> Michael, does my guess about the source of corruption make sense? Is > that pattern of every fourth byte counting up 00 ... 1f something the > the HCA would write during initialization of ICM? Yes. -- MST