problems with Via KT133(A) chipsets

[John Duff]

For those of you who have machines with mobos based on VIA's KT133 and KT133A
chipset (with the VT82C686A and VT82C686B south bridges, respectively), you may 
be interested in reading this.  All others may delete this...

It seems that recently it was more or less confirmed that there is a nasty
bug in the 686B chipset.  Copying large amounts of data between IDE interfaces
(using drives in DMA mode and at ATA 66/100) can result in silent data 
corruption.  There are some proposed fixes for Windows available, and Via
announced it is looking into BIOS fixes as well.  As for linux fixes, I
scanned the kernel mailing list archives, and while problems with Via
chipsets seem to have been known for some time, I didn't see any cure-all
fix mentioned.  Running the drives in slower PIO mode seems to be the only
solid fix for now.  This problem affects mainly 2.4.x kernels, but there
is some mention of 2.2.x possibly having problems as well.

There have also been some minor problems mentioned with the 686A south bridge in
the KT133 chipsets, but so far no one has reported that the 686A has the
same bug mentioned above that the 686B has.

Here are some links to descriptions of the problem with he 686B:

http://home.tiscalinet.de/au-ja/review-kt133a-1-en.html

http://www.theregister.co.uk/content/3/18267.html

-John