all 3 comments

sorted by: hot top controversial new old
[–] 14 points 2 months ago

Usama Arif found that with the three patches to just probe BMI2 CPU support only once, it yield a 71% reduction in decompression time and a 18% reduction in compression time for a crypto_acomp benchmark.

Wow

  • source
  • [–] 4 points 2 months ago (1 child)

    Is it that slow outside of a VM? Either way, this seems like a very sensible patch.

  • source
  • hideshow 1 child comment
  • [–] 4 points 2 months ago* (last edited 2 months ago)

    It's only slightly faster. cpuid is a serialising instruction, which means all the instructions before it have to complete, removing any gains from pipelining and out of order execution. However for a VM this also triggers a VM exit, where the hypervisor takes over to modify the values for the guest.

  • source
  • parent