Quoting Olivier Thauvin <[EMAIL PROTECTED]>:

> Le Jeudi 14 Ao�t 2003 01:48, Svetoslav Slavtchev a �crit :
> > Hi,
> > i'm tracking really strange file system corruption
> > (both reiserfs & xfs) on soft-raid-5 and lvm
> > using 4 ibm drives as masters on HPT374
> >
> > now the more strange part is, that this used to work perfectly
> > on my previous mainboard Epox 8K5A3+, but the board stoped working
> > (unsuccefull CPU update) and i had to upgrade to epox 8K9A3+.
> >
> > Since the update i'm seeing this fs corruption,
> > and the difference between both boards is only
> > KT333(old board) <-> KT400 (new board)
> > may be newer HPT374 BIOS(1.24) on the new board
> > and a number of KT400 issues with ACPI/IO-APIC
> > which seems to be mostly resolved with recent kernels.
> >
> > so i#m looking for any kind of pointers for debuging,
> > and i just saw that in the mdk kernel taskfile is enabled,
> > i remmebered that a while ago taskfile was marked "eating data"
> >
> > have i missed smth? when was it activated in the -mdk kernel ?
> > any pointers for debugging the fs corruption?
> >
> > and about the corruption:
> > rpm -ivh kernel-2.4.22.x.mdk.src.rpm
> > for i in 1 2 3; do rpmbuild -ba kernel-2.4.spec;done
> > would end on the 2nd or 3rd run with smth like:
> > ......
> > linux-2.4.21.tar.bz2 obsolated header
> > the archive seems brocken, you might want to try recoverring it with ....
> > ......
> >
> 
> I have somethings here with kernel 2.4.22-pre9 vanilla, Mandrake 9.1.
> I am using lvm, Reiserfs, 130MB.
> I got some "udma interrupt lost on hdx" in my log.
> 
> But IDE TASK is disable here.

130MB ? RAM

i have the problem with 2.4.21-ac1, -ac4, 2.4.21.4mdk, 6mdk, 2.4.22.0.1mdk

svetljo



Reply via email to