From e9959f0f37160e1f5351af828cc981712b5066c1 Mon Sep 17 00:00:00 2001 From: KAMEZAWA Hiroyuki Date: Wed, 24 Nov 2010 12:57:09 -0800 Subject: mm/page_alloc.c: fix build_all_zonelist() where percpu_alloc() is wrongly called under stop_machine_run() During memory hotplug, build_allzonelists() may be called under stop_machine_run(). In this function, setup_zone_pageset() is called. But it's bug because it will do page allocation under stop_machine_run(). Here is a report from Alok Kataria. BUG: sleeping function called from invalid context at kernel/mutex.c:94 in_atomic(): 0, irqs_disabled(): 1, pid: 4, name: migration/0 Pid: 4, comm: migration/0 Not tainted 2.6.35.6-45.fc14.x86_64 #1 Call Trace: [] __might_sleep+0xeb/0xf0 [] mutex_lock+0x24/0x50 [] pcpu_alloc+0x6d/0x7ee [] ? load_balance+0xbe/0x60e [] ? rt_se_boosted+0x21/0x2f [] ? dequeue_rt_stack+0x18b/0x1ed [] __alloc_percpu+0x10/0x12 [] setup_zone_pageset+0x38/0xbe [] ? build_zonelists_node.clone.58+0x79/0x8c [] __build_all_zonelists+0x419/0x46c [] ? cpu_stopper_thread+0xb2/0x198 [] stop_machine_cpu_stop+0x8e/0xc5 [] ? stop_machine_cpu_stop+0x0/0xc5 [] cpu_stopper_thread+0x108/0x198 [] ? schedule+0x5b2/0x5cc [] ? cpu_stopper_thread+0x0/0x198 [] kthread+0x7f/0x87 [] kernel_thread_helper+0x4/0x10 [] ? kthread+0x0/0x87 [] ? kernel_thread_helper+0x0/0x10 Built 5 zonelists in Node order, mobility grouping on. Total pages: 289456 Policy zone: Normal This patch tries to fix the issue by moving setup_zone_pageset() out from stop_machine_run(). It's obviously not necessary to be called under stop_machine_run(). [akpm@linux-foundation.org: remove unneeded local] Reported-by: Alok Kataria Signed-off-by: KAMEZAWA Hiroyuki Cc: Tejun Heo Cc: Petr Vandrovec Cc: Pekka Enberg Reviewed-by: Christoph Lameter Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- mm/page_alloc.c | 14 +++++--------- 1 file changed, 5 insertions(+), 9 deletions(-) diff --git a/mm/page_alloc.c b/mm/page_alloc.c index 07a654486f75..e4092704c1a9 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -3008,14 +3008,6 @@ static __init_refok int __build_all_zonelists(void *data) build_zonelist_cache(pgdat); } -#ifdef CONFIG_MEMORY_HOTPLUG - /* Setup real pagesets for the new zone */ - if (data) { - struct zone *zone = data; - setup_zone_pageset(zone); - } -#endif - /* * Initialize the boot_pagesets that are going to be used * for bootstrapping processors. The real pagesets for @@ -3064,7 +3056,11 @@ void build_all_zonelists(void *data) } else { /* we have to stop all cpus to guarantee there is no user of zonelist */ - stop_machine(__build_all_zonelists, data, NULL); +#ifdef CONFIG_MEMORY_HOTPLUG + if (data) + setup_zone_pageset((struct zone *)data); +#endif + stop_machine(__build_all_zonelists, NULL, NULL); /* cpuset refresh routine should be here */ } vm_total_pages = nr_free_pagecache_pages(); -- cgit v1.2.3