Merge tag 'libcrypto-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/ebiggers/linux

Pull crypto library updates from Eric Biggers:

- Add a RISC-V optimized implementation of Poly1305.

Merge tag 'libcrypto-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/ebiggers/linux

Pull crypto library updates from Eric Biggers:

- Add a RISC-V optimized implementation of Poly1305. This code was
written by Andy Polyakov and contributed by Zhihang Shao.

- Migrate the MD5 code into lib/crypto/, and add KUnit tests for MD5.

Yes, it's still the 90s, and several kernel subsystems are still
using MD5 for legacy use cases. As long as that remains the case,
it's helpful to clean it up in the same way as I've been doing for
other algorithms.

Later, I plan to convert most of these users of MD5 to use the new
MD5 library API instead of the generic crypto API.

- Simplify the organization of the ChaCha, Poly1305, BLAKE2s, and
Curve25519 code.

Consolidate these into one module per algorithm, and centralize the
configuration and build process. This is the same reorganization that
has already been successful for SHA-1 and SHA-2.

- Remove the unused crypto_kpp API for Curve25519.

- Migrate the BLAKE2s and Curve25519 self-tests to KUnit.

- Always enable the architecture-optimized BLAKE2s code.

* tag 'libcrypto-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/ebiggers/linux: (38 commits)
crypto: md5 - Implement export_core() and import_core()
wireguard: kconfig: simplify crypto kconfig selections
lib/crypto: tests: Enable Curve25519 test when CRYPTO_SELFTESTS
lib/crypto: curve25519: Consolidate into single module
lib/crypto: curve25519: Move a couple functions out-of-line
lib/crypto: tests: Add Curve25519 benchmark
lib/crypto: tests: Migrate Curve25519 self-test to KUnit
crypto: curve25519 - Remove unused kpp support
crypto: testmgr - Remove curve25519 kpp tests
crypto: x86/curve25519 - Remove unused kpp support
crypto: powerpc/curve25519 - Remove unused kpp support
crypto: arm/curve25519 - Remove unused kpp support
crypto: hisilicon/hpre - Remove unused curve25519 kpp support
lib/crypto: tests: Add KUnit tests for BLAKE2s
lib/crypto: blake2s: Consolidate into single C translation unit
lib/crypto: blake2s: Move generic code into blake2s.c
lib/crypto: blake2s: Always enable arch-optimized BLAKE2s code
lib/crypto: blake2s: Remove obsolete self-test
lib/crypto: x86/blake2s: Reduce size of BLAKE2S_SIGMA2
lib/crypto: chacha: Consolidate into single module
...

show more ...

lib/crypto: md5: Add MD5 and HMAC-MD5 library functions

Add library functions for MD5, including HMAC support. The MD5
implementation is derived from crypto/md5.c. This closely mirrors the
corresp

lib/crypto: md5: Add MD5 and HMAC-MD5 library functions

Add library functions for MD5, including HMAC support. The MD5
implementation is derived from crypto/md5.c. This closely mirrors the
corresponding SHA-1 and SHA-2 changes.

Like SHA-1 and SHA-2, support for architecture-optimized MD5
implementations is included. I originally proposed dropping those, but
unfortunately there is an AF_ALG user of the PowerPC MD5 code
(https://lore.kernel.org/r/c4191597-341d-4fd7-bc3d-13daf7666c41@csgroup.eu/),
and dropping that code would be viewed as a performance regression. We
don't add new software algorithm implementations purely for AF_ALG, as
escalating to kernel mode merely to do calculations that could be done
in userspace is inefficient and is completely the wrong design. But
since this one already existed, it gets grandfathered in for now. An
objection was also raised to dropping the SPARC64 MD5 code because it
utilizes the CPU's direct support for MD5, although it remains unclear
that anyone is using that. Regardless, we'll keep these around for now.

Note that while MD5 is a legacy algorithm that is vulnerable to
practical collision attacks, it still has various in-kernel users that
implement legacy protocols. Switching to a simple library API, which is
the way the code should have been organized originally, will greatly
simplify their code. For example:

MD5:
drivers/md/dm-crypt.c (for lmk IV generation)
fs/nfsd/nfs4recover.c
fs/ecryptfs/
fs/smb/client/
net/{ipv4,ipv6}/ (for TCP-MD5 signatures)

HMAC-MD5:
fs/smb/client/
fs/smb/server/

(Also net/sctp/ if it continues using HMAC-MD5 for cookie generation.
However, that use case has the flexibility to upgrade to a more modern
algorithm, which I'll be proposing instead.)

As usual, the "md5" and "hmac(md5)" crypto_shash algorithms will also be
reimplemented on top of these library functions. For "hmac(md5)" this
will provide a faster, more streamlined implementation.

Link: https://lore.kernel.org/r/20250805222855.10362-2-ebiggers@kernel.org
Signed-off-by: Eric Biggers <ebiggers@kernel.org>

show more ...