Issue with Radeon Mobility and suspend on T42 (and other thinkpads)

[Stephen Ryan]

On Sun, 13 Mar 2005 09:49:54 -0500, thully at umich.edu <thully at umich.edu> wrote:
> For some time, I have had trouble with my ThinkPad T42 and suspend-to-RAM.
> It suspends/resumes fine, but it uses about 5-7% of its battery per hour in
> suspend.  I've been conversing w/mjg59 on this topic, and the problem is known
> (radeon card isn't powered off properly) - however the solution  has been a bit
> tricky.  It can be solved with a kernel patch, but this means using radeonfb
> which is a Bad Thing (and not viable as a fix for Ubuntu).  Does anyone else
> out there have this problem with this machine?  If so, any chance that you can
> look at the solution mjg59 posted at http://www.srcf.ucam.org/~mjg59/radeon/
> and look at the code for problems (currently, this doesn't work for me).
> 
> To enable this, you will need to:
> 1) Compile the contents of the tarball
> 2) Move radeontool to /usr/local/bin
> 3) Add radeontool power off to the end of /etc/acpi/prepare.sh, and radeontool
> power on to /etc/acpi/resume.sh right after the #!/bin/bash line
> 
> This bug (and the kernel patch for radeonfb) can be found at
> http://bugme.osdl.org/show_bug.cgi?id=3022


I have a T40 (2373-8CU, Radeon M 7500) which is affected by this bug. 
It is fixed by the radeonfb kernel patch; it is not affected by this
version of radeontool.  FWIW, I can still see a bunch of flashing
lines on the screen after running "radeontool power off", which I
don't recall having seen before, either with the almost-stock warty
configuration (I have Daniel Stone's X40 support ACPI scripts
installed), or with the patched 2.6.11-rc2 kernel.

I'll try to take another look at this after work, but the last time I
tried writing software that directly accessed a piece of hardware was
back when the C> prompt was a common sight, so it's rather unlikely
that I'll make much progress on my own.