WARNING: at /build/buildd/linux-2.6.26/drivers/usb/serial/usb-serial.c:322 serial_write_room+0x72/0x80 [usbserial]()

The Ubuntu Kernel Team is planning to move to the 2.6.27 kernel for the upcoming Intrepid Ibex 8.10 release. As a result, the kernel team would appreciate it if you could please test this newer 2.6.27 Ubuntu kernel. There are one of two ways you should be able to test:

1) If you are comfortable installing packages on your own, the linux-image-2.6.27-* package is currently available for you to install and test.

--or--

2) The upcoming Alpha5 for Intrepid Ibex 8.10 will contain this newer 2.6.27 Ubuntu kernel. Alpha5 is set to be released Thursday Sept 4. Please watch http://www.ubuntu.com/testing for Alpha5 to be announced. You should then be able to test via a LiveCD.

Please let us know immediately if this newer 2.6.27 kernel resolves the bug reported here or if the issue remains. More importantly, please open a new bug report for each new bug/regression introduced by the 2.6.27 kernel and tag the bug report with 'linux-2.6.27'. Also, please specifically note if the issue does or does not appear in the 2.6.26 kernel. Thanks again, we really appreicate your help and feedback.

I am still seeing this in 2.6.27-2.3.

This warning seems to indicate that the .write_room function is being called when the device is not open. It's not clear to me why that condition justifies a warning, though. Other tty devices don't seem to care so much.

Seems to exist on Debian as well: http://archive.netbsd.se/?ml=debian-arm&a=2008-08&m=8365675

Still happens on 2.6.27-5.8

The driver behind this is sierra.c. The kernel bug as well as the debian bug are related to option.c. I change the kernel bug to one matching better.

There seem to be debug options for usb-serial and sierra. Could you add dmesg output of the warning with

options usb-serial debug=1
options sierra debug=1

set?

Debug output attached.

Some of the things I noticed: open and close seem balanced. The interesting part is that in the first case the outdat callbacks for the written data are called and all indat callbacks see urb's with error status. In the failure case there outdat callbacks are not called before the close and two of the three indat callbacks see urb's without an error status and then try to write data.
Compared to the option driver, the sierra driver allocates the outbound urb's dynamically and therfor cannot kill them on close.

Does this sierra.ko (for i386-generic 2.6.27-7.11) give better results?

On Thu, Oct 16, 2008 at 09:50:12PM -0000, Stefan Bader wrote:
> Does this sierra.ko (for i386-generic 2.6.27-7.11) give better results?
>
> ** Attachment added: "sierra.ko"
> http://launchpadlibrarian.net/18619719/sierra.ko

As shown in the bug description, I'm running a 64-bit kernel, so I can't use
this module. Can you attach the patch you used and/or build me a matching
module?

--
 - mdz

Sorry, I am attaching the module with the correct architecture.

This patch was used for the modules.

On Mon, Oct 20, 2008 at 08:15:24AM -0000, Stefan Bader wrote:
> Sorry, I am attaching the module with the correct architecture.

With this module, I'm no longer able to reproduce the warning with my
previous test case (echoing a command to /dev/ttyUSB0). I can't currently
test whether the device still works properly otherwise, as I will need to
swap a SIM card into it.

--
 - mdz

Matt, I got a intrepid kernel including the proposed fix in my PPA at https://launchpad.net/~stefan-bader-canonical/+archive. Would it be possible for you to try this and check whether this fixes the warning while still keep the device working? Thanks.

On Thu, Jan 08, 2009 at 02:44:03PM -0000, Stefan Bader wrote:
> Matt, I got a intrepid kernel including the proposed fix in my PPA at
> https://launchpad.net/~stefan-bader-canonical/+archive. Would it be
> possible for you to try this and check whether this fixes the warning
> while still keep the device working? Thanks.

I've since upgraded to Jaunty, so I'm running 2.6.28-4.6-generic. Does that
package include this patch as well?

--
 - mdz

> I've since upgraded to Jaunty, so I'm running 2.6.28-4.6-generic. Does that
> package include this patch as well?

No, not yet. The driver itself has also changed slightly around there. But if
you could verify whether the warning will still be issued with Jaunty, that
would help too. If yes, I forward-port the patch-ords!

On Fri, Jan 09, 2009 at 09:26:13PM -0000, Stefan Bader wrote:
> > I've since upgraded to Jaunty, so I'm running 2.6.28-4.6-generic. Does that
> > package include this patch as well?
>
> No, not yet. The driver itself has also changed slightly around there. But if
> you could verify whether the warning will still be issued with Jaunty, that
> would help too. If yes, I forward-port the patch-ords!

I can still reproduce on 2.6.28-4.6-generic with:

echo "at" > /dev/ttyUSB0

--
 - mdz

I uploaded some test kernels based on 2.6.28-11.42 to http://people.ubuntu.com/~smb/bug258804/
Could you try with one of those and see whether the warning is gone and the functionality seems to be ok? Thanks.

I'm no longer able to reproduce this on 2.6.28-11.42 using my test case in https://bugs.edge.launchpad.net/ubuntu/+source/linux/+bug/258804/comments/19

This means it would be difficult to tell whether your test kernels fix the problem. Do you have any other ideas for how I could trigger the bug?

Right, I might need to check whether there has been some other means of
protecting that case. I try to see whether there is another way of getting
there but as it is related to the device being used (opened) or not I think
there would be no other way of doing so.