From mboxrd@z Thu Jan 1 00:00:00 1970 Message-ID: <43C8AB11.3070405@cornell.edu> Date: Sat, 14 Jan 2006 00:41:05 -0700 From: Ivan Gyurdiev MIME-Version: 1.0 To: SELinux List CC: Stephen Smalley , Daniel J Walsh Subject: Re: [SEMANAGE(tool)] Fix many issues References: <43C89A43.3020105@cornell.edu> <43C8A1FF.8090605@cornell.edu> In-Reply-To: <43C8A1FF.8090605@cornell.edu> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Sender: owner-selinux@tycho.nsa.gov List-Id: selinux@tycho.nsa.gov Ivan Gyurdiev wrote: > >> >> Further notes: >> - Multiple roles don't seem to work - I get the usage() printout >> - Modifying with -R doesn't work properly - it adds roles on top of >> the current ones, when it should clear the current role set first - >> you can use the del_role function, or better, the set_roles function >> (not sure if it will work, let joshua know if the swig typemaps >> create problems for that case). >> - I would recommend a message be printed out after a delete() that >> leaves a policy object behind (do exists check again), that says: >> "Deleted local modifications to , falling back to policy >> default". This describes what the delete function really does in that >> case. > Another thing - I am still not clear on memory management vs python. > I don't see a single thing being freed in the python semanage - is > that a problem? Valgrind certainly prints a full 10 screens of errors > that complain about uninitialized value and invalid read on > PyObject_Free, but I'm not sure I understand what's going on. I doubt that python can manage memory for precompiled code that it calls. I would guess it should be freeing things. Another issue that I hadn't noticed is that the rc value on all exists functions isn't checked anywhere. -- This message was distributed to subscribers of the selinux mailing list. If you no longer wish to subscribe, send mail to majordomo@tycho.nsa.gov with the words "unsubscribe selinux" without quotes as the message.