linux-kernel - Re: [RFC PATCH] crypto: prevent helper ciphers from being allocated by users

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:	Tue, 17 Mar 2015 12:40:12 +0100
From:	Stephan Mueller <smueller@...onox.de>
To:	Herbert Xu <herbert@...dor.apana.org.au>
Cc:	linux-crypto@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [RFC PATCH] crypto: prevent helper ciphers from being allocated by users

Am Dienstag, 17. März 2015, 22:23:50 schrieb Herbert Xu:

Hi Herbert,

>On Fri, Mar 13, 2015 at 10:09:21PM +0100, Stephan Mueller wrote:
>> +struct crypto_tfm *__crypto_alloc_tfm_safe(struct crypto_alg *alg,
>> u32 type, +					   u32 mask)
>> +{
>> +	/*
>> +	 * Prevent all ciphers from being loaded which have a 
cra_priority
>> +	 * of 0. Those cipher implementations are helper ciphers and
>> +	 * are not intended for general consumption.
>> +	 *
>> +	 * The only exceptions are the compression algorithms which
>> +	 * have no priority.
>> +	 */
>> +	if (!alg->cra_priority &&
>> +	    ((alg->cra_flags & CRYPTO_ALG_TYPE_MASK) !=
>> +	      CRYPTO_ALG_TYPE_PCOMPRESS) &&
>> +	    ((alg->cra_flags & CRYPTO_ALG_TYPE_MASK) !=
>> +	      CRYPTO_ALG_TYPE_COMPRESS))
>> +		return ERR_PTR(-ENOENT);
>
>How about adding a flag to all these internal algorithms and then
>change crypto_alg_mod_lookup to disable that flag by default?

The issue with flags is the following: first we have to think about 
whether we want a black list or white list approach. Your suggestion 
implies a black list. Black lists for ensuring security is not good IMHO 
as it has a tendency to miss cases. This especially applies to this area 
where we have already an indicator for internal ciphers: the prio is so 
low that it will never ever be selected based on the name. Now, adding a 
flag means that we mark such an internal cipher twice.

Therefore, I would not opt for such a flag (at least for a black list) 
and stay with the prio approach.

Ciao
Stephan
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/