From mboxrd@z Thu Jan 1 00:00:00 1970 Message-ID: <43BBD715.8070603@cornell.edu> Date: Wed, 04 Jan 2006 09:09:25 -0500 From: Ivan Gyurdiev MIME-Version: 1.0 To: SELinux List CC: Stephen Smalley Subject: Re: [SEMANAGE] [SETSEBOOL] Clone record on set/add/modify References: <43BBD6C0.6020108@cornell.edu> In-Reply-To: <43BBD6C0.6020108@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: > Hi, Steven pointed out (off-list) that libsemanage vs libsepol > behavior is inconsistent with respect to set/add/modify. > In libsemanage, the object being added becomes "managed" by the > library, while this is not true in libsepol. We decided it was better > to have the client control the object passed in, and be responsible > for freeing it. > > This patch makes the appropriate changes (which include reversal of > some of the code from the other patches) to implement that. > It also includes a bugfix for the error path of > database_policydb.c:list, pointed out by Russel Coker. > By the way, how does python handle this - does it manage memory management automatically, or do we still need to call the free functions.... because.. I don't see any free functions being called in the semanage tool - not even for the key. -- 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.