[2.6.33-rc2] ACPI problems on HP nc6400 notebook

[2.6.33-rc2] ACPI problems on HP nc6400 notebook

[Sedat Dilek]

Hi,

I build linux-2.6.33-rc2 on a Debian/sid 32-bit host.
Unfortunately, this kernel falls in an unusable state (sort of loop)
and does not a normal startup.

With this setting (see below), I hoped to catch the early
boot-messages (but, no success):

# cat /etc/default/bootlogd
# Run bootlogd at startup ?
BOOTLOGD_ENABLE=Yes

Finally, I could take a picture [1].
It seems to be that latest ACPI patches are breaking my system (speculating).

Q: How can I log such early boot (error) messages to give a detailled log?

Hope this helps fixing the problem.

Kind Regards,
- Sedat -


[1] http://files.iniza.org/linux-kernel/DSCN2134.JPG

Re: [2.6.33-rc2] ACPI problems on HP nc6400 notebook

[Justin P. Mattock]

On 12/24/09 17:11, Sedat Dilek wrote:
> Hi,
>
> I build linux-2.6.33-rc2 on a Debian/sid 32-bit host.
> Unfortunately, this kernel falls in an unusable state (sort of loop)
> and does not a normal startup.
>
> With this setting (see below), I hoped to catch the early
> boot-messages (but, no success):
>
> # cat /etc/default/bootlogd
> # Run bootlogd at startup ?
> BOOTLOGD_ENABLE=Yes
>
> Finally, I could take a picture [1].
> It seems to be that latest ACPI patches are breaking my system (speculating).
>
> Q: How can I log such early boot (error) messages to give a detailled log?
>
> Hope this helps fixing the problem.
>
> Kind Regards,
> - Sedat -
>
>
> [1] http://files.iniza.org/linux-kernel/DSCN2134.JPG
> --
> 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/
>

you can try(if you have the port)
ohci1394_dma=early or the usb
early logging located in kernel hacking.

but if this thing is hitting way early it
might be a bit more tricky.

if you can try  a bisect from the bad
kernel to the good.

Justin P. Mattock

Re: [2.6.33-rc2] ACPI problems on HP nc6400 notebook

[Sedat Dilek]

Hi,

as I suspected it is an ACPI issue while I tried to boot with "acpi=off".

I have taken new screenshots and added dmesg (acpi=off) and all
messages in /var/log/* [1].

It is possible to load acpi kernel-module afterwards?
Or it is not possible whenn loading with acpi=off?

@Len Brown:
There are two photos showing thermal problems.
Notebook did shut down in the night.
But this was before 2.6.33-rc2.

I will not try a bisect (visting currently my brother and working with
touchpad really sucks), but I will revert the wmi-related (wmi/hp-wmi)
commits.
There seems problems with wmi on other systems w/ 2.6.33-rc2 [2].

Kind Regards,
- Sedat -

[1] http://files.iniza.org/linux-kernel/BUG_2.6.33-rc2/
[2] http://marc.info/?l=linux-kernel&m=126171951030608&w=2

On Fri, Dec 25, 2009 at 2:22 AM, Justin P. Mattock
<justinmattock@gmail.com> wrote:
> On 12/24/09 17:11, Sedat Dilek wrote:
>>
>> Hi,
>>
>> I build linux-2.6.33-rc2 on a Debian/sid 32-bit host.
>> Unfortunately, this kernel falls in an unusable state (sort of loop)
>> and does not a normal startup.
>>
>> With this setting (see below), I hoped to catch the early
>> boot-messages (but, no success):
>>
>> # cat /etc/default/bootlogd
>> # Run bootlogd at startup ?
>> BOOTLOGD_ENABLE=Yes
>>
>> Finally, I could take a picture [1].
>> It seems to be that latest ACPI patches are breaking my system
>> (speculating).
>>
>> Q: How can I log such early boot (error) messages to give a detailled log?
>>
>> Hope this helps fixing the problem.
>>
>> Kind Regards,
>> - Sedat -
>>
>>
>> [1] http://files.iniza.org/linux-kernel/DSCN2134.JPG
>> --
>> 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/
>>
>
> you can try(if you have the port)
> ohci1394_dma=early or the usb
> early logging located in kernel hacking.
>
> but if this thing is hitting way early it
> might be a bit more tricky.
>
> if you can try  a bisect from the bad
> kernel to the good.
>
> Justin P. Mattock
>

Re: [2.6.33-rc2] ACPI problems on HP nc6400 notebook

[Sedat Dilek]

[CCed linux-acpi ML, maciej.rutecki@gmail.com and landgraf@ru.ru]
(see [2] and [3])

GOTCHA!

I reverted patches introduced in:

commit 2969cd482ebe9c5b82d55917b0cd41175eff042a:
"Merge branch 'misc-2.6.33' into release"

...and now I could boot into 2.6.33-rc2.

Here the revert-acpi-patches:

$ ls debian/patches/revert-acpi-patches/
0001-Revert-hp-wmi-Fix-two-memleaks.patch
0002-Revert-acer-wmi-msi-wmi-Remove-needless-DMI-MODULE_A.patch
0003-Revert-dell-wmi-do-not-keep-driver-loaded-on-unsuppo.patch
0004-Revert-wmi-Free-the-allocated-acpi-objects-through-w.patch
0005-Revert-drivers-platform-x86-acerhdf.c-check-BIOS-inf.patch
0006-Revert-acerhdf-add-new-BIOS-versions.patch
0007-Revert-acerhdf-limit-modalias-matching-to-supported.patch
0008-Revert-toshiba_acpi-convert-to-seq_file.patch
0009-Revert-asus_acpi-convert-to-seq_file.patch
0010-Revert-ACPI-do-not-select-ACPI_DOCK-from-ATA_ACPI.patch
0011-Revert-MAINTAINERS-add-maintainer-for-msi-wmi-driver.patch
0012-Revert-fujitu-laptop-fix-tests-of-acpi_evaluate_inte.patch
0013-Revert-arch-x86-kernel-cpu-cpufreq-acpi-cpufreq.c-av.patch

In [1] you can find revert-acpi-patchset, dmesg & lspci outputs of my
new kernel.

Marciej and Werner can you please verify or dig deeper by partially
excluding the single reverts?
Thank you.

Kind Regards,
- Sedat -

[1] http://files.iniza.org/linux-kernel/BUG_2.6.33-rc2/
[2] http://marc.info/?l=linux-kernel&m=126171951030608&w=2
[3] http://marc.info/?l=linux-kernel&m=126173586309724&w=2

On Fri, Dec 25, 2009 at 9:48 AM, Sedat Dilek <sedat.dilek@googlemail.com> wrote:
> Hi,
>
> as I suspected it is an ACPI issue while I tried to boot with "acpi=off".
>
> I have taken new screenshots and added dmesg (acpi=off) and all
> messages in /var/log/* [1].
>
> It is possible to load acpi kernel-module afterwards?
> Or it is not possible whenn loading with acpi=off?
>
> @Len Brown:
> There are two photos showing thermal problems.
> Notebook did shut down in the night.
> But this was before 2.6.33-rc2.
>
> I will not try a bisect (visting currently my brother and working with
> touchpad really sucks), but I will revert the wmi-related (wmi/hp-wmi)
> commits.
> There seems problems with wmi on other systems w/ 2.6.33-rc2 [2].
>
> Kind Regards,
> - Sedat -
>
> [1] http://files.iniza.org/linux-kernel/BUG_2.6.33-rc2/
> [2] http://marc.info/?l=linux-kernel&m=126171951030608&w=2
>
> On Fri, Dec 25, 2009 at 2:22 AM, Justin P. Mattock
> <justinmattock@gmail.com> wrote:
>> On 12/24/09 17:11, Sedat Dilek wrote:
>>>
>>> Hi,
>>>
>>> I build linux-2.6.33-rc2 on a Debian/sid 32-bit host.
>>> Unfortunately, this kernel falls in an unusable state (sort of loop)
>>> and does not a normal startup.
>>>
>>> With this setting (see below), I hoped to catch the early
>>> boot-messages (but, no success):
>>>
>>> # cat /etc/default/bootlogd
>>> # Run bootlogd at startup ?
>>> BOOTLOGD_ENABLE=Yes
>>>
>>> Finally, I could take a picture [1].
>>> It seems to be that latest ACPI patches are breaking my system
>>> (speculating).
>>>
>>> Q: How can I log such early boot (error) messages to give a detailled log?
>>>
>>> Hope this helps fixing the problem.
>>>
>>> Kind Regards,
>>> - Sedat -
>>>
>>>
>>> [1] http://files.iniza.org/linux-kernel/DSCN2134.JPG
>>> --
>>> 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/
>>>
>>
>> you can try(if you have the port)
>> ohci1394_dma=early or the usb
>> early logging located in kernel hacking.
>>
>> but if this thing is hitting way early it
>> might be a bit more tricky.
>>
>> if you can try  a bisect from the bad
>> kernel to the good.
>>
>> Justin P. Mattock
>>
>

Re: [2.6.33-rc2] ACPI problems on HP nc6400 notebook

[Sedat Dilek]

[-- Attachment #1: Type: text/plain, Size: 5249 bytes --]

While dell-wmi and hp-wmi is concerned I only applied this patch (see
file attachment):

    0001-Revert-wmi-Free-the-allocated-acpi-objects-through-w.patch

This fixes ACPI/WMI problems onmy HP nc6400 notebook.

----- Please REVERT the following commit in 2.6.33-rc2 -----

commit 3e9b988e4edf065d39c1343937f717319b1c1065
Refs: v2.6.33-rc1-10-g3e9b988
Author:     Anisse Astier <anisse@astier.eu>
AuthorDate: Fri Dec 4 10:10:09 2009 +0100
Commit:     Len Brown <len.brown@intel.com>
CommitDate: Thu Dec 24 00:42:00 2009 -0500

    wmi: Free the allocated acpi objects through wmi_get_event_data

    These function allocate an acpi object by calling wmi_get_event_data, which
    then calls acpi_evaluate_object, and it is not freed afterwards.

    And kernel doc is fixed for parameters of wmi_get_event_data.

    Signed-off-by: Anisse Astier <anisse@astier.eu>
    Acked-by: Randy Dunlap <randy.dunlap@oracle.com>
    Acked-by: Carlos Corbacho <carlos@strangeworlds.co.uk>
    Signed-off-by: Len Brown <len.brown@intel.com>
[...]

- Sedat -

On Fri, Dec 25, 2009 at 12:15 PM, Sedat Dilek
<sedat.dilek@googlemail.com> wrote:
> [CCed linux-acpi ML, maciej.rutecki@gmail.com and landgraf@ru.ru]
> (see [2] and [3])
>
> GOTCHA!
>
> I reverted patches introduced in:
>
> commit 2969cd482ebe9c5b82d55917b0cd41175eff042a:
> "Merge branch 'misc-2.6.33' into release"
>
> ...and now I could boot into 2.6.33-rc2.
>
> Here the revert-acpi-patches:
>
> $ ls debian/patches/revert-acpi-patches/
> 0001-Revert-hp-wmi-Fix-two-memleaks.patch
> 0002-Revert-acer-wmi-msi-wmi-Remove-needless-DMI-MODULE_A.patch
> 0003-Revert-dell-wmi-do-not-keep-driver-loaded-on-unsuppo.patch
> 0004-Revert-wmi-Free-the-allocated-acpi-objects-through-w.patch
> 0005-Revert-drivers-platform-x86-acerhdf.c-check-BIOS-inf.patch
> 0006-Revert-acerhdf-add-new-BIOS-versions.patch
> 0007-Revert-acerhdf-limit-modalias-matching-to-supported.patch
> 0008-Revert-toshiba_acpi-convert-to-seq_file.patch
> 0009-Revert-asus_acpi-convert-to-seq_file.patch
> 0010-Revert-ACPI-do-not-select-ACPI_DOCK-from-ATA_ACPI.patch
> 0011-Revert-MAINTAINERS-add-maintainer-for-msi-wmi-driver.patch
> 0012-Revert-fujitu-laptop-fix-tests-of-acpi_evaluate_inte.patch
> 0013-Revert-arch-x86-kernel-cpu-cpufreq-acpi-cpufreq.c-av.patch
>
> In [1] you can find revert-acpi-patchset, dmesg & lspci outputs of my
> new kernel.
>
> Marciej and Werner can you please verify or dig deeper by partially
> excluding the single reverts?
> Thank you.
>
> Kind Regards,
> - Sedat -
>
> [1] http://files.iniza.org/linux-kernel/BUG_2.6.33-rc2/
> [2] http://marc.info/?l=linux-kernel&m=126171951030608&w=2
> [3] http://marc.info/?l=linux-kernel&m=126173586309724&w=2
>
> On Fri, Dec 25, 2009 at 9:48 AM, Sedat Dilek <sedat.dilek@googlemail.com> wrote:
>> Hi,
>>
>> as I suspected it is an ACPI issue while I tried to boot with "acpi=off".
>>
>> I have taken new screenshots and added dmesg (acpi=off) and all
>> messages in /var/log/* [1].
>>
>> It is possible to load acpi kernel-module afterwards?
>> Or it is not possible whenn loading with acpi=off?
>>
>> @Len Brown:
>> There are two photos showing thermal problems.
>> Notebook did shut down in the night.
>> But this was before 2.6.33-rc2.
>>
>> I will not try a bisect (visting currently my brother and working with
>> touchpad really sucks), but I will revert the wmi-related (wmi/hp-wmi)
>> commits.
>> There seems problems with wmi on other systems w/ 2.6.33-rc2 [2].
>>
>> Kind Regards,
>> - Sedat -
>>
>> [1] http://files.iniza.org/linux-kernel/BUG_2.6.33-rc2/
>> [2] http://marc.info/?l=linux-kernel&m=126171951030608&w=2
>>
>> On Fri, Dec 25, 2009 at 2:22 AM, Justin P. Mattock
>> <justinmattock@gmail.com> wrote:
>>> On 12/24/09 17:11, Sedat Dilek wrote:
>>>>
>>>> Hi,
>>>>
>>>> I build linux-2.6.33-rc2 on a Debian/sid 32-bit host.
>>>> Unfortunately, this kernel falls in an unusable state (sort of loop)
>>>> and does not a normal startup.
>>>>
>>>> With this setting (see below), I hoped to catch the early
>>>> boot-messages (but, no success):
>>>>
>>>> # cat /etc/default/bootlogd
>>>> # Run bootlogd at startup ?
>>>> BOOTLOGD_ENABLE=Yes
>>>>
>>>> Finally, I could take a picture [1].
>>>> It seems to be that latest ACPI patches are breaking my system
>>>> (speculating).
>>>>
>>>> Q: How can I log such early boot (error) messages to give a detailled log?
>>>>
>>>> Hope this helps fixing the problem.
>>>>
>>>> Kind Regards,
>>>> - Sedat -
>>>>
>>>>
>>>> [1] http://files.iniza.org/linux-kernel/DSCN2134.JPG
>>>> --
>>>> 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/
>>>>
>>>
>>> you can try(if you have the port)
>>> ohci1394_dma=early or the usb
>>> early logging located in kernel hacking.
>>>
>>> but if this thing is hitting way early it
>>> might be a bit more tricky.
>>>
>>> if you can try  a bisect from the bad
>>> kernel to the good.
>>>
>>> Justin P. Mattock
>>>
>>
>

[-- Attachment #2: 0001-Revert-wmi-Free-the-allocated-acpi-objects-through-w.patch --]
[-- Type: text/x-diff, Size: 1761 bytes --]

From f2366a0b009de3fc7be9967e9e65a12b666b05c1 Mon Sep 17 00:00:00 2001
From: Sedat Dilek <sedat.dilek@gmail.com>
Date: Fri, 25 Dec 2009 09:54:31 +0100
Subject: [PATCH] Revert "wmi: Free the allocated acpi objects through wmi_get_event_data"

This reverts commit 3e9b988e4edf065d39c1343937f717319b1c1065.
---
 drivers/platform/x86/dell-wmi.c |    1 -
 drivers/platform/x86/hp-wmi.c   |    2 --
 drivers/platform/x86/wmi.c      |    4 ++--
 3 files changed, 2 insertions(+), 5 deletions(-)

diff --git a/drivers/platform/x86/dell-wmi.c b/drivers/platform/x86/dell-wmi.c
index 6561dfc..67f3fe7 100644
--- a/drivers/platform/x86/dell-wmi.c
+++ b/drivers/platform/x86/dell-wmi.c
@@ -238,7 +238,6 @@ static void dell_wmi_notify(u32 value, void *context)
 			input_sync(dell_wmi_input_dev);
 		}
 	}
-	kfree(obj);
 }
 
 
diff --git a/drivers/platform/x86/hp-wmi.c b/drivers/platform/x86/hp-wmi.c
index db10c5d..63c3e65 100644
--- a/drivers/platform/x86/hp-wmi.c
+++ b/drivers/platform/x86/hp-wmi.c
@@ -381,8 +381,6 @@ static void hp_wmi_notify(u32 value, void *context)
 	} else
 		printk(KERN_INFO "HP WMI: Unknown key pressed - %x\n",
 			eventcode);
-
-	kfree(obj);
 }
 
 static int __init hp_wmi_input_setup(void)
diff --git a/drivers/platform/x86/wmi.c b/drivers/platform/x86/wmi.c
index 9f93d6c..e425a86 100644
--- a/drivers/platform/x86/wmi.c
+++ b/drivers/platform/x86/wmi.c
@@ -540,8 +540,8 @@ EXPORT_SYMBOL_GPL(wmi_remove_notify_handler);
 /**
  * wmi_get_event_data - Get WMI data associated with an event
  *
- * @event: Event to find
- * @out: Buffer to hold event data. out->pointer should be freed with kfree()
+ * @event - Event to find
+ * &out - Buffer to hold event data
  *
  * Returns extra data associated with an event in WMI.
  */
-- 
1.6.5.7

Re: [2.6.33-rc2] ACPI problems on HP nc6400 notebook

[Maciej Rutecki]

2009/12/25 Sedat Dilek <sedat.dilek@googlemail.com>:
> While dell-wmi and hp-wmi is concerned I only applied this patch (see
> file attachment):
>
>    0001-Revert-wmi-Free-the-allocated-acpi-objects-through-w.patch
>
> This fixes ACPI/WMI problems onmy HP nc6400 notebook.
>
> ----- Please REVERT the following commit in 2.6.33-rc2 -----
>
> commit 3e9b988e4edf065d39c1343937f717319b1c1065
> Refs: v2.6.33-rc1-10-g3e9b988
> Author:     Anisse Astier <anisse@astier.eu>
> AuthorDate: Fri Dec 4 10:10:09 2009 +0100
> Commit:     Len Brown <len.brown@intel.com>
> CommitDate: Thu Dec 24 00:42:00 2009 -0500
>
>    wmi: Free the allocated acpi objects through wmi_get_event_data
>
>    These function allocate an acpi object by calling wmi_get_event_data, which
>    then calls acpi_evaluate_object, and it is not freed afterwards.
>
>    And kernel doc is fixed for parameters of wmi_get_event_data.
>
>    Signed-off-by: Anisse Astier <anisse@astier.eu>
>    Acked-by: Randy Dunlap <randy.dunlap@oracle.com>
>    Acked-by: Carlos Corbacho <carlos@strangeworlds.co.uk>
>    Signed-off-by: Len Brown <len.brown@intel.com>

Tested-by Maciej Rutecki <maciej.rutecki@gmail.com>

> [...]
>
> - Sedat -
>
> On Fri, Dec 25, 2009 at 12:15 PM, Sedat Dilek
> <sedat.dilek@googlemail.com> wrote:
>> [CCed linux-acpi ML, maciej.rutecki@gmail.com and landgraf@ru.ru]
>> (see [2] and [3])
>>
>> GOTCHA!
>>
>> I reverted patches introduced in:
>>
>> commit 2969cd482ebe9c5b82d55917b0cd41175eff042a:
>> "Merge branch 'misc-2.6.33' into release"
>>
>> ...and now I could boot into 2.6.33-rc2.
>>
>> Here the revert-acpi-patches:
>>
>> $ ls debian/patches/revert-acpi-patches/
>> 0001-Revert-hp-wmi-Fix-two-memleaks.patch
>> 0002-Revert-acer-wmi-msi-wmi-Remove-needless-DMI-MODULE_A.patch
>> 0003-Revert-dell-wmi-do-not-keep-driver-loaded-on-unsuppo.patch
>> 0004-Revert-wmi-Free-the-allocated-acpi-objects-through-w.patch
>> 0005-Revert-drivers-platform-x86-acerhdf.c-check-BIOS-inf.patch
>> 0006-Revert-acerhdf-add-new-BIOS-versions.patch
>> 0007-Revert-acerhdf-limit-modalias-matching-to-supported.patch
>> 0008-Revert-toshiba_acpi-convert-to-seq_file.patch
>> 0009-Revert-asus_acpi-convert-to-seq_file.patch
>> 0010-Revert-ACPI-do-not-select-ACPI_DOCK-from-ATA_ACPI.patch
>> 0011-Revert-MAINTAINERS-add-maintainer-for-msi-wmi-driver.patch
>> 0012-Revert-fujitu-laptop-fix-tests-of-acpi_evaluate_inte.patch
>> 0013-Revert-arch-x86-kernel-cpu-cpufreq-acpi-cpufreq.c-av.patch
>>
>> In [1] you can find revert-acpi-patchset, dmesg & lspci outputs of my
>> new kernel.
>>
>> Marciej and Werner can you please verify or dig deeper by partially
>> excluding the single reverts?
>> Thank you.


Patch solves the problem.


Thanks
-- 
Maciej Rutecki
http://www.maciek.unixy.pl