Skip site navigation (1)Skip section navigation (2)
Date:      Tue, 8 Dec 2009 21:20:16 +0100
From:      Marius Strobl <marius@alchemy.franken.de>
To:        Anton Shterenlikht <mexas@bristol.ac.uk>
Cc:        freebsd-sparc64@freebsd.org
Subject:   Re: trouble installing to sun blade 1500
Message-ID:  <20091208202016.GM6562@alchemy.franken.de>
In-Reply-To: <20091203131609.GA86854@mech-cluster241.men.bris.ac.uk>
References:  <20091201154904.GA11679@mech-cluster241.men.bris.ac.uk> <alpine.LNX.2.00.0912011723060.21559@servww6.ww.uni-erlangen.de> <20091201165752.GA12095@mech-cluster241.men.bris.ac.uk> <20091201215130.GA64974@alchemy.franken.de> <20091202124137.GA20168@mech-cluster241.men.bris.ac.uk> <20091202125915.GA20276@mech-cluster241.men.bris.ac.uk> <20091202192134.GA6562@alchemy.franken.de> <20091203131609.GA86854@mech-cluster241.men.bris.ac.uk>

next in thread | previous in thread | raw e-mail | index | archive | help
On Thu, Dec 03, 2009 at 01:16:09PM +0000, Anton Shterenlikht wrote:
> On Wed, Dec 02, 2009 at 08:21:34PM +0100, Marius Strobl wrote:
> > On Wed, Dec 02, 2009 at 12:59:15PM +0000, Anton Shterenlikht wrote:
> > > On Wed, Dec 02, 2009 at 12:41:37PM +0000, Anton Shterenlikht wrote:
> > > > On Tue, Dec 01, 2009 at 10:51:30PM +0100, Marius Strobl wrote:
> > > > 
> > > > > > > > I get to OK prompt, type "boot cdrom", but the process
> > > > > > > > stops after a few seconds with:
> > > > > > > >
> > > > > > > > 	jumping to kernel entry at 0xc0088000.
> > > > > > > >
> > > > > 
> > > > > You're using an unsupported framebuffer (see the release notes).
> > > > 
> > > > yes, I missed that, thank you
> > > > 
> > > > > This unfortunately isn't sufficient information and my Blade 1500
> > > > > doesn't exhibit this problem. Could you please put the full dmesg
> > > > > output online somewhere (this should be easy to capture when
> > > > > using a serial console)? Can you boot to single user?
> > > > 
> > > > a multi-user boot dmesg:
> > > > 	http://seis.bris.ac.uk/~mexas/dmesg-blade1500-8.0
> > > > 
> > > > a single user boot went ok, but some disk problems:
> > > > 	http://seis.bris.ac.uk/~mexas/dmesg-blade1500-8.0-single-user
> > > 
> > > just to add on that. After fixing disk with fsck, boot into single
> > > user mode is fine, and no further disk problems reported.
> > > But multi-user mode still panics, just as before.
> > > 
> > 
> > Hrm, the only noticeable difference between your Blade 1500
> > and mine is that on yours the generic vgapci(4) attaches to
> > the otherwise unsupported framebuffer. I gave such a setup
> > a try but that didn't trigger a panic (this actually wasn't
> > likely to). Could you please build and install a debugging
> > kernel in single user mode (a `mount -a` should re-mount /
> > read-write), i.e. a kernel with the following options:
> > makeoptions DEBUG=-g
> > options KDB
> > options DDB
> > If that kernel panics you should end up in the debugger
> > prompt. Please issue a `bt` there and provide its output.
> 
> done:
> 
> 
> Trying to mount root from ufs:/dev/ad0a
> Setting hostuuid: .
> Setting hostid: 0xd41d8cd9.
> Entropy harvesting: interrupts ethernet point_to_pointpanic: trap: memory addres
> s not aligned
> cpuid = 0
> KDB: enter: panic
> [thread pid 50 tid 100062 ]
> Stopped at      kdb_enter+0x80: ta              %xcc, 1
> db>
> db> bt
> Tracing pid 50 tid 100062 td 0xfffff80001532380
> panic() at panic+0x20c
> trap() at trap+0x4d0
> -- memory address not aligned sfar=0xfddfa187 sfsr=0x40029 %o7=0xc06d6ed8 --
> stack_capture() at stack_capture+0x114
> stack_save_td() at stack_save_td+0x60
> sysctl_kern_proc_kstack() at sysctl_kern_proc_kstack+0x330
> sysctl_root() at sysctl_root+0x1c4
> userland_sysctl() at userland_sysctl+0x208
> __sysctl() at __sysctl+0x70
> syscall() at syscall+0x2e8
> -- syscall (202, FreeBSD ELF64, __sysctl) %o7=0x1016b0 --
> userland() at 0x4046bac8
> user trace: trap %o7=0x1016b0
> pc 0x4046bac8, sp 0x7fdffffd001
> pc 0x101fb8, sp 0x7fdffffd961
> pc 0x102ba4, sp 0x7fdffffdae1
> pc 0x100f70, sp 0x7fdffffe451
> pc 0x40208d94, sp 0x7fdffffe511
> done
> db>
> 
> Is this any help?
> 

Thanks, I think I have fixed this in r200272.

Marius




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20091208202016.GM6562>