Actions
Bug #5765
closedosmo-bts-trx crashes when using gsmtap-sapi pdtch
Start date:
11/11/2022
Due date:
% Done:
0%
Spec Reference:
Description
config looks like this:
e1_input e1_line 0 driver ipa e1_line 0 port 0 no e1_line 0 keepalive cpu-sched policy rr 5 phy 0 osmotrx ip local 127.0.0.1 osmotrx ip remote 127.0.0.1 osmotrx base-port local 5800 osmotrx base-port remote 5700 osmotrx fn-advance 2 osmotrx rts-advance 3 instance 0 osmotrx rx-gain 1 osmotrx tx-attenuation oml bts 0 band GSM900 ipa unit-id 1 0 oml remote-ip 127.0.0.1 rtp jitter-buffer 100 rtp port-range 16384 17407 paging queue-size 200 paging lifetime 0 gsmtap-remote-host 127.0.0.1 min-qual-rach 50 min-qual-norm -5 max-ber10k-rach 1707 supp-meas-info toa256 smscb queue-max-length 15 smscb queue-target-length 2 smscb queue-hysteresis 2 osmux use off local-ip 127.0.0.1 batch-factor 4 batch-size 1472 port 1984 dummy-padding off trx 0 power-ramp max-initial 0 mdBm power-ramp step-size 1000000 mdB power-ramp step-interval 1 ms-power-control osmo ta-control interval 0 phy 0 instance 0 end
I can enable gsmtap for various other SAPI without any crash, but as soon as I enable gsmtap-sapi for pdtch, osmo-bts-trx crashes:
OsmoBTS(config)# bts 0 OsmoBTS(bts)# gsmtap-sapi enable-all Enable all kinds of messages (all SAPI) disable-all Disable all kinds of messages (all SAPI) bcch BCCH ccch CCCH rach RACH agch AGCH pch PCH sdcch SDCCH tch/f TCH/F tch/h TCH/H pacch PACCH pdtch PDTCH ptcch PTCCH cbch CBCH sacch SACCH OsmoBTS(bts)# gsmtap-sapi bcch OsmoBTS(bts)# OsmoBTS(bts)# gsmtap-sapi ptcch OsmoBTS(bts)# OsmoBTS(bts)# gsmtap-sapi pacch OsmoBTS(bts)# OsmoBTS(bts)# gsmtap-sapi pdtch OsmoBTS(bts)# OsmoBTS(bts)# Connection closed by foreign host
This is of course not particularly helpful when wanting to debug GPRS issues...
Related issues
Updated by laforge over 1 year ago
- Subject changed from osmo-bts-trx crashes when using @gsmtap-sapi pdtch@ to osmo-bts-trx crashes when using gsmtap-sapi pdtch
Updated by laforge over 1 year ago
for the record, this was OsmoBTS version 1.5.0.27-ae6a-dirty
- not sure what the dirty part refers to, was observed on a customer device. I doubt they have gsmtap specific changes.
Updated by fixeria over 1 year ago
- Related to Bug #5645: osmo-bts-trx crashes when using gsmtap option -i added
Updated by fixeria over 1 year ago
- Status changed from New to Feedback
- Assignee set to laforge
laforge wrote in #note-2:
for the record, this was
OsmoBTS version 1.5.0.27-ae6a-dirty
- not sure what the dirty part refers to
Probably some uncommitted modifications on top of the ae6a? In any case, this patch:
commit ae6ae89a199556293f4cfca26efbfadd828e3252 Author: Philipp Maier <pmaier@sysmocom.de> Date: Mon Sep 26 13:01:30 2022 +0200 pcu_sock: fix sourcecode formatting Change-Id: Ib88173363fc11ad44a5bdbcd85ea8ca075db7997
is behind the patches by dexter fixing some GSMTAP related segfault (see #5645):
commit 0087a1137ebaa870cad23fec8d19a51c6843de56 Author: Philipp Maier <pmaier@sysmocom.de> Date: Thu Oct 6 18:21:11 2022 +0200 measurement: do not call msgb_l3len without checking The function lchan_meas_handle_sacch() calls msgb_l3len without checking if l3h is even populated. Lets add a check to be sure. Change-Id: Ie5a9fe1ba880e68edb74f5f4ca559ac191330d4f commit 59e147e7ad89561957f7d3a4c8bfcce721c514ed Author: Philipp Maier <pmaier@sysmocom.de> Date: Thu Oct 6 18:18:22 2022 +0200 l1sap: do not call msgb_l2hlen without checking We request the length using msgb_l2len() in two locations where whe cannot be sure that l2h is populated. Lets check this first. Change-Id: Ie13d0724f183ff240714dcdbd24e5a21b4276bfe Related: OS#5645
Actions