Message-ID: <55D0945A.2040505@internode.on.net>
Date:	Sun, 16 Aug 2015 23:17:06 +0930
From:	Arthur Marsh <arthur.marsh@...ernode.on.net>
To:	Oleg Nesterov <oleg@...hat.com>, Al Viro <viro@...iv.linux.org.uk>
CC:	Dave Chinner <david@...morbit.com>,
	Dave Hansen <dave.hansen@...ux.intel.com>,
	Jan Kara <jack@...e.cz>,
	"Paul E. McKenney" <paulmck@...ux.vnet.ibm.com>,
	Peter Zijlstra <peterz@...radead.org>,
	linux-fsdevel@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH v3 0/8] change sb_writers to use percpu_rw_semaphore

Oleg Nesterov wrote on 15/08/15 02:49:
> On 08/13, Jan Kara wrote:
>>
>> Regarding the routing, ideally Al Viro should take these as a VFS
>> maintainer.
>
> Al, could you take these patches?
>
> Only cosmetic changes in V3 to address the comments from Jan, I
> preserved his acks.
>
> In case you missed all the spam I sent before, let me repeat that
> the awful (and currently unneeded) 7/8 will be reverted later. We
> need it to ensure that other percpu_rw_semaphore changes routed
> via another tree won't break fs/super.c. After that we will add
> rcu_sync_dtor(s_writers->rw_sem) into deactivate_locked_super()
> and revert this horror.
>
> 3/8 documents the lockdep problems we currently have. This is fixed
> by the patch below but it depends on xfs ILOCK fixes from Dave, so
> I will send it later. Plus another patch which removes the "trylock"
> hack in __sb_start_write().
>
> Oleg.

Would these patches address what I've seen in the last day or so using 
Linus' git head kernel and seeing problems like:

[    0.000000] Initializing cgroup subsys cpuset
[    0.000000] Initializing cgroup subsys cpu
[    0.000000] Initializing cgroup subsys cpuacct
[    0.000000] Linux version 4.2.0-rc6+ (root@...toria) (gcc version 
5.2.1 20150808 (Debian 5.2.1-15) ) #11 SMP PREEMPT Sun Aug 16 07:27:00 
ACST 2015
...
[ 6000.096107] INFO: task basename:7796 blocked for more than 120 seconds.
[ 6000.096116]       Not tainted 4.2.0-rc6+ #11
[ 6000.096120] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" 
disables this message.
[ 6000.096123] basename        D e7b5b180     0  7796   6936 0x00000000
[ 6000.096132]  c0379a84 00000086 c11127a5 e7b5b180 e7b5b5ec 2e0a5fb9 
00000557 f5f0b310
[ 6000.096143]  f330b180 e7b5b180 c037a000 f5f0b300 7fffffff c0379a90 
c155b740 00000000
[ 6000.096154]  c0379b04 c155fa1d 00000046 c11127a5 00000246 00000000 
c0379ab0 c10a569b
[ 6000.096164] Call Trace:
[ 6000.096174]  [<c11127a5>] ? __delayacct_blkio_start+0x15/0x20
[ 6000.096179]  [<c155b740>] schedule+0x30/0x80
[ 6000.096184]  [<c155fa1d>] schedule_timeout+0x2cd/0x5c0
[ 6000.096188]  [<c11127a5>] ? __delayacct_blkio_start+0x15/0x20
[ 6000.096193]  [<c10a569b>] ? trace_hardirqs_on+0xb/0x10
[ 6000.096198]  [<c10d701c>] ? ktime_get+0xac/0x1a0
[ 6000.096202]  [<c11127a5>] ? __delayacct_blkio_start+0x15/0x20
[ 6000.096206]  [<c155ad99>] io_schedule_timeout+0x89/0xf0
[ 6000.096209]  [<c155beb0>] ? bit_wait+0x40/0x40
[ 6000.096213]  [<c155bed5>] bit_wait_io+0x25/0x50
[ 6000.096216]  [<c155bbc9>] __wait_on_bit+0x49/0x70
[ 6000.096219]  [<c155beb0>] ? bit_wait+0x40/0x40
[ 6000.096223]  [<c155bc4d>] out_of_line_wait_on_bit+0x5d/0x70
[ 6000.096226]  [<c155beb0>] ? bit_wait+0x40/0x40
[ 6000.096230]  [<c109b210>] ? autoremove_wake_function+0x40/0x40
[ 6000.096236]  [<c11ed5be>] bh_submit_read+0x7e/0x90
[ 6000.096265]  [<f8321354>] ext4_get_branch+0xa4/0x110 [ext4]
[ 6000.096286]  [<f8321f14>] ext4_ind_map_blocks+0xd4/0xe30 [ext4]
[ 6000.096291]  [<c10a6290>] ? __lock_acquire+0x910/0x16a0
[ 6000.096295]  [<c10a6290>] ? __lock_acquire+0x910/0x16a0
[ 6000.096300]  [<c155ed53>] ? down_read+0x33/0x50
[ 6000.096315]  [<f82ddc9d>] ext4_map_blocks+0x29d/0x4f0 [ext4]
[ 6000.096319]  [<c10a548b>] ? mark_held_locks+0x5b/0x90
[ 6000.096323]  [<c10a55ec>] ? trace_hardirqs_on_caller+0x12c/0x1d0
[ 6000.096337]  [<f82db052>] ? ext4_readpages+0x32/0x40 [ext4]
[ 6000.096358]  [<f832d37b>] ext4_mpage_readpages+0x30b/0x8c0 [ext4]
[ 6000.096372]  [<f82db052>] ? ext4_readpages+0x32/0x40 [ext4]
[ 6000.096377]  [<c1156930>] ? __alloc_pages_nodemask+0x9c0/0xa40
[ 6000.096383]  [<c107ed46>] ? preempt_count_sub+0x26/0x70
[ 6000.096397]  [<f82db052>] ext4_readpages+0x32/0x40 [ext4]
[ 6000.096411]  [<f82db020>] ? do_journal_get_write_access+0xb0/0xb0 [ext4]
[ 6000.096416]  [<c115c376>] __do_page_cache_readahead+0x2e6/0x370
[ 6000.096420]  [<c115c233>] ? __do_page_cache_readahead+0x1a3/0x370
[ 6000.096426]  [<c114fb85>] filemap_fault+0x505/0x570
[ 6000.096430]  [<c117bb6f>] ? __do_fault+0x2f/0x80
[ 6000.096435]  [<c117bb6f>] __do_fault+0x2f/0x80
[ 6000.096439]  [<c1560ca7>] ? _raw_spin_unlock+0x27/0x50
[ 6000.096443]  [<c117f412>] handle_mm_fault+0xb22/0x11d0
[ 6000.096448]  [<c104aa7a>] __do_page_fault+0x16a/0x500
[ 6000.096452]  [<c104ae10>] ? __do_page_fault+0x500/0x500
[ 6000.096456]  [<c104ae31>] do_page_fault+0x21/0x30
[ 6000.096460]  [<c156282b>] error_code+0x5f/0x64
[ 6000.096464]  [<c104ae10>] ? __do_page_fault+0x500/0x500
[ 6000.096468] 2 locks held by basename/7796:
[ 6000.096470]  #0:  (&mm->mmap_sem){++++++}, at: [<c104aa25>] 
__do_page_fault+0x115/0x500
[ 6000.096479]  #1:  (&ei->i_data_sem){++++..}, at: [<f82ddd9b>] 
ext4_map_blocks+0x39b/0x4f0 [ext4]
[ 6000.096500] INFO: task hddtemp:7797 blocked for more than 120 seconds.
[ 6000.096503]       Not tainted 4.2.0-rc6+ #11
[ 6000.096505] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" 
disables this message.
[ 6000.096508] hddtemp         D e896d100     0  7797   5140 0x00000000
[ 6000.096514]  c02c3a84 00000086 e896d588 e896d100 e896d56c 00000001 
c02c3a84 f5f0b310
[ 6000.096525]  c176fb00 e896d100 c02c4000 f5f0b300 7fffffff c02c3a90 
c155b740 00000000
[ 6000.096535]  c02c3b04 c155fa1d 00000046 c11127a5 00000246 00000000 
c02c3ab0 c10a569b
[ 6000.096546] Call Trace:
[ 6000.096550]  [<c155b740>] schedule+0x30/0x80
[ 6000.096554]  [<c155fa1d>] schedule_timeout+0x2cd/0x5c0
[ 6000.096558]  [<c11127a5>] ? __delayacct_blkio_start+0x15/0x20
[ 6000.096562]  [<c10a569b>] ? trace_hardirqs_on+0xb/0x10
[ 6000.096566]  [<c10d701c>] ? ktime_get+0xac/0x1a0
[ 6000.096569]  [<c11127a5>] ? __delayacct_blkio_start+0x15/0x20
[ 6000.096574]  [<c155ad99>] io_schedule_timeout+0x89/0xf0
[ 6000.096577]  [<c109ad07>] ? prepare_to_wait_exclusive+0x47/0x80
[ 6000.096581]  [<c155beb0>] ? bit_wait+0x40/0x40
[ 6000.096584]  [<c155bed5>] bit_wait_io+0x25/0x50
[ 6000.096587]  [<c155bd12>] __wait_on_bit_lock+0x32/0x80
[ 6000.096591]  [<c155bdbd>] out_of_line_wait_on_bit_lock+0x5d/0x70
[ 6000.096595]  [<c155beb0>] ? bit_wait+0x40/0x40
[ 6000.096598]  [<c109b210>] ? autoremove_wake_function+0x40/0x40
[ 6000.096602]  [<c11ea166>] bh_uptodate_or_lock+0x66/0x70
[ 6000.096623]  [<f8321349>] ext4_get_branch+0x99/0x110 [ext4]
[ 6000.096643]  [<f8321f14>] ext4_ind_map_blocks+0xd4/0xe30 [ext4]
[ 6000.096647]  [<c10a6290>] ? __lock_acquire+0x910/0x16a0
[ 6000.096651]  [<c10a6290>] ? __lock_acquire+0x910/0x16a0
[ 6000.096656]  [<c155ed53>] ? down_read+0x33/0x50
[ 6000.096671]  [<f82ddc9d>] ext4_map_blocks+0x29d/0x4f0 [ext4]
[ 6000.096675]  [<c10a548b>] ? mark_held_locks+0x5b/0x90
[ 6000.096679]  [<c10a55ec>] ? trace_hardirqs_on_caller+0x12c/0x1d0
[ 6000.096693]  [<f82db052>] ? ext4_readpages+0x32/0x40 [ext4]
[ 6000.096713]  [<f832d37b>] ext4_mpage_readpages+0x30b/0x8c0 [ext4]
[ 6000.096727]  [<f82db052>] ? ext4_readpages+0x32/0x40 [ext4]
[ 6000.096732]  [<c1156930>] ? __alloc_pages_nodemask+0x9c0/0xa40
[ 6000.096747]  [<f82db052>] ext4_readpages+0x32/0x40 [ext4]
[ 6000.096761]  [<f82db020>] ? do_journal_get_write_access+0xb0/0xb0 [ext4]
[ 6000.096766]  [<c115c376>] __do_page_cache_readahead+0x2e6/0x370
[ 6000.096770]  [<c115c233>] ? __do_page_cache_readahead+0x1a3/0x370
[ 6000.096775]  [<c114fb85>] filemap_fault+0x505/0x570
[ 6000.096779]  [<c117bb6f>] ? __do_fault+0x2f/0x80
[ 6000.096783]  [<c117bb6f>] __do_fault+0x2f/0x80
[ 6000.096787]  [<c1560ca7>] ? _raw_spin_unlock+0x27/0x50
[ 6000.096791]  [<c117f412>] handle_mm_fault+0xb22/0x11d0
[ 6000.096796]  [<c104aa7a>] __do_page_fault+0x16a/0x500
[ 6000.096800]  [<c104ae10>] ? __do_page_fault+0x500/0x500
[ 6000.096803]  [<c104ae31>] do_page_fault+0x21/0x30
[ 6000.096807]  [<c156282b>] error_code+0x5f/0x64
[ 6000.096811]  [<c104ae10>] ? __do_page_fault+0x500/0x500
[ 6000.096815] 2 locks held by hddtemp/7797:
[ 6000.096817]  #0:  (&mm->mmap_sem){++++++}, at: [<c104aa25>] 
__do_page_fault+0x115/0x500
[ 6000.096825]  #1:  (&ei->i_data_sem){++++..}, at: [<f82ddd9b>] 
ext4_map_blocks+0x39b/0x4f0 [ext4]


--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

Powered by blists - more mailing lists