[PATCH] Register sysfs file for hotplugged new node

When new node becomes enable by hot-add, new sysfs file must be created for
new node.  So, if new node is enabled by add_memory(), register_one_node() is
called to create it.  In addition, I386's arch_register_node() and a part of
register_nodes() of powerpc are consolidated to register_one_node() as a
generic_code().

This is tested by Tiger4(IPF) with node hot-plug emulation.

Signed-off-by: Keiichiro Tokunaga <tokuanga.keiich@jp.fujitsu.com>
Signed-off-by: Yasunori Goto <y-goto@jp.fujitsu.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

8 files changed, 48 insertions, 63 deletions

diff --git a/arch/i386/kernel/topology.c b/arch/i386/kernel/topology.c
index 2963552..1eecc2e 100644
--- a/arch/i386/kernel/topology.c
+++ b/arch/i386/kernel/topology.c
@@ -38,7 +38,7 @@ int arch_register_cpu(int num){
 #ifdef CONFIG_NUMA
 	int node = cpu_to_node(num);
 	if (node_online(node))
-		parent = &node_devices[node].node;
+		parent = &node_devices[parent_node(node)];
 #endif /* CONFIG_NUMA */
 
 	/*
@@ -61,7 +61,7 @@ void arch_unregister_cpu(int num) {
 #ifdef CONFIG_NUMA
 	int node = cpu_to_node(num);
 	if (node_online(node))
-		parent = &node_devices[node].node;
+		parent = &node_devices[parent_node(node)];
 #endif /* CONFIG_NUMA */
 
 	return unregister_cpu(&cpu_devices[num].cpu, parent);
@@ -74,16 +74,13 @@ EXPORT_SYMBOL(arch_unregister_cpu);
 
 #ifdef CONFIG_NUMA
 #include <linux/mmzone.h>
-#include <asm/node.h>
-
-struct i386_node node_devices[MAX_NUMNODES];
 
 static int __init topology_init(void)
 {
 	int i;
 
 	for_each_online_node(i)
-		arch_register_node(i);
+		register_one_node(i);
 
 	for_each_present_cpu(i)
 		arch_register_cpu(i);
diff --git a/arch/ia64/kernel/topology.c b/arch/ia64/kernel/topology.c
index 879edb5..42cb05b 100644
--- a/arch/ia64/kernel/topology.c
+++ b/arch/ia64/kernel/topology.c
@@ -26,9 +26,6 @@
 #include <asm/numa.h>
 #include <asm/cpu.h>
 
-#ifdef CONFIG_NUMA
-static struct node *sysfs_nodes;
-#endif
 static struct ia64_cpu *sysfs_cpus;
 
 int arch_register_cpu(int num)
@@ -36,7 +33,7 @@ int arch_register_cpu(int num)
 	struct node *parent = NULL;
 	
 #ifdef CONFIG_NUMA
-	parent = &sysfs_nodes[cpu_to_node(num)];
+	parent = &node_devices[cpu_to_node(num)];
 #endif /* CONFIG_NUMA */
 
 #if defined (CONFIG_ACPI) && defined (CONFIG_HOTPLUG_CPU)
@@ -59,7 +56,7 @@ void arch_unregister_cpu(int num)
 
 #ifdef CONFIG_NUMA
 	int node = cpu_to_node(num);
-	parent = &sysfs_nodes[node];
+	parent = &node_devices[node];
 #endif /* CONFIG_NUMA */
 
 	return unregister_cpu(&sysfs_cpus[num].cpu, parent);
@@ -74,17 +71,11 @@ static int __init topology_init(void)
 	int i, err = 0;
 
 #ifdef CONFIG_NUMA
-	sysfs_nodes = kzalloc(sizeof(struct node) * MAX_NUMNODES, GFP_KERNEL);
-	if (!sysfs_nodes) {
-		err = -ENOMEM;
-		goto out;
-	}
-
 	/*
 	 * MCD - Do we want to register all ONLINE nodes, or all POSSIBLE nodes?
 	 */
 	for_each_online_node(i) {
-		if ((err = register_node(&sysfs_nodes[i], i, 0)))
+		if ((err = register_one_node(i)))
 			goto out;
 	}
 #endif
diff --git a/arch/powerpc/kernel/sysfs.c b/arch/powerpc/kernel/sysfs.c
index 5bc2585..338491d 100644
--- a/arch/powerpc/kernel/sysfs.c
+++ b/arch/powerpc/kernel/sysfs.c
@@ -304,23 +304,12 @@ static struct notifier_block sysfs_cpu_nb = {
 /* NUMA stuff */
 
 #ifdef CONFIG_NUMA
-static struct node node_devices[MAX_NUMNODES];
-
 static void register_nodes(void)
 {
 	int i;
 
-	for (i = 0; i < MAX_NUMNODES; i++) {
-		if (node_online(i)) {
-			int p_node = parent_node(i);
-			struct node *parent = NULL;
-
-			if (p_node != i)
-				parent = &node_devices[p_node];
-
-			register_node(&node_devices[i], i, parent);
-		}
-	}
+	for (i = 0; i < MAX_NUMNODES; i++)
+		register_one_node(i);
 }
 
 int sysfs_add_device_to_node(struct sys_device *dev, int nid)
diff --git a/drivers/base/node.c b/drivers/base/node.c
index c80c3ae..cbd0f62 100644
--- a/drivers/base/node.c
+++ b/drivers/base/node.c
@@ -190,6 +190,31 @@ void unregister_node(struct node *node)
 	sysdev_unregister(&node->sysdev);
 }
 
+struct node node_devices[MAX_NUMNODES];
+
+int register_one_node(int nid)
+{
+	int error = 0;
+
+	if (node_online(nid)) {
+		int p_node = parent_node(nid);
+		struct node *parent = NULL;
+
+		if (p_node != nid)
+			parent = &node_devices[p_node];
+
+		error = register_node(&node_devices[nid], nid, parent);
+	}
+
+	return error;
+
+}
+
+void unregister_one_node(int nid)
+{
+	unregister_node(&node_devices[nid]);
+}
+
 static int __init register_node_type(void)
 {
 	return sysdev_class_register(&node_class);
diff --git a/include/asm-i386/cpu.h b/include/asm-i386/cpu.h
index e7252c2..b1bc7b1 100644
--- a/include/asm-i386/cpu.h
+++ b/include/asm-i386/cpu.h
@@ -7,8 +7,6 @@
 #include <linux/nodemask.h>
 #include <linux/percpu.h>
 
-#include <asm/node.h>
-
 struct i386_cpu {
 	struct cpu cpu;
 };
diff --git a/include/asm-i386/node.h b/include/asm-i386/node.h
deleted file mode 100644
index e13c6ff..0000000
--- a/include/asm-i386/node.h
+++ /dev/null
@@ -1,29 +0,0 @@
-#ifndef _ASM_I386_NODE_H_
-#define _ASM_I386_NODE_H_
-
-#include <linux/device.h>
-#include <linux/mmzone.h>
-#include <linux/node.h>
-#include <linux/topology.h>
-#include <linux/nodemask.h>
-
-struct i386_node {
-	struct node node;
-};
-extern struct i386_node node_devices[MAX_NUMNODES];
-
-static inline int arch_register_node(int num){
-	int p_node;
-	struct node *parent = NULL;
-
-	if (!node_online(num))
-		return 0;
-	p_node = parent_node(num);
-
-	if (p_node != num)
-		parent = &node_devices[p_node].node;
-
-	return register_node(&node_devices[num].node, num, parent);
-}
-
-#endif /* _ASM_I386_NODE_H_ */
diff --git a/include/linux/node.h b/include/linux/node.h
index 254dc3d..1e53475 100644
--- a/include/linux/node.h
+++ b/include/linux/node.h
@@ -26,8 +26,12 @@ struct node {
 	struct sys_device	sysdev;
 };
 
+extern struct node node_devices[];
+
 extern int register_node(struct node *, int, struct node *);
 extern void unregister_node(struct node *node);
+extern int register_one_node(int nid);
+extern void unregister_one_node(int nid);
 
 #define to_node(sys_device) container_of(sys_device, struct node, sysdev)
 
diff --git a/mm/memory_hotplug.c b/mm/memory_hotplug.c
index f13783e..ea40388 100644
--- a/mm/memory_hotplug.c
+++ b/mm/memory_hotplug.c
@@ -256,9 +256,19 @@ int add_memory(int nid, u64 start, u64 size)
 	if (ret < 0)
 		goto error;
 
-	/* we online node here. we have no error path from here. */
+	/* we online node here. we can't roll back from here. */
 	node_set_online(nid);
 
+	if (new_pgdat) {
+		ret = register_one_node(nid);
+		/*
+		 * If sysfs file of new node can't create, cpu on the node
+		 * can't be hot-added. There is no rollback way now.
+		 * So, check by BUG_ON() to catch it reluctantly..
+		 */
+		BUG_ON(ret);
+	}
+
 	/* register this memory as resource */
 	register_memory_resource(start, size);