acpi=off vs. blacklist

acpi=off vs. blacklist

[Ph. Marek]

Hello everybody!

I have two machines which freeze during boot; this seems to be ACPI-related, 
because with acpi=off they start.

The first one is the HP DC 7700; some summary can be seen on
  http://forums1.itrc.hp.com/service/forums/questionanswer.do?threadId=1081719
It hangs after a line
  ACPI assume root bridge[\sp_.pci0] bus is 0

The second one is a notebook; a Roda Rocky III+ RK886EX, which just shows a 
blank screen.

As we drive many machines from the same kernel image via PXE, I tried to 
insert them into drivers/acpi/blacklist.c; but that doesn't help, either - 
the notebook hangs as before, no matter whether I set is_critical_error or 
not. 
 [ I suppose I wrote the entries wrong; is there some HOWTO? I used
   acpidump/acpixtract/acpitbl, copied other blacklist entries, and took
   "OEM Table ID", "OEMID", "OEM REVISION". Help? ]


Using the "ACPI debug" kernel option showed no more messages.


Help, please? What can/should I do now?


Regards,

Phil

Re: acpi=off vs. blacklist

[Gabriel C]

Ph. Marek wrote:
[ added linux-acpi to CC ]
> Hello everybody!
> 
> I have two machines which freeze during boot; this seems to be ACPI-related, 
> because with acpi=off they start.
> 
> The first one is the HP DC 7700; some summary can be seen on
>   http://forums1.itrc.hp.com/service/forums/questionanswer.do?threadId=1081719
> It hangs after a line
>   ACPI assume root bridge[\sp_.pci0] bus is 0
> 
> The second one is a notebook; a Roda Rocky III+ RK886EX, which just shows a 
> blank screen.
> 
> As we drive many machines from the same kernel image via PXE, I tried to 
> insert them into drivers/acpi/blacklist.c; but that doesn't help, either - 
> the notebook hangs as before, no matter whether I set is_critical_error or 
> not. 
>  [ I suppose I wrote the entries wrong; is there some HOWTO? I used
>    acpidump/acpixtract/acpitbl, copied other blacklist entries, and took
>    "OEM Table ID", "OEMID", "OEM REVISION". Help? ]
> 
> 
> Using the "ACPI debug" kernel option showed no more messages.
> 
> 
> Help, please? What can/should I do now?
> 
> 
> Regards,
> 
> Phil
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/
> 

Re: acpi=off vs. blacklist

[Ph. Marek]

On Dienstag, 31. Juli 2007, Gabriel C wrote:
> [ added linux-acpi to CC ]

Sorry, forgot to mention:
I'd like to use 2.6.20.1, but I already tested with 2.6.22.1.

IIUC if there's a problem with the ACPI tables, the blacklist won't help, 
because to know whether the ACPI blacklist applies ACPI tables must be 
parsed :-/ ?


Regards,

Phil

Re: acpi=off vs. blacklist (HP DC 7700)

[Len Brown]

On Tuesday 31 July 2007 06:35, Ph. Marek wrote:
> Hello everybody!
> 
> I have two machines which freeze during boot; this seems to be ACPI-related, 
> because with acpi=off they start.
> 
> The first one is the HP DC 7700; some summary can be seen on
>   http://forums1.itrc.hp.com/service/forums/questionanswer.do?threadId=1081719
> It hangs after a line
>   ACPI assume root bridge[\sp_.pci0] bus is 0

Looks like this box has been broken since December, 2006!
Not booting unless "acpi=off" is a serious failure.

> The second one is a notebook; a Roda Rocky III+ RK886EX, which just shows a 
> blank screen.

Please get Linux up and running on the two boxes
using whatever means are at your disposal (some mentioned
in the thread above are acpi=off, pci=noacpi,
pci=conf1, noapic hda=noprobe hdc=noprobe)

Download the latest stable kernel.org kernel (2.6.22.1 today)
Make sure that the latest kernel still fails unless
you apply the workaround(s).

Then open two a bug report for each machine here:
http://bugzilla.kernel.org/enter_bug.cgi?product=ACPI

Attach the output from acpidump, as well as the
complete dmesg and lspci -vv output,
and paste the contents of /proc/interrupts.
Then, if you can capture the console output taken from the actual failure
case, that would be ideal

> As we drive many machines from the same kernel image via PXE, I tried to 
> insert them into drivers/acpi/blacklist.c; but that doesn't help, either - 
> the notebook hangs as before, no matter whether I set is_critical_error or 
> not. 
>  [ I suppose I wrote the entries wrong; is there some HOWTO? I used
>    acpidump/acpixtract/acpitbl, copied other blacklist entries, and took
>    "OEM Table ID", "OEMID", "OEM REVISION". Help? ]

Lets see if we can find out what the problem is before we
blacklist the machines.

thanks,
-Len

Re: acpi=off vs. blacklist (HP DC 7700)

[Ph. Marek]

On Dienstag, 31. Juli 2007, Len Brown wrote:
> Please get Linux up and running on the two boxes
> using whatever means are at your disposal 
...
> Then open two a bug report for each machine here: 
> http://bugzilla.kernel.org/enter_bug.cgi?product=ACPI
Done, see
  http://bugzilla.kernel.org/show_bug.cgi?id=8833
for the HP DC 7700 and
  http://bugzilla.kernel.org/show_bug.cgi?id=8834
for the Rocky notebook.


> Lets see if we can find out what the problem is before we
> blacklist the machines.
Ok, suits me fine.


Regards,

Phil