On Thu, 2007-05-17 at 15:27 -0700, Andrew Morton wrote: > Lovely. So it was actually generating wrong code on all > sizeof(size_t)!=sizeof(int) architectures.
I was trying to work out which architectures would actually be affected. Probably not many. If it's in a register it probably doesn't matter -- the size only matters if you put it on the stack. And it would have to be a 64-bit architecture, none of which (iirc) are so register-starved that this particular argument would be on the stack anyway. So I'm not actually sure it will really generate wrong code on any architecture we support -- not that this excuses it in any way :) > If only we could find some way in which all callers of a function as > well as its definition can see the same declaration? Well, building with --combine helps. :) Admittedly it doesn't necessarily catch _all_ callers -- only those which we actually compile together. That does cover _most_ of the times we do crap like this without a prototype in a header file though, I suspect. As Geert points out, sparse can warn about non-static functions which aren't prototyped; if we start getting anal about that, it might help. -- dwmw2 - To unsubscribe from this list: send the line "unsubscribe netdev" in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
