Am Fri, 23 Jul 2010 15:25:32 +0800 schrieb Holger Hans Peter Freyther holger@freyther.de:

On 07/23/2010 02:52 PM, Luca Bertoncello wrote:

...

Segmentation fault

I can not comment on the other bits (besides Nordin having had issues as well), the segfault is an issue I introduced. In the coredump uploaded to the On Waves servers it is crashing in my drop OML when RSL failed code and I plan to fix that once I have my private stuff settled.

I tried a very old version I have in a old PC (compiled on 16/10/2009). Now there is no segmentation fault, but it does not work...

I get:

netzing@debian503:~/openbsc/openbsc/src$ ./ipaccess-config -u 151/0/0 -o 192.168.1.18 -r 192.168.1.108 ipaccess-config (C) 2009 by Harald Welte This is FREE SOFTWARE with ABSOLUTELY NO WARRANTY

Trying to connect to ip.access BTS ... BTS disappeared, dead socket Lost some E1 TEI link

and it waits until the end of the eternity (I suppose). Is it possible, that the problem is on the BTS? How can I be sure?

Thanks a lot!

Am Fri, 23 Jul 2010 10:02:57 +0200 schrieb Nordin bouchtaoui@gmail.com:

First of all, how did you configure? You wrote:

lucabert at Luca https://lists.gnumonks.org/mailman/listinfo/openbsc:/tmp$ ipaccess-config -u 151/0/0 -o 192.168.1.18 192.168.1.108 ipaccess-config (C) 2009-2010 by Harald Welte and others This is FREE SOFTWARE with ABSOLUTELY NO WARRANTY

Trying to connect to ip.access BTS ... Lost some E1 TEI link Lost some E1 TEI link Segmentation fault

I think it's best to split it in two steps: ./ipaccess-config -u 151/0/0 bts-ip ./ipaccess-config -o bsc-ip bts-ip

I tried it, too! Same error... :(

You need also to power off than power on and than try it again.

And I tried this, too. Unfortunately no better results...

My BTS kind of got lock as TRX (as a slave of master BTS) and can't send BCCH anymore. This problem occured since I tried to modify the bts-id, that's the middle number of the id.

I recommend you not to do the same, otherwise your bts won't work properly.

OK, thanks for the advice, but I first need to configure the BTS to speech with my PC and my OpenBSC...

Thanks

Am Fri, 23 Jul 2010 16:40:24 +0800 schrieb Holger Hans Peter Freyther holger@freyther.de:

As usual for being able to debug problems:

Segmentation Fault -> Attach a Backtrace Some Protocol Error -> Attach a Packetdump.

You're right! Please excuse me...

Attached the PCAP file and follow the backtrace:

lucabert@Luca:~/BSC/openbsc/openbsc/src/ipaccess$ gdb ./ipaccess-config /tmp/core GNU gdb 6.8-debian Copyright (C) 2008 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 "i486-linux-gnu"...

warning: Can't read pathname for load map: Input/output error. Reading symbols from /usr/local/lib/libosmocore.so.0...done. Loaded symbols for /usr/local/lib/libosmocore.so.0 Reading symbols from /lib/tls/i686/cmov/libdl.so.2...done. Loaded symbols for /lib/tls/i686/cmov/libdl.so.2 Reading symbols from /usr/lib/libdbi.so.0...done. Loaded symbols for /usr/lib/libdbi.so.0 Reading symbols from /lib/tls/i686/cmov/libcrypt.so.1...done. Loaded symbols for /lib/tls/i686/cmov/libcrypt.so.1 Reading symbols from /lib/tls/i686/cmov/libc.so.6...done. Loaded symbols for /lib/tls/i686/cmov/libc.so.6 Reading symbols from /lib/ld-linux.so.2...done. Loaded symbols for /lib/ld-linux.so.2 Reading symbols from /lib/tls/i686/cmov/libm.so.6...done. Loaded symbols for /lib/tls/i686/cmov/libm.so.6 Core was generated by `ipaccess-config -u 151/0/0 -o 192.168.1.18 192.168.1.108'. Program terminated with signal 11, Segmentation fault. [New process 27115] #0 bsc_unregister_fd (fd=0x80c59f4) at ../include/osmocore/linuxlist.h:107 107 next->prev = prev; (gdb)

Thanks

Am Fri, 23 Jul 2010 10:58:05 +0200 schrieb Sylvain Munaut 246tnt@gmail.com:

That's not much of a backtrace

Please, use the 'bt' command to get the full backtrace. Using the 'bt full' command if even better. And to further improve things you could also make sure the libosmocore and openbsc were compiled with full debugging symbol by including -ggdb if the CFLAGS/LDFLAGS

OK, here more information:

(gdb) bt full #0 bsc_unregister_fd (fd=0x80c59f4) at ../include/osmocore/linuxlist.h:107 No locals. #1 0x08055c67 in ipaccess_drop_rsl (trx=0x806d880) at input/ipaccess.c:427 ts = (struct e1inp_ts *) 0x80c548c #2 0x08055d3f in ipaccess_drop_oml (bts=0x806cc38) at input/ipaccess.c:368 trx = (struct gsm_bts_trx *) 0x806d880 ts = <value optimized out> line = (struct e1inp_line *) 0x80c4ef8 #3 0x08055d97 in ipaccess_drop (ts=0x0, bfd=<value optimized out>) at input/ipaccess.c:400 link = (struct e1inp_sign_link *) 0x0 bts_nr = 0 #4 0x080563ca in ipaccess_fd_cb (bfd=0x80c5470, what=<value optimized out>) at input/ipaccess.c:451 line = <value optimized out> ts_nr = 1 rc = <value optimized out> #5 0xb76e8073 in bsc_select_main (polling=0) at select.c:119 flags = 1 ufd = (struct bsc_fd *) 0x80c5470 tmp = (struct bsc_fd *) 0xb76f93a0 readset = {__fds_bits = {0 <repeats 32 times>}} writeset = {__fds_bits = {0 <repeats 32 times>}} exceptset = {__fds_bits = {0 <repeats 32 times>}} work = 0 rc = <value optimized out> no_time = {tv_sec = 0, tv_usec = 0} #6 0x0804af42 in main (argc=6, argv=0xbffbde84) at ipaccess-config.c:861 bts = (struct gsm_bts *) 0x806cc38 sin = {sin_family = 2, sin_port = 48651, sin_addr = {s_addr = 1812048064}, sin_zero = "\000\000\000\000\000\000\000"} rc = 135027188 option_index = 0 stream_id = <value optimized out> stderr_target = <value optimized out> long_options = {{name = 0x8061169 "unit-id", has_arg = 1, flag = 0x0, val = 117}, {name = 0x8061171 "oml-ip", has_arg = 1, flag = 0x0, val = 111}, { name = 0x8061178 "ip-address", has_arg = 1, flag = 0x0, val = 105}, {name = 0x8061183 "ip-gateway", has_arg = 1, flag = 0x0, val = 103}, {name = 0x806118e "restart", has_arg = 0, flag = 0x0, val = 114}, {name = 0x8061196 "nvram-flags", has_arg = 1, flag = 0x0, val = 110}, {name = 0x80611a2 "nvattr-set", has_arg = 1, flag = 0x0, val = 83}, {name = 0x80611ad "nvattr-unset", has_arg = 1, flag = 0x0, val = 85}, {name = 0x80611ba "help", has_arg = 0, flag = 0x0, val = 104}, {name = 0x80611bf "listen", has_arg = 1, flag = 0x0, val = 108}, {name = 0x80611c6 "stream-id", has_arg = 1, flag = 0x0, val = 115}, {name = 0x80611d0 "software", has_arg = 1, flag = 0x0, val = 100}, { name = 0x806106d "firmware", has_arg = 1, flag = 0x0, val = 102}, {name = 0x80611d9 "write-firmware", has_arg = 0, flag = 0x0, val = 119}, {name = 0x0, has_arg = 0, flag = 0x0, val = 0}}

Thanks

On 07/23/2010 04:47 PM, Luca Bertoncello wrote:

Attached the PCAP file and follow the backtrace:

ip.src == 192.168.1.108 || ip.dst == 192.168.1.108

doesn't really show much information...

Your PC opens the connection, it is acked The BTS sends sends a IPA Identity ACK, we send one. The BTS closes the connections.

Now the question is why, and I don't know. :)

Could is the DSC1800 KO? The device is NOT new, but it comes after more experiments from a university...

Is there a possibility to check, if the DSC1800 has a problem?

Did you do a full factory reset with a reset dongle ?

You can try the ipaccess-telnet utility (search the ml or the wiki for details) and see what the console says (if anything).

Sylvain

...

...

Is there a possibility to check, if the DSC1800 has a problem?

Did you do a full factory reset with a reset dongle ?

No, I didn't... What is this "reset dongle"?

See http://lists.gnumonks.org/pipermail/openbsc/2009-August/000804.html

The dongle can be emulated with a carefully placed conductive pin (screwdriver, needed, probe, whatever you have laying around :)

Sylvain

...

http://lists.gnumonks.org/pipermail/openbsc/2009-August/000804.html

The dongle can be emulated with a carefully placed conductive pin (screwdriver, needed, probe, whatever you have laying around :)

Unfortunately I don't have a dongle and I don't have the possibility to make one...

That's what I noted above, you don't _need_ a dongle ... you just need to follow the procedure and instead of the dongle, just short the pin 9 & 10 of TIB-IN with anything metallic/conductive.

Sylvain

...

That's what I noted above, you don't _need_ a dongle ... you just need to follow the procedure and instead of the dongle, just short the pin 9 & 10 of TIB-IN with anything metallic/conductive.

OK, I understand, but without a USB-connector it is very hard to do without damages...

???

Have you _looked_ at the connector. It has nothing to do with USB and 9/10 are the twi pins on the exterior ... can't miss them (unless you're half-blind or have any other condition obviously ...)

I don't want to try. The BTS has to give back to the university in the next days...

well, there is always the 'It was like that when I got it' excuse :)

On 23-7-2010 12:48, Luca Bertoncello wrote:

Am Fri, 23 Jul 2010 12:41:28 +0200 schrieb Sylvain Munaut246tnt@gmail.com:

...

...

OK, I understand, but without a USB-connector it is very hard to do without damages...

???

Have you _looked_ at the connector. It has nothing to do with USB and 9/10 are the twi pins on the exterior ... can't miss them (unless you're half-blind or have any other condition obviously ...)

Then I didn't understood which port do you mean...

Now I found the connector you mean. It's like a RJ45, but with 10 pins. Isn't it?

Now which are the PINs 9 and 10? The last two LEFT or the last two RIGHT?

Thanks!

----------------------------- | U U U U U U U U U U | | | | | | 0 .. 9 10 | | | | | |___________ ___________| |_____|

I hope this helps :)

Than you haven't shortcutted the two pins well, cause if you did, after 30 seconds it flashes red light. I had that in the first time too, until I found a way to shortcut it well. I used a screwdriver, just try it again, it should work. Don't worry, it won't damage a lot :p

On 23-7-2010 13:09, Luca Bertoncello wrote:

Am Fri, 23 Jul 2010 13:03:53 +0200 schrieb Nordinbouchtaoui@gmail.com:

...

...

Now I found the connector you mean. It's like a RJ45, but with 10 pins. Isn't it?

Now which are the PINs 9 and 10? The last two LEFT or the last two RIGHT?

Thanks!

---

| U U U U U U U U U U | | | | | | 0 .. 9 10 | | | | | |___________ ___________| |_____|

I hope this helps :)

Unfortunately not...

I tried to short PIN9+10, and wait until the LED flashes (more than 2 mins), then I powered off and on the BTS.

It does not seems it was resetted...

Other suggestions?

On 23-7-2010 13:26, Luca Bertoncello wrote:

Am Fri, 23 Jul 2010 13:15:32 +0200 schrieb Nordinbouchtaoui@gmail.com:

...

Than you haven't shortcutted the two pins well, cause if you did, after 30 seconds it flashes red light. I had that in the first time too, until I found a way to shortcut it well. I used a screwdriver, just try it again, it should work. Don't worry, it won't damage a lot :p

Unfortunately not... I just tried again. After ~30 seconds the LED flashes, but as I powered on it has the same configured data and I can't configure it...

Are you sure? You can use ipaccess-find to see if the id is modified. If I remember it will be something like 65xxx/0/0

Now I have this idea: I supply the power using PoE. Maybe have I to supply the power not with PoE?

I don't see any difference, If you see the BTS's LED flashes, it's working. But maybe you should use the normal power supplier that comes with the BTS. It's possible the PoE you use is not strong enough.

Thanks

On Fri, Jul 23, 2010 at 11:53:52AM +0200, Luca Bertoncello wrote:

Am Fri, 23 Jul 2010 11:49:23 +0200 schrieb Sylvain Munaut 246tnt@gmail.com:

...

See http://lists.gnumonks.org/pipermail/openbsc/2009-August/000804.html

The dongle can be emulated with a carefully placed conductive pin (screwdriver, needed, probe, whatever you have laying around :)

Unfortunately I don't have a dongle and I don't have the possibility to make one...

Luca, it is not the responsibility of this community mailing list to fix your problems at work. If NETZING has not bought a reset dongle and has no ability to manufacture one (which I seriously doubt, I have seen the NETZING electronics lab before) - then you will simply have to tell NETZING that this device will have to be bought or manufactured.

If you are on a short deadline than I am sorry to hear this, but you will have to solve the problem somehow, and it is not the responsibility of the people on this list to do so.

Also, please spend more time analyzing problems and investigating possible resolutions. Your follow-up mails regarding a USB connector are showing that you have neither studied the ip.access documentation (that NETZING has), nor our wiki in detail before contacting the list.

Please use this list after you have exhausted all other resources.

Regards, Harald.