The SHA512 code we adopted from the OpenSSL project uses a rather
peculiar way to take the address of the round constant table: it
takes the address of the sha256_block_data_order() routine, and
substracts a constant known quantity to arrive at the base of the
table, which is emitted by the same assembler code right before
the routine's entry point.
However, recent versions of binutils have helpfully changed the
behavior of references emitted via an ADR instruction when running
in Thumb2 mode: it now takes the Thumb execution mode bit into
account, which is bit 0 af the address. This means the produced
table address also has bit 0 set, and so we end up with an address
value pointing 1 byte past the start of the table, which results
in crashes such as
Unable to handle kernel paging request at virtual address bf825000
pgd = 42f44b11
[bf825000] *pgd=80000040206003, *pmd=5f1bd003, *pte=00000000
Internal error: Oops: 207 [#1] PREEMPT SMP THUMB2
Modules linked in: sha256_arm(+) sha1_arm_ce sha1_arm ...
CPU: 7 PID: 396 Comm: cryptomgr_test Not tainted 5.0.0-rc6+ #144
Hardware name: QEMU KVM Virtual Machine, BIOS 0.0.0 02/06/2015
PC is at sha256_block_data_order+0xaaa/0xb30 [sha256_arm]
LR is at __this_module+0x17fd/0xffffe800 [sha256_arm]
pc : [<bf820bca>] lr : [<bf824ffd>] psr: 800b0033
sp : ebc8bbe8 ip : faaabe1c fp : 2fdd3433
r10: 4c5f1692 r9 : e43037df r8 : b04b0a5a
r7 : c369d722 r6 : 39c3693e r5 : 7a013189 r4 : 1580d26b
r3 : 8762a9b0 r2 : eea9c2cd r1 : 3e9ab536 r0 : 1dea4ae7
Flags: Nzcv IRQs on FIQs on Mode SVC_32 ISA Thumb Segment user
Control: 70c5383d Table: 6b8467c0 DAC: dbadc0de
Process cryptomgr_test (pid: 396, stack limit = 0x69e1fe23)
Stack: (0xebc8bbe8 to 0xebc8c000)
...
unwind: Unknown symbol address bf820bca
unwind: Index not found bf820bca
Code: 441a ea80 40f9 440a (f85e) 3b04
---[ end trace e560cce92700ef8a ]---
Given that this affects older kernels as well, in case they are built
with a recent toolchain, apply a minimal backportable fix, which is
to emit another non-code label at the start of the routine, and
reference that instead. (This is similar to the current upstream state
of this file in OpenSSL)
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
The SHA256 code we adopted from the OpenSSL project uses a rather
peculiar way to take the address of the round constant table: it
takes the address of the sha256_block_data_order() routine, and
substracts a constant known quantity to arrive at the base of the
table, which is emitted by the same assembler code right before
the routine's entry point.
However, recent versions of binutils have helpfully changed the
behavior of references emitted via an ADR instruction when running
in Thumb2 mode: it now takes the Thumb execution mode bit into
account, which is bit 0 af the address. This means the produced
table address also has bit 0 set, and so we end up with an address
value pointing 1 byte past the start of the table, which results
in crashes such as
Unable to handle kernel paging request at virtual address bf825000
pgd = 42f44b11
[bf825000] *pgd=80000040206003, *pmd=5f1bd003, *pte=00000000
Internal error: Oops: 207 [#1] PREEMPT SMP THUMB2
Modules linked in: sha256_arm(+) sha1_arm_ce sha1_arm ...
CPU: 7 PID: 396 Comm: cryptomgr_test Not tainted 5.0.0-rc6+ #144
Hardware name: QEMU KVM Virtual Machine, BIOS 0.0.0 02/06/2015
PC is at sha256_block_data_order+0xaaa/0xb30 [sha256_arm]
LR is at __this_module+0x17fd/0xffffe800 [sha256_arm]
pc : [<bf820bca>] lr : [<bf824ffd>] psr: 800b0033
sp : ebc8bbe8 ip : faaabe1c fp : 2fdd3433
r10: 4c5f1692 r9 : e43037df r8 : b04b0a5a
r7 : c369d722 r6 : 39c3693e r5 : 7a013189 r4 : 1580d26b
r3 : 8762a9b0 r2 : eea9c2cd r1 : 3e9ab536 r0 : 1dea4ae7
Flags: Nzcv IRQs on FIQs on Mode SVC_32 ISA Thumb Segment user
Control: 70c5383d Table: 6b8467c0 DAC: dbadc0de
Process cryptomgr_test (pid: 396, stack limit = 0x69e1fe23)
Stack: (0xebc8bbe8 to 0xebc8c000)
...
unwind: Unknown symbol address bf820bca
unwind: Index not found bf820bca
Code: 441a ea80 40f9 440a (f85e) 3b04
---[ end trace e560cce92700ef8a ]---
Given that this affects older kernels as well, in case they are built
with a recent toolchain, apply a minimal backportable fix, which is
to emit another non-code label at the start of the routine, and
reference that instead. (This is similar to the current upstream state
of this file in OpenSSL)
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
The x86, arm, and arm64 asm implementations of crct10dif are very
difficult to understand partly because many of the comments, labels, and
macros are named incorrectly: the lengths mentioned are usually off by a
factor of two from the actual code. Many other things are unnecessarily
convoluted as well, e.g. there are many more fold constants than
actually needed and some aren't fully reduced.
This series therefore cleans up all these implementations to be much
more maintainable. I also made some small optimizations where I saw
opportunities, resulting in slightly better performance.
This patch cleans up the arm version.
(Also moved the constants to .rodata as suggested by Ard Biesheuvel.)
Acked-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Remove some code that is no longer called now that we make sure never
to invoke the SIMD routine with less that 16 bytes of input.
Reviewed-by: Eric Biggers <ebiggers@kernel.org>
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
The SIMD routine ported from x86 used to have a special code path
for inputs < 16 bytes, which got lost somewhere along the way.
Instead, the current glue code aligns the input pointer to permit
the NEON routine to use special versions of the vld1 instructions
that assume 16 byte alignment, but this could result in inputs of
less than 16 bytes to be passed in. This not only fails the new
extended tests that Eric has implemented, it also results in the
code reading past the end of the input, which could potentially
result in crashes when dealing with less than 16 bytes of input
at the end of a page which is followed by an unmapped page.
So update the glue code to only invoke the NEON routine if the
input is at least 16 bytes.
Reported-by: Eric Biggers <ebiggers@kernel.org>
Reviewed-by: Eric Biggers <ebiggers@kernel.org>
Fixes: 1d481f1cd8 ("crypto: arm/crct10dif - port x86 SSE implementation to ARM")
Cc: <stable@vger.kernel.org> # v4.10+
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
In commit 54a702f705 ("kbuild: mark $(targets) as .SECONDARY and
remove .PRECIOUS markers"), I missed one important feature of the
.SECONDARY target:
.SECONDARY with no prerequisites causes all targets to be
treated as secondary.
... which agrees with the policy of Kbuild.
Let's move it to scripts/Kbuild.include, with no prerequisites.
Note:
If an intermediate file is generated by $(call if_changed,...), you
still need to add it to "targets" so its .*.cmd file is included.
The arm/arm64 crypto files are generated by $(call cmd,shipped),
so they do not need to be added to "targets", but need to be added
to "clean-files" so "make clean" can properly clean them away.
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Add an ARM NEON implementation of NHPoly1305, an ε-almost-∆-universal
hash function used in the Adiantum encryption mode. For now, only the
NH portion is actually NEON-accelerated; the Poly1305 part is less
performance-critical so is just implemented in C.
Signed-off-by: Eric Biggers <ebiggers@google.com>
Reviewed-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Now that the 32-bit ARM NEON implementation of ChaCha20 and XChaCha20
has been refactored to support varying the number of rounds, add support
for XChaCha12. This is identical to XChaCha20 except for the number of
rounds, which is 12 instead of 20.
XChaCha12 is faster than XChaCha20 but has a lower security margin,
though still greater than AES-256's since the best known attacks make it
through only 7 rounds. See the patch "crypto: chacha - add XChaCha12
support" for more details about why we need XChaCha12 support.
Reviewed-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
In preparation for adding XChaCha12 support, rename/refactor the NEON
implementation of ChaCha20 to support different numbers of rounds.
Reviewed-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Add an XChaCha20 implementation that is hooked up to the ARM NEON
implementation of ChaCha20. This is needed for use in the Adiantum
encryption mode; see the generic code patch,
"crypto: chacha20-generic - add XChaCha20 support", for more details.
We also update the NEON code to support HChaCha20 on one block, so we
can use that in XChaCha20 rather than calling the generic HChaCha20.
This required factoring the permutation out into its own macro.
Reviewed-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
To improve responsivesess, disable preemption for each step of the walk
(which is at most PAGE_SIZE) rather than for the entire
encryption/decryption operation.
Suggested-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
In preparation for adding XChaCha12 support, rename/refactor
chacha20-generic to support different numbers of rounds. The
justification for needing XChaCha12 support is explained in more detail
in the patch "crypto: chacha - add XChaCha12 support".
The only difference between ChaCha{8,12,20} are the number of rounds
itself; all other parts of the algorithm are the same. Therefore,
remove the "20" from all definitions, structures, functions, files, etc.
that will be shared by all ChaCha versions.
Also make ->setkey() store the round count in the chacha_ctx (previously
chacha20_ctx). The generic code then passes the round count through to
chacha_block(). There will be a ->setkey() function for each explicitly
allowed round count; the encrypt/decrypt functions will be the same. I
decided not to do it the opposite way (same ->setkey() function for all
round counts, with different encrypt/decrypt functions) because that
would have required more boilerplate code in architecture-specific
implementations of ChaCha and XChaCha.
Reviewed-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Acked-by: Martin Willi <martin@strongswan.org>
Signed-off-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Remove asm/hwcap.h which is included more than once
Signed-off-by: Brajeswar Ghosh <brajeswar.linux@gmail.com>
Acked-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Make the ARM scalar AES implementation closer to constant-time by
disabling interrupts and prefetching the tables into L1 cache. This is
feasible because due to ARM's "free" rotations, the main tables are only
1024 bytes instead of the usual 4096 used by most AES implementations.
On ARM Cortex-A7, the speed loss is only about 5%. The resulting code
is still over twice as fast as aes_ti.c. Responsiveness is potentially
a concern, but interrupts are only disabled for a single AES block.
Note that even after these changes, the implementation still isn't
necessarily guaranteed to be constant-time; see
https://cr.yp.to/antiforgery/cachetiming-20050414.pdf for a discussion
of the many difficulties involved in writing truly constant-time AES
software. But it's valuable to make such attacks more difficult.
Much of this patch is based on patches suggested by Ard Biesheuvel.
Suggested-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Eric Biggers <ebiggers@google.com>
Reviewed-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
The table id (second) argument to MODULE_DEVICE_TABLE is often
referenced otherwise. This is not the case for CPU features. This
leads to a warning when building the kernel with Clang:
arch/arm/crypto/crc32-ce-glue.c:239:33: warning: variable
'crc32_cpu_feature' is not needed and will not be emitted
[-Wunneeded-internal-declaration]
static const struct cpu_feature crc32_cpu_feature[] = {
^
Avoid warnings by using __maybe_unused, similar to commit 1f318a8baf
("modules: mark __inittest/__exittest as __maybe_unused").
Fixes: 2a9faf8b7e ("crypto: arm/crc32 - enable module autoloading based on CPU feature bits")
Signed-off-by: Stefan Agner <stefan@agner.ch>
Acked-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Optimize ChaCha20 NEON performance by:
- Implementing the 8-bit rotations using the 'vtbl.8' instruction.
- Streamlining the part that adds the original state and XORs the data.
- Making some other small tweaks.
On ARM Cortex-A7, these optimizations improve ChaCha20 performance from
about 12.08 cycles per byte to about 11.37 -- a 5.9% improvement.
There is a tradeoff involved with the 'vtbl.8' rotation method since
there is at least one CPU (Cortex-A53) where it's not fastest. But it
seems to be a better default; see the added comment. Overall, this
patch reduces Cortex-A53 performance by less than 0.5%.
Signed-off-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Speed up the GHASH algorithm based on 64-bit polynomial multiplication
by adding support for 4-way aggregation. This improves throughput by
~85% on Cortex-A53, from 1.7 cycles per byte to 0.9 cycles per byte.
When combined with AES into GCM, throughput improves by ~25%, from
3.8 cycles per byte to 3.0 cycles per byte.
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
These are unused, undesired, and have never actually been used by
anybody. The original authors of this code have changed their mind about
its inclusion. While originally proposed for disk encryption on low-end
devices, the idea was discarded [1] in favor of something else before
that could really get going. Therefore, this patch removes Speck.
[1] https://marc.info/?l=linux-crypto-vger&m=153359499015659
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Acked-by: Eric Biggers <ebiggers@google.com>
Cc: stable@vger.kernel.org
Acked-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Almost all files in the kernel are either plain text or UTF-8 encoded. A
couple however are ISO_8859-1, usually just a few characters in a C
comments, for historic reasons.
This converts them all to UTF-8 for consistency.
Link: http://lkml.kernel.org/r/20180724111600.4158975-1-arnd@arndb.de
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Acked-by: Simon Horman <horms@verge.net.au> [IPVS portion]
Acked-by: Jonathan Cameron <Jonathan.Cameron@huawei.com> [IIO]
Acked-by: Michael Ellerman <mpe@ellerman.id.au> [powerpc]
Acked-by: Rob Herring <robh@kernel.org>
Cc: Joe Perches <joe@perches.com>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Samuel Ortiz <sameo@linux.intel.com>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Rob Herring <robh+dt@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The 4-way ChaCha20 NEON code implements 16-bit rotates with vrev32.16,
but the one-way code (used on remainder blocks) implements it with
vshl + vsri, which is slower. Switch the one-way code to vrev32.16 too.
Signed-off-by: Eric Biggers <ebiggers@google.com>
Acked-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Some ahash algorithms set .cra_type = &crypto_ahash_type. But this is
redundant with the C structure type ('struct ahash_alg'), and
crypto_register_ahash() already sets the .cra_type automatically.
Apparently the useless assignment has just been copy+pasted around.
So, remove the useless assignment from all the ahash algorithms.
This patch shouldn't change any actual behavior.
Signed-off-by: Eric Biggers <ebiggers@google.com>
Acked-by: Gilad Ben-Yossef <gilad@benyossef.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Many ahash algorithms set .cra_flags = CRYPTO_ALG_TYPE_AHASH. But this
is redundant with the C structure type ('struct ahash_alg'), and
crypto_register_ahash() already sets the type flag automatically,
clearing any type flag that was already there. Apparently the useless
assignment has just been copy+pasted around.
So, remove the useless assignment from all the ahash algorithms.
This patch shouldn't change any actual behavior.
Signed-off-by: Eric Biggers <ebiggers@google.com>
Acked-by: Gilad Ben-Yossef <gilad@benyossef.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Many shash algorithms set .cra_flags = CRYPTO_ALG_TYPE_SHASH. But this
is redundant with the C structure type ('struct shash_alg'), and
crypto_register_shash() already sets the type flag automatically,
clearing any type flag that was already there. Apparently the useless
assignment has just been copy+pasted around.
So, remove the useless assignment from all the shash algorithms.
This patch shouldn't change any actual behavior.
Signed-off-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Building the kernel with CONFIG_THUMB2_KERNEL=y and
CONFIG_CRYPTO_SPECK_NEON set fails with the following errors:
arch/arm/crypto/speck-neon-core.S: Assembler messages:
arch/arm/crypto/speck-neon-core.S:419: Error: r13 not allowed here -- `bic sp,#0xf'
arch/arm/crypto/speck-neon-core.S:423: Error: r13 not allowed here -- `bic sp,#0xf'
arch/arm/crypto/speck-neon-core.S:427: Error: r13 not allowed here -- `bic sp,#0xf'
arch/arm/crypto/speck-neon-core.S:431: Error: r13 not allowed here -- `bic sp,#0xf'
The problem is that the 'bic' instruction can't operate on the 'sp'
register in Thumb2 mode. Fix it by using a temporary register. This
isn't in the main loop, so the performance difference is negligible.
This also matches what aes-neonbs-core.S does.
Reported-by: Stefan Agner <stefan@agner.ch>
Fixes: ede9622162 ("crypto: arm/speck - add NEON-accelerated implementation of Speck-XTS")
Signed-off-by: Eric Biggers <ebiggers@google.com>
Acked-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Reviewed-by: Stefan Agner <stefan@agner.ch>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Several source files have been taken from OpenSSL. In some of them a
comment that "permission to use under GPL terms is granted" was
included below a contradictory license statement. In several cases,
there was no indication that the license of the code was compatible
with the GPLv2.
This change clarifies the licensing for all of these files. I've
confirmed with the author (Andy Polyakov) that a) he has licensed the
files with the GPLv2 comment under that license and b) that he's also
happy to license the other files under GPLv2 too. In one case, the
file is already contained in his CRYPTOGAMS bundle, which has a GPLv2
option, and so no special measures are needed.
In all cases, the license status of code has been clarified by making
the GPLv2 license prominent.
The .S files have been regenerated from the updated .pl files.
This is a comment-only change. No code is changed.
Signed-off-by: Adam Langley <agl@chromium.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
GNU Make automatically deletes intermediate files that are updated
in a chain of pattern rules.
Example 1) %.dtb.o <- %.dtb.S <- %.dtb <- %.dts
Example 2) %.o <- %.c <- %.c_shipped
A couple of makefiles mark such targets as .PRECIOUS to prevent Make
from deleting them, but the correct way is to use .SECONDARY.
.SECONDARY
Prerequisites of this special target are treated as intermediate
files but are never automatically deleted.
.PRECIOUS
When make is interrupted during execution, it may delete the target
file it is updating if the file was modified since make started.
If you mark the file as precious, make will never delete the file
if interrupted.
Both can avoid deletion of intermediate files, but the difference is
the behavior when Make is interrupted; .SECONDARY deletes the target,
but .PRECIOUS does not.
The use of .PRECIOUS is relatively rare since we do not want to keep
partially constructed (possibly corrupted) targets.
Another difference is that .PRECIOUS works with pattern rules whereas
.SECONDARY does not.
.PRECIOUS: $(obj)/%.lex.c
works, but
.SECONDARY: $(obj)/%.lex.c
has no effect. However, for the reason above, I do not want to use
.PRECIOUS which could cause obscure build breakage.
The targets specified as .SECONDARY must be explicit. $(targets)
contains all targets that need to include .*.cmd files. So, the
intermediates you want to keep are mostly in there. Therefore, mark
$(targets) as .SECONDARY. It means primary targets are also marked
as .SECONDARY, but I do not see any drawback for this.
I replaced some .SECONDARY / .PRECIOUS markers with 'targets'. This
will make Kbuild search for non-existing .*.cmd files, but this is
not a noticeable performance issue.
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Acked-by: Frank Rowand <frowand.list@gmail.com>
Acked-by: Ingo Molnar <mingo@kernel.org>
The decision to rebuild .S_shipped is made based on the relative
timestamps of .S_shipped and .pl files but git makes this essentially
random. This means that the perl script might run anyway (usually at
most once per checkout), defeating the whole purpose of _shipped.
Fix by skipping the rule unless explicit make variables are provided:
REGENERATE_ARM_CRYPTO or REGENERATE_ARM64_CRYPTO.
This can produce nasty occasional build failures downstream, for example
for toolchains with broken perl. The solution is minimally intrusive to
make it easier to push into stable.
Another report on a similar issue here: https://lkml.org/lkml/2018/3/8/1379
Signed-off-by: Leonard Crestez <leonard.crestez@nxp.com>
Cc: <stable@vger.kernel.org>
Reviewed-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Acked-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Add an ARM NEON-accelerated implementation of Speck-XTS. It operates on
128-byte chunks at a time, i.e. 8 blocks for Speck128 or 16 blocks for
Speck64. Each 128-byte chunk goes through XTS preprocessing, then is
encrypted/decrypted (doing one cipher round for all the blocks, then the
next round, etc.), then goes through XTS postprocessing.
The performance depends on the processor but can be about 3 times faster
than the generic code. For example, on an ARMv7 processor we observe
the following performance with Speck128/256-XTS:
xts-speck128-neon: Encryption 107.9 MB/s, Decryption 108.1 MB/s
xts(speck128-generic): Encryption 32.1 MB/s, Decryption 36.6 MB/s
In comparison to AES-256-XTS without the Cryptography Extensions:
xts-aes-neonbs: Encryption 41.2 MB/s, Decryption 36.7 MB/s
xts(aes-asm): Encryption 31.7 MB/s, Decryption 30.8 MB/s
xts(aes-generic): Encryption 21.2 MB/s, Decryption 20.9 MB/s
Speck64/128-XTS is even faster:
xts-speck64-neon: Encryption 138.6 MB/s, Decryption 139.1 MB/s
Note that as with the generic code, only the Speck128 and Speck64
variants are supported. Also, for now only the XTS mode of operation is
supported, to target the disk and file encryption use cases. The NEON
code also only handles the portion of the data that is evenly divisible
into 128-byte chunks, with any remainder handled by a C fallback. Of
course, other modes of operation could be added later if needed, and/or
the NEON code could be updated to handle other buffer sizes.
The XTS specification is only defined for AES which has a 128-bit block
size, so for the GF(2^64) math needed for Speck64-XTS we use the
reducing polynomial 'x^64 + x^4 + x^3 + x + 1' given by the original XEX
paper. Of course, when possible users should use Speck128-XTS, but even
that may be too slow on some processors; Speck64-XTS can be faster.
Signed-off-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Move the AES inverse S-box to the .rodata section
where it is safe from abuse by speculation.
Signed-off-by: Jinbum Park <jinb.park7@gmail.com>
Acked-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
We need to consistently enforce that keyed hashes cannot be used without
setting the key. To do this we need a reliable way to determine whether
a given hash algorithm is keyed or not. AF_ALG currently does this by
checking for the presence of a ->setkey() method. However, this is
actually slightly broken because the CRC-32 algorithms implement
->setkey() but can also be used without a key. (The CRC-32 "key" is not
actually a cryptographic key but rather represents the initial state.
If not overridden, then a default initial state is used.)
Prepare to fix this by introducing a flag CRYPTO_ALG_OPTIONAL_KEY which
indicates that the algorithm has a ->setkey() method, but it is not
required to be called. Then set it on all the CRC-32 algorithms.
The same also applies to the Adler-32 implementation in Lustre.
Also, the cryptd and mcryptd templates have to pass through the flag
from their underlying algorithm.
Cc: stable@vger.kernel.org
Signed-off-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Fix ptr_ret.cocci warnings:
arch/arm/crypto/aes-neonbs-glue.c:184:1-3: WARNING: PTR_ERR_OR_ZERO can be used
arch/arm/crypto/aes-neonbs-glue.c:261:1-3: WARNING: PTR_ERR_OR_ZERO can be used
Use PTR_ERR_OR_ZERO rather than if(IS_ERR(...)) + PTR_ERR
Generated by: scripts/coccinelle/api/ptr_ret.cocci
Signed-off-by: Vasyl Gomonovych <gomonovych@gmail.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Many source files in the tree are missing licensing information, which
makes it harder for compliance tools to determine the correct license.
By default all files without license information are under the default
license of the kernel, which is GPL version 2.
Update the files which contain no license information with the 'GPL-2.0'
SPDX license identifier. The SPDX identifier is a legally binding
shorthand, which can be used instead of the full boiler plate text.
This patch is based on work done by Thomas Gleixner and Kate Stewart and
Philippe Ombredanne.
How this work was done:
Patches were generated and checked against linux-4.14-rc6 for a subset of
the use cases:
- file had no licensing information it it.
- file was a */uapi/* one with no licensing information in it,
- file was a */uapi/* one with existing licensing information,
Further patches will be generated in subsequent months to fix up cases
where non-standard license headers were used, and references to license
had to be inferred by heuristics based on keywords.
The analysis to determine which SPDX License Identifier to be applied to
a file was done in a spreadsheet of side by side results from of the
output of two independent scanners (ScanCode & Windriver) producing SPDX
tag:value files created by Philippe Ombredanne. Philippe prepared the
base worksheet, and did an initial spot review of a few 1000 files.
The 4.13 kernel was the starting point of the analysis with 60,537 files
assessed. Kate Stewart did a file by file comparison of the scanner
results in the spreadsheet to determine which SPDX license identifier(s)
to be applied to the file. She confirmed any determination that was not
immediately clear with lawyers working with the Linux Foundation.
Criteria used to select files for SPDX license identifier tagging was:
- Files considered eligible had to be source code files.
- Make and config files were included as candidates if they contained >5
lines of source
- File already had some variant of a license header in it (even if <5
lines).
All documentation files were explicitly excluded.
The following heuristics were used to determine which SPDX license
identifiers to apply.
- when both scanners couldn't find any license traces, file was
considered to have no license information in it, and the top level
COPYING file license applied.
For non */uapi/* files that summary was:
SPDX license identifier # files
---------------------------------------------------|-------
GPL-2.0 11139
and resulted in the first patch in this series.
If that file was a */uapi/* path one, it was "GPL-2.0 WITH
Linux-syscall-note" otherwise it was "GPL-2.0". Results of that was:
SPDX license identifier # files
---------------------------------------------------|-------
GPL-2.0 WITH Linux-syscall-note 930
and resulted in the second patch in this series.
- if a file had some form of licensing information in it, and was one
of the */uapi/* ones, it was denoted with the Linux-syscall-note if
any GPL family license was found in the file or had no licensing in
it (per prior point). Results summary:
SPDX license identifier # files
---------------------------------------------------|------
GPL-2.0 WITH Linux-syscall-note 270
GPL-2.0+ WITH Linux-syscall-note 169
((GPL-2.0 WITH Linux-syscall-note) OR BSD-2-Clause) 21
((GPL-2.0 WITH Linux-syscall-note) OR BSD-3-Clause) 17
LGPL-2.1+ WITH Linux-syscall-note 15
GPL-1.0+ WITH Linux-syscall-note 14
((GPL-2.0+ WITH Linux-syscall-note) OR BSD-3-Clause) 5
LGPL-2.0+ WITH Linux-syscall-note 4
LGPL-2.1 WITH Linux-syscall-note 3
((GPL-2.0 WITH Linux-syscall-note) OR MIT) 3
((GPL-2.0 WITH Linux-syscall-note) AND MIT) 1
and that resulted in the third patch in this series.
- when the two scanners agreed on the detected license(s), that became
the concluded license(s).
- when there was disagreement between the two scanners (one detected a
license but the other didn't, or they both detected different
licenses) a manual inspection of the file occurred.
- In most cases a manual inspection of the information in the file
resulted in a clear resolution of the license that should apply (and
which scanner probably needed to revisit its heuristics).
- When it was not immediately clear, the license identifier was
confirmed with lawyers working with the Linux Foundation.
- If there was any question as to the appropriate license identifier,
the file was flagged for further research and to be revisited later
in time.
In total, over 70 hours of logged manual review was done on the
spreadsheet to determine the SPDX license identifiers to apply to the
source files by Kate, Philippe, Thomas and, in some cases, confirmation
by lawyers working with the Linux Foundation.
Kate also obtained a third independent scan of the 4.13 code base from
FOSSology, and compared selected files where the other two scanners
disagreed against that SPDX file, to see if there was new insights. The
Windriver scanner is based on an older version of FOSSology in part, so
they are related.
Thomas did random spot checks in about 500 files from the spreadsheets
for the uapi headers and agreed with SPDX license identifier in the
files he inspected. For the non-uapi files Thomas did random spot checks
in about 15000 files.
In initial set of patches against 4.14-rc6, 3 files were found to have
copy/paste license identifier errors, and have been fixed to reflect the
correct identifier.
Additionally Philippe spent 10 hours this week doing a detailed manual
inspection and review of the 12,461 patched files from the initial patch
version early this week with:
- a full scancode scan run, collecting the matched texts, detected
license ids and scores
- reviewing anything where there was a license detected (about 500+
files) to ensure that the applied SPDX license was correct
- reviewing anything where there was no detection but the patch license
was not GPL-2.0 WITH Linux-syscall-note to ensure that the applied
SPDX license was correct
This produced a worksheet with 20 files needing minor correction. This
worksheet was then exported into 3 different .csv files for the
different types of files to be modified.
These .csv files were then reviewed by Greg. Thomas wrote a script to
parse the csv files and add the proper SPDX tag to the file, in the
format that the file expected. This script was further refined by Greg
based on the output to detect more types of files automatically and to
distinguish between header and source .c files (which need different
comment types.) Finally Greg ran the script using the .csv files to
generate the patches.
Reviewed-by: Kate Stewart <kstewart@linuxfoundation.org>
Reviewed-by: Philippe Ombredanne <pombredanne@nexb.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
For the final round, avoid the expanded and padded lookup tables
exported by the generic AES driver. Instead, for encryption, we can
perform byte loads from the same table we used for the inner rounds,
which will still be hot in the caches. For decryption, use the inverse
AES Sbox directly, which is 4x smaller than the inverse lookup table
exported by the generic driver.
This should significantly reduce the Dcache footprint of our code,
which makes the code more robust against timing attacks. It does not
introduce any additional module dependencies, given that we already
rely on the core AES module for the shared key expansion routines.
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Implement a NEON fallback for systems that do support NEON but have
no support for the optional 64x64->128 polynomial multiplication
instruction that is part of the ARMv8 Crypto Extensions. It is based
on the paper "Fast Software Polynomial Multiplication on ARM Processors
Using the NEON Engine" by Danilo Camara, Conrado Gouvea, Julio Lopez and
Ricardo Dahab (https://hal.inria.fr/hal-01506572)
On a 32-bit guest executing under KVM on a Cortex-A57, the new code is
not only 4x faster than the generic table based GHASH driver, it is also
time invariant. (Note that the existing vmull.p64 code is 16x faster on
this core).
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
There are quite a number of occurrences in the kernel of the pattern
if (dst != src)
memcpy(dst, src, walk.total % AES_BLOCK_SIZE);
crypto_xor(dst, final, walk.total % AES_BLOCK_SIZE);
or
crypto_xor(keystream, src, nbytes);
memcpy(dst, keystream, nbytes);
where crypto_xor() is preceded or followed by a memcpy() invocation
that is only there because crypto_xor() uses its output parameter as
one of the inputs. To avoid having to add new instances of this pattern
in the arm64 code, which will be refactored to implement non-SIMD
fallbacks, add an alternative implementation called crypto_xor_cpy(),
taking separate input and output arguments. This removes the need for
the separate memcpy().
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Make the module autoloadable by tying it to the CPU feature bits that
describe whether the optional instructions it relies on are implemented
by the current CPU.
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Make the module autoloadable by tying it to the CPU feature bit that
describes whether the optional instructions it relies on are implemented
by the current CPU.
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Make the module autoloadable by tying it to the CPU feature bit that
describes whether the optional instructions it relies on are implemented
by the current CPU.
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Make the module autoloadable by tying it to the CPU feature bit that
describes whether the optional instructions it relies on are implemented
by the current CPU.
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Make the module autoloadable by tying it to the CPU feature bit that
describes whether the optional instructions it relies on are implemented
by the current CPU.
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Currently, the bit sliced NEON AES code for ARM has a link time
dependency on the scalar ARM asm implementation, which it uses as a
fallback to perform CBC encryption and the encryption of the initial
XTS tweak.
The bit sliced NEON code is both fast and time invariant, which makes
it a reasonable default on hardware that supports it. However, the
ARM asm code it pulls in is not time invariant, and due to the way it
is linked in, cannot be overridden by the new generic time invariant
driver. In fact, it will not be used at all, given that the ARM asm
code registers itself as a cipher with a priority that exceeds the
priority of the fixed time cipher.
So remove the link time dependency, and allocate the fallback cipher
via the crypto API. Note that this requires this driver's module_init
call to be replaced with late_initcall, so that the (possibly generic)
fallback cipher is guaranteed to be available when the builtin test
is performed at registration time.
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
The accelerated CRC32 module for ARM may use either the scalar CRC32
instructions, the NEON 64x64 to 128 bit polynomial multiplication
(vmull.p64) instruction, or both, depending on what the current CPU
supports.
However, this also requires support in binutils, and as it turns out,
versions of binutils exist that support the vmull.p64 instruction but
not the crc32 instructions.
So refactor the Makefile logic so that this module only gets built if
binutils has support for both.
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Acked-by: Jon Hunter <jonathanh@nvidia.com>
Tested-by: Jon Hunter <jonathanh@nvidia.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Annotate a vmov instruction with an explicit element size of 32 bits.
This is inferred by recent toolchains, but apparently, older versions
need some help figuring this out.
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
The ARM bit sliced AES core code uses the IV buffer to pass the final
keystream block back to the glue code if the input is not a multiple of
the block size, so that the asm code does not have to deal with anything
except 16 byte blocks. This is done under the assumption that the outgoing
IV is meaningless anyway in this case, given that chaining is no longer
possible under these circumstances.
However, as it turns out, the CCM driver does expect the IV to retain
a value that is equal to the original IV except for the counter value,
and even interprets byte zero as a length indicator, which may result
in memory corruption if the IV is overwritten with something else.
So use a separate buffer to return the final keystream block.
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Remove the unnecessary alignmask: it is much more efficient to deal with
the misalignment in the core algorithm than relying on the crypto API to
copy the data to a suitably aligned buffer.
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Remove the unnecessary alignmask: it is much more efficient to deal with
the misalignment in the core algorithm than relying on the crypto API to
copy the data to a suitably aligned buffer.
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
The GNU assembler for ARM version 2.22 or older fails to infer the
element size from the vmov instructions, and aborts the build in
the following way;
.../aes-neonbs-core.S: Assembler messages:
.../aes-neonbs-core.S:817: Error: bad type for scalar -- `vmov q1h[1],r10'
.../aes-neonbs-core.S:817: Error: bad type for scalar -- `vmov q1h[0],r9'
.../aes-neonbs-core.S:817: Error: bad type for scalar -- `vmov q1l[1],r8'
.../aes-neonbs-core.S:817: Error: bad type for scalar -- `vmov q1l[0],r7'
.../aes-neonbs-core.S:818: Error: bad type for scalar -- `vmov q2h[1],r10'
.../aes-neonbs-core.S:818: Error: bad type for scalar -- `vmov q2h[0],r9'
.../aes-neonbs-core.S:818: Error: bad type for scalar -- `vmov q2l[1],r8'
.../aes-neonbs-core.S:818: Error: bad type for scalar -- `vmov q2l[0],r7'
Fix this by setting the element size explicitly, by replacing vmov with
vmov.32.
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
The ARMv8-M architecture introduces 'tt' and 'ttt' instructions,
which means we can no longer use 'tt' as a register alias on recent
versions of binutils for ARM. So replace the alias with 'ttab'.
Fixes: 81edb42629 ("crypto: arm/aes - replace scalar AES cipher")
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
This replaces the unwieldy generated implementation of bit-sliced AES
in CBC/CTR/XTS modes that originated in the OpenSSL project with a
new version that is heavily based on the OpenSSL implementation, but
has a number of advantages over the old version:
- it does not rely on the scalar AES cipher that also originated in the
OpenSSL project and contains redundant lookup tables and key schedule
generation routines (which we already have in crypto/aes_generic.)
- it uses the same expanded key schedule for encryption and decryption,
reducing the size of the per-key data structure by 1696 bytes
- it adds an implementation of AES in ECB mode, which can be wrapped by
other generic chaining mode implementations
- it moves the handling of corner cases that are non critical to performance
to the glue layer written in C
- it was written directly in assembler rather than generated from a Perl
script
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
This replaces the scalar AES cipher that originates in the OpenSSL project
with a new implementation that is ~15% (*) faster (on modern cores), and
reuses the lookup tables and the key schedule generation routines from the
generic C implementation (which is usually compiled in anyway due to
networking and other subsystems depending on it).
Note that the bit sliced NEON code for AES still depends on the scalar cipher
that this patch replaces, so it is not removed entirely yet.
* On Cortex-A57, the performance increases from 17.0 to 14.9 cycles per byte
for 128-bit keys.
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
This is a straight port to ARM/NEON of the x86 SSE3 implementation
of the ChaCha20 stream cipher. It uses the new skcipher walksize
attribute to process the input in strides of 4x the block size.
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
This patch reverts the following commits:
8621caa0d48096667273
I should not have applied them because they had already been
obsoleted by a subsequent patch series. They also cause a build
failure because of the subsequent commit 9ae433bc79.
Fixes: 9ae433bc79 ("crypto: chacha20 - convert generic and...")
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
This is a straight port to ARM/NEON of the x86 SSE3 implementation
of the ChaCha20 stream cipher.
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
This is a combination of the the Intel algorithm implemented using SSE
and PCLMULQDQ instructions from arch/x86/crypto/crc32-pclmul_asm.S, and
the new CRC32 extensions introduced for both 32-bit and 64-bit ARM in
version 8 of the architecture. Two versions of the above combo are
provided, one for CRC32 and one for CRC32C.
The PMULL/NEON algorithm is faster, but operates on blocks of at least
64 bytes, and on multiples of 16 bytes only. For the remaining input,
or for all input on systems that lack the PMULL 64x64->128 instructions,
the CRC32 instructions will be used.
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
This is a transliteration of the Intel algorithm implemented
using SSE and PCLMULQDQ instructions that resides in the file
arch/x86/crypto/crct10dif-pcl-asm_64.S, but simplified to only
operate on buffers that are 16 byte aligned (but of any size)
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
The variable aes_simd_algs should be static. In fact if it isn't
it causes build errors when multiple copies of aes-ce-glue.c are
built into the kernel.
Fixes: da40e7a4ba ("crypto: aes-ce - Convert to skcipher")
Reported-by: kbuild test robot <fengguang.wu@intel.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
The CBC encryption routine should use the encryption round keys, not
the decryption round keys.
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
This patch adds one more missing SIMD select for AES_ARM_BS. It
also changes selects on ALGAPI to BLKCIPHER.
Fixes: 211f41af53 ("crypto: aesbs - Convert to skcipher")
Reported-by: Horia Geantă <horia.geanta@nxp.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
The skcipher conversion for ARM missed the select on CRYPTO_SIMD,
causing build failures if SIMD was not otherwise enabled.
Fixes: da40e7a4ba ("crypto: aes-ce - Convert to skcipher")
Fixes: 211f41af53 ("crypto: aesbs - Convert to skcipher")
Reported-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
The AES key schedule generation is mostly endian agnostic, with the
exception of the rotation and the incorporation of the round constant
at the start of each round. So implement a big endian specific version
of that part to make the whole routine big endian compatible.
Fixes: 86464859cc ("crypto: arm - AES in ECB/CBC/CTR/XTS modes using ARMv8 Crypto Extensions")
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
The AES-CTR glue code avoids calling into the blkcipher API for the
tail portion of the walk, by comparing the remainder of walk.nbytes
modulo AES_BLOCK_SIZE with the residual nbytes, and jumping straight
into the tail processing block if they are equal. This tail processing
block checks whether nbytes != 0, and does nothing otherwise.
However, in case of an allocation failure in the blkcipher layer, we
may enter this code with walk.nbytes == 0, while nbytes > 0. In this
case, we should not dereference the source and destination pointers,
since they may be NULL. So instead of checking for nbytes != 0, check
for (walk.nbytes % AES_BLOCK_SIZE) != 0, which implies the former in
non-error conditions.
Fixes: 86464859cc ("crypto: arm - AES in ECB/CBC/CTR/XTS modes using ARMv8 Crypto Extensions")
Cc: stable@vger.kernel.org
Reported-by: xiakaixu <xiakaixu@huawei.com>
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
The fact that the internal synchrous hash implementation is called
"ghash" like the publicly visible one is causing the testmgr code
to misidentify it as an algorithm that requires testing at boottime.
So rename it to "__ghash" to prevent this.
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Since commit 8996eafdcb ("crypto: ahash - ensure statesize is non-zero"),
all ahash drivers are required to implement import()/export(), and must have
a non-zero statesize. Fix this for the ARM Crypto Extensions GHASH
implementation.
Fixes: 8996eafdcb ("crypto: ahash - ensure statesize is non-zero")
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
The ARMv7 NEON module is explicitly built in ARM mode, which is not
supported by the Thumb2 kernel. So remove the explicit override, and
leave it up to the build environment to decide whether the core SHA1
routines are assembled as ARM or as Thumb2 code.
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
This patch fixes an old bug where requests can be reordered because
some are processed by cryptd while others are processed directly
in softirq context.
The fix is to always postpone to cryptd if there are currently
requests outstanding from the same tfm.
This patch also removes the redundant use of cryptd in the async
init function as init never touches the FPU.
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Pull crypto update from Herbert Xu:
"Here is the crypto update for 4.6:
API:
- Convert remaining crypto_hash users to shash or ahash, also convert
blkcipher/ablkcipher users to skcipher.
- Remove crypto_hash interface.
- Remove crypto_pcomp interface.
- Add crypto engine for async cipher drivers.
- Add akcipher documentation.
- Add skcipher documentation.
Algorithms:
- Rename crypto/crc32 to avoid name clash with lib/crc32.
- Fix bug in keywrap where we zero the wrong pointer.
Drivers:
- Support T5/M5, T7/M7 SPARC CPUs in n2 hwrng driver.
- Add PIC32 hwrng driver.
- Support BCM6368 in bcm63xx hwrng driver.
- Pack structs for 32-bit compat users in qat.
- Use crypto engine in omap-aes.
- Add support for sama5d2x SoCs in atmel-sha.
- Make atmel-sha available again.
- Make sahara hashing available again.
- Make ccp hashing available again.
- Make sha1-mb available again.
- Add support for multiple devices in ccp.
- Improve DMA performance in caam.
- Add hashing support to rockchip"
* 'linus' of git://git.kernel.org/pub/scm/linux/kernel/git/herbert/crypto-2.6: (116 commits)
crypto: qat - remove redundant arbiter configuration
crypto: ux500 - fix checks of error code returned by devm_ioremap_resource()
crypto: atmel - fix checks of error code returned by devm_ioremap_resource()
crypto: qat - Change the definition of icp_qat_uof_regtype
hwrng: exynos - use __maybe_unused to hide pm functions
crypto: ccp - Add abstraction for device-specific calls
crypto: ccp - CCP versioning support
crypto: ccp - Support for multiple CCPs
crypto: ccp - Remove check for x86 family and model
crypto: ccp - memset request context to zero during import
lib/mpi: use "static inline" instead of "extern inline"
lib/mpi: avoid assembler warning
hwrng: bcm63xx - fix non device tree compatibility
crypto: testmgr - allow rfc3686 aes-ctr variants in fips mode.
crypto: qat - The AE id should be less than the maximal AE number
lib/mpi: Endianness fix
crypto: rockchip - add hash support for crypto engine in rk3288
crypto: xts - fix compile errors
crypto: doc - add skcipher API documentation
crypto: doc - update AEAD AD handling
...
Commit 28856a9e52 missed the addition of the crypto/xts.h include file
for different architecture-specific AES implementations.
Signed-off-by: Stephan Mueller <smueller@chronox.de>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
The patch centralizes the XTS key check logic into the service function
xts_check_key which is invoked from the different XTS implementations.
With this, the XTS implementations in ARM, ARM64, PPC and S390 have now
a sanity check for the XTS keys similar to the other arches.
In addition, this service function received a check to ensure that the
key != the tweak key which is mandated by FIPS 140-2 IG A.9. As the
check is not present in the standards defining XTS, it is only enforced
in FIPS mode of the kernel.
Signed-off-by: Stephan Mueller <smueller@chronox.de>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
ECB modes don't use an initialization vector. The kernel
/proc/crypto interface doesn't reflect this properly.
Acked-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Jeremy Linton <jeremy.linton@arm.com>
Signed-off-by: Will Deacon <will.deacon@arm.com>
This trims off a couple of instructions of the total size of the
core AES transform by reordering the final branch in the AES-192
code path with the rounds that are performed regardless of whether
the branch is taken or not. Other than the slight size reduction,
this has no performance benefit.
Fix up a comment regarding the prototype of this function while
we're at it.
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
This replaces the SHA-512 NEON module with the faster and more
versatile implementation from the OpenSSL project. It consists
of both a NEON and a generic ASM version of the core SHA-512
transform, where the NEON version reverts to the ASM version
when invoked in non-process context.
This patch is based on the OpenSSL upstream version b1a5d1c65208
of sha512-armv4.pl, which can be found here:
https://git.openssl.org/gitweb/?p=openssl.git;h=b1a5d1c65208
Performance relative to the generic implementation (measured
using tcrypt.ko mode=306 sec=1 running on a Cortex-A57 under
KVM):
input size block size asm neon old neon
16 16 1.39 2.54 2.21
64 16 1.32 2.33 2.09
64 64 1.38 2.53 2.19
256 16 1.31 2.28 2.06
256 64 1.38 2.54 2.25
256 256 1.40 2.77 2.39
1024 16 1.29 2.22 2.01
1024 256 1.40 2.82 2.45
1024 1024 1.41 2.93 2.53
2048 16 1.33 2.21 2.00
2048 256 1.40 2.84 2.46
2048 1024 1.41 2.96 2.55
2048 2048 1.41 2.98 2.56
4096 16 1.34 2.20 1.99
4096 256 1.40 2.84 2.46
4096 1024 1.41 2.97 2.56
4096 4096 1.41 3.01 2.58
8192 16 1.34 2.19 1.99
8192 256 1.40 2.85 2.47
8192 1024 1.41 2.98 2.56
8192 4096 1.41 2.71 2.59
8192 8192 1.51 3.51 2.69
Acked-by: Jussi Kivilinna <jussi.kivilinna@iki.fi>
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Pull crypto update from Herbert Xu:
"Here is the crypto update for 4.1:
New interfaces:
- user-space interface for AEAD
- user-space interface for RNG (i.e., pseudo RNG)
New hashes:
- ARMv8 SHA1/256
- ARMv8 AES
- ARMv8 GHASH
- ARM assembler and NEON SHA256
- MIPS OCTEON SHA1/256/512
- MIPS img-hash SHA1/256 and MD5
- Power 8 VMX AES/CBC/CTR/GHASH
- PPC assembler AES, SHA1/256 and MD5
- Broadcom IPROC RNG driver
Cleanups/fixes:
- prevent internal helper algos from being exposed to user-space
- merge common code from assembly/C SHA implementations
- misc fixes"
* git://git.kernel.org/pub/scm/linux/kernel/git/herbert/crypto-2.6: (169 commits)
crypto: arm - workaround for building with old binutils
crypto: arm/sha256 - avoid sha256 code on ARMv7-M
crypto: x86/sha512_ssse3 - move SHA-384/512 SSSE3 implementation to base layer
crypto: x86/sha256_ssse3 - move SHA-224/256 SSSE3 implementation to base layer
crypto: x86/sha1_ssse3 - move SHA-1 SSSE3 implementation to base layer
crypto: arm64/sha2-ce - move SHA-224/256 ARMv8 implementation to base layer
crypto: arm64/sha1-ce - move SHA-1 ARMv8 implementation to base layer
crypto: arm/sha2-ce - move SHA-224/256 ARMv8 implementation to base layer
crypto: arm/sha256 - move SHA-224/256 ASM/NEON implementation to base layer
crypto: arm/sha1-ce - move SHA-1 ARMv8 implementation to base layer
crypto: arm/sha1_neon - move SHA-1 NEON implementation to base layer
crypto: arm/sha1 - move SHA-1 ARM asm implementation to base layer
crypto: sha512-generic - move to generic glue implementation
crypto: sha256-generic - move to generic glue implementation
crypto: sha1-generic - move to generic glue implementation
crypto: sha512 - implement base layer for SHA-512
crypto: sha256 - implement base layer for SHA-256
crypto: sha1 - implement base layer for SHA-1
crypto: api - remove instance when test failed
crypto: api - Move alg ref count init to crypto_check_alg
...
Old versions of binutils (before 2.23) do not yet understand the
crypto-neon-fp-armv8 fpu instructions, and an attempt to build these
files results in a build failure:
arch/arm/crypto/aes-ce-core.S:133: Error: selected processor does not support ARM mode `vld1.8 {q10-q11},[ip]!'
arch/arm/crypto/aes-ce-core.S:133: Error: bad instruction `aese.8 q0,q8'
arch/arm/crypto/aes-ce-core.S:133: Error: bad instruction `aesmc.8 q0,q0'
arch/arm/crypto/aes-ce-core.S:133: Error: bad instruction `aese.8 q0,q9'
arch/arm/crypto/aes-ce-core.S:133: Error: bad instruction `aesmc.8 q0,q0'
Since the affected versions are still in widespread use, and this breaks
'allmodconfig' builds, we should try to at least get a successful kernel
build. Unfortunately, I could not come up with a way to make the Kconfig
symbol depend on the binutils version, which would be the nicest solution.
Instead, this patch uses the 'as-instr' Kbuild macro to find out whether
the support is present in the assembler, and otherwise emits a non-fatal
warning indicating which selected modules could not be built.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Link: http://storage.kernelci.org/next/next-20150410/arm-allmodconfig/build.log
Fixes: 864cbeed4a ("crypto: arm - add support for SHA1 using ARMv8 Crypto Instructions")
[ard.biesheuvel:
- omit modules entirely instead of building empty ones if binutils is too old
- update commit log accordingly]
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
The sha256 assembly implementation can deal with all architecture levels
from ARMv4 to ARMv7-A, but not with ARMv7-M. Enabling it in an
ARMv7-M kernel results in this build failure:
arm-linux-gnueabi-ld: error: arch/arm/crypto/sha256_glue.o: Conflicting architecture profiles M/A
arm-linux-gnueabi-ld: failed to merge target specific data of file arch/arm/crypto/sha256_glue.o
This adds a Kconfig dependency to prevent the code from being disabled
for ARMv7-M.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
This removes all the boilerplate from the existing implementation,
and replaces it with calls into the base layer.
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
This removes all the boilerplate from the existing implementation,
and replaces it with calls into the base layer.
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
This removes all the boilerplate from the existing implementation,
and replaces it with calls into the base layer.
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
This removes all the boilerplate from the existing implementation,
and replaces it with calls into the base layer.
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
This removes all the boilerplate from the existing implementation,
and replaces it with calls into the base layer.
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Flag all ARMv8 AES helper ciphers as internal ciphers to prevent
them from being called by normal users.
Signed-off-by: Stephan Mueller <smueller@chronox.de>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Flag all NEON bit sliced AES helper ciphers as internal ciphers to
prevent them from being called by normal users.
Signed-off-by: Stephan Mueller <smueller@chronox.de>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Flag all GHASH ARMv8 vmull.p64 helper ciphers as internal ciphers
to prevent them from being called by normal users.
Signed-off-by: Stephan Mueller <smueller@chronox.de>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
This fixes a bug in the new v8 Crypto Extensions GHASH code
that only manifests itself in big-endian mode.
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
This implements the GHASH hash algorithm (as used by the GCM AEAD
chaining mode) using the AArch32 version of the 64x64 to 128 bit
polynomial multiplication instruction (vmull.p64) that is part of
the ARMv8 Crypto Extensions.
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
This implements the ECB, CBC, CTR and XTS asynchronous block ciphers
using the AArch32 versions of the ARMv8 Crypto Extensions for AES.
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
This implements the SHA-224/256 secure hash algorithm using the AArch32
versions of the ARMv8 Crypto Extensions for SHA2.
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
This implements the SHA1 secure hash algorithm using the AArch32
versions of the ARMv8 Crypto Extensions for SHA1.
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
This moves all Kconfig symbols defined in crypto/Kconfig that depend
on CONFIG_ARM to a dedicated Kconfig file in arch/arm/crypto, which is
where the code that implements those features resides as well.
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
This updates the bit sliced AES module to the latest version in the
upstream OpenSSL repository (e620e5ae37bc). This is needed to fix a
bug in the XTS decryption path, where data chunked in a certain way
could trigger the ciphertext stealing code, which is not supposed to
be active in the kernel build (The kernel implementation of XTS only
supports round multiples of the AES block size of 16 bytes, whereas
the conformant OpenSSL implementation of XTS supports inputs of
arbitrary size by applying ciphertext stealing). This is fixed in
the upstream version by adding the missing #ifndef XTS_CHAIN_TWEAK
around the offending instructions.
The upstream code also contains the change applied by Russell to
build the code unconditionally, i.e., even if __LINUX_ARM_ARCH__ < 7,
but implemented slightly differently.
Cc: stable@vger.kernel.org
Fixes: e4e7f10bfc ("ARM: add support for bit sliced AES using NEON instructions")
Reported-by: Adrian Kotelba <adrian.kotelba@gmail.com>
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Tested-by: Milan Broz <gmazyland@gmail.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Memset on a local variable may be removed when it is called just before the
variable goes out of scope. Using memzero_explicit defeats this
optimization. A simplified version of the semantic patch that makes this
change is as follows: (http://coccinelle.lip6.fr/)
// <smpl>
@@
identifier x;
type T;
@@
{
... when any
T x[...];
... when any
when exists
- memset
+ memzero_explicit
(x,
-0,
...)
... when != x
when strict
}
// </smpl>
This change was suggested by Daniel Borkmann <dborkman@redhat.com>
Signed-off-by: Julia Lawall <Julia.Lawall@lip6.fr>
Acked-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
This prefixes all crypto module loading with "crypto-" so we never run
the risk of exposing module auto-loading to userspace via a crypto API,
as demonstrated by Mathias Krause:
https://lkml.org/lkml/2013/3/4/70
Signed-off-by: Kees Cook <keescook@chromium.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
This tweaks the SHA-1 NEON code slightly so it works correctly under big
endian, and removes the Kconfig condition preventing it from being
selected if CONFIG_CPU_BIG_ENDIAN is set.
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Pull ARM updates from Russell King:
"Included in this update:
- perf updates from Will Deacon:
The main changes are callchain stability fixes from Jean Pihet and
event mapping and PMU name rework from Mark Rutland
The latter is preparatory work for enabling some code re-use with
arm64 in the future.
- updates for nommu from Uwe Kleine-König:
Two different fixes for the same problem making some ARM nommu
configurations not boot since 3.6-rc1. The problem is that
user_addr_max returned the biggest available RAM address which
makes some copy_from_user variants fail to read from XIP memory.
- deprecate legacy OMAP DMA API, in preparation for it's removal.
The popular drivers have been converted over, leaving a very small
number of rarely used drivers, which hopefully can be converted
during the next cycle with a bit more visibility (and hopefully
people popping out of the woodwork to help test)
- more tweaks for BE systems, particularly with the kernel image
format. In connection with this, I've cleaned up the way we
generate the linker script for the decompressor.
- removal of hard-coded assumptions of the kernel stack size, making
everywhere depend on the value of THREAD_SIZE_ORDER.
- MCPM updates from Nicolas Pitre.
- Make it easier for proper CPU part number checks (which should
always include the vendor field).
- Assembly code optimisation - use the "bx" instruction when
returning from a function on ARMv6+ rather than "mov pc, reg".
- Save the last kernel misaligned fault location and report it via
the procfs alignment file.
- Clean up the way we create the initial stack frame, which is a
repeated pattern in several different locations.
- Support for 8-byte get_user(), needed for some DRM implementations.
- mcs locking from Will Deacon.
- Save and restore a few more Cortex-A9 registers (for errata
workarounds)
- Fix various aspects of the SWP emulation, and the ELF hwcap for the
SWP instruction.
- Update LPAE logic for pte_write and pmd_write to make it more
correct.
- Support for Broadcom Brahma15 CPU cores.
- ARM assembly crypto updates from Ard Biesheuvel"
* 'for-linus' of git://ftp.arm.linux.org.uk/~rmk/linux-arm: (53 commits)
ARM: add comments to the early page table remap code
ARM: 8122/1: smp_scu: enable SCU standby support
ARM: 8121/1: smp_scu: use macro for SCU enable bit
ARM: 8120/1: crypto: sha512: add ARM NEON implementation
ARM: 8119/1: crypto: sha1: add ARM NEON implementation
ARM: 8118/1: crypto: sha1/make use of common SHA-1 structures
ARM: 8113/1: remove remaining definitions of PLAT_PHYS_OFFSET from <mach/memory.h>
ARM: 8111/1: Enable erratum 798181 for Broadcom Brahma-B15
ARM: 8110/1: do CPU-specific init for Broadcom Brahma15 cores
ARM: 8109/1: mm: Modify pte_write and pmd_write logic for LPAE
ARM: 8108/1: mm: Introduce {pte,pmd}_isset and {pte,pmd}_isclear
ARM: hwcap: disable HWCAP_SWP if the CPU advertises it has exclusives
ARM: SWP emulation: only initialise on ARMv7 CPUs
ARM: SWP emulation: always enable when SMP is enabled
ARM: 8103/1: save/restore Cortex-A9 CP15 registers on suspend/resume
ARM: 8098/1: mcs lock: implement wfe-based polling for MCS locking
ARM: 8091/2: add get_user() support for 8 byte types
ARM: 8097/1: unistd.h: relocate comments back to place
ARM: 8096/1: Describe required sort order for textofs-y (TEXT_OFFSET)
ARM: 8090/1: add revision info for PL310 errata 588369 and 727915
...
Common SHA-1 structures are defined in <crypto/sha.h> for code sharing.
This patch changes SHA-1/ARM glue code to use these structures.
Acked-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Jussi Kivilinna <jussi.kivilinna@iki.fi>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Fix the same alignment bug as in arm64 - we need to pass residue
unprocessed bytes as the last argument to blkcipher_walk_done.
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Cc: stable@vger.kernel.org # 3.13+
Acked-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
ARMv6 and greater introduced a new instruction ("bx") which can be used
to return from function calls. Recent CPUs perform better when the
"bx lr" instruction is used rather than the "mov pc, lr" instruction,
and this sequence is strongly recommended to be used by the ARM
architecture manual (section A.4.1.1).
We provide a new macro "ret" with all its variants for the condition
code which will resolve to the appropriate instruction.
Rather than doing this piecemeal, and miss some instances, change all
the "mov pc" instances to use the new macro, with the exception of
the "movs" instruction and the kprobes code. This allows us to detect
the "mov pc, lr" case and fix it up - and also gives us the possibility
of deploying this for other registers depending on the CPU selection.
Reported-by: Will Deacon <will.deacon@arm.com>
Tested-by: Stephen Warren <swarren@nvidia.com> # Tegra Jetson TK1
Tested-by: Robert Jarzmik <robert.jarzmik@free.fr> # mioa701_bootresume.S
Tested-by: Andrew Lunn <andrew@lunn.ch> # Kirkwood
Tested-by: Shawn Guo <shawn.guo@freescale.com>
Tested-by: Tony Lindgren <tony@atomide.com> # OMAPs
Tested-by: Gregory CLEMENT <gregory.clement@free-electrons.com> # Armada XP, 375, 385
Acked-by: Sekhar Nori <nsekhar@ti.com> # DaVinci
Acked-by: Christoffer Dall <christoffer.dall@linaro.org> # kvm/hyp
Acked-by: Haojian Zhuang <haojian.zhuang@gmail.com> # PXA3xx
Acked-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com> # Xen
Tested-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de> # ARMv7M
Tested-by: Simon Horman <horms+renesas@verge.net.au> # Shmobile
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Building a multi-arch kernel results in:
arch/arm/crypto/built-in.o: In function `aesbs_xts_decrypt':
sha1_glue.c:(.text+0x15c8): undefined reference to `bsaes_xts_decrypt'
arch/arm/crypto/built-in.o: In function `aesbs_xts_encrypt':
sha1_glue.c:(.text+0x1664): undefined reference to `bsaes_xts_encrypt'
arch/arm/crypto/built-in.o: In function `aesbs_ctr_encrypt':
sha1_glue.c:(.text+0x184c): undefined reference to `bsaes_ctr32_encrypt_blocks'
arch/arm/crypto/built-in.o: In function `aesbs_cbc_decrypt':
sha1_glue.c:(.text+0x19b4): undefined reference to `bsaes_cbc_encrypt'
This code is already runtime-conditional on NEON being supported, so
there's no point compiling it out depending on the minimum build
architecture.
Acked-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Bit sliced AES gives around 45% speedup on Cortex-A15 for encryption
and around 25% for decryption. This implementation of the AES algorithm
does not rely on any lookup tables so it is believed to be invulnerable
to cache timing attacks.
This algorithm processes up to 8 blocks in parallel in constant time. This
means that it is not usable by chaining modes that are strictly sequential
in nature, such as CBC encryption. CBC decryption, however, can benefit from
this implementation and runs about 25% faster. The other chaining modes
implemented in this module, XTS and CTR, can execute fully in parallel in
both directions.
The core code has been adopted from the OpenSSL project (in collaboration
with the original author, on cc). For ease of maintenance, this version is
identical to the upstream OpenSSL code, i.e., all modifications that were
required to make it suitable for inclusion into the kernel have been made
upstream. The original can be found here:
http://git.openssl.org/gitweb/?p=openssl.git;a=commit;h=6f6a6130
Note to integrators:
While this implementation is significantly faster than the existing table
based ones (generic or ARM asm), especially in CTR mode, the effects on
power efficiency are unclear as of yet. This code does fundamentally more
work, by calculating values that the table based code obtains by a simple
lookup; only by doing all of that work in a SIMD fashion, it manages to
perform better.
Cc: Andy Polyakov <appro@openssl.org>
Acked-by: Nicolas Pitre <nico@linaro.org>
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Put the struct definitions for AES keys and the asm function prototypes in a
separate header and export the asm functions from the module.
This allows other drivers to use them directly.
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Patch 638591c enabled building the AES assembler code in Thumb2 mode.
However, this code used arithmetic involving PC rather than adr{l}
instructions to generate PC-relative references to the lookup tables,
and this needs to take into account the different PC offset when
running in Thumb mode.
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Acked-by: Nicolas Pitre <nico@linaro.org>
Cc: stable@vger.kernel.org
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Make the SHA1 asm code ABI conformant by making sure all stack
accesses occur above the stack pointer.
Origin:
http://git.openssl.org/gitweb/?p=openssl.git;a=commit;h=1a9d60d2
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Acked-by: Nicolas Pitre <nico@linaro.org>
Cc: stable@vger.kernel.org
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
This patch fixes aes-armv4.S and sha1-armv4-large.S to work
natively in Thumb. This allows ARM/Thumb interworking workarounds
to be removed.
I also take the opportunity to convert some explicit assembler
directives for exported functions to the standard
ENTRY()/ENDPROC().
For the code itself:
* In sha1_block_data_order, use of TEQ with sp is deprecated in
ARMv7 and not supported in Thumb. For the branches back to
.L_00_15 and .L_40_59, the TEQ is converted to a CMP, under the
assumption that clobbering the C flag here will not cause
incorrect behaviour.
For the first branch back to .L_20_39_or_60_79 the C flag is
important, so sp is moved temporarily into another register so
that TEQ can be used for the comparison.
* In the AES code, most forms of register-indexed addressing with
shifts and rotates are not permitted for loads and stores in
Thumb, so the address calculation is done using a separate
instruction for the Thumb case.
The resulting code is unlikely to be optimally scheduled, but it
should not have a large impact given the overall size of the code.
I haven't run any benchmarks.
Signed-off-by: Dave Martin <dave.martin@linaro.org>
Tested-by: David McCullough <ucdevel@gmail.com> (ARM only)
Acked-by: David McCullough <ucdevel@gmail.com>
Acked-by: Nicolas Pitre <nico@linaro.org>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Add assembler versions of AES and SHA1 for ARM platforms. This has provided
up to a 50% improvement in IPsec/TCP throughout for tunnels using AES128/SHA1.
Platform CPU SPeed Endian Before (bps) After (bps) Improvement
IXP425 533 MHz big 11217042 15566294 ~38%
KS8695 166 MHz little 3828549 5795373 ~51%
Signed-off-by: David McCullough <ucdevel@gmail.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
