On Mon, 2009-07-27 at 11:51 +0200, Jan Damborsky wrote: > Hi Alex, > > Alexander Eremin wrote: > >> Hi Alex, > >> > >> test_td/libtd consumes libdiskmgt library for > >> obtaining information about drives - libdiskmgt > >> enumerates device tree in order to > >> determine what is available. > >> > >> Could you please check if drives which are > >> not reported by test_td have links in /dev/ directory > >> populated before and after you invoke format(1M) ? > >> > >> Thank you, > >> Jan > >> > > Thanks Jan, > > Think I found cause - if I ever do 'ls /dev', then test_td can see second > > disk. Format looks for /dev in > > http://src.opensolaris.org/source/xref/onnv/onnv-gate/usr/src/cmd/format/startup.c#do_search, > > therefore has no problems. > > For strange /dev in snv114 I reported already > > (http://opensolaris.org/jive/thread.jspa?messageID=383550). Now I don't > > see here same problem using snv116. > > As far as LiveCD or AI (Automated Installer) is concerned, devfsadm(1M) > is explicitly called to make sure device tree is populated appropriately: > > http://src.opensolaris.org/source/xref/caiman/slim_source/usr/src/cmd/slim-install/svc/live-devices-local#42 > > We encountered similar issue you saw with test_td, when we initially > omitted this step in AI: > http://defect.opensolaris.org/bz/show_bug.cgi?id=5886 > > Cheers, > Jan Yes, I'm using devfsadm in live-devices-local, and right c6t0d0.. entries are presented in /dev, but some advanced process like 'cd /dev' or 'ls /dev' is needed in my case. But this not libtd problem, so thanks for help, Jan. --
::alhazred
