On 2 May 2014 05:54, Slava Pestov <sp@xxxxxxxxx> wrote:
> On Thu, May 1, 2014 at 2:38 AM, Daniel Smedegaard Buus
> <danielbuus@xxxxxxxxx> wrote:
>> On Wed, Apr 30, 2014 at 7:24 PM, Darrick J. Wong
>> <darrick.wong@xxxxxxxxxx> wrote:
>>>
>>> I haven't spent time on figuring out the other source of load average. Kent
>>> didn't seem to like the patch to convert the bcache_writeback thread to
>>> interruptible sleep (I recall he said it was 'wrong', but didn't elaborate).
>>
>> Sorry to hear that... Would be really nice to be able to go back to
>> normal load. And I cannot revert to an older kernel, as I need
>> 3.15-rc2 or greater to fix a different problem concerning Oracle Java
> Hi Daniel and Darrick,
>
> I mailed a patch that attempts to fix the uninterruptible issue while
> taking Kent's feedback regarding your earlier patch into account.
> Please test it out and let me know what you think.
Hi Slava,
Apologies for the delays. I rebuilt 3.14.4 with your bcache patch [1]
and the 'bcache_writeback blocked for more than 120 seconds' don't
occur, though when the bcache threads are torn down during reboot, we
crash [2] at:
static void cached_dev_free(struct closure *cl)
{
struct cached_dev *dc = container_of(cl, struct cached_dev, disk.cl);
cancel_delayed_work_sync(&dc->writeback_rate_update);
kthread_stop(dc->writeback_thread);
dc->writeback_thread is clearly zero, as likely the struct cached_dev
was freed already.
Many thanks,
Daniel
[1] http://www.spinics.net/lists/linux-bcache/msg02464.html
-- [2]
BUG: unable to handle kernel NULL pointer dereference at 0000000000000010
IP: [<ffffffff8108bff6>] kthread_stop+0x16/0xe0
PGD 0
Oops: 0002 [#1] SMP
Modules linked in: microcode(F) nfsd(F) nfs_acl(F) auth_rpcgss(F)
nfs(F) fscache(F) lockd(F) sunrpc(F) joydev(F) ipmi_si(F)
ipmi_msghandler(F) psmouse(F) serio_raw(F) video(F) mac_hid(F)
lpc_ich(F) lp(F) parport(F) btrfs(F) raid6_pq(F) bcache(F) xor(F)
hid_generic(F) usbhid(F) hid(F) e1000e(F) ptp(F) pps_core(F) ahci(F)
CPU: 2 PID: 27 Comm: kworker/2:0 Tainted: GF 3.14.4-bcachefix+ #3
Hardware name: Supermicro X9SCL/X9SCM/X9SCL/X9SCM, BIOS 2.0a 06/08/2012
Workqueue: events cached_dev_free [bcache]
task: ffff8804097363c0 ti: ffff8804097d2000 task.ti: ffff8804097d2000
RIP: 0010:[<ffffffff8108bff6>] [<ffffffff8108bff6>] kthread_stop+0x16/0xe0
RSP: 0018:ffff8804097d3df0 EFLAGS: 00010292
RAX: 0000000fffffff00 RBX: ffff8804034e0010 RCX: 000000007fffffff
RDX: 0000000000000296 RSI: 000000007fffffff RDI: 0000000000000000
RBP: ffff8804097d3e08 R08: 20100d3800400000 R09: 0080000000000000
R10: dfef7acc030e0010 R11: 0000000000000400 R12: 0000000000000000
R13: ffff8804034e0010 R14: 0000000000000000 R15: 0000000000000080
FS: 0000000000000000(0000) GS:ffff88041fd00000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 0000000000000010 CR3: 0000000001c0e000 CR4: 00000000000407e0
Stack:
ffff8804034e0010 ffff88041fd13d80 ffff8804034e0010 ffff8804097d3e20
ffffffffa00b2dd5 ffff880409503580 ffff8804097d3e68 ffffffff81084482
000000001fd13d98 ffff88041fd17f00 ffff88041fd13d98 ffff8804095035b0
Call Trace:
[<ffffffffa00b2dd5>] cached_dev_free+0x25/0x100 [bcache]
[<ffffffff81084482>] process_one_work+0x182/0x450
[<ffffffff81085241>] worker_thread+0x121/0x410
[<ffffffff81085120>] ? rescuer_thread+0x3e0/0x3e0
[<ffffffff8108bdc2>] kthread+0xd2/0xf0
[<ffffffff8108bcf0>] ? kthread_create_on_node+0x190/0x190
[<ffffffff8173c67c>] ret_from_fork+0x7c/0xb0
[<ffffffff8108bcf0>] ? kthread_create_on_node+0x190/0x190
Code: e8 20 ff ff ff 48 89 df be 00 02 00 00 e8 63 10 01 00 5b 5d c3
66 66 66 66 90 55 48 89 e5 41 55 41 54 49 89 fc 53 66 66 66 66 90 <f0>
41 ff 44 24 10 49 8b 9c 24 a0 04 00 00 48 85 db 74 21 f0 80
RIP [<ffffffff8108bff6>] kthread_stop+0x16/0xe0
RSP <ffff8804097d3df0>
CR2: 0000000000000010
--
Daniel J Blueman
--
To unsubscribe from this list: send the line "unsubscribe linux-bcache" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
