Hello folks, In this series, I present a couple of fixes for kvmclock. In patch 1, a guest-side fix is proposed for a problem that is biting us for quite a while now. the tsc inside VMs does not seem to be that good, (up to now, only single-socket nehalems were stable enough), and we're seeing small (but nevertheless wrong) time warps inside SMP guests. I am proposing the fix to reside on common code in pvclock.c, but would be good to hear Jeremy on this.
On the other 3 patches, I change kvmclock MSR numbers in a compatible fashion. Both MSR sets will be supported for a while. Patch 5 adds documentation about kvmclock, which to date, we lacked. Glauber Costa (5): Add a global synchronization point for pvclock change msr numbers for kvmclock Try using new kvm clock msrs export new cpuid KVM_CAP add documentation about kvmclock Documentation/kvm/kvmclock.txt | 138 +++++++++++++++++++++++++++++++++++++++ arch/x86/include/asm/kvm_para.h | 12 +++- arch/x86/kernel/kvmclock.c | 68 +++++++++++++------ arch/x86/kernel/pvclock.c | 23 +++++++ arch/x86/kvm/x86.c | 13 ++++- include/linux/kvm.h | 1 + 6 files changed, 231 insertions(+), 24 deletions(-) create mode 100644 Documentation/kvm/kvmclock.txt -- To unsubscribe from this list: send the line "unsubscribe kvm" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
