| Message ID | 20230729013535.1070024-1-seanjc@google.com (mailing list archive) |
|---|---|
| Headers | show |
| Series | drm/i915/gvt: KVM: KVMGT fixes and page-track cleanups | expand |
On Fri, 28 Jul 2023 18:35:06 -0700, Sean Christopherson wrote: > Fix a handful of minor bugs in KVMGT, and overhaul KVM's page-track APIs > to provide a leaner and cleaner interface. The motivation for this > series is to (significantly) reduce the number of KVM APIs that KVMGT > uses, with a long-term goal of making all kvm_host.h headers KVM-internal. > > If there are no objections or issues, my plan is to take this through the > KVM tree for 6.6 (I had it ready early last week, and then forgot to actually > post v4, /facepalm). > > [...] Applied to kvm-x86 mmu, thanks! [01/29] drm/i915/gvt: Verify pfn is "valid" before dereferencing "struct page" https://github.com/kvm-x86/linux/commit/865327865164 [02/29] drm/i915/gvt: remove interface intel_gvt_is_valid_gfn https://github.com/kvm-x86/linux/commit/823ab2ea8429 [03/29] drm/i915/gvt: Verify hugepages are contiguous in physical address space https://github.com/kvm-x86/linux/commit/e27395fb1b87 [04/29] drm/i915/gvt: Don't try to unpin an empty page range https://github.com/kvm-x86/linux/commit/6a718c54c2ee [05/29] drm/i915/gvt: Put the page reference obtained by KVM's gfn_to_pfn() https://github.com/kvm-x86/linux/commit/f969ecabe30b [06/29] drm/i915/gvt: Explicitly check that vGPU is attached before shadowing https://github.com/kvm-x86/linux/commit/537eef32e720 [07/29] drm/i915/gvt: Error out on an attempt to shadowing an unknown GTT entry type https://github.com/kvm-x86/linux/commit/c94811471997 [08/29] drm/i915/gvt: Don't rely on KVM's gfn_to_pfn() to query possible 2M GTT https://github.com/kvm-x86/linux/commit/f018c319cc2f [09/29] drm/i915/gvt: Use an "unsigned long" to iterate over memslot gfns https://github.com/kvm-x86/linux/commit/4879a4370304 [10/29] drm/i915/gvt: Drop unused helper intel_vgpu_reset_gtt() https://github.com/kvm-x86/linux/commit/ac5e77621712 [11/29] drm/i915/gvt: Protect gfn hash table with vgpu_lock https://github.com/kvm-x86/linux/commit/49a83e190b5b [12/29] KVM: x86/mmu: Move kvm_arch_flush_shadow_{all,memslot}() to mmu.c https://github.com/kvm-x86/linux/commit/2f502998b046 [13/29] KVM: x86/mmu: Don't rely on page-track mechanism to flush on memslot change https://github.com/kvm-x86/linux/commit/e2fe84fb5eae [14/29] KVM: x86/mmu: Don't bounce through page-track mechanism for guest PTEs https://github.com/kvm-x86/linux/commit/f1c58cdb8e04 [15/29] KVM: drm/i915/gvt: Drop @vcpu from KVM's ->track_write() hook https://github.com/kvm-x86/linux/commit/3f8eb1d7d3ee [16/29] KVM: x86: Reject memslot MOVE operations if KVMGT is attached https://github.com/kvm-x86/linux/commit/aa611a99adb4 [17/29] drm/i915/gvt: Don't bother removing write-protection on to-be-deleted slot https://github.com/kvm-x86/linux/commit/a41e34b05da7 [18/29] KVM: x86: Add a new page-track hook to handle memslot deletion https://github.com/kvm-x86/linux/commit/cc49e12d8d3b [19/29] drm/i915/gvt: switch from ->track_flush_slot() to ->track_remove_region() https://github.com/kvm-x86/linux/commit/b9ae8a09f357 [20/29] KVM: x86: Remove the unused page-track hook track_flush_slot() https://github.com/kvm-x86/linux/commit/1265fb534fa1 [21/29] KVM: x86/mmu: Move KVM-only page-track declarations to internal header https://github.com/kvm-x86/linux/commit/c87966b313cc [22/29] KVM: x86/mmu: Use page-track notifiers iff there are external users https://github.com/kvm-x86/linux/commit/b5e33f265acd [23/29] KVM: x86/mmu: Drop infrastructure for multiple page-track modes https://github.com/kvm-x86/linux/commit/2431c9ab231a [24/29] KVM: x86/mmu: Rename page-track APIs to reflect the new reality https://github.com/kvm-x86/linux/commit/4b42f39917c1 [25/29] KVM: x86/mmu: Assert that correct locks are held for page write-tracking https://github.com/kvm-x86/linux/commit/21e0e1efd880 [26/29] KVM: x86/mmu: Bug the VM if write-tracking is used but not enabled https://github.com/kvm-x86/linux/commit/6d6ff9e6db15 [27/29] KVM: x86/mmu: Drop @slot param from exported/external page-track APIs https://github.com/kvm-x86/linux/commit/ca181aa50724 [28/29] KVM: x86/mmu: Handle KVM bookkeeping in page-track APIs, not callers https://github.com/kvm-x86/linux/commit/eae2d71635a1 [29/29] drm/i915/gvt: Drop final dependencies on KVM internal details https://github.com/kvm-x86/linux/commit/2d28b1230c1c -- https://github.com/kvm-x86/linux/tree/next https://github.com/kvm-x86/linux/tree/fixes
Fix a handful of minor bugs in KVMGT, and overhaul KVM's page-track APIs to provide a leaner and cleaner interface. The motivation for this series is to (significantly) reduce the number of KVM APIs that KVMGT uses, with a long-term goal of making all kvm_host.h headers KVM-internal. If there are no objections or issues, my plan is to take this through the KVM tree for 6.6 (I had it ready early last week, and then forgot to actually post v4, /facepalm). Thanks much for all the help! v4: - Collect tags. [Yongwei, Zhi, Yan] - Add a patch to fix a benign (other than a WARN) bug where KVMGT would attempt to unpin an empty range. [Yan] - Move the check for an attached vGPU all the way up to shadow_ppgtt_mm(). [Zhi] v3: - https://lore.kernel.org/all/20230513003600.818142-1-seanjc@google.com - Collect reviewed/tested tags (I apologize if I missed any, I manually gathered them this time due to a goof in my workflow). [Yan] - Drop check on max KVM paging size from KVMGT. [Yan] - Drop the explicit change on THP pages, and instead validate that the pfns (not struct page pointers) are contiguous. [Yan] - Fix buggy intel_gvt_dma_map_guest_page() usage by eliminating a helper for shadowing 2MiB GTT entries. [Yan] - Move kvm_arch_flush_shadow_{all,memslot}() to mmu.c instead of exposing kvm_mmu_zap_all_fast() outside of mmu.c. [Yan] - Fix an alignment goof in hlist_for_each_entry_srcu() usage. [Yan] - Wrap full definition of external page track structures with CONFIG_KVM_EXTERNAL_WRITE_TRACKING. [Yan] v2: - https://lore.kernel.org/all/20230311002258.852397-1-seanjc@google.com - Reuse vgpu_lock to protect gfn hash instead of introducing a new (and buggy) mutext. [Yan] - Remove a spurious return from kvm_page_track_init(). [Yan] - Take @kvm directly in the inner __kvm_page_track_write(). [Yan] - Delete the gfn sanity check that relies on kvm_is_visible_gfn() instead of providing a dedicated interface. [Yan] v1: https://lore.kernel.org/lkml/20221223005739.1295925-1-seanjc@google.com Sean Christopherson (24): drm/i915/gvt: Verify pfn is "valid" before dereferencing "struct page" drm/i915/gvt: Verify hugepages are contiguous in physical address space drm/i915/gvt: Put the page reference obtained by KVM's gfn_to_pfn() drm/i915/gvt: Explicitly check that vGPU is attached before shadowing drm/i915/gvt: Error out on an attempt to shadowing an unknown GTT entry type drm/i915/gvt: Don't rely on KVM's gfn_to_pfn() to query possible 2M GTT drm/i915/gvt: Use an "unsigned long" to iterate over memslot gfns drm/i915/gvt: Drop unused helper intel_vgpu_reset_gtt() drm/i915/gvt: Protect gfn hash table with vgpu_lock KVM: x86/mmu: Move kvm_arch_flush_shadow_{all,memslot}() to mmu.c KVM: x86/mmu: Don't rely on page-track mechanism to flush on memslot change KVM: x86/mmu: Don't bounce through page-track mechanism for guest PTEs KVM: drm/i915/gvt: Drop @vcpu from KVM's ->track_write() hook KVM: x86: Reject memslot MOVE operations if KVMGT is attached drm/i915/gvt: Don't bother removing write-protection on to-be-deleted slot KVM: x86/mmu: Move KVM-only page-track declarations to internal header KVM: x86/mmu: Use page-track notifiers iff there are external users KVM: x86/mmu: Drop infrastructure for multiple page-track modes KVM: x86/mmu: Rename page-track APIs to reflect the new reality KVM: x86/mmu: Assert that correct locks are held for page write-tracking KVM: x86/mmu: Bug the VM if write-tracking is used but not enabled KVM: x86/mmu: Drop @slot param from exported/external page-track APIs KVM: x86/mmu: Handle KVM bookkeeping in page-track APIs, not callers drm/i915/gvt: Drop final dependencies on KVM internal details Yan Zhao (5): drm/i915/gvt: remove interface intel_gvt_is_valid_gfn drm/i915/gvt: Don't try to unpin an empty page range KVM: x86: Add a new page-track hook to handle memslot deletion drm/i915/gvt: switch from ->track_flush_slot() to ->track_remove_region() KVM: x86: Remove the unused page-track hook track_flush_slot() arch/x86/include/asm/kvm_host.h | 16 +- arch/x86/include/asm/kvm_page_track.h | 73 +++----- arch/x86/kvm/mmu.h | 2 + arch/x86/kvm/mmu/mmu.c | 51 +++-- arch/x86/kvm/mmu/page_track.c | 256 +++++++++++++------------- arch/x86/kvm/mmu/page_track.h | 58 ++++++ arch/x86/kvm/x86.c | 22 +-- drivers/gpu/drm/i915/gvt/gtt.c | 102 ++-------- drivers/gpu/drm/i915/gvt/gtt.h | 1 - drivers/gpu/drm/i915/gvt/gvt.h | 3 +- drivers/gpu/drm/i915/gvt/kvmgt.c | 120 +++++------- drivers/gpu/drm/i915/gvt/page_track.c | 10 +- 12 files changed, 322 insertions(+), 392 deletions(-) create mode 100644 arch/x86/kvm/mmu/page_track.h base-commit: fdf0eaf11452d72945af31804e2a1048ee1b574c