Re: Strange "flush" process bahaviour

[Vyacheslav Dubeyko <slava-yeENwD64cLxBDgjK7y7TUQ@public.gmane.org>]

On Sun, 2013-03-31 at 15:26 +0200, Piotr Szymaniak wrote:

[snip]
> 
> Hrm, maybe this is btrfs realated then? But I suppose I don't have any
> btrfs on the second machine with flush issues.
> 

If you detected issue with flush-8:0 kernel thread then it is a NILFS2
related issue, as I understand. Because flush-8:0 processes /dev/sda
drive and you have NILFS2 partition on /dev/sda3 (/dev/sda3 on / type
nilfs2 (rw,noatime,nodiratime,gcpid=16150)).

> I had a system freeze yesterday and right after reboot flush is back.
> So it looks like I can reproduce it (something makes the flush thing
> while booting. Maybe I can catch some debug info there?). Right now I'm
> going back to 3.4.x series kernel.
> 

The visible symptoms of the issue (flush kernel thread "abnormal"
activity) is not a reason of the issue. Such "abnormal" activity of
flush kernel thread takes place because in nilfs_mdt_write_page() and
nilfs_write_page() can be called with wbc->for_kupdate or
wbc->for_background flags. It is made simply
redirty_page_for_writepage() for such case.

The real issue hides in other NILFS2 subsystem. And only clear
understanding of file system operation that initiates such "abnormal"
activity of flush kernel thread can be a basis for further debugging and
investigation of the issue.

With the best regards,
Vyacheslav Dubeyko.

> 
> Piotr Szymaniak.


--
To unsubscribe from this list: send the line "unsubscribe linux-nilfs" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html