From: Yang Zhang <yang.z.zhang@xxxxxxxxx>
APIC virtualization is a new feature which can eliminate most of VM exit
when vcpu handle a interrupt:
APIC register virtualization:
APIC read access doesn't cause APIC-access VM exits.
APIC write becomes trap-like.
Virtual interrupt delivery:
Virtual interrupt delivery avoids KVM to inject vAPIC interrupts
manually, which is fully taken care of by the hardware.
The first patch adds APIC register virtualization supporting.
The second patch adds virtual x2apic mode supporting since it is required by APICv when guest uses msr based way to access APIC.
The third patch adds virtual interrupt delivery supporting.
Please refer to Intel SDM volume 3, chapter 29 for more details.
Changes v12 to v13:
* Remove unnecessary check when set virtualized apic access
* Use vm_need_tpr_shadow() instead read vmcs to check tpr.
* Check irqchip_in_kernel when set msr bitmap.
* Correct comment format
* Remove unnecessary callback when set eoi exit bitmap.
* Disable vid when irqchip is in userspace.
* Rename vmx_vcpu_has_apicv to vmx_vm_has_apicv.
* Rebased on top of KVM upstream
Changes v11 to v12:
* Check irqchip in kernel when enabling apicv, if using userspace irq chip,
apicv cannot be used and must be disabled.
* Rename some fucntion to more descriptive name.
* Move ioapic entry pase logic to lapic.c
* Rebased on top of KVM upstream
Changes v10 to v11:
* Use two new msr bitmaps for guest that enabling x2apic mode:
Since msr bitmap is shared by all guests, it will break guest that
not using x2apic when updating the global msr bitmap. To solve this,
we use two new msr bitmap for guest which using x2apic.
Changes v9 to v10:
* Enable virtualize x2apic mode when guest is using x2apic and apicv:
There is no point to enable x2apic mode when apicv is disabled.
* Grep ioapic_lock when traversing ioapic entry to set eoi exit bitmap
* Rebased on top of KVM upstream
Changes v8 to v9:
* Update eoi exit bitmap by vcpu itself.
* Enable virtualize x2apic mode when guest is using x2apic.
* Rebase on top of KVM upstream
Changes v7 to v8:
* According Marcelo's suggestion, add comments for irr_pending and isr_count,
since the two valiables have different meaning when using apicv.
* Set highest bit in vISR to SVI after migation.
* Use spinlock to access eoi exit bitmap synchronously.
* Enable virtualize x2apic mode when guest is using x2apic
* Rebased on top of KVM upstream.
Yang Zhang (3):
x86, apicv: add APICv register virtualization support
x86, apicv: add virtual x2apic support
x86, apicv: add virtual interrupt delivery support
arch/ia64/kvm/lapic.h | 6 +
arch/x86/include/asm/kvm_host.h | 6 +
arch/x86/include/asm/vmx.h | 14 ++
arch/x86/kvm/irq.c | 56 ++++++-
arch/x86/kvm/lapic.c | 140 +++++++++++++----
arch/x86/kvm/lapic.h | 34 ++++
arch/x86/kvm/svm.c | 24 +++
arch/x86/kvm/vmx.c | 327 ++++++++++++++++++++++++++++++++++++---
arch/x86/kvm/x86.c | 23 +++-
include/linux/kvm_host.h | 3 +
virt/kvm/ioapic.c | 39 +++++
virt/kvm/ioapic.h | 4 +
virt/kvm/irq_comm.c | 25 +++
virt/kvm/kvm_main.c | 5 +
14 files changed, 647 insertions(+), 59 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
