l...@gnu.org (Ludovic Courtès) writes: > Pádraig Brady <p...@draigbrady.com> skribis: > >> The speed of md5 and sha* hashes has lagged a bit in gnulib. >> So to address that and to take advantage of the architecture >> specific assembly used in libcrypto, the attached gnulib patch >> allows projects to configure --with-openssl to use that if >> available or fall back to the existing internal routines. > > Any idea how libcrypto compares to what libgcrypt and Nettle provide?
I' afraid openssl/libcrypto is a bit faster than nettle on sha1 and md5. Current benchmark (on my office machine, intel core i5, 3.4 GHz): Algorithm mode Mbyte/s md2 update 8.32 md4 update 854.80 md5 update 577.86 openssl md5 update 721.54 sha1 update 548.50 openssl sha1 update 775.10 sha224 update 270.92 sha256 update 271.31 sha384 update 436.46 sha512 update 437.23 sha3_224 update 293.78 sha3_256 update 275.69 sha3_384 update 211.33 sha3_512 update 146.43 I haven't paid much attention to x86_64 assembly for sha1 and md5 (there is x86_64 assembly for sha1, which is a direct translation of an earlier version of the x86 assembly, and there's no md5-assembly at all for x86_64). It's faster then openssl for some other algorithms. For use in coreutils, I really think the minimalistic api in nettle should be easy to use. Regards, /Niels (nettle maintainer) -- Niels Möller. PGP-encrypted email is preferred. Keyid C0B98E26. Internet email is subject to wholesale government surveillance.