Messages in this thread | | | From | (Eric W. Biederman) | Date | Thu, 10 Jul 2008 13:25:33 -0700 | Subject | Re: [RFC 00/15] x86_64: Optimize percpu accesses |
| |
Jeremy Fitzhardinge <jeremy@goop.org> writes:
> Secondly, I depend on it. I register a percpu structure with Xen to share > per-vcpu specific information (interrupt mask, time info, runstate stats, etc).
Note. That I expect at least something like this is interesting in the context per cpu device queues. However except possibly for Xen that implies allocating DMA addressable memory and going through that API, which will keep device drivers from using per cpu memory that way even if the allocating something for each cpu?
Eric
| |