On 9/25/15 12:12 AM, Bandan Das wrote:
Wanpeng Li <wanpeng.li@xxxxxxxxxxx> writes:
Introduce __vmx_flush_tlb() to handle specific vpid.
Signed-off-by: Wanpeng Li <wanpeng.li@xxxxxxxxxxx>
---
arch/x86/kvm/vmx.c | 21 +++++++++++++--------
1 file changed, 13 insertions(+), 8 deletions(-)
diff --git a/arch/x86/kvm/vmx.c b/arch/x86/kvm/vmx.c
index 794c529..7188c5e 100644
--- a/arch/x86/kvm/vmx.c
+++ b/arch/x86/kvm/vmx.c
@@ -1343,13 +1343,13 @@ static void loaded_vmcs_clear(struct loaded_vmcs *loaded_vmcs)
__loaded_vmcs_clear, loaded_vmcs, 1);
}
-static inline void vpid_sync_vcpu_single(struct vcpu_vmx *vmx)
+static inline void vpid_sync_vcpu_single(int vpid)
{
- if (vmx->vpid == 0)
+ if (vpid == 0)
return;
if (cpu_has_vmx_invvpid_single())
- __invvpid(VMX_VPID_EXTENT_SINGLE_CONTEXT, vmx->vpid, 0);
+ __invvpid(VMX_VPID_EXTENT_SINGLE_CONTEXT, vpid, 0);
}
static inline void vpid_sync_vcpu_global(void)
@@ -1358,10 +1358,10 @@ static inline void vpid_sync_vcpu_global(void)
__invvpid(VMX_VPID_EXTENT_ALL_CONTEXT, 0, 0);
}
-static inline void vpid_sync_context(struct vcpu_vmx *vmx)
+static inline void vpid_sync_context(int vpid)
{
if (cpu_has_vmx_invvpid_single())
- vpid_sync_vcpu_single(vmx);
+ vpid_sync_vcpu_single(vpid);
else
vpid_sync_vcpu_global();
}
Not sure myself what's the right thing to do but this may be undesirable
in a nested environment. Assuming the processor supports global invalidation
only, this seems like a easy way for the nested guest to invalidate *all*
mappings - even the L1 specific mappings.
Indeed, however, there's no easy way to handle the w/o single
invalidation case, we can improve it if you have any idea, otherwise, it
can be left to further optimization. :-)
Regards,
Wanpeng Li
--
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
