[PATCH] any value in numerous uses of "register" storage class?

[Rob Landley]

On Sunday 26 March 2006 9:41 am, Robert P. J. Day wrote:
>   given the efficiency of optimizing compilers these days, is there
> much value in the plethora of variables defined with the "register"
> storage class?  in any event, we can at least get rid of the few that
> have been commented out:

Yeah, removing commented out code is generally an easy sell.  (When they get 
around to enabling it, they can put it back in.  Meanwhile, it doesn't belong 
in the tree.)

Applied.

For any others you'd like to remove, bench and do make sizes to see if they 
make a difference.  In theory it shouldn't make any difference on a 
reasonably modern compiler, but in practice it may.  (It's a hint.)

I vaguely remember that qemu had a problem moving from gcc 3.x to gcc 4.x for 
reasons related to this, but don't know the details.  (qemu 0.8.0 apparently 
won't build on gcc 4.0 for this reason.  I haven't tried it...)

Rob
-- 
Never bet against the cheap plastic solution.