acpi

classic Classic list List threaded Threaded
7 messages Options
Reply | Threaded
Open this post in threaded view
|

acpi

Marco Peereboom-3
Jordan Hargrave (your friendly IPMI developer) has been plugging away at ACPI.
He has written an ASL parser and an AML interpreter.  Jordan also wrote a
userland tool that dumps the ACPI tables and all kinds of other things.  We can
then replay these dumps to test the ASL parser and AML interpreter.  In order
to be able to test the AML interpreter we need as many ACPI dumps as possible.
This is where *you* come in...

You can find the binary and source at:
http://www.peereboom.us/acpi.tgz
I added the source code + silly makefile so that you can roll your own
acpiscan.  Simply delete the binary I provided (compiled on 3.8-current) and
run "make".

As root run the following command:
./acpiscan -save descriptive_name_of_box

example: ./acpiscan -save dell_pe1850

This will generate 3 files:
dell_pe1850.DSDT
dell_pe1850.FACP
dell_pe1850.RSDT

Yes, acpiscan WILL core dump at the end.  Its supposed to!

Tar and compress these files and send them to jordan@ and marco@.

Thanks,
/marco

Reply | Threaded
Open this post in threaded view
|

Re: acpi

Martin Ekendahl
I'm a bit embarrassed to say, but I have a dedicated colo which I
bought, and I don't know the specifics of it (other than generic specs).
I'd would guess it's hand built. Whats a good line to take from the
dmesg to name my files? Maybe the pci interrupts?

pcibios0: PCI Interrupt Router at 000:17:0 ("VIA VT82C596A ISA" rev 0x00)

or something similar?

-Martin


Marco Peereboom wrote:

> Jordan Hargrave (your friendly IPMI developer) has been plugging away at ACPI.
> He has written an ASL parser and an AML interpreter.  Jordan also wrote a
> userland tool that dumps the ACPI tables and all kinds of other things.  We can
> then replay these dumps to test the ASL parser and AML interpreter.  In order
> to be able to test the AML interpreter we need as many ACPI dumps as possible.
> This is where *you* come in...
>
> You can find the binary and source at:
> http://www.peereboom.us/acpi.tgz
> I added the source code + silly makefile so that you can roll your own
> acpiscan.  Simply delete the binary I provided (compiled on 3.8-current) and
> run "make".
>
> As root run the following command:
> ./acpiscan -save descriptive_name_of_box
>
> example: ./acpiscan -save dell_pe1850
>
> This will generate 3 files:
> dell_pe1850.DSDT
> dell_pe1850.FACP
> dell_pe1850.RSDT
>
> Yes, acpiscan WILL core dump at the end.  Its supposed to!
>
> Tar and compress these files and send them to jordan@ and marco@.
>
> Thanks,
> /marco

Reply | Threaded
Open this post in threaded view
|

Re: acpi

David Gwynne
On 09/11/2005, at 3:25 PM, Martin Ekendahl wrote:

> I'm a bit embarrassed to say, but I have a dedicated colo which I  
> bought, and I don't know the specifics of it (other than generic  
> specs). I'd would guess it's hand built. Whats a good line to take  
> from the dmesg to name my files? Maybe the pci interrupts?

call it unknown_via or similar and include the dmesg as another file  
in the tarball. thats probably a good thing to do anyway, regardless  
of whether you know what type of machine it is or not.

dlg

Reply | Threaded
Open this post in threaded view
|

Re: acpi

Jordan Hargrave
In reply to this post by Marco Peereboom-3
Thanks for all the ACPI dumps everyone!  I think I have enough of a sample
set now, no more, please! :)
Hmm.. the acpiscan should not be core dumping; there maybe an invalid
address in the ACPI table?

  --jordan hargrave

>From: Marco Peereboom <[hidden email]>
>To: [hidden email]
>CC: [hidden email], [hidden email]
>Subject: acpi
>Date: Tue, 8 Nov 2005 22:54:47 -0600
>MIME-Version: 1.0
>Received: from cvs.openbsd.org ([199.185.137.3]) by mc9-f17.hotmail.com
>with Microsoft SMTPSVC(6.0.3790.211); Tue, 8 Nov 2005 20:54:59 -0800
>Received: from shear.ucar.edu (shear.ucar.edu [192.43.244.163])by
>cvs.openbsd.org (8.13.4/8.12.1) with ESMTP id
>jA94uxhk002386(version=TLSv1/SSLv3 cipher=DHE-DSS-AES256-SHA bits=256
>verify=FAIL);Tue, 8 Nov 2005 21:57:00 -0700 (MST)
>Received: from dev.peereboom.us (adsl-67-64-89-177.dsl.austtx.swbell.net
>[67.64.89.177])by shear.ucar.edu (8.13.4/8.13.3) with ESMTP id
>jA94smZY004756(version=TLSv1/SSLv3 cipher=DHE-DSS-AES256-SHA bits=256
>verify=NO);Tue, 8 Nov 2005 21:54:49 -0700 (MST)
>Received: from dev.peereboom.us ([hidden email]
>[127.0.0.1])by dev.peereboom.us (8.13.4/8.13.4) with ESMTP id
>jA94smfi004639;Tue, 8 Nov 2005 22:54:48 -0600 (CST)
>Received: (from marco@localhost)by dev.peereboom.us (8.13.4/8.13.4/Submit)
>id jA94slxo031124;Tue, 8 Nov 2005 22:54:47 -0600 (CST)
>X-Message-Info: 6sSXyD95QpUGy/uYCdX58EJJ8oGL00uqpZIyqW0sOks=
>User-Agent: Mutt/1.5.11
>Return-Path: [hidden email]
>X-OriginalArrivalTime: 09 Nov 2005 04:54:59.0740 (UTC)
>FILETIME=[BCE27DC0:01C5E4E9]
>
>Jordan Hargrave (your friendly IPMI developer) has been plugging away at
>ACPI.
>He has written an ASL parser and an AML interpreter.  Jordan also wrote a
>userland tool that dumps the ACPI tables and all kinds of other things.  We
>can
>then replay these dumps to test the ASL parser and AML interpreter.  In
>order
>to be able to test the AML interpreter we need as many ACPI dumps as
>possible.
>This is where *you* come in...
>
>You can find the binary and source at:
>http://www.peereboom.us/acpi.tgz
>I added the source code + silly makefile so that you can roll your own
>acpiscan.  Simply delete the binary I provided (compiled on 3.8-current)
>and
>run "make".
>
>As root run the following command:
>./acpiscan -save descriptive_name_of_box
>
>example: ./acpiscan -save dell_pe1850
>
>This will generate 3 files:
>dell_pe1850.DSDT
>dell_pe1850.FACP
>dell_pe1850.RSDT
>
>Yes, acpiscan WILL core dump at the end.  Its supposed to!
>
>Tar and compress these files and send them to jordan@ and marco@.
>
>Thanks,
>/marco

Reply | Threaded
Open this post in threaded view
|

Re: acpi

Marco Peereboom
In reply to this post by Marco Peereboom-3
Alright we have enough dumps for now so please stop eating all my bandwidth :-)

On Tue, Nov 08, 2005 at 10:54:47PM -0600, Marco Peereboom wrote:

> Jordan Hargrave (your friendly IPMI developer) has been plugging away at ACPI.
> He has written an ASL parser and an AML interpreter.  Jordan also wrote a
> userland tool that dumps the ACPI tables and all kinds of other things.  We can
> then replay these dumps to test the ASL parser and AML interpreter.  In order
> to be able to test the AML interpreter we need as many ACPI dumps as possible.
> This is where *you* come in...
>
> You can find the binary and source at:
> http://www.peereboom.us/acpi.tgz
> I added the source code + silly makefile so that you can roll your own
> acpiscan.  Simply delete the binary I provided (compiled on 3.8-current) and
> run "make".
>
> As root run the following command:
> ./acpiscan -save descriptive_name_of_box
>
> example: ./acpiscan -save dell_pe1850
>
> This will generate 3 files:
> dell_pe1850.DSDT
> dell_pe1850.FACP
> dell_pe1850.RSDT
>
> Yes, acpiscan WILL core dump at the end.  Its supposed to!
>
> Tar and compress these files and send them to jordan@ and marco@.
>
> Thanks,
> /marco

Reply | Threaded
Open this post in threaded view
|

Re: acpi

Justin Krejci
In reply to this post by Marco Peereboom-3
On Wednesday 09 November 2005 10:52 pm, Jordan Hargrave wrote:
> Thanks for all the ACPI dumps everyone!  I think I have enough of a sample
> set now, no more, please! :)
> Hmm.. the acpiscan should not be core dumping; there maybe an invalid
> address in the ACPI table?
>
>   --jordan hargrave
>

It did not core dump on me.

Reply | Threaded
Open this post in threaded view
|

Re: acpi

Andrew Dalgleish
On Thu, Nov 10, 2005 at 08:24:35AM -0600, Justin Krejci wrote:
> It did not core dump on me.

Same here.