Re: [RFC PATCH 02/23] kvm: Switch KVM_CAP_READONLY_MEM to a per-VM ioctl()

[Isaku Yamahata]

On Tue, Feb 16, 2021 at 08:56:45AM +0100,
Philippe Mathieu-Daudé <philmd@redhat.com> wrote:

> Hi Isaku,
> 
> On 2/16/21 3:12 AM, Isaku Yamahata wrote:
> > Switch to making a VM ioctl() call for KVM_CAP_READONLY_MEM, which may
> > be conditional on VM type in recent versions of KVM, e.g. when TDX is
> > supported.
> > 
> > Signed-off-by: Isaku Yamahata <isaku.yamahata@intel.com>
> > ---
> >  accel/kvm/kvm-all.c | 2 +-
> >  1 file changed, 1 insertion(+), 1 deletion(-)
> > 
> > diff --git a/accel/kvm/kvm-all.c b/accel/kvm/kvm-all.c
> > index 47516913b7..351c25a5cb 100644
> > --- a/accel/kvm/kvm-all.c
> > +++ b/accel/kvm/kvm-all.c
> > @@ -2164,7 +2164,7 @@ static int kvm_init(MachineState *ms)
> >      }
> >  
> >      kvm_readonly_mem_allowed =
> > -        (kvm_check_extension(s, KVM_CAP_READONLY_MEM) > 0);
> > +        (kvm_vm_check_extension(s, KVM_CAP_READONLY_MEM) > 0);
> 
> Can this check with "recent KVM" be a problem with older ones?
> 
> Maybe for backward compatibility we need:
> 
>           = (kvm_vm_check_extension(s, KVM_CAP_READONLY_MEM) > 0) ||
>             (kvm_check_extension(s, KVM_CAP_READONLY_MEM) > 0);

Agreed. That's safer and it's difficult to check the very old version of kenel
and non-x86 arch.

Thanks,

-- 
Isaku Yamahata <isaku.yamahata@gmail.com>