Re: oops in bk pull (oct 03)

[David S. Miller]

   From: Linus Torvalds <[e-mail]>
   Date: Fri, 4 Oct 2002 18:02:15 -0700 (PDT)
   
   On Fri, 4 Oct 2002, Alexander Viro wrote:
   > Hell knows.  The only explanation I see (and that's not worth much) is that
   > we somehow confuse the chipset and get crapped on something like next cache
   > miss.
   
   I don't see any better explanation right now, so I guess we just revert 
   that thing.
   
The people seeing this don't happen to be on Serverworks chipsets
are they?

I've seen a bug on serverworks where back to back PCI config
space operations can cause some to be lost or corrupted.

Another theory is that some device just dislikes being given
a 0 in one of it's base registers, but somehow ~0 is ok :-)