Re: [PATCH] VMX: Return to userspace on invalid state emulation failure

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On 08/26/2009 01:07 PM, Mohammed Gamal wrote:
On Wed, Aug 26, 2009 at 12:02 PM, Avi Kivity<avi@xxxxxxxxxx>  wrote:
On 08/25/2009 01:37 AM, Mohammed Gamal wrote:
Return to userspace instead of repeatedly trying to emulate
instructions that have already failed

Signed-off-by: Mohammed Gamal<m.gamal005@xxxxxxxxx>
---
  arch/x86/kvm/vmx.c |    6 +++++-
  1 files changed, 5 insertions(+), 1 deletions(-)

diff --git a/arch/x86/kvm/vmx.c b/arch/x86/kvm/vmx.c
index 6b57eed..c559bb7 100644
--- a/arch/x86/kvm/vmx.c
+++ b/arch/x86/kvm/vmx.c
@@ -3337,6 +3337,8 @@ static void handle_invalid_guest_state(struct
kvm_vcpu *vcpu)

                if (err != EMULATE_DONE) {
                        kvm_report_emulation_failure(vcpu, "emulation
failure");
+                       vcpu->run->exit_reason = KVM_EXIT_INTERNAL_ERROR;
+                       vcpu->run->internal.suberror =
KVM_INTERNAL_ERROR_EMULATION;
                        break;
                }

@@ -3607,7 +3609,9 @@ static void vmx_vcpu_run(struct kvm_vcpu *vcpu)
                vmx->entry_time = ktime_get();

        /* Handle invalid guest state instead of entering VMX */
-       if (vmx->emulation_required&&    emulate_invalid_guest_state) {
+       if (vmx->emulation_required&&    emulate_invalid_guest_state
+&&    !(vcpu->run->exit_reason == KVM_EXIT_INTERNAL_ERROR&&
+                 vcpu->run->internal.suberror ==
KVM_INTERNAL_ERROR_EMULATION)) {
                handle_invalid_guest_state(vcpu);
                return;
        }

Still suffers from the same problem.  You don't always update
vcpu->run->exit_reason, so you can't test it.  Best to return a value from
handle_invalid_guest_state() (the standard return codes for exit handlers
are 1 for return-to-guest, 0 for return-to-host, and -errno to return with
an error).

I was thinking of the same idea since I was also concerned about
vcpu->run->exit_reason not being updated. But how can we interpret the
return values of handle_invalid_guest_state() inside vmx_vcpu_run()
since it doesn't have a return value. Or would it be better to move
handle_invalid_guest_state() to the standard vmx exit handlers?

We can move the call to vmx_handle_exit(). We have a check for emulate_invalid_guest_state there anyway. I don't think it should be a standard exit handler since there is no exit_reason for it.

--
error compiling committee.c: too many arguments to function

--
To unsubscribe from this list: send the line "unsubscribe kvm" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[Index of Archives]     [KVM ARM]     [KVM ia64]     [KVM ppc]     [Virtualization Tools]     [Spice Development]     [Libvirt]     [Libvirt Users]     [Linux USB Devel]     [Linux Audio Users]     [Yosemite Questions]     [Linux Kernel]     [Linux SCSI]     [XFree86]
  Powered by Linux