[OpenSIPS-Users] Could someone review my pending OpenSIPS patches?

Peter Lemenkov lemenkov at gmail.com
Tue Jun 26 10:12:41 CEST 2012


Hello.
Right now I've got more than a dozen pending opensips-related patches
in my queue and this complicates packaging for Fedora/EPEL. Could
someone, please, take a look at them  and see which patches are ready
to merge and which ones needs love.

I already opened a necessary tickets. Namely these ones:

* https://sourceforge.net/tracker/?func=detail&aid=3535148&group_id=232389&atid=1086412

This patchset is the simplest one - just a small fix and
non-destructive additions related to a RPM packaging wrt to a systemd.

* https://sourceforge.net/tracker/?func=detail&aid=3535151&group_id=232389&atid=1086412

This ticket contains fixes for a proper memory freeing of a column
data in a db library. Unless I missed something the memory resources
allocated to rows aree freed while memory resources for columns are
not freed entirely. Also the type of a resource should be set to NULL
after freeing. Please take a closer look at this since nor me neither
my colleague are familiar with OpenSIPS internals well enough.

* https://sourceforge.net/tracker/?func=detail&aid=3535153&group_id=232389&atid=1086412

This one needs a very very closer look from a opensips-expert. There
is definitely something wrong within a usrloc module -
https://github.com/lemenkov/OpenSIPS/commit/e675790 . Take a look -
although types of a vals2[3] and vals2[4] are set to DB_INT you
assigned a value to a bitmap_val members. if these objects are really
of a DB_INT type then you must assign these valuet to a int_val
members instead. However there is another possible approach - fix type
of these objects from DB_INT to DB_BITMAP. Please, take a look in the
meantime.

Also there is a first set of a patches agaist db_oracle module which
is entirely broken now. My colleague already made it working with a
few additional patches some of which are quite big and his work is
still ongoing. On a contrary these three patches looks quite safe and
simple so why not to apply them?

* https://sourceforge.net/tracker/?func=detail&aid=3535168&group_id=232389&atid=1086412



-- 
With best regards, Peter Lemenkov.



More information about the Users mailing list