* [PATCH] KVM: nVMX: Set segment infomation of L1 when L2 exits
@ 2013-07-11 9:06 Arthur Chunqi Li
2013-07-11 9:43 ` Gleb Natapov
0 siblings, 1 reply; 3+ messages in thread
From: Arthur Chunqi Li @ 2013-07-11 9:06 UTC (permalink / raw)
To: kvm; +Cc: gleb, pbonzini, jan.kiszka, Arthur Chunqi Li
When L2 exits to L1, segment infomations of L1 are not set correctly.
According to Intel SDM 27.5.2(Loading Host Segment and Descriptor
Table Registers), segment base/limit/access right of L1 should be
set to some designed value when L2 exits to L1. This patch fixes
this.
Signed-off-by: Arthur Chunqi Li <yzt356@gmail.com>
---
arch/x86/kvm/vmx.c | 26 ++++++++++++++++++++++++++
1 file changed, 26 insertions(+)
diff --git a/arch/x86/kvm/vmx.c b/arch/x86/kvm/vmx.c
index 1200e4e..3cf702b 100644
--- a/arch/x86/kvm/vmx.c
+++ b/arch/x86/kvm/vmx.c
@@ -8028,6 +8028,32 @@ static void load_vmcs12_host_state(struct kvm_vcpu *vcpu,
vmcs_write16(GUEST_GS_SELECTOR, vmcs12->host_gs_selector);
vmcs_write16(GUEST_TR_SELECTOR, vmcs12->host_tr_selector);
+ /* Set L1 segment info according to Intel SDM
+ 27.5.2 Loading Host Segment and Descriptor-Table Registers */
+ vmcs_writel(GUEST_CS_BASE, 0x0);
+ vmcs_writel(GUEST_SS_BASE, 0x0);
+ vmcs_writel(GUEST_DS_BASE, 0x0);
+ vmcs_writel(GUEST_ES_BASE, 0x0);
+ vmcs_write32(GUEST_CS_LIMIT, 0xFFFFFFFF);
+ vmcs_write32(GUEST_SS_LIMIT, 0xFFFFFFFF);
+ vmcs_write32(GUEST_DS_LIMIT, 0xFFFFFFFF);
+ vmcs_write32(GUEST_ES_LIMIT, 0xFFFFFFFF);
+ vmcs_write32(GUEST_FS_LIMIT, 0xFFFFFFFF);
+ vmcs_write32(GUEST_GS_LIMIT, 0xFFFFFFFF);
+ vmcs_write32(GUEST_TR_LIMIT, 0x67);
+ vmcs_write32(GUEST_ES_AR_BYTES, 0xc093);
+ if (vmcs12->vm_exit_controls & VM_EXIT_HOST_ADDR_SPACE_SIZE)
+ vmcs_write32(GUEST_CS_AR_BYTES, 0xa09b);
+ else
+ vmcs_write32(GUEST_CS_AR_BYTES, 0xc09b);
+ vmcs_write32(GUEST_SS_AR_BYTES, 0xc093);
+ vmcs_write32(GUEST_DS_AR_BYTES, 0xc093);
+ vmcs_write32(GUEST_FS_AR_BYTES, 0xc093);
+ vmcs_write32(GUEST_GS_AR_BYTES, 0xc093);
+ vmcs_write32(GUEST_TR_AR_BYTES, 0x8b);
+ vmcs_writel(GUEST_RFLAGS, 0x2);
+
+
if (vmcs12->vm_exit_controls & VM_EXIT_LOAD_IA32_PAT)
vmcs_write64(GUEST_IA32_PAT, vmcs12->host_ia32_pat);
if (vmcs12->vm_exit_controls & VM_EXIT_LOAD_IA32_PERF_GLOBAL_CTRL)
--
1.7.9.5
^ permalink raw reply related [flat|nested] 3+ messages in thread
* Re: [PATCH] KVM: nVMX: Set segment infomation of L1 when L2 exits
2013-07-11 9:06 [PATCH] KVM: nVMX: Set segment infomation of L1 when L2 exits Arthur Chunqi Li
@ 2013-07-11 9:43 ` Gleb Natapov
2013-07-11 10:15 ` Arthur Chunqi Li
0 siblings, 1 reply; 3+ messages in thread
From: Gleb Natapov @ 2013-07-11 9:43 UTC (permalink / raw)
To: Arthur Chunqi Li; +Cc: kvm, pbonzini, jan.kiszka
On Thu, Jul 11, 2013 at 05:06:04PM +0800, Arthur Chunqi Li wrote:
> When L2 exits to L1, segment infomations of L1 are not set correctly.
> According to Intel SDM 27.5.2(Loading Host Segment and Descriptor
> Table Registers), segment base/limit/access right of L1 should be
> set to some designed value when L2 exits to L1. This patch fixes
> this.
>
Does this fix your test case?
> Signed-off-by: Arthur Chunqi Li <yzt356@gmail.com>
> ---
> arch/x86/kvm/vmx.c | 26 ++++++++++++++++++++++++++
> 1 file changed, 26 insertions(+)
>
> diff --git a/arch/x86/kvm/vmx.c b/arch/x86/kvm/vmx.c
> index 1200e4e..3cf702b 100644
> --- a/arch/x86/kvm/vmx.c
> +++ b/arch/x86/kvm/vmx.c
> @@ -8028,6 +8028,32 @@ static void load_vmcs12_host_state(struct kvm_vcpu *vcpu,
> vmcs_write16(GUEST_GS_SELECTOR, vmcs12->host_gs_selector);
> vmcs_write16(GUEST_TR_SELECTOR, vmcs12->host_tr_selector);
>
> + /* Set L1 segment info according to Intel SDM
> + 27.5.2 Loading Host Segment and Descriptor-Table Registers */
> + vmcs_writel(GUEST_CS_BASE, 0x0);
> + vmcs_writel(GUEST_SS_BASE, 0x0);
> + vmcs_writel(GUEST_DS_BASE, 0x0);
> + vmcs_writel(GUEST_ES_BASE, 0x0);
> + vmcs_write32(GUEST_CS_LIMIT, 0xFFFFFFFF);
> + vmcs_write32(GUEST_SS_LIMIT, 0xFFFFFFFF);
> + vmcs_write32(GUEST_DS_LIMIT, 0xFFFFFFFF);
> + vmcs_write32(GUEST_ES_LIMIT, 0xFFFFFFFF);
> + vmcs_write32(GUEST_FS_LIMIT, 0xFFFFFFFF);
> + vmcs_write32(GUEST_GS_LIMIT, 0xFFFFFFFF);
> + vmcs_write32(GUEST_TR_LIMIT, 0x67);
> + vmcs_write32(GUEST_ES_AR_BYTES, 0xc093);
> + if (vmcs12->vm_exit_controls & VM_EXIT_HOST_ADDR_SPACE_SIZE)
> + vmcs_write32(GUEST_CS_AR_BYTES, 0xa09b);
> + else
> + vmcs_write32(GUEST_CS_AR_BYTES, 0xc09b);
> + vmcs_write32(GUEST_SS_AR_BYTES, 0xc093);
> + vmcs_write32(GUEST_DS_AR_BYTES, 0xc093);
> + vmcs_write32(GUEST_FS_AR_BYTES, 0xc093);
> + vmcs_write32(GUEST_GS_AR_BYTES, 0xc093);
> + vmcs_write32(GUEST_TR_AR_BYTES, 0x8b);
Please use AR_ defines for AR_BYTES fields (AR_S_MASK etc).
> + vmcs_writel(GUEST_RFLAGS, 0x2);
> +
> +
> if (vmcs12->vm_exit_controls & VM_EXIT_LOAD_IA32_PAT)
> vmcs_write64(GUEST_IA32_PAT, vmcs12->host_ia32_pat);
> if (vmcs12->vm_exit_controls & VM_EXIT_LOAD_IA32_PERF_GLOBAL_CTRL)
> --
> 1.7.9.5
--
Gleb.
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: [PATCH] KVM: nVMX: Set segment infomation of L1 when L2 exits
2013-07-11 9:43 ` Gleb Natapov
@ 2013-07-11 10:15 ` Arthur Chunqi Li
0 siblings, 0 replies; 3+ messages in thread
From: Arthur Chunqi Li @ 2013-07-11 10:15 UTC (permalink / raw)
To: Gleb Natapov; +Cc: kvm, Paolo Bonzini, Jan Kiszka
On Thu, Jul 11, 2013 at 5:43 PM, Gleb Natapov <gleb@redhat.com> wrote:
> On Thu, Jul 11, 2013 at 05:06:04PM +0800, Arthur Chunqi Li wrote:
>> When L2 exits to L1, segment infomations of L1 are not set correctly.
>> According to Intel SDM 27.5.2(Loading Host Segment and Descriptor
>> Table Registers), segment base/limit/access right of L1 should be
>> set to some designed value when L2 exits to L1. This patch fixes
>> this.
>>
> Does this fix your test case?
Besides this bug, there are also some bugs in my test case. Now all
things run well.
>
>> Signed-off-by: Arthur Chunqi Li <yzt356@gmail.com>
>> ---
>> arch/x86/kvm/vmx.c | 26 ++++++++++++++++++++++++++
>> 1 file changed, 26 insertions(+)
>>
>> diff --git a/arch/x86/kvm/vmx.c b/arch/x86/kvm/vmx.c
>> index 1200e4e..3cf702b 100644
>> --- a/arch/x86/kvm/vmx.c
>> +++ b/arch/x86/kvm/vmx.c
>> @@ -8028,6 +8028,32 @@ static void load_vmcs12_host_state(struct kvm_vcpu *vcpu,
>> vmcs_write16(GUEST_GS_SELECTOR, vmcs12->host_gs_selector);
>> vmcs_write16(GUEST_TR_SELECTOR, vmcs12->host_tr_selector);
>>
>> + /* Set L1 segment info according to Intel SDM
>> + 27.5.2 Loading Host Segment and Descriptor-Table Registers */
>> + vmcs_writel(GUEST_CS_BASE, 0x0);
>> + vmcs_writel(GUEST_SS_BASE, 0x0);
>> + vmcs_writel(GUEST_DS_BASE, 0x0);
>> + vmcs_writel(GUEST_ES_BASE, 0x0);
>> + vmcs_write32(GUEST_CS_LIMIT, 0xFFFFFFFF);
>> + vmcs_write32(GUEST_SS_LIMIT, 0xFFFFFFFF);
>> + vmcs_write32(GUEST_DS_LIMIT, 0xFFFFFFFF);
>> + vmcs_write32(GUEST_ES_LIMIT, 0xFFFFFFFF);
>> + vmcs_write32(GUEST_FS_LIMIT, 0xFFFFFFFF);
>> + vmcs_write32(GUEST_GS_LIMIT, 0xFFFFFFFF);
>> + vmcs_write32(GUEST_TR_LIMIT, 0x67);
>> + vmcs_write32(GUEST_ES_AR_BYTES, 0xc093);
>> + if (vmcs12->vm_exit_controls & VM_EXIT_HOST_ADDR_SPACE_SIZE)
>> + vmcs_write32(GUEST_CS_AR_BYTES, 0xa09b);
>> + else
>> + vmcs_write32(GUEST_CS_AR_BYTES, 0xc09b);
>> + vmcs_write32(GUEST_SS_AR_BYTES, 0xc093);
>> + vmcs_write32(GUEST_DS_AR_BYTES, 0xc093);
>> + vmcs_write32(GUEST_FS_AR_BYTES, 0xc093);
>> + vmcs_write32(GUEST_GS_AR_BYTES, 0xc093);
>> + vmcs_write32(GUEST_TR_AR_BYTES, 0x8b);
> Please use AR_ defines for AR_BYTES fields (AR_S_MASK etc).
>
>> + vmcs_writel(GUEST_RFLAGS, 0x2);
>> +
>> +
>> if (vmcs12->vm_exit_controls & VM_EXIT_LOAD_IA32_PAT)
>> vmcs_write64(GUEST_IA32_PAT, vmcs12->host_ia32_pat);
>> if (vmcs12->vm_exit_controls & VM_EXIT_LOAD_IA32_PERF_GLOBAL_CTRL)
>> --
>> 1.7.9.5
>
> --
> Gleb.
--
Arthur Chunqi Li
Department of Computer Science
School of EECS
Peking University
Beijing, China
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2013-07-11 10:15 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2013-07-11 9:06 [PATCH] KVM: nVMX: Set segment infomation of L1 when L2 exits Arthur Chunqi Li
2013-07-11 9:43 ` Gleb Natapov
2013-07-11 10:15 ` Arthur Chunqi Li
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.