There are a couple of fairly severe problems with NMI on AMD, both triggered with nmi_watchdog=1 in the guest and kvm ftrace in the host. One of the bug leads to guest userspace crashes via spurious setting of EFLAGS.TF, while the other leads to guest kernel hangs looping on the NMI handler's IRET instruction. I believe ftrace only affects timing here, and is not a real requirement to reproduce the bug. See https://bugzilla.redhat.com/show_bug.cgi?id=612436 for the original report. I will try to write unit tests for both issues. Avi Kivity (2): KVM: Fix race between nmi injection and enabling nmi window KVM: SVM: check for progress after IRET interception arch/x86/kvm/svm.c | 10 +++++++++- arch/x86/kvm/x86.c | 4 +++- include/linux/kvm_host.h | 1 + 3 files changed, 13 insertions(+), 2 deletions(-) -- To unsubscribe from this list: send the line "unsubscribe kvm" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
