Date: Fri, 13 Sep 2019 16:31:32 -0300
From: Alvaro Herrera <alvherre@2ndquadrant.com>
To: Tom Lane <tgl@sss.pgh.pa.us>
Cc: pgsql-committers@lists.postgresql.org
Subject: Re: pgsql: Fix progress reporting of CLUSTER / VACUUM FULL
Message-ID: <20190913193132.GA18625@alvherre.pgsql>
MIME-Version: 1.0
Content-Type: text/plain; charset=iso-8859-1
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
In-Reply-To: <19795.1568400476@sss.pgh.pa.us>
User-Agent: Mutt/1.9.4 (2018-02-28)
Precedence: bulk

On 2019-Sep-13, Tom Lane wrote:

> Not a new problem of this patch, exactly, but:
>  
>  /* Reindex options */
>  #define REINDEXOPT_VERBOSE 1 << 0  /* print progress info */
> +#define REINDEXOPT_REPORT_PROGRESS 1 << 1  /* report pgstat progress */
> 
> Surely these macro definitions are incredibly dangerous due to their
> lack of parentheses.
> 
> I'd initially thought that we already had bugs in existing usages like
> 
> 	if (options & REINDEXOPT_VERBOSE)
> 
> After consulting a nearby C reference I see that we're accidentally
> saved by << having higher priority than &, but this is not safely-
> maintainable code.  Expressions like "~REINDEXOPT_VERBOSE" would not
> do what one expects.

Fixed back to 9.5, where this macro appeared.  I was unable to come up
with a way to search for other occurrences of the same problem :-(

Thanks,

-- 
Álvaro Herrera                https://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services