Horton

Re: Error on pgbench logs

Fabien COELHO <coelho@cri.ensmp.fr>

View thread View on postgr.es
From: Fabien COELHO <coelho@cri.ensmp.fr>
To: Yugo NAGATA <nagata@sraoss.co.jp>
Cc: Michael Paquier <michael@paquier.xyz>, Kyotaro Horiguchi <horikyota.ntt@gmail.com>, rulyox@gmail.com, PostgreSQL Developers <pgsql-hackers@lists.postgresql.org>
Date: 2021-06-12T21:32:54Z
Lists: pgsql-hackers 
> +		while ((next = agg->start_time + agg_interval * INT64CONST(1000000)) <= now)
>
> I can find the similar code to convert "seconds" to "us" using casting like
>
> end_time = threads[0].create_time + (int64) 1000000 * duration;
>
> or
>
> next_report = last_report + (int64) 1000000 * progress;
>
> Is there a reason use INT64CONST instead of (int64)? Do these imply the same effect?

I guess that the macros does 1000000LL or something similar to directly 
create an int64 constant. It is necessary if the constant would overflow a 
usual 32 bits C integer, whereas the cast is sufficient if there is no 
overflow. Maybe I c/should have used the previous approach.

> Sorry, if this is a dumb question...

Nope.

-- 
Fabien.

Commits

  1. Fix pgbench timestamp bugs.

  2. pgbench: Improve time logic.

  3. Make [U]INT64CONST safe for use in #if conditions.

  4. Teach libpq to detect integer overflow in the row count of a PGresult.