The soft dependency on cryptomgr is only needed in algapi because
if algapi isn't present then no algorithms can be loaded.  This
also fixes the case where api is built-in but algapi is built as
a module as the soft dependency would otherwise get lost.

Fixes: 8ab23d54 ("crypto: api - Add softdep on cryptomgr")
Reported-by: NJan Beulich <jbeulich@suse.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>
Tested-by: NJan Beulich <jbeulich@suse.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

When the crypto manager is disabled, we need to explicitly set
the crypto algorithms' tested status so that they can be used.

Fixes: cad439fc ("crypto: api - Do not create test larvals if...")
Reported-by: NGeert Uytterhoeven <geert@linux-m68k.org>
Reported-by: NIdo Schimmel <idosch@idosch.org>
Reported-by: NGuenter Roeck <linux@roeck-us.net>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>
Tested-by: NIdo Schimmel <idosch@nvidia.com>
Tested-by: NGeert Uytterhoeven <geert@linux-m68k.org>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

The delayed boot-time testing patch created a dependency loop
between api.c and algapi.c because it added a crypto_alg_tested
call to the former when the crypto manager is disabled.

We could instead avoid creating the test larvals if the crypto
manager is disabled.  This avoids the dependency loop as well
as saving some unnecessary work, albeit in a very unlikely case.
Reported-by: NNathan Chancellor <nathan@kernel.org>
Reported-by: NNaresh Kamboju <naresh.kamboju@linaro.org>
Reported-by: Nkernel test robot <lkp@intel.com>
Fixes: adad556e ("crypto: api - Fix built-in testing dependency failures")
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

When complex algorithms that depend on other algorithms are built
into the kernel, the order of registration must be done such that
the underlying algorithms are ready before the ones on top are
registered.  As otherwise they would fail during the self-test
which is required during registration.

In the past we have used subsystem initialisation ordering to
guarantee this.  The number of such precedence levels are limited
and they may cause ripple effects in other subsystems.

This patch solves this problem by delaying all self-tests during
boot-up for built-in algorithms.  They will be tested either when
something else in the kernel requests for them, or when we have
finished registering all built-in algorithms, whichever comes
earlier.
Reported-by: NVladis Dronov <vdronov@redhat.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

According to the advice of Eric and Herbert, type CRYPTOA_U32
has been unused for over a decade, so remove the code related to
CRYPTOA_U32.

After removing CRYPTOA_U32, the type of the variable attrs can be
changed from union to struct.
Signed-off-by: NLiu Shixin <liushixin2@huawei.com>
Reviewed-by: NEric Biggers <ebiggers@google.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

The flag CRYPTO_ALG_ASYNC is "inherited" in the sense that when a
template is instantiated, the template will have CRYPTO_ALG_ASYNC set if
any of the algorithms it uses has CRYPTO_ALG_ASYNC set.

We'd like to add a second flag (CRYPTO_ALG_ALLOCATES_MEMORY) that gets
"inherited" in the same way.  This is difficult because the handling of
CRYPTO_ALG_ASYNC is hardcoded everywhere.  Address this by:

  - Add CRYPTO_ALG_INHERITED_FLAGS, which contains the set of flags that
    have these inheritance semantics.

  - Add crypto_algt_inherited_mask(), for use by template ->create()
    methods.  It returns any of these flags that the user asked to be
    unset and thus must be passed in the 'mask' to crypto_grab_*().

  - Also modify crypto_check_attr_type() to handle computing the 'mask'
    so that most templates can just use this.

  - Make crypto_grab_*() propagate these flags to the template instance
    being created so that templates don't have to do this themselves.

Make crypto/simd.c propagate these flags too, since it "wraps" another
algorithm, similar to a template.

Based on a patch by Mikulas Patocka <mpatocka@redhat.com>
(https://lore.kernel.org/r/alpine.LRH.2.02.2006301414580.30526@file01.intranet.prod.int.rdu2.redhat.com).
Signed-off-by: NEric Biggers <ebiggers@google.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

Add crypto_enqueue_request_head function that enqueues a
request in front of queue.
This will be used in crypto-engine, on error path. In case a request
was not executed by hardware, enqueue it back in front of queue (to
keep the order of requests).
Signed-off-by: NIuliana Prodan <iuliana.prodan@nxp.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

There are two problems in crypto_spawn_alg.  First of all it may
return spawn->alg even if spawn->dead is set.  This results in a
double-free as detected by syzbot.

Secondly the setting of the DYING flag is racy because we hold
the read-lock instead of the write-lock.  We should instead call
crypto_shoot_alg in a safe manner by gaining a refcount, dropping
the lock, and then releasing the refcount.

This patch fixes both problems.

Reported-by: syzbot+fc0674cde00b66844470@syzkaller.appspotmail.com
Fixes: 4f87ee11 ("crypto: api - Do not zap spawn->alg")
Fixes: 73669cc5 ("crypto: api - Fix race condition in...")
Cc: <stable@vger.kernel.org>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

Currently after any algorithm is registered and tested, there's an
unnecessary request_module("cryptomgr") even if it's already loaded.
Also, CRYPTO_MSG_ALG_LOADED is sent twice, and thus if the algorithm is
"crct10dif", lib/crc-t10dif.c replaces the tfm twice rather than once.

This occurs because CRYPTO_MSG_ALG_LOADED is sent using
crypto_probing_notify(), which tries to load "cryptomgr" if the
notification is not handled (NOTIFY_DONE).  This doesn't make sense
because "cryptomgr" doesn't handle this notification.

Fix this by using crypto_notify() instead of crypto_probing_notify().

Fixes: dd8b083f ("crypto: api - Introduce notifier for new crypto algorithms")
Cc: <stable@vger.kernel.org> # v4.20+
Cc: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: NEric Biggers <ebiggers@google.com>
Reviewed-by: NMartin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

Now that all templates provide a ->create() method which creates an
instance, installs a strongly-typed ->free() method directly to it, and
registers it, the older ->alloc() and ->free() methods in
'struct crypto_template' are no longer used.  Remove them.
Signed-off-by: NEric Biggers <ebiggers@google.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

Now that crypto_init_spawn() is only called by crypto_grab_spawn(),
simplify things by moving its functionality into crypto_grab_spawn().

In the process of doing this, also be more consistent about when the
spawn and instance are updated, and remove the crypto_spawn::dropref
flag since now it's always set.
Signed-off-by: NEric Biggers <ebiggers@google.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

Remove lots of helper functions that were previously used for
instantiating crypto templates, but are now unused:

- crypto_get_attr_alg() and similar functions looked up an inner
  algorithm directly from a template parameter.  These were replaced
  with getting the algorithm's name, then calling crypto_grab_*().

- crypto_init_spawn2() and similar functions initialized a spawn, given
  an algorithm.  Similarly, these were replaced with crypto_grab_*().

- crypto_alloc_instance() and similar functions allocated an instance
  with a single spawn, given the inner algorithm.  These aren't useful
  anymore since crypto_grab_*() need the instance allocated first.
Signed-off-by: NEric Biggers <ebiggers@google.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

Currently, crypto_spawn::inst is first used temporarily to pass the
instance to crypto_grab_spawn().  Then crypto_init_spawn() overwrites it
with crypto_spawn::next, which shares the same union.  Finally,
crypto_spawn::inst is set again when the instance is registered.

Make this less convoluted by just passing the instance as an argument to
crypto_grab_spawn() instead.
Signed-off-by: NEric Biggers <ebiggers@google.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

To allow further simplifying template ->create() functions, make
crypto_grab_spawn() handle an ERR_PTR() name by passing back the error.

For most templates, this will allow the result of crypto_attr_alg_name()
to be passed directly to crypto_grab_*(), rather than first having to
assign it to a variable [where it can then potentially be misused, as it
was in the rfc7539 template prior to commit 5e27f38f ("crypto:
chacha20poly1305 - set cra_name correctly")] and check it for error.
Signed-off-by: NEric Biggers <ebiggers@google.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

Make crypto_drop_spawn() do nothing when the spawn hasn't been
initialized with an algorithm yet.  This will allow simplifying error
handling in all the template ->create() functions, since on error they
will be able to just call their usual "free instance" function, rather
than having to handle dropping just the spawns that have been
initialized so far.

This does assume the spawn starts out zero-filled, but that's always the
case since instances are allocated with kzalloc().  And some other code
already assumes this anyway.
Signed-off-by: NEric Biggers <ebiggers@google.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

This patch changes crypto_grab_spawn to retain the reference count
on the algorithm.  This is because the caller needs to access the
algorithm parameters and without the reference count the algorithm
can be freed at any time.

The reference count will be subsequently dropped by the crypto API
once the instance has been registered.  The helper crypto_drop_spawn
will also conditionally drop the reference count depending on whether
it has been registered.

Note that the code is actually added to crypto_init_spawn.  However,
unless the caller activates this by setting spawn->dropref beforehand
then nothing happens.  The only caller that sets dropref is currently
crypto_grab_spawn.

Once all legacy users of crypto_init_spawn disappear, then we can
kill the dropref flag.

Internally each instance will maintain a list of its spawns prior
to registration.  This memory used by this list is shared with
other fields that are only used after registration.  In order for
this to work a new flag spawn->registered is added to indicate
whether spawn->inst can be used.

Fixes: d6ef2f19 ("crypto: api - Add crypto_grab_spawn primitive")
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

Some of the algorithm unregistration functions return -ENOENT when asked
to unregister a non-registered algorithm, while others always return 0
or always return void.  But no users check the return value, except for
two of the bulk unregistration functions which print a message on error
but still always return 0 to their caller, and crypto_del_alg() which
calls crypto_unregister_instance() which always returns 0.

Since unregistering a non-registered algorithm is always a kernel bug
but there isn't anything callers should do to handle this situation at
runtime, let's simplify things by making all the unregistration
functions return void, and moving the error message into
crypto_unregister_alg() and upgrading it to a WARN().
Signed-off-by: NEric Biggers <ebiggers@google.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

When CONFIG_CRYPTO_MANAGER_EXTRA_TESTS=y, the first lookup of an
algorithm that needs to be instantiated using a template will always get
the generic implementation, even when an accelerated one is available.

This happens because the extra self-tests for the accelerated
implementation allocate the generic implementation for comparison
purposes, and then crypto_alg_tested() for the generic implementation
"fulfills" the original request (i.e. sets crypto_larval::adult).

This patch fixes this by only fulfilling the original request if
we are currently the best outstanding larval as judged by the
priority.  If we're not the best then we will ask all waiters on
that larval request to retry the lookup.

Note that this patch introduces a behaviour change when the module
providing the new algorithm is unregistered during the process.
Previously we would have failed with ENOENT, after the patch we
will instead redo the lookup.

Fixes: 9a8a6b3f ("crypto: testmgr - fuzz hashes against...")
Fixes: d435e10e ("crypto: testmgr - fuzz skciphers against...")
Fixes: 40153b10 ("crypto: testmgr - fuzz AEADs against...")
Reported-by: NEric Biggers <ebiggers@google.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>
Reviewed-by: NEric Biggers <ebiggers@google.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

This patch explains the logic behind crypto_remove_spawns and its
underling crypto_more_spawns.
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

Currently when a spawn is removed we will zap its alg field.
This is racy because the spawn could belong to an unregistered
instance which may dereference the spawn->alg field.

This patch fixes this by keeping spawn->alg constant and instead
adding a new spawn->dead field to indicate that a spawn is going
away.
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

The function crypto_spawn_alg is racy because it drops the lock
before shooting the dying algorithm.  The algorithm could disappear
altogether before we shoot it.

This patch fixes it by moving the shooting into the locked section.

Fixes: 6bfd4809 ("[CRYPTO] api: Added spawns")
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

We need to check whether spawn->alg is NULL under lock as otherwise
the algorithm could be removed from under us after we have checked
it and found it to be non-NULL.  This could cause us to remove the
spawn from a non-existent list.

Fixes: 7ede5a5b ("crypto: api - Fix crypto_drop_spawn crash...")
Cc: <stable@vger.kernel.org>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

Now that all users of the deprecated ablkcipher interface have been
moved to the skcipher interface, ablkcipher is no longer used and
can be removed.
Reviewed-by: NEric Biggers <ebiggers@kernel.org>
Signed-off-by: NArd Biesheuvel <ardb@kernel.org>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

Now that all algorithms explicitly set cra_driver_name, make it required
for algorithm registration and remove the code that generated a default
cra_driver_name.

Also add an explicit check that cra_name is set too, since that's
obviously required too, yet it didn't seem to be checked anywhere.
Signed-off-by: NEric Biggers <ebiggers@google.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

Based on 1 normalized pattern(s):

  this program is free software you can redistribute it and or modify
  it under the terms of the gnu general public license as published by
  the free software foundation either version 2 of the license or at
  your option any later version

extracted by the scancode license scanner the SPDX license identifier

  GPL-2.0-or-later

has been chosen to replace the boilerplate/reference in 3029 file(s).
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Reviewed-by: NAllison Randal <allison@lohutok.net>
Cc: linux-spdx@vger.kernel.org
Link: https://lkml.kernel.org/r/20190527070032.746973796@linutronix.deSigned-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Remove the crypto_tfm_in_queue() function, which is unused.
Signed-off-by: NEric Biggers <ebiggers@google.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

This patch add a helper to (un)register a array of templates. The
following patches will use this helper to simplify the code.
Signed-off-by: NXiongfeng Wang <xiongfeng.wang@linaro.org>
Reviewed-by: NEric Biggers <ebiggers@google.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

It took me a while to notice the bug where the adiantum template left
crypto_spawn::inst == NULL, because this only caused problems in certain
cases where algorithms are dynamically loaded/unloaded.

More improvements are needed, but for now make crypto_init_spawn()
reject this case and WARN(), so this type of bug will be noticed
immediately in the future.

Note: I checked all callers and the adiantum template was the only place
that had this wrong.  So this WARN shouldn't trigger anymore.
Signed-off-by: NEric Biggers <ebiggers@google.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

Now that all "blkcipher" templates have been converted to "skcipher",
crypto_alloc_instance() is no longer used.  And it's not useful any
longer as it creates an old-style weakly typed instance rather than a
new-style strongly typed instance.  So remove it, and now that the name
is freed up rename crypto_alloc_instance2() to crypto_alloc_instance().
Signed-off-by: NEric Biggers <ebiggers@google.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

This patch add the crypto_stats_init() function.
This will permit to remove some ifdef from __crypto_register_alg().
Signed-off-by: NCorentin Labbe <clabbe@baylibre.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

Since now all crypto stats are on their own structures, it is now
useless to have the algorithm name in the err_cnt member.
Signed-off-by: NCorentin Labbe <clabbe@baylibre.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

Like for userspace, this patch splits stats into multiple structures,
one for each algorithm class.
Signed-off-by: NCorentin Labbe <clabbe@baylibre.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

All crypto_stats functions use the struct xxx_request for feeding stats,
but in some case this structure could already be freed.

For fixing this, the needed parameters (len and alg) will be stored
before the request being executed.
Fixes: cac5818c ("crypto: user - Implement a generic crypto statistics")
Reported-by: Nsyzbot <syzbot+6939a606a5305e9e9799@syzkaller.appspotmail.com>
Signed-off-by: NCorentin Labbe <clabbe@baylibre.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

All the 32-bit fields need to be 64-bit.  In some cases, UINT32_MAX crypto
operations can be done in seconds.
Reported-by: NEric Biggers <ebiggers@kernel.org>
Signed-off-by: NCorentin Labbe <clabbe@baylibre.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

Even if CRYPTO_STATS is set to n, some part of CRYPTO_STATS are
compiled.
This patch made all part of crypto_user_stat uncompiled in that case.
Signed-off-by: NCorentin Labbe <clabbe@baylibre.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

This patch implement a generic way to get statistics about all crypto
usages.
Signed-off-by: NCorentin Labbe <clabbe@baylibre.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

Introduce a facility that can be used to receive a notification
callback when a new algorithm becomes available. This can be used by
existing crypto registrations to trigger a switch from a software-only
algorithm to a hardware-accelerated version.

A new CRYPTO_MSG_ALG_LOADED state is introduced to the existing crypto
notification chain, and the register/unregister functions are exported
so they can be called by subsystems outside of crypto.
Signed-off-by: NMartin K. Petersen <martin.petersen@oracle.com>
Suggested-by: NHerbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: NMartin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

In the quest to remove all stack VLA usage from the kernel[1], this
exposes a new general upper bound on crypto blocksize and alignmask
(higher than for the existing cipher limits) for VLA removal,
and introduces new checks.

At present, the highest cra_alignmask in the kernel is 63. The highest
cra_blocksize is 144 (SHA3_224_BLOCK_SIZE, 18 8-byte words). For the
new blocksize limit, I went with 160 (20 8-byte words).

[1] https://lkml.kernel.org/r/CA+55aFzCG-zNmZwX4A2FQpadafLfEzK6CC=qPXydAacU1RqZWA@mail.gmail.comSigned-off-by: NKees Cook <keescook@chromium.org>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

In preparation for the removal of VLAs[1] from crypto code.
We create 2 new compile-time constants: all ciphers implemented
in Linux have a block size less than or equal to 16 bytes and
the most demanding hw require 16 bytes alignment for the block
buffer.
We also enforce these limits in crypto_check_alg when a new
cipher is registered.

[1] http://lkml.kernel.org/r/CA+55aFzCG-zNmZwX4A2FQpadafLfEzK6CC=qPXydAacU1RqZWA@mail.gmail.comSigned-off-by: NSalvatore Mesoraca <s.mesoraca16@gmail.com>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>

This patch reverts commit 9c521a20 ("crypto: api - remove
instance when test failed") and fixes the underlying problem
in a different way.

To recap, prior to the reverted commit, an instance that fails
a self-test is kept around.  However, it would satisfy any new
lookups against its name and therefore the system may accumlulate
an unbounded number of failed instances for the same algorithm
name.

The reverted commit fixed it by unregistering the instance.  Hoever,
this still does not prevent the creation of the same failed instance
over and over again each time the name is looked up.

This patch fixes it by keeping the failed instance around, just as
we would if it were a normal algorithm.  However, the lookup code
has been udpated so that we do not attempt to create another
instance as long as this failed one is still registered.  Of course,
you could still force a new creation by deleting the instance from
user-space.

A new error (ELIBBAD) has been commandeered for this purpose and
will be returned when all registered algorithm of a given name
have failed the self-test.
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>