linux-kernel - Re: [PATCH] Smack: check for SMACK xattr max size in smack_inode

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <49E60B2D.5080508@schaufler-ca.com>
Date:	Wed, 15 Apr 2009 09:28:29 -0700
From:	Casey Schaufler <casey@...aufler-ca.com>
To:	Etienne Basset <etienne.basset@...ericable.fr>
CC:	LSM <linux-security-module@...r.kernel.org>,
	Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
	Casey Schaufler <casey@...aufler-ca.com>
Subject: Re: [PATCH] Smack: check for SMACK xattr max size in  smack_inode_setxattr

Etienne Basset wrote:
> Casey Schaufler wrote:
>   
>> Etienne Basset wrote:
>>     
>>> the following patch move the check for SMACK xattr size >= SMK_LABELLEN
>>> from  smack_inode_post_setxattr (which cannot return an error to the user)
>>> to smack_inode_setxattr (which can return an error).
>>>
>>> without patch a SMACK setxattr with size >= SMK_LABELLEN returns success although it doesn't 
>>>   
>>>       
>> How about an early call to smk_import() to determine if the
>> label if legitimate, rather than checking the length? That
>> could save grief for other "invalid label" issues.
>>     
>
> Hi,
>
> both are needed. And we'll need one call in setxattr to check the validity and one in
> postsetxattr/ inode_setsecurity (to set the security blob)
> (which make me think we should split smk_import in 2,  smk_check/smk_import_nocheck or whatever)
>
> The harder issue is that smk_import can change the value/size of the xattr
> and smk_import can fail for 2 reasons : ENOMEM or EINVAL (not sure this is an issue though)
> (we should really split)
>
> root@...enne-desktop:/tmp# attr -S -s SMACK64 -V '123////' toto
> Attribute "SMACK64" set to a 7 byte value for toto:
> 123////
> root@...enne-desktop:/tmp# attr -S -g SMACK64 toto
> Attribute "SMACK64" had a 4 byte value for toto:
> 123
>
> maybe it would be nicer to fail than change the label under the foot of the-user-who-hasnt-read-the-doc?
>   

Clumsy attribute setting (trailing spaces, NULLs, etc) is just too easy.
I like the existing behavior.

> Etienne
>   
>>> Signed-off-by: Etienne Basset <etienne.basset@...ericable.fr>
>>> ---
>>> diff --git a/security/smack/smack_lsm.c b/security/smack/smack_lsm.c
>>> index 9215149..da6954d 100644
>>> --- a/security/smack/smack_lsm.c
>>> +++ b/security/smack/smack_lsm.c
>>> @@ -610,7 +610,8 @@ static int smack_inode_setxattr(struct dentry *dentry, const char *name,
>>>  		if (!capable(CAP_MAC_ADMIN))
>>>  			rc = -EPERM;
>>>  		/* a label cannot be void and cannot begin with '-' */
>>> -		if (size == 0 || (size > 0 && ((char *)value)[0] == '-'))
>>> +		if (size == 0 || size >= SMK_LABELLEN ||
>>> +		   (size > 0 && ((char *)value)[0] == '-'))
>>>  			rc = -EINVAL;
>>>  	} else
>>>  		rc = cap_inode_setxattr(dentry, name, value, size, flags);
>>> @@ -644,9 +645,6 @@ static void smack_inode_post_setxattr(struct dentry *dentry, const char *name,
>>>  	if (strcmp(name, XATTR_NAME_SMACK))
>>>  		return;
>>>  
>>> -	if (size >= SMK_LABELLEN)
>>> -		return;
>>> -
>>>  	isp = dentry->d_inode->i_security;
>>>  
>>>  	/*
>>>
>>> --
>>> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
>>> the body of a message to majordomo@...r.kernel.org
>>> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>>> Please read the FAQ at  http://www.tux.org/lkml/
>>>
>>>
>>>   
>>>       
>>     
>
>
>   

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/