45fd07dc33
nat: Attempt to assign the BSC Timeslot based on a free list
...
Do attempt to not reassign an endpoint immediately but go
to the next free one.
2010-08-29 16:35:26 +08:00
39cd32e650
nat: Remove the CRCX value from the nat
...
Assume that if the MSC has assigned a timeslot/multiplex it will
also be used for the MGCP. So we just assume that it was allocated
on the BSC as well... in the worse case we will send a DLCX downstream
but it should be fine.
2010-08-29 16:35:25 +08:00
c021fbe603
nat: Reindent the code, no functional change
2010-08-29 16:35:25 +08:00
f4b3439206
nat: Make the code work in terms of endpoints instead of timeslot/multiplex
...
We are going to have more than one trunk, so all code hardcoding the
multiplex to zero must go. Avoid this kind of problem by saving the
MGCP endpoint number and comparing that.
2010-08-29 16:35:24 +08:00
591003328d
nat: Use the bsc_timeslot when talking to the BSC downstream
2010-08-29 16:35:24 +08:00
82049d8141
mgcp: Add method to go back from endpoint to multiplex/timeslot
2010-08-29 16:35:23 +08:00
cb8c35cd51
mgcp: Speculative mgcp fix...
...
We really have 32 channels per multiplex... so use the right
number... or at least it seems we do have 32.
2010-08-29 16:35:22 +08:00
Harald Welte
9ae28a128a
[gprs] Add option parsing to SGSN
2010-08-27 09:26:44 +02:00
Harald Welte
0e6be0e8f8
debian: add init script, /etc/default file and example config files
2010-08-27 09:09:40 +02:00
Harald Welte
2c67f47dba
debian: make more dependencies explicit in 'control' file
2010-08-27 09:09:40 +02:00
7a2c86b67d
vty: Provide a is_config_node for the VTY and use it for the bsc
...
Right now only bsc_hack and osmo_bsc need to have a custom
config handling as they use the subscr/oml nodes.
2010-08-26 16:55:37 +08:00
Harald Welte
2c869efcd3
Introduce '-D' commandline option to daemonize processes
...
This uses the osmo_daemonize() function of libosmocore >= 0.1.18,
and is now implemented for bac_nat, osmo-bsc, bsc_hack, osmo-gbproxy
and bsc_mgcp. This means only osmo-sgsn is missing, which currently
has no option parsing at all.
2010-08-25 19:43:54 +02:00
Harald Welte
305e2906aa
debian: don't buld arch-independent packages with executables in them
2010-08-25 13:57:22 +02:00
Harald Welte
ba43969832
debian: don't use override_dh_* for lenny compatibility
2010-08-25 13:54:58 +02:00
Harald Welte
5c55c4901d
debian: add build dependency to pkg-config
2010-08-25 12:59:29 +02:00
Harald Welte
11e79bdbbb
debian: add runtime dependency to libdbd-sqlite3
2010-08-25 12:58:32 +02:00
Harald Welte
3edb60de7e
debian: add build dependency to libdbi0-dev
2010-08-25 12:58:32 +02:00
4dc025b4a8
bsc_init: Compare it with the right enum type
...
In both cases the int value was zero, so it actually worked
2010-08-25 12:33:57 +08:00
Harald Welte
640d72ab3b
Add 'debian' directory containing debian packaging information
2010-08-24 15:33:04 +02:00
Harald Welte
2a90a25c93
Add gprs_sndcp.h to noinst_HEADERS
2010-08-24 13:59:08 +02:00
e0009f10ca
gsm48: Add method to send a separate facility message...
...
Send a facility messages with notifySS/CNAP as payload
to the phone... the function is not called from anywhere
in the tree yet but I use it locally.
2010-08-12 01:44:57 +08:00
Harald Welte
c1fa794ed5
[GPRS NS] Fix memory leak in gprs_ns_sendmsg() error path
...
When gprs_ns_sendmsg() succeeds in sending the message, we free()d the
msgb after transmitting it on the socket. However, if the NS-VC is
blocked or some other error condition exists, we returned an error
code but didn't free the msgb.
This resulted in an error leak which is now being addressed.
2010-08-09 21:15:40 +08:00
5508fe3320
mgcp: Add a vty command to free a given endpoint.
2010-08-08 16:39:57 +08:00
9e9392d5b3
mgcp: Use transient errors in more cases...
2010-08-08 16:24:48 +08:00
e378cb1341
mgcp: Reduce the log level to debug for these calls.
2010-08-06 20:22:23 +08:00
da340638fd
nat: Do not report paging to every BTS...
2010-08-06 20:15:40 +08:00
7d47601a35
nat: In case the downstream fails on a CRCX, ask it to do a DLCX
...
If the CRCX is failing, we will send a DLCX downstream and the next
time the callagent tries to do a CRCX we will be more lucky. This is
for the case where we have an endp allocated.
2010-08-06 19:16:34 +08:00
39a97e24fb
mgcp: Remember if the endpoint was allocated...
...
Do not use the CI_UNUSED to decide if an endpoint is allocated
but introduce a new flag. This way only the CRCX and free_endp
play with the allocated field.
2010-08-06 18:03:11 +08:00
869e38eced
mgcp: Add a callback to inform the client that an endpoint got reallocated
2010-08-06 17:54:27 +08:00
c2ff55475e
nat: Use the state information for the callback.
2010-08-06 09:07:20 +08:00
5b2726e6b5
nat: Keep the state (CRCX,MDCX,DLCX) of the operation..
...
This will allow to have additional debug information and to
use the state inside the forwarding code.
2010-08-06 09:07:07 +08:00
b84b5f61ac
nat: Free the endpoint when we get an empty CI.
...
In case the BSC is sending us an empty/unused CI we have
a bug, but for now let us free the endpoint and figure out
which response we are getting.
2010-08-06 08:34:46 +08:00
4634013cde
mgcp: Make the CI uint32_t all the way to avoid mismatch
2010-08-06 08:29:37 +08:00
9c31cfc3a2
nat: Provide information of when a msg does not contain a CI
...
When we fail to find a CI in a message that should contain it
then print the message so we can analyze the issue and improve
the code later on.
2010-08-06 08:19:05 +08:00
52ccf6a480
mgcp: Fix discovering the RTCP port with the more strict check.
...
If we have found the BTS and we receive data on the RTCP port
from the IP of the BTS we will set our RTCP port and forward it
to the network and hope it will be useful.
2010-08-06 07:49:57 +08:00
0d44eb6a24
mgcp: Determine the proto properly...
2010-08-06 07:49:57 +08:00
e2170a9f03
bsc_msc: Remove the except code as it is wrong...
...
Remove wrong code that is luckily not called. We would end up
in a reconnect and attempt to bsc_fd_register the same socket
again. I am removing this part of the code as it is not used
and it would need to know if the fd has ever been registered
or not...
2010-08-06 07:49:57 +08:00
7fe2a3d293
mgcp: Add instrumentation code to find a possible port leak/bsc-fd corruption
2010-08-06 07:18:22 +08:00
414bf401c7
mgcp: Move the mgcp_free_endp into the right path for the CRCX failure
2010-08-06 07:05:13 +08:00
6f36e92cab
mgcp: Fix a filedescriptor leak in case the bind is failing.
2010-08-06 03:00:17 +08:00
0e01d09500
mgcp: Enable the tap after configuring it...
2010-08-06 02:05:15 +08:00
260d6ed02c
mgcp: Add a call tap feature to forward audio to another forward port
...
For debugging it is useful to forward (tee) UDP packets to another
system and use gstreamer to inspect the rtp stream. This is untested
code and might contain bugs.... and of course only tap your own calls.
2010-08-06 01:12:21 +08:00
6357a8e311
mgcp: Patch RTP packets again if that is allowed.
2010-08-05 20:08:22 +08:00
a39ddd5654
mgcp: Get the for network/for bts flag right.
2010-08-05 19:35:44 +08:00
ea97fbf1a1
mgcp: Be more strict on the source addr/source port of the bts
...
Once we have discovered the bts we will not accept data from
anything else. The call will drop if the BTS is changing the
ip address of the nat anyway.
2010-08-05 19:35:44 +08:00
b97c50db0a
mgcp: Only discover the bts once, the extra check got lost
2010-08-05 19:35:44 +08:00
d5e6c2374b
nat: Use the mgcp_config_alloc to have defaults initialized
...
Initialize the net_base properly and benefit from future
config changes but also reset certain defauls that we do
not want at the nat.
2010-08-05 19:17:25 +08:00
f138f917ec
mgcp: Allow to dynamically allocate ports from a range..
...
Allow to switch to a dynamic port allocator and not reuse
the ports for a long time... This should help with a crazy
network sending two streams at the same time.
2010-08-05 08:08:17 +08:00
1be9f2fddc
mgcp: Allow to have a different port allocation mode
2010-08-05 07:20:09 +08:00
15e7389cf0
mgcp: Prepare to have different port allocation strategies.
2010-08-05 07:10:56 +08:00
Holger Hans Peter Freyther
Harald Welte