[RFC PATCH v2 0/9] Add LSM access controls and auditing to io_uring

[Steve Grubb]

On Monday, October 4, 2021 9:27:33 AM EDT Paul Moore wrote:
> On Mon, Oct 4, 2021 at 8:40 AM Richard Guy Briggs <rgb at redhat.com> wrote:
> > On 2021-10-03 19:21, Paul Moore wrote:
> > > On Sat, Oct 2, 2021 at 9:16 AM Steve Grubb <sgrubb at redhat.com> wrote:
> > > > On Thursday, September 9, 2021 8:58:58 PM EDT Richard Guy Briggs 
wrote:
> > > > > > I spent some time this morning/afternoon playing with the
> > > > > > io_uring
> > > > > > audit filtering capability and with your audit userspace
> > > > > > ghau-iouring-filtering.v1.0 branch it appears to work correctly. 
> > > > > > Yes,
> > > > > > the userspace tooling isn't quite 100% yet (e.g. `auditctl -l`
> > > > > > doesn't
> > > > > > map the io_uring ops correctly), but I know you mentioned you
> > > > > > have a
> > > > > > number of fixes/improvements still as a work-in-progress there so
> > > > > > I'm
> > > > > > not too concerned.  The important part is that the kernel pieces
> > > > > > look
> > > > > > to be working correctly.
> > > > > 
> > > > > Ok, I have squashed and pushed the audit userspace support for
> > > > > iouring:
> > > > > 
> > > > > https://github.com/rgbriggs/audit-userspace/commit/e8bd8d2ea8adcaa7
> > > > > 58024cb
> > > > > 9b8fa93895ae35eea
> > > > > https://github.com/linux-audit/audit-userspace/compare/master...rgb
> > > > > riggs:g
> > > > > 
> > > > > hak-iouring-filtering.v2.1 There are test rpms for f35 here:
> > > > >         http://people.redhat.com/~rbriggs/ghak-iouring/git-e8bd8d2-> > > > >         fc35/
> > > > > 
> > > > > userspace v2 changelog:
> > > > > - check for watch before adding perm
> > > > > - update manpage to include filesystem filter
> > > > > - update support for the uring filter list: doc, -U op, op names
> > > > > - add support for the AUDIT_URINGOP record type
> > > > > - add uringop support to ausearch
> > > > > - add uringop support to aureport
> > > > > - lots of bug fixes
> > > > > 
> > > > > "auditctl -a uring,always -S ..." will now throw an error and
> > > > > require  "-U" instead.
> > > > 
> > > > OK, now that the bug fix release is out of the way, let's start
> > > > merging this into user space. I think we should start with the code
> > > > that let's auditd write the record correctly and then the auditctl
> > > > piece that inserts the rule into the kernel. Those should be easy to
> > > > merge.
> > > > 
> > > > I see one section of code that mirrors all of the operations in
> > > > ioring.h. I thought that Paul only wanted to audit some of the
> > > > operations and not all of them. Did that change? Are we really going
> > > > to allow auditing reads on ioring?> > 
> > > Only certain io_uring operations are audited, you can see the patch
> > > here in the selinux/next tree (look for the io_op_defs struct changes
> > > and the "audit_skip" field):
> > > 
> > > https://git.kernel.org/pub/scm/linux/kernel/git/pcmoore/selinux.git/com
> > > mit/?h=next&id=5bd2182d58e9d9c6279b7a8a2f9b41add0e7f9cb> 
> > I understood that was the default, but that it could be changed by
> > userspace configuration?  Now that I say this, I realize there is no API
> > to do so, but that could be added without breaking anything.
> 
> The approach implemented in the patch currently living in selinux/next
> was a carefully arranged compromise with the io_uring devs (see all of
> the on-list discussions); it is unlikely to change.

Right. That's what I remember. So, I think that we should comment out the 
items in the ioring lookup table that are not auditable so that we can 
provide a meaningful warning to users of auditctl.

-Steve