> If the hypothesis is that test and development kernels would catch the > alignment problems, then there shouldn't be any in the production > kernels right? However, if there _are_ any out there in production > don't we want them to become visible? Even on processors which can > handle them it is still better to not have the unaligned acesses in the > first place right? I was thinking of the case where someone loads a module that has not had the benefit of being developed in an environment where all the unaligned accesses could be easily seen and removed. It seems (to me) that the user of that system would be happier to have the flow of "unaligned access" messages on the console reduced by letting the h/w handle the ones that it could. It is definitely better to avoid the unaligned access completely, but I'm not sure this is possible in practice. -Tony -- To unsubscribe from this list: send the line "unsubscribe linux-ia64" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
|