[ltp] T23 Suspend2Ram and Suspend2Disk issues (savage framebuffer and tpb)

[Tino Keitel]

On Wed, Feb 22, 2006 at 07:44:15 -0500, Kevin Locke wrote:
> On Wed, 2006-02-22 at 09:12 +0100, Tino Keitel wrote:
> > Can I use savagefb and suspend/resume successfully using ACPI S3
> > suspend at all, or is this a general problem?
> 
> On my TP21 compiling in the savagefb module will prevent both S3 and S1
> ACPI sleep.  I had asked about this (and some other savagefb problems) 
> before[1] and after a lot of searching, my conclusion is that the
> savagefb driver is quite troubled.  A response that I received on another
> savagefb related bug[2] suggested that now vesafb is the de facto fb
> module and that effectively savagefb is deprecated (I don't consider
> this answer to be authoritative, but it does provide another example of
> the extent of feelings against savagefb).

I switched to vesafb, using vga=791. Suspend to RAM works now. However,
when I switch from X to the console, all I get is garbage.

Since I only enabled the framebuffer to get a suspend2 progress
display, it looks like I have to disable it again. It isn't possible to
get working suspend to RAM and a usable framebuffer after running X.

> > > I also reactivated tpb (the ThinkPad OSD utility). When I suspend to
> > > disk while the ThinkLight is switched on, X seems to hang after resume.
> > > I can see a garbled tpb display in this case. I suppose that tpb tries
> > > to show that the ThinkLight is now switched off, which causes the hang.
> > 
> > Is anyone able to run tpb with the OSD enabled, switch on the
> > ThinkLight, suspend to disk, and resume?
> 
> It worked for me (although the "ThinkLight off" text was off-color for
> its first message after resume).  If you are using IBM-ACPI, as a
> workaround, you might want to add "echo -n off > /proc/acpi/ibm/light"
> to your suspend script.  If you are using the hibernate utility to

This could do the trick. However, it won't fix ACPI S3.

Nasty bugs...

Regards,
Tino