diff options
| author | Sheng Yang | 2010-06-30 06:25:15 +0200 |
|---|---|---|
| committer | Avi Kivity | 2010-08-01 09:47:21 +0200 |
| commit | f5f48ee15c2ee3e44cf429e34b16c6fa9b900246 (patch) | |
| tree | 75496197219d9aeedd3317fa007cc3b2e414c5da /arch/x86/kvm/svm.c | |
| parent | KVM: Document KVM specific review items (diff) | |
| download | kernel-qcow2-linux-f5f48ee15c2ee3e44cf429e34b16c6fa9b900246.tar.gz kernel-qcow2-linux-f5f48ee15c2ee3e44cf429e34b16c6fa9b900246.tar.xz kernel-qcow2-linux-f5f48ee15c2ee3e44cf429e34b16c6fa9b900246.zip | |
KVM: VMX: Execute WBINVD to keep data consistency with assigned devices
Some guest device driver may leverage the "Non-Snoop" I/O, and explicitly
WBINVD or CLFLUSH to a RAM space. Since migration may occur before WBINVD or
CLFLUSH, we need to maintain data consistency either by:
1: flushing cache (wbinvd) when the guest is scheduled out if there is no
wbinvd exit, or
2: execute wbinvd on all dirty physical CPUs when guest wbinvd exits.
Signed-off-by: Yaozu (Eddie) Dong <eddie.dong@intel.com>
Signed-off-by: Sheng Yang <sheng@linux.intel.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
Diffstat (limited to 'arch/x86/kvm/svm.c')
| -rw-r--r-- | arch/x86/kvm/svm.c | 7 |
1 files changed, 7 insertions, 0 deletions
diff --git a/arch/x86/kvm/svm.c b/arch/x86/kvm/svm.c index 587b99d37d44..56c9b6bd7655 100644 --- a/arch/x86/kvm/svm.c +++ b/arch/x86/kvm/svm.c @@ -3424,6 +3424,11 @@ static bool svm_rdtscp_supported(void) return false; } +static bool svm_has_wbinvd_exit(void) +{ + return true; +} + static void svm_fpu_deactivate(struct kvm_vcpu *vcpu) { struct vcpu_svm *svm = to_svm(vcpu); @@ -3508,6 +3513,8 @@ static struct kvm_x86_ops svm_x86_ops = { .rdtscp_supported = svm_rdtscp_supported, .set_supported_cpuid = svm_set_supported_cpuid, + + .has_wbinvd_exit = svm_has_wbinvd_exit, }; static int __init svm_init(void) |