Looking through the kernel list, I see a couple things from Andre on
this, early in the year... (thread starts at
http://www.uwsg.iu.edu/hypermail/linux/kernel/0001.3/0940.html)
One, he says that mixing Maxtor and WD drives on a bus causes timing
problems(?!) So... just out of curiosity, what drives are in the box?
Also, Andre says to the guy on the list:
"Also you have not enable PIIX tuning."
CONFIG_PIIX_TUNING is not set in the .configs in our tree, although
CONFIG_BLK_DEV_PIIX is set. Config.help for CONFIG_BLK_DEV_PIIX says
"If you say Y here, you should also say Y to 'PIIXn Tuning support,'
below." so that looks like an error on our part.
What's the version string on the test11 kernel that works? In some of
the boot kernels we produced, both of the above options were NOT set on
the BOOT kernel.
I'll generate some new RPMs with both of the above set - is the box in
good enough shape to try a new RPM?
-Eric
Rajagopal Ananthanarayanan wrote:
>
> We are seeing some problems, likely related to 2.4.0.
> When creating an xfs root on an IDE drive with the
> test2 CD the installed kernel (2.4.0+xfs) does not
> get past mounting root. The system keeps spewing
> messages like "hda lost interrupt". The same system
> doesn't have problems with the same drive when running
> the boot kernel based on test11. A further datapoint is
> that if the root is made from ext2 the problem is not seen.
>
> Has anyone successfully installed the test2 CD on an IDE drive?
> I've also seen some reports relating to IDE & interrupts
> on the l-k mailing list, but couldn't really say distill
> any concise workaround for this problem. Does anyone have
> a good handle on what's going on? Please let me know if you
> need additional information.
>
> thanks!
|