[<prev] [next>] [thread-next>] [day] [month] [year] [list]
Message-Id: <20200416025034.29780-1-songmuchun@bytedance.com>
Date: Thu, 16 Apr 2020 10:50:34 +0800
From: Muchun Song <songmuchun@...edance.com>
To: akpm@...ux-foundation.org, Markus.Elfring@....de, david@...hat.com,
ktkhai@...tuozzo.com, yang.shi@...ux.alibaba.com
Cc: linux-mm@...ck.org, linux-kernel@...r.kernel.org,
Muchun Song <songmuchun@...edance.com>,
Xiongchun Duan <duanxiongchun@...edance.com>
Subject: [PATCH v4] mm/ksm: Fix NULL pointer dereference when KSM zero page is enabled
The find_mergeable_vma can return NULL. In this case, it leads
to a crash when we access vm_mm(its offset is 0x40) later in
write_protect_page. And this case did happen on our server. The
following call trace is captured in kernel 4.19 with the following
patch applied and KSM zero page enabled on our server.
commit e86c59b1b12d ("mm/ksm: improve deduplication of zero pages with colouring")
So add a vma check to fix it.
--------------------------------------------------------------------------
BUG: unable to handle kernel NULL pointer dereference at 0000000000000040
Oops: 0000 [#1] SMP NOPTI
CPU: 9 PID: 510 Comm: ksmd Kdump: loaded Tainted: G OE 4.19.36.bsk.9-amd64 #4.19.36.bsk.9
RIP: 0010:try_to_merge_one_page+0xc7/0x760
Code: 24 58 65 48 33 34 25 28 00 00 00 89 e8 0f 85 a3 06 00 00 48 83 c4
60 5b 5d 41 5c 41 5d 41 5e 41 5f c3 48 8b 46 08 a8 01 75 b8 <49>
8b 44 24 40 4c 8d 7c 24 20 b9 07 00 00 00 4c 89 e6 4c 89 ff 48
RSP: 0018:ffffadbdd9fffdb0 EFLAGS: 00010246
RAX: ffffda83ffd4be08 RBX: ffffda83ffd4be40 RCX: 0000002c6e800000
RDX: 0000000000000000 RSI: ffffda83ffd4be40 RDI: 0000000000000000
RBP: ffffa11939f02ec0 R08: 0000000094e1a447 R09: 00000000abe76577
R10: 0000000000000962 R11: 0000000000004e6a R12: 0000000000000000
R13: ffffda83b1e06380 R14: ffffa18f31f072c0 R15: ffffda83ffd4be40
FS: 0000000000000000(0000) GS:ffffa0da43b80000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 0000000000000040 CR3: 0000002c77c0a003 CR4: 00000000007626e0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
PKRU: 55555554
Call Trace:
? follow_page_pte+0x36d/0x5e0
ksm_scan_thread+0x115e/0x1960
? remove_wait_queue+0x60/0x60
kthread+0xf5/0x130
? try_to_merge_with_ksm_page+0x90/0x90
? kthread_create_worker_on_cpu+0x70/0x70
ret_from_fork+0x1f/0x30
--------------------------------------------------------------------------
Fixes: e86c59b1b12d ("mm/ksm: improve deduplication of zero pages with colouring")
Signed-off-by: Muchun Song <songmuchun@...edance.com>
Co-developed-by: Xiongchun Duan <duanxiongchun@...edance.com>
Reviewed-by: David Hildenbrand <david@...hat.com>
Reviewed-by: Kirill Tkhai <ktkhai@...tuozzo.com>
---
Change in v4:
1. Update commit message.
2. If the vma is out of date, just exit.
Change in v3:
1. Update "Signed-off-by" to "Co-developed-by"
2. Update commit message
Change in v2:
1. Update commit message.
2. Update patch subject from:
"mm/ksm: Fix kernel NULL pointer dereference at 0000000000000040"
to:
"mm/ksm: Fix NULL pointer dereference when KSM zero page is enabled"
mm/ksm.c | 11 +++++++++--
1 file changed, 9 insertions(+), 2 deletions(-)
diff --git a/mm/ksm.c b/mm/ksm.c
index a558da9e71770..15339538da299 100644
--- a/mm/ksm.c
+++ b/mm/ksm.c
@@ -2112,8 +2112,15 @@ static void cmp_and_merge_page(struct page *page, struct rmap_item *rmap_item)
down_read(&mm->mmap_sem);
vma = find_mergeable_vma(mm, rmap_item->address);
- err = try_to_merge_one_page(vma, page,
- ZERO_PAGE(rmap_item->address));
+ if (vma)
+ err = try_to_merge_one_page(vma, page,
+ ZERO_PAGE(rmap_item->address));
+ else
+ /**
+ * If the vma is out of date, we do not need to
+ * continue.
+ */
+ err = 0;
up_read(&mm->mmap_sem);
/*
* In case of failure, the page was not really empty, so we
--
2.11.0
Powered by blists - more mailing lists