> > 3. use run-time checks all over the place, of the
> > "sizeof(dma_addr_t)==sizeof(u64)" kind, which adds unnecessary
> > overhead to
> > all platforms.
>
> Actually, these aren't technically run time checks. Although the cpp
> can't be used for sizeof(), the compiler (at least gcc) does seem to
> have enough sense to optimise away if(..) branches it has enough
> information to know won't be taken at compile time.
>
> As long as this optimisation works, I think the if(sizeof(..)) checks
> are fine for this.
Well, yes and no. Indeed those checks are optimized away, but as a result
of using them most data-access macros must be converted to inline
functions. And, last I heard at least, gcc was optimizing inline functions
much worse than preprocessor macros.
There are various other things that are made easier by a preprocessor
directive -- constructing the right data structures, for instance.
However, if such a patch is ultimately not getting accepted, these checks
is probably what I'll end up using...
Thanks,
Ion
--
It is better to keep your mouth shut and be thought a fool,
than to open it and remove all doubt.
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/