When I backed out of using the generic sg chaining (as it isn't currently
portable) and introduced scatterwalk_sg_chain/scatterwalk_sg_next I left
out the sg_is_last check in the latter.  This causes it to potentially
dereference beyond the end of the sg array.

As most uses of scatterwalk_sg_next are bound by an overall length, this
only affected the chaining code in authenc and eseqiv. Thanks to Patrick
McHardy for identifying this problem.

This patch also clears the "last" bit on the head of the chained list as
it's no longer last.  This also went missing in scatterwalk_sg_chain and
is present in sg_chain.

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

The key expansion routine could be get little more generic, become
a kernel doc entry and then get exported.

Signed-off-by: Sebastian Siewior <sebastian@breakpoint.cc>
Tested-by: Stefan Hellermann <stefan@the2masters.de>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

The previous patch to move chainiv and eseqiv into blkcipher created
a section mismatch for the chainiv exit function which was also called
from __init.  This patch removes the __exit marking on it.

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

For compatibility with dm-crypt initramfs setups it is useful to merge
chainiv/seqiv into the crypto_blkcipher module.  Since they're required
by most algorithms anyway this is an acceptable trade-off.

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

As Andrew Morton correctly points out, we need to explicitly include
sched.h as we use the function cond_resched in crypto/scatterwalk.h.

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

This patch makes chainiv avoid spinning by postponing requests on lock
contention if the user allows the use of asynchronous algorithms.  If
a synchronous algorithm is requested then we behave as before.

This should improve IPsec performance on SMP when two CPUs attempt to
transmit over the same SA.  Currently one of them will spin doing nothing
waiting for the other CPU to finish its encryption.  This patch makes it
postpone the request and get on with other work.

If only one CPU is transmitting for a given SA, then we will process
the request synchronously as before.

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

This patch adds a null blkcipher algorithm called ecb(cipher_null) for
backwards compatibility.  Previously the null algorithm when used by
IPsec copied the data byte by byte.  This new algorithm optimises that
to a straight memcpy which lets us better measure inherent overheads in
our IPsec code.

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

This patch makes crypto_alloc_aead always return algorithms that is
capable of generating their own IVs through givencrypt and givdecrypt.
All existing AEAD algorithms already do.  New ones must either supply
their own or specify a generic IV generator with the geniv field.

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

This patch creates the infrastructure to help the construction of IV
generator templates that wrap around AEAD algorithms by adding an IV
generator to them.  This is useful for AEAD algorithms with no built-in
IV generator or to replace their built-in generator.

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

This patch finally makes the givencrypt/givdecrypt operations available
to users by adding crypto_aead_givencrypt and crypto_aead_givdecrypt.
A suite of helpers to allocate and fill in the request is also available.

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