KVM: arm64: Require in-kernel irqchip for PMU support
authorChristoffer Dall <christoffer.dall@linaro.org>
Tue, 27 Sep 2016 01:51:47 +0000 (18:51 -0700)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Sun, 16 Oct 2016 15:50:37 +0000 (17:50 +0200)
commit6b30b92d592f7da0e1a653dc951479cff43fd82b
treeecb5d5ec9e04a402e9fc074220c1ac4790b67c22
parent7398f669a4a0c912dc9951e2f914707e1ac87e53
KVM: arm64: Require in-kernel irqchip for PMU support

commit 6fe407f2d18a4f94216263f91cb7d1f08fa5887c upstream.

If userspace creates a PMU for the VCPU, but doesn't create an in-kernel
irqchip, then we end up in a nasty path where we try to take an
uninitialized spinlock, which can lead to all sorts of breakages.

Luckily, QEMU always creates the VGIC before the PMU, so we can
establish this as ABI and check for the VGIC in the PMU init stage.
This can be relaxed at a later time if we want to support PMU with a
userspace irqchip.

Cc: Shannon Zhao <shannon.zhao@linaro.org>
Acked-by: Marc Zyngier <marc.zyngier@arm.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Documentation/virtual/kvm/devices/vcpu.txt
virt/kvm/arm/pmu.c