cpuidle-haltpoll: Enable kvm guest polling when dedicated physical CPUs are available

mainline inclusion
from mainline-5.4
commit 1328edca
category: feature
bugzilla: https://bugzilla.openeuler.org/show_bug.cgi?id=34


CVE: NA

The downside of guest side polling is that polling is performed even
with other runnable tasks in the host. However, even if poll in kvm
can aware whether or not other runnable tasks in the same pCPU, it
can still incur extra overhead in over-subscribe scenario. Now we can
just enable guest polling when dedicated pCPUs are available.

Signed-off-by: Wanpeng Li <wanpengli@tencent.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Yubo Miao <miaoyubo@huawei.com>
Signed-off-by: Xiangyou Xie <xiexiangyou@huawei.com>
Reviewed-by: Hailiang Zhang <zhang.zhanghailiang@huawei.com>
Reviewed-by: Hanjun Guo <guohanjun@huawei.com>
Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Signed-off-by: Jiajun Chen <chenjiajun8@huawei.com>
Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>

arch/x86/kernel/kvm.c

@@ -703,6 +703,7 @@ unsigned int kvm_arch_para_hints(void)
 {
 	return cpuid_edx(kvm_cpuid_base() | KVM_CPUID_FEATURES);
 }
+EXPORT_SYMBOL_GPL(kvm_arch_para_hints);
 
 static uint32_t __init kvm_detect(void)
 {

drivers/cpuidle/cpuidle-haltpoll.c

@@ -97,7 +97,8 @@ static int __init haltpoll_init(void)
 
 	cpuidle_poll_state_init(drv);
 
-	if (!kvm_para_available())
+	if (!kvm_para_available() ||
+		!kvm_para_has_hint(KVM_HINTS_REALTIME))
 		return -ENODEV;
 
 	ret = cpuidle_register_driver(drv);