Conversation

Notices

1. clacke (clacke@social.heldscal.la)'s status on Thursday, 04-Jan-2018 05:04:10 EST clacke
   #meltdown #spectre #lkml
   
   https://lkml.org/lkml/2018/1/3/780
   1. clacke (clacke@social.heldscal.la)'s status on Thursday, 04-Jan-2018 05:08:57 EST clacke

      in reply to
      This thread is about a rather invasive change, that makes sure no speculative indirect calls are ever made inside the kernel, IIIC basically flushing the pipeline at any indirect branch?
      1. Oneesan succubus (lain@pleroma.soykaf.com)'s status on Thursday, 04-Jan-2018 05:12:55 EST Oneesan succubus

         in reply to
         @clacke RIP performance
         1. clacke (clacke@social.heldscal.la)'s status on Thursday, 04-Jan-2018 05:18:25 EST clacke

            in reply to
            @lain In the 4.15-rc6 thread, people are sharing some benchmarks.
            
            > I can confirm, I've just run some tests on haproxy on a core i7-4790K
            > and I'm observing a performance loss of ~17%, making the connection
            > rate go down from 245k/s to 204k/s. It's indeed quite significant for
            > such use cases, even though I think it might reasonably be absorbed by
            > usual noise in most use cases.
            
            These comments are about the PTI, which is definitely going in. The other things I shared I believe to be about potential patches, which I read (caveat) as possibly making PTI unnecessary, by making the kernel code more resilient to the attack.
            
            https://lkml.org/lkml/2018/1/3/281