x86: sched: provide arch implementations using aperf/mperf

APERF/MPERF support for cpu_power.

APERF/MPERF is arch defined to be a relative scale of work capacity
per logical cpu, this is assumed to include SMT and Turbo mode.

APERF/MPERF are specified to both reset to 0 when either counter
wraps, which is highly inconvenient, since that'll give a blimp when
that happens. The manual specifies writing 0 to the counters after
each read, but that's 1) too expensive, and 2) destroys the
possibility of sharing these counters with other users, so we live
with the blimp - the other existing user does too.

Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Dinakar Guniguntala <dino@in.ibm.com>
Cc: John Stultz <johnstul@us.ibm.com>
Cc: Darren Hart <dvhltc@us.ibm.com>
Cc: John Kacur <jkacur@redhat.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>

3 files changed, 63 insertions, 1 deletions

diff --git a/arch/x86/kernel/cpu/Makefile b/arch/x86/kernel/cpu/Makefile
index c1f253dac155..8dd30638fe44 100644
--- a/arch/x86/kernel/cpu/Makefile
+++ b/arch/x86/kernel/cpu/Makefile
@@ -13,7 +13,7 @@ CFLAGS_common.o		:= $(nostackp)
 
 obj-y			:= intel_cacheinfo.o addon_cpuid_features.o
 obj-y			+= proc.o capflags.o powerflags.o common.o
-obj-y			+= vmware.o hypervisor.o
+obj-y			+= vmware.o hypervisor.o sched.o
 
 obj-$(CONFIG_X86_32)	+= bugs.o cmpxchg.o
 obj-$(CONFIG_X86_64)	+= bugs_64.o
diff --git a/arch/x86/kernel/cpu/sched.c b/arch/x86/kernel/cpu/sched.c
new file mode 100644
index 000000000000..2e59c7162222
--- /dev/null
+++ b/arch/x86/kernel/cpu/sched.c
@@ -0,0 +1,58 @@
+#include <linux/sched.h>
+#include <linux/math64.h>
+#include <linux/percpu.h>
+#include <linux/irqflags.h>
+
+#include <asm/cpufeature.h>
+#include <asm/processor.h>
+
+static DEFINE_PER_CPU(struct aperfmperf, old_aperfmperf);
+
+static unsigned long scale_aperfmperf(void)
+{
+	struct aperfmperf cur, val, *old = &__get_cpu_var(old_aperfmperf);
+	unsigned long ratio = SCHED_LOAD_SCALE;
+	unsigned long flags;
+
+	local_irq_save(flags);
+	get_aperfmperf(&val);
+	local_irq_restore(flags);
+
+	cur = val;
+	cur.aperf -= old->aperf;
+	cur.mperf -= old->mperf;
+	*old = val;
+
+	cur.mperf >>= SCHED_LOAD_SHIFT;
+	if (cur.mperf)
+		ratio = div_u64(cur.aperf, cur.mperf);
+
+	return ratio;
+}
+
+unsigned long arch_scale_freq_power(struct sched_domain *sd, int cpu)
+{
+	/*
+	 * do aperf/mperf on the cpu level because it includes things
+	 * like turbo mode, which are relevant to full cores.
+	 */
+	if (boot_cpu_has(X86_FEATURE_APERFMPERF))
+		return scale_aperfmperf();
+
+	/*
+	 * maybe have something cpufreq here
+	 */
+
+	return default_scale_freq_power(sd, cpu);
+}
+
+unsigned long arch_scale_smt_power(struct sched_domain *sd, int cpu)
+{
+	/*
+	 * aperf/mperf already includes the smt gain
+	 */
+	if (boot_cpu_has(X86_FEATURE_APERFMPERF))
+		return SCHED_LOAD_SCALE;
+
+	return default_scale_smt_power(sd, cpu);
+}
diff --git a/include/linux/sched.h b/include/linux/sched.h
index de9f89e184a4..c231a2467a83 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -1047,6 +1047,10 @@ partition_sched_domains(int ndoms_new, struct cpumask *doms_new,
 }
 #endif	/* !CONFIG_SMP */
 
+
+unsigned long default_scale_freq_power(struct sched_domain *sd, int cpu);
+unsigned long default_scale_smt_power(struct sched_domain *sd, int cpu);
+
 struct io_context;			/* See blkdev.h */