Re: uninitialized values in revised prepared xact code

From: Heikki Linnakangas <hlinnakangas(at)vmware(dot)com>
To: Andres Freund <andres(at)2ndquadrant(dot)com>
Cc: Bruce Momjian <bruce(at)momjian(dot)us>, <pgsql-hackers(at)postgresql(dot)org>
Subject: Re: uninitialized values in revised prepared xact code
Date: 2014-10-24 15:09:40
Message-ID: 544A6BB4.2060202@vmware.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

On 10/24/2014 05:40 PM, Andres Freund wrote:
> On 2014-10-24 17:13:49 +0300, Heikki Linnakangas wrote:
>> Yeah. The padding bytes in TwoPhaseFileHeader were not initialized.
>>
>> That's simple enough to fix, but when I run valgrind, I get a lot whole
>> bunch of similar messages. A few are from pgstat: the padding bytes in the
>> pgstat messages are not initialized. One comes from
>> write_relcache_init_file(); again I believe it's padding bytes being
>> uninitialized (in FormData_pg_attribute). And one from the XLogInsert from
>> heap_insert; there's an uninitialized padding byte in xl_heap_insert. And so
>> forth.. Is it worthwhile to hunt down all of these? If there aren't many
>> more than these, it probably is worth it, but I fear this might be an
>> endless effort. Have we been clean of these warnings at any point in the
>> past?
>
> Did you use the valgrind suppression file in src/tools? It suppresses
> some "known harmless" cases.

Ah, I did not. With the file, the original warning that started this
thread is gone; you added a suppression for it in commit
9a0a12f683235d3e10b873baba974f6414297a7e.

- Heikki

In response to

Browse pgsql-hackers by date

  From Date Subject
Next Message Robert Haas 2014-10-24 15:25:23 Re: Deferring some AtStart* allocations?
Previous Message Andres Freund 2014-10-24 14:40:04 Re: uninitialized values in revised prepared xact code