[Nut-upsdev] Bug#462468: nut: megatec_usb ups not detected since 2.2.1 update anymore

[Arnaud Quette]

2008/1/26, Arjen de Korte <arjen at de-korte.org>:
> [...]
>
> >> The output of lsusb and ls -lR /dev/bus/usb is the same for the two
> >> versions.
> >
> > thanks for your quick feedback Sven.
> >
> > @Carlos: there is a clear regression here.
>
> Poor Carlos, he is getting all the flak, while the problem is in the
> megatec_usb.c... :-)

right, sorry for this Carlos!

I still have a hard link on the pattern megatec <=> Carlos, and I've
not updated it for the "_usb" version ;-)

@Carlos: can you be the general interface for megatec, both serial and usb?
since you're the NUT reference for megatec...

> > I'm thinking of a recent timeout problem (was 1000 instead of 5000).
> > Do you see something?
>
> Increasing the timeout value here, is the worst fix possible. Under no
> circumstance should drivers hang around that long in upsdrv_updateinfo().

that is the case for usbhid-ups: the various functions that accept a
TIMEOUT use 4 sec...
And 1 second can be too short.

> The problem here is that none of the subdrivers in 'megatec_usb.c' in
> nut-2.2.1 handle the return codes from the libusb functions properly. By
> sending the EAGAIN error code, they clearly indicate that the command
> needs to be retried, instead of declaring it a failure. This might be
> handled a little better from r1211 onwards, which added a the possibility
> to reconnect. I'm not sure if reconnecting is the proper way to handle
> this though. It might be better to just retry the command the next time.

has this been backported to Testing?

Arnaud
-- 
Linux / Unix Expert R&D - MGE Office Protection Systems - http://www.mgeops.com
Network UPS Tools (NUT) Project Leader - http://www.networkupstools.org/
Debian Developer - http://people.debian.org/~aquette/
Free Software Developer - http://arnaud.quette.free.fr/