[breezy] No /dev/ttyUSB* is being created

[Derek Broughton]

Mickel-John Bunink wrote:

> On Mon, 2005-09-26 at 10:22 -0300, Derek Broughton wrote:
>> 
>> I don't know if that's the correct kernel name, but you should try
>> reading the "writing_udev_rules" file that comes with udev - check out
>> udevtest. That'll help give you some idea of what's happening in udev...
> 
>> Do:
>>  sudo modprobe visor
>> 
> 
> Right. So far I've modprobed visor, and got no errors back, so that
> would mean the kernel would support my palm. Oh, b.t.w. I've got a palm
> m500. I've also tried making a node /dev/ttyUSB0 as well
> as /dev/ttyUSB1. Both did not work so I removed them just in case udev
> wants the nodes.
> 
> Getting nowhere manually I've looked into udev.
> udevtest /dev/ttyUSB0 returned:
> 
> version 050
> looking at '/dev/ttyUSB0'
> sysfs_open_class_device_path failed

Reread the doc - the message is a pretty good hint, actually.  The argument
to udevtest is a path under /sys/class
> 
> udevtesting /dev/ttyUSB1 returned thesame (looking at /dev/ttyUSB1
> offcourse). I don't know if I used udevtest correctly, but the man page
> doesn't make things much clearer also.

Look in the documentation files
- /usr/share/doc/udev/writing_udev_rules/index.html

> The only thing left to do (afaik) is making the udev rules myself. From
> what I can figure out of the documentation I need the serial number.

Why would you need that?  Push the sync button, and watch syslog for the
hotplug events - that's going to have everything you need.

> Also, this seems a lot of work for just syncing with a palm. Is this
> normal for any new hardware? I can't call this very userfriendly. Is
> there maybe a frontend that can automate detecting hardware and
> configuring udev rules?

Well, for most people, it Just Works.  discover and the builtin configs for
hotplug & udev do the job.  You've got a configuration that doesn't fit. 
When you get it working, if you post what you had to do to the appropriate
bugzilla, the next person won't have to do that.
-- 
derek