[prev in list] [next in list] [prev in thread] [next in thread] 

List:       linux-bcache
Subject:    Re: [BUG] attach backing device to cache_set hung
From:       Coly Li <colyli () suse ! de>
Date:       2018-08-22 11:04:13
Message-ID: eedfb93c-9153-bc23-8a8c-105b3027985f () suse ! de
[Download RAW message or body]

On 2018/8/22 6:52 PM, shenghui wrote:
> 
> Hi,
> 
> I used sda6 as a backing device and had it attached to cache_set already.
> 
> I first run detach and it worked. Then I tried to re-attach sda6 to the cache_set, \
>                 but the command blocked.
> ----------------------------------------
> /sys/block/sda/sda6/bcache# echo 1 > detach
> /sys/block/sda/sda6/bcache# echo a514972f-3a2e-42b0-a0db-d99dcde607d9 > attach
> [hung here, and ps showed state in 'D+']
> 
> 
> In dmesg, there is:
> 
> [ 1209.534265] INFO: task bash:1575 blocked for more than 120 seconds.
> [ 1209.534277]       Not tainted 4.18.0+ #4
> [ 1209.534282] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this \
> message. [ 1209.534289] bash            D    0  1575   1574 0x00000004
> [ 1209.534300] Call Trace:
> [ 1209.534319]  __schedule+0x1f2/0x740
> [ 1209.534330]  ? __switch_to_asm+0x33/0x4c
> [ 1209.534339]  schedule+0x27/0x70
> [ 1209.534347]  rwsem_down_write_failed+0x135/0x2f0
> [ 1209.534357]  call_rwsem_down_write_failed+0x9/0xc
> [ 1209.534365]  down_write+0x25/0x40
> [ 1209.534386]  bch_cached_dev_attach+0x409/0x4f0 [bcache]
> [ 1209.534403]  ? bch_keybuf_init+0x70/0x70 [bcache]
> [ 1209.534420]  bch_cached_dev_store+0x76a/0x800 [bcache]
> [ 1209.534438]  ? bch_cache_set_internal_show+0x20/0x20 [bcache]
> [ 1209.534447]  sysfs_kf_write+0x35/0x50
> [ 1209.534455]  ? sysfs_kf_bin_read+0xb0/0xb0
> [ 1209.534462]  kernfs_fop_write+0xf1/0x180
> [ 1209.534470]  ? kernfs_seq_start+0x90/0x90
> [ 1209.534480]  __vfs_write+0x33/0x1a0
> [ 1209.534491]  ? apparmor_file_permission+0x16/0x20
> [ 1209.534500]  ? security_file_permission+0x2c/0xb0
> [ 1209.534508]  ? rw_verify_area+0x5b/0x110
> [ 1209.534516]  vfs_write+0x94/0x1c0
> [ 1209.534525]  ksys_write+0x49/0xb0
> [ 1209.534533]  sys_write+0x16/0x20
> [ 1209.534542]  do_fast_syscall_32+0x7f/0x19c
> [ 1209.534552]  entry_SYSENTER_32+0x6b/0xbe
> [ 1209.534559] EIP: 0xb7f38d11
> [ 1209.534570] Code: Bad RIP value.
> [ 1209.534577] EAX: ffffffda EBX: 00000001 ECX: 00d602f0 EDX: 00000025
> [ 1209.534584] ESI: 00d602f0 EDI: 00000025 EBP: b7eead80 ESP: bfc5a290
> [ 1209.534591] DS: 007b ES: 007b FS: 0000 GS: 0033 SS: 007b EFLAGS: 00000246

Hi Shenghui,

Could you please to try the patch just posted by Shan ? I think his fix
addresses this issue.

Thanks.

Coly Li


[prev in list] [next in list] [prev in thread] [next in thread]