KVM: Take vcpu->mutex outside vcpu_load
authorChristoffer Dall <christoffer.dall@linaro.org>
Mon, 4 Dec 2017 20:35:23 +0000 (21:35 +0100)
committerPaolo Bonzini <pbonzini@redhat.com>
Thu, 14 Dec 2017 08:26:49 +0000 (09:26 +0100)
commitec7660ccdd2b71d8c7f0243f8590253713e9b75d
treec745134c653e1f9a1d08decdd179f4fd742dd024
parent8eb73e2d410f00d383023fe41c0c25c6195b7389
KVM: Take vcpu->mutex outside vcpu_load

As we're about to call vcpu_load() from architecture-specific
implementations of the KVM vcpu ioctls, but yet we access data
structures protected by the vcpu->mutex in the generic code, factor
this logic out from vcpu_load().

x86 is the only architecture which calls vcpu_load() outside of the main
vcpu ioctl function, and these calls will no longer take the vcpu mutex
following this patch.  However, with the exception of
kvm_arch_vcpu_postcreate (see below), the callers are either in the
creation or destruction path of the VCPU, which means there cannot be
any concurrent access to the data structure, because the file descriptor
is not yet accessible, or is already gone.

kvm_arch_vcpu_postcreate makes the newly created vcpu potentially
accessible by other in-kernel threads through the kvm->vcpus array, and
we therefore take the vcpu mutex in this case directly.

Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
Reviewed-by: Cornelia Huck <cohuck@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
arch/x86/kvm/vmx.c
arch/x86/kvm/x86.c
include/linux/kvm_host.h
virt/kvm/kvm_main.c