The patch titled
Subject: mm/hotplug: correctly setup fallback zonelists when creating new pgdat
has been added to the -mm tree. Its filename is
mm-hotplug-correctly-setup-fallback-zonelists-when-creating-new-pgdat.patch
Before you just go and hit "reply", please:
a) Consider who else should be cc'ed
b) Prefer to cc a suitable mailing list as well
c) Ideally: find the original patch on the mailing list and do a
reply-to-all to that, adding suitable additional cc's
*** Remember to use Documentation/SubmitChecklist when testing your code ***
The -mm tree is included into linux-next and is updated
there every 3-4 working days
------------------------------------------------------
From: Jiang Liu <jiang.liu@xxxxxxxxxx>
Subject: mm/hotplug: correctly setup fallback zonelists when creating new pgdat
When hotadd_new_pgdat() is called to create new pgdat for a new node, a
fallback zonelist should be created for the new node. There's code to try
to achieve that in hotadd_new_pgdat() as below:
/*
* The node we allocated has no zone fallback lists. For avoiding
* to access not-initialized zonelist, build here.
*/
mutex_lock(&zonelists_mutex);
build_all_zonelists(pgdat, NULL);
mutex_unlock(&zonelists_mutex);
But it doesn't work as expected. When hotadd_new_pgdat() is called, the
new node is still in offline state because node_set_online(nid) hasn't
been called yet. And build_all_zonelists() only builds zonelists for
online nodes as:
for_each_online_node(nid) {
pg_data_t *pgdat = NODE_DATA(nid);
build_zonelists(pgdat);
build_zonelist_cache(pgdat);
}
Though we hope to create zonelist for the new pgdat, but it doesn't. So
add a new parameter "pgdat" the build_all_zonelists() to build pgdat for
the new pgdat too.
Signed-off-by: Jiang Liu <liuj97@xxxxxxxxx>
Signed-off-by: Xishi Qiu <qiuxishi@xxxxxxxxxx>
Cc: Mel Gorman <mgorman@xxxxxxx>
Cc: Michal Hocko <mhocko@xxxxxxx>
Cc: Minchan Kim <minchan@xxxxxxxxxx>
Cc: Rusty Russell <rusty@xxxxxxxxxxxxxxx>
Cc: Yinghai Lu <yinghai@xxxxxxxxxx>
Cc: Tony Luck <tony.luck@xxxxxxxxx>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@xxxxxxxxxxxxxx>
Cc: KOSAKI Motohiro <kosaki.motohiro@xxxxxxxxxxxxxx>
Cc: David Rientjes <rientjes@xxxxxxxxxx>
Cc: Keping Chen <chenkeping@xxxxxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
---
include/linux/mmzone.h | 2 +-
init/main.c | 2 +-
kernel/cpu.c | 2 +-
mm/memory_hotplug.c | 4 ++--
mm/page_alloc.c | 17 ++++++++++++-----
5 files changed, 17 insertions(+), 10 deletions(-)
diff -puN include/linux/mmzone.h~mm-hotplug-correctly-setup-fallback-zonelists-when-creating-new-pgdat include/linux/mmzone.h
--- a/include/linux/mmzone.h~mm-hotplug-correctly-setup-fallback-zonelists-when-creating-new-pgdat
+++ a/include/linux/mmzone.h
@@ -720,7 +720,7 @@ typedef struct pglist_data {
#include <linux/memory_hotplug.h>
extern struct mutex zonelists_mutex;
-void build_all_zonelists(void *data);
+void build_all_zonelists(pg_data_t *pgdat, struct zone *zone);
void wakeup_kswapd(struct zone *zone, int order, enum zone_type classzone_idx);
bool zone_watermark_ok(struct zone *z, int order, unsigned long mark,
int classzone_idx, int alloc_flags);
diff -puN init/main.c~mm-hotplug-correctly-setup-fallback-zonelists-when-creating-new-pgdat init/main.c
--- a/init/main.c~mm-hotplug-correctly-setup-fallback-zonelists-when-creating-new-pgdat
+++ a/init/main.c
@@ -506,7 +506,7 @@ asmlinkage void __init start_kernel(void
setup_per_cpu_areas();
smp_prepare_boot_cpu(); /* arch-specific boot-cpu hooks */
- build_all_zonelists(NULL);
+ build_all_zonelists(NULL, NULL);
page_alloc_init();
printk(KERN_NOTICE "Kernel command line: %s\n", boot_command_line);
diff -puN kernel/cpu.c~mm-hotplug-correctly-setup-fallback-zonelists-when-creating-new-pgdat kernel/cpu.c
--- a/kernel/cpu.c~mm-hotplug-correctly-setup-fallback-zonelists-when-creating-new-pgdat
+++ a/kernel/cpu.c
@@ -416,7 +416,7 @@ int __cpuinit cpu_up(unsigned int cpu)
if (pgdat->node_zonelists->_zonerefs->zone == NULL) {
mutex_lock(&zonelists_mutex);
- build_all_zonelists(NULL);
+ build_all_zonelists(NULL, NULL);
mutex_unlock(&zonelists_mutex);
}
#endif
diff -puN mm/memory_hotplug.c~mm-hotplug-correctly-setup-fallback-zonelists-when-creating-new-pgdat mm/memory_hotplug.c
--- a/mm/memory_hotplug.c~mm-hotplug-correctly-setup-fallback-zonelists-when-creating-new-pgdat
+++ a/mm/memory_hotplug.c
@@ -513,7 +513,7 @@ int __ref online_pages(unsigned long pfn
zone->present_pages += onlined_pages;
zone->zone_pgdat->node_present_pages += onlined_pages;
if (need_zonelists_rebuild)
- build_all_zonelists(zone);
+ build_all_zonelists(NULL, zone);
else
zone_pcp_update(zone);
@@ -562,7 +562,7 @@ static pg_data_t __ref *hotadd_new_pgdat
* to access not-initialized zonelist, build here.
*/
mutex_lock(&zonelists_mutex);
- build_all_zonelists(NULL);
+ build_all_zonelists(pgdat, NULL);
mutex_unlock(&zonelists_mutex);
return pgdat;
diff -puN mm/page_alloc.c~mm-hotplug-correctly-setup-fallback-zonelists-when-creating-new-pgdat mm/page_alloc.c
--- a/mm/page_alloc.c~mm-hotplug-correctly-setup-fallback-zonelists-when-creating-new-pgdat
+++ a/mm/page_alloc.c
@@ -3031,7 +3031,7 @@ int numa_zonelist_order_handler(ctl_tabl
user_zonelist_order = oldval;
} else if (oldval != user_zonelist_order) {
mutex_lock(&zonelists_mutex);
- build_all_zonelists(NULL);
+ build_all_zonelists(NULL, NULL);
mutex_unlock(&zonelists_mutex);
}
}
@@ -3414,10 +3414,17 @@ static __init_refok int __build_all_zone
{
int nid;
int cpu;
+ pg_data_t *self = data;
#ifdef CONFIG_NUMA
memset(node_load, 0, sizeof(node_load));
#endif
+
+ if (self && !node_online(self->node_id)) {
+ build_zonelists(self);
+ build_zonelist_cache(self);
+ }
+
for_each_online_node(nid) {
pg_data_t *pgdat = NODE_DATA(nid);
@@ -3462,7 +3469,7 @@ static __init_refok int __build_all_zone
* Called with zonelists_mutex held always
* unless system_state == SYSTEM_BOOTING.
*/
-void __ref build_all_zonelists(void *data)
+void __ref build_all_zonelists(pg_data_t *pgdat, struct zone *zone)
{
set_zonelist_order();
@@ -3474,10 +3481,10 @@ void __ref build_all_zonelists(void *dat
/* we have to stop all cpus to guarantee there is no user
of zonelist */
#ifdef CONFIG_MEMORY_HOTPLUG
- if (data)
- setup_zone_pageset((struct zone *)data);
+ if (zone)
+ setup_zone_pageset(zone);
#endif
- stop_machine(__build_all_zonelists, NULL, NULL);
+ stop_machine(__build_all_zonelists, pgdat, NULL);
/* cpuset refresh routine should be here */
}
vm_total_pages = nr_free_pagecache_pages();
_
Subject: Subject: mm/hotplug: correctly setup fallback zonelists when creating new pgdat
Patches currently in -mm which might be from jiang.liu@xxxxxxxxxx are
linux-next.patch
memory-hotplug-fix-invalid-memory-access-caused-by-stale-kswapd-pointer.patch
memory-hotplug-fix-invalid-memory-access-caused-by-stale-kswapd-pointer-fix.patch
mm-hotplug-correctly-setup-fallback-zonelists-when-creating-new-pgdat.patch
mm-hotplug-correctly-add-new-zone-to-all-other-nodes-zone-lists.patch
mm-hotplug-free-zone-pageset-when-a-zone-becomes-empty.patch
mm-hotplug-mark-memory-hotplug-code-in-page_allocc-as-__meminit.patch
--
To unsubscribe from this list: send the line "unsubscribe mm-commits" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
