Skip to contentSkip to conversation

Remove genfscon rule in selinux.if

PATCH New James Carter 3 participants 4 messages Aug 24, 2010 – Aug 25, 2010

lore
Share & tools
lore
#0 lore raw

PATCH 1/6 Remove genfscon rule in selinux.if

This is obviously not a solution. The problem here is that m4 is being used to perform string concatenation. The argument, which is a boolean, is not being used like a boolean and this is a problem when you are inferring data types.

The interface is not being used, so ignoring it doesn't cause a problem for now.

---

 policy/modules/kernel/selinux.if |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
policy/modules/kernel/selinux.if +1 -1
index f8b357c..c1d0d98 100644
@@ -40,7 +40,7 @@ interface(`selinux_labeled_boolean',`
 
        # because of this statement, any module which
        # calls this interface must be in the base module:
-       genfscon selinuxfs /booleans/$2 gen_context(system_u:object_r:$1,s0)
+       #genfscon selinuxfs /booleans/$2 gen_context(system_u:object_r:$1,s0)
 ')
 
 ########################################
Signature · 2 lines
James Carter <[email protected]>
National Security Agency
#1 lore raw

PATCH 1/6 Remove genfscon rule in selinux.if

Quoted 7 lines · James Carter

This is obviously not a solution. The problem here is that m4 is being used to perform string concatenation. The argument, which is a boolean, is not being used like a boolean and this is a problem when you are inferring data types.

The interface is not being used, so ignoring it doesn't cause a problem for now.

If I'm not mistaken, Dan uses this in the Fedora policy. We've also used this in internal Tresys projects.

Quoted 18 lines
---
  policy/modules/kernel/selinux.if |    2 +-
  1 file changed, 1 insertion(+), 1 deletion(-)
policy/modules/kernel/selinux.if +1 -1
index f8b357c..c1d0d98 100644
@@ -40,7 +40,7 @@ interface(`selinux_labeled_boolean',`
 
         # because of this statement, any module which
         # calls this interface must be in the base module:
-       genfscon selinuxfs /booleans/$2 gen_context(system_u:object_r:$1,s0)
+       #genfscon selinuxfs /booleans/$2 gen_context(system_u:object_r:$1,s0)
  ')
 
  ########################################
Signature · 3 lines
Chris PeBenito
Tresys Technology, LLC
www.tresys.com | oss.tresys.com
#2 lore raw

PATCH 1/6 Remove genfscon rule in selinux.if

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Quoted 22 lines · Christopher J. PeBenito
Quoted 7 lines · James Carter

This is obviously not a solution. The problem here is that m4 is being used to perform string concatenation. The argument, which is a boolean, is not being used like a boolean and this is a problem when you are inferring data types.

The interface is not being used, so ignoring it doesn't cause a problem for now.

If I'm not mistaken, Dan uses this in the Fedora policy. We've also used this in internal Tresys projects.

Quoted 10 lines
---
  policy/modules/kernel/selinux.if |    2 +-
  1 file changed, 1 insertion(+), 1 deletion(-)
policy/modules/kernel/selinux.if +0 -0
index f8b357c..c1d0d98 100644
@@ -40,7 +40,7 @@ interface(`selinux_labeled_boolean',`
 
Quoted 10 lines

policy/modules/kernel/selinux.if interface(`selinux_labeled_boolean',` · line 46

Quoted 8 lines
policy/modules/kernel/selinux.if +1 -1
         # because of this statement, any module which
         # calls this interface must be in the base module:
-       genfscon selinuxfs /booleans/$2 gen_context(system_u:object_r:$1,s0)
+       #genfscon selinuxfs /booleans/$2 gen_context(system_u:object_r:$1,s0)
  ')
 
  ########################################

I used it but until we can define booleans in modules it is fairly useless. -----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.16 (GNU/Linux)
Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org/

iEYEARECAAYFAkx1IQEACgkQrlYvE4MpobPzZACgqtxlXjXMcl5Dv8CJHfAlLULq drAAoOJn7pieDHIqJ6zOB7LqRNtSWP7q
=W9Bp
-----END PGP SIGNATURE-----

#3 lore raw

PATCH 1/6 Remove genfscon rule in selinux.if

In reply to Christopher J. PeBenito
Quoted 6 lines · Christopher J. PeBenito
Quoted 5 lines · James Carter

This is obviously not a solution. The problem here is that m4 is being used to perform string concatenation. The argument, which is a boolean, is not being used like a boolean and this is a problem when you are inferring data types.

Quoted 6 lines
Quoted 2 lines

The interface is not being used, so ignoring it doesn't cause a problem for now.

If I'm not mistaken, Dan uses this in the Fedora policy. We've also used this in internal Tresys projects.

I was afraid that someone was using it.

I guess we will have to support some sort of string concatenation in CIL.

Quoted 20 lines
Quoted 18 lines
---
  policy/modules/kernel/selinux.if |    2 +-
  1 file changed, 1 insertion(+), 1 deletion(-)
policy/modules/kernel/selinux.if +1 -1
index f8b357c..c1d0d98 100644
@@ -40,7 +40,7 @@ interface(`selinux_labeled_boolean',`
 
         # because of this statement, any module which
         # calls this interface must be in the base module:
-       genfscon selinuxfs /booleans/$2 gen_context(system_u:object_r:$1,s0)
+       #genfscon selinuxfs /booleans/$2 gen_context(system_u:object_r:$1,s0)
  ')
 
  ########################################
Signature · 2 lines
James Carter <[email protected]>
National Security Agency