| Message ID | 1375701493-18751-1-git-send-email-yzt356@gmail.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
----- Original Message ----- > From: "Arthur Chunqi Li" <yzt356@gmail.com> > To: kvm@vger.kernel.org > Cc: "jan kiszka" <jan.kiszka@web.de>, gleb@redhat.com, pbonzini@redhat.com, "Arthur Chunqi Li" <yzt356@gmail.com> > Sent: Monday, August 5, 2013 1:18:13 PM > Subject: [PATCH v2] KVM: nVMX: Advertise IA32_PAT in VM exit control > > Advertise VM_EXIT_SAVE_IA32_PAT and VM_EXIT_LOAD_IA32_PAT. > > Signed-off-by: Arthur Chunqi Li <yzt356@gmail.com> > --- > > ChangeLog to v1: > Add host features check. Format is the same as "entry controls". > > arch/x86/kvm/vmx.c | 10 ++++++---- > 1 file changed, 6 insertions(+), 4 deletions(-) > > diff --git a/arch/x86/kvm/vmx.c b/arch/x86/kvm/vmx.c > index 396572d..fd451c1 100644 > --- a/arch/x86/kvm/vmx.c > +++ b/arch/x86/kvm/vmx.c > @@ -2191,14 +2191,16 @@ static __init void nested_vmx_setup_ctls_msrs(void) > * If bit 55 of VMX_BASIC is off, bits 0-8 and 10, 11, 13, 14, 16 and > * 17 must be 1. > */ > + rdmsr(MSR_IA32_VMX_EXIT_CTLS, > + nested_vmx_exit_ctls_low, nested_vmx_exit_ctls_high); > nested_vmx_exit_ctls_low = VM_EXIT_ALWAYSON_WITHOUT_TRUE_MSR; > + nested_vmx_exit_ctls_high &= > + VM_EXIT_LOAD_IA32_PAT | VM_EXIT_SAVE_IA32_PAT; > + nested_vmx_exit_ctls_high |= VM_EXIT_ALWAYSON_WITHOUT_TRUE_MSR; > /* Note that guest use of VM_EXIT_ACK_INTR_ON_EXIT is not supported. */ > #ifdef CONFIG_X86_64 > - nested_vmx_exit_ctls_high = VM_EXIT_HOST_ADDR_SPACE_SIZE; > -#else > - nested_vmx_exit_ctls_high = 0; > + nested_vmx_exit_ctls_high |= VM_EXIT_HOST_ADDR_SPACE_SIZE; I think you should clear the bit if !X86_64 (and not do anything if X86_64, the bit should be already set in MSR_IA32_VMX_EXIT_CTLS). Paolo > #endif > - nested_vmx_exit_ctls_high |= VM_EXIT_ALWAYSON_WITHOUT_TRUE_MSR; > > /* entry controls */ > rdmsr(MSR_IA32_VMX_ENTRY_CTLS, > -- > 1.7.9.5 > > -- To unsubscribe from this list: send the line "unsubscribe kvm" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Mon, Aug 05, 2013 at 08:03:31AM -0400, Paolo Bonzini wrote: > > > ----- Original Message ----- > > From: "Arthur Chunqi Li" <yzt356@gmail.com> > > To: kvm@vger.kernel.org > > Cc: "jan kiszka" <jan.kiszka@web.de>, gleb@redhat.com, pbonzini@redhat.com, "Arthur Chunqi Li" <yzt356@gmail.com> > > Sent: Monday, August 5, 2013 1:18:13 PM > > Subject: [PATCH v2] KVM: nVMX: Advertise IA32_PAT in VM exit control > > > > Advertise VM_EXIT_SAVE_IA32_PAT and VM_EXIT_LOAD_IA32_PAT. > > > > Signed-off-by: Arthur Chunqi Li <yzt356@gmail.com> > > --- > > > > ChangeLog to v1: > > Add host features check. Format is the same as "entry controls". > > > > arch/x86/kvm/vmx.c | 10 ++++++---- > > 1 file changed, 6 insertions(+), 4 deletions(-) > > > > diff --git a/arch/x86/kvm/vmx.c b/arch/x86/kvm/vmx.c > > index 396572d..fd451c1 100644 > > --- a/arch/x86/kvm/vmx.c > > +++ b/arch/x86/kvm/vmx.c > > @@ -2191,14 +2191,16 @@ static __init void nested_vmx_setup_ctls_msrs(void) > > * If bit 55 of VMX_BASIC is off, bits 0-8 and 10, 11, 13, 14, 16 and > > * 17 must be 1. > > */ > > + rdmsr(MSR_IA32_VMX_EXIT_CTLS, > > + nested_vmx_exit_ctls_low, nested_vmx_exit_ctls_high); > > nested_vmx_exit_ctls_low = VM_EXIT_ALWAYSON_WITHOUT_TRUE_MSR; > > + nested_vmx_exit_ctls_high &= > > + VM_EXIT_LOAD_IA32_PAT | VM_EXIT_SAVE_IA32_PAT; > > + nested_vmx_exit_ctls_high |= VM_EXIT_ALWAYSON_WITHOUT_TRUE_MSR; > > /* Note that guest use of VM_EXIT_ACK_INTR_ON_EXIT is not supported. */ > > #ifdef CONFIG_X86_64 > > - nested_vmx_exit_ctls_high = VM_EXIT_HOST_ADDR_SPACE_SIZE; > > -#else > > - nested_vmx_exit_ctls_high = 0; > > + nested_vmx_exit_ctls_high |= VM_EXIT_HOST_ADDR_SPACE_SIZE; > > I think you should clear the bit if !X86_64 (and not do anything if > X86_64, the bit should be already set in MSR_IA32_VMX_EXIT_CTLS). > Yes, we do not support 64 bit guests on 32bit host. It means that 32 bit host is essentially "processors that do not support Intel 64 architecture", so the control should be zero. -- Gleb. -- To unsubscribe from this list: send the line "unsubscribe kvm" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
diff --git a/arch/x86/kvm/vmx.c b/arch/x86/kvm/vmx.c index 396572d..fd451c1 100644 --- a/arch/x86/kvm/vmx.c +++ b/arch/x86/kvm/vmx.c @@ -2191,14 +2191,16 @@ static __init void nested_vmx_setup_ctls_msrs(void) * If bit 55 of VMX_BASIC is off, bits 0-8 and 10, 11, 13, 14, 16 and * 17 must be 1. */ + rdmsr(MSR_IA32_VMX_EXIT_CTLS, + nested_vmx_exit_ctls_low, nested_vmx_exit_ctls_high); nested_vmx_exit_ctls_low = VM_EXIT_ALWAYSON_WITHOUT_TRUE_MSR; + nested_vmx_exit_ctls_high &= + VM_EXIT_LOAD_IA32_PAT | VM_EXIT_SAVE_IA32_PAT; + nested_vmx_exit_ctls_high |= VM_EXIT_ALWAYSON_WITHOUT_TRUE_MSR; /* Note that guest use of VM_EXIT_ACK_INTR_ON_EXIT is not supported. */ #ifdef CONFIG_X86_64 - nested_vmx_exit_ctls_high = VM_EXIT_HOST_ADDR_SPACE_SIZE; -#else - nested_vmx_exit_ctls_high = 0; + nested_vmx_exit_ctls_high |= VM_EXIT_HOST_ADDR_SPACE_SIZE; #endif - nested_vmx_exit_ctls_high |= VM_EXIT_ALWAYSON_WITHOUT_TRUE_MSR; /* entry controls */ rdmsr(MSR_IA32_VMX_ENTRY_CTLS,
Advertise VM_EXIT_SAVE_IA32_PAT and VM_EXIT_LOAD_IA32_PAT. Signed-off-by: Arthur Chunqi Li <yzt356@gmail.com> --- ChangeLog to v1: Add host features check. Format is the same as "entry controls". arch/x86/kvm/vmx.c | 10 ++++++---- 1 file changed, 6 insertions(+), 4 deletions(-)