Mailing List Archive

rsyslog 1.19.11 released
Hi all,

in the effort to fix the segfault bug, I have just released 1.19.11 of
rsyslog. There is a single, minimal, change in that release: an
additional compiler option is set. In short, the "-pthread" option was
missing and it may be the reason for the segfault. The longer story is
here:

http://rgerhards.blogspot.com/2007/11/did-i-find-rsyslog-bug.html

I released a new version to make sure new users will find the fixed
result set. Setting that option definitely doesn't hurt. The question
now is if it helps.

So... if you experience the segfault AND you are NOT running the
instrumented build created by Peter Vrabec, I'd appreciate if you could
update and let me know the results.

If you run Peter's version AND have experienced and reported a segfault,
I'll also kindly ask you to upgrade to the new version and see what
happens.

If you run Peter's version AND have not yet experienced a segfault with
it, I'd prefer if you could continue running it, at least for a while.
Even though I may have found the bug, there is still no solid evidence
and a report from Peter's version may help provide that evidence.

If you DO NOT experience a segfault, there is no point in upgrading.
Save yourself the time ;)

Download location:
http://www.rsyslog.com/Downloads-req-getit-lid-63.phtml

Many thanks to all for your collaboration on this issue. Your help is
deeply appreciated.

Many thanks,
Rainer
rsyslog 1.19.11 released [ In reply to ]
rsyslog 1.19.11 released [ In reply to ]
2007/11/29, Rainer Gerhards <rgerhards at hq.adiscon.com>:
> Hi all,
>
> in the effort to fix the segfault bug, I have just released 1.19.11 of
> rsyslog. There is a single, minimal, change in that release: an
> additional compiler option is set. In short, the "-pthread" option was
> missing and it may be the reason for the segfault. The longer story is

There is a small problem:
pthreads_libs, as used in rsyslogd_LDADD, is not set anymore.
The attached patch fixes that.
I also used a separate pthreads_cflags variable which is assigned to
rsyslogd_CPPFLAGS. It's cleaner than overwriting the global CFLAGS
variable.

Please review.

Cheers,
Michael


--
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?
rsyslog 1.19.11 released [ In reply to ]
Ah, excellent. Once more I have to admit that I am not a tools guy ;)

And, as usual, the mailing list is very picky - the attachment was
stripped. Can you send it via private mail, I can host it on
rsyslog.com.

Rainer

> -----Original Message-----
> From: rsyslog-bounces at lists.adiscon.com [mailto:rsyslog-
> bounces at lists.adiscon.com] On Behalf Of Michael Biebl
> Sent: Friday, November 30, 2007 11:15 AM
> To: rsyslog-users
> Subject: Re: [rsyslog] rsyslog 1.19.11 released
>
> 2007/11/29, Rainer Gerhards <rgerhards at hq.adiscon.com>:
> > Hi all,
> >
> > in the effort to fix the segfault bug, I have just released 1.19.11
> of
> > rsyslog. There is a single, minimal, change in that release: an
> > additional compiler option is set. In short, the "-pthread" option
> was
> > missing and it may be the reason for the segfault. The longer story
> is
>
> There is a small problem:
> pthreads_libs, as used in rsyslogd_LDADD, is not set anymore.
> The attached patch fixes that.
> I also used a separate pthreads_cflags variable which is assigned to
> rsyslogd_CPPFLAGS. It's cleaner than overwriting the global CFLAGS
> variable.
>
> Please review.
>
> Cheers,
> Michael
>
>
> --
> Why is it that all of the instruments seeking intelligent life in the
> universe are pointed away from Earth?
rsyslog 1.19.11 released [ In reply to ]
2007/11/30, Rainer Gerhards <rgerhards at hq.adiscon.com>:
> And, as usual, the mailing list is very picky - the attachment was
> stripped. Can you send it via private mail, I can host it on
> rsyslog.com.

Oh, sorry. I forgot...
Here it is:
http://debs.michaelbiebl.de/rsyslog/0001-pthreads-cflags-proper.patch

Cheers,
MIchael

--
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?
rsyslog 1.19.11 released [ In reply to ]
I just noticed one other thing: the ommysql.c is being generated by its
own makefile (we changed that a few releases down the road). However,
that means that makefile must also be compiled with -pthread -
otherwise, run time library calls in it may fail on reentrancy. I am
right now changing this. I will not re-issue the release. Will do
together with the other things as 1.19.12, but let's wait for some
feedback first...

Rainer

> -----Original Message-----
> From: rsyslog-bounces at lists.adiscon.com [mailto:rsyslog-
> bounces at lists.adiscon.com] On Behalf Of Michael Biebl
> Sent: Friday, November 30, 2007 12:02 PM
> To: rsyslog-users
> Subject: Re: [rsyslog] rsyslog 1.19.11 released
>
> 2007/11/30, Rainer Gerhards <rgerhards at hq.adiscon.com>:
> > And, as usual, the mailing list is very picky - the attachment was
> > stripped. Can you send it via private mail, I can host it on
> > rsyslog.com.
>
> Oh, sorry. I forgot...
> Here it is:
> http://debs.michaelbiebl.de/rsyslog/0001-pthreads-cflags-proper.patch
>
> Cheers,
> MIchael
>
> --
> Why is it that all of the instruments seeking intelligent life in the
> universe are pointed away from Earth?
> _______________________________________________
> rsyslog mailing list
> http://lists.adiscon.net/mailman/listinfo/rsyslog