lkml.org 
[lkml]   [2021]   [Mar]   [30]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
From
SubjectRe: [Question] Is there a race window between swapoff vs synchronous swap_readpage
Date
Yu Zhao <yuzhao@google.com> writes:

> On Mon, Mar 29, 2021 at 9:44 PM Huang, Ying <ying.huang@intel.com> wrote:
>>
>> Miaohe Lin <linmiaohe@huawei.com> writes:
>>
>> > On 2021/3/30 9:57, Huang, Ying wrote:
>> >> Hi, Miaohe,
>> >>
>> >> Miaohe Lin <linmiaohe@huawei.com> writes:
>> >>
>> >>> Hi all,
>> >>> I am investigating the swap code, and I found the below possible race window:
>> >>>
>> >>> CPU 1 CPU 2
>> >>> ----- -----
>> >>> do_swap_page
>> >>> skip swapcache case (synchronous swap_readpage)
>> >>> alloc_page_vma
>> >>> swapoff
>> >>> release swap_file, bdev, or ...
>> >>> swap_readpage
>> >>> check sis->flags is ok
>> >>> access swap_file, bdev or ...[oops!]
>> >>> si->flags = 0
>> >>>
>> >>> The swapcache case is ok because swapoff will wait on the page_lock of swapcache page.
>> >>> Is this will really happen or Am I miss something ?
>> >>> Any reply would be really grateful. Thanks! :)
>> >>
>> >> This appears possible. Even for swapcache case, we can't guarantee the
>> >
>> > Many thanks for reply!
>> >
>> >> swap entry gotten from the page table is always valid too. The
>> >
>> > The page table may change at any time. And we may thus do some useless work.
>> > But the pte_same() check could handle these races correctly if these do not
>> > result in oops.
>> >
>> >> underlying swap device can be swapped off at the same time. So we use
>> >> get/put_swap_device() for that. Maybe we need similar stuff here.
>> >
>> > Using get/put_swap_device() to guard against swapoff for swap_readpage() sounds
>> > really bad as swap_readpage() may take really long time. Also such race may not be
>> > really hurtful because swapoff is usually done when system shutdown only.
>> > I can not figure some simple and stable stuff out to fix this. Any suggestions or
>> > could anyone help get rid of such race?
>>
>> Some reference counting on the swap device can prevent swap device from
>> swapping-off. To reduce the performance overhead on the hot-path as
>> much as possible, it appears we can use the percpu_ref.
>
> Hi,
>
> I've been seeing crashes when testing the latest kernels with
> stress-ng --class vm -a 20 -t 600s --temp-path /tmp
>
> I haven't had time to look into them yet:
>
> DEBUG_VM:
> BUG: unable to handle page fault for address: ffff905c33c9a000
> Call Trace:
> get_swap_pages+0x278/0x590
> get_swap_page+0x1ab/0x280
> add_to_swap+0x7d/0x130
> shrink_page_list+0xf84/0x25f0
> reclaim_pages+0x313/0x430
> madvise_cold_or_pageout_pte_range+0x95c/0xaa0

If my understanding were correct, two bugs are reported? One above and
one below? If so, and the above one is reported firstly. Can you share
the full bug message reported in dmesg?

Can you convert the call trace to source line? And the commit of the
kernel? Or the full kconfig? So I can build it by myself.

Best Regards,
Huang, Ying

> KASAN:
> ==================================================================
> BUG: KASAN: slab-out-of-bounds in __frontswap_store+0xc9/0x2e0
> Read of size 8 at addr ffff88901f646f18 by task stress-ng-mrema/31329
> CPU: 2 PID: 31329 Comm: stress-ng-mrema Tainted: G S I L
> 5.12.0-smp-DEV #2
> Call Trace:
> dump_stack+0xff/0x165
> print_address_description+0x81/0x390
> __kasan_report+0x154/0x1b0
> ? __frontswap_store+0xc9/0x2e0
> ? __frontswap_store+0xc9/0x2e0
> kasan_report+0x47/0x60
> kasan_check_range+0x2f3/0x340
> __kasan_check_read+0x11/0x20
> __frontswap_store+0xc9/0x2e0
> swap_writepage+0x52/0x80
> pageout+0x489/0x7f0
> shrink_page_list+0x1b11/0x2c90
> reclaim_pages+0x6ca/0x930
> madvise_cold_or_pageout_pte_range+0x1260/0x13a0
>
> Allocated by task 16813:
> ____kasan_kmalloc+0xb0/0xe0
> __kasan_kmalloc+0x9/0x10
> __kmalloc_node+0x52/0x70
> kvmalloc_node+0x50/0x90
> __se_sys_swapon+0x353a/0x4860
> __x64_sys_swapon+0x5b/0x70
>
> The buggy address belongs to the object at ffff88901f640000
> which belongs to the cache kmalloc-32k of size 32768
> The buggy address is located 28440 bytes inside of
> 32768-byte region [ffff88901f640000, ffff88901f648000)
> The buggy address belongs to the page:
> page:0000000032d23e33 refcount:1 mapcount:0 mapping:0000000000000000
> index:0x0 pfn:0x101f640
> head:0000000032d23e33 order:4 compound_mapcount:0 compound_pincount:0
> flags: 0x400000000010200(slab|head)
> raw: 0400000000010200 ffffea00062b8408 ffffea000a6e9008 ffff888100040300
> raw: 0000000000000000 ffff88901f640000 0000000100000001 000000000000000
> page dumped because: kasan: bad access detected
>
> Memory state around the buggy address:
> ffff88901f646e00: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
> ffff88901f646e80: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
> >ffff88901f646f00: 00 00 00 fc fc fc fc fc fc fc fc fc fc fc fc fc
> ^
> ffff88901f646f80: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc
> ffff88901f647000: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc
> ==================================================================
>
> Relevant config options I could think of:
>
> CONFIG_MEMCG_SWAP=y
> CONFIG_THP_SWAP=y
> CONFIG_ZSWAP=y

\
 
 \ /
  Last update: 2021-03-30 08:58    [W:0.057 / U:0.308 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site