Support creating the ARM VGIC device through the KVM_CREATE_DEVICE
ioctl, which can then later be leveraged to use the
KVM_{GET/SET}_DEVICE_ATTR, which is useful both for setting addresses in
a more generic API than the ARM-specific one and is useful for
save/restore of VGIC state.

Adds KVM_CAP_DEVICE_CTRL to ARM capabilities.

Note that we change the check for creating a VGIC from bailing out if
any VCPUs were created, to bailing out if any VCPUs were ever run.  This
is an important distinction that shouldn't break anything, but allows
creating the VGIC after the VCPUs have been created.
Acked-by: NMarc Zyngier <marc.zyngier@arm.com>
Signed-off-by: NChristoffer Dall <christoffer.dall@linaro.org>

Rework the VGIC initialization slightly to allow initialization of the
vgic cpu-specific state even if the irqchip (the VGIC) hasn't been
created by user space yet.  This is safe, because the vgic data
structures are already allocated when the CPU is allocated if VGIC
support is compiled into the kernel.  Further, the init process does not
depend on any other information and the sacrifice is a slight
performance degradation for creating VMs in the no-VGIC case.

The reason is that the new device control API doesn't mandate creating
the VGIC before creating the VCPU and it is unreasonable to require user
space to create the VGIC before creating the VCPUs.

At the same time move the irqchip_in_kernel check out of
kvm_vcpu_first_run_init and into the init function to make the per-vcpu
and global init functions symmetric and add comments on the exported
functions making it a bit easier to understand the init flow by only
looking at vgic.c.
Acked-by: NMarc Zyngier <marc.zyngier@arm.com>
Signed-off-by: NChristoffer Dall <christoffer.dall@linaro.org>

Initialize the cntvoff at kvm_init_vm time, not before running the VCPUs
at the first time because that will overwrite any potentially restored
values from user space.

Cc: Andre Przywara <andre.przywara@linaro.org>
Acked-by: NMarc Zynger <marc.zyngier@arm.com>
Signed-off-by: NChristoffer Dall <christoffer.dall@linaro.org>

For bytemaps each IRQ field is 1 byte wide, so we pack 4 irq fields in
one word and since there are 32 private (per cpu) irqs, we have 8
private u32 fields on the vgic_bytemap struct.  We shift the offset from
the base of the register group right by 2, giving us the word index
instead of the field index.  But then there are 8 private words, not 4,
which is also why we subtract 8 words from the offset of the shared
words.
Signed-off-by: NChristoffer Dall <christoffer.dall@linaro.org>
Signed-off-by: NMarc Zyngier <marc.zyngier@arm.com>
Signed-off-by: NGleb Natapov <gleb@redhat.com>

All the code in handle_mmio_cfg_reg() assumes the offset has
been shifted right to accomodate for the 2:1 bit compression,
but this is only done when getting the register address.

Shift the offset early so the code works mostly unchanged.
Reported-by: NZhaobo (Bob, ERC) <zhaobo@huawei.com>
Signed-off-by: NMarc Zyngier <marc.zyngier@arm.com>
Signed-off-by: NGleb Natapov <gleb@redhat.com>

vgic_get_target_reg is quite complicated, for no good reason.
Actually, it is fairly easy to write it in a much more efficient
way by using the target CPU array instead of the bitmap.
Signed-off-by: NMarc Zyngier <marc.zyngier@arm.com>
Signed-off-by: NGleb Natapov <gleb@redhat.com>

As KVM/arm64 is looming on the horizon, it makes sense to move some
of the common code to a single location in order to reduce duplication.

The code could live anywhere. Actually, most of KVM is already built
with a bunch of ugly ../../.. hacks in the various Makefiles, so we're
not exactly talking about style here. But maybe it is time to start
moving into a less ugly direction.

The include files must be in a "public" location, as they are accessed
from non-KVM files (arch/arm/kernel/asm-offsets.c).

For this purpose, introduce two new locations:
- virt/kvm/arm/ : x86 and ia64 already share the ioapic code in
  virt/kvm, so this could be seen as a (very ugly) precedent.
- include/kvm/  : there is already an include/xen, and while the
  intent is slightly different, this seems as good a location as
  any

Eventually, we should probably have independant Makefiles at every
levels (just like everywhere else in the kernel), but this is just
the first step.
Signed-off-by: NMarc Zyngier <marc.zyngier@arm.com>
Signed-off-by: NGleb Natapov <gleb@redhat.com>

The virtual GIC is supposed to be 4kB aligned. On a 64kB page
system, comparing the alignment to PAGE_SIZE is wrong.

Use SZ_4K instead.
Signed-off-by: NMarc Zyngier <marc.zyngier@arm.com>

Now that the maintenance interrupt handling is actually out of the
handler itself, the code becomes quite racy as we can get preempted
while we process the state.

Wrapping this code around the distributor lock ensures that we're not
preempted and relatively race-free.
Signed-off-by: NMarc Zyngier <marc.zyngier@arm.com>

The VGIC doesn't guarantee that an EOIed LR that has been configured
to generate a maintenance interrupt will appear as empty.

While the code recovers from this situation, it is better to clean
the LR and flag it as empty so it can be quickly recycled.
Signed-off-by: NMarc Zyngier <marc.zyngier@arm.com>

It is now possible to select CONFIG_KVM_ARM_TIMER to enable the
KVM architected timer support.
Reviewed-by: NWill Deacon <will.deacon@arm.com>
Signed-off-by: NChristoffer Dall <c.dall@virtualopensystems.com>
Signed-off-by: NMarc Zyngier <marc.zyngier@arm.com>

Add the init code for the hypervisor, the virtual machine, and
the virtual CPUs.

An interrupt handler is also wired to allow the VGIC maintenance
interrupts, used to deal with level triggered interrupts and LR
underflows.

A CPU hotplug notifier is registered to disable/enable the interrupt
as requested.
Reviewed-by: NWill Deacon <will.deacon@arm.com>
Signed-off-by: NChristoffer Dall <c.dall@virtualopensystems.com>
Signed-off-by: NMarc Zyngier <marc.zyngier@arm.com>

Plug the interrupt injection code. Interrupts can now be generated
from user space.
Reviewed-by: NWill Deacon <will.deacon@arm.com>
Signed-off-by: NChristoffer Dall <c.dall@virtualopensystems.com>
Signed-off-by: NMarc Zyngier <marc.zyngier@arm.com>

An interrupt may have been disabled after being made pending on the
CPU interface (the classic case is a timer running while we're
rebooting the guest - the interrupt would kick as soon as the CPU
interface gets enabled, with deadly consequences).

The solution is to examine already active LRs, and check the
interrupt is still enabled. If not, just retire it.
Reviewed-by: NWill Deacon <will.deacon@arm.com>
Signed-off-by: NChristoffer Dall <c.dall@virtualopensystems.com>
Signed-off-by: NMarc Zyngier <marc.zyngier@arm.com>

Add VGIC virtual CPU interface code, picking pending interrupts
from the distributor and stashing them in the VGIC control interface
list registers.
Reviewed-by: NWill Deacon <will.deacon@arm.com>
Signed-off-by: NChristoffer Dall <c.dall@virtualopensystems.com>
Signed-off-by: NMarc Zyngier <marc.zyngier@arm.com>

Add the GIC distributor emulation code. A number of the GIC features
are simply ignored as they are not required to boot a Linux guest.
Reviewed-by: NWill Deacon <will.deacon@arm.com>
Signed-off-by: NChristoffer Dall <c.dall@virtualopensystems.com>
Signed-off-by: NMarc Zyngier <marc.zyngier@arm.com>

User space defines the model to emulate to a guest and should therefore
decide which addresses are used for both the virtual CPU interface
directly mapped in the guest physical address space and for the emulated
distributor interface, which is mapped in software by the in-kernel VGIC
support.
Reviewed-by: NWill Deacon <will.deacon@arm.com>
Signed-off-by: NChristoffer Dall <c.dall@virtualopensystems.com>
Signed-off-by: NMarc Zyngier <marc.zyngier@arm.com>

Wire the basic framework code for VGIC support and the initial in-kernel
MMIO support code for the VGIC, used for the distributor emulation.
Reviewed-by: NWill Deacon <will.deacon@arm.com>
Signed-off-by: NChristoffer Dall <c.dall@virtualopensystems.com>
Signed-off-by: NMarc Zyngier <marc.zyngier@arm.com>