[PATCH V3] security/selinux: Always allow FIOCLEX and FIONCLEX

[PATCH V3] security/selinux: Always allow FIOCLEX and FIONCLEX

[Richard Haines]

These ioctls are equivalent to fcntl(fd, F_SETFD, flags), which SELinux
always allows too.  Furthermore, a failed FIOCLEX could result in a file
descriptor being leaked to a process that should not have access to it.

As this patch removes access controls, a policy capability needs to be
enabled in policy to always allow these ioctls.

Based-on-patch-by: Demi Marie Obenour <demiobenour@gmail.com>
Signed-off-by: Richard Haines <richard_c_haines@btinternet.com>
---
V2 Change: Control via a policy capability.
V3 Change: Update switch check.

 security/selinux/hooks.c                   | 6 ++++++
 security/selinux/include/policycap.h       | 1 +
 security/selinux/include/policycap_names.h | 3 ++-
 security/selinux/include/security.h        | 7 +++++++
 4 files changed, 16 insertions(+), 1 deletion(-)

diff --git a/security/selinux/hooks.c b/security/selinux/hooks.c
index 5b6895e4f..d369c2d82 100644
--- a/security/selinux/hooks.c
+++ b/security/selinux/hooks.c
@@ -3745,6 +3745,12 @@ static int selinux_file_ioctl(struct file *file, unsigned int cmd,
 					    CAP_OPT_NONE, true);
 		break;
 
+	case FIOCLEX:
+	case FIONCLEX:
+		if (!selinux_policycap_ioctl_skip_cloexec())
+			error = ioctl_has_perm(cred, file, FILE__IOCTL, (u16) cmd);
+		break;
+
 	/* default case assumes that the command will go
 	 * to the file's ioctl() function.
 	 */
diff --git a/security/selinux/include/policycap.h b/security/selinux/include/policycap.h
index 2ec038efb..44d73dc32 100644
--- a/security/selinux/include/policycap.h
+++ b/security/selinux/include/policycap.h
@@ -11,6 +11,7 @@ enum {
 	POLICYDB_CAPABILITY_CGROUPSECLABEL,
 	POLICYDB_CAPABILITY_NNP_NOSUID_TRANSITION,
 	POLICYDB_CAPABILITY_GENFS_SECLABEL_SYMLINKS,
+	POLICYDB_CAPABILITY_IOCTL_CLOEXEC,
 	__POLICYDB_CAPABILITY_MAX
 };
 #define POLICYDB_CAPABILITY_MAX (__POLICYDB_CAPABILITY_MAX - 1)
diff --git a/security/selinux/include/policycap_names.h b/security/selinux/include/policycap_names.h
index b89289f09..ebd64afe1 100644
--- a/security/selinux/include/policycap_names.h
+++ b/security/selinux/include/policycap_names.h
@@ -12,7 +12,8 @@ const char *selinux_policycap_names[__POLICYDB_CAPABILITY_MAX] = {
 	"always_check_network",
 	"cgroup_seclabel",
 	"nnp_nosuid_transition",
-	"genfs_seclabel_symlinks"
+	"genfs_seclabel_symlinks",
+	"ioctl_skip_cloexec"
 };
 
 #endif /* _SELINUX_POLICYCAP_NAMES_H_ */
diff --git a/security/selinux/include/security.h b/security/selinux/include/security.h
index ac0ece013..8a789c22b 100644
--- a/security/selinux/include/security.h
+++ b/security/selinux/include/security.h
@@ -219,6 +219,13 @@ static inline bool selinux_policycap_genfs_seclabel_symlinks(void)
 	return READ_ONCE(state->policycap[POLICYDB_CAPABILITY_GENFS_SECLABEL_SYMLINKS]);
 }
 
+static inline bool selinux_policycap_ioctl_skip_cloexec(void)
+{
+	struct selinux_state *state = &selinux_state;
+
+	return READ_ONCE(state->policycap[POLICYDB_CAPABILITY_IOCTL_CLOEXEC]);
+}
+
 struct selinux_policy_convert_data;
 
 struct selinux_load_state {
-- 
2.35.1

Re: [PATCH V3] security/selinux: Always allow FIOCLEX and FIONCLEX

[Paul Moore]

On Thu, Feb 24, 2022 at 5:24 AM Richard Haines
<richard_c_haines@btinternet.com> wrote:
>
> These ioctls are equivalent to fcntl(fd, F_SETFD, flags), which SELinux
> always allows too.  Furthermore, a failed FIOCLEX could result in a file
> descriptor being leaked to a process that should not have access to it.
>
> As this patch removes access controls, a policy capability needs to be
> enabled in policy to always allow these ioctls.
>
> Based-on-patch-by: Demi Marie Obenour <demiobenour@gmail.com>
> Signed-off-by: Richard Haines <richard_c_haines@btinternet.com>
> ---
> V2 Change: Control via a policy capability.
> V3 Change: Update switch check.
>
>  security/selinux/hooks.c                   | 6 ++++++
>  security/selinux/include/policycap.h       | 1 +
>  security/selinux/include/policycap_names.h | 3 ++-
>  security/selinux/include/security.h        | 7 +++++++
>  4 files changed, 16 insertions(+), 1 deletion(-)

This looks good to me, but before I merge this are the SELinux
userspace folks okay with the policy capability's name and enum value?

> diff --git a/security/selinux/hooks.c b/security/selinux/hooks.c
> index 5b6895e4f..d369c2d82 100644
> --- a/security/selinux/hooks.c
> +++ b/security/selinux/hooks.c
> @@ -3745,6 +3745,12 @@ static int selinux_file_ioctl(struct file *file, unsigned int cmd,
>                                             CAP_OPT_NONE, true);
>                 break;
>
> +       case FIOCLEX:
> +       case FIONCLEX:
> +               if (!selinux_policycap_ioctl_skip_cloexec())
> +                       error = ioctl_has_perm(cred, file, FILE__IOCTL, (u16) cmd);
> +               break;
> +
>         /* default case assumes that the command will go
>          * to the file's ioctl() function.
>          */
> diff --git a/security/selinux/include/policycap.h b/security/selinux/include/policycap.h
> index 2ec038efb..44d73dc32 100644
> --- a/security/selinux/include/policycap.h
> +++ b/security/selinux/include/policycap.h
> @@ -11,6 +11,7 @@ enum {
>         POLICYDB_CAPABILITY_CGROUPSECLABEL,
>         POLICYDB_CAPABILITY_NNP_NOSUID_TRANSITION,
>         POLICYDB_CAPABILITY_GENFS_SECLABEL_SYMLINKS,
> +       POLICYDB_CAPABILITY_IOCTL_CLOEXEC,
>         __POLICYDB_CAPABILITY_MAX
>  };
>  #define POLICYDB_CAPABILITY_MAX (__POLICYDB_CAPABILITY_MAX - 1)
> diff --git a/security/selinux/include/policycap_names.h b/security/selinux/include/policycap_names.h
> index b89289f09..ebd64afe1 100644
> --- a/security/selinux/include/policycap_names.h
> +++ b/security/selinux/include/policycap_names.h
> @@ -12,7 +12,8 @@ const char *selinux_policycap_names[__POLICYDB_CAPABILITY_MAX] = {
>         "always_check_network",
>         "cgroup_seclabel",
>         "nnp_nosuid_transition",
> -       "genfs_seclabel_symlinks"
> +       "genfs_seclabel_symlinks",
> +       "ioctl_skip_cloexec"
>  };
>
>  #endif /* _SELINUX_POLICYCAP_NAMES_H_ */
> diff --git a/security/selinux/include/security.h b/security/selinux/include/security.h
> index ac0ece013..8a789c22b 100644
> --- a/security/selinux/include/security.h
> +++ b/security/selinux/include/security.h
> @@ -219,6 +219,13 @@ static inline bool selinux_policycap_genfs_seclabel_symlinks(void)
>         return READ_ONCE(state->policycap[POLICYDB_CAPABILITY_GENFS_SECLABEL_SYMLINKS]);
>  }
>
> +static inline bool selinux_policycap_ioctl_skip_cloexec(void)
> +{
> +       struct selinux_state *state = &selinux_state;
> +
> +       return READ_ONCE(state->policycap[POLICYDB_CAPABILITY_IOCTL_CLOEXEC]);
> +}
> +
>  struct selinux_policy_convert_data;
>
>  struct selinux_load_state {
> --
> 2.35.1

-- 
paul-moore.com

Re: [PATCH V3] security/selinux: Always allow FIOCLEX and FIONCLEX

[Ondrej Mosnacek]

On Fri, Feb 25, 2022 at 1:34 AM Paul Moore <paul@paul-moore.com> wrote:
> On Thu, Feb 24, 2022 at 5:24 AM Richard Haines
> <richard_c_haines@btinternet.com> wrote:
> >
> > These ioctls are equivalent to fcntl(fd, F_SETFD, flags), which SELinux
> > always allows too.  Furthermore, a failed FIOCLEX could result in a file
> > descriptor being leaked to a process that should not have access to it.
> >
> > As this patch removes access controls, a policy capability needs to be
> > enabled in policy to always allow these ioctls.
> >
> > Based-on-patch-by: Demi Marie Obenour <demiobenour@gmail.com>
> > Signed-off-by: Richard Haines <richard_c_haines@btinternet.com>
> > ---
> > V2 Change: Control via a policy capability.
> > V3 Change: Update switch check.
> >
> >  security/selinux/hooks.c                   | 6 ++++++
> >  security/selinux/include/policycap.h       | 1 +
> >  security/selinux/include/policycap_names.h | 3 ++-
> >  security/selinux/include/security.h        | 7 +++++++
> >  4 files changed, 16 insertions(+), 1 deletion(-)
>
> This looks good to me, but before I merge this are the SELinux
> userspace folks okay with the policy capability's name and enum value?

Since you mention it... I would suggest naming the enum
POLICYDB_CAPABILITY_IOCTL_SKIP_CLOEXEC to match the display name. Yes,
it becomes awkwardly long, but e.g.
POLICYDB_CAPABILITY_GENFS_SECLABEL_SYMLINKS is already longer than
that, so I'd prefer more descriptiveness over brevity.

(IMHO the POLICYDB_CAPABILITY_ prefix is ridiculously long for no
reason and we should simply shorten it (just POLCAP_ would be
perfectly fine, IMHO) instead of trying to abbreviate the rest. Of
course, this doesn't have to be done now - I'm taking a note to myself
to splice in such rename next time I add a new capability, if not
earlier.)

--
Ondrej Mosnacek
Software Engineer, Linux Security - SELinux kernel
Red Hat, Inc.

Re: [PATCH V3] security/selinux: Always allow FIOCLEX and FIONCLEX

[Paul Moore]

On Fri, Feb 25, 2022 at 5:02 AM Ondrej Mosnacek <omosnace@redhat.com> wrote:
> (IMHO the POLICYDB_CAPABILITY_ prefix is ridiculously long for no
> reason and we should simply shorten it (just POLCAP_ would be
> perfectly fine, IMHO) instead of trying to abbreviate the rest. Of
> course, this doesn't have to be done now - I'm taking a note to myself
> to splice in such rename next time I add a new capability, if not
> earlier.)

Yeah, that prefix is probably longer than it needs to be.  Since we're
at -rc5 right now, with -rc6 just a few days away, I'll go ahead and
toss out a patch to shorten the names next week.  While I have very
mixed feelings on style/formatting-only patches, things like this
should go into selinux/next at the end of the -rcX cycle so as not to
cause unnecessary conflicts for people making "real" changes :)

-- 
paul-moore.com