Comment 35 for bug 204868

There's still an open hal task here. Unfortunately there is no stack trace, lshal, or other output which would indicate what the problem is. I presume that info.linux.driver was not set?

There is no chance that info.linux.driver can be present and NULL, see hald/linux/osspec.c, hal_util_set_driver(). If the sysfs path of a device is invalid/not existing/NULL, then the property is not set at all. So I assume what happened was that NM did libhal_device_get_property_string("info.linux.driver") and used the return value without checking for NULL, or using libhal_device_property_exists().

Please correct me if I am wrong and reopen.