MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII;
 format=flowed
Content-Transfer-Encoding: 7bit
Date: Fri, 10 Aug 2018 17:59:29 +0300
From: Marina Polyakova <m.polyakova@postgrespro.ru>
To: Arthur Zakirov <a.zakirov@postgrespro.ru>
Cc: Fabien COELHO <coelho@cri.ensmp.fr>, pgsql-hackers@postgresql.org,
 Alvaro Herrera <alvherre@2ndquadrant.com>, Ildus Kurbangaliev
 <i.kurbangaliev@postgrespro.ru>, Teodor Sigaev <teodor@sigaev.ru>
Subject: Re: [HACKERS] WIP Patch: Pgbench Serialization and deadlock errors
Organization: Postgres Professional
In-Reply-To: <20180810141918.GA5450@zakirov.localdomain>
References: <alpine.DEB.2.21.1807111435250.27883@lancre>
 <976a25a46420c93866ada4051976b1be@postgrespro.ru>
 <alpine.DEB.2.21.1807112124210.27883@lancre>
 <f153ffd55f481d714ee44afc99ecc67e@postgrespro.ru>
 <alpine.DEB.2.21.1807121801250.27883@lancre>
 <25d7b1f518c218a8c335bdfc6c858ea1@postgrespro.ru>
 <alpine.DEB.2.21.1808071823540.13466@lancre>
 <fcc2512cdc9e6bc49d3b489181f454da@postgrespro.ru>
 <20180810125327.GA2374@zakirov.localdomain>
 <177cba37dface8918a58fda2188fa1ce@postgrespro.ru>
 <20180810141918.GA5450@zakirov.localdomain>
Message-ID: <fe86b444c131176ea0e32f179747abe1@postgrespro.ru>
User-Agent: Roundcube Webmail/1.2.4
Precedence: bulk

On 10-08-2018 17:19, Arthur Zakirov wrote:
> On Fri, Aug 10, 2018 at 04:46:04PM +0300, Marina Polyakova wrote:
>> > +1 from me to keep initial name "pgbench_error". "pgbench_log" for new
>> > function looks nice to me. I think it is better than just "log",
>> > because "log" may conflict with natural logarithmic function (see "man 3
>> > log").
>> 
>> Do you think that pgbench_log (or another whose name speaks only about
>> logging) will look good, for example, with FATAL? Because this means 
>> that
>> the logging function also processes errors and calls exit(1) if 
>> necessary..
> 
> Yes, why not. "_log" just means that you want to log some message with
> the specified log level. Moreover those messages sometimes aren't 
> error:
> 
> pgbench_error(LOG, "starting vacuum...");

"pgbench_log" is already used as the default filename prefix for 
transaction logging.

>> > I agree with Fabien. Calling pgbench_error() inside pgbench_error()
>> > could be dangerous. I think "fmt" checking could be removed, or we may
>> > use Assert()
>> 
>> I would like not to use Assert in this case because IIUC they are 
>> mostly
>> used for testing.
> 
> I'd vote to remove this check at all. I don't see any place where it is
> possible to call pgbench_error() passing empty "fmt".

pgbench_error(..., "%s", PQerrorMessage(con)); ?

-- 
Marina Polyakova
Postgres Professional: http://www.postgrespro.com
The Russian Postgres Company