Message ID | 20240725175232.337266-1-mlevitsk@redhat.com (mailing list archive) |
---|---|
Headers | show |
Series | Fix for a very old KVM bug in the segment cache | expand |
У чт, 2024-07-25 у 13:52 -0400, Maxim Levitsky пише: > Hi, > > Recently, while trying to understand why the pmu_counters_test > selftest sometimes fails when run nested I stumbled > upon a very interesting and old bug: > > It turns out that KVM caches guest segment state, > but this cache doesn't have any protection against concurrent use. > > This usually works because the cache is per vcpu, and should > only be accessed by vCPU thread, however there is an exception: > > If the full preemption is enabled in the host kernel, > it is possible that vCPU thread will be preempted, for > example during the vmx_vcpu_reset. > > vmx_vcpu_reset resets the segment cache bitmask and then initializes > the segments in the vmcs, however if the vcpus is preempted in the > middle of this code, the kvm_arch_vcpu_put is called which > reads SS's AR bytes to determine if the vCPU is in the kernel mode, > which caches the old value. > > Later vmx_vcpu_reset will set the SS's AR field to the correct value > in vmcs but the cache still contains an invalid value which > can later for example leak via KVM_GET_SREGS and such. > > In particular, kvm selftests will do KVM_GET_SREGS, > and then KVM_SET_SREGS, with a broken SS's AR field passed as is, > which will lead to vm entry failure. > > This issue is not a nested issue, and actually I was able > to reproduce it on bare metal, but due to timing it happens > much more often nested. The only requirement for this to happen > is to have full preemption enabled in the kernel which runs the selftest. > > pmu_counters_test reproduces this issue well, because it creates > lots of short lived VMs, but the issue as was noted > about is not related to pmu. > > To paritally fix this issue, call vmx_segment_cache_clear > after we done with segment register setup in vmx_vcpu_reset. > > V2: incorporated Paolo's suggestion of having > vmx_write_segment_cache_start/end functions (thanks!) > > V3: reverted to a partial fix. > > Best regards, > Maxim Levitsky > > Maxim Levitsky (2): > KVM: nVMX: use vmx_segment_cache_clear > VMX: reset the segment cache after segment initialization in > vmx_vcpu_reset > > arch/x86/kvm/vmx/nested.c | 3 ++- > arch/x86/kvm/vmx/vmx.c | 10 +++------- > arch/x86/kvm/vmx/vmx.h | 5 +++++ > 3 files changed, 10 insertions(+), 8 deletions(-) > > -- > 2.26.3 > > Any update? Best regards, Maxim Levitsky
On Fri, Aug 02, 2024, mlevitsk@redhat.com wrote:
> Any update?
I am planning on jumpin into review/merge mode on Monday.
On Thu, 25 Jul 2024 13:52:30 -0400, Maxim Levitsky wrote: > Recently, while trying to understand why the pmu_counters_test > selftest sometimes fails when run nested I stumbled > upon a very interesting and old bug: > > It turns out that KVM caches guest segment state, > but this cache doesn't have any protection against concurrent use. > > [...] Applied patch 1 to kvm-x86 vmx, thanks! [1/2] KVM: nVMX: use vmx_segment_cache_clear https://github.com/kvm-x86/linux/commit/41ab0d59faa9 -- https://github.com/kvm-x86/linux/tree/next