NetBSD-Bugs archive
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]
install/44569: sysinst doesn't understand dkwedges
>Number: 44569
>Category: install
>Synopsis: sysinst doesn't understand dkwedges
>Confidential: no
>Severity: serious
>Priority: medium
>Responsible: install-manager
>State: open
>Class: sw-bug
>Submitter-Id: net
>Arrival-Date: Mon Feb 14 17:25:00 +0000 2011
>Originator: Taylor R Campbell <campbell+netbsd%mumble.net@localhost>
>Release: NetBSD 5.1_STABLE
>Organization:
>Environment:
System: NetBSD smalltalk.local 5.1_STABLE NetBSD 5.1_STABLE (RIADEBUG) #0: Tue
Feb 1 20:28:45 UTC 2011
root%smalltalk.local@localhost:/home/riastradh/netbsd/5/obj/sys/arch/i386/compile/RIADEBUG
i386
Architecture: i386
Machine: i386
>Description:
Sysinst doesn't understand dkwedges, so if the kernel
autodetects dkwedges on a disk on which one wants to install
NetBSD, sysinst will ignore the dkwedges, try to edit the
underlying disk, and fail because the kernel won't allow that.
>How-To-Repeat:
Boot the installer on a machine whose wd0 has a gpt. Watch the
kernel autodetect dkwedges. Run sysinst. Choose to install on
wd0 because that is the only option sysinst gives. Watch
sysinst fail to open /dev/rwd0d during newfs because the kernel
doesn't let you do that after it has already autodetected
dkwedges.
>Fix:
Yes, please! Ideally, sysinst would be able to create gpts and
maybe ponies too, but at least it should check for dkwedges and
offer to install on the dkwedges rather than the disk in this
case. It's not enough just to disable dkwedge autodetection in
the install kernel, and to let sysinst scribble over the gpt
and use an mbr instead -- that would leave sysinst just as
useless on the machines for which this matters, such as Intel
Macs, which, as I understand it, insist on gpts.
Home |
Main Index |
Thread Index |
Old Index