blob: d80b46456e585e283c63d190b01e197b1106cb62 (
plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
|
From 6929d59ae0314653a70193d22cad68cb911ed84a Mon Sep 17 00:00:00 2001
From: Carsten Emde <c.emde@osadl.org>
Date: Tue, 21 Jul 2009 18:58:20 +0200
Subject: [PATCH] mm: memcontrol preemption fix
commit e580f50631b1d82949f8a1c72f0e953cef2df07c in tip.
Cures the following On RT:
BUG: using smp_processor_id() in preemptible [00000000] code: init/1
caller is mem_cgroup_update_mapped_file_stat+0x3c/0x68
Pid: 1, comm: init Not tainted 2.6.31-rc3-rt1 #1
Call Trace:
[<c0759ec5>] ? printk+0x14/0x17
[<c059c932>] debug_smp_processor_id+0xa6/0xbc
[<c04c137a>] mem_cgroup_update_mapped_file_stat+0x3c/0x68
[<c04b43cd>] page_add_file_rmap+0x2e/0x31
[<c04ae3aa>] __do_fault+0x2be/0x3c3
[<c04ae83b>] handle_mm_fault+0x38c/0x874
[<c0488643>] ? ftrace_now+0x36/0x42
[<c075e4a6>] do_page_fault+0x2b2/0x2c8
[<c075e1f4>] ? do_page_fault+0x0/0x2c8
[<c075c923>] error_code+0x6b/0x70
[<c075e1f4>] ? do_page_fault+0x0/0x2c8
RT needs to disable preemption explicitely here.
[PG: change get_cpu() -> preempt_disable() due to upstream c62b1a3b31]
Signed-off-by: Carsten Emde <carsten.emde@osadl.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
---
mm/memcontrol.c | 5 ++++-
1 files changed, 4 insertions(+), 1 deletions(-)
diff --git a/mm/memcontrol.c b/mm/memcontrol.c
index 8a79a6f..ddcedda 100644
--- a/mm/memcontrol.c
+++ b/mm/memcontrol.c
@@ -1363,8 +1363,10 @@ void mem_cgroup_update_file_mapped(struct page *page, int val)
goto done;
/*
- * Preemption is already disabled. We can use __this_cpu_xxx
+ * Preemption is already disabled. We can use __this_cpu_xxx,
+ * but that's not true for RT !
*/
+ preempt_disable();
if (val > 0) {
__this_cpu_inc(mem->stat->count[MEM_CGROUP_STAT_FILE_MAPPED]);
SetPageCgroupFileMapped(pc);
@@ -1372,6 +1374,7 @@ void mem_cgroup_update_file_mapped(struct page *page, int val)
__this_cpu_dec(mem->stat->count[MEM_CGROUP_STAT_FILE_MAPPED]);
ClearPageCgroupFileMapped(pc);
}
+ preempt_enable();
done:
unlock_page_cgroup(pc);
--
1.7.0.4
|
