On Thu, 14 Jul 2022 07:37:25 +0100,
Reiji Watanabe <reijiw@xxxxxxxxxx> wrote:
>
> Hi Marc,
>
> On Wed, Jul 6, 2022 at 10:05 AM Marc Zyngier <maz@xxxxxxxxxx> wrote:
> >
> > We carry a legacy interface to set the base addresses for GICv2.
> > As this is currently plumbed into the same handling code as
> > the modern interface, it limits the evolution we can make there.
> >
> > Add a helper dedicated to this handling, with a view of maybe
> > removing this in the future.
> >
> > Signed-off-by: Marc Zyngier <maz@xxxxxxxxxx>
> > ---
> > arch/arm64/kvm/arm.c | 11 ++-------
> > arch/arm64/kvm/vgic/vgic-kvm-device.c | 32 +++++++++++++++++++++++++++
> > include/kvm/arm_vgic.h | 1 +
> > 3 files changed, 35 insertions(+), 9 deletions(-)
> >
> > diff --git a/arch/arm64/kvm/arm.c b/arch/arm64/kvm/arm.c
> > index 83a7f61354d3..bf39570c0aef 100644
> > --- a/arch/arm64/kvm/arm.c
> > +++ b/arch/arm64/kvm/arm.c
> > @@ -1414,18 +1414,11 @@ void kvm_arch_flush_remote_tlbs_memslot(struct kvm *kvm,
> > static int kvm_vm_ioctl_set_device_addr(struct kvm *kvm,
> > struct kvm_arm_device_addr *dev_addr)
> > {
> > - unsigned long dev_id, type;
> > -
> > - dev_id = (dev_addr->id & KVM_ARM_DEVICE_ID_MASK) >>
> > - KVM_ARM_DEVICE_ID_SHIFT;
> > - type = (dev_addr->id & KVM_ARM_DEVICE_TYPE_MASK) >>
> > - KVM_ARM_DEVICE_TYPE_SHIFT;
> > -
> > - switch (dev_id) {
> > + switch (FIELD_GET(KVM_ARM_DEVICE_ID_MASK, dev_addr->id)) {
> > case KVM_ARM_DEVICE_VGIC_V2:
> > if (!vgic_present)
> > return -ENXIO;
> > - return kvm_vgic_addr(kvm, type, &dev_addr->addr, true);
> > + return kvm_set_legacy_vgic_v2_addr(kvm, dev_addr);
> > default:
> > return -ENODEV;
> > }
> > diff --git a/arch/arm64/kvm/vgic/vgic-kvm-device.c b/arch/arm64/kvm/vgic/vgic-kvm-device.c
> > index fbbd0338c782..0dfd277b9058 100644
> > --- a/arch/arm64/kvm/vgic/vgic-kvm-device.c
> > +++ b/arch/arm64/kvm/vgic/vgic-kvm-device.c
> > @@ -41,6 +41,38 @@ static int vgic_check_type(struct kvm *kvm, int type_needed)
> > return 0;
> > }
> >
> > +int kvm_set_legacy_vgic_v2_addr(struct kvm *kvm, struct kvm_arm_device_addr *dev_addr)
> > +{
> > + struct vgic_dist *vgic = &kvm->arch.vgic;
> > + int r;
> > +
> > + mutex_lock(&kvm->lock);
> > + switch (FIELD_GET(KVM_ARM_DEVICE_ID_MASK, dev_addr->id)) {
>
> Shouldn't this be KVM_ARM_DEVICE_TYPE_MASK (not KVM_ARM_DEVICE_ID_MASK) ?
Damn, you just ruined my attempt at deprecating this API ;-). More
seriously, thanks for catching this one!
M.
--
Without deviation from the norm, progress is not possible.
