[ibm-acpi-devel] [ltp] Re: thinkpad-acpi release 0.17-20071002 uploaded to ibm-acpi.sf.net

[Henrique de Moraes Holschuh]

On Wed, 03 Oct 2007, Chris Hanson wrote:
> Henrique de Moraes Holschuh wrote:
> > Better use this one instead...
> 
> The patch made no difference; it still gets stuck generating EINVAL.

:-(

> I forgot to mention that, although the /proc and /sys entries are unable
> to change the brightness, the hotkeys work fine, and the correct
> brightness level is shown by actual_brightness and brightness.

So the driver is always reading the correct brightness from hardware, but it
is still going bonkers when trying to set the brightness. Hmm.  Weird that I
cannot reproduce this, no matter how much I try.  Is there a possibility of
doing a quick try with just 2.6.22.9 + thinkpad-acpi without the other
patches, just to make sure it is not interacting in surprising ways with
something else?

I test it here with 2.6.22.9 + tp-smapi and cfs v22, plus updated ipw2200
and some other drivers.  But RT changes a damn great deal in the kernel, and
some of the other patches you mentioned are also relatively intrusive, and
might be triggering the issue.

Also, what if you remove the video driver temporarily, does it work?  Also,
using the keys, do you have 8 or 16 brightness levels?

If that doesn't help, the only thing that comes to mind is to get the driver
source, and sprinkle a number of "printk(IBM_ERR "debug: I am at point
foo\n");" on the source to trace code flow (or, if you are up to it, using
an in-kernel debugger, but THAT is really intrusive and I can't guide you
through it).

-- 
  "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