From mboxrd@z Thu Jan  1 00:00:00 1970
From: Paolo Bonzini <pbonzini@redhat.com>
Subject: Re: [PATCH v5] KVM: nVMX: Fully support of nested VMX preemption
 timer
Date: Thu, 26 Sep 2013 22:44:45 +0200
Message-ID: <52449CBD.4010603@redhat.com>
References: <1379319104-10266-1-git-send-email-yzt356@gmail.com> <52444CF6.1020102@redhat.com> <52446C8D.8030000@web.de> <52447335.9090601@redhat.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
Cc: Jan Kiszka <jan.kiszka@web.de>,
	Arthur Chunqi Li <yzt356@gmail.com>, kvm@vger.kernel.org,
	gleb@redhat.com, "Zhang, Yang Z" <yang.z.zhang@intel.com>
To: unlisted-recipients:; (no To-header on input)
Return-path: <kvm-owner@vger.kernel.org>
Received: from mail-qc0-f178.google.com ([209.85.216.178]:32841 "EHLO
	mail-qc0-f178.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org
	with ESMTP id S1752951Ab3IZUoa (ORCPT <rfc822;kvm@vger.kernel.org>);
	Thu, 26 Sep 2013 16:44:30 -0400
Received: by mail-qc0-f178.google.com with SMTP id r5so1140850qcx.37
        for <kvm@vger.kernel.org>; Thu, 26 Sep 2013 13:44:29 -0700 (PDT)
In-Reply-To: <52447335.9090601@redhat.com>
Sender: kvm-owner@vger.kernel.org
List-ID: <kvm.vger.kernel.org>

Il 26/09/2013 19:47, Paolo Bonzini ha scritto:
> 
> If I only apply this hunk, which disables the preemption timer while
> in L1:
> 
> @@ -8396,6 +8375,8 @@ static void nested_vmx_vmexit(struct kvm_vcpu *vcpu)
> 
>         load_vmcs12_host_state(vcpu, vmcs12);
> 
> +       vmcs_write32(PIN_BASED_VM_EXEC_CONTROL, vmx_pin_based_exec_ctrl(vmx));
> +
>         /* Update TSC_OFFSET if TSC was changed while L2 ran */
>         vmcs_write64(TSC_OFFSET, vmx->nested.vmcs01_tsc_offset);
> 
> then the testcase works for somewhat larger values of the preemption timer
> (up to ~1500000 TSC cycles), but then fails.

I mean if I apply it on top of current kvm/next, without Arthur's patch.

If I apply the hunk on top of Arthur's patch nothing changes and the
timer testcase starts breaking around ~65000 TSC cycles.

It is a bit problematic that adding printks changes something, so that
the test starts passing.  I haven't tried tracepoints yet.

Jan, which L1 is using the preemption timer?  Any reason why you added
it?  I wonder if it isn't better to revert it, since it is quite broken.

Paolo