linux-kernel - Re: mm: race in put_and_wait_on_page

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <fc11de02-9644-1087-9ab6-1537594b924b@lca.pw>
Date:   Mon, 4 Feb 2019 20:37:19 -0500
From:   Qian Cai <cai@....pw>
To:     Hugh Dickins <hughd@...gle.com>, Artem Savkov <asavkov@...hat.com>
Cc:     Baoquan He <bhe@...hat.com>,
        Andrea Arcangeli <aarcange@...hat.com>,
        Michal Hocko <mhocko@...e.com>,
        Vlastimil Babka <vbabka@...e.cz>,
        Linus Torvalds <torvalds@...ux-foundation.org>,
        Andrew Morton <akpm@...ux-foundation.org>,
        linux-kernel@...r.kernel.org, linux-mm@...ck.org
Subject: Re: mm: race in put_and_wait_on_page_locked()



On 2/4/19 3:42 PM, Hugh Dickins wrote:
> On Mon, 4 Feb 2019, Artem Savkov wrote:
> 
>> Hi Hugh,
>>
>> Your recent patch 9a1ea439b16b "mm: put_and_wait_on_page_locked() while
>> page is migrated" seems to have introduced a race into page migration
>> process. I have a host that eagerly reproduces the following BUG under
>> stress:
>>
>> [  302.847402] page:f000000000021700 count:0 mapcount:0 mapping:c0000000b2710bb0 index:0x19
>> [  302.848096] xfs_address_space_operations [xfs] 
>> [  302.848100] name:"libc-2.28.so" 
>> [  302.848244] flags: 0x3ffff800000006(referenced|uptodate)
>> [  302.848521] raw: 003ffff800000006 5deadbeef0000100 5deadbeef0000200 0000000000000000
>> [  302.848724] raw: 0000000000000019 0000000000000000 00000001ffffffff c0000000bc0b1000
>> [  302.848919] page dumped because: VM_BUG_ON_PAGE(page_ref_count(page) == 0)
>> [  302.849076] page->mem_cgroup:c0000000bc0b1000
>> [  302.849269] ------------[ cut here ]------------
>> [  302.849397] kernel BUG at include/linux/mm.h:546!
>> [  302.849586] Oops: Exception in kernel mode, sig: 5 [#1]
>> [  302.849711] LE SMP NR_CPUS=2048 NUMA pSeries
>> [  302.849839] Modules linked in: pseries_rng sunrpc xts vmx_crypto virtio_balloon xfs libcrc32c virtio_net net_failover virtio_console failover virtio_blk
>> [  302.850400] CPU: 3 PID: 8759 Comm: cc1 Not tainted 5.0.0-rc4+ #36
>> [  302.850571] NIP:  c00000000039c8b8 LR: c00000000039c8b4 CTR: c00000000080a0e0
>> [  302.850758] REGS: c0000000b0d7f7e0 TRAP: 0700   Not tainted  (5.0.0-rc4+)
>> [  302.850952] MSR:  8000000000029033 <SF,EE,ME,IR,DR,RI,LE>  CR: 48024422  XER: 00000000
>> [  302.851150] CFAR: c0000000003ff584 IRQMASK: 0 
>> [  302.851150] GPR00: c00000000039c8b4 c0000000b0d7fa70 c000000001bcca00 0000000000000021 
>> [  302.851150] GPR04: c0000000b044c628 0000000000000007 55555555555555a0 c000000001fc3760 
>> [  302.851150] GPR08: 0000000000000007 0000000000000000 c0000000b0d7c000 c0000000b0d7f5ff 
>> [  302.851150] GPR12: 0000000000004400 c00000003fffae80 0000000000000000 0000000000000000 
>> [  302.851150] GPR16: 0000000000000000 0000000000000000 0000000000000000 0000000000000000 
>> [  302.851150] GPR20: c0000000689f5aa8 c00000002a13ee48 0000000000000000 c000000001da29b0 
>> [  302.851150] GPR24: c000000001bf7d80 c0000000689f5a00 0000000000000000 0000000000000000 
>> [  302.851150] GPR28: c000000001bf9e80 c0000000b0d7fab8 0000000000000001 f000000000021700 
>> [  302.852914] NIP [c00000000039c8b8] put_and_wait_on_page_locked+0x398/0x3d0
>> [  302.853080] LR [c00000000039c8b4] put_and_wait_on_page_locked+0x394/0x3d0
>> [  302.853235] Call Trace:
>> [  302.853305] [c0000000b0d7fa70] [c00000000039c8b4] put_and_wait_on_page_locked+0x394/0x3d0 (unreliable)
>> [  302.853540] [c0000000b0d7fb10] [c00000000047b838] __migration_entry_wait+0x178/0x250
>> [  302.853738] [c0000000b0d7fb50] [c00000000040c928] do_swap_page+0xd78/0xf60
>> [  302.853997] [c0000000b0d7fbd0] [c000000000411078] __handle_mm_fault+0xbf8/0xe80
>> [  302.854187] [c0000000b0d7fcb0] [c000000000411548] handle_mm_fault+0x248/0x450
>> [  302.854379] [c0000000b0d7fd00] [c000000000078ca4] __do_page_fault+0x2d4/0xdf0
>> [  302.854877] [c0000000b0d7fde0] [c0000000000797f8] do_page_fault+0x38/0xf0
>> [  302.855057] [c0000000b0d7fe20] [c00000000000a7c4] handle_page_fault+0x18/0x38
>> [  302.855300] Instruction dump:
>> [  302.855432] 4bfffcf0 60000000 3948ffff 4bfffd20 60000000 60000000 3c82ff36 7fe3fb78 
>> [  302.855689] fb210068 38843b78 48062f09 60000000 <0fe00000> 60000000 3b400001 3b600001 
>> [  302.855950] ---[ end trace a52140e0f9751ae0 ]---
>>
>> What seems to be happening is migrate_page_move_mapping() calling
>> page_ref_freeze() on another cpu somewhere between __migration_entry_wait()
>> taking a reference and wait_on_page_bit_common() calling page_put().
> 
> Thank you for reporting, Artem.
> 
> And see the mm thread https://marc.info/?l=linux-mm&m=154821775401218&w=2
> 
> That was on arm64, you are on power I think: both point towards xfs
> (Cai could not reproduce it on ext4), but that should not be taken too
> seriously - it could just be easier to reproduce on one than the other.

Agree, although I have never been able to trigger it for ext4 running LTP
migrate_pages03 exclusively overnight (500+ iterations) and spontaneously for a
few weeks now. It might just be lucky.