On Sun, Mar 13, 2022 at 09:11:01PM -0600, Jason A. Donenfeld wrote:
> The lib/crypto libraries live in lib because they are used by various
> drivers of the kernel. In contrast, the various helper functions in
> crypto are there because they're used exclusively by the crypto API. The
> SM3 and SM4 helper functions were erroniously moved into lib/crypto/
> instead of crypto/, even though there are no in-kernel users outside of
> the crypto API of those functions. This commit moves them into crypto/.
>
> Cc: Herbert Xu <herbert@xxxxxxxxxxxxxxxxxxx>
> Cc: Tianjia Zhang <tianjia.zhang@xxxxxxxxxxxxxxxxx>
> Cc: Eric Biggers <ebiggers@xxxxxxxxxx>
> Signed-off-by: Jason A. Donenfeld <Jason@xxxxxxxxx>
> ---
> arch/arm64/crypto/Kconfig | 4 ++--
> crypto/Kconfig | 18 ++++++++++++------
> crypto/Makefile | 6 ++++--
> {lib/crypto => crypto}/sm3.c | 0
> {lib/crypto => crypto}/sm4.c | 0
> lib/crypto/Kconfig | 6 ------
> lib/crypto/Makefile | 6 ------
> 7 files changed, 18 insertions(+), 22 deletions(-)
> rename {lib/crypto => crypto}/sm3.c (100%)
> rename {lib/crypto => crypto}/sm4.c (100%)
>
> diff --git a/arch/arm64/crypto/Kconfig b/arch/arm64/crypto/Kconfig
> index 2a965aa0188d..454621a20eaa 100644
> --- a/arch/arm64/crypto/Kconfig
> +++ b/arch/arm64/crypto/Kconfig
> @@ -45,13 +45,13 @@ config CRYPTO_SM3_ARM64_CE
> tristate "SM3 digest algorithm (ARMv8.2 Crypto Extensions)"
> depends on KERNEL_MODE_NEON
> select CRYPTO_HASH
> - select CRYPTO_LIB_SM3
> + select CRYPTO_SM3
>
> config CRYPTO_SM4_ARM64_CE
> tristate "SM4 symmetric cipher (ARMv8.2 Crypto Extensions)"
> depends on KERNEL_MODE_NEON
> select CRYPTO_ALGAPI
> - select CRYPTO_LIB_SM4
> + select CRYPTO_SM4
>
> config CRYPTO_GHASH_ARM64_CE
> tristate "GHASH/AES-GCM using ARMv8 Crypto Extensions"
> diff --git a/crypto/Kconfig b/crypto/Kconfig
> index d6d7e84bb7f8..517525d7d12e 100644
> --- a/crypto/Kconfig
> +++ b/crypto/Kconfig
> @@ -274,7 +274,7 @@ config CRYPTO_ECRDSA
>
> config CRYPTO_SM2
> tristate "SM2 algorithm"
> - select CRYPTO_LIB_SM3
> + select CRYPTO_SM3
> select CRYPTO_AKCIPHER
> select CRYPTO_MANAGER
> select MPILIB
> @@ -1005,9 +1005,12 @@ config CRYPTO_SHA3
> http://keccak.noekeon.org/
>
> config CRYPTO_SM3
> + tristate
> +
> +config CRYPTO_SM3_GENERIC
> tristate "SM3 digest algorithm"
> select CRYPTO_HASH
> - select CRYPTO_LIB_SM3
> + select CRYPTO_SM3
> help
> SM3 secure hash function as defined by OSCCA GM/T 0004-2012 SM3).
> It is part of the Chinese Commercial Cryptography suite.
This patch generally looks good, but perhaps CRYPTO_SM3 and CRYPTO_SM3_GENERIC
should be merged? Having separate options provides the ability to enable the
architecture implementation without the generic implementation, so arguably it's
the right thing to do; however, that's not what the other algorithms in crypto/
do. For example, if you enable CONFIG_CRYPTO_SHA3, you get the generic SHA-3
algorithms as well as the SHA-3 library functions. I.e., there's no
CONFIG_CRYPTO_SHA3_GENERIC.
IMO, being consistent is more important than providing the ability to omit the
generic implementation.
- Eric
