RE: [PATCH v1 11/23] KVM: x86: Add kvm_is_fred_enabled()

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



> > >> FRED is enabled when CR4.FRED = IA32_EFER.LMA = 1. Any reason to
> > >> omit the check about long mode?
> > >
> > >It won' t allow CR4.FRED to be set if not in long mode, I don't
> > >expect it at runtime.  Or you have one?
> >
> > I was thinking about a very contrived case:
> >
> > 1. the CPU enters 64-bit long mode and sets CR4.FRED 2. the CPU
> > switches out of 64-bit long mode

Actually, SDM3 Section 10.8.5 Initializing IA-32e Mode says: 64-bit mode
consistency checks fail on attempts to enable or disable IA-32e mode
while paging is enabled.  In another word, the CPU allows software to
modify IA32_EFER.LME only when CR0.PG = 0 (i.e., only in legacy mode).
Thus, attempts to do so when CR0.PG = 1 will cause #GP.

Remember FRED only works with 64-bit kernel, which can be entered only
if paging is enabled.  As such, to clear IA32_EFER.LME, OS/VMM needs to
turn off paging first.  And to turn off paging, 64-bit mode needs to be
turned off (to enter compatibility mode in ring 0 because it's not allowed
to disable paging in 64-bit mode).  Thus CR4.FRED needs be cleared first.

So CR4.FRED fully indicates whether FRED is enabled or not, which is
why FRED spec 5.0 section 9.5 says so.  But we can put more explanation
there.





[Index of Archives]     [KVM ARM]     [KVM ia64]     [KVM ppc]     [Virtualization Tools]     [Spice Development]     [Libvirt]     [Libvirt Users]     [Linux USB Devel]     [Linux Audio Users]     [Yosemite Questions]     [Linux Kernel]     [Linux SCSI]     [XFree86]

  Powered by Linux