x86: Enable NMI on all cpus on UV

Enable NMI on all cpus in UV system and add an NMI handler
to dump_stack on each cpu.

By default on x86 all the cpus except the boot cpu have NMI
masked off.  This patch enables NMI on all cpus in UV system
and adds an NMI handler to dump_stack on each cpu.  This
way if a system hangs we can NMI the machine and get a
backtrace from all the cpus.

Version 2: Use x86_platform driver mechanism for nmi init, per
           Ingo's suggestion.

Version 3: Clean up Ingo's nits.

Signed-off-by: Russ Anderson <rja@sgi.com>
LKML-Reference: <20100226164912.GA24439@sgi.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>

3 files changed, 48 insertions, 0 deletions

diff --git a/arch/x86/kernel/apic/x2apic_uv_x.c b/arch/x86/kernel/apic/x2apic_uv_x.c
index 6ef2899eb86..4b8dbb25614 100644
--- a/arch/x86/kernel/apic/x2apic_uv_x.c
+++ b/arch/x86/kernel/apic/x2apic_uv_x.c
@@ -21,6 +21,7 @@
 #include <linux/init.h>
 #include <linux/io.h>
 #include <linux/pci.h>
+#include <linux/kdebug.h>
 
 #include <asm/uv/uv_mmrs.h>
 #include <asm/uv/uv_hub.h>
@@ -41,6 +42,7 @@ static enum uv_system_type uv_system_type;
 static u64 gru_start_paddr, gru_end_paddr;
 int uv_min_hub_revision_id;
 EXPORT_SYMBOL_GPL(uv_min_hub_revision_id);
+static DEFINE_SPINLOCK(uv_nmi_lock);
 
 static inline bool is_GRU_range(u64 start, u64 end)
 {
@@ -74,6 +76,7 @@ static int __init uv_acpi_madt_oem_check(char *oem_id, char *oem_table_id)
 	if (!strcmp(oem_id, "SGI")) {
 		nodeid = early_get_nodeid();
 		x86_platform.is_untracked_pat_range =  uv_is_untracked_pat_range;
+		x86_platform.nmi_init = uv_nmi_init;
 		if (!strcmp(oem_table_id, "UVL"))
 			uv_system_type = UV_LEGACY_APIC;
 		else if (!strcmp(oem_table_id, "UVX"))
@@ -596,6 +599,46 @@ void __cpuinit uv_cpu_init(void)
 		set_x2apic_extra_bits(uv_hub_info->pnode);
 }
 
+/*
+ * When NMI is received, print a stack trace.
+ */
+int uv_handle_nmi(struct notifier_block *self, unsigned long reason, void *data)
+{
+	if (reason != DIE_NMI_IPI)
+		return NOTIFY_OK;
+	/*
+	 * Use a lock so only one cpu prints at a time
+	 * to prevent intermixed output.
+	 */
+	spin_lock(&uv_nmi_lock);
+	pr_info("NMI stack dump cpu %u:\n", smp_processor_id());
+	dump_stack();
+	spin_unlock(&uv_nmi_lock);
+
+	return NOTIFY_STOP;
+}
+
+static struct notifier_block uv_dump_stack_nmi_nb = {
+	.notifier_call	= uv_handle_nmi
+};
+
+void uv_register_nmi_notifier(void)
+{
+	if (register_die_notifier(&uv_dump_stack_nmi_nb))
+		printk(KERN_WARNING "UV NMI handler failed to register\n");
+}
+
+void uv_nmi_init(void)
+{
+	unsigned int value;
+
+	/*
+	 * Unmask NMI on all cpus
+	 */
+	value = apic_read(APIC_LVT1) | APIC_DM_NMI;
+	value &= ~APIC_LVT_MASKED;
+	apic_write(APIC_LVT1, value);
+}
 
 void __init uv_system_init(void)
 {
@@ -717,6 +760,7 @@ void __init uv_system_init(void)
 
 	uv_cpu_init();
 	uv_scir_register_cpu_notifier();
+	uv_register_nmi_notifier();
 	proc_mkdir("sgi_uv", NULL);
 
 	/* register Legacy VGA I/O redirection handler */
diff --git a/arch/x86/kernel/smpboot.c b/arch/x86/kernel/smpboot.c
index 678d0b8c26f..838a118876c 100644
--- a/arch/x86/kernel/smpboot.c
+++ b/arch/x86/kernel/smpboot.c
@@ -320,6 +320,7 @@ notrace static void __cpuinit start_secondary(void *unused)
 	unlock_vector_lock();
 	ipi_call_unlock();
 	per_cpu(cpu_state, smp_processor_id()) = CPU_ONLINE;
+	x86_platform.nmi_init();
 
 	/* enable local interrupts */
 	local_irq_enable();
diff --git a/arch/x86/kernel/x86_init.c b/arch/x86/kernel/x86_init.c
index ccd179dec36..ee5746c9462 100644
--- a/arch/x86/kernel/x86_init.c
+++ b/arch/x86/kernel/x86_init.c
@@ -76,10 +76,13 @@ struct x86_cpuinit_ops x86_cpuinit __cpuinitdata = {
 	.setup_percpu_clockev		= setup_secondary_APIC_clock,
 };
 
+static void default_nmi_init(void) { };
+
 struct x86_platform_ops x86_platform = {
 	.calibrate_tsc			= native_calibrate_tsc,
 	.get_wallclock			= mach_get_cmos_time,
 	.set_wallclock			= mach_set_rtc_mmss,
 	.iommu_shutdown			= iommu_shutdown_noop,
 	.is_untracked_pat_range		= is_ISA_range,
+	.nmi_init			= default_nmi_init
 };