Re: [COMMITTERS] pgsql: Add PF_PRINTF_ATTRIBUTE to on_exit_msg_fmt. - Mailing list pgsql-hackers

From Heikki Linnakangas
Subject Re: [COMMITTERS] pgsql: Add PF_PRINTF_ATTRIBUTE to on_exit_msg_fmt.
Date
Msg-id 51515396.3090302@vmware.com
Whole thread Raw
In response to Re: [COMMITTERS] pgsql: Add PF_PRINTF_ATTRIBUTE to on_exit_msg_fmt.  (Tom Lane <tgl@sss.pgh.pa.us>)
Responses Re: [COMMITTERS] pgsql: Add PF_PRINTF_ATTRIBUTE to on_exit_msg_fmt.
List pgsql-hackers
On 26.03.2013 02:02, Tom Lane wrote:
> Heikki Linnakangas<hlinnakangas@vmware.com>  writes:
>> On 25.03.2013 15:36, Tom Lane wrote:
>>> Heikki Linnakangas<heikki.linnakangas@iki.fi>   writes:
>>>> Add PF_PRINTF_ATTRIBUTE to on_exit_msg_fmt.
>>>> Per warning from -Wmissing-format-attribute.
>
>>> Hm, this is exactly what I removed yesterday, because it makes the build
>>> fail outright on old gcc:
>
>> The attached seems to work. With this patch, on_exit_msg_func() is gone.
>> There's a different implementation of exit_horribly for pg_dumpall and
>> pg_dump/restore. In pg_dumpall, it just calls vwrite_msg(). In
>> pg_dump/restore's version, the logic from parallel_exit_msg_func() is
>> moved directly to exit_horribly().
>
> Seems probably reasonable, though if we're taking exit_horribly out of
> dumputils.c, meseems it ought not be declared in dumputils.h anymore.
> Can we put that declaration someplace else, rather than commenting it
> with an apology?

Ugh, the patch I posted doesn't actually work, because dumputils.c is 
also used in psql and some scripts, so you get a linker error in those. 
psql and scripts don't use exit_horribly or many of the other functions 
in dumputils.c, so I think we should split dumputils.c into two parts 
anyway. fmtId and the other functions that are used by psql in one file, 
and the functions that are only shared between pg_dumpall and pg_dump in 
another. Then there's also functions that are used by pg_dump and 
pg_restore, but not pg_dumpall or psql.

I'll try moving things around a bit...

- Heikki



pgsql-hackers by date:

Previous
From: Amit Kapila
Date:
Subject: Ideas for improving Concurrency Tests
Next
From: Albe Laurenz
Date:
Subject: Re: adding support for zero-attribute unique/etc keys