixing MSVC's inability to detect elog(ERROR) does not return

Peter Eisentraut <peter@eisentraut.org>

View thread View on postgr.es
From: Peter Eisentraut <peter@eisentraut.org>
To: David Rowley <dgrowleyml@gmail.com>
Cc: PostgreSQL Developers <pgsql-hackers@lists.postgresql.org>
Date: 2025-09-30T05:24:49Z
Lists: pgsql-hackers

Attachments

On 27.09.25 12:43, David Rowley wrote:
> On Wed, 17 Sept 2025 at 15:52, David Rowley <dgrowleyml@gmail.com> wrote:
>> Updated patch attached.  Thanks for the review.
> 
> Now pushed and awaiting buildfarm feedback.

Cool, seems to work.  I also tried it on CI by removing a few "silence 
compiler warning" lines.

Quick follow-up: How about we rename pg_builtin_integer_constant_p to 
pg_integer_constant_p, since it's not actually built-in?  See attached 
patch.

Commits

  1. Rename pg_builtin_integer_constant_p to pg_integer_constant_p

  2. Teach MSVC that elog/ereport ERROR doesn't return