crypto: arm/chacha-neon - add missing counter increment

Commit 86cd97ec4b943af3 ("crypto: arm/chacha-neon - optimize for non-block
size multiples") refactored the chacha block handling in the glue code in
a way that may result in the counter increment to be omitted when calling
chacha_block_xor_neon() to process a full block. This violates the skcipher
API, which requires that the output IV is suitable for handling more input
as long as the preceding input has been presented in round multiples of the
block size. Also, the same code is exposed via the chacha library interface
whose callers may actually rely on this increment to occur even for final
blocks that are smaller than the chacha block size.

So increment the counter after calling chacha_block_xor_neon().

Fixes: 86cd97ec4b943af3 ("crypto: arm/chacha-neon - optimize for non-block size multiples")
Reported-by: Eric Biggers <ebiggers@kernel.org>
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

crypto: arm/chacha-neon - optimize for non-block size multiples

The current NEON based ChaCha implementation for ARM is optimized for
multiples of 4x the ChaCha block size (64 bytes). This makes sense for
block encryption, but given that ChaCha is also often used in the
context of networking, it makes sense to consider arbitrary length
inputs as well.

For example, WireGuard typically uses 1420 byte packets, and performing
ChaCha encryption involves 5 invocations of chacha_4block_xor_neon()
and 3 invocations of chacha_block_xor_neon(), where the last one also
involves a memcpy() using a buffer on the stack to process the final
chunk of 1420 % 64 == 12 bytes.

Let's optimize for this case as well, by letting chacha_4block_xor_neon()
deal with any input size between 64 and 256 bytes, using NEON permutation
instructions and overlapping loads and stores. This way, the 140 byte
tail of a 1420 byte input buffer can simply be processed in one go.

This results in the following performance improvements for 1420 byte
blocks, without significant impact on power-of-2 input sizes. (Note
that Raspberry Pi is widely used in combination with a 32-bit kernel,
even though the core is 64-bit capable)

Cortex-A8 (BeagleBone) : 7%
Cortex-A15 (Calxeda Midway) : 21%
Cortex-A53 (Raspberry Pi 3) : 3%
Cortex-A72 (Raspberry Pi 4) : 19%

Cc: Eric Biggers <ebiggers@google.com>
Cc: "Jason A . Donenfeld" <Jason@zx2c4.com>
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

crypto: arch/lib - limit simd usage to 4k chunks

The initial Zinc patchset, after some mailing list discussion, contained
code to ensure that kernel_fpu_enable would not be kept on for more than
a 4k chunk, since it disables preemption. The choice of 4k isn't totally
scientific, but it's not a bad guess either, and it's what's used in
both the x86 poly1305, blake2s, and nhpoly1305 code already (in the form
of PAGE_SIZE, which this commit corrects to be explicitly 4k for the
former two).

Ard did some back of the envelope calculations and found that
at 5 cycles/byte (overestimate) on a 1ghz processor (pretty slow), 4k
means we have a maximum preemption disabling of 20us, which Sebastian
confirmed was probably a good limit.

Unfortunately the chunking appears to have been left out of the final
patchset that added the glue code. So, this commit adds it back in.

Fixes: 84e03fa39fbe ("crypto: x86/chacha - expose SIMD ChaCha routine as library function")
Fixes: b3aad5bad26a ("crypto: arm64/chacha - expose arm64 ChaCha routine as library function")
Fixes: a44a3430d71b ("crypto: arm/chacha - expose ARM ChaCha routine as library function")
Fixes: d7d7b8535662 ("crypto: x86/poly1305 - wire up faster implementations for kernel")
Fixes: f569ca164751 ("crypto: arm64/poly1305 - incorporate OpenSSL/CRYPTOGAMS NEON implementation")
Fixes: a6b803b3ddc7 ("crypto: arm/poly1305 - incorporate OpenSSL/CRYPTOGAMS NEON implementation")
Fixes: ed0356eda153 ("crypto: blake2s - x86_64 SIMD implementation")
Cc: Eric Biggers <ebiggers@google.com>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: stable@vger.kernel.org
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Reviewed-by: Ard Biesheuvel <ardb@kernel.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

crypto: arm/chacha - fix build failured when kernel mode NEON is disabled

When the ARM accelerated ChaCha driver is built as part of a configuration
that has kernel mode NEON disabled, we expect the compiler to propagate
the build time constant expression IS_ENABLED(CONFIG_KERNEL_MODE_NEON) in
a way that eliminates all the cross-object references to the actual NEON
routines, which allows the chacha-neon-core.o object to be omitted from
the build entirely.

Unfortunately, this fails to work as expected in some cases, and we may
end up with a build error such as

chacha-glue.c:(.text+0xc0): undefined reference to `chacha_4block_xor_neon'

caused by the fact that chacha_doneon() has not been eliminated from the
object code, even though it will never be called in practice.

Let's fix this by adding some IS_ENABLED(CONFIG_KERNEL_MODE_NEON) tests
that are not strictly needed from a logical point of view, but should
help the compiler infer that the NEON code paths are unreachable in
those cases.

Fixes: b36d8c09e710c71f ("crypto: arm/chacha - remove dependency on generic ...")
Reported-by: Russell King <linux@armlinux.org.uk>
Cc: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

crypto: arch - conditionalize crypto api in arch glue for lib code

For glue code that's used by Zinc, the actual Crypto API functions might
not necessarily exist, and don't need to exist either. Before this
patch, there are valid build configurations that lead to a unbuildable
kernel. This fixes it to conditionalize those symbols on the existence
of the proper config entry.

Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Acked-by: Ard Biesheuvel <ardb@kernel.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

crypto: arm/chacha - expose ARM ChaCha routine as library function

Expose the accelerated NEON ChaCha routine directly as a symbol
export so that users of the ChaCha library API can use it directly.

Given that calls into the library API will always go through the
routines in this module if it is enabled, switch to static keys
to select the optimal implementation available (which may be none
at all, in which case we defer to the generic implementation for
all invocations).

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

crypto: arm/chacha - remove dependency on generic ChaCha driver

Instead of falling back to the generic ChaCha skcipher driver for
non-SIMD cases, use a fast scalar implementation for ARM authored
by Eric Biggers. This removes the module dependency on chacha-generic
altogether, which also simplifies things when we expose the ChaCha
library interface from this module.

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