Date:	Mon, 23 Jun 2008 21:54:52 +1000
From:	Nick Piggin <nickpiggin@...oo.com.au>
To:	"Ryan Hope" <rmh3093@...il.com>,
	"Paul E. McKenney" <paulmck@...ux.vnet.ibm.com>
Cc:	"Peter Zijlstra" <peterz@...radead.org>, linux-mm@...r.kernel.org,
	LKML <linux-kernel@...r.kernel.org>
Subject: Re: [BUG] Lockless patches cause hardlock under heavy IO

On Monday 23 June 2008 13:51, Ryan Hope wrote:
> well i get the hardlock on -mm with out using reiser4, i am pretty
> sure is swap related

The guys seeing hangs don't use PREEMPT_RCU, do they?

In my swapping tests, I found -mm3 to be stable with classic RCU, but
on a hunch, I tried PREEMPT_RCU and it crashed a couple of times rather
quickly. First crash was in find_get_pages so I suspected lockless
pagecache doing something subtly wrong with the RCU API, but I just got
another crash in __d_lookup:

BUG: unable to handle kernel paging request at ffff81004a139f38
IP: [<ffffffff802bb82c>] __d_lookup+0x8c/0x160
PGD 8063 PUD 7fc3f163 PMD 7df50163 PTE 800000004a139160
Oops: 0000 [1] PREEMPT SMP DEBUG_PAGEALLOC
last sysfs file: /sys/devices/system/cpu/cpu7/cache/index2/shared_cpu_map
CPU 0
Modules linked in: brd
Pid: 29563, comm: cc1 Not tainted 2.6.26-rc5-mm3 #467
RIP: 0010:[<ffffffff802bb82c>]  [<ffffffff802bb82c>] __d_lookup+0x8c/0x160
RSP: 0018:ffff81004bf7dba8  EFLAGS: 00010282
RAX: 0000000000000007 RBX: ffff81004a139f38 RCX: 0000000000000000
RDX: ffff810028057808 RSI: 0000000000000000 RDI: ffff81004bf7a880
RBP: ffff81004bf7dbf8 R08: 0000000000000001 R09: 0000000000000001
R10: 0000000000000000 R11: 0000000000000001 R12: ffff81004a139ef8
R13: 0000000073885cf7 R14: ffff810070f53ef8 R15: ffff81004bf7dca8
FS:  00002abe0a1decf0(0000) GS:ffffffff80779dc0(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: ffff81004a139f38 CR3: 0000000057569000 CR4: 00000000000006e0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
Process cc1 (pid: 29563, threadinfo ffff81004bf7c000, task ffff81004bf7a880)
Stack:  0000000100000001 0000000000000007 ffff810070f53f00 00000007000041ed
 ffff810001ce2013 ffff81004bf7dca8 00000000000041ed ffff81004bf7de48
 ffff81004bf7dca8 ffff81004bf7dcb8 ffff81004bf7dc48 ffffffff802af2b5
Call Trace:
 [<ffffffff802af2b5>] do_lookup+0x35/0x230
 [<ffffffff80312d60>] ? ext3_permission+0x10/0x20
 [<ffffffff802b0cbb>] __link_path_walk+0x39b/0x10a0
 [<ffffffff802b1a26>] path_walk+0x66/0xd0
 [<ffffffff802b1cde>] do_path_lookup+0x9e/0x240
 [<ffffffff802b21d7>] __path_lookup_intent_open+0x67/0xd0
 [<ffffffff802b224c>] path_lookup_open+0xc/0x10
 [<ffffffff802b31ba>] do_filp_open+0xaa/0x9f0
 [<ffffffff805445f0>] ? _spin_unlock+0x30/0x60
 [<ffffffff802a467d>] ? get_unused_fd_flags+0xed/0x140
 [<ffffffff802a4746>] do_sys_open+0x76/0x100
 [<ffffffff802a47fb>] sys_open+0x1b/0x20
 [<ffffffff8020b90b>] system_call_after_swapgs+0x7b/0x80

This path is completely independent of the pagecache, but it does
also use RCU, so I suspect PREEMPT_RCU is freeing things before
the proper grace period. These are showing up as oopses for me
because I have DEBUG_PAGEALLOC set, but if you don't have that set
then you'll get much more subtle corruption.

Here is the find_get_pages bug FYI:
BUG: unable to handle kernel paging request at ffff8100c7997de0
IP: [<ffffffff802732ee>] find_get_pages+0xce/0x130
PGD 8063 PUD 7fa6e163 PMD cfa64163 PTE 80000000c7997163
Oops: 0000 [1] PREEMPT SMP DEBUG_PAGEALLOC
last sysfs file: /sys/devices/system/cpu/cpu7/cache/index2/shared_cpu_map
CPU 0
Modules linked in: brd
Pid: 446, comm: kswapd0 Not tainted 2.6.26-rc5-mm3 #465
RIP: 0010:[<ffffffff802732ee>]  [<ffffffff802732ee>] find_get_pages+0xce/0x130
RSP: 0000:ffff81007e4cbbf0  EFLAGS: 00010246
RAX: ffff8100c7997de0 RBX: ffff81007e4cbc90 RCX: 0000000000000001
RDX: 0000000000000002 RSI: 0000000000000001 RDI: ffffe2000447f080
RBP: ffff81007e4cbc30 R08: ffffe2000447f088 R09: 0000000000000004
R10: 0000000000000040 R11: 0000000000000040 R12: 0000000000000000
R13: ffff81007e4cbc90 R14: ffff8100c7996e18 R15: 0000000000000000
240 97   7184  1FS:  00002b774a14ccf0(0000) GS:ffffffff807e5dc0(0000) 
knlGS:0000
000000000000
CS:  0010 DS: 0018 ES: 0018 CR0: 000000008005003b
2204  25364 4164CR2: ffff8100c7997de0 CR3: 0000000000201000 CR4: 
00000000000006e
0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
88    0    8    DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 
000000000000040
0
Process kswapd0 (pid: 446, threadinfo ffff81007e4ca000, task ffff81007e4d2a00)
 ffff81007e4cbcb0 0000000e00000000000000437 65 35  0  0
                                   ffff81007e4cbc80
 0000000000000080 0000000000000052 0000000000000000 ffffffffffffffff
 ffff81007e4cbc50 ffffffff8027dcdf 0000000000000000 ffff8100c7996c28
Call Trace:
 [<ffffffff8027dcdf>] pagevec_lookup+0x1f/0x30
 [<ffffffff8027ef43>] __invalidate_mapping_pages+0x83/0x1b0
 [<ffffffff8027f07b>] invalidate_mapping_pages+0xb/0x10
 [<ffffffff802be2a3>] shrink_icache_memory+0x293/0x2a0
 [<ffffffff80281632>] ? shrink_slab+0x32/0x220
 [<ffffffff8028172d>] shrink_slab+0x12d/0x220
 [<ffffffff8028202a>] kswapd+0x53a/0x670
 [<ffffffff8027f830>] ? isolate_pages_global+0x0/0x280
 [<ffffffff805a1ada>] ? thread_return+0xa6/0x3bc
 [<ffffffff802513f0>] ? autoremove_wake_function+0x0/0x40
 [<ffffffff80281af0>] ? kswapd+0x0/0x670
 [<ffffffff80251059>] kthread+0x49/0x80
 [<ffffffff8020c878>] child_rip+0xa/0x12
 [<ffffffff8020bf63>] ? restore_args+0x0/0x30
 [<ffffffff80251010>] ? kthread+0x0/0x80
 [<ffffffff8020c86e>] ? child_rip+0x0/0x12

If you're not using PREEMPT_RCU, then I'm stumped for the moment. You'll
have to send .configs over...
 
--
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