On Mon, 2017-12-11 at 16:46 +0000, Daniel P. Berrange wrote: > On Mon, Dec 11, 2017 at 05:40:36PM +0100, Andrea Bolognani wrote: > > Some ARM platforms, such as the original Raspberry Pi, report the > > CPU frequency in the BogoMIPS field of /proc/cpuinfo, so libvirt > > parsed that field and returned it through its API. > > > > However, not only many more boards don't report any value there, > > but several - including ARMv8-based server hardware, and even the > > more recent Raspberry Pi 3 - use this field as originally intended: > > to report the BogoMIPS value instead of the CPU frequency. > > > > Since we have no way of detecting how the field is being used, > > it's better to report no information at all rather than something > > ludicrous like "your shiny 96-core aarch64 virtualization host's > > CPUs are running at a whopping 100 MHz". > > Can we perhaps get useful freq data from sysfs instead ? I know my > x86 machines report freq there, but I'm unsure if that reporting > is x86-specific or not. The plan is to start using dmidecode(8) to retrieve the CPU frequency, since AIUI it's already doing some work to locate the information and process it and it would be quite pointless to duplicate all that in libvirt. Plus we're already using the tool for other stuff. I think that would only work on server-grade aarch64 hardware, though, because most other ARM hardware will simply not expose the information we need; even in that case, not reporting any CPU frequency is better than reporting a completely bogus value. So basically reporting the actual CPU frequency when possible is in my TODO list[1], but I believe there's value in merging this series as-is since it already improves significantly on the status quo. [1] https://bugzilla.redhat.com/show_bug.cgi?id=1206353 -- Andrea Bolognani / Red Hat / Virtualization -- libvir-list mailing list libvir-list@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/libvir-list
