[dm-devel] [PATCH] dm-crypt: make printing of the key constant-time

[dm-devel] [PATCH] dm-crypt: make printing of the key constant-time

[Mikulas Patocka]

The device mapper dm-crypt target is using scnprintf("%02x", cc->key[i]) to
report the current key to userspace. However, this is not constant-time
operation and it may leak information about the key via timing, via cache
access patterns or via the branch predictor.

This patch changes it to use "%c" instead. We introduce a function
hex2asc. hex2asc converts a number in the range 0 ... 15 to an ascii
character and it is coded in such a way that it contains no branches and
no memory accesses.

Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Cc; stable@vger.kernel.org

---
 drivers/md/dm-crypt.c |    7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)

Index: linux-2.6/drivers/md/dm-crypt.c
===================================================================
--- linux-2.6.orig/drivers/md/dm-crypt.c	2022-04-24 19:44:14.000000000 +0200
+++ linux-2.6/drivers/md/dm-crypt.c	2022-04-24 19:54:13.000000000 +0200
@@ -3439,6 +3439,11 @@ static int crypt_map(struct dm_target *t
 	return DM_MAPIO_SUBMITTED;
 }
 
+static char hex2asc(unsigned char c)
+{
+	return c + '0' + ((9 - c) >> 4 & 0x27);
+}
+
 static void crypt_status(struct dm_target *ti, status_type_t type,
 			 unsigned status_flags, char *result, unsigned maxlen)
 {
@@ -3459,7 +3464,7 @@ static void crypt_status(struct dm_targe
 				DMEMIT(":%u:%s", cc->key_size, cc->key_string);
 			else
 				for (i = 0; i < cc->key_size; i++)
-					DMEMIT("%02x", cc->key[i]);
+					DMEMIT("%c%c", hex2asc(cc->key[i] >> 4), hex2asc(cc->key[i] & 0xf));
 		} else
 			DMEMIT("-");
 
--
dm-devel mailing list
dm-devel@redhat.com
https://listman.redhat.com/mailman/listinfo/dm-devel

[dm-devel] [PATCH v2] dm-crypt: make printing of the key constant-time

[Mikulas Patocka]

This is the second version of this patch. It casts the value to unsigned 
before doing the shift, because right shift of a negative number is not 
defined.

Mikulas



From: Mikulas Patocka <mpatocka@redhat.com>

The device mapper dm-crypt target is using scnprintf("%02x", cc->key[i]) to
report the current key to userspace. However, this is not constant-time
operation and it may leak information about the key via timing, via cache
access patterns or via the branch predictor.

This patch changes it to use "%c" instead. We introduce a function
hex2asc. hex2asc converts a number in the range 0 ... 15 to an ascii
character and it is coded in such a way that it contains no branches and
no memory accesses.

Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Cc; stable@vger.kernel.org

---
 drivers/md/dm-crypt.c |    7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)

Index: linux-2.6/drivers/md/dm-crypt.c
===================================================================
--- linux-2.6.orig/drivers/md/dm-crypt.c	2022-04-25 14:41:15.000000000 +0200
+++ linux-2.6/drivers/md/dm-crypt.c	2022-04-25 14:42:06.000000000 +0200
@@ -3439,6 +3439,11 @@ static int crypt_map(struct dm_target *t
 	return DM_MAPIO_SUBMITTED;
 }
 
+static char hex2asc(unsigned char c)
+{
+	return c + '0' + ((unsigned)(9 - c) >> 4 & 0x27);
+}
+
 static void crypt_status(struct dm_target *ti, status_type_t type,
 			 unsigned status_flags, char *result, unsigned maxlen)
 {
@@ -3459,7 +3464,7 @@ static void crypt_status(struct dm_targe
 				DMEMIT(":%u:%s", cc->key_size, cc->key_string);
 			else
 				for (i = 0; i < cc->key_size; i++)
-					DMEMIT("%02x", cc->key[i]);
+					DMEMIT("%c%c", hex2asc(cc->key[i] >> 4), hex2asc(cc->key[i] & 0xf));
 		} else
 			DMEMIT("-");
 
--
dm-devel mailing list
dm-devel@redhat.com
https://listman.redhat.com/mailman/listinfo/dm-devel

Re: [dm-devel] [PATCH v2] dm-crypt: make printing of the key constant-time

[Milan Broz]

On 25/04/2022 14:53, Mikulas Patocka wrote:
> This is the second version of this patch. It casts the value to unsigned
> before doing the shift, because right shift of a negative number is not
> defined.
> 
> Mikulas
> 
> 
> 
> From: Mikulas Patocka <mpatocka@redhat.com>
> 
> The device mapper dm-crypt target is using scnprintf("%02x", cc->key[i]) to
> report the current key to userspace. However, this is not constant-time
> operation and it may leak information about the key via timing, via cache
> access patterns or via the branch predictor.
> 
> This patch changes it to use "%c" instead. We introduce a function
> hex2asc. hex2asc converts a number in the range 0 ... 15 to an ascii
> character and it is coded in such a way that it contains no branches and
> no memory accesses.
> 
> Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
> Cc; stable@vger.kernel.org

I have run some tests with it, also cryptsetup testsuite is fine.
If it helps, you can add

Tested-by: Milan Broz <gmazyland@gmail.com>

Milan



> 
> ---
>   drivers/md/dm-crypt.c |    7 ++++++-
>   1 file changed, 6 insertions(+), 1 deletion(-)
> 
> Index: linux-2.6/drivers/md/dm-crypt.c
> ===================================================================
> --- linux-2.6.orig/drivers/md/dm-crypt.c	2022-04-25 14:41:15.000000000 +0200
> +++ linux-2.6/drivers/md/dm-crypt.c	2022-04-25 14:42:06.000000000 +0200
> @@ -3439,6 +3439,11 @@ static int crypt_map(struct dm_target *t
>   	return DM_MAPIO_SUBMITTED;
>   }
>   
> +static char hex2asc(unsigned char c)
> +{
> +	return c + '0' + ((unsigned)(9 - c) >> 4 & 0x27);
> +}
> +
>   static void crypt_status(struct dm_target *ti, status_type_t type,
>   			 unsigned status_flags, char *result, unsigned maxlen)
>   {
> @@ -3459,7 +3464,7 @@ static void crypt_status(struct dm_targe
>   				DMEMIT(":%u:%s", cc->key_size, cc->key_string);
>   			else
>   				for (i = 0; i < cc->key_size; i++)
> -					DMEMIT("%02x", cc->key[i]);
> +					DMEMIT("%c%c", hex2asc(cc->key[i] >> 4), hex2asc(cc->key[i] & 0xf));
>   		} else
>   			DMEMIT("-");
>   
> 

--
dm-devel mailing list
dm-devel@redhat.com
https://listman.redhat.com/mailman/listinfo/dm-devel

Re: [dm-devel] [PATCH v2] dm-crypt: make printing of the key constant-time

[Demi Marie Obenour]

[-- Attachment #1: Type: text/plain, Size: 321 bytes --]

Nit: did you mean CC: stable@vger.kernel.org (colon instead of
semicolon)?  Also, would it be better to avoid using the formatting
built-in to DMEMIT()?  Finally, are there any architectures for which
gcc or clang make this non-constant-time?
-- 
Sincerely,
Demi Marie Obenour (she/her/hers)
Invisible Things Lab

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

Re: [dm-devel] [PATCH v2] dm-crypt: make printing of the key constant-time

[Demi Marie Obenour]

[-- Attachment #1.1: Type: text/plain, Size: 321 bytes --]

Nit: did you mean CC: stable@vger.kernel.org (colon instead of
semicolon)?  Also, would it be better to avoid using the formatting
built-in to DMEMIT()?  Finally, are there any architectures for which
gcc or clang make this non-constant-time?
-- 
Sincerely,
Demi Marie Obenour (she/her/hers)
Invisible Things Lab

[-- Attachment #1.2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

[-- Attachment #2: Type: text/plain, Size: 98 bytes --]

--
dm-devel mailing list
dm-devel@redhat.com
https://listman.redhat.com/mailman/listinfo/dm-devel