Date: Mon, 07 Nov 2005 19:00:10 +0000 From: Graham J Lee <leeg@teaching.physics.ox.ac.uk> To: John Baldwin <jhb@freebsd.org> Cc: Peter Grehan <grehan@freebsd.org>, freebsd-ppc@freebsd.org Subject: Re: 7.0-snap available Message-ID: <436FA43A.1040102@teaching.physics.ox.ac.uk> In-Reply-To: <200511071141.16445.jhb@freebsd.org> References: <436CB6EB.4080402@freebsd.org> <436E442C.4090507@teaching.physics.ox.ac.uk> <436E8E6D.80008@freebsd.org> <200511071141.16445.jhb@freebsd.org>
next in thread | previous in thread | raw e-mail | index | archive | help
John Baldwin wrote: > On Sunday 06 November 2005 06:14 pm, Peter Grehan wrote: > >> >> I think once the 'auto' option was selected, it isn't possible to roll >>back to using partitions underneath those slices. I even suspect the >>same would happen on i386. > > > No, it isn't. I often do an auto to see what it sets, then delete the > selections, bump up /'s size for a test machine (lots of kernels) and then > use auto again to setup the other partitions. I've done this for amd64 and > sparc64 recently and it worked fine. sysinstall tries to invoke swapon on > swap partitions after it newfs's everything and it seems that swapon() is > failing, probably because GEOM thinks that something else has ad1s5 open > still. Might be a bug in the ppc libdisk bits maybe? > Actually, Peter's suggestion to just go ahead and use my custom scheme worked. I've never had the failure he described on x86 though - I always proceed by selecting the auto options then tweaking to suit. But it's now unpacking base (into my monolithic / slice :-)), I'll post a dmesg to my website when I get the installation done. Cheers, Graham. -- Graham Lee UNIX Systems Manager, Oxford Physics Practical Course http://users.ox.ac.uk/~wadh1342
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?436FA43A.1040102>