Re: [OpenSIPS-Users] opensips 3.2.2 Segfault on debian 11 bullseye
Yes, I've updated to the latest version and can confirm the crash no longer happens. Thanks. -- Regards, William Jin From: Liviu Chircu Sent: Wednesday, 29 September 2021 8:01 PM To: William Jin ; OpenSIPS users mailling list Subject: Re: [OpenSIPS-Users] opensips 3.2.2 Segfault on debian 11 bullseye On 17.09.2021 07:42, William Jin wrote: Sep 17 14:33:23 ip-10-20-12-22 /usr/sbin/opensips[188225]: CRITICAL:core:sig_usr: segfault in process pid: 188225, id: 15 Hi William, Not sure if you noticed the identical thread below (with James Fox, Rob Dyck and Max) but this crash has been fixed in latest 3.2! Cheers, -- Liviu Chircu www.twitter.com/liviuchircu<http://www.twitter.com/liviuchircu> | www.opensips-solutions.com<http://www.opensips-solutions.com> OpenSIPS eBootcamp 2021 | www.opensips.org/training<http://www.opensips.org/training> ___ Users mailing list Users@lists.opensips.org http://lists.opensips.org/cgi-bin/mailman/listinfo/users
Re: [OpenSIPS-Users] opensips 3.2.2 Segfault on debian 11 bullseye
On 17.09.2021 07:42, William Jin wrote: Sep 17 14:33:23 ip-10-20-12-22 /usr/sbin/opensips[188225]: CRITICAL:core:sig_usr: segfault in process pid: 188225, id: 15 Hi William, Not sure if you noticed the identical thread below (with James Fox, Rob Dyck and Max) but this crash has been fixed in latest 3.2! Cheers, -- Liviu Chircu www.twitter.com/liviuchircu | www.opensips-solutions.com OpenSIPS eBootcamp 2021 | www.opensips.org/training ___ Users mailing list Users@lists.opensips.org http://lists.opensips.org/cgi-bin/mailman/listinfo/users
Re: [OpenSIPS-Users] opensips 3.2.2 Segfault on debian 11 bullseye
Merged here, thanks for reporting:
https://github.com/OpenSIPS/opensips/commit/3f9fb0923207e320a27926433f50e0c39a1c643c
(master)
https://github.com/OpenSIPS/opensips/commit/83fee1c7a42f6d01b8d770088728791a03727b6
(3.2)
-Max
On Mon, Sep 27, 2021 at 12:46 PM Rob Dyck wrote:
> Thank you
> With the patch applied I can confirm that the users are registering
> without
> incident..
> Rob
>
> On Monday, September 27, 2021 10:52:15 A.M. PDT Maxim Sobolev wrote:
> > Hi Rob / Liviu,
> >
> > I browsed quickly through the code and I think the following clause may
> be
> > a culprit:
> >
> > if (calc_ha1) {
> > /* Only plaintext passwords are stored in database,
> > * we have to calculate HA1 */
> > cprms.creds.open = &(const struct
> digest_auth_credential){
> > .realm = *_domain, .user = _username->whole, .passwd
> =
> > result};
> > cprms.use_hashed = 0;
> > }
> >
> > Compiler might deallocate / overwrite struct digest_auth_credential after
> > exiting that block causing subsequent call to auth_api.calc_HA1() to
> access
> > bogus pointer.
> >
> > Rob, can you try applying the following commit and recompile/reinstall
> the
> > module in question and see if it helps?
> >
> >
> https://github.com/sippy/opensips/commit/fea6a1d60d70f64971dff3ec2dc83f7ddc0
> > 0389d
> >
> > Thanks!
> >
> > -Max
> >
> > On Mon, Sep 27, 2021 at 12:48 AM Liviu Chircu
> wrote:
> > > On 27.09.2021 03:56, Rob Dyck wrote:
> > > > I am seeing the same. opensips-3.2.2 compiled from git source on
> Fedora.
> > >
> > > Thank you for the help, gents! Let's see if I can reproduce it...
> > > should be fairly straightforward.
> > >
> > > Best,
> > >
> > > --
> > > Liviu Chircu
> > > www.twitter.com/liviuchircu | www.opensips-solutions.com
> > >
> > >
> > > ___
> > > Users mailing list
> > > Users@lists.opensips.org
> > > http://lists.opensips.org/cgi-bin/mailman/listinfo/users
>
>
>
>
>
___
Users mailing list
Users@lists.opensips.org
http://lists.opensips.org/cgi-bin/mailman/listinfo/users
Re: [OpenSIPS-Users] opensips 3.2.2 Segfault on debian 11 bullseye
Thank you
With the patch applied I can confirm that the users are registering without
incident..
Rob
On Monday, September 27, 2021 10:52:15 A.M. PDT Maxim Sobolev wrote:
> Hi Rob / Liviu,
>
> I browsed quickly through the code and I think the following clause may be
> a culprit:
>
> if (calc_ha1) {
> /* Only plaintext passwords are stored in database,
> * we have to calculate HA1 */
> cprms.creds.open = &(const struct digest_auth_credential){
> .realm = *_domain, .user = _username->whole, .passwd =
> result};
> cprms.use_hashed = 0;
> }
>
> Compiler might deallocate / overwrite struct digest_auth_credential after
> exiting that block causing subsequent call to auth_api.calc_HA1() to access
> bogus pointer.
>
> Rob, can you try applying the following commit and recompile/reinstall the
> module in question and see if it helps?
>
> https://github.com/sippy/opensips/commit/fea6a1d60d70f64971dff3ec2dc83f7ddc0
> 0389d
>
> Thanks!
>
> -Max
>
> On Mon, Sep 27, 2021 at 12:48 AM Liviu Chircu wrote:
> > On 27.09.2021 03:56, Rob Dyck wrote:
> > > I am seeing the same. opensips-3.2.2 compiled from git source on Fedora.
> >
> > Thank you for the help, gents! Let's see if I can reproduce it...
> > should be fairly straightforward.
> >
> > Best,
> >
> > --
> > Liviu Chircu
> > www.twitter.com/liviuchircu | www.opensips-solutions.com
> >
> >
> > ___
> > Users mailing list
> > Users@lists.opensips.org
> > http://lists.opensips.org/cgi-bin/mailman/listinfo/users
___
Users mailing list
Users@lists.opensips.org
http://lists.opensips.org/cgi-bin/mailman/listinfo/users
Re: [OpenSIPS-Users] opensips 3.2.2 Segfault on debian 11 bullseye
On 27.09.2021 20:52, Maxim Sobolev wrote: and see if it helps? Umm... it definitely helped me in a very interesting way! [1] [1]: https://github.com/OpenSIPS/opensips/commit/0fadc0a6cb130d40fba6cf36bb1399d45d0496aa Cheers! -- Liviu Chircu www.twitter.com/liviuchircu | www.opensips-solutions.com ___ Users mailing list Users@lists.opensips.org http://lists.opensips.org/cgi-bin/mailman/listinfo/users
Re: [OpenSIPS-Users] opensips 3.2.2 Segfault on debian 11 bullseye
Hi Rob / Liviu,
I browsed quickly through the code and I think the following clause may be
a culprit:
if (calc_ha1) {
/* Only plaintext passwords are stored in database,
* we have to calculate HA1 */
cprms.creds.open = &(const struct digest_auth_credential){
.realm = *_domain, .user = _username->whole, .passwd =
result};
cprms.use_hashed = 0;
}
Compiler might deallocate / overwrite struct digest_auth_credential after
exiting that block causing subsequent call to auth_api.calc_HA1() to access
bogus pointer.
Rob, can you try applying the following commit and recompile/reinstall the
module in question and see if it helps?
https://github.com/sippy/opensips/commit/fea6a1d60d70f64971dff3ec2dc83f7ddc00389d
Thanks!
-Max
On Mon, Sep 27, 2021 at 12:48 AM Liviu Chircu wrote:
> On 27.09.2021 03:56, Rob Dyck wrote:
> > I am seeing the same. opensips-3.2.2 compiled from git source on Fedora.
>
> Thank you for the help, gents! Let's see if I can reproduce it...
> should be fairly straightforward.
>
> Best,
>
> --
> Liviu Chircu
> www.twitter.com/liviuchircu | www.opensips-solutions.com
>
>
> ___
> Users mailing list
> Users@lists.opensips.org
> http://lists.opensips.org/cgi-bin/mailman/listinfo/users
>
___
Users mailing list
Users@lists.opensips.org
http://lists.opensips.org/cgi-bin/mailman/listinfo/users
Re: [OpenSIPS-Users] opensips 3.2.2 Segfault on debian 11 bullseye
On 27.09.2021 03:56, Rob Dyck wrote: I am seeing the same. opensips-3.2.2 compiled from git source on Fedora. Thank you for the help, gents! Let's see if I can reproduce it... should be fairly straightforward. Best, -- Liviu Chircu www.twitter.com/liviuchircu | www.opensips-solutions.com ___ Users mailing list Users@lists.opensips.org http://lists.opensips.org/cgi-bin/mailman/listinfo/users
Re: [OpenSIPS-Users] opensips 3.2.2 Segfault on debian 11 bullseye
ol table info available. > > #23 0xba5f6e00 in udp_start_processes () > > No symbol table info available. > > #24 0xba472ae4 in main () > > No symbol table info available. > > > > > > Client is performing MD5 authentication using the following header > (anonymised): > Authorization: Digest username="username", realm="realm", > nonce="+P1CdVJHGkKtXhTn4vsZUwM66e8TYcsFS0V0cIYYcrsA", uri="sip:ipaddress", > response="32charhash", cnonce="32charhash", qop=auth, nc=0001 > > > Any help would be appreciated. > > > > Regards, > > > > James > > > > From: Liviu Chircu > Sent: Saturday, 25 September 2021 1:03 AM > To: James Fox ; OpenSIPS users mailling list > Subject: Re: [OpenSIPS-Users] opensips 3.2.2 > Segfault on debian 11 bullseye > > > On 24.09.2021 17:53, James Fox via Users wrote: > > Previously this script was working with Opensips 3.1.1. I’ve manually > upgraded the subscriber database schema to include the new column: > ha1_sha512t256. > I’ll work on getting a crash dump. Any ideas in the meantime would be > welcome. > Hi, > > This bug was fixed ~2 months ago, and I'm pretty sure that both 3.2.0 or > 3.1.1 minors date *before* the fix. So please update your repo, grab the > latest minor release for each of them and the bug should go away. > Best, > > -- > Liviu Chircu > www.twitter.com/liviuchircu <http://www.twitter.com/liviuchircu> | > www.opensips-solutions.com <http://www.opensips-solutions.com> ___ Users mailing list Users@lists.opensips.org http://lists.opensips.org/cgi-bin/mailman/listinfo/users
Re: [OpenSIPS-Users] opensips 3.2.2 Segfault on debian 11 bullseye
Hi Liviu, Thanks for your response. I’m still experiencing the segfault after upgrading to 3.2.2. I’ve managed to output a stack trace from the core dump. I believe it shows the segfault occurring on MD5Update(). Reading symbols from /usr/sbin/opensips... (No debugging symbols found in /usr/sbin/opensips) [New LWP 92955] [Thread debugging using libthread_db enabled] Using host libthread_db library "/lib/aarch64-linux-gnu/libthread_db.so.1". Core was generated by `/usr/sbin/opensips -P /run/opensips/opensips.pid -f /etc/opensips/opensips.cfg'. Program terminated with signal SIGSEGV, Segmentation fault. --Type for more, q to quit, c to continue without paging--c #0 0xba47c7c8 in ?? () (gdb) bt full #0 0xba47c7c8 in ?? () No symbol table info available. #1 0xba47d2cc in MD5Update () No symbol table info available. #2 0xac85c8fc in ?? () from /usr/lib/aarch64-linux-gnu/opensips/modules/auth.so No symbol table info available. #3 0xac858678 in ?? () from /usr/lib/aarch64-linux-gnu/opensips/modules/auth.so No symbol table info available. #4 0xac5aec20 in ?? () from /usr/lib/aarch64-linux-gnu/opensips/modules/auth_db.so No symbol table info available. #5 0xac5afb28 in www_authorize () from /usr/lib/aarch64-linux-gnu/opensips/modules/auth_db.so No symbol table info available. #6 0xba491d1c in do_action () No symbol table info available. #7 0xba4944a8 in run_action_list () No symbol table info available. #8 0xba4daf9c in ?? () No symbol table info available. #9 0xba4da94c in eval_expr () --Type for more, q to quit, c to continue without paging--c No symbol table info available. #10 0xba4daa14 in eval_expr () No symbol table info available. #11 0xba491c84 in do_action () No symbol table info available. #12 0xba4944a8 in run_action_list () No symbol table info available. #13 0xba493b28 in do_action () No symbol table info available. #14 0xba4944a8 in run_action_list () No symbol table info available. #15 0xba492fac in do_action () No symbol table info available. #16 0xba4944a8 in run_action_list () No symbol table info available. #17 0xba493b28 in do_action () No symbol table info available. #18 0xba4944a8 in run_action_list () No symbol table info available. #19 0xba494814 in run_top_route () No symbol table info available. #20 0xba4a36ec in receive_msg () No symbol table info available. #21 0xba61af20 in ?? () No symbol table info available. #22 0xba5f2248 in ?? () No symbol table info available. #23 0xba5f6e00 in udp_start_processes () No symbol table info available. #24 0xba472ae4 in main () No symbol table info available. Client is performing MD5 authentication using the following header (anonymised): Authorization: Digest username="username", realm="realm", nonce="+P1CdVJHGkKtXhTn4vsZUwM66e8TYcsFS0V0cIYYcrsA", uri="sip:ipaddress", response="32charhash", cnonce="32charhash", qop=auth, nc=0001 Any help would be appreciated. Regards, James From: Liviu Chircu Sent: Saturday, 25 September 2021 1:03 AM To: James Fox ; OpenSIPS users mailling list Subject: Re: [OpenSIPS-Users] opensips 3.2.2 Segfault on debian 11 bullseye On 24.09.2021 17:53, James Fox via Users wrote: Previously this script was working with Opensips 3.1.1. I’ve manually upgraded the subscriber database schema to include the new column: ha1_sha512t256. I’ll work on getting a crash dump. Any ideas in the meantime would be welcome. Hi, This bug was fixed ~2 months ago, and I'm pretty sure that both 3.2.0 or 3.1.1 minors date *before* the fix. So please update your repo, grab the latest minor release for each of them and the bug should go away. Best, -- Liviu Chircu www.twitter.com/liviuchircu <http://www.twitter.com/liviuchircu> | www.opensips-solutions.com <http://www.opensips-solutions.com> ___ Users mailing list Users@lists.opensips.org http://lists.opensips.org/cgi-bin/mailman/listinfo/users
Re: [OpenSIPS-Users] opensips 3.2.2 Segfault on debian 11 bullseye
On 24.09.2021 17:53, James Fox via Users wrote: Previously this script was working with Opensips 3.1.1. I’ve manually upgraded the subscriber database schema to include the new column: ha1_sha512t256. I’ll work on getting a crash dump. Any ideas in the meantime would be welcome. Hi, This bug was fixed ~2 months ago, and I'm pretty sure that both 3.2.0 or 3.1.1 minors date *before* the fix. So please update your repo, grab the latest minor release for each of them and the bug should go away. Best, -- Liviu Chircu www.twitter.com/liviuchircu | www.opensips-solutions.com ___ Users mailing list Users@lists.opensips.org http://lists.opensips.org/cgi-bin/mailman/listinfo/users
Re: [OpenSIPS-Users] opensips 3.2.2 Segfault on debian 11 bullseye
And I just enabled log_level=4 It connects to db (mysql 8.0.26-1debian11) and converting password string to val, then segfault Sep 17 14:33:23 ip-10-20-12-22 /usr/sbin/opensips[188225]: DBG:db_mysql:db_mysql_do_prepared_query: conn=0x7f6d33facd80 (tail=140107000235208) MC=0x7f6d33fac560 Sep 17 14:33:23 ip-10-20-12-22 /usr/sbin/opensips[188225]: DBG:db_mysql:db_mysql_do_prepared_query: new query=|select password,rpid from subscriber where username=?| Sep 17 14:33:23 ip-10-20-12-22 /usr/sbin/opensips[188225]: DBG:db_mysql:re_init_statement: query is , ptr=(nil) Sep 17 14:33:23 ip-10-20-12-22 /usr/sbin/opensips[188225]: DBG:db_mysql:mysql_raise_event: MySQL status has not changed: connected Sep 17 14:33:23 ip-10-20-12-22 /usr/sbin/opensips[188225]: DBG:db_mysql:db_mysql_do_prepared_query: new statement(0x7f6d33faf978) on connection: (0x7f6d33facd80) 0x7f6d33fac4c8 Sep 17 14:33:23 ip-10-20-12-22 /usr/sbin/opensips[188225]: DBG:db_mysql:db_mysql_do_prepared_query: set values for the statement run Sep 17 14:33:23 ip-10-20-12-22 /usr/sbin/opensips[188225]: DBG:db_mysql:db_mysql_val2bind: added val (0): len=10; type=254; is_null=0 Sep 17 14:33:23 ip-10-20-12-22 /usr/sbin/opensips[188225]: DBG:db_mysql:db_mysql_do_prepared_query: doing BIND_PARAM in... Sep 17 14:33:23 ip-10-20-12-22 /usr/sbin/opensips[188225]: DBG:db_mysql:mysql_raise_event: MySQL status has not changed: connected Sep 17 14:33:23 ip-10-20-12-22 /usr/sbin/opensips[188225]: DBG:db_mysql:db_mysql_do_prepared_query: prepared statement has 2 columns in result Sep 17 14:33:23 ip-10-20-12-22 /usr/sbin/opensips[188225]: DBG:db_mysql:db_mysql_do_prepared_query: doing to BIND_PARAM out ... Sep 17 14:33:23 ip-10-20-12-22 /usr/sbin/opensips[188225]: DBG:core:db_new_result: allocate 48 bytes for result set at 0x7f6d33fadf68 Sep 17 14:33:23 ip-10-20-12-22 /usr/sbin/opensips[188225]: DBG:db_mysql:db_mysql_get_columns: 2 columns returned from the query Sep 17 14:33:23 ip-10-20-12-22 /usr/sbin/opensips[188225]: DBG:core:db_allocate_columns: allocate 56 bytes for result columns at 0x7f6d33faf060 Sep 17 14:33:23 ip-10-20-12-22 /usr/sbin/opensips[188225]: DBG:db_mysql:db_mysql_get_columns: RES_NAMES(0x7f6d33faf070)[0]=[password] Sep 17 14:33:23 ip-10-20-12-22 /usr/sbin/opensips[188225]: DBG:db_mysql:db_mysql_get_columns: use DB_STRING result type Sep 17 14:33:23 ip-10-20-12-22 /usr/sbin/opensips[188225]: DBG:db_mysql:db_mysql_get_columns: RES_NAMES(0x7f6d33faf080)[1]=[rpid] Sep 17 14:33:23 ip-10-20-12-22 /usr/sbin/opensips[188225]: DBG:db_mysql:db_mysql_get_columns: use DB_STRING result type Sep 17 14:33:23 ip-10-20-12-22 /usr/sbin/opensips[188225]: DBG:core:db_allocate_rows: allocate 80 bytes for result rows and values at 0x7f6d33fb0450 Sep 17 14:33:23 ip-10-20-12-22 /usr/sbin/opensips[188225]: DBG:db_mysql:db_mysql_str2val: converting STRING [my@password] Sep 17 14:33:23 ip-10-20-12-22 /usr/sbin/opensips[188225]: CRITICAL:core:sig_usr: segfault in process pid: 188225, id: 15 -- Regards, William Jin From: Users on behalf of William Jin Sent: Friday, 17 September 2021 2:19 PM To: Liviu Chircu ; OpenSIPS users mailling list Subject: Re: [OpenSIPS-Users] opensips 3.2.2 Segfault on debian 11 bullseye And just further investigation, looks like it's crashed only after I register my test user. (which can register on debian 10 before upgraded to debian 11 using zoiper softphone) And the backtrack also indicate the crash from auth_db?? -- Regards, William Jin From: Users on behalf of William Jin Sent: Friday, 17 September 2021 9:12 AM To: Liviu Chircu ; OpenSIPS users mailling list Subject: Re: [OpenSIPS-Users] opensips 3.2.2 Segfault on debian 11 bullseye Hi Liviu, I attached the backtrace in the init email. I've attached it again. -- Regards, William Jin From: Liviu Chircu Sent: Thursday, 16 September 2021 5:12 PM To: OpenSIPS users mailling list ; William Jin Subject: Re: [OpenSIPS-Users] opensips 3.2.2 Segfault on debian 11 bullseye On 16.09.2021 02:16, William Jin wrote: > Anyone has any chance to look into this? Hi William, There is not enough info to look into anything, really: * how are you reproducing this crash? can you share a basic opensips.cfg that causes it? * are there any useful logs _before_ those "signal 11" logs? * can you obtain a backtrace out of this crash? See this guide [1] for some help here [1]: https://www.opensips.org/Documentation/TroubleShooting-Crash Best, -- Liviu Chircu www.twitter.com/liviuchircu<http://www.twitter.com/liviuchircu> | www.opensips-solutions.com<http://www.opensips-solutions.com> ___ Users mailing list Users@lists.opensips.org http://lists.opensips.org/cgi-bin/mailman/listinfo/users
Re: [OpenSIPS-Users] opensips 3.2.2 Segfault on debian 11 bullseye
And just further investigation, looks like it's crashed only after I register my test user. (which can register on debian 10 before upgraded to debian 11 using zoiper softphone) And the backtrack also indicate the crash from auth_db?? -- Regards, William Jin From: Users on behalf of William Jin Sent: Friday, 17 September 2021 9:12 AM To: Liviu Chircu ; OpenSIPS users mailling list Subject: Re: [OpenSIPS-Users] opensips 3.2.2 Segfault on debian 11 bullseye Hi Liviu, I attached the backtrace in the init email. I've attached it again. -- Regards, William Jin From: Liviu Chircu Sent: Thursday, 16 September 2021 5:12 PM To: OpenSIPS users mailling list ; William Jin Subject: Re: [OpenSIPS-Users] opensips 3.2.2 Segfault on debian 11 bullseye On 16.09.2021 02:16, William Jin wrote: > Anyone has any chance to look into this? Hi William, There is not enough info to look into anything, really: * how are you reproducing this crash? can you share a basic opensips.cfg that causes it? * are there any useful logs _before_ those "signal 11" logs? * can you obtain a backtrace out of this crash? See this guide [1] for some help here [1]: https://www.opensips.org/Documentation/TroubleShooting-Crash Best, -- Liviu Chircu www.twitter.com/liviuchircu<http://www.twitter.com/liviuchircu> | www.opensips-solutions.com<http://www.opensips-solutions.com> ___ Users mailing list Users@lists.opensips.org http://lists.opensips.org/cgi-bin/mailman/listinfo/users
Re: [OpenSIPS-Users] opensips 3.2.2 Segfault on debian 11 bullseye
Hi Liviu,
I attached the backtrace in the init email.
I've attached it again.
--
Regards,
William Jin
From: Liviu Chircu
Sent: Thursday, 16 September 2021 5:12 PM
To: OpenSIPS users mailling list ; William Jin
Subject: Re: [OpenSIPS-Users] opensips 3.2.2 Segfault on debian 11 bullseye
On 16.09.2021 02:16, William Jin wrote:
> Anyone has any chance to look into this?
Hi William,
There is not enough info to look into anything, really:
* how are you reproducing this crash? can you share a basic
opensips.cfg that causes it?
* are there any useful logs _before_ those "signal 11" logs?
* can you obtain a backtrace out of this crash? See this guide [1] for
some help here
[1]: https://www.opensips.org/Documentation/TroubleShooting-Crash
Best,
--
Liviu Chircu
www.twitter.com/liviuchircu<http://www.twitter.com/liviuchircu> |
www.opensips-solutions.com<http://www.opensips-solutions.com>
gdb /usr/sbin/opensips /tmp/core
GNU gdb (Debian 10.1-1.7) 10.1.90.20210103-git
Copyright (C) 2021 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.
Type "show copying" and "show warranty" for details.
This GDB was configured as "x86_64-linux-gnu".
Type "show configuration" for configuration details.
For bug reporting instructions, please see:
<https://www.gnu.org/software/gdb/bugs/>.
Find the GDB manual and other documentation resources online at:
<http://www.gnu.org/software/gdb/documentation/>.
For help, type "help".
Type "apropos word" to search for commands related to "word"...
Reading symbols from /usr/sbin/opensips...
Reading symbols from
/usr/lib/debug/.build-id/68/43b0716a68e5fab40b5fba966d9bce8d0e224f.debug...
warning: Can't open file /dev/zero (deleted) during file-backed mapping note
processing
[New LWP 10586]
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1".
Core was generated by `/usr/sbin/opensips -P /run/opensips/opensips.pid -f
/etc/opensips/opensips.cfg'.
Program terminated with signal SIGSEGV, Segmentation fault.
#0 0x559ddc5502a0 in memcpy (__len=64, __src=0xaf01ff,
__dest=0x7ffd97072be8) at
/usr/include/x86_64-linux-gnu/bits/string_fortified.h:34
34return __builtin___memcpy_chk (__dest, __src, __len, __bos0 (__dest));
(gdb) bt full
#0 0x559ddc5502a0 in memcpy (__len=64, __src=0xaf01ff,
__dest=0x7ffd97072be8) at
/usr/include/x86_64-linux-gnu/bits/string_fortified.h:34
No locals.
#1 MD5_memcpy (len=64, input=0xaf01ff , output=0x7ffd97072be8 "\240-\a\227\375\177") at md5.c:329
No locals.
#2 MD5Update (context=0x7ffd97072bd0, input=0xaf01ff , inputLen=127) at md5.c:152
i =
index = 0
partLen = 64
#3 0x7f67bad9a788 in digest_calc_HA1 (crd=0x7ffd97072d60,
sess_key=0x7ffd97072de0) at dauth_calc_md5.c:48
Md5Ctx = {state = {1732584193, 4023233417, 2562383102, 271733878},
count = {1016, 0},
buffer =
"\240-\a\227\375\177\000\000(-\a\227\375\177\000\000\000\305L\321\027`\200P\220-\a\227\375\177\000\000\200\355\003\277g\177\000\000(-\a\227\375\177\000\000\310\344\003\277g\177\000\000(-\a\227\375\177\000"}
HA1 = "(-\a\227\375\177\000\000z\247\244\272g\177\000"
#4 0x7f67bad9476d in auth_calc_HA1 (params=0x7ffd97072d40,
sess_key=0x7ffd97072de0) at api.c:341
digest_calc = 0x7f67bada2800
__FUNCTION__ = "auth_calc_HA1"
#5 0x7f67baa8e936 in get_ha1 (res=0x7ffd97072d28, _ha1=0x7ffd97072de0,
_table=, _domain=0x7ffd97072f88, digest=0x7f67bf040f58) at
authorize.c:169
cprms = {use_hashed = 0, alg = ALG_MD5, creds = {open = 0x7ffd97072d60,
ha1 = 0x7ffd97072d60}, nonce = 0x7f67bf040f98, cnonce = 0x7f67bf040fe0}
vals = {{type = DB_STR, nul = 0, free = 0, val = {int_val = -594864196,
bigint_val = 94136498329532, double_val = 4.6509609844413654e-310, time_val =
94136498329532,
string_val = 0x559ddc8b17bc
"1122334455\",realm=\"sip1.mydomain.com\",nonce=\"KprTqedqhYxO7vVNGoDdw6U2mvUQ1v6Jd7liqtOCO94A\",uri=\"sip:sip1.mydomain.com;transport=UDP\",response=\"46f25ec4472d495ed41f46677a1eb647\",algorithm=MD5"...,
str_val = {
s = 0x559ddc8b17bc
"1122334455\",realm=\"sip1.mydomain.com\",nonce=\"KprTqedqhYxO7vVNGoDdw6U2mvUQ1v6Jd7liqtOCO94A\",uri=\"sip:sip1.mydomain.com;transport=UDP\",response=\"46f25ec4472d495ed41f46677a1eb647\",algorithm=MD5"...,
len = 10}, blob_val = {
s = 0x559ddc8b17bc
"1122334455\",realm=\"sip1.mydomain.com\",nonce=\"KprTqedqhYxO
Re: [OpenSIPS-Users] opensips 3.2.2 Segfault on debian 11 bullseye
On 16.09.2021 02:16, William Jin wrote: Anyone has any chance to look into this? Hi William, There is not enough info to look into anything, really: * how are you reproducing this crash? can you share a basic opensips.cfg that causes it? * are there any useful logs _before_ those "signal 11" logs? * can you obtain a backtrace out of this crash? See this guide [1] for some help here [1]: https://www.opensips.org/Documentation/TroubleShooting-Crash Best, -- Liviu Chircu www.twitter.com/liviuchircu | www.opensips-solutions.com ___ Users mailing list Users@lists.opensips.org http://lists.opensips.org/cgi-bin/mailman/listinfo/users
Re: [OpenSIPS-Users] opensips 3.2.2 Segfault on debian 11 bullseye
Anyone has any chance to look into this? -- Regards, William Jin From: Users on behalf of William Jin Sent: Thursday, 9 September 2021 12:32 PM To: users@lists.opensips.org Subject: [OpenSIPS-Users] opensips 3.2.2 Segfault on debian 11 bullseye We currently upgraded our Debian server from 10(buster) to 11(bullseye), After the upgrade, the Opensips keep crashing with a segfault. Sep 9 10:35:04 ip-10-20-12-22 /usr/sbin/opensips[8700]: CRITICAL:core:sig_usr: segfault in process pid: 8700, id: 16 Sep 9 10:35:04 ip-10-20-12-22 /usr/sbin/opensips[8683]: INFO:core:handle_sigs: child process 8700 exited by a signal 11 Sep 9 10:35:04 ip-10-20-12-22 /usr/sbin/opensips[8683]: INFO:core:handle_sigs: core was generated Sep 9 10:35:04 ip-10-20-12-22 /usr/sbin/opensips[8683]: INFO:core:handle_sigs: terminating due to SIGCHLD We are using the apt repo opensips, tried both stable and nightly version, same result. opensips -V version: opensips 3.2.2 (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, BUF_SIZE 65535 poll method support: poll, epoll, sigio_rt, select. git revision: dc45f112e main.c compiled on with gcc 10 DBG attached PS: I've replaced my number to 1122334455, domain to sip1.mydomain.com -- Regards, William Jin ___ Users mailing list Users@lists.opensips.org http://lists.opensips.org/cgi-bin/mailman/listinfo/users
