| Message ID | 20190826045520.92153-1-aik@ozlabs.ru |
|---|---|
| State | Accepted |
| Headers | show |
| Series | [kernel] KVM: PPC: Fix incorrect guest-to-user-translation error handling | expand |
On Mon, 2019-08-26 at 04:55:20 UTC, Alexey Kardashevskiy wrote: > H_PUT_TCE_INDIRECT handlers receive a page with up to 512 TCEs from > a guest. Although we verify correctness of TCEs before we do anything > with the existing tables, there is a small window when a check in > kvmppc_tce_validate might pass and right after that the guest alters > the page with TCEs which can cause early exit from the handler and > leave srcu_read_lock(&vcpu->kvm->srcu) (virtual mode) or lock_rmap(rmap) > (real mode) locked. > > This fixes the bug by jumping to the common exit code with an appropriate > unlock. > > Fixes: 121f80ba68f1 ("KVM: PPC: VFIO: Add in-kernel acceleration for VFIO") > Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> Applied to powerpc topic/ppc-kvm, thanks. https://git.kernel.org/powerpc/c/4f916593be9da38c5cf0d3a5c386b57beb70f422 cheers
diff --git a/arch/powerpc/kvm/book3s_64_vio.c b/arch/powerpc/kvm/book3s_64_vio.c index e99a14798ab0..c4b606fe73eb 100644 --- a/arch/powerpc/kvm/book3s_64_vio.c +++ b/arch/powerpc/kvm/book3s_64_vio.c @@ -660,8 +660,10 @@ long kvmppc_h_put_tce_indirect(struct kvm_vcpu *vcpu, } tce = be64_to_cpu(tce); - if (kvmppc_tce_to_ua(vcpu->kvm, tce, &ua)) - return H_PARAMETER; + if (kvmppc_tce_to_ua(vcpu->kvm, tce, &ua)) { + ret = H_PARAMETER; + goto unlock_exit; + } list_for_each_entry_lockless(stit, &stt->iommu_tables, next) { ret = kvmppc_tce_iommu_map(vcpu->kvm, stt, diff --git a/arch/powerpc/kvm/book3s_64_vio_hv.c b/arch/powerpc/kvm/book3s_64_vio_hv.c index f50bbeedfc66..b4f20f13b860 100644 --- a/arch/powerpc/kvm/book3s_64_vio_hv.c +++ b/arch/powerpc/kvm/book3s_64_vio_hv.c @@ -556,8 +556,10 @@ long kvmppc_rm_h_put_tce_indirect(struct kvm_vcpu *vcpu, unsigned long tce = be64_to_cpu(((u64 *)tces)[i]); ua = 0; - if (kvmppc_rm_tce_to_ua(vcpu->kvm, tce, &ua, NULL)) - return H_PARAMETER; + if (kvmppc_rm_tce_to_ua(vcpu->kvm, tce, &ua, NULL)) { + ret = H_PARAMETER; + goto unlock_exit; + } list_for_each_entry_lockless(stit, &stt->iommu_tables, next) { ret = kvmppc_rm_tce_iommu_map(vcpu->kvm, stt,
H_PUT_TCE_INDIRECT handlers receive a page with up to 512 TCEs from a guest. Although we verify correctness of TCEs before we do anything with the existing tables, there is a small window when a check in kvmppc_tce_validate might pass and right after that the guest alters the page with TCEs which can cause early exit from the handler and leave srcu_read_lock(&vcpu->kvm->srcu) (virtual mode) or lock_rmap(rmap) (real mode) locked. This fixes the bug by jumping to the common exit code with an appropriate unlock. Fixes: 121f80ba68f1 ("KVM: PPC: VFIO: Add in-kernel acceleration for VFIO") Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> --- arch/powerpc/kvm/book3s_64_vio.c | 6 ++++-- arch/powerpc/kvm/book3s_64_vio_hv.c | 6 ++++-- 2 files changed, 8 insertions(+), 4 deletions(-)