Re: relayd crash
On 2013/08/23 04:03, Bogdan Andu wrote: $ sudo gdb ./relayd GNU gdb 6.3 Copyright 2004 Free Software Foundation, Inc. GDB is free software, covered by the GNU General Public License, and you are welcome to change it and/or distribute copies of it under certain conditions. Type show copying to see the conditions. There is absolutely no warranty for GDB. Type show warranty for details. This GDB was configured as amd64-unknown-openbsd5.3... (gdb) attach 32389 Attaching to program: /usr/src/usr.sbin/relayd-current/relayd, process 32389 Loaded symbols for /usr/src/usr.sbin/relayd-current/relayd Reading symbols from /usr/lib/libevent.so.3.1...done. Loaded symbols for /usr/lib/libevent.so.3.1 Reading symbols from /usr/lib/libssl.so.19.0...done. Loaded symbols for /usr/lib/libssl.so.19.0 Reading symbols from /usr/lib/libcrypto.so.22.0...done. Loaded symbols for /usr/lib/libcrypto.so.22.0 Reading symbols from /usr/lib/libutil.so.11.4...done. Loaded symbols for /usr/lib/libutil.so.11.4 Reading symbols from /usr/lib/libc.so.66.2...done. Loaded symbols for /usr/lib/libc.so.66.2 Reading symbols from /usr/libexec/ld.so...done. Loaded symbols for /usr/libexec/ld.so [Switching to thread 1032389] 0x05462804581a in kevent () at stdin:2 2 stdin: No such file or directory. in stdin Current language: auto; currently asm (gdb) continue Continuing. Program received signal SIGSEGV, Segmentation fault. A backtrace from this one would be helpful, that will hopefully show the chain of functions which called event_queue_insert().. event_queue_insert (base=0x0, ev=0x54621afd208, queue=8) at /usr/src/ lib/libevent/event.c:997 997 base-event_count++; Current language: auto; currently c (gdb) Continuing. But why this child crashes in /usr/src/lib/libevent/event.c, line 997 ? It is attempting to dereference a null pointer. event_queue_insert (base=0x0, ev=0x54621afd208, queue=8) at /usr/src/lib/libevent/event.c:997 997 base-event_count++; ^^
Re: relayd crash
or directory. in stdin (gdb) Continuing. Program exited normally. (gdb) The program is not being run. (gdb) The program is not being run. CHILD #1: $ sudo gdb ./relayd GNU gdb 6.3 Copyright 2004 Free Software Foundation, Inc. GDB is free software, covered by the GNU General Public License, and you are welcome to change it and/or distribute copies of it under certain conditions. Type show copying to see the conditions. There is absolutely no warranty for GDB. Type show warranty for details. This GDB was configured as amd64-unknown-openbsd5.3... (gdb) attach 32389 Attaching to program: /usr/src/usr.sbin/relayd-current/relayd, process 32389 Loaded symbols for /usr/src/usr.sbin/relayd-current/relayd Reading symbols from /usr/lib/libevent.so.3.1...done. Loaded symbols for /usr/lib/libevent.so.3.1 Reading symbols from /usr/lib/libssl.so.19.0...done. Loaded symbols for /usr/lib/libssl.so.19.0 Reading symbols from /usr/lib/libcrypto.so.22.0...done. Loaded symbols for /usr/lib/libcrypto.so.22.0 Reading symbols from /usr/lib/libutil.so.11.4...done. Loaded symbols for /usr/lib/libutil.so.11.4 Reading symbols from /usr/lib/libc.so.66.2...done. Loaded symbols for /usr/lib/libc.so.66.2 Reading symbols from /usr/libexec/ld.so...done. Loaded symbols for /usr/libexec/ld.so [Switching to thread 1032389] 0x05462804581a in kevent () at stdin:2 2 stdin: No such file or directory. in stdin Current language: auto; currently asm (gdb) continue Continuing. Program received signal SIGSEGV, Segmentation fault. event_queue_insert (base=0x0, ev=0x54621afd208, queue=8) at /usr/src/lib/libevent/event.c:997 997 base-event_count++; Current language: auto; currently c (gdb) Continuing. Program terminated with signal SIGSEGV, Segmentation fault. The program no longer exists. (gdb) The program is not being run. (gdb) The program is not being run. CHILD #2: $ sudo gdb ./relayd GNU gdb 6.3 Copyright 2004 Free Software Foundation, Inc. GDB is free software, covered by the GNU General Public License, and you are welcome to change it and/or distribute copies of it under certain conditions. Type show copying to see the conditions. There is absolutely no warranty for GDB. Type show warranty for details. This GDB was configured as amd64-unknown-openbsd5.3... (gdb) attach 10140 Attaching to program: /usr/src/usr.sbin/relayd-current/relayd, process 10140 Loaded symbols for /usr/src/usr.sbin/relayd-current/relayd Reading symbols from /usr/lib/libevent.so.3.1...done. Loaded symbols for /usr/lib/libevent.so.3.1 Reading symbols from /usr/lib/libssl.so.19.0...done. Loaded symbols for /usr/lib/libssl.so.19.0 Reading symbols from /usr/lib/libcrypto.so.22.0...done. Loaded symbols for /usr/lib/libcrypto.so.22.0 Reading symbols from /usr/lib/libutil.so.11.4...done. Loaded symbols for /usr/lib/libutil.so.11.4 Reading symbols from /usr/lib/libc.so.66.2...done. Loaded symbols for /usr/lib/libc.so.66.2 Reading symbols from /usr/libexec/ld.so...done. Loaded symbols for /usr/libexec/ld.so [Switching to thread 1010140] 0x05462804581a in kevent () at stdin:2 2 stdin: No such file or directory. in stdin Current language: auto; currently asm (gdb) continue Continuing. Program received signal SIGPIPE, Broken pipe. 0x05462805eb8a in write () at stdin:2 2 stdin: No such file or directory. in stdin (gdb) Continuing. Program received signal SIGTERM, Terminated. 0x05462804581a in kevent () at stdin:2 2 stdin: No such file or directory. in stdin (gdb) Continuing. Program received signal SIGTERM, Terminated. 0x05462804581a in kevent () at stdin:2 2 in stdin (gdb) Continuing. Program exited normally. (gdb) The program is not being run. It seems that child with pid 32389 receives a SIGSEGV signal, amking the other threads to crash. But why this child crashes in /usr/src/lib/libevent/event.c, line 997 ? If the message is unreadable I can re-send it in attachement From: Chris Cappuccio ch...@nmedia.net To: Bogdan Andu bo...@yahoo.com Cc: Stuart Henderson s...@spacehopper.org; misc@openbsd.org misc@openbsd.org Sent: Friday, August 23, 2013 8:45 AM Subject: Re: relayd crash You can't debug relayd without attaching to all of the processes (you must use multiple simultaneous gdb sessions) Bogdan Andu [bo...@yahoo.com] wrote: ok, I checked out relayd -current, compiled with debug symbols, launched gdb and attached to pfe pid : pwd: /usr/src/usr.sbin/relayd $ sudo gdb ./relayd GNU gdb 6.3 Copyright 2004 Free Software Foundation, Inc. GDB is free software, covered by the GNU General Public License, and you are welcome to change it and/or distribute copies of it under certain conditions. Type show copying to see the conditions. There is absolutely no warranty for GDB.? Type show warranty for details. This GDB was configured as amd64-unknown-openbsd5.3... (gdb) attach 27726 Attaching to program
Re: relayd crash
Here are the results of running relayd in debug mode, attached to pid of pfe child: pwd: /usr/src/usr.sbin/relayd $ sudo gdb ./relayd GNU gdb 6.3 Copyright 2004 Free Software Foundation, Inc. GDB is free software, covered by the GNU General Public License, and you are welcome to change it and/or distribute copies of it under certain conditions. Type show copying to see the conditions. There is absolutely no warranty for GDB. Type show warranty for details. This GDB was configured as amd64-unknown-openbsd5.3... (gdb) attach 28300 Attaching to program: /usr/src/usr.sbin/relayd/relayd, process 28300 Loaded symbols for /usr/src/usr.sbin/relayd/relayd Reading symbols from /usr/lib/libevent.so.3.1...done. Loaded symbols for /usr/lib/libevent.so.3.1 Reading symbols from /usr/lib/libssl.so.19.0...done. Loaded symbols for /usr/lib/libssl.so.19.0 Reading symbols from /usr/lib/libcrypto.so.22.0...done. Loaded symbols for /usr/lib/libcrypto.so.22.0 Reading symbols from /usr/lib/libutil.so.11.4...done. Loaded symbols for /usr/lib/libutil.so.11.4 Reading symbols from /usr/lib/libc.so.66.2...done. Loaded symbols for /usr/lib/libc.so.66.2 Reading symbols from /usr/libexec/ld.so...done. Loaded symbols for /usr/libexec/ld.so [Switching to thread 1028300] 0x1c765263481a in kevent () at stdin:2 2      stdin: No such file or directory.        in stdin Current language: auto; currently asm (gdb) continue Continuing. Program exited normally. (gdb) The program is not being run. (gdb) The program is not being run. (gdb) The program is not being run. (gdb) The program is not being run. (gdb) The program is not being run. (gdb) The program is not being run. (gdb) The program is not being run. (gdb) The program is not being run. (gdb) The program is not being run. (gdb) The program is not being run. (gdb) The program is not being run. (gdb) The program is not being run. (gdb) The program is not being run. (gdb) The program is not being run. (gdb) The program is not being run. (gdb) The program is not being run. (gdb) The program is not being run. (gdb) The program is not being run. (gdb) The program is not being run. (gdb) The program is not being run. (gdb) But the daemon relayd crashed. Bogdan From: Stuart Henderson s...@spacehopper.org To: Bogdan Andu bo...@yahoo.com Cc: misc@openbsd.org misc@openbsd.org Sent: Wednesday, August 21, 2013 6:06 PM Subject: Re: relayd crash On 2013/08/21 06:32, Bogdan Andu wrote: [Switching to thread 1027726] 0x029e250da81a in kevent () at stdin:2 2   stdin: No such file or directory.     in stdin Current language: auto; currently asm (gdb) oh, I forgot, you will need to continue here ;) launched qualys scan and now waiting for crash to occur - will take 2 hours. Unfortunately the relayd patch branch failed with the same error, and that is why I tried relayd -current I'll come back with crash results ââââââââââââââââââââââââââ ââââââââââââââââââââââââââ âââââââââââââââââââ From: Stuart Henderson s...@spacehopper.org To: misc@openbsd.org Sent: Wednesday, August 21, 2013 3:51 PM Subject: Re: relayd crash On 2013-08-21, Bogdan Andu bo...@yahoo.com wrote: I tried relayd -current and crashed with the error: pwd: /usr/src/usr.sbin/relayd $ sudo ./relayd -f /etc/relayd.conf -d . A LOT OF LOGS relay www_ssl, session 1368 (1 active), 0, 64.39.111.90 - :0, SSL accept error relay www_ssl, session 1369 (1 active), 0, 64.39.111.90 - 127.0.0.1:8080, done pfe exiting, pid 8360 lost child: relay terminated; signal 11 hce exiting, pid 18726 parent terminating, pid 2704 Build it with debug symbols (make clean, make DEBUG=-g install) and re-run it, then you can attach gdb to the pfe process (use ps to find the correct pid and 'gdb /usr/sbin/relayd' and 'attach pid'. Now when you trigger the crash you should be able to get a backtrace.
Re: relayd crash
You can't debug relayd without attaching to all of the processes (you must use multiple simultaneous gdb sessions) Bogdan Andu [bo...@yahoo.com] wrote: ok, I checked out relayd -current, compiled with debug symbols, launched gdb and attached to pfe pid : pwd: /usr/src/usr.sbin/relayd $ sudo gdb ./relayd GNU gdb 6.3 Copyright 2004 Free Software Foundation, Inc. GDB is free software, covered by the GNU General Public License, and you are welcome to change it and/or distribute copies of it under certain conditions. Type show copying to see the conditions. There is absolutely no warranty for GDB.? Type show warranty for details. This GDB was configured as amd64-unknown-openbsd5.3... (gdb) attach 27726 Attaching to program: /usr/src/usr.sbin/relayd/relayd, process 27726 Loaded symbols for /usr/src/usr.sbin/relayd/relayd Reading symbols from /usr/lib/libevent.so.3.1...done. Loaded symbols for /usr/lib/libevent.so.3.1 Reading symbols from /usr/lib/libssl.so.19.0...done. Loaded symbols for /usr/lib/libssl.so.19.0 Reading symbols from /usr/lib/libcrypto.so.22.0...done. Loaded symbols for /usr/lib/libcrypto.so.22.0 Reading symbols from /usr/lib/libutil.so.11.4...done. Loaded symbols for /usr/lib/libutil.so.11.4 Reading symbols from /usr/lib/libc.so.66.2...done. Loaded symbols for /usr/lib/libc.so.66.2 Reading symbols from /usr/libexec/ld.so...done. Loaded symbols for /usr/libexec/ld.so [Switching to thread 1027726] 0x029e250da81a in kevent () at stdin:2 2?? stdin: No such file or directory. ??? in stdin Current language:? auto; currently asm (gdb) launched qualys scan and now waiting for crash to occur - will take 2 hours. Unfortunately the relayd patch branch failed with the same error, and that is why I tried relayd -current I'll come back with crash results From: Stuart Henderson s...@spacehopper.org To: misc@openbsd.org Sent: Wednesday, August 21, 2013 3:51 PM Subject: Re: relayd crash On 2013-08-21, Bogdan Andu bo...@yahoo.com wrote: I tried relayd -current and crashed with the error: pwd: /usr/src/usr.sbin/relayd $ sudo ./relayd -f /etc/relayd.conf -d . A LOT OF LOGS relay www_ssl, session 1368 (1 active), 0, 64.39.111.90 - :0, SSL accept error relay www_ssl, session 1369 (1 active), 0, 64.39.111.90 - 127.0.0.1:8080, done pfe exiting, pid 8360 lost child: relay terminated; signal 11 hce exiting, pid 18726 parent terminating, pid 2704 Build it with debug symbols (make clean, make DEBUG=-g install) and re-run it, then you can attach gdb to the pfe process (use ps to find the correct pid and 'gdb /usr/sbin/relayd' and 'attach pid'. Now when you trigger the crash you should be able to get a backtrace. -- Semicolons; use them
Re: relayd crash
Bogdan Andu bog09 at yahoo.com writes: machine is OpenBSD 5.3/amd64 GENERIC.MP relayd on 5.3 is buggy. you either need newer version, or backport a fix.
Re: relayd crash
From: Alexey E. Suslikov alexey.susli...@gmail.com To: misc@openbsd.org Sent: Wednesday, August 21, 2013 11:40 AM Subject: Re: relayd crash Bogdan Andu bog09 at yahoo.com writes: machine is OpenBSD 5.3/amd64 GENERIC.MP relayd on 5.3 is buggy. you either need newer version, or backport a fix. so, between the following two options what is the best (given the fact the machine is OpenBSD 5.3)? 1. checkout in 5.3 from current, compile and install on 5.3; 2. checkout in 5.3 relayd with tag OPENBSD_5_2, compile and install in 5.3 Bogdan
Re: relayd crash
Bogdan Andu bog09 at yahoo.com writes: From: Alexey E. Suslikov alexey.suslikov at gmail.com To: misc at openbsd.org Sent: Wednesday, August 21, 2013 11:40 AM Subject: Re: relayd crash Bogdan Andu bog09 at yahoo.com writes: machine is OpenBSD 5.3/amd64 GENERIC.MP relayd on 5.3 is buggy. you either need newer version, or backport a fix. so, between the following two options what is the best (given the fact the machine is OpenBSD 5.3)? 1. checkout in 5.3 from current, compile and install on 5.3; -current is not directly compatible with older releases due to time_t switch. 2. checkout in 5.3 relayd with tag OPENBSD_5_2, compile and install in 5.3 haven't tried by myself, cause I follow -current in most cases. search archives for relayd crash, I remember people cherry picking the diff.
Re: relayd crash
On 2013 Aug 21 (Wed) at 02:16:32 -0700 (-0700), Bogdan Andu wrote: : : From: Alexey E. Suslikov alexey.susli...@gmail.com :To: misc@openbsd.org :Sent: Wednesday, August 21, 2013 11:40 AM :Subject: Re: relayd crash : : :Bogdan Andu bog09 at yahoo.com writes: : : machine is OpenBSD 5.3/amd64 GENERIC.MP : :relayd on 5.3 is buggy. you either need newer version, :or backport a fix. : :so, between the following two options what is the best (given the fact the machine is OpenBSD 5.3)? : :1. checkout in 5.3 from current, compile and install on 5.3; :2. checkout in 5.3 relayd with tag OPENBSD_5_2, compile and install in 5.3 : :Bogdan : Checkout the stable branch on OPENBSD_5_3 as it already has a fix for relayd. http://www.openbsd.org/stable.html -- A long-forgotten loved one will appear soon. Buy the negatives at any price.
Re: relayd crash
I tried relayd -current and crashed with the error: pwd: /usr/src/usr.sbin/relayd $ sudo ./relayd -f /etc/relayd.conf -d . A LOT OF LOGS relay www_ssl, session 1368 (1 active), 0, 64.39.111.90 - :0, SSL accept error relay www_ssl, session 1369 (1 active), 0, 64.39.111.90 - 127.0.0.1:8080, done pfe exiting, pid 8360 lost child: relay terminated; signal 11 hce exiting, pid 18726 parent terminating, pid 2704 now I am trying relayd from patch branch - OPENBSD_5_3 and see what happens relaunched the free qualys scan I let you know the results when the scan finishes From: Peter Hessler phess...@theapt.org To: Bogdan Andu bo...@yahoo.com Cc: Alexey E. Suslikov alexey.susli...@gmail.com; misc@openbsd.org misc@openbsd.org Sent: Wednesday, August 21, 2013 1:45 PM Subject: Re: relayd crash On 2013 Aug 21 (Wed) at 02:16:32 -0700 (-0700), Bogdan Andu wrote: : : From: Alexey E. Suslikov alexey.susli...@gmail.com :To: misc@openbsd.org :Sent: Wednesday, August 21, 2013 11:40 AM :Subject: Re: relayd crash : : :Bogdan Andu bog09 at yahoo.com writes: : : machine is OpenBSD 5.3/amd64 GENERIC.MP : :relayd on 5.3 is buggy. you either need newer version, :or backport a fix. : :so, between the following two options what is the best (given the fact the machine is OpenBSD 5.3)? : :1. checkout in 5.3 from current, compile and install on 5.3; :2. checkout in 5.3 relayd with tag OPENBSD_5_2, compile and install in 5.3 : :Bogdan : Checkout the stable branch on OPENBSD_5_3 as it already has a fix for relayd. http://www.openbsd.org/stable.html -- A long-forgotten loved one will appear soon. Buy the negatives at any price.
Re: relayd crash
On 2013-08-21, Bogdan Andu bo...@yahoo.com wrote: I tried relayd -current and crashed with the error: pwd: /usr/src/usr.sbin/relayd $ sudo ./relayd -f /etc/relayd.conf -d . A LOT OF LOGS relay www_ssl, session 1368 (1 active), 0, 64.39.111.90 - :0, SSL accept error relay www_ssl, session 1369 (1 active), 0, 64.39.111.90 - 127.0.0.1:8080, done pfe exiting, pid 8360 lost child: relay terminated; signal 11 hce exiting, pid 18726 parent terminating, pid 2704 Build it with debug symbols (make clean, make DEBUG=-g install) and re-run it, then you can attach gdb to the pfe process (use ps to find the correct pid and 'gdb /usr/sbin/relayd' and 'attach pid'. Now when you trigger the crash you should be able to get a backtrace.
Re: relayd crash
ok, I checked out relayd -current, compiled with debug symbols, launched gdb and attached to pfe pid : pwd: /usr/src/usr.sbin/relayd $ sudo gdb ./relayd GNU gdb 6.3 Copyright 2004 Free Software Foundation, Inc. GDB is free software, covered by the GNU General Public License, and you are welcome to change it and/or distribute copies of it under certain conditions. Type show copying to see the conditions. There is absolutely no warranty for GDB. Type show warranty for details. This GDB was configured as amd64-unknown-openbsd5.3... (gdb) attach 27726 Attaching to program: /usr/src/usr.sbin/relayd/relayd, process 27726 Loaded symbols for /usr/src/usr.sbin/relayd/relayd Reading symbols from /usr/lib/libevent.so.3.1...done. Loaded symbols for /usr/lib/libevent.so.3.1 Reading symbols from /usr/lib/libssl.so.19.0...done. Loaded symbols for /usr/lib/libssl.so.19.0 Reading symbols from /usr/lib/libcrypto.so.22.0...done. Loaded symbols for /usr/lib/libcrypto.so.22.0 Reading symbols from /usr/lib/libutil.so.11.4...done. Loaded symbols for /usr/lib/libutil.so.11.4 Reading symbols from /usr/lib/libc.so.66.2...done. Loaded symbols for /usr/lib/libc.so.66.2 Reading symbols from /usr/libexec/ld.so...done. Loaded symbols for /usr/libexec/ld.so [Switching to thread 1027726] 0x029e250da81a in kevent () at stdin:2 2 stdin: No such file or directory. in stdin Current language: auto; currently asm (gdb) launched qualys scan and now waiting for crash to occur - will take 2 hours. Unfortunately the relayd patch branch failed with the same error, and that is why I tried relayd -current I'll come back with crash results From: Stuart Henderson s...@spacehopper.org To: misc@openbsd.org Sent: Wednesday, August 21, 2013 3:51 PM Subject: Re: relayd crash On 2013-08-21, Bogdan Andu bo...@yahoo.com wrote: I tried relayd -current and crashed with the error: pwd: /usr/src/usr.sbin/relayd $ sudo ./relayd -f /etc/relayd.conf -d . A LOT OF LOGS relay www_ssl, session 1368 (1 active), 0, 64.39.111.90 - :0, SSL accept error relay www_ssl, session 1369 (1 active), 0, 64.39.111.90 - 127.0.0.1:8080, done pfe exiting, pid 8360 lost child: relay terminated; signal 11 hce exiting, pid 18726 parent terminating, pid 2704 Build it with debug symbols (make clean, make DEBUG=-g install) and re-run it, then you can attach gdb to the pfe process (use ps to find the correct pid and 'gdb /usr/sbin/relayd' and 'attach pid'. Now when you trigger the crash you should be able to get a backtrace.
Re: relayd crash
On 2013/08/21 06:32, Bogdan Andu wrote: [Switching to thread 1027726] 0x029e250da81a in kevent () at stdin:2 2 stdin: No such file or directory. in stdin Current language: auto; currently asm (gdb) oh, I forgot, you will need to continue here ;) launched qualys scan and now waiting for crash to occur - will take 2 hours. Unfortunately the relayd patch branch failed with the same error, and that is why I tried relayd -current I'll come back with crash results ━━━ From: Stuart Henderson s...@spacehopper.org To: misc@openbsd.org Sent: Wednesday, August 21, 2013 3:51 PM Subject: Re: relayd crash On 2013-08-21, Bogdan Andu bo...@yahoo.com wrote: I tried relayd -current and crashed with the error: pwd: /usr/src/usr.sbin/relayd $ sudo ./relayd -f /etc/relayd.conf -d . A LOT OF LOGS relay www_ssl, session 1368 (1 active), 0, 64.39.111.90 - :0, SSL accept error relay www_ssl, session 1369 (1 active), 0, 64.39.111.90 - 127.0.0.1:8080, done pfe exiting, pid 8360 lost child: relay terminated; signal 11 hce exiting, pid 18726 parent terminating, pid 2704 Build it with debug symbols (make clean, make DEBUG=-g install) and re-run it, then you can attach gdb to the pfe process (use ps to find the correct pid and 'gdb /usr/sbin/relayd' and 'attach pid'. Now when you trigger the crash you should be able to get a backtrace.
Re: relayd crash
yes. we waiting for tests to finish - I launched 3 concurrent tests. tomorrow I'll give you the results. From: Stuart Henderson s...@spacehopper.org To: Bogdan Andu bo...@yahoo.com Cc: misc@openbsd.org misc@openbsd.org Sent: Wednesday, August 21, 2013 6:06 PM Subject: Re: relayd crash On 2013/08/21 06:32, Bogdan Andu wrote: [Switching to thread 1027726] 0x029e250da81a in kevent () at stdin:2 2Â Â Â stdin: No such file or directory. Â Â Â Â in stdin Current language:Â auto; currently asm (gdb) oh, I forgot, you will need to continue here ;) launched qualys scan and now waiting for crash to occur - will take 2 hours. Unfortunately the relayd patch branch failed with the same error, and that is why I tried relayd -current I'll come back with crash results ââââââââââââââââââââââââââ ââââââââââââââââââââââââââ âââââââââââââââââââ From: Stuart Henderson s...@spacehopper.org To: misc@openbsd.org Sent: Wednesday, August 21, 2013 3:51 PM Subject: Re: relayd crash On 2013-08-21, Bogdan Andu bo...@yahoo.com wrote: I tried relayd -current and crashed with the error: pwd: /usr/src/usr.sbin/relayd $ sudo ./relayd -f /etc/relayd.conf -d . A LOT OF LOGS relay www_ssl, session 1368 (1 active), 0, 64.39.111.90 - :0, SSL accept error relay www_ssl, session 1369 (1 active), 0, 64.39.111.90 - 127.0.0.1:8080, done pfe exiting, pid 8360 lost child: relay terminated; signal 11 hce exiting, pid 18726 parent terminating, pid 2704 Build it with debug symbols (make clean, make DEBUG=-g install) and re-run it, then you can attach gdb to the pfe process (use ps to find the correct pid and 'gdb /usr/sbin/relayd' and 'attach pid'. Now when you trigger the crash you should be able to get a backtrace.
Re: Relayd crash on reload
Le 16/07/2013 15:53, Mathieu BLANC a écrit : Hi ! I have read several mails/bug in the mailing list about reloading relayd. But i didn't understand if all the bugs were fixed or not ? [...] If i launch the daemon with relayd -d -vvv, and relayctl reload, i have this error : parent_sig_handler: reload requested with SIGHUP parent_reload: level 0 config file /etc/relayd.conf relayd in free(): error: chunk is already free 0x139b3e2e6400 init_filter: filter init done init_tables: created 0 tables flush_table: flushed table std_https1 kill_tables: deleted 1 tables flush_rulesets: flushed rules pfe exiting, pid 1270 lost child: hce terminated; signal 6 relay exiting, pid 18390 relay exiting, pid 17435 relay exiting, pid 579 relay exiting, pid 7823 relay exiting, pid 25910 parent terminating, pid 23068 I know that relayd can't be reloaded if we have relays with SSL, but i've just a redirect (with check https...). I've tested with check tcp (and not check https /), and it seems that reload is OK with this ! Am I missing something ? Thanks in advance ! Sebastian (benoit-li...@fb12.de) asked me to test this patch : http://www.openbsd.org/cgi-bin/cvsweb/src/usr.sbin/relayd/ssl.c.diff?r1=1.18;r2=1.19;sortby=date I've tried this morning, and relayctl reload doesn't crash anymore. Thanks to him ;) -- Mathieu
Re: Relayd crash on reload
Mathieu BLANC mathieu.blanc at smile.fr writes: I have read several mails/bug in the mailing list about reloading relayd. But i didn't understand if all the bugs were fixed or not ? you need to try out -current relayd first.