Re: [PATCH] Added backlight driver for Acer Aspire 4736

[Pradeep Subrahmanion <subrahmanion.pradeep@gmail.com>]

On Thu, 2012-03-15 at 16:05 +0800, joeyli wrote:
> Hi Pradeep, 
> 
> 於 三，2012-03-14 於 11:47 +0530，Pradeep Subrahmanion 提到：
> > On Wed, Mar 14, 2012 at 11:21 AM, joeyli <jlee@suse.com> wrote:
> > > 於 三，2012-03-14 於 08:13 +0530，Pradeep Subrahmanion 提到：
> > >> Hi Joey ,
> > >>
> > >> > Per my understood, EC firmware should change brightness but didn't do
> > >> > that, another
> > >> > way is touch i915 register in _BCM.
> > >>
> > >>   how do we do this ? you mean change the _BCM implementation ?
> > >
> > > "BIOS guy" should do something like this:
> > >
> > >  Method (AINT, 2, NotSerialized)
> > > {
> > > ...
> > >        If (LEqual (Arg0, One))
> > >        {
> > >            Store (Divide (Multiply (Arg1, 0xFF), 0x64, ), BCLP)
> > >            Or (BCLP, 0x80000000, BCLP)         <== touch BCLP register
> > >            Store (0x02, ASLC)
> > >        }
> > >
> > >    Method (_BCM, 1, NotSerialized)
> > >    {
> > >        If (LAnd (LGreaterEqual (Arg0, Zero), LLessEqual (Arg0, 0x64)))
> > >        {
> > >            AINT (One, Arg0)     <== call AINT method
> > >            Store (Arg0, BRTL)
> > >        }
> > >    }
> > >
> > >
> > > Just for reference, they should do that when EC didn't wire to
> > > backlight.
> > >
> > >> >
> > >> > Acer machine provide a broken _BCM implementation and they didn't test
> > >> > it.
> > >> >
> > >> > > > > By ' ACPI interface' , I mean 'acpi_video0' inside the
> > >> > > > > /sys/class/backlight. I havn't tried the /sys/class/backlight interface
> > >> > > > > directly . I will try that also.
> > >> > > >
> > >> > > > So writing values into /sys/class/backlight/acpi_video0/brightness does
> > >> > > > nothing?
> > >> > >
> > >> > >
> > >> > > No change in value when writing
> > >> > > to /sys/class/backlight/acpi_video0/brightness.
> > >> > >
> > >> > > Another thing is that when i did boot with acpi_backlight = 'acer_wmi' ,
> > >> > > in new kernel (3.3.0-rc7) , it shows following messages ,
> > >> > >
> > >> > > [    8.350825] wmi: Mapper loaded
> > >> > > [   10.363975] acer_wmi: Acer Laptop ACPI-WMI Extras
> > >> > > [   10.396186] acer_wmi: Function bitmap for Communication Device: 0x91
> > >> > > [   10.396385] acer_wmi: Brightness must be controlled by generic video
> > >> > > driver
> > >> > >
> > >> > > Also there was no interface inside /sys/class/backlight for acer_wmi.
> > >> > >
> > >> >
> > >> > Yes, acer_wmi support backlight control with AMW0 interface, your
> > >> > machine didn't have AMW0 interface.
> > >> >
> > >> > Normally, backlight should control by standard acpi interface.
> > >> >
> > >> > > I also tried writing directly to Embedded controller register .But no
> > >> > > change.
> > >> >
> > >> > The machine has broken _BCM method, because EC should do something after
> > >> > _BCM changed EC register.
> > >>
> > >> Thanks ,
> > >>
> > >> Pradeep Subrahmanion
> > >
> > > Why they didn't find _BCM not work?
> > >
> > > My guess is:
> > >
> > > Because the backlight control is through WDDM driver on Windows platform
> > > but not through standard ACPI method _BCM. They only test Windows
> > > platform, so, they didn't find _BCM broken.
> > >
> > > And, they also didn't really follow Microsoft WDDM spec:
> > >
> > >  http://msdn.microsoft.com/en-us/windows/hardware/gg487382.aspx
> > >
> > > Per spec,
> > > ODM should keep _BCM works fine for any other OS didn't support WDDM
> > > driver, but they didn't.
> > >
> > > At last year, I told Acer PM one time for this issue, they said will
> > > check but finally didn't response me.
> > >
> > 
> > >
> > > Thanks
> > > Joey Lee
> > >
> > 
> > So touching the PCI LBB register is the only feasible solution now
> > (even though it may not be a clean  method) ?
> > 
> > Thanks,
> > Pradeep Subrahmanion
> 
> That will be better leave LBB register only touched by i915 driver.
> 
> If 'acpi_backlight=vendor' works to you, maybe we can add a quirk to
> video_detect.c. 
> You can try the following patch.

Thanks . I tried your patch .acpi_backlight=vendor allows me to control
brightness with hot key.But there are problems with it like increasing
brightness after  maximum level causes blank screen.So I am trying it
sort it out.
 
Thank you , 

Pradeep Subrahmanion

> 
> 
> Thanks a lot!
> Joey Lee
> 
> 
> >From 038bd3c4e53b7195f34e9d46c999b8dcb279da5e Mon Sep 17 00:00:00 2001
> From: "Lee, Chun-Yi" <jlee@suse.com>
> Date: Thu, 15 Mar 2012 16:03:45 +0800
> Subject: [PATCH] acer-wmi: Add quirk table for video backlight vendor mode
> 
> Add quirk table for video backlight vendor mode
> 
> Signed-off-by: "Lee, Chun-Yi" <jlee@suse.com>
> ---
>  drivers/acpi/video_detect.c |   27 +++++++++++++++++++++++++++
>  1 files changed, 27 insertions(+), 0 deletions(-)
> 
> diff --git a/drivers/acpi/video_detect.c b/drivers/acpi/video_detect.c
> index f3f0fe7..acb15d6 100644
> --- a/drivers/acpi/video_detect.c
> +++ b/drivers/acpi/video_detect.c
> @@ -132,6 +132,32 @@ find_video(acpi_handle handle, u32 lvl, void *context, void **rv)
>  	return AE_OK;
>  }
>  
> +static int video_set_backlight_vendor(const struct dmi_system_id *d)
> +{
> +	acpi_video_support |= ACPI_VIDEO_BACKLIGHT_DMI_VENDOR;
> +	return 0;
> +}
> +
> +static const struct dmi_system_id video_vendor_dmi_table[] = {
> +	{
> +	 .callback = video_set_backlight_vendor,
> +	 .ident = "Acer Aspire 4736",
> +	 .matches = {
> +		DMI_MATCH(DMI_BOARD_VENDOR, "Acer"),
> +		DMI_MATCH(DMI_PRODUCT_NAME, "Aspire 4736"),
> +		},
> +	},
> +	{
> +	 .callback = video_set_backlight_vendor,
> +	 .ident = "Acer TravelMate 4750",
> +	 .matches = {
> +		DMI_MATCH(DMI_BOARD_VENDOR, "Acer"),
> +		DMI_MATCH(DMI_PRODUCT_NAME, "TravelMate 4750"),
> +		},
> +	},
> +	{}
> +};
> +
>  /*
>   * Returns the video capabilities of a specific ACPI graphics device
>   *
> @@ -164,6 +190,7 @@ long acpi_video_get_capabilities(acpi_handle graphics_handle)
>  		 *		ACPI_VIDEO_BACKLIGHT_DMI_VENDOR;
>  		 *}
>  		 */
> +		dmi_check_system(video_vendor_dmi_table);
>  	} else {
>  		status = acpi_bus_get_device(graphics_handle, &tmp_dev);
>  		if (ACPI_FAILURE(status)) {