x86/cpuidle: get accurate C0 value with xenpm tool

When checking the ACPI funciton of C-status, after 100 seconds sleep,
the sampling value of C0 status from the xenpm tool decreases.
Because C0=NOW()-C1-C2-C3-C4, when NOW() value is during idle time,
NOW() value is bigger than last C-status update time, and C0 value
is also bigger than ture value. if margin of the second error cannot
make up for margin of the first error, the value of C0 would decrease.

Signed-off-by: Huaitong Han <huaitong.han@intel.com>
Reviewed-by: Jan Beulich <jbeulich@suse.com>

diff --git a/xen/arch/x86/acpi/cpu_idle.c b/xen/arch/x86/acpi/cpu_idle.c
index d6613c915c793885a6c26594f3bc802fb7f93522..95676dd7941bd9f45e8e8899ffaef22cffbe6d79 100644 (file)
--- a/xen/arch/x86/acpi/cpu_idle.c
+++ b/xen/arch/x86/acpi/cpu_idle.c
@@ -253,11 +253,32 @@ static char* acpi_cstate_method_name[] =
     "HALT"
 };
 
+static uint64_t get_stime_tick(void) { return (uint64_t)NOW(); }
+static uint64_t stime_ticks_elapsed(uint64_t t1, uint64_t t2) { return t2 - t1; }
+static uint64_t stime_tick_to_ns(uint64_t ticks) { return ticks; }
+
+static uint64_t get_acpi_pm_tick(void) { return (uint64_t)inl(pmtmr_ioport); }
+static uint64_t acpi_pm_ticks_elapsed(uint64_t t1, uint64_t t2)
+{
+    if ( t2 >= t1 )
+        return (t2 - t1);
+    else if ( !(acpi_gbl_FADT.flags & ACPI_FADT_32BIT_TIMER) )
+        return (((0x00FFFFFF - t1) + t2 + 1) & 0x00FFFFFF);
+    else
+        return ((0xFFFFFFFF - t1) + t2 +1);
+}
+
+uint64_t (*__read_mostly cpuidle_get_tick)(void) = get_acpi_pm_tick;
+static uint64_t (*__read_mostly ticks_elapsed)(uint64_t, uint64_t)
+    = acpi_pm_ticks_elapsed;
+
 static void print_acpi_power(uint32_t cpu, struct acpi_processor_power *power)
 {
-    uint32_t i, idle_usage = 0;
-    uint64_t res, idle_res = 0;
-    u32 usage;
+    uint64_t idle_res = 0, idle_usage = 0;
+    uint64_t last_state_update_tick, current_tick, current_stime;
+    uint64_t usage[ACPI_PROCESSOR_MAX_POWER] = { 0 };
+    uint64_t res_tick[ACPI_PROCESSOR_MAX_POWER] = { 0 };
+    unsigned int i;
     u8 last_state_idx;
 
     printk("==cpu%d==\n", cpu);
@@ -265,28 +286,37 @@ static void print_acpi_power(uint32_t cpu, struct acpi_processor_power *power)
     printk("active state:\t\tC%d\n", last_state_idx);
     printk("max_cstate:\t\tC%d\n", max_cstate);
     printk("states:\n");
-    
+
+    spin_lock_irq(&power->stat_lock);
+    current_tick = cpuidle_get_tick();
+    current_stime = NOW();
     for ( i = 1; i < power->count; i++ )
     {
-        spin_lock_irq(&power->stat_lock);      
-        res = tick_to_ns(power->states[i].time);
-        usage = power->states[i].usage;
-        spin_unlock_irq(&power->stat_lock);
+        res_tick[i] = power->states[i].time;
+        usage[i] = power->states[i].usage;
+    }
+    last_state_update_tick = power->last_state_update_tick;
+    spin_unlock_irq(&power->stat_lock);
+
+    res_tick[last_state_idx] += ticks_elapsed(last_state_update_tick, current_tick);
+    usage[last_state_idx]++;
 
-        idle_usage += usage;
-        idle_res += res;
+    for ( i = 1; i < power->count; i++ )
+    {
+        idle_usage += usage[i];
+        idle_res += tick_to_ns(res_tick[i]);
 
         printk((last_state_idx == i) ? "   *" : "    ");
         printk("C%d:\t", i);
         printk("type[C%d] ", power->states[i].type);
         printk("latency[%03d] ", power->states[i].latency);
-        printk("usage[%08d] ", usage);
+        printk("usage[%08"PRIu64"] ", usage[i]);
         printk("method[%5s] ", acpi_cstate_method_name[power->states[i].entry_method]);
-        printk("duration[%"PRId64"]\n", res);
+        printk("duration[%"PRIu64"]\n", tick_to_ns(res_tick[i]));
     }
     printk((last_state_idx == 0) ? "   *" : "    ");
-    printk("C0:\tusage[%08d] duration[%"PRId64"]\n",
-           idle_usage, NOW() - idle_res);
+    printk("C0:\tusage[%08"PRIu64"] duration[%"PRIu64"]\n",
+           usage[0] + idle_usage, current_stime - idle_res);
 
     print_hw_residencies(cpu);
 }
@@ -314,25 +344,6 @@ static int __init cpu_idle_key_init(void)
 }
 __initcall(cpu_idle_key_init);
 
-static uint64_t get_stime_tick(void) { return (uint64_t)NOW(); }
-static uint64_t stime_ticks_elapsed(uint64_t t1, uint64_t t2) { return t2 - t1; }
-static uint64_t stime_tick_to_ns(uint64_t ticks) { return ticks; }
-
-static uint64_t get_acpi_pm_tick(void) { return (uint64_t)inl(pmtmr_ioport); }
-static uint64_t acpi_pm_ticks_elapsed(uint64_t t1, uint64_t t2)
-{
-    if ( t2 >= t1 )
-        return (t2 - t1);
-    else if ( !(acpi_gbl_FADT.flags & ACPI_FADT_32BIT_TIMER) )
-        return (((0x00FFFFFF - t1) + t2 + 1) & 0x00FFFFFF);
-    else
-        return ((0xFFFFFFFF - t1) + t2 +1);
-}
-
-uint64_t (*__read_mostly cpuidle_get_tick)(void) = get_acpi_pm_tick;
-static uint64_t (*__read_mostly ticks_elapsed)(uint64_t, uint64_t)
-    = acpi_pm_ticks_elapsed;
-
 /*
  * The bit is set iff cpu use monitor/mwait to enter C state
  * with this flag set, CPU can be waken up from C state
@@ -487,6 +498,17 @@ bool_t errata_c6_eoi_workaround(void)
     return (fix_needed && cpu_has_pending_apic_eoi());
 }
 
+void update_last_cx_stat(struct acpi_processor_power *power,
+                         struct acpi_processor_cx *cx, uint64_t ticks)
+{
+    ASSERT(!local_irq_is_enabled());
+
+    spin_lock(&power->stat_lock);
+    power->last_state = cx;
+    power->last_state_update_tick = ticks;
+    spin_unlock(&power->stat_lock);
+}
+
 void update_idle_stats(struct acpi_processor_power *power,
                        struct acpi_processor_cx *cx,
                        uint64_t before, uint64_t after)
@@ -502,6 +524,8 @@ void update_idle_stats(struct acpi_processor_power *power,
         power->last_residency = tick_to_ns(sleep_ticks) / 1000UL;
         cx->time += sleep_ticks;
     }
+    power->last_state = &power->states[0];
+    power->last_state_update_tick = after;
 
     spin_unlock(&power->stat_lock);
 }
@@ -558,7 +582,6 @@ static void acpi_processor_idle(void)
     if ( (cx->type == ACPI_STATE_C3) && errata_c6_eoi_workaround() )
         cx = power->safe_state;
 
-    power->last_state = cx;
 
     /*
      * Sleep:
@@ -575,6 +598,9 @@ static void acpi_processor_idle(void)
             t1 = cpuidle_get_tick();
             /* Trace cpu idle entry */
             TRACE_4D(TRC_PM_IDLE_ENTRY, cx->idx, t1, exp, pred);
+
+            update_last_cx_stat(power, cx, t1);
+
             /* Invoke C2 */
             acpi_idle_do_entry(cx);
             /* Get end time (ticks) */
@@ -604,6 +630,8 @@ static void acpi_processor_idle(void)
         /* Trace cpu idle entry */
         TRACE_4D(TRC_PM_IDLE_ENTRY, cx->idx, t1, exp, pred);
 
+        update_last_cx_stat(power, cx, t1);
+
         /*
          * disable bus master
          * bm_check implies we need ARB_DIS
@@ -1173,7 +1201,9 @@ int pmstat_get_cx_stat(uint32_t cpuid, struct pm_cx_stat *stat)
 {
     struct acpi_processor_power *power = processor_powers[cpuid];
     uint64_t idle_usage = 0, idle_res = 0;
-    uint64_t usage[ACPI_PROCESSOR_MAX_POWER], res[ACPI_PROCESSOR_MAX_POWER];
+    uint64_t last_state_update_tick, current_stime, current_tick;
+    uint64_t usage[ACPI_PROCESSOR_MAX_POWER] = { 0 };
+    uint64_t res[ACPI_PROCESSOR_MAX_POWER] = { 0 };
     unsigned int i, nr, nr_pc = 0, nr_cc = 0;
 
     if ( power == NULL )
@@ -1186,7 +1216,6 @@ int pmstat_get_cx_stat(uint32_t cpuid, struct pm_cx_stat *stat)
         return 0;
     }
 
-    stat->last = power->last_state ? power->last_state->idx : 0;
     stat->idle_time = get_cpu_idle_time(cpuid);
     nr = min(stat->nr, power->count);
 
@@ -1194,9 +1223,12 @@ int pmstat_get_cx_stat(uint32_t cpuid, struct pm_cx_stat *stat)
     if ( pm_idle_save == NULL )
     {
         stat->nr = 2;
+        stat->last = power->last_state ? power->last_state->idx : 0;
 
         usage[1] = idle_usage = 1;
         res[1] = idle_res = stat->idle_time;
+
+        current_stime = NOW();
     }
     else
     {
@@ -1204,13 +1236,24 @@ int pmstat_get_cx_stat(uint32_t cpuid, struct pm_cx_stat *stat)
 
         stat->nr = power->count;
 
+        spin_lock_irq(&power->stat_lock);
+        current_tick = cpuidle_get_tick();
+        current_stime = NOW();
         for ( i = 1; i < nr; i++ )
         {
-            spin_lock_irq(&power->stat_lock);
             usage[i] = power->states[i].usage;
-            res[i] = tick_to_ns(power->states[i].time);
-            spin_unlock_irq(&power->stat_lock);
+            res[i] = power->states[i].time;
+        }
+        last_state_update_tick = power->last_state_update_tick;
+        stat->last = power->last_state ? power->last_state->idx : 0;
+        spin_unlock_irq(&power->stat_lock);
 
+        usage[stat->last]++;
+        res[stat->last] += ticks_elapsed(last_state_update_tick, current_tick);
+
+        for ( i = 1; i < nr; i++ )
+        {
+            res[i] = tick_to_ns(res[i]);
             idle_usage += usage[i];
             idle_res += res[i];
         }
@@ -1243,8 +1286,8 @@ int pmstat_get_cx_stat(uint32_t cpuid, struct pm_cx_stat *stat)
 #undef PUT_xC
     }
 
-    usage[0] = idle_usage;
-    res[0] = NOW() - idle_res;
+    usage[0] += idle_usage;
+    res[0] = current_stime - idle_res;
 
     if ( copy_to_guest(stat->triggers, usage, nr) ||
          copy_to_guest(stat->residencies, res, nr) )

diff --git a/xen/arch/x86/cpu/mwait-idle.c b/xen/arch/x86/cpu/mwait-idle.c
index 162e19b66d1a3ab2f274fb61386f9a2ead92563e..91b76ec6b5e9e70eb45811af56b7cc4f3199f7e6 100644 (file)
--- a/xen/arch/x86/cpu/mwait-idle.c
+++ b/xen/arch/x86/cpu/mwait-idle.c
@@ -564,7 +564,6 @@ static void mwait_idle(void)
                return;
        }
 
-       power->last_state = cx;
        eax = cx->address;
        cstate = ((eax >> MWAIT_SUBSTATE_SIZE) & MWAIT_CSTATE_MASK) + 1;
 
@@ -583,6 +582,8 @@ static void mwait_idle(void)
        before = cpuidle_get_tick();
        TRACE_4D(TRC_PM_IDLE_ENTRY, cx->type, before, exp, pred);
 
+       update_last_cx_stat(power, cx, before);
+
        if (cpu_is_haltable(cpu))
                mwait_idle_with_hints(eax, MWAIT_ECX_INTERRUPT_BREAK);
 
@@ -593,15 +594,13 @@ static void mwait_idle(void)
        TRACE_6D(TRC_PM_IDLE_EXIT, cx->type, after,
                irq_traced[0], irq_traced[1], irq_traced[2], irq_traced[3]);
 
+       /* Now back in C0. */
        update_idle_stats(power, cx, before, after);
        local_irq_enable();
 
        if (!(lapic_timer_reliable_states & (1 << cstate)))
                lapic_timer_on();
 
-       /* Now back in C0. */
-       power->last_state = &power->states[0];
-
        sched_tick_resume();
        cpufreq_dbs_timer_resume();
 

diff --git a/xen/include/asm-x86/cpuidle.h b/xen/include/asm-x86/cpuidle.h
index 4d70677d09a91693cdd978c9153e1c6f9005ad1b..46e614bb506392806dcaf83f925ccf97a2e27c6c 100644 (file)
--- a/xen/include/asm-x86/cpuidle.h
+++ b/xen/include/asm-x86/cpuidle.h
@@ -23,6 +23,8 @@ void acpi_dead_idle(void);
 void trace_exit_reason(u32 *irq_traced);
 void update_idle_stats(struct acpi_processor_power *,
                        struct acpi_processor_cx *, uint64_t, uint64_t);
+void update_last_cx_stat(struct acpi_processor_power *,
+                         struct acpi_processor_cx *, uint64_t);
 
 /*
  * vcpu is urgent if vcpu is polling event channel

diff --git a/xen/include/xen/cpuidle.h b/xen/include/xen/cpuidle.h
index b7b9e8c72bec79367352847ac28abc2b46ded2d3..342f4fed98202a5bd68f575a9854de75c22f5301 100644 (file)
--- a/xen/include/xen/cpuidle.h
+++ b/xen/include/xen/cpuidle.h
@@ -66,6 +66,7 @@ struct acpi_processor_power
     struct acpi_processor_cx *last_state;
     struct acpi_processor_cx *safe_state;
     void *gdata; /* governor specific data */
+    u64 last_state_update_tick;
     u32 last_residency;
     u32 count;
     spinlock_t stat_lock;