Move emergency handlers header file into the support folder.
Refactored signal handling:
- Remove the dependency with the running static variable in the header file.
- Move implementations down to cc files.
- Allow specifying a new signal handler that will be used to stop the applications.
- Move signal handling files to support.
* Started logger replacement in epc main.
* Replaced loggers in nas and s1ap.
* Replace loggers in nas statics.
* Replaced loggers in s1ap ctx mngmt.
* Replace loggers in s1ap erab mngmt.
* Replaced loggers in s1ap mngmt proc.
* Replaced loggers in s1ap nas transport.
* Replaced loggers in s1ap paging.
* Replaced loggers in mme.
* Replaced loggers in mme gtpc.
* Started removing logger injection in init functions and cleaned up argument passing to nas statics.
* Remove logger injection from mme and s1ap.
* Replaced loggers in the hss class.
* Replaced loggers in gtpc.
* Replaced loggers in GTPU and SPGW.
* Replaced loggers in MBMS.
* Refactor NR resource allocation classes. Use DCI instead of grant for dummy PDSCH UE/eNB test
* Minor refactors in NR workers and ue_dl
* Fix include issues
* fix compilation issues
the signal handler is the same for all three apps. The "running" flag
as well as the file_logger object are in the common header in order
to allow the signal handler to flush the file if the alarm goes off.
* Remove boost string
* Make sure that int/enc algos can be specified in capital case in the MME.
* Use == instead of compare for comparing strings with const chars
Co-authored-by: Pedro Alvarez <pedro.alvarez@softwareradiosystems.com>
If a process exits and has a controlling terminal, everything under
it gets a SIGHUP by the kernel. That's the case for instance when
running a srsUE over ssh with some specific parameters. As a result,
when the ssh client is killed, SIGHUP is sent to srsUE, which will be
killed immediatelly without proper release, which means in turn that log
buffers are not flushed.
Let's do proper shutdown when SIGHUP is received so that everyone can
get log files.
SIGKILL cannot be intercepted by apps, so it's useless adding a signal
handler for them. I'm pretty sure the signal() function is erroring in
this scenarios. From signal() man page:
"The signals SIGKILL and SIGSTOP cannot be caught or ignored."
This entails:
* Adding S1AP paging messaging to the MME.
* Adding the Downlink Data Notification, Donlink Data Notification Acknoledgment, and Downlink Data Notification Failure Messages.
* Adding the logic at the GTP-U to detect that Downlink Data notification needs to be sent.
* Proper handling of the new GTP-C messages at the MME GTPC and SPGW GTPC classes
* Add queuing mechanisms at the GTP-U while waiting for paging response. Queue is flushed when paging fails (timeout or other).
* Make sure eNB's SCTP information is properly stored.
* Make sure UE's GUTI information is properly stored.
* Adding timer creation at NAS using timerfd_create
* Adding the ability to add and remove the timer file descriptor to the MME main select loop
* Adding the ability to the MME for the NAS to check if a timer is running
* Adding the timer expiry handling functionality at the MME
* added config option for tun dev names vs fixed names added config option for M1-U interface addr
* Added options to sgi_mb tun name.
* Added some comments to mbms.conf.example
* Added some comments to example configs to explain the new values.