When comparing MAC hashes, AEAD authentication tags, or other hash
values in the context of authentication or integrity checking, it
is important not to leak timing information to a potential attacker,
i.e. when communication happens over a network.

Bytewise memory comparisons (such as memcmp) are usually optimized so
that they return a nonzero value as soon as a mismatch is found. E.g,
on x86_64/i5 for 512 bytes this can be ~50 cyc for a full mismatch
and up to ~850 cyc for a full match (cold). This early-return behavior
can leak timing information as a side channel, allowing an attacker to
iteratively guess the correct result.

This patch adds a new method crypto_memneq ("memory not equal to each
other") to the crypto API that compares memory areas of the same length
in roughly "constant time" (cache misses could change the timing, but
since they don't reveal information about the content of the strings
being compared, they are effectively benign). Iow, best and worst case
behaviour take the same amount of time to complete (in contrast to
memcmp).

Note that crypto_memneq (unlike memcmp) can only be used to test for
equality or inequality, NOT for lexicographical order. This, however,
is not an issue for its use-cases within the crypto API.

We tried to locate all of the places in the crypto API where memcmp was
being used for authentication or integrity checking, and convert them
over to crypto_memneq.

crypto_memneq is declared noinline, placed in its own source file,
and compiled with optimizations that might increase code size disabled
("Os") because a smart compiler (or LTO) might notice that the return
value is always compared against zero/nonzero, and might then
reintroduce the same early-return optimization that we are trying to
avoid.

Using #pragma or __attribute__ optimization annotations of the code
for disabling optimization was avoided as it seems to be considered
broken or unmaintained for long time in GCC [1]. Therefore, we work
around that by specifying the compile flag for memneq.o directly in
the Makefile. We found that this seems to be most appropriate.

As we use ("Os"), this patch also provides a loop-free "fast-path" for
frequently used 16 byte digests. Similarly to kernel library string
functions, leave an option for future even further optimized architecture
specific assembler implementations.

This was a joint work of James Yonan and Daniel Borkmann. Also thanks
for feedback from Florian Weimer on this and earlier proposals [2].

  [1] http://gcc.gnu.org/ml/gcc/2012-07/msg00211.html
  [2] https://lkml.org/lkml/2013/2/10/131

Signed-off-by: James Yonan <james@openvpn.net>
Signed-off-by: Daniel Borkmann <dborkman@redhat.com>
Cc: Florian Weimer <fw@deneb.enyo.de>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

Pull crypto update from Herbert Xu:

 - XTS mode optimisation for twofish/cast6/camellia/aes on x86

 - AVX2/x86_64 implementation for blowfish/twofish/serpent/camellia

 - SSSE3/AVX/AVX2 optimisations for sha256/sha512

 - Added driver for SAHARA2 crypto accelerator

 - Fix for GMAC when used in non-IPsec secnarios

 - Added generic CMAC implementation (including IPsec glue)

 - IP update for crypto/atmel

 - Support for more than one device in hwrng/timeriomem

 - Added Broadcom BCM2835 RNG driver

 - Misc fixes

* git://git.kernel.org/pub/scm/linux/kernel/git/herbert/crypto-2.6: (59 commits)
  crypto: caam - fix job ring cleanup code
  crypto: camellia - add AVX2/AES-NI/x86_64 assembler implementation of camellia cipher
  crypto: serpent - add AVX2/x86_64 assembler implementation of serpent cipher
  crypto: twofish - add AVX2/x86_64 assembler implementation of twofish cipher
  crypto: blowfish - add AVX2/x86_64 implementation of blowfish cipher
  crypto: tcrypt - add async cipher speed tests for blowfish
  crypto: testmgr - extend camellia test-vectors for camellia-aesni/avx2
  crypto: aesni_intel - fix Kconfig problem with CRYPTO_GLUE_HELPER_X86
  crypto: aesni_intel - add more optimized XTS mode for x86-64
  crypto: x86/camellia-aesni-avx - add more optimized XTS code
  crypto: cast6-avx: use new optimized XTS code
  crypto: x86/twofish-avx - use optimized XTS code
  crypto: x86 - add more optimized XTS-mode for serpent-avx
  xfrm: add rfc4494 AES-CMAC-96 support
  crypto: add CMAC support to CryptoAPI
  crypto: testmgr - add empty test vectors for null ciphers
  crypto: testmgr - add AES GMAC test vectors
  crypto: gcm - fix rfc4543 to handle async crypto correctly
  crypto: gcm - make GMAC work when dst and src are different
  hwrng: timeriomem - added devicetree hooks
  ...

If the gcm cipher used by rfc4543 does not complete request immediately,
the authentication tag is not copied to destination buffer. Patch adds
correct async logic for this case.

Signed-off-by: Jussi Kivilinna <jussi.kivilinna@iki.fi>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

The GMAC code assumes that dst==src, which causes problems when trying to add
rfc4543(gcm(aes)) test vectors.

So fix this code to work when source and destination buffer are different.

Signed-off-by: Jussi Kivilinna <jussi.kivilinna@iki.fi>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

rfc4543(gcm(*)) code for GMAC assumes that assoc scatterlist always contains
only one segment and only makes use of this first segment. However ipsec passes
assoc with three segments when using 'extended sequence number' thus in this
case rfc4543(gcm(*)) fails to function correctly. Patch fixes this issue.

Reported-by: Chaoxing Lin <Chaoxing.Lin@ultra-3eti.com>
Tested-by: Chaoxing Lin <Chaoxing.Lin@ultra-3eti.com>
Cc: stable@vger.kernel.org
Signed-off-by: Jussi Kivilinna <jussi.kivilinna@iki.fi>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

Replace PTR_ERR followed by ERR_PTR by ERR_CAST, to be more concise.

The semantic patch that makes this change is as follows:
(http://coccinelle.lip6.fr/)

// <smpl>
@@
expression err,x;
@@
-       err = PTR_ERR(x);
        if (IS_ERR(x))
-                return ERR_PTR(err);
+                return ERR_CAST(x);
// </smpl>

Signed-off-by: Julia Lawall <Julia.Lawall@lip6.fr>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

Use scatterwalk_crypto_chain in favor of locally defined chaining functions.

Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

This patch adds the RFC4543 (GMAC) wrapper for GCM similar to the
existing RFC4106 wrapper. The main differences between GCM and GMAC are
the contents of the AAD and that the plaintext is empty for the latter.

Signed-off-by: Tobias Brunner <tobias@strongswan.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

The flow of the complete function (xxx_done) in gcm.c is as follow:

void complete(struct crypto_async_request *areq, int err)
{
	struct aead_request *req = areq->data;

	if (!err) {
		err = async_next_step();
		if (err == -EINPROGRESS || err == -EBUSY)
			return;
	}

	complete_for_next_step(areq, err);
}

But *areq may be destroyed in async_next_step(), this makes
complete_for_next_step() can not work properly. To fix this, one of
following methods is used for each complete function.

- Add a __complete() for each complete(), which accept struct
  aead_request *req instead of areq, so avoid using areq after it is
  destroyed.

- Expand complete_for_next_step().

The fixing method is based on the idea of Herbert Xu.

Signed-off-by: Huang Ying <ying.huang@intel.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

Remove the dedicated GHASH implementation in GCM, and uses the GHASH
digest algorithm instead. This will make GCM uses hardware accelerated
GHASH implementation automatically if available.

ahash instead of shash interface is used, because some hardware
accelerated GHASH implementation needs asynchronous interface.

Signed-off-by: Huang Ying <ying.huang@intel.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>