Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]
Groups > linux.kernel > #1320450 > unrolled thread
| Started by | Borislav Petkov <bp@alien8.de> |
|---|---|
| First post | 2016-01-28 10:10 +0100 |
| Last post | 2016-01-28 16:30 +0100 |
| Articles | 6 — 3 participants |
Back to article view | Back to linux.kernel
This discussion starts older than the indexed window; earlier articles aren't shown. The article labeled Started by
below is the oldest one visible, not the original post.
Re: [PATCH v4] perf/x86/amd/power: Add AMD accumulated power reporting mechanism Borislav Petkov <bp@alien8.de> - 2016-01-28 10:10 +0100
Re: [PATCH] perf/x86/amd/power: Add AMD accumulated power reporting kbuild test robot <lkp@intel.com> - 2016-01-28 10:50 +0100
Re: [PATCH] perf/x86/amd/power: Add AMD accumulated power reporting kbuild test robot <lkp@intel.com> - 2016-01-28 11:10 +0100
Re: [PATCH v4] perf/x86/amd/power: Add AMD accumulated power reporting mechanism Borislav Petkov <bp@alien8.de> - 2016-01-28 13:50 +0100
Re: [PATCH v4] perf/x86/amd/power: Add AMD accumulated power reporting mechanism Borislav Petkov <bp@alien8.de> - 2016-01-28 16:00 +0100
Re: [PATCH v4] perf/x86/amd/power: Add AMD accumulated power reporting mechanism Peter Zijlstra <peterz@infradead.org> - 2016-01-28 16:30 +0100
| From | Borislav Petkov <bp@alien8.de> |
|---|---|
| Date | 2016-01-28 10:10 +0100 |
| Subject | Re: [PATCH v4] perf/x86/amd/power: Add AMD accumulated power reporting mechanism |
| Message-ID | <qVQKu-1al-13@gated-at.bofh.it> |
On Thu, Jan 28, 2016 at 02:38:51PM +0800, Huang Rui wrote:
> Introduce an AMD accumlated power reporting mechanism for Carrizo
> (Family 15h, Model 60h) processor that should be used to calculate the
> average power consumed by a processor during a measurement interval.
> The feature of accumulated power mechanism is indicated by CPUID
> Fn8000_0007_EDX[12].
...
> Suggested-by: Peter Zijlstra <peterz@infradead.org>
> Suggested-by: Ingo Molnar <mingo@kernel.org>
> Suggested-by: Borislav Petkov <bp@suse.de>
> Signed-off-by: Huang Rui <ray.huang@amd.com>
> Cc: Guenter Roeck <linux@roeck-us.net>
...
> +static int power_cpu_exit(int cpu)
> +{
> + struct power_pmu *pmu = per_cpu(amd_power_pmu, cpu);
> + int ret = 0;
> + int target = nr_cpumask_bits;
> +
> + cpumask_copy(pmu->mask, topology_sibling_cpumask(cpu));
> +
> + cpumask_clear_cpu(cpu, &cpu_mask);
> + cpumask_clear_cpu(cpu, pmu->mask);
> +
> + if (!cpumask_and(pmu->tmp_mask, pmu->mask, cpu_online_mask))
> + goto out;
> +
> + /*
> + * Find a new CPU on same compute unit, if was set in cpumask
> + * and still some CPUs on compute unit. Then move on to the
> + * new CPU.
> + */
Uuh, boy, I *think* I know what you mean here but I'm not sure. Please
explain.
Anyway, I went through it and fixed a bunch of minor issues like
comments formatting and formulation, code formatting and text
streamlining. I couldn't spot anything strange except that MAX_CUS thing
which I think we should remove completely.
The meat of the code is for Peter/Ingo to review though.
Please use this version below for future changes:
---
From: Huang Rui <ray.huang@amd.com>
Date: Thu, 28 Jan 2016 14:38:51 +0800
Subject: [PATCH] perf/x86/amd/power: Add AMD accumulated power reporting
mechanism
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Introduce an AMD accumlated power reporting mechanism for the Carrizo
(Family 15h, Model 60h) processor that can be used to calculate the
average power consumed by a processor during a measurement interval. The
feature support is indicated by CPUID Fn8000_0007_EDX[12].
This feature will be implemented both in hwmon and perf. The current
design provides one event to report per package/processor power
consumption by counting each compute unit power value.
Here the gory details of how the computation is done:
---------------------------------------------------------------------
* Tsample: compute unit power accumulator sample period
* Tref: the PTSC counter period (PTSC: performance timestamp counter)
* N: the ratio of compute unit power accumulator sample period to the
PTSC period
* Jmax: max compute unit accumulated power which is indicated by
MSR_C001007b[MaxCpuSwPwrAcc]
* Jx/Jy: compute unit accumulated power which is indicated by
MSR_C001007a[CpuSwPwrAcc]
* Tx/Ty: the value of performance timestamp counter which is indicated
by CU_PTSC MSR_C0010280[PTSC]
* PwrCPUave: CPU average power
i. Determine the ratio of Tsample to Tref by executing CPUID Fn8000_0007.
N = value of CPUID Fn8000_0007_ECX[CpuPwrSampleTimeRatio[15:0]].
ii. Read the full range of the cumulative energy value from the new
MSR MaxCpuSwPwrAcc.
Jmax = value returned.
iii. At time x, software reads CpuSwPwrAcc and samples the PTSC.
Jx = value read from CpuSwPwrAcc and Tx = value read from PTSC.
iv. At time y, software reads CpuSwPwrAcc and samples the PTSC.
Jy = value read from CpuSwPwrAcc and Ty = value read from PTSC.
v. Calculate the average power consumption for a compute unit over
time period (y-x). Unit of result is uWatt:
if (Jy < Jx) // Rollover has occurred
Jdelta = (Jy + Jmax) - Jx
else
Jdelta = Jy - Jx
PwrCPUave = N * Jdelta * 1000 / (Ty - Tx)
----------------------------------------------------------------------
Simple example:
root@hr-zp:/home/ray/tip# ./tools/perf/perf stat -a -e 'power/power-pkg/' make -j4
CHK include/config/kernel.release
CHK include/generated/uapi/linux/version.h
CHK include/generated/utsrelease.h
CHK include/generated/timeconst.h
CHK include/generated/bounds.h
CHK include/generated/asm-offsets.h
CALL scripts/checksyscalls.sh
CHK include/generated/compile.h
SKIPPED include/generated/compile.h
Building modules, stage 2.
Kernel: arch/x86/boot/bzImage is ready (#40)
MODPOST 4225 modules
Performance counter stats for 'system wide':
183.44 mWatts power/power-pkg/
341.837270111 seconds time elapsed
root@hr-zp:/home/ray/tip# ./tools/perf/perf stat -a -e 'power/power-pkg/' sleep 10
Performance counter stats for 'system wide':
0.18 mWatts power/power-pkg/
10.012551815 seconds time elapsed
Suggested-by: Peter Zijlstra <peterz@infradead.org>
Suggested-by: Ingo Molnar <mingo@kernel.org>
Suggested-by: Borislav Petkov <bp@suse.de>
Signed-off-by: Huang Rui <ray.huang@amd.com>
Cc: Aaron Lu <aaron.lu@intel.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Andreas Herrmann <herrmann.der.user@googlemail.com>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Aravind Gopalakrishnan <Aravind.Gopalakrishnan@amd.com>
Cc: Arnaldo Carvalho de Melo <acme@kernel.org>
Cc: Fengguang Wu <fengguang.wu@intel.com>
Cc: Frédéric Weisbecker <fweisbec@gmail.com>
Cc: Guenter Roeck <linux@roeck-us.net>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Jacob Shin <jacob.w.shin@gmail.com>
Cc: John Stultz <john.stultz@linaro.org>
Cc: Kan Liang <kan.liang@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Robert Richter <rric@kernel.org>
Cc: spg_linux_kernel@amd.com
Cc: Suravee Suthikulpanit <suravee.suthikulpanit@amd.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: x86-ml <x86@kernel.org>
Link: http://lkml.kernel.org/r/20150831160622.GA29830@nazgul.tnic
Link: http://lkml.kernel.org/r/1453963131-2013-1-git-send-email-ray.huang@amd.com
Signed-off-by: Borislav Petkov <bp@suse.de>
---
arch/x86/kernel/cpu/Makefile | 1 +
arch/x86/kernel/cpu/perf_event_amd_power.c | 489 +++++++++++++++++++++++++++++
2 files changed, 490 insertions(+)
create mode 100644 arch/x86/kernel/cpu/perf_event_amd_power.c
diff --git a/arch/x86/kernel/cpu/Makefile b/arch/x86/kernel/cpu/Makefile
index faa7b5204129..ffc96503d610 100644
--- a/arch/x86/kernel/cpu/Makefile
+++ b/arch/x86/kernel/cpu/Makefile
@@ -34,6 +34,7 @@ obj-$(CONFIG_PERF_EVENTS) += perf_event.o
ifdef CONFIG_PERF_EVENTS
obj-$(CONFIG_CPU_SUP_AMD) += perf_event_amd.o perf_event_amd_uncore.o
+obj-$(CONFIG_CPU_SUP_AMD) += perf_event_amd_power.o
ifdef CONFIG_AMD_IOMMU
obj-$(CONFIG_CPU_SUP_AMD) += perf_event_amd_iommu.o
endif
diff --git a/arch/x86/kernel/cpu/perf_event_amd_power.c b/arch/x86/kernel/cpu/perf_event_amd_power.c
new file mode 100644
index 000000000000..ff6893620828
--- /dev/null
+++ b/arch/x86/kernel/cpu/perf_event_amd_power.c
@@ -0,0 +1,489 @@
+/*
+ * Performance events - AMD Processor Power Reporting Mechanism
+ *
+ * Copyright (C) 2016 Advanced Micro Devices, Inc.
+ *
+ * Author: Huang Rui <ray.huang@amd.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/module.h>
+#include <linux/slab.h>
+#include <linux/perf_event.h>
+#include <asm/cpu_device_id.h>
+#include "perf_event.h"
+
+#define MSR_F15H_CU_PWR_ACCUMULATOR 0xc001007a
+#define MSR_F15H_CU_MAX_PWR_ACCUMULATOR 0xc001007b
+#define MSR_F15H_PTSC 0xc0010280
+
+/* Event code: LSB 8 bits, passed in attr->config any other bit is reserved. */
+#define AMD_POWER_EVENT_MASK 0xFFULL
+
+#define MAX_CUS 8
+
+/*
+ * Accumulated power status counters.
+ */
+#define AMD_POWER_PKG_ID 0
+#define AMD_POWER_EVENTSEL_PKG 1
+
+/*
+ * The ratio of compute unit power accumulator sample period to the
+ * PTSC period.
+ */
+static unsigned int cpu_pwr_sample_ratio;
+static unsigned int cores_per_cu;
+static unsigned int cu_num;
+
+/* Maximum accumulated power of a compute unit. */
+static u64 max_cu_acc_power;
+
+struct power_pmu {
+ raw_spinlock_t lock;
+ struct pmu *pmu;
+ local64_t cpu_sw_pwr_ptsc;
+
+ /*
+ * These two cpumasks are used for avoiding the allocations on the
+ * CPU_STARTING phase because power_cpu_prepare() will be called with
+ * IRQs disabled.
+ */
+ cpumask_var_t mask;
+ cpumask_var_t tmp_mask;
+};
+
+static struct pmu pmu_class;
+
+/*
+ * Accumulated power represents the sum of each compute unit's (CU) power
+ * consumption. On any core of each CU we read the total accumulated power from
+ * MSR_F15H_CU_PWR_ACCUMULATOR. cpu_mask represents CPU bit map of all cores
+ * which are picked to measure the power for the CUs they belong to.
+ */
+static cpumask_t cpu_mask;
+
+static DEFINE_PER_CPU(struct power_pmu *, amd_power_pmu);
+
+static u64 event_update(struct perf_event *event, struct power_pmu *pmu)
+{
+ struct hw_perf_event *hwc = &event->hw;
+ u64 prev_raw_count, new_raw_count, prev_ptsc, new_ptsc;
+ u64 delta, tdelta;
+
+again:
+ prev_raw_count = local64_read(&hwc->prev_count);
+ prev_ptsc = local64_read(&pmu->cpu_sw_pwr_ptsc);
+ rdmsrl(event->hw.event_base, new_raw_count);
+ rdmsrl(MSR_F15H_PTSC, new_ptsc);
+
+ if (local64_cmpxchg(&hwc->prev_count, prev_raw_count,
+ new_raw_count) != prev_raw_count) {
+ cpu_relax();
+ goto again;
+ }
+
+ /*
+ * Calculate the CU power consumption over a time period, the unit of
+ * final value (delta) is micro-Watts. Then add it to the event count.
+ */
+ if (new_raw_count < prev_raw_count) {
+ delta = max_cu_acc_power + new_raw_count;
+ delta -= prev_raw_count;
+ } else
+ delta = new_raw_count - prev_raw_count;
+
+ delta *= cpu_pwr_sample_ratio * 1000;
+ tdelta = new_ptsc - prev_ptsc;
+
+ do_div(delta, tdelta);
+ local64_add(delta, &event->count);
+
+ return new_raw_count;
+}
+
+static void __pmu_event_start(struct power_pmu *pmu, struct perf_event *event)
+{
+ u64 ptsc, counts;
+
+ if (WARN_ON_ONCE(!(event->hw.state & PERF_HES_STOPPED)))
+ return;
+
+ event->hw.state = 0;
+
+ rdmsrl(MSR_F15H_PTSC, ptsc);
+ local64_set(&pmu->cpu_sw_pwr_ptsc, ptsc);
+ rdmsrl(event->hw.event_base, counts);
+ local64_set(&event->hw.prev_count, counts);
+}
+
+static void pmu_event_start(struct perf_event *event, int mode)
+{
+ struct power_pmu *pmu = __this_cpu_read(amd_power_pmu);
+
+ raw_spin_lock(&pmu->lock);
+ __pmu_event_start(pmu, event);
+ raw_spin_unlock(&pmu->lock);
+}
+
+static void pmu_event_stop(struct perf_event *event, int mode)
+{
+ struct power_pmu *pmu = __this_cpu_read(amd_power_pmu);
+ struct hw_perf_event *hwc = &event->hw;
+
+ raw_spin_lock(&pmu->lock);
+
+ /* Mark event as deactivated and stopped. */
+ if (!(hwc->state & PERF_HES_STOPPED))
+ hwc->state |= PERF_HES_STOPPED;
+
+ /* Check if software counter update is necessary. */
+ if ((mode & PERF_EF_UPDATE) && !(hwc->state & PERF_HES_UPTODATE)) {
+ /*
+ * Drain the remaining delta count out of an event
+ * that we are disabling:
+ */
+ event_update(event, pmu);
+ hwc->state |= PERF_HES_UPTODATE;
+ }
+
+ raw_spin_unlock(&pmu->lock);
+}
+
+static int pmu_event_add(struct perf_event *event, int mode)
+{
+ struct power_pmu *pmu = __this_cpu_read(amd_power_pmu);
+ struct hw_perf_event *hwc = &event->hw;
+
+ raw_spin_lock(&pmu->lock);
+
+ hwc->state = PERF_HES_UPTODATE | PERF_HES_STOPPED;
+
+ if (mode & PERF_EF_START)
+ __pmu_event_start(pmu, event);
+
+ raw_spin_unlock(&pmu->lock);
+
+ return 0;
+}
+
+static void pmu_event_del(struct perf_event *event, int flags)
+{
+ pmu_event_stop(event, PERF_EF_UPDATE);
+}
+
+static int pmu_event_init(struct perf_event *event)
+{
+ u64 cfg = event->attr.config & AMD_POWER_EVENT_MASK;
+ int ret = 0;
+
+ /* Only look at AMD power events. */
+ if (event->attr.type != pmu_class.type)
+ return -ENOENT;
+
+ /* Unsupported modes and filters. */
+ if (event->attr.exclude_user ||
+ event->attr.exclude_kernel ||
+ event->attr.exclude_hv ||
+ event->attr.exclude_idle ||
+ event->attr.exclude_host ||
+ event->attr.exclude_guest ||
+ /* no sampling */
+ event->attr.sample_period)
+ return -EINVAL;
+
+ if (cfg != AMD_POWER_EVENTSEL_PKG)
+ return -EINVAL;
+
+ event->hw.event_base = MSR_F15H_CU_PWR_ACCUMULATOR;
+ event->hw.config = cfg;
+ event->hw.idx = AMD_POWER_PKG_ID;
+
+ return ret;
+}
+
+static void pmu_event_read(struct perf_event *event)
+{
+ struct power_pmu *pmu = __this_cpu_read(amd_power_pmu);
+
+ event_update(event, pmu);
+}
+
+static ssize_t
+get_attr_cpumask(struct device *dev, struct device_attribute *attr, char *buf)
+{
+ return cpumap_print_to_pagebuf(true, buf, &cpu_mask);
+}
+
+static DEVICE_ATTR(cpumask, S_IRUGO, get_attr_cpumask, NULL);
+
+static struct attribute *pmu_attrs[] = {
+ &dev_attr_cpumask.attr,
+ NULL,
+};
+
+static struct attribute_group pmu_attr_group = {
+ .attrs = pmu_attrs,
+};
+
+/*
+ * Currently it only supports to report the power of each
+ * processor/package.
+ */
+EVENT_ATTR_STR(power-pkg, power_pkg, "event=0x01");
+
+EVENT_ATTR_STR(power-pkg.unit, power_pkg_unit, "mWatts");
+
+/* Convert the count from micro-Watts to milli-Watts. */
+EVENT_ATTR_STR(power-pkg.scale, power_pkg_scale, "1.000000e-3");
+
+
+static struct attribute *events_attr[] = {
+ EVENT_PTR(power_pkg),
+ EVENT_PTR(power_pkg_unit),
+ EVENT_PTR(power_pkg_scale),
+ NULL,
+};
+
+static struct attribute_group pmu_events_group = {
+ .name = "events",
+ .attrs = events_attr,
+};
+
+PMU_FORMAT_ATTR(event, "config:0-7");
+
+static struct attribute *formats_attr[] = {
+ &format_attr_event.attr,
+ NULL,
+};
+
+static struct attribute_group pmu_format_group = {
+ .name = "format",
+ .attrs = formats_attr,
+};
+
+static const struct attribute_group *attr_groups[] = {
+ &pmu_attr_group,
+ &pmu_format_group,
+ &pmu_events_group,
+ NULL,
+};
+
+static struct pmu pmu_class = {
+ .attr_groups = attr_groups,
+ /* system-wide only */
+ .task_ctx_nr = perf_invalid_context,
+ .event_init = pmu_event_init,
+ .add = pmu_event_add,
+ .del = pmu_event_del,
+ .start = pmu_event_start,
+ .stop = pmu_event_stop,
+ .read = pmu_event_read,
+};
+
+static int power_cpu_exit(int cpu)
+{
+ struct power_pmu *pmu = per_cpu(amd_power_pmu, cpu);
+ int target = nr_cpumask_bits;
+ int ret = 0;
+
+ cpumask_copy(pmu->mask, topology_sibling_cpumask(cpu));
+
+ cpumask_clear_cpu(cpu, &cpu_mask);
+ cpumask_clear_cpu(cpu, pmu->mask);
+
+ if (!cpumask_and(pmu->tmp_mask, pmu->mask, cpu_online_mask))
+ goto out;
+
+ /*
+ * Find a new CPU on the same compute unit, if was set in cpumask
+ * and still some CPUs on compute unit. Then move on to the new CPU.
+ */
+ target = cpumask_any(pmu->tmp_mask);
+ if (target < nr_cpumask_bits && target != cpu)
+ cpumask_set_cpu(target, &cpu_mask);
+
+ WARN_ON(cpumask_empty(&cpu_mask));
+
+out:
+ /*
+ * Migrate event and context to new CPU.
+ */
+ if (target < nr_cpumask_bits)
+ perf_pmu_migrate_context(pmu->pmu, cpu, target);
+
+ return ret;
+
+}
+
+static int power_cpu_init(int cpu)
+{
+ struct power_pmu *pmu = per_cpu(amd_power_pmu, cpu);
+
+ if (!pmu)
+ return 0;
+
+ if (!cpumask_and(pmu->mask, topology_sibling_cpumask(cpu), &cpu_mask))
+ cpumask_set_cpu(cpu, &cpu_mask);
+
+ return 0;
+}
+
+static int power_cpu_prepare(int cpu)
+{
+ struct power_pmu *pmu = per_cpu(amd_power_pmu, cpu);
+ int phys_id = topology_physical_package_id(cpu);
+ int ret = 0;
+
+ if (pmu)
+ return 0;
+
+ if (phys_id < 0)
+ return -EINVAL;
+
+ pmu = kzalloc_node(sizeof(*pmu), GFP_KERNEL, cpu_to_node(cpu));
+ if (!pmu)
+ return -ENOMEM;
+
+ if (!zalloc_cpumask_var(&pmu->mask, GFP_KERNEL)) {
+ ret = -ENOMEM;
+ goto out;
+ }
+
+ if (!zalloc_cpumask_var(&pmu->tmp_mask, GFP_KERNEL)) {
+ ret = -ENOMEM;
+ goto out1;
+ }
+
+ raw_spin_lock_init(&pmu->lock);
+
+ pmu->pmu = &pmu_class;
+
+ per_cpu(amd_power_pmu, cpu) = pmu;
+
+ return 0;
+
+out1:
+ free_cpumask_var(pmu->mask);
+out:
+ kfree(pmu);
+
+ return ret;
+}
+
+static void power_cpu_kfree(int cpu)
+{
+ struct power_pmu *pmu = per_cpu(amd_power_pmu, cpu);
+
+ if (!pmu)
+ return;
+
+ free_cpumask_var(pmu->mask);
+ free_cpumask_var(pmu->tmp_mask);
+ kfree(pmu);
+
+ per_cpu(amd_power_pmu, cpu) = NULL;
+}
+
+static int
+power_cpu_notifier(struct notifier_block *self, unsigned long action, void *hcpu)
+{
+ unsigned int cpu = (long)hcpu;
+
+ switch (action & ~CPU_TASKS_FROZEN) {
+ case CPU_UP_PREPARE:
+ if (power_cpu_prepare(cpu))
+ return NOTIFY_BAD;
+ break;
+ case CPU_STARTING:
+ if (power_cpu_init(cpu))
+ return NOTIFY_BAD;
+ break;
+ case CPU_DEAD:
+ power_cpu_kfree(cpu);
+ break;
+ case CPU_DOWN_PREPARE:
+ if (power_cpu_exit(cpu))
+ return NOTIFY_BAD;
+ break;
+ default:
+ break;
+ }
+
+ return NOTIFY_OK;
+}
+
+static const struct x86_cpu_id cpu_match[] = {
+ { .vendor = X86_VENDOR_AMD, .family = 0x15 },
+ {},
+};
+
+static int __init amd_power_pmu_init(void)
+{
+ int i, ret;
+ u64 tmp;
+
+ if (!x86_match_cpu(cpu_match))
+ return 0;
+
+ if (!boot_cpu_has(X86_FEATURE_ACC_POWER))
+ return -ENODEV;
+
+ cores_per_cu = amd_get_cores_per_cu();
+ cu_num = boot_cpu_data.x86_max_cores / cores_per_cu;
+
+ if (WARN_ON_ONCE(cu_num > MAX_CUS))
+ return -EINVAL;
+
+ cpu_pwr_sample_ratio = cpuid_ecx(0x80000007);
+
+ if (rdmsrl_safe(MSR_F15H_CU_MAX_PWR_ACCUMULATOR, &tmp)) {
+ pr_err("Failed to read max compute unit power accumulator MSR\n");
+ return -ENODEV;
+ }
+ max_cu_acc_power = tmp;
+
+ cpu_notifier_register_begin();
+
+ /* Choose one online core of each compute unit. */
+ for (i = 0; i < boot_cpu_data.x86_max_cores; i += cores_per_cu) {
+ WARN_ON(cpumask_empty(topology_sibling_cpumask(i)));
+ cpumask_set_cpu(cpumask_any(topology_sibling_cpumask(i)), &cpu_mask);
+ }
+
+ for_each_present_cpu(i) {
+ ret = power_cpu_prepare(i);
+ if (ret) {
+ /* Unwind on [0 ... i-1] CPUs. */
+ while (i--)
+ power_cpu_kfree(i);
+ goto out;
+ }
+ ret = power_cpu_init(i);
+ if (ret) {
+ /* Unwind on [0 ... i] CPUs. */
+ while (i >= 0)
+ power_cpu_kfree(i--);
+ goto out;
+ }
+ }
+
+ __perf_cpu_notifier(power_cpu_notifier);
+
+ ret = perf_pmu_register(&pmu_class, "power", -1);
+ if (WARN_ON(ret)) {
+ pr_warn("AMD Power PMU registration failed\n");
+ goto out;
+ }
+
+ pr_info("AMD Power PMU detected, %d compute units\n", cu_num);
+
+out:
+ cpu_notifier_register_done();
+
+ return ret;
+}
+device_initcall(amd_power_pmu_init);
--
2.3.5
--
Regards/Gruss,
Boris.
ECO tip #101: Trim your mails when you reply.
[toc] | [next] | [standalone]
| From | kbuild test robot <lkp@intel.com> |
|---|---|
| Date | 2016-01-28 10:50 +0100 |
| Subject | Re: [PATCH] perf/x86/amd/power: Add AMD accumulated power reporting |
| Message-ID | <qVRnd-1sZ-31@gated-at.bofh.it> |
| In reply to | #1320450 |
[Multipart message — attachments visible in raw view] — view raw
Hi Borislav,
[auto build test WARNING on tip/x86/core]
[also build test WARNING on v4.5-rc1 next-20160128]
[if your patch is applied to the wrong git tree, please drop us a note to help improving the system]
url: https://github.com/0day-ci/linux/commits/Borislav-Petkov/perf-x86-amd-power-Add-AMD-accumulated-power-reporting/20160128-170527
config: i386-randconfig-s0-201604 (attached as .config)
reproduce:
# save the attached .config to linux build tree
make ARCH=i386
All warnings (new ones prefixed by >>):
In file included from arch/x86/include/asm/alternative.h:158:0,
from arch/x86/include/asm/bitops.h:16,
from include/linux/bitops.h:36,
from include/linux/kernel.h:10,
from include/linux/list.h:8,
from include/linux/module.h:9,
from arch/x86/kernel/cpu/perf_event_amd_power.c:13:
arch/x86/kernel/cpu/perf_event_amd_power.c: In function 'amd_power_pmu_init':
arch/x86/kernel/cpu/perf_event_amd_power.c:432:20: error: 'X86_FEATURE_ACC_POWER' undeclared (first use in this function)
if (!boot_cpu_has(X86_FEATURE_ACC_POWER))
^
arch/x86/include/asm/cpufeature.h:319:24: note: in definition of macro 'cpu_has'
(__builtin_constant_p(bit) && REQUIRED_MASK_BIT_SET(bit) ? 1 : \
^
>> arch/x86/kernel/cpu/perf_event_amd_power.c:432:7: note: in expansion of macro 'boot_cpu_has'
if (!boot_cpu_has(X86_FEATURE_ACC_POWER))
^
arch/x86/kernel/cpu/perf_event_amd_power.c:432:20: note: each undeclared identifier is reported only once for each function it appears in
if (!boot_cpu_has(X86_FEATURE_ACC_POWER))
^
arch/x86/include/asm/cpufeature.h:319:24: note: in definition of macro 'cpu_has'
(__builtin_constant_p(bit) && REQUIRED_MASK_BIT_SET(bit) ? 1 : \
^
>> arch/x86/kernel/cpu/perf_event_amd_power.c:432:7: note: in expansion of macro 'boot_cpu_has'
if (!boot_cpu_has(X86_FEATURE_ACC_POWER))
^
arch/x86/kernel/cpu/perf_event_amd_power.c:435:17: error: implicit declaration of function 'amd_get_cores_per_cu' [-Werror=implicit-function-declaration]
cores_per_cu = amd_get_cores_per_cu();
^
cc1: some warnings being treated as errors
vim +/boot_cpu_has +432 arch/x86/kernel/cpu/perf_event_amd_power.c
416 return NOTIFY_OK;
417 }
418
419 static const struct x86_cpu_id cpu_match[] = {
420 { .vendor = X86_VENDOR_AMD, .family = 0x15 },
421 {},
422 };
423
424 static int __init amd_power_pmu_init(void)
425 {
426 int i, ret;
427 u64 tmp;
428
429 if (!x86_match_cpu(cpu_match))
430 return 0;
431
> 432 if (!boot_cpu_has(X86_FEATURE_ACC_POWER))
433 return -ENODEV;
434
435 cores_per_cu = amd_get_cores_per_cu();
436 cu_num = boot_cpu_data.x86_max_cores / cores_per_cu;
437
438 if (WARN_ON_ONCE(cu_num > MAX_CUS))
439 return -EINVAL;
440
---
0-DAY kernel test infrastructure Open Source Technology Center
https://lists.01.org/pipermail/kbuild-all Intel Corporation
[toc] | [prev] | [next] | [standalone]
| From | kbuild test robot <lkp@intel.com> |
|---|---|
| Date | 2016-01-28 11:10 +0100 |
| Subject | Re: [PATCH] perf/x86/amd/power: Add AMD accumulated power reporting |
| Message-ID | <qVRGy-1R3-19@gated-at.bofh.it> |
| In reply to | #1320450 |
[Multipart message — attachments visible in raw view] — view raw
Hi Borislav,
[auto build test WARNING on tip/x86/core]
[also build test WARNING on v4.5-rc1 next-20160128]
[if your patch is applied to the wrong git tree, please drop us a note to help improving the system]
url: https://github.com/0day-ci/linux/commits/Borislav-Petkov/perf-x86-amd-power-Add-AMD-accumulated-power-reporting/20160128-170527
config: x86_64-randconfig-i0-01270829 (attached as .config)
reproduce:
# save the attached .config to linux build tree
make ARCH=x86_64
All warnings (new ones prefixed by >>):
In file included from include/uapi/linux/stddef.h:1:0,
from include/linux/stddef.h:4,
from include/uapi/linux/posix_types.h:4,
from include/uapi/linux/types.h:13,
from include/linux/types.h:5,
from include/linux/list.h:4,
from include/linux/module.h:9,
from arch/x86/kernel/cpu/perf_event_amd_power.c:13:
arch/x86/kernel/cpu/perf_event_amd_power.c: In function 'amd_power_pmu_init':
arch/x86/kernel/cpu/perf_event_amd_power.c:432:20: error: 'X86_FEATURE_ACC_POWER' undeclared (first use in this function)
if (!boot_cpu_has(X86_FEATURE_ACC_POWER))
^
include/linux/compiler.h:147:28: note: in definition of macro '__trace_if'
if (__builtin_constant_p((cond)) ? !!(cond) : \
^
>> arch/x86/kernel/cpu/perf_event_amd_power.c:432:2: note: in expansion of macro 'if'
if (!boot_cpu_has(X86_FEATURE_ACC_POWER))
^
arch/x86/include/asm/cpufeature.h:338:27: note: in expansion of macro 'cpu_has'
#define boot_cpu_has(bit) cpu_has(&boot_cpu_data, bit)
^
arch/x86/kernel/cpu/perf_event_amd_power.c:432:7: note: in expansion of macro 'boot_cpu_has'
if (!boot_cpu_has(X86_FEATURE_ACC_POWER))
^
arch/x86/kernel/cpu/perf_event_amd_power.c:432:20: note: each undeclared identifier is reported only once for each function it appears in
if (!boot_cpu_has(X86_FEATURE_ACC_POWER))
^
include/linux/compiler.h:147:28: note: in definition of macro '__trace_if'
if (__builtin_constant_p((cond)) ? !!(cond) : \
^
>> arch/x86/kernel/cpu/perf_event_amd_power.c:432:2: note: in expansion of macro 'if'
if (!boot_cpu_has(X86_FEATURE_ACC_POWER))
^
arch/x86/include/asm/cpufeature.h:338:27: note: in expansion of macro 'cpu_has'
#define boot_cpu_has(bit) cpu_has(&boot_cpu_data, bit)
^
arch/x86/kernel/cpu/perf_event_amd_power.c:432:7: note: in expansion of macro 'boot_cpu_has'
if (!boot_cpu_has(X86_FEATURE_ACC_POWER))
^
arch/x86/kernel/cpu/perf_event_amd_power.c:435:17: error: implicit declaration of function 'amd_get_cores_per_cu' [-Werror=implicit-function-declaration]
cores_per_cu = amd_get_cores_per_cu();
^
cc1: some warnings being treated as errors
vim +/if +432 arch/x86/kernel/cpu/perf_event_amd_power.c
416 return NOTIFY_OK;
417 }
418
419 static const struct x86_cpu_id cpu_match[] = {
420 { .vendor = X86_VENDOR_AMD, .family = 0x15 },
421 {},
422 };
423
424 static int __init amd_power_pmu_init(void)
425 {
426 int i, ret;
427 u64 tmp;
428
429 if (!x86_match_cpu(cpu_match))
430 return 0;
431
> 432 if (!boot_cpu_has(X86_FEATURE_ACC_POWER))
433 return -ENODEV;
434
435 cores_per_cu = amd_get_cores_per_cu();
436 cu_num = boot_cpu_data.x86_max_cores / cores_per_cu;
437
438 if (WARN_ON_ONCE(cu_num > MAX_CUS))
439 return -EINVAL;
440
---
0-DAY kernel test infrastructure Open Source Technology Center
https://lists.01.org/pipermail/kbuild-all Intel Corporation
[toc] | [prev] | [next] | [standalone]
| From | Borislav Petkov <bp@alien8.de> |
|---|---|
| Date | 2016-01-28 13:50 +0100 |
| Message-ID | <qVUbo-3sP-19@gated-at.bofh.it> |
| In reply to | #1320450 |
On Thu, Jan 28, 2016 at 06:01:43PM +0800, Huang Rui wrote:
> For example: Carrizo has four CPU cores and two compute units (CUs).
> CPU0 and CPU1 belongs to CU0, CPU2 and CPU3 belongs to CU1.
>
> At normal initialization, cpu_mask should be "0,2". That means OS
> choose CPU0 in CU0 and CPU2 in CU1 to measure the CU0 and CU1's power
> consumption. If we make the CPU2 offline at runtime, OS need try to
> find another CPU in same compute unit (Here is CU1, only CPU3 can be
> picked). Then OS will move on to the CPU3 to measure CU1's power
> consumption instead of CPU2.
So basically you want to simply say:
"Find another CPU on the same compute unit and set it in the mask of
CPUs on which we do the measurements."
Which reminds me: that cpu_mask thing is insufficiently named - it
should be called measuring_cpus_mask or so.
Btw, the kbuild robot errors come from the fact that there are changes
to cpufeature.h which I didn't mention when applying your patches. So
I've pushed the whole pile here:
http://git.kernel.org/cgit/linux/kernel/git/bp/bp.git/log/?h=tip-perf
Please use that branch instead.
Thanks.
--
Regards/Gruss,
Boris.
ECO tip #101: Trim your mails when you reply.
[toc] | [prev] | [next] | [standalone]
| From | Borislav Petkov <bp@alien8.de> |
|---|---|
| Date | 2016-01-28 16:00 +0100 |
| Message-ID | <qVWdb-4Sp-1@gated-at.bofh.it> |
| In reply to | #1320633 |
On Thu, Jan 28, 2016 at 01:42:09PM +0100, Borislav Petkov wrote:
> http://git.kernel.org/cgit/linux/kernel/git/bp/bp.git/log/?h=tip-perf
>
> Please use that branch instead.
I'm pasting next version here, I'm not hoping that kbuild robot will be
able to paste this email and know *not* to build it directly but use the
branch above instead :-)
Ok, next version, I killed the MAX_CUS and cu_num thing:
> #define MAX_CUS 8
>
> ...
>
> cores_per_cu = amd_get_cores_per_cu();
> cu_num = boot_cpu_data.x86_max_cores / cores_per_cu;
>
> if (WARN_ON_ONCE(cu_num > MAX_CUS))
> return -EINVAL;
as it is completely arbitrary. What's wrong with having a multi-socket
machine where CUs are more than 8 counting in the whole fabric? Nothing,
AFAICT.
---
From: Huang Rui <ray.huang@amd.com>
Date: Thu, 28 Jan 2016 14:38:51 +0800
Subject: [PATCH] perf/x86/amd/power: Add AMD accumulated power reporting
mechanism
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Introduce an AMD accumlated power reporting mechanism for the Carrizo
(Family 15h, Model 60h) processor that can be used to calculate the
average power consumed by a processor during a measurement interval. The
feature support is indicated by CPUID Fn8000_0007_EDX[12].
This feature will be implemented both in hwmon and perf. The current
design provides one event to report per package/processor power
consumption by counting each compute unit power value.
Here the gory details of how the computation is done:
---------------------------------------------------------------------
* Tsample: compute unit power accumulator sample period
* Tref: the PTSC counter period (PTSC: performance timestamp counter)
* N: the ratio of compute unit power accumulator sample period to the
PTSC period
* Jmax: max compute unit accumulated power which is indicated by
MSR_C001007b[MaxCpuSwPwrAcc]
* Jx/Jy: compute unit accumulated power which is indicated by
MSR_C001007a[CpuSwPwrAcc]
* Tx/Ty: the value of performance timestamp counter which is indicated
by CU_PTSC MSR_C0010280[PTSC]
* PwrCPUave: CPU average power
i. Determine the ratio of Tsample to Tref by executing CPUID Fn8000_0007.
N = value of CPUID Fn8000_0007_ECX[CpuPwrSampleTimeRatio[15:0]].
ii. Read the full range of the cumulative energy value from the new
MSR MaxCpuSwPwrAcc.
Jmax = value returned.
iii. At time x, software reads CpuSwPwrAcc and samples the PTSC.
Jx = value read from CpuSwPwrAcc and Tx = value read from PTSC.
iv. At time y, software reads CpuSwPwrAcc and samples the PTSC.
Jy = value read from CpuSwPwrAcc and Ty = value read from PTSC.
v. Calculate the average power consumption for a compute unit over
time period (y-x). Unit of result is uWatt:
if (Jy < Jx) // Rollover has occurred
Jdelta = (Jy + Jmax) - Jx
else
Jdelta = Jy - Jx
PwrCPUave = N * Jdelta * 1000 / (Ty - Tx)
----------------------------------------------------------------------
Simple example:
root@hr-zp:/home/ray/tip# ./tools/perf/perf stat -a -e 'power/power-pkg/' make -j4
CHK include/config/kernel.release
CHK include/generated/uapi/linux/version.h
CHK include/generated/utsrelease.h
CHK include/generated/timeconst.h
CHK include/generated/bounds.h
CHK include/generated/asm-offsets.h
CALL scripts/checksyscalls.sh
CHK include/generated/compile.h
SKIPPED include/generated/compile.h
Building modules, stage 2.
Kernel: arch/x86/boot/bzImage is ready (#40)
MODPOST 4225 modules
Performance counter stats for 'system wide':
183.44 mWatts power/power-pkg/
341.837270111 seconds time elapsed
root@hr-zp:/home/ray/tip# ./tools/perf/perf stat -a -e 'power/power-pkg/' sleep 10
Performance counter stats for 'system wide':
0.18 mWatts power/power-pkg/
10.012551815 seconds time elapsed
Suggested-by: Peter Zijlstra <peterz@infradead.org>
Suggested-by: Ingo Molnar <mingo@kernel.org>
Suggested-by: Borislav Petkov <bp@suse.de>
Signed-off-by: Huang Rui <ray.huang@amd.com>
Cc: Aaron Lu <aaron.lu@intel.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Andreas Herrmann <herrmann.der.user@googlemail.com>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Aravind Gopalakrishnan <Aravind.Gopalakrishnan@amd.com>
Cc: Arnaldo Carvalho de Melo <acme@kernel.org>
Cc: Fengguang Wu <fengguang.wu@intel.com>
Cc: Frédéric Weisbecker <fweisbec@gmail.com>
Cc: Guenter Roeck <linux@roeck-us.net>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Jacob Shin <jacob.w.shin@gmail.com>
Cc: John Stultz <john.stultz@linaro.org>
Cc: Kan Liang <kan.liang@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Robert Richter <rric@kernel.org>
Cc: spg_linux_kernel@amd.com
Cc: Suravee Suthikulpanit <suravee.suthikulpanit@amd.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: x86-ml <x86@kernel.org>
Link: http://lkml.kernel.org/r/20150831160622.GA29830@nazgul.tnic
Link: http://lkml.kernel.org/r/1453963131-2013-1-git-send-email-ray.huang@amd.com
Signed-off-by: Borislav Petkov <bp@suse.de>
---
arch/x86/kernel/cpu/Makefile | 1 +
arch/x86/kernel/cpu/perf_event_amd_power.c | 482 +++++++++++++++++++++++++++++
2 files changed, 483 insertions(+)
create mode 100644 arch/x86/kernel/cpu/perf_event_amd_power.c
diff --git a/arch/x86/kernel/cpu/Makefile b/arch/x86/kernel/cpu/Makefile
index faa7b5204129..ffc96503d610 100644
--- a/arch/x86/kernel/cpu/Makefile
+++ b/arch/x86/kernel/cpu/Makefile
@@ -34,6 +34,7 @@ obj-$(CONFIG_PERF_EVENTS) += perf_event.o
ifdef CONFIG_PERF_EVENTS
obj-$(CONFIG_CPU_SUP_AMD) += perf_event_amd.o perf_event_amd_uncore.o
+obj-$(CONFIG_CPU_SUP_AMD) += perf_event_amd_power.o
ifdef CONFIG_AMD_IOMMU
obj-$(CONFIG_CPU_SUP_AMD) += perf_event_amd_iommu.o
endif
diff --git a/arch/x86/kernel/cpu/perf_event_amd_power.c b/arch/x86/kernel/cpu/perf_event_amd_power.c
new file mode 100644
index 000000000000..b0df1594b095
--- /dev/null
+++ b/arch/x86/kernel/cpu/perf_event_amd_power.c
@@ -0,0 +1,482 @@
+/*
+ * Performance events - AMD Processor Power Reporting Mechanism
+ *
+ * Copyright (C) 2016 Advanced Micro Devices, Inc.
+ *
+ * Author: Huang Rui <ray.huang@amd.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/module.h>
+#include <linux/slab.h>
+#include <linux/perf_event.h>
+#include <asm/cpu_device_id.h>
+#include "perf_event.h"
+
+#define MSR_F15H_CU_PWR_ACCUMULATOR 0xc001007a
+#define MSR_F15H_CU_MAX_PWR_ACCUMULATOR 0xc001007b
+#define MSR_F15H_PTSC 0xc0010280
+
+/* Event code: LSB 8 bits, passed in attr->config any other bit is reserved. */
+#define AMD_POWER_EVENT_MASK 0xFFULL
+
+/*
+ * Accumulated power status counters.
+ */
+#define AMD_POWER_PKG_ID 0
+#define AMD_POWER_EVENTSEL_PKG 1
+
+/*
+ * The ratio of compute unit power accumulator sample period to the
+ * PTSC period.
+ */
+static unsigned int cpu_pwr_sample_ratio;
+static unsigned int cores_per_cu;
+
+/* Maximum accumulated power of a compute unit. */
+static u64 max_cu_acc_power;
+
+struct power_pmu {
+ raw_spinlock_t lock;
+ struct pmu *pmu;
+ local64_t cpu_sw_pwr_ptsc;
+
+ /*
+ * These two cpumasks are used for avoiding the allocations on the
+ * CPU_STARTING phase because power_cpu_prepare() will be called with
+ * IRQs disabled.
+ */
+ cpumask_var_t mask;
+ cpumask_var_t tmp_mask;
+};
+
+static struct pmu pmu_class;
+
+/*
+ * Accumulated power represents the sum of each compute unit's (CU) power
+ * consumption. On any core of each CU we read the total accumulated power from
+ * MSR_F15H_CU_PWR_ACCUMULATOR. cpu_mask represents CPU bit map of all cores
+ * which are picked to measure the power for the CUs they belong to.
+ */
+static cpumask_t cpu_mask;
+
+static DEFINE_PER_CPU(struct power_pmu *, amd_power_pmu);
+
+static u64 event_update(struct perf_event *event, struct power_pmu *pmu)
+{
+ struct hw_perf_event *hwc = &event->hw;
+ u64 prev_raw_count, new_raw_count, prev_ptsc, new_ptsc;
+ u64 delta, tdelta;
+
+again:
+ prev_raw_count = local64_read(&hwc->prev_count);
+ prev_ptsc = local64_read(&pmu->cpu_sw_pwr_ptsc);
+ rdmsrl(event->hw.event_base, new_raw_count);
+ rdmsrl(MSR_F15H_PTSC, new_ptsc);
+
+ if (local64_cmpxchg(&hwc->prev_count, prev_raw_count,
+ new_raw_count) != prev_raw_count) {
+ cpu_relax();
+ goto again;
+ }
+
+ /*
+ * Calculate the CU power consumption over a time period, the unit of
+ * final value (delta) is micro-Watts. Then add it to the event count.
+ */
+ if (new_raw_count < prev_raw_count) {
+ delta = max_cu_acc_power + new_raw_count;
+ delta -= prev_raw_count;
+ } else
+ delta = new_raw_count - prev_raw_count;
+
+ delta *= cpu_pwr_sample_ratio * 1000;
+ tdelta = new_ptsc - prev_ptsc;
+
+ do_div(delta, tdelta);
+ local64_add(delta, &event->count);
+
+ return new_raw_count;
+}
+
+static void __pmu_event_start(struct power_pmu *pmu, struct perf_event *event)
+{
+ u64 ptsc, counts;
+
+ if (WARN_ON_ONCE(!(event->hw.state & PERF_HES_STOPPED)))
+ return;
+
+ event->hw.state = 0;
+
+ rdmsrl(MSR_F15H_PTSC, ptsc);
+ local64_set(&pmu->cpu_sw_pwr_ptsc, ptsc);
+ rdmsrl(event->hw.event_base, counts);
+ local64_set(&event->hw.prev_count, counts);
+}
+
+static void pmu_event_start(struct perf_event *event, int mode)
+{
+ struct power_pmu *pmu = __this_cpu_read(amd_power_pmu);
+
+ raw_spin_lock(&pmu->lock);
+ __pmu_event_start(pmu, event);
+ raw_spin_unlock(&pmu->lock);
+}
+
+static void pmu_event_stop(struct perf_event *event, int mode)
+{
+ struct power_pmu *pmu = __this_cpu_read(amd_power_pmu);
+ struct hw_perf_event *hwc = &event->hw;
+
+ raw_spin_lock(&pmu->lock);
+
+ /* Mark event as deactivated and stopped. */
+ if (!(hwc->state & PERF_HES_STOPPED))
+ hwc->state |= PERF_HES_STOPPED;
+
+ /* Check if software counter update is necessary. */
+ if ((mode & PERF_EF_UPDATE) && !(hwc->state & PERF_HES_UPTODATE)) {
+ /*
+ * Drain the remaining delta count out of an event
+ * that we are disabling:
+ */
+ event_update(event, pmu);
+ hwc->state |= PERF_HES_UPTODATE;
+ }
+
+ raw_spin_unlock(&pmu->lock);
+}
+
+static int pmu_event_add(struct perf_event *event, int mode)
+{
+ struct power_pmu *pmu = __this_cpu_read(amd_power_pmu);
+ struct hw_perf_event *hwc = &event->hw;
+
+ raw_spin_lock(&pmu->lock);
+
+ hwc->state = PERF_HES_UPTODATE | PERF_HES_STOPPED;
+
+ if (mode & PERF_EF_START)
+ __pmu_event_start(pmu, event);
+
+ raw_spin_unlock(&pmu->lock);
+
+ return 0;
+}
+
+static void pmu_event_del(struct perf_event *event, int flags)
+{
+ pmu_event_stop(event, PERF_EF_UPDATE);
+}
+
+static int pmu_event_init(struct perf_event *event)
+{
+ u64 cfg = event->attr.config & AMD_POWER_EVENT_MASK;
+ int ret = 0;
+
+ /* Only look at AMD power events. */
+ if (event->attr.type != pmu_class.type)
+ return -ENOENT;
+
+ /* Unsupported modes and filters. */
+ if (event->attr.exclude_user ||
+ event->attr.exclude_kernel ||
+ event->attr.exclude_hv ||
+ event->attr.exclude_idle ||
+ event->attr.exclude_host ||
+ event->attr.exclude_guest ||
+ /* no sampling */
+ event->attr.sample_period)
+ return -EINVAL;
+
+ if (cfg != AMD_POWER_EVENTSEL_PKG)
+ return -EINVAL;
+
+ event->hw.event_base = MSR_F15H_CU_PWR_ACCUMULATOR;
+ event->hw.config = cfg;
+ event->hw.idx = AMD_POWER_PKG_ID;
+
+ return ret;
+}
+
+static void pmu_event_read(struct perf_event *event)
+{
+ struct power_pmu *pmu = __this_cpu_read(amd_power_pmu);
+
+ event_update(event, pmu);
+}
+
+static ssize_t
+get_attr_cpumask(struct device *dev, struct device_attribute *attr, char *buf)
+{
+ return cpumap_print_to_pagebuf(true, buf, &cpu_mask);
+}
+
+static DEVICE_ATTR(cpumask, S_IRUGO, get_attr_cpumask, NULL);
+
+static struct attribute *pmu_attrs[] = {
+ &dev_attr_cpumask.attr,
+ NULL,
+};
+
+static struct attribute_group pmu_attr_group = {
+ .attrs = pmu_attrs,
+};
+
+/*
+ * Currently it only supports to report the power of each
+ * processor/package.
+ */
+EVENT_ATTR_STR(power-pkg, power_pkg, "event=0x01");
+
+EVENT_ATTR_STR(power-pkg.unit, power_pkg_unit, "mWatts");
+
+/* Convert the count from micro-Watts to milli-Watts. */
+EVENT_ATTR_STR(power-pkg.scale, power_pkg_scale, "1.000000e-3");
+
+
+static struct attribute *events_attr[] = {
+ EVENT_PTR(power_pkg),
+ EVENT_PTR(power_pkg_unit),
+ EVENT_PTR(power_pkg_scale),
+ NULL,
+};
+
+static struct attribute_group pmu_events_group = {
+ .name = "events",
+ .attrs = events_attr,
+};
+
+PMU_FORMAT_ATTR(event, "config:0-7");
+
+static struct attribute *formats_attr[] = {
+ &format_attr_event.attr,
+ NULL,
+};
+
+static struct attribute_group pmu_format_group = {
+ .name = "format",
+ .attrs = formats_attr,
+};
+
+static const struct attribute_group *attr_groups[] = {
+ &pmu_attr_group,
+ &pmu_format_group,
+ &pmu_events_group,
+ NULL,
+};
+
+static struct pmu pmu_class = {
+ .attr_groups = attr_groups,
+ /* system-wide only */
+ .task_ctx_nr = perf_invalid_context,
+ .event_init = pmu_event_init,
+ .add = pmu_event_add,
+ .del = pmu_event_del,
+ .start = pmu_event_start,
+ .stop = pmu_event_stop,
+ .read = pmu_event_read,
+};
+
+static int power_cpu_exit(int cpu)
+{
+ struct power_pmu *pmu = per_cpu(amd_power_pmu, cpu);
+ int target = nr_cpumask_bits;
+ int ret = 0;
+
+ cpumask_copy(pmu->mask, topology_sibling_cpumask(cpu));
+
+ cpumask_clear_cpu(cpu, &cpu_mask);
+ cpumask_clear_cpu(cpu, pmu->mask);
+
+ if (!cpumask_and(pmu->tmp_mask, pmu->mask, cpu_online_mask))
+ goto out;
+
+ /*
+ * Find a new CPU on the same compute unit, if was set in cpumask
+ * and still some CPUs on compute unit. Then move on to the new CPU.
+ */
+ target = cpumask_any(pmu->tmp_mask);
+ if (target < nr_cpumask_bits && target != cpu)
+ cpumask_set_cpu(target, &cpu_mask);
+
+ WARN_ON(cpumask_empty(&cpu_mask));
+
+out:
+ /*
+ * Migrate event and context to new CPU.
+ */
+ if (target < nr_cpumask_bits)
+ perf_pmu_migrate_context(pmu->pmu, cpu, target);
+
+ return ret;
+
+}
+
+static int power_cpu_init(int cpu)
+{
+ struct power_pmu *pmu = per_cpu(amd_power_pmu, cpu);
+
+ if (!pmu)
+ return 0;
+
+ if (!cpumask_and(pmu->mask, topology_sibling_cpumask(cpu), &cpu_mask))
+ cpumask_set_cpu(cpu, &cpu_mask);
+
+ return 0;
+}
+
+static int power_cpu_prepare(int cpu)
+{
+ struct power_pmu *pmu = per_cpu(amd_power_pmu, cpu);
+ int phys_id = topology_physical_package_id(cpu);
+ int ret = 0;
+
+ if (pmu)
+ return 0;
+
+ if (phys_id < 0)
+ return -EINVAL;
+
+ pmu = kzalloc_node(sizeof(*pmu), GFP_KERNEL, cpu_to_node(cpu));
+ if (!pmu)
+ return -ENOMEM;
+
+ if (!zalloc_cpumask_var(&pmu->mask, GFP_KERNEL)) {
+ ret = -ENOMEM;
+ goto out;
+ }
+
+ if (!zalloc_cpumask_var(&pmu->tmp_mask, GFP_KERNEL)) {
+ ret = -ENOMEM;
+ goto out1;
+ }
+
+ raw_spin_lock_init(&pmu->lock);
+
+ pmu->pmu = &pmu_class;
+
+ per_cpu(amd_power_pmu, cpu) = pmu;
+
+ return 0;
+
+out1:
+ free_cpumask_var(pmu->mask);
+out:
+ kfree(pmu);
+
+ return ret;
+}
+
+static void power_cpu_kfree(int cpu)
+{
+ struct power_pmu *pmu = per_cpu(amd_power_pmu, cpu);
+
+ if (!pmu)
+ return;
+
+ free_cpumask_var(pmu->mask);
+ free_cpumask_var(pmu->tmp_mask);
+ kfree(pmu);
+
+ per_cpu(amd_power_pmu, cpu) = NULL;
+}
+
+static int
+power_cpu_notifier(struct notifier_block *self, unsigned long action, void *hcpu)
+{
+ unsigned int cpu = (long)hcpu;
+
+ switch (action & ~CPU_TASKS_FROZEN) {
+ case CPU_UP_PREPARE:
+ if (power_cpu_prepare(cpu))
+ return NOTIFY_BAD;
+ break;
+ case CPU_STARTING:
+ if (power_cpu_init(cpu))
+ return NOTIFY_BAD;
+ break;
+ case CPU_DEAD:
+ power_cpu_kfree(cpu);
+ break;
+ case CPU_DOWN_PREPARE:
+ if (power_cpu_exit(cpu))
+ return NOTIFY_BAD;
+ break;
+ default:
+ break;
+ }
+
+ return NOTIFY_OK;
+}
+
+static const struct x86_cpu_id cpu_match[] = {
+ { .vendor = X86_VENDOR_AMD, .family = 0x15 },
+ {},
+};
+
+static int __init amd_power_pmu_init(void)
+{
+ int i, ret;
+ u64 tmp;
+
+ if (!x86_match_cpu(cpu_match))
+ return 0;
+
+ if (!boot_cpu_has(X86_FEATURE_ACC_POWER))
+ return -ENODEV;
+
+ cores_per_cu = amd_get_cores_per_cu();
+
+ cpu_pwr_sample_ratio = cpuid_ecx(0x80000007);
+
+ if (rdmsrl_safe(MSR_F15H_CU_MAX_PWR_ACCUMULATOR, &tmp)) {
+ pr_err("Failed to read max compute unit power accumulator MSR\n");
+ return -ENODEV;
+ }
+ max_cu_acc_power = tmp;
+
+ cpu_notifier_register_begin();
+
+ /* Choose one online core of each compute unit. */
+ for (i = 0; i < boot_cpu_data.x86_max_cores; i += cores_per_cu) {
+ WARN_ON(cpumask_empty(topology_sibling_cpumask(i)));
+ cpumask_set_cpu(cpumask_any(topology_sibling_cpumask(i)), &cpu_mask);
+ }
+
+ for_each_present_cpu(i) {
+ ret = power_cpu_prepare(i);
+ if (ret) {
+ /* Unwind on [0 ... i-1] CPUs. */
+ while (i--)
+ power_cpu_kfree(i);
+ goto out;
+ }
+ ret = power_cpu_init(i);
+ if (ret) {
+ /* Unwind on [0 ... i] CPUs. */
+ while (i >= 0)
+ power_cpu_kfree(i--);
+ goto out;
+ }
+ }
+
+ __perf_cpu_notifier(power_cpu_notifier);
+
+ ret = perf_pmu_register(&pmu_class, "power", -1);
+ if (WARN_ON(ret)) {
+ pr_warn("AMD Power PMU registration failed\n");
+ goto out;
+ }
+
+ pr_info("AMD Power PMU detected.\n");
+
+out:
+ cpu_notifier_register_done();
+
+ return ret;
+}
+device_initcall(amd_power_pmu_init);
--
2.3.5
--
Regards/Gruss,
Boris.
ECO tip #101: Trim your mails when you reply.
[toc] | [prev] | [next] | [standalone]
| From | Peter Zijlstra <peterz@infradead.org> |
|---|---|
| Date | 2016-01-28 16:30 +0100 |
| Message-ID | <qVWGd-5of-3@gated-at.bofh.it> |
| In reply to | #1320450 |
On Thu, Jan 28, 2016 at 10:03:15AM +0100, Borislav Petkov wrote:
> +
> +struct power_pmu {
> + raw_spinlock_t lock;
Now that the list is gone, what does this thing protect?
> + struct pmu *pmu;
This member seems superfluous, there's only the one possible value.
> + local64_t cpu_sw_pwr_ptsc;
> +
> + /*
> + * These two cpumasks are used for avoiding the allocations on the
> + * CPU_STARTING phase because power_cpu_prepare() will be called with
> + * IRQs disabled.
> + */
> + cpumask_var_t mask;
> + cpumask_var_t tmp_mask;
> +};
> +
> +static struct pmu pmu_class;
> +
> +/*
> + * Accumulated power represents the sum of each compute unit's (CU) power
> + * consumption. On any core of each CU we read the total accumulated power from
> + * MSR_F15H_CU_PWR_ACCUMULATOR. cpu_mask represents CPU bit map of all cores
> + * which are picked to measure the power for the CUs they belong to.
> + */
> +static cpumask_t cpu_mask;
> +
> +static DEFINE_PER_CPU(struct power_pmu *, amd_power_pmu);
> +
> +static u64 event_update(struct perf_event *event, struct power_pmu *pmu)
> +{
Is there ever a case where @pmu != __this_cpu_read(power_pmu) ?
> + struct hw_perf_event *hwc = &event->hw;
> + u64 prev_raw_count, new_raw_count, prev_ptsc, new_ptsc;
> + u64 delta, tdelta;
> +
> +again:
> + prev_raw_count = local64_read(&hwc->prev_count);
> + prev_ptsc = local64_read(&pmu->cpu_sw_pwr_ptsc);
> + rdmsrl(event->hw.event_base, new_raw_count);
Is hw.event_base != MSR_F15H_CU_PWR_ACCUMULATOR possible?
> + rdmsrl(MSR_F15H_PTSC, new_ptsc);
Also, I suspect this doesn't do what you expect it to do.
We measure per-event PWR_ACC deltas, but per CPU PTSC values. These do
not match when there's more than 1 event on the CPU.
I would suggest adding a new struct to the hw_perf_event union with the
two u64 deltas like:
struct { /* amd_power */
u64 pwr_acc;
u64 ptsc;
};
And track these values per-event.
> +
> + if (local64_cmpxchg(&hwc->prev_count, prev_raw_count,
> + new_raw_count) != prev_raw_count) {
> + cpu_relax();
> + goto again;
> + }
> +
> + /*
> + * Calculate the CU power consumption over a time period, the unit of
> + * final value (delta) is micro-Watts. Then add it to the event count.
> + */
> + if (new_raw_count < prev_raw_count) {
> + delta = max_cu_acc_power + new_raw_count;
> + delta -= prev_raw_count;
> + } else
> + delta = new_raw_count - prev_raw_count;
> +
> + delta *= cpu_pwr_sample_ratio * 1000;
> + tdelta = new_ptsc - prev_ptsc;
> +
> + do_div(delta, tdelta);
> + local64_add(delta, &event->count);
Then this division can be redone on the total values, that looses less
precision over-all.
> +
> + return new_raw_count;
> +}
[toc] | [prev] | [standalone]
Back to top | Article view | linux.kernel
csiph-web