PROBLEM: Acer D255E Backlight out for 10 seconds with ACPI

PROBLEM: Acer D255E Backlight out for 10 seconds with ACPI

[Dan]

Hello,

Apologies if this was sent to the wrong place.

With an Acer AOD255E netbook, the backlight goes out for exactly 10
seconds before returning while booting.

This is a backlight issue and not KMS, as I can shine a flashlight at
the screen and still see boot text scrolling.

With "acpi=off" appended to the kernel line, this no longer occurs (Fn
key backlight controls still work with this appended). With every
other acpi option I have tried, this still occurs though. This
includes "acpi_backlight" and "acpi_display_output". With
"acpi_backlight=vendor" the entry in /sys/class/backlight changes from
acpi_video0 to acer_wmi, but backlight still goes out.

Otherwise backlight works correctly.

Kernel version (I have reproduced it with several other older kernels
as well on different distributions):
Linux version 3.5.4-1-ARCH (tobias@T-POWA-LX) (gcc version 4.7.1
20120721 (prerelease) (GCC) ) #1 SMP PREEMPT Sat Sep 15 08:12:04 CEST
2012

I compiled the system information from /proc and lspci here (You can
also find a dmesg comparision between a regular boot and one with acpi
disabled here):
http://pastebin.com/u/vxl

I'll also point you to the Ubuntu bug report I filed awhile back while
using that. It contains information on the exact same system (running
Ubuntu rather than Arch Linux):
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1034222

Thanks for your time.

Re: PROBLEM: Acer D255E Backlight out for 10 seconds with ACPI

[joeyli]

於 三，2012-09-26 於 16:48 -0400，Dan 提到：
> Hello,
> 
> Apologies if this was sent to the wrong place.
> 
> With an Acer AOD255E netbook, the backlight goes out for exactly 10
> seconds before returning while booting.
> 
> This is a backlight issue and not KMS, as I can shine a flashlight at
> the screen and still see boot text scrolling.
> 
> With "acpi=off" appended to the kernel line, this no longer occurs (Fn
> key backlight controls still work with this appended). With every
> other acpi option I have tried, this still occurs though. This
> includes "acpi_backlight" and "acpi_display_output". With
> "acpi_backlight=vendor" the entry in /sys/class/backlight changes from
> acpi_video0 to acer_wmi, but backlight still goes out.
> 
> Otherwise backlight works correctly. 

Could you please try this kernel parameter?
video.use_bios_initial_backlight=0


Thanks
Joey Lee

--
To unsubscribe from this list: send the line "unsubscribe linux-acpi" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: PROBLEM: Acer D255E Backlight out for 10 seconds with ACPI

[Dan]

On Wed, Sep 26, 2012 at 7:19 PM, joeyli <jlee@suse.com> wrote:
> 於 三，2012-09-26 於 16:48 -0400，Dan 提到：
>> Hello,
>>
>> Apologies if this was sent to the wrong place.
>>
>> With an Acer AOD255E netbook, the backlight goes out for exactly 10
>> seconds before returning while booting.
>>
>> This is a backlight issue and not KMS, as I can shine a flashlight at
>> the screen and still see boot text scrolling.
>>
>> With "acpi=off" appended to the kernel line, this no longer occurs (Fn
>> key backlight controls still work with this appended). With every
>> other acpi option I have tried, this still occurs though. This
>> includes "acpi_backlight" and "acpi_display_output". With
>> "acpi_backlight=vendor" the entry in /sys/class/backlight changes from
>> acpi_video0 to acer_wmi, but backlight still goes out.
>>
>> Otherwise backlight works correctly.
>
> Could you please try this kernel parameter?
> video.use_bios_initial_backlight=0
>
>
> Thanks
> Joey Lee
>

Same results with that kernel parameter applied.

Just to make it clear when this is happening:
The BIOS screen and bootloader work fine. This occurs during the boot
process while the systemd text is scrolling (or with the splash screen
in the case of Ubuntu). I have not been able to pin down any event
during the boot process that triggers it. It seems to be based on time
after the kernel loads. I can tell by the difference booting from a
HDD vs a SSD. With a HDD, it will occur earlier in the process (close
to the time init is called) since booting takes longer. With an SSD,
the backlight goes out close to the time the login screen appears.

Once the backlight is out, it will stay out for exactly 10 seconds
before returning no matter what is on the screen.
--
To unsubscribe from this list: send the line "unsubscribe linux-acpi" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: PROBLEM: Acer D255E Backlight out for 10 seconds with ACPI

[joeyli]

Hi Dan, 

於 四，2012-09-27 於 14:09 -0400，Dan 提到：
> On Wed, Sep 26, 2012 at 7:19 PM, joeyli <jlee@suse.com> wrote:
> > 於 三，2012-09-26 於 16:48 -0400，Dan 提到：
> >> Hello,
> >>
> >> Apologies if this was sent to the wrong place.
> >>
> >> With an Acer AOD255E netbook, the backlight goes out for exactly 10
> >> seconds before returning while booting.
> >>
> >> This is a backlight issue and not KMS, as I can shine a flashlight at
> >> the screen and still see boot text scrolling.
> >>
> >> With "acpi=off" appended to the kernel line, this no longer occurs (Fn
> >> key backlight controls still work with this appended). With every

It's a clue for that problem is happen in acpi, but sometimes the Fn key
on notebook is changed by BIOS, need check DSDT. 

> >> other acpi option I have tried, this still occurs though. This
> >> includes "acpi_backlight" and "acpi_display_output". With
> >> "acpi_backlight=vendor" the entry in /sys/class/backlight changes from
> >> acpi_video0 to acer_wmi, but backlight still goes out.
> >>
> >> Otherwise backlight works correctly.
> >
> > Could you please try this kernel parameter?
> > video.use_bios_initial_backlight=0
> >
> >
> > Thanks
> > Joey Lee
> >
> 
> Same results with that kernel parameter applied.
> 

This parameter works when _BQC used indexed value but not direct use
brightness value.
One possible thing is BIOS really set the initial index or brightness
level to lowest level.

> Just to make it clear when this is happening:
> The BIOS screen and bootloader work fine. This occurs during the boot
> process while the systemd text is scrolling (or with the splash screen
> in the case of Ubuntu). I have not been able to pin down any event
> during the boot process that triggers it. It seems to be based on time
> after the kernel loads. I can tell by the difference booting from a
> HDD vs a SSD. With a HDD, it will occur earlier in the process (close
> to the time init is called) since booting takes longer. With an SSD,
> the backlight goes out close to the time the login screen appears.
> 
> Once the backlight is out, it will stay out for exactly 10 seconds
> before returning no matter what is on the screen.
> 

What's your value of max brightness? Please run:
 # cat /sys/class/backlight/acpi_video0/max_brightness

Please also manually change brightness level after your system boot
finished, and look at does it changed or have problem like reversed?
 # echo 10 > /sys/class/backlight/acpi_video0/brightness

And, 
Please attached your acpidump:
 # acpidump > acpidump.dat


Thanks
Joey Lee


--
To unsubscribe from this list: send the line "unsubscribe linux-acpi" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: PROBLEM: Acer D255E Backlight out for 10 seconds with ACPI

[joeyli]

於 六，2012-09-29 於 22:42 -0400，Dan 提到：
> On Fri, Sep 28, 2012 at 7:58 PM, joeyli <jlee@suse.com> wrote:
> > Hi Dan,
> >
> > 於 四，2012-09-27 於 14:09 -0400，Dan 提到：
> >> On Wed, Sep 26, 2012 at 7:19 PM, joeyli <jlee@suse.com> wrote:
> >> > 於 三，2012-09-26 於 16:48 -0400，Dan 提到：
> >> >> Hello,
> >> >>
> >> >> Apologies if this was sent to the wrong place.
> >> >>
> >> >> With an Acer AOD255E netbook, the backlight goes out for exactly 10
> >> >> seconds before returning while booting.
> >> >>
> >> >> This is a backlight issue and not KMS, as I can shine a flashlight at
> >> >> the screen and still see boot text scrolling.
> >> >>
> >> >> With "acpi=off" appended to the kernel line, this no longer occurs (Fn
> >> >> key backlight controls still work with this appended). With every
> >
> > It's a clue for that problem is happen in acpi, but sometimes the Fn key
> > on notebook is changed by BIOS, need check DSDT.
> >
> >> >> other acpi option I have tried, this still occurs though. This
> >> >> includes "acpi_backlight" and "acpi_display_output". With
> >> >> "acpi_backlight=vendor" the entry in /sys/class/backlight changes from
> >> >> acpi_video0 to acer_wmi, but backlight still goes out.
> >> >>
> >> >> Otherwise backlight works correctly.
> >> >
> >> > Could you please try this kernel parameter?
> >> > video.use_bios_initial_backlight=0
> >> >
> >> >
> >> > Thanks
> >> > Joey Lee
> >> >
> >>
> >> Same results with that kernel parameter applied.
> >>
> >
> > This parameter works when _BQC used indexed value but not direct use
> > brightness value.
> > One possible thing is BIOS really set the initial index or brightness
> > level to lowest level.
> >
> >> Just to make it clear when this is happening:
> >> The BIOS screen and bootloader work fine. This occurs during the boot
> >> process while the systemd text is scrolling (or with the splash screen
> >> in the case of Ubuntu). I have not been able to pin down any event
> >> during the boot process that triggers it. It seems to be based on time
> >> after the kernel loads. I can tell by the difference booting from a
> >> HDD vs a SSD. With a HDD, it will occur earlier in the process (close
> >> to the time init is called) since booting takes longer. With an SSD,
> >> the backlight goes out close to the time the login screen appears.
> >>
> >> Once the backlight is out, it will stay out for exactly 10 seconds
> >> before returning no matter what is on the screen.
> >>
> >
> > What's your value of max brightness? Please run:
> >  # cat /sys/class/backlight/acpi_video0/max_brightness
> >
> > Please also manually change brightness level after your system boot
> > finished, and look at does it changed or have problem like reversed?
> >  # echo 10 > /sys/class/backlight/acpi_video0/brightness
> >
> > And,
> > Please attached your acpidump:
> >  # acpidump > acpidump.dat
> >
> >
> > Thanks
> > Joey Lee
> >
> >
> 
> # cat /sys/class/backlight/acpi_video0/max_brightness
> 9
> 
> Echoing values into brightness works as expected. Higher values
> increase the brightness, lower values descrease it. Echoing 0 into it
> results in a dim screen, but not a blank one.
> 
> I should also mention my findings are consistent with this person's:
> https://bbs.archlinux.org/viewtopic.php?pid=941871#p941871
> 
> He found that the brightness level in
> /sys/class/backlight/acpi_video0/brightness remains the same during
> the blackout by echoing it out at random times during the boot process
> (I had found this out awhile back but failed to remember and mention
> it in the original message. Sorry about that!).
> 
> Acpidump is attached.
> 

I traced your DSDT at this morning, didn't see problem in _BCL, _BCM and
_BQC. 
Per DSDT, the _BQC on this machine used brightness value but not index,
so 'video.use_bios_initial_backlight=0' should works to you for set the
backlight to max_level when acpi/video driver loaded.

And, the value of max_level in package of _BCL is 100, there have no 0
value in level list:

    Method (_BCL, 0, NotSerialized)
    {
        Return (Package (0x0C)
        {
            0x46,               /* level on AC */
            0x28,               /* level on Battery */
            0x0A,               /* 10 */
            0x14,               /* 20 */
            0x1E,
            0x28,
            0x32,
            0x3C,
            0x46,
            0x50,
            0x5A,
            0x64                /* 100 */

That strange for 'video.use_bios_initial_backlight=0' doesn't work to
you for set the backlight level to maximum.
Could you please share the acpi debug log? Please add the following
kernel parameter when system boot. Please also check the kernel config
need CONFIG_ACPI_DEBUG=y:

	acpi.debug_level=0x0000000F acpi.debug_layer=0xffffffff log_buf_len=5M

Please attached on your dmesg or boot message after system boot finished. Hope we
can find any hint from acpi debug log.


Thanks a lot!
Joey Lee

--
To unsubscribe from this list: send the line "unsubscribe linux-acpi" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: PROBLEM: Acer D255E Backlight out for 10 seconds with ACPI

[joeyli]

於 三，2012-10-03 於 08:26 -0400，Dan 提到：
> >> Acpidump is attached.
> >>
> >
> > I traced your DSDT at this morning, didn't see problem in _BCL, _BCM
> and
> > _BQC.
> > Per DSDT, the _BQC on this machine used brightness value but not
> index,
> > so 'video.use_bios_initial_backlight=0' should works to you for set
> the
> > backlight to max_level when acpi/video driver loaded.
> >
> > And, the value of max_level in package of _BCL is 100, there have no
> 0
> > value in level list:
> >
> >     Method (_BCL, 0, NotSerialized)
> >     {
> >         Return (Package (0x0C)
> >         {
> >             0x46,               /* level on AC */
> >             0x28,               /* level on Battery */
> >             0x0A,               /* 10 */
> >             0x14,               /* 20 */
> >             0x1E,
> >             0x28,
> >             0x32,
> >             0x3C,
> >             0x46,
> >             0x50,
> >             0x5A,
> >             0x64                /* 100 */
> >
> > That strange for 'video.use_bios_initial_backlight=0' doesn't work
> to
> > you for set the backlight level to maximum.
> > Could you please share the acpi debug log? Please add the following
> > kernel parameter when system boot. Please also check the kernel
> config
> > need CONFIG_ACPI_DEBUG=y:
> >
> >         acpi.debug_level=0x0000000F acpi.debug_layer=0xffffffff
> log_buf_len=5M
> >
> > Please attached on your dmesg or boot message after system boot
> finished. Hope we
> > can find any hint from acpi debug log.
> >
> >
> > Thanks a lot!
> > Joey Lee
> >
> 
> I've attached the resulting dmesg log. I also set the loglevel to 8
> during this so I could watch the messages on the console. The last
> message I could see before the backlight went out was at 5.93 seconds.
> I tried looking through the log around there, but it's over my head.
> 
> Thank you,
> Dan
> 

Thanks for your acpi debug log.

Per acpi debug log, the DD02 device have backlight control capability.
The default brightness value is 30, acpi/video driver set brightness
value to 100 then set it back to 30 for determine does _BQC use index
but not brightness level:

[    3.113825] ACPI: Execute Method [\_SB_.PCI0.OVGA.DD02._BCL] (Node ffff88007b741c58)
[    3.115980] ACPI: Execute Method [\_SB_.PCI0.OVGA.DD02._BQC] (Node ffff88007b741ca8)
[    3.118310] exregion-0302 [00] ex_system_io_space_han: System-IO (width 32) R/W 1 Address=0000000000000080
[    3.220145]    utils-0286 [00] evaluate_integer      : Return value [30]				<=== default brightness value is 30(default value set by BIOS)
[    3.222315] ACPI: Execute Method [\_SB_.PCI0.OVGA.DD02._BCM] (Node ffff88007b741c80)
[    3.225422] exregion-0302 [00] ex_system_io_space_han: System-IO (width 8) R/W 1 Address=0000000000000072
[    3.227947] exregion-0302 [00] ex_system_io_space_han: System-IO (width 8) R/W 1 Address=0000000000000073
[    3.230424] ACPI: Execute Method [\_SB_.PCI0.OVGA.DD02._BQC] (Node ffff88007b741ca8)
[    3.233123] exregion-0302 [00] ex_system_io_space_han: System-IO (width 32) R/W 1 Address=0000000000000080
[    3.333487]    utils-0286 [00] evaluate_integer      : Return value [100]				<=== set to 100 for determine _BQC_use_index
[    3.336019] ACPI: Execute Method [\_SB_.PCI0.OVGA.DD02._BCM] (Node ffff88007b741c80)
[    3.339263] exregion-0302 [00] ex_system_io_space_han: System-IO (width 8) R/W 1 Address=0000000000000072
[    3.342167] exregion-0302 [00] ex_system_io_space_han: System-IO (width 8) R/W 1 Address=0000000000000073
[    3.345001]    video-0734 [00] video_init_brightness : found 10 brightness levels
[    3.347899]    utils-0286 [00] evaluate_integer      : Return value [131072]
[    3.350956] ACPI: Execute Method [\_SB_.PCI0.OVGA.DD02._BQC] (Node ffff88007b741ca8)
[    3.354598] exregion-0302 [00] ex_system_io_space_han: System-IO (width 32) R/W 1 Address=0000000000000080
[    3.456824]    utils-0286 [00] evaluate_integer      : Return value [30]				<=== set brightness value back to 30
[    3.460059] acpi device:28: registered as cooling_device0

The above acpi/video's behavior is normal, didn't see strange thing.
Please add 'video.use_bios_initial_backlight=0' then grab acpi debug log
again. You can find the similar log of DD02 device, this time the latest
'Return value' should be 100 like: Return value [100].

If 'video.use_bios_initial_backlight=0' works fine, the brightness
should set to 100%, means lightest.


Another dubious message in your dmesg is not from ACPI, it's a message
from i915 driver:
[    2.083603] i915: fixme: max PWM is zero

That means ODM set PWM register value default to zero, this register
used to control the the intensity of the backlight. I am still looking
what's the side effect that was caused by 'PWM = 0'.


Thanks a lot!
Joey Lee

--
To unsubscribe from this list: send the line "unsubscribe linux-acpi" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: PROBLEM: Acer D255E Backlight out for 10 seconds with ACPI

[Dan]

On Thu, Oct 4, 2012 at 11:01 AM, joeyli <jlee@suse.com> wrote:
> 於 三，2012-10-03 於 08:26 -0400，Dan 提到：
>> >> Acpidump is attached.
>> >>
>> >
>> > I traced your DSDT at this morning, didn't see problem in _BCL, _BCM
>> and
>> > _BQC.
>> > Per DSDT, the _BQC on this machine used brightness value but not
>> index,
>> > so 'video.use_bios_initial_backlight=0' should works to you for set
>> the
>> > backlight to max_level when acpi/video driver loaded.
>> >
>> > And, the value of max_level in package of _BCL is 100, there have no
>> 0
>> > value in level list:
>> >
>> >     Method (_BCL, 0, NotSerialized)
>> >     {
>> >         Return (Package (0x0C)
>> >         {
>> >             0x46,               /* level on AC */
>> >             0x28,               /* level on Battery */
>> >             0x0A,               /* 10 */
>> >             0x14,               /* 20 */
>> >             0x1E,
>> >             0x28,
>> >             0x32,
>> >             0x3C,
>> >             0x46,
>> >             0x50,
>> >             0x5A,
>> >             0x64                /* 100 */
>> >
>> > That strange for 'video.use_bios_initial_backlight=0' doesn't work
>> to
>> > you for set the backlight level to maximum.
>> > Could you please share the acpi debug log? Please add the following
>> > kernel parameter when system boot. Please also check the kernel
>> config
>> > need CONFIG_ACPI_DEBUG=y:
>> >
>> >         acpi.debug_level=0x0000000F acpi.debug_layer=0xffffffff
>> log_buf_len=5M
>> >
>> > Please attached on your dmesg or boot message after system boot
>> finished. Hope we
>> > can find any hint from acpi debug log.
>> >
>> >
>> > Thanks a lot!
>> > Joey Lee
>> >
>>
>> I've attached the resulting dmesg log. I also set the loglevel to 8
>> during this so I could watch the messages on the console. The last
>> message I could see before the backlight went out was at 5.93 seconds.
>> I tried looking through the log around there, but it's over my head.
>>
>> Thank you,
>> Dan
>>
>
> Thanks for your acpi debug log.
>
> Per acpi debug log, the DD02 device have backlight control capability.
> The default brightness value is 30, acpi/video driver set brightness
> value to 100 then set it back to 30 for determine does _BQC use index
> but not brightness level:
>
> [    3.113825] ACPI: Execute Method [\_SB_.PCI0.OVGA.DD02._BCL] (Node ffff88007b741c58)
> [    3.115980] ACPI: Execute Method [\_SB_.PCI0.OVGA.DD02._BQC] (Node ffff88007b741ca8)
> [    3.118310] exregion-0302 [00] ex_system_io_space_han: System-IO (width 32) R/W 1 Address=0000000000000080
> [    3.220145]    utils-0286 [00] evaluate_integer      : Return value [30]                             <=== default brightness value is 30(default value set by BIOS)
> [    3.222315] ACPI: Execute Method [\_SB_.PCI0.OVGA.DD02._BCM] (Node ffff88007b741c80)
> [    3.225422] exregion-0302 [00] ex_system_io_space_han: System-IO (width 8) R/W 1 Address=0000000000000072
> [    3.227947] exregion-0302 [00] ex_system_io_space_han: System-IO (width 8) R/W 1 Address=0000000000000073
> [    3.230424] ACPI: Execute Method [\_SB_.PCI0.OVGA.DD02._BQC] (Node ffff88007b741ca8)
> [    3.233123] exregion-0302 [00] ex_system_io_space_han: System-IO (width 32) R/W 1 Address=0000000000000080
> [    3.333487]    utils-0286 [00] evaluate_integer      : Return value [100]                            <=== set to 100 for determine _BQC_use_index
> [    3.336019] ACPI: Execute Method [\_SB_.PCI0.OVGA.DD02._BCM] (Node ffff88007b741c80)
> [    3.339263] exregion-0302 [00] ex_system_io_space_han: System-IO (width 8) R/W 1 Address=0000000000000072
> [    3.342167] exregion-0302 [00] ex_system_io_space_han: System-IO (width 8) R/W 1 Address=0000000000000073
> [    3.345001]    video-0734 [00] video_init_brightness : found 10 brightness levels
> [    3.347899]    utils-0286 [00] evaluate_integer      : Return value [131072]
> [    3.350956] ACPI: Execute Method [\_SB_.PCI0.OVGA.DD02._BQC] (Node ffff88007b741ca8)
> [    3.354598] exregion-0302 [00] ex_system_io_space_han: System-IO (width 32) R/W 1 Address=0000000000000080
> [    3.456824]    utils-0286 [00] evaluate_integer      : Return value [30]                             <=== set brightness value back to 30
> [    3.460059] acpi device:28: registered as cooling_device0
>
> The above acpi/video's behavior is normal, didn't see strange thing.
> Please add 'video.use_bios_initial_backlight=0' then grab acpi debug log
> again. You can find the similar log of DD02 device, this time the latest
> 'Return value' should be 100 like: Return value [100].
>
> If 'video.use_bios_initial_backlight=0' works fine, the brightness
> should set to 100%, means lightest.
>
>
> Another dubious message in your dmesg is not from ACPI, it's a message
> from i915 driver:
> [    2.083603] i915: fixme: max PWM is zero
>
> That means ODM set PWM register value default to zero, this register
> used to control the the intensity of the backlight. I am still looking
> what's the side effect that was caused by 'PWM = 0'.
>
>
> Thanks a lot!
> Joey Lee
>

Output is as expected with 'video.use_bios_initial_backlight=0'

[    1.658690] ACPI: Execute Method [\_SB_.PCI0.OVGA.DD02._BCL] (Node
ffff88007b741c58)
[    1.658756] ACPI: Execute Method [\_SB_.PCI0.OVGA.DD02._BQC] (Node
ffff88007b741ca8)
[    1.659083] exregion-0302 [00] ex_system_io_space_han: System-IO
(width 32) R/W 1 Address=0000000000000080
[    1.756824]    utils-0286 [00] evaluate_integer      : Return value [30]
[    1.756839] ACPI: Execute Method [\_SB_.PCI0.OVGA.DD02._BCM] (Node
ffff88007b741c80)
[    1.757720] exregion-0302 [00] ex_system_io_space_han: System-IO
(width 8) R/W 1 Address=0000000000000072
[    1.757788] exregion-0302 [00] ex_system_io_space_han: System-IO
(width 8) R/W 1 Address=0000000000000073
[    1.757823] ACPI: Execute Method [\_SB_.PCI0.OVGA.DD02._BQC] (Node
ffff88007b741ca8)
[    1.758093] exregion-0302 [00] ex_system_io_space_han: System-IO
(width 32) R/W 1 Address=0000000000000080
[    1.856768]    utils-0286 [00] evaluate_integer      : Return value [100]
[    1.856782] ACPI: Execute Method [\_SB_.PCI0.OVGA.DD02._BCM] (Node
ffff88007b741c80)
[    1.857607] exregion-0302 [00] ex_system_io_space_han: System-IO
(width 8) R/W 1 Address=0000000000000072
[    1.857680] exregion-0302 [00] ex_system_io_space_han: System-IO
(width 8) R/W 1 Address=0000000000000073
[    1.857710]    video-0734 [00] video_init_brightness : found 10
brightness levels
[    1.857735]    utils-0286 [00] evaluate_integer      : Return value [131072]
[    1.857851] ACPI: Execute Method [\_SB_.PCI0.OVGA.DD02._BQC] (Node
ffff88007b741ca8)
[    1.858179] exregion-0302 [00] ex_system_io_space_han: System-IO
(width 32) R/W 1 Address=0000000000000080
[    1.956845]    utils-0286 [00] evaluate_integer      : Return value [100]
[    1.956967] acpi device:28: registered as cooling_device0

Looks like this is happening independently of what the brightness
level is set to. 30% is what I usually have it set to (corresponds to
2 in /sys/class/backlight/acpi_video0/brightness). With
'video.use_bios_initial_backlight=0' The brightness is reset to max,
but the blackout still occurs later in the boot process.

I followed up on that other message from the i915 driver. I applied
this patch and rebuilt the kernel: https://lkml.org/lkml/2012/9/14/377

This should work around it from my (low level of) understanding. I
received the message "i915 doesn't seem to be connected to backlight"
in dmesg as expected, but still the backlight goes out.

Thank you for your help on this.

- Dan
--
To unsubscribe from this list: send the line "unsubscribe linux-acpi" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html