kernel mode : mmu

[Martins3]

Something related to emualted:

• mapped and unmapped
  • kvm_mips_gpa_pte_to_gva_unmapped, kvm_mips_gpa_pte_to_gva_mapped
  • kvm_mips_handle_kseg0_tlb_fault <= kvm_trap_emul_gva_fault <= kvm_get_inst <= used by emulate.c
• kseg0
  • kvm_mips_handle_kseg0_tlb_fault
• emulated
  • x: kvm_trap_emul_gva_fault
    • x: kvm_get_inst
  • kvm_mips_handle_mapped_seg_tlb_fault : caller is emulate

Boring stuff

• move mmu notifier to mmu.c
• kvm_pgd_alloc
• examine transparent hugepage
• dune_arch_vcpu_put : some effort to complete the code
• kvm_mips_handle_commpage_tlb_fault : now that commpage is used for tlb, should dune remove it ?
• remove kvm_host.c
• remove dirty bit tracking in mmu.c

vcpu, ept and guest page table

• A sthread fork will create two guest page table
• A Linux fork with VM clone will create two
• A dune_enter doesn't create user space percpu
  • dune_enter create a new vcpu and ept

kvm_memory_slot

• maybe we should take advantage of it
• move memory_slot related code into one place

[Martins3]

• can vz_vm substitude kvm in mmu.c ?
• it's kvm_flush_remote_tlbs necessary ?
  • vz_vm::tlbs_dirty : why we need it ? (No, only x86 use it, in the paging_tmpl.h)
  • There is no doubt local TLB flush is obligatory
  • flush remote is necessary if ept is shared by multiple vcpu
  • Should dune support remote tlb flush
  • What's the difference between two separate ept and one sharing ept?
    • ept mapping is the same
    • does dune create two ept with fork ? (yes, but shouldn't) vmx_launch -> vmx_create_vcpu -> vmx_init_ept will create new cpu.
    • should dune create a new vcpu for every process?
      • it's possible, but need some efforts like sthread, and set a limitation on two processes where they can't run simultaneously