On Mon, Dec 06, 2021, Vitaly Kuznetsov wrote: > Sean Christopherson <seanjc@xxxxxxxxxx> writes: > > I objected to the patch[*], but looking back at the dates, it appears that I did > > so after the patch was queued and my comments were never addressed. > > I'll see if I can reproduce this with a selftest. The fix is likely just: > > > > diff --git a/arch/x86/kvm/vmx/vmx.c b/arch/x86/kvm/vmx/vmx.c > > index dc4909b67c5c..927a7c43b73b 100644 > > --- a/arch/x86/kvm/vmx/vmx.c > > +++ b/arch/x86/kvm/vmx/vmx.c > > @@ -6665,10 +6665,6 @@ static fastpath_t vmx_vcpu_run(struct kvm_vcpu *vcpu) > > * consistency check VM-Exit due to invalid guest state and bail. > > */ > > if (unlikely(vmx->emulation_required)) { > > - > > - /* We don't emulate invalid state of a nested guest */ > > - vmx->fail = is_guest_mode(vcpu); > > - > > vmx->exit_reason.full = EXIT_REASON_INVALID_STATE; > > vmx->exit_reason.failed_vmentry = 1; > > kvm_register_mark_available(vcpu, VCPU_EXREG_EXIT_INFO_1); > > > > [*] https://lore.kernel.org/all/YWDWPbgJik5spT1D@xxxxxxxxxx/ Boom. VCPU_RUN exits with KVM_EXIT_INTERNAL_ERROR. diff --git a/tools/testing/selftests/kvm/x86_64/vmx_close_while_nested_test.c b/tools/testing/selftests/kvm/x86_64/vmx_close_while_nested_test.c index 2835a17f1b7a..4f77c5d7c7b9 100644 --- a/tools/testing/selftests/kvm/x86_64/vmx_close_while_nested_test.c +++ b/tools/testing/selftests/kvm/x86_64/vmx_close_while_nested_test.c @@ -27,6 +27,11 @@ enum { /* The virtual machine object. */ static struct kvm_vm *vm; +static void l2_guest_infinite_loop(void) +{ + while (1); +} + static void l2_guest_code(void) { /* Exit to L0 */ @@ -53,6 +58,9 @@ static void l1_guest_code(struct vmx_pages *vmx_pages) int main(int argc, char *argv[]) { vm_vaddr_t vmx_pages_gva; + struct kvm_sregs sregs; + struct kvm_regs regs; + int r; nested_vmx_check_supported(); @@ -83,4 +91,17 @@ int main(int argc, char *argv[]) TEST_FAIL("Unknown ucall %lu", uc.cmd); } } + + memset(®s, 0, sizeof(regs)); + vcpu_regs_get(vm, VCPU_ID, ®s); + regs.rip = (u64)l2_guest_infinite_loop; + vcpu_regs_set(vm, VCPU_ID, ®s); + + memset(&sregs, 0, sizeof(sregs)); + vcpu_sregs_get(vm, VCPU_ID, &sregs); + sregs.tr.unusable = 1; + vcpu_sregs_set(vm, VCPU_ID, &sregs); + + r = _vcpu_run(vm, VCPU_ID); + TEST_ASSERT(0, "Unexpected return from L2, r = %d, exit_reason = %d", r, vcpu_state(vm, VCPU_ID)->exit_reason); } ------------[ cut here ]------------ WARNING: CPU: 6 PID: 273926 at arch/x86/kvm/vmx/nested.c:4565 nested_vmx_vmexit+0xd59/0xdb0 [kvm_intel] CPU: 6 PID: 273926 Comm: vmx_close_while Not tainted 5.15.2-7cc36c3e14ae-pop #279 Hardware name: ASUS Q87M-E/Q87M-E, BIOS 1102 03/03/2014 RIP: 0010:nested_vmx_vmexit+0xd59/0xdb0 [kvm_intel] Call Trace: vmx_leave_nested+0x30/0x40 [kvm_intel] nested_vmx_free_vcpu+0x16/0x20 [kvm_intel] vmx_free_vcpu+0x4b/0x60 [kvm_intel] kvm_arch_vcpu_destroy+0x40/0x160 [kvm] kvm_vcpu_destroy+0x1d/0x50 [kvm] kvm_arch_destroy_vm+0xc1/0x1c0 [kvm] kvm_put_kvm+0x187/0x2a0 [kvm] kvm_vm_release+0x1d/0x30 [kvm] __fput+0x95/0x250 task_work_run+0x5f/0x90 do_exit+0x3c8/0xab0 do_group_exit+0x47/0xb0 __x64_sys_exit_group+0x14/0x20 do_syscall_64+0x3b/0xc0 entry_SYSCALL_64_after_hwframe+0x44/0xae