Hypervisor-Managed Linear Address Translation

Hypervisor-Managed Linear Address Translation (HLAT) is active when the “enable HLAT” VM-execution control is 1. The processor looks up the HLAT if, during a guest linear address translation, the guest linear address matches the Protected Linear Range. The lookup from guest linear addresses to the guest physical address and attributes is determined by a set of HLAT paging structures.

The guest paging structure managed by the guest OS specifies the ordinary translation of a guest linear address to the guest physical address and attributes that the guest ring-0 software has programmed, whereas HLAT specifies the alternate translation of the guest linear address to guest physical address and attributes that the Secure Kernel and VMM seek to enforce. A logical processor uses HLAT to translate guest linear addresses only when those guest linear addresses are used to access memory (both for code fetch and data load/store) and the guest linear addresses match the PLR programmed by the VMM/Secure Kernel.

HLAT specifications and functional descriptions are included in the Intel® Architecture Instruction Set Extensions Programming Reference. Available at:

https://software.intel.com/en-us/download/intel-architecture-instruction-set-extensions-programming-reference