diff mbox series

[4/4] ima: support fs-verity file digest based signatures

Message ID 20211129170057.243127-5-zohar@linux.ibm.com (mailing list archive)
State Superseded
Headers show
Series ima: support fs-verity signatures stored as | expand

Commit Message

Mimi Zohar Nov. 29, 2021, 5 p.m. UTC
Instead of calculating a file hash and verifying the signature stored
in the security.ima xattr against the calculated file hash, verify the
signature of the fs-verity's file digest.  The fs-verity file digest is
a hash that includes the Merkle tree root hash.

Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
---
 security/integrity/ima/ima_api.c | 20 ++++++++++++++++++++
 1 file changed, 20 insertions(+)

Comments

Lakshmi Ramasubramanian Nov. 30, 2021, 5:56 a.m. UTC | #1
Hi Mimi,

On 11/29/2021 9:00 AM, Mimi Zohar wrote:
> Instead of calculating a file hash and verifying the signature stored
> in the security.ima xattr against the calculated file hash, verify the
> signature of the fs-verity's file digest.  The fs-verity file digest is
> a hash that includes the Merkle tree root hash.
This patch is reading the fs-verity signature for the given file using 
the new function fsverity_measure() that was defined in [Patch 1/4]. Is 
it also verifying the fs-verity signature here?

> 
> Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
> ---
>   security/integrity/ima/ima_api.c | 20 ++++++++++++++++++++
>   1 file changed, 20 insertions(+)
> 
> diff --git a/security/integrity/ima/ima_api.c b/security/integrity/ima/ima_api.c
> index 179c7f0364c2..ee1701f8c0f3 100644
> --- a/security/integrity/ima/ima_api.c
> +++ b/security/integrity/ima/ima_api.c
> @@ -16,6 +16,7 @@
>   #include <linux/xattr.h>
>   #include <linux/evm.h>
>   #include <linux/iversion.h>
> +#include <linux/fsverity.h>
>   
>   #include "ima.h"
>   
> @@ -205,6 +206,23 @@ int ima_get_action(struct user_namespace *mnt_userns, struct inode *inode,
>   				allowed_algos);
>   }
>   
> +static int ima_collect_verity_measurement(struct integrity_iint_cache *iint,
> +					  struct ima_digest_data *hash)
> +{
> +	u8 verity_digest[FS_VERITY_MAX_DIGEST_SIZE];
> +	enum hash_algo verity_alg;
> +	int rc;
> +
> +	rc = fsverity_measure(iint->inode, verity_digest, &verity_alg);
nit: fsverity_collect_measurement() may be more appropriate for this 
function (defined in [PATCH 1/4]).

thanks,
  -lakshmi

> +	if (rc)
> +		return -EINVAL;
> +	if (hash->algo != verity_alg)
> +		return -EINVAL;
> +	hash->length = hash_digest_size[verity_alg];
> +	memcpy(hash->digest, verity_digest, hash->length);
> +	return 0;
> +}
> +
>   /*
>    * ima_collect_measurement - collect file measurement
>    *
> @@ -256,6 +274,8 @@ int ima_collect_measurement(struct integrity_iint_cache *iint,
>   
>   	if (buf)
>   		result = ima_calc_buffer_hash(buf, size, &hash.hdr);
> +	else if (veritysig)
> +		result = ima_collect_verity_measurement(iint, &hash.hdr);
>   	else
>   		result = ima_calc_file_hash(file, &hash.hdr);
>   
>
Mimi Zohar Nov. 30, 2021, 1:36 p.m. UTC | #2
Hi Lakshmi, Eric,

On Mon, 2021-11-29 at 21:56 -0800, Lakshmi Ramasubramanian wrote:
> Hi Mimi,
> 
> On 11/29/2021 9:00 AM, Mimi Zohar wrote:
> > Instead of calculating a file hash and verifying the signature stored
> > in the security.ima xattr against the calculated file hash, verify the
> > signature of the fs-verity's file digest.  The fs-verity file digest is
> > a hash that includes the Merkle tree root hash.

> This patch is reading the fs-verity signature for the given file using 
> the new function fsverity_measure() that was defined in [Patch 1/4]. Is 
> it also verifying the fs-verity signature here?

Yes, the signature stored in the security.ima xattr may be a file hash,
a regular file signature, or a signature of the fs-verity file digest. 
The signature is verified like any other signature stored as an xattr.

>  
> > +static int ima_collect_verity_measurement(struct integrity_iint_cache *iint,
> > +					  struct ima_digest_data *hash)
> > +{
> > +	u8 verity_digest[FS_VERITY_MAX_DIGEST_SIZE];
> > +	enum hash_algo verity_alg;
> > +	int rc;
> > +
> > +	rc = fsverity_measure(iint->inode, verity_digest, &verity_alg);
> nit: fsverity_collect_measurement() may be more appropriate for this 
> function (defined in [PATCH 1/4]).

From an IMA perspective it certainly would be a better function name,
but this function may be used by other kernel subsystems.  Eric
suggested renaming the function as fsverity_get_digest(), as opposed to
fsverity_read_digest().   get/put are normally used to bump a reference
count or to get/release a lock.   Perhaps a combination like
fsverity_collect_digest() would be acceptable.

thanks,

Mimi
diff mbox series

Patch

diff --git a/security/integrity/ima/ima_api.c b/security/integrity/ima/ima_api.c
index 179c7f0364c2..ee1701f8c0f3 100644
--- a/security/integrity/ima/ima_api.c
+++ b/security/integrity/ima/ima_api.c
@@ -16,6 +16,7 @@ 
 #include <linux/xattr.h>
 #include <linux/evm.h>
 #include <linux/iversion.h>
+#include <linux/fsverity.h>
 
 #include "ima.h"
 
@@ -205,6 +206,23 @@  int ima_get_action(struct user_namespace *mnt_userns, struct inode *inode,
 				allowed_algos);
 }
 
+static int ima_collect_verity_measurement(struct integrity_iint_cache *iint,
+					  struct ima_digest_data *hash)
+{
+	u8 verity_digest[FS_VERITY_MAX_DIGEST_SIZE];
+	enum hash_algo verity_alg;
+	int rc;
+
+	rc = fsverity_measure(iint->inode, verity_digest, &verity_alg);
+	if (rc)
+		return -EINVAL;
+	if (hash->algo != verity_alg)
+		return -EINVAL;
+	hash->length = hash_digest_size[verity_alg];
+	memcpy(hash->digest, verity_digest, hash->length);
+	return 0;
+}
+
 /*
  * ima_collect_measurement - collect file measurement
  *
@@ -256,6 +274,8 @@  int ima_collect_measurement(struct integrity_iint_cache *iint,
 
 	if (buf)
 		result = ima_calc_buffer_hash(buf, size, &hash.hdr);
+	else if (veritysig)
+		result = ima_collect_verity_measurement(iint, &hash.hdr);
 	else
 		result = ima_calc_file_hash(file, &hash.hdr);