Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Message-ID: <6209669358d038b30e5fe20ba571f93241b5248d.camel@redhat.com>
Subject: Re: [PATCH] sctp: Make sha1 as default algorithm if fips is enabled
From:   Paolo Abeni <pabeni@redhat.com>
To:     Ashwin Dayanand Kamat <kashwindayan@vmware.com>,
        Vlad Yasevich <vyasevich@gmail.com>,
        Neil Horman <nhorman@tuxdriver.com>,
        Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>,
        "David S . Miller" <davem@davemloft.net>,
        Eric Dumazet <edumazet@google.com>,
        Jakub Kicinski <kuba@kernel.org>, linux-sctp@vger.kernel.org,
        netdev@vger.kernel.org, linux-kernel@vger.kernel.org
Cc:     srivatsab@vmware.com, srivatsa@csail.mit.edu, amakhalov@vmware.com,
        vsirnapalli@vmware.com, akaher@vmware.com
Date:   Thu, 22 Dec 2022 10:22:24 +0100
In-Reply-To: <1671513037-8958-1-git-send-email-kashwindayan@vmware.com>
References: <1671513037-8958-1-git-send-email-kashwindayan@vmware.com>
Content-Type: text/plain; charset="UTF-8"
User-Agent: Evolution 3.42.4 (3.42.4-2.fc35) 
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Precedence: bulk

On Tue, 2022-12-20 at 10:40 +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.
> 
> Signed-off-by: Ashwin Dayanand Kamat <kashwindayan@vmware.com>

I don't know the fips standard in details, but it feel strange that you
get fips compliance _disabling_ the encryption. Can you please point
which part of the standard states it?

Since this is fix, you should also provide a suitable fixes tag. When
you will repost additionally include the target tree name (net) into
the subject, thanks!

> ---
>  net/sctp/protocol.c | 16 ++++++++--------
>  1 file changed, 8 insertions(+), 8 deletions(-)
> 
> diff --git a/net/sctp/protocol.c b/net/sctp/protocol.c
> index 909a89a..b6e9810 100644
> --- a/net/sctp/protocol.c
> +++ b/net/sctp/protocol.c
> @@ -34,6 +34,7 @@
>  #include <linux/memblock.h>
>  #include <linux/highmem.h>
>  #include <linux/slab.h>
> +#include <linux/fips.h>
>  #include <net/net_namespace.h>
>  #include <net/protocol.h>
>  #include <net/ip.h>
> @@ -1321,14 +1322,13 @@ static int __net_init sctp_defaults_init(struct net *net)
>  	/* Whether Cookie Preservative is enabled(1) or not(0) */
>  	net->sctp.cookie_preserve_enable 	= 1;
>  
> -	/* Default sctp sockets to use md5 as their hmac alg */
> -#if defined (CONFIG_SCTP_DEFAULT_COOKIE_HMAC_MD5)
> -	net->sctp.sctp_hmac_alg			= "md5";
> -#elif defined (CONFIG_SCTP_DEFAULT_COOKIE_HMAC_SHA1)
> -	net->sctp.sctp_hmac_alg			= "sha1";
> -#else
> -	net->sctp.sctp_hmac_alg			= NULL;
> -#endif
> +	/* Default sctp sockets to use md5 as default only if fips is not enabled */
> +	if (!fips_enabled && IS_ENABLED(CONFIG_SCTP_DEFAULT_COOKIE_HMAC_MD5))
> +		net->sctp.sctp_hmac_alg = "md5";
> +	else if (IS_ENABLED(CONFIG_SCTP_DEFAULT_COOKIE_HMAC_SHA1))
> +		net->sctp.sctp_hmac_alg = "sha1";
> +	else
> +		net->sctp.sctp_hmac_alg = NULL;

It looks like the listener can still fail if fips mode is enabled after
that the netns is initialized. I think it would be better to take
action in sctp_listen_start() and buming a ratelimited notice the
selected hmac is changed due to fips.

Thanks,

Paolo

>  
>  	/* Max.Burst		    - 4 */
>  	net->sctp.max_burst			= SCTP_DEFAULT_MAX_BURST;