Re: Problem with udev & Vodafone UMTS card

[Greg KH <greg@kroah.com>]

On Wed, Feb 02, 2005 at 05:13:26AM +0100, Kay Sievers wrote:
> On Wed, Feb 02, 2005 at 02:44:05AM +0100, Marc Packenius wrote:
> > Greg KH <greg@kroah.com> writes:
> > > On Mon, Jan 24, 2005 at 05:17:53PM +0100, Kay Sievers wrote:
> > >> On Mon, 2005-01-24 at 16:58 +0100, Marc Packenius wrote:
> > >> > Kay Sievers <kay.sievers@vrfy.org> writes:
> > >> > > On Mon, 2005-01-24 at 15:15 +0100, Marc Packenius wrote:
> > >> > >> after installing udev on a Debian/sid-powered Toshiba Satellite
> > >> > >> notebook, the Vodafone UMTS card stopped working.
> > >> > >> 
> > >> > >> This pc-card includes a USB controller and feeds three serial character
> > >> > >> devices when installed properly. They're called "/dev/ttyUSB[0-2]".
> > >> > >> 
> > >> > >> When using udev, those device nodes are not created properly, instead
> > >> > >> the dev directory shows files with the filename "/dev/<NULL>[0-2]", e.g. 
> > >> > >> "/dev/<NULL>0".
> > >> > 
> > >> > >> => /var/log/daemon.log <=
> > >> > >> Jan 24 14:59:35 mudshark udev[3157]: creating device node '/dev/<NULL>0'
> > >> > >> Jan 24 14:59:35 mudshark udev[3169]: creating device node '/dev/<NULL>1'
> > >> > >> Jan 24 14:59:35 mudshark udev[3179]: creating device node '/dev/<NULL>2'
> > >> 
> > >> That's a kernel bug and not a udev issue. You may try to work around
> > >> this by placing:
> > >> KERNEL="<NULL>*", NAME="ttyUSB%n"
> > >> 
> > >> in a /etc/udev/rules.d/ file.
> > >> 
> > >> Greg, any idea what is causing this in the serial driver?
> > 
> > > No I do not.  What kernel version is this?  What driver is this (the
> > > usb-serial generic driver, or some other usb-serial driver?)
> > 
> > That's with kernels that I built locally on a Debian/sid system. I can
> > reproduce the behaviour with Debian-patched 2.6.8 to 2.6.10 and vanilla
> > 2.6.10.
> > 
> > The driver is usb-serial generic. Inserting the card and loading the
> > driver using
> > 
> >   modprobe -v usbserial vendor=0xaf0 product=0x5000
> 
> This patch may fix it. We call the probe function before the driver
> struct is initialized. You may be able to work around this bug by
> loading the module without the card inserted.
> 
> Signed-off-by: Kay Sievers <kay.sievers@vrfy.org>

Marc, does this fix your issue?

thanks,

greg k-h


-------------------------------------------------------
This SF.Net email is sponsored by: IntelliVIEW -- Interactive Reporting
Tool for open source databases. Create drag-&-drop reports. Save time
by over 75%! Publish reports on the web. Export to DOC, XLS, RTF, etc.
Download a FREE copy at http://www.intelliview.com/go/osdn_nl
_______________________________________________
Linux-hotplug-devel mailing list  http://linux-hotplug.sourceforge.net
Linux-hotplug-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linux-hotplug-devel