Hypercall / HSM Upcall¶
The hypercall/upcall is used to request services between the Guest VM and the hypervisor. The hypercall is from the Guest VM to hypervisor, and the upcall is from the hypervisor to the Guest VM. The hypervisor currently supports hypercall APIs for VM management, I/O request distribution, interrupt injection, PCI assignment, guest memory mapping, power management, and secure world switch.
There are some restrictions for hypercall and upcall:
Only specific VMs (currently the Service VM and the VM with trusty enabled) can invoke hypercalls. A VM that cannot invoke hypercalls will get
#UD(i.e. invalid opcode exception).
Only ring 0 hypercalls from the guest VM are handled by the hypervisor; otherwise, the hypervisor will inject
#GP(0)(i.e. generation protection exception with error code
0) to the Guest VM.
Each VM is permitted to invoke a certain subset of hypercalls. Currently a VM with trusty enabled is allowed to invoke trusty hypercalls, and the Service VM is allowed to invoke the other hypercalls. A VM that invokes an unpermitted hypercall will get the return value
-EINVAL. see Secure Hypervisor Interface for a detailed description.
The hypervisor needs to protect the critical resources such as global VM and VCPU structures for VM and VCPU management hypercalls.
Upcall is only used for the Service VM.
HV and Service VM both use the same vector (0xF3) reserved as x86 platform IPI vector for HV notification to the Service VM. This upcall is necessary whenever there is device emulation requirement to the Service VM. The upcall vector (0xF3) is injected to Service VM vCPU0. The Service VM will register the IRQ handler for vector (0xF3) and notify the I/O emulation module in the Service VM once the IRQ is triggered. View the detailed upcall process at IPI Management