Re: BUG #5235: Segmentation fault under high load through JDBC

From: Craig Ringer <craig(at)postnewspapers(dot)com(dot)au>
To: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
Cc: Oleg Yurchenko <oleg(at)fts(dot)ee>, pgsql-bugs(at)postgresql(dot)org
Subject: Re: BUG #5235: Segmentation fault under high load through JDBC
Date: 2009-12-08 12:19:34
Message-ID: 4B1E4456.8070502@postnewspapers.com.au
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-bugs

On 8/12/2009 8:09 AM, Tom Lane wrote:
> "Oleg Yurchenko"<oleg(at)fts(dot)ee> writes:
>> Program terminated with signal 11, Segmentation fault.
>> #0 0x0839380b in MemoryContextAllocZeroAligned (context=0x2d9eaf70,
>> size=16) at mcxt.c:559
>> 559 mcxt.c: No such file or directory.
>> in mcxt.c
>> [New Thread 28b01140 (LWP 100115)]
>
>> #(gdb)bt
>> #14185 0x081c1024 in ExecTargetList (targetlist=0x2b61a158,
>> econtext=0x2b619330, values=0x2b61a0c8, isnull=0x2b61a0d8 "",
>> itemIsDone=0x2b61a170, isDone=0xbfbfe4f0) at execQual.c:5007
>> #14186 0x081c14cd in ExecProject (projInfo=0x2b61a0e8, isDone=0xbfbfe4f0) at
>> execQual.c:5222
>
> So where are the 14184 intermediate call levels?

To generate a full backtrace, disable paging and enable logging with:

(gdb) set pagination off
(gdb) set logging file debuglog.txt
(gdb) set logging on
(gdb) bt

Gzip the resulting 'debuglog.txt' and put it somewhere accessible, then
post a link to this mailing list.

(If your PostgreSQL was built with full debugging support, rather than
just having debuginfo packages installed for a distro version, run 'gdb
full' instead of 'gdb' to get stack frame details).

Alternately, it might be helpful enough to just go back toward the
bottom of the stack and dump the first few hundred frames:

(gdb) set pagination off
(gdb) set logging file debuglog.txt
(gdb) set logging on
(gdb) bt -200

... then paste the result directly into this.

--
Craig Ringer

In response to

Responses

Browse pgsql-bugs by date

  From Date Subject
Next Message Hiroshi Saito 2009-12-08 14:27:18 Re: BUG #5099: When MetaData is acquired, it becomes an SQL error.
Previous Message Robert Haas 2009-12-08 04:10:01 Re: BUG #5210: error in intidb process when installing on japanese