public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
From: Thomas Gleixner <tglx@linutronix.de>
To: LKML <linux-kernel@vger.kernel.org>
Cc: x86@kernel.org, Peter Zijlstra <peterz@infradead.org>,
	Sebastian Siewior <bigeasy@linutronix.de>,
	Yasuaki Ishimatsu <yasu.isimatu@gmail.com>
Subject: [patch 1/3] perf/x86/intel/rapl: Make package handling more robust
Date: Tue, 31 Jan 2017 23:58:38 +0100	[thread overview]
Message-ID: <20170131230141.212593966@linutronix.de> (raw)
In-Reply-To: 20170131225837.646254998@linutronix.de

[-- Attachment #1: perf-x86-intel-rapl--Make-package-handling-more-robust.patch --]
[-- Type: text/plain, Size: 4910 bytes --]

The package management code in RAPL relies on package mapping being
available before a CPU is started. This changed with commit 9d85eb9119f4
because the ACPI/BIOS information turned out to be unreliable, but that
left RAPL in broken state. This was not noticed because on a regular boot
all CPUs are online before RAPL is initialized.

A possible fix would be to reintroduce the mess which allocates a package
data structure in CPU prepare and when it turns out to already exist in
starting throw it away later in the CPU online callback. But that's a
horrible hack and not required at all because RAPL becomes functional for
perf only in the CPU online callback. That's correct because user space is
not yet informed about the CPU being onlined, so nothing caan rely on RAPL
being available on that particular CPU.

Move the allocation to the CPU online callback and simplify the hotplug
handling. At this point the package mapping is established and correct.

This also adds a missing check for available package data in the
event_init() function.

Fixes: 9d85eb9119f4 ("x86/smpboot: Make logical package management more robust")
Reported-by: Yasuaki Ishimatsu <yasu.isimatu@gmail.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Peter Zijlstra <peterz@infradead.org>
---
 arch/x86/events/intel/rapl.c |   60 ++++++++++++++++++-------------------------
 include/linux/cpuhotplug.h   |    1 
 2 files changed, 26 insertions(+), 35 deletions(-)

--- a/arch/x86/events/intel/rapl.c
+++ b/arch/x86/events/intel/rapl.c
@@ -161,7 +161,13 @@ static u64 rapl_timer_ms;
 
 static inline struct rapl_pmu *cpu_to_rapl_pmu(unsigned int cpu)
 {
-	return rapl_pmus->pmus[topology_logical_package_id(cpu)];
+	unsigned int pkgid = topology_logical_package_id(cpu);
+
+	/*
+	 * The unsigned check also catches the '-1' return value for non
+	 * existent mappings in the topology map.
+	 */
+	return pkgid < rapl_pmus->maxpkg ? rapl_pmus->pmus[pkgid] : NULL;
 }
 
 static inline u64 rapl_read_counter(struct perf_event *event)
@@ -402,6 +408,8 @@ static int rapl_pmu_event_init(struct pe
 
 	/* must be done before validate_group */
 	pmu = cpu_to_rapl_pmu(event->cpu);
+	if (!pmu)
+		return -EINVAL;
 	event->cpu = pmu->cpu;
 	event->pmu_private = pmu;
 	event->hw.event_base = msr;
@@ -585,6 +593,20 @@ static int rapl_cpu_online(unsigned int
 	struct rapl_pmu *pmu = cpu_to_rapl_pmu(cpu);
 	int target;
 
+	if (!pmu) {
+		pmu = kzalloc_node(sizeof(*pmu), GFP_KERNEL, cpu_to_node(cpu));
+		if (!pmu)
+			return -ENOMEM;
+
+		raw_spin_lock_init(&pmu->lock);
+		INIT_LIST_HEAD(&pmu->active_list);
+		pmu->pmu = &rapl_pmus->pmu;
+		pmu->timer_interval = ms_to_ktime(rapl_timer_ms);
+		rapl_hrtimer_init(pmu);
+
+		rapl_pmus->pmus[topology_logical_package_id(cpu)] = pmu;
+	}
+
 	/*
 	 * Check if there is an online cpu in the package which collects rapl
 	 * events already.
@@ -598,27 +620,6 @@ static int rapl_cpu_online(unsigned int
 	return 0;
 }
 
-static int rapl_cpu_prepare(unsigned int cpu)
-{
-	struct rapl_pmu *pmu = cpu_to_rapl_pmu(cpu);
-
-	if (pmu)
-		return 0;
-
-	pmu = kzalloc_node(sizeof(*pmu), GFP_KERNEL, cpu_to_node(cpu));
-	if (!pmu)
-		return -ENOMEM;
-
-	raw_spin_lock_init(&pmu->lock);
-	INIT_LIST_HEAD(&pmu->active_list);
-	pmu->pmu = &rapl_pmus->pmu;
-	pmu->timer_interval = ms_to_ktime(rapl_timer_ms);
-	pmu->cpu = -1;
-	rapl_hrtimer_init(pmu);
-	rapl_pmus->pmus[topology_logical_package_id(cpu)] = pmu;
-	return 0;
-}
-
 static int rapl_check_hw_unit(bool apply_quirk)
 {
 	u64 msr_rapl_power_unit_bits;
@@ -803,29 +804,21 @@ static int __init rapl_pmu_init(void)
 	/*
 	 * Install callbacks. Core will call them for each online cpu.
 	 */
-
-	ret = cpuhp_setup_state(CPUHP_PERF_X86_RAPL_PREP, "perf/x86/rapl:prepare",
-				rapl_cpu_prepare, NULL);
-	if (ret)
-		goto out;
-
 	ret = cpuhp_setup_state(CPUHP_AP_PERF_X86_RAPL_ONLINE,
 				"perf/x86/rapl:online",
 				rapl_cpu_online, rapl_cpu_offline);
 	if (ret)
-		goto out1;
+		goto out;
 
 	ret = perf_pmu_register(&rapl_pmus->pmu, "power", -1);
 	if (ret)
-		goto out2;
+		goto out1;
 
 	rapl_advertise();
 	return 0;
 
-out2:
-	cpuhp_remove_state(CPUHP_AP_PERF_X86_RAPL_ONLINE);
 out1:
-	cpuhp_remove_state(CPUHP_PERF_X86_RAPL_PREP);
+	cpuhp_remove_state(CPUHP_AP_PERF_X86_RAPL_ONLINE);
 out:
 	pr_warn("Initialization failed (%d), disabled\n", ret);
 	cleanup_rapl_pmus();
@@ -836,7 +829,6 @@ module_init(rapl_pmu_init);
 static void __exit intel_rapl_exit(void)
 {
 	cpuhp_remove_state_nocalls(CPUHP_AP_PERF_X86_RAPL_ONLINE);
-	cpuhp_remove_state_nocalls(CPUHP_PERF_X86_RAPL_PREP);
 	perf_pmu_unregister(&rapl_pmus->pmu);
 	cleanup_rapl_pmus();
 }
--- a/include/linux/cpuhotplug.h
+++ b/include/linux/cpuhotplug.h
@@ -10,7 +10,6 @@ enum cpuhp_state {
 	CPUHP_PERF_X86_PREPARE,
 	CPUHP_PERF_X86_UNCORE_PREP,
 	CPUHP_PERF_X86_AMD_UNCORE_PREP,
-	CPUHP_PERF_X86_RAPL_PREP,
 	CPUHP_PERF_BFIN,
 	CPUHP_PERF_POWER,
 	CPUHP_PERF_SUPERH,

  reply	other threads:[~2017-01-31 23:03 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2017-01-31 22:58 [patch 0/3] perf/x86/intel: Fix RAPL and UNCORE hotplug issues Thomas Gleixner
2017-01-31 22:58 ` Thomas Gleixner [this message]
2017-02-01  9:43   ` [tip:perf/urgent] perf/x86/intel/rapl: Make package handling more robust tip-bot for Thomas Gleixner
2017-01-31 22:58 ` [patch 2/3] perf/x86/intel/uncore: Cleanup hotplug conversion fallout Thomas Gleixner
2017-02-01  9:43   ` [tip:perf/urgent] perf/x86/intel/uncore: Clean up " tip-bot for Thomas Gleixner
2017-01-31 22:58 ` [patch 3/3] perf/x86/intel/uncore: Make package handling more robust Thomas Gleixner
2017-02-01  9:44   ` [tip:perf/urgent] " tip-bot for Thomas Gleixner
2017-02-01 19:52 ` [patch 0/3] perf/x86/intel: Fix RAPL and UNCORE hotplug issues Yasuaki Ishimatsu

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20170131230141.212593966@linutronix.de \
    --to=tglx@linutronix.de \
    --cc=bigeasy@linutronix.de \
    --cc=linux-kernel@vger.kernel.org \
    --cc=peterz@infradead.org \
    --cc=x86@kernel.org \
    --cc=yasu.isimatu@gmail.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox