MD5 is not FIPS compliant. But still md5 was used as the
default algorithm for sctp if fips was enabled.
Due to this, listen() system call in ltp tests was
failing for sctp in fips environment, with below error message.
[ 6397.892677] sctp: failed to load transform for md5: -2
Fix is to not assign md5 as default algorithm for sctp
if fips_enabled is true. Instead make sha1 as default algorithm.
The issue fixes ltp testcase failure "cve-2018-5803 sctp_big_chunk"
Signed-off-by: Ashwin Dayanand Kamat <[email protected]>
---
v3:
* Resolved hunk failures.
* Changed the ratelimited notice to be more meaningful.
* Used ternary condition for if/else condtion.
v2:
* The listener can still fail if fips mode is enabled after
that the netns is initialized.
* Fixed this in sctp_listen_start() as suggested by
Paolo Abeni <[email protected]>
---
net/sctp/socket.c | 10 ++++++++++
1 file changed, 10 insertions(+)
diff --git a/net/sctp/socket.c b/net/sctp/socket.c
index cda8c2874691..d7cde9cc706e 100644
--- a/net/sctp/socket.c
+++ b/net/sctp/socket.c
@@ -49,6 +49,7 @@
#include <linux/poll.h>
#include <linux/init.h>
#include <linux/slab.h>
+#include <linux/fips.h>
#include <linux/file.h>
#include <linux/compat.h>
#include <linux/rhashtable.h>
@@ -8501,6 +8502,15 @@ static int sctp_listen_start(struct sock *sk, int backlog)
struct crypto_shash *tfm = NULL;
char alg[32];
+ if (fips_enabled && !strcmp(sp->sctp_hmac_alg, "md5")) {
+ sp->sctp_hmac_alg = IS_ENABLED(CONFIG_SCTP_DEFAULT_COOKIE_HMAC_SHA1) ?
+ "sha1" : NULL;
+
+ net_info_ratelimited("changing the hmac algorithm to %s "
+ "as md5 is not supported when fips is enabled",
+ sp->sctp_hmac_alg);
+ }
+
/* Allocate HMAC for generating cookie. */
if (!sp->hmac && sp->sctp_hmac_alg) {
sprintf(alg, "hmac(%s)", sp->sctp_hmac_alg);
--
2.39.0
+ Xin Long
On Thu, Jun 01, 2023 at 11:47:54PM +0530, Ashwin Dayanand Kamat wrote:
> MD5 is not FIPS compliant. But still md5 was used as the
> default algorithm for sctp if fips was enabled.
> Due to this, listen() system call in ltp tests was
> failing for sctp in fips environment, with below error message.
>
> [ 6397.892677] sctp: failed to load transform for md5: -2
>
> Fix is to not assign md5 as default algorithm for sctp
> if fips_enabled is true. Instead make sha1 as default algorithm.
> The issue fixes ltp testcase failure "cve-2018-5803 sctp_big_chunk"
>
> Signed-off-by: Ashwin Dayanand Kamat <[email protected]>
> ---
> v3:
> * Resolved hunk failures.
> * Changed the ratelimited notice to be more meaningful.
> * Used ternary condition for if/else condtion.
> v2:
> * The listener can still fail if fips mode is enabled after
> that the netns is initialized.
> * Fixed this in sctp_listen_start() as suggested by
> Paolo Abeni <[email protected]>
FWIIW, this seems reasonable to me.
Reviewed-by: Simon Horman <[email protected]>
On Fri, Jun 2, 2023 at 11:15 AM Simon Horman <[email protected]> wrote:
>
> + Xin Long
>
> On Thu, Jun 01, 2023 at 11:47:54PM +0530, Ashwin Dayanand Kamat wrote:
> > MD5 is not FIPS compliant. But still md5 was used as the
> > default algorithm for sctp if fips was enabled.
> > Due to this, listen() system call in ltp tests was
> > failing for sctp in fips environment, with below error message.
> >
> > [ 6397.892677] sctp: failed to load transform for md5: -2
> >
> > Fix is to not assign md5 as default algorithm for sctp
> > if fips_enabled is true. Instead make sha1 as default algorithm.
> > The issue fixes ltp testcase failure "cve-2018-5803 sctp_big_chunk"
Hi, Ashwin,
I have the same question as Paolo about "this patch gets fips compliance
_disabling_ the encryption", is it from any standard?
If not, can't you fix the ltp testcase for fips environment by sysctl?
or set 'CONFIG_SCTP_DEFAULT_COOKIE_HMAC_SHA1=y' instead in config.
Sorry if I don't understand this well. You're trying to avoid SCTP code
calling crypto_alloc_shash(MD5), right? What about other places
where it may also do it in kernel? (where ltp just doesn't cover)
I don't think it makes sense to let SCTP have some code reply on
FIPS only to make ltp testcase happy, while we can actually fix it
in ltp by "sysctl".
Thanks.
> >
> > Signed-off-by: Ashwin Dayanand Kamat <[email protected]>
> > ---
> > v3:
> > * Resolved hunk failures.
> > * Changed the ratelimited notice to be more meaningful.
> > * Used ternary condition for if/else condtion.
> > v2:
> > * The listener can still fail if fips mode is enabled after
> > that the netns is initialized.
> > * Fixed this in sctp_listen_start() as suggested by
> > Paolo Abeni <[email protected]>
>
> FWIIW, this seems reasonable to me.
>
> Reviewed-by: Simon Horman <[email protected]>
> On 03-Jun-2023, at 4:46 AM, Xin Long <[email protected]> wrote:
>
> !! External Email
>
> On Fri, Jun 2, 2023 at 11:15 AM Simon Horman <[email protected]> wrote:
>>
>> + Xin Long
>>
>> On Thu, Jun 01, 2023 at 11:47:54PM +0530, Ashwin Dayanand Kamat wrote:
>>> MD5 is not FIPS compliant. But still md5 was used as the
>>> default algorithm for sctp if fips was enabled.
>>> Due to this, listen() system call in ltp tests was
>>> failing for sctp in fips environment, with below error message.
>>>
>>> [ 6397.892677] sctp: failed to load transform for md5: -2
>>>
>>> Fix is to not assign md5 as default algorithm for sctp
>>> if fips_enabled is true. Instead make sha1 as default algorithm.
>>> The issue fixes ltp testcase failure "cve-2018-5803 sctp_big_chunk"
> Hi, Ashwin,
>
> I have the same question as Paolo about "this patch gets fips compliance
> _disabling_ the encryption", is it from any standard?
>
> If not, can't you fix the ltp testcase for fips environment by sysctl?
> or set 'CONFIG_SCTP_DEFAULT_COOKIE_HMAC_SHA1=y' instead in config.
>
> Sorry if I don't understand this well. You're trying to avoid SCTP code
> calling crypto_alloc_shash(MD5), right? What about other places
> where it may also do it in kernel? (where ltp just doesn't cover)
>
> I don't think it makes sense to let SCTP have some code reply on
> FIPS only to make ltp testcase happy, while we can actually fix it
> in ltp by "sysctl".
>
> Thanks.
>
Hi Xi,
Thanks for your inputs. I have sent the patches to ltp upstream to handle the fix from ltp codebase.
Regards,
Ashwin Kamat
>>>
>>> Signed-off-by: Ashwin Dayanand Kamat <[email protected]>
>>> ---
>>> v3:
>>> * Resolved hunk failures.
>>> * Changed the ratelimited notice to be more meaningful.
>>> * Used ternary condition for if/else condtion.
>>> v2:
>>> * The listener can still fail if fips mode is enabled after
>>> that the netns is initialized.
>>> * Fixed this in sctp_listen_start() as suggested by
>>> Paolo Abeni <[email protected]>
>>
>> FWIIW, this seems reasonable to me.
>>
>> Reviewed-by: Simon Horman <[email protected]>
>
> !! External Email: This email originated from outside of the organization. Do not click links or open attachments unless you recognize the sender.