Hi William, thank you for catching this problem. The changes look good to me.
As far as '-d <device>' feature is concerned, I would suggest to completely remove it from the code, as it has not been maintained for a while (e.g. list of packages to be installed/removed is not up to date) and introduces unnecessary code duplication. Since it is definitely out of the scope of this fix, could you please file bug for this issue ? Thank you, Jan On 01/19/10 09:30 AM, William Schumann wrote: > http://cr.opensolaris.org/~wmsch/bug-13954/ > http://defect.opensolaris.org/bz/show_bug.cgi?id=13954 > > The iSCSI install process works like this: > - the AI manifest is read, and iSCSI parameters are input > - the iSCSI initiator mounts the target, which appears as a new device > in /dev/dsk > - target discovery runs, picking up the iSCSI device and automatically > marks it as the install device > > So, manifest processing must proceed TD for iSCSI. > > Tested iSCSI/non-iSCSI installations - passed > Tested unsupported -d <device> on auto-install command line - failed > due to changes preceding this bug. Will deal with this case later, > based on actual usage of the auto-install -d option > _______________________________________________ > caiman-discuss mailing list > caiman-discuss at opensolaris.org > http://mail.opensolaris.org/mailman/listinfo/caiman-discuss