Nut upsdev - Jun 2007 - Possible error in drivers/main-hal.c

Looking at the above sources, it looks like the device_path is
configured twice:

509 device_path = xstrdup("auto"); /*getenv
("HAL_PROP_HIDDEV_DEVICE"); */

626 device_path = xstrdup(argv[0]);

Other than the fact that we leak a few bytes of memory here by using
xstrdup() twice without free'ing in between, I don't think this is what
we intended to do. It looks to me that the socket will now be named
'usbhid-ups-usbhid-ups', which would be ugly? Since I don't use HAL,
I
can't verify this, but I think we really want it to look like
'usbhid-ups-auto'.

Best regards, Arjen

2007/6/6, Arjen de Korte <nut+devel at
de-korte.org>:
> Looking at the above sources, it looks like the device_path is
> configured twice:
>
> 509 device_path = xstrdup("auto"); /*getenv
("HAL_PROP_HIDDEV_DEVICE"); */
>
> 626 device_path = xstrdup(argv[0]);
>
> Other than the fact that we leak a few bytes of memory here by using
> xstrdup() twice without free'ing in between, I don't think this is
what
> we intended to do. It looks to me that the socket will now be named
> 'usbhid-ups-usbhid-ups', which would be ugly? Since I don't use
HAL, I
> can't verify this, but I think we really want it to look like
> 'usbhid-ups-auto'.
good catch, thanks Arjen.
fixed on Testing and trunk.

there is still a cleanup to be done here, but I'm now getting more
visibility on where we'll go with the HAL bridge.

Arnaud
-- 
Linux / Unix Expert - MGE UPS SYSTEMS - R&D Dpt
Network UPS Tools (NUT) Project Leader - http://www.networkupstools.org/
Debian Developer - http://people.debian.org/~aquette/
OpenSource Developer - http://arnaud.quette.free.fr/