event injection

• vcpu_vmx::idt_vectoring_info; 在普通模式下的

• vmcs12->idt_vectoring_info_field : 在 nested 模式下

• 在 vmcs 中间，存在 IDT-vectoring information
  • 文档:
    • 25.9.3 Information for VM Exits That Occur During Event Delivery
    • 27.5 EVENT INJECTION
    • 28.2.2 Information for VM Exits Due to Vectored Events
    • 28.2.4 Information for VM Exits During Event Delivery

中间指出，在处理一个 event delivery 的时候， 结果造成了另一个 vmexit, 那么这个 vmexit 出来的时候， 在 vmcs 中间的 idt_vectoring_info 就会存储下来当时正在 deliver 的 event，具体描述如下:

• A VMExit can occur during event-delivery
• Example: Write of exception frame to stack triggers EPT_VIOLATION
• CPU saves the event which attempted to deliver in vmcs->idt_vectoring_info
• On guest->host:
  1. KVM checks if vmcs->idt_vectoring_info valid
  2. If valid, queue injected event in struct kvm_vcpu_arch and set KVM_REQ_EVENT
• KVM_REQ_EVENT will evaluate injected event on next entry to guest

如果错误注入的时候，发生了 vmexit ，那么重新注入。 具体工作在 __vmx_complete_interrupts 中进行

机制是 event deliver 的时候出现 vmexit， 到时候从 vmexit 的位置重新进入到 deliver 的 event 对应的 handler 不就完成了， 这个操作看来是重新进行一次 handler 执行

What if VMExit occurs during event-delivery to L2?

• 需要重复注入才对吧

关键参考

• https://events19.linuxfoundation.org/wp-content/uploads/2017/12/Improving-KVM-x86-Nested-Virtualization-Liran-Alon-Oracle.pdf

• https://lore.kernel.org/kvm/20241015195227.GA18617@dev-dsk-iorlov-1b-d2eae488.eu-west-1.amazon.com/T/#m0e2931815a55992f1e3ffe3f1724f8948c7d3fdc

Currently, the situation when guest accesses MMIO during event delivery is handled differently in VMX and SVM: on VMX KVM returns internal error with suberror = KVM_INTERNAL_ERROR_DELIVERY_EV, when SVM simply goes into infinite loop trying to deliver an event again and again.

Such a situation could happen when the exception occurs with guest IDTR (or GDTR) descriptor base pointing to an MMIO address.

从这里的描述，更加处理中断的时候，遇到了错误

• https://bugzilla.kernel.org/show_bug.cgi?id=218267

其实也是符合预期的

时钟中断，触发 ept violation ，中断的处理过程，一些物理地址不存在，所以，

因为 external interrupt

本站所有文章转发 CSDN 将按侵权追究法律责任，其它情况随意。