[PATCH] remove CONFIG_XFS_SECURITY

[PATCH] remove CONFIG_XFS_SECURITY

[Eric Sandeen]

Is there any point to this option?  Sure, it disables the ability
to set security attributes at runtime, but it doesn't slim down 
any code.

Any reason to not remove it, and always allow security attributes
to be set?

Signed-off-by: Eric Sandeen <sandeen@sandeen.net>

---

Index: linux-2.6.24-rc3/fs/xfs/Kconfig
===================================================================
--- linux-2.6.24-rc3.orig/fs/xfs/Kconfig
+++ linux-2.6.24-rc3/fs/xfs/Kconfig
@@ -35,18 +35,6 @@ config XFS_QUOTA
 	  with or without the generic quota support enabled (CONFIG_QUOTA) -
 	  they are completely independent subsystems.
 
-config XFS_SECURITY
-	bool "XFS Security Label support"
-	depends on XFS_FS
-	help
-	  Security labels support alternative access control models
-	  implemented by security modules like SELinux.  This option
-	  enables an extended attribute namespace for inode security
-	  labels in the XFS filesystem.
-
-	  If you are not using a security module that requires using
-	  extended attributes for inode security labels, say N.
-
 config XFS_POSIX_ACL
 	bool "XFS POSIX ACL support"
 	depends on XFS_FS
Index: linux-2.6.24-rc3/fs/xfs/linux-2.6/xfs_super.h
===================================================================
--- linux-2.6.24-rc3.orig/fs/xfs/linux-2.6/xfs_super.h
+++ linux-2.6.24-rc3/fs/xfs/linux-2.6/xfs_super.h
@@ -50,13 +50,8 @@ extern void xfs_qm_exit(void);
 # define set_posix_acl_flag(sb)	do { } while (0)
 #endif
 
-#ifdef CONFIG_XFS_SECURITY
-# define XFS_SECURITY_STRING	"security attributes, "
-# define ENOSECURITY		0
-#else
-# define XFS_SECURITY_STRING
-# define ENOSECURITY		EOPNOTSUPP
-#endif
+/* Used to be "configurable" so keep it around. */
+#define XFS_SECURITY_STRING	"security attributes, "
 
 #ifdef CONFIG_XFS_RT
 # define XFS_REALTIME_STRING	"realtime, "
Index: linux-2.6.24-rc3/fs/xfs/xfs_attr.c
===================================================================
--- linux-2.6.24-rc3.orig/fs/xfs/xfs_attr.c
+++ linux-2.6.24-rc3/fs/xfs/xfs_attr.c
@@ -2651,7 +2651,7 @@ attr_secure_capable(
 	bhv_vnode_t	*vp,
 	cred_t		*cred)
 {
-	return -ENOSECURITY;
+	return 0;
 }
 
 STATIC int

Re: [PATCH] remove CONFIG_XFS_SECURITY

[Christoph Hellwig]

On Sun, Jan 06, 2008 at 09:36:39PM -0600, Eric Sandeen wrote:
> Is there any point to this option?  Sure, it disables the ability
> to set security attributes at runtime, but it doesn't slim down 
> any code.
> 
> Any reason to not remove it, and always allow security attributes
> to be set?

I suspect the reason it is there currently is because the other
filesystems have similar option.   Removing it sounds perfectly
fine to me.

Re: [PATCH] remove CONFIG_XFS_SECURITY

[Eric Sandeen]

Eric Sandeen wrote:
> Is there any point to this option?  Sure, it disables the ability
> to set security attributes at runtime, but it doesn't slim down 
> any code.
> 
> Any reason to not remove it, and always allow security attributes
> to be set?

Ack? Nak?  Comments?

-Eric

> Signed-off-by: Eric Sandeen <sandeen@sandeen.net>
> 
> ---
> 
> Index: linux-2.6.24-rc3/fs/xfs/Kconfig
> ===================================================================
> --- linux-2.6.24-rc3.orig/fs/xfs/Kconfig
> +++ linux-2.6.24-rc3/fs/xfs/Kconfig
> @@ -35,18 +35,6 @@ config XFS_QUOTA
>  	  with or without the generic quota support enabled (CONFIG_QUOTA) -
>  	  they are completely independent subsystems.
>  
> -config XFS_SECURITY
> -	bool "XFS Security Label support"
> -	depends on XFS_FS
> -	help
> -	  Security labels support alternative access control models
> -	  implemented by security modules like SELinux.  This option
> -	  enables an extended attribute namespace for inode security
> -	  labels in the XFS filesystem.
> -
> -	  If you are not using a security module that requires using
> -	  extended attributes for inode security labels, say N.
> -
>  config XFS_POSIX_ACL
>  	bool "XFS POSIX ACL support"
>  	depends on XFS_FS
> Index: linux-2.6.24-rc3/fs/xfs/linux-2.6/xfs_super.h
> ===================================================================
> --- linux-2.6.24-rc3.orig/fs/xfs/linux-2.6/xfs_super.h
> +++ linux-2.6.24-rc3/fs/xfs/linux-2.6/xfs_super.h
> @@ -50,13 +50,8 @@ extern void xfs_qm_exit(void);
>  # define set_posix_acl_flag(sb)	do { } while (0)
>  #endif
>  
> -#ifdef CONFIG_XFS_SECURITY
> -# define XFS_SECURITY_STRING	"security attributes, "
> -# define ENOSECURITY		0
> -#else
> -# define XFS_SECURITY_STRING
> -# define ENOSECURITY		EOPNOTSUPP
> -#endif
> +/* Used to be "configurable" so keep it around. */
> +#define XFS_SECURITY_STRING	"security attributes, "
>  
>  #ifdef CONFIG_XFS_RT
>  # define XFS_REALTIME_STRING	"realtime, "
> Index: linux-2.6.24-rc3/fs/xfs/xfs_attr.c
> ===================================================================
> --- linux-2.6.24-rc3.orig/fs/xfs/xfs_attr.c
> +++ linux-2.6.24-rc3/fs/xfs/xfs_attr.c
> @@ -2651,7 +2651,7 @@ attr_secure_capable(
>  	bhv_vnode_t	*vp,
>  	cred_t		*cred)
>  {
> -	return -ENOSECURITY;
> +	return 0;
>  }
>  
>  STATIC int
> 
> 

Re: [PATCH] remove CONFIG_XFS_SECURITY

[Timothy Shimmin]

Eric Sandeen wrote:
> Eric Sandeen wrote:
>> Is there any point to this option?  Sure, it disables the ability
>> to set security attributes at runtime, but it doesn't slim down 
>> any code.
>>
>> Any reason to not remove it, and always allow security attributes
>> to be set?
> 
> Ack? Nak?  Comments?
> 
Fine by me. I'm not sure of the point either.
However, don't need to modify the attr_secure_capable function,
might as well delete it and
just use fs_noerr for the capable hook field.

--Tim

(BTW, will check in attr2 fixes soon - need to write a qa test ;-)


> -Eric
> 
>> Signed-off-by: Eric Sandeen <sandeen@sandeen.net>
>>
>> ---
>>
>> Index: linux-2.6.24-rc3/fs/xfs/Kconfig
>> ===================================================================
>> --- linux-2.6.24-rc3.orig/fs/xfs/Kconfig
>> +++ linux-2.6.24-rc3/fs/xfs/Kconfig
>> @@ -35,18 +35,6 @@ config XFS_QUOTA
>>  	  with or without the generic quota support enabled (CONFIG_QUOTA) -
>>  	  they are completely independent subsystems.
>>  
>> -config XFS_SECURITY
>> -	bool "XFS Security Label support"
>> -	depends on XFS_FS
>> -	help
>> -	  Security labels support alternative access control models
>> -	  implemented by security modules like SELinux.  This option
>> -	  enables an extended attribute namespace for inode security
>> -	  labels in the XFS filesystem.
>> -
>> -	  If you are not using a security module that requires using
>> -	  extended attributes for inode security labels, say N.
>> -
>>  config XFS_POSIX_ACL
>>  	bool "XFS POSIX ACL support"
>>  	depends on XFS_FS
>> Index: linux-2.6.24-rc3/fs/xfs/linux-2.6/xfs_super.h
>> ===================================================================
>> --- linux-2.6.24-rc3.orig/fs/xfs/linux-2.6/xfs_super.h
>> +++ linux-2.6.24-rc3/fs/xfs/linux-2.6/xfs_super.h
>> @@ -50,13 +50,8 @@ extern void xfs_qm_exit(void);
>>  # define set_posix_acl_flag(sb)	do { } while (0)
>>  #endif
>>  
>> -#ifdef CONFIG_XFS_SECURITY
>> -# define XFS_SECURITY_STRING	"security attributes, "
>> -# define ENOSECURITY		0
>> -#else
>> -# define XFS_SECURITY_STRING
>> -# define ENOSECURITY		EOPNOTSUPP
>> -#endif
>> +/* Used to be "configurable" so keep it around. */
>> +#define XFS_SECURITY_STRING	"security attributes, "
>>  
>>  #ifdef CONFIG_XFS_RT
>>  # define XFS_REALTIME_STRING	"realtime, "
>> Index: linux-2.6.24-rc3/fs/xfs/xfs_attr.c
>> ===================================================================
>> --- linux-2.6.24-rc3.orig/fs/xfs/xfs_attr.c
>> +++ linux-2.6.24-rc3/fs/xfs/xfs_attr.c
>> @@ -2651,7 +2651,7 @@ attr_secure_capable(
>>  	bhv_vnode_t	*vp,
>>  	cred_t		*cred)
>>  {
>> -	return -ENOSECURITY;
>> +	return 0;
>>  }
>>  
>>  STATIC int
>>
>>
> 

[PATCH V2] remove CONFIG_XFS_SECURITY

[Eric Sandeen]

Timothy Shimmin wrote:

> Eric Sandeen wrote:
>   
>> Eric Sandeen wrote:
>>     
>>> Is there any point to this option?  Sure, it disables the ability
>>> to set security attributes at runtime, but it doesn't slim down 
>>> any code.
>>>
>>> Any reason to not remove it, and always allow security attributes
>>> to be set?
>>>       
>> Ack? Nak?  Comments?
>>
>>     
> Fine by me. I'm not sure of the point either.
> However, don't need to modify the attr_secure_capable function,
> might as well delete it and
> just use fs_noerr for the capable hook field.
>   
Oh, I see, it's normally cast... ok.

-----------------------------

There is no point to the CONFIG_XFS_SECURITY option; it
disables the ability to set security attributes at runtime,
but it does not actually slim down or remove any code for
runtime.  Just remove it and always allow security attributes
to be set.

Signed-off-by: Eric Sandeen <sandeen@sandeen.net>
---

Index: linux-2.6-xfs/fs/xfs/Kconfig
===================================================================
--- linux-2.6-xfs.orig/fs/xfs/Kconfig
+++ linux-2.6-xfs/fs/xfs/Kconfig
@@ -48,18 +48,6 @@ config XFS_DMAPI
 
 	  If unsure, say N.
 
-config XFS_SECURITY
-	bool "XFS Security Label support"
-	depends on XFS_FS
-	help
-	  Security labels support alternative access control models
-	  implemented by security modules like SELinux.  This option
-	  enables an extended attribute namespace for inode security
-	  labels in the XFS filesystem.
-
-	  If you are not using a security module that requires using
-	  extended attributes for inode security labels, say N.
-
 config XFS_POSIX_ACL
 	bool "XFS POSIX ACL support"
 	depends on XFS_FS
Index: linux-2.6-xfs/fs/xfs/linux-2.6/xfs_super.h
===================================================================
--- linux-2.6-xfs.orig/fs/xfs/linux-2.6/xfs_super.h
+++ linux-2.6-xfs/fs/xfs/linux-2.6/xfs_super.h
@@ -28,13 +28,7 @@
 # define set_posix_acl_flag(sb)	do { } while (0)
 #endif
 
-#ifdef CONFIG_XFS_SECURITY
-# define XFS_SECURITY_STRING	"security attributes, "
-# define ENOSECURITY		0
-#else
-# define XFS_SECURITY_STRING
-# define ENOSECURITY		EOPNOTSUPP
-#endif
+#define XFS_SECURITY_STRING	"security attributes, "
 
 #ifdef CONFIG_XFS_RT
 # define XFS_REALTIME_STRING	"realtime, "
Index: linux-2.6-xfs/fs/xfs/xfs_attr.c
===================================================================
--- linux-2.6-xfs.orig/fs/xfs/xfs_attr.c
+++ linux-2.6-xfs/fs/xfs/xfs_attr.c
@@ -2647,14 +2647,6 @@ attr_trusted_capable(
 }
 
 STATIC int
-attr_secure_capable(
-	bhv_vnode_t	*vp,
-	cred_t		*cred)
-{
-	return -ENOSECURITY;
-}
-
-STATIC int
 attr_system_set(
 	bhv_vnode_t *vp, char *name, void *data, size_t size, int xflags)
 {
@@ -2724,7 +2716,7 @@ struct attrnames attr_secure = {
 	.attr_get	= attr_generic_get,
 	.attr_set	= attr_generic_set,
 	.attr_remove	= attr_generic_remove,
-	.attr_capable	= attr_secure_capable,
+	.attr_capable	= (attrcapable_t)fs_noerr,
 };
 
 struct attrnames attr_user = {

Re: [PATCH V2] remove CONFIG_XFS_SECURITY

[Christoph Hellwig]

> +#define XFS_SECURITY_STRING	"security attributes, "

This one should just be killed completely, as it's not an optional
feature anymore.

Re: [PATCH V2] remove CONFIG_XFS_SECURITY

[Eric Sandeen]

Christoph Hellwig wrote:
>> +#define XFS_SECURITY_STRING	"security attributes, "
> 
> This one should just be killed completely, as it's not an optional
> feature anymore.
> 

My only concern is that if people are used to seeing it, they'll assume
that somehow xfs now doesn't support it now, or was built without it, or ...

*shrug*

-Eric

Re: [PATCH V2] remove CONFIG_XFS_SECURITY

[Timothy Shimmin]

Eric Sandeen wrote:
> Christoph Hellwig wrote:
>>> +#define XFS_SECURITY_STRING	"security attributes, "
>> This one should just be killed completely, as it's not an optional
>> feature anymore.
>>
> 
> My only concern is that if people are used to seeing it, they'll assume
> that somehow xfs now doesn't support it now, or was built without it, or ...
> 
> *shrug*
> 
> -Eric
Yeah, I kind of see your point there.
I think I'd prefer to keep it too.

--Tim