| commit | 456135a605dd7addd6f4add007d4650bcbabfec6 | [log] [tgz] |
|---|---|---|
| author | Harald Welte <laforge@osmocom.org> | Tue Jul 18 14:39:57 2023 +0200 |
| committer | Harald Welte <laforge@osmocom.org> | Tue Jul 18 14:40:30 2023 +0200 |
| tree | 853aea5e99ccb9fbca9f11b742a3092e79c451ab | |
| parent | dc4b5d6522e49e35095e45428903227458e520de [diff] |
e1_recorder: Fix potential null-dereference Fixes: CID#307523: Dereference null return value (NULL_RETURNS) Change-Id: I481f0695f971f5cd2e77b8c9c62d423a70e0840d
(C) 2016 by Harald Welte laforge@gnumonks.org
The idea of this program is to be able to build a "poor mans E1/T1 recorder" purposes of data analysis, without any special equipment.
This approach is more risky than a purely passive, hardware based approach as that of the Osmocom e1-tracer released at https://osmocom.org/projects/e1-t1-adapter/wiki/E1_tracer
If you can, it is strongly recommended to use the purely passive, high-impedance tap approach of e1-tracer and not the poor-man's software proxy approach presented in osmo-e1-recorder.
To do so, two E1 cards are used as some kind of proxy for the E1 communication. Recording of a single E1 link always requires two E1 interface cards, one for each direction.
Recording can be performed either
All timeslots will be opened in "raw" mode, making sure the recording will work whether or not there is HDLC-based signalling (MTP or LAPD), PCM voice, TRAU frames or anything else on the line.
Recording will be done on a per-timeslot basis, dumping the raw bytes read for this timeslot into a file.
New files are started regularly, after reaching a pre-determined file size limit. File names contain RTC time stamping and timeslot number.
Later possible extensions could include automatic detection of the payload and a more intelligent storage format (e.g. in case of HDLC based signalling).