Received-SPF: pass (google.com: domain of linux-nfs+bounces-162-linux.lists.archive=gmail.com@vger.kernel.org designates 147.75.199.223 as permitted sender) client-ip=147.75.199.223;
Message-ID: <b6c51351be3913be197492469a13980ab379e412.camel@huaweicloud.com>
Subject: Re: [PATCH v5 23/23] integrity: Switch from rbtree to LSM-managed
 blob for integrity_iint_cache
From: Roberto Sassu <roberto.sassu@huaweicloud.com>
To: Paul Moore <paul@paul-moore.com>
Cc: viro@zeniv.linux.org.uk, brauner@kernel.org, chuck.lever@oracle.com, 
 jlayton@kernel.org, neilb@suse.de, kolga@netapp.com, Dai.Ngo@oracle.com, 
 tom@talpey.com, jmorris@namei.org, serge@hallyn.com, zohar@linux.ibm.com, 
 dmitry.kasatkin@gmail.com, dhowells@redhat.com, jarkko@kernel.org, 
 stephen.smalley.work@gmail.com, eparis@parisplace.org,
 casey@schaufler-ca.com,  mic@digikod.net, linux-fsdevel@vger.kernel.org,
 linux-kernel@vger.kernel.org,  linux-nfs@vger.kernel.org,
 linux-security-module@vger.kernel.org,  linux-integrity@vger.kernel.org,
 keyrings@vger.kernel.org,  selinux@vger.kernel.org, Roberto Sassu
 <roberto.sassu@huawei.com>
Date: Wed, 29 Nov 2023 13:27:51 +0100
In-Reply-To: <CAHC9VhTTKac1o=RnQadu2xqdeKH8C_F+Wh4sY=HkGbCArwc8JQ@mail.gmail.com>
References: <20231107134012.682009-24-roberto.sassu@huaweicloud.com>
	 <17befa132379d37977fc854a8af25f6d.paul@paul-moore.com>
	 <2084adba3c27a606cbc5ed7b3214f61427a829dd.camel@huaweicloud.com>
	 <CAHC9VhTTKac1o=RnQadu2xqdeKH8C_F+Wh4sY=HkGbCArwc8JQ@mail.gmail.com>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
User-Agent: Evolution 3.44.4-0ubuntu2 
Precedence: bulk
MIME-Version: 1.0

On Mon, 2023-11-20 at 16:06 -0500, Paul Moore wrote:
> On Mon, Nov 20, 2023 at 3:16=E2=80=AFAM Roberto Sassu
> <roberto.sassu@huaweicloud.com> wrote:
> > On Fri, 2023-11-17 at 15:57 -0500, Paul Moore wrote:
> > > On Nov  7, 2023 Roberto Sassu <roberto.sassu@huaweicloud.com> wrote:
> > > >=20
> > > > Before the security field of kernel objects could be shared among L=
SMs with
> > > > the LSM stacking feature, IMA and EVM had to rely on an alternative=
 storage
> > > > of inode metadata. The association between inode metadata and inode=
 is
> > > > maintained through an rbtree.
> > > >=20
> > > > Because of this alternative storage mechanism, there was no need to=
 use
> > > > disjoint inode metadata, so IMA and EVM today still share them.
> > > >=20
> > > > With the reservation mechanism offered by the LSM infrastructure, t=
he
> > > > rbtree is no longer necessary, as each LSM could reserve a space in=
 the
> > > > security blob for each inode. However, since IMA and EVM share the
> > > > inode metadata, they cannot directly reserve the space for them.
> > > >=20
> > > > Instead, request from the 'integrity' LSM a space in the security b=
lob for
> > > > the pointer of inode metadata (integrity_iint_cache structure). The=
 other
> > > > reason for keeping the 'integrity' LSM is to preserve the original =
ordering
> > > > of IMA and EVM functions as when they were hardcoded.
> > > >=20
> > > > Prefer reserving space for a pointer to allocating the integrity_ii=
nt_cache
> > > > structure directly, as IMA would require it only for a subset of in=
odes.
> > > > Always allocating it would cause a waste of memory.
> > > >=20
> > > > Introduce two primitives for getting and setting the pointer of
> > > > integrity_iint_cache in the security blob, respectively
> > > > integrity_inode_get_iint() and integrity_inode_set_iint(). This wou=
ld make
> > > > the code more understandable, as they directly replace rbtree opera=
tions.
> > > >=20
> > > > Locking is not needed, as access to inode metadata is not shared, i=
t is per
> > > > inode.
> > > >=20
> > > > Signed-off-by: Roberto Sassu <roberto.sassu@huawei.com>
> > > > Reviewed-by: Casey Schaufler <casey@schaufler-ca.com>
> > > > Reviewed-by: Mimi Zohar <zohar@linux.ibm.com>
> > > > ---
> > > >  security/integrity/iint.c      | 71 +++++-------------------------=
----
> > > >  security/integrity/integrity.h | 20 +++++++++-
> > > >  2 files changed, 29 insertions(+), 62 deletions(-)
> > > >=20
> > > > diff --git a/security/integrity/iint.c b/security/integrity/iint.c
> > > > index 882fde2a2607..a5edd3c70784 100644
> > > > --- a/security/integrity/iint.c
> > > > +++ b/security/integrity/iint.c
> > > > @@ -231,6 +175,10 @@ static int __init integrity_lsm_init(void)
> > > >     return 0;
> > > >  }
> > > >=20
> > > > +struct lsm_blob_sizes integrity_blob_sizes __ro_after_init =3D {
> > > > +   .lbs_inode =3D sizeof(struct integrity_iint_cache *),
> > > > +};
> > >=20
> > > I'll admit that I'm likely missing an important detail, but is there
> > > a reason why you couldn't stash the integrity_iint_cache struct
> > > directly in the inode's security blob instead of the pointer?  For
> > > example:
> > >=20
> > >   struct lsm_blob_sizes ... =3D {
> > >     .lbs_inode =3D sizeof(struct integrity_iint_cache),
> > >   };
> > >=20
> > >   struct integrity_iint_cache *integrity_inode_get(inode)
> > >   {
> > >     if (unlikely(!inode->isecurity))
> > >       return NULL;
> > >     return inode->i_security + integrity_blob_sizes.lbs_inode;
> > >   }
> >=20
> > It would increase memory occupation. Sometimes the IMA policy
> > encompasses a small subset of the inodes. Allocating the full
> > integrity_iint_cache would be a waste of memory, I guess?
>=20
> Perhaps, but if it allows us to remove another layer of dynamic memory
> I would argue that it may be worth the cost.  It's also worth
> considering the size of integrity_iint_cache, while it isn't small, it
> isn't exactly huge either.
>=20
> > On the other hand... (did not think fully about that) if we embed the
> > full structure in the security blob, we already have a mutex available
> > to use, and we don't need to take the inode lock (?).
>=20
> That would be excellent, getting rid of a layer of locking would be signi=
ficant.
>=20
> > I'm fully convinced that we can improve the implementation
> > significantly. I just was really hoping to go step by step and not
> > accumulating improvements as dependency for moving IMA and EVM to the
> > LSM infrastructure.
>=20
> I understand, and I agree that an iterative approach is a good idea, I
> just want to make sure we keep things tidy from a user perspective,
> i.e. not exposing the "integrity" LSM when it isn't required.

Ok, I went back to it again.

I think trying to separate integrity metadata is premature now, too
many things at the same time.

I started to think, does EVM really need integrity metadata or it can
work without?

The fact is that CONFIG_IMA=3Dn and CONFIG_EVM=3Dy is allowed, so we have
the same problem now. What if we make IMA the one that manages
integrity metadata, so that we can remove the 'integrity' LSM?

So, no embedding the full structure in the security blob now, move
integrity_inode_free() and integrity_kernel_module_request() to IMA,
call integrity_iintcache_init() from IMA.

EVM verification of new files would fail without IMA, but it would be
the same now.

Also, evm_verifyxattr() would only work with IMA, as it assumes that
the latter creates integrity metadata and passes them as argument.

Regarding the LSM order, I would take Casey's suggestion of introducing
LSM_ORDER_REALLY_LAST, for EVM.

Thanks

Roberto