Non-x86 folks, as with v3, patches 1 and 2 are likely the only thing of interest here. They make kvm_arch_vcpu_async_ioctl() mandatory and then rename it to kvm_arch_vcpu_unlocked_ioctl().
As for the x86 side... Clean up the TDX post-populate paths (and many tangentially related paths) to address locking issues between gmem and TDX's post-populate hook[*], and within KVM itself (KVM doesn't ensure full mutual exclusivity between paths that for all intents and purposes the TDX-Module requires to be serialized). I apologize if I missed any trailers or feedback, I think I got everything... [*] http://lore.kernel.org/all/[email protected] v4: - Collect reviews/acks. - Add a lockdep assertion in kvm_tdp_mmu_map_private_pfn(). [Yan] - Wrap kvm_tdp_mmu_map_private_pfn() with CONFIG_KVM_GUEST_MEMFD=y. [test bot] - Improve (or add) comments. [Kai, and probably others] - s/spte/mirror_spte to make it clear what's being passed in - Update set_external_spte() to take @mirror_spte as well. [Yan] - Move the KVM_BUG_ON() on tdh_mr_extend() failure to the end. [Rick] - Take "all" the locks in tdx_vm_ioctl(). [Kai] - WARN if KVM attempts to map SPTEs into an invalid root. [Yan] - Use tdx_flush_vp_on_cpu() instead of tdx_disassociate_vp() when freeing a vCPU in VCPU_TD_STATE_UNINITIALIZED state. [Yan] v3: - https://lore.kernel.org/all/[email protected] - Collect more reviews. - Add the async_ioctl() => unlocked_ioctl() patches, and use the "unlocked" variant in the TDX vCPU sub-ioctls so they can take kvm->lock outside of vcpu->mutex. - Add a patch to document that vcpu->mutex is taken *outside* kvm->slots_lock. - Add the tdx_vm_state_guard CLASS() to take kvm->lock, all vcpu->mutex locks, and kvm->slots_lock, in order to make tdx_td_init(), tdx_td_finalize(), tdx_vcpu_init_mem_region(), and tdx_vcpu_init() mutually exclusive with each other, and mutually exclusvie with basically anything that can result in contending one of the TDX-Module locks (can't remember which one). - Refine the changelog for the "Drop PROVE_MMU=y" patch. [Binbin] v2: - Collect a few reviews (and ignore some because the patches went away). [Rick, Kai, Ira] - Move TDH_MEM_PAGE_ADD under mmu_lock and drop nr_premapped. [Yan, Rick] - Force max_level = PG_LEVEL_4K straightaway. [Yan] - s/kvm_tdp_prefault_page/kvm_tdp_page_prefault. [Rick] - Use Yan's version of "Say no to pinning!". [Yan, Rick] - Tidy up helpers and macros to reduce boilerplate and copy+pate code, and to eliminate redundant/dead code (e.g. KVM_BUG_ON() the same error multiple times). - KVM_BUG_ON() if TDH_MR_EXTEND fails (I convinced myself it can't). v1: https://lore.kernel.org/all/[email protected] Sean Christopherson (26): KVM: Make support for kvm_arch_vcpu_async_ioctl() mandatory KVM: Rename kvm_arch_vcpu_async_ioctl() to kvm_arch_vcpu_unlocked_ioctl() KVM: TDX: Drop PROVE_MMU=y sanity check on to-be-populated mappings KVM: x86/mmu: Add dedicated API to map guest_memfd pfn into TDP MMU KVM: x86/mmu: WARN if KVM attempts to map into an invalid TDP MMU root Revert "KVM: x86/tdp_mmu: Add a helper function to walk down the TDP MMU" KVM: x86/mmu: Rename kvm_tdp_map_page() to kvm_tdp_page_prefault() KVM: TDX: Return -EIO, not -EINVAL, on a KVM_BUG_ON() condition KVM: TDX: Fold tdx_sept_drop_private_spte() into tdx_sept_remove_private_spte() KVM: x86/mmu: Drop the return code from kvm_x86_ops.remove_external_spte() KVM: TDX: WARN if mirror SPTE doesn't have full RWX when creating S-EPT mapping KVM: TDX: Avoid a double-KVM_BUG_ON() in tdx_sept_zap_private_spte() KVM: TDX: Use atomic64_dec_return() instead of a poor equivalent KVM: TDX: Fold tdx_mem_page_record_premap_cnt() into its sole caller KVM: TDX: ADD pages to the TD image while populating mirror EPT entries KVM: TDX: Fold tdx_sept_zap_private_spte() into tdx_sept_remove_private_spte() KVM: TDX: Combine KVM_BUG_ON + pr_tdx_error() into TDX_BUG_ON() KVM: TDX: Derive error argument names from the local variable names KVM: TDX: Assert that mmu_lock is held for write when removing S-EPT entries KVM: TDX: Add macro to retry SEAMCALLs when forcing vCPUs out of guest KVM: TDX: Add tdx_get_cmd() helper to get and validate sub-ioctl command KVM: TDX: Convert INIT_MEM_REGION and INIT_VCPU to "unlocked" vCPU ioctl KVM: TDX: Use guard() to acquire kvm->lock in tdx_vm_ioctl() KVM: TDX: Don't copy "cmd" back to userspace for KVM_TDX_CAPABILITIES KVM: TDX: Guard VM state transitions with "all" the locks KVM: TDX: Bug the VM if extending the initial measurement fails Yan Zhao (2): KVM: TDX: Drop superfluous page pinning in S-EPT management KVM: TDX: Fix list_add corruption during vcpu_load() arch/arm64/kvm/arm.c | 6 + arch/loongarch/kvm/Kconfig | 1 - arch/loongarch/kvm/vcpu.c | 4 +- arch/mips/kvm/Kconfig | 1 - arch/mips/kvm/mips.c | 4 +- arch/powerpc/kvm/Kconfig | 1 - arch/powerpc/kvm/powerpc.c | 4 +- arch/riscv/kvm/Kconfig | 1 - arch/riscv/kvm/vcpu.c | 4 +- arch/s390/kvm/Kconfig | 1 - arch/s390/kvm/kvm-s390.c | 4 +- arch/x86/include/asm/kvm-x86-ops.h | 1 + arch/x86/include/asm/kvm_host.h | 7 +- arch/x86/kvm/mmu.h | 3 +- arch/x86/kvm/mmu/mmu.c | 87 +++- arch/x86/kvm/mmu/tdp_mmu.c | 50 +-- arch/x86/kvm/vmx/main.c | 9 + arch/x86/kvm/vmx/tdx.c | 659 ++++++++++++++--------------- arch/x86/kvm/vmx/tdx.h | 8 +- arch/x86/kvm/vmx/x86_ops.h | 1 + arch/x86/kvm/x86.c | 13 + include/linux/kvm_host.h | 14 +- virt/kvm/Kconfig | 3 - virt/kvm/kvm_main.c | 6 +- 24 files changed, 468 insertions(+), 424 deletions(-) base-commit: 4cc167c50eb19d44ac7e204938724e685e3d8057 -- 2.51.1.930.gacf6e81ea2-goog
