Nut upsdev - Mar 2006 - APC 1500RM / NUT 2.0.3 / newhidups / Solaris 10 [long]

Folks, while I've got the above combination to compile and run under
Solaris 10, it seems that there are still problems in the libusb/ugen
driver stack from Sun and or the APC Smart UPS 1500, so things aren't
actually working.

It seems that the APC Smart UPS 1500RM doesn't return the HID descriptor
length information in 0x21 as noted in a previous thread, but in my
case, it also doesn't return it in 0x02 either.

I've hacked the newhidups code so it just reads as much as it can, and
if the vendor codes match my ups, it allows things to proceed which
would otherwise return a short-read.

With this, the newhidups driver starts, recognises the UPS and gets to -
   Waiting for notifications...
     ok
   Notification: (3 bytes) => 07 0C 01

then sits and does nothing. It looks like it's stuck in kernel space,
so I'm guessing it's waiting for something from the UPS.

Lastly, the Solaris libusb document
(/usr/sfw/share/doc/libusb/libusb.txt) notes that the function
usb_get_descriptor_by_endpoint is 'Not implemented for Sun plugins',
and by my reckoning, this is what NUT uses quite a bit - the libusb
code for  usb_get_descriptor_by_endpoint seems to just use a
usb_control_msg, in the same format as NUT uses in newhidups.

So. on the off chance someone comes along again asking for newhidups
support on Solaris and/or an APC Smart UPS 1500RM, then people at
least know that compilation is only the first hurdle - hopefully
Sun will sort out the rest of the ugen/libusb driver stack and
it may all come down to the APC Smart UPS not having the correct
USB implementation.

If anyone can throw any other ideas into the pot on this one, just
let me know.

Many regards

marty

(heading back to serial ports for now!)


Checking device (051D/0002) (/dev/usb/51d.2/0)
- VendorID: 051d
- ProductID: 0002
- Manufacturer: American Power Conversion
- Product: Smart-UPS 1500 RM FW:617.3.I USB FW:1.5
- Serial Number: AS0444320421
- Bus: /dev/usb



-- 
Marty Lee                         e: marty@upstart-solutions.co.uk
Technical Director                v: +44 131 476 3775
Upstart Solutions Ltd             f: +44 871 433 8922
Scotland, UK                      w: http://www.upstart-solutions.co.uk

> It seems that the APC Smart UPS 1500RM doesn't return the HID
descriptor
> length information in 0x21 as noted in a previous thread, but in my
> case, it also doesn't return it in 0x02 either.
We have 750RM, 1000RM and 1500RM units on USB on Linux (FC3).  I could 
probably obtain get_descriptor output from all three, if it would be useful 
for comparison, and possibly also from a 3000RM.
-- 
Dr George D M Ross, School of Informatics, University of Edinburgh
    Kings Buildings, Mayfield Road, Edinburgh, Scotland, EH9 3JZ
Mail: gdmr@inf.ed.ac.uk   Voice: +44 131 650 5147   Fax: +44 131 667 7209
 PGP: 1024D/AD758CC5  B91E D430 1E0D 5883 EF6A  426C B676 5C2B AD75 8CC5


-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 239 bytes
Desc: not available
Url :
http://lists.alioth.debian.org/pipermail/nut-upsdev/attachments/20060308/26550389/attachment.pgp

Marty Lee wrote:
> 
> 
> Folks, while I've got the above combination to compile and run under
> Solaris 10, it seems that there are still problems in the libusb/ugen
> driver stack from Sun and or the APC Smart UPS 1500, so things aren't
> actually working.
> 
> It seems that the APC Smart UPS 1500RM doesn't return the HID
descriptor
> length information in 0x21 as noted in a previous thread, but in my
> case, it also doesn't return it in 0x02 either.
Is there a working 'lsusb' (or equivalent) for Solaris? What does it
show for the report descriptor length?

-- Peter