Currently for skcipher IV generators they must provide givencrypt
as that is the whole point.  We are currently replacing skcipher
IV generators with explicit IV generators.  In order to maintain
backwards compatibility, we need to allow the IV generators to
still function as a normal skcipher when the RNG Is not present
(e.g., in the initramfs during boot).  IOW everything but givencrypt
and givdecrypt will still work but those two will fail.

Therefore this patch assigns a default givencrypt that simply
returns an error should it be NULL.
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

Newer templates use tmpl->create and have a NULL tmpl->alloc.  So
we must use tmpl->create if it is set.
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

When the user explicitly states that they don't care whether the
algorithm has been tested (type = CRYPTO_ALG_TESTED and mask = 0),
there is a corner case where we may erroneously return ENOENT.

This patch fixes it by correcting the logic in the test.
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

Modify crypto drivers to use the generic SG helper since
both of them are equivalent and the one from crypto is redundant.

See also:
  468577ab   reverted in
  b2ab4a57Signed-off-by: NCristian Stoica <cristian.stoica@freescale.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

Fixed style errors reported by checkpatch.

WARNING: Missing a blank line after declarations
+       u8 *end_page = (u8 *)(((unsigned long)(start + len - 1)) & PAGE_MASK);
+       return max(start, end_page);

WARNING: line over 80 characters
+               scatterwalk_start(&walk->out, scatterwalk_sg_next(walk->out.sg));

WARNING: Missing a blank line after declarations
+               int err = ablkcipher_copy_iv(walk, tfm, alignmask);
+               if (err)

ERROR: do not use assignment in if condition
+       if ((err = crypto_register_instance(tmpl, inst))) {
Signed-off-by: NJoshua I. James <joshua@cybercrimetech.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

Previously we would use eseqiv on all async ciphers in all cases,
and sync ciphers if we have more than one CPU.  This meant that
chainiv is only used in the case of sync ciphers on a UP machine.

As chainiv may aid attackers by making the IV predictable, even
though this risk itself is small, the above usage pattern causes
it to further leak information about the host.

This patch addresses these issues by using eseqiv even if we're
on a UP machine.
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>
Acked-by: NSteffen Klassert <steffen.klassert@secunet.com>
Acked-by: NDavid S. Miller <davem@davemloft.net>

Three errors resulting in kernel memory disclosure:

1/ The structures used for the netlink based crypto algorithm report API
are located on the stack. As snprintf() does not fill the remainder of
the buffer with null bytes, those stack bytes will be disclosed to users
of the API. Switch to strncpy() to fix this.

2/ crypto_report_one() does not initialize all field of struct
crypto_user_alg. Fix this to fix the heap info leak.

3/ For the module name we should copy only as many bytes as
module_name() returns -- not as much as the destination buffer could
hold. But the current code does not and therefore copies random data
from behind the end of the module name, as the module name is always
shorter than CRYPTO_MAX_ALG_NAME.

Also switch to use strncpy() to copy the algorithm's name and
driver_name. They are strings, after all.
Signed-off-by: NMathias Krause <minipli@googlemail.com>
Cc: Steffen Klassert <steffen.klassert@secunet.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

These macros contain a hidden goto, and are thus extremely error
prone and make code hard to audit.
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

We lookup algorithms with crypto_alg_mod_lookup() when instantiating via
crypto_add_alg(). However, algorithms that are wrapped by an IV genearator
(e.g. aead or genicv type algorithms) need special care. The userspace
process hangs until it gets a timeout when we use crypto_alg_mod_lookup()
to lookup these algorithms. So export the lookup functions for these
algorithms and use them in crypto_add_alg().
Signed-off-by: NSteffen Klassert <steffen.klassert@secunet.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

The report functions use NLA_PUT so we need to ensure that NET
is enabled.
Reported-by: NLuis Henriques <henrix@camandro.org>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

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

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

Signed-off-by: NDavidlohr Bueso <dave@gnu.org>
Acked-by: NDavid S. Miller <davem@davemloft.net>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

Stanse found a potential NULL dereference in ablkcipher_next_slow.
Even though kmalloc fails, its retval is dereferenced later. Return
from that function properly earlier.
Signed-off-by: NJiri Slaby <jslaby@suse.cz>
Acked-by: NDavid S. Miller <davem@davemloft.net>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

These are akin to the blkcipher_walk helpers.

The main differences in the async variant are:

1) Only physical walking is supported.  We can't hold on to
   kmap mappings across the async operation to support virtual
   ablkcipher_walk operations anyways.

2) Bounce buffers used for async more need to be persistent and
   freed at a later point in time when the async op completes.
   Therefore we maintain a list of writeback buffers and require
   that the ablkcipher_walk user call the 'complete' operation
   so we can copy the bounce buffers out to the real buffers and
   free up the bounce buffer chunks.

These interfaces will be used by the new Niagara2 crypto driver.
Signed-off-by: NDavid S. Miller <davem@davemloft.net>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

Signed-off-by: NRichard Hartmann <richih.mailinglist@gmail.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

Recently we switched to using eseqiv on SMP machines in preference
over chainiv.  However, eseqiv does not support stream ciphers so
they should still default to chainiv.

This patch applies the same check as done by eseqiv to weed out
the stream ciphers.  In particular, all algorithms where the IV
size is not equal to the block size will now default to chainiv.
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

As it stands we use chainiv for sync algorithms and eseqiv for
async algorithms.  However, when there is more than one CPU
chainiv forces all processing to be serialised which is usually
not what you want.  Also, the added overhead of eseqiv isn't that
great.

Therefore this patch changes the default sync geniv on SMP machines
to eseqiv.  For the odd situation where the overhead is unacceptable
then chainiv is still available as an option.

Note that on UP machines chainiv is still preferred over eseqiv
for sync algorithms.
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

When a sync givcipher algorithm is requested, if an async version
of the same algorithm already exists, then we will loop forever
without ever constructing the sync version based on a blkcipher.

This is because we did not include the requested type/mask when
getting a larval for the geniv algorithm that is to be constructed.
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

When an skcipher constructed through crypto_givcipher_default fails
its selftest, we'll loop forever trying to construct new skcipher
objects but failing because it already exists.

The crux of the issue is that once a givcipher fails the selftest,
we'll ignore it on the next run through crypto_skcipher_lookup and
attempt to construct a new givcipher.

We should instead return an error to the caller if we find a
givcipher that has failed the test.
Signed-off-by: NHerbert 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: NHerbert Xu <herbert@gondor.apana.org.au>

This patch adds an async field to /proc/crypto for ablkcipher and aead
algorithms.
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

This patch makes crypto_alloc_ablkcipher/crypto_grab_skcipher always
return algorithms that are capable of generating their own IVs through
givencrypt and givdecrypt.  Each algorithm may specify its default IV
generator through the geniv field.

For algorithms that do not set the geniv field, the blkcipher layer will
pick a default.  Currently it's chainiv for synchronous algorithms and
eseqiv for asynchronous algorithms.  Note that if these wrappers do not
work on an algorithm then that algorithm must specify its own geniv or
it can't be used at all.
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

This patch creates the infrastructure to help the construction of givcipher
templates that wrap around existing blkcipher/ablkcipher algorithms by adding
an IV generator to them.
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

This patch introduces the geniv field which indicates the default IV
generator for each algorithm.  It should point to a string that is not
freed as long as the algorithm is registered.
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

Different block cipher modes have different requirements for intialisation
vectors.  For example, CBC can use a simple randomly generated IV while
modes such as CTR must use an IV generation mechanisms that give a stronger
guarantee on the lack of collisions.  Furthermore, disk encryption modes
have their own IV generation algorithms.

Up until now IV generation has been left to the users of the symmetric
key cipher API.  This is inconvenient as the number of block cipher modes
increase because the user needs to be aware of which mode is supposed to
be paired with which IV generation algorithm.

Therefore it makes sense to integrate the IV generation into the crypto
API.  This patch takes the first step in that direction by creating two
new ablkcipher operations, givencrypt and givdecrypt that generates an
IV before performing the actual encryption or decryption.

The operations are currently not exposed to the user.  That will be done
once the underlying functionality has actually been implemented.

It also creates the underlying givcipher type.  Algorithms that directly
generate IVs would use it instead of ablkcipher.  All other algorithms
(including all existing ones) would generate a givcipher algorithm upon
registration.  This givcipher algorithm will be constructed from the geniv
string that's stored in every algorithm.  That string will locate a template
which is instantiated by the blkcipher/ablkcipher algorithm in question to
give a givcipher algorithm.
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

Note: From now on the collective of ablkcipher/blkcipher/givcipher will
be known as skcipher, i.e., symmetric key cipher.  The name blkcipher has
always been much of a misnomer since it supports stream ciphers too.

This patch adds the function crypto_grab_skcipher as a new way of getting
an ablkcipher spawn.  The problem is that previously we did this in two
steps, first getting the algorithm and then calling crypto_init_spawn.

This meant that each spawn user had to be aware of what type and mask to
use for these two steps.  This is difficult and also presents a problem
when the type/mask changes as they're about to be for IV generators.

The new interface does both steps together just like crypto_alloc_ablkcipher.

As a side-effect this also allows us to be stronger on type enforcement
for spawns.  For now this is only done for ablkcipher but it's trivial
to extend for other types.

This patch also moves the type/mask logic for skcipher into the helpers
crypto_skcipher_type and crypto_skcipher_mask.

Finally this patch introduces the function crypto_require_sync to determine
whether the user is specifically requesting a sync algorithm.
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

Since not everyone needs a queue pointer and those who need it can
always get it from the context anyway the queue pointer in the
common alg object is redundant.
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

This patch ensures that kernel.h and slab.h are included for
the setkey_unaligned function.  It also breaks a couple of
long lines.
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

setkey_unaligned() commited in ca7c3938
overwrites unallocated memory in the following memset() because
I used the wrong buffer length.
Signed-off-by: NSebastian Siewior <sebastian@breakpoint.cc>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

Evgeniy's hifn driver and probably mine don't use ablkcipher->queue at all. 
The show method of ablkcipher will access this field without checking if it 
is valid.
Signed-off-by: NSebastian Siewior <linux-crypto@ml.breakpoint.cc>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

setkey() in {cipher,blkcipher,ablkcipher,hash}.c does not respect the
requested alignment by the algorithm. This patch fixes it. The extra
memory is allocated by kmalloc() with GFP_ATOMIC flag.
Signed-off-by: NSebastian Siewior <linux-crypto@ml.breakpoint.cc>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

This patch adds the mid-level interface for asynchronous block ciphers.
It also includes a generic queueing mechanism that can be used by other
asynchronous crypto operations in future.
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>