https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=242747
--- Comment #33 from John Baldwin <j...@freebsd.org> --- I think the original question of if Epyc uses hardware AES is resolved (it does use AESNI). However, the performance issue that prompted the question is still open, which is that we are creating too many worker threads and/or not distributing the work optimally across threads when multiple GELI providers exist. -- You are receiving this mail because: You are the assignee for the bug.
