If we don't accept this patch, then we will need to modify all the existing protocols, as they all use the duplicate names (do a grep "Address Address"). The only reason they compile is that they get lucky at this check and the Address the symbol table finds first is the one it is looking for. In some other protocols (or, I suspect, even with different python versions), it isn't as lucky.
I'm not sure I follow the simple compiler argument. With or without this patch, the compiler is still performs type matching and name lookup. I don't really see why this complicates that. -Derek On Jan 22, 2013, at 6:37 PM, Nilay wrote: > On Thu, January 17, 2013 2:28 pm, Hower, Derek wrote: >> Most protocols, by convention, use uppercase names for members of TBEs, >> messages, etc. >> > > My opinion is that the compiler should do as little work as possible. > Unless allowing such convention makes the compiler's job simpler, I would > vote against having this convention. > > -- > Nilay > > _______________________________________________ gem5-dev mailing list gem5-dev@gem5.org http://m5sim.org/mailman/listinfo/gem5-dev
