Skip site navigation (1)Skip section navigation (2)
Date:      Sat, 5 Dec 2009 01:34:57 +0100 (CET)
From:      Gheorghe Ardelean <ardelean@ww.uni-erlangen.de>
To:        Anton Shterenlikht <mexas@bristol.ac.uk>
Cc:        freebsd-sparc64@freebsd.org
Subject:   Re: trouble installing to sun blade 1500
Message-ID:  <alpine.LNX.2.00.0912050130150.10606@servww6.ww.uni-erlangen.de>
In-Reply-To: <20091205001122.GA54581@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> <20091203141711.GA43641@mech-cluster241.men.bris.ac.uk> <20091205001122.GA54581@mech-cluster241.men.bris.ac.uk>

next in thread | previous in thread | raw e-mail | index | archive | help
On Sat, 5 Dec 2009, Anton Shterenlikht wrote:

> On Thu, Dec 03, 2009 at 02:17:11PM +0000, Anton Shterenlikht wrote:
>> 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?
>>
>> 7.2 seems to be fine:
>>
>> http://seis.bris.ac.uk/~mexas/freebsd/
>>
>> I'll try to go from 7.2 up, perhaps straight to current,
>> will build a debugging kernel and report.
>>
>> One question: now that I installed fbsd, I can connect
>> keyboard, mouse and screen and use it as a desktop, can't I.
>> The serial line limitation is just for the installation
>> program isn't it?
>
> just to report that 8.0 installs fine on a Silver Blade 1500.
> See dmesg at: http://seis.bris.ac.uk/~mexas/freebsd/

Anton,

if you want to have your dmesgs added to our wiki,
I think Pietro can make it happen (I see we don't have any from a 
blade1500)! Here is the current list:
http://wiki.freebsd.org/FreeBSD/sparc64/dmesgs

Regards,

Gheorghe.




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?alpine.LNX.2.00.0912050130150.10606>