KVM: VMX: Save/restore rflags.vm correctly in real mode

Currently we set eflags.vm unconditionally when entering real mode emulation
through virtual-8086 mode, and clear it unconditionally when we enter protected
mode.  The means that the following sequence

  KVM_SET_REGS  (rflags.vm=1)
  KVM_SET_SREGS (cr0.pe=1)

Ends up with rflags.vm clear due to KVM_SET_SREGS triggering enter_pmode().

Fix by shadowing rflags.vm (and rflags.iopl) correctly while in real mode:
reads and writes to those bits access a shadow register instead of the actual
register.

Signed-off-by: Avi Kivity <avi@redhat.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>

diff --git a/arch/x86/kvm/vmx.c b/arch/x86/kvm/vmx.c
index 686492ed3079b6a9ab101a2298f76e8483840f88..bc933cfb4e66d7fff8af14a8a4d8beac2ce6fed3 100644 (file)
--- a/arch/x86/kvm/vmx.c
+++ b/arch/x86/kvm/vmx.c
@@ -77,6 +77,8 @@ module_param(emulate_invalid_guest_state, bool, S_IRUGO);
 #define KVM_PMODE_VM_CR4_ALWAYS_ON (X86_CR4_PAE | X86_CR4_VMXE)
 #define KVM_RMODE_VM_CR4_ALWAYS_ON (X86_CR4_VME | X86_CR4_PAE | X86_CR4_VMXE)
 
+#define RMODE_GUEST_OWNED_EFLAGS_BITS (~(X86_EFLAGS_IOPL | X86_EFLAGS_VM))
+
 /*
  * These 2 parameters are used to config the controls for Pause-Loop Exiting:
  * ple_gap:    upper bound on the amount of time between two successive
@@ -131,7 +133,7 @@ struct vcpu_vmx {
        } host_state;
        struct {
                int vm86_active;
-               u8 save_iopl;
+               ulong save_rflags;
                struct kvm_save_segment {
                        u16 selector;
                        unsigned long base;
@@ -818,18 +820,23 @@ static void vmx_fpu_deactivate(struct kvm_vcpu *vcpu)
 
 static unsigned long vmx_get_rflags(struct kvm_vcpu *vcpu)
 {
-       unsigned long rflags;
+       unsigned long rflags, save_rflags;
 
        rflags = vmcs_readl(GUEST_RFLAGS);
-       if (to_vmx(vcpu)->rmode.vm86_active)
-               rflags &= ~(unsigned long)(X86_EFLAGS_IOPL | X86_EFLAGS_VM);
+       if (to_vmx(vcpu)->rmode.vm86_active) {
+               rflags &= RMODE_GUEST_OWNED_EFLAGS_BITS;
+               save_rflags = to_vmx(vcpu)->rmode.save_rflags;
+               rflags |= save_rflags & ~RMODE_GUEST_OWNED_EFLAGS_BITS;
+       }
        return rflags;
 }
 
 static void vmx_set_rflags(struct kvm_vcpu *vcpu, unsigned long rflags)
 {
-       if (to_vmx(vcpu)->rmode.vm86_active)
+       if (to_vmx(vcpu)->rmode.vm86_active) {
+               to_vmx(vcpu)->rmode.save_rflags = rflags;
                rflags |= X86_EFLAGS_IOPL | X86_EFLAGS_VM;
+       }
        vmcs_writel(GUEST_RFLAGS, rflags);
 }
 
@@ -1483,8 +1490,8 @@ static void enter_pmode(struct kvm_vcpu *vcpu)
        vmcs_write32(GUEST_TR_AR_BYTES, vmx->rmode.tr.ar);
 
        flags = vmcs_readl(GUEST_RFLAGS);
-       flags &= ~(X86_EFLAGS_IOPL | X86_EFLAGS_VM);
-       flags |= (vmx->rmode.save_iopl << IOPL_SHIFT);
+       flags &= RMODE_GUEST_OWNED_EFLAGS_BITS;
+       flags |= vmx->rmode.save_rflags & ~RMODE_GUEST_OWNED_EFLAGS_BITS;
        vmcs_writel(GUEST_RFLAGS, flags);
 
        vmcs_writel(GUEST_CR4, (vmcs_readl(GUEST_CR4) & ~X86_CR4_VME) |
@@ -1557,8 +1564,7 @@ static void enter_rmode(struct kvm_vcpu *vcpu)
        vmcs_write32(GUEST_TR_AR_BYTES, 0x008b);
 
        flags = vmcs_readl(GUEST_RFLAGS);
-       vmx->rmode.save_iopl
-               = (flags & X86_EFLAGS_IOPL) >> IOPL_SHIFT;
+       vmx->rmode.save_rflags = flags;
 
        flags |= X86_EFLAGS_IOPL | X86_EFLAGS_VM;