As Avi recently mentioned, the new standard mechanism for exposing features
is KVM_GET_SUPPORTED_CPUID, not spanning CAPs. For some reason async pf missed that.
So expose async_pf here.
Signed-off-by: Glauber Costa <[email protected]>
CC: Gleb Natapov <[email protected]>
CC: Avi Kivity <[email protected]>
---
arch/x86/kvm/x86.c | 1 +
1 files changed, 1 insertions(+), 0 deletions(-)
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index f1e4025..e6a1cd8 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -2418,6 +2418,7 @@ static void do_cpuid_ent(struct kvm_cpuid_entry2 *entry, u32 function,
entry->eax = (1 << KVM_FEATURE_CLOCKSOURCE) |
(1 << KVM_FEATURE_NOP_IO_DELAY) |
(1 << KVM_FEATURE_CLOCKSOURCE2) |
+ (1 << KVM_FEATURE_ASYNC_PF) |
(1 << KVM_FEATURE_CLOCKSOURCE_STABLE_BIT);
entry->ebx = 0;
entry->ecx = 0;
--
1.7.2.3
On 03/23/2011 06:40 PM, Glauber Costa wrote:
> As Avi recently mentioned, the new standard mechanism for exposing features
> is KVM_GET_SUPPORTED_CPUID, not spanning CAPs. For some reason async pf missed that.
>
> So expose async_pf here.
>
Applied, thanks.
--
error compiling committee.c: too many arguments to function