Bug#695492: [wheezy] Ctrl-C on program accessing CIFS mount breaks/hangs mount
On Sat, Dec 08, 2012 at 10:11:22PM -0800, Jonathan Nieder wrote:
> Hi John,
>
> John Darrah wrote:
>
> > When using a program like 'find' or 'cp' in an operation involving a CIFS mounted network
> > location and then typing a Ctrl-C to terminate it will cause -all- CIFS mount points to
> > become unusable. To fix this, you just unmount them and then re-mount again. This is repeatable
> > as often as I choose.
>
> That's beautiful. Thanks for a clear report.
>
> Can you say a little more about the server hosting this share (Windows
> Server or samba? What version? etc).
>
This is the target system:
OS Name Microsoft® Windows Server® 2008 Standard Version 6.0.6001 Service Pack 1 Build 6001
Processor Intel(R) Xeon(R) CPU E5405 @ 2.00GHz, 1995 Mhz, 4 Core(s), 4 Logical Processor(s)
BIOS Version/Date Dell Inc. 2.3.1, 4/29/2008
SMBIOS Version 2.5
> > I might also note this behavior has been around since I ran 2.6 series.
>
> If you remember which version introduced it, that could help in
> tracking this down. (In other words, what are the newest working and
> oldest non-working versions you've tried?)
>
Unfortunately, I can't use my current server as a test case
but I will attempt to duplicate this behavior in a VM or on
my old retired server. This may take a couple of weeks as
weekends is the only time I have to work on this... but I
will get it done. :-)
>
> There has been work on the upstream CIFS driver recently, including
> making that message more precise. Could you try 3.6.y from
> experimental and let us know how it goes?
>
Yes, I will try a 3.6.Y kernel in my testing.
> If it's fixed, the next step will be to try to find which patch fixed
> it and apply the same to wheezy. If it's broken, we can take this
> upstream. So either result is progress.
>
I'm sorry I have been quiet on this so long. I had assumed
that I was triggering some sort of debug case as I could not
ever find any reference on the net or in bug reports. I just
assumed it was some config error on my part. As most of the
processes on this machine are non-interactive, I almost
never triggered this issue and so never pursued any
resolution.
I'll try not to be so apathetic in the future. :-)
-- john
Reply to: