Re: ACPI video.c brightness handler conflicts with toshiba_acpi

[Andrey Borzenkov]

On Wednesday 29 October 2008, Thomas Renninger wrote:
> On Tuesday 28 October 2008 17:22:12 Andrey Borzenkov wrote:
> > On Tuesday 28 October 2008, Rafael J. Wysocki wrote:
> > > On Tuesday, 28 of October 2008, Thomas Renninger wrote:
> > > > On Monday 27 October 2008 17:52:55 Andrey Borzenkov wrote:
> > > > > On Monday 08 September 2008, Thomas Renninger wrote:
> > > > > > On Saturday 06 September 2008 09:19:59 am Andrey Borzenkov wrote:
> > > > > > > On Saturday 06 September 2008, Andrey Borzenkov wrote:
> > > > > > > > I have now two different devices that refer to the same
> > > > > > > > hardware:
> > > > > > > >
> > > > > > > > lrwxrwxrwx 1 root root 0 2008-09-06 11:04 acpi_video0 ->
> > > > > > > > ../../devices/virtual/backlight/acpi_video0/ lrwxrwxrwx 1 root
> > > > > > > > root 0 2008-09-06 11:04 toshiba ->
> > > > > > > > ../../devices/virtual/backlight/toshiba/
> > > > > > > >
> > > > > > > > Unfortunately, due to ACPI implementation the acpi_video0 one
> > > > > > > > is much inferior (as it provides only effectively two levels
> > > > > > > > instead of 8); and user level tools are apparently quite
> > > > > > > > confused which one to select.
> > > > > > > >
> > > > > > > > Is there any mechanism that would allow tochiba_acpi to claim
> > > > > > > > brightness for internal LCD screen that video would not attempt
> > > > > > > > to grab it too?
> > > > > > > >
> > > > > > > > Of course manually disabling brightness handling in video is
> > > > > > > > always possible,
> > > > > > >
> > > > > > > Actually it is not. brightness_switch_enabled only disables event
> > > > > > > handling; it still resets actual brightness on loading and
> > > > > > > creates sysfs files to confuse user space.
> > > > > > >
> > > > > > > > still is nice for this to be handled automatically.
> > > > > >
> > > > > > It is in latest ACPI test branch queued for 2.6.28.
> > > > >
> > > > > It is still not in rc2; is it scheduled for 2.6.28 or delayed
> > > > > further?
> > > >
> > > > Right, it's not there.
> > > > Len, what happened with the:
> > > > "Check for ACPI backlight support otherwise use vendor ACPIdrivers -
> > > > version 4" patches?
> > > >
> > > > They are pretty essential for video.ko.
> > > > Otherwise graphics devices may get double poked (through vendor_acpi.ko
> > > > drivers) or may register for graphics cards which are not there.
> > >
> > > Do you have a pointer to those patches?
> >
> > So far I am running with these commits from
> > git://git.kernel.org/pub/scm/linux/kernel/git/ak/linux-acpi-2.6:
> This tree does not exist anymore...
> 
> Andrey, do you still have a local copy?
> Could you use git format-patch and repost the whole series
> (should be about 11 patches) to Len and CC me.

Do you have at least subjects for the patches? I have only
been using the quoted below, so I am not sure which other to pick.
Here is what git log --author=Renninger gives me:

Thu Aug 28 17:26:10 2008 +0200 ACPI video: Debug interface used for error message cleanup
Fri Aug 1 17:37:55 2008 +0200 Check for ACPI backlight support otherwise use vendor ACPI drivers
Fri Aug 1 17:37:54 2008 +0200 ACPI: video: Ignore devices that aren't present in hardware
Mon Jul 23 14:44:41 2007 +0200 ACPI: autoload modules - Create __mod_acpi_device_table symbol for all ACPI drivers
Mon Jun 26 23:58:43 2006 -0400 ACPI: Enable ACPI error messages w/o CONFIG_ACPI_DEBUG
Tue Nov 8 05:27:00 2005 -0500 [ACPI] Fix Null pointer deref in video/lcd/brightness

If you have open git repository, I could probably just push my
linux-acpi branch to you?

> It would be some work for me to recollect everything.
> 
> I hope Len is reading that soon and can give a comment why this was
> in Andi's, but did not get into his ACPI git tree...
> 
> This should directly go into ACPI test branch again.
> It would be great if it could still get cherry-picked into .28.
> 
> Thanks,
> 
>      Thomas
> 
> Norbert: I just saw your brightness complains about your Sony on the acpi 
> list. You should start testing with the patchset we are talking about here.
> video.ko is rather broken without them.
> 
>      Thomas
> 
> > commit f43d728731c691772ddc29e50d25c68a859935b5
> > Author: Thomas Renninger <trenn@xxxxxxx>
> > Date:   Fri Aug 1 17:37:55 2008 +0200
> >
> >     Check for ACPI backlight support otherwise use vendor ACPI drivers
> >
> >
> > commit ad9ed8385ed6ec5be8da7094db911c824258ceec
> > Author: Thomas Renninger <trenn@xxxxxxx>
> > Date:   Fri Aug 1 17:37:54 2008 +0200
> >
> >     ACPI: video: Ignore devices that aren't present in hardware
> >
> > I have been waiting for in-tree inclusion to submit patches for
> > toshiba_acpi against stable base.
> 
> 
>

Attachment: signature.asc
Description: This is a digitally signed message part.