Re: ch_malloc of 8388608 bytes failed (ITS#2270)

[joseph.tingiris@cox.net]

Thanks for the update.  I will apply the patch and inform the list as to its
effectiveness.

I wanted to add another bit of information that may be useful to other Sun
administrators using OpenLDAP 2.1.x on Sun Fire 3800, 4800, 6000, 10000, and
15k servers.  Sun released a libc, libthread, and malloc patch (108827-40,
release 02/07/03) that I applied to this same machine, recently.  What
caught my eye were these bugs that are primarily applicable to the sunfire
series (I wasn't seeing slapd crash on multiprocessor 280R, 420R, etc) but
only on the SC systems.

4193683 malloc() with very big requested array size works incorrectly
4225913 malloc does not set errno to ENOMEM when not enough memory
4314913 fix for Bug 4042527 leaves part of heap unusable, so frees cannot
combine.

Now, after the patch, the behavior of the slapd crashes are different.  I'm
getting a core file, now, that's nice.  And, slapd is not throwing the
ch_malloc failed message, anymore, it just crashes without a log message,
now.  Tonight, I plan on fixing back-bdb/idl.c with the recommended change
to Openldap, recompiling with the new Solaris (patched) libraries, and then
I will continue to monitor its progress.

Thanks again for all the help,

Joseph

----- Original Message -----
From: "Kurt D. Zeilenga" <Kurt@OpenLDAP.org>
To: <joseph.tingiris@cox.net>
Cc: <openldap-its@OpenLDAP.org>
Sent: Wednesday, February 12, 2003 4:44 PM
Subject: Re: ch_malloc of 8388608 bytes failed (ITS#2270)


> At 11:57 AM 2/11/2003, joseph.tingiris@cox.net wrote:
> >I'm just curious why this case is suspended?
>
> Because it waiting for me to say:
>
>   This issue was discussed on the software list.  Two
>   fixes were discussed.  A fix based upon Howard's
>   comment:
>     http://www.openldap.org/lists/openldap-software/200302/msg00228.html
>   has been applied to HEAD and OPENLDAP_REL_ENG_2_1.
>
>   Please test.
>
> and then change the status to Feedback...
>
>