Jerry Gilliam wrote: > I am submitting the following self-review case, > with requested release binding of minor.
Perhaps, with IPS being the way of the future, it might not matter. But: shouldn't this be patch binding? Ie it will be unusable in svr4 packaging scripts unless it is backported, because the patch delivering this change will be required for liveupgrade? Isn't this behavior already accomplished (perhaps unintentionally) by specifying "-b /" as a basedir? --Mark > Problem > ------- > > The add_drv(1M) command to add a device driver to the system > provides a '-n' option to modify the various binding files only, > without causing the driver to be loaded and configured on the > running system. > > Developers have requested that update_drv(1M) and rem_drv(1M) > also support this occasionally useful ability. > > > Proposal > -------- > > Add support for the '-n' argument to update_drv(1M) and rem_drv(1M), > to modify a system's binding files without communicating the > change to the running kernel. > > Man page changes included. > > > NAME > update_drv - modify device driver attributes > > SYNOPSIS > + update_drv [-f | -v] [-n] driver_module > > > + update_drv [-b basedir] [-f | -v] [-n] -a [-m 'permission'] > [-i 'identify-name'] [-P 'privilege'] [-p 'policy'] driver_module > > > + update_drv [-b basedir] [-f | -v] [-n] -d [-m 'permission'] > [-i 'identify-name'] [-P 'privilege'] [-p 'policy'] driver_module > > > OPTIONS > The following options are supported: > > <.... > > + -n Do not try to load and attach > + device_driver, just modify the sys- > + tem configuration files for the > + device_driver. > + > > > > NAME > rem_drv - remove a device driver from the system > > SYNOPSIS > + rem_drv [-b basedir] [-n] device_driver > > + -n Do not try to detach and unload > + device_driver, just modify the sys- > + tem configuration files for the > + device_driver. > + >
