linux-kernel - Re: [PATCH v4 1/9] capability: introduce new capable flag NODENYAUDIT

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives
Hash Suite for Android: free password hash cracker in your pocket
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CAJ2a_DesiD+LU-aWOEWRkyc0rcmZ0Za5i6-rZX-kHP2GzQyuFg@mail.gmail.com>
Date:   Wed, 31 May 2023 20:34:11 +0200
From:   Christian Göttsche <cgzones@...glemail.com>
To:     "Serge E. Hallyn" <serge@...lyn.com>
Cc:     selinux@...r.kernel.org, Paul Moore <paul@...l-moore.com>,
        John Johansen <john.johansen@...onical.com>,
        James Morris <jmorris@...ei.org>,
        Stephen Smalley <stephen.smalley.work@...il.com>,
        Eric Paris <eparis@...isplace.org>,
        Christian Brauner <brauner@...nel.org>,
        Casey Schaufler <casey@...aufler-ca.com>,
        Dave Chinner <dchinner@...hat.com>,
        Nathan Lynch <nathanl@...ux.ibm.com>,
        Al Viro <viro@...iv.linux.org.uk>,
        Roberto Sassu <roberto.sassu@...wei.com>,
        Micah Morton <mortonm@...omium.org>,
        Frederick Lawler <fred@...udflare.com>,
        Günther Noack <gnoack3000@...il.com>,
        linux-kernel@...r.kernel.org, apparmor@...ts.ubuntu.com,
        linux-security-module@...r.kernel.org, bpf@...r.kernel.org
Subject: Re: [PATCH v4 1/9] capability: introduce new capable flag NODENYAUDIT

On Wed, 31 May 2023 at 16:08, Serge E. Hallyn <serge@...lyn.com> wrote:
>
> On Wed, May 31, 2023 at 09:07:34AM -0500, Serge E. Hallyn wrote:
> > On Thu, May 11, 2023 at 04:25:24PM +0200, Christian Göttsche wrote:
> > > Introduce a new capable flag, CAP_OPT_NODENYAUDIT, to not generate
> > > an audit event if the requested capability is not granted.  This will be
> > > used in a new capable_any() functionality to reduce the number of
> > > necessary capable calls.
> > >
> > > Handle the flag accordingly in AppArmor and SELinux.
> > >
> > > Suggested-by: Paul Moore <paul@...l-moore.com>
> > > Signed-off-by: Christian Göttsche <cgzones@...glemail.com>
> >
> > Reviewed-by: Serge Hallyn <serge@...lyn.com>
>
> Sorry, obviously I should have removed this, until the comment below was
> answered :)
>
> > > ---
> > >  include/linux/security.h       |  2 ++
> > >  security/apparmor/capability.c |  8 +++++---
> > >  security/selinux/hooks.c       | 14 ++++++++------
> > >  3 files changed, 15 insertions(+), 9 deletions(-)
> > >
> > > diff --git a/include/linux/security.h b/include/linux/security.h
> > > index e2734e9e44d5..629c775ec297 100644
> > > --- a/include/linux/security.h
> > > +++ b/include/linux/security.h
> > > @@ -67,6 +67,8 @@ struct watch_notification;
> > >  #define CAP_OPT_NOAUDIT BIT(1)
> > >  /* If capable is being called by a setid function */
> > >  #define CAP_OPT_INSETID BIT(2)
> > > +/* If capable should audit the security request for authorized requests only */
> > > +#define CAP_OPT_NODENYAUDIT BIT(3)
> > >
> > >  /* LSM Agnostic defines for security_sb_set_mnt_opts() flags */
> > >  #define SECURITY_LSM_NATIVE_LABELS 1
> > > diff --git a/security/apparmor/capability.c b/security/apparmor/capability.c
> > > index 326a51838ef2..98120dd62ca7 100644
> > > --- a/security/apparmor/capability.c
> > > +++ b/security/apparmor/capability.c
> > > @@ -108,7 +108,8 @@ static int audit_caps(struct common_audit_data *sa, struct aa_profile *profile,
> > >   * profile_capable - test if profile allows use of capability @cap
> > >   * @profile: profile being enforced    (NOT NULL, NOT unconfined)
> > >   * @cap: capability to test if allowed
> > > - * @opts: CAP_OPT_NOAUDIT bit determines whether audit record is generated
> > > + * @opts: CAP_OPT_NOAUDIT/CAP_OPT_NODENYAUDIT bit determines whether audit
> > > + * record is generated
> > >   * @sa: audit data (MAY BE NULL indicating no auditing)
> > >   *
> > >   * Returns: 0 if allowed else -EPERM
> > > @@ -126,7 +127,7 @@ static int profile_capable(struct aa_profile *profile, int cap,
> > >     else
> > >             error = -EPERM;
> > >
> > > -   if (opts & CAP_OPT_NOAUDIT) {
> > > +   if ((opts & CAP_OPT_NOAUDIT) || ((opts & CAP_OPT_NODENYAUDIT) && error)) {
> > >             if (!COMPLAIN_MODE(profile))
> > >                     return error;
> > >             /* audit the cap request in complain mode but note that it
> > > @@ -142,7 +143,8 @@ static int profile_capable(struct aa_profile *profile, int cap,
> > >   * aa_capable - test permission to use capability
> > >   * @label: label being tested for capability (NOT NULL)
> > >   * @cap: capability to be tested
> > > - * @opts: CAP_OPT_NOAUDIT bit determines whether audit record is generated
> > > + * @opts: CAP_OPT_NOAUDIT/CAP_OPT_NODENYAUDIT bit determines whether audit
> > > + * record is generated
> > >   *
> > >   * Look up capability in profile capability set.
> > >   *
> > > diff --git a/security/selinux/hooks.c b/security/selinux/hooks.c
> > > index 79b4890e9936..0730edf2f5f1 100644
> > > --- a/security/selinux/hooks.c
> > > +++ b/security/selinux/hooks.c
> > > @@ -1571,7 +1571,7 @@ static int cred_has_capability(const struct cred *cred,
> > >     u16 sclass;
> > >     u32 sid = cred_sid(cred);
> > >     u32 av = CAP_TO_MASK(cap);
> > > -   int rc;
> > > +   int rc, rc2;
> > >
> > >     ad.type = LSM_AUDIT_DATA_CAP;
> > >     ad.u.cap = cap;
> > > @@ -1590,11 +1590,13 @@ static int cred_has_capability(const struct cred *cred,
> > >     }
> > >
> > >     rc = avc_has_perm_noaudit(sid, sid, sclass, av, 0, &avd);
> > > -   if (!(opts & CAP_OPT_NOAUDIT)) {
> > > -           int rc2 = avc_audit(sid, sid, sclass, av, &avd, rc, &ad);
> > > -           if (rc2)
> > > -                   return rc2;
> > > -   }
> > > +   if ((opts & CAP_OPT_NOAUDIT) || ((opts & CAP_OPT_NODENYAUDIT) && rc))
> > > +           return rc;
> >
> > Hm, if the caller passes only CAP_OPT_NODENYAUDIT, and rc == 0, then
> > you will audit the allow.  Is that what you want, or did you want, or
> > did you want CAP_OPT_NODENYAUDIT to imply CAP_OPT_NOAUDIT?
> >

The new option should cause to issue an audit event if (and only if)
the requested capability is in effect for the current task. If the
task does not have the capability no audit event should be issued.

The new option should not imply CAP_OPT_NOAUDIT since we want an audit
event in the case the capability is in effect.

I admit the naming is a bit confusing as CAP_OPT_NODENYAUDIT as well
as the commit description contains a double negation (while the inline
comment for the macro definition does not).

Do you prefer naming the constant CAP_OPT_ALLOWAUDIT or CAP_OPT_AUDIT_ON_ALLOW?

> > > +
> > > +   rc2 = avc_audit(sid, sid, sclass, av, &avd, rc, &ad);
> > > +   if (rc2)
> > > +           return rc2;
> > > +
> > >     return rc;
> > >  }
> > >
> > > --
> > > 2.40.1