| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <202209200811.UMs7ibme-lkp@intel.com>
Date: Tue, 20 Sep 2022 08:40:37 +0800
From: kernel test robot <lkp@...el.com>
To: Isaku Yamahata <isaku.yamahata@...el.com>
Cc: kbuild-all@...ts.01.org, linux-kernel@...r.kernel.org,
Kai Huang <kai.huang@...el.com>
Subject: [intel-tdx:kvm-upstream-workaround 72/314]
arch/x86/kvm/mmu/tdp_mmu.c:641:17: sparse: sparse: incorrect type in
argument 1 (different address spaces)
tree: https://github.com/intel/tdx.git kvm-upstream-workaround
head: 552dd80c48f67ca01bcdd10667e0c11efd375177
commit: e31d260a88d02b51769ca56edd75a84f2d2cbf97 [72/314] KVM: x86/tdp_mmu: Support TDX private mapping for TDP MMU
config: x86_64-randconfig-s022-20220919
compiler: gcc-11 (Debian 11.3.0-5) 11.3.0
reproduce:
# apt-get install sparse
# sparse version: v0.6.4-39-gce1a6720-dirty
# https://github.com/intel/tdx/commit/e31d260a88d02b51769ca56edd75a84f2d2cbf97
git remote add intel-tdx https://github.com/intel/tdx.git
git fetch --no-tags intel-tdx kvm-upstream-workaround
git checkout e31d260a88d02b51769ca56edd75a84f2d2cbf97
# save the config file
mkdir build_dir && cp config build_dir/.config
make W=1 C=1 CF='-fdiagnostic-prefix -D__CHECK_ENDIAN__' O=build_dir ARCH=x86_64 SHELL=/bin/bash arch/x86/kvm/ block/ drivers/acpi/ drivers/ata/ drivers/auxdisplay/ drivers/block/ drivers/comedi/ drivers/hwmon/ drivers/infiniband/core/ drivers/infiniband/hw/qib/ drivers/net/ drivers/nvme/host/ drivers/platform/x86/ drivers/rapidio/ drivers/scsi/ drivers/staging/media/av7110/ drivers/thermal/intel/int340x_thermal/ drivers/tty/ drivers/vfio/ drivers/vhost/ drivers/watchdog/ drivers/xen/ fs/crypto/ fs/ext2/ fs/ext4/ fs/fat/ fs/quota/ fs/reiserfs/ ipc/ kernel/ net/appletalk/ net/ax25/ net/bluetooth/ net/can/ net/core/ net/dccp/ net/decnet/ net/ipv4/bpfilter/ net/llc/ net/mptcp/ net/netfilter/ net/netrom/ net/qrtr/ net/rfkill/ net/rose/ net/rxrpc/ net/sctp/ net/tls/ net/x25/ security/apparmor/ sound/core/ sound/pci/asihpi/ virt/
If you fix the issue, kindly add following tag where applicable
Reported-by: kernel test robot <lkp@...el.com>
sparse warnings: (new ones prefixed by >>)
>> arch/x86/kvm/mmu/tdp_mmu.c:641:17: sparse: sparse: incorrect type in argument 1 (different address spaces) @@ expected unsigned long long [usertype] *sptep @@ got unsigned long long [noderef] [usertype] __rcu * @@
arch/x86/kvm/mmu/tdp_mmu.c:641:17: sparse: expected unsigned long long [usertype] *sptep
arch/x86/kvm/mmu/tdp_mmu.c:641:17: sparse: got unsigned long long [noderef] [usertype] __rcu *
>> arch/x86/kvm/mmu/tdp_mmu.c:833:28: sparse: sparse: incorrect type in argument 1 (different address spaces) @@ expected unsigned long long [usertype] *sptep @@ got unsigned long long [noderef] [usertype] __rcu *[usertype] sptep @@
arch/x86/kvm/mmu/tdp_mmu.c:833:28: sparse: expected unsigned long long [usertype] *sptep
arch/x86/kvm/mmu/tdp_mmu.c:833:28: sparse: got unsigned long long [noderef] [usertype] __rcu *[usertype] sptep
>> arch/x86/kvm/mmu/tdp_mmu.c:1341:25: sparse: sparse: incorrect type in argument 1 (different address spaces) @@ expected unsigned long long [usertype] *sptep @@ got unsigned long long [noderef] [usertype] __rcu *[addressable] [usertype] sptep @@
arch/x86/kvm/mmu/tdp_mmu.c:1341:25: sparse: expected unsigned long long [usertype] *sptep
arch/x86/kvm/mmu/tdp_mmu.c:1341:25: sparse: got unsigned long long [noderef] [usertype] __rcu *[addressable] [usertype] sptep
arch/x86/kvm/mmu/tdp_mmu.c:1606:9: sparse: sparse: incorrect type in argument 1 (different address spaces) @@ expected unsigned long long [usertype] *sptep @@ got unsigned long long [noderef] [usertype] __rcu *[usertype] sptep @@
arch/x86/kvm/mmu/tdp_mmu.c:1606:9: sparse: expected unsigned long long [usertype] *sptep
arch/x86/kvm/mmu/tdp_mmu.c:1606:9: sparse: got unsigned long long [noderef] [usertype] __rcu *[usertype] sptep
arch/x86/kvm/mmu/tdp_mmu.c:406:9: sparse: sparse: context imbalance in 'tdp_mmu_unlink_sp' - different lock contexts for basic block
arch/x86/kvm/mmu/tdp_mmu.c:721:49: sparse: sparse: incorrect type in argument 1 (different address spaces) @@ expected unsigned long long [usertype] *sptep @@ got unsigned long long [noderef] [usertype] __rcu *[usertype] sptep @@
arch/x86/kvm/mmu/tdp_mmu.c:721:49: sparse: expected unsigned long long [usertype] *sptep
arch/x86/kvm/mmu/tdp_mmu.c:721:49: sparse: got unsigned long long [noderef] [usertype] __rcu *[usertype] sptep
>> arch/x86/kvm/mmu/tdp_mmu.c:747:42: sparse: sparse: incorrect type in argument 1 (different address spaces) @@ expected unsigned long long [noderef] [usertype] __rcu *[addressable] [usertype] sptep @@ got unsigned long long [usertype] *sptep @@
arch/x86/kvm/mmu/tdp_mmu.c:747:42: sparse: expected unsigned long long [noderef] [usertype] __rcu *[addressable] [usertype] sptep
arch/x86/kvm/mmu/tdp_mmu.c:747:42: sparse: got unsigned long long [usertype] *sptep
arch/x86/kvm/mmu/tdp_mmu.c: note: in included file (through include/linux/rculist.h, include/linux/pid.h, include/linux/sched.h, ...):
include/linux/rcupdate.h:737:9: sparse: sparse: context imbalance in '__tdp_mmu_zap_root' - unexpected unlock
arch/x86/kvm/mmu/tdp_mmu.c:721:49: sparse: sparse: incorrect type in argument 1 (different address spaces) @@ expected unsigned long long [usertype] *sptep @@ got unsigned long long [noderef] [usertype] __rcu *[usertype] sptep @@
arch/x86/kvm/mmu/tdp_mmu.c:721:49: sparse: expected unsigned long long [usertype] *sptep
arch/x86/kvm/mmu/tdp_mmu.c:721:49: sparse: got unsigned long long [noderef] [usertype] __rcu *[usertype] sptep
>> arch/x86/kvm/mmu/tdp_mmu.c:747:42: sparse: sparse: incorrect type in argument 1 (different address spaces) @@ expected unsigned long long [noderef] [usertype] __rcu *[addressable] [usertype] sptep @@ got unsigned long long [usertype] *sptep @@
arch/x86/kvm/mmu/tdp_mmu.c:747:42: sparse: expected unsigned long long [noderef] [usertype] __rcu *[addressable] [usertype] sptep
arch/x86/kvm/mmu/tdp_mmu.c:747:42: sparse: got unsigned long long [usertype] *sptep
arch/x86/kvm/mmu/tdp_mmu.c:721:49: sparse: sparse: incorrect type in argument 1 (different address spaces) @@ expected unsigned long long [usertype] *sptep @@ got unsigned long long [noderef] [usertype] __rcu *[usertype] sptep @@
arch/x86/kvm/mmu/tdp_mmu.c:721:49: sparse: expected unsigned long long [usertype] *sptep
arch/x86/kvm/mmu/tdp_mmu.c:721:49: sparse: got unsigned long long [noderef] [usertype] __rcu *[usertype] sptep
>> arch/x86/kvm/mmu/tdp_mmu.c:747:42: sparse: sparse: incorrect type in argument 1 (different address spaces) @@ expected unsigned long long [noderef] [usertype] __rcu *[addressable] [usertype] sptep @@ got unsigned long long [usertype] *sptep @@
arch/x86/kvm/mmu/tdp_mmu.c:747:42: sparse: expected unsigned long long [noderef] [usertype] __rcu *[addressable] [usertype] sptep
arch/x86/kvm/mmu/tdp_mmu.c:747:42: sparse: got unsigned long long [usertype] *sptep
arch/x86/kvm/mmu/tdp_mmu.c:721:49: sparse: sparse: incorrect type in argument 1 (different address spaces) @@ expected unsigned long long [usertype] *sptep @@ got unsigned long long [noderef] [usertype] __rcu *[usertype] sptep @@
arch/x86/kvm/mmu/tdp_mmu.c:721:49: sparse: expected unsigned long long [usertype] *sptep
arch/x86/kvm/mmu/tdp_mmu.c:721:49: sparse: got unsigned long long [noderef] [usertype] __rcu *[usertype] sptep
>> arch/x86/kvm/mmu/tdp_mmu.c:747:42: sparse: sparse: incorrect type in argument 1 (different address spaces) @@ expected unsigned long long [noderef] [usertype] __rcu *[addressable] [usertype] sptep @@ got unsigned long long [usertype] *sptep @@
arch/x86/kvm/mmu/tdp_mmu.c:747:42: sparse: expected unsigned long long [noderef] [usertype] __rcu *[addressable] [usertype] sptep
arch/x86/kvm/mmu/tdp_mmu.c:747:42: sparse: got unsigned long long [usertype] *sptep
arch/x86/kvm/mmu/tdp_mmu.c:721:49: sparse: sparse: incorrect type in argument 1 (different address spaces) @@ expected unsigned long long [usertype] *sptep @@ got unsigned long long [noderef] [usertype] __rcu *[usertype] sptep @@
arch/x86/kvm/mmu/tdp_mmu.c:721:49: sparse: expected unsigned long long [usertype] *sptep
arch/x86/kvm/mmu/tdp_mmu.c:721:49: sparse: got unsigned long long [noderef] [usertype] __rcu *[usertype] sptep
>> arch/x86/kvm/mmu/tdp_mmu.c:747:42: sparse: sparse: incorrect type in argument 1 (different address spaces) @@ expected unsigned long long [noderef] [usertype] __rcu *[addressable] [usertype] sptep @@ got unsigned long long [usertype] *sptep @@
arch/x86/kvm/mmu/tdp_mmu.c:747:42: sparse: expected unsigned long long [noderef] [usertype] __rcu *[addressable] [usertype] sptep
arch/x86/kvm/mmu/tdp_mmu.c:747:42: sparse: got unsigned long long [usertype] *sptep
include/linux/rcupdate.h:737:9: sparse: sparse: context imbalance in 'tdp_mmu_alloc_sp_for_split' - unexpected unlock
arch/x86/kvm/mmu/tdp_mmu.c:721:49: sparse: sparse: incorrect type in argument 1 (different address spaces) @@ expected unsigned long long [usertype] *sptep @@ got unsigned long long [noderef] [usertype] __rcu *[usertype] sptep @@
arch/x86/kvm/mmu/tdp_mmu.c:721:49: sparse: expected unsigned long long [usertype] *sptep
arch/x86/kvm/mmu/tdp_mmu.c:721:49: sparse: got unsigned long long [noderef] [usertype] __rcu *[usertype] sptep
>> arch/x86/kvm/mmu/tdp_mmu.c:747:42: sparse: sparse: incorrect type in argument 1 (different address spaces) @@ expected unsigned long long [noderef] [usertype] __rcu *[addressable] [usertype] sptep @@ got unsigned long long [usertype] *sptep @@
arch/x86/kvm/mmu/tdp_mmu.c:747:42: sparse: expected unsigned long long [noderef] [usertype] __rcu *[addressable] [usertype] sptep
arch/x86/kvm/mmu/tdp_mmu.c:747:42: sparse: got unsigned long long [usertype] *sptep
arch/x86/kvm/mmu/tdp_mmu.c:721:49: sparse: sparse: incorrect type in argument 1 (different address spaces) @@ expected unsigned long long [usertype] *sptep @@ got unsigned long long [noderef] [usertype] __rcu *[usertype] sptep @@
arch/x86/kvm/mmu/tdp_mmu.c:721:49: sparse: expected unsigned long long [usertype] *sptep
arch/x86/kvm/mmu/tdp_mmu.c:721:49: sparse: got unsigned long long [noderef] [usertype] __rcu *[usertype] sptep
>> arch/x86/kvm/mmu/tdp_mmu.c:747:42: sparse: sparse: incorrect type in argument 1 (different address spaces) @@ expected unsigned long long [noderef] [usertype] __rcu *[addressable] [usertype] sptep @@ got unsigned long long [usertype] *sptep @@
arch/x86/kvm/mmu/tdp_mmu.c:747:42: sparse: expected unsigned long long [noderef] [usertype] __rcu *[addressable] [usertype] sptep
arch/x86/kvm/mmu/tdp_mmu.c:747:42: sparse: got unsigned long long [usertype] *sptep
vim +641 arch/x86/kvm/mmu/tdp_mmu.c
520
521 /**
522 * __handle_changed_spte - handle bookkeeping associated with an SPTE change
523 * @kvm: kvm instance
524 * @as_id: the address space of the paging structure the SPTE was a part of
525 * @gfn: the base GFN that was mapped by the SPTE
526 * @old_spte: The value of the SPTE before the change
527 * @new_spte: The value of the SPTE after the change
528 * @role: the role of the PT the SPTE is part of in the paging structure
529 * @shared: This operation may not be running under the exclusive use of
530 * the MMU lock and the operation must synchronize with other
531 * threads that might be modifying SPTEs.
532 *
533 * Handle bookkeeping that might result from the modification of a SPTE.
534 * This function must be called for all TDP SPTE modifications.
535 */
536 static void __handle_changed_spte(struct kvm *kvm, int as_id, gfn_t gfn,
537 u64 old_spte, u64 new_spte,
538 union kvm_mmu_page_role role, bool shared)
539 {
540 bool is_private = kvm_mmu_page_role_is_private(role);
541 int level = role.level;
542 bool was_present = is_shadow_present_pte(old_spte);
543 bool is_present = is_shadow_present_pte(new_spte);
544 bool was_last = is_last_spte(old_spte, level);
545 bool is_last = is_last_spte(new_spte, level);
546 bool was_leaf = was_present && was_last;
547 bool is_leaf = is_present && is_last;
548 kvm_pfn_t old_pfn = spte_to_pfn(old_spte);
549 kvm_pfn_t new_pfn = spte_to_pfn(new_spte);
550 bool pfn_changed = old_pfn != new_pfn;
551 struct kvm_spte_change change = {
552 .gfn = gfn,
553 .level = level,
554 .old = {
555 .pfn = old_pfn,
556 .is_present = was_present,
557 .is_last = was_last,
558 },
559 .new = {
560 .pfn = new_pfn,
561 .is_present = is_present,
562 .is_last = is_last,
563 },
564 };
565
566 WARN_ON(level > PT64_ROOT_MAX_LEVEL);
567 WARN_ON(level < PG_LEVEL_4K);
568 WARN_ON(gfn & (KVM_PAGES_PER_HPAGE(level) - 1));
569
570 /*
571 * If this warning were to trigger it would indicate that there was a
572 * missing MMU notifier or a race with some notifier handler.
573 * A present, leaf SPTE should never be directly replaced with another
574 * present leaf SPTE pointing to a different PFN. A notifier handler
575 * should be zapping the SPTE before the main MM's page table is
576 * changed, or the SPTE should be zeroed, and the TLBs flushed by the
577 * thread before replacement.
578 */
579 if (was_leaf && is_leaf && pfn_changed) {
580 pr_err("Invalid SPTE change: cannot replace a present leaf\n"
581 "SPTE with another present leaf SPTE mapping a\n"
582 "different PFN!\n"
583 "as_id: %d gfn: %llx old_spte: %llx new_spte: %llx level: %d",
584 as_id, gfn, old_spte, new_spte, level);
585
586 /*
587 * Crash the host to prevent error propagation and guest data
588 * corruption.
589 */
590 BUG();
591 }
592
593 if (old_spte == new_spte)
594 return;
595
596 trace_kvm_tdp_mmu_spte_changed(as_id, gfn, level, old_spte, new_spte);
597
598 if (is_leaf)
599 check_spte_writable_invariants(new_spte);
600
601 /*
602 * The only times a SPTE should be changed from a non-present to
603 * non-present state is when an MMIO entry is installed/modified/
604 * removed. In that case, there is nothing to do here.
605 */
606 if (!was_present && !is_present) {
607 /*
608 * If this change does not involve a MMIO SPTE or removed SPTE,
609 * it is unexpected. Log the change, though it should not
610 * impact the guest since both the former and current SPTEs
611 * are nonpresent.
612 */
613 if (WARN_ON(!is_mmio_spte(kvm, old_spte) &&
614 !is_mmio_spte(kvm, new_spte) &&
615 !is_removed_spte(new_spte)))
616 pr_err("Unexpected SPTE change! Nonpresent SPTEs\n"
617 "should not be replaced with another,\n"
618 "different nonpresent SPTE, unless one or both\n"
619 "are MMIO SPTEs, or the new SPTE is\n"
620 "a temporary removed SPTE.\n"
621 "as_id: %d gfn: %llx old_spte: %llx new_spte: %llx level: %d",
622 as_id, gfn, old_spte, new_spte, level);
623 return;
624 }
625
626 if (is_leaf != was_leaf)
627 kvm_update_page_stats(kvm, level, is_leaf ? 1 : -1);
628
629 if (was_leaf && is_dirty_spte(old_spte) &&
630 (!is_present || !is_dirty_spte(new_spte) || pfn_changed))
631 kvm_set_pfn_dirty(old_pfn);
632
633 /*
634 * Recursively handle child PTs if the change removed a subtree from
635 * the paging structure. Note the WARN on the PFN changing without the
636 * SPTE being converted to a hugepage (leaf) or being zapped. Shadow
637 * pages are kernel allocations and should never be migrated.
638 */
639 if (was_present && !was_leaf &&
640 (is_leaf || !is_present || WARN_ON_ONCE(pfn_changed))) {
> 641 KVM_BUG_ON(is_private != is_private_sptep(spte_to_child_pt(old_spte, level)),
642 kvm);
643 handle_removed_pt(kvm, spte_to_child_pt(old_spte, level), shared);
644 }
645
646 /*
647 * Special handling for the private mapping. We are either
648 * setting up new mapping at middle level page table, or leaf,
649 * or tearing down existing mapping.
650 *
651 * This is after handling lower page table by above
652 * handle_remove_tdp_mmu_page(). S-EPT requires to remove S-EPT tables
653 * after removing childrens.
654 */
655 if (is_private &&
656 /* Ignore change of software only bits. e.g. host_writable */
657 (was_leaf != is_leaf || was_present != is_present || pfn_changed)) {
658 void *private_spt = NULL;
659
660 if (is_present && !is_leaf) {
661 struct kvm_mmu_page *sp = to_shadow_page(pfn_to_hpa(new_pfn));
662
663 private_spt = kvm_mmu_private_spt(sp);
664 KVM_BUG_ON(!private_spt, kvm);
665 KVM_BUG_ON(sp->role.level + 1 != level, kvm);
666 KVM_BUG_ON(sp->gfn != gfn, kvm);
667 }
668 change.private_spt = private_spt;
669
670 static_call(kvm_x86_handle_changed_private_spte)(kvm, &change);
671 }
672 }
673
674 static void handle_changed_spte(struct kvm *kvm, int as_id, gfn_t gfn,
675 u64 old_spte, u64 new_spte,
676 union kvm_mmu_page_role role, bool shared)
677 {
678 __handle_changed_spte(kvm, as_id, gfn, old_spte, new_spte, role, shared);
679 handle_changed_spte_acc_track(old_spte, new_spte, role.level);
680 handle_changed_spte_dirty_log(kvm, as_id, gfn, old_spte,
681 new_spte, role.level);
682 }
683
684 /*
685 * tdp_mmu_set_spte_atomic - Set a TDP MMU SPTE atomically
686 * and handle the associated bookkeeping. Do not mark the page dirty
687 * in KVM's dirty bitmaps.
688 *
689 * If setting the SPTE fails because it has changed, iter->old_spte will be
690 * refreshed to the current value of the spte.
691 *
692 * @kvm: kvm instance
693 * @iter: a tdp_iter instance currently on the SPTE that should be set
694 * @new_spte: The value the SPTE should be set to
695 * Return:
696 * * 0 - If the SPTE was set.
697 * * -EBUSY - If the SPTE cannot be set. In this case this function will have
698 * no side-effects other than setting iter->old_spte to the last
699 * known value of the spte.
700 */
701 static inline int tdp_mmu_set_spte_atomic(struct kvm *kvm,
702 struct tdp_iter *iter,
703 u64 new_spte)
704 {
705 /*
706 * For conventional page table, the update flow is
707 * - update STPE with atomic operation
708 * - handle changed SPTE. __handle_changed_spte()
709 * NOTE: __handle_changed_spte() (and functions) must be safe against
710 * concurrent update. It is an exception to zap SPTE. See
711 * tdp_mmu_zap_spte_atomic().
712 *
713 * For private page table, callbacks are needed to propagate SPTE
714 * change into the protected page table. In order to atomically update
715 * both the SPTE and the protected page tables with callbacks, utilize
716 * freezing SPTE.
717 * - Freeze the SPTE. Set entry to REMOVED_SPTE.
718 * - Trigger callbacks for protected page tables. __handle_changed_spte()
719 * - Unfreeze the SPTE. Set the entry to new_spte.
720 */
721 bool freeze_spte = is_private_sptep(iter->sptep) && !is_removed_spte(new_spte);
722 u64 tmp_spte = freeze_spte ? REMOVED_SPTE : new_spte;
723 u64 *sptep = rcu_dereference(iter->sptep);
724
725 /*
726 * The caller is responsible for ensuring the old SPTE is not a REMOVED
727 * SPTE. KVM should never attempt to zap or manipulate a REMOVED SPTE,
728 * and pre-checking before inserting a new SPTE is advantageous as it
729 * avoids unnecessary work.
730 */
731 WARN_ON_ONCE(iter->yielded || is_removed_spte(iter->old_spte));
732
733 lockdep_assert_held_read(&kvm->mmu_lock);
734
735 /*
736 * Note, fast_pf_fix_direct_spte() can also modify TDP MMU SPTEs and
737 * does not hold the mmu_lock.
738 */
739 if (!try_cmpxchg64(sptep, &iter->old_spte, tmp_spte))
740 return -EBUSY;
741
742 __handle_changed_spte(kvm, iter->as_id, iter->gfn, iter->old_spte,
743 new_spte, sptep_to_sp(sptep)->role, true);
744 handle_changed_spte_acc_track(iter->old_spte, new_spte, iter->level);
745
746 if (freeze_spte)
> 747 __kvm_tdp_mmu_write_spte(sptep, new_spte);
748
749 return 0;
750 }
751
752 static inline int tdp_mmu_zap_spte_atomic(struct kvm *kvm,
753 struct tdp_iter *iter)
754 {
755 int ret;
756
757 /*
758 * Freeze the SPTE by setting it to a special,
759 * non-present value. This will stop other threads from
760 * immediately installing a present entry in its place
761 * before the TLBs are flushed.
762 */
763 ret = tdp_mmu_set_spte_atomic(kvm, iter, REMOVED_SPTE);
764 if (ret)
765 return ret;
766
767 kvm_flush_remote_tlbs_with_address(kvm, iter->gfn,
768 KVM_PAGES_PER_HPAGE(iter->level));
769
770 /*
771 * No other thread can overwrite the removed SPTE as they must either
772 * wait on the MMU lock or use tdp_mmu_set_spte_atomic() which will not
773 * overwrite the special removed SPTE value. No bookkeeping is needed
774 * here since the SPTE is going from non-present to non-present. Use
775 * the raw write helper to avoid an unnecessary check on volatile bits.
776 *
777 * Set non-present value to SHADOW_NONPRESENT_VALUE, rather than 0.
778 * It is because when TDX is enabled, TDX module always
779 * enables "EPT-violation #VE", so KVM needs to set
780 * "suppress #VE" bit in EPT table entries, in order to get
781 * real EPT violation, rather than TDVMCALL. KVM sets
782 * SHADOW_NONPRESENT_VALUE (which sets "suppress #VE" bit) so it
783 * can be set when EPT table entries are zapped.
784 */
785 __kvm_tdp_mmu_write_spte(iter->sptep, SHADOW_NONPRESENT_VALUE);
786
787 return 0;
788 }
789
790
791 /*
792 * __tdp_mmu_set_spte - Set a TDP MMU SPTE and handle the associated bookkeeping
793 * @kvm: KVM instance
794 * @as_id: Address space ID, i.e. regular vs. SMM
795 * @sptep: Pointer to the SPTE
796 * @old_spte: The current value of the SPTE
797 * @new_spte: The new value that will be set for the SPTE
798 * @gfn: The base GFN that was (or will be) mapped by the SPTE
799 * @level: The level _containing_ the SPTE (its parent PT's level)
800 * @record_acc_track: Notify the MM subsystem of changes to the accessed state
801 * of the page. Should be set unless handling an MMU
802 * notifier for access tracking. Leaving record_acc_track
803 * unset in that case prevents page accesses from being
804 * double counted.
805 * @record_dirty_log: Record the page as dirty in the dirty bitmap if
806 * appropriate for the change being made. Should be set
807 * unless performing certain dirty logging operations.
808 * Leaving record_dirty_log unset in that case prevents page
809 * writes from being double counted.
810 *
811 * Returns the old SPTE value, which _may_ be different than @old_spte if the
812 * SPTE had voldatile bits.
813 */
814 static u64 __tdp_mmu_set_spte(struct kvm *kvm, int as_id, tdp_ptep_t sptep,
815 u64 old_spte, u64 new_spte, gfn_t gfn, int level,
816 bool record_acc_track, bool record_dirty_log)
817 {
818 union kvm_mmu_page_role role;
819
820 lockdep_assert_held_write(&kvm->mmu_lock);
821
822 /*
823 * No thread should be using this function to set SPTEs to or from the
824 * temporary removed SPTE value.
825 * If operating under the MMU lock in read mode, tdp_mmu_set_spte_atomic
826 * should be used. If operating under the MMU lock in write mode, the
827 * use of the removed SPTE should not be necessary.
828 */
829 WARN_ON(is_removed_spte(old_spte) || is_removed_spte(new_spte));
830
831 old_spte = kvm_tdp_mmu_write_spte(sptep, old_spte, new_spte, level);
832
> 833 role = sptep_to_sp(sptep)->role;
834 role.level = level;
835 __handle_changed_spte(kvm, as_id, gfn, old_spte, new_spte, role, false);
836
837 if (record_acc_track)
838 handle_changed_spte_acc_track(old_spte, new_spte, level);
839 if (record_dirty_log)
840 handle_changed_spte_dirty_log(kvm, as_id, gfn, old_spte,
841 new_spte, level);
842 return old_spte;
843 }
844
--
0-DAY CI Kernel Test Service
https://01.org/lkp
View attachment "config" of type "text/plain" (166514 bytes)
Powered by blists - more mailing lists