* Re: [PATCH] swap: fix oops during block io poll in swapin path
[not found] <1501064703-5888-1-git-send-email-penguin-kernel@I-love.SAKURA.ne.jp>
@ 2017-07-26 16:33 ` Shaohua Li
2017-07-26 21:31 ` Andrew Morton
0 siblings, 1 reply; 3+ messages in thread
From: Shaohua Li @ 2017-07-26 16:33 UTC (permalink / raw)
To: Tetsuo Handa, akpm
Cc: Tim Chen, Huang Ying, Jens Axboe, Hugh Dickins, linux-mm
On Wed, Jul 26, 2017 at 07:25:03PM +0900, Tetsuo Handa wrote:
> When a thread is OOM-killed during swap_readpage() operation, an oops
> occurs because end_swap_bio_read() is calling wake_up_process() based on
> an assumption that the thread which called swap_readpage() is still alive.
>
> ----------
> [ 167.408563] Out of memory: Kill process 525 (polkitd) score 0 or sacrifice child
> [ 167.410592] Killed process 525 (polkitd) total-vm:528128kB, anon-rss:0kB, file-rss:4kB, shmem-rss:0kB
> [ 167.415666] oom_reaper: reaped process 525 (polkitd), now anon-rss:0kB, file-rss:0kB, shmem-rss:0kB
> [ 167.460471] general protection fault: 0000 [#1] SMP DEBUG_PAGEALLOC
> [ 167.462303] Modules linked in: nf_conntrack_netbios_ns nf_conntrack_broadcast ip6t_rpfilter ipt_REJECT nf_reject_ipv4 ip6t_REJECT nf_reject_ipv6 xt_conntrack ip_set nfnetlink ebtable_nat ebtable_broute bridge stp llc ip6table_nat nf_conntrack_ipv6 nf_defrag_ipv6 nf_nat_ipv6 ip6table_mangle ip6table_raw iptable_nat nf_conntrack_ipv4 nf_defrag_ipv4 nf_nat_ipv4 nf_nat nf_conntrack iptable_mangle iptable_raw ebtable_filter ebtables ip6table_filter ip6_tables iptable_filter coretemp ppdev pcspkr vmw_balloon sg shpchp vmw_vmci parport_pc parport i2c_piix4 ip_tables xfs libcrc32c sd_mod sr_mod cdrom ata_generic pata_acpi vmwgfx ahci libahci drm_kms_helper ata_piix syscopyarea sysfillrect sysimgblt fb_sys_fops mptspi scsi_transport_spi ttm e1000 mptscsih drm mptbase i2c_core libata serio_raw
> [ 167.476975] CPU: 0 PID: 0 Comm: swapper/0 Not tainted 4.13.0-rc2-next-20170725 #129
> [ 167.479002] Hardware name: VMware, Inc. VMware Virtual Platform/440BX Desktop Reference Platform, BIOS 6.00 07/31/2013
> [ 167.481523] task: ffffffffb7c16500 task.stack: ffffffffb7c00000
> [ 167.483240] RIP: 0010:__lock_acquire+0x151/0x12f0
> [ 167.484808] RSP: 0018:ffffa01f39e03c50 EFLAGS: 00010002
> [ 167.486659] RAX: 6b6b6b6b6b6b6b6b RBX: 0000000000000000 RCX: 0000000000000000
> [ 167.488996] RDX: 0000000000000001 RSI: 0000000000000000 RDI: ffffa01f350d0bb8
> [ 167.491392] RBP: ffffa01f39e03d10 R08: ffffffffb709fefb R09: 0000000000000001
> [ 167.493375] R10: 0000000000000000 R11: ffffffffb7c16500 R12: 0000000000000001
> [ 167.495316] R13: 0000000000000000 R14: 0000000000000000 R15: ffffa01f350d0bb8
> [ 167.497253] FS: 0000000000000000(0000) GS:ffffa01f39e00000(0000) knlGS:0000000000000000
> [ 167.499384] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
> [ 167.501090] CR2: 00007f5ab0e3a9d0 CR3: 000000012dee5000 CR4: 00000000001606f0
> [ 167.503124] Call Trace:
> [ 167.504271] <IRQ>
> [ 167.505331] ? free_debug_processing+0x25d/0x3b0
> [ 167.506807] ? __slab_free+0x9f/0x280
> [ 167.508075] ? __slab_free+0x9f/0x280
> [ 167.509339] lock_acquire+0x59/0x80
> [ 167.510582] ? lock_acquire+0x59/0x80
> [ 167.511872] ? try_to_wake_up+0x3b/0x410
> [ 167.513133] _raw_spin_lock_irqsave+0x3b/0x4f
> [ 167.514449] ? try_to_wake_up+0x3b/0x410
> [ 167.515693] try_to_wake_up+0x3b/0x410
> [ 167.516857] ? mempool_free_slab+0x12/0x20
> [ 167.518068] ? mempool_free+0x26/0x80
> [ 167.519291] wake_up_process+0x10/0x20
> [ 167.520763] end_swap_bio_read+0x6f/0xf0
> [ 167.522229] bio_endio+0x92/0xb0
> [ 167.523324] blk_update_request+0x88/0x270
> [ 167.524642] scsi_end_request+0x32/0x1c0
> [ 167.525864] scsi_io_completion+0x209/0x680
> [ 167.527040] scsi_finish_command+0xd4/0x120
> [ 167.528210] scsi_softirq_done+0x120/0x140
> [ 167.529369] __blk_mq_complete_request_remote+0xe/0x10
> [ 167.530809] flush_smp_call_function_queue+0x51/0x120
> [ 167.532109] generic_smp_call_function_single_interrupt+0xe/0x20
> [ 167.533597] smp_trace_call_function_single_interrupt+0x22/0x30
> [ 167.535049] smp_call_function_single_interrupt+0x9/0x10
> [ 167.536391] call_function_single_interrupt+0xa7/0xb0
> [ 167.537821] </IRQ>
> [ 167.538670] RIP: 0010:native_safe_halt+0x6/0x10
> [ 167.539895] RSP: 0018:ffffffffb7c03df8 EFLAGS: 00000202 ORIG_RAX: ffffffffffffff04
> [ 167.541602] RAX: ffffffffb7c16500 RBX: ffffffffb7c16500 RCX: 0000000000000000
> [ 167.543201] RDX: 0000000000000001 RSI: 0000000000000001 RDI: ffffffffb7c16500
> [ 167.544894] RBP: ffffffffb7c03df8 R08: 0000000000000001 R09: 0000000000000000
> [ 167.546497] R10: 0000000000000000 R11: 0000000000000000 R12: ffffffffb7e4fa20
> [ 167.548023] R13: ffffffffb7c16500 R14: 0000000000000000 R15: 0000000000000000
> [ 167.549561] ? trace_hardirqs_on+0xd/0x10
> [ 167.550593] default_idle+0xe/0x20
> [ 167.551521] arch_cpu_idle+0xa/0x10
> [ 167.552496] default_idle_call+0x1e/0x30
> [ 167.553783] do_idle+0x187/0x200
> [ 167.554875] cpu_startup_entry+0x6e/0x70
> [ 167.556023] rest_init+0xd0/0xe0
> [ 167.556921] start_kernel+0x456/0x477
> [ 167.557875] ? early_idt_handler_array+0x120/0x120
> [ 167.559018] x86_64_start_reservations+0x24/0x26
> [ 167.560104] x86_64_start_kernel+0xf7/0x11a
> [ 167.561131] secondary_startup_64+0xa5/0xa5
> [ 167.562169] Code: c3 49 81 3f 20 9e 0b b8 41 bc 00 00 00 00 44 0f 45 e2 83 fe 01 0f 87 62 ff ff ff 89 f0 49 8b 44 c7 08 48 85 c0 0f 84 52 ff ff ff <f0> ff 80 98 01 00 00 8b 3d 5a 49 c4 01 45 8b b3 18 0c 00 00 85
> [ 167.565895] RIP: __lock_acquire+0x151/0x12f0 RSP: ffffa01f39e03c50
> [ 167.567280] ---[ end trace 6c441db499169b1e ]---
> [ 167.568400] Kernel panic - not syncing: Fatal exception in interrupt
> [ 167.569907] Kernel Offset: 0x36000000 from 0xffffffff81000000 (relocation range: 0xffffffff80000000-0xffffffffbfffffff)
> [ 167.572108] ---[ end Kernel panic - not syncing: Fatal exception in interrupt
> ----------
>
> Fix it by holding a reference to the thread.
Ok, so the task is killed in the page fault retry time check, thanks!
Reviewed-by: Shaohua Li <shli@fb.com>
> Signed-off-by: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
> Fixes: 23955622ff8d231b ("swap: add block io poll in swapin path")
> ---
> mm/page_io.c | 3 +++
> 1 file changed, 3 insertions(+)
>
> diff --git a/mm/page_io.c b/mm/page_io.c
> index b6c4ac38..844c18c 100644
> --- a/mm/page_io.c
> +++ b/mm/page_io.c
> @@ -22,6 +22,7 @@
> #include <linux/frontswap.h>
> #include <linux/blkdev.h>
> #include <linux/uio.h>
> +#include <linux/sched/task.h>
> #include <asm/pgtable.h>
>
> static struct bio *get_swap_bio(gfp_t gfp_flags,
> @@ -136,6 +137,7 @@ static void end_swap_bio_read(struct bio *bio)
> WRITE_ONCE(bio->bi_private, NULL);
> bio_put(bio);
> wake_up_process(waiter);
> + put_task_struct(waiter);
> }
>
> int generic_swapfile_activate(struct swap_info_struct *sis,
> @@ -378,6 +380,7 @@ int swap_readpage(struct page *page, bool do_poll)
> goto out;
> }
> bdev = bio->bi_bdev;
> + get_task_struct(current);
> bio->bi_private = current;
> bio_set_op_attrs(bio, REQ_OP_READ, 0);
> count_vm_event(PSWPIN);
> --
> 1.8.3.1
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: [PATCH] swap: fix oops during block io poll in swapin path
2017-07-26 16:33 ` [PATCH] swap: fix oops during block io poll in swapin path Shaohua Li
@ 2017-07-26 21:31 ` Andrew Morton
2017-07-26 21:48 ` Andrew Morton
0 siblings, 1 reply; 3+ messages in thread
From: Andrew Morton @ 2017-07-26 21:31 UTC (permalink / raw)
To: Shaohua Li
Cc: Tetsuo Handa, Tim Chen, Huang Ying, Jens Axboe, Hugh Dickins, linux-mm
On Wed, 26 Jul 2017 09:33:50 -0700 Shaohua Li <shli@fb.com> wrote:
> On Wed, Jul 26, 2017 at 07:25:03PM +0900, Tetsuo Handa wrote:
> > When a thread is OOM-killed during swap_readpage() operation, an oops
> > occurs because end_swap_bio_read() is calling wake_up_process() based on
> > an assumption that the thread which called swap_readpage() is still alive.
> >
>
> ...
>
> >
> > Fix it by holding a reference to the thread.
>
> Ok, so the task is killed in the page fault retry time check, thanks!
>
> Reviewed-by: Shaohua Li <shli@fb.com>
>
The original patch didn't appear in my inbox and marc.info doesn't
appear to have received it either. Can we please have a resend?
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: [PATCH] swap: fix oops during block io poll in swapin path
2017-07-26 21:31 ` Andrew Morton
@ 2017-07-26 21:48 ` Andrew Morton
0 siblings, 0 replies; 3+ messages in thread
From: Andrew Morton @ 2017-07-26 21:48 UTC (permalink / raw)
To: Shaohua Li, Tetsuo Handa, Tim Chen, Huang Ying, Jens Axboe,
Hugh Dickins, linux-mm
On Wed, 26 Jul 2017 14:31:53 -0700 Andrew Morton <akpm@linux-foundation.org> wrote:
> On Wed, 26 Jul 2017 09:33:50 -0700 Shaohua Li <shli@fb.com> wrote:
>
> > On Wed, Jul 26, 2017 at 07:25:03PM +0900, Tetsuo Handa wrote:
> > > When a thread is OOM-killed during swap_readpage() operation, an oops
> > > occurs because end_swap_bio_read() is calling wake_up_process() based on
> > > an assumption that the thread which called swap_readpage() is still alive.
> > >
> >
> > ...
> >
> > >
> > > Fix it by holding a reference to the thread.
> >
> > Ok, so the task is killed in the page fault retry time check, thanks!
> >
> > Reviewed-by: Shaohua Li <shli@fb.com>
> >
>
> The original patch didn't appear in my inbox and marc.info doesn't
> appear to have received it either. Can we please have a resend?
Jens sent me a copy (thanks).
A get_task_struct in the middle of readpage is very strange-looking.
So a comment is needed. This?
--- a/mm/page_io.c~swap-fix-oops-during-block-io-poll-in-swapin-path-fix
+++ a/mm/page_io.c
@@ -380,6 +380,10 @@ int swap_readpage(struct page *page, boo
goto out;
}
bdev = bio->bi_bdev;
+ /*
+ * Keep this task valid during swap readpage because the oom killer may
+ * attempt to access it in the page fault retry time check.
+ */
get_task_struct(current);
bio->bi_private = current;
bio_set_op_attrs(bio, REQ_OP_READ, 0);
_
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2017-07-26 21:48 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
[not found] <1501064703-5888-1-git-send-email-penguin-kernel@I-love.SAKURA.ne.jp>
2017-07-26 16:33 ` [PATCH] swap: fix oops during block io poll in swapin path Shaohua Li
2017-07-26 21:31 ` Andrew Morton
2017-07-26 21:48 ` Andrew Morton
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).