Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Message-ID: <b6a0475c67e975cf564e4b78830ae0e598c584ac.camel@linux.ibm.com>
Subject: Re: [PATCH] ima: require signed IMA policy when UEFI secure boot is
 enabled
From:   Mimi Zohar <zohar@linux.ibm.com>
To:     Coiby Xu <coxu@redhat.com>
Cc:     linux-integrity@vger.kernel.org,
        Dmitry Kasatkin <dmitry.kasatkin@gmail.com>,
        Paul Moore <paul@paul-moore.com>,
        James Morris <jmorris@namei.org>,
        "Serge E. Hallyn" <serge@hallyn.com>,
        "open list:SECURITY SUBSYSTEM" 
        <linux-security-module@vger.kernel.org>,
        open list <linux-kernel@vger.kernel.org>
Date:   Thu, 20 Jul 2023 10:12:35 -0400
In-Reply-To: <kglacif5fwh3byrjkmlzygpnaecdb2zxh6hbglnvdbsfu7hbxm@gc3mjr37mnpb>
References: <20230703115442.129725-1-coxu@redhat.com>
         <b61fedf214cbe72de063a3bf516dd72f80595219.camel@linux.ibm.com>
         <kglacif5fwh3byrjkmlzygpnaecdb2zxh6hbglnvdbsfu7hbxm@gc3mjr37mnpb>
Content-Type: text/plain; charset="ISO-8859-15"
Mime-Version: 1.0
Content-Transfer-Encoding: 7bit
Precedence: bulk

On Fri, 2023-07-14 at 09:29 +0800, Coiby Xu wrote:
> On Tue, Jul 04, 2023 at 08:57:10AM -0400, Mimi Zohar wrote:
> >On Mon, 2023-07-03 at 19:54 +0800, Coiby Xu wrote:
> >> With the introduction of the .machine keyring for UEFI-based systems,
> >> users are able to add custom CAs keys via MOK. This allow users to sign
> >> their own IMA polices. For the sake of security, mandate signed IMA
> >> policy when UEFI secure boot is enabled.
> >>
> >> Suggested-by: Mimi Zohar <zohar@linux.ibm.com>
> >> Signed-off-by: Coiby Xu <coxu@redhat.com>

With commit 099f26f22f58 ("integrity: machine keyring CA
configuration") it is now possible to require signed IMA policies
without having to recompile the kernel.  Please note this change might
affect existing users/tests.

> >> ---
> >>  security/integrity/ima/ima_efi.c | 3 +++
> >>  1 file changed, 3 insertions(+)
> >>
> >> diff --git a/security/integrity/ima/ima_efi.c b/security/integrity/ima/ima_efi.c
> >> index 9db66fe310d4..bb2881759505 100644
> >> --- a/security/integrity/ima/ima_efi.c
> >> +++ b/security/integrity/ima/ima_efi.c
> >> @@ -58,6 +58,9 @@ static const char * const sb_arch_rules[] = {
> >>  #if !IS_ENABLED(CONFIG_MODULE_SIG)
> >>  	"appraise func=MODULE_CHECK appraise_type=imasig",
> >>  #endif
> >> +#if IS_ENABLED(CONFIG_INTEGRITY_MACHINE_KEYRING) && IS_ENABLED(CONFIG_IMA_KEYRINGS_PERMIT_SIGNED_BY_BUILTIN_OR_SECONDARY)
> >> +	"appraise func=POLICY_CHECK appraise_type=imasig",
> >> +#endif /* CONFIG_INTEGRITY_MACHINE_KEYRING && IMA_KEYRINGS_PERMIT_SIGNED_BY_BUILTIN_OR_SECONDARY */
> >>  	"measure func=MODULE_CHECK",
> >>  	NULL
> >>  };
> >
> >Thanks, Coiby.
> 
> You are welcome!
> 
> >
> >Using IS_ENABLED() is not wrong, but unnecessary.  IS_BUILTIN()
> >suffices.
> 
> Thanks for the reminding! When I was going to apply this suggestion, I
> noticed ima_efi.c uses IS_ENABLED for all configuration items i.e.
> CONFIG_MODULE_SIG and CONFIG_KEXEC_SIG which are all of bool type. Would
> you like me to switch all IS_ENABLEs to IS_BUILTIN or still use
> IS_ENABLED? While IS_BUILTIN is exclusively for bool type, it's not as
> intuitive as IS_ENABLED. So it's not easy for me to make a choice.

Sure, for consistency with the other rules IS_ENABLED is fine.

thanks,

Mimi