aboutsummaryrefslogtreecommitdiff
diff options
context:
space:
mode:
authorLinus Torvalds <torvalds@linux-foundation.org>2009-09-11 13:15:55 -0700
committerLinus Torvalds <torvalds@linux-foundation.org>2009-09-11 13:15:55 -0700
commit989aa44a5f215427085ccfe65c1e378e3399c9eb (patch)
treecd1993f2187e17e0e8d46d4fe12848e7be21fe96
parent4004f02d7af9bb1f5fd993fba60ed5bd0f5397a9 (diff)
parent47cab6a722d44c71c4f8224017ef548522243cf4 (diff)
Merge branch 'core-debug-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip
* 'core-debug-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip: debug lockups: Improve lockup detection, fix generic arch fallback debug lockups: Improve lockup detection
-rw-r--r--arch/sparc/include/asm/irq_64.h4
-rw-r--r--arch/sparc/kernel/process_64.c4
-rw-r--r--arch/x86/include/asm/nmi.h4
-rw-r--r--arch/x86/kernel/apic/nmi.c20
-rw-r--r--drivers/char/sysrq.c19
-rw-r--r--include/linux/nmi.h19
-rw-r--r--kernel/rcutree.c7
7 files changed, 56 insertions, 21 deletions
diff --git a/arch/sparc/include/asm/irq_64.h b/arch/sparc/include/asm/irq_64.h
index 1934f2cbf51..a0b443cb3c1 100644
--- a/arch/sparc/include/asm/irq_64.h
+++ b/arch/sparc/include/asm/irq_64.h
@@ -89,8 +89,8 @@ static inline unsigned long get_softint(void)
return retval;
}
-void __trigger_all_cpu_backtrace(void);
-#define trigger_all_cpu_backtrace() __trigger_all_cpu_backtrace()
+void arch_trigger_all_cpu_backtrace(void);
+#define arch_trigger_all_cpu_backtrace arch_trigger_all_cpu_backtrace
extern void *hardirq_stack[NR_CPUS];
extern void *softirq_stack[NR_CPUS];
diff --git a/arch/sparc/kernel/process_64.c b/arch/sparc/kernel/process_64.c
index 4041f94e772..18d67854a1b 100644
--- a/arch/sparc/kernel/process_64.c
+++ b/arch/sparc/kernel/process_64.c
@@ -251,7 +251,7 @@ static void __global_reg_poll(struct global_reg_snapshot *gp)
}
}
-void __trigger_all_cpu_backtrace(void)
+void arch_trigger_all_cpu_backtrace(void)
{
struct thread_info *tp = current_thread_info();
struct pt_regs *regs = get_irq_regs();
@@ -304,7 +304,7 @@ void __trigger_all_cpu_backtrace(void)
static void sysrq_handle_globreg(int key, struct tty_struct *tty)
{
- __trigger_all_cpu_backtrace();
+ arch_trigger_all_cpu_backtrace();
}
static struct sysrq_key_op sparc_globalreg_op = {
diff --git a/arch/x86/include/asm/nmi.h b/arch/x86/include/asm/nmi.h
index c86e5ed4af5..e63cf7d441e 100644
--- a/arch/x86/include/asm/nmi.h
+++ b/arch/x86/include/asm/nmi.h
@@ -45,8 +45,8 @@ extern int proc_nmi_enabled(struct ctl_table *, int , struct file *,
void __user *, size_t *, loff_t *);
extern int unknown_nmi_panic;
-void __trigger_all_cpu_backtrace(void);
-#define trigger_all_cpu_backtrace() __trigger_all_cpu_backtrace()
+void arch_trigger_all_cpu_backtrace(void);
+#define arch_trigger_all_cpu_backtrace arch_trigger_all_cpu_backtrace
static inline void localise_nmi_watchdog(void)
{
diff --git a/arch/x86/kernel/apic/nmi.c b/arch/x86/kernel/apic/nmi.c
index b3025b43b63..db7220220d0 100644
--- a/arch/x86/kernel/apic/nmi.c
+++ b/arch/x86/kernel/apic/nmi.c
@@ -39,7 +39,7 @@
int unknown_nmi_panic;
int nmi_watchdog_enabled;
-static cpumask_var_t backtrace_mask;
+static cpumask_t backtrace_mask __read_mostly;
/* nmi_active:
* >0: the lapic NMI watchdog is active, but can be disabled
@@ -138,7 +138,6 @@ int __init check_nmi_watchdog(void)
if (!prev_nmi_count)
goto error;
- alloc_cpumask_var(&backtrace_mask, GFP_KERNEL|__GFP_ZERO);
printk(KERN_INFO "Testing NMI watchdog ... ");
#ifdef CONFIG_SMP
@@ -415,14 +414,17 @@ nmi_watchdog_tick(struct pt_regs *regs, unsigned reason)
}
/* We can be called before check_nmi_watchdog, hence NULL check. */
- if (backtrace_mask != NULL && cpumask_test_cpu(cpu, backtrace_mask)) {
+ if (cpumask_test_cpu(cpu, &backtrace_mask)) {
static DEFINE_SPINLOCK(lock); /* Serialise the printks */
spin_lock(&lock);
printk(KERN_WARNING "NMI backtrace for cpu %d\n", cpu);
+ show_regs(regs);
dump_stack();
spin_unlock(&lock);
- cpumask_clear_cpu(cpu, backtrace_mask);
+ cpumask_clear_cpu(cpu, &backtrace_mask);
+
+ rc = 1;
}
/* Could check oops_in_progress here too, but it's safer not to */
@@ -552,14 +554,18 @@ int do_nmi_callback(struct pt_regs *regs, int cpu)
return 0;
}
-void __trigger_all_cpu_backtrace(void)
+void arch_trigger_all_cpu_backtrace(void)
{
int i;
- cpumask_copy(backtrace_mask, cpu_online_mask);
+ cpumask_copy(&backtrace_mask, cpu_online_mask);
+
+ printk(KERN_INFO "sending NMI to all CPUs:\n");
+ apic->send_IPI_all(NMI_VECTOR);
+
/* Wait for up to 10 seconds for all CPUs to do the backtrace */
for (i = 0; i < 10 * 1000; i++) {
- if (cpumask_empty(backtrace_mask))
+ if (cpumask_empty(&backtrace_mask))
break;
mdelay(1);
}
diff --git a/drivers/char/sysrq.c b/drivers/char/sysrq.c
index 5d7a02f63e1..50eecfe1d72 100644
--- a/drivers/char/sysrq.c
+++ b/drivers/char/sysrq.c
@@ -24,6 +24,7 @@
#include <linux/sysrq.h>
#include <linux/kbd_kern.h>
#include <linux/proc_fs.h>
+#include <linux/nmi.h>
#include <linux/quotaops.h>
#include <linux/perf_counter.h>
#include <linux/kernel.h>
@@ -222,12 +223,20 @@ static DECLARE_WORK(sysrq_showallcpus, sysrq_showregs_othercpus);
static void sysrq_handle_showallcpus(int key, struct tty_struct *tty)
{
- struct pt_regs *regs = get_irq_regs();
- if (regs) {
- printk(KERN_INFO "CPU%d:\n", smp_processor_id());
- show_regs(regs);
+ /*
+ * Fall back to the workqueue based printing if the
+ * backtrace printing did not succeed or the
+ * architecture has no support for it:
+ */
+ if (!trigger_all_cpu_backtrace()) {
+ struct pt_regs *regs = get_irq_regs();
+
+ if (regs) {
+ printk(KERN_INFO "CPU%d:\n", smp_processor_id());
+ show_regs(regs);
+ }
+ schedule_work(&sysrq_showallcpus);
}
- schedule_work(&sysrq_showallcpus);
}
static struct sysrq_key_op sysrq_showallcpus_op = {
diff --git a/include/linux/nmi.h b/include/linux/nmi.h
index 29af2d5df09..b752e807add 100644
--- a/include/linux/nmi.h
+++ b/include/linux/nmi.h
@@ -28,8 +28,23 @@ static inline void acpi_nmi_disable(void) { }
static inline void acpi_nmi_enable(void) { }
#endif
-#ifndef trigger_all_cpu_backtrace
-#define trigger_all_cpu_backtrace() do { } while (0)
+/*
+ * Create trigger_all_cpu_backtrace() out of the arch-provided
+ * base function. Return whether such support was available,
+ * to allow calling code to fall back to some other mechanism:
+ */
+#ifdef arch_trigger_all_cpu_backtrace
+static inline bool trigger_all_cpu_backtrace(void)
+{
+ arch_trigger_all_cpu_backtrace();
+
+ return true;
+}
+#else
+static inline bool trigger_all_cpu_backtrace(void)
+{
+ return false;
+}
#endif
#endif
diff --git a/kernel/rcutree.c b/kernel/rcutree.c
index 7717b95c202..9c5fa9fc57e 100644
--- a/kernel/rcutree.c
+++ b/kernel/rcutree.c
@@ -35,6 +35,7 @@
#include <linux/rcupdate.h>
#include <linux/interrupt.h>
#include <linux/sched.h>
+#include <linux/nmi.h>
#include <asm/atomic.h>
#include <linux/bitops.h>
#include <linux/module.h>
@@ -469,6 +470,8 @@ static void print_other_cpu_stall(struct rcu_state *rsp)
}
printk(" (detected by %d, t=%ld jiffies)\n",
smp_processor_id(), (long)(jiffies - rsp->gp_start));
+ trigger_all_cpu_backtrace();
+
force_quiescent_state(rsp, 0); /* Kick them all. */
}
@@ -479,12 +482,14 @@ static void print_cpu_stall(struct rcu_state *rsp)
printk(KERN_ERR "INFO: RCU detected CPU %d stall (t=%lu jiffies)\n",
smp_processor_id(), jiffies - rsp->gp_start);
- dump_stack();
+ trigger_all_cpu_backtrace();
+
spin_lock_irqsave(&rnp->lock, flags);
if ((long)(jiffies - rsp->jiffies_stall) >= 0)
rsp->jiffies_stall =
jiffies + RCU_SECONDS_TILL_STALL_RECHECK;
spin_unlock_irqrestore(&rnp->lock, flags);
+
set_need_resched(); /* kick ourselves to get things going. */
}