From 024d83cadc6b2af027e473720f3c3da97496c318 Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Sun, 12 Aug 2018 20:41:45 +0200 Subject: [PATCH] KVM: x86: SVM: Call x86_spec_ctrl_set_guest/host() with interrupts disabled MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Mikhail reported the following lockdep splat: WARNING: possible irq lock inversion dependency detected CPU 0/KVM/10284 just changed the state of lock: 000000000d538a88 (&st->lock){+...}, at: speculative_store_bypass_update+0x10b/0x170 but this lock was taken by another, HARDIRQ-safe lock in the past: (&(&sighand->siglock)->rlock){-.-.} and interrupts could create inverse lock ordering between them. Possible interrupt unsafe locking scenario: CPU0 CPU1 ---- ---- lock(&st->lock); local_irq_disable(); lock(&(&sighand->siglock)->rlock); lock(&st->lock); lock(&(&sighand->siglock)->rlock); *** DEADLOCK *** The code path which connects those locks is: speculative_store_bypass_update() ssb_prctl_set() do_seccomp() do_syscall_64() In svm_vcpu_run() speculative_store_bypass_update() is called with interupts enabled via x86_virt_spec_ctrl_set_guest/host(). This is actually a false positive, because GIF=0 so interrupts are disabled even if IF=1; however, we can easily move the invocations of x86_virt_spec_ctrl_set_guest/host() into the interrupt disabled region to cure it, and it's a good idea to keep the GIF=0/IF=1 area as small and self-contained as possible. Fixes: 1f50ddb4f418 ("x86/speculation: Handle HT correctly on AMD") Reported-by: Mikhail Gavrilov Signed-off-by: Thomas Gleixner Tested-by: Mikhail Gavrilov Cc: Joerg Roedel Cc: Paolo Bonzini Cc: Radim Krčmář Cc: Matthew Wilcox Cc: Borislav Petkov Cc: Konrad Rzeszutek Wilk Cc: Tom Lendacky Cc: kvm@vger.kernel.org Cc: x86@kernel.org Cc: stable@vger.kernel.org Signed-off-by: Paolo Bonzini --- arch/x86/kvm/svm.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/arch/x86/kvm/svm.c b/arch/x86/kvm/svm.c index 73e27a98456f..6276140044d0 100644 --- a/arch/x86/kvm/svm.c +++ b/arch/x86/kvm/svm.c @@ -5586,8 +5586,6 @@ static void svm_vcpu_run(struct kvm_vcpu *vcpu) clgi(); - local_irq_enable(); - /* * If this vCPU has touched SPEC_CTRL, restore the guest's value if * it's non-zero. Since vmentry is serialising on affected CPUs, there @@ -5596,6 +5594,8 @@ static void svm_vcpu_run(struct kvm_vcpu *vcpu) */ x86_spec_ctrl_set_guest(svm->spec_ctrl, svm->virt_spec_ctrl); + local_irq_enable(); + asm volatile ( "push %%" _ASM_BP "; \n\t" "mov %c[rbx](%[svm]), %%" _ASM_BX " \n\t" @@ -5718,12 +5718,12 @@ static void svm_vcpu_run(struct kvm_vcpu *vcpu) if (unlikely(!msr_write_intercepted(vcpu, MSR_IA32_SPEC_CTRL))) svm->spec_ctrl = native_read_msr(MSR_IA32_SPEC_CTRL); - x86_spec_ctrl_restore_host(svm->spec_ctrl, svm->virt_spec_ctrl); - reload_tss(vcpu); local_irq_disable(); + x86_spec_ctrl_restore_host(svm->spec_ctrl, svm->virt_spec_ctrl); + vcpu->arch.cr2 = svm->vmcb->save.cr2; vcpu->arch.regs[VCPU_REGS_RAX] = svm->vmcb->save.rax; vcpu->arch.regs[VCPU_REGS_RSP] = svm->vmcb->save.rsp;