KVM: VMX: Require KVM_SET_TSS_ADDR being called prior to running a VCPU
authorJan Kiszka <jan.kiszka@siemens.com>
Fri, 15 Mar 2013 07:38:56 +0000 (08:38 +0100)
committerMarcelo Tosatti <mtosatti@redhat.com>
Mon, 18 Mar 2013 16:48:15 +0000 (13:48 -0300)
Very old user space (namely qemu-kvm before kvm-49) didn't set the TSS
base before running the VCPU. We always warned about this bug, but no
reports about users actually seeing this are known. Time to finally
remove the workaround that effectively prevented to call vmx_vcpu_reset
while already holding the KVM srcu lock.

Reviewed-by: Gleb Natapov <gleb@redhat.com>
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
arch/x86/kvm/vmx.c

index 17a693868458c46d7bc1de5b5c8c368275077f54..ad978a6c282e896145d2424dd4f799591218d5ba 100644 (file)
@@ -2898,22 +2898,6 @@ static void enter_pmode(struct kvm_vcpu *vcpu)
        vmx->cpl = 0;
 }
 
-static gva_t rmode_tss_base(struct kvm *kvm)
-{
-       if (!kvm->arch.tss_addr) {
-               struct kvm_memslots *slots;
-               struct kvm_memory_slot *slot;
-               gfn_t base_gfn;
-
-               slots = kvm_memslots(kvm);
-               slot = id_to_memslot(slots, 0);
-               base_gfn = slot->base_gfn + slot->npages - 3;
-
-               return base_gfn << PAGE_SHIFT;
-       }
-       return kvm->arch.tss_addr;
-}
-
 static void fix_rmode_seg(int seg, struct kvm_segment *save)
 {
        const struct kvm_vmx_segment_field *sf = &kvm_vmx_segment_fields[seg];
@@ -2964,19 +2948,15 @@ static void enter_rmode(struct kvm_vcpu *vcpu)
 
        /*
         * Very old userspace does not call KVM_SET_TSS_ADDR before entering
-        * vcpu. Call it here with phys address pointing 16M below 4G.
+        * vcpu. Warn the user that an update is overdue.
         */
-       if (!vcpu->kvm->arch.tss_addr) {
+       if (!vcpu->kvm->arch.tss_addr)
                printk_once(KERN_WARNING "kvm: KVM_SET_TSS_ADDR need to be "
                             "called before entering vcpu\n");
-               srcu_read_unlock(&vcpu->kvm->srcu, vcpu->srcu_idx);
-               vmx_set_tss_addr(vcpu->kvm, 0xfeffd000);
-               vcpu->srcu_idx = srcu_read_lock(&vcpu->kvm->srcu);
-       }
 
        vmx_segment_cache_clear(vmx);
 
-       vmcs_writel(GUEST_TR_BASE, rmode_tss_base(vcpu->kvm));
+       vmcs_writel(GUEST_TR_BASE, vcpu->kvm->arch.tss_addr);
        vmcs_write32(GUEST_TR_LIMIT, RMODE_TSS_SIZE - 1);
        vmcs_write32(GUEST_TR_AR_BYTES, 0x008b);
 
@@ -3623,7 +3603,7 @@ static int init_rmode_tss(struct kvm *kvm)
        int r, idx, ret = 0;
 
        idx = srcu_read_lock(&kvm->srcu);
-       fn = rmode_tss_base(kvm) >> PAGE_SHIFT;
+       fn = kvm->arch.tss_addr >> PAGE_SHIFT;
        r = kvm_clear_guest_page(kvm, fn, 0, PAGE_SIZE);
        if (r < 0)
                goto out;
@@ -4190,9 +4170,7 @@ static void vmx_vcpu_reset(struct kvm_vcpu *vcpu)
                vmcs_write16(VIRTUAL_PROCESSOR_ID, vmx->vpid);
 
        vmx->vcpu.arch.cr0 = X86_CR0_NW | X86_CR0_CD | X86_CR0_ET;
-       vcpu->srcu_idx = srcu_read_lock(&vcpu->kvm->srcu);
        vmx_set_cr0(&vmx->vcpu, kvm_read_cr0(vcpu)); /* enter rmode */
-       srcu_read_unlock(&vcpu->kvm->srcu, vcpu->srcu_idx);
        vmx_set_cr4(&vmx->vcpu, 0);
        vmx_set_efer(&vmx->vcpu, 0);
        vmx_fpu_activate(&vmx->vcpu);