> The return value semantics of nxm_read_field() are surprising in that > it just masks off the bits not being read. If I had not read the code > for the function, I would have guessed that it also shifted the bits > so that the least-significant bit being read was the least-significant > bit in the return value. If the existing behavior is actually the > most convenient, it'd be good for the function's comment to mention > it.
Turns out this is a bug. I sent out another version of that patch, plus a patch that adds test. We may want to consider backporting it. > > Instead of nxm_read_field__() and nxm_read_field(), I'd be inclined to > keep nxm_read_field()'s existing name and name the new function > nxm_read_subfield() or nxm_read_field_bits(). It sounds more > descriptive to me. > > _______________________________________________ dev mailing list dev@openvswitch.org http://openvswitch.org/mailman/listinfo/dev
