Full kern.log is to big for attachment, you can find it under:
http://oban.stoch.uni-linz.ac.at/kern.log

Reply-To: James.Bottomley@HansenPartnership.com

[Cc of more relevant fs lists added]
On Thu, 2012-08-16 at 12:40 +0000, bugzilla-daemon@bugzilla.kernel.org
wrote:
> https://bugzilla.kernel.org/show_bug.cgi?id=46031
> 
>            Summary: kswapd0 moving to uninterruptible sleep (STAT D)
>            Product: IO/Storage
>            Version: 2.5
>     Kernel Version: 3.5.2
>           Platform: All
>         OS/Version: Linux
>               Tree: Mainline
>             Status: NEW
>           Severity: normal
>           Priority: P1
>          Component: SCSI
>         AssignedTo: linux-scsi@vger.kernel.org
>         ReportedBy: Markus.Hetzmannseder@jku.at
>         Regression: No
> 
> 
> Hi,
> 
> I have a hangup problem with my litle server. The Hardware is a Dell
> Poweredge
> SC1430 with mirrored harddrives conntected on the PERC 5/i Adapter, it uses
> the
> megaraid/megasas scsi driver.
> 
> The problem occurs specially at heavy diskIO like update of the file name
> database.
> 
> The system is running in x86_PAE mode with 8GB RAM installed. So far I have
> tried out kernel 3.1.4 3.6.0-rc1 and now running 3.5.2 version.
> 
> According to kernel.log its allways the kswapd0 process which starts to hang
> in
> STAT D mode. After that more and more processes are hitting STAT D and the
> system is getting practically unusable. In that state a login over the
> network
> is still possible. A normal reboot is not working anymore (keeps waiting to
> kill some processes) only a reboot -f is doing the job.
> 
> When the error accurs the /proc/sys/kernel/tainted has state 512
> 
> In the attachment I add all the kern.log output I got so far.
> 
> In the kern.log I see something like this:
> 
> -----------------------------------------------------------------
> Aug 16 11:49:57 servername kernel: [ 7361.062388] WARNING: at
> fs/jbd/journal.c:469 __log_start_commit+0x6b/0x7e()
> Aug 16 11:49:57 servername kernel: [ 7361.062391] Hardware name: PowerEdge
> SC1430
> Aug 16 11:49:57 servername kernel: [ 7361.062393] jbd: bad log_start_commit:
> 2168023832 2168023832 0 0

This would seem to be the root cause unless there's some clue in the log
you cut (the fs people may ask for more).

This either indicates a bug in the ext4 journalling layer or that
there's some type of hardware problem like memory corruption.  Hopefully
the fs people can provide more insight.  It looks like kswapd is trying
to do direct reclaim and the write fails for some reason (which could be
the HW problem) and then it doesn't recover from the error.

James


> Aug 16 11:49:57 servername kernel: [ 7361.062395] Modules linked in: ppdev lp
> bluetooth rfkill mperf cpufreq_conservative cpufreq_userspace
> cpufreq_powersave
> cpufreq_stats nfsd exportfs nfs lockd fscache auth_rpcgss nfs_acl sunrpc fuse
> loop psmouse lpc_ich mfd_core i5000_edac edac_core serio_raw evdev tpm_tis
> pcspkr tpm shpchp hid_generic coretemp rng_core dcdbas tpm_bios i5k_amb
> pci_hotplug microcode parport_pc processor button parport thermal_sys usbhid
> hid uhci_hcd sg sr_mod tg3 cdrom ehci_hcd libphy usbcore usb_common sd_mod
> crc_t10dif [last unloaded: scsi_wait_scan]
> Aug 16 11:49:57 servername kernel: [ 7361.062454] Pid: 46, comm: kswapd0 Not
> tainted 3.5.2 #1
> Aug 16 11:49:57 servername kernel: [ 7361.062456] Call Trace:
> Aug 16 11:49:57 servername kernel: [ 7361.062464]  [<c1023a3d>] ?
> warn_slowpath_common+0x6a/0x7b
> Aug 16 11:49:57 servername kernel: [ 7361.062468]  [<c11575ae>] ?
> __log_start_commit+0x6b/0x7e
> Aug 16 11:49:57 servername kernel: [ 7361.062472]  [<c1023ab4>] ?
> warn_slowpath_fmt+0x28/0x2c
> Aug 16 11:49:57 servername kernel: [ 7361.062476]  [<c11575ae>] ?
> __log_start_commit+0x6b/0x7e
> Aug 16 11:49:57 servername kernel: [ 7361.062480]  [<c1157625>] ?
> log_start_commit+0x1b/0x22
> Aug 16 11:49:57 servername kernel: [ 7361.062484]  [<c110fa0a>] ?
> ext3_evict_inode+0xbe/0x1cc
> Aug 16 11:49:57 servername kernel: [ 7361.062489]  [<c10d4a6a>] ?
> evict+0x8a/0x126
> Aug 16 11:49:57 servername kernel: [ 7361.062492]  [<c10d4e72>] ?
> dispose_list+0x2e/0x37
> Aug 16 11:49:57 servername kernel: [ 7361.062496]  [<c10d50fa>] ?
> prune_icache_sb+0x27f/0x287
> Aug 16 11:49:57 servername kernel: [ 7361.062501]  [<c10c5f21>] ?
> prune_super+0xa2/0xf5
> Aug 16 11:49:57 servername kernel: [ 7361.062506]  [<c109f8bb>] ?
> shrink_slab+0x1b7/0x254
> Aug 16 11:49:57 servername kernel: [ 7361.062509]  [<c10a16fe>] ?
> kswapd+0x54f/0x805
> Aug 16 11:49:57 servername kernel: [ 7361.062515]  [<c103ad7d>] ?
> wake_up_bit+0x56/0x56
> Aug 16 11:49:57 servername kernel: [ 7361.062519]  [<c10a11af>] ?
> try_to_free_pages+0xd5/0xd5
> Aug 16 11:49:57 servername kernel: [ 7361.062522]  [<c103aa1f>] ?
> kthread+0x68/0x6d
> Aug 16 11:49:57 servername kernel: [ 7361.062526]  [<c103a9b7>] ?
> kthread_freezable_should_stop+0x45/0x45
> Aug 16 11:49:57 servername kernel: [ 7361.062531]  [<c1346b7e>] ?
> kernel_thread_helper+0x6/0xd
> Aug 16 11:49:57 servername kernel: [ 7361.062534] ---[ end trace
> 7f2284fed89c7a03 ]---
> Aug 16 12:33:17 servername kernel: [ 9960.684081] INFO: task acroread:3117
> blocked for more than 120 seconds.
> Aug 16 12:33:17 servername kernel: [ 9960.684116] "echo 0 >
> /proc/sys/kernel/hung_task_timeout_secs" disables this message.
> Aug 16 12:33:17 servername kernel: [ 9960.684162] acroread        D 00000000  
>  0  3117   3115 0x00000000
> Aug 16 12:33:17 servername kernel: [ 9960.684179]  f0ef69a0 00200082 00000001
> 00000000 c6b6ddac 00000002 39abe377 c1514dc0
> Aug 16 12:33:17 servername kernel: [ 9960.684186]  c6b6ddac c2c0dd38 c1514dc0
> c1514dc0 f0ef69a0 c1514dc0 0101b7ba 00000020
> Aug 16 12:33:17 servername kernel: [ 9960.684192]  c10d7899 c2c0ddb0 009e8d67
> 00000000 da7ff09c c6b6ddac 0000000b ce221700
> Aug 16 12:33:17 servername kernel: [ 9960.684199] Call Trace:
> Aug 16 12:33:17 servername kernel: [ 9960.684210]  [<c10d7899>] ?
> mntput_no_expire+0x15/0xf1
> Aug 16 12:33:17 servername kernel: [ 9960.684215]  [<c1110214>] ?
> search_dirblock+0x5f/0x93
> Aug 16 12:33:17 servername kernel: [ 9960.684221]  [<c103aeef>] ?
> prepare_to_wait+0x14/0x52
> Aug 16 12:33:17 servername kernel: [ 9960.684225]  [<c10d4106>] ?
> __wait_on_freeing_inode+0x6e/0x88
> Aug 16 12:33:17 servername kernel: [ 9960.684229]  [<c103ada6>] ?
> autoremove_wake_function+0x29/0x29
> Aug 16 12:33:17 servername kernel: [ 9960.684232]  [<c10d4155>] ?
> find_inode_fast+0x35/0x6d
> Aug 16 12:33:17 servername kernel: [ 9960.684236]  [<c10d54a8>] ?
> iget_locked+0x2f/0xd5
> Aug 16 12:33:17 servername kernel: [ 9960.684240]  [<c110ce15>] ?
> ext3_iget+0x18/0x332
> Aug 16 12:33:17 servername kernel: [ 9960.684243]  [<c1111e0c>] ?
> ext3_lookup+0x5d/0x9b
> Aug 16 12:33:17 servername kernel: [ 9960.684248]  [<c10cb8b8>] ?
> __lookup_hash+0x8f/0xa8
> Aug 16 12:33:17 servername kernel: [ 9960.684251]  [<c10cb8fd>] ?
> lookup_slow+0x2c/0x78
> Aug 16 12:33:17 servername kernel: [ 9960.684255]  [<c10cccde>] ?
> walk_component+0x48/0xe8
> Aug 16 12:33:17 servername kernel: [ 9960.684259]  [<c10cdc9a>] ?
> path_lookupat+0xa4/0x2a6
> Aug 16 12:33:17 servername kernel: [ 9960.684264]  [<c109a79a>] ?
> free_hot_cold_page_list+0x4a/0x60
> Aug 16 12:33:17 servername kernel: [ 9960.684268]  [<c10cdeb7>] ?
> do_path_lookup+0x1b/0x85
> Aug 16 12:33:17 servername kernel: [ 9960.684271]  [<c10ce88c>] ?
> user_path_at_empty+0x3d/0x65
> Aug 16 12:33:17 servername kernel: [ 9960.684277]  [<c10adb55>] ?
> handle_mm_fault+0x118/0x129
> Aug 16 12:33:17 servername kernel: [ 9960.684281]  [<c10ce8bf>] ?
> user_path_at+0xb/0xe
> Aug 16 12:33:17 servername kernel: [ 9960.684284]  [<c10c75ab>] ?
> vfs_fstatat+0x3d/0x63
> Aug 16 12:33:17 servername kernel: [ 9960.684287]  [<c10c768d>] ?
> vfs_stat+0x10/0x12
> Aug 16 12:33:17 servername kernel: [ 9960.684290]  [<c10c769e>] ?
> sys_stat64+0xf/0x23
> Aug 16 12:33:17 servername kernel: [ 9960.684295]  [<c1343c4b>] ?
> spurious_fault+0xe5/0xe5
> Aug 16 12:33:17 servername kernel: [ 9960.684299]  [<c1346613>] ?
> sysenter_do_call+0x12/0x22
> Aug 16 12:35:17 servername kernel: [10080.684102] INFO: task acroread:3117
> blocked for more than 120 seconds.
> Aug 16 12:35:17 servername kernel: [10080.684138] "echo 0 >
> /proc/sys/kernel/hung_task_timeout_secs" disables this message.
> Aug 16 12:35:17 servername kernel: [10080.684183] acroread        D 00000000  
>  0  3117   3115 0x00000000
> Aug 16 12:35:17 servername kernel: [10080.684200]  f0ef69a0 00200082 00000001
> 00000000 c6b6ddac 00000002 39abe377 c1514dc0
> Aug 16 12:35:17 servername kernel: [10080.684207]  c6b6ddac c2c0dd38 c1514dc0
> c1514dc0 f0ef69a0 c1514dc0 0101b7ba 00000020
> Aug 16 12:35:17 servername kernel: [10080.684214]  c10d7899 c2c0ddb0 009e8d67
> 00000000 da7ff09c c6b6ddac 0000000b ce221700
> Aug 16 12:35:17 servername kernel: [10080.684220] Call Trace:
> Aug 16 12:35:17 servername kernel: [10080.684231]  [<c10d7899>] ?
> mntput_no_expire+0x15/0xf1
> Aug 16 12:35:17 servername kernel: [10080.684237]  [<c1110214>] ?
> search_dirblock+0x5f/0x93
> Aug 16 12:35:17 servername kernel: [10080.684243]  [<c103aeef>] ?
> prepare_to_wait+0x14/0x52
> Aug 16 12:35:17 servername kernel: [10080.684247]  [<c10d4106>] ?
> __wait_on_freeing_inode+0x6e/0x88
> Aug 16 12:35:17 servername kernel: [10080.684251]  [<c103ada6>] ?
> autoremove_wake_function+0x29/0x29
> Aug 16 12:35:17 servername kernel: [10080.684254]  [<c10d4155>] ?
> find_inode_fast+0x35/0x6d
> Aug 16 12:35:17 servername kernel: [10080.684258]  [<c10d54a8>] ?
> iget_locked+0x2f/0xd5
> Aug 16 12:35:17 servername kernel: [10080.684261]  [<c110ce15>] ?
> ext3_iget+0x18/0x332
> Aug 16 12:35:17 servername kernel: [10080.684265]  [<c1111e0c>] ?
> ext3_lookup+0x5d/0x9b
> Aug 16 12:35:17 servername kernel: [10080.684269]  [<c10cb8b8>] ?
> __lookup_hash+0x8f/0xa8
> Aug 16 12:35:17 servername kernel: [10080.684273]  [<c10cb8fd>] ?
> lookup_slow+0x2c/0x78
> Aug 16 12:35:17 servername kernel: [10080.684276]  [<c10cccde>] ?
> walk_component+0x48/0xe8
> Aug 16 12:35:17 servername kernel: [10080.684280]  [<c10cdc9a>] ?
> path_lookupat+0xa4/0x2a6
> Aug 16 12:35:17 servername kernel: [10080.684285]  [<c109a79a>] ?
> free_hot_cold_page_list+0x4a/0x60
> Aug 16 12:35:17 servername kernel: [10080.684289]  [<c10cdeb7>] ?
> do_path_lookup+0x1b/0x85
> Aug 16 12:35:17 servername kernel: [10080.684292]  [<c10ce88c>] ?
> user_path_at_empty+0x3d/0x65
> Aug 16 12:35:17 servername kernel: [10080.684298]  [<c10adb55>] ?
> handle_mm_fault+0x118/0x129
> Aug 16 12:35:17 servername kernel: [10080.684302]  [<c10ce8bf>] ?
> user_path_at+0xb/0xe
> Aug 16 12:35:17 servername kernel: [10080.684305]  [<c10c75ab>] ?
> vfs_fstatat+0x3d/0x63
> Aug 16 12:35:17 servername kernel: [10080.684308]  [<c10c768d>] ?
> vfs_stat+0x10/0x12
> Aug 16 12:35:17 servername kernel: [10080.684311]  [<c10c769e>] ?
> sys_stat64+0xf/0x23
> Aug 16 12:35:17 servername kernel: [10080.684316]  [<c1343c4b>] ?
> spurious_fault+0xe5/0xe5
> Aug 16 12:35:17 servername kernel: [10080.684320]  [<c1346613>] ?
> sysenter_do_call+0x12/0x22
> --------------------------------------------------------------
> 
> Any hints how to get the system back in a stable mode?
> 
> Markus
>

Thx for the help on it. As far as i could strip down the reasion for the errors, its deffinitly something hardware related. I think there is something wrong with the motherboard. 

Markus