On 07/19/2010 06:31 PM, Gleb Natapov wrote:
If guest indicates that it can handle async pf in kernel mode too send
it, but only if interrupt are enabled.
Reviewed-by: Rik van Riel<r...@redhat.com>
Signed-off-by: Gleb Natapov<g...@redhat.com>
---
arch/x86/kvm/mmu.c | 8 +++++++-
1 files changed, 7 insertions(+), 1 deletions(-)
diff --git a/arch/x86/kvm/mmu.c b/arch/x86/kvm/mmu.c
index 12d1a7b..ed87b1c 100644
--- a/arch/x86/kvm/mmu.c
+++ b/arch/x86/kvm/mmu.c
@@ -2361,7 +2361,13 @@ static bool can_do_async_pf(struct kvm_vcpu *vcpu)
if (!vcpu->arch.apf_data || kvm_event_needs_reinjection(vcpu))
return false;
- return !!kvm_x86_ops->get_cpl(vcpu);
+ if (vcpu->arch.apf_send_user_only)
+ return !!kvm_x86_ops->get_cpl(vcpu);
cpl is not a bool. Compare it with 0.
+
+ if (!kvm_x86_ops->interrupt_allowed(vcpu))
+ return false;
+
+ return true;
}
Should have commented before, but get_cpl() is not accurate when doing
nested virtualization. When L1 intercepts page faults, being in L2 is
equivalent to CPL 3. But we need to get the apf information to L1 somehow.
--
error compiling committee.c: too many arguments to function
--
To unsubscribe from this list: send the line "unsubscribe kvm" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html