From: Xiaoyao Li <xiaoyao.li@linux.intel.com>
To: Paolo Bonzini <pbonzini@redhat.com>
Cc: linux-kernel <linux-kernel@vger.kernel.org>, x86 <x86@kernel.org>,
kvm@vger.kernel.org
Subject: Re: [PATCH v4 17/17] kvm: vmx: Emulate TEST_CTL MSR
Date: Mon, 11 Mar 2019 23:10:58 +0800 [thread overview]
Message-ID: <b7a879103eb409bea3f2664d7e7a8b1787768cf2.camel@linux.intel.com> (raw)
In-Reply-To: <7a10bb11-e9bf-f49c-6575-25c3da08cfac@redhat.com>
On Mon, 2019-03-11 at 14:31 +0100, Paolo Bonzini wrote:
> On 09/03/19 03:31, Xiaoyao Li wrote:
> > Hi, Paolo,
> >
> > Do you have any comments on this patch?
> >
> > We are preparing v5 patches for split lock detection, if you have any
> > comments
> > about this one, please let me know.
>
> No, my only comment is that it should be placed _before_ the other two
> for bisectability. I think I have already sent that small remark.
>
> Thanks,
>
> Paolo
I cannot find the small remark you sent before. Maybe I missed something.
But I'am confused why it should be placed _before_ the other two. This patch
just use the vmx->core_capability that defined it the previous patch.
> > Thanks,
> > Xiaoyao
> >
> > On Fri, 2019-03-01 at 18:45 -0800, Fenghua Yu wrote:
> > > From: Xiaoyao Li <xiaoyao.li@linux.intel.com>
> > >
> > > A control bit (bit 29) in TEST_CTL MSR 0x33 will be introduced in
> > > future x86 processors. When bit 29 is set, the processor causes #AC
> > > exception for split locked accesses at all CPL.
> > >
> > > Please check the latest Intel Software Developer's Manual
> > > for more detailed information on the MSR and the split lock bit.
> > >
> > > 1. Since the kernel chooses to enable AC split lock by default, which
> > > means if we don't emulate TEST_CTL MSR for guest, guest will run with
> > > this feature enable while does not known it. Thus existing guests with
> > > buggy firmware (like OVMF) and old kernels having the cross cache line
> > > issues will fail the boot due to #AC.
> > >
> > > So we should emulate TEST_CTL MSR, and set it zero to disable AC split
> > > lock by default. Whether and when to enable it is left to guest firmware
> > > and guest kernel.
> > >
> > > 2. Host and guest can enable AC split lock independently, so using
> > > msr autoload to switch it during VM entry/exit.
> > >
> > > Signed-off-by: Xiaoyao Li <xiaoyao.li@linux.intel.com>
> > > Signed-off-by: Fenghua Yu <fenghua.yu@intel.com>
> > > ---
> > > arch/x86/kvm/vmx/vmx.c | 35 +++++++++++++++++++++++++++++++++++
> > > arch/x86/kvm/vmx/vmx.h | 1 +
> > > 2 files changed, 36 insertions(+)
> > >
> > > diff --git a/arch/x86/kvm/vmx/vmx.c b/arch/x86/kvm/vmx/vmx.c
> > > index 3e03c6e1e558..c0c5e8621afa 100644
> > > --- a/arch/x86/kvm/vmx/vmx.c
> > > +++ b/arch/x86/kvm/vmx/vmx.c
> > > @@ -1659,6 +1659,12 @@ static int vmx_get_msr(struct kvm_vcpu *vcpu,
> > > struct
> > > msr_data *msr_info)
> > > u32 index;
> > >
> > > switch (msr_info->index) {
> > > + case MSR_TEST_CTL:
> > > + if (!msr_info->host_initiated &&
> > > + !(vmx->core_capability & CORE_CAP_SPLIT_LOCK_DETECT))
> > > + return 1;
> > > + msr_info->data = vmx->msr_test_ctl;
> > > + break;
> > > #ifdef CONFIG_X86_64
> > > case MSR_FS_BASE:
> > > msr_info->data = vmcs_readl(GUEST_FS_BASE);
> > > @@ -1805,6 +1811,14 @@ static int vmx_set_msr(struct kvm_vcpu *vcpu,
> > > struct
> > > msr_data *msr_info)
> > > u32 index;
> > >
> > > switch (msr_index) {
> > > + case MSR_TEST_CTL:
> > > + if (!(vmx->core_capability & CORE_CAP_SPLIT_LOCK_DETECT))
> > > + return 1;
> > > +
> > > + if (data & ~TEST_CTL_ENABLE_SPLIT_LOCK_DETECT)
> > > + return 1;
> > > + vmx->msr_test_ctl = data;
> > > + break;
> > > case MSR_EFER:
> > > ret = kvm_set_msr_common(vcpu, msr_info);
> > > break;
> > > @@ -4108,6 +4122,9 @@ static void vmx_vcpu_setup(struct vcpu_vmx *vmx)
> > >
> > > vmx->arch_capabilities = kvm_get_arch_capabilities();
> > >
> > > + /* disable AC split lock by default */
> > > + vmx->msr_test_ctl = 0;
> > > +
> > > vm_exit_controls_init(vmx, vmx_vmexit_ctrl());
> > >
> > > /* 22.2.1, 20.8.1 */
> > > @@ -4145,6 +4162,7 @@ static void vmx_vcpu_reset(struct kvm_vcpu *vcpu,
> > > bool
> > > init_event)
> > >
> > > vmx->rmode.vm86_active = 0;
> > > vmx->spec_ctrl = 0;
> > > + vmx->msr_test_ctl = 0;
> > >
> > > vcpu->arch.microcode_version = 0x100000000ULL;
> > > vmx->vcpu.arch.regs[VCPU_REGS_RDX] = get_rdx_init_val();
> > > @@ -6344,6 +6362,21 @@ static void atomic_switch_perf_msrs(struct vcpu_vmx
> > > *vmx)
> > > msrs[i].host, false);
> > > }
> > >
> > > +static void atomic_switch_msr_test_ctl(struct vcpu_vmx *vmx)
> > > +{
> > > + u64 host_msr_test_ctl;
> > > +
> > > + if (!boot_cpu_has(X86_FEATURE_SPLIT_LOCK_DETECT))
> > > + return;
> > > +
> > > + rdmsrl(MSR_TEST_CTL, host_msr_test_ctl);
> > > + if (host_msr_test_ctl == vmx->msr_test_ctl)
> > > + clear_atomic_switch_msr(vmx, MSR_TEST_CTL);
> > > + else
> > > + add_atomic_switch_msr(vmx, MSR_TEST_CTL, vmx->msr_test_ctl,
> > > + host_msr_test_ctl, false);
> > > +}
> > > +
> > > static void vmx_arm_hv_timer(struct vcpu_vmx *vmx, u32 val)
> > > {
> > > vmcs_write32(VMX_PREEMPTION_TIMER_VALUE, val);
> > > @@ -6585,6 +6618,8 @@ static void vmx_vcpu_run(struct kvm_vcpu *vcpu)
> > >
> > > atomic_switch_perf_msrs(vmx);
> > >
> > > + atomic_switch_msr_test_ctl(vmx);
> > > +
> > > vmx_update_hv_timer(vcpu);
> > >
> > > /*
> > > diff --git a/arch/x86/kvm/vmx/vmx.h b/arch/x86/kvm/vmx/vmx.h
> > > index cc22379991f3..e8831609c6c3 100644
> > > --- a/arch/x86/kvm/vmx/vmx.h
> > > +++ b/arch/x86/kvm/vmx/vmx.h
> > > @@ -191,6 +191,7 @@ struct vcpu_vmx {
> > > u64 msr_guest_kernel_gs_base;
> > > #endif
> > >
> > > + u64 msr_test_ctl;
> > > u64 core_capability;
> > > u64 arch_capabilities;
> > > u64 spec_ctrl;
>
>
next prev parent reply other threads:[~2019-03-11 15:14 UTC|newest]
Thread overview: 65+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-03-02 2:44 [PATCH v4 00/17] x86/split_lock: Enable #AC exception for split locked accesses Fenghua Yu
2019-03-02 2:44 ` [PATCH v4 01/17] x86/common: Align cpu_caps_cleared and cpu_caps_set to unsigned long Fenghua Yu
2019-03-04 8:33 ` Paolo Bonzini
2019-03-04 10:17 ` Peter Zijlstra
2019-03-04 10:48 ` Paolo Bonzini
2019-03-04 12:44 ` Peter Zijlstra
2019-03-04 13:13 ` Paolo Bonzini
2019-03-02 2:44 ` [PATCH v4 02/17] drivers/net/b44: Align pwol_mask to unsigned long for better performance Fenghua Yu
2019-03-04 10:00 ` Peter Zijlstra
2019-03-04 14:45 ` Fenghua Yu
2019-03-04 15:27 ` Peter Zijlstra
2019-03-02 2:44 ` [PATCH v4 03/17] wlcore: Align reg_ch_conf_pending and tmp_ch_bitmap " Fenghua Yu
2019-03-04 10:11 ` Peter Zijlstra
2019-03-04 10:46 ` Paolo Bonzini
2019-03-04 12:41 ` Peter Zijlstra
2019-03-04 13:09 ` Paolo Bonzini
2019-03-04 13:30 ` Peter Zijlstra
2019-03-04 14:40 ` Fenghua Yu
2019-03-04 15:54 ` Paolo Bonzini
2019-03-02 2:44 ` [PATCH v4 04/17] x86/split_lock: Align x86_capability to unsigned long to avoid split locked access Fenghua Yu
2019-03-04 18:52 ` Dave Hansen
2019-03-04 19:15 ` Fenghua Yu
2019-03-04 19:29 ` Dave Hansen
2019-03-04 20:08 ` Peter Zijlstra
2019-03-04 20:12 ` Peter Zijlstra
2019-03-02 2:44 ` [PATCH v4 05/17] x86/cpufeatures: Enumerate IA32_CORE_CAPABILITIES MSR Fenghua Yu
2019-03-04 18:53 ` Dave Hansen
2019-03-04 18:55 ` Yu, Fenghua
2019-03-04 19:01 ` Dave Hansen
2019-03-02 2:45 ` [PATCH v4 06/17] x86/msr-index: Define IA32_CORE_CAPABILITY MSR and #AC exception for split lock bit Fenghua Yu
2019-03-02 2:45 ` [PATCH v4 07/17] x86/split_lock: Enumerate #AC for split lock by MSR IA32_CORE_CAPABILITY Fenghua Yu
2019-03-04 18:58 ` Dave Hansen
2019-03-04 18:59 ` Fenghua Yu
2019-03-04 19:19 ` Dave Hansen
2019-03-02 2:45 ` [PATCH v4 08/17] x86/clearcpuid: Support multiple clearcpuid options Fenghua Yu
2019-03-02 2:45 ` [PATCH v4 09/17] x86/clearcpuid: Support feature flag string in kernel option clearcpuid Fenghua Yu
2019-03-02 2:45 ` [PATCH v4 10/17] x86/clearcpuid: Apply cleared feature bits that are forced set before Fenghua Yu
2019-03-02 2:45 ` [PATCH v4 11/17] x86/clearcpuid: Clear CPUID bit in CPUID faulting Fenghua Yu
2019-03-04 22:04 ` kbuild test robot
2019-03-05 7:27 ` kbuild test robot
2019-03-02 2:45 ` [PATCH v4 12/17] Change document for kernel option clearcpuid Fenghua Yu
2019-03-02 2:45 ` [PATCH v4 13/17] x86/split_lock: Handle #AC exception for split lock Fenghua Yu
2019-03-02 2:45 ` [PATCH v4 14/17] x86/split_lock: Add a sysfs interface to allow user to enable or disable split lock detection on all CPUs during run time Fenghua Yu
2019-03-02 2:45 ` [PATCH v4 15/17] kvm: x86: Report CORE_CAPABILITY on GET_SUPPORTED_CPUID Fenghua Yu
2019-03-04 8:38 ` Paolo Bonzini
2019-03-04 10:47 ` Xiaoyao Li
2019-03-04 10:49 ` Paolo Bonzini
2019-03-04 11:10 ` Xiaoyao Li
2019-03-04 11:14 ` Paolo Bonzini
2019-03-04 11:21 ` Xiaoyao Li
2019-03-05 7:03 ` Xiaoyao Li
2019-03-02 2:45 ` [PATCH v4 16/17] kvm: x86: Add support IA32_CORE_CAPABILITY MSR Fenghua Yu
2019-03-04 8:42 ` Paolo Bonzini
2019-03-04 12:32 ` Xiaoyao Li
2019-03-08 6:10 ` Xiaoyao Li
2019-03-08 7:54 ` Paolo Bonzini
2019-03-08 8:03 ` Xiaoyao Li
2019-03-02 2:45 ` [PATCH v4 17/17] kvm: vmx: Emulate TEST_CTL MSR Fenghua Yu
2019-03-09 2:31 ` Xiaoyao Li
2019-03-11 13:31 ` Paolo Bonzini
2019-03-11 15:10 ` Xiaoyao Li [this message]
2019-03-11 15:21 ` Paolo Bonzini
2019-03-11 16:58 ` Xiaoyao Li
2019-03-04 21:55 ` [PATCH v4 00/17] x86/split_lock: Enable #AC exception for split locked accesses Konrad Rzeszutek Wilk
2019-03-05 0:06 ` Fenghua Yu
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=b7a879103eb409bea3f2664d7e7a8b1787768cf2.camel@linux.intel.com \
--to=xiaoyao.li@linux.intel.com \
--cc=kvm@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=pbonzini@redhat.com \
--cc=x86@kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).