[Raring,Saucy,SRU,1/1] UBUNTU: SAUCE: Work around broken ACPI backlight on Dell Inspiron 5537

Add a quirk for this machine so backlight can be adjusted again.

BugLink: http://bugs.launchpad.net/bugs/1231305
Signed-off-by: Anthony Wong <anthony.wong@canonical.com>
---
 drivers/acpi/blacklist.c |    8 ++++++++
 1 file changed, 8 insertions(+)

On Tue, Nov 05, 2013 at 07:58:27PM +0800, Anthony Wong wrote:
> Add a quirk for this machine so backlight can be adjusted again.
> 
> BugLink: http://bugs.launchpad.net/bugs/1231305
> Signed-off-by: Anthony Wong <anthony.wong@canonical.com>
> ---
>  drivers/acpi/blacklist.c |    8 ++++++++
>  1 file changed, 8 insertions(+)
> 
> diff --git a/drivers/acpi/blacklist.c b/drivers/acpi/blacklist.c
> index 6023e51..d630bcd 100644
> --- a/drivers/acpi/blacklist.c
> +++ b/drivers/acpi/blacklist.c
> @@ -345,6 +345,14 @@ static struct dmi_system_id acpi_osi_dmi_table[] __initdata = {
>  		     DMI_MATCH(DMI_PRODUCT_VERSION, "ThinkPad Edge E330"),
>  		},
>  	},
> +	{
> +	.callback = dmi_disable_osi_win8,
> +	.ident = "Dell Inspiron 5537",
> +	.matches = {
> +		     DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."),
> +		     DMI_MATCH(DMI_PRODUCT_NAME, "Inspiron 5537"),
> +		},
> +	},
>  
>  	/*
>  	 * BIOS invocation of _OSI(Linux) is almost always a BIOS bug.

Is this going upstream?  I don't see this in Trusty either?

Otherwise it seems ok, matches the standard pattern and looks very
limited in scope to one machine.  So:

Acked-by: Andy Whitcroft <apw@canonical.com>

-apw

On Tue, Nov 05, 2013 at 01:47:40PM +0000, Andy Whitcroft wrote:
> On Tue, Nov 05, 2013 at 07:58:27PM +0800, Anthony Wong wrote:
> > Add a quirk for this machine so backlight can be adjusted again.
> > 
> > BugLink: http://bugs.launchpad.net/bugs/1231305
> > Signed-off-by: Anthony Wong <anthony.wong@canonical.com>
> > ---
> >  drivers/acpi/blacklist.c |    8 ++++++++
> >  1 file changed, 8 insertions(+)
> > 
> > diff --git a/drivers/acpi/blacklist.c b/drivers/acpi/blacklist.c
> > index 6023e51..d630bcd 100644
> > --- a/drivers/acpi/blacklist.c
> > +++ b/drivers/acpi/blacklist.c
> > @@ -345,6 +345,14 @@ static struct dmi_system_id acpi_osi_dmi_table[] __initdata = {
> >  		     DMI_MATCH(DMI_PRODUCT_VERSION, "ThinkPad Edge E330"),
> >  		},
> >  	},
> > +	{
> > +	.callback = dmi_disable_osi_win8,
> > +	.ident = "Dell Inspiron 5537",
> > +	.matches = {
> > +		     DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."),
> > +		     DMI_MATCH(DMI_PRODUCT_NAME, "Inspiron 5537"),
> > +		},
> > +	},
> >  
> >  	/*
> >  	 * BIOS invocation of _OSI(Linux) is almost always a BIOS bug.
> 
> Is this going upstream?  I don't see this in Trusty either?

linux-next has a patch series that's supposed to be the real "fix" for
these backlights. Atm this seems to amount to a module parameter that
tells acpi_video not to register its backlight, with plans to add a
blacklist later to default the parameter to 1 for these broken machines.

Point being that I'm not sure whether they're still taking patches to
expand the OSI blacklist for these sorts of machines or if they're going
to expect them to go in some other blacklist. Once we know which way
they're going in 3.13 we should try (again) to upstream fixes for all
the machines we've blacklisted in sauce patches for this problem.

This seems perfectly reasonable for raring/saucy however.

Acked-by: Seth Forshee <seth.forshee@canonical.com>

On 5 November 2013 21:47, Andy Whitcroft <apw@canonical.com> wrote:

> On Tue, Nov 05, 2013 at 07:58:27PM +0800, Anthony Wong wrote:
> > Add a quirk for this machine so backlight can be adjusted again.
> >
> > BugLink: http://bugs.launchpad.net/bugs/1231305
> > Signed-off-by: Anthony Wong <anthony.wong@canonical.com>
> > ---
> >  drivers/acpi/blacklist.c |    8 ++++++++
> >  1 file changed, 8 insertions(+)
> >
> > diff --git a/drivers/acpi/blacklist.c b/drivers/acpi/blacklist.c
> > index 6023e51..d630bcd 100644
> > --- a/drivers/acpi/blacklist.c
> > +++ b/drivers/acpi/blacklist.c
> > @@ -345,6 +345,14 @@ static struct dmi_system_id acpi_osi_dmi_table[]
> __initdata = {
> >                    DMI_MATCH(DMI_PRODUCT_VERSION, "ThinkPad Edge E330"),
> >               },
> >       },
> > +     {
> > +     .callback = dmi_disable_osi_win8,
> > +     .ident = "Dell Inspiron 5537",
> > +     .matches = {
> > +                  DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."),
> > +                  DMI_MATCH(DMI_PRODUCT_NAME, "Inspiron 5537"),
> > +             },
> > +     },
> >
> >       /*
> >        * BIOS invocation of _OSI(Linux) is almost always a BIOS bug.
>
> Is this going upstream?  I don't see this in Trusty either?
>
> Otherwise it seems ok, matches the standard pattern and looks very
> limited in scope to one machine.  So:
>
> Acked-by: Andy Whitcroft <apw@canonical.com>
>
> -apw
>

Although upstream has accepted OSI blacklisting several machines like what
we are doing here, it doesn't seem they actually like it [1].
As Seth pointed out there's a more acceptable patch series on the horizon,
agree to wait for 3.13 to decide how to fix Trusty.

[1] https://groups.google.com/d/msg/linux.kernel/SoCcSOFE1Og/ZDDXX-ovpFQJ

Thanks,
Anthony