After discussion with Avi, we decided to take this chance to extend the shared MSR framework, discard the limitation of "All CPU should have same value for the MSRs". But the limitation of "The MSRs' value shouldn't be modified after they were read" still apply. I have tested this patchset using 2.6.32. It can boot well with "rdtscp" feature(-cpu core2duo,+rdtscp) after patching, and a simple "rdtscp" userspace program can running in the guest(it's too simple that can only prove the instruction won't cause #UD, can't guarantee the result, so I think it's not necessary to get it in the unit test). In the past, this would result in guest kernel crash on vsyscall initialization (due to access MSR_TSC_AUX result in #GP fault). -- regards Yang, Sheng -- arch/x86/include/asm/kvm_host.h | 1 + arch/x86/include/asm/msr-index.h | 1 + arch/x86/include/asm/msr.h | 4 +- arch/x86/include/asm/vmx.h | 1 + arch/x86/kvm/svm.c | 6 +++ arch/x86/kvm/vmx.c | 33 ++++++++++++++++--- arch/x86/kvm/x86.c | 67 +++++++++++++++++++++++++------------- 7 files changed, 83 insertions(+), 30 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
