[Vm-dev] Interesting discussion about a Intel Skylake performance regression with the PAUSE instruction
Marcel Taeumel
marcel.taeumel at hpi.de
Sun Jun 14 10:34:50 UTC 2020
Hmm... does the OSVM do any spin-wait tricks to avoid context switches?
Best,
Marcel
Am 13.06.2020 09:18:26 schrieb Bruce O'Neel <bruce.oneel at pckswarms.ch>:
Skylake takes PAUSE about 10 times more seriously then did previous Intel CPUs. Surprises happen..
[https://aloiskraus.wordpress.com/2018/06/16/why-skylakex-cpus-are-sometimes-50-slower-how-intel-has-broken-existing-code/](https://aloiskraus.wordpress.com/2018/06/16/why-skylakex-cpus-are-sometimes-50-slower-how-intel-has-broken-existing-code/)
Skylake takes PAUSE about 10 times more seriously then did previous Intel CPUs. Surprises happen..
https://aloiskraus.wordpress.com/2018/06/16/why-skylakex-cpus-are-sometimes-50-slower-how-intel-has-broken-existing-code/ [https://aloiskraus.wordpress.com/2018/06/16/why-skylakex-cpus-are-sometimes-50-slower-how-intel-has-broken-existing-code/]
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.squeakfoundation.org/pipermail/vm-dev/attachments/20200614/4f4b1b2b/attachment.html>
More information about the Vm-dev
mailing list