aboutsummaryrefslogtreecommitdiffstats
AgeCommit message (Collapse)AuthorFilesLines
2010-05-12Increase the version number.on-waves/0.3.99.10Holger Hans Peter Freyther1-1/+1
2010-05-12chan: After sending the GSM04.08 RR Release, reset the subscriber and waitHolger Hans Peter Freyther1-0/+5
After we send the SACH DEACTIVATE the BTS will get back to us with a Release Indication which will trigger the RF Channel Release handling. This is why we can return here, but we need to put the subscriber reference to make sure to not end in a infinite loop. This and the previous change fix the USSD issue for me.
2010-05-12bsc_msc_ip: Assign a dummy gsm_subscriber to send a SACH DEACTIVATEHolger Hans Peter Freyther1-0/+18
This is part of fixing USSD delivered to the MS. Currently only MT services would end up with a GSM Subscriber assigned. The LCHAN code is using the GSM Subscriber to figure out if a SACH DEACTIVATE should be send to the MS. Add code to always assign a GSM Subscriber.
2010-05-12bssap: Use libosmocore for message creation.Holger Hans Peter Freyther1-15/+1
2010-05-12bssap: Use libosmocore to create GSM0808 ResetHolger Hans Peter Freyther2-17/+2
2010-05-12bssap: Start to libosmocore for gsm0808 message creation.Holger Hans Peter Freyther1-32/+2
2010-05-12ipaccess: Wait for the BASEBAND_TRANSCEIVER and then bootstrap OMLHolger Hans Peter Freyther1-17/+16
Currently we are connecting to the BTS and once the OML is established we are bootstrapping the OML. This does not work for a multi TRX setup as we will need to use a trx_nr != 0 for it. Change the code to wait for a message (in this case NM OC_BASEBAND_TRANSC) to detect the trx_nr used by the BTS and then use that TRX to bootstrap the network. I have tested setting the unit id on a single and multi trx system for the first and second trx.
2010-05-12abis: Pass the abis_om_obj_inst in the nm_state_event..Holger Hans Peter Freyther6-7/+12
2010-05-12nat: Have a recycle timer that removes unconfirmed SCCP connections.Holger Hans Peter Freyther1-0/+32
The MSC does not respond to a SCCP CR with Paging Response as GSM payload, when the response comes in 'too late'. Prevent the MUX having stale connections and start removing old connections every 20 minutes.
2010-05-12nat: When we fail to reallocate... also close down the MGCP partHolger Hans Peter Freyther1-0/+1
Give the BSC a chanche to close down MGCP ports as well.
2010-05-12nat: Store the creation time of a sccp connection.Holger Hans Peter Freyther4-2/+8
Generate it when creating the connection but also when reusing an existing connection.
2010-05-11bsc_msc_ip: Use constants for ?/0/1.Holger Hans Peter Freyther1-2/+2
2010-05-11Increase the version... as we have new commandson-waves/0.3.99.9Holger Hans Peter Freyther1-1/+1
2010-05-11bsc_msc_ip: Add an extra command to show the MSC status.Holger Hans Peter Freyther1-0/+28
2010-05-11bsc_msc_ip: Move the MSC connection into the structureHolger Hans Peter Freyther2-17/+19
2010-05-11nat: Print the MSC status with a new vty command.Holger Hans Peter Freyther3-15/+33
2010-05-11misc: Make sure PACKAGE_VERSION is getting defined with a useful content.Holger Hans Peter Freyther1-3/+2
PACKAGE_VERSION is used by the copyright message.
2010-05-11bsc_msc_ip: Add a test mode to send messages to the MSC.Holger Hans Peter Freyther1-1/+34
Check if the MSC likes paging responses when it has not recently send out a paging request.
2010-05-11gsm0408: Use counter_inc to increment the counter.Holger Hans Peter Freyther1-1/+1
2010-05-05bsc_msc: Add a connection timeout for the MSC.Holger Hans Peter Freyther2-0/+13
When no one is listening our connection would get stuck in the SYN_SENT state and we would be there forever.
2010-05-05Version bump for testing it on the targeton-waves/0.3.99.8Holger Hans Peter Freyther1-1/+1
2010-05-05nat: Using the right fd can be a good idea as wellHolger Hans Peter Freyther1-1/+1
2010-05-05Another version... another tryon-waves/0.3.99.7Holger Hans Peter Freyther1-1/+1
2010-05-05nat: Fix bad bug, make sure the fd is not overwritten..Holger Hans Peter Freyther1-11/+11
The adding of the innocent looking code was actually overwrote the fd and then stupid things happened. Rename variables to avoid that. rc,ret should be scratch variables...
2010-05-05Bump the version for TCP_NODELAY.on-waves/0.3.99.6Holger Hans Peter Freyther1-1/+1
2010-05-05bsc_msc_ip: Attempt to disable nagleHolger Hans Peter Freyther1-0/+7
Use TCP_NODELAY on the connection to the MSC. We want small messages to be send immediately.
2010-05-05nat: Use TCP_NODELAY for the connection to the BSC.Holger Hans Peter Freyther1-1/+7
We do not want to use NAGLE for the BSC connection.
2010-05-05Bump versionon-waves/0.3.99.5Holger Hans Peter Freyther1-1/+1
Configurable timeout.
2010-05-05bsc_msc_ip: Add VTY code for ping/pong timeout.Holger Hans Peter Freyther1-0/+21
2010-05-05bsc_msc_ip: Do not send a ping when the timeout is negativeHolger Hans Peter Freyther1-0/+3
2010-05-05bsc_msc_ip: Make the ping/pong timeouts configurableHolger Hans Peter Freyther3-2/+6
Take the timeouts from the struct.
2010-05-05nat: Improve log messages. Refer to ip and fd.Holger Hans Peter Freyther1-3/+6
2010-05-05nat: Make ping/pong timeout configurable.Holger Hans Peter Freyther4-3/+47
2010-05-03Work with later libosmocore.on-waves/0.3.99.4Holger Hans Peter Freyther1-1/+1
2010-05-03move gsm48_construct_ra() to libosmocoreHarald Welte1-27/+0
2010-05-03'struct gprs_ra_id' is now defined in libosmocoreHarald Welte1-8/+0
2010-05-03Bump version..on-waves/0.3.99.3Holger Hans Peter Freyther1-1/+1
* ping pong and possible crash fix.
2010-05-03bsc: Speculative crash fix.Holger Hans Peter Freyther1-0/+1
Make sure the sccp_cc_timeout is stopped when we delete the associated data. There is one crash report that indicates that we have a pending timer that is inside freed memory. A crash could have occured when the connection to the MSC was lost while have unconfirmed connections.
2010-05-03nat/bsc: Send PONG on PING, send PING from the BSC tooHolger Hans Peter Freyther2-6/+70
We do want to send PING/PONG in both ways to have a heartbeat on the TCP connection. When switching over to SCTP we can rely on the builtin heartbeat functionality.
2010-05-02nat: Send a IPA PING down the stream and wait for the pong.Holger Hans Peter Freyther2-0/+59
We will send a ping every 20 seconds and if we have no pong within 5 seconds we will close down the BSC connection and wait for a reconnect. We will start this after having authenticated the BSC and we stop the timer when destructing the BSC connection.
2010-05-02nat: Allow to only show statistics for a given BSC Cfg.Holger Hans Peter Freyther1-1/+9
2010-05-02nat: Do not allow a BSC to send auth messages twice.Holger Hans Peter Freyther1-0/+6
2010-05-02nat: Fix vty output for connected BSCsHolger Hans Peter Freyther1-1/+1
2010-05-01bssap: Store the link_id in the new msgb->cb.Holger Hans Peter Freyther1-4/+6
Work with a new version of libosmocore that gets rid of additional pointers.
2010-05-01remove any reference to 'struct gsm_bts_link'Holger Hans Peter Freyther2-7/+0
2010-05-01gsm_04_11.c: Use msgb->l4h instead of sms->smsh, as the latter is goneHarald Welte1-1/+1
2010-05-01configure.in: Bump the version due mem leak fixes.on-waves/0.3.99.2Holger Hans Peter Freyther1-1/+1
The nat is stable and in testing..
2010-05-01nat: Improve log message and refer to the BSC that was lost.Holger Hans Peter Freyther1-2/+6
2010-05-01nat: Fix memory leak... in MGCP forwardingHolger Hans Peter Freyther1-0/+1
The code needs to be refactored but this is fixing the leak for now. We used to forward everything to the BSC but now we handle the DLCX locally and this means we need to clear the patched message. We should refactor it to not generate the patched msg until a lot later.
2010-04-30[vty] Free the matched at the end of the routine.Holger Hans Peter Freyther1-4/+0
Remove the return from the case labels and cleanup at the end matched array at the end of the routine.