From 8767556995adf9a10b49fb0c2098b7aeb40ee64c Mon Sep 17 00:00:00 2001 From: Peter Shier Date: Thu, 11 Oct 2018 11:46:46 -0700 Subject: [PATCH] KVM: nVMX: unconditionally cancel preemption timer in free_nested (CVE-2019-7221) commit ecec76885bcfe3294685dc363fd1273df0d5d65f upstream. Bugzilla: 1671904 There are multiple code paths where an hrtimer may have been started to emulate an L1 VMX preemption timer that can result in a call to free_nested without an intervening L2 exit where the hrtimer is normally cancelled. Unconditionally cancel in free_nested to cover all cases. Embargoed until Feb 7th 2019. Signed-off-by: Peter Shier Reported-by: Jim Mattson Reviewed-by: Jim Mattson Reported-by: Felix Wilhelm Cc: stable@kernel.org Message-Id: <20181011184646.154065-1-pshier@google.com> Signed-off-by: Paolo Bonzini Signed-off-by: Greg Kroah-Hartman --- arch/x86/kvm/vmx.c | 1 + 1 file changed, 1 insertion(+) diff --git a/arch/x86/kvm/vmx.c b/arch/x86/kvm/vmx.c index cd51f320a4600..3467783fe49c6 100644 --- a/arch/x86/kvm/vmx.c +++ b/arch/x86/kvm/vmx.c @@ -6295,6 +6295,7 @@ static void free_nested(struct vcpu_vmx *vmx) if (!vmx->nested.vmxon) return; + hrtimer_cancel(&vmx->nested.preemption_timer); vmx->nested.vmxon = false; nested_release_vmcs12(vmx); if (enable_shadow_vmcs) -- 2.47.2