[ltp] Re: [ibm-acpi-devel] RFC: removal of video output switching from thinkpad-acpi

[Jerone Young]

At least for me on my Thinkpad T61 & T41 I primary use xrandr. I don't
even bother with the switching button anymore (and now with xrandr it
may be obsolete). As long as the button sends some ACPI id when
pressed,  then userspace can do whatever. I think it will turn into
xrandr doing all the magic.


Hmm while your on radical changes .. what about moving /proc/acpi/ibm
to /proc/acpi/thinkpad .. at least while /proc/acpi exists.

My T61 has an intel 965 graphics card and the T41 has an ati card.

On Feb 9, 2008 6:39 PM, Henrique de Moraes Holschuh <hmh@hmh.eng.br> wrote:
> This is a request for comments.  Nothing is set in stone yet.
>
>
> About the thinkpad-acpi video output switching:
>
> The video output port switching in thinkpad-acpi has always been a major
> hack.  It works by hinting to the firmware what ports we want to have
> enabled, then asking it to switch to the next port (whatever might it be).
>
> Enabling and disabling video ports is sane, but it seldom works (especially
> when X.org is running).
>
> The whole thing doesn't work well on many thinkpads.  It is not very
> reliable on some of the thinkpads it works well on.  And I have no idea how
> well it works on most thinkpads, except for the vague idea I can get from
> the ibm-acpi.sf.net tables (which doesn't take new X.org into account), and
> a few emails I got with some information that did not go to the webpages
> yet.
>
>
> About video output switching in the generic ACPI video driver:
>
> It is gone.  ACPI defines how it should work, but the vendors almost never
> get it right.  So, it was removed from video.c very recently.
>
> Most thinkpads don't have a real generic ACPI video interface, it piggybacks
> on IBM's weird proprietary ACPI hooks for this (which thinkpad-acpi uses),
> and as (AFAIK) Windows doesn't use it either, it might not be really well
> tested.  AFAIK, only recent thinkpads have standard ACPI hooks for video
> control, and all of those are much better using X.org to deal with the video
> output ports (see below).
>
>
> About video output switching in userspace:
>
> Newer RandR 1.2 X.org drivers can do it, and they can do it very well.  ATI
> Radeon does it nearly perfectly, and so do the Intel drivers.  Any remaining
> bugs on that area are probably going to be gone soon.
>
> As I type this on a Debian testing machine with a backport of the latest git
> xserver-xorg-video-ati code, I have LVDS, VGA, DVI and S-Video output
> control working through RandR (DVI goes through the port replicator).
>
>
> The questions I need answered:
>
> I am inclined to NOT port any of the thinkpad-acpi video output switching
> interfaces to sysfs, which means they will be gone if/when /proc/acpi dies.
> IMHO, it looks like we better work on userspace utilities to deal with it,
> or move it to kernel framebuffer drivers.
>
> Also, I am considering removing the entire thing from thinkpad-acpi one year
> from now, regardless of the fate of /proc/acpi.
>
> Would the removal of this functionality impact your use of your thinkpad?
> Which model of thinkpad? How well does the video output stuff work on
> console mode? and under X.org 7.2 or 7.3?
>
> --
>   "One disk to rule them all, One disk to find them. One disk to bring
>   them all and in the darkness grind them. In the Land of Redmond
>   where the shadows lie." -- The Silicon Valley Tarot
>   Henrique Holschuh
>
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Microsoft
> Defy all challenges. Microsoft(R) Visual Studio 2008.
> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> _______________________________________________
> ibm-acpi-devel mailing list
> ibm-acpi-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/ibm-acpi-devel
>