From: Daniel Wagner <wagi@monom.org>
Date: Tue, 17 Feb 2015 09:37:44 +0100
Subject: thermal: Defer thermal wakups to threads
Git-repo: git://git.kernel.org/pub/scm/linux/kernel/git/rt/linux-rt-devel.git
Git-commit: 86ef0d4ca9dfde58c5667b42151604f744668c72
Patch-mainline: Queued in subsystem maintainer repository
References: SLE Realtime Extension

On RT the spin lock in pkg_temp_thermal_platfrom_thermal_notify will
call schedule while we run in irq context.

[<ffffffff816850ac>] dump_stack+0x4e/0x8f
[<ffffffff81680f7d>] __schedule_bug+0xa6/0xb4
[<ffffffff816896b4>] __schedule+0x5b4/0x700
[<ffffffff8168982a>] schedule+0x2a/0x90
[<ffffffff8168a8b5>] rt_spin_lock_slowlock+0xe5/0x2d0
[<ffffffff8168afd5>] rt_spin_lock+0x25/0x30
[<ffffffffa03a7b75>] pkg_temp_thermal_platform_thermal_notify+0x45/0x134 [x86_pkg_temp_thermal]
[<ffffffff8103d4db>] ? therm_throt_process+0x1b/0x160
[<ffffffff8103d831>] intel_thermal_interrupt+0x211/0x250
[<ffffffff8103d8c1>] smp_thermal_interrupt+0x21/0x40
[<ffffffff8169415d>] thermal_interrupt+0x6d/0x80

Let's defer the work to a kthread.

Signed-off-by: Daniel Wagner <daniel.wagner@bmw-carit.de>
[bigeasy: reoder init/denit position. TODO: flush swork on exit]
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Mike Galbraith <mgalbraith@suse.de>
---
 drivers/thermal/x86_pkg_temp_thermal.c |   52 +++++++++++++++++++++++++++++++--
 1 file changed, 49 insertions(+), 3 deletions(-)

--- a/drivers/thermal/x86_pkg_temp_thermal.c
+++ b/drivers/thermal/x86_pkg_temp_thermal.c
@@ -29,6 +29,7 @@
 #include <linux/pm.h>
 #include <linux/thermal.h>
 #include <linux/debugfs.h>
+#include <linux/swork.h>
 #include <asm/cpu_device_id.h>
 #include <asm/mce.h>
 
@@ -329,7 +330,7 @@ static void pkg_thermal_schedule_work(in
 	schedule_delayed_work_on(cpu, work, ms);
 }
 
-static int pkg_thermal_notify(u64 msr_val)
+static void pkg_thermal_notify_work(struct swork_event *event)
 {
 	int cpu = smp_processor_id();
 	struct zone_device *zonedev;
@@ -348,9 +349,47 @@ static int pkg_thermal_notify(u64 msr_va
 	}
 
 	spin_unlock_irqrestore(&pkg_temp_lock, flags);
+}
+
+#ifdef CONFIG_PREEMPT_RT_FULL
+static struct swork_event notify_work;
+
+static int pkg_thermal_notify_work_init(void)
+{
+	int err;
+
+	err = swork_get();
+	if (err)
+		return err;
+
+	INIT_SWORK(&notify_work, pkg_thermal_notify_work);
 	return 0;
 }
 
+static void pkg_thermal_notify_work_cleanup(void)
+{
+	swork_put();
+}
+
+static int pkg_thermal_notify(u64 msr_val)
+{
+	swork_queue(&notify_work);
+	return 0;
+}
+
+#else  /* !CONFIG_PREEMPT_RT_FULL */
+
+static int pkg_thermal_notify_work_init(void) { return 0; }
+
+static void pkg_thermal_notify_work_cleanup(void) {  }
+
+static int pkg_thermal_notify(u64 msr_val)
+{
+	pkg_thermal_notify_work(NULL);
+	return 0;
+}
+#endif /* CONFIG_PREEMPT_RT_FULL */
+
 static int pkg_temp_thermal_device_add(unsigned int cpu)
 {
 	int id = topology_logical_die_id(cpu);
@@ -515,10 +554,15 @@ static int __init pkg_temp_thermal_init(
 	if (!x86_match_cpu(pkg_temp_thermal_ids))
 		return -ENODEV;
 
+	if (!pkg_thermal_notify_work_init())
+		return -ENODEV;
+
 	max_id = topology_max_packages() * topology_max_die_per_package();
 	zones = kcalloc(max_id, sizeof(struct zone_device *), GFP_KERNEL);
-	if (!zones)
-		return -ENOMEM;
+	if (!zones) {
+		ret = -ENOMEM;
+		goto err;
+	}
 
 	ret = cpuhp_setup_state(CPUHP_AP_ONLINE_DYN, "thermal/x86_pkg:online",
 				pkg_thermal_cpu_online,	pkg_thermal_cpu_offline);
@@ -536,6 +580,7 @@ static int __init pkg_temp_thermal_init(
 	return 0;
 
 err:
+	pkg_thermal_notify_work_cleanup();
 	kfree(zones);
 	return ret;
 }
@@ -549,6 +594,7 @@ static void __exit pkg_temp_thermal_exit
 	cpuhp_remove_state(pkg_thermal_hp_state);
 	debugfs_remove_recursive(debugfs);
 	kfree(zones);
+	pkg_thermal_notify_work_cleanup();
 }
 module_exit(pkg_temp_thermal_exit)