If (r Commit d98f26073beb ("target/i386: kvm: add VMX migration blocker") This commit fixįixes: d98f26073beb ("target/i386: kvm: add VMX migration blocker")ġ file changed, 5 insertions(+), 2 deletions(-)ĭiff -git a/target/i386/kvm.c b/target/i386/kvm.c However, that commit missed adding deletion of the migration blocker inĬase init of vCPU failed. ` (9 subsequent siblings) 10 siblings, 1 reply 20+ messages in threadĬc: pbonzini, mtosatti, rth, ehabkost, kvm, jmattson, maran.wilson,Ĭommit d98f26073beb ("target/i386: kvm: add VMX migration blocker")Īdded migration blocker for vCPU exposed with Intel VMX because QEMUĭoesn't yet contain code to support migration of nested virtualization 16:21 ` KVM: Introduce kvm_arch_destroy_vcpu() Liran Alon 16:21 : target/i386: kvm: Add support for save and restore of nested state Liran 16:21 ` Liran Alon * target/i386: kvm: Delete VMX migration blocker on vCPU init failure.Is lacking one of the KVM caps required to migrate nested workloads. * Add migration blocker in case vCPU is exposed with VMX/SVM and kernel
* Wrapped definition of kernel integer vmstate macros with #ifdef CONFIG_LINUX. * Modify linux-headers to not have SVM stubs and use constant for vmcs12 size. * Change detection of VMX/SVM to not consider CPU vendor. * Add delete of nested migration blocker in case vCPU init fail. * Signal if vCPU is in guest-mode in hflags as original intention by Paolo. * Don't rely on CR4 or MSR_EFER to know if it is required to * Add migration blocker for vCPU exposed with AMD SVM. * Use kvm_arch_destroy_vcpu() to free nested_state.
* Converted max_nested_state_len to int instead of uint32_t. * Add cpu_pre_save() logic to convert pending exception to injected * support forward & backwards migration compatability. * Use vmstate subsections to save/restore nested state in order to properly * Use explicit fields for struct kvm_nested_state data portion. * Add support for KVM_CAP_EXCEPTION_PAYLOAD. * Add patch to fix bug when re-inject #DB to guest.
This new KVM capabilityĪllows userspace to properly distingiush between pending and injecting exceptions.ġ0th patch changes the nested virtualization migration blocker to onlyīe added when kernel lack support for one of the capabilities requiredįor correct nested migration. _u16).Ĩth patch adds support for saving and restoring nested state in order to migrateĩth patch add support for KVM_CAP_EXCEPTION_PAYLOAD. The updated struct now have explicit fields for the data portion.ħrd patch add vmstate support for saving/restoring kernel integer types (e.g. Until now it was blocked only for Intel VMX.Ħrd patch updates linux-headers to have updated struct kvm_nested_state. It was intercepted by KVM when guest is debugged.ĥth patch adds migration blocker for vCPU exposed with either Intel VMX KVM_CAP_NESTED_STATE") which was created for this purpose.ġst patch add a missed cleanup for deleting VMX migration blocker inĢst patch introduce kvm_arch_destroy_vcpu() to perform per-vCPUĭestruction logic that is arch-dependent.ģst patch is just refactoring to use symbolic constants instead of hard-codedĤst patch fixes QEMU to update DR6 when QEMU re-inject #DB to guest after IOCTLs introduced in KVM commit 8fcc4b5923af ("kvm: nVMX: Introduce This series aims to add support for QEMU to be able to migrate VMs thatĪre running nested hypervisors. ` (10 more replies) 0 siblings, 11 replies 20+ messages in threadĬc: pbonzini, mtosatti, rth, ehabkost, kvm, jmattson, maran.wilson, dgilbert 16:21 ` target/i386: kvm: Delete VMX migration blocker on vCPU init failure Liran Alon : target/i386: kvm: Add support for save and restore of nested state KVM Archive on help / color / mirror / Atom feed * : target/i386: kvm: Add support for save and restore of nested state 16:21 Liran Alon
0 kommentar(er)
