On Fri, 2010-11-12 at 11:20 +0200, Michael S. Tsirkin wrote: > On Thu, Nov 11, 2010 at 07:56:13PM -0700, Alex Williamson wrote: > > Now that common PCI code doesn't have a hangup on capabilities > > being contiguous, > > Hmm, this comment confused me : there's no requirement of > contigious allocations in current code in pci.c, is there?
Exactly, but the code used to have cap.start and cap.length, which implied it was contiguous. Since those were removed in 5/8, we don't need to worry about where the physical capabilities land in config space. Thanks, Alex
