From: Yasunori Goto <y-goto@jp.fujitsu.com>
To: Andrew Morton <akpm@osdl.org>
Cc: Linux Kernel ML <linux-kernel@vger.kernel.org>,
"Luck, Tony" <tony.luck@intel.com>, linux-mm <linux-mm@kvack.org>
Subject: [Patch 002/003] pgdat allocation and update for ia64 of memory hotplug. (update pgdat address array)
Date: Tue, 02 May 2006 20:35:54 +0900 [thread overview]
Message-ID: <20060502203247.CF1A.Y-GOTO@jp.fujitsu.com> (raw)
In-Reply-To: <20060502201614.CF14.Y-GOTO@jp.fujitsu.com>
As I mentioned previous patches,
ia64 has copies of information of pgdat address array on each node
as per node data.
At v2 of node_add, this function used stop_machine_run() to update them.
(I wished that they were copied safety as much as possible.)
But, in this patch, this arrays are just copied simply, and
set node_online_map bit after completion of pgdat initialization.
So, kernel must touch NODE_DATA() macro after checking
node_online_map(). (Current code has already done it.)
This is more simple way for just hot-add.....
Note : It will be problem when hot-remove will occur,
because, even if online_map bit is set, kernel may
touch NODE_DATA() due to race condition. :-(
Signed-off-by: Yasunori Goto <y-goto@jp.fujitsu.com>
arch/ia64/mm/discontig.c | 24 +++++++++++++++++++-----
include/asm-ia64/nodedata.h | 12 ++++++++++++
include/linux/memory_hotplug.h | 4 +---
3 files changed, 32 insertions(+), 8 deletions(-)
Index: pgdat12/arch/ia64/mm/discontig.c
===================================================================
--- pgdat12.orig/arch/ia64/mm/discontig.c 2006-04-28 10:24:56.000000000 +0900
+++ pgdat12/arch/ia64/mm/discontig.c 2006-04-28 10:31:49.000000000 +0900
@@ -308,6 +308,17 @@ static void __init reserve_pernode_space
}
}
+static void __meminit scatter_node_data(void)
+{
+ pg_data_t **dst;
+ int node;
+
+ for_each_online_node(node){
+ dst = LOCAL_DATA_ADDR(pgdat_list[node])->pg_data_ptrs;
+ memcpy(dst, pgdat_list, sizeof(pgdat_list));
+ }
+}
+
/**
* initialize_pernode_data - fixup per-cpu & per-node pointers
*
@@ -320,11 +331,8 @@ static void __init initialize_pernode_da
{
int cpu, node;
- /* Copy the pg_data_t list to each node and init the node field */
- for_each_online_node(node) {
- memcpy(mem_data[node].node_data->pg_data_ptrs, pgdat_list,
- sizeof(pgdat_list));
- }
+ scatter_node_data();
+
#ifdef CONFIG_SMP
/* Set the node_data pointer for each per-cpu struct */
for (cpu = 0; cpu < NR_CPUS; cpu++) {
@@ -783,3 +791,9 @@ void __init paging_init(void)
zero_page_memmap_ptr = virt_to_page(ia64_imva(empty_zero_page));
}
+
+void arch_refresh_nodedata(int update_node, pg_data_t *update_pgdat)
+{
+ pgdat_list[update_node] = update_pgdat;
+ scatter_node_data();
+}
Index: pgdat12/include/asm-ia64/nodedata.h
===================================================================
--- pgdat12.orig/include/asm-ia64/nodedata.h 2006-04-28 10:24:51.000000000 +0900
+++ pgdat12/include/asm-ia64/nodedata.h 2006-04-28 10:27:40.000000000 +0900
@@ -47,6 +47,18 @@ struct ia64_node_data {
*/
#define NODE_DATA(nid) (local_node_data->pg_data_ptrs[nid])
+/*
+ * LOCAL_DATA_ADDR - This is to calculate the address of other node's
+ * "local_node_data" at hot-plug phase. The local_node_data
+ * is pointed by per_cpu_page. Kernel usually use it for
+ * just executing cpu. However, when new node is hot-added,
+ * the addresses of local data for other nodes are necessary
+ * to update all of them.
+ */
+#define LOCAL_DATA_ADDR(pgdat) \
+ ((struct ia64_node_data *)((u64)(pgdat) + \
+ L1_CACHE_ALIGN(sizeof(struct pglist_data))))
+
#endif /* CONFIG_NUMA */
#endif /* _ASM_IA64_NODEDATA_H */
Index: pgdat12/include/linux/memory_hotplug.h
===================================================================
--- pgdat12.orig/include/linux/memory_hotplug.h 2006-04-28 10:24:51.000000000 +0900
+++ pgdat12/include/linux/memory_hotplug.h 2006-04-28 10:31:49.000000000 +0900
@@ -91,9 +91,7 @@ static inline pg_data_t *arch_alloc_node
static inline void arch_free_nodedata(pg_data_t *pgdat)
{
}
-static inline void arch_refresh_nodedata(int nid, pg_data_t *pgdat)
-{
-}
+extern void arch_refresh_nodedata(int nid, pg_data_t *pgdat);
#else /* CONFIG_HAVE_ARCH_NODEDATA_EXTENSION */
--
Yasunori Goto
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
next prev parent reply other threads:[~2006-05-02 11:35 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
2006-05-02 11:30 [Patch 000/003] pgdat allocation and update for ia64 of memory hotplug Yasunori Goto
2006-05-02 11:35 ` [Patch 001/003] pgdat allocation and update for ia64 of memory hotplug.(hold pgdat address at system running) Yasunori Goto
2006-05-02 11:35 ` Yasunori Goto [this message]
2006-05-02 11:35 ` [Patch 003/003] pgdat allocation and update for ia64 of memory hotplug.(allocate pgdat and per node data) Yasunori Goto
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=20060502203247.CF1A.Y-GOTO@jp.fujitsu.com \
--to=y-goto@jp.fujitsu.com \
--cc=akpm@osdl.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=tony.luck@intel.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