[Nut-upsdev] FreeBSD, NUT OmniVS1500
clepple at gmail.com
Sat Sep 5 13:16:20 UTC 2009
On Sep 4, 2009, at 11:01 PM, Charles Sprickman wrote:
> On Fri, 4 Sep 2009, Charles Lepple wrote:
>> Hi Charles,
>> On Sep 4, 2009, at 9:00 PM, Charles Sprickman wrote:
>>> USB error: could not set alt intf 0/0: Invalid argument
>> Interesting that you should mention this. It came up just the other
>> day regarding Mac OS X:
>> That error should go away if you comment out the
>> usb_set_altinterface() line in drivers/libusb.c.
> Sure does!
OK. I will make sure that other drivers don't break with the removal
of that line, but it sounds like that wasn't blocking the rest of the
>>> USB error: error reading from interrupt endpoint /dev/ugen0.1:
>>> temporarily unavailable
>> Not so sure about this one. Does this version of FreeBSD have
>> separate device files for /dev/ugen0 and /dev/ugen0.1?
> [root at miko /home/spork]# ls -l /dev/ugen0*
> crw-rw---- 1 root nut 0, 70 Sep 4 22:09 /dev/ugen0
> crw-rw---- 1 root nut 0, 71 Sep 4 22:09 /dev/ugen0.1
> Still get the other errors though regarding "resource temporarily
> unavailable". Google's not showing me much other than one thread
> about some usb-based outboard I/O kit having a similar issue:
/usr/include/errno.h:#define EAGAIN 35 /* Resource temporarily
I seem to remember seeing something about EAGAIN in the kernel code
for 6.2 or 6.3. Maybe I will have time later this weekend to look at
One of the issues with most of these Tripp Lite devices is that they
specify a polling interval of 10 ms, but they really aren't designed
to be polled that quickly.
I did have a branch to deal with Tripp Lite on FreeBSD about two years
ago, but now that I look at it, this uses uhid instead of libusb/ugen:
Not sure if it's worth playing with that again, since we don't have a
good way to merge that back into the mainline, and toggling uhid
support is not terribly quick.
More information about the Nut-upsdev