Herewith some backtraces:
#0 __strlen_sse2 () at ../sysdeps/x86_64/multiarch/../strlen.S:120
No locals.
#1 0x7f3e6c2b25d3 in _IO_vfprintf_internal (s=s@entry=0x557c28170060,
format=format@entry=0x557c27af4378 "CRITICAL:core:%s: freeing already freed %s
pointer (%p), first free: %s:
Hi everyone
I fixed my core_dump issue and have uploaded the output of a 'bt full' here:
https://github.com/OpenSIPS/opensips/issues/2667
HTH
Mark.
On Thu, 18 Nov 2021 at 21:38, Vlad Patrascu wrote:
> Hi Gregory,
>
> It is difficult to draw any conclusions without a backtrace so I'm hoping
Hi Gregory,
It is difficult to draw any conclusions without a backtrace so I'm
hoping you'll be able to get the proper core dumps again in order to
investigate this further.
Regards,
--
Vlad Patrascu
OpenSIPS Core Developer
http://www.opensips-solutions.com
On 12.11.2021 14:44, Gregory
Sorry, I do have the core files, however, my /usr/sbin/opensips is now
3.2.2 and the core dumps were taken against 3.2.3.
I should have thought to get the backtrace before downgrading.
This is a production server, however, I will try and schedule some
downtime and then restore 3.2.3 to
Sadly I have not got my core dumps working yet.
However, I have just cloned & compiled the 3.2 branch and I can confirm
that the issue remains.
opensips -V
version: opensips 3.2.3 (x86_64/linux)
flags: STATS: On, DISABLE_NAGLE, USE_MCAST, SHM_MMAP, PKG_MALLOC, Q_MALLOC,
F_MALLOC, HP_MALLOC,
Do you have a core dump with the backtrace when it dies?
There are a few fixes in the nightly releases around some TLS things, and
we've found these to be a better choice for SSL performance.
Thanks,
Andrew
On Fri, 12 Nov 2021 at 02:30, Gregory Massel wrote:
> Thanks.
>
> This report lists it
Thanks.
This report lists it as a bug in WolfSSL, however, given that I'm got
similar errors (although not a segfault) when changing to OpenSSL, it
would appear more likely that it's in proto_tls or tls_mgm.
--Greg
On 2021-11-11 17:06, Mark Farmer wrote:
Confirmed here too, 3.2.2 is
Confirmed here too, 3.2.2 is running fine.
There does seem to be a bug report about this:
https://github.com/OpenSIPS/opensips/issues/2667
On Thu, 11 Nov 2021 at 14:33, Gregory Massel wrote:
> I've managed to downgrade to 3.2.2 and all is working again. So this
> appears to be a bug
My version info:
opensips -V
version: opensips 3.2.3 (x86_64/linux)
flags: STATS: On, DISABLE_NAGLE, USE_MCAST, SHM_MMAP, PKG_MALLOC, Q_MALLOC,
F_MALLOC, HP_MALLOC, DBG_MALLOC, FAST_LOCK-ADAPTIVE_WAIT
ADAPTIVE_WAIT_LOOPS=1024, MAX_RECV_BUFFER_SIZE 262144, MAX_LISTEN 16,
MAX_URI_SIZE 1024,
Weirdly I started seeing this yesterday and I have not found a solution yet.
I am trying to get a core dump which is not working for some reason but
this is the sanitized log from my server:
INFO:tls_wolfssl:_wolfssl_tls_async_connect: new TLS connection to
52.114.75.24:5061 established
I've managed to downgrade to 3.2.2 and all is working again. So this
appears to be a bug introduced in 3.2.3.
Environment is:
Linux msteams 5.4.0-90-generic #101~18.04.1-Ubuntu SMP Fri Oct 22
09:25:04 UTC 2021 x86_64 x86_64 x86_64 GNU/Linux
The downgrade/fix was with:
apt-get install
I'm running OpenSIPS 3.2.3 and it keeps bombing out on SSL connections.
With WolfSSL it segfaults:
Nov 11 11:52:04 msteams /usr/sbin/opensips[15322]:
NOTICE:tls_wolfssl:verify_callback: depth = 1, verify success
Nov 11 11:52:04 msteams /usr/sbin/opensips[15322]:
12 matches
Mail list logo