On 2013-08-12 7:04 PM, David Daney wrote: > That's a mighty thin changelog there. I thought it was obvious that the lines I removed contain no useful code at all :) > You are changing the semantics in the > asm/mach-cavium-octeon/dma-coherence.h case. I'm just removing a fallback BUG() that could never be reached before my change either. As for the other platforms, I verified that there's no useful code in these functions anywhere in the tree. > Have you verified that all in-tree cases really are NOPs? Yes. - Felix
