On 06/02/2016 01:59 PM, Michal Hocko wrote:
[CCing Ebru]
On Thu 02-06-16 13:50:46, Michal Hocko wrote:
[CCing Andrea and Kirill]
Hmm, thinking about it little bit more it might be related to "mm, thp:
make swapin readahead under down_read of mmap_sem". I didn't get to look
closer at the patch but maybe revalidate after mmap sem is dropped is
not sufficient.
I've noticed locking imbalance in the patch as pointed out in the
Sergey's thread about deadlock [1]
I imagine it's possible that the wrong mmap_sem handling can also
violate critical section protections and result in this report?
[1] http://marc.info/?i=0c47a3a0-5530-b257-1c1f-28ed44ba97e6%40suse.cz%3E
On Thu 02-06-16 20:14:38, Tetsuo Handa wrote:
FYI, I hit this bug while compiling kernel. Is this known issue?
----------------------------------------
[ 2893.482222] vma ffff880014150428 start 00002afed7db3000 end 00002afed89bc000
next ffff8800106b7de8 prev ffff880014150a58 mm ffff88007a9e8d40
prot 8000000000000025 anon_vma ffff880016c19d18 vm_ops (null)
pgoff 2afed7db3 file (null) private_data (null)
flags: 0x100073(read|write|mayread|maywrite|mayexec|account)
[ 2893.490801] ------------[ cut here ]------------
[ 2893.492087] kernel BUG at mm/rmap.c:1253!
[ 2893.493240] invalid opcode: 0000 [#1] PREEMPT SMP DEBUG_PAGEALLOC
[ 2893.494789] Modules linked in: ip6t_rpfilter ip6t_REJECT nf_reject_ipv6 nf_conntrack_ipv6 nf_defrag_ipv6 ipt_REJECT nf_reject_ipv4 nf_conntrack_ipv4 nf_defrag_ipv4 xt_conntrack nf_conntrack ebtable_nat ebtable_broute bridge stp llc ebtable_filter ebtables ip6table_mangle ip6table_raw ip6table_filter ip6_tables iptable_mangle iptable_raw iptable_filter coretemp pcspkr sg vmw_vmci i2c_piix4 ip_tables sd_mod ata_generic pata_acpi serio_raw ata_piix vmwgfx mptspi ahci drm_kms_helper syscopyarea libahci scsi_transport_spi sysfillrect sysimgblt mptscsih fb_sys_fops libata e1000 ttm mptbase drm i2c_core
[ 2893.509843] CPU: 0 PID: 50 Comm: khugepaged Not tainted 4.7.0-rc1-next-20160602 #431
[ 2893.512105] Hardware name: VMware, Inc. VMware Virtual Platform/440BX Desktop Reference Platform, BIOS 6.00 07/31/2013
[ 2893.515024] task: ffff88007c036440 ti: ffff88007c040000 task.ti: ffff88007c040000
[ 2893.517340] RIP: 0010:[<ffffffff811ca46c>] [<ffffffff811ca46c>] page_add_new_anon_rmap+0x13c/0x180
[ 2893.519977] RSP: 0018:ffff88007c043ce8 EFLAGS: 00010246
[ 2893.521917] RAX: 0000000000000149 RBX: ffffea00001b0000 RCX: 0000000000000000
[ 2893.524218] RDX: 0000000000000000 RSI: ffffffff819e92ea RDI: 00000000ffffffff
[ 2893.526451] RBP: ffff88007c043d08 R08: 0000000000000001 R09: 0000000000000001
[ 2893.528822] R10: 0000000000000001 R11: 000000000000058e R12: ffff880014150428
[ 2893.531306] R13: 00002afed8a00000 R14: 0000000000000200 R15: ffff880014150428
[ 2893.533699] FS: 0000000000000000(0000) GS:ffff88007f800000(0000) knlGS:0000000000000000
[ 2893.536250] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[ 2893.538309] CR2: 00002acb71513220 CR3: 0000000001c06000 CR4: 00000000001406f0
[ 2893.540777] Stack:
[ 2893.542230] 8000000006c000e7 ffffea0001e55000 0000000000000001 ffffea00001b0000
[ 2893.544867] ffff88007c043e60 ffffffff811fb292 0000000000000000 0000000000000000
[ 2893.547490] ffff88007c036440 ffff88007c036440 ffff88007c036440 ffff88007c036440
[ 2893.550048] Call Trace:
[ 2893.551498] [<ffffffff811fb292>] khugepaged+0x1552/0x25c0
[ 2893.553487] [<ffffffff810bfda0>] ? prepare_to_wait_event+0xf0/0xf0
[ 2893.555705] [<ffffffff811f9d40>] ? vmf_insert_pfn_pmd+0x170/0x170
[ 2893.557981] [<ffffffff81093b0e>] kthread+0xee/0x110
[ 2893.560040] [<ffffffff8172a17f>] ret_from_fork+0x1f/0x40
[ 2893.562302] [<ffffffff81093a20>] ? kthread_create_on_node+0x220/0x220
[ 2893.564690] Code: e8 2a e9 ff ff 5b 41 5c 41 5d 41 5e 5d c3 48 8b 43 20 a8 01 0f 85 37 ff ff ff c7 43 18 00 00 00 00 eb 9b 4c 89 e7 e8 f4 d1 fe ff <0f> 0b 48 83 e8 01 e9 07 ff ff ff 48 c7 c6 40 6b 9b 81 48 89 df
[ 2893.572649] RIP [<ffffffff811ca46c>] page_add_new_anon_rmap+0x13c/0x180
[ 2893.574980] RSP <ffff88007c043ce8>
[ 2893.583817] ---[ end trace 994b25e4ac8d495c ]---
[ 2893.585665] note: khugepaged[50] exited with preempt_count 1
----------------------------------------
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@xxxxxxxxx. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>
--
Michal Hocko
SUSE Labs
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@xxxxxxxxx. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>