[PATCH v11 2/4] smack: Set the SMACK64TRANSMUTE xattr in smack_inode_init_security()

[Roberto Sassu]

On 6/9/2023 9:35 PM, Mimi Zohar wrote:
> Hi Roberto,
> 
> On Sat, 2023-06-03 at 21:15 +0200, Roberto Sassu wrote:
>> From: Roberto Sassu <roberto.sassu at huawei.com>
>>
>> With the newly added ability of LSMs to supply multiple xattrs, set
>> SMACK64TRASMUTE in smack_inode_init_security(), instead of d_instantiate().
>> Do it by incrementing SMACK_INODE_INIT_XATTRS to 2 and by calling
>> lsm_get_xattr_slot() a second time, if the transmuting conditions are met.
>>
>> The LSM infrastructure passes all xattrs provided by LSMs to the
>> filesystems through the initxattrs() callback, so that filesystems can
>> store xattrs in the disk.
>>
>> After the change, the SMK_INODE_TRANSMUTE inode flag is always set by
>> d_instantiate() after fetching SMACK64TRANSMUTE from the disk. Before it
>> was done by smack_inode_post_setxattr() as result of the __vfs_setxattr()
>> call.
>>
>> Removing __vfs_setxattr() also prevents invalidating the EVM HMAC, by
>> adding a new xattr without checking and updating the existing HMAC.
>>
>> Signed-off-by: Roberto Sassu <roberto.sassu at huawei.com>
> 
> Just a few comments/nits inline.
> 
>> ---
>>   security/smack/smack.h     |  2 +-
>>   security/smack/smack_lsm.c | 43 +++++++++++++++++++++++---------------
>>   2 files changed, 27 insertions(+), 18 deletions(-)
>>
>> diff --git a/security/smack/smack.h b/security/smack/smack.h
>> index aa15ff56ed6..041688e5a77 100644
>> --- a/security/smack/smack.h
>> +++ b/security/smack/smack.h
>> @@ -128,7 +128,7 @@ struct task_smack {
>>   
>>   #define	SMK_INODE_INSTANT	0x01	/* inode is instantiated */
>>   #define	SMK_INODE_TRANSMUTE	0x02	/* directory is transmuting */
>> -#define	SMK_INODE_CHANGED	0x04	/* smack was transmuted */
>> +#define	SMK_INODE_CHANGED	0x04	/* smack was transmuted (unused) */
>>   #define	SMK_INODE_IMPURE	0x08	/* involved in an impure transaction */
>>   
>>   /*
>> diff --git a/security/smack/smack_lsm.c b/security/smack/smack_lsm.c
>> index a1c30275692..b67d901ee74 100644
>> --- a/security/smack/smack_lsm.c
>> +++ b/security/smack/smack_lsm.c
>> @@ -52,7 +52,14 @@
>>   #define SMK_RECEIVING	1
>>   #define SMK_SENDING	2
>>   
>> -#define SMACK_INODE_INIT_XATTRS 1
>> +/*
>> + * Smack uses multiple xattrs.
>> + * SMACK64 - for access control,
>> + * SMACK64TRANSMUTE - label initialization,
>> + * Not saved on files - SMACK64IPIN and SMACK64IPOUT,
>> + * Must be set explicitly - SMACK64EXEC and SMACK64MMAP
>> + */
>> +#define SMACK_INODE_INIT_XATTRS 2
>>   
>>   #ifdef SMACK_IPV6_PORT_LABELING
>>   static DEFINE_MUTEX(smack_ipv6_lock);
>> @@ -935,7 +942,6 @@ static int smack_inode_init_security(struct inode *inode, struct inode *dir,
>>   				     struct xattr *xattrs, int *xattr_count)
>>   {
>>   	struct task_smack *tsp = smack_cred(current_cred());
>> -	struct inode_smack *issp = smack_inode(inode);
>>   	struct smack_known *skp = smk_of_task(tsp);
>>   	struct smack_known *isp = smk_of_inode(inode);
>>   	struct smack_known *dsp = smk_of_inode(dir);
>> @@ -963,6 +969,8 @@ static int smack_inode_init_security(struct inode *inode, struct inode *dir,
>>   		if ((tsp->smk_task == tsp->smk_transmuted) ||
>>   		    (may > 0 && ((may & MAY_TRANSMUTE) != 0) &&
>>   		     smk_inode_transmutable(dir))) {
>> +			struct xattr *xattr_transmute;
>> +
> 
> Variables should be defined at the beginning of the function.

Casey asked to declare the variable in this block.

> Is there a reason for beginning the function with "if (xattr) {"
> instead "if (!xattr) return 0;".  This causes unnecessary indenting.

I revisited this part and made few fixes:

https://lore.kernel.org/linux-security-module/20230607123612.2791303-1-roberto.sassu@huaweicloud.com/

Patch 3 should address your concern.

>>   			/*
>>   			 * The caller of smack_dentry_create_files_as()
>>   			 * should have overridden the current cred, so the
>> @@ -971,7 +979,16 @@ static int smack_inode_init_security(struct inode *inode, struct inode *dir,
>>   			 */
>>   			if (tsp->smk_task != tsp->smk_transmuted)
>>   				isp = dsp;
>> -			issp->smk_flags |= SMK_INODE_CHANGED;
>> +			xattr_transmute = lsm_get_xattr_slot(xattrs, xattr_count);
>> +			if (xattr_transmute) {
>> +				xattr_transmute->value = kmemdup(TRANS_TRUE,
>> +						TRANS_TRUE_SIZE, GFP_NOFS);
> 
> script/checkpatch --strict complains here.

Thanks, I didn't know about it.

It seems that they are more stylistic things. Probably, not worth to 
respin the patch set just for those (unless you prefer I do it).

Thanks

Roberto

>> +				if (xattr_transmute->value == NULL)
>> +					return -ENOMEM;
>> +
>> +				xattr_transmute->value_len = TRANS_TRUE_SIZE;
>> +				xattr_transmute->name = XATTR_SMACK_TRANSMUTE;
>> +			}
>>   		}
>>   
>>   		xattr->value = kstrdup(isp->smk_known, GFP_NOFS);
>> @@ -3518,20 +3535,12 @@ static void smack_d_instantiate(struct dentry *opt_dentry, struct inode *inode)
>>   			 * If there is a transmute attribute on the
>>   			 * directory mark the inode.
>>   			 */
>> -			if (isp->smk_flags & SMK_INODE_CHANGED) {
>> -				isp->smk_flags &= ~SMK_INODE_CHANGED;
>> -				rc = __vfs_setxattr(&nop_mnt_idmap, dp, inode,
>> -					XATTR_NAME_SMACKTRANSMUTE,
>> -					TRANS_TRUE, TRANS_TRUE_SIZE,
>> -					0);
>> -			} else {
>> -				rc = __vfs_getxattr(dp, inode,
>> -					XATTR_NAME_SMACKTRANSMUTE, trattr,
>> -					TRANS_TRUE_SIZE);
>> -				if (rc >= 0 && strncmp(trattr, TRANS_TRUE,
>> -						       TRANS_TRUE_SIZE) != 0)
>> -					rc = -EINVAL;
>> -			}
>> +			rc = __vfs_getxattr(dp, inode,
>> +					    XATTR_NAME_SMACKTRANSMUTE, trattr,
>> +					    TRANS_TRUE_SIZE);
>> +			if (rc >= 0 && strncmp(trattr, TRANS_TRUE,
>> +					       TRANS_TRUE_SIZE) != 0)
>> +				rc = -EINVAL;
>>   			if (rc >= 0)
>>   				transflag = SMK_INODE_TRANSMUTE;
>>   		}
> 
>