You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
From #129 (comment) instead of the expensive STCK, the less expensive STCKF could be used (if it meets our needs, namely being not easy to predict). Additionally, performance measurements should be taken.
PRNG support (atm using rdrand/rdseed on x86) should be improved with the PRNO instruction (see #129 (comment)) on s390x (and potentially on arm & riscv using some special instructions as well).
Also, various platforms have extended hash & cipher support now (x86 has sha2 instructions, arm has multiple of them, s390x has poly1305, ...).
The text was updated successfully, but these errors were encountered:
Further improvements you mention: STCKF, PRNO, POLY1305/CHACHA20 instructions would be nice to have, but we at robur are at the moment busy with other projects (and don't have customer demand for improved s390x) support.
If you have (customer) demand for better random numbers (using PRNO) or improved ciphers, please don't hesitate to contact us.
As I mentioned in the earlier issue, IBM is interested to sponsor this work through financial bounties.
From #129 (comment) instead of the expensive STCK, the less expensive STCKF could be used (if it meets our needs, namely being not easy to predict). Additionally, performance measurements should be taken.
PRNG support (atm using rdrand/rdseed on x86) should be improved with the PRNO instruction (see #129 (comment)) on s390x (and potentially on arm & riscv using some special instructions as well).
Also, various platforms have extended hash & cipher support now (x86 has sha2 instructions, arm has multiple of them, s390x has poly1305, ...).
The text was updated successfully, but these errors were encountered: