Re: ext3 kernel 2.4.13

[Oguzhan Eris <eris cooper edu>]

> > Compiled everything into the kernel.  It produces the same exact error
> > output.  Any ideas?
> > 
> 
> Never seen this before.  It's unlikely to be an ext3 problem.
> It seems that when ext3 tries to read the superblock of hda1,
> the IDE driver decides that one or both of blocks 2&3 lie
> outside the bounds defined in the partition table.
> 
> My *guess* would be that an earlier attempt to read the partition
> table went wrong, perhaps by picking up stale cache data.  People
> have been playing with code which could do that, lately
> 
> What happens if you do all the same steps, only with ext2?
> 

I change the mount options  or specify rootfstype=ext2 as a kernel
argument and it craps out the same way actually.. So I guess it isn't an
ext3 issue directly.  I'm gonna try again by removing the ext3 patches
from the 2.4.13 and just going with an ext2 mount.  

The reason I actually want to try out this latest kernel is because my
kernel currently doesn't allow dma on my ide chipset/hardrive  

I had to put  hdparm -d 0 /dev/hda   in the init scripts to resolve a 

hda: dma_intr: status=0x51 { DriveReady SeekComplete Error }
hda: dma_intr: error=0x84 { DriveStatusError BadCRC }

its a via chipset and my old 2.2 kernel was fine with it all.  when I got
Mandrake 8.1 these errors started popping up and the hard drive eventually
developed bad sectors and needed to be replaced. So now I run without dma
and things are good although slow.  There have been some ide changes since
the 2.4.8 kernel so I was hoping that the 2.4.13 would fix it, but no
luck.   I know this isn't the place to write about this but I just wanted
to explain it.   Who knows, maybe it's all a chipset driver error.

I will post my results with ext2 on an unpatched 2.4.13 kernel

Thanks for your help btw.

Oguzhan Eris