intel_pstate: Take core C0 time into account for core busy calculation

Take non-idle time into account when calculating core busy time.
This ensures that intel_pstate will notice a decrease in load.

References: https://bugzilla.kernel.org/show_bug.cgi?id=66581
Cc: 3.10+ <stable@vger.kernel.org> # 3.10+
Signed-off-by: Dirk Brandewie <dirk.j.brandewie@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>

diff --git a/drivers/cpufreq/intel_pstate.c b/drivers/cpufreq/intel_pstate.c
index 7e257b2336025c3b2afd6faff9c0ae091fc7228f..79606f473f481b516069f66ce3c8a17ee60b053c 100644 (file)
--- a/drivers/cpufreq/intel_pstate.c
+++ b/drivers/cpufreq/intel_pstate.c
@@ -57,6 +57,7 @@ struct sample {
        int32_t core_pct_busy;
        u64 aperf;
        u64 mperf;
+       unsigned long long tsc;
        int freq;
 };
 
@@ -96,6 +97,7 @@ struct cpudata {
 
        u64     prev_aperf;
        u64     prev_mperf;
+       unsigned long long prev_tsc;
        int     sample_ptr;
        struct sample samples[SAMPLE_COUNT];
 };
@@ -548,30 +550,41 @@ static inline void intel_pstate_calc_busy(struct cpudata *cpu,
                                        struct sample *sample)
 {
        u64 core_pct;
-       core_pct = div64_u64(int_tofp(sample->aperf * 100),
-                            sample->mperf);
-       sample->freq = fp_toint(cpu->pstate.max_pstate * core_pct * 1000);
+       u64 c0_pct;
 
-       sample->core_pct_busy = core_pct;
+       core_pct = div64_u64(sample->aperf * 100, sample->mperf);
+
+       c0_pct = div64_u64(sample->mperf * 100, sample->tsc);
+       sample->freq = fp_toint(
+               mul_fp(int_tofp(cpu->pstate.max_pstate),
+                       int_tofp(core_pct * 1000)));
+
+       sample->core_pct_busy = mul_fp(int_tofp(core_pct),
+                               div_fp(int_tofp(c0_pct + 1), int_tofp(100)));
 }
 
 static inline void intel_pstate_sample(struct cpudata *cpu)
 {
        u64 aperf, mperf;
+       unsigned long long tsc;
 
        rdmsrl(MSR_IA32_APERF, aperf);
        rdmsrl(MSR_IA32_MPERF, mperf);
+       tsc = native_read_tsc();
 
        cpu->sample_ptr = (cpu->sample_ptr + 1) % SAMPLE_COUNT;
        cpu->samples[cpu->sample_ptr].aperf = aperf;
        cpu->samples[cpu->sample_ptr].mperf = mperf;
+       cpu->samples[cpu->sample_ptr].tsc = tsc;
        cpu->samples[cpu->sample_ptr].aperf -= cpu->prev_aperf;
        cpu->samples[cpu->sample_ptr].mperf -= cpu->prev_mperf;
+       cpu->samples[cpu->sample_ptr].tsc -= cpu->prev_tsc;
 
        intel_pstate_calc_busy(cpu, &cpu->samples[cpu->sample_ptr]);
 
        cpu->prev_aperf = aperf;
        cpu->prev_mperf = mperf;
+       cpu->prev_tsc = tsc;
 }
 
 static inline void intel_pstate_set_sample_time(struct cpudata *cpu)