x86/timer: Force PIT initialization when !X86_FEATURE_ARAT

KVM guests with commit 4a08095c7df8 ("x86/timer: Skip PIT initialization on
modern chipsets") applied to guest kernel have been observed to have
unusually higher CPU usage with symptoms of increase in vm exits for HLT
and MSW_WRITE (MSR_IA32_TSCDEADLINE).

This is caused by older QEMUs lacking support for X86_FEATURE_ARAT.  lapic
clock retains CLOCK_EVT_FEAT_C3STOP and nohz stays inactive.  There's no
usable broadcast device either.

Do the PIT initialization if guest CPU lacks X86_FEATURE_ARAT.  On real
hardware it shouldn't matter as ARAT and DEADLINE come together.

Fixes: 4a08095c7df8 ("x86/timer: Skip PIT initialization on modern chipsets")
Signed-off-by: Jan Stancek <jstancek@redhat.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>

diff --git a/arch/x86/kernel/apic/apic.c b/arch/x86/kernel/apic/apic.c
index dba2828b779a57abe32892600be3fdc021a5254a..f91b3ff9dc03b98adb1096d0d6f69c3631f5a7e6 100644 (file)
--- a/arch/x86/kernel/apic/apic.c
+++ b/arch/x86/kernel/apic/apic.c
@@ -834,6 +834,10 @@ bool __init apic_needs_pit(void)
        if (!boot_cpu_has(X86_FEATURE_APIC))
                return true;
 
+       /* Virt guests may lack ARAT, but still have DEADLINE */
+       if (!boot_cpu_has(X86_FEATURE_ARAT))
+               return true;
+
        /* Deadline timer is based on TSC so no further PIT action required */
        if (boot_cpu_has(X86_FEATURE_TSC_DEADLINE_TIMER))
                return false;