Re: Warning: pnp 00:0b: can't evaluate _CRS: 12311

[Sedat Dilek]

On Tue, Apr 24, 2012 at 12:04 AM, Bjorn Helgaas <bhelgaas@xxxxxxxxxx> wrote:
> On Mon, Apr 23, 2012 at 6:03 AM, Sedat Dilek <sedat.dilek@xxxxxxxxxxxxxx> wrote:
>> On Mon, Apr 23, 2012 at 11:48 AM, Sedat Dilek
>> <sedat.dilek@xxxxxxxxxxxxxx> wrote:
>>> [ Please CC me I am not subscribed to this ML ]
>>>
>>> Hi,
>>>
>>> with a self-compiled linux-3.4-rc4 I am seeing this bugs/failures/warnings:
>>>
>>> [ Â Â0.241572] [Firmware Bug]: ACPI: BIOS _OSI(Linux) query ignored
>>>
>>> [ Â Â0.279385] Âpci0000:00: ACPI _OSC request failed (AE_ERROR),
>>> returned control mask: 0x1d
>>>
>>> [ Â Â0.293699] system 00:0a: Plug and Play ACPI device, IDs PNP0c02 (active)
>>> [ Â Â0.294032] ACPI Error: Invalid/unsupported resource descriptor:
>>> Type 0x00 (20120320/utresrc-650)
>>> [ Â Â0.294037] pnp 00:0b: can't evaluate _CRS: 12311
>>>
>>> The same outputs I have seen also with an unsupported drm-intel-next
>>> kernel from Ubuntu/PPA archives [1].
>>>
>>> What does ist mean? Is it harmless? Buggy firmware/bios?
>>>
>>> Do you need more background infos (dmesg, kernel-config, etc.)?
>>>
>>> Regards,
>>> - Sedat -
>>>
>>> [1] http://kernel.ubuntu.com/~kernel-ppa/mainline/drm-intel-next/current/
>>
>> I have upgraded this Samsung series-5 notebook from BIOS version 05XK
>> to 06XK, but the pnp/_CRS warning still remains.
>
> Resource type 0 is "reserved" per the ACPI spec, so this is likely a
> BIOS defect. ÂHowever, we might be able to handle it better than we
> currently do.
>
> I would expect all versions of Linux would report this error -- can
> you confirm that or identify a version that didn't complain? ÂAlso,
> can you collect an acpidump (see
> http://www.lesswatts.org/projects/acpi/utilities.php)?
>
> In acpi_ut_validate_resource(), we return AE_AML_INVALID_RESOURCE_TYPE
> (0x3017 == 12311), which explains the messages you see. ÂThe comments
> in acpi_ut_walk_aml_resources() suggest that we can't continue because
> the length may be bogus if the type is invalid. ÂBut I suspect that if
> we just treat it as a zero-length descriptor and ignore it, things
> will probably work, and that's likely what Windows does.
>
> Bjorn

[ Removed Adam Belay <abelay@xxxxxxx> (email-address does not exist anymore) ]

Yeah, I asked today on #ubuntu-kernel and people there were very
helpful and gave some good explanations (AE_AML_INVALID_RESOURCE_TYPE
etc.).

The acpidump.log I pastebin-ed today, so here it is :-).

The kernel shipped with Ubuntu/precise does not show the pnp/_CSR
issue, this is linux-image-3.2.0-23-generic (3.2.0-23.36).

Hope this helps.

- Sedat -

Attachment: acpidump.log
Description: Binary data