WARNING - OLD ARCHIVES

This is an archived copy of the Xen.org mailing list, which we have preserved to ensure that existing links to archives are not broken. The live archive, which contains the latest emails, can be found at http://lists.xen.org/
   
 
 
Xen 
 
Home Products Support Community News
 
   
 

xen-devel

Re: [Xen-devel] tapdisk2 dont remove in kernel 2.6.31.14

To: Roberto Scudeller <beto.rvs@xxxxxxxxx>
Subject: Re: [Xen-devel] tapdisk2 dont remove in kernel 2.6.31.14
From: Daniel Stodden <daniel.stodden@xxxxxxxxxx>
Date: Wed, 23 Mar 2011 18:53:48 -0700
Cc: "xen-devel@xxxxxxxxxxxxxxxxxxx" <xen-devel@xxxxxxxxxxxxxxxxxxx>, Pasi
Delivery-date: Wed, 23 Mar 2011 18:54:38 -0700
Envelope-to: www-data@xxxxxxxxxxxxxxxxxxx
In-reply-to: <20110324001603.GA32595@xxxxxxxxxxx>
List-help: <mailto:xen-devel-request@lists.xensource.com?subject=help>
List-id: Xen developer discussion <xen-devel.lists.xensource.com>
List-post: <mailto:xen-devel@lists.xensource.com>
List-subscribe: <http://lists.xensource.com/mailman/listinfo/xen-devel>, <mailto:xen-devel-request@lists.xensource.com?subject=subscribe>
List-unsubscribe: <http://lists.xensource.com/mailman/listinfo/xen-devel>, <mailto:xen-devel-request@lists.xensource.com?subject=unsubscribe>
References: <AANLkTimBjRT9KTW9iosuJQWkEPEKv1QoFxbVu3Am33aD@xxxxxxxxxxxxxx> <20110324001603.GA32595@xxxxxxxxxxx>
Sender: xen-devel-bounces@xxxxxxxxxxxxxxxxxxx
On Wed, 2011-03-23 at 20:16 -0400, Pasi Kärkkäinen wrote:
> On Wed, Mar 23, 2011 at 03:02:57PM -0300, Roberto Scudeller wrote:
> >    Hi all,
> > 
> >    I try to use kernel 2.6.31.14 from the git with xen 4.0.2-rc3-pre.
> > 
> 
> Hello,
> 
> You really shouldn't use 2.6.31 kernel.. it's not maintained anymore
> and pretty much everyone has moved to 2.6.32 (or newer) kernels.
> 
> So please switch to xen/stable-2.6.32.x branch of xen.git.
> 2.6.31 has many known bugs, that haven't been fixed.

In case that's not possible -- which is as bad, bad, bad as Pasi
indicates, so chose the answer carefully for yourself  -- could you
check if that tapdisk which refuses to exit happens to be spinning?

I.e. fire up top, and look if there's a tapdisk2 striving for 100% CPU
utilization. strace -p <pidof tapdisk> would probably show a really
tight select(2) loop.

We only broke the compat stuff once, iirc.

Daniel

> -- Pasi
> 
> >    I create tapdevice:
> >    # tapdisk2 -n aio:/root/discoteste
> >    /dev/xen/blktap-2/tapdev0
> > 
> >    And I try remove this device:
> >    # echo "1" > /sys/class/blktap2/blktap0/remove
> >    And I cancel this:
> >    ^C-bash: echo: write error: Resource temporarily unavailable
> > 
> >    In /var/log/messages:
> >    Mar 23 14:09:23 tapdisk2[11178]: Created /dev/xen/blktap-2/control device
> >    Mar 23 14:09:23 tapdisk2[11178]: Created /dev/xen/blktap-2/blktap0 device
> >    Mar 23 14:09:23 tapdisk2[11178]: Created /dev/xen/blktap-2/tapdev0 device
> >    Mar 23 14:09:23 tapdisk2[11178]: new interface: ring: 251, device: 253,
> >    minor: 0
> >    Mar 23 14:09:23 tapdisk2[11178]: I/O queue driver: lio
> >    Mar 23 14:09:23 kernel: [ 3900.429698] blktap_sysfs_create: adding
> >    attributes for dev ffff88006f57f400
> >    Mar 23 14:09:23 kernel: [ 3900.429763] blktap_control_allocate_tap:
> >    allocated tap ffff88006a9f9800
> >    Mar 23 14:09:23 tapdisk2[11178]: block-aio open('/root/discoteste')
> >    Mar 23 14:09:23 tapdisk2[11178]: open(/root/discoteste) with O_DIRECT
> >    Mar 23 14:09:23 tapdisk2[11178]: Image size:      pre sector_shift
> >    [10736369664]     post sector_shift [20969472]
> >    Mar 23 14:09:23 tapdisk2[11178]: opened image /root/discoteste (1 users,
> >    state: 0x00000001, type: 0)
> >    Mar 23 14:09:23 tapdisk2[11178]: VBD CHAIN:
> >    Mar 23 14:09:23 tapdisk2[11178]: /root/discoteste: 0
> >    Mar 23 14:09:23 kernel: [ 3900.431130] blktap_ring_open: opening device
> >    blktap0
> >    Mar 23 14:09:23 kernel: [ 3900.431137] blktap_ring_open: opened device 0
> >    Mar 23 14:09:23 kernel: [ 3900.431227] blktap_ring_mmap: blktap: mapping
> >    pid is 11178
> >    Mar 23 14:09:23 kernel: [ 3900.431267] blktap_validate_params:
> >    aio:/root/discoteste: capacity: 20969472, sector-size: 512
> >    Mar 23 14:09:23 kernel: [ 3900.431294] blktap_validate_params:
> >    aio:/root/discoteste: capacity: 20969472, sector-size: 512
> >    Mar 23 14:09:23 kernel: [ 3900.431316] blktap_device_create: minor 0
> >    sectors 20969472 sector-size 512
> >    Mar 23 14:09:23 kernel: [ 3900.431645] blktap_device_create: creation of
> >    253:0: 0
> >    Mar 23 14:09:56 kernel: [ 3931.416247] blktap_device_destroy: destroy
> >    device 0 users 0
> > 
> >    When a DomU is destroy ( or shutdown), this tapdisk2 process keeps:
> >    # ps -ef |grep tapdisk
> >    root     11178     1 98 14:09 pts/0    00:47:44 tapdisk2 -n
> >    aio:/root/discoteste
> > 
> >    When I kill the tapdisk2 process this device is removed.
> > 
> >    Is there a patch for this?
> > 
> >    Thanks for any help.
> > 
> >    --
> >    Roberto Scudeller
> 
> > _______________________________________________
> > Xen-devel mailing list
> > Xen-devel@xxxxxxxxxxxxxxxxxxx
> > http://lists.xensource.com/xen-devel
> 
> 
> _______________________________________________
> Xen-devel mailing list
> Xen-devel@xxxxxxxxxxxxxxxxxxx
> http://lists.xensource.com/xen-devel



_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-devel

<Prev in Thread] Current Thread [Next in Thread>