El 24/07/15 a les 19.36, Konrad Rzeszutek Wilk ha escrit: > On Fri, Jul 24, 2015 at 06:54:09PM +0200, Roger Pau Monné wrote: >> I have the feeling that we are over engineering this interface. IMHO we >> should only allow the user to set the control registers, efer (on amd64) >> and the GP registers. Segment selectors would be set by Xen to point to >> a flat segment suitable for the mode the guest has selected. I think >> that should be enough to get the guest OS into it's entry point, and >> then it can do whatever it wants. > > If you are doing that why not use the old interface?
Because the current structure contains quite a lot of PV fields, like the failsafe_callback, the GDT... Also the old structure is limited in a way that you can only use the x86-<bitness> format if your kernel is compiled with <bitness>, and we want to allow the vCPU to start in any mode, regardless of the mode the kernel is compiled with. >> >> If that's not suitable, then my second option would be to allow the >> guest to set the base, limit and AR bytes of the selectors, but not load >> a GDT. > > They should be able to do any of those operations as a normal HVM > guest I would think? Yes, hence I would like to limit setting anything that's not essential to boot. _______________________________________________ Xen-devel mailing list Xen-devel@lists.xen.org http://lists.xen.org/xen-devel