Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 21 Jul 2008 14:17:29 -0700
From:      Jo Rhett <hostmaster@netconsonance.com>
To:        Peter Wemm <peter@wemm.org>
Cc:        FreeBSD Stable <freebsd-stable@freebsd.org>
Subject:   chipset causing locks.
Message-ID:  <C73C133D-8CE4-4F67-999E-E6F925E719CA@netconsonance.com>
In-Reply-To: <e7db6d980807182329of3cb3a6xd2a83ca928be1c2e@mail.gmail.com>
References:  <C278655C-4FFB-4A8E-9501-2B84283E324D@netconsonance.com> <20080711164939.GA10238@lava.net> <C1103C22-6A2B-4E08-9AAF-EA95DF088647@netconsonance.com> <e7db6d980807182329of3cb3a6xd2a83ca928be1c2e@mail.gmail.com>

next in thread | previous in thread | raw e-mail | index | archive | help
Thanks for the note.  No, just a coincidence.   The chipset is a VIA  
ProSavageDDR KM266.

But thanks for bringing that up ;-)

FWIW, as others have speculated enabling more logging from GEOM  
produced nothing.  It does appear to be a hardware failure of some sort.

On Jul 18, 2008, at 11:29 PM, Peter Wemm wrote:
> On Wed, Jul 16, 2008 at 2:42 PM, Jo Rhett <hostmaster@netconsonance.com 
> > wrote:
>>> On Fri, Jul 11, 2008 at 12:59:33AM -0700, Jo Rhett wrote:
>>>>
>>>> Every time it is rebuilding ad0.   Every single boot in the last  
>>>> two
>>>> weeks.
>>
>> On Jul 11, 2008, at 9:49 AM, Clifton Royston wrote:
>>>
>>> That just means that it halted without a proper shutdown.  If it
>>> crashes, the mirror isn't stopped properly, so it's marked dirty,  
>>> so it
>>> must rebuild it.  It is the precise analogy of finding all the file
>>> systems dirty on boot and fscking them, following a crash.
>>
>>
>> Thanks for the clarification.  Dang, I hoped I was on to something.
>
> This is really off on a tangent, but I thought I'd mention it on the
> off-chance that it fit your problem.
>
> Recently there have been grumblings about heat problems with certain
> nvidia chipsets on consumer boards.  Apparently, there is some process
> issue, if you believe trade rags like theinquirer.net etc.  Apparently
> there is some issue with heat damage over time.  Consumer motherboards
> with passive cooled (no fan) heat pipes etc seem to be particularly
> vulnerable.  I use the word "apparently" because it is far from a
> verified fact.
>
> However, I've got two motherboards, one running freebsd, one running
> windows, with nvidia chipsets.  Both used to be fine with onboard IDE
> activity.  Both now use raid controllers so the IDE interfaces have
> been idle for a good year or so.
>
> Something came up and I had to use the IDE interfaces for a lot of
> data transfer.  Suddenly, both machines are flakey.  The windows
> machine blue screens under load.  My freebsd box just "turns off"
> (motherboard appears to power off, but the power supply is on still).
> The same happens when I use a linux boot disk, so I know its not
> FreeBSD's fault.
>
> The common factor seems to be that the motherboards are now about a
> year and a half old.  They both have the same nvidia south bridge that
> theinquirer.net was trashing.   Both used to work fine, now have
> problems with IDE.  and now I recalled the article and started
> wondering...
>
> Do you, by any wildly remote chance, have an nvidia based motherboard?
>
> I believe the fault I'm seeing is the system asserting a fatal error
> by doing a HT ECC flood to halt everything.
>
> -- 
> Peter Wemm - peter@wemm.org; peter@FreeBSD.org; peter@yahoo-inc.com;  
> KI6FJV
> "All of this is for nothing if we don't go to the stars" - JMS/B5
> "If Java had true garbage collection, most programs would delete
> themselves upon execution." -- Robert Sewell

-- 
Jo Rhett
Net Consonance : consonant endings by net philanthropy, open source  
and other randomness





Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?C73C133D-8CE4-4F67-999E-E6F925E719CA>