From mboxrd@z Thu Jan 1 00:00:00 1970 From: David Howells Date: Thu, 06 Jun 2019 22:38:01 +0000 Subject: Re: [RFC][PATCH 00/10] Mount, FS, Block and Keyrings notifications [ver #3] Message-Id: <30567.1559860681@warthog.procyon.org.uk> MIME-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: 8bit List-Id: References: <155981411940.17513.7137844619951358374.stgit@warthog.procyon.org.uk> <3813.1559827003@warthog.procyon.org.uk> <8382af23-548c-f162-0e82-11e308049735@tycho.nsa.gov> <0eb007c5-b4a0-9384-d915-37b0e5a158bf@schaufler-ca.com> <07e92045-2d80-8573-4d36-643deeaff9ec@schaufler-ca.com> <23611.1559855827@warthog.procyon.org.uk> In-Reply-To: To: Andy Lutomirski Cc: dhowells@redhat.com, Andy Lutomirski , Casey Schaufler , Stephen Smalley , Al Viro , Greg Kroah-Hartman , USB list , raven@themaw.net, Linux FS Devel , Linux API , linux-block@vger.kernel.org, keyrings@vger.kernel.org, LSM List , LKML , Paul Moore Andy Lutomirski wrote: > I mean: are there cases where some action generates a notification but does > not otherwise have an effect visible to the users who can receive the > notification. It looks like the answer is probably “no”, which is good. mount_notify(). You can get a notification that someone altered the mount topology (eg. by mounting something). A process receiving a notification could then use fsinfo(), say, to reread the mount topology tree, find out where the new mount is and wander over there to have a look - assuming they have the permissions for pathwalk to succeed. David