Sylvain Munaut
adea4f1bf8
db: Use proper %llu for subscriber ID in format string.
...
This was causing weird crashes when running in 32 bit linux.
Thanks to horiz0n for taking the time to debug this with me on IRC.
Written-by: horiz0n <cscan@gmx.net>
Signed-off-by: Sylvain Munaut <tnt@246tNt.com>
2010-07-03 22:24:13 +02:00
Harald Welte
c48ac4750d
[SGSN] Some IEs in ATTACH REQUEST can be longer in 24.008
...
The original length checks were based on 04.08, but 24.008 allows
longer IEs.
2010-07-03 21:20:06 +02:00
Harald Welte
4b6ac1edb9
[SGSN] Initial work on GMM authentication and ciphering
2010-07-03 11:09:06 +02:00
Harald Welte
3e2e159707
[SGSN] LLC: only copy Kc if algorithm != GPRS_ALGO_GEA0
...
The caller currently has set the Kc pointer to NULL, this fix avoids
crashing at the callee.
2010-07-02 22:56:55 +02:00
Harald Welte
d6f582b352
[SGSN] use GPRS_ALGO_GEA0 instead of '0'
2010-07-02 22:56:55 +02:00
Harald Welte
1b8827ae5b
[SGSN] LLC: Make sure we calculate FCS after decryption
2010-07-02 22:55:39 +02:00
Harald Welte
d07b4f99ce
[SGSN] Use libosmocore GPRS encryption plugins from LLC layer
...
This adds the bits that call into libosmocore (and its plugins)
to implement GPRS (LLC) encryption.
2010-07-02 22:55:39 +02:00
Harald Welte
496aee463f
[SGSN] Call libosmocore helper to load cipher plugins
2010-07-02 22:54:55 +02:00
Harald Welte
6ed4950d5b
[SGSN] remove bogus debug statement
2010-07-02 22:52:33 +02:00
Harald Welte
23c13760ca
[SGSN] SM: Properly parse APN and PCO Optional IE in PDP CTX ACT REQ
2010-07-02 22:52:04 +02:00
Harald Welte
c5d4a0ced1
[SGSN] Print AP name and PDP address of each PDP context
2010-07-02 22:51:16 +02:00
Harald Welte
6add7c2009
[SGSN] Correctly pass IMSI of MM ctx to GTP/GGSN
2010-07-02 22:47:39 +02:00
Harald Welte
3d6815abd1
[SGSN] SNDCP: Subtract SNDCP header from fragment length
...
We don't want to store too many bytes as we would otherwise create
too large re-assembled packets..
2010-07-02 17:16:07 +02:00
Harald Welte
36f1217f32
[SGSN] SNDCP: SN-PDUs need a 16bit length field
...
The BSSGP LLC-PDU IE can have up ot 16bit length, so subtracting
the LLC header the SN-PDU can still have almost a length field
close to 0xffff
2010-07-02 16:44:24 +02:00
Harald Welte
8b705f2533
[SGSN] SNDCP: re-set total number of fragment length
2010-07-02 16:18:59 +02:00
Harald Welte
18d1452a92
[SGSN] Send poper SM and GMM STATUS in case we have no MM CTX
2010-07-02 16:18:18 +02:00
Harald Welte
ab4094ca6b
[SNDCP] some more defragmentation debug statements
2010-07-02 16:01:47 +02:00
Harald Welte
60da7d4f88
[SGSN] SNDCP defrag: use correct length, don't include LLC FCS
...
The LLC layer tells us the PDU length, and we have to use it
in SNDCP rather than to re-calculate it if we want to avoid copying
the CRC24 into the defrag elements.
2010-07-02 15:45:12 +02:00
Harald Welte
8f0c0a3a3c
[SGSN] SNDCP: Actually copy the payload data in the defrag queue entry
...
Prior to this patch our defragmented messages always consisted of 00-bytes
2010-07-02 10:29:06 +02:00
Harald Welte
11ff3e8ec2
[SGSN] Remove old dead code
2010-07-02 10:17:24 +02:00
Harald Welte
19f9e30eb9
[SGSN] Only increment PDP CTX ACT count after check for retransmission
2010-07-02 10:15:49 +02:00
Harald Welte
b37515e86b
[SGSN] Account for re-transmitted PDP CTX ACT REQ
...
In case our PDP CTX ACT ACK gets lost on the Gb or Um link,
the MS re-transmits the PDP CTX ACT REQ. We have to ACK it
again instead of REJecting it..
2010-07-02 10:12:58 +02:00
Harald Welte
993697c8d2
[SGSN] SNDCP: Fix off-by-one error missing last segment in defrag
...
When we defragment the segments from the defrage queue, we have
to iterate all the way up to (and including) the last segment number
that we have received.
2010-07-02 10:11:42 +02:00
Harald Welte
b87bc86006
[SGSN] SNDCP: more verbose defragmentation debugging
2010-07-01 20:29:20 +02:00
Harald Welte
8911cef81d
[SGSN] Fix segfault when passing re-assembled SN-PDU to GMM
...
sgsn_rx_sndcp_ud_ind() can no longer make the assumption that msgb_bcid() is
valid, as this is only true for an un-fragmented SN-PDU. So instead,
we now store the RAID in the SNDCP Entity and pass it as an explicit
argument to sgsn_rx_sndcp_ud_ind().
2010-07-01 19:56:19 +02:00
Harald Welte
951a12cf4e
[SGSN] SNDCP: fix off-by-one error during defragmentation
2010-07-01 15:09:45 +02:00
Harald Welte
362aea0f4d
[SGSN] SNDCP: Initialize defrag list header
2010-07-01 12:31:10 +02:00
Harald Welte
fd29a1f6cc
[SGSN] GMM: Complete value_string for GMM cause values
2010-07-01 12:26:59 +02:00
Harald Welte
65d96783df
[SGSN] SNDCP: use llist_for_each_entry_safe() when cleaning frag queue
2010-07-01 12:19:02 +02:00
Harald Welte
875840c603
[SGSN] Fix TLLI (re)assignments
...
Once The TLLI (or P-TMSI of which it is derived) change has been
confirmed by the MS, we need to unassign the old TLLI but keep
the new TLLI _without_ re-setting the LLC entity structure such
as VUsend /VUrecv counters.
2010-07-01 11:54:31 +02:00
Harald Welte
937a9ecfcd
[SGSN] Fix segfault when doing PS PAGING
...
The 'bssgp_paging_info' contains a pointer to the P-TMSI, and not
the P-TMSI itself. The reason is that it is an optional BSSGP IE.
2010-07-01 11:53:25 +02:00
Harald Welte
f78a3b2a22
[SGSN] Add VTY interface for SNDCP
2010-07-01 10:57:44 +02:00
cacbc73d5e
osmo_bsc: Add option parsing, vty dummy
...
Add dummy files and option parsing for the Osmo BSC code.
2010-06-30 15:11:30 +08:00
ab5135e36a
bsc: Move meas_rep.c into the libbsc.c as it is used from bsc_vty.c
2010-06-30 15:11:30 +08:00
5ccab1020c
osmo_bsc: Add empty osmo_bsc_main.c and hook it into the build
2010-06-30 15:11:29 +08:00
ec4bfdc435
bsc_api: Move the API init into the main function.
...
The BSC API will not be used until the first iteration through
the event loop so it is safe to call it after the bootstrap.
2010-06-30 15:11:29 +08:00
2f4dbebb20
osmo-bsc: Start adding code for the osmo_bsc..
...
This adds the unix domain socket rf control protocol into
the new subdirectory.
2010-06-30 15:11:29 +08:00
5950236b5e
* Fix null ptr dereference and sms memleak in case the recipient of an sms sent via vty is not attached. Store the sms in the database in this case for later delivery.
...
The problem is that sms_from_text returns NULL in case the
subscriber is not attached which a) leaks memory of the
previously allocated sms and b) runs into a null ptr
dereference in _send_sms_str().
There may be a better solution than this but this is the
easiest way of noticing and taking action I could find
without changing return values of sms_from_text.
2010-06-30 09:10:04 +02:00
Harald Welte
bd17b39fd2
[BSC] VTY: Fix parsing of uppercase hex digits
2010-06-30 09:01:35 +02:00
Harald Welte
43ef6483db
[SGSN] BSSGP: Print TLLI as hex value like everwhere else
2010-06-30 09:01:35 +02:00
d9c9f07c2c
handover: lchan_free will only free the local resources.
...
lchan_free only free's the local resource of the BSC but
does not release the channel at the BTS. Use lchan_release
to properly release the channel. This code assumes that the
timeout happens after a CHAN ACT ACK/NACK otherwise we have
some problems. The comment indicates that this is the case.
2010-06-30 13:04:13 +08:00
f2553a6c3a
handover: Call bsc_handover_clear from gsm0808_clear
...
The bsc_handover_clear will release an in-progress handover
and free the lchana and the data associated with this handover
2010-06-30 13:02:08 +08:00
0610947f4c
misc: Move handover_logic.c into the libbsc.a
...
We are implementing the intra BSC handover so we should have the
code in the BSC as well.
2010-06-30 12:59:20 +08:00
88519eaaef
msc: Add includes to silence a compiler warning.
2010-06-30 12:59:20 +08:00
e071ab70e9
handover: Attempt to use the same gsm_subscriber_conn for new/old channel
...
The transaction should not know on which lchan we are operating
as this can change due handover. Add untested code to share the
subscriber connection of the new and old lchan and move the pointer
in case of success/failure. Also on a clear command we will free
any resources allocated...
This code is not tested and needs to be debugged, but it should
have the right structure. I am going to fix a potential memleak
in the next commit.
2010-06-30 12:59:20 +08:00
d06516557a
bsc_api: Avoid use after free kind of error on the lchan.
2010-06-30 12:17:35 +08:00
2486663844
ussd: Release the MSC connection as fast as possible..
2010-06-30 12:15:19 +08:00
0379c6d386
abis_rsl: Add full MA again until we are confident to remove it
...
The spec seems to say we do not need to include the full MA,
even for Phase1 phones but that is not so clear...
2010-06-30 12:06:20 +08:00
e38bd6caa3
abis_rsl: Change code to generate Channel Identification
...
1.) memset the gsm48_chan_desc to avoid sending dummy data
2.) According to the GSM08.58 9.3.5 the Mobile Allocation
shall be included but the empty (by setting the length
to zero).
3.) use msgb_tlv_put and calculate the length via the l3h
msgb_l3len and assign it.
2010-06-30 12:01:22 +08:00
06abe9f383
ussd.c: Make sure text is memset to zero..
...
Make the strstr fail fast, make sure the text is null terminated...
2010-06-30 11:59:30 +08:00
Sylvain Munaut
Harald Welte
Holger Hans Peter Freyther