Re: [Q] Static checker fun

Jerry> First, note that window_char_height is called directly Jerry> (i.e., without the macro) in both redisplay.c and window.c, Jerry> so the abstraction is already broken. Sure, but this is a very local abstraction: only for that function, in fact. Jerry> Second, it was *after* I expanded CURCHARSIZE that the Jerry> static checker then complained about CHARSIZE. Weird, huh? Jerry> I guess it was working its way backward up the dataflow Jerry> path and complained about the first item it saw that seemed Jerry> wrong to it. OK. I just thought that if this was one of the tools the shop you're visiting is developing, they might like to know about the anomoly. Jerry> If you would like me to put those macros back, that's fine. Jerry> As you note, a good optimizing compiler should do the job, Jerry> anyway. I think it's a good idea. The original author (Ben, I suppose) thought that all those "things" were somehow "the same". IMO we should either change none or change all. Jerry> And that CURSIZE macro is bizarre. Would you object to me Jerry> "fixing" it to be an rvalue? As it stands, it's violating Jerry> the C99 aliasing rules, apparently for no reason. Please do! I spent about 5 minutes checking to make sure that the lvalue-ness was bogus. Let's not put anybody else through that. If lvalue-ness becomes useful, surely that person will know how to rewrite the macro for lvalue-ness. I wonder if that might not confuse checkers and optimizers, too, although they can probably do what I did, just in 4 nanoseconds. :-) -- School of Systems and Information Engineering http://turnbull.sk.tsukuba.ac.jp University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN Ask not how you can "do" free software business; ask what your business can "do for" free software.