OpenBSC November 2015

openbsc@lists.osmocom.org

18 participants
50 discussions

OpenBSC compile error --> tests --> GTP hub

by Sipos Csaba

Dear Neels, Today just grabbed the latest OpenBSC master, and got this compile error: Making all in gtphub make[3]: Entering directory `/root/new_osmocom/openbsc/openbsc/tests/gtphub' CC gtphub_test.o CC ../../src/gprs/gtphub.o ../../src/gprs/gtphub.c:2191:1: fatal error: opening dependency file .deps/../../src/gprs/gtphub.Tpo: No such file or directory } ^ compilation terminated. make[3]: *** [../../src/gprs/gtphub.o] Error 1 make[3]: Leaving directory `/root/new_osmocom/openbsc/openbsc/tests/gtphub' make[2]: *** [all-recursive] Error 1 make[2]: Leaving directory `/root/new_osmocom/openbsc/openbsc/tests' make[1]: *** [all-recursive] Error 1 make[1]: Leaving directory `/root/new_osmocom/openbsc/openbsc' make: *** [all] Error 2 Do you have any idea why is this happening? Regards, Csaba

9 years, 6 months

[PATCH 1/3] gtphub: lose obsolete comment.

by Neels Hofmeyr

The del_cb is now also used for ares (GGSN resolution) timeouts, and expiry is anyway separated from nr_map, so this comment is void. Sponsored-by: On-Waves ehi --- openbsc/include/openbsc/gtphub.h | 2 -- 1 file changed, 2 deletions(-) diff --git a/openbsc/include/openbsc/gtphub.h b/openbsc/include/openbsc/gtphub.h index 49d685c..68cf90e 100644 --- a/openbsc/include/openbsc/gtphub.h +++ b/openbsc/include/openbsc/gtphub.h @@ -248,8 +248,6 @@ int expiry_tick(struct expiry *exq, time_t now); * Mapping allocation and a del_cb are provided by the caller. If del_cb is * NULL, no deallocation will be done (allowing statically allocated entries). */ -/* TODO at some point I thought the allocation & del_cb complexity was - * needed/helpful, but by now it seems like overkill. Maybe lose that again. */ typedef int nr_t; -- 2.1.4

9 years, 6 months

Meas_web and meas_json

by Sipos Csaba

Hi Alexander, Just wanted to thank you for the web version of meas_vis utility, and wanted to confirm that it works very well with Nokia Site family (E1 based) and the USRP B200 too :-) If someone missed it, the web version can be found here: https://github.com/fairwaves/meas_web It would make life easier if the meas_json utility would make it to the master branch: http://cgit.osmocom.org/openbsc/log/?h=achemeris/meas_json Is there a reason it is not being merged? Regards, Csaba

9 years, 6 months

<0004> abis_rsl.c:1423 BTS 0 CHAN RQD: no resources for SDCCH 0xf

by Sipos Csaba

Dear List, I am in the process of further updating the following Wiki article: The goal is to have a clean, up to date manual which tells in detail how to set up OpenBSC. In the light of recent changes I decided to change this already working setup to use the "master" branches of every involved Osmocom project. It is finally possible to compile and run every project after the proper dependencies are met. But some new problem popped up I haven't seen before. After everything is started up (OpenBSC, GGSN, SGSN, LCR, TRX, and PCU), when a phone tries to RACH, this is what I get in the BSC log: Sat Nov 28 14:09:08 2015 <0004> abis_rsl.c:1423 BTS 0 CHAN RQD: no resources for SDCCH 0xf Sat Nov 28 14:09:19 2015 <0004> abis_rsl.c:1423 BTS 0 CHAN RQD: no resources for SDCCH 0x6 Sat Nov 28 14:09:30 2015 <0004> abis_rsl.c:1423 BTS 0 CHAN RQD: no resources for SDCCH 0x6 On the BTS side, this is what I get: <0006> scheduler.c:1186 Received Access Burst on RACH fn=9868 toa=5.88 <0006> scheduler.c:1192 Received bad AB frame at fn=9868 (25/51) <0006> scheduler.c:1186 Received Access Burst on RACH fn=52459 toa=-1.50 <0006> scheduler.c:1192 Received bad AB frame at fn=52459 (31/51) <0006> scheduler.c:1186 Received Access Burst on RACH fn=67956 toa=-0.06 <0006> scheduler.c:1192 Received bad AB frame at fn=67956 (24/51) Because I used exactly the same config files I used before when it all worked, I analyzed the logs a littlebit more. In the BTS log, during the OML setup this is what I found: <0001> oml.c:802 ADM state already was Unlocked <0001> oml.c:256 OC=BASEBAND-TRANSCEIVER INST=(00,00,ff) OPER STATE NULL -> Enabled <0001> oml.c:217 OC=BASEBAND-TRANSCEIVER INST=(00,00,ff) Tx STATE CHG REP <000b> trx_if.c:366 Response message 'RSP ERR 1' does not match command message 'CMD SETBSIC 62' !!!!!! <0001> oml.c:1059 Rx IPA RSL CONNECT IP=127.0.0.1 PORT=3003 STREAM=0x00 <0011> input/ipa.c:129 connection done. <0011> input/ipaccess.c:693 received ID get <0000> rsl.c:279 Tx RSL RF RESource INDication Maybe the recent changes in the BSIC/TSC handling part is causing this? Anyway it seems that the BSIC I set in the BSC is the correct one the BTS tries to configure during the timeslot, transceiver and BTS setup. I attached the corresponding OML trace, and the logs. If someone has any ideas, please let me know. Regards, Csaba

9 years, 7 months

[PATCH 1/8] debug: Add DSS category for supplementary service

by Sergey.Kostanbaev

From: Sergey Kostanbaev <sergey.kostanbaev(a)gmail.com> This series of patches add support for external USSD processing via simplified MAP-like protocol over SUP socket. As an example of a such external application simple USSD to SIP proxy is provided (it targets TS 124 390) --- openbsc/include/openbsc/debug.h | 1 + openbsc/src/libcommon/debug.c | 6 ++++++ 2 files changed, 7 insertions(+) diff --git a/openbsc/include/openbsc/debug.h b/openbsc/include/openbsc/debug.h index 19d8fc2..10b0f46 100644 --- a/openbsc/include/openbsc/debug.h +++ b/openbsc/include/openbsc/debug.h @@ -33,6 +33,7 @@ enum { DCTRL, DSMPP, DFILTER, + DSS, Debug_LastEntry, }; diff --git a/openbsc/src/libcommon/debug.c b/openbsc/src/libcommon/debug.c index 7fb3ecb..6536944 100644 --- a/openbsc/src/libcommon/debug.c +++ b/openbsc/src/libcommon/debug.c @@ -165,6 +165,12 @@ static const struct log_info_cat default_categories[] = { .description = "BSC/NAT IMSI based filtering", .enabled = 1, .loglevel = LOGL_DEBUG, }, + [DSS] = { + .name = "DSS", + .description = "Layer3 Supplementary Service (SS)", + .color = "\033[0;33m", + .enabled = 1, .loglevel = LOGL_DEBUG, + }, }; enum log_filter { -- 1.9.1

9 years, 7 months

[PATCH 1/3] mncc: Implement helper functions to convert MNCC cause to a string.

by Alexander Chemeris

This is the first patch of the series. The first two patches are improvements of the log output and are independent of the 3rd patch in terms of code - proposed logging improvements just make it easier to debug issues with MNCC cause codes. The code is also a part of the achemeris/mncc_cause_fixes_master branch. -- Regards, Alexander Chemeris. CEO, Fairwaves, Inc. https://fairwaves.co

9 years, 7 months

[PATCH 1/4] core: RELEASE COMPLETE can be without any payload

by Sergey.Kostanbaev

From: Sergey Kostanbaev <Sergey.Kostanbaev(a)gmail.com> This series of patches supposed to add support of USSD-menu style message handling with language support. --- src/gsm/gsm0480.c | 6 ------ 1 file changed, 6 deletions(-) diff --git a/src/gsm/gsm0480.c b/src/gsm/gsm0480.c index 952604b..9bd54fa 100644 --- a/src/gsm/gsm0480.c +++ b/src/gsm/gsm0480.c @@ -248,12 +248,6 @@ int gsm0480_decode_ss_request(const struct gsm48_hdr *hdr, uint16_t len, struct ss_request *req) { int rc = 0; - - if (len < sizeof(*hdr) + 2) { - LOGP(0, LOGL_DEBUG, "SS Request is too short.\n"); - return 0; - } - if ((hdr->proto_discr & 0x0f) == GSM48_PDISC_NC_SS) { req->transaction_id = hdr->proto_discr & 0x70; rc = parse_ss(hdr, len, req); -- 1.9.1

9 years, 7 months

[PATCH 3/3] mncc: Send "Dest OOO" cause in case of a link radio failure.

by Alexander Chemeris

Previously we were sending a generic "Resource unavailable" cause code making it impossible to distinguish real error cases from a regular radio link failure. This was the reason for many "unknown" call errors we've seen at Rhizomatica installations. Now they are properly classified as non-erroneous call failures. The code is also a part of the achemeris/mncc_cause_fixes_master branch. -- Regards, Alexander Chemeris. CEO, Fairwaves, Inc. https://fairwaves.co

9 years, 7 months

[PATCH 2/3] mncc: Log CC cause when sending a primitive to MNCC.

by Alexander Chemeris

The code is also a part of the achemeris/mncc_cause_fixes_master branch. -- Regards, Alexander Chemeris. CEO, Fairwaves, Inc. https://fairwaves.co

9 years, 7 months

HAVE_LIBCARES

by Neels Hofmeyr

About below commit: Sorry / Thanks! I forgot to move gtphub into the HAVE_LIBCARES conditional when I added ares... The gtphub tests wrap the ares functions away, so they build without libc-ares. But does it make sense to enable the gtphub test when the binary isn't built? ~Neels > commit 97b6bfa996ed1745d27c5598186c4583f7e90861 > Author: Harald Welte <laforge(a)gnumonks.org> > Date: Sat Nov 21 13:16:08 2015 +0100 > > Fix compilation with no libc-ares present on the system > > This build failure was introduced with the OAP and gtphub changes. > > diff --git a/openbsc/src/gprs/Makefile.am b/openbsc/src/gprs/Makefile.am > index 5212c67..5e115eb 100644 > --- a/openbsc/src/gprs/Makefile.am > +++ b/openbsc/src/gprs/Makefile.am > @@ -11,9 +11,8 @@ noinst_HEADERS = gprs_sndcp.h > bin_PROGRAMS = osmo-gbproxy > > if HAVE_LIBGTP > -bin_PROGRAMS += osmo-gtphub > if HAVE_LIBCARES > -bin_PROGRAMS += osmo-sgsn > +bin_PROGRAMS += osmo-sgsn osmo-gtphub > endif > endif > > diff --git a/openbsc/tests/Makefile.am b/openbsc/tests/Makefile.am > index 995660a..04b8e34 100644 > --- a/openbsc/tests/Makefile.am > +++ b/openbsc/tests/Makefile.am > @@ -13,7 +13,10 @@ SUBDIRS += smpp > endif > > if HAVE_LIBGTP > -SUBDIRS += sgsn oap gtphub > +SUBDIRS += gtphub > +if HAVE_LIBCARES > +SUBDIRS += sgsn oap > +endif > endif >

9 years, 7 months

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

OpenBSC November 2015