intel_pstate: Add X86_FEATURE_APERFMPERF to cpu match parameters.
authorDirk Brandewie <dirk.j.brandewie@intel.com>
Mon, 6 Jan 2014 18:59:16 +0000 (10:59 -0800)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Wed, 15 Jan 2014 23:31:43 +0000 (15:31 -0800)
commit 6cbd7ee10e2842a3d1f9b60abede1c8f3d1f1130 upstream.

KVM environments do not support APERF/MPERF MSRs. intel_pstate cannot
operate without these registers.

The previous validity checks in intel_pstate_msrs_not_valid() are
insufficent in nested KVMs.

References: https://bugzilla.redhat.com/show_bug.cgi?id=1046317
Signed-off-by: Dirk Brandewie <dirk.j.brandewie@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
drivers/cpufreq/intel_pstate.c

index 99d8ab5..d5dc567 100644 (file)
@@ -518,7 +518,8 @@ static void intel_pstate_timer_func(unsigned long __data)
 }
 
 #define ICPU(model, policy) \
-       { X86_VENDOR_INTEL, 6, model, X86_FEATURE_ANY, (unsigned long)&policy }
+       { X86_VENDOR_INTEL, 6, model, X86_FEATURE_APERFMPERF,\
+                       (unsigned long)&policy }
 
 static const struct x86_cpu_id intel_pstate_cpu_ids[] = {
        ICPU(0x2a, default_policy),