ASUS A8N-VM unbootable (panics) with ACPI in 6.0-RELEASE]

[mls at quaggaspace.org]

> From: Adriaan de Groot <groot at kde.org>
> To: freebsd-acpi at freebsd.org
> Date: Fri, 30 Dec 2005 11:53:07 +0100
> Subject: Re: ASUS A8N-VM unbootable (panics) with ACPI in 6.0-RELEASE
> On Friday 30 December 2005 05:20, Justin Settle wrote:
>> So my nice shiny new ASUS A8N-VM came in the mail today, and well, the
>> system kernel panics if I try to boot it with ACPI.  This happens in
>> both the i386 and AMD64 6.0-RELEASE installs.  The issue is the same one
>> in this pr: http://www.freebsd.org/cgi/query-pr.cgi?pr=standards/90871
>
> See my recent posts re: A8N-VM CSM with the same issue. The DSDT table is so
> fundamentally broken (but has a correct checksum, so it's not an issue of one
> nibble falling over during manufacturing) that you can't use it on anything
> except Windows (which ignores the AML spec saying that names consist of
> characters A-Z and _). In any case, what I did to fix this was:

Actually I had seen some of your posts prior to me buying the board.  I didn't
see the one that said to stay away from it (doh!).  So now I'm at work reading
up on the CSM board since more people seem to have that one and the problems
are the same.  Lovely fun with brand new hardware.  Anyway I read in a few
places that ASUS has been working off-and-on on a new bios.  Thats my hope
anyway.

> 1) Boot 6.0-R x86 in ACPI disabled mode
> 2) Use acpidump to get the DSDT
> 3) Hex edit the DSDT (you can't disassemlble it because it's so broken)
> 4) Set that modified DSDT for use as per Handbook section 11.rumtiddlypo
> 5) Reboot into 6.0-R x86 with ACPI
> 6) Profit!
>
> It seems to be a speedy and stable machine from there on in, at least when
> doing only one thing (portinstall kde) at a time.
>
> If you like, I can post the modified DSDT on a website somewhere.

Would it be possible for you to post it somewhere just to see what you 
did? Hexediting dstl tables would be a new thing for me :).

>> The system then panics, waits 15 seconds and reboots.  If I boot the
>> system without ACPI, it boots up, however my ata controller drops to
>> ata-33.
>
> Also with the fixed DSDT, ata-33. nForce430 is still too leading-edge. Do you
> get the nForce430 NIC listed at all? Mine doesn't even show up in pciconf
> -lv.

I actually didn't try the nve at all.  Having read previous messages I 
knew that
would be an uphill battle.  The machine is working nicely with an xl plugged
into its PCI slot.  The ATA is a bummer.  I may try and got with a gentoo
install and see if it'll work there.  My understanding there is that nvidia's
nforce drivers will allow the ata to work at full speed.  However, the linux
folks also have to hack up the kernel to get around the buggy asus acpi
character.

>> please let me know.  I'm sure if this is fixable on the driver end or if
>> I need to bug ASUS to fix it in the bios but this seemed like a good
>> place to start.
>
> I figured that the first BIOS update (to 0506) would fix the problem, but no.
> What's worse, it also made the machine lock up in x86 no-ACPI mode.

Lovely.  I think when I get home this evening I'm going to tinker with the
machine some more and while I'm at it write an email to asus support.  Rumors
of a new bios have been around for a bit but I want add my voice to the number
of people who have asked to get this resolved.

Thanks for your help on this,
-Justin