[LinuxBIOS] Eventually a RAM problem for slow booting Epia-M

Christian Sühs chris at suehsi.de
Tue Dec 20 20:34:21 CET 2005

> Chage it to 9.  9 is printk_spew
> However I just looked at the epia-m auto.c main() and the serial port
> is not enabled until _after_ enable_smbus().  So something in your
> smbus may be hanging things up until it times out.
> Try this:
> in epia-m/auto.c comment out line 124, enable_smbus() and re-run. 
> With loglevel 9 you should get the "In auto.c main()" message pretty
> quick.  If that works then try moveing enable_smbus() to after the
> serial init.
> If that crashes or hangs then you will have to leave it alone and
> debug code prior to main() and find out where its looping to cause the
> pause.

Ok, first I have commented out enable_smbus(), for that there are 
unreadable characters at the console and it takes the same time --> the 
false way

Then I recognized the lines above in auto.c, which disable the Firewire 
stuff, but The Epia-ML does not have Firewire, ok it is the same chipset 
and the firewire is integrated in the chip, but voila it works.

I have commented out line 119-123 and all works fine on high speed :D
Is there a better way?

And now, the next problem.

It is possible to get the vesafb on boottime. Because, there is no 
longer a bootsplash :(

Stefan means, I need an CLE266 fb kerneldriver and it should work, but 
via offered a kernel modul, only

Know anybody a kernel fb driver for that chipset.


More information about the coreboot mailing list