Commit 1cfff4d9 authored by Jiří Paleček's avatar Jiří Paleček Committed by Paolo Bonzini
Browse files

kvm: Nested KVM MMUs need PAE root too 

On AMD processors, in PAE 32bit mode, nested KVM instances don't
work. The L0 host get a kernel OOPS, which is related to
arch.mmu->pae_root being NULL.

The reason for this is that when setting up nested KVM instance,
arch.mmu is set to &arch.guest_mmu (while normally, it would be
&arch.root_mmu). However, the initialization and allocation of
pae_root only creates it in root_mmu. KVM code (ie. in
mmu_alloc_shadow_roots) then accesses arch.mmu->pae_root, which is the
unallocated arch.guest_mmu->pae_root.

This fix just allocates (and frees) pae_root in both guest_mmu and
root_mmu (and also lm_root if it was allocated). The allocation is
subject to previous restrictions ie. it won't allocate anything on
64-bit and AFAIK not on Intel.

Fixes: https://bugzilla.kernel.org/show_bug.cgi?id=203923


Fixes: 14c07ad8 ("x86/kvm/mmu: introduce guest_mmu")
Signed-off-by: default avatarJiri Palecek <jpalecek@web.de>
Tested-by: default avatarJiri Palecek <jpalecek@web.de>
Signed-off-by: default avatarPaolo Bonzini <pbonzini@redhat.com>
parent c8848cee

arch/x86/kvm/mmu.c

+22 −8
Original line number Diff line number Diff line
@@ -5595,13 +5595,13 @@ slot_handle_leaf(struct kvm *kvm, struct kvm_memory_slot *memslot, 
				 PT_PAGE_TABLE_LEVEL, lock_flush_tlb);

}



static void free_mmu_pages(struct kvm_vcpu *vcpu)

static void free_mmu_pages(struct kvm_mmu *mmu)

{

	free_page((unsigned long)vcpu->arch.mmu->pae_root);

	free_page((unsigned long)vcpu->arch.mmu->lm_root);

	free_page((unsigned long)mmu->pae_root);

	free_page((unsigned long)mmu->lm_root);

}



static int alloc_mmu_pages(struct kvm_vcpu *vcpu)

static int alloc_mmu_pages(struct kvm_vcpu *vcpu, struct kvm_mmu *mmu)

{

	struct page *page;

	int i;
@@ -5622,9 +5622,9 @@ static int alloc_mmu_pages(struct kvm_vcpu *vcpu) 
	if (!page)

		return -ENOMEM;



	vcpu->arch.mmu->pae_root = page_address(page);

	mmu->pae_root = page_address(page);

	for (i = 0; i < 4; ++i)

		vcpu->arch.mmu->pae_root[i] = INVALID_PAGE;

		mmu->pae_root[i] = INVALID_PAGE;



	return 0;

}
@@ -5632,6 +5632,7 @@ static int alloc_mmu_pages(struct kvm_vcpu *vcpu) 
int kvm_mmu_create(struct kvm_vcpu *vcpu)

{

	uint i;

	int ret;



	vcpu->arch.mmu = &vcpu->arch.root_mmu;

	vcpu->arch.walk_mmu = &vcpu->arch.root_mmu;
@@ -5649,7 +5650,19 @@ int kvm_mmu_create(struct kvm_vcpu *vcpu) 
		vcpu->arch.guest_mmu.prev_roots[i] = KVM_MMU_ROOT_INFO_INVALID;



	vcpu->arch.nested_mmu.translate_gpa = translate_nested_gpa;

	return alloc_mmu_pages(vcpu);



	ret = alloc_mmu_pages(vcpu, &vcpu->arch.guest_mmu);

	if (ret)

		return ret;



	ret = alloc_mmu_pages(vcpu, &vcpu->arch.root_mmu);

	if (ret)

		goto fail_allocate_root;



	return ret;

 fail_allocate_root:

	free_mmu_pages(&vcpu->arch.guest_mmu);

	return ret;

}



static void kvm_mmu_invalidate_zap_pages_in_memslot(struct kvm *kvm,
@@ -6074,7 +6087,8 @@ unsigned long kvm_mmu_calculate_default_mmu_pages(struct kvm *kvm) 
void kvm_mmu_destroy(struct kvm_vcpu *vcpu)

{

	kvm_mmu_unload(vcpu);

	free_mmu_pages(vcpu);

	free_mmu_pages(&vcpu->arch.root_mmu);

	free_mmu_pages(&vcpu->arch.guest_mmu);

	mmu_free_memory_caches(vcpu);

}

CRA Git | Maintained and supported by SUSTech CRA and CCSE