[tlinux-users:07339] Re: Ubuntu on Satellite M35x-S3112 [ Update on Toshiba L10 suppot ]

I'm brand new at this but based on my experience with a Toshiba Libretto
70CT, I would bet that when the system 'hibernates', the BIOS writes the
'machine' out to a section of the hard disk and that stomps all over
your formatted drive.  Usually the Toshiba BIOS use the very last
cylinders of the hard drive.  Its just a guess but you might check it.
Jim 

-----Original Message-----
From: Fred Thiele [mailto:[email protected]] 
Sent: Sunday, October 16, 2005 2:17 AM
To: [email protected]
Subject: [tlinux-users:07318] Re: Ubuntu on Satellite M35x-S3112 [
Update on Toshiba L10 suppot ]

* Serhan Danis <[email protected]> [2005-10-16 06:07:58 +0300]:

> 
> hi, i've edited the acpi_sbs..bla..bla for the 2.6.13 kernel. 
> the whole tarball's here:
> http://www.gsulinux.org/~fenix/src/acpi_sbs-20050120.tar.gz
>         .
>         |-- COPYING
>         |-- Makefile
>         |-- README
>         |-- TODO
>         |-- acpi-ec-2.6.13.diff 
>         |-- acpi_sbs-20050120-Toshiba-L10-Patch.diff
>         |-- acpi-sbs.c
>         |-- i2c-acpi-ec.c
>         `-- i2c-acpi-ec.h

Nice to see my L10-patch appear here ;-). Well I gave the sbs4linux
project of Richard Townsend another try.
https://sourceforge.net/projects/sbs-linux/

I managed to get it working in the 2.6.13 Kernel too and I'll update my
webpage at http://www.minet.uni-jena.de/~ferdy/l10.html ASAP to contain
a step-by-step documentation on how to get it working. Although the
patches are written for some ACER-Modells, the patch did also work with
my Toshiba. One drawback mentioned by Rich still applies (the
spinlock-patch isn't applyable to the 2.6.13-sources): You loose
interupts when the kernel accesses the sbs-systems which means that e.g.
keypresses get lost. I'll try to rework that patch for the 2.6.13. If
I've got good news I'll let you know here first ;-)

> On Sat October 15 2005 20:10, steve wrote:
> > I've read on the web that other Toshiba users have experienced 
> > similar problems.  I installed Ubuntu Breezy Badger on a Satellite, 
> > and whenever I hibernate it stops working.  Not entirely, but sudo 
> > makes a terminal window become unresponsive, and it won't shut down;

> > it freezes when it's trying to deal with network connections.
> >
> > /var/log/messages tells me
> > ACPI: Processor [CPU0] (supports 8 throttling states) Attempting 
> > manual resume

> > swsusp: Suspend partition has wrong signature?

This means that something is wrong with your swap partition. You should
choose a partition of at least the size of your available memory. Then
you can do a mkswap /dev/hdXY to initialize it. In the kernel's section
"default resume partition" also says"/dev/hdXY". After this recompile
the kernel. And finally you put in your boot options (grub.conf or
lilo.conf) the line "resume=/dev/hdXY".  That's it. For further
documentation consider reading the
/usr/src/linux/Documentation/power/swsuspend.txt documentation on how to
suspend to disk (hibernate) your system right. 

I messed around with suspend2 in the kernel 2.6.13 these days, but i got
no good results. As there are problems for resuming Direct Rendering
Interface (DRI) in Xorg for both suspend and suspend2 (after resuming
you no longer have 3d-acceleration using dri-kernel interface) I've
choosen to have suspend(1) running until a better support for the
dri-stuff is known for suspend2. At least suspend works flawlessly on my
Sat. L10. I will also try to document the procedure in my words at my
site (see above).

--
Best regards,

Fred Thiele                    GnuPG-Key: 0x4F6BE238 <[email protected]>