From: "Lorenzo Stoakes (ARM)" <ljs@kernel.org> mainline inclusion from mainline-v7.4-rc1 commit b477bcdfa0342145084948a079b606e058066d80 category: feature bugzilla: https://atomgit.com/openeuler/kernel/issues/10374 CVE: NA Reference: https://github.com/torvalds/linux/commit/b477bcdfa0342145084948a079b606e0580... -------------------------------- commit b477bcdfa0342145084948a079b606e058066d80 upstream kvm_handle_guest_abort() establishes a kvm_s2_fault_desc data structure, s2fd, to store and propagate state to either pkvm_mem_abort(), gmem_abort() or user_mem_abort() handlers. Each of these, however, examines the state of the stage 2 MMU via vcpu->arch.hw_mmu. Introduce an s2fd->mmu field to abstract this and propagate it to callers. Similar to adding the esr field, this allows injection of synthetic faults with the ultimate intention of implementing stage 2 page table pre-faulting. No functional change intended. [Backport note] Adapted for OLK-6.6: - The stage-2 MMU is propagated through the s2fd->mmu field added by this patch. - kvm_handle_guest_abort() caches vcpu->arch.hw_mmu in a local mmu variable and uses it for the IPA-limit check and the canonical page-table selection, matching the upstream intent. - The PVM shadow-stage-1 page-table selection is left unchanged. Reviewed-by: Oliver Upton <oupton@kernel.org> Signed-off-by: Lorenzo Stoakes (ARM) <ljs@kernel.org> Reviewed-by: Fuad Tabba <fuad.tabba@linux.dev> Tested-by: Fuad Tabba <fuad.tabba@linux.dev> Link: https://patch.msgid.link/20260923-kvm-arm-prefault-v4-5-d4b0b4dfa8c3@kernel.... Signed-off-by: Marc Zyngier <maz@kernel.org> Signed-off-by: Yanan Wang <wangyanan55@huawei.com> --- arch/arm64/kvm/mmu.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) diff --git a/arch/arm64/kvm/mmu.c b/arch/arm64/kvm/mmu.c index eab7ebd0a1d2..5279399d3120 100644 --- a/arch/arm64/kvm/mmu.c +++ b/arch/arm64/kvm/mmu.c @@ -1603,6 +1603,7 @@ struct kvm_s2_fault_desc { struct kvm_memory_slot *memslot; unsigned long hva; unsigned long esr; + struct kvm_s2_mmu *mmu; }; static int user_mem_abort(const struct kvm_s2_fault_desc *s2fd, @@ -1615,6 +1616,7 @@ static int user_mem_abort(const struct kvm_s2_fault_desc *s2fd, struct kvm_memory_slot *memslot = s2fd->memslot; unsigned long hva = s2fd->hva; unsigned long esr = s2fd->esr; + struct kvm_s2_mmu *mmu = s2fd->mmu; int ret = 0; bool write_fault, writable, force_pte = false; bool exec_fault, mte_allowed; @@ -1805,7 +1807,7 @@ static int user_mem_abort(const struct kvm_s2_fault_desc *s2fd, read_lock(&kvm->mmu_lock); pgt = result ? (kernel_shadow ? vcpu_to_pvm(vcpu)->skern_mmu->mmu.pgt : - vcpu_to_pvm(vcpu)->suser_mmu->mmu.pgt) : vcpu->arch.hw_mmu->pgt; + vcpu_to_pvm(vcpu)->suser_mmu->mmu.pgt) : mmu->pgt; if (mmu_invalidate_retry(kvm, mmu_seq)) goto out_unlock; if (result && mmu_invalidate_retry(kvm, walk_seq)) @@ -1963,6 +1965,7 @@ int kvm_handle_guest_abort(struct kvm_vcpu *vcpu) unsigned long shadow_seq = 0; unsigned long walk_seq = 0; unsigned long esr = kvm_vcpu_get_esr(vcpu); + struct kvm_s2_mmu *mmu = vcpu->arch.hw_mmu; fault_status = esr & ESR_ELx_FSC_TYPE; @@ -1989,7 +1992,7 @@ int kvm_handle_guest_abort(struct kvm_vcpu *vcpu) } /* Falls between the IPA range and the PARange? */ - if (fault_ipa >= BIT_ULL(vcpu->arch.hw_mmu->pgt->ia_bits)) { + if (fault_ipa >= BIT_ULL(mmu->pgt->ia_bits)) { fault_ipa |= kvm_vcpu_get_hfar(vcpu) & GENMASK(11, 0); if (is_iabt) @@ -2144,6 +2147,7 @@ int kvm_handle_guest_abort(struct kvm_vcpu *vcpu) .memslot = memslot, .hva = hva, .esr = esr, + .mmu = mmu, }; ret = user_mem_abort(&s2fd, result, fault_status, shadow_seq, walk_seq); -- 2.43.0