On Monday 25 February 2008 8:47:15 am James Morris wrote: > On Tue, 26 Feb 2008, James Morris wrote: > > See updated version below, with the sk pointer (to allow the callee > > to take the rcu read lock). > > Actually, we shouldn't need that lock there, and can directly call > selinux_netlbl_sk_security_reset, right? I'm looking at this right now ... based on the idea that you can safely read/write 32-bit values without a lock I believe the locking for sk_security_struct could use a bit of a tweak. I'm still looking/thinking but I believe we can remove the rcu lock entirely and just rework selinux_netlbl_inode_permission() a little bit. If you can spare a few hours I'll post a RFC patch ... -- paul moore linux security @ hp -- This message was distributed to subscribers of the selinux mailing list. If you no longer wish to subscribe, send mail to majordomo@xxxxxxxxxxxxx with the words "unsubscribe selinux" without quotes as the message.
