linux-kernel - [PATCH] KVM: x86/mmu: Treat NX as a valid SPTE bit for NPT

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

Hash Suite for Android: free password hash cracker in your pocket

[<prev] [next>] [thread-next>] [day] [month] [year] [list]

Message-Id: <20220723013029.1753623-1-seanjc@google.com>
Date:   Sat, 23 Jul 2022 01:30:29 +0000
From:   Sean Christopherson <seanjc@...gle.com>
To:     Sean Christopherson <seanjc@...gle.com>,
        Paolo Bonzini <pbonzini@...hat.com>
Cc:     kvm@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: [PATCH] KVM: x86/mmu: Treat NX as a valid SPTE bit for NPT

Treat the NX bit as valid when using NPT, as KVM will set the NX bit when
the NX huge page mitigation is enabled (mindblowing) and trigger the WARN
that fires on reserved SPTE bits being set.

KVM has required NX support for SVM since commit b26a71a1a5b9 ("KVM: SVM:
Refuse to load kvm_amd if NX support is not available") for exactly this
reason, but apparently it never occurred to anyone to actually test NPT
with the mitigation enabled.

  ------------[ cut here ]------------
  spte = 0x800000018a600ee7, level = 2, rsvd bits = 0x800f0000001fe000
  WARNING: CPU: 152 PID: 15966 at arch/x86/kvm/mmu/spte.c:215 make_spte+0x327/0x340 [kvm]
  Hardware name: Google, Inc. Arcadia_IT_80/Arcadia_IT_80, BIOS 10.48.0 01/27/2022
  RIP: 0010:make_spte+0x327/0x340 [kvm]
  Call Trace:
   <TASK>
   tdp_mmu_map_handle_target_level+0xc3/0x230 [kvm]
   kvm_tdp_mmu_map+0x343/0x3b0 [kvm]
   direct_page_fault+0x1ae/0x2a0 [kvm]
   kvm_tdp_page_fault+0x7d/0x90 [kvm]
   kvm_mmu_page_fault+0xfb/0x2e0 [kvm]
   npf_interception+0x55/0x90 [kvm_amd]
   svm_invoke_exit_handler+0x31/0xf0 [kvm_amd]
   svm_handle_exit+0xf6/0x1d0 [kvm_amd]
   vcpu_enter_guest+0xb6d/0xee0 [kvm]
   ? kvm_pmu_trigger_event+0x6d/0x230 [kvm]
   vcpu_run+0x65/0x2c0 [kvm]
   kvm_arch_vcpu_ioctl_run+0x355/0x610 [kvm]
   kvm_vcpu_ioctl+0x551/0x610 [kvm]
   __se_sys_ioctl+0x77/0xc0
   __x64_sys_ioctl+0x1d/0x20
   do_syscall_64+0x44/0xa0
   entry_SYSCALL_64_after_hwframe+0x46/0xb0
   </TASK>
  ---[ end trace 0000000000000000 ]---

Cc: stable@...r.kernel.org
Signed-off-by: Sean Christopherson <seanjc@...gle.com>
---
 arch/x86/kvm/mmu/mmu.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/arch/x86/kvm/mmu/mmu.c b/arch/x86/kvm/mmu/mmu.c
index 8e477333a263..3e1317325e1f 100644
--- a/arch/x86/kvm/mmu/mmu.c
+++ b/arch/x86/kvm/mmu/mmu.c
@@ -4735,7 +4735,7 @@ reset_tdp_shadow_zero_bits_mask(struct kvm_mmu *context)
 
 	if (boot_cpu_is_amd())
 		__reset_rsvds_bits_mask(shadow_zero_check, reserved_hpa_bits(),
-					context->root_role.level, false,
+					context->root_role.level, true,
 					boot_cpu_has(X86_FEATURE_GBPAGES),
 					false, true);
 	else

base-commit: 1a4d88a361af4f2e91861d632c6a1fe87a9665c2
-- 
2.37.1.359.gd136c6c3e2-goog