Re: pg_dump on Postgres 9.1 - Mailing list pgsql-admin

From Tom Lane
Subject Re: pg_dump on Postgres 9.1
Date
Msg-id 10180.1344024118@sss.pgh.pa.us
Whole thread Raw
In response to Re: pg_dump on Postgres 9.1  (Alanoly Andrews <alanolya@invera.com>)
List pgsql-admin
Alanoly Andrews <alanolya@invera.com> writes:
> On this issue, instead of going for a newer version of xlc, as suggested, I opted to get a newer version of the
Postgressource code, 9.1.4. After compiling it with the same xlc version, I found that pg_dump works as expected. So,
theproblem appears to be somewhere in the 9.1 source code, at least for binaries created from it for AIX (6.1). 

Hm, maybe this 9.1.2 fix?

Author: Tom Lane <tgl@sss.pgh.pa.us>
Branch: master Release: REL9_2_BR [cf22e851b] 2011-11-10 16:09:01 -0500
Branch: REL9_1_STABLE Release: REL9_1_2 [febda37fd] 2011-11-10 16:09:16 -0500

    Avoid platform-dependent infinite loop in pg_dump.

    If malloc(0) returns NULL, the binary search in findSecLabels() will
    probably go into an infinite loop when there are no security labels,
    because NULL-1 is greater than NULL after wraparound.

    (We've seen this pathology before ... I wonder whether there's a way to
    detect the class of bugs automatically?)

    Diagnosis and patch by Steve Singer, cosmetic adjustments by me


            regards, tom lane

pgsql-admin by date:

Previous
From: "Kevin Grittner"
Date:
Subject: Re: PostgreSQL oom_adj postmaster process to -17
Next
From: Scott Marlowe
Date:
Subject: Re: PostgreSQL oom_adj postmaster process to -17