KVM: x86/mmu: Initialize kvm_page_fault's pfn and hva to error values
authorSean Christopherson <seanjc@google.com>
Wed, 28 Feb 2024 02:41:46 +0000 (18:41 -0800)
committerPaolo Bonzini <pbonzini@redhat.com>
Tue, 7 May 2024 15:59:23 +0000 (11:59 -0400)
commitf3310e622f77979f13a36f6bfaf0252eecc7b9f6
tree76d25631b2ab51ee9adf8ccf58f3f55727275b6f
parent36d4492765fe74d69f91b0f2dae7340c03665649
KVM: x86/mmu: Initialize kvm_page_fault's pfn and hva to error values

Explicitly set "pfn" and "hva" to error values in kvm_mmu_do_page_fault()
to harden KVM against using "uninitialized" values.  In quotes because the
fields are actually zero-initialized, and zero is a legal value for both
page frame numbers and virtual addresses.  E.g. failure to set "pfn" prior
to creating an SPTE could result in KVM pointing at physical address '0',
which is far less desirable than KVM generating a SPTE with reserved PA
bits set and thus effectively killing the VM.

Signed-off-by: Sean Christopherson <seanjc@google.com>
Reviewed-by: Kai Huang <kai.huang@intel.com>
Message-ID: <20240228024147.41573-16-seanjc@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
arch/x86/kvm/mmu/mmu_internal.h